Ps Genes E.coli

Front 1

clpX

Back 1

ATPase and specificity subunit of ClpX-ClpP ATP-dependent serine protease

ClpX is an ATP-dependent molecular chaperone that serves as a substrate-specifying adapter for the ClpP serine protease in the ClpXP and ClpAXP protease complexes.

ClpX protects the lambda O protein from heat-induced aggregation, disassembles lambda aggregates and enhances lambda DNA binding. ATP binding is required for all these effects, and disaggregation requires ATP hydrolysis [ Wawrzynow95 ]. ClpX also converts inactive, dimeric TrfA into its monomeric form (capable of initiating replication of plasmid RK2) in an ATP-dependent manner [ Konieczny97 ].

ClpX is required for normal replication of Mu transposase [ MhammediAl94 ]. ClpX catalyzes the ATP-dependent release of MuA from its active transposase tetramer form, allowing recruitment of host factors necessary for post-recombination steps in Mu transposition [ Levchenko95 , Kruklitis96 ]. ClpX is also able to globally unfold MuA monomers. ClpX recognizes a ten amino acid peptide from the carboxy-terminus of MuA when it is revealed by MuB. ClpX will recognize other proteins with this tag artificially attached [ Levchenko97 ].

SspB binding stimulates ClpX ATPase activity [ Wah02 ].

ClpX is a hexamer of ClpX monomers, stabilized by ATP binding and capable of capping ClpP tetradecamers [ Grimaud98 ]. While the ClpX ATP-binding site is necessary for oligomerization and binding to ClpP, both processes continue in the absence of ATP [ Banecki01 ]. ATP-bound ClpX is protease resistant [ Singh01a ]. The carboxy-terminus of ClpX is required for interaction with ClpP, as is the tripeptide IGF, though the latter is dispensible for ClpX chaperone activities [ Kim01e , Singh01a ]. Mutations in the interface between the carboxy-terminus of each subunit and the ATPase domain of its neighbor prevent disassembly of bound substrate [ Joshi03 ]. Substrate recognition requires that three of the six monomers have bound ATP, while the other three may have bound ATP or ADP [ Hersch05 ].

Each ClpX monomer has two PDZ domains the bind to the carboxy-terminus of target proteins. These domains show up as disordered sequence in NMR and are unstable when expressed independently [ Levchenko97a , Smith99b ]. ClpX also has an ATP-binding site motif and a zinc-binding domain, the latter being a member of the treble clef zinc finger family, involved in macromolecular interactions [ Gottesman93 , Donaldson03 ].

ClpX is required for adaptation to and extended viability in stationary phase, as well as growth in SDS [ Weichart03 , Rajagopal02 ].

ClpX can be expressed without ClpP [ Yoo94 ]

Front 2

prfH

Back 2

predicted peptide chain release factor

PrfH shows sequence similarity to the peptide chain release factors RF1 and RF2 [ Pel92 ].

Based on the orthologous region in other sequenced bacterial genomes, a deletion event appears to have removed approximately 1000 bp, encompassing the 3' end of ykfJ and the 5' end of prfH. Thus, both ykfJ and prfH may be pseudogenes. An intact PrfH may function as a peptide chain release factor. The co-occurance of ykfJ and prfH appears to be evolutionarily conserved and may thus provide insight into the function of PrfH [ Baranov06 ]

Front 3

ppiD

Back 3

periplasmic folding helper protein

PpiD shows similarity to peptidyl-prolyl cis-trans isomerases, but contrary to an earlier report [ Dartigalon98 ], does not have catalytic activity [ Weininger10 ].

The ppiD gene encodes a peptidyl-prolyl isomerase (PPIase) [ Dartigalon98 ]. PpiD activity is a chaperone required for wild-type outer membrane protein folding [ Dartigalon98 ].

PpiD activity is periplasmic, and PpiD has a segment within the cytoplasmic membrane [ Dartigalon98 ]. PpiD has been found as a homodimer and homotrimer in the membrane [ Stenberg05 ]. PpiD forms a complex with YfgM [ Maddalo11 ]. ppiD was initially isolated as a multicopy suppressor of surA mutant phenotypes, and a ppiD surA double mutation was reported to cause synthetic lethality [ Dartigalon98 ]. In addition, [ Dartigalon98 ] reported that a ppiD mutation causes an outer membrane protein folding defect and a periplasmic stress phenotype. However, later reports showed that In contrast to this, other mutations involving four PPIases present in Escherichia coli (FkpA, PpiA, PpiD, and SurA) showed that three triple mutants have growth defects, while one does not [ Justice05 ]. The quadruple mutant had the most pronounced defect as well as increased antibiotic sensitivity compared with other surA mutants [ Justice05 ]. No single or double mutants were shown to have growth defects in this study [ Justice05 ].

A mutation causes retardation of outer membrane protein translocation into the periplasm [ Antonoaea08 ].

Regulation has been described [ Dartigalon98 ].

Locations: periplasmic space, inner membrane [Stenberg05 ]

Front 4

ppiB

Back 4

PpiB is a peptidyl-prolyl cis-trans-isomerase (PPIase), catalyzing the conformational isomerization of prolyl residues in peptides. Cis-trans isomerization of prolyl peptide bonds is a slow step in protein folding, and thus PpiB is thought to facilitate proper protein folding.

PpiB was shown to improve the efficiency of bovine protein disulfide isomerase as a catalyst of oxidative protein folding [ Schonbrunn92 ] and to catalyze the refolding of denatured type III collagen [ Compton92 ]. Unlike the eukaryotic cyclophilins, PpiB activity is only inhibited by high concentrations of cyclosporin A [ Compton92 ] or FK506 [ Hayano91 ].

Crystal structures of PpiB in two different conformations have been determined [ Konno96 , Edwards97a ]. Refolding kinetics of the protein have been studied [ Ikura00 ]

Front 5

rna
Ribonuclease I (RNase I)

Back 5

Ribonuclease I (RNase I) is an endonuclease that cleaves phosphodiester bonds in RNA, yielding nucleoside 3'-phosphates and 3'-phosphooligonucleotides [ Spahr61 ]. RNase I is partially reponsible for the degradation of total and ribosomal RNA during both normal and nutrient starvation conditions, especially during carbon starvation [ Kaplan74 , Kaplan75 , Cohen77 ]. RNase I is specifically required for the breakdown of 23 S RNA, though it is not required for degradation of 16 S RNA or very small (4 S) RNA fragments that result from breakdown of larger RNA [ Kaplan75a , Kaplan75 ]. RNase I degradation of the 50 S ribosome releases the ribosomal proteins L4, L10 and L7/12 in addition to cleaving the 23 S RNA to yield a smaller product [ Raziuddin79 ]. Polyamines stimulate the activity of RNase I against synthetic polynucleotides in vitro [ Kumagai77 ].

RNase I is a periplasmic protein that can be released by spheroblasting or treatment of cells with N-dodecyldiethanolamine. This release allows subsequent enhanced breakdown of rRNA by RNase I [ Neu64 , NEU64 , NEU64a , Neu65 , Abrell71 , Lambert76 ]. RNase I appears to remain with the membrane fraction in disrupted cells [ Kaplan76 ].

Some structural analysis of RNase I has been completed. Its conformation energy is 11.5 kcal/mol at pH 7.5, and its T(m) is 64 degrees C at pH 4.0 [ Padmanabha01 ]. It has a 23-residue amino-terminal signal sequence which is cleaved and likely allows its transport to the periplasm [ Meador90 ]. Preliminary crystallization of RNase I has been done, with visualization of the structure at greater than three angstrom resolution [ Lim93 ].

Mutants lacking RNase I or other ribonuclease activities have reduced DNA degradation, possibly due to interaction between excess RNA and DNA endonuclease I [ Wright71 ]. RNase I is also required for full recovery from starvation, as cell viability studies show a direct correlation between recovery from starvation and the ability to degrade RNA [ Kaplan75a ]. A method has been developed to screen for mutants in RNase I by checking for a delay in β-galactosidase expression during amino acid starvation [ Kaplan73a ]

Front 6

miaB

Back 6

isopentenyl-adenosine A37 tRNA methylthiolase MiaB

MiaB is an isopentenyl-adenosine tRNA methylthiolase that assists in the prevention of frameshift mutations.

MiaB is required for the methylthiolation of N-6-isopentyl adenosine-37 in tRNA to yield 2-methylthio-N-6-isopentyl adenosine-37 tRNA [ Esberg95 , Esberg99 , Pierrel02 ]. This reaction has been characterized in vitro with the MiaB homolog in Thermotoga maritima, which can complement miaB nulls. In this reaction, the anticodon stem loop of the tRNA is the substrate [ Pierrel04 ]. The sulfur in this reaction appears to come from the MiaB protein itself, although how that sulfur is regenerated is unknown [ Pierrel04 ].

MiaB is a monomeric iron-sulfur cluster protein, with 2.5-3 iron and 3-3.5 sulfur atoms per protein. Under reducing or anaerobic conditions, MiaB contains 4Fe-4S clusters, but under aerobic conditions 2Fe-2S and 3Fe-4S clusters are found [ Pierrel02 ].

MiaB has been identified by an iterative profile search as a member of the large Radical SAM protein superfamily, some of whose members generate radicals by reductive cleavage of S-adenosylmethionine [ Sofia01 ].

miaB mutation leads to increased frameshifting by certain tRNAs [ Urbonavici01 ]. However, unlike mutants in the associated gene miaA, miaB mutants do not have a mutator phenotype or a drop in tet resistance [ Zhao01a , Taylor98 ]

Front 7

rimK

Back 7

ribosomal protein S6 modification protein

RimK is required for wild-type post-translational addition of C-terminal glutamic acid residues to ribosomal protein S6 (RpsF) [ Kang89 ].

RimK has a sequence motif that is predicted to mediate interaction with RNA [ Koonin94a ].

A rimK mutant exhibits a defect in the wild-type post-translational addition of C-terminal glutamic acid residues to ribosomal protein S6 (RpsF) [ Kang89 ]. An RpsF-E130K mutant protein is refractory to RimK-mediated modification [ Kang89 ]

Front 8

clpA

Back 8

ClpA ATP-dependent protease specificity component and chaperone

ClpA is an ATP-dependent molecular chaperone that serves as a substrate-specifying adapter for the ClpP serine protease in the ClpAP and ClpAXP protease complexes.

In its capacity as a chaperone, ClpA activates the RepA replication initiator protein of plasmid F1 in an ATP-dependent manner, converting it from a dimer to a monomer [ Wickner94 ]. This activity requires interaction between ClpA and the amino-terminus of RepA [ Hoskins00 ]. Should the RepA amino-terminus be blocked, ClpA can still interact with it as long as there is an accessible amino acid tract at the RepA carboxy-terminus [ Hoskins06 ].

ClpA lacking its own amino-terminal domain is still able to function as both chaperone and protease adaptor, though it is less effective in both roles than the wild-type protein [ Lo01 ].

Each ClpA monomer has two domains, leading to a double-stacked ring structure in the complete ClpA hexamer [ Beuron98 ]. The putative substrate-recognition domain of ClpA is stable and folds independently, unlike the matching domains in ClpB and ClpX [ Smith99 ]. Each ClpA monomer has two AAA+ modules (consensus ATP-binding sites), the first of which interacts with the amino-terminal domain of the protein [ Gottesman90 , Guo02 ]. A lysine mutation in either ATP-binding site prevents the ATP-dependent formation of the ClpA hexamer, as well as disrupting ATPase activity and removing the ability to activate ClpP. Mutants in the second ATP-binding site were still able to stimulate degradation of some shorter peptide substrates requiring nucleotide binding but not hydrolysis [ Singh94 , Seol95 ].

The ClpA hexamer forms in an ATP-dependent manner [ Kessel95 ]. Successful formation of the hexamer and subsequent interaction with ClpP requires the carboxy-terminus of ClpA. In its ATP-bound state, ClpA is protease resistant [ Singh01 ].

ClpA is required for the ATP-dependent degradation of certain substrates by ClpP, including some abnormal proteins and the in vitro test substrate casein [ Katayama88 , Hinnerwisc05 ]. ClpA binds to the SsrA degradation peptide tag, with one tag binding per ClpA hexamer. This interaction does not depend on either ATP-binding domain in ClpA [ Piszczek05 ]. The amino-terminal domain of ClpA is required for binding nonspecific protein substrates that have not been tagged with SsrA [ Xia ]. NEM inhibits ClpA function by introducing a bulky alkyl group but not by directly binding to a catalytic residue [ Seol97a ].

ClpA is required for maximal growth in SDS, normal adaptation to and extended viability in stationary phase and for activity of bacteriophage Mu [ Rajagopal02 , Weichart03 , Shapiro93 ].

A 65 kD truncated internal transcript from clpA prevents degradation of ClpA by ClpAP [ Seol95a ].

ClpS binds to the ClpA amino-terminus and affects the specificity of protein degradation by the ClpAP chaperone-protease complex, possibly by interfering with interactions between substrate and ClpA [ Dougan02 ]. ClpS stimulates ClpAP recognition and degradation of aggregated protein substrates while it inhibits degradation of non-aggregated substrates including ClpA [ Dougan02 ].

A crystal structure of the ClpA N terminus shows a zinc binding site [ Guo02a ]. Crystal structures of ClpS bound to the N-terminal region of ClpA are presented at 2.3 Å [ Guo02a ], 2.5 Å [ Zeth02 ], and 3.3 Å [ Guo02a ] resolution.
The clpA gene contains a sequence for an internal translational initiation and therefore encode two polypeptides with different sizes (ClpA65 and ClpA84)

Front 9

infA

Back 9

IF-1 is one of three translation initiation factors in E. coli and is essential for viability [ Cummings94 , Baba06 ]. Its functional role has not been fully elucidated; a collection of mutants was generated to attempt to determine the function of IF-1 [ Croitoru04 ]. It was suggested that the essential function of IF-1 and IF-3 may be to minimize the fraction of ribosomes lacking an initiator tRNA [ Antoun06a ].

IF-1 binds to the ribosomal A site [ Moazed95 ], which may suggest a function in translational fidelity; such a function could not be shown [ Croitoru05 ]. Certain mutations in 16S rRNA disrupt binding of IF-1 to the 30S subunit [ Dahlquist00 ], and a model where IF-1 modulates specific conformational change during initiation has been suggested [ Dahlquist00 ]. IF-1 enhances the dissociation rate of 70S ribosomes apparently by stimulating the acitivity of IF-3 [ DottavioMa79 , GrunbergMa75 , Pon84 ]. IF-1 stimulates the association of IF-2 with the 30S subunit [ Stringer77 , Moreno99 ] and the ability of IF-2 to stimulate template-dependent binding of aminoacyl-tRNAs to the ribosome [ Canonaco86 ]. IF-1 is released from the ribosome during association of the 30S and 50S subunits [ Celano88 ].

The solution structure of IF-1 has been determined, and residues that are involved in interactions with the 30S ribosomal subunit have been identified. The structure appears similar to the oligomer-binding (OB) fold family of proteins [ Sette97 ].

Expression of infA is growth rate-controlled at the level of transcription [ Cummings91 ]. Transcription of infA is increased by cold shock via activation of the distal promoter [ Ko06 ]

Front 10

rpsA

Back 10

The S1 protein is essential in E. coli; it is a component of the ribosome and is likely required for translation of most mRNAs [ Sorensen98 ]. During translation initiation, mRNAs are selected and bound to the ribosome with the help of two components: the conserved 3' end of 16S rRNA is complementary to the Shine-Dalgarno region of the typical mRNA, while the S1 protein binds to the leader sequence of mRNAs, upstream of the Shine-Dalgarno region [ Boni91 , Komarova02 ]. The S1 protein is not required for the translation of leaderless mRNAs [ Tedin97 , Moll02 ]. Association of S1 with the ribosome is unstable [ Subramania77 ], and the S2 protein is required for binding of S1 to the 30S subunit of the ribosome [ Moll02 ].

The S1 protein is the largest of the ribosomal proteins and assumes a complex elongated shape; it is located at the junction of the head, platform, and main body of the 30S subunit [ Sengupta01 ].

The translation initiation region of rpsA mRNA, which encodes the S1 protein, lacks a canonical Shine-Dalgarno region, but nevertheless supports high levels of translation [ Boni00 ]. Expression of S1 is autoregulated; the S1 protein acts as a specific repressor of translation of its own mRNA [ Christians81 , Skouv90 ]. Supported by various types of evidence including site-directed mutagenesis, footprinting and phylogenetic studies, a model was proposed in which the rpsA leader region folds into three stem-loop structures which enable efficient translation; binding of free S1 protein disturbs the conformation of the rpsA mRNA and thus specifically represses translation [ Boni01 , Tchufistov03 ].

S1 proteins carrying a C-terminal truncation due to the ssyF29 mutation lack the R4 RNA binding domain and are able to stimulate translation of the rpsA mRNA. This autoregulation requires the presence of at least 90 nucleotides upstream of the start codon [ Boni00 ]. In contrast, the N-terminus of the S1 protein was reported to be sufficient for repressing the expression of rpsA [ Skouv90 ]. It has since been proposed that repression may have been due to full-length S1 protein that had been displaced from the ribosome by high concentrations of the N-terminal domain [ Boni00 ].

The S1 protein has been suggested to play a role in trans-translation by binding transfer-messenger RNA (tmRNA) and delivering it to stalled ribosomes [ Bordeau02 , Karzai01 , Wower00a ]. An alternative model suggests that S1 binds to both mRNA and tmRNA molecules indiscriminately and may not play a direct role in tmRNA-mediated tagging of incompletely translated polypeptides [ McGinness04 ].

The S1 protein has been reported to bind to RNA polymerase and stimulate transcription of a number of promoters [ Sukhodolet03 ].

ssyF: "suppressor of secY24(Ts)" [ Shiba86 ]

Front 11

aat

Back 11

L/F-transferase (Aat) attaches leucine and phenylalanine to exposed amino-terminal arginines and lysines on proteins. This is a key step in the ClpAP-dependent N-end rule degradation pathway.

L/F-transferase catalyzes the transfer of leucine and phenylalanine from charged tRNA to amino-terminal lysines and arginines on substrate proteins [ Momose66 , Leibowitz69 , Kaji65 , Leibowitz70 , Kuno03 ]. When assayed with small peptides, Aat only modifies peptides bearing amino-terminal L-arginine and L-lysine, although the simple dipeptide D-lysyl-D-valine is not a substrate [ Soffer73 ]. Testing with purified Aat and actual protein substrates shows that Aat will only transfer to amino-terminal, rather than internal, arginines in proteins [ Leibowitz71 , Leibowitz71a ]. Though addition of both leucine and phenylalanine occurs in vitro, an in vivo experiment with arginine-β-galactosidase yielded only leucine addition [ Shrader93 ]. In one case, purified L/F-transferase has been shown to transfer methionine onto the amino-terminus of substrate peptides [ Scarpulla76 ]. L/F-transferase has also been shown to transfer to an unidentified 12-kilodalton protein component of the 30S ribosome [ Leibowitz71a ].

L/F-transferase is required for the degradation of some substrates for the N-end rule pathway, by which a protein undergoes rapid degradation by the ClpAP protease if its amino-terminal residue is arginine, lysine, leucine, phenylalanine, tyrosine or tryptophan. Arginine and lysine are considered secondary destabilizing residues, as the addition of leucine or phenylalanine by L/F-transferase is required to allow degradation [ Tobias91 ].

The chaperone protein GroEL copurifies with L/F-transferase. Recombinant L/F-transferase with no RNA present is as active as purified wild-type protein which may contain some RNA [ Abramochki95 ].

A number of structural studies have been carried out on L/F-transferase. Crystal structures of isolated Aat to 2.4 Å and 1.6 Å resolution and of Aat bound to the aminoacyl-tRNA analog puromycin to 2.8 Å resolution show that Aat comprises two compact domains. Both domains are involved in puromycin, and presumably tRNA, binding via a hydrophobic pocket that is required for Aat activity [ Suto06 , Dong07 ]. Crystal structures of Aat complexed with a phenylalanyl-tRNA analog with or without an associated peptide substrate show that the phenylalanine side chain binds in the hydrophobic pocket as suggested by the puromycin analysis. In addition, this work shows that except for the amino-terminal residue, recognition of peptide substrates by L/F-transferase is sequence independent [ Watanabe07 ]. Truncation of the Aat amino-terminus by 33 or 78 residues results in a protein that is capable of measureable transferase activity and that has wild type substrate specificity [ Ichetovkin97 ].

Mutants lacking L/F-transferase have diminished activities of L-phenylalanyl-transfer ribonucleic acid synthetase and tryptophanase and accumulate high levels of enterochelin during iron limitation. Neither synthetase nor tryptophanase are substrates for L/F-transferase, however [ Deutch77 ]. When grown on glycerol, L/F-transferase mutants are partial proline auxotrophs and show a fourfold increase in proline oxidase activity [ Deutch75 ]. This increase in activity is due to an increase in the amount of proline oxidase, rather than in the enzymatic activity of individual proline oxidases [ Scarpulla79 ]

Front 12

hyaC

Back 12

hydrogenase 1, b-type cytochrome subunit

The hyaC gene product is very hydrophobic, rich in aromatic residues, and has four putative hydrophobic membrane-spanning regions [ Menon90 ].

An in-frame deletion in the hyaC gene results in wild-type levels of hydrogenase 1 activity, although resulting in the appearance of multiple forms of the enzyme [ Menon91 ]

Front 13

rpmF

Back 13

50S ribosomal subunit protein L32

Front 14

pepT

Back 14

peptidase T

Peptidase T is a tripeptidase that is upregulated in sessile bacteria growing on biofilms [ Sussman70 , Simmonds76 , PrigentCom99 ]

Front 15

lolB

Back 15

outer membrane lipoprotein, localization of lipoproteins in the outer membrane

LolB is an outer membrane lipoprotein which is generally required for LolA-dependent localization of lipoproteins (including LolB itself) to the outer membrane [ Matsuyama97 , Yokota99 ].

LolB is essential for growth [ Matsuyama97 ]. Depletion of LolB causes the accumulation of LolA-lipoprotein complexes in the periplasm and of mature lipoproteins in the inner membrane [ Tanaka01a ]. Site-directed mutagenesis has identified possible functions for several conserved Trp residues within LolB [ Wada04a ]. LolB is expressed constitutively [ McNicholas97a ].

The LolB (HemM) protein was originally thought to be involved in the biosynthesis of delta-aminolevulinic acid [ Ikemi92 ].

A crystal structure has been solved at 1.9 A resolution; the LolB protein contains a hydrophobic cavity which may be a binding site for the lipid moiety of lipoproteins [ Takeda03 ]. In vivo photo cross-linking between LolA and mLolB suggests that the internal surface of the LolA hydrophobic cavity interacts with the external surface of the mLolB hydrophobic cavity and a 'mouth-to-mouth' model for the transfer of lipoproteins has been proposed [ Okuda09 ]

Front 16

prfA

Back 16

Release factor 1 (RF1) is one of two class 1 codon-specific factors in E. coli that facilitate the release of the growing polypeptide chain at stop codons. RF1 recognizes the termination codons UAG and UAA [ Capecchi67 , Scolnick68 , Capecchi70 , Beaudet70 , Martin88 ]. Termination is highly precise; RF1 is able to discriminate against related sense codons by 3 to 6 orders of magnitude and without requiring energy-driven error correction [ Freistroff00 ]. The discriminator site consists of a tripeptide which appears to be functionally equivalent to the anticodon of tRNA [ Ito00a ]. An E. coli cell is variously estimated to contain between 500 and 4900 molecules of RF1 and reaches the highest RF1 content at a high growth rate [ Klein71 , Adamski94 ].

RF1-mediated termination is sensitive to the codon-anticodon interaction of the wobble base at the last amino acid residue of the peptide chain and the tRNA at the ribosomal P-site [ Zhang96c ]. Interactions between RF1 and the ribosome have been mapped [ Moffat91 , Van03b ]. The universally conserved tripeptide sequence GGQ is thought to interact with the peptidyl-transferase center of the ribosome and may be functionally equivalent to the CCA end of tRNA [ Zavialov02 , Mora03 ]. RF1 is post-translationally modified by N5 methylation by PrmC in vitro, most likely at position Gln235 within the conserved GGQ sequence [ Nakahigash02 , HeurgueHam02 ]. A crystal structure of the complex between RF1, PrmC, and S-adenosyl-homocysteine, the product of the methyl transfer reaction, has been solved at 3.1 Å resolution [ Graille05 ]. A small-angle X-ray scattering solution structure of RF1 alone shows an open structure, similar to its structure when RF1 is associated with the ribosome [ Vestergaar05 ].

Certain mutations in prfA lead to suppression of the UAG and UAA stop codons and are conditionally lethal [ Ryden84 ]. An allele of rpsL suppresses this phenotype, indicating that RF1 and L7/L12 interact [ Zhang94a ]. The N-terminal domain of RF1 is not required for its activity [ Mora03a ].

Expression of RF1 increases with the growth rate; the regulation is controlled at the hemA1 promoter [ Dahlgren04 ]

Front 17

prmC

Back 17

PrmC is a protein-(glutamine-N5) methyltransferase that shows activity toward polypeptide chain release factors RF1 and RF2 [ Nakahigash02 , HeurgueHam02 ]. PrmC was initially reported to be a putative protoporphyrinogen oxidase, based on genetic experiments [ Nakayashik95 ].

The crystal structure of PrmC is reported at 3.2 A resolution [ Yang04a ]. The N terminus may determine methyltransferase substrate specificity [ Bujnicki99 ], and the structure of the putative substrate binding domains of the E. coli and Thermotoga maritima proteins are similar despite their lack of sequence similarity [ Yang04a ]. The C terminus of the Thermotoga maritima protein contains the active site, and the structure illustrates similarities to DNA methyltransferases within this catalytic region [ Schubert03 ]. The catalytic mechanism is discussed [ Schubert03 ].

A mutant shows a defect in nonsense codon recognition [ Nakahigash02 ], defects in ribosome progression [ Nakahigash02 ], and a growth defect [ Nakayashik95 ] that is suppressed by mutations in polypeptide chain release factor RF2 [ Nakahigash02 , HeurgueHam02 ]. A mutant shows a defect in heme synthesis from 5-aminolevulinic acid [ Nakayashik95 ]. A mutant exhibits a global transcription pattern characteristic of anaerobic growth, which may decrease oxidative stress and therefore account for the genetic results originally thought to be indicative of a PrmC role in heme synthesis [ Nakahigash02 ].

The PrmC/HemK family of proteins shows similarity to gamma subfamily S-adenosyl-methionine-dependent adenine-specific DNA methyltransferases [ Bujnicki99 ]. PrmB and PrmC have sequence similarity to each other; PrmB is active toward ribosomal protein L3, whereas PrmC is active toward polypeptide chain release factors RF1 and RF2 [ HeurgueHam02 ]

Front 18

hns

Back 18

H-NS protein for "Histone-like nucleoid structuring protein," is a nucleoid-associated multifunctional protein that is capable of condensing [ Dame00 ] and supercoiling DNA [ Zimmerman06 , Dame05 , McLeod01 , Tupper94 ]. This protein acts as a pleiotropic transcriptional factor with a strong preference for horizontally acquired genes among the 250 loci to which it binds [ Yamada90 , Oshima06 ]. H-NS functions almost exclusively as a transcriptional repressor, although there is no clear evidence that this regulator is an activator. Currently, no inducer for this regulator has been reported in the literature, although Reush et al. proposed that this regulator can form a complex with a short chain of polyhydroxybutyrate [ Reusch02 ]. New genes may be identified by high-throughput analysis [ Uyar09 ],

H-NS plays an important role in the regulation of many genes in response to environmental changes and adaptation to stress; therefore, it is capable of controlling its own synthesis [ Falconi93 , Falconi96 ]. It also regulates transcription of many other genes that participate in a variety of cellular functions, including genes involved in the following processes or responses: biogenesis of flagella [ Bertin94 , Soutourina99 , Landini02 ], transcription control of the type I fimbria structural genes [ Donato99 , Donato97 , Olsen98 , Olsen94 , Schembri98 ], acid resistance [ Shin05 ], the functional glutamic acid-dependent system [ De99a ], osmotically inducible genes [ Bouvier98 ], the glutamate decarboxylase-dependent acid resistance system [ Giangrossi05 , Ma02 , Tramonti02 , Hommais01 ], osmotic control [ Lucht94 , Rajkumari01 ], the type II secretion pathway [ Francetic00 ], carbon sources [ Rimsky90 ], genes involved in the RNA component of the small subunit (30S subunit) [ Afflerbach98 , Gralla05 ], and proteases [ Forns05 ], among others.

H-NS is capable of inducing severe bends in the DNA, interacting with a large number of DNA regions that contain a planar curvature [ Yamada90 , Yamada91 , Jauregui03 ]. It has been suggested that H-NS binds strongly to sites carrying a 10-bp AT-rich consensus sequence, which functions as a nucleation site for the formation of a repressive higher-order nucleoprotein complex [ Lang07 , Sette09 ]. H-NS binds to intergenic regions as well as regions within genes, but not all genes that bind H-NS are affected by this protein, a fact that is in agreement with the primary role assigned to HN-S in maintenance of nucleoid structure. Currently, there are different models for the formation of DNA-H-NS-DNA bridges which show that this protein binds in tandem to sequences in the genome, forming multimers [ Dame05 , Dorman07 , Luijsterbu06 ].

As expected for a gene involved in the modulation of many cellular processes, the expression of hns is regulated by several systems and at different levels. At the transcription level, hns is autoregulated, and it is controlled by different transcription factors. hns is induced by high hydrostatic pressure [ Welch93 ] and DNA synthesis [ Free95 ]. At the posttranscriptional level, it is subject to regulation by the sRNAs Hfq and DsrA [ Lease00 , Brescia03 , Majdalani05a , Brescia04 , Repoila03 ].

It is a DNA-binding protein with similarity to StpA [ Shi94 , Zhang92 ] and these two proteins can have similar functions [ Ali99 , Azam00 ]. It has an approximately fivefold-lower affinity for DNA than StpA and has a major preference for curved DNA [ Sonnenfiel01 ].

Expression of stpA from a plasmid can complement an hns mutant phenotype and StpA is able to repress and activate a subset of H-NS-regulated genes, but the specific mechanisms remain to be determined [ Shi94 , Sonnenfiel01 , Sonden96 , Zhang96 , Uyar09 , Sonden96 ]. A dominant negative form of StpA can disrupt H-NS activity and vice versa, and H-NS can interact with StpA at two distinct domains to form heterodimers in vitro; also, there is evidence that these proteins can form homodimers [ Johansson01 , Williams96a , Dorman99 , Williams96a ]. For this reason, in the absence of H-NS the StpA protein is rapidly degraded in a Lon protease-dependent manner [ Johansson01 , Johansson99 ]; protection from proteolytic degradation appears to be mediated by a direct interaction between StpA and H-NS [ Johansson01 ]. On the other hand H-NS also may form heterotrimeric complexes with Hha and YdgT [ Madrid07 , Paytubi04 ].

H-NS is a small protein and it is an abundant nucleic acid protein in the genome, with about 20,000 copies per cell. This regulator belongs to the histone-like family of transcriptional regulators and the structure of the protein consists of two structured domains which are separated by a flexible linker [ Dorman04 ]. The N-terminal domain is required for oligomerization and it is involved in protein-protein interactions, while the purified C-terminal domain is involved in DNA binding [ Dorman04 , Dorman99 , Williams96a , Rimsky04 , Sette09 ]

Front 19

oppF

Back 19

OppF is an ATP-binding component of the oligopeptide ABC transporter and the murein tripeptide ABC transporter

Front 20

oppC

Back 20

OppC is an integral membrane component of the oligopeptide ABC transporter and the murein tripeptide ABC transporter

Front 21

sohB

Back 21

predicted inner membrane peptidase

SohB might be a periplasmic protease, based on sequence analysis and the genetic interaction between sohB and htrA [ Baird91 ].

Multicopy expression of sohB suppresses the heat sensitivity of an htrA mutant [ Baird91 ].

SohB is subject to posttranslational processing of a signal sequence; the precursor is 39 kDa and the mature protein is 37 kDa [ Baird91 ].

SohB has similarity to the signal peptidase Protease IV [ Baird91 ].

SohB: "suppressor of htrA" [ Baird91 ]

Front 22

rnb

Back 22

Ribonuclease II (RNase II) is an exonuclease that cleaves RNA from the 3' end to produce ribonucleoside 5'-monophosphates. RNase II is one of five exonucleases (RNase II, Rnase D, RNase BN, RNase T, RNase PH) that can process tRNA and is responsible for the first step in the conversion of tyrosine tRNA to its final form [ Reuven93 , Kitamura77 ]. It has been shown in vitro to cleave long 3' tRNA sequences to yield 2-4 nucleotide intermediates for subsequent processing [ Li94a ]. It also degrades the polycistronic tryptophan operon mRNA from its primary transcription termination site (trp t') to a secondary termination hairpin (trp t) to yield the mature tryptophan mRNA [ Mott85 ]. RNase II is responsible for a signficant portion of the degradation of artificial hammerhead ribozymes in vivo [ Wang96g ]. RNase II also contributes to the degradation of smaller RNA during carbon starvation, though it has no role in 23 S or 16 S RNA breakdown [ Kaplan74 , Kaplan75 ]. In concert with PNPase, RNase II degrades the major 3' cleavage product of the antisense RNA CopA, though it can also protect the 3' end from PNPase degradation [ Soderbom98 ]. Temperature-sensitive double mutants in RNase II and PNPase are inviable at the nonpermissive temperature and accumulate mRNA fragments of 100 to 1,500 nucleotides in length [ Donovan86 ]. Triple mutants in RNase II, PNPase and ams show a three- to four-fold increase in the half life of pulse-labeled RNA [ Arraiano88 ]. RNase II degrades rpsO mRNA following removal of the transcript's 3' stem-loop structure by RNase E, but protects it from degradation by other factors if the 3' stem-loop is present [ Hajnsdorf94 ]. RNase II also plays a role in bacteriophage T4 ribonucleic acid metabolism [ Birenbaum80 ].

RNase II degrades the poly(A) tails of mRNA. In triple mutants lacking RNase II, PNPase and RNase E, the length and abundance of mRNA poly(A) tails increases dramatically [ OHara95 ]. Double mutants in RNase II and PNPase show a twenty- to sixty-fold increase in appearance of poly(A) RNA and a three- to four-fold increase in the length of poly(A) tails relative to wild type, but less than a two-fold change in non-poly(A) RNA levels. Poly(A) synthesis is also higher in mutant cells, indicating that the exonucleases tested both degrade polyadenylated mRNA and reduce its rate of synthesis [ Cao97a ]. RNase II removes poly(A) tails from rpsO mRNA. The resulting deadenylated rpsO mRNA is more stable than oligoadenylated rpsO mRNA [ Marujo00 ]. RNase II is responsible for most degradation of of poly(A) tails associated with 23 S RNA and can be overexpressed to prevent toxicity due to excessive polyadenylation, which decreases the half-life of mRNA [ Mohanty00 , Mohanty99 ]. Hfq blocks the degradation of poly(A) tails by RNase II [ Folichon03 ]. Polyadenylation of the mRNA for the ribosomal protein S20 allows RNase II and PNPase to degrade it [ Coburn96 ].

RNase II catalyzes the direct removal of ribonucleoside 5'-monophosphates from the 3' end of oligoribonucleotides [ Wade61 , Wade61a , Sekiguchi63 , Spahr63 , Spahr64 ]. It degrades substrates with 3'-hydroxyl or 2',3'-cyclic phosphate ends somewhat more readily than those with 3'-phosphate or 2'-phosphate ends, and degrades substrate at a rate of seventy nucleotides per second at 37 degrees C. Its enzymatic activity has been extensively characterized with regards to substrate and reaction conditions [ Cannistrar94 ]. RNase II preferentially degrades longer ribonucleotides and does so processively, leaving a short oligonucleotide and little evidence of intermediates [ Nossal68 ]. It has been shown that the RNase II catalytic site binds the 3' end of substrate RNA and its anchor site binds 15-25 nucleotides from the 3' end, with both bindings needed to maintain the enzyme-RNA complex. The enzyme remains fixed at the anchor point on the RNA, processively cleaving twelve nucleotides from the 3' end and then dissociating [ Cannistrar99 ]. An Asp-209-Asn mutant of RNase II binds but does not hydrolyze substrate [ Amblar05 ].

Stable stem-loops block RNase II, forcing it to dissociate from the substrate RNA [ Spickler00 , Coburn96a ]. Stable stem-loops and other secondary structure elements have been shown to block RNase II degradation of the Tn10/IS10 antisense RNA RNA-OUT, rpsO mRNA and glyA mRNA, though stem-loop structures alone do not appear to stall RNase II long enough in vitro to account for their effectiveness in stabilizing mRNA in vivo [ Pepe94 , Hajnsdorf94 , Plamann90 , McLaren91 ]. RNase II degrades the tryptophan operon mRNA from its primary transcription termination site (trp t') down to a secondary termination hairpin (trp t), resulting in the final, mature 3' end of tryptophan operon mRNA [ Mott85 ].

In addition to its exonuclease activity, one study claims a significant endonuclease function for RNase II [ Spahr64 ]. A study based on a thermolabile RNase II appears to show that RNase II is not responsible for the majority of ribonucleoside 5'-monophosphate formation [ Cohen77 ].

ATP has been reported to be a competitive inhibitor of RNase II [ Venkov71 ]. This was subsequently demonstrated to be an experimental artifact caused by the presence of adenylate kinase, which works in concert with nucleoside monophosphate kinases to convert 5'-mononucleosides to insoluble nucleoside diposphates, thus resulting in an apparent diminishing of enzymatic output [ Ko73 , Holmes73 ].

Crystal structures of RNase II in the bound and unbound states have been determined to 2.74 and 2.44 Å resolution, respectively [ Frazao06 , McVey06 , Zuo06 ]. Based on these structures, Rnase II appears to bind RNAs at two locations, and to constrain catalytic access to single-stranded RNA via a narrow, basic channel. Based on subsequent truncation mutants, the RNase II S1 domain is important for stable protein-RNA complexes, the amino-terminal domain prevents overbinding to poly(A) RNA, and there is a third RNA-binding domain in the amino-terminal portion of the catalytic domain [ Amblar06 ].

PNPase degrades rnb mRNA, limiting RNase II expression. PNP expression is increased in an RNase II null strain, suggesting that this mRNA-based regulation is reciprocal [ Zilhao96 ]. RNaseE and RNase III also affect RNase II expression, the latter indirectly and the former directly by cleaving rnb mRNA [ Zilhao95 ]. Gmr regulates RNase II by modulating its degradation. In a gmr deletion mutant, RNase II is three times more abundant than in a wild type background [ Cairrao01 ].

Strains deleted in five exoribonucleases (RNase II, RNase D, RNase BN, RNase T, RNase PH) are inviable, but any one of them can support growth with no accumulation of precursor RNA or disruption of RNA synthesis [ Kelly92 , Zaniewski84 ]. Mutants in RNase II do show reduced DNA degradation, possibly due to competitive inhibition of endonuclease I by RNA [ Wright71 ]

Front 23

abgB

Back 23

p-aminobenzoyl-glutamate hydrolase subunit B

Front 24

abgA

Back 24

p-aminobenzoyl-glutamate hydrolase subunit A

Transcription of the potential abgABT operon may be regulated by the predicted transcriptional regulator AbgR [ Hussein98 ]

Front 25

lhr

Back 25

Lhr codes for the longest known protein in E. coli. It has conserved motifs from helicase superfamily II, and its amino-terminus is similar to eukaryotic DEAD family helicases, though no DNA- or RNA-stimulated Lhr ATPase activity has so far been detected [ Reuven95 ]

Front 26

relE

Back 26

Qin prophage; toxin of the RelE-RelB toxin-antitoxin system and cofactor to enhance the repressor activity of RelB

RelE is the toxin of the RelE-RelB toxin-antitoxin system [ Gotfredsen98 ]. RelE inhibits protein translation by catalyzing cleavage of mRNA in the A site of the ribosome in response to amino acid starvation [ Pedersen03 , Christense03 , Neubauer09 , Hurley11 ].

RelE is involved in regulation of cellular protein translation under nutritional stress conditions [ Christense01 , Pedersen03 , Christense03 ]. RelE-mediated translation inhibition is reported to cause reversible inhibition of cell growth [ Pedersen02 ]. The activity of tmRNA counteracts RelE translation inhibition [ Pedersen03 , Christense03 ]. RelE and RF1 functionally interact [ DiagoNavar09 ].

RelE and RelB physically interact [ Galvani01 ]. When cells are starved of amino acids, Lon protease degrades RelB; RelB degradation frees RelE and derepresses transcription of relBE. RelE accumulates in excess compared with its RelB antitoxin, and this free RelE causes translation inhibition [ Christense01 ]. Conversely, RelB binding to RelE protects RelB from proteolytic degradation [ Cherny07 , Overgaard09 ]. The structure of the RelE-RelB complex and cooperative binding to the relBE promoter region has been studied; at low levels, RelE enhances the interaction between RelB and the promoter, while excess RelE destabilizes promoter binding of the complex [ Li08 , Overgaard08 ].

Solution structures of a RelE mutant together with a RelB peptide have been determined; RelB appears to induce conformational changes in the RelE active site, inactivating it [ Li09b ]. Crystal structures of RelE alone and bound to a heterologous ribosome have elucidated the reaction mechanism of RelE. RelE binds to the ribosomal A site and cleaves mRNA after the second nucleotide of the codon; cleavage is ribosome-dependent [ Neubauer09 ]. The in vivo frequency and codon specificity of RelE cleavage has been mapped; contrary to earlier in vitro results, no sequence preference can be seen in vivo. RelE appears to cleave within the first 100 codons in the coding region of mRNAs [ Hurley11 ].

Mutations in the relB locus were initially identified by a "delayed relaxed" phenotype, characterized by a ~10 minute lag period followed by continued stable RNA synthesis in response to amino acid starvation [ Lavalle65 , Diderichse77 ]. This phenotype is due to destabilization of RelB, resulting in hyperactivation of RelE [ Christense04 ].

relBE is overexpressed in persister cells. Ectopic expression of RelE increases the number of persister cells that survive normally lethal antibiotic treatment [ Keren04 ]. Conversely, deletion of relBE was reported to increase survival after antibiotic treatment [ KolodkinGa09 ].

Phylogenetic analysis of toxin-antitoxin systems has been performed [ Anantharam03 , Pandey05 ]

Front 27

dcp

Back 27

dipeptidyl carboxypeptidase II

Dipeptidyl carboxypeptidase is a peptidase capable of cleaving peptide bonds in amino-blocked small peptide substrates [ Yaron72 , Deutch78 , Henrich93 ]. It is required for growth when only amino-blocked peptides such as N-acetyl-alanylalanylalanine and N-benzoyl-glycylhistidylleucine are available as carbon sources [ Deutch78 , Henrich93 ].

A fraction of the total pool of dipeptidyl carboxypeptidase is in the periplasm [ Deutch78 ]

Front 28

pqqL

Back 28

PqqL is a putative zinc peptidase by similarity [ Turlin96 ]

Front 29

pheS

Back 29

phenylalanyl-tRNA synthetase α-chain

The α subunit of PheRS contains the phenylalanine binding site [ Hennecke75 , Lavrik82 ] within the conserved motif 2 and motif 3 of the protein [ Kast91 , Kast91a ]. It interacts with the 3'-adenosine of tRNAPhe [ Hountondji87 ].

Isolated α subunits exist primarily as dimers [ Bobkova91 ]

Front 30

rpmI

Back 30

50S ribosomal subunit protein L35

The L35 protein is a component of the 50S subunit of the ribosome [ Wada87 ].

L35 can be crosslinked to the spiramycin derivative dihydrospiramycin and may thus be located near the peptidyl transferase center [ Bischof95 ]

Front 31

rnd

Back 31

RNase D is an exonuclease involved in the 3' ribonucleolytic processing of precursor tRNA. Though RNase D appears to be a minor player in this task and is not required for viability or proper tRNA processing, it can support such processing in the absence of other exonucleases (RNase II, RNase BN, RNase T and RNase PH) [ Reuven93 , Kelly92 , Zaniewski84 ].

In vitro, RNase D cleaves tRNA nonprocessively from the 3' end to yield mononucleotides and active tRNA. Cleavage of tRNA slows at the CCA nucleotide sequence, allowing aminoacylation of the tRNA that prevents additional cleavage [ Cudny80 ]. RNase D activity is very dependent on the structure of the 3' end of the target tRNA, with cleavage of altered tRNA proceeding much faster than cleavage of wild type, and no cleavage of tRNA bearing a 3' phosphate [ Ghosh78 , Cudny81 ]. When overexpressed in vivo, RNase D cleaves the 3' end of tRNA, eventually cleaving past the CCA sequence to yield damaged tRNA that cannot be fixed by tRNA nucleotidyltransferase [ Zhang88a ].

Translation of RNase D depends on a stem-loop structure followed by eight uridines upstream of the Shine-Delgarno sequence [ Zhang89 , Zhang92 ].

Overexpression of RNase D leads to slow growth [ Zhang88b ]

Front 32

ptrB

Back 32

Protease II is a serine protease that hydrolyzes peptide bonds following arginines and lysines.

Protease II cleaves after arginines and lysines [ Pacaud75 ]. Its specificity for artificial peptide substrates has been compared with oligopeptidase A [ Yan06 ]

Front 33

torZ

Back 33

trimethylamine N-oxide reductase III, TorZ subunit

TorZ is the catalytic subunit of TMAO reductase III. The TorZ protein was first identified as a biotin sulfoxide (BSO) reductase, and the gene was named bisZ [ delCampill96 ]. Recent work has shown that the reductase is more efficient with TMAO as a substrate than BSO. TorZ is exported to the periplasm via the Tat pathway [ Gon00 ]

Front 34

argS

Back 34

arginyl-tRNA synthetase

ArgS is a member of the family of aminoacyl tRNA synthetases, which interpret the genetic code by covalently linking amino acids to their specific tRNA molecules. The reaction is driven by ATP hydrolysis. ArgS belongs to the Class I aminoacyl tRNA synthetases; apart from sequence motifs within the active site, the different enzymes show little similarity in their primary amino acid sequences.

The argS gene has been cloned; the protein coding region uses the unusual initiation codon GUG and has a codon usage pattern which is typical for highly expressed genes [ Eriani89 ]. Several mutations in the argS gene have been described. The MA5002 mutant encodes a serine in place of arginine at position 134, close to the active site. The purified enzyme shows defects in enzymatic activity and Km value for ATP [ Eriani90b ]. The lov-1 mutation confers a slow growth phenotype as well as mecillinam resistance, which appears to be dependent on the RelA-mediated stringent response [ Vinella92 ]. Similarly, an argS mutant conferring novobiocin resistance has been isolated [ Jovanovic99a ]. A mutant deleting the arginine residue at position 245 was constructed, and the resulting protein was purified by renaturation from inclusion bodies; the specific activity of the purified enzyme was 0.3 % of the native enzyme [ Wu98a ].

The crystal structure of the ArgS enzyme has been determined at 2.8 Å and 3.1 Å resolutions [ Zhou97b ]. Conformational changes induced by interactions with tRNA substrates have been studied [ Yao04 , Yao03 , Gu96 ], and specific interations with nucleotide residues in some tRNA species have been determined [ Kiga01 ]

Front 35

rplY

Back 35

The L25 protein is a component of the 50S subunit of the ribosome and binds to 5S rRNA.

Binding of L25 to 5S rRNA has been studied in detail [ Spierer78 , Huber84 , Speek82 , Kime83a , Moore83 , AbdelMegui83 , Kime84a , Kime84b , Szymkowiak85 , Shpanchenk96 ]. Solution structures of L25 alone and in complex with a segment of 5S rRNA has been determined [ Stoldt98 , Stoldt99 ], and a crystal structure of L25 bound to a fragment of 5S rRNA has been solved at 1.8 Å resolution [ Lu00 ].

A truncated form of L25 can be detected in 2-D gels [ Wasinger98 ]

Front 36

napF

Back 36

ferredoxin-type protein

The napF gene encodes a predicted 3Fe-4S iron sulfur protein. It is not essential for the activity of periplasmic nitrate reductase (Nap), but stimulates its activity [ Grove96 , Grove96a , Cole96 , Potter99a ].

NapF does not appear to be involved in the electron transfer from menaquinol or ubiquinol to Nap [ Brondijk02 ]. Loss of NapF alone causes a growth defect under anaerobic conditions on glycerol/nitrate medium; concurrent loss of NapG and NapH suppresses that defect. NapF may therefore play a role in energy conservation rather than a direct role in nitrate reduction [ Brondijk02 ]

Front 37

glpC

Back 37

glycerol-3-phosphate dehydrogenase (anaerobic), small subunit

Front 38

prmB

Back 38

N5-glutamine methyltransferase

PrmB is an N(5)-glutamine methyltransferase that methylates ribosomal protein L3 [ HeurgueHam02 ]. PrmB was shown to be involved in methylation of ribosomal protein L3 [ Lhoest77 ].

A prmB mutant that lacks methylation of L3 has a cold-sensitive growth phenotype and accumulates abnormal ribosomal particles [ Colson79 , Lhoest81 ]

PrmB and PrmC have sequence similarity to each other; PrmB is active toward ribosomal protein L3, whereas PrmC is active toward polypeptide chain release factors RF1 and RF2 [ HeurgueHam02 ]

Front 39

ppk

Back 39

polyphosphate kinase

The role of polyphosphate (poly(Pi)) in E. coli is still not fully understood; it may function in energy storage. Polyphosphate kinase (PPK) has several enzymatic activities. It catalyzes the synthesis of poly(Pi) through the transfer of a phosphoryl group of ATP to a poly(Pi) polymer. This reaction is readily reversible. PPK is most active with poly(Pi) substrates of chain lengths greater than 132 phosphoryl units. Activity decreases with decreasing chain length. In the reverse reaction ATP is synthesized from poly(Pi). PPK can also act as a nucleoside diphosphate kinase (NDK), converting GDP, CDP and UDP to their respective nucleoside triphosphates, in the reverse of the poly(Pi) synthesis reaction. Finally, GDP acts on a subterminal linkage of poly(Pi), transferring a pyrophosphoryl group and generating linear guanosine tetraphosphate (ppppG). PPK can also autophosphorylate. These various activities differ in their biochemical optima, subunit organization and in responses to chemical agents.

PPK has also been identified as a component of the E. coli RNA degradosome [ Blum97 ].

Mutants lacking PPK do not survive stationary phase. After a period of amino acid starvation PPK is important to the cell to increase intracellular protein degradation to provide amino acids for synthesis of new enzymes.

PPK can combine with adenylate kinase to catalyze the formation of ADP from AMP and poly(Pi). This polyP:AMP phosphotransferase (PAP) activity requires both enzymes acting together. [ Neidhardt96 , Ahn90 , Haeusler92 , Akiyama92 , Crooke94 , Kumble96 , Shiba97 , Kuroda97a , Rao96 , Van97a , Kuroda99 , Tzeng00 , Ishige00 ]

The basic functional unit of homotetrameric polyphosphate kinase is a dimer. However the enzyme has diverse functions involving different subunit organizations and conformations. When synthesizing ppppG the enzyme is a trimer and when autophosphorylating it is a tetramer. [ Tzeng00 ] Crystal structures have been determined for polyphosphate kinase on its own and binding AMP-PNP [ Zhu05a ]

Front 40

rnc

Back 40

RNase III is an endonuclease that cleaves double-stranded RNA to yield 5'-phosphates and 3'-hydroxyls [ Robertson75 ]. It is required for processing of ribosomal RNA (rRNA) and phage mRNA, for the regulation of a number of genes and for proper function of regulatory antisense RNAs, usually cleaving dsRNA created by the formation of stem structures within single-stranded RNA.

RNase III is a key enzyme in rRNA processing, cleaving 30 S precursor rRNA to yield 23 S, 17 S and 5 S rRNA, though it has been suggested based on degradation experiments that the 30 S RNA is not actually a precursor to these rRNAs [ Ginsburg75 , Nikolaev74 , Gegenheime75 ]. The 23 S rRNA sequence is flanked by RNase III cleavage sites and is cleaved at multiple points at its 5' end and one at its 3' end [ Bram80 , Sirdeshmuk85 , Stark85 ]. Some of the 5' cleavage of 23 S rRNA occurs even in the absence of RNase III [ Sirdeshmuk85a ]. Regions flanking the 16 S rRNA are predicted to form a 26-bp double helical stem at the base of a loop containing the 16 S rRNA. Rnase III cleaves at both sites that combine to form the stem [ Young78 ]. Electronic microscopy confirms that both the 16 S and 23 S rRNA sequences form loops with stems made from flanking regions that coincide with RNase III cleavage sites [ Edlind80 ]. 5 S rRNA is seperated from the 30 S precursor as a 7 S fragment, which is then cleaved at its 3' end by RNase III and finally by RNase E to yield the final 5 S rRNA [ Szeberenyi84 ]. In the absence of RNase III activity, mature 16 S rRNA can still be found, but no mature 23 S rRNA is detectable and unprocessed 23 S rRNA with extraneous sequence is incorporated in 50 S ribosomes [ King84 ]. This unprocessed 30 S rRNA can contribute elements to multiple ribosomes, resulting in pairs or occasionaly trios of 50 S ribosomes linked together by unprocessed rRNA [ Clark84 ].

Many types of phage RNA require RNase III processing. RNase III limits synthesis of the lambda lysogeny protein Int, possibly due to processing at RNase III sites within the int mRNA and in the nearby sib regulatory region [ Belfort80 , Schindler81 , Guarneros82 , Plunkett89 ]. RNase III also cleaves the lambda early N leader RNA, removing the NUTL binding site and thus preventing negative autoregulatory binding of N protein to NUTL, resulting in a 200-fold increase in N translation [ Wilson02 ]. RNase III processes T4 species I RNA and may be required for proper transcriptional termination as well [ Paddock76 ]. RNase III cleaves at multiple sites within the T7 phage genome, including the initiator RNA A3t and gene 1.2, which is subject to 3' processing [ Gross87 , Saito81 ]. RNase III cleavage of T7 RNA has been extensively evaluated at the R.1 site [ Li93 ]. Cleavage at R1.1 depends not on specific sequence information by on the asymmetry of the 4/5 loop at that site [ CalinJagem03 ]. A mutant with a bulge-helix-bulge motif replacing this asymmetric loop is bound but not hydrolyzed by RNase III [ CalinJagem03a ]. RNase III processes high molecular weight T3 RNA polymerase transcripts into discrete late T3 RNAs [ Majumder77 ]. MS2 phage RNA with an artificially-inserted RNase III stem-loop target undergoes selective pressure from RNase III, evolving variants with base mismatches, bulges and shortened stems, all of which are inviable RNase II targets [ Klovins97 ].

Polynucleotide phosphorylase (PNP) regulation depends on RNase III. The half life of pnp mRNA increases from 1.5 minutes to from 8 to 40 minutes in the absence of RNase III, leading to an up to eleven-fold increase in the amount of pnp mRNA [ Takata87 , Portier87 , Takata89 ]. Cleavage of the pnp transcript occurs in the 5' leader of the pnp mRNA, cleaving a stem-loop to leave a short 3' overhang [ Regnier86 , Jarrige01 , Portier87 , Takata89 , RobertLe92 ]. Following 5' cleavage by RNase III, the remaining mRNA is processed by RNase E [ Hajnsdorf94 ].

RNase III is also involved in processing quite a few other cellular mRNAs. It processes the end of the ribosomal protein-RNA polymerase rplJL-rpoBC operon [ Barry80 ]. Deletions around an RNase III site in between rplL and rpoB in the same operon substantially reduce the translation efficiency of β mRNA, though there is no observable effect of the absence of RNase III [ Dennis84 ]. RNase III cleaves in the leader of the secEnusG transcript and in the L12-β intercistronic region in the rplKAJLrpoBC transcript. Loss of RNase III stabilizes both mRNAs 1-2 fold without changing steady-state level, indicating a compensatory decrease in transcription [ Chow94 ]. Variants of the rpsO mRNA that are destabilized by polyadenylation can be cleaved by RNase III. The RNase III-cleaved mRNA is no longer destabilized by polyadenylation [ HaugelNiel96 ]. The 10 Sa precursor RNA is processed twice by RNase III before undergoing a final processing step to become the tmRNA that places the ssrA tag on incompletely translated proteins [ Srivastava90 , Makarov92 , Srivastava92 ]. RNase III cleaves the trp and lac operons and can inactivate lac α-peptide mRNA in vitro [ Shen81 , Shen82 ]. Primary transcripts of the metY-nusA-infB, rnc-era-recO and rpsO-pnp operons are all cleaved by RNase III in their 5' noncoding leader regions. All three mRNAs are degraded much more rapidly following processing [ Regnier90 ]. RNase III cleavage of the metY-nusA-infB operon also releases the metY minor initiator tRNA [ Regnier89 ]. RNase III cleaves the 5' untranslated region of both the rnc-era-recO and rnc-era transcripts, reducing their stability [ Matsunaga96 ]. The second stem-loop in the untranslated region is required for this cleavage [ Matsunaga96a ]. Cleavage of rnc increases its degradation rate ten fold [ Matsunaga97 ]. Three RNase III sites in the sdhCDAB-sucABCD intergenic region contribute to extreme instability of the operon transcript [ Cunningham98 ]. RNase III cleavage in the 5' untranslated region of the speF-potE transcript increases translational efficiency and thus enhances expression of ornithine decarboxylase [ Kashiwagi94 ]. Translation from the -292 start site of the adhE gene, which is needed for fermentative growth on glucose, requires RNase III. This may be due to a predicted intramolecular base pairing in the 5' untranslated region which would be expected to block ribosome binding [ MembrilloH99 , Aristarkho96 ]. RNase III cleaves DicF RNA to a precursor stage that is then cleaved by RNase E to its final form as a trans-acting cell division inhibitor [ Faubladier90 ]. Finally, bolA RNA is stabilized in the absence of RNase III [ Santos97 ].

Some antisense RNA regulatory systems are RNase III targets. The antisense RNA-OUT, which regulates IS10 transposition by binding the transposase mRNA RNA-IN, folds into a stable stem-loop structure that is RNase III resistant unless certain point mutations are introduced [ Case89 ]. However, RNase III cleavage is required for destabilization of RNA-IN following binding by RNA-OUT [ Case90 ]. Synthesis of the plasmid IncFII replication protein RepA is controlled by binding of the antisense RNA CopA to the leader region of RepA mRNA (CopT). RNase III cleaves the CopA/CopT duplex, limiting RepA expression [ Blomberg90 ]. Unpaired bases within the stem-loop in CopA reduce its suitability as a substrate for RNase III, limiting this regulatory cleavage [ Hjalt95 ]. Unstable antisense RNAs target for RNase III cleavage truncated mRNAs from the hok/sok (plasmid R1), srnB (F plasmid) and pnd (plasmid R438) plasmid toxicity systems. Absence of the truncated mRNAs leads to rapid degradation of the stable full-length mRNAs that code for toxins [ Gerdes92 ]. Sok antisense RNA forms a three-way junction with hok mRNA (which forms an internal stem-loop). This sok binding, along with the presence of a transcriptional terminator hairpin within sok, allows RNase III cleavage to occur [ Franch99 ].

RNAI, required for replication of the plasmid ColE1, is cleaved by RNase III, with polyadenylated RNAI collecting in the absence of RNase III [ Binnie99 ]. The plasmid R1 gene 19 is also controlled by RNase III, which cleaves within the coding region of gene 19 [ Koraimann93 ]. The copy number of miniR6-5 but not miniF plasmids is reduced 2-fold in the absence of RNase III [ Ely81 ].

Though Rnase III is dispensible for growth, mutants lacking its activity do not grow at 45 degrees C and are nonmotile [ Takiff89 , Apirion75 , Apirion78 ]. Suppressor mutations that cancel this temperature sensitivity have been discovered [ Apirion76 ]. mRNA is broken down equally quickly in cells with or without RNase III, but during carbon starvation RNA decays faster in cells lacking RNase III [ Apirion76a , Freire06 ]. In strains mutated in multiple ribonucleases, lack of RNase III led to more rapid decay of RNA generally [ Babitzke93 ]. In mutants lacking RNase III function, β-galactosidase induction from the lac operon is twice as slow as normal and yield ten times less enzyme. In vitro, β-galactosidase RNA from the mutant is three times less competent for translation initiation [ Talkad78 ]. Mutants in RNase III supress cold sensitive suhB mutants [ Inada95 ].

RNase III is a dimer, though a single functional active site is sufficient for catalytic activity [ Dunn76 , March90 , Conrad02 ]. The dominant-negative point mutant rnc70, which binds but does not cleave dsRNA, does not necessarily exert its negative effect on wild type through formation of mixed dimers [ Dasgupta98 ].

RNase III has two modules, an amino-terminal 150-residue catalytic domain and a carboxy-terminal 70-residue recognition module that is homologous to other dsRNA binding domains. By NMR, RNase III has an α-beta-beta-beta-alpha topology, with a three-stranded beta sheet packing on two alpha helices [ Kharrat95 ]. Both domains play a part in substrate specificity and cleavage site selection, though truncated RNase III lacking the RNA binding domain still cleaves small substrates in vitro and is specific for dsRNA [ Conrad01 , Sun01 ]. Changing the interdomain spacer from nine to twenty residues has no effect on RNase III activity [ Conrad01 ]. The RNase III catalytic module has one processing center with two mRNA cleavage sites, generating products with two nucleotide 3' overhangs [ Zhang04 ]. Cleavage involves incorporation of a solvent oxygen atom into the 5'-phosphate of the RNA product, making cleavage irreversible. The catalytic rate depends on the divalent metal ion used in the reaction. A two-step change in fluorescence may indicate conformational change during distinct substrate binding and catalytic steps [ Campbell02 ]. Experiments suggest that two magnesium ions are involved in catalysis [ Sun05 ].

Though there are no canonical motifs for RNase III cleavage, certain base pairings are unfavorable, disrupting proper RNase III binding and thus preventing subsequent cleavage. Relative to the cleavage site, GC/CG is never seen at -5, is rarely seen at -4 and -6 and is never seen at -12. UA is never seen at -11. Addition of these unfavored base pairs in the stated locations in T7 R1.1 RNA disrupted RNase III binding. The three base-pair antideterminant sequence from tRNASec, which prevents interaction with incorrect aminoacyl-tRNA-synthetases, also disrupts RNase III binding when placed in the critical -5 and -12 positions, but not when placed in the intervening region [ Zhang97 ]. RNase III cleavage does not depend on tertiary RNA-RNA interactions or a conserved CUU/GAA base pair sequence [ Chelladura91 ]. RNase III recognizes RNA duplexes longer than eleven base pairs with little specificity [ Lamontagne04 ]. RNase III is still able to cleave at the expected site in R1.1 RNA even when a phosphorothioate internucleotide linkage is present [ Nicholson88 ].

RNase III activity is stimulated four-fold following infection by T7 phage. The T7 protein kinase gp0.7 PK phosphorylates RNase III at a serine, stimulating processing of T7 early and late mRNAs [ Mayer83 , Robertson94 ]

Front 41

truD

Back 41

TruD is responsible for biosynthesis of the pseudouridine13 modification of tRNA(Glu) [ Kaya03 ]. The enzyme has been purified, and activity has been observed in vitro [ Kaya03 ].

A truD mutant lacks tRNA(Glu) pseudouridine13 [ Kaya03 ]. A conserved Asp is required for catalysis [ Kaya03 ].

TruD is a member of a family of pseudouridine synthases that is widely conserved, yet unrelated to the families of pseudouridine synthases characterized previously [ Kaya03 ]. The crystal structure showed that the protein consists of two domains, a catalytic domain with a fold similar to other pseudouridine synthases, and an insertion domain with a novel fold [ Kaya04 , Ericsson04 , Ericsson04a , Hoang04 ]

Front 42

iap

Back 42

alkaline phosphatase isozyme conversion protein

Alkaline phosphatase isozyme conversion protein is responsible for the stepwise removal of the two amino-terminal arginines from alkaline phosphatase isozyme 1, creating isozymes 2 and 3 in the process [ Nakata77 ]. This is presumably an aminopeptidase activity, as it can be inhibited by protease inhibitors [ Nakata79 , Nakata84 ].

Alkaline phosphatase isozyme conversion protein has a characteristic signal peptide at its amino-terminus and can be found as both full-length and signal-peptide-cleaved forms in the cell. It is a membrane-associated protein that acts in the periplasm [ Ishino87 ].

Efficieny of secretion of alkaline phosphatase isozyme conversion protein is reduced when phosphatidylethanolamine is depleted, resulting in less effective modification of alkaline phosphatase [ Golovastov02 ]

Front 43

prfB

Back 43

peptide chain release factor RF2

Release factor 2 (RF2) is one of two class 1 codon-specific factors in E. coli that facilitate the release of the growing polypeptide chain at stop codons. RF2 recognizes the termination codons UGA and UAA [ Capecchi67 , Scolnick68 , Capecchi70 , Beaudet70 ]. Termination is highly precise; RF2 is able to discriminate against related sense codons by 3 to 6 orders of magnitude and without requiring energy-driven error correction [ Freistroff00 ]. The discriminator site consists of a tripeptide which appears to be functionally equivalent to the anticodon of tRNA [ Ito00 ]. An E. coli cell is variously estimated to contain between 700 and 24,900 molecules of RF2 and reaches the highest RF2 content at a high growth rate [ Klein71 , Adamski94 ].

A +1 translational frameshift allows translation of the prfB mRNA past a premature in-frame stop codon at position 26; the frameshift was confirmed by both DNA and peptide sequencing data [ Craigen85 ] as well as in vitro [ Craigen86 ]. The natural frameshift efficiency was variously estimated to be between 18% [ Mikuni91 ] and 48% [ Curran93 ] and strongly depends on the identity of the codon preceding the in-frame stop codon [ Curran93 ]. The rate of frame-shifting increases in a prfB mutant, indicating autoregulation of the process [ Mikuni91 , Donly90 ]. Translation termination, frameshifting, and readthrough by a suppressor tRNA directly compete at the prfB frameshift site [ Adamski93 ]. High-level frameshifting is coupled with premature release of the E site tRNA from the ribosome [ Marquez04 ].

Interactions between RF2 and the ribosome have been mapped [ Moffat91 ]. The universally conserved tripeptide sequence GGQ is thought to interact with the peptidyl-transferase center of the ribosome and may be functionally equivalent to the CCA end of tRNA [ Zavialov02 , Mora03 ]. RF2 is post-translationally modified by N5 methylation at Gln252 within the GGQ sequence; methylation increases the termination efficiency of the protein. Overproduced RF2 lacks this modification, which may be one cause for the lethality of overproducing RF2 [ DincbasRen00 , Nakahigash02 ]. Interactions of the conserved motifs within RF2 with the ribosome have been determined [ Scarlett03 ]. The N-terminal domain of RF2 is required for normal activity [ Mora03a ].

Certain mutations in prfB lead to suppression of the UGA stop codon and are conditionally lethal [ Kawakami88 ]. The RF2* mutant (E167K) facilitates termination at all three stop codons [ Ito98 ].

A crystal structure of RF2 has been solved at 1.8 Å resolution; the structure is not similar to that of eukaryotic peptide release factor [ Vestergaar01 ]. A single-particle cryo-electron microscopy structure of RF2 in a post-termination complex with the ribosome shows that RF2 adopts a different conformation on the ribosome which is consistent with functional data [ Rawat03a , Klaholz03 ].

Reviews: [ Craigen87 , Kisselev03 ]
A +1 translational frameshift allows translation of the prfB mRNA past a premature in-frame stop codon at position 26; the frameshift was confirmed by both DNA and peptide sequencing data [ Craigen85 ], as well as in vitro testing [ Craigen86 ]. The frameshifting event is encoded as a 1 bp "intron" in EcoCyc, which allows us to show the correct protein sequence; however, no splicing occurs.

Front 44

pepP

Back 44

proline aminopeptidase P II

Aminopeptidase P is an exopeptidase that cleaves the amino-terminal residue from polypeptides, as long as the following residue is a proline [ Yoshimoto88 , Yoshimoto94 ].

Aminopeptidase P comprises four PepP monomers, arranged as a dimer of dimers [ Yoshimoto88 , Wilce98 ]. Kinetic and structural analysis of PepP has identified residues comprising the active site as well as parts of its structure that are required for substrate specificity [ Graham06 ]

Front 45

hybD

Back 45

predicted maturation peptidase for hydrogenase 2

Based on its similarity to the processing enzyme for hydrogenase 3, HycI, HybD is predicted to be an endopeptidase involved in the maturation of the large subunit of hydrogenase 2 (HybC) [ Rossmann95 ]. It is expected to cleave off a 15 amino acid peptide from the carboxy-terminus of the immature HybC protein. The crystal structure of HybD has been determined at 2.2 Å resolution [ Fritsche99 ]

Front 46

hybO

Back 46

hydrogenase 2, small subunit

HybO is the small subunit of hydrogenase 2, and it contains three Fe-S centers [ Sargent98a ]. Hydrogenase 2 is associated with the periplasmic side of the cytoplasmic membrane [ Ballantine86 , Rodrigue96 ]. HybO contains a twin-arginine signal sequence which is required for membrane targeting by the Tat system [ Sargent98a ].

HybC and HybO are coordinately assembled and processed; the presence of both subunits, nickel acquisition and the subsequent processing of HybC are required for export of both subunits by the Tat system [ Rodrigue96 , Rodrigue99 ]

Front 47

cca

Back 47

fused tRNA nucleotidyltransferase / 2',3'-cyclic phosphodiesterase / 2' nucleotidase and phosphatase

tRNA nucleotidyltransferase catalyzes the addition of CCA to the 3' ends of tRNAs in a template-independent manner. Phosphatase, nucleotidase, and phosphodiesterase activities of CCA were discovered in a high-throughput screen of purified proteins, and these activities may act together to repair the 3' end of tRNA [ Kuznetsova05 ].

Polymerization of the CCA sequence does not appear to involve translocation of the enzyme [ Shi98 ]. A model for the specific addition of CCA involving binding sites for ATP and CTP has been proposed [ Tomari00 , Seth02 ].

The N-terminal domain of the enzyme contains the nucleotidyltransferase activity. The C-terminal domain carries an HD motif found in a family of metal-dpendent phosphohydrolases and displays a variety of phosphohydrolase activities in the absence of tRNA [ Yakunin04 ]. The C terminus of the protein contains an "anchor domain" of 27 amino acids that defines the addition of CCA versus poly(A), the activity of poly(A) polymerase [ Betat04 ].

tRNA nucleotidyltransferase is not essential for viability of E. coli; a cca null mutant has a slow growth phenotype [ Zhu87a ]. Poly(A) polymerase I and polynucleotide phosphorylase can partially substitute for tRNA nucleotidyltransferase in the repair of the CCA end of tRNAs [ Reuven97 ]

Front 48

pnp

Back 48

polynucleotide phosphorylase

Polynucleotide phosphorylase (PNPase) is a 3' to 5' exonuclease and a 3'-terminal oligonucleotide polymerase. It degrades various mRNAs, is involved in cold shock regulation, is a part of tRNA maturation and degradation, adds heteropolymeric tails to some RNAs and is a component of the degradosome, a multienzyme complex that carries out RNA degradation.

PNPase is involved in general mRNA degradation. Loss of PNPase leads to an increase in steady-state levels of mRNA, as well as increasing mRNA half lives in the absence of the 3' exonuclease RNase II [ Mohanty03 , Kinscherf75 ]. PNPase also has a role in mRNA degradation during carbon starvation, where it may be required for breakdown of small rRNA fragments produced by other RNases [ Kaplan74 , Kaplan75 ].

A number of specific PNPase substrates have been identified. PNPase is involved in degradation of lac mRNA, rnb mRNA, mRNA coding for ribosomal protein S20, and the RNA-OUT antisense RNA [ HarEl79 , Pepe94 , Mackie89 , Zilhao96 ]. It also degrades sok antisense RNA and thrS and rpsO mRNA following cleavage by RNase E [ Dam97 , Nogueira01 , Braun96 , Hajnsdorf94 ]. PNPase binds to but does not degrade RNA containing 8-oxoguanine [ Hayakawa01 ].

PNPase-mediated degradation is required for regulation of the cold shock response. PNPase degrades a number of mRNAs induced by cold shock, including those coding for CspA, RbfA, CsdA, RpoE, RseA, Rnr and many others [ Yamanaka01b , Cairrao03 , Polissi03 ]. The isolated PNPase S1 RNA-binding domain can complement a deletion in four cold-shock genes [ Xia01a ].

The 3' to 5' processive cleavage of RNA by PNPase depends on the composition and structure of the 3' end of the substrate [ Plamann90 , Cisneros96 ]. RhlB and poly(A) polymerase I (PAP I) in concert with the degradosome are required for PNPase-mediated degradation of cistrons with 3' REP-stabilizers [ Khemici04 ].

Binding of the protein Hfq to poly(A) tracts prevents PNPase degradation of these tails in vitro [ Folichon03 ]. RNA with 3' stem-loops are resistant to degradation by pure PNPase or whole degradosome in vitro, but addition of even a short poly(A) or mixed nucleotide tail overcomes this block [ Causton94 , Blum99 , Lisitsky99 ]. Polyadenylation similarly destabilizes rpsO mRNA against degradation by RNase E, RNase II and PNPase, and is required for sok RNA degradation [ Hajnsdorf95 , Hajnsdorf96 , Dam97 ]. Both 3' adenylation and 5' phosphorylation affect the rate of degradation of RNA I [ Xu95g ]. PNPase itself modulates polyadenylation several RNAs [ Mohanty00 ].

PNPase is involved in tRNA processing and maintenance. Though purified PNPase is incapable of completely processing tRNA in vitro, it is effective, along with RNase II, in trimming long 3' trailing sequences to yield 2-4 nucleotide intermediates which will be trimmed by RNases T and PH [ Deutscher88 , Li94 ]. PNPase is also partially required for repair of 3'-terminal CCA sequences in tRNAs in the absence of tRNA nucleotidyltransferase [ Reuven97 ]. PNPase is also involved in the degradation of mutant tRNA, in a process that is enhanced by polyadenylation by PAP I [ Li02f ].

PNPase also catalyzes the "reverse" reaction, converting nucleoside diphosphates into polyribonucleotides [ Littauer57 , Gillam78 , Gillam80 ]. PNPase generates heteropolymeric tails on RNA and is responsible for residual polyadenylation detected in PAP I deficient strains [ Mohanty00a ]. Hfq, which binds to the 3' end of RNA and prevents PNPase-mediated degradation, also prevents PNPase-mediated addition of nucleosides to bound RNA, while promoting PAP I activity [ Folichon05 ].

PNPase is a trimer of Pnp monomers [ Portier75 , Soreq77 ]. Each Pnp monomer has two RNA-binding sites, KH and S1, that are dispensible for strict catalytic function but are required for Pnp autoregulation, growth at low temperature, and the generation of oligonucleotides [ Jarrige02 , MatusOrteg07 , Guissani76 ]. The S1 domain is a five-stranded antiparallel β barrel with conserved residues on one face forming the RNA binding site [ Bycroft97 ].

PNPase is subject to autoregulation at the mRNA level. RNase III cleaves a stem-loop in the pnp mRNA leader sequence, following which PNPase binds and degrades the 5' half of the cleaved duplex [ Portier87 , Takata89 , Jarrige01 , RobertLe92 , Takata87 , Carzaniga09 ]. PNPase autoregulation also decreases as general RNA polyadenylation increases and following a shift to cold temperatures [ Mohanty02 , Mathy01 , Zangrossi00 , Beran01 ].

Strains lacking both PNPase and RNase II activity are inviable and collect mRNA fragments 100-1,500 nucleotides long [ Donovan86 ]. In a triple mutant in pnp, rnb and rne, mRNA degradation slows three- to fourfold and the length and number of poly(A) tails increases [ Arraiano88 , OHara95 ]. In a pnp mutant lacking RNase PH function, the 50S ribosomal subunit and 23S rRNA is degraded [ Zhou97a ].

Even in the absence of the degradosome scaffold RNase E, PNPase and the helicase RhlB interact. In vitro, RhlB unwinding of dsRNA allows PNPase degradation to occur [ Liou02a ].

PNPase is required to prevent phage P4 superinfection. This prevention requires binding of CI antisense RNA to sequences on nascent P4 transcripts; PNPase processes CI RNA [ Piazza96 ]

Front 49

infB

Back 49

IF-2 is one of three translation initiation factors in E. coli; it ensures the correct binding of fMet-tRNAfMet in the ribosomal P site.

EF-2 has ribosome-dependent GTPase activity [ Dubnoff72 ] with approximately 10 times higher affinity for GDP than for GTP [ Pon85 ]. Domain IV of EF-2 is the GTP-binding domain.

IF-2 exists in three isoforms, IF-2α (IF2-1), IF-2β (IF2-2), and IF-2β' (IF-2γ, IF2-3) [ Miller73 , Nyengaard91 , MorelDevil90 , Plumbridge85 , Sacerdot92 ], which are generated by the use of alternative in-frame translation initiation codons: IF2-1 is the full-length protein, GUG at position 158 is used for IF2-2, and AUG at position 165 is used for IF2-3 [ Mortensen95 ]. Both the α and β forms are required for optimal growth; the isoforms thus have acquired specialized, although not essential functions [ Sacerdot92 ].

A solution structure of the N-terminal domain I of IF-2 (IF2N) has been solved [ Laursen03 ]. A flexible linker region connects this domain to the conserved C-terminal domains of IF-2 [ Laursen04 ]. A cryo-EM reconstruction of the initiation complex shows a conformation of IF-2 that is different from that seen in the crystal structure of the Methanobacterium thermoautotrophicum protein [ Allen05a ]

Front 50

sspB

Back 50

ClpXP protease specificity-enhancing factor

The SspB protein is a specificity-enhancing factor for the ClpXP protease [ Levchenko00 ]. When protein synthesis is stalled, incomplete proteins that are produced are tagged with the small SsrA peptide. The ribosome-associated SspB protein binds to the SsrA tag and enhances degradation of the tagged peptide by the ClpXP protease [ Levchenko00 ].

The SspB protein forms a homodimer with two independent binding sites for SsrA-tagged proteins [ Wah02 ]. It also binds to ClpX and stimulates its ATPase activity [ Wah02 ]. The dimerization and SsrA binding domain resides in the amino terminal 110-120 residues of SspB, while the C-terminal 40-50 residues are required for association with ClpXP and stimulation of its ATPase activity [ Wah03 ]. Efficient ClpX hexamer binding and substrate delivery requires both C-terminal domains of the SspB dimer [ Bolon04 ]. Interactions between binding of SspB and ClpX to the SsrA tag have been described [ Hersch04 ].

Protein degradation substrates regulated by SspB have been identified and include RseA, which contains an SspB binding site that is unrelated to the SsrA tag sequence [ Flynn04 ]. Degradation of the RseA cytoplasmic fragment is the last step in the proteolytic cascade leading to the induction of the sigma E extracytoplasmic stress regulon.

Crystal structures of SspB alone and in complex with the SsrA peptide tag have been determined at 2.2 and 2.9 A resolution, respectively [ Song03 ], and a crystal structure of SspB in a complex with its recognition peptide in RseA has been determined at 1.8 A resolution [ Levchenko05 ]. The crystal structures reveal diversity in the recognition of different target proteins.

The level of SspB protein remains constant throughout the transition from exponential growth to early stationary phase [ Farrell05 ].

ssp: "stringent starvation protein"

Front 51

tufA

Back 51

Elongation factor Tu (EF-Tu) is the most abundant protein in E. coli. In its GTP-bound (active) form, EF-Tu binds aminoacylated tRNAs to form the so-called ternary complex. At the decoding site of the ribosome, the ternary complex is "tested" for a codon-anticodon match; if the proper aminoacyl-tRNA has been found, GTP is hydrolyzed and EF-Tu and GDP dissociate from the ribosome, while the aminoacyl-tRNA remains bound to the ribosome.

Proteomic analyses indicate that EF-Tu is maximally expressed during stress response [ Muela08 ].

In E. coli, EF-Tu is encoded by two genes, tufA and tufB.

Front 52

rplD

Back 52

The L4 protein is a component of the 50S subunit of the ribosome and also regulates the expression of S10 operon at both the transcriptional and posttranscriptional level. The functions of L4 in ribosome assembly, attenuation, and translational regulation of the operon are separable [ Freedman87 , Li96c ].

Addition of L4 to an in vitro protein synthesis reaction inhibits the synthesis of the promoter-proximal proteins in the S10 operon, suggesting that L4 may be an inhibitor of translation [ Yates80 ]. In vivo, overproduction of L4 was found to reduce the synthesis of S10 operon mRNA [ Zengel80 ]. The steady-state growth rate-dependent control of the S10 operon expression requires additional factors [ Lindahl90 ].

L4 stimulates premature termination (transcription attenuation) at a NusA-dependent terminator site 30 bases upstream of the first structural gene of the S10 operon, and termination requires the function of NusA [ Lindahl83 , Zengel90 ]. The attenuator hairpin region is sufficient for NusA-dependent pausing, and a second hairpin region immediately upstream of the attenuator hairpin is necessary for L4 to prolong the pause [ Zengel92 , Sha95 , Zengel96 ]. Structural and sequence requirements for L4-mediated transcription termination within the S10 leader region have been determined [ Zengel02 ]; the first 150 bases of the S10 leader region are sufficient for L4-mediated attenuation control [ Zengel90a ], and a 64-nucleotide sequence is required for L4 interaction with the S10 leader mRNA [ Stelzl03 ]. Binding of L4 to the leader region induces structural changes in the mRNA [ Stelzl03 ]. A region of 110 bases within domain I of 23S rRNA competes with paused transcription complex for binding of L4 [ Zengel91 , Zengel93 ].

L4 interacts with the 5' segment of 23S rRNA [ Spierer75 , Gulle88 , Maly80 , Urlaub95 , Thiede98 ]. The L4 and L24 binding sites in 23S rRNA localize to a small fragment [ Stelzl00 ] and may be a key element for rRNA folding in the early assembly pathway for the ribosome [ Stelzl01 ]. The L4 binding region within domain I of 23S rRNA has three-dimensional structural similarity to its binding region on the S10 operon leader mRNA where it inhibits its own translation [ Ostergaard98 , Stelzl03 ].

L4 is a component of the binding site for erythromycin on the ribosome [ Arevalo88 ]. An L4 mutant is resistant to erythromycin [ Apirion67 , Wittmann73 ]; this mutant has a cold-sensitive growth defect, and 50S subunit assembly is defective [ Chittum94 ]. L4, L5, and L21 bind to erythromycin cooperatively [ Pye90 ]. The extended loop region of L4 (amino acids 40-88) contributes to the lining of the peptide exit tunnel in the ribosome and is contacted by the growing peptide chain [ Houben05 ]. While deletions of this region of L4 do not appear to affect function of L4 [ Zengel03 ], the K63E point mutation that results in erythromycin resistance alters the decoding accuracy of the ribosome [ OConnor04 ]

Front 53

rplW

Back 53

The L23 protein is a component of the 50S subunit of the ribosome and provides a chaperone docking site that links protein biosynthesis with protein folding. L23 is essential for growth of E. coli [ Kramer02 ].

L23 crosslinks to Trigger Factor (TF), a protein that interacts with nascent polypeptides on the ribosome, and is essential for the association of TF with the ribosome [ Kramer02 , Ullers03 ]. It also crosslinks to the Ffh component of the Signal Recognition Particle (SRP) [ Gu03a , Ullers03 ]. Kd values for binding of TF and SRP to the ribosome under various conditions have been estimated, and TF and SRP are thought to have separate binding sites on L23 [ Raine04 ]. L23 can also be crosslinked to a nascent peptide chain [ Houben05 ].

L23 is an early assembly protein that interacts with 23S rRNA [ Marquardt79 , Osswald90 , Egebjerg91 , Chistyakov89 , Tumminia94 , Thiede98 , Skold81 , Vester84 , Thiede98 ]. L23 crosslinks to tRNA in the A site [ Graifer89 ]. L23 can also be crosslinked to L34 [ Walleczek89 , Walleczek89a ] and L29 [ Walleczek89a ].

L23 is photoaffinity labeled by puromycin, thus placing it within the A-site domain of the peptidyl transferase center [ Cooperman75 , Jaynes78 , Weitzmann85 ]. Puromycin-crosslinked L23 interferes with tRNA binding, but 50S subunits containing it retain peptidyl transferase activity [ Weitzmann90 ]

Front 54

rpmJ

Back 54

The L36 protein is a component of the 50S subunit of the ribosome [ Wada87 ]. L36 is highly conserved in bacteria, mitochondria and chloroplasts, but not present in archaea and eucarya [ Maeder05 ].

Ribosomes lacking L36 are correctly assembled. However, chemical protection experiments suggest that rRNA tertiary interactions are disrupted in ribosomes lacking L36, thus arguing that L36 plays a significant role in organizing 23S rRNA structure [ Maeder05 ]. L36 has ben shown to crosslink to 23S rRNA [ Urlaub95 ].

Deletion of rpmJ causes a temperature-sensitive growth defect [ Ikegami05 , Maeder05 ]; the rpmJ deletion led to decreased expression of secY, located immediately upstream of rpmJ. rpmJ does not appear to be essential for protein synthesis [ Ikegami05 ]

Front 55

rpoA

Back 55

RpoA is the α subunit of the RNA polymerase core enzyme. It consists of two domains connected by a flexible linker.

The RpoA amino-terminus is both necessary and sufficient for dimerization of RpoA and subsequent assembly of the RNA polymerase core complex [ Zhang98h ]. The amino-terminus has been analyzed both by NMR and via a 2.5 Å resolution cystral structure [ Otomo00 , Zhang98h ].

The amino-terminal and carboxy-terminal domains of RpoA are connected by a flexible linker, which has been shown to affect transcription in a promoter-dependent fashion [ Jeon97 , Fujita00 , Meng00 ].

The carboxy-terminal domain of RpoA is involved in antitermination, rho-dependent termination, and is a target for interactions with transcription termination/antitermination L factor that control termination and pausing [ Schauer96 , Kainz98 ]. Interaction with the RpoA carboxy-terminal domain activates RNA binding by transcription termination/antitermination L factor [ Mah00 ]. The RpoA carboxy-terminal domain is also required for some kinds of transcriptional activation and plays a role in some transcriptional initiation [ Zou97 , Burns99 ]

Front 56

rplQ

Back 56

The L17 protein is a component of the 50S subunit of the ribosome.

L17 interacts directly with tRNA [ Metspalu81 ] and can be crosslinked to 23S rRNA [ Osswald90 ]. L17 can also be crosslinked to the spiramycin derivative dihydrospiramycin and may thus be located near the peptidyl transferase center [ Bischof95 ]

Front 57

fis

Back 57

Fis, "factor for inversion stimulation", is a small DNA-binding and bending protein whose main role appears to be the organization and maintenance of nucleoid structure through direct DNA binding and by modulating gyrase [ Cho08a , Schneider01 ] and topoisomerase I production [ WeinsteinF07 ], as well as regulation of other proteins that modulate the nucleoid structure, such as CRP, HNS, and HU. Fis directly modulates several cellular processes, such as transcription, chromosomal replication, DNA inversion, phage integration/excision, and DNA transposition [ Finkel92 , Travers01 ].

As a transcriptional regulator, Fis regulates the expression of many genes involved in translation (rRNA and tRNA genes), virulence, biofilm formation, energy metabolism, stress response, central intermediary metabolism, amino acid biosynthesis, transport, cell structure, carbon compound metabolism, amino acid metabolism, nucleotide metabolism, motility, and chemotaxis [ Bradley07 , Finkel92 , Sheikh01 ]. A transcriptome analysis has shown that transcription of approximately 21% of genes is modulated directly or indirectly by Fis [ Cho08a ].

Fis, together with proteins such as HNS, HU, IHF, and Dps, is one of the largest components of the nucleoid. A ChIP-chip analysis has shown that Fis binds to 894 DNA regions in the genome, which results in two sites of Fis per supercoiling domain. These regions include both intergenic regions and regions within genes, and not all genes to which Fis binds are affected by this protein, a fact that is in agreement with the primary role assigned to Fis in maintenance of nucleoid structure [ Cho08a ].

Under optimal growth conditions, Fis is the dominant DNA-binding protein in the cell. Up to 60 000 copies of the protein can be found in a single cell under log phase, but it is nearly imperceptible in stationary phase (<100 molecules per cell) [ Ali99 ]. Fis bends the DNA between 40 and 90° [ Finkel92 ]. This bending stabilizes the DNA looping to regulate transcription and to promote DNA compaction [ Skoko06 , Travers98 ].

The crystal structure of Fis has been solved by several research groups [ Cheng00a , Yuan91 , Kostrewa91 , Safo97 ], and they have shown that this protein has an α-helical core of four helices (A to D). The C-terminal domain contains the C and D helices, forming a helix-turn-helix motif, characteristic of many DNA-binding proteins. The N-terminal domain has the A helix and a flexible β-hairpin arm, which are involved in DNA inversion; this region appears to make contact with the DNA invertase Hin in the invertasome structure [ Safo97 , Finkel92 ].

As expected for a gene involved in the modulation of many cellular processes, expression of fis is regulated by several systems and at different levels. At the transcription level, Fis is autoregulated, induced by high supercoiling levels [ Schneider00a ], and regulated by both growth rate-dependent and stringent control systems that require the presence of a GC motif downstream of the -10 region [ Ninnemann92 ].

Transcription of fis is also regulated by the availability of the nucleotide triphosphate CTP, which is the nucleotide with which transcription of fis is initiated and whose largest concentration is seen during log phase, a fact that correlates with the pattern of fis expression. If transcription of this gene is initiated with an A or a G instead of a C, transcription is induced at the same level in early stationary phase rather than in log phase [ Walker04 ].

A promoter that does not transcribe any ORF is located upstream of the promoter that transcribes fis and in the opposite direction. Under RNA polymerase limitation, both promoters compete for the protein. This might be a type of fis regulation expression, but there is no evidence to support this claim [ Nasser01 ].

DksA, a protein that is associated with RNA polymerase in regulating transcription, inhibits transcription of fis by increasing the inhibitory effects of ppGpp, decreasing the lifetime of the RNA polymerase-fis promoter complex, and increasing the sensitivity to the CTP nucleotide [ Mallik06 ].

BipA is a protein required for fis translation; this protein appears to destabilize the strong interaction between the 5� end of the untranslated region of the fis mRNA and the ribosome [ Owens04 ].

In a specific way and with tight binding affinity, Fis recognizes and binds as a dimer to sites that have poorly related sequences. However, a very degenerated consensus sequence has been proposed for Fis sites, with only a subset of common nucleotides. This consensus sequence shows a core binding site of 15 bp with partial dyad symmetry [ Finkel92 ]. This sequence presents only four highly conserved nucleotides, a G and a C at the 1st and at the 15th nucleotide, respectively, and a pyrimidine (A or G) and a purine (T or C) at the 5th and 11th positions, respectively. The central region commonly presents an AT-rich sequence [ Shao08a , Hengen97 ]. Some of these sites have internal sequences that are potential methylation targets, 5�-GATC-3�, suggesting that under some circumstances Fis binding is controlled by methylation [ Weinreich92 ]

Front 58

prfC

Back 58

Release factor 3 (RF3) is a ribosome-dependent GTPase that stimulates the release of RF1 and RF2 from the ribosome after peptide chain termination. The action requires nucleotide exchange and hydrolysis of GTP.

It was initially found that RF3 stimulates the release of the growing polypeptide chain at stop codons; unlike RF1 and RF2, it is not codon-specific [ Capecchi69 , Goldstein70 ]. Rather, RF3 stimulates the activities of RF1 and RF2 [ McCaughan84 , Mikuni94 ]. It was reported that RF3 increases the affinity of RF2 to the UGA termination complex, but not that of RF1 to the UAG termination complex [ Grentzmann95 ]. However, later experiments showed that RF3 and GTP catalyzes the removal of either RF1 or RF2 from the ribosome after termination [ Freistroff97 ].

In its free form, RF3 binds GDP much more strongly than GTP; thus, RF3 enters the ribosome in its GDP-bound form. The ribosomal complex containing RF1or RF2 and hydrolysis of peptidyl-tRNA by RF1 or RF2 stimulates nucleotide exchange in RF3. RF3-GTP then stimulates the release of RF1 or RF2 from the ribosome. GTP hydrolysis subsequently allows release of RF3 itself from the ribosome [ Zavialov01 ]. Removal of the terminated polypeptide chain is the essential step regulating the sequence of steps during termination [ Zavialov02 ]. The N-terminal domain of RF1 and RF2 is required for interaction with RF3 and stimulating nucleotide exchange [ Mora03a ].

Cryo-electron microscopy studies of RF3 trapped on the ribosome show two different states of RF3 binding and involving a large conformational rearrangement of the ribosome itself [ Klaholz04 ].

A transposon insertion mutant in prfC leads to suppression of UGA stop codons; this phenotype allowed cloning of the prfC gene [ Mikuni94 ]. Another transposon insertion mutant suppresses a leaky miaA mutation [ Connolly91 , Grentzmann94 ]. Inactivation of prfC stimulates synthesis of RF2 [ Grentzmann95 ].

Tos: "transposon-induced opal suppressor" [ Mikuni94 ]

Front 59

kptA

Back 59

KptA exhibits tRNA 2'-phosphotransferase activity; however, this substrate is not likely to have physiological relevance in E. coli [ Spinelli99 ],[ Steiger01 ].

KptA has similarity to Saccharomyces cerevisiae Tpt1p, which is a tRNA splicing enzyme [ Spinelli99 ], [ Steiger01 ]

Front 60

rplI

Back 60

The L9 protein is a component of the 50S subunit of the ribosome.

A solution structure determined by cryo-EM showed that the L9 protein consists of N-terminal and C-terminal domains connected by a long α helical domain that is wrapped around the L1 stalk [ Matadeen99 ].

The binding region of L9 on domain V of 23S rRNA was mapped [ Ostergaard98 ]. The N- and C-terminal domains of L9 interact with two separate sites in domain V of 23S rRNA [ Adamski96 , Lieberman00 ]. Crosslinking of tRNA in the A site of the ribosome to L9 depends on the state of the tRNA [ Graifer89 ]

Front 61

rnr

Back 61

RNase R is a ribonuclease that has been implicated in rRNA maturation, as well as mRNA regulation during stationary phase.

RNAse R is a processive 3' to 5' exoribonuclease [ Cheng02 ]. Although it can cleave many RNA substrates, it shows greatest activity toward rRNA [ Cheng02 ].

RNAse R may be involved in the maturation of the small stable RNA SsrA [ Cairrao03 ]. It also downregulates ompA mRNA abundance [ Andrade06 ].

Rnr expression is induced 7- to 8-fold by cold shock, and is also induced during stationary phase growth [ Cairrao03 , Andrade06 ].

Rnr of enteroinvasive E. coli, Shigella flexneri, and other Shigella is important for the virulence of these organisms [ Tobe92 ]. Shigella flexneri Rnr plays a posttranscriptional role in the production of virulence proteins [ Tobe92 ]

Front 62

efp

Back 62

EF-P is an elongation factor that stimulates the synthesis of peptide bonds [ Glick75 , Glick79 , Ganoza85 , Green85 , Baxter87 , Ganoza00 ]. It may function to facilitate proper positioning of fMet-tRNAfMet for initial peptide bond formation [ Aoki08 , Blaha09 ]. The EF-P binding site on the ribosome may overlap the peptidyltransferase center [ Aoki97 ]; interactions between EF-P and 70S ribosomes have been mapped [ Aoki08 ].

EF-P is essential for viability [ Aoki97a ]. E. coli contains approximately one molecule of EF-P per 10 ribosomes [ An80 ]. The EF-P-stimulated synthesis of peptide bonds is one of several targets of the oxazolidinone class of antibiotics [ Aoki02 ].

EF-P may form a complex in solution; its molecular weight as measured on a Sephadex column is 50 kD [ Glick75 ]. The activity of EF-P is dependent on Mg2+ ions [ Glick75 ]. EF-P is phosphorylated in bacteriophage T7-infected cells [ Robertson94 ]. The Lys34 residue was found to be modified; the modification was thought to be a spermidine residue [ Aoki08 ], but was later predicted to be a β-lysine residue [ Bailly10 ]. Further experiments showed that Lys34 is modified with a lysyl residue by YjeA, and that this modified form is the active form in vivo [ Yanagisawa10 ]

Front 63

fdhF

Back 63

Formate dehydrogenase-H is one of three membrane-associated formate dehydrogenase isoenzymes in E. coli [ Sawers94 ]. All are functional in the anaerobic metabolism of the organism. Formate dehydrogenase-H (FDH-H) is located in the cytoplasm and together with hydrogenase-3, FDH-H forms the formate-hydrogen lyase complex [ Axley90 , Sawers94 ].

The enzyme is oxygen sensitive and contains selenium as selenocysteine incorporated cotranslationally at the position of an in-frame UGA stop codon in the FdhF open reading frame [ Zinoni86 ].

A crystal structure of FDH-H has been solved at 2.3 Å resolution, confirming the presence of a [4Fe-4S] cluster, coordination of the Mo cofactor by selenocysteine, and the position of the binding site for the inhibitor nitrate [ Boyington97 ].

Expression of fdhF is induced by formate and the absence of external electron acceptors, and is repressed by nitrate, nitrite, trimethylamine N-oxide, and oxygen [ Wu87 , Pecher83 , Abaibou97 ]. Formate can overcome repression by nitrate but not by oxygen [ Pecher83 ].

Inhibition of DNA gyrase enhances expression of fdhF [ Axley88 ].

Cells grown anaerobically in the presence of nitrate express FDH-N, a membrane-bound nitrate reductase linked formate dehydrogenase, which oxidizes formate and transfers the resulting electrons to nitrate reductase

Front 64

pepE

Back 64

PepE hydrolyzes peptides bond in dipeptides where the first amino acid is aspartate [ Zhang98c ]

Front 65

tufB

Back 65

Elongation factor Tu (EF-Tu) is the most abundant protein in Escherichia coli. In its GTP-bound (active) form, EF-Tu binds aminoacylated tRNAs to form the so-called ternary complex. At the decoding site of the ribosome, the ternary complex is "tested" for a codon-anticodon match; if the proper aminoacyl-tRNA has been found, GTP is hydrolyzed and EF-Tu and GDP dissociate from the ribosome, while the aminoacyl-tRNA remains bound to the ribosome.

In E. coli, EF-Tu is encoded by two genes, tufA and tufB

Front 66

rpmE

Back 66

The L31 protein is a component of the 50S subunit of the ribosome.

L31 can be isolated in a complex with 5S rRNA and L5, L18, and L25 [ Fanning81 ] and can be crosslinked to tRNA in the P site [ Graifer89 ]. L31 may only be loosely associated with the ribosome [ Eistetter99 ]

Front 67

rnpA

Back 67

RnpB RNA and RnpA protein together compose ribonuclease P (RNAse P) [ Kole79 , Kole81 ]. RNAse P acts in processing of 4.5S RNA [ Bothwell76 ] and tRNA precursor molecules [ Schedl75 ]. The RnpB RNA subunit alone exhibits catalytic activity toward tRNA precursors, but requires the protein subunit for efficient activity toward an 4.5S RNA precursor substrate [ GuerrierTa83 , GuerrierTa84 , PeckMiller91 ]. RNAse P also exhibits activity toward C4 repressor RNA of bacteriophages P1 and P7, which may indicate a role for RNAse P in lysogeny/lysis regulation [ Hartmann95 ], and toward CI RNA, which is involved in immunity to bacteriophage P4 [ Briani02 ].

Activity of the RNAse P RNA and holoenzyme [ Schedl75 , Chang75b , Bikoff75 , Bikoff75a , Sakano75 , Kirsebom89 , Ilgen76 , Burgin90 , GuerrierTa83 , GuerrierTa84 , Orellana86 , GuerrierTa86 , GuerrierTa88 , Vioque89 , Surratt90 , PeckMiller91 , Kazakov91 , Nomura88 , Tallsjo93 , Hardt95 , Kufel94 , Hartmann95 , Steitz93 , Warnecke96 , Lazard98 , Brannvall98 , Hori01 , Li02 , Kirsebom92 , Mikkelsen99 , Warnecke00 , Pomeranz00 , Busch00 , Tanaka01e , Eubank02 , Kaye02 , Persson03 ] and substrate interaction and specificity [ Schmidt76 , Bothwell76a , Bothwell76a , Thurlow91 , Subbarao84 , Shlyapniko86 , Kirsebom89 , GuerrierTa84a , Holm92 , Green88a , McClain87 , Burkard88 , Reilly86 , Baer88 , Vold88 , Burkard88a , GuerrierTa88 , Samuelsson88 , GuerrierTa89 , Knap90 , Surratt90a , Surratt90 , Krupp91 , Perreault92 , Svard92 , Schlegl92 , Furdon83 , Gurevitz83 , Nolan93 , Gaur93 , Svard93 , Kleineidam93 , Kirsebom93 , GuerrierTa93 , Hardt93 , Hardt93a , Westhof94 , LaGrandeur94 , Oh94 , Kirsebom94 , Hardt95a , Meinnel95b , Yan94 , Tallsjo96 , Svard96 , Kufel96 , Gaur96 , Liu96e , Kufel96a , Lazard98 , Brannvall98 , Park00a , Brannvall02 , Brannvall03 , Hardt96 , Harris97a , Christian98 , Heide99 , Siew99 , Pascual99 , Heide01 , Heide01a , Hansen01c , Zahler03 , Ando03 , Ando03a , Tanaka03b , Nagai03 , Ando03b ] have been examined in detail.

The association of RnpB RNA and RnpA protein has been examined [ Vioque88 , Baer89 , Talbot94 , Talbot94a , Gopalan97 , Gopalan99 , Kim00b , Biswas00 , Sharkady01 , Tsai03 ]. The enzyme complex exhibits 1:1 stoichiometry of RnpB RNA to RnpA protein [ Talbot94 ].

RNAse P fractionates with the inner membrane [ Srivastava91a ]. Overproduced RNAse P enzyme is also localized to the inner membrane, whereas overproduced RnpB RNA is localized to the cytoplasm [ Miczak91 ].

Structural properties of RnpA have been investigated [ Gopalan97a ]. The N terminus plays a role in the catalytic activity of RNAse P [ Kim00b ]. The RNR motif, centrally located in the RnpA protein, and the C-terminal region, are both required for interaction with the RnpB RNA [ Kim00b ]. RnpA is very basic [ Hansen85 ]. The RnpA protein has been crystallized for structural studies [ Choe03 ]. RnpA may interact with additional, non-RnpB RNAs [ Lee02d ].

An rnpA49 mutation results in a recessive heat sensitivity phenotype [ Apirion80 ] and reduced RNAse P Kcat [ Kirsebom89 ]. The rnp-241 also causes heat sensitivity [ Hansen85 ]. An rnpA49 mutant harboring a ColE1-type episome exhibits buildup of RNA I-derived RNAs [ Jung92 ]. R46H (C5A49) mutant protein shows defects in interaction with RnpB RNA [ Baer89 ].

Mutations in RnpA affect RNAse P substrate specificity [ Gopalan97 ]. The large-scale gene expression profile of an rnpA mutant has been evaluated [ Li03e ]. Mutations in RnbP RNA and in the RnpA protein subunit of RNAse P exhibit different effects on tRNA precursor processing [ Kirsebom88 ].

The gene encoding the RNAse P protein subunit has been isolated from Bacillus subtilis [ Ogasawara85 ], Borrelia burgdorferi [ Old93 ], Synechocystis sp. PCC 6803 [ Pascual96 ], Thermotoga maritima [ Paul01a ], Mycoplasma capricolum [ Miyata93a ], Proteus mirabilis [ Skovgaard90 ], Micrococcus luteus [ Fujita90 ], Buchnera aphidicola [ Lai92 , Hassan96 , Clark98 ], Streptomyces coelicolor [ Calcutt92 ], Streptomyces bikiniensis var. zorbonensis [ Morse92 ], Mycobacterium leprae [ Fsihi96 ], Xanthomonas campestris pv. campestris 17 [ Yen02 ], Thermus thermophilus [ Feltens03 ], and Thermus filiformis [ Feltens03 ]. The RnpA protein exhibits antigenic similarity to proteins from human cells and from Bacillus subtilis [ Mamula89 ]. E. coli rnpA mutant phenotypes are functionally complemented by RnpA from Synechocystis sp. PCC 6803 [ Pascual96 ] or Streptomyces bikiniensis var. zorbonensis [ Morse92 ], or by Brevibacterium albidum arginine tRNA [ Kim97a ]

Front 68

rph

Back 68

Rph has ribonuclease PH (RNase PH) activity [ Ost91 , Kelly92a ]. RNase PH is a divalent cation-dependent [ Kelly92a ] and phosphate-dependent exonuclease that acts in processing of the tRNA 3' terminus [ Deutscher88 , Kelly92a , Kelly92b ]. In the E. coli K-12 strains MG1655 (represented in EcoCyc) and W3110, the rph gene contains a frame shift mutation resulting in a shortened open reading frame. The truncated protein lacks poly(A) phosphorylase activity; the strains also show a defect due to PyrE deficiency [ Jensen93 ]. Further descriptions of RNase PH here refer to the wild-type version of the enzyme present in other E. coli K-12 strains.

RNase PH is multimeric [ Jensen92 , Kelly92a ]. The enzyme activity has been studied in vitro [ Jensen92 , Kelly92a ]. The reaction is reversible in vitro, but probably unidirectional in vivo [ Ost90 , Kelly92a ].

Rph exhibits nonspecific DNA-binding activity [ Jensen92 ].

An rph mutation causes synthetic phenotypes, including synthetic lethality, in combination with various other RNase mutations [ Kelly92b ]. Quintuple mutation of exoribonucleases RNase II, D, BN, T, and PH is lethal, whereas quadruple mutants are viable, but not wild-type [ Kelly92 , Reuven93 ]. Double mutation of polynucleotide phosphorylase (PNPase) and RNase PH causes ribosomal defects [ Zhou97 ].

Rph has helix-turn-helix sequence motifs [ Jensen92 ]

Front 69

rpmG

Back 69

The L33 protein is a component of the 50S subunit of the ribosome.

L33 is located within 21 Å of nucleotide C2475 of 23S rRNA, near the peptidyltransferase center [ Muralikris95 ]. L33 had previously been shown to crosslink to 23S rRNA [ Osswald90 ] and to tRNA in the P site [ Bausk85 , Graifer89 , Podkowinsk89 , Sumpter90 , Podkowinsk91 , Mitchell93a , Osswald95 ] and E site [ Osswald95 ]. L33 also crosslinks to L1 and L27 [ Walleczek89 , Redl89 , Walleczek89a ].

The initiating methionine of L33 is lost [ Arnold99 ] or may be methylated [ Chang77 ]; L33 is methylated at the first alanine residue [ Cammack65 , Chang77 , Arnold99 ].

A strain containing an IS element insertion in rpmG shows no major defect in ribosome assembly [ Coleman93 ]; L33 appears to have no significant role in ribosome synthesis or function [ Maguire97 , Maguire97a , Maguire97b ]

Front 70

selA

Back 70

Selenocysteine synthase is encoded by the selA gene [ Forchhamme91 ]. Selenocysteine synthase contains 10 SelA subunits arranged in two rings [ Engelhardt92 ]; the complex exhibits a molecular weight of approximately 600 kDa [ Forchhamme91 ]. Pyridoxal 5-phosphate is present at a stoichiometry of one per monomer [ Forchhamme91 ]. The seryl-tRNA(Sec UCA) substrate is present at a stoichiometry of one per two monomers [ Forchhamme91a ]. The reaction mechanism is described [ Forchhamme91a ].

A selA or selD mutant exhibits a defect in selenocysteine formation, whereas a selB mutant does not [ Leinfelder89 ].

Cloning, sequencing, and protein purification is described [ Forchhamme91 ]. Extracts of cells overproducing SelA and SelD exhibit in vitro production of selenocysteinyl-tRNA(Ser)(UCA) from seryl-tRNA(UCA) [ Leinfelder90 ].

Regulation has been described [ Sawers91 ]

Front 71

prlC

Back 71

Oligopeptidase A is a cytoplasmic protease with broad specificity. It was originally identified as a suppressor of signal sequence mutants and thus may be involved in degradation of signal peptides [ Emr82 , Trun87 ]. Oligoeptidase A appears shortly before DNA synthesis in at least some strains and may be involved in the onset of DNA replication, as treatment of cells with an inhibitor of Oligopeptidase A prevents cell division [ Kato92 , Irisawa93 ]. Oligopeptidase A may also be partially reponsible for degradation of signal peptides [ Novak88 ].

Oligopeptidase A catalyzes broad-specificity hydrolysis of peptide bonds. In hydrolysis of bonds in signal sequence peptides, some specificity is seen for cleavage near alanine and glycine [ Kato92 , Irisawa93 , Novak88 ]

Front 72

rbbA

Back 72

ribosome-associated ATPase

RbbA is associated with the 30S subunit of the ribosome and has ribosome-dependent ATPase activity [ Kiel99 , Kiel01 ]. Stimulation of protein synthesis by RbbA is dependent on ATP hydrolysis, and the ATPase activity is inhibited by hygromycin [ Kiel01 , Ganoza01 ]. The effect of hygromycin may be due to its ability to release RbbA from the ribosome [ Ganoza01 ].

RbbA specifically interacts with EF-Tu and crosslinks to the 30S ribosomal subunit protein S1 [ Kiel01 ]. The E-site base A937 of 16S rRNA is protected by RbbA in the intact 70S ribosome [ Xu06g ].

RbbA contains two ATP-binding domains in its N terminus [ Kiel99 ]. The ATP-binding domains together with five predicted transmembrane helices led to the initial functional prediction as the ATP-binding component of an ABC transporter [ Saurin99 ]. A truncated form of RbbA which does not contain the predicted transmembrane domains still stimulates protein synthesis and has ribosome-dependent ATPase activity [ Xu06g ].

RbbA cross-reacts with antibody raised against fungal elongation factor EF-3 [ Kiel99 ]

Front 73

glpG

Back 73

GlpG is a protease that may be involved in intramembrane proteolysis.

GlpG cleaves between Ser and Asp in a region of high hydrophobicity in a model substrate in vitro [ Maegawa05 ]. GlpG appears to recognize substrates through their transmembrane regions [ Maegawa05 ].

GlpG is a basic, polytopic membrane protein associated with the inner membrane [ Zeng96 , Maegawa05 ]. It appears to belong to the rhomboid family of membrane proteases [ Maegawa05 ]. Crystal structures of GlpG to 2.1 and 2.3 Å resolution show that is has six transmembrane domains, and that its active site is within the membrane-inserted portion of the protein, accessible through a lateral opening that is blocked by a loop structure [ Wang06f , BenShem07 ]. A crystal structure of GlpG with this loop moved to expose the catalytic site has also been determined [ Wang07e ].

The gating mechanism of GlpG has been analysed by mutational analyses both in vitro, with purified protein [ Baker07 ], and in vivo within the membrane of live E. coli cells [ Urban08a ]. Both studies indicate that transmembrane segment five is the substrate gate. Conversely, mutations in other parts of the protease, including the membrane-inserted L1 loop previously thought to be the gate, decrease enzyme activity, and may represent the open form of GlpG [ Urban08a ]. Molecular dynamics simulation of wild type and mutant GlpG showed that mutations in the transmembrane five helix affected the dynamics and structure of the L1 loop suggesting that these two domains are dynamically coupled [ Bondar09 ].

Regulation has been described [ Schweizer86a , Zeng96 , Yang98a ]

Front 74

trpS

Back 74

tryptophanyl-tRNA synthetase

Tryptophanyl-tRNA synthetase (TrpRS) is a member of the family of aminoacyl-tRNA synthetases, which interpret the genetic code by covalently linking amino acids to their specific tRNA molecules. The reaction is driven by ATP hydrolysis. TrpRS belongs to the Class I aminoacyl tRNA synthetases [ Eriani90 , Landes95 ].

The enzyme purified from E. coli B is a homodimer in solution [ Joseph71 , Joseph71a ]. The dimer has two binding sites for both tryptophan and tRNATrp [ Muench76 ].

Specificity determinants within tRNATrp that are important for recognition by TrpRS have been identified. The anticodon and the G73 discriminator base are the major identity determinants [ Himeno91 ], and C35 is recognized as well [ Rogers92 ]. The apparent affinity of TrpRS for tryptophan appears to be dependent on the tRNA [ Ibba96 ]. The identity of tRNATrp predominantly affects the rate of transfer of tryptophan from the TrpRS-tryptophanyl adenylate to the tRNA [ Ibba99 ]. TrpRS also aminoacylates tRNATrp with D-tryptophan; the resulting D-Trp-tRNATrp can be hydrolyzed by D-Tyr-tRNATyr deacylase [ Soutourina00 ]. A strain that can incorporate 4-fluorotryptophan in place of tryptophan into proteins has been isolated and contains, among others, mutations in TrpRS [ Bacher01 ].

The sequential processes involved in the tRNA charging reaction have been studied [ Andrews85 , Merle86 , Lloyd95 ]. Mutagenesis of the Thr17 residue in the TIGN motif indicates that Thr17 is important for binding of substrate in the transition state [ Chan94 ], and mutagenesis of Lys195 in the conserved KMSKS motif indicates that Lys195 may interact with ATP in the transition state [ Chan95 ]. Mutants that are auxotrophic for tryptophan and map within trpS have been identified; the mutations are located within the conserved KMSKS motif, near the active site, or lining a proposed dimerization interface, supporting a role for dimerization of the enzyme in catalysis [ Sever96 ].

TrpRS activity increases with the growth rate; the regulation is at the level of trpS transcription [ Hall82 ]

Front 75

ppiA

Back 75

peptidyl-prolyl cis-trans isomerase A (rotamase A)

PpiA is a peptidyl-prolyl cis-trans-isomerase (PPIase), catalyzing the conformational isomerization of prolyl residues in peptides. Cis-trans isomerization of prolyl peptide bonds is a slow step in protein folding, and thus PpiA is thought to facilitate proper protein folding.

PpiA was shown to catalyze the refolding of denatured type III collagen [ Compton92 ]. PpiA is a homolog of the human enzyme cyclophilin; unlike that enzyme, PpiA activity is only inhibited by high concentrations of cyclosporin A [ Liu90a , Compton92 ] or FK506 [ Hayano91 ]. An F112W mutant, changing to the tryptophan residue conserved in eukaryotic cylcophilins, is more sensitive to inhibition by cyclosporin A [ Liu91 ] and binds cylcosporin A in a configuration similar to the human enzyme [ Fejzo94 ].

Solution structures of wild type [ Clubb94 ] and mutant [ Fejzo94 ] PpiA as well as crystal structures of a mutant form of PpiA in a complex with a peptide containing the trans form of proline have been determined [ Konno04 ].

A ppiA null mutant shows no apparent growth defect [ Kleerebeze95 ]. A strain containing null mutations in all four known periplasmic peptidyl-prolyl cis-trans-isomerases, ppiA, ppiD, fkpA, and surA, is viable, but shows a decreased growth rate and increased antibiotic susceptibility [ Justice05 ]. Expression of PpiA is regulated by cAMP-CRP, CytR [ Norregaard94 ], and the Cpx two-component system [ Pogliano97 ]

Front 76

fic

Back 76

Reports disagree about whether a fic null mutant is viable [ Kawamukai88 ] or inviable [ Komano91 ]. A fic null mutant exhibits a PAB (p-aminobenzoate) or folate requirement [ Komano91 ]. A fic-1 mutant exhibits heat-sensitive CRP-dependent cell filamentation in the presence of cAMP [ Utsumi82 ]. Filamentation of the fic-1 mutant is associated with production of a 40 kD membrane protein [ Utsumi83 ], and the phenotype is rescued by NaCl [ Kawamukai88 ] or by folate [ Komano91 ]. The fic-1 allele is not a null mutation, and a null mutant does not exhibit heat-sensitive, cAMP-induced filamentation [ Kawamukai88 ]. The fic-1 allele encodes a G55R change [ Kawamukai89 ].

Fic has similarity to E. coli FtsA [ Kawamukai89 ], Saccharomyces cerevisiae Cdc28p [ Kawamukai89 ], and Schizosaccharomyces pombe Cdc2p [ Kawamukai89 ].

Fic: "filamentation induced by cAMP" [ Utsumi82 ]

Front 77

fusA

Back 77

Elongation factor G is an essential protein that facilitates the translocation of the ribosome by one codon along the mRNA molecule. The activity requires GTP hydrolysis [ Wintermeye01 ]. EF-G is also involved in ribosome disassembly and recycling [ Kaji01 ]

Front 78

rsgA =yjeQ

Back 78

Based on genetic interactions, RgsA/YjeQ is thought to play a role in late ribosome biogenesis [ Campbell08 ].

YjeQ exhibits GTPase activity [ Daigle02 ], which is stimulated 160-fold by association of the YjeQ protein with stochiometric amounts of the 30S subunit of the ribosome [ Daigle04 ]. Addition of aminoglycosides that bind to the A site of the 30S ribosomal subunit inhibit that effect [ Himeno04 ].

YjeQ is a member of a family of P-loop-containing GTPases with an unusual arrangement of GTPase motifs [ Daigle02 ]. The N terminus has an OB-fold RNA-binding domain, the central region comprises the GTPase motifs, and the C terminus has a potential zinc knuckle domain [ Daigle02 ]. The YjeQ protein copurifies with ribosomes at a ratio of 1:200. Recombinant YjeQ interacts most strongly with the 30S subunit in the presence of 5'-guanylylimidodiphosphate (GMP-PNP), a nonhydrolyzable GTP analog [ Daigle04 ].

The protein was thought to be essential for growth [ Arigoni98 ], but a deletion mutant was shown to have a slow growth phenotype and accumulate 30S and 50S ribosomal subunits [ Himeno04 , Campbell08 ]. Multicopy suppressors of the slow growth phenotype and synthetic lethal mutants have been isolated [ Campbell08 ].

An S221A mutation within the central GTPase motif causes a catalytic defect [ Daigle02 ]. N-terminal truncation mutants showed that the OB-fold region is essential for ribosome interaction and GTPase stimulation; the N-terminal amino acids 1-20 are essential for GMP-PNP-dependent interaction with the 30S subunit [ Daigle04 ].

RsgA: "ribosome small subunit-dependent GTPase A" [ Himeno04 ]

Front 79

bipA

Back 79

BipA is a member of the ribosome-binding GTPase superfamily. The variety of phenotypes of a bipA deletion, as well as the genetic interaction with the RluC 23S rRNA pseudouridine synthase suggest that BipA plays a role in modulating the structure or function of the ribosome [ Krishnan08 ].

BipA is necessary for growth under low temperature conditions [ Pfennig01 , Grant01b ]. Deletion of rluC suppresses the cold-sensitive phenotype of a bipA mutant [ Krishnan08 ].

A bipA mutation suppresses the defect in lipopolysaccharide core biosynthesis, the SDS sensitivity, and the defect in mouse intestinal colonization of a waaQ mutant [ Moller03 ]. BipA may be involved in positive regulation of colanic acid synthesis [ Krishnan08 ]. A strain containing a transposon insertion in bipA is more sensitive to chloramphenicol than wild-type [ Duo08 ].

Whereas BipA of E. coli K-12 is reported not to be tyrosine phosphorylated [ Freestone98 , Pfennig01 ], BipA of enteropathogenic E. coli [ Farris98 , Freestone98 ] and wall-less L-form E. coli [ Freestone98a ] is phosphorylated on tyrosine.

BipA has been characterized in pathogenic bacteria [ Qi95 , Farris98 , Freestone98 , Farris98a , Barker00 , Rowe00 , Grant03b , Scott03b ]. Salmonella enterica serovar Typhimurium BipA is a ribosome-associated GTPase [ Qi95 , deLivron08 ]. BipA of enteropathogenic E. coli is a regulatory protein [ Freestone98 , Rowe00 , Farris98 , Grant03b ] involved in pathogenesis [ Farris98 , Grant03b ] and is required for wild-type translation [ Freestone98 ]

Front 80

trmE=mnmE

Back 80

GTP-binding protein with a role in modification of tRNA
MnmE is required for wild-type 5-methylaminomethyl-2-thiouridine modification of tRNA [ Elseviers84 ]. Together with MnmG, MnmE is thus involved in maintenance of the correct reading frame [ Brierley97 , Urbonavici01 , Bregeon01 , Urbonavici03 ].

MnmE also appears to play a role in oxidation of thiophene and furan compounds [ Alam91 ] and regulates glutamate-dependent acid resistance [ Gong04 ].

MnmE is a GTP-binding protein that also exhibits GTPase activity, showing rapid GTP hydrolysis and low nucleotide affinity. The nucleotide binding and hydrolysis activities are localized within the central 17 kDa GTPase domain [ Cabedo99 ]. The GTPase activity as well as the Cys451 residue in the C-terminal domain are required for the wild-type tRNA modification function [ Yim03 ], but not sufficient [ MartinezVi05 ]. Dimerization of the GTPase domain is potassium ion-dependent; subsequent GTP hydrolysis activity is dependent on dimerization [ Scrima06 ]. Low pH inhibits the GTP hydrolysis activity [ Monleon07 ]. Unlike other GTPases, MnmE does not appear to use an "arginine finger" for catalysis [ Scrima06 , Monleon07 ].

Solution and crystal structures of the G-domain of MnmE have been solved [ Scrima06 , Monleon07 ]. MnmE can homomultimerize and localizes to the cytoplasm, showing some association with the cytoplasmic membrane [ Cabedo99 ]. MnmE interacts specifically with MnmG [ Yim06 ].

Viability of an mnmE mutation is dependent on the strain background [ Cabedo99 ]. mnmE mutants are defective in the tRNA modification 5-methylaminomethyl-2-thiouridine; tRNA anticodons that are modified with 5-methylaminomethyl-2-thiouridine in the wild type show 2-thiouridine modification in the mutant, and mutants exhibit a defect in UAG readthrough [ Elseviers84 ]. Unexpectedly, the hypomodified tRNALys of an mnmE mutant leads to decreased misreading of the anticodon [ Hagervall98 ].

Expression of mnmE is increased during stationary phase, but independent of the stationary phase sigma factor RpoS. Expression is also subject to catabolite repression and is decreased in the absence of oxygen [ Zabel00 ]

Front 81

der

Back 81

50S ribosomal subunit stability factor

The Der protein is a GTPase that is ubiquitously conserved in eubacteria [ Bharat06 ]. It is associated with the large subunit of the ribosome and is required for its stability [ Hwang06 , Bharat06 ]. In E. coli, Der is essential for growth [ Hwang01 , Baba06 , Hwang06 , Bharat06 ].

GTP hydrolysis appears to regulate the specificity of interactions of Der with ribosomal subunits [ Tomar09 ]. The GAP-like protein YihI interacts with Der and activates its GTPase activity [ Hwang10 ]. The KH-like C-terminal domain of Der plays a role in recognition of the ribosome [ Hwang10a ].

The Der protein appears to be evenly distributed throughout the cytoplasm [ Watt07 ].

Mutants containing substitutions in either of the two predicted GTP binding domains of Der are not able to complement the lethal phenotype of the der null mutant at low temperature and are defective in ribosome biogenesis [ Hwang06 , Bharat06 ]. Depletion of Der results in the accumulation of 23S and 16S rRNA precursors and accumulation of aberrant 50S ribosomal subunits at low concentrations of Mg2+ [ Hwang06 ]. Overexpression of RelA suppresses the growth defect of a der mutant [ Hwang08 ].

Overexpression of der suppresses the growth and ribosome-related defects of a mutant lacking the RrmJ rRNA methyltransferase [ Tan02 ]. Both the GTP binding domains and the C-terminal KH-like domain of Der are required for suppression [ Hwang10a ].

Der: "double-Era-like domains" [ Hwang01 ]

Front 82

elaD

Back 82

ElaD is a deubiquitinating protease.

ElaD is a highly specific, moderately active deubiquitinating enzyme [ Catic07 ]

Front 83

rbn

Back 83

RNase BN, also identified as binuclear zinc phosphodiesterase, cleaves the 3'-terminal portion of tRNAs as well as various short unstructured RNAs.

RNase BN carries out the exoribonucleolytic cleavage of the 3'-terminus of tRNA with terminal nucleotide substitutions or deletions. It has demonstrated particular affinity for tRNA-CU and tRNA-CA, but not intact tRNA-CCA [ Callahan00 , Ezraty05a , Takaku04 ]. Substrate identification depends on a region of RNase BN known as the ZiPD exosite [ Schilling05 ].

RNase BN has also been shown to cleave unstructured RNA [ Shibata06 ].

RNase BN is a member of the metallo-beta-lactamase family and functions as a dimer [ Vogel02 , Callahan00 ]. A crystal structure of RNase BN has been determined to 2.9 Å resolution [ Kostelecky06 ].

Deletion of elaC alone has no effect on growth, but loss of the RNases BN, II, T, PH and D results in inviability [ Schilling04 ]. RNase BN alone can restore some viability [ Kelly92b ].

Site-directed mutagenesis has elucidated a potential zinc binding site, though other studies point to divalent cobalt or magnesium as being required for tRNase activity [ Vogel04 , Callahan00 ].

Prior to its identification as RNase BN, the activity of ElaC was gauged in vitro using the nonphysiological substrate bis-(p-nitrophenyl)phosphate in the presence of divalent zinc cation. Kinetic parameters for this reaction have been determined [ Vogel02 ].

The gene coding for RNase BN was originally incorrectly identified as yihY [ Callahan96 ]

Front 84

sppA

Back 84

Protease IV is an endopeptidase that degrades cleaved lipoprotein signal peptide [ Ichihara84 ]. It may not be the only protease carrying out this function, as signal peptide is still degraded, albeit slowly, in its absence [ Suzuki87a ]. Protease IV functions as a tetramer [ Ichihara86 ] with a novel bowl shaped architecture [ Kim08d ]. Analysis of the peptide fragments generated from cleavage of prolipoprotein signal peptide by purified SppA suggests that it cleaves predominantly in the hydrophobic segment of the signal peptide [ Novak88 ].

SppA has one transmembrane segment (spanning residues 29-45) and the carboxy-terminal domain is located in the periplasm [ Wang08c ].

Deletion of sppA did not affect degradation of the signal peptides of fusion proteins in vivo [ Saito11 ]

Front 85

hlpA =skp

Back 85

Skp is a periplasmic protein chaperone that binds to unfolded outer membrane proteins. There are at least two pathways for periplasmic chaperone activity, one involving Skp and DegP, and another involving SurA. The SurA pathway is the primary pathway for assembly of OMPs, while the DegP/Skp pathway is important for rescuing proteins that fall off of the SurA pathway.

Skp prevents folding of OmpA in solution [ Bulieris03 ]. Skp and LPS improve insertion of OmpA into phospholipid bilayers [ Bulieris03 ]. Skp interacts with OmpA and PhoE at 1:1 ratios as they cross the inner membrane, and the resulting complexes can be identified from the membrane fraction [ Schafer99 , Harms01 ].

A skp or skp degP mutation causes induction of the unfolded protein-mediated σE stress response [ Missiakas96 , Sklar07a ]. A skp mutant exhibits decreased abundance of outer membrane proteins, compared to wild type [ Chen96b ]. A skp null mutant exhibits mild membrane defects and is unable to efficiently release OmpA from the inner membrane [ Schafer99 ]. Skp suppresses the protein export defect of a secA mutant in vitro with respect to membrane incorporation of LamB and OmpA [ Thome90 ]. A skp degP double mutant exhibits heat sensitivity and periplasmic buildup of denatured proteins [ Schafer99 ]. A skp surA double mutant exhibits a growth defect with formation of filaments [ Rizzitello01 ], accumulates unfolded OmpA, and prevents proper formation of LamB trimers in the outer membrane [ Sklar07a ]. A degP surA double mutation is lethal [ Rizzitello01 ]. A surA mutant forms LamB trimers more slowly than a skp degP mutant [ Sklar07a ]. Depletion of SurA reduced OMPs in the outer membrane, but a skp degP double mutant did not have significantly reduced amounts of OMPs in the outer membrane [ Sklar07a ]. Overproduction of Skp increases production of some recombinant proteins [ Bothmann98 , Strachan99 , Mavrangelo01 , Levy01 , Lin08a ] or proteins produced for phage display [ Bothmann98 , Hayhurst99 ]. Expression of skp also reduces extracytoplasmic stress associated with expression of recombinant outer membrane proteins [ Narayanan ].

Crystal structures of Skp have been solved [ Walton04 , Korndorfer04 , Schlapschy04 ]. Skp is periplasmic and soluble [ Thome91 , Chen96b ] and forms stable homo-trimers [ Schlapschy04 ]. Wild-type Skp translocation to the periplasm requires the SecA [ Thome91 , Ernst94 ] and SecY [ Thome91 ] proteins, but not SecB [ Ernst94 ]. Wild-type Skp translocation also requires ATP [ Thome91 ] and the proton gradient [ Thome91 , Ernst94 ]. Skp is subject to post-translational processing [ Holck88 , Thome91 ]. A 20-residue N-terminal leader is removed to generate the mature species [ Holck88 ]. Skp exhibits interactions with OmpA, OmpC, OmpF, LamB, PhoE OmpG, and YaeT outer membrane proteins at a ratio of one Skp trimer per OMP protein [ Chen96b , De99c , Harms01 , Qu07 ]. Proteomic analyses have identified 30 other interacting proteins, especially from the outer membrane, among these FadL and BtuB, and from the periplasm, MalE and OppA [ Jarchow08 ]. Skp shows a conformational change to protease resistance upon interaction with phospholipids in the presence of divalent cations by inserting into the membrane, and this change is inhibited in the presence of divalent cations and lipopolysaccharides [ De99c ]. Nuclear magnetic resonance of Skp with bound OmpA shows that the OmpA βbarrel is maintained in an unfolded state within the Skp cavity whereas the folded periplasmic domain of OmpA protrudes outside of the cavity [ Walton09 ]. The interaction of Skp with OmpA has been studied using site directed fluorescence spectroscopy [ Qu09 ].

Expression of skp is regulated by σE as well as by the CpxAR two-component response regulator [ Dartigalon01 , Rhodius05 ].

There was some confusion among the neighboring genes skp and lpdX/firA in early studies [ Dicker91 , Aasland88 , Thome90 ].

HlpA: "histone-like protein" / HLP-I: "histone-like protein I" / Skp: "seventeen kDa protein"

Front 86

groL = groEL

Back 86

GroEL, chaperone Hsp60, peptide-dependent ATPase, heat shock protein

GroL protein synthesis was stimulated both by heat shock in vivo and by the heat shock-specific RNA polymerase Esigma32 [ Chuang93 ]

Front 87

groS = groES

Back 87

GroES, chaperone binds to Hsp60 in pres. Mg-ATP, suppressing its ATPase activity

Front 88

hslR

Back 88

Hsp15 is an abundant, highly conserved heat shock protein with DNA and RNA binding activity [ Korber99 ]. In vivo, Hsp15 specifically interacts with the 50S ribosomal subunit when it contains a nascent polypeptide chain [ Korber00 , Jiang09 ] and is involved in recycling such complexes [ Jiang09 ].

The crystal structure of Hsp15 has been determined at 2 Å resolution, revealing a novel RNA binding motif that is widely shared among stress proteins, ribosomal proteins and tRNA synthetases [ Staker00 ]. The structure of the 50S ribosomal subunit-nascent chain tRNA complex alone and with Hsp15 has been reconstructed using cryo-electron microscopy [ Jiang09 ].

The levels of hslR mRNA and Hsp15 protein are strongly upregulated after heat shock treatment [ Chuang93 , Korber99 ].

HslR: "heat shock locus R" [ Chuang93 ]

Front 89

rlmE

Back 89

23S rRNA 2'-O-ribose U2552 methyltransferase

RlmE is the methyltransferase responsible for methylation of 23S rRNA at the 2'-O position of the ribose at the universally conserved U2552 nucleotide [ Caldas00 ]. In vitro, the enzyme is active on ribosomes and the 50S ribosomal subunit, but not free rRNA [ Caldas00 , Bugl00 ].

A crystal structure of RlmE has been solved at 1.5 Å resolution [ Bugl00 ]. Site-directed mutagenesis has identified possible active site and substrate binding residues, and a reaction mechanism has been proposed [ Hager02 , Hager04 ].

A mutant strain lacking RlmE has a decreased growth rate at all temperatures tested and shows reduced protein synthesis activity and accumulation of free ribosomal subunits [ Caldas00a , Bugl00 ]. Overexpression of the small GTPases Obg, Der [ Tan02 , Hwang10 ] or CtgA [ Jiang06 ] suppresses the ribosomal assembly or stability defect of an rlmE mutant without restoring the methylation of U2552 [ Tan02 ]. An rlmE-deficient strain shows a decrease in -1 and +1 frameshifting and a decrease in UAA and UGA stop codon readthrough, suggesting that the U2552 base may interact with aminoacyl-tRNAs at the ribosomal A site [ Widerak05 ].

An rlmE mutant is more sensitive to lincomycin [ Caldas00a ], clindamycin, hygromycin A and sparsomycin [ Toh08 ] than wild-type

Front 90

rlmD

Back 90

RlmD is the methyltransferase responsible for methylation of 23S rRNA at the C5 position of the U1939 nucleotide [ Agarwalla02 , Madsen03 ]. In vitro, the enzyme methylates full-length 23S rRNA as well as a 70 nt fragment containing nucleotides 1930-1969 [ Agarwalla02 ].

Crystal structures of apo-RlmD and a ternary complex have been solved at 1.95 and 2.15 Å resolution, suggesting active site residues and a mechanism for base selectivity [ Lee04 , Lee05 ]. Since methyltransferase reactions do not involve a redox step, the presence of a [4Fe-4S] iron-sulfur cluster was unexpected. The iron-sulfur cluster was hypothesized to provide a mechanism for regulating RlmD activity under oxidative stress conditions [ Agarwalla04 ].

RumA: RNA uridine methyltransferase [ Agarwalla02 ]

Front 91

rlmC

Back 91

RlmC is the methyltransferase responsible for methylation of 23S rRNA at the C5 position of the U747 nucleotide [ Madsen03 ].

An rlmC mutant shows a specific defect in methylation of position U747 in 23S rRNA [ Madsen03 ].

RumB: RNA uridine methyltransferase [ Madsen03 ]

Front 92

rsmA= ksgA

Back 92

KsgA is the methyltransferase responsible for dimethylation of 16S rRNA at the two adjacent adenosine bases A1518 and A1519 [ Poldermans79 ]. In vitro, the enzyme is active on 30S ribosomal subunits, but not the fully assembled 70S ribosome [ Poldermans79 ]. KsgA may play a role in biogenesis of the small subunit of the ribosome [ Connolly08 ].

The evolutionary relationship and functional divergence of this enzyme and its homologs in eukaryotes has been studied [ Cotney06 ]. KsgA as well as the dimethylation modification it catalyzes are nearly universally conserved, suggesting an important function. KsgA is only able to methylate 30S ribosomal subunits in their translationally inactive conformation [ Desai06 ]; it interacts with the decoding site of the 30S subunit, and interactions of 30S with KsgA and IF3 appear to be mutually exclusive. A checkpoint model where binding of KsgA keeps immature 30S subunits from entering the translational cycle has been suggested [ Xu08 ].

Methylation of the two adenosine residues is independent of each other [ Cunningham90 , VilaSanjur99 ]. Recognition of the 3' terminal hairpin of 16S rRNA and methylation activity of KsgA have been studied [ Formenoy94 ].

Surprisingly, KsgA also exhibits cytosine-DNA glycosylase activity and may play a role in protection of DNA against oxidative stress [ ZhangAkiya09 ].

A crystal structure of KsgA has been solved at 2.1 Å resolution [ OFarrell04 ].

Mutation of ksgA causes resistance to kasugamycin (an inhibitor of translation initiation), but no substantial growth defect [ Sparling70 , Dabbs80 , Leveque90 ]. The lack of methylation at A1519, but not A1518, appears to be responsible for the kasugamycin resistance phenotype of a ksgA mutant [ VilaSanjur99 ]. A ksgA deletion enhances the slow growth phenotype of an rsgA mutation [ Campbell08 ]. ksgA was also identified as a multicopy suppressor for a cold-sensitive mutant of era, which encodes an essential GTP-binding protein [ Lu98 ]. The methyltransferase and multicopy suppressor activities of KsgA are separable, indicating a possible second function of KsgA [ Inoue07 ]. High-level overexpression of KsgA is toxic, and the cells show increased sensitivity to acid shock [ Inoue07 ]. A ksgA deletion strain is cold sensitive and shows an altered ribosome profile and altered 16S rRNA processing [ Connolly08 ]. AksgA point mutant lacking catalytic activity has a dominant negative effect on growth and ribosome formation [ Connolly08 ].

The KsgA protein binds specifically to its own mRNA and may regulate its own translation [ vanGemen89 ]. Growth rate positively regulates ksgA expression [ Pease02 ].

KsgA: "kasugamycin resistance"

Front 93

nlpD

Back 93

NlpD in Escherichia coli is similar in sequence to the LppB lipoprotein outer membrane antigen of Haemophilus somnus, a putative virulence determinant. In E.coli, the nlpD gene product has been shown to incorporate radiolabeled palmitic acid and to accumulate an uncleaved precursor in the presence of globomycin, an inhibitor of the lipoprotein signal peptidase, giving strong evidence that NlpD is a lipoprotein. Insertion deletion of nlpD results in decreased stationary-phase survival after 7 days [ Ichikawa94 ].

NlpD is thought to function in cell division. An NlpD-Mcherry fusion protein localises to the septal ring division site. envC nlpD double null mutants show significant defects in cell separation resulting in very long chain formation due to a defect in septal peptidoglycan splitting [ Uehara09 ]

Front 94

surA

Back 94

SurA in Escherichia coli is a periplasmic peptidyl-prolyl isomerase which is necessary for the proper folding of outer membrane proteins (OMP) including OmpA, OmpF and LamB. There are at least two pathways for periplasmic chaperone activity, one involving Skp and DegP, and another involving SurA. It has been proposed that the SurA pathway is the primary pathway for assembly of OMPs, while the DegP/Skp pathway is important for rescuing proteins that fall off of the SurA pathway. However recent proteonomic studies show that only 8 of 23 identified βbarrel proteins were negatively affected in a surA knockout mutant [ Vertommen09 ]. The decreased OMP included FadL, FecA, LptD, FhuA, OmpX as well as the major OMP OmpA, OmpF and LamB.

SurA can be found in soluble form and associated with the outer membrane but not the inner membrane, suggesting it acts in early periplasmic and late outer membrane-association steps [ Hennecke05 ]. The crystal structure of SurA has been determined to a 3.0 Å resolution [ Bitto02 ]. Purified SurA has low PPIase activity [ Rouviere96 ]. SurA has partial overlapping substrate specificity with PpiD [ Stymest08 ]. The N-terminal domain and C-terminal tail are required for SurA activity, and removal of the parvulin-like domains results in slightly reduced activity [ Behrens01 ]. In vivo cross-linking and coimmunoprecipitation studies suggest that SurA interacts directly with BamA in the OM independently of YfgH [ Sklar07a , Vuong08 ]. SurA increases the efficiency of OmpT assembly by the BamABCDSmpA Outer Membrane Protein Assembly Complex in vitro [ Hagan10 ] Peptide-binding assays and crystallization of peptide-bound SurA proteins suggest that SurA recognizes the aromatic-random-aromatic peptide motif common in beta barrel structures of integral outer membrane proteins and does not require proline [ Webb01 , Bitto03 , Hennecke05 , Xu07c , Alcock08 ].

A surA mutation is lethal during stationary phase when pH becomes elevated [ Lazar98 , Tormo90 ]. A rpoS surA mutation becomes lethal during stationary phase in LB due to elevated pH, but rpoS expression compensates for loss of surA during statinary phase [ Lazar98 ]. The activity of SurA may be necessary for proper assembly of cell-wall synthesizing proteins at high pH because the phenotype of a surA rpoS mutant is similar to that of a ftsI rpoS mutant [ Lazar98 ]. Mutation of surA activates the σE stress response [ Rouviere96 , Missiakas96 ]. Mutation of SurA has been shown to increase sensitivity to antibiotics due to defects in membrane integrity [ Justice05 ]. A surA null mutant has reduced steady state levels of LamB, OmpA, OmpC, and OmpF, but is not impaired in TolC assembly [ Werner03 , Dartigalon98 , Rouviere96 , Lazar96 ]. A surA mutant is defective in converting unfolded LamB into folded LamB monomer and is indistinguishable from yfgL mutants in this respect [ Ureta07 , Rouviere96 ], though there does not appear to be significant functional overlap of SurA and YfgL [ Charlson06 ]. A yfgL surA double mutation is severely impaired for growth [ Onufryk05 ]. A skp surA double mutant grown in rich medium exhibits a growth defect with formation of filaments [ Rizzitello01 , Onufryk05 ], accumulates unfolded OmpA, and prevents proper formation of LamB trimers in the outer membrane [ Sklar07a ]. degP surA, nlpB surA, and yraP surA double mutations are lethal or severely impaired in rich media [ Rizzitello01 , Onufryk05 ]. A surA mutant forms LamB trimers more slowly than a skp degP mutant [ Sklar07a ]. Depletion of SurA reduced OMPs in the outer membrane, but a skp degP double mutant did not have significantly reduced amounts of OMPs in the outer membrane [ Sklar07a , Onufryk05 ]. surA deletion mutants are also defective for biofilm formation [ Niba07 ] and are defective for growth in absence of aromatic amino acids [ Smith07a ]. RseA is unstable in surA mutants resulting in activation of the σE stress resopnse [ Ades99 ]. A surA yfgA double deletion had a synthetic lethal phenotype [ Niba07 ]. A surA null mutation results in RcsF-dependent activation of the RcsCDB signal transduction system [ CastanieCo06 ]. A surA deletion had reduced numbers of multidrug tolerant persistors upon exposure to ofloxacin [ Hansen08a ]. Mutation studies involving four PPIases present in Escherichia coli (FkpA, PpiA, PpiD, and SurA) showed that three triple mutants have growth defects, while one does not [ Justice05 ]. The quadruple mutant had the most pronounced defect as well as increased antibiotic sensitivity compared with other surA mutants [ Justice05 ]. No single or double mutants were shown to have growth defects in this study [ Justice05 ]. A ppiD surA mutation was lethal in another study [ Dartigalon98 ]. PpiD was found to be a multicopy suppressor of a surA mutation [ Dartigalon98 ]. Mutation of surA also resulted in reduced piliation in cells expressing the P and type 1 pilus operons suggesting a role for SurA and possibly the other PPIases in survival outside the laboratory or in pathogenesis [ Justice05 ]. Removal of the PPIase activity of SurA by mutation does not result in loss of SurA activity related to outer membrane incorporation of proteins [ Behrens01 ].

Expression of surA increases in cultures of high density [ Cuny05 ] and is lowest at neutral pH but higher at high and low pH [ Maurer05 ].

Reviews: [ Han06 , Schleiff05 , Miot04 , Behrens02 , Muller01a ]

Front 95

degP

Back 95

Protease Do, or DegP, is a periplasmic serine protease required for survival at high temperatures [ Lipinska89 , Strauch89 , Seol91 ]. DegP degrades abnormal proteins in the periplasm, including mutant proteins, oxidatively damaged proteins and aggregated proteins [ Strauch88 , Strauch89 , SkorkoGlon99 , Laskowska96 ]. DegP has been specifically shown to degrade the mutant periplasmic protein MalS, as well as unassembled subunits from protein complexes, including HflK, LamB and PapA [ Spiess99 , Kihara98 , Misra91 , Jones02 ].

DegP also proteolyzes a range of other proteins that may not be quality control substrates, such as the DNA methyltransferare Ada, various forms of the colicin A lysis protein and the replication initiation inhibitor IciA [ Lee90 , Cavard89 , Cavard95 , Yoo93 ]. DegP also binds to the ssrA-encoded degradation tag, though this PDZ-domain-mediated interaction does not appear to allow DegP proteolysis of tagged proteins [ Spiers02 ]. Finally, strains lacking DegP are more susceptible to the cationic antimicrobial peptide Lactoferricin B, indicating a possible role for DegP in degradation of that molecule [ Ulvatne02 ].

DegP also has an independent chaperone activity that functions even in proteolytically inactive mutants of DegP [ Spiess99 ]. This chaperone activity is required for survival in the case of disrupted outer membrane assembly, preventing buildup of toxic aggregates [ Misra00 ]. There may be some redundancy between DegP and the chaperones Skp and SurA [ Rizzitello01 ].

DegP is a six-membered ring-shaped structure with a central cavity which contains its proteolytic sites [ Swamy83 , Kim99 ]. The hexamer is built from a pair of staggered trimeric rings, with the proteolytic cavity accessible from the sides rather than the ends [ Krojer02 ]. There are two PDZ domains in each monomer which are required for this assembly, and which may be involved in opening and closing the lateral openings [ Sassoon99 ]. Binding of substrate to the PDZ1 domain induces oligomer conversion from a resting hexameric state to a higher order active complex [ Krojer10 , Merdanovic10 ]. The PDZ1 domain anchors substrate, facilitating its presentation to the proteolytic domain [ Krojer08 ]. DegP is a processive protease - cleaving its substrate into peptides with a mean size of 13-15 residues [ Krojer08 ]. The PDZ1 domain is required for protease activity and for binding of unfolded proteins, while the PDZ2 domain is primarily required for maintaining a hexameric configuration [ Iwanczyk07 ]. The inner cavity also has several hydrophobic patches, which may be involved in its chaperone function [ Krojer02 ].

Hexameric DegP assembles into large catalytically active spherical structures around its substrate [ Krojer08a , Jiang08 ]. The spherical multimers exhibit proteolytic and chaperone-like activity [ Shen09 ]. A model polypeptide substrate binds each DegP subunit at two sites in the crystal structure of a DegP dodecamer [ Kim11 ]. Substrate binding drives the formation of proteolytically active dodecamers and larger cages of 18, 24 and 30 subunits while substrate cleavage promotes cage disassembly [ Kim11 ].

DegP's proteolytic activity is increased at high temperatures but drops dramatically at low temperatures, leaving its chaperone function unaffected [ SkorkoGlon95 , Spiess99 ]. DegP interacts with phosphatidylglycerol on the periplasmic face of the inner membrane, undergoing a conformational change that correlates with the temperature dependence of its proteolytic capacity [ SkorkoGlon97 ].

The mature form of DegP is derived by cleavage of its first twenty-six amino acids by leader peptidase [ Lipinska90 , Lipinska89 ]. Targeting of DegP to the Sec-translocase for transport across the inner membrane is SecB-dependent [ Baars06 ].

DegP is a member of the HtrA (high temperature requirement) family of proteases which combine a protease domain with one or more PDZ domains and function as higher order oligomers [ Kim05 ].

DegP is downregulated during low osmolarity [ Forns05 ]

Front 96

ftsZ

Back 96

Assembly of FtsZ into a ring structure (the Z ring, [ Bi91 ]) at the future cell division site is the earliest known event in cell division. FtsZ is the most highly conserved of the proteins that eventually comprise the septal ring structure; homologs of FtsZ are nearly universally present in bacteria as well as in many archaea, some chloroplasts and a few mitochondria [ Vaughan04 ].

FtsZ is essential [ Dai91 ]; it binds GTP and has Mg2+-dependent GTPase activity [ RayChaudhu92 , deBoer92 ]. Assembly of FtsZ into protein filaments is dynamic and regulated by GTP hydrolysis [ Mukherjee94 ], resembling tubulin [ Mukherjee98 ]. Turnover of FtsZ within the Z ring is extremely rapid, with the rate-limiting step for turnover likely to be GTP hydrolysis [ Anderson04 , Romberg04 ].

The position of the FtsZ ring structure marks the cell division site and is serving as the assembly point to which other proteins of the cell division machinery are recruited. Based on studies of mutants, FtsA and ZipA initially associate with FtsZ, followed by addition of FtsEX, FtsK, FtsQ, FtsL/FtsB, FtsW, FtsI, FtsN and AmiC, apparently in this defined order.

The question of how FtsZ itself is positioned precisely at mid-cell, enabling a normal cell division event, has not been definitively answered yet. Both nucleoid occlusion and the Min system appear to play a role in division site selection; the hypothesis has been discussed in detail [ Norris04 ]. Nucleoid occlusion describes a process by which the presence of the nucleoid inhibits Z ring formation at that site. When the nucleoid structure is perturbed by a block in transcription, nucleoid occlusion is affected; thus, the process may depend on the specific organization of the nucleoid [ Sun04 ]. Both SulA and MinC are negative regulators of Z ring assembly [ Bi93 ]. SulA interacts directly with FtsZ [ Higashitan95 , Cordell03 ] and inhibits GTPase activity and polymerization in vitro [ Mukherjee98a , Trusca98 ] and FtsZ ring formation in vivo [ Justice00 ]. MinC also interacts with FtsZ directly and prevents FtsZ polymerization without inhibiting its GTPase activity [ Hu99 , Pichoff01 ]. The MinCD proteins oscillate between the two cell poles; this behavior may allow Z ring formation at mid-cell because the time-integrated concentration of MinCD is lowest at mid-cell. A conflicting report asserts that MinCD does not block Z ring formation, but instead blocks FtsA association with the Z ring [ Justice00 ].

In studies using GFP-labeled FtsZ, FtsZ outside of the Z ring was found to move rapidly in a helix-like pattern along the cell, similar to the movements of the Min proteins. The presence of a dynamic, helical cytoskeleton was proposed [ Thanedar04 ]. Supporting this hypothesis, FtsZ was found to be involved in maintaining cell shape [ Varma04 ].

The domain structure of FtsZ has been described [ Vaughan04 ]. A highly conserved central domain is structurally and functionally homologous to tubulin; it contains the dimerization domain [ Di99 ]. The C-terminal core domain consists of 12 amino acids essential for FtsA and ZipA binding [ Ma99a ].

The crystal structure of a C-terminal fragment of FtsZ in complex with ZipA has been solved [ Mosyak00 ].

Compounds with activity against E. coli FtsZ, with potential utility as broad-spectrum antimicrobials, have been recently isolated and characterized [ Margalit04 ].

Regulation of FtsZ expression and activity is complex and has been summarized in [ Addinall02 ].

Selected reviews: [ Weiss04 , Romberg03 , Errington03 ]

Front 97

acnB

Back 97

bifunctional aconitate hydratase 2 and 2-methylisocitrate dehydratase

There are two aconitases in E. coli, both of which catalyze the reversible isomerization of citrate and iso-citrate via cis-aconitate. AcnB also plays a role in the methylcitrate cycle for degradation of propionate, where it is responsible for hydration of 2-methyl-cis-aconitate to (2R,3S)-2-methylisocitrate [ Brock02 ]. The apo form of AcnB is able to bind mRNA and enhances translation of AcnB [ Tang99 ].

AcnB appears to function as the main catabolic enzyme, while the main role of AcnA appears to be as a maintenance or survival enzyme during nutritional or oxidative stress [ Cunningham97 ]. The AcnB enzyme is less stable, has a lower affinity for citrate and is active over a more narrow pH range than the AcnA enzyme [ Jordan99a , Varghese03 ]. Unlike AcnA, AcnB is sensitive to oxidation in vivo [ Brock02 , Varghese03 ]. AcnB rapidly loses catalytic activity when the iron concentration is low [ Varghese03 ].

The N-terminal region of AcnB mediates the formation of AcnB homodimers in the presence of Fe2+; the 4Fe-4S cluster or catalytic activity is not required for dimer formation. In the absence of Fe2+, the same region is able to bind to mRNA [ Tang05 ]. AcnB also interacts weakly with isocitrate dehydrogenase [ Tsuchiya08 ]. The catalytically inactive AcnB apo-protein, lacking its iron-sulfur cluster, has a negative effect on SodA synthesis in vitro [ Tang02 ].

A crystal structure of AcnB has been solved at 2.4 Å resolution [ Williams02 ].

An acnB mutant does not grow on acetate as the sole source of carbon, grows poorly on other carbon sources such as glucose and pyruvate [ Gruer97 ], contains high levels of citrate, and excretes substantial amounts of citrate into the medium [ Varghese03 ]. An acnB mutant is more sensitive to peroxide stress than wild type and shows increased SodA synthesis [ Tang02 ].

Expression of acnB increases early in exponential phase and decreases during entry into stationary phase [ Gruer97 , Cunningham97 ].

Reviews: [ Gruer97a , Kiley03 ]

Front 98

pcnB

Back 98

Poly(A) polymerase I is responsible for the polyadenylation of 3' ends of RNA molecules.

Poly(A) polymerase polyadenylates the vast majority of mRNA transcripts [ Mohanty06 ]. Unlike in eukaryotes, increased polyadenylation of mRNAs leads to decreased mRNA half-life [ Mohanty99 , Mohanty06 ]. Rho-independent transcription terminators appear to serve as targeting signals for polyadenylation [ Mohanty06 ]. The Hfq protein appears to be involved in the recognition of 3' termini of RNA by poly(A) polymerase I [ Le03 ].

Intracellular levels of poly(A) polymerase I as well as the level of pcnB transcription vary inversely with growth rate [ Jasiecki03 ]. Overexpression of poly(A) polymerase I is toxic and leads to slowed growth [ Mohanty99 , Mohanty06 ]. Use of AUU as the translational start codon results in InfC discrimination (as with production of the IF-3 translation initiation factor) and results in low levels of poly(A) polymerase I in the cells [ Binns02 ].

A his-tagged version of poly(A) polymerase I showed reduced activity following phosphorylation of its his tag. In other proteins, this sometimes correlates with the protein itself being regulated via phosphorylation [ Jasiecki06 ]

Front 99

map

Back 99

methionine aminopeptidase

All known proteins in E. coli use N-formyl methionine as the first amino acid in a peptide chain. Amino-terminal maturation involves two enzymes, a deformylase which removes the formyl group, and methionine aminopeptidase (MAP), which catalyzes the removal of the deformylated methionine residue [ Neidhardt96 , BenBassat87 ]. The activity of MAP is dependent on the identity of the second, third and fourth amino acid residues of the target protein [ Hirel89 , Frottin06 , BenBassat87 ]; the substrate specificity has been analyzed in detail [ Xiao10 ]. The most preferred amino acid residue in the position following the fMet is alanine.

The map gene encoding methionine aminopeptidase is essential for growth in E. coli [ Chang89 ]. E. coli MAP is a type-I enzyme and is a potential antibiotic target; selective inhibitors have been designed [ Swierczek05 , Wang09 , Mitra09 ].

Crystal structures of the protein itself and in complex with various inhibitors have been reported [ Roderick93 , Lowther99 , Ye04 , Huang06a , Evdokimov07 , Huang07 , Ma07b , Wang08a ], and the catalytic mechanism was studied in cocrystals with various substrate and transition state analogs [ Lowther99a , Ye06 ].

MAP contains two metal binding sites; reports differ on whether a single metal ion is sufficient for catalysis [ Cosper01 , Ye06 , Huang07 , Chai09 ], or both are required [ Hu07 , Mitra08 ]. The physiologically relevant metal cofactor of MAP is most likely Fe2+ rather than Co2+ [ Chai08 ]. Studies with inhibitors that were active in vitro, but not in vivo, indicated that high metal concentrations in in vitro assays may have led to artefacts [ Schiffmann05 , Schiffmann06 ].

The activity of enzymes containing mutations in predicted active site and metal-binding residues has been measured [ Chiu99 , Copik03 , Li04 , Watterson08 , Mitra09a , Mitra08 ]

Front 100

rpsB

Back 100

The S2 protein, a component of the 30S subunit of the ribosome, is essential in E. coli [ Bollen79 ]. S2 is required for S1 binding to the ribosome [ Moll02 ]. Overexpression of csdA, a DEAD-box RNA helicase, supresses the defect of a temperature-sensitive allele of rpsB [ Toone91 ] by restoring assembly of both S1 and S2 with the ribosome at the non-permissive temperature [ Moll02 ].

Expression of S2 is autoregulated, involving conserved elements located in the 5' UTR of the rpsB-tsf mRNA. Regulation may also involve the S1 protein [ Aseev08 ].

Purified S2 protein binds to inorganic polyphosphate (polyP). In the presence of polyP, S2 forms a complex with the Lon protease and is degraded by it [ Kuroda01 , Nishii05 ]. S2 was also shown to crosslink to IF3 [ Cooperman81 ].

A low-resolution cryo-electron microscopy map of the ribosome containing S2 has been analyzed [ Gao03 ]

Front 101

clpP

Back 101

ClpP is a serine protease with a chymotrypsin-like activity that is a part of the ClpAP, ClpAPX and ClpXP protease complexes [ Arribas93 , Wang97e ].

The ClpP protease is a tetradecamer, consisting of two heptamers of ClpP subunits stacked head-to-head [ Kessel95 , Shin96 ]. ClpP has an axial pore large enough to accept unfolded polypeptide chains, leading into a central cavity that contains fourteen serine protease active sites [ Flanagan95 , Wang98j ]. This ring structure is required for proper protease function [ Thompson98a ]. Serine-111 and histidine-136 are also required for protease function [ Maurizi90 ]. The interface between the two heptameric rings can switch between two different conformations; limiting this switching via crosslinking slows substrate release [ Sprangers05 ].

Translocation of polypeptide substrates into ClpP is directional, with the carboxy-terminus going first [ Reid01 ].

ClpP degrades the antitoxin proteins Phd and MazE from the toxin/antitoxin pairs phd-doc (from plasmid prophage P1) and mazEF (from the rel plasmid). The lysogenically expressed lambda protein lambdarexB inhibits this proteolysis [ EngelbergK98 ].
Lambda protein gpW mutants with hydrophobic tails are degraded in a ClpP-dependent manner [ Maxwell00 ].
ClpP is required for normal adaptation to and extended viability in stationary phase, and for growth in SDS [ Weichart03 , Rajagopal02 ].
ClpP is a heat shock protein expressed in a sigma 32-dependent manner [ Kroh90 ]. It has a 14-amino acid leader peptide which is cleaved intermolecularly by another ClpP without any requirement for associated ClpA [ Maurizi90a , Maurizi90 ]

Front 102

cyoA

Back 102

CyoA is subunit II of the cytochrome bo terminal oxidase complex encoded by cyoABCDE. Crosslinking studies suggested that subunit II functions as a ubiquinone binding site of the cytochrome bo terminal oxidase complex [ Welter94 ]. However, the crystal structure of the entire cytochrome bo terminal oxidase complex suggests that a potential ubiquinone binding site is instead located in the membrane domain of subunit I [ Abramson00 ].
The CyoA polypeptide contains two transmembrane helices [ Chepuri90a ]. CyoA is a lipoprotein; during maturation, the protein is modified by attachment of fatty acids and protease cleavage at C25; however, the posttranslational modification is not essential for assembly or activity of the cytochrome bo terminal oxidase complex [ Ma97 , Brokx04 ]. In vitro models indicate that posttranslational modifications to CyoA occur after membrane insertion. The same models also demonstrate that YidC, the Sec translocon (Sec YEG) and SecA are required for efficient insertion of cyoA into the membrane [ duPlessis06 ].
Protease mapping assays, using strains with inactivated or depleted components of the Sec translocon, SecA and YidC, indicate that membrane integration and assembly occur in separate, sequential steps; insertion of the large periplasmic C-terminal segment requires both the Sec translocon and SecA whereas YidC is sufficient for insertion of the N-terminal domain [ vanBloois06 ].
Mutation of residues within the signal peptide and first hydrophobic domain of CyoA, resulting in a non-neutral overall charge of the first periplamic domain, confirm results obtained in in vitro studies [ duPlessis06 ] and show that the overall neutral charge in the first periplasmic domain of CyoA is required for membrane insertion [ Celebi08 ].
The three-dimensional structure of the periplasmic fragment of CyoA has been determined to 2.3 Å resolution [ vanderOost93 , Wilmanns95 ]. A crystal structure of the entire cytochrome bo terminal oxidase complex containing CyoA has been determined at 3.5 Å resolution [ Abramson00 ].
Under anaerobic conditions, cyoA is repressed 140-fold compared to growth under aerobic conditions. This regulation is in part due to repression by Fnr [ Cotter90 ]

Front 103

cyoD

Back 103

CyoD is subunit IV of the cytochrome bo terminal oxidase complex encoded by cyoABCDE. Although this subunit's function is unknown, it is necessary for a functional enzyme [ Neidhardt96 ].

The CyoD polypeptide contains three transmembrane helices [ Chepuri90a ]. Deletion and cross-linking studies have suggested that subunit IV interacts with subunits I and III [ Saiki96 ], which is confirmed by the crystal structure of the entire cytochrome bo terminal oxidase complex that has been determined at 3.5 Å resolution [ Abramson00 ]

Front 104

tilS

Back 104

TilS is a tRNAIle-lysidine synthetase, the enzyme responsible for modifying the wobble base of the CAU anticodon of tRNAIle at the keto group in position 2 of C34 such that it exhibits proper recognition of the AUA codon rather than the AUG codon [ Soma03 ]. This modification is necessary and sufficient for correct charging of the tRNA with isoleucine (not methionine) [ Muramatsu88 , Soma03 , Salowe09 ].
TilS recognizes and modifies the precursor form of tRNAIle, thus ensuring the fidelity and avoiding errors in translation. Unmodified C34 is only found in pre-tRNAIle; mature tRNAIle contains the modified L34 base. Thus, it is not possible to generate mischarged Met-tRNAIleCAU [ Nakanishi09 ].
TilS recognizes the antidodon loop, the anticodon stem, and the acceptor stem of tRNA, allowing it to discriminate between tRNAIle and tRNAMet, which are structurally similar and share the same anticodon loop [ Ikeuchi05 ]. Various analogs of L-lysine have been tested as inhibitors or alternative substrates for the enzyme [ Salowe09 ].
tilS is an essential gene [ Soma03 , Baba06 ]. A tilS reduction-of-function mutation causes defects in translation of AUA codons [ Soma03 ].
TilS: tRNA(Ile)-lysidine synthetase [ Soma03 ]

Front 105

queA

Back 105

S-adenosylmethionine:tRNA ribosyltransferase-isomerase (QueA) catalyzes addition of the 2,3-epoxy-4,5-dihydroxycyclopentane ring of epoxyqueuosine (oQ) to preQ1. This is the penultimate step in the formation of the queuosine (Q) anticodon loop modification in tRNA(Asp), tRNA(Asn), tRNA(His), and tRNA(Tyr) [ Slany93 ].

The enzyme transfers the ribosyl moiety of SAM to preQ1 and isomerizes it to the epoxycyclopentane residue of oQ [ Slany93 , Slany94 ]. A chemical reaction mechanism has been proposed [ Kinzie00 ]. Inhibition studies are consistent with a fully ordered sequential bi-ter kinetic mechanism in which preQ1-tRNA binds first followed by SAM, with product release in the order adenine, methionine, and oQ-tRNA [ Van03 ]. Recognition of the tRNA substrate is mediated via the anticodon loop region [ Mueller95a , Van03c ].

A queA mutant accumulates preQ1-modified tRNAs [ Okada78 , Reuter91 ]. The growth rate of the queA mutant is approximately half of that of the wild-type parent strain [ Okada78 ]

Front 106

tig

Back 106

In Escherichia coli Trigger Factor (TF), the tig gene product, is one of three major (along with DnaK and GroEL) chaperones which cooperate in the folding of newly synthesized cytosolic proteins [ Gragerov92 , Crooke88 , Hesterkamp96 ]. The majority of nascent polypeptides interact first with TF [ Deuerling99 ] at the ribosome, where TF binds at proteins L23/L29 at the polypeptide exit site [ Kramer02 ]. TF is thought to interact primarily with short nascent chains [ Valent95 ] and its function overlaps with that of DnaK, which interacts with longer nascent chains downstream of TF [ Teter99 ]. Although deletion mutants of either tig or dnaK are viable, their combined deletion is lethal at temperatures above 30 degrees [ Deuerling99 ]. In studies using biochemical assays [ Hesterkamp96 ] TF was found to be capable of cross-linking to virtually all nascent and cytoplasmic proteins and to perform ATP-independent chaperone-like functions [ Scholz97 ]. TF substrates have been identified by TF copurification and TF resolvable aggregation. Many of the substrate are subunits of multimeric complexes including many ribosomal proteins and a role for TF in the biogenesis of protein complexes has been suggested [ MartinezHa09 ].
E. coli TF consists of three domains [ Ferbitz04 ] all of which interact with the nascent polypeptide during translation [ Lakshmipat07 ]. The N-terminal domain contains the TF signature motif which mediates ribosome binding [ Kramer02 , Kristensen03 ]. The middle domain displays peptidyl-prolyl cis/trans isomerase (PPIase) activity and is dispensable for the in vivo chaperone activity of TF [ Genevaux04 ]. The C-terminal domain is essential for chaperone activity [ Merz06 , Zeng06 ]

Front 107

cyoE

Back 107

The CyoE protein, heme O synthase, catalyzes the synthesis of heme O, which is essential for the catalytic fuctions of the cytochrome bo oxidase complex [ Saiki92 , Saiki93a ].

At one time it was thought that CyoE was a fifth subunit of the cytochrome bo oxidase, but it was shown that a 28 kDa polypeptide which co-purifies with the cytochrome bo oxidase complex appears even in a cyoE deletion strain [ Saiki93 ]

Front 108

cyoC

Back 108

CyoC is subunit III of the cytochrome bo terminal oxidase complex encoded by cyoABCDE.

The CyoC polypeptide contains five transmembrane helices [ Chepuri90a ]. A crystal structure of the entire cytochrome bo terminal oxidase complex containing CyoC has been determined at 3.5 Å resolution [ Abramson00 ]

Front 109

lon

Back 109

Lon is an ATP-dependent protease responsible for degradation of misfolded proteins as well as a number of rapidly degraded regulatory proteins. Key regulatory proteins that are Lon substrates include the cell division regulator SulA [ Schoemaker84 , Higashitan97 ], the capsule synthesis regulator RcsA [ TorresCaba87 ] and possibly TER components involved in blocking septation sites during the SOS response [ Dopazo87 ]. Lon is required for degradation of misfolded proteins and the prevention of aggregate formation [ Chung81 , Ryzhavskai , Laskowska96a ]. In the absence of Lon function, aggregation triples [ Rosen02 ]. At least some of this degradation of misfolded proteins depends on the chaperone DnaK [ Sherman92b ].
*Lon also degrades the lamba phage N and Xis proteins, with degradation of the latter promoting lysogeny over lysis [ Maurizi87 , Leffers98 ]. Other substrates included HU1 in the absence of its partner HU2, HemA and DAM methylase [ Bonnefoy89 , Wang99e , Calmann03 ].
*Lon degrades the antitoxin protein in many toxin/antitoxin protein pairs, including both plasmid and chromosomal versions. Lon proteolysis of the antitoxin protein in plasmid-encoded pairs is required for plasmid maintenance, as the antitoxin has a shorter half life in lon+ cells than the toxin, thus requiring the continued presence of the plasmid for cell survival. Plasmid-encoded antitoxin substrates include CcdA from F plasmid, relBP307 and PasA [ Van96b , Van94b , Gronlund99 , Smith98b ]. Lon proteolyzes chromosomal toxin/antoxin pairs as well, including RelB and YoeB [ Christense01 , Christense04a ]. This degradation of chromosomal pairs may regulate part of the starvation stress response, as the breakdown of RelB leaves RelE, which suppresses translation [ Christense01 ]. Starvation-induced transcription of chpA also depends on Lon [ Christense03a ].
*Lon is an ATP-dependent protease with chymotrypsin-like specificity based on a Serine (679)-Lysine (722) dyad [ Charette81 , Waxman85 , Botos04 , Nishii05 ]. Lon has one proteolytic and four ATP-binding sites, two high affinity, the other two low affinity [ Chin88 , Menon87 ]. Detailed kinetic analysis shows that the two types of ATP sites use ATP at different rates as well [ Vineyard06 ]. Lon's protease function depends on its ATPase activity; both require Mg2+, two ATP molecules are used per peptide bond hydrolyzed and loss of ATPase functions leads to concomitant loss of peptidase function [ Menon87 , Menon87a , Fischer94 , Waxman82 ]. ATPase activity continues in mutants that are unable to proteolyze [ Pohl76 ]. Though its ATPase activity is required for protein degradation, Lon is capable of breaking down small peptides in the absence of ATP or ATPase function [ Goldberg85 , Rasulova98 ]. The isolated ATPase domain undergoes conformational change in response to ADP and ATP binding [ Vasilyeva02 ].
*Lon has an independent protein-binding domain in addition to its proteolytic domain. This domain binds unfolded proteins [ Chin88 ]. Protein binding substantially stimulates peptide degradation and ATPase activity, the latter even in mutants incapable of peptidase function [ Waxman86 , Pohl76 ].
*Lon binds DNA via its DNA-binding domain [ Charette81 , Chin88 ]. Addition of DNA, especially denatured DNA, stimulates substrate proteolysis in vitro, as well as stimulating ATPase activity even in the absence of substrate [ Chung82 , Charette84 ]. Lon may have specificity for promoter regions, explaining how it targets regulatory proteins [ Fu97 ].
*Lon can form a complex with inorganic polyphosphate, allowing subsequent degradation of ribosomal proteins, including S2, L9 and L13 [ Kuroda01 ]. Lon's DNA-binding domain binds polyphosphate with greater affinity than DNA [ Nomura04 ]. Lon complexed with polyphosphate may be an octamer instead of a tetramer [ Nishii05 ]

Front 110

lon

Back 110

DNA-binding, ATP-dependent protease La
Lon is an ATP-dependent protease responsible for degradation of misfolded proteins as well as a number of rapidly degraded regulatory proteins. Key regulatory proteins that are Lon substrates include the cell division regulator SulA [ Schoemaker84 , Higashitan97 ], the capsule synthesis regulator RcsA [ TorresCaba87 ] and possibly TER components involved in blocking septation sites during the SOS response [ Dopazo87 ]. Lon is required for degradation of misfolded proteins and the prevention of aggregate formation [ Chung81 , Ryzhavskai , Laskowska96 ]. In the absence of Lon function, aggregation triples [ Rosen02 ]. At least some of this degradation of misfolded proteins depends on the chaperone DnaK [ Sherman92 ].

Lon also degrades the lamba phage N and Xis proteins, with degradation of the latter promoting lysogeny over lysis [ Maurizi87 , Leffers98 ]. Other substrates included HU1 in the absence of its partner HU2, HemA and DAM methylase [ Bonnefoy89 , Wang99b , Calmann03 ].

Lon degrades the antitoxin protein in many toxin/antitoxin protein pairs, including both plasmid and chromosomal versions. Lon proteolysis of the antitoxin protein in plasmid-encoded pairs is required for plasmid maintenance, as the antitoxin has a shorter half life in lon+ cells than the toxin, thus requiring the continued presence of the plasmid for cell survival. Plasmid-encoded antitoxin substrates include CcdA from F plasmid, relBP307 and PasA [ Van96a , Van94 , Gronlund99 , Smith98 ]. Lon proteolyzes chromosomal toxin/antoxin pairs as well, including RelB and YoeB [ Christense01 , Christense04 ]. This degradation of chromosomal pairs may regulate part of the starvation stress response, as the breakdown of RelB leaves RelE, which suppresses translation [ Christense01 ]. Starvation-induced transcription of chpA also depends on Lon [ Christense03a ].

Lon is an ATP-dependent protease with chymotrypsin-like specificity based on a Serine (679)-Lysine (722) dyad [ Charette81 , Waxman85 , Botos04 , Nishii05 ]. Lon has one proteolytic and four ATP-binding sites, two high affinity, the other two low affinity [ Chin88 , Menon87 ]. Detailed kinetic analysis shows that the two types of ATP sites use ATP at different rates as well [ Vineyard06 ]. Lon's protease function depends on its ATPase activity; both require Mg2+, two ATP molecules are used per peptide bond hydrolyzed and loss of ATPase functions leads to concomitant loss of peptidase function [ Menon87 , Menon87a , Fischer94 , Waxman82 ]. ATPase activity continues in mutants that are unable to proteolyze [ Pohl76 ]. Though its ATPase activity is required for protein degradation, Lon is capable of breaking down small peptides in the absence of ATP or ATPase function [ Goldberg85 , Rasulova98 ]. The isolated ATPase domain undergoes conformational change in response to ADP and ATP binding [ Vasilyeva02 ].

Lon has an independent protein-binding domain in addition to its proteolytic domain. This domain binds unfolded proteins [ Chin88 ]. Protein binding substantially stimulates peptide degradation and ATPase activity, the latter even in mutants incapable of peptidase function [ Waxman86 , Pohl76 ].

Lon binds DNA via its DNA-binding domain [ Charette81 , Chin88 ]. Addition of DNA, especially denatured DNA, stimulates substrate proteolysis in vitro, as well as stimulating ATPase activity even in the absence of substrate [ Chung82 , Charette84 ]. Lon may have specificity for promoter regions, explaining how it targets regulatory proteins [ Fu97 ].

Lon can form a complex with inorganic polyphosphate, allowing subsequent degradation of ribosomal proteins, including S2, L9 and L13 [ Kuroda01a ]. Lon's DNA-binding domain binds polyphosphate with greater affinity than DNA [ Nomura04 ]. Lon complexed with polyphosphate may be an octamer instead of a tetramer [ Nishii05 ]

Front 111

allC

Back 111

allantoate amidohydrolase
E. coli is able to utilize allantoin as a sole nitrogen source under anaerobic conditions [ Cusa99 ]. Following S-allantoin ring opening, allantoate amidohydrolase converts allantoate to S-ureidoglycine , liberating one molecule each of ammonia and CO2 [ Werner10 , Serventi09 ]. Before the identification of S-ureidoglycine aminohydrolase , allantoate amidohydrolase was thought to catalyze both hydrolysis reactions of the allantoin degradation pathway, generating S-(-)-ureidoglycolate directly [ Cusa99 , Agarwal07 ].

The crystal structure of allantoate amidohydrolase has been solved at 2.25 Å resolution. It exists as a homodimer in solution and the crystal, with each polypeptide chain folded into two domains - a large catalytic domain containing two metal ions, sulfate, and substrate binding sites, and a smaller dimerization domain. Due to the overall similarity of the protein to the di-zinc-dependent metallopeptidases, the metal ions were thought to be zinc. The sulfate ion may be an allosteric effector. A reaction mechanism was proposed [ Agarwal07 ].

Allantoate amidohydrolase activity is induced by growth on allantoin as the sole source of nitrogen [ Cusa99 ]

Front 112

ompT

Back 112

OmpT is an outer membrane protease with specificity for paired basic residues [ Sugimura88 ]; detailed studies on substrate specificity have been performed [ Dekker01 , Okuno02 , McCarter04 ]. Cellular localization studies have shown that OmpT localizes to the cellular poles [ Lai04 ]. Targeting of OmpT to the Sec-translocase for transport across the inner membrane is SecB-dependent [ Baars06 ]. In vitro studies have investigated the strategies used by a number of outer-membrane proteins, including OmpT, to efficiently fold into the membrane [ Burgess08 ].

OmpT is active under extreme denaturing conditions and shows a preference for denatured substrates [ White95a ]. It is responsible for hydrolysis of the antimicrobial peptide protamine [ Stumpe98 ] and is a virulence determinant in urinary tract infections [ Kanamaru03 ].

OmpT has been shown to be responsible for cleavage of the endonuclease colicin E2 (ColE2), a bacteriocidal protein, and the associated cognate immunity protein (Im2) in the presence of BtuB. Specifically, OmpT cleaves the C-terminal DNase domain of ColE2 [ Duche09 ].

Kinetic parameters of the purified protein have been determined [ Kramer00 ].

Based on information from a crystal structure, a novel catalytic mechanism was suggested [ Vandeputte01 ]. Site-directed mutagenesis studies had previously identified essential active site residues Ser99 and His212 [ Kramer00a ]; mutational studies of additional residues support a novel catalytic mechanism [ Kramer01 ].

OmpT expression and activity increases in response to heat shock as well as overexpression of recombinant proteins [ Gill00 , Jurgen00 , Gill00a ]. Expression is regulated by the EvgA/EvgS two-component system [ Eguchi04 ]

Front 113

glnS

Back 113

glutaminyl-tRNA synthetase
Glutaminyl-tRNA synthetase (GlnRS) is a member of the family of aminoacyl tRNA synthetases, which interpret the genetic code by covalently linking amino acids to their specific tRNA molecules. The reaction is driven by ATP hydrolysis. GlnRS belongs to the Class I aminoacyl tRNA synthetases [ Eriani90a , Landes95 ]; apart from sequence motifs within the active site, the different enzymes show little similarity in their primary amino acid sequences.

Substrate recognition and discrimination by GlnRS has been studied extensively. Binding of tRNA and ATP to GlnRS is cooperative, and transfer of the aminoacyl adenylate to the tRNA is the rate determining step [ Lloyd95 ]. Interactions between the tRNA identity nucleotides and GlnRS modulate the affinity of the enzyme for glutamine [ Ibba96 ], and the terminal adenosine base of tRNAGln mediates amino acid recognition [ Liu98f ]. tRNA binding to GlnRS results in conformational changes in the active site [ Sherlin03 ]; induced-fit changes in the active site structure by both amino acid and tRNA binding may contribute to enzyme specificity [ Uter05 ]. Reports differ on whether the properties of the amino acid binding site allow the enzyme to specifically recognize glutamine and select against glutamate [ Rath98 ] or function as a negative determinant, binding glutamate in a non-productive orientation [ Bullock03 ]. The glutaminyl adenylate intermediate is hydrolyzed by the GlnRS-tRNAGln complex and is analogous to pre-transfer editing reactions catalyzed by editing aminoacyl-tRNA synthetases like IleRS [ GruicSovul05 ]. A substrate-assisted catalytic mechanism is supported by single-turnover kinetic analysis [ Uter06 ]. Molecular dynamics simulations of tRNA binding to GlnRS have been performed [ Yamasaki07 ].

Various crystal structures of GlnRS have been solved [ Rould89 , Rould91 , Arnez94 , Perona93 , Arnez96 , Bullock00 , Bullock03 , Sherlin03 ].

GlnRS mutants that allow recognition and mischarging of su+3 tRNATyr and others have been identified [ Inokuchi84 , Uemura88 , WeygandDur93 , Rogers94 , WeygandDur94 , Sherman96 , Sherman96a , Hong96 , Kitabatake96 ], and their relaxed specificity has been interpreted in light of their location within the crystal structure [ Perona89 , Steitz90 , Arnez96 ]. Overexpression of wild type GlnRS also leads to mischarging of su+3 tRNATyr [ Hoben84 ]; concomitant overexpression of the cognate tRNAGln2 abolishes this effect, indicating that proper balance between the tRNA and the aminoacyl-tRNA synthetase is critical [ Swanson88 ]. A "minimal" GlnRS enzyme retains catalytic activity [ Schwob93 ].

Specificity determinants within tRNA that are important for recognition by GlnRS have been identified; see, for example, [ Rogers92 , Hayase92 , Wright93 , McClain93a , Sherlin00 ] and references therein.

GlnRS enzyme levels increase with increasing growth rate [ Cheung85 ]. GlnRS expression is regulated at the transcriptional as well as posttranscriptional levels [ Cheung85 , Kuriki90 , Faxen91 ].

The tRNAGln/GlnRS pair has been engineered for incorporation of unnatural amino acids into proteins [ Liu97f , Liu97g ].

Reviews: [ Freist97a , Ibba96a , Ibba95 , Rogers93a , EnglischPe91 ]
The glnS gene expression could be repressed in a direct or indirect way by the OmpR protein. Since, using gene arrray system, a repression of this gene was observed as a result of an up mutation in the ompR gene [ Brombacher03 ]

Front 114

pth

Back 114

Peptidyl-tRNA hydrolase (Pth) catalyzes the removal of peptides from peptidyl-tRNAs that have been prematurely released from the ribosome. The enzyme is essential for growth in E. coli [ Menninger79 , Singh04a ]. lysV, encoding a tRNALys, was found to be a multicopy suppressor of a temperature sensitive pth mutant, indicating that the pth mutant phenotype is due to starvation for tRNALys [ HeurgueHam96 , VivancoDom06 ]. In general, Pth relieves growth inhibition by freeing tRNAs from aborted translation products [ Tenson99 , Dincbas99 ]. Pth is also required for growth of bacteriophage lambda due to translation of minigene ORFs in the lambda bar region [ Henderson76 , Ontiveros97 , Valadez01 , Oviedo04 ].

The crystal structure of Pth has been solved at 1.2 Å resolution [ Schmitt97 , Schmitt97a ]. Pth is monomeric. Predicted active site residues were confirmed by site-directed mutagenesis [ Schmitt97a ]. The His20 residue is essential for catalytic activity [ Goodall04 ]. The catalytic center has been mapped by NMR [ Giorgi11 ].

The integrity of ychF, the downstream open reading frame in the pth operon, influences the stability of the transcript. Lower RNase E activity results in overproduction of Pth, and thus it is proposed that RNase E processing within ychF causes mRNA degradation [ CruzVera02 ].

A temperature-sensitive pth mutant was isolated [ Atherly72 , Menninger73 ]; the mutant protein is unstable in vivo [ CruzVera00 ]. The mutant allele can revert rapidly by gene duplication [ Menez01 ]. Overexpression of tmRNA can rescue the defect of the pthts allele [ Singh04a ]. The rap mutant of Pth [ Henderson76 ] is highly sensitive to the length of the peptidyl-tRNA [ HeurgueHam00 ]

Front 115

sdhC

Back 115

succinate dehydrogenase membrane protein
SdhC is one of two membrane proteins in the four subunit succinate dehydrogenase (SQR) enzyme. SdhC and SdhD are the large and small subunits of cytochrome b556, respectively [ Nakamura96a ]. The quinone binding (Qp) site resides in the interface between SdhB, SdhC and SdhD [ Tran06a ].

The b556 type heme bridges both membrane subunits [ Maklashina99 , Nakamura96a ]. Mutation of key heme binding residues in SdhC and SdhD does not affect proper assembly or physiological function of the complex [ Tran07 ].

Despite similar function, hydrophobicity, and protein size, the SdhC and SdhD subunits of succinate dehydrogenase do not share significant sequence identity with the corresponding membrane-binding subunits of fumarate reductase, FrdC and FrdD [ Wood84 ]

Front 116

cydB

Back 116

cytochrome bd-I terminal oxidase subunit II
CydB is required for binding the heme b595 component and one heme d (iron-chlorin) component of cytochrome bd-I [ Newton91 ].

A cydB mutant has a temperature sensitive growth phenotype [ Wall92 ]

Front 117

rhlE

Back 117

ATP-dependent RNA helicase
RhlE is a ribosome-associated factor that may be involved in ribosome maturation [ Jain08 ].

RhlE is a member of the DEAD-box-containing ATP-dependent RNA helicase family [ Ohmori94 ]. Its ATPase activity is stimulated by both long RNAs and short oligoribonucleotides [ Bizebard04 ]. Of the three E. coli DEAD-box RNA helicases CsdA, SrmB and RhlE, only RhlE can unwind substrates with short or no single-stranded extensions, but the enzyme shows low processivity [ Bizebard04 ].

RhlE physically interacts with poly(A) polymerase I [ Raynal99 ] and RNase E [ Khemici04a ]. RhlE can functionally replace RhlB within the degradosome [ Khemici04a ].

An rhlE mutant does not have any obvious growth defect [ Ohmori94 ]. Overexpression of RhlE can substitute for the essential function of CsdA during cold shock [ Awano07 , Jain08 ], but exacerbates the cold-sensitive growth defect of a srmB mutant [ Jain08 ]. Conversely, the cold-sensitive growth defect of an rhlE csdA double mutant is enhanced, while that of an rhlE srmB double mutant is alleviated [ Jain08 ]. Both CsdA and SrmB are involved in ribosome maturation, indicating that RhlE may play a similar role [ Jain08 ]

Front 118

ftsK

Back 118

essential cell division protein FtsK
FtsK is an essential cell division protein linking cell division and chromosome segregation [ Kennedy08 ].

FtsK colocalizes with FtsZ to the septal ring structure; localization is dependent on FtsZ, FtsA and ZipA, but not FtsI and FtsQ [ Yu98a , Wang98i , Pichoff02 ]. Conversely, FtsQ, FtsL and FtsI require FtsK for localization to the Z ring [ Chen01b ]. The FtsK protein domains involved in the interactions with other cell division proteins have been mapped [ Grenga08 ]. When FtsK is overexpressed, Z ring formation and cell division are inhibited [ Draper98 ].

Although the hierarchy of dependency in the assembly of cell division proteins is largely linear, recent results showed that assembly of the cell division machinery is complex [ Goehring05 , Goehring06 ]. The requirement for FtsK for localization of FtsI is indirect [ Geissler05 ]. Overexpression of FtsN partially restores localization of FtsL and FtsQ in the absence of FtsK [ Goehring07a ].

The FtsK protein is very large, and its membrane and cytoplasmic domains appear to have to separable functions during cell division [ Bigot04 ]. The N-terminal domain of FtsK, spanning ~200 amino acids, is sufficient for targeting FtsK to the septum [ Yu98a ] and for its function in cell division [ Draper98 , Wang98i ]. This domain shares some functional overlap with FtsQ, FtsB, FtsA, ZipA and FtsN and may be involved in the stability of the division protein machinery [ Geissler05 ]. It contains four transmembrane helices linking two periplasmic loops, one of which contains a zinc metalloprotease consensus sequence that is essential for function of FtsK [ Dorazi00 ]. As long as it is targeted to the cell division septum, a truncated form of FtsK that lacks all transmembrane segments still functions in the resolution of chromosome dimers, indicating that DNA does not need to be transported through a pore formed by the transmembrane helices [ Dubarry10 ].

Following the N-terminal membrane domain is a ~500 residue cytoplasmic linker region which may function in stabilizing the interactions of FtsK with other cell division proteins [ Bigot04 ]. Two regions within this domain, 179-331 and 332-641, independently interact with FtsZ, FtsQ, FtsL and FtsI and were found to be required for normal septation [ Dubarry10a ].

The C-terminal domain of ~500 amino acids is cytoplasmic and involved in septation and chromosome partitioning [ Liu98g , Yu98b , Steiner99a ]. It can be separated into three domains, α, β and γ [ Massey06 ]. The αβ domain contains a nucleotide binding motif belonging to the AAA family of ATPases [ Begg95 ] and has ATP-dependent DNA translocase activity [ Aussel02 , Saleh04 ], which has been observed in single molecule assays [ Pease05 , Saleh05 ]. The translocation step size is ~2 bp per ATP [ Graham10 ]. The chromosomal domain within which FtsK acts has been identified [ Corre05 ]. DNA translocation by FtsK is directional and is guided by octameric sequences (known as KOPS - FtsK Orienting Polar Sequences - in E. coli) [ Bigot05 , Levy05 , Bigot06 ]. The γ domain is a DNA-binding winged-helix domain that recognizes KOPS [ Sivanathan06 ]. KOPS appear to act only as FtsK loading sites and are not read during DNA translocation [ Graham10 ]. Failure to recognize KOPS has little effect in wild type E. coli, but more serious consequences in cell populations where chromosome dimers occur more frequently [ Sivanathan09 ].

Although FtsK can displace proteins from DNA, DNA translocation by FtsK stops at XerCD-dif sites [ Graham10a ]. FtsK's DNA translocation activity and its ability to displace roadblocks on DNA can be separated [ Crozat10 ].

Recombination between sister chromosomes causes formation of chromosome dimers, which must be resolved by XerCD-mediated recombination between dif sites. The C-terminal domain of FtsK is required for this activity, activating the recombinase and actively positioning the dif sites [ Aussel02 , Capiaux02 , Massey04 , Yates06 ]. In particular, the γ regulatory subdomain activates the formation of a Holliday junction intermediate by XerD; it is proposed that this activation of unlinking is normally coupled to the translocation function of FtsK [ Grainge11 ].

Newly replicated chromosomes are topologically linked and must be decatenated by Topo IV before they can be physically separated. FtsK and Topo IV interact physically via the ParC subunit of Topo IV. Functionally, the C-terminal domain of FtsK stimulates the decatenation activity of Topo IV on positively supercoiled DNA [ Espeli03a , Bigot10 ], but not the activity of DNA gyrase [ Espeli03a ]. Unlinking of chromosomes can also be accomplished by step-wise XerCD-dif recombination in the presence of chromosomal translocation by FtsK [ Grainge07 ]. Unexpectedly, FtsK antagonizes Ter-induced recombination [ Louarn07 ].

The crystal structure of a monomeric point mutant of the motor domain has been solved at 2.7 Å resolution [ Massey06 ], and a solution structure of the γ winged-helix domain was obtained [ Sivanathan06 ].

The cell division defect of an ftsK ts mutant, but not a deletion mutant, can be suppressed specifically by deletion of dacA, which encodes PBP5 (peptidoglycan-modifying D-alanine:D-alanine carboxypeptidase) [ Begg95 , Draper98 ]. A deletion of ftsK can be partially suppressed by overproduction of FtsN [ Draper98 ].

ftsK expression is increased as part of the SOS response, conferring increased resistance to DNA damage [ Wang98i ].

Selected reviews: [ Crozat10a , Sherratt10 , Allemand09 , Bigot07 , Strick06 , Lesterlin04 , Weiss04a , Sherratt04 , Errington03 , Donachie02 ]

Front 119

lolA

Back 119

periplasmic chaperone, effects translocation of lipoproteins from inner membrane to outer

Front 120

dmsA

Back 120

dimethyl sulfoxide reductase, chain A
This subunit contains the active site and the molybdenum cofactor [ Rothery91 ].

DmsA contains a twin-arginine leader peptide which targets the protein to the membrane, although DmsA does not appear to be exported to the periplasm. The leader peptide is also essential for expression of DmsA and stability of the DmsAB dimer, and is cleaved between residues 45 and 46 [ Bilous88a , Sambasivar00 ]

Front 121

rmf

Back 121

ribosome modulation factor
Rmf is a ribosome modulation factor that reversibly converts active 70S ribosomes to a dimeric form (100S ribosomes), which appears during the transition from exponential growth to stationary phase and is associated with a decrease in overall translation activity [ Wada90 ]. Rmf specifically associates with 100S ribosomes in a 1:1 ratio and does not associate with 70S, 50S, and 30S ribosomal particles [ Wada95 ]. Rmf binds near the ribosomal proteins S13, L13, and L2, close to the peptidyl-tRNA binding site [ Yoshida02c ], and protects certain bases in the 23S rRNA, including A2451, which is thought to be involved in the peptidyl transferase activity [ Yoshida04a ]. This suggests a possible mechanism for Rmf-dependent inactivation of translation. In late stationary phase, the ribosome dimers dissociate, which is followed by disassembly of the 70S ribosomes and loss of viability [ Wada00 ].

Purified Rmf protein can cause dimerization of 70S ribosomes in vitro, inhibiting protein synthesis and binding of aminoacyl-tRNA to ribosomes [ Wada95 ].

An rmf mutant shows reduced viability during stationary phase and does not contain ribosome dimers [ Yamagishi93 ]. An rmf mutant is also heat sensitive in stationary phase [ Niven04 ]. An rmf ompC double mutant and an rmf ompC rpoS triple mutant show even further reduced cell viability; this synthetic phenotype may be due to decreased levels of Mg2+ [ Apirakaram98 , Samuel02 ].

Expression of rmf is induced at the transition from exponential growth to stationary phase or during slow growth; rmf expression appears to be inversely proportional to the growth rate [ Yamagishi93 ]. rmf expression is positively regulated by the stringent starvation factor ppGpp [ Izutsu01a ]. The rmf mRNA is extremely stable during stationary phase, with a half life of approximately 24 minutes in early and 120 minutes in late stationary phase; transfer of the cells into fresh medium leads to an immediate drop in half life to approximately 5 minutes. rmf mRNA degradation requires RNA polymerase activity, RNase E, and PcnB. Transfer of stationary phase cells also causes Rmf protein levels to drop rapidly, and inactive ribosome dimers dissociate into active 70S ribosomes [ Aiso05 ]. rmf was one of only three genes whose expression changed under all stress conditions tested by [ Moen09 ]

Front 122

rimJ

Back 122

ribosomal-protein-S5-alanine N-acetyltransferase
RimJ is an alanine acetyltransferase that is specific for ribosomal protein S5 [ Cumberlidg79 , Yoshikawa87 ].

RimJ has overall similarity to RimL [ Tanaka89b ] and C-terminal similarity to RimI [ Yoshikawa87 ]. In a rimJ mutant, the ribosomal protein S5 is not acetylated at the N terminus [ Cumberlidg79 ].

RimJ acts in transcriptional regulation of the pap pilin operon by environmental conditions in uropathogenic E. coli [ WhiteZiegl92 , WhiteZiegl02 ].

Tcp: "thermoregulatory control of pap" [ WhiteZiegl90 ]

Front 123

pth

Back 123

peptidyl-tRNA hydrolase
Peptidyl-tRNA hydrolase (Pth) catalyzes the removal of peptides from peptidyl-tRNAs that have been prematurely released from the ribosome. The enzyme is essential for growth in E. coli [ Menninger79 , Singh04a ]. lysV, encoding a tRNALys, was found to be a multicopy suppressor of a temperature sensitive pth mutant, indicating that the pth mutant phenotype is due to starvation for tRNALys [ HeurgueHam96 , VivancoDom06 ]. In general, Pth relieves growth inhibition by freeing tRNAs from aborted translation products [ Tenson99 , Dincbas99 ]. Pth is also required for growth of bacteriophage lambda due to translation of minigene ORFs in the lambda bar region [ Henderson76 , Ontiveros97 , Valadez01 , Oviedo04 ].

The crystal structure of Pth has been solved at 1.2 Å resolution [ Schmitt97 , Schmitt97a ]. Pth is monomeric. Predicted active site residues were confirmed by site-directed mutagenesis [ Schmitt97a ]. The His20 residue is essential for catalytic activity [ Goodall04 ]. The catalytic center has been mapped by NMR [ Giorgi11 ].

The integrity of ychF, the downstream open reading frame in the pth operon, influences the stability of the transcript. Lower RNase E activity results in overproduction of Pth, and thus it is proposed that RNase E processing within ychF causes mRNA degradation [ CruzVera02 ].

A temperature-sensitive pth mutant was isolated [ Atherly72 , Menninger73 ]; the mutant protein is unstable in vivo [ CruzVera00 ]. The mutant allele can revert rapidly by gene duplication [ Menez01 ]. Overexpression of tmRNA can rescue the defect of the pthts allele [ Singh04a ]. The rap mutant of Pth [ Henderson76 ] is highly sensitive to the length of the peptidyl-tRNA [ HeurgueHam00 ]

Front 124

narG

Back 124

nitrate reductase A, α subunit
The α subunit of nitrate reductase A is the actual site of nitrate reduction and also contains the molybdenum cofactor [ Blasco92 ]. In addition, a novel [4Fe-4S] cluster with unusual coordination and a high-spin ground state was detected in the crystal structure [ Bertero03 , Jormakka04 , Rothery04 ]

Front 125

dbpA

Back 125

ATP-dependent RNA helicase, specific for 23S rRNA
DbpA is a 3'->5' RNA helicase [ Diges05 ] that plays a role in the late stage of biogenesis of the 50S subunit of the ribosome [ Sharpe09 ].

DbpA interacts with 23S rRNA [ FullerPace93 , Tsu98 , Pugh99 ] and exhibits RNA-dependent ATPase activity that shows specificity for 23S rRNA that is not incorporated within the ribosome [ Tsu98 , Del09 ]. Footprinting of DbpA protein on a fragment of 23S rRNA indicates binding at multiple sites [ Karginov04 ], but the primary binding site appears to be hairpin 92 of the 23S rRNA, which is part of the peptidyltransferase center of the ribosome [ Nicol95 , Tsu01 ]. A dominant negative R331A active site mutant accumulates an incompletely assembled 45S particle which can stimulate the ATPase activity of wild-type DbpA [ Sharpe09 ].

The binding of ATP and 23S rRNA shows cooperativity [ Polach02 ]. Kinetics and equilibria of ATP and ADP binding have been measured [ Talavera05 ], and the rRNA-activated ATPase cycle mechanism has been investigated [ Henn08 ]. Measurements of the kinetics of rRNA unwinding showed that hydrolysis of a single ATP molecule is sufficient for unwinding of 8 bp of duplex RNA; strand displacement occurs after ATP hydrolysis. The release of phosphate weakens binding of DbpA to rRNA, thus facilitating its release from the enzyme [ Henn10 ]. The active form of DbpA is a monomer, binding to a single RNA molecule [ Talavera06 ].

DbpA has a DEAD box motif, which is characteristic of RNA-dependent ATPases/helicases [ Iggo90 ]. The ATPase activity [ Tsu98 , Tsu01 , Boddeker97 , Nicol95 , FullerPace93 ], helicase activity [ Diges01 , Henn01 , Pugh99 , Boddeker97 ], and interactions of DbpA with the substrates and products [ Henn02 , Polach02 , Tsu01 , Henn01 , Pugh99 , Tsu98 , Nicol95 , FullerPace93 ] have been characterized in detail.

Translation of DbpA is unusual in that it initiates at a GUG codon [ FullerPace93 ].

Deletion of dbpA does not alter growth or the ribosome profile of the cell [ Iost06 , Peil08 ]. Site-directed mutants were designed to generate a dominant negative dbpA mutant. A mutation in the proposed "arginine finger" of motif VI, R331A, leads to loss of helicase activity and low ATPase activity; expression of this mutant confers a dominant slow growth and cold sensitive phenotype [ Elles08 ]

Front 126

hrpA

Back 126

ATP-dependent helicase
HrpA has sequence similarity to DEAH-box RNA helicases [ Moriya95 ]. HrpA is involved in the post-transcriptional processing of the daa operon mRNA, which encodes proteins involved in fimbrial biogenesis of an enteropathogenic E. coli strain. Activity requires the predicted nucleotide triphosphate binding and hydrolysis functions [ Koo04a ]. The ATPase activity of His-tagged partially purified HrpA has been measured [ Jain06 ].

HrpA is not essential for growth [ Moriya95 ].
HrpA: "DEAH-family RNA helicase-like protein" [ Moriya95 ]

Front 127

narZ

Back 127

nitrate reductase Z, α subunit
The α subunit is the actual site of nitrate reduction and also contains the molybdenum cofactor [ Blasco92 ]

Front 128

infC

Back 128

IF-3 is one of three translation initiation factors in E. coli [ Sabol70 ]. It was also initially identified as a ribosomal dissociation factor [ Subramania68 , Subramania70 ].

The interaction sites of IF-3 with the 30S subunit have been mapped [ Dallas01 ]. The A790G and U789C mutations in 16S rRNA decreases translation fidelity, which may be due to decreased affinity of the 30S subunit for IF-3 [ Qin07 , Qin09a ]. The m2G966 and m5C967 residues of 16S rRNA appear to be important for interaction with IF-3 [ Saraiya08 ].

After the 30S subunit has dissociated from the post-termination ribosome in a process that requires RRF, EF-G, and GTP hydrolysis, IF-3 binds and stimulates dissociation of deacylated tRNA [ Karimi99 , Peske05 , Hirokawa05 ] and destabilizes binding of all tested tRNAs to the 30S subunit [ Antoun06 ]. IF-3 also destabilizes incorrect initiation ternary complexes [ Risuleo76 ]. The presence of fMet-tRNAfMet increases the rate of dissociation of IF-3 from the 30S subunit and subsequently increases the rate of docking to the 50S subunit [ vanderHofs78 , Antoun06a ]. In the presence of IF-3, the 30S preinitiation complex can not dock to the 50S subunit [ Antoun06a ]. The path of IF-3 binding to the 30S subunit and its subsequent release upon 30S association with the 50S subunit has been probed [ Fabbretti07 ].

IF-3 also appears to play a role in recycling of stalled ribosomal complexes [ Singh05c ]. IF-3 appears to facilitate ribosome recycling factor-mediated processing of stalled ribosomes [ Singh08a ].

infC, the gene encoding IF-3, contains the unusual initiation codon AUU [ Sacerdot82 ]. Expression of infC is negatively autoregulated at the level of translation [ Butler86 ]; this regulation is dependent on the presence of the AUU initiation codon [ Butler87 ]. The mechanism of IF-3 autoregulation was originally thought to be based on its ability to differentiate between typical and atypical initiation codons and discriminate against the atypical initiation codons [ Sacerdot96 , Sussman96 ] by recognition of codon-anticodon complementarity [ Meinnel99 ] as well as unique features of the initiator tRNA [ OConnor01a ]. However, recent experiments suggest that discrimination against the AUU start codon in the presence of IF-3 is a kinetic effect [ Antoun06 ].

IF-3 consists of two independent domains that are connected via a flexible hydrophilic linker peptide [ Fortier94 , Hua98 ]. The solution structures of the full-length protein [ Moreau97 ] as well as that of the N-terminal [ Garcia95 ] and C-terminal domains [ Garcia95a ] were determined by NMR. Sites of interaction of both domains with the 30S subunit of the ribosome were determined by NMR [ Sette99 ]. The C-terminal domain retains all activities of the full-length IF-3 [ Garcia95a , Petrelli01 ], while the N-terminal domain provides additional binding energy for the interaction with the 30S subunit [ Petrelli01 ]. A physical link between the two domains is required for full function of IF-3 [ deCock99 ]. The phenotypes of a point mutation, Y75N, in the N-terminal domain indicate that this domain is involved in start codon discrimination, initiator tRNA selection, and inhibition of leaderless mRNA translation [ Maar08 ].

Together with IF-1 and CspA, IF-3 plays a role in translational bias during cold acclimation [ Giuliodori04 , Giuliodori07 ]. infC transcription and IF-3 synthesis increases during cold shock, and IF-3 increases the rate of initiator tRNA binding to ribosomes programmed with cold shock mRNAs [ Giuliodori07 ].

Overexpression of IF-3 decreases the efficiency of translation re-initiation for the second open reading frame in translationally coupled polycistronic mRNA [ Yoo08 ].

Front 129

htpX

Back 129

heat shock protein, integral membrane protein
HtpX is a heat shock protein that may be involved in degrading misfolded proteins [ Kornitzer91 ]. Expression of HtpX is controlled by the CpxR-CpxA signal system, which senses abnormal proteins. The Cpx pathway has been shown to become activated by abnormal folding of proteins at the periplasmic surface of the cytoplasmic membrane [ Hung01 ]. Overexpression of HtpX leads to enhanced degradation of partial-length puromycyl peptides [ Shimohata02 , Kornitzer91 ].

The putative zinc metalloproteinase active site of HtpX is located on the cytosolic side of the inner membrane. Mutations in this site abrogate HtpX function [ Shimohata02 ]. HptX is integrated into the inner membrane with two N-terminally located transmembrane segments [ Shimohata02 ]

Front 130

nuoL

Back 130

ubiquinone oxidoreductase, membrane subunit L
NuoL is part of the inner membrane component of NADH dehydrogenase I [ Leif95 , Baranova07 ]. The protein has 14 transmembrane helices and a 110 Å long amphipathic α-helix that spans almost the entire length of the membrane domain [ Efremov10 ]. Transmembrane helices were assigned to locations in the crystal structure using Fourier transform analysis [ Vik11 ]. A crystal structure of the membrane component at higher resolution has allowed better identification of the unusual arrangement of the transmembrane helices [ Efremov11 ].

NuoL was proposed to be located at the distal end of the membrane arm of NDH-1 [ Holt03 , Baranova07 ] and to function in proton translocation [ Friedrich98 ]. Proton translocation may be facilitated by movement of its long amphipathic α-helix, which may result from conformational changes at the interface between the membrane and peripheral domains [ Efremov10 ]. A hypothesis for the involvement of NuoL in proton pumping has been proposed [ Ohnishi10a ]. Site-directed mutagenesis of residues that are conserved between NuoL and Na+/H+ antiporters as well as studies using specific inhibitors indicate that NuoL is involved in the indirect coupling mechanism for proton pumping [ NakamaruOg10 ]. When the NDH-1 complex is reconstituted with NuoL subunits that are truncated at various points in the amphipathic helix, proton translocation is diminished, but not abolished, arguing for at least two coupling sites for proton translocation, with NuoL being essential for the translocation of 2H+/2e- [ Steimle11 ]. However, an additional report using C-terminally truncated forms of NuoL indicated that intact NuoL is required for correct assembly of NDH-1 [ TorresBace11 ]. Further site-directed mutagenesis of the three antiporter-like subunits NuoL, M and N indicate that these subunits have a common role in NDH-1 [ Michel11 ].

The specific functional complementation by NuoL of a B. subtilis ΔmrpA Na+ channel mutant indicates that NuoL may transport Na+ as well [ Moparthi11 ]. A C-terminally truncated fragment of NuoL was previously shown to function as a Na+ pump [ Steuber03 , Gemperli07 ].
NuoL: "NADH:ubiquinone oxidoreductase" [ Calhoun93 ]

Front 131

nuoI

Back 131

NADH:ubiquinone oxidoreductase, chain I
NuoI is part of the connecting fragment of NADH dehydrogenase I [ Leif95 ].

Based on sequence similarity, this subunit was predicted to contain the two 4Fe-4S clusters N6a and N6b [ Weidner93 , Friedrich98 , Rasmussen01 ]. The location and identity of EPR spectra for the N4 and N5 Fe-S clusters were subject of some controversy. The 4Fe-4S cluster N4, located on the NuoG subunit, was thought to be identical to either N6a or N6b [ Yakovlev07 ]. Recent reevaluation of the data [ Ohnishi08 ] and mutational analysis of the N5 His(Cys)3 ligands confirmed the location of both N4 and N5 in the NuoG subunit [ NakamaruOg08 ].

In the presence of NADH or reducing agents, crosslinking between NuoB and NuoI in the intact Complex I is abolished, indicating a conformational change involving the hydrophilic subunits in the presence of NADH [ Berrisford08 ].
NuoI: "NADH:ubiquinone oxidoreductase" [ Calhoun93 ]

Front 132

iscS

Back 132

cysteine desulfurase
Cysteine desulfurase (IscS) catalyzes the transfer of sulfur and selenium from cysteine and selenocysteine to a range of recipients. It is critical for addition of sulfur to tRNA, for synthesis and repair of iron-sulfur (Fe-S) clusters, and for generation of a number of other sulfur- and selenium-dependent proteins.

IscS is a cysteine desulfurase that catalyzes the conversion of cysteine into alanine and sulfur via intermediate formation of a cysteine persulfide. Although the mechanism appears to differ for selenium, IscS can similarly convert L-selenocysteine to alanine [ Mihara00 ].

IscS provides sulfur and selenium for modification of several positions on a number of tRNAs, and is responsible for 95% of all sulfur in cellular tRNA [ Lauhon02 ]. Donation of sulfur for 4-thiouridine in tRNA occurs via transfer to the intermediate protein, ThiI [ Kambampati99 , Kambampati00 ]. Another intermediate protein, MnmA, is similarly required for IscS-dependent 2-thiouridine biosynthesis [ Kambampati03 ]. 2-thiouridine synthesis also requires TusA, a protein that activates IscS desulfurase activity and accepts activated sulfur from it as a cysteine persulfide [ Ikeuchi06 ]. No intermediate protein has yet been identified for the IscS-dependent synthesis of 2-selenouridines on tRNA [ Mihara02a ].

In addition to its role in tRNA modification, IscS is critical for iron-sulfur (Fe-S) cluster formation and repair [ Takahashi99 , Nakamura99 , Schwartz00 , Tokumoto01 , Rogers03 ]. IscS transfers sulfur to the coregulated "scaffold" protein IscU, on which Fe-S clusters are assembled [ Agar00 ]. The two proteins form a disulfide linkage between Cys-63 of IscU and Cys-328 of IscS. This interaction between IscU and IscS activates IscS up to 6-fold, and is dependent on the presence of IscU Cys-63 and the carboxy-terminal portion of IscS [ Kato02 , Urbina01 , Lauhon04a ]. Although E. coli contains at least two other cysteine desulfurases that appear to be involved in Fe-S cluster formation ( selenocysteine lyase and cysteine sulfinate desulfinase ). IscS is the only one that interacts with the scaffold protein IscU [ Kurihara03 ]. Although multiple FE-S cluster synthesis systems occur in E. coli, an inability to complement loss of the isc system suggests that IscS and its partners represent the constitutive "housekeeping" cluster synthesis group [ Outten04 ]. Iron for Fe-S synthesis is donated by IscA and CyaY, the latter having been shown to interact directly with IscS as well [ Ding04 , Ding05 , Layer06 ].

IscS is also a key element in repair of Fe-S clusters damaged by oxidation. IscS given ferrous ion and cysteine can repair Fe-S clusters damaged by nitric oxide, even in the absence of new protein synthesis [ Yang02 , Rogers03 ]. The mechanistic role of IscS in this repair process is unclear, as the oxidative damage actually affects the cluster's iron content, rather than removing any sulfur [ Djaman04 ].

Beyond these roles in tRNA synthesis and Fe-S cluster formation, IscS is implicated as a key player in many other synthetic pathways. It donates sulfur to ThiS, which then routes the sulfur into thiazole formation as a part of thiamine biosynthesis, in a process stimulated by ThiI [ Lauhon00 ]. IscS stimulates the biotin synthase reaction, is required for formation of the selenocysteine-containing protein formate dehydrogenase H, and may be involved in NAD synthesis as well [ Kiyasu00 , Lauhon00 ].

IscS is also the most important of the cysteine desulfurases in their role of conferring tellurite resistance [ Rojas05 ].

A 2.1 Å-resolution crystal structure of IscS has been determined [ Urbina02 , CuppVicker03 ]

Front 133

rimM

Back 133

ribosome maturation protein
RimM may play a role in the maturation of the 30S ribosomal subunit, or possibly in initiation of translation [ Bylund98 , Bylund97 ]. RimM does not appear to be required for wild-type 16S rRNA processing [ Lovgren04 ].

RimM has affinity for the 30S ribosomal subunit, but not the 70S ribosome [ Bylund97 ]. Two conserved tyrosine residues in the proposed PRC β-barrel domain of RimM are responsible for specific interactions with the 30S subunit ribosomal protein S19 [ Lovgren04 ].

A rimM deletion mutant accumulates 17S rRNA [ Bylund98 ], and the levels of polysomes decreases while levels of free 30S and 50S subunits increase [ Lovgren04 ]. rimM mutants exhibits defects in growth and a slower rate of translation, compared to wild type, and these phenotypes are suppressed by some rpsM (30S ribosomal subunit protein S13) and rpsS (30S ribosomal subunit protein S19) mutations [ Bylund97 , Lovgren04 ] or by RbfA overproduction [ Bylund98 , Bylund01 ]. Mutations causing decreased RimM interaction with the 30S ribosomal subunit result in a growth defect that is suppressed by mutations that increase RimM production [ Lovgren01a ].

Transcription of the operon containing the rimM gene increases with the growth rate [ Wikstrom88 ]. A sequence element located between codons 18 and 50 of rimM appears to reduce the efficiency of rimM translation [ Wikstrom89 ]. An mRNA secondary structure which sequesters the rimM Shine-Dalgarno sequence is thought to be responsible for this posttranscriptional regulation [ Wikstrom92 ]

Front 134

ffh

Back 134

protein component of the signal recognition particle (SRP)

Front 135

deaD

Back 135

DeaD, DEAD-box RNA helicase
The DeaD protein is an RNA helicase that participates in the assembly of the large, but not the small subunit of the ribosome [ Charollais04 , Kitahara09 ] and is involved in RNA degradation under low temperature growth conditions [ PrudhommeG04 , Awano07 ]. DeaD may destabilize mRNA secondary structures in the translation initiation region of mRNAs [ Butland07 ].

Reports differed on whether [ Bizebard04 , Turner07 ] or not [ Jones96 , Lu99 ] the RNA helicase activity of purified DeaD requires ATP hydrolysis. The DEAD motif is required for efficient in vitro ATPase and helicase acivities and for in vivo function at 15°C [ Turner07 , Awano07 ]. DeaD is able to destabilize secondary structure in the translation initiation region of mRNAs, facilitating translation [ Butland07 ].

The DeaD protein was found to be associated with a pre-50S ribosomal particle [ Charollais04 ] and interacts with poly(A) polymerase I [ Raynal99 ], ObgE [ Sato05 ] and RNase E [ PrudhommeG04 ], as well as a number of ribosomal and other proteins [ Butland07 ]. DeaD is able to replace the function of the RNA helicase RhlB in the degradosome [ PrudhommeG04 ]. DeaD can stabilize overexpressed mRNAs [ Iost94 ], including cspA mRNA [ Brandi99 ], or may be involved in selective degradation of cold shock protein mRNAs [ Yamanaka01a ].

RhlE as well as CspA and RNase R complement the cold-sensitive phenotype of a deaD mutant [ Awano07 , Jain08 ]. An rhlE mutation exacerbates the growth defect of a deaD mutant at 20°C [ Jain08 ].

Deletion of the deaD gene causes a growth defect at low temperature [ Jones96 , Charollais04 ]. Expression of DeaD is induced by cold shock [ Jones96 ]. DeaD has been isolated as a multicopy suppressor of the cold-sensitive phenotype of the smbA2 mutation [ Yamanaka94 ]. It also suppresses the effects of a temperature sensitive mutation in ribosomal protein S2 (rpsB(ts)) [ Toone91 , Moll02 ] by restoring assembly of both S1 and S2 with the ribosome at the non-permissive temperature [ Moll02 ]. A deaD mutant accumulates unprocessed 23S rRNA and shows an altered ribosome profile [ Jain08 ], accumulating ribosome large subunit assembly intermediates [ Peil08 ].

CsdA: "cold-shock DEAD-box protein A" [ Jones96 ]

RhlD: "RNA helicase-like" [ Kalman91a ]

Front 136

rpsO

Back 136

30S ribosomal subunit protein S15
The S15 protein is a component of the 30S subunit of the ribosome and also functions in the post-transcriptional regulation of its own expression.

The S15 protein binds to 16S rRNA in the absence of other ribosomal proteins [ Zimmermann72 , Zimmermann75 , Gregory84 ]. Nucleotides essential for the S15-16S rRNA interaction have been determined by mutagenesis [ Stark84 , Serganov01 ] and nuclease protection [ Wiener88 , Mougel88 ]. Binding of S8 to 16S rRNA influences the central domain organisation and affects the rRNA environment of S15 [ Jagannatha03 ].

In addition to its function in the ribosome, the ribosomal protein S15 binds to its own mRNA, stabilizing a pseudoknot secondary structure and impeding translation initiation [ Portier90 , Philippe90 , Portier90a , Philippe94 , Benard94 , Philippe95 , Benard98 ]. S15 appears to prevent the formation of a functional ternary 30S-mRNA-tRNA(fMet) complex, trapping the ribosome in a preinitiation complex [ Philippe93 ]. rpsO mRNA and 16S rRNA compete for binding to S15 [ Philippe94 ]; common structural determinants between the mRNA and rRNA bindig sites have been investigated, showing that there is limited similarity between the two targets [ Serganov02 , Mathy04 ].

Ribosomes lacking S15 can suppress the rpoH11 mutation [ Yano89 ]. A mutation in rpsO (secC) suppresses a secA(Ts) allele [ FerroNovic84 ]. S15 appears to be required for the optimal synthesis of lipoprotein [ Watanabe88 ].

Processing and degradation of rpsO mRNA have been studied extensively; see for example [ Le02 , Marujo03 , Folichon03 , Folichon05a ] and references therein.

Front 137

rbfA

Back 137

30S ribosome binding factor
RbfA can be found associated with the 30S subunit of the ribosome [ Dammel95 ] and is essential for efficient processing of the 16S rRNA [ Bylund98 ]. Following cold shock, a larger fraction of 30S ribosomal subunits contain RbfA [ Xia03 ].

Cells lacking RbfA show a cold-sensitive growth phenotype [ Dammel95 ] and constitutive induction of the cold shock response [ Jones96a ]. Overexpression of RbfA allows for suppression of the dominant cold-sensitive C23U mutation in 16S rRNA [ Dammel95 ] and suppresses the slow growth rate of a rimM deletion mutant [ Bylund98 ].

Cold shock induction of rbfA expression may occur through transcription antitermiation mediated by CspA and other cold shock-induced proteins [ Bae00 ].

A solution structure of RbfA has been determined, and a potential RNA-binding site was identified [ Huang03c ]

Front 138

rng

Back 138

ribonuclease G (RNAse G)
The rng gene encodes RNAse G [ Li99d , Wachi99 ], which acts in maturation of the 5' end of 16S rRNA [ Li99d , Wachi99 ]. Rng is also involved in RNA turnover [ Tock00 , Umitsuki01 , Kaga02a ]. A large-scale study of RNAs affected by RNAse G and RNAse E is presented [ Lee02a ].

RNAse G activity is distributed and is not directional with respect to the RNA substrate, in contrast to RNAse E activity [ Feng02 ]. The substrate sequence specificity of the enzyme has been examined [ Tock00 ]. The preferred RNA substrate has a 5' monophosphate [ Tock00 ]; the 5' monophosphate appears to enhance the catalytic potency of the enzyme rather than improve substrate binding [ Jiang04 ].

The enzyme is homodimeric [ Briant03 ]. Cysteines that are involved in multimerization have been identified [ Briant03 ].

The C-terminal HSR2 region, conserved between RNAses E and G, is important for function [ Wachi01 ]. The N terminus of the protein has been experimentally identified [ Briant03 ].

An rng mutant exhibits defects in bacteriophage T4 as well as host RNA maturation [ Miczak83 ]. An rng mutant shows a defect in maturation of the 5' end of 16S rRNA and accumulates a 16.3 S precursor form of this RNA [ Li99d , Wachi99 ]. An rng mutant shows greater stability of adhE mRNA and greater abundance of adhE-encoded fermentative alcohol dehydrogenase, compared to wild type [ Umitsuki01 ], and a mutation causing this defect can be separated from rng mutations that cause the 16S rRNA processing defect [ Wachi01 ]. An rng mutant also shows greater stability of eno mRNA and greater abundance of enolase, compared to wild type [ Kaga02a ]. Overproduction causes abnormalities in cell morphology including formation of chains of cells with cytoplasmic axial filaments [ Okada94a , Okada94b ].

An rne rng double mutant exhibits a complete defect in maturation of the 5' end of 16S rRNA [ Li99d ]. An ams1 (RNAse E) rng double mutant exhibits greater heat sensitivity than an ams1 single mutant [ Wachi97 ]. The heat sensitivity of an rne mutant is partially suppressed by overproduction of Rng [ Wachi97 ]. Rng overproduction does not fully replace RNAse E function in an rne mutant [ Jiang00 , Lee02a , Ow03 ]. Modified RNase G with extensions at the amino or carboxy terminus can support growth of an RNase E-deficient E. coli cell [ Deana04 ].

Rng has similarity to E. coli RNAse E (Rne) [ Okada94b , Wachi97 ]. Rng has similarity to Streptomyces coelicolor RNase ES, which can functionally complement rng or rne mutant phenotypes in E. coli [ Inagawa03 ]

Front 139

prmA

Back 139

methyltransferase for 50S ribosomal subunit protein L11
PrmA is the methyltransferase responsible for the multiple methylation of ribosomal protein L11.

prmA mutants are deficient in the methylation of ribosomal protein L11 [ Colson77 , Vanet94 ]. A strain containing a prmA null mutation is viable and shows no growth defect, indicating that lack of methylation of L11 does not affect its function in the ribosome [ Vanet94 ].
PrmA: "protein methylation" [ Colson77 ]

Front 140

def

Back 140

peptide deformylase
Peptide deformylase releases the formyl group from the amino terminal methionine residue of most nascent proteins [ Neidhardt96 , Meinnel95 ]. It interacts directly with the ribosome at the ribosomal exit tunnel [ BingelErle08 ].

Peptide deformylase is a member of the zinc metalloprotease family, defining a new subfamily [ Meinnel95b , Meinnel96 ]. The physiological metal ion is Fe2+ [ Groche98 , Rajagopala98 ]. Systematic in vitro experiments showed that peptide deformylase requires at least a dipeptide as substrate [ Rajagopala97a ] and has the expected broad substrate specificity, while the rate of deformylation varies between different substrates [ Hu99a , Ragusa99 ]. The enzyme is essential in E. coli [ Mazel94 ].

A reaction mechanism involving the conserved E133 residue has been proposed [ Rajagopala00 ], and theoretical studies of the catalytic mechanism and metal-ion dependence of the enzyme were performed [ Wu07b , Dong08 ]. The C-terminal domain was originally shown to be disordered in solution and not essential for activity of the enzyme [ Meinnel96a ]. However, it was shown to interact directly with the large subunit of the ribosome, and a strain containing peptide deformylase that lacks the C-terminal helix has a severely diminished growth rate [ BingelErle08 ].

NMR solution structures [ Meinnel96 , Dardel98 , OConnell99 , OConnell99 , Amero09 , Larue09 ] and crystal structures [ Chan97 , Becker98 , Becker98a , Hao99 , Guilloteau02 , Smith03b , Jain05 , Yen10 ] have been reported. A crystal structure of the C-terminal ribosome-binding helix together with the ribosome show L22 as the major site of interaction [ BingelErle08 ].

Peptide deformylase provides an attractive target for the development of novel antibacterial agents [ Yuan06a ]; see for example: [ Boularot07 , Chikhi06 , Shen08 ]. The enzyme was long thought to be absent from eukaryotic organisms, but genome sequences have revealed its presence even in Homo sapiens, where it appears to be an evolutionary remnant (and thus remains an excellent drug target) [ Nguyen03 ]

Front 141

rpsD

Back 141

30S ribosomal subunit protein S4
The S4 protein, a component of the 30S subunit of the ribosome, functions in the assembly of the 30S ribosomal subunit, the mRNA helicase activity of the ribosome, the regulation of translation of a subset of ribosomal proteins, and transcription antitermination of rRNA operons.
S4 interacts directly with helical elements at the 5' domain of the 16S rRNA [ Stern86 , Powers95 , Vartikar89 ]. The ability of both S4 and S7 to bind 16S rRNA by themselves indicates that they function as initiator proteins for the assembly of the 30S subunit of the ribosome. The S20, S16, S15, S6, and S8 subunits appear to depend on S4 for assembly [ Nowotny88a ].
The S4 protein is involved in the regulation of translation of the other ribosomal proteins encoded by the α operon, RpsM (S13), RpsK (S11), RplQ (L17) and S4 itself [ Yates80 , Thomas87a , JinksRober82 ]. The α operon leader region is required for translational repression by S4 [ Thomas87a ]; S4 specifically interacts with a double pseudoknot structure which overlaps with the ribosome binding site and initiation codon for RpsM [ Tang89 , Tang90 ]. There may be a second binding site for S4 upstream of the RplQ open reading frame [ Meek84 ]. The same protein domain appears to be responsible for both mRNA and rRNA binding [ Baker95a , Conrad87 ].
S4 can also act as a general transcription antitermination factor similar to NusA; it associates with RNA polymerase and is involved in rRNA operon antitermination [ Torres01 ].
S4 influences translational fidelity [ Topisirovi77 ]. Certain mutations in rpsD confer a "ribosomal ambiguity" phenotype, which is characterized by decreased growth rate, increased streptomycin sensitivity, and increased errors in translation [ Zimmermann71 , Andersson83 , Andersson82 , Olsson79 ]. Cells carrying the rpsD14 allele have a mutator phenotype [ Balashov03 ]. The ribosome was found to have mRNA helicase activity, and mutations in the S3 and S4 subunits impair this activity [ Takyar05 ].
ramA: "ribosomal ambiguity"

Front 142

rpsE

Back 142

The S5 protein is a component of the 30S subunit of the ribosome. It was suggested that S5 is positioned to have access to the interface between the 30S and 50S subunits of the ribosome [ Culver99 ].
The N-terminal half of S5 contains the sites of mutations that confer resistance to spectinomycin [ Piepersber75 , Dekio69 , DeWilde73 ] and binds non-specifically to helix 34 of the 16S rRNA [ Heilek96 , Stern88a ]. S5 may be involved in modulating the conformation of the 16S rRNA [ Lodmell97 ]. S5 can be cross-linked to mRNA [ RinkeAppel91 ] and tRNA [ Graifer89 ].
S5 influences translational fidelity. Certain mutations in rpsE confer a "ribosomal ambiguity" phenotype, which is characterized by decreased growth rate, increased streptomycin sensitivity, and increased errors in translation [ Ito73 , Hasenbank73 ].
The S5 protein is acetylated at the N-terminus [ WittmannLi78 , Arnold99 ]; mutants in the alanine acetyltransferase enzyme, RimJ, are temperature sensitive [ Cumberlidg79 , Yoshikawa87 ].
spcA: "spectinomycin"

Front 143

ftsX

Back 143

FtsX is the putative membrane component of an ATP-binding cassette (ABC) transporter [ deLeeuw99 ].

FtsE and FtsX localize to the cell division site; localization is dependent on FtsZ, FtsA and ZipA, but not FtsK, FtsQ, FtsL and FtsI [ Schmidt04 ]. FtsEX is important for assembly or stability of the septal ring under low-salt growth conditions [ Schmidt04 ]

Front 144

ftsE

Back 144

cell division protein FtsE
FtsE is the putative ATP-binding protein component of an ATP-binding cassette (ABC) transporter. FtsE dimerizes and associates with the inner membrane via interaction with FtsX, an integral membrane protein [ deLeeuw99 ].

FtsE and FtsX localize to the cell division site; localization is dependent on FtsZ, FtsA and ZipA, but not FtsK, FtsQ, FtsL and FtsI [ Schmidt04 ]. FtsEX is important for assembly or stability of the septal ring under low-salt growth conditions [ Schmidt04 ].

An ftsE null mutant is only viable on high salt medium [ deLeeuw99 ]. ftsE and ftsX are located in an operon with the ftsY gene. ftsY encodes a GTPase which acts as a receptor for the SRP (signal recognition particle) involved in protein targeting [ Gill87 , Miller94 ]. It has been suggested that FtsE and FtsX are not to be required for SRP-mediated targeting [ deLeeuw99 ], although it has been reported that ftsE mutants are affected in translocation of potassium ion pump proteins into the cytoplasmic membrane [ Ukai98 ]

Front 145

ftsY

Back 145

SRP receptor

Front 146

secB

Back 146

SecB is a cytoplasmic protein involved in the Sec secretion pathway. SecB is believed to function as a molecular chaperone that maintains newly synthesized precursor proteins in their unfolded, translocation-competent state and delivers them to the membrane-embedded translocon.

Protein purification studies have shown that SecB codes for a cytoplasmic homotetramer required for the translocation of preproteins into inverted vesicles of the E.coli plasma membrane [ Watanabe89a ]. SecB has an affinity for a number of periplasmic and outer-membrane proteins in their unfolded forms. Some examples of these, shown through SecB disruption experiments, are MalE, LamB, OmpA, OmpF, OppA, PhoE, MBP, DegP, FhuA, FkpA, OmpT, OmpX, TolB, TolC, YbgF, YcgK, YgiW and YncE [ Baars06 ]. In SecB- cells, much of the synthesized MalE (40%) was permanently trapped in the cytoplasm. SecB has been shown to bind MBP in its extendend or molten globule-like state, preventing tertiary structures from forming and allowing transfer of the protein across the membrane without having to expend energy disrupting the tertiary structures [ Bechtluft07 ]. Electron paramagnetic resonance experiments as well as protein binding sites suggests that polypeptides are wrapped around SecB [ Crane06 ]. SecB can block the aggregation of MBP and preMBP in vitro and can promote the disaggregation of preformed MBP in soluble aggregates [ Kulothunga09 ].

The requirements for SecB assistance in protein transport are unknown, however it has been shown in signal peptide exchange experiments that the signal sequence is not the determining factor in nascent envelope protein/SecB interaction [ Collier90 ] in SecB-dependent preproteins. These results strongly suggest that SecB interacts with the mature moiety of the preproteins and has an anti-folding characteristic which enables efficient targeting to the SecA/EGY complex. SecB most likely is able to deliver its precursor protein to the inner membrane by interacting with the final 22 residues of SecA which can itself be found bound to the inner membrane [ Fekkes97 ], in a complex with SecEGY.

Kinetic studies [ Miller02c ] using purified SecA and SecB indicates that the SecA-SecB interaction increases SecA ATPase activity.

X-ray crystallography studies on SecB have been performed and the structure has been solved to 2.35 Å resolution [ Dekker03 ]

Front 147

atpC

Back 147

ATP synthase, F1 complex, ε subunit
The epsilon subunit appears to play an important role in coupling the catalytic site events with proton translocation in association with the gamma subunit. The coupling involves conformational changes and probable translocations of one or both subunits. This subunit is also required for binding of the F-1 complex to the F-O complex. [ Tang96 , Senior90 ]

Front 148

atpA

Back 148

The α-subunit plays an essential role in the catalytic mechanism of the enzyme and in the binding and coupling between the F1 and F0 complexes. The α-subunit also contains an adenine-specific binding site which is noncatalytic, nonregulatory and not essential for enzyme assembly in vitro. Its function has not yet been determined. The α-subunit complex is a homotrimer [ Rao88 ].

A hydrogen-bonding network is formed at the closed α/β-subunit interface of F1 [ Abrahams94 ]. Elimination of this network results in a severely impaired enzyme. A possible role for the hydrogen-bonding network in coupling of ATP synthesis/hydrolysis and rotation has been proposed [ Mao08 ]. The role of conserved residues surrounding the catalytic site has been studied [ Li09a ]

Front 149

dsbA

Back 149

protein disulfide oxidoreductase
The dsbA gene codes for a protein that is a disulfide catalyst. The protein itself has a disulfide bond that is transferred catalytically to folding proteins in the periplasm. The disulfide oxidoreductase is capable of oxidizing proteins very rapidly. The oxidoreductase requires the dsbB protein for reoxidation [ Bardwell94 , Grauschopf95 , Kishigami95 , Guilhot95 , Akiyama92 , Metheringh95 ].

Oxidative folding is thought to occur via formation of intermediate mixed disulfide complexes between DsbA and its substrates. A DsbA mutant has been identified which slows down the resolution of DsbA/substrate compexes and allows characterisation of these intermediates [ Kadokura04 ]. Disulfide linked complexes formed between DsbA and newly synthesized PhoA have been identified [ Kadokura09 ].

The crystal structure of DsbA in complex with a peptide residue from the Shigella flexneri SigA autotransporter protein has been determined at 1.9Å resolution [ Paxman09 ]

Front 150

fdoI

Back 150

formate dehydrogenase-O, γ subunit
By similarity to the paralogous γ subunit of formate dehydrogenase-N, FdnI, FdoI is the heme-containing membrane subunit of formate dehydrogenase-O.
Both the N- and C-terminus of FdoI appear to be located in the cytoplasm [ Benoit98 ]

Front 151

rraA

Back 151

ribonuclease E inhibitor protein A
RraA inhibits ribonuclease E (RNase E, Rne) activity by binding to and masking the C-terminal RNA binding domain of RNase E. The interaction of RraA with the degradosome is facilitated by protein-RNA remodeling via the ATPase activity of RhlB [ Gorna10 ].
RraA physically interacts with RNase E, but does not interact with the RNA substrates [ Lee03a ]. High-affinity binding of RraA to RNase E requires the C-terminal domain (CTD) of RNase E [ Lee03a , Gao06 ]. RraA interacts with both RNA-binding sites of RNase E and interferes with their interaction with RNA [ Gorna10 ]. RraA also interacts with the RhlB helicase component of the degradosome, and a ternary complex of RraA, RNase E and RhlB can be observed [ Gorna10 ]. Binding of RraA or RraB, a second modulator of RNase E activity, differently affect the composition of the degradosome [ Gao06 ].
A crystal structure of RraA is presented at 2.0 Å resolution. RraA forms a homotrimer; the complex is shaped like a ring with a hole of 12 Å across. RraA is structurally related to a family of aldolases [ Monzingo03 ].
The regulatory interaction between RraA and RNase E and their orthologs appears to be evolutionarily conserved [ Yeom08 , Yeom08a , Lee09g ].
Transcription of rraA is σS-dependent and increased upon entry into stationary phase. The stability of rraA mRNA itself is dependent on the activity of RNase E [ Zhao06b ]. Overproduction of RraA causes pleiotropic phenotypes due to increased abundance of RNAs that are usually substrates of RNase E [ Lee03a ]. Overexpression of RraA rescues cells overexpressing RNase E from growth arrest [ Yeom06 ].
RraA was originally mis-annotated as a SAM-dependent methyltransferase predicted to act in menaquinone biosynthesis, and given the name MenG (Hudspeth et al., unpublished; GenBank record U56082); however, the protein was found to lack both structural or functional indications of methyltransferase activity [ Lee03a , Monzingo03 ]

Front 152

hslV = clpQ

Back 152

HslV hexamer
HslV is the peptidase component of the HslVU protease, which is composed of HslU and HslV [ Rohrwild96 , Yoo96 ]. This ATP-stimulated protease exhibits activity similar to that of the chymotrypsin-like activity of the eukaryotic proteasome [ Rohrwild96 ]. HslV exhibits weak peptidase activity in the absence of HslU [ Seol97 ].
The HslVU protease plays a role in clearing the defective peptides produced in the presence of puromycin [ Missiakas96 ]. HslVU is capable of filling the role of the Lon protease under some conditions [ Wu99 ]. HslVU degrades the Lon substrate SulA [ Seong99 ] and exhibits activity toward DnaA204 mutant protein [ Slominska03 ].
Crystal structures of HslV [ Bochtler97 ] and of the HslVU complex are presented [ Bochtler00 , Sousa00 , Song00 , Wang , Wang01e , Bochtler01 , Wang03b , Kwon03 ]. HslU and HslV form ring shaped complexes [ Rohrwild96 ] of protein hexamers [ Kessel96 ] that stack into a four-ring cylinder with HslU rings on each end and HslV rings in the center [ Rohrwild97 ]. HslU and HslV coimmunoprecipitate, whereas the association is labile under chromatographic conditions [ Rohrwild96 ]. Interactions among HslU and HslV are discussed in detail [ Yoo96 , Yoo97 , Huang97 , Yoo96 , Rohrwild97 , Song00 , Seong02 , Lee03b , Kwon03 ]. The role of ATP binding and hydrolysis in complex formation and activity is discussed [ Yoo96 , Yoo97a , Shin96 , Huang97 , Rohrwild96 , Yoo96 , Seong99 , Song00 , Wang01f ]. Stimulation of HslU-mediated ATP hydrolysis by poly-L-lysine stimulates the peptidase activity of HslV within the HslVU complex [ Yoo96a ]. The presence of a protein substrate also stimulates HslVU protease activity [ Seol97 ]. The activity of the HslVU complex is discussed in detail [ Bogyo97 , Huang97 , Wang , Wang98f ]. HslV is modified by an inhibitor of cysteine proteases [ Bogyo97 ].
Mutagenesis studies of HslV have identified the catalytic threonine residue and have identified residues involved in HslU interaction [ Yoo97 ]. A C159S or C159A mutation eliminates peptidase activity, disrupts association with HslU, and inhibits N-ethylmaleimide-mediated dissociation of HslV complexes [ Yoo98 ]. Overproduction of HslU and HslV causes resistance to nitrofurantoin and to UV irradiation in a lon mutant background [ Khattar97 ]. Genetic interactions between hslVU and lon are discussed; Lon functions can be carried out by HslVU [ Wu99 ].
HslV has 52% identity to Bacillus subtilis CodW [ Kang01 ]. HslV has some structural similarity to ClpP [ Wang98f ]. HslV has similarity to Leptospira borgpetersenii serovar hardjobovis HslV [ Lin01 ], to Leishmania infantum HslV [ Couvreur02 ], and to a Lactobacillus leichmannii protein [ Becker96 ]. HslV-related proteins are detected in Leishmania, Trypanosoma, Plasmodium, Magnetospirillum magnetotacticum, and Enterococcus faecium [ Gille03 ]. E. coli HslV activity is stimulated by Thermotoga maritima HslU [ Song03 ].
Regulation has been described [ Chuang93a , Peruski94 , Nakasono00 ]. Transcription [ Chuang93a , Peruski94 ] and protein abundance [ Nakasono00 ] are induced by heat shock

Front 153

argE

Back 153

acetylornithine deacetylase
E. coli acetylornithine deacetylase, encoded by the ArgE gene is a homodimer that catalyses the deacylation of N2-acetyl-L-ornithine to yield ornithine and acetate [ Meinnel92 ]. Ornithine is an obligatory intermediate in the arginine biosynthetic pathway and a branch point in the synthesis of polyamines [ JavidMajd00 ]. It is a metalloenzyme that is activated by cobalt and inorganic phosphate [ Boyen92 ]. The enzyme contains two metal binding sites: a high-affinity site that shows a preference for Zn(II) and a second lower affinity site that shows a strong preference for Co(II) [ JavidMajd00 ]. ArgE can also be activated by Mn(II) ions [ McGregor05 ]. Two Mn(II) ions bind to ArgE to form a dinuclear site in ArgE [ McGregor07 ]

Front 154

rplK

Back 154

The L11 protein is a component of the 50S subunit of the ribosome.

L11 is posttranslationally modified by methylation [ Chang75a ]. The N-terminal alanine residue is methylated to N-trimethylalanine [ Lederer77 , Dognin80 ], and two lysine residues are methylated to Nε, Nε, Nε-trimethyllysine [ Dognin80a , Jerez80 , Dognin80 , Arnold99 ]. Methylation of L11 does not appear to be essential for its function [ Vanet94 ].

Ribosome core particles lacking L11 do not display EF-G-dependent GTPase activity [ Schrier75 ]. L11 participates in joining the 30S initiation complex to the 50S subunit [ Naaktgebor76 , Gotz89 ]. The N-terminal domain of L11 participates in the formation of the arc-like connection between EF-G and L7/L12 during tRNA translocation [ Datta05 ].

L11 directly interacts with 23S rRNA [ Littlechil77 ]. Its binding site has been identified [ Schmidt81a ] and is adjacent to the L8 complex binding site in 23S rRNA [ Beauclerk84 , Egebjerg90 ]. Binding of the L8 complex and L11 appears to be cooperative [ Rosendahl93 , Rosendahl95 ]. L11 recognizes and stabilizes the three-dimensional structure of its binding site in 23S rRNA [ Ryan91 , Blyn00 , Bowen05 , Maeder06 ]. A crystal structure of L11 bound to a 58 nucleotide fragment of 23S rRNA has been determined; the C-terminal domain of L11 binds RNA tightly, while the N-terminal domain makes only limited contacts with RNA [ Wimberly99 ].

The L11 binding region of 23S rRNA appears to be important for translation termination [ Van02b ]. The N-terminal domain of L11 is critical for modulation of release factor binding [ Tate83 , McCaughan84 , Tate84 , Tate86 , Van03a ]. Deletion of the amino terminus reduces the termination efficiency of RF1, but not RF2, thus increasing nonsense suppression at UAG codons, while changes in L11 reduce dissociation of RF2 from ribosome, causing a decrease in nonsense suppression at UGA codons [ Bouakaz06 ]. Strains lacking L11 exhibit UAG stop codon suppression, defective growth, and high-temperature lethality; strains lacking only the N-terminal domain of L11 are only defective in UAG-dependent termination [ Van03a ]. However, later experiments show that L11 affects RF1 and RF2 activity similarly [ Sato06 ].

L11 is required for binding of the thiazole antibiotic thiostrepton to the ribosome [ Highland75 ].

L11 plays a role in regulating the stringent response; an rplK (relC) mutant has a relaxed phenotype [ Friesen74 , Parker76 ]. Both a proline-rich helix in the N-terminal domain of L11 and the linker region between the N- and C-terminal domains are required for regulating the activity of (p)ppGpp synthetase I (RelA) [ Yang01 , Jenvert07 ], although no direct interaction between the two proteins has been detected [ Yang01 ]. ppGpp synthesis by RelA requires both uncharged tRNA at the A site of the ribosome and the presence of L11 [ Wendrich02 ]. The L11 N-terminal domain alone can activate RelA in the presence of ribosomes and uncharged tRNA [ Jenvert07 ]

Front 155

rplA

Back 155

The L1 protein is a component of the 50S subunit of the ribosome and also functions in the post-transcriptional regulation of the ribosomal protein genes encoded in the L11 operon. Ribosomes lacking L1 show a lower translation activity than wild type [ Subramania80 ] and are defective in binding of aminoacylated tRNA [ Sander83 ]. L1 has been identified within a 3-D map of the 70S ribosome constructed by cryo-electron microscopy [ Malhotra98 ].

L1 interacts with a region in the 5' end of 23S rRNA [ Branlant76 , Branlant80 , Egebjerg91 ]. It also can be crosslinked to a region near the central fold of aminoacylated tRNA in the P and E site [ Podkowinsk89 , Rosen93 , Osswald95 ]. L1 is located within 21 Å of nucleotide C2475 of 23S rRNA, near the peptidyltransferase center [ Muralikris95 ].

L1 is a translational repressor of the synthesis of L11 and L1, the proteins encoded by the L11 operon [ Brot81 , Thomas87a , Stoffler81 , Yates80 , Dabbs81 ]. Synthesis of L1 is regulated by translational coupling to the synthesis of L11 [ Yates81a , Baughman83 ]. The target site for L1 binding to the mRNA is near the translation initiation site of L11 [ Yates81a , Baughman83 , Thomas87b ], and the presence of 23S rRNA relieves translational inhibition by L1 [ Yates81a ]. The predicted secondary structure of the L1 binding region within 23S rRNA and rplKA mRNA is similar [ Branlant80 , Gourse81 ] and has been studied experimentally [ Baughman84 , Kearney87 , Said88 , Drygin00 ].

Both the growth rate control and stringent control of the synthesis of ribosomal proteins L11 and L1 are resulting from translational regulation by L1 [ Cole86 ]

Front 156

frdB = fnr

Back 156

FNR DNA-binding transcriptional dual regulator
FNR is the primary transcriptional regulator that mediates the transition from aerobic to anaerobic growth through the regulation of hundreds of genes. Generally, this protein activates genes involved in anaerobic metabolism and represses genes involved in aerobic metabolism [ Salmon03 , Kang05 ]. FNR also regulates the transcription of many genes with other functions, such as acid resistance, chemotaxis, cell structure, and molecular biosynthesis, among others [ Salmon03 , Kang05 ].

The cellular concentration of FNR is similar under both anaerobic and aerobic growth [ Sutton04 ], but its activity is regulated directly by oxygen. Under anaerobiosis, FNR acquires a [4Fe-4S] cluster that causes a conformational change and dimerization of the protein that causes it to become activated [ Moore01 ]. Purification of [4Fe-4S]-FNR in an O2-free environment has been described [ Yan09 ]. The presence of O2 results in inactivation of FNR via oxidation of this [4Fe-4S] cluster into the [2Fe-2S] cluster [ Jervis09 , Sutton04 , Khoroshilo97 ] and the disassembly of the dimer [ Lazazzera96 ]. After prolonged O2 exposure, the [2Fe-2S] cluster is destroyed, and apo-FNR, which lacks an Fe-S cluster, is the primary form of FNR under aerobiosis [ Reinhart08 , Sutton04a ]. Nitric oxide is also able to inactivate FNR through the nitrosylation of the [4Fe-4S] cluster [ CruzRamos02 ].

Under aerobiosis, the apo-FNR monomer is exposed and can be degraded by the ClpXP protease [ Mettert05 ]. Two motifs of FNR appear to be necessary for this degradation, one located in the N-terminal region and the other in the C-terminal region [ Mettert05 ].

The activated FNR conformation is able to bind a specific palindromic sequence of DNA with the consensus sequence TTGATNNNNATCAA [ Gerasimova01 , Eiglmeier89 ]. The G and the first T of each FNR half-site appears to interact with FNR residues Glu-209 and Ser-212 [ Spiro90 ]. FNR activates the transcription from class I and class II promoters, in which the FNR-binding sites are located around -61, -71, -82, or -92 in class I and around -41.5 in class II [ Wing95 ]. In class I promoters the second N and in class II promoters the third N of the consensus FNR site tend to be an A-T [ Scott03a ].

During activation, three activating regions (AR) are surface exposed for contact with the polymerase and promote transcription. The AR1 region contacts with the α subunit [ Lee00c ], the AR2 region makes contact with the &sigma70 [ Blake02 ], and the AR3 contacts the α NTD of the RNA polymerase [ Lamberg02 ]. The group amino acids Thr-118 with Ser-187; Lys-49 with Lys50; and Asp86 and Ile81 with Gly-85 are important for AR1, AR2, and AR3, respectively, to make contact with the RNA polymerase [ Lee00c , Blake02 , Lamberg02 ]. The monomer in the FNR dimer that makes contact with the RNA polymerase and the activating region in FNR that makes the contact depends on the class of promoter, class I or class II [ Lee00c , Blake02 , Lamberg02 ].

FNR belongs to the CRP/FNR superfamily of transcription factors whose members are widely distributed in bacteria [ Korner03 ]. These proteins have an N-terminal sensory domain, a C-terminal helix-turn-helix DNA-binding domain, and a dimerization motif in between [ Korner03 ]. The sensory domain of FNR contains five cysteine residues, four of which are essential for linking the [4Fe-4S] cluster [ Green93a ].

Under anaerobic growth conditions, transcription of the fnr gene is negatively autoregulated [ Mettert07 ].

FNR was named for the mutant defect in "fumarate and nitrate reduction" [ Lambden76 ]

Front 157

miaA

Back 157

tRNA(i6A37) synthase
Dimethylallyl diphosphate:tRNA dimethylallyltransferase (DMAPP-tRNA transferase, MiaA) catalyzes the first step in the pathway for hypermodification of the A37 base of certain tRNAs such as tRNATyr and tRNAPhe. The enzyme transfers the dimethylallyl moiety of DMAPP to the N6 position of A37, which is adjacent to the anticodon [ Rosenbaum72 , Bartz70 ]. This modification is important in preventing frameshifts and other mutations [ Urbonavici01 ]. Further thiomethylation of the A37 position is dependent on the presence of the isopentenyl modification [ Vold79 ].

MiaA can act on isolated tRNA stem-loops, as long as the helix stem and the A36-A37-A38 motif are both present [ Motorin97 , Soderberg00 ]. Co-crystal structures of MiaA in complex with tRNAPhe have been solved, showing a mutually induced fit mechanism of recognition involving structural rearrangement of the anticodon loop of the tRNA [ Seif09 , Chimnaronk09a ]. No substantial change in the structure of the catalytic domain of MiaA is seen upon tRNA binding [ Chimnaronk09a ]. Both reports rely for comparison on apo-MiaA structures of the protein from other organisms. An ordered sequential mechanism of substrate binding and catalysis has been suggested [ Moore97 , Moore00 ].

Modification of tRNAs by MiaA is important in preventing various mutations. miaA mutants show increased +1 (but not -1) frameshifting as well as GC to TA transversions, the latter requiring active recombination [ Urbonavici01 , Urbonavici03 , Connolly91a , Zhao01a ]. miaA mutations also increase the general rate of spontaneous mutations and alter readthrough and suppression of nonsense codons [ Connolly89 , Petrullo83 ].

In miaA mutants, the altered efficiency of translation by tRNAs lacking the dimethylallyl modification affects regulation by attenuation of several operons, including pheA, trp, pheMST and tnaCA [ Gowrishank82 , Springer83 , Vacher84 , Gollnick90 , Pages90 , Landick90a ]. A mutation in miaA inhibits growth and, when combined with a mutation in rpsL, leads to streptomycin dependence [ Diaz87 , Diaz86 ]. A miaA mutation is lethal in a strain lacking functional tRNA6Leu [ Nakayashik98 ].

The miaA gene is part of a superoperon with complex patterns of transcription and mechanisms of regulation [ Tsui94a ]. The superoperon includes a number of heat shock promoters, and miaA is essential for growth at high temperatures [ Tsui96 ]

Front 158

rpsF

Back 158

The S6 protein is a component of the 30S subunit of the ribosome. S6 interacts with the central domain of 16S rRNA [ Gregory84 , Stern88a ].

The S6 protein contains glutamate residues at the C-terminus, only two of which are encoded by the rpsF gene [ Reeh79 , Schnier86 ]; up to four additional glutamate residues are added post-translationally by the RimK enzyme [ Kang89 ]. This form of S6 accumulates when the soxR regulon is activated [ Greenberg90 ]. In bacteriophage T7-infected cells, S6 is phosphorylated [ Robertson94 ].

A class of mutations in rpsF supresses the temperature-sensitive growth defect of certain dnaG alleles [ Britton97 ]

Front 159

pepA

Back 159

aminopeptidase A/I and DNA-binding transcriptional repressor

Front 160

dnaG

Back 160

DNA primase
DNA primase catalyzes the synthesis of RNA primers on single-stranded template DNA [ Rowen78 ]. These primers then serve as the starting point for DNA synthesis by DNA polymerase III [ Bouche75 ].

Using single-stranded DNA in vitro, primase and NTPs are sufficient to produce RNA primers [ Swart93 , vanderEnde85 , Swart95 ]. In practice, primase also relies on ssDNA-binding protein (SSB) to stabilize its interaction with the primer. The Chi subunit of DNA polymerase III interacts with SSB near the primer, displacing DNA primase and allowing loading of the DNA polymerase III beta sliding clamp [ Yuzhakov99 , Sun98a ]. Primase binds three distinct sites during priming, one of them as far as 115 nucleotides distant from the start of primer synthesis [ Sims80 ]. In the case of lagging strand synthesis, primase dissociates from DNA each time an Okazaki fragment is completed and then repeats this binding process to begin priming of the next fragment [ Wu92a ].

During replication, primase follows DNA helicase as the helicase processively unwinds DNA at the replication fork, putting down primers on the lagging strand in its wake [ McMacken77 ]. Three DnaG monomers bind per helicase hexamer via the DnaG carboxy-terminal domain [ Oakley05 ]. This physical interaction is required for optimal primer synthesis, stimulating primase function 5,000 fold [ Lu96a , Zhang02g , Johnson00c ]. In addition, interaction with helicase reduces the specificity of the primer initiation site from CTG to PyPyPu [ Bhattachar00 , Yoda91 ]. The degree of interaction between primase and helicase sets the length of Okazaki fragments [ Tougu96 ]. More generally, the availability of primase controls how frequently new fragments are started and thus how long they are, and therefore the ssDNA-binding capability of helicase is required to direct primase to new template DNA [ Wu92b , Zechner92 , Mitkova03 ]. Initiation of bidirectional replication from oriC also requires the helicase-primase interaction, as this controls proper leading-strand primer placement [ Hiasa99 , Hiasa94 ]. Primer length is constrained by the addition of DnaC, which loads helicase onto DNA [ Mitkova03 ]. The presence of DNA polymerase III, which follows helicase, also limits primer length to about 10 nucleotides [ Zechner92a ].

Two primases bind at the site of primer initiation, although they do not appear to form a dimer. This differs from the stoichiometry of interaction with helicase discussed above [ Khopde02 , Stayton83 ].

Primase has been subject to structural and cofactor analysis. Its catalytic activity is located in the amino-terminus, although the helicase-interacting carboxy-terminal domain is also required for priming at the replication fork [ Tougu94 , Loscha04 ]. The catalytic center has been evaluated by crosslinking as well as via crystal structures at 2.9, 1.7 and 1.6 Å resolution, which reveal that the core has a TOPRIM domain similar to DNA topoisomerases [ Mustaev95 , Podobnik00 , Keck00 ]. The core catalytic domain contains several metal-coordinating residues [ Keck00 , Godson00 ]. Primase appears to bind two magnesium ions and has been shown to contain about one gram of zinc per mol of purified protein [ Urlacher95 , Stamford92 ]. Deletion of a putative zinc-binding region in the amino-terminus inactivates primase [ Stamford92 ]. Zinc in primase prevents formation of deleterious disulfide bonds between the cysteines that coordinate the zinc ions [ Griep96 ]. The binding of zinc and the effect of magnesium abundance on zinc binding has been examined in detail [ Powers99 ].

Primase is required for initiation of DNA replication in plasmids and phages. It is a component of the primosome, a complex that initiates DNA replication in phiX174 phage in vitro [ Ng96a , Ng96 ]. Other phages that depend on primase include T4 and alpha 3 [ Lilien82 , Benz80 ]. Plasmids that require primase to initiate replication include Mu, R1, mini RK2 and pSC101 [ Kruklitis94 , Ortega86 , Masai89 , Pinkney88 , Ely85 ]. Sequence and structure comparisons have been made across phage and plasmid priming sites [ Tanaka94 ].

In vitro, primase can polymerize RNA from the 3' end of a DNA oligomer primer [ Sun98b ].

Primase is regulated posttranscriptionally by RNase E, which cleaves its mRNA, and by its use of rare codons, which may serve to keep its expression lower than other cotranscribed genes [ Yajnik93 , Konigsberg83a ]

Front 161

pnp

Back 161

Polynucleotide phosphorylase (PNPase) is a 3' to 5' exonuclease and a 3'-terminal oligonucleotide polymerase. It degrades various mRNAs, is involved in cold shock regulation, is a part of tRNA maturation and degradation, adds heteropolymeric tails to some RNAs and is a component of the degradosome, a multienzyme complex that carries out RNA degradation.

PNPase is involved in general mRNA degradation. Loss of PNPase leads to an increase in steady-state levels of mRNA, as well as increasing mRNA half lives in the absence of the 3' exonuclease RNase II [ Mohanty03 , Kinscherf75 ]. PNPase also has a role in mRNA degradation during carbon starvation, where it may be required for breakdown of small rRNA fragments produced by other RNases [ Kaplan74 , Kaplan75 ].

A number of specific PNPase substrates have been identified. PNPase is involved in degradation of lac mRNA, rnb mRNA, mRNA coding for ribosomal protein S20, and the RNA-OUT antisense RNA [ HarEl79 , Pepe94 , Mackie89 , Zilhao96 ]. It also degrades sok antisense RNA and thrS and rpsO mRNA following cleavage by RNase E [ Dam97 , Nogueira01 , Braun96 , Hajnsdorf94a ]. PNPase binds to but does not degrade RNA containing 8-oxoguanine [ Hayakawa01 ].

PNPase-mediated degradation is required for regulation of the cold shock response. PNPase degrades a number of mRNAs induced by cold shock, including those coding for CspA, RbfA, CsdA, RpoE, RseA, Rnr and many others [ Yamanaka01 , Cairrao03 , Polissi03 ]. The isolated PNPase S1 RNA-binding domain can complement a deletion in four cold-shock genes [ Xia01 ].

The 3' to 5' processive cleavage of RNA by PNPase depends on the composition and structure of the 3' end of the substrate [ Plamann90 , Cisneros96 ]. RhlB and poly(A) polymerase I (PAP I) in concert with the degradosome are required for PNPase-mediated degradation of cistrons with 3' REP-stabilizers [ Khemici04 ].

Binding of the protein Hfq to poly(A) tracts prevents PNPase degradation of these tails in vitro [ Folichon03 ]. RNA with 3' stem-loops are resistant to degradation by pure PNPase or whole degradosome in vitro, but addition of even a short poly(A) or mixed nucleotide tail overcomes this block [ Causton94 , Blum99 , Lisitsky99 ]. Polyadenylation similarly destabilizes rpsO mRNA against degradation by RNase E, RNase II and PNPase, and is required for sok RNA degradation [ Hajnsdorf95 , Hajnsdorf96 , Dam97 ]. Both 3' adenylation and 5' phosphorylation affect the rate of degradation of RNA I [ Xu95b ]. PNPase itself modulates polyadenylation several RNAs [ Mohanty00 ].

PNPase is involved in tRNA processing and maintenance. Though purified PNPase is incapable of completely processing tRNA in vitro, it is effective, along with RNase II, in trimming long 3' trailing sequences to yield 2-4 nucleotide intermediates which will be trimmed by RNases T and PH [ Deutscher88 , Li94 ]. PNPase is also partially required for repair of 3'-terminal CCA sequences in tRNAs in the absence of tRNA nucleotidyltransferase [ Reuven97 ]. PNPase is also involved in the degradation of mutant tRNA, in a process that is enhanced by polyadenylation by PAP I [ Li02c ].

PNPase also catalyzes the "reverse" reaction, converting nucleoside diphosphates into polyribonucleotides [ Littauer57 , Gillam78 , Gillam80 ]. PNPase generates heteropolymeric tails on RNA and is responsible for residual polyadenylation detected in PAP I deficient strains [ Mohanty00a ]. Hfq, which binds to the 3' end of RNA and prevents PNPase-mediated degradation, also prevents PNPase-mediated addition of nucleosides to bound RNA, while promoting PAP I activity [ Folichon05 ].

PNPase is a trimer of Pnp monomers [ Portier75 , Soreq77 ]. Each Pnp monomer has two RNA-binding sites, KH and S1, that are dispensible for strict catalytic function but are required for Pnp autoregulation, growth at low temperature, and the generation of oligonucleotides [ Jarrige02 , MatusOrteg07 , Guissani76 ]. The S1 domain is a five-stranded antiparallel β barrel with conserved residues on one face forming the RNA binding site [ Bycroft97 ].

PNPase is subject to autoregulation at the mRNA level. RNase III cleaves a stem-loop in the pnp mRNA leader sequence, following which PNPase binds and degrades the 5' half of the cleaved duplex [ Portier87 , Takata89 , Jarrige01 , RobertLe92 , Takata87 , Carzaniga09 ]. PNPase autoregulation also decreases as general RNA polyadenylation increases and following a shift to cold temperatures [ Mohanty02 , Mathy01 , Zangrossi00 , Beran01 ].

Strains lacking both PNPase and RNase II activity are inviable and collect mRNA fragments 100-1,500 nucleotides long [ Donovan86 ]. In a triple mutant in pnp, rnb and rne, mRNA degradation slows three- to fourfold and the length and number of poly(A) tails increases [ Arraiano88 , OHara95 ]. In a pnp mutant lacking RNase PH function, the 50S ribosomal subunit and 23S rRNA is degraded [ Zhou97a ].

Even in the absence of the degradosome scaffold RNase E, PNPase and the helicase RhlB interact. In vitro, RhlB unwinding of dsRNA allows PNPase degradation to occur [ Liou02 ].

PNPase is required to prevent phage P4 superinfection. This prevention requires binding of CI antisense RNA to sequences on nascent P4 transcripts; PNPase processes CI RNA [ Piazza96 ]

Front 162

sspB

Back 162

ClpXP protease specificity-enhancing factor
The SspB protein is a specificity-enhancing factor for the ClpXP protease [ Levchenko00 ]. When protein synthesis is stalled, incomplete proteins that are produced are tagged with the small SsrA peptide. The ribosome-associated SspB protein binds to the SsrA tag and enhances degradation of the tagged peptide by the ClpXP protease [ Levchenko00 ].

The SspB protein forms a homodimer with two independent binding sites for SsrA-tagged proteins [ Wah02 ]. It also binds to ClpX and stimulates its ATPase activity [ Wah02 ]. The dimerization and SsrA binding domain resides in the amino terminal 110-120 residues of SspB, while the C-terminal 40-50 residues are required for association with ClpXP and stimulation of its ATPase activity [ Wah03 ]. Efficient ClpX hexamer binding and substrate delivery requires both C-terminal domains of the SspB dimer [ Bolon04 ]. Interactions between binding of SspB and ClpX to the SsrA tag have been described [ Hersch04 ].

Protein degradation substrates regulated by SspB have been identified and include RseA, which contains an SspB binding site that is unrelated to the SsrA tag sequence [ Flynn04 ]. Degradation of the RseA cytoplasmic fragment is the last step in the proteolytic cascade leading to the induction of the sigma E extracytoplasmic stress regulon.

Crystal structures of SspB alone and in complex with the SsrA peptide tag have been determined at 2.2 and 2.9 A resolution, respectively [ Song03 ], and a crystal structure of SspB in a complex with its recognition peptide in RseA has been determined at 1.8 A resolution [ Levchenko05 ]. The crystal structures reveal diversity in the recognition of different target proteins.

The level of SspB protein remains constant throughout the transition from exponential growth to early stationary phase [ Farrell05 ]

Front 163

srmB

Back 163

SrmB is a DEAD-box protein with RNA helicase activity that facilitates an early step in the assembly of the 50S subunit of the ribosome [ Charollais03 ]. The SrmB protein was found to be associated with a pre-50S ribosomal particle [ Charollais03 ] and forms a complex with ribosomal proteins L4, L24, and the region of 23S rRNA that interacts with L4 and L24 [ Trubetskoy09 ].

The ATPase and helicase activities of SrmB are stimulated by long single-stranded RNAs [ Bizebard04 , Worrall08 ]. SrmB can directly interact with poly(A) polymerase I [ Raynal99 ] and can stabilize overexpressed mRNAs [ Iost94 ].

Deletion of the srmB gene causes a growth defect at low temperature [ Charollais03 ], while overexpression supresses a temperature sensitive lethal mutation in ribosomal protein L24 [ Nishi88 ]. The C-terminal domain of SrmB is not required for function [ Trubetskoy09 ]

Front 164

rna

Back 164

Ribonuclease I (RNase I) is an endonuclease that cleaves phosphodiester bonds in RNA, yielding nucleoside 3'-phosphates and 3'-phosphooligonucleotides [ Spahr61 ]. RNase I is partially reponsible for the degradation of total and ribosomal RNA during both normal and nutrient starvation conditions, especially during carbon starvation [ Kaplan74 , Kaplan75 , Cohen77 ]. RNase I is specifically required for the breakdown of 23 S RNA, though it is not required for degradation of 16 S RNA or very small (4 S) RNA fragments that result from breakdown of larger RNA [ Kaplan75a , Kaplan75 ]. RNase I degradation of the 50 S ribosome releases the ribosomal proteins L4, L10 and L7/12 in addition to cleaving the 23 S RNA to yield a smaller product [ Raziuddin79 ]. Polyamines stimulate the activity of RNase I against synthetic polynucleotides in vitro [ Kumagai77 ].

RNase I is a periplasmic protein that can be released by spheroblasting or treatment of cells with N-dodecyldiethanolamine. This release allows subsequent enhanced breakdown of rRNA by RNase I [ Neu64 , NEU64 , NEU64a , Neu65 , Abrell71 , Lambert76 ]. RNase I appears to remain with the membrane fraction in disrupted cells [ Kaplan76 ].

Some structural analysis of RNase I has been completed. Its conformation energy is 11.5 kcal/mol at pH 7.5, and its T(m) is 64 degrees C at pH 4.0 [ Padmanabha01 ]. It has a 23-residue amino-terminal signal sequence which is cleaved and likely allows its transport to the periplasm [ Meador90 ]. Preliminary crystallization of RNase I has been done, with visualization of the structure at greater than three angstrom resolution [ Lim93 ].

Mutants lacking RNase I or other ribonuclease activities have reduced DNA degradation, possibly due to interaction between excess RNA and DNA endonuclease I [ Wright71 ]. RNase I is also required for full recovery from starvation, as cell viability studies show a direct correlation between recovery from starvation and the ability to degrade RNA [ Kaplan75a ]. A method has been developed to screen for mutants in RNase I by checking for a delay in β-galactosidase expression during amino acid starvation [ Kaplan73a ]

Front 165

rhlE

Back 165

RhlE is a ribosome-associated factor that may be involved in ribosome maturation [ Jain08 ].

RhlE is a member of the DEAD-box-containing ATP-dependent RNA helicase family [ Ohmori94 ]. Its ATPase activity is stimulated by both long RNAs and short oligoribonucleotides [ Bizebard04 ]. Of the three E. coli DEAD-box RNA helicases CsdA, SrmB and RhlE, only RhlE can unwind substrates with short or no single-stranded extensions, but the enzyme shows low processivity [ Bizebard04 ].

RhlE physically interacts with poly(A) polymerase I [ Raynal99 ] and RNase E [ Khemici04 ]. RhlE can functionally replace RhlB within the degradosome [ Khemici04 ].

An rhlE mutant does not have any obvious growth defect [ Ohmori94 ]. Overexpression of RhlE can substitute for the essential function of CsdA during cold shock [ Awano07 , Jain08 ], but exacerbates the cold-sensitive growth defect of a srmB mutant [ Jain08 ]. Conversely, the cold-sensitive growth defect of an rhlE csdA double mutant is enhanced, while that of an rhlE srmB double mutant is alleviated [ Jain08 ]. Both CsdA and SrmB are involved in ribosome maturation, indicating that RhlE may play a similar role [ Jain08 ]

Front 166

rmf

Back 166

Rmf is a ribosome modulation factor that reversibly converts active 70S ribosomes to a dimeric form (100S ribosomes), which appears during the transition from exponential growth to stationary phase and is associated with a decrease in overall translation activity [ Wada90 ]. Rmf specifically associates with 100S ribosomes in a 1:1 ratio and does not associate with 70S, 50S, and 30S ribosomal particles [ Wada95 ]. Rmf binds near the ribosomal proteins S13, L13, and L2, close to the peptidyl-tRNA binding site [ Yoshida02a ], and protects certain bases in the 23S rRNA, including A2451, which is thought to be involved in the peptidyl transferase activity [ Yoshida04a ]. This suggests a possible mechanism for Rmf-dependent inactivation of translation. In late stationary phase, the ribosome dimers dissociate, which is followed by disassembly of the 70S ribosomes and loss of viability [ Wada00 ].

Purified Rmf protein can cause dimerization of 70S ribosomes in vitro, inhibiting protein synthesis and binding of aminoacyl-tRNA to ribosomes [ Wada95 ].

An rmf mutant shows reduced viability during stationary phase and does not contain ribosome dimers [ Yamagishi93 ]. An rmf mutant is also heat sensitive in stationary phase [ Niven04 ]. An rmf ompC double mutant and an rmf ompC rpoS triple mutant show even further reduced cell viability; this synthetic phenotype may be due to decreased levels of Mg2+ [ Apirakaram98 , Samuel02 ].

Expression of rmf is induced at the transition from exponential growth to stationary phase or during slow growth; rmf expression appears to be inversely proportional to the growth rate [ Yamagishi93 ]. rmf expression is positively regulated by the stringent starvation factor ppGpp [ Izutsu01a ]. The rmf mRNA is extremely stable during stationary phase, with a half life of approximately 24 minutes in early and 120 minutes in late stationary phase; transfer of the cells into fresh medium leads to an immediate drop in half life to approximately 5 minutes. rmf mRNA degradation requires RNA polymerase activity, RNase E, and PcnB. Transfer of stationary phase cells also causes Rmf protein levels to drop rapidly, and inactive ribosome dimers dissociate into active 70S ribosomes [ Aiso05 ]. rmf was one of only three genes whose expression changed under all stress conditions tested by [ Moen09 ]

Front 167

pcnB

Back 167

poly(A) polymerase I
Poly(A) polymerase I is responsible for the polyadenylation of 3' ends of RNA molecules.

Poly(A) polymerase polyadenylates the vast majority of mRNA transcripts [ Mohanty06 ]. Unlike in eukaryotes, increased polyadenylation of mRNAs leads to decreased mRNA half-life [ Mohanty99 , Mohanty06 ]. Rho-independent transcription terminators appear to serve as targeting signals for polyadenylation [ Mohanty06 ]. The Hfq protein appears to be involved in the recognition of 3' termini of RNA by poly(A) polymerase I [ Le03 ].

Intracellular levels of poly(A) polymerase I as well as the level of pcnB transcription vary inversely with growth rate [ Jasiecki03 ]. Overexpression of poly(A) polymerase I is toxic and leads to slowed growth [ Mohanty99 , Mohanty06 ]. Use of AUU as the translational start codon results in InfC discrimination (as with production of the IF-3 translation initiation factor) and results in low levels of poly(A) polymerase I in the cells [ Binns02 ].

A his-tagged version of poly(A) polymerase I showed reduced activity following phosphorylation of its his tag. In other proteins, this sometimes correlates with the protein itself being regulated via phosphorylation [ Jasiecki06 ]

Front 168

hfq

Back 168

HF-I, host factor for RNA phage Q β replication
Hfq is an RNA-binding protein that stimulates RNA-RNA pairing and affects many cellular processes, similar to mammalian Sm/Sm-like proteins [ Zhang02b , Moller02a ]. Hfq interacts with many small RNA species [ Wassarman01 ], and exhibits RNA chaperone activity [ Moll03 ]. The Hfq complex is associated with an ATPase activity [ Sukhodolet03 ].

Hfq is a host factor (HF-I) required for wild-type bacteriophage Q beta replication [ Kajitani91 , Su97 ]. Hfq is required for wild-type translation of RpoS [ Muffler96 ]. Hfq interacts with the oxyS RNA to negatively regulate RpoS translation [ Zhang98 ]. Hfq stimulates binding by oxyS RNA [ Zhang02b ] and spf RNA [ Moller02a ] to their RNA targets. Hfq binds ompA mRNA and is required for the regulation of its stability by the micA antisense RNA, which also binds to ompA and blocks ribosome binding [ Vytvytska98 , Vytvytska00 , Udekwu05 ]. Hfq binds to dsrA RNA [ Sledjeski01 , Brescia03 ] and is important for its activity [ Sledjeski01 ], and also binds the antisense regulatory ryhB RNA and is necessary for its activity [ Masse02 ]. Hfq is required for protection of ryhB RNA [ Masse03 , Moll03a ], dsrA RNA [ Moll03a ], and ompA mRNA [ Moll03a ] from degradation by RNaseE. Hfq binds to sodB mRNA, altering its structure to allow binding of ryhB and leading to the degradation of both ryhB and sodB RNAs [ Geissmann04 ].

Hfq is required for wild-type regulation of FtsZ abundance and cell division at stationary phase [ Takada99 ]. Hfq interacts with poly(A) RNA [ Hajnsdorf00 ] and increases the processivity of poly(A) polymerase I [ Hajnsdorf00 , Le03 ]. Hfq also interacts with RNA polymerase [ Sukhodolet03 ].

A crystal structure of Hfq is presented [ Vassilieva03 , Sauter03 ].

Hfq forms a homohexamer [ Zhang02b , Moller02a , Arluison02 , Sauter03 ]. Hfq is a cytoplasmic [ Azam00 ] ribosome-associated protein [ Kajitani94 , Azam00 ]. Hfq exhibits nonspecific DNA interactions [ Takada97 , Azam99 ].
An hfq mutant exhibits pleiotropic phenotypes [ Tsui94 , Muffler97 , Wachi99a , Takada99 ].
Hfq: "host factor for phage Q beta"

Front 169

yciH

Back 169

translation initiation factor
YciH has similarity to eIF1, a eukaryotic translation initiation factor which is functionally equivalent to the eubacterial IF3, with the exception of the proofreading function of the IF3 C-terminal domain [ Cort99 , Lomakin06 ]. YciH is able to perform some of the functions of IF3, indicating that eIF1 and YciH may have a common evolutionary origin [ Lomakin06 ].
Purified YciH is able to bind to both the 30S subunit of eubacterial ribosomes and the 40S subunit of eukaryotic ribosomes. Like IF3, YciH reduces ribosomal complex formation at initiation sites with codon-anticodon mismatches and discriminates against initiation complexes formed with mutant tRNAs [ Lomakin06 ].
A solution structure shows that YciH is a member of a new superfamily of alpha-beta plait domain proteins [ Cort99 ]

Front 170

appC

Back 170

cytochrome bd-II terminal oxidase subunit I
The appC encoded protein is subunit 1 of E. coli cytochrome bd-II terminal oxidase.

Front 171

cytochrome bo terminal oxidase

Back 171

The E. coli K-12 genome contains gene clusters for 3 cytochrome oxidase enzymes - cytochrome bo oxidase (CyoABCD), cytochrome bd-I oxidase (CydAB) and cytochrome bd-II oxidase (AppCD). The three enzymes function as the major terminal oxidases in the aerobic respiratory chain of E. coli. Cytochrome bo oxidase genes (cyoABCD) are expressed when oxygen levels are high while cytochrome bd-I oxidase genes are expressed under oxygen limited conditions. Both enymes contribute to the generation of a proton motive force (PMF), cytochrome bo oxidase functions as a proton pump whilst cytochrome bd-I does not [ Puustinen91 ]. Cytochrome bd-II does not contribute to the generation of a PMF and may function to uncouple catabolism from ATP synthesis.

The cytochrome bo terminal oxidase catalyzes the two-electron oxidation of ubiquinol within the membrane and the four-electron reduction of molecular oxygen to water. In the cell the enzyme functions as a proton pump, with a net movement of 2H+/e- across the cytoplasmic membrane, thereby generating a proton-motive force [ Puustinen91 ]. Expression of the cyo operon is negatively regulated by Fnr and the ArcA/ArcB two component system under anaerobic conditions [ Iuchi90a , Cotter90 ]. Expression varies depending on the carbon source used for growth - being highest on non-fermentable carbon sources and lowest on glucose [ Cotter90 ]. Expression is induced by iron limitation [ Cotter92a ].

Cytochrome bo terminal oxidase consists of four subunits encoded by the cyoB, cyoA, cyoC and cyoD genes, all of which are necessary for a functional enzyme. Analyses of intermediates in the assembly of cytochrome bo oxidase indicate that assembly of the complex is an ordered process whereby subunit III and IV assemble first, followed by subunit I and finally subunit II [ Stenberg07 ]. The cytochrome bo complex is similar to the aa3-type family of cytochrome c oxidases [ Chepuri90a ].

A crystal structure of the entire cytochrome bo terminal oxidase complex has been determined at 3.5 Å resolution [ Abramson00 ]

Front 172

cyoD

Back 172

cytochrome bo terminal oxidase subunit IV
CyoD is subunit IV of the cytochrome bo terminal oxidase complex encoded by cyoABCDE. Although this subunit's function is unknown, it is necessary for a functional enzyme [ Neidhardt96 ].

The CyoD polypeptide contains three transmembrane helices [ Chepuri90 ]. Deletion and cross-linking studies have suggested that subunit IV interacts with subunits I and III [ Saiki96 ], which is confirmed by the crystal structure of the entire cytochrome bo terminal oxidase complex that has been determined at 3.5 Å resolution [ Abramson00 ]

Front 173

cyoA

Back 173

cytochrome bo terminal oxidase subunit II
CyoA is subunit II of the cytochrome bo terminal oxidase complex encoded by cyoABCDE. Crosslinking studies suggested that subunit II functions as a ubiquinone binding site of the cytochrome bo terminal oxidase complex [ Welter94 ]. However, the crystal structure of the entire cytochrome bo terminal oxidase complex suggests that a potential ubiquinone binding site is instead located in the membrane domain of subunit I [ Abramson00 ].

The CyoA polypeptide contains two transmembrane helices [ Chepuri90 ]. CyoA is a lipoprotein; during maturation, the protein is modified by attachment of fatty acids and protease cleavage at C25; however, the posttranslational modification is not essential for assembly or activity of the cytochrome bo terminal oxidase complex [ Ma97 , Brokx04 ]. In vitro models indicate that posttranslational modifications to CyoA occur after membrane insertion. The same models also demonstrate that YidC, the Sec translocon (Sec YEG) and SecA are required for efficient insertion of cyoA into the membrane [ duPlessis06 ].

Protease mapping assays, using strains with inactivated or depleted components of the Sec translocon, SecA and YidC, indicate that membrane integration and assembly occur in separate, sequential steps; insertion of the large periplasmic C-terminal segment requires both the Sec translocon and SecA whereas YidC is sufficient for insertion of the N-terminal domain [ vanBloois06 ].

Mutation of residues within the signal peptide and first hydrophobic domain of CyoA, resulting in a non-neutral overall charge of the first periplamic domain, confirm results obtained in in vitro studies [ duPlessis06 ] and show that the overall neutral charge in the first periplasmic domain of CyoA is required for membrane insertion [ Celebi08 ].

The three-dimensional structure of the periplasmic fragment of CyoA has been determined to 2.3 Å resolution [ vanderOost93 , Wilmanns95 ]. A crystal structure of the entire cytochrome bo terminal oxidase complex containing CyoA has been determined at 3.5 Å resolution [ Abramson00 ].

Under anaerobic conditions, cyoA is repressed 140-fold compared to growth under aerobic conditions. This regulation is in part due to repression by Fnr [ Cotter90 ]

Front 174

atpC

Back 174

ATP synthase, F1 complex, ε subunit
The epsilon subunit appears to play an important role in coupling the catalytic site events with proton translocation in association with the gamma subunit. The coupling involves conformational changes and probable translocations of one or both subunits. This subunit is also required for binding of the F-1 complex to the F-O complex. [ Tang96a , Senior90 ]

Front 175

atpA

Back 175

The α-subunit of the ATP synthase plays an essential role in the catalytic mechanism of the enzyme and in the binding and coupling between the F1 and F0 complexes. The α-subunit also contains an adenine-specific binding site which is noncatalytic, nonregulatory and not essential for enzyme assembly in vitro. Its function has not yet been determined. The α-subunit complex is a homotrimer [ Rao88 ].

A hydrogen-bonding network is formed at the closed α/β-subunit interface of F1 [ Abrahams94 ]. Elimination of this network results in a severely impaired enzyme. A possible role for the hydrogen-bonding network in coupling of ATP synthesis/hydrolysis and rotation has been proposed [ Mao08 ]. The role of conserved residues surrounding the catalytic site has been studied [ Li09 ]

Front 176

ftsE

Back 176

FtsE is the putative ATP-binding protein component of an ATP-binding cassette (ABC) transporter. FtsE dimerizes and associates with the inner membrane via interaction with FtsX, an integral membrane protein [ deLeeuw99 ].

FtsE and FtsX localize to the cell division site; localization is dependent on FtsZ, FtsA and ZipA, but not FtsK, FtsQ, FtsL and FtsI [ Schmidt04 ]. FtsEX is important for assembly or stability of the septal ring under low-salt growth conditions [ Schmidt04 ].

An ftsE null mutant is only viable on high salt medium [ deLeeuw99 ]. ftsE and ftsX are located in an operon with the ftsY gene. ftsY encodes a GTPase which acts as a receptor for the SRP (signal recognition particle) involved in protein targeting [ Gill87 , Miller94 ]. It has been suggested that FtsE and FtsX are not to be required for SRP-mediated targeting [ deLeeuw99 ], although it has been reported that ftsE mutants are affected in translocation of potassium ion pump proteins into the cytoplasmic membrane [ Ukai98 ]

Front 177

rplD

Back 177

The L4 protein is a component of the 50S subunit of the ribosome and also regulates the expression of S10 operon at both the transcriptional and posttranscriptional level. The functions of L4 in ribosome assembly, attenuation, and translational regulation of the operon are separable [ Freedman87 , Li96d ].

Addition of L4 to an in vitro protein synthesis reaction inhibits the synthesis of the promoter-proximal proteins in the S10 operon, suggesting that L4 may be an inhibitor of translation [ Yates80a ]. In vivo, overproduction of L4 was found to reduce the synthesis of S10 operon mRNA [ Zengel80 ]. The steady-state growth rate-dependent control of the S10 operon expression requires additional factors [ Lindahl90 ].

L4 stimulates premature termination (transcription attenuation) at a NusA-dependent terminator site 30 bases upstream of the first structural gene of the S10 operon, and termination requires the function of NusA [ Lindahl83 , Zengel90 ]. The attenuator hairpin region is sufficient for NusA-dependent pausing, and a second hairpin region immediately upstream of the attenuator hairpin is necessary for L4 to prolong the pause [ Zengel92 , Sha95 , Zengel96 ]. Structural and sequence requirements for L4-mediated transcription termination within the S10 leader region have been determined [ Zengel02 ]; the first 150 bases of the S10 leader region are sufficient for L4-mediated attenuation control [ Zengel90a ], and a 64-nucleotide sequence is required for L4 interaction with the S10 leader mRNA [ Stelzl03 ]. Binding of L4 to the leader region induces structural changes in the mRNA [ Stelzl03 ]. A region of 110 bases within domain I of 23S rRNA competes with paused transcription complex for binding of L4 [ Zengel91 , Zengel93 ].

L4 interacts with the 5' segment of 23S rRNA [ Spierer75 , Gulle88 , Maly80 , Urlaub95 , Thiede98 ]. The L4 and L24 binding sites in 23S rRNA localize to a small fragment [ Stelzl00 ] and may be a key element for rRNA folding in the early assembly pathway for the ribosome [ Stelzl01 ]. The L4 binding region within domain I of 23S rRNA has three-dimensional structural similarity to its binding region on the S10 operon leader mRNA where it inhibits its own translation [ Ostergaard98 , Stelzl03 ].

L4 is a component of the binding site for erythromycin on the ribosome [ Arevalo88 ]. An L4 mutant is resistant to erythromycin [ Apirion67 , Wittmann73 ]; this mutant has a cold-sensitive growth defect, and 50S subunit assembly is defective [ Chittum94 ]. L4, L5, and L21 bind to erythromycin cooperatively [ Pye90 ]. The extended loop region of L4 (amino acids 40-88) contributes to the lining of the peptide exit tunnel in the ribosome and is contacted by the growing peptide chain [ Houben05 ]. While deletions of this region of L4 do not appear to affect function of L4 [ Zengel03 ], the K63E point mutation that results in erythromycin resistance alters the decoding accuracy of the ribosome [ OConnor04 ]

Front 178

rpsE

Back 178

The S5 protein is a component of the 30S subunit of the ribosome. It was suggested that S5 is positioned to have access to the interface between the 30S and 50S subunits of the ribosome [ Culver99 ].

The N-terminal half of S5 contains the sites of mutations that confer resistance to spectinomycin [ Piepersber75 , Dekio69 , DeWilde73 ] and binds non-specifically to helix 34 of the 16S rRNA [ Heilek96a , Stern88a ]. S5 may be involved in modulating the conformation of the 16S rRNA [ Lodmell97 ]. S5 can be cross-linked to mRNA [ RinkeAppel91 ] and tRNA [ Graifer89 ].

S5 influences translational fidelity. Certain mutations in rpsE confer a "ribosomal ambiguity" phenotype, which is characterized by decreased growth rate, increased streptomycin sensitivity, and increased errors in translation [ Ito73 , Hasenbank73 ].

The S5 protein is acetylated at the N-terminus [ WittmannLi78 , Arnold99 ]; mutants in the alanine acetyltransferase enzyme, RimJ, are temperature sensitive [ Cumberlidg79 , Yoshikawa87 ].

Front 179

rpsD

Back 179

The S4 protein, a component of the 30S subunit of the ribosome, functions in the assembly of the 30S ribosomal subunit, the mRNA helicase activity of the ribosome, the regulation of translation of a subset of ribosomal proteins, and transcription antitermination of rRNA operons.

S4 interacts directly with helical elements at the 5' domain of the 16S rRNA [ Stern86 , Powers95a , Vartikar89 ]. The ability of both S4 and S7 to bind 16S rRNA by themselves indicates that they function as initiator proteins for the assembly of the 30S subunit of the ribosome. The S20, S16, S15, S6, and S8 subunits appear to depend on S4 for assembly [ Nowotny88 ].

The S4 protein is involved in the regulation of translation of the other ribosomal proteins encoded by the α operon, RpsM (S13), RpsK (S11), RplQ (L17) and S4 itself [ Yates80 , Thomas87 , JinksRober82 ]. The α operon leader region is required for translational repression by S4 [ Thomas87 ]; S4 specifically interacts with a double pseudoknot structure which overlaps with the ribosome binding site and initiation codon for RpsM [ Tang89 , Tang90 ]. There may be a second binding site for S4 upstream of the RplQ open reading frame [ Meek84 ]. The same protein domain appears to be responsible for both mRNA and rRNA binding [ Baker95a , Conrad87 ].

S4 can also act as a general transcription antitermination factor similar to NusA; it associates with RNA polymerase and is involved in rRNA operon antitermination [ Torres01a ].

S4 influences translational fidelity [ Topisirovi77 ]. Certain mutations in rpsD confer a "ribosomal ambiguity" phenotype, which is characterized by decreased growth rate, increased streptomycin sensitivity, and increased errors in translation [ Zimmermann71 , Andersson83 , Andersson82a , Olsson79 ]. Cells carrying the rpsD14 allele have a mutator phenotype [ Balashov03 ]. The ribosome was found to have mRNA helicase activity, and mutations in the S3 and S4 subunits impair this activity [ Takyar05 ]

Front 180

rlmE

Back 180

RlmE is the methyltransferase responsible for methylation of 23S rRNA at the 2'-O position of the ribose at the universally conserved U2552 nucleotide [ Caldas00a ]. In vitro, the enzyme is active on ribosomes and the 50S ribosomal subunit, but not free rRNA [ Caldas00a , Bugl00 ].

A crystal structure of RlmE has been solved at 1.5 Å resolution [ Bugl00 ]. Site-directed mutagenesis has identified possible active site and substrate binding residues, and a reaction mechanism has been proposed [ Hager02 , Hager04 ].

A mutant strain lacking RlmE has a decreased growth rate at all temperatures tested and shows reduced protein synthesis activity and accumulation of free ribosomal subunits [ Caldas00 , Bugl00 ]. Overexpression of the small GTPases Obg, Der [ Tan02 , Hwang10a ] or CtgA [ Jiang06a ] suppresses the ribosomal assembly or stability defect of an rlmE mutant without restoring the methylation of U2552 [ Tan02 ]. An rlmE-deficient strain shows a decrease in -1 and +1 frameshifting and a decrease in UAA and UGA stop codon readthrough, suggesting that the U2552 base may interact with aminoacyl-tRNAs at the ribosomal A site [ Widerak05 ].

An rlmE mutant is more sensitive to lincomycin [ Caldas00 ], clindamycin, hygromycin A and sparsomycin [ Toh08 ] than wild-type

Front 181

rna

Back 181

RNase I, cleaves phosphodiester bond between any two nucleotides
Ribonuclease I (RNase I) is an endonuclease that cleaves phosphodiester bonds in RNA, yielding nucleoside 3'-phosphates and 3'-phosphooligonucleotides [ Spahr61 ]. RNase I is partially reponsible for the degradation of total and ribosomal RNA during both normal and nutrient starvation conditions, especially during carbon starvation [ Kaplan74 , Kaplan75 , Cohen77 ]. RNase I is specifically required for the breakdown of 23 S RNA, though it is not required for degradation of 16 S RNA or very small (4 S) RNA fragments that result from breakdown of larger RNA [ Kaplan75a , Kaplan75 ]. RNase I degradation of the 50 S ribosome releases the ribosomal proteins L4, L10 and L7/12 in addition to cleaving the 23 S RNA to yield a smaller product [ Raziuddin79 ]. Polyamines stimulate the activity of RNase I against synthetic polynucleotides in vitro [ Kumagai77 ].

RNase I is a periplasmic protein that can be released by spheroblasting or treatment of cells with N-dodecyldiethanolamine. This release allows subsequent enhanced breakdown of rRNA by RNase I [ Neu64 , NEU64 , NEU64a , Neu65 , Abrell71 , Lambert76 ]. RNase I appears to remain with the membrane fraction in disrupted cells [ Kaplan76 ].

Some structural analysis of RNase I has been completed. Its conformation energy is 11.5 kcal/mol at pH 7.5, and its T(m) is 64 degrees C at pH 4.0 [ Padmanabha01 ]. It has a 23-residue amino-terminal signal sequence which is cleaved and likely allows its transport to the periplasm [ Meador90 ]. Preliminary crystallization of RNase I has been done, with visualization of the structure at greater than three angstrom resolution [ Lim93 ].

Mutants lacking RNase I or other ribonuclease activities have reduced DNA degradation, possibly due to interaction between excess RNA and DNA endonuclease I [ Wright71 ]. RNase I is also required for full recovery from starvation, as cell viability studies show a direct correlation between recovery from starvation and the ability to degrade RNA [ Kaplan75a ]. A method has been developed to screen for mutants in RNase I by checking for a delay in β-galactosidase expression during amino acid starvation [ Kaplan73 ]

Front 182

rlmN

Back 182

RlmN is the methyltransferase responsible for methylation of 23S rRNA at the C2 position of the A2503 nucleotide [ Toh08a ]. The enzyme can utilize free 23S rRNA as the substrate, but not the fully assembled large ribosomal subunit [ Yan10 ].

RlmN belongs to the Cfr/RlmN family of the "radical SAM" superfamily of proteins [ Kozbial05 , Kaminska10 ]. The reaction requires two molecules of SAM, one as the methyl group donor, and the second as the source of the 5'-deoxyadenosyl radical [ Yan11a ]. The cysteine residues of the 4Fe-4S cluster-binding motif are essential for activity of RlmN [ Yan10 ]. The reaction proceeds via a ping-pong mechanism whereby the conserved C355 residue or RlmN is methylated using SAM as a donor in a typical SN2 displacement reaction, followed by reductive cleavage of a second molecule of SAM, which generates the 5'-deoxyadenosyl radical that is required for transfering the methyl residue to the C2 position of A2503 [ Grove11 ].

Crystal structures of RlmN in the absence and presence of SAM have been solved. The location of the bound SAM supports the proposed ping-pong reaction mechanism [ Boal11 ].

In an rlmN null mutant strain, A2503 is not methylated. Although the null mutant alone does not show any obvious growth defect, it is outcompeted by wild type in a growth competition experiment [ Toh08a ]. An rlmN null mutant shows slightly increased susceptibility to certain peptidyl transferase-targeting antibiotics [ Toh08a , Toh08 ]

Front 183

clpX

Back 183

ClpX is an ATP-dependent molecular chaperone that serves as a substrate-specifying adapter for the ClpP serine protease in the ClpXP and ClpAXP protease complexes.

ClpX protects the lambda O protein from heat-induced aggregation, disassembles lambda aggregates and enhances lambda DNA binding. ATP binding is required for all these effects, and disaggregation requires ATP hydrolysis [ Wawrzynow95 ]. ClpX also converts inactive, dimeric TrfA into its monomeric form (capable of initiating replication of plasmid RK2) in an ATP-dependent manner [ Konieczny97 ].

ClpX is required for normal replication of Mu transposase [ MhammediAl94 ]. ClpX catalyzes the ATP-dependent release of MuA from its active transposase tetramer form, allowing recruitment of host factors necessary for post-recombination steps in Mu transposition [ Levchenko95 , Kruklitis96 ]. ClpX is also able to globally unfold MuA monomers. ClpX recognizes a ten amino acid peptide from the carboxy-terminus of MuA when it is revealed by MuB. ClpX will recognize other proteins with this tag artificially attached [ Levchenko97 ].

SspB binding stimulates ClpX ATPase activity [ Wah02 ].

ClpX is a hexamer of ClpX monomers, stabilized by ATP binding and capable of capping ClpP tetradecamers [ Grimaud98 ]. While the ClpX ATP-binding site is necessary for oligomerization and binding to ClpP, both processes continue in the absence of ATP [ Banecki01 ]. ATP-bound ClpX is protease resistant [ Singh01 ]. The carboxy-terminus of ClpX is required for interaction with ClpP, as is the tripeptide IGF, though the latter is dispensible for ClpX chaperone activities [ Kim01f , Singh01 ]. Mutations in the interface between the carboxy-terminus of each subunit and the ATPase domain of its neighbor prevent disassembly of bound substrate [ Joshi03a ]. Substrate recognition requires that three of the six monomers have bound ATP, while the other three may have bound ATP or ADP [ Hersch05 ].

Each ClpX monomer has two PDZ domains the bind to the carboxy-terminus of target proteins. These domains show up as disordered sequence in NMR and are unstable when expressed independently [ Levchenko97a , Smith99b ]. ClpX also has an ATP-binding site motif and a zinc-binding domain, the latter being a member of the treble clef zinc finger family, involved in macromolecular interactions [ Gottesman93 , Donaldson03 ].

ClpX is required for adaptation to and extended viability in stationary phase, as well as growth in SDS [ Weichart03 , Rajagopal02 ].

ClpX can be expressed without ClpP [ Yoo94 ]

Front 184

ClpXP

Back 184

ClpXP is a serine protease complex responsible for the ATP-dependent degradation of a wide range of proteins [ Gottesman93 , Wojtkowiak93 ].

ClpXP degrades the altered Mu immunity repressor, Vir. When Vir is present, the normal immunity repressor, Rep, becomes more vulnerable to ClpXP-mediated degradation as well [ Welty97 ]. ClpXP can also degrade MuA, although it does not degrade it all, allowing ClpX to act in its chaperone capacity to assist MuA function [ Jones98 , MhammediAl94 ].

ClpXP is partially responsible for degradation of proteins with the SsrA degradation tag, including SsrA-tagged lambda repressor [ Bohn02 , Gottesman98 ]. ClpXP degrades stably folded SsrA proteins efficiently, but only poorly degrades proteins bearing SsrA tags artificially attached in the middle of their sequences via cysteine linkages [ Kenniston04 ].

ClpXP can degrade DNA-bound lambda O protein when transcription is possible, otherwise, it is stable [ Zylicz98 ]. ClpXP-mediated degradation of lambda O protein can affect the lysis/lysogeny decision under certain growth conditions [ Wojtkowiak93 , Czyz01 ].

ClpXP is also required for degradation of the starvation-induced proteins Dps and sigma S during exponential growth [ Stephani03 , Schweder96 ].

Several other ClpXP substrates have been discovered. ClpXP degrades variants of the restriction enzyme EcoKI that have impaired enzymatic function, the mutagenically active protein UmuD' when it is in a heterodimer with unmodified UmuD and the antitoxin Phd from the Doc-Phd toxin/antitoxin pair (from plasmid prophage P1) [ ONeill01 , Frank96 , Lehnherr95 ].

Putative ClpXP substrates were found by trapping with inactivated ClpP. Potential substrates included some with SsrA-like tails (crl, dksA, fnr, iscR, rplJ, rplU, gcp, pepB, katE, nrdH, tpx, chew, cysA, exbB, acnB, aldA, glpD, glyA, IldD and ycbW), MuA-like carboxy-terminal motifs (paaA, pncB, ribB, ybaQ), novel amino-terminal binding motifs (crl, dksA, fnr, lexA, rpoS, rplE, rplJ, rplK, rplS, rplU, tufB, dps, katE, nrdH, tpx, insH, chew, cysA, gatA, ompA, secA, aceA, atpD, cysD, dada, fabB, gapA, gatY, gatZ, glcB, glyA, iscS, lipA, moaA, pncB, tnaA, udp, ybaQ and ycbW) and no specific binding motifs (rseA, rplN, lon, clpX, dnaK, groEL, ftsZ, iscU, yebO and ygaT) [ Flynn03 ].

SspB binds to SsrA-tagged proteins via its amino-terminal domain and enhances their degradation by ClpXP through carboxy-terminal binding to ClpXP [ Levchenko00 , Wah03 ]. SspB alone is sufficient to allow interaction with ClpXP. A protein that has been covalently linked to SspB becomes a ClpXP substrate even in the absence of an SsrA tag [ Bolon04 ]. ClpX and SspB bind to overlapping parts of the SsrA tag, weakening the direct SsrA-ClpX interaction. The SspB-ClpX interaction overcomes this weakening effect [ Hersch04 ]. Trapping experiments based on SspB show that RseA, which is cleaved from the membrane and binds to sigma E as an inhibitor during stress interacts with SspB and is degraded by ClpXP, thus releasing sigma E [ Flynn04 ]. Sigma S degradation by ClpXP requires the adaptor RssB, which binds to Region 2.5 of sigma S, allowing binding of ClpX at the amino-terminus and subsequent degradation by ClpXP [ Muffler96 , Zhou01a , Studemann03 ]. Each ClpX hexamer has three SspB binding domains to match up with two ClpXP binding domains per SspB dimer, so only one SspB dimer can function with a given ClpX hexamer at a time [ Bolon04a ]. UmuD operates in a manner similar to SspB, binding to the ClpX amino-terminus and serving as a substrate tether for UmuD' [ Neher03 ].

ClpXP consists of a ClpP tetradecamer capped at one or both ends by ClpX hexamers [ Grimaud98 ].

Substrates bind to the distal surface of ClpX, and then are passed off to the inner cavity of ClpP to be degraded, in a process that is driven by ATP and modulated by ClpXP protease specificity-enhancing factor [ Ortega00 , Thibault06 ]. This process involves both static and dynamic contacts between ClpX and ClpP [ Martin07 ]. The initial ClpX-mediated denaturation of substrate is the rate-limiting step in degradation of a well-folded protein, such as SsrA-tagged GFP [ Kim00b ].

ClpXP is required for limitation of lambda phage early DNA replication during slow growth [ Wegrzyn ].

ClpXP is required for acquisition of the genes encoding the restriction enzymes EcoKI and EcoAI by conjugation or transformation [ Makovets98 ].

Despite lambda O initiator protein being a ClpXP substrate, lambda replication does not depend on ClpXP levels [ Szalewska94 ]

Front 185

rsmA = ksgA

Back 185

KsgA is the methyltransferase responsible for dimethylation of 16S rRNA at the two adjacent adenosine bases A1518 and A1519 [ Poldermans79 ]. In vitro, the enzyme is active on 30S ribosomal subunits, but not the fully assembled 70S ribosome [ Poldermans79 ]. KsgA may play a role in biogenesis of the small subunit of the ribosome [ Connolly08 ].

The evolutionary relationship and functional divergence of this enzyme and its homologs in eukaryotes has been studied [ Cotney06 ]. KsgA as well as the dimethylation modification it catalyzes are nearly universally conserved, suggesting an important function. KsgA is only able to methylate 30S ribosomal subunits in their translationally inactive conformation [ Desai06 ]; it interacts with the decoding site of the 30S subunit, and interactions of 30S with KsgA and IF3 appear to be mutually exclusive. A checkpoint model where binding of KsgA keeps immature 30S subunits from entering the translational cycle has been suggested [ Xu08 ].

Methylation of the two adenosine residues is independent of each other [ Cunningham90 , VilaSanjur99 ]. Recognition of the 3' terminal hairpin of 16S rRNA and methylation activity of KsgA have been studied [ Formenoy94 ].

Surprisingly, KsgA also exhibits cytosine-DNA glycosylase activity and may play a role in protection of DNA against oxidative stress [ ZhangAkiya09 ].

A crystal structure of KsgA has been solved at 2.1 Å resolution [ OFarrell04 ].

Mutation of ksgA causes resistance to kasugamycin (an inhibitor of translation initiation), but no substantial growth defect [ Sparling70 , Dabbs80 , Leveque90 ]. The lack of methylation at A1519, but not A1518, appears to be responsible for the kasugamycin resistance phenotype of a ksgA mutant [ VilaSanjur99 ]. A ksgA deletion enhances the slow growth phenotype of an rsgA mutation [ Campbell08 ]. ksgA was also identified as a multicopy suppressor for a cold-sensitive mutant of era, which encodes an essential GTP-binding protein [ Lu98 ]. The methyltransferase and multicopy suppressor activities of KsgA are separable, indicating a possible second function of KsgA [ Inoue07a ]. High-level overexpression of KsgA is toxic, and the cells show increased sensitivity to acid shock [ Inoue07a ]. A ksgA deletion strain is cold sensitive and shows an altered ribosome profile and altered 16S rRNA processing [ Connolly08 ]. AksgA point mutant lacking catalytic activity has a dominant negative effect on growth and ribosome formation [ Connolly08 ].

The KsgA protein binds specifically to its own mRNA and may regulate its own translation [ vanGemen89 ]. Growth rate positively regulates ksgA expression [ Pease02 ]

Front 186

map

Back 186

methionine aminopeptidase
All known proteins in E. coli use N-formyl methionine as the first amino acid in a peptide chain. Amino-terminal maturation involves two enzymes, a deformylase which removes the formyl group, and methionine aminopeptidase (MAP), which catalyzes the removal of the deformylated methionine residue [ Neidhardt96 , BenBassat87 ]. The activity of MAP is dependent on the identity of the second, third and fourth amino acid residues of the target protein [ Hirel89 , Frottin06 , BenBassat87 ]; the substrate specificity has been analyzed in detail [ Xiao10 ]. The most preferred amino acid residue in the position following the fMet is alanine.

The map gene encoding methionine aminopeptidase is essential for growth in E. coli [ Chang89 ]. E. coli MAP is a type-I enzyme and is a potential antibiotic target; selective inhibitors have been designed [ Swierczek05 , Wang09a , Mitra09 ].

Crystal structures of the protein itself and in complex with various inhibitors have been reported [ Roderick93 , Lowther99 , Ye04a , Huang06b , Evdokimov07 , Huang07 , Ma07b , Wang08d ], and the catalytic mechanism was studied in cocrystals with various substrate and transition state analogs [ Lowther99a , Ye06 ].

MAP contains two metal binding sites; reports differ on whether a single metal ion is sufficient for catalysis [ Cosper01 , Ye06 , Huang07 , Chai09 ], or both are required [ Hu07 , Mitra08 ]. The physiologically relevant metal cofactor of MAP is most likely Fe2+ rather than Co2+ [ Chai08 ]. Studies with inhibitors that were active in vitro, but not in vivo, indicated that high metal concentrations in in vitro assays may have led to artefacts [ Schiffmann05 , Schiffmann06 ].

The activity of enzymes containing mutations in predicted active site and metal-binding residues has been measured [ Chiu99 , Copik03 , Li04c , Watterson08 , Mitra09a , Mitra08 ]

Front 187

rmf

Back 187

ribosome modulation factor
Rmf is a ribosome modulation factor that reversibly converts active 70S ribosomes to a dimeric form (100S ribosomes), which appears during the transition from exponential growth to stationary phase and is associated with a decrease in overall translation activity [ Wada90 ]. Rmf specifically associates with 100S ribosomes in a 1:1 ratio and does not associate with 70S, 50S, and 30S ribosomal particles [ Wada95 ]. Rmf binds near the ribosomal proteins S13, L13, and L2, close to the peptidyl-tRNA binding site [ Yoshida02e ], and protects certain bases in the 23S rRNA, including A2451, which is thought to be involved in the peptidyl transferase activity [ Yoshida04b ]. This suggests a possible mechanism for Rmf-dependent inactivation of translation. In late stationary phase, the ribosome dimers dissociate, which is followed by disassembly of the 70S ribosomes and loss of viability [ Wada00 ].

Purified Rmf protein can cause dimerization of 70S ribosomes in vitro, inhibiting protein synthesis and binding of aminoacyl-tRNA to ribosomes [ Wada95 ].

An rmf mutant shows reduced viability during stationary phase and does not contain ribosome dimers [ Yamagishi93 ]. An rmf mutant is also heat sensitive in stationary phase [ Niven04 ]. An rmf ompC double mutant and an rmf ompC rpoS triple mutant show even further reduced cell viability; this synthetic phenotype may be due to decreased levels of Mg2+ [ Apirakaram98 , Samuel02 ].

Expression of rmf is induced at the transition from exponential growth to stationary phase or during slow growth; rmf expression appears to be inversely proportional to the growth rate [ Yamagishi93 ]. rmf expression is positively regulated by the stringent starvation factor ppGpp [ Izutsu01a ]. The rmf mRNA is extremely stable during stationary phase, with a half life of approximately 24 minutes in early and 120 minutes in late stationary phase; transfer of the cells into fresh medium leads to an immediate drop in half life to approximately 5 minutes. rmf mRNA degradation requires RNA polymerase activity, RNase E, and PcnB. Transfer of stationary phase cells also causes Rmf protein levels to drop rapidly, and inactive ribosome dimers dissociate into active 70S ribosomes [ Aiso05 ]. rmf was one of only three genes whose expression changed under all stress conditions tested by [ Moen09 ]

Front 188

raiA = yfiA

Back 188

stationary phase translation inhibitor and ribosome stability factor
YfiA interferes with translation elongation, specifically by occluding the access of aminoacyl-tRNA to the ribosomal A site [ Agafonov01 ]. However, at physiological Mg2+ concentrations during in vitro poly(U) translation, YfiA inhibits misincorporation of leucine more strongly than incorporation of phenylalanine, the appropriate amino acid, suggesting an anti-miscoding activity [ Agafonov04 ].

YfiA stabilizes the association of the subunits of the 70S ribosome [ Agafonov99 ]. YfiA associates with 70S ribosomes at stationary phase, and may participate in the ribosome dimerization and modulation of activity that is observed at stationary phase [ Maki00 ]. YfiA binds to an exposed surface of the 30S ribosomal subunit such that the 50S ribosomal subunit occludes YfiA exposure in the context of the 70S ribosome [ Agafonov99 ]. YfiA binds to 70S ribosomes while YhbH binds to 100S ribosome dimers at stationary phase [ Maki00 ]. YfiA and YhbH exhibit a low amount of association with 70S ribosomes during log phase growth [ Maki00 ]. Binding to 16S ribosomal RNA is not detected [ Rak02 ].

YfiA has similarity to YhbH [ Maki00 ]. YfiA has structural similarity to Haemophilus influenzae HI0257 and to Drosophila Staufen protein [ Rak02 ].

NMR structural characterization has been performed [ Ye02a , Kalinin02 , Rak02 ]. Protein purification has been described [ Agafonov99 ].

RaiA: "ribosome-associated inhibitor" [ Agafonov01 ]

Front 189

hpf

Back 189

predicted ribosome-associated, sigma 54 modulation protein
YhbH binds to 100S ribosome dimers while YfiA binds to 70S ribosomes at stationary phase [ Maki00 ]. YfiA and YhbH exhibit a low amount of association with 70S ribosomes during log phase growth [ Maki00 ].

YhbH has similarity to YfiA [ Maki00 ].

Regulation has been described [ DeLisa01 ]. The yhbH gene is induced by the autoinducer 2 (AI-2) quorum sensing pheromone [ DeLisa01 ]

Front 190

frdB

Back 190

fumarate reductase iron-sulfur protein
This is one of two catalytic subunits of the four subunit enzyme. This subunit contains three iron-sulfur clusters: a 4Fe-4S, a 3Fe-4S and a 2Fe-2S.

This subunit has 38% identity to the succinate dehydrogenase iron-sulfur cluster subunit, SdhB [ Darlison84a ]

Front 191

fnr

Back 191

FNR is the primary transcriptional regulator that mediates the transition from aerobic to anaerobic growth through the regulation of hundreds of genes. Generally, this protein activates genes involved in anaerobic metabolism and represses genes involved in aerobic metabolism [ Salmon03 , Kang05 ]. FNR also regulates the transcription of many genes with other functions, such as acid resistance, chemotaxis, cell structure, and molecular biosynthesis, among others [ Salmon03 , Kang05 ].

The cellular concentration of FNR is similar under both anaerobic and aerobic growth [ Sutton04 ], but its activity is regulated directly by oxygen. Under anaerobiosis, FNR acquires a [4Fe-4S] cluster that causes a conformational change and dimerization of the protein that causes it to become activated [ Moore01 ]. Purification of [4Fe-4S]-FNR in an O2-free environment has been described [ Yan09 ]. The presence of O2 results in inactivation of FNR via oxidation of this [4Fe-4S] cluster into the [2Fe-2S] cluster [ Jervis09 , Sutton04 , Khoroshilo97 ] and the disassembly of the dimer [ Lazazzera96 ]. After prolonged O2 exposure, the [2Fe-2S] cluster is destroyed, and apo-FNR, which lacks an Fe-S cluster, is the primary form of FNR under aerobiosis [ Reinhart08 , Sutton04a ]. Nitric oxide is also able to inactivate FNR through the nitrosylation of the [4Fe-4S] cluster [ CruzRamos02 ].

Under aerobiosis, the apo-FNR monomer is exposed and can be degraded by the ClpXP protease [ Mettert05 ]. Two motifs of FNR appear to be necessary for this degradation, one located in the N-terminal region and the other in the C-terminal region [ Mettert05 ].

The activated FNR conformation is able to bind a specific palindromic sequence of DNA with the consensus sequence TTGATNNNNATCAA [ Gerasimova01 , Eiglmeier89 ]. The G and the first T of each FNR half-site appears to interact with FNR residues Glu-209 and Ser-212 [ Spiro90 ]. FNR activates the transcription from class I and class II promoters, in which the FNR-binding sites are located around -61, -71, -82, or -92 in class I and around -41.5 in class II [ Wing95 ]. In class I promoters the second N and in class II promoters the third N of the consensus FNR site tend to be an A-T [ Scott03a ].

During activation, three activating regions (AR) are surface exposed for contact with the polymerase and promote transcription. The AR1 region contacts with the α subunit [ Lee00d ], the AR2 region makes contact with the &sigma70 [ Blake02 ], and the AR3 contacts the α NTD of the RNA polymerase [ Lamberg02 ]. The group amino acids Thr-118 with Ser-187; Lys-49 with Lys50; and Asp86 and Ile81 with Gly-85 are important for AR1, AR2, and AR3, respectively, to make contact with the RNA polymerase [ Lee00d , Blake02 , Lamberg02 ]. The monomer in the FNR dimer that makes contact with the RNA polymerase and the activating region in FNR that makes the contact depends on the class of promoter, class I or class II [ Lee00d , Blake02 , Lamberg02 ].

FNR belongs to the CRP/FNR superfamily of transcription factors whose members are widely distributed in bacteria [ Korner03 ]. These proteins have an N-terminal sensory domain, a C-terminal helix-turn-helix DNA-binding domain, and a dimerization motif in between [ Korner03 ]. The sensory domain of FNR contains five cysteine residues, four of which are essential for linking the [4Fe-4S] cluster [ Green93 ].

Under anaerobic growth conditions, transcription of the fnr gene is negatively autoregulated [ Mettert07 ].

FNR was named for the mutant defect in "fumarate and nitrate reduction" [ Lambden76 ]

Front 192

bipA

Back 192

protein possibly involved in ribosome structure or function
BipA is a member of the ribosome-binding GTPase superfamily. The variety of phenotypes of a bipA deletion, as well as the genetic interaction with the RluC 23S rRNA pseudouridine synthase suggest that BipA plays a role in modulating the structure or function of the ribosome [ Krishnan08 ].

BipA is necessary for growth under low temperature conditions [ Pfennig01 , Grant01b ]. Deletion of rluC suppresses the cold-sensitive phenotype of a bipA mutant [ Krishnan08 ].

A bipA mutation suppresses the defect in lipopolysaccharide core biosynthesis, the SDS sensitivity, and the defect in mouse intestinal colonization of a waaQ mutant [ Moller03 ]. BipA may be involved in positive regulation of colanic acid synthesis [ Krishnan08 ]. A strain containing a transposon insertion in bipA is more sensitive to chloramphenicol than wild-type [ Duo08 ].

Whereas BipA of E. coli K-12 is reported not to be tyrosine phosphorylated [ Freestone98a , Pfennig01 ], BipA of enteropathogenic E. coli [ Farris98 , Freestone98a ] and wall-less L-form E. coli [ Freestone98 ] is phosphorylated on tyrosine.

BipA has been characterized in pathogenic bacteria [ Qi95a , Farris98 , Freestone98a , Farris98a , Barker00 , Rowe00 , Grant03a , Scott03b ]. Salmonella enterica serovar Typhimurium BipA is a ribosome-associated GTPase [ Qi95a , deLivron08 ]. BipA of enteropathogenic E. coli is a regulatory protein [ Freestone98a , Rowe00 , Farris98 , Grant03a ] involved in pathogenesis [ Farris98 , Grant03a ] and is required for wild-type translation [ Freestone98a ].

BipA: "BPI-inducible protein A" [ Qi95a ]

Front 193

rpe

Back 193

ribulose-5-phosphate 3-epimerase
Ribulose-5-phosphate 3-epimerase (Rpe) is an enzyme of the non-oxidative branch of the pentose phosphate pathway.

Rpe requires ferrous iron for activity and is vulnerable to damage by H2O2 due to Fenton chemistry. Mn2+, Co2+ and Zn2+ can substitute for Fe2+ to varying degrees, and Rpe containing these alternative cations is not vulnerable to H2O2. Induction of the managnese transporter can protect Rpe from H2O2 damage [ Sobota11 ].

An rpe mutant strain does not grow on minimal medium containing either ribose or xylose, but grows when both sugars are present [ Lyngstadaa98 ]. rpe mutant strains have a growth defect in complex medium and a more severe growth defect in minimal medium containing glycolytic carbon sources or gluconate [ Lyngstadaa95 , Sakakibara97 , Lyngstadaa98 ]. A mutation in rpe, drsE30, supresses the temperature-sensitive growth defect of the dnaR130 mutant allele [ Sakakibara97 ]

Front 194

ffs

Back 194

The 4.5S RNA and Ffh together form the signal recognition particle (SRP) [ Ribes90 , Poritz90 , Luirink92 ], which binds to the signal sequence [ Luirink92 ] and targets the nascent protein to the SRP receptor, FtsY, [ Miller94 ]. The SRP is involved in integration of nascent proteins into the membrane [ Macfarlane95 , deGier96 , deGier98 , Koch99 , Tian00 ].

The interaction between 4.5S RNA and Ffh, [ Wood92 , Lentzen94 , Lentzen96 , Zheng97 , Suzuma99 ], the GTP-dependent [ Miller94 ] interaction between the SRP and the SRP receptor [ Miller94 , Peluso00 , Jagath01 ], and the GTP hydrolysis cycle [ Miller94 , Powers95 , Moser97 , Jagath98 , Peluso01 ] have been examined in detail. The 4.5S RNA catalyses the interaction of Ffh and the SRP receptor only when signal sequence is bound [ Bradshaw09 ].

Independently of Ffh, 4.5S RNA binds to elongation factor G (FusA) and to ribosomes [ Brown87 , Nakamura99 , Suzuma99 , Ishibashi99 , RinkeAppel02 ], and 4.5S RNA is proposed to modulate (specifically to release [ Ishibashi99 ]) the interaction of elongation factor G with the ribosome [ Nakamura99 , Ishibashi99 ].

RNase P processing of the precursor of 4.5S RNA has been examined [ Bothwell76 , GuerrierTa83 , GuerrierTa84 , Liu94b , Cole99 ]. Processing by other RNases has been observed [ Li98 ]. The 4.5S RNA structure has been studied in detail [ Bourgaize84 , Struck90 , Lentzen96 , Jovine00 , Jovine00a , Deng03 ].

Any mutant lacking 4.5S RNA is inviable [ Brown84 ]. Depletion of 4.5S RNA leads to induction of the heat shock response [ Bourgaize90 , Poritz90 ], lambda phage induction [ Bourgaize90 ], defects in translation [ Brown84 , Bourgaize87 , Jensen94 ], buildup of pre-beta-lactamase [ Ribes90 ], decreased abundance of maltose binding protein [ Ribes90 ], and cell inviability [ Brown84 ]. Mutants exhibit defects in membrane protein assembly [ Tian00 , Tian02 ]. The 4.5S RNAdl1 allele is a dominant mutation that causes induction of the heat shock response, defects in translocation of a subset of proteins, aberrant cell envelope morphology, and eventual translation defects and cell inviability [ Poritz90 ]. Overexpression of 4.5S RNA suppresses the heat sensitivity of a ffh-10(Ts) mutant [ Park02 ]. The inviability of a 4.5S RNA mutant is suppressed by fusA mutations [ Brown87 ]. The phenotypes of a 4.5S RNA mutant are partially suppressed by mutations in 16S or 23S rRNA [ Brunelli02 ].

The viability of a 4.5S RNA mutant is functionally complemented by expression of RNAs from various bacterial species [ Brown89 , Struck90 ], including Haemophilus influenzae [ Jenkins98 ], by archaebacterial [ Brown91 ] or human SRP7S RNA [ Ribes90 , Brown91 ], by a Mycoplasma pneumoniae RNA [ Simoneau92 ]. The viability of a Bacillus subtilis scRNA mutant is functionally complemented by expression of E. coli 4.5S RNA [ Nakamura92 ], but the B. subtilis sporulation defect is not [ Nakamura95 ]. Exogenous SRP54 protein exhibits binding to E. coli 4.5S RNA [ Romisch90 , Zopf90 ].

Regulation has been described [ Dong96 ]

Front 195

degP

Back 195

serine protease Do
Protease Do, or DegP, is a periplasmic serine protease required for survival at high temperatures [ Lipinska89 , Strauch89a , Seol91a ]. DegP degrades abnormal proteins in the periplasm, including mutant proteins, oxidatively damaged proteins and aggregated proteins [ Strauch88 , Strauch89a , SkorkoGlon99 , Laskowska96 ]. DegP has been specifically shown to degrade the mutant periplasmic protein MalS, as well as unassembled subunits from protein complexes, including HflK, LamB and PapA [ Spiess99 , Kihara98 , Misra91 , Jones02 ].

DegP also proteolyzes a range of other proteins that may not be quality control substrates, such as the DNA methyltransferare Ada, various forms of the colicin A lysis protein and the replication initiation inhibitor IciA [ Lee90a , Cavard89 , Cavard95 , Yoo93a ]. DegP also binds to the ssrA-encoded degradation tag, though this PDZ-domain-mediated interaction does not appear to allow DegP proteolysis of tagged proteins [ Spiers02 ]. Finally, strains lacking DegP are more susceptible to the cationic antimicrobial peptide Lactoferricin B, indicating a possible role for DegP in degradation of that molecule [ Ulvatne02 ].

DegP also has an independent chaperone activity that functions even in proteolytically inactive mutants of DegP [ Spiess99 ]. This chaperone activity is required for survival in the case of disrupted outer membrane assembly, preventing buildup of toxic aggregates [ Misra00 ]. There may be some redundancy between DegP and the chaperones Skp and SurA [ Rizzitello01 ].

DegP is a six-membered ring-shaped structure with a central cavity which contains its proteolytic sites [ Swamy83 , Kim99a ]. The hexamer is built from a pair of staggered trimeric rings, with the proteolytic cavity accessible from the sides rather than the ends [ Krojer02 ]. There are two PDZ domains in each monomer which are required for this assembly, and which may be involved in opening and closing the lateral openings [ Sassoon99 ]. Binding of substrate to the PDZ1 domain induces oligomer conversion from a resting hexameric state to a higher order active complex [ Krojer10 , Merdanovic10 ]. The PDZ1 domain anchors substrate, facilitating its presentation to the proteolytic domain [ Krojer08 ]. DegP is a processive protease - cleaving its substrate into peptides with a mean size of 13-15 residues [ Krojer08 ]. The PDZ1 domain is required for protease activity and for binding of unfolded proteins, while the PDZ2 domain is primarily required for maintaining a hexameric configuration [ Iwanczyk07 ]. The inner cavity also has several hydrophobic patches, which may be involved in its chaperone function [ Krojer02 ].

Hexameric DegP assembles into large catalytically active spherical structures around its substrate [ Krojer08a , Jiang08c ]. The spherical multimers exhibit proteolytic and chaperone-like activity [ Shen09 ]. A model polypeptide substrate binds each DegP subunit at two sites in the crystal structure of a DegP dodecamer [ Kim11b ]. Substrate binding drives the formation of proteolytically active dodecamers and larger cages of 18, 24 and 30 subunits while substrate cleavage promotes cage disassembly [ Kim11b ].

DegP's proteolytic activity is increased at high temperatures but drops dramatically at low temperatures, leaving its chaperone function unaffected [ SkorkoGlon95 , Spiess99 ]. DegP interacts with phosphatidylglycerol on the periplasmic face of the inner membrane, undergoing a conformational change that correlates with the temperature dependence of its proteolytic capacity [ SkorkoGlon97 ].

The mature form of DegP is derived by cleavage of its first twenty-six amino acids by leader peptidase [ Lipinska90 , Lipinska89 ]. Targeting of DegP to the Sec-translocase for transport across the inner membrane is SecB-dependent [ Baars06 ].

DegP is a member of the HtrA (high temperature requirement) family of proteases which combine a protease domain with one or more PDZ domains and function as higher order oligomers [ Kim05e ].

DegP is downregulated during low osmolarity [ Forns05 ]

Front 196

ftsZ

Back 196

Assembly of FtsZ into a ring structure (the Z ring, [ Bi91 ]) at the future cell division site is the earliest known event in cell division. FtsZ is the most highly conserved of the proteins that eventually comprise the septal ring structure; homologs of FtsZ are nearly universally present in bacteria as well as in many archaea, some chloroplasts and a few mitochondria [ Vaughan04 ].

FtsZ is essential [ Dai91 ]; it binds GTP and has Mg2+-dependent GTPase activity [ RayChaudhu92 , deBoer92 ]. Assembly of FtsZ into protein filaments is dynamic and regulated by GTP hydrolysis [ Mukherjee94 ], resembling tubulin [ Mukherjee98 ]. Turnover of FtsZ within the Z ring is extremely rapid, with the rate-limiting step for turnover likely to be GTP hydrolysis [ Anderson04 , Romberg04 ].

The position of the FtsZ ring structure marks the cell division site and is serving as the assembly point to which other proteins of the cell division machinery are recruited. Based on studies of mutants, FtsA and ZipA initially associate with FtsZ, followed by addition of FtsEX, FtsK, FtsQ, FtsL/FtsB, FtsW, FtsI, FtsN and AmiC, apparently in this defined order.

The question of how FtsZ itself is positioned precisely at mid-cell, enabling a normal cell division event, has not been definitively answered yet. Both nucleoid occlusion and the Min system appear to play a role in division site selection; the hypothesis has been discussed in detail [ Norris04 ]. Nucleoid occlusion describes a process by which the presence of the nucleoid inhibits Z ring formation at that site. When the nucleoid structure is perturbed by a block in transcription, nucleoid occlusion is affected; thus, the process may depend on the specific organization of the nucleoid [ Sun04 ]. Both SulA and MinC are negative regulators of Z ring assembly [ Bi93 ]. SulA interacts directly with FtsZ [ Higashitan95 , Cordell03 ] and inhibits GTPase activity and polymerization in vitro [ Mukherjee98a , Trusca98 ] and FtsZ ring formation in vivo [ Justice00 ]. MinC also interacts with FtsZ directly and prevents FtsZ polymerization without inhibiting its GTPase activity [ Hu99b , Pichoff01 ]. The MinCD proteins oscillate between the two cell poles; this behavior may allow Z ring formation at mid-cell because the time-integrated concentration of MinCD is lowest at mid-cell. A conflicting report asserts that MinCD does not block Z ring formation, but instead blocks FtsA association with the Z ring [ Justice00 ].

In studies using GFP-labeled FtsZ, FtsZ outside of the Z ring was found to move rapidly in a helix-like pattern along the cell, similar to the movements of the Min proteins. The presence of a dynamic, helical cytoskeleton was proposed [ Thanedar04 ]. Supporting this hypothesis, FtsZ was found to be involved in maintaining cell shape [ Varma04 ].

The domain structure of FtsZ has been described [ Vaughan04 ]. A highly conserved central domain is structurally and functionally homologous to tubulin; it contains the dimerization domain [ Di99 ]. The C-terminal core domain consists of 12 amino acids essential for FtsA and ZipA binding [ Ma99 ].

The crystal structure of a C-terminal fragment of FtsZ in complex with ZipA has been solved [ Mosyak00 ].

Compounds with activity against E. coli FtsZ, with potential utility as broad-spectrum antimicrobials, have been recently isolated and characterized [ Margalit04 ].

Regulation of FtsZ expression and activity is complex and has been summarized in [ Addinall02 ]

Front 197

lon

Back 197

Lon is an ATP-dependent protease responsible for degradation of misfolded proteins as well as a number of rapidly degraded regulatory proteins. Key regulatory proteins that are Lon substrates include the cell division regulator SulA [ Schoemaker84 , Higashitan97 ], the capsule synthesis regulator RcsA [ TorresCaba87 ] and possibly TER components involved in blocking septation sites during the SOS response [ Dopazo87 ]. Lon is required for degradation of misfolded proteins and the prevention of aggregate formation [ Chung81 , Ryzhavskai , Laskowska96 ]. In the absence of Lon function, aggregation triples [ Rosen02 ]. At least some of this degradation of misfolded proteins depends on the chaperone DnaK [ Sherman92a ].

Lon also degrades the lamba phage N and Xis proteins, with degradation of the latter promoting lysogeny over lysis [ Maurizi87 , Leffers98 ]. Other substrates included HU1 in the absence of its partner HU2, HemA and DAM methylase [ Bonnefoy89 , Wang99d , Calmann03 ].

Lon degrades the antitoxin protein in many toxin/antitoxin protein pairs, including both plasmid and chromosomal versions. Lon proteolysis of the antitoxin protein in plasmid-encoded pairs is required for plasmid maintenance, as the antitoxin has a shorter half life in lon+ cells than the toxin, thus requiring the continued presence of the plasmid for cell survival. Plasmid-encoded antitoxin substrates include CcdA from F plasmid, relBP307 and PasA [ Van96a , Van94a , Gronlund99 , Smith98b ]. Lon proteolyzes chromosomal toxin/antoxin pairs as well, including RelB and YoeB [ Christense01 , Christense04 ]. This degradation of chromosomal pairs may regulate part of the starvation stress response, as the breakdown of RelB leaves RelE, which suppresses translation [ Christense01 ]. Starvation-induced transcription of chpA also depends on Lon [ Christense03 ].

Lon is an ATP-dependent protease with chymotrypsin-like specificity based on a Serine (679)-Lysine (722) dyad [ Charette81 , Waxman85 , Botos04 , Nishii05 ]. Lon has one proteolytic and four ATP-binding sites, two high affinity, the other two low affinity [ Chin88 , Menon87 ]. Detailed kinetic analysis shows that the two types of ATP sites use ATP at different rates as well [ Vineyard06 ]. Lon's protease function depends on its ATPase activity; both require Mg2+, two ATP molecules are used per peptide bond hydrolyzed and loss of ATPase functions leads to concomitant loss of peptidase function [ Menon87 , Menon87a , Fischer94 , Waxman82 ]. ATPase activity continues in mutants that are unable to proteolyze [ Pohl76 ]. Though its ATPase activity is required for protein degradation, Lon is capable of breaking down small peptides in the absence of ATP or ATPase function [ Goldberg85 , Rasulova98 ]. The isolated ATPase domain undergoes conformational change in response to ADP and ATP binding [ Vasilyeva02 ].

Lon has an independent protein-binding domain in addition to its proteolytic domain. This domain binds unfolded proteins [ Chin88 ]. Protein binding substantially stimulates peptide degradation and ATPase activity, the latter even in mutants incapable of peptidase function [ Waxman86 , Pohl76 ].

Lon binds DNA via its DNA-binding domain [ Charette81 , Chin88 ]. Addition of DNA, especially denatured DNA, stimulates substrate proteolysis in vitro, as well as stimulating ATPase activity even in the absence of substrate [ Chung82 , Charette84 ]. Lon may have specificity for promoter regions, explaining how it targets regulatory proteins [ Fu97 ].

Lon can form a complex with inorganic polyphosphate, allowing subsequent degradation of ribosomal proteins, including S2, L9 and L13 [ Kuroda01 ]. Lon's DNA-binding domain binds polyphosphate with greater affinity than DNA [ Nomura04 ]. Lon complexed with polyphosphate may be an octamer instead of a tetramer [ Nishii05 ]

Front 198

dsbA

Back 198

protein disulfide oxidoreductase
The dsbA gene codes for a protein that is a disulfide catalyst. The protein itself has a disulfide bond that is transferred catalytically to folding proteins in the periplasm. The disulfide oxidoreductase is capable of oxidizing proteins very rapidly. The oxidoreductase requires the dsbB protein for reoxidation [ Bardwell94 , Grauschopf95 , Kishigami95 , Guilhot95 , Akiyama92a , Metheringh95 ].

Oxidative folding is thought to occur via formation of intermediate mixed disulfide complexes between DsbA and its substrates. A DsbA mutant has been identified which slows down the resolution of DsbA/substrate compexes and allows characterisation of these intermediates [ Kadokura04 ]. Disulfide linked complexes formed between DsbA and newly synthesized PhoA have been identified [ Kadokura09 ].

The crystal structure of DsbA in complex with a peptide residue from the Shigella flexneri SigA autotransporter protein has been determined at 1.9Å resolution [ Paxman09 ]

Front 199

ompH, skp, hlpA

Back 199

Skp is a periplasmic protein chaperone that binds to unfolded outer membrane proteins. There are at least two pathways for periplasmic chaperone activity, one involving Skp and DegP, and another involving SurA. The SurA pathway is the primary pathway for assembly of OMPs, while the DegP/Skp pathway is important for rescuing proteins that fall off of the SurA pathway.

Skp prevents folding of OmpA in solution [ Bulieris03 ]. Skp and LPS improve insertion of OmpA into phospholipid bilayers [ Bulieris03 ]. Skp interacts with OmpA and PhoE at 1:1 ratios as they cross the inner membrane, and the resulting complexes can be identified from the membrane fraction [ Schafer99 , Harms01 ].

A skp or skp degP mutation causes induction of the unfolded protein-mediated σE stress response [ Missiakas96 , Sklar07a ]. A skp mutant exhibits decreased abundance of outer membrane proteins, compared to wild type [ Chen96a ]. A skp null mutant exhibits mild membrane defects and is unable to efficiently release OmpA from the inner membrane [ Schafer99 ]. Skp suppresses the protein export defect of a secA mutant in vitro with respect to membrane incorporation of LamB and OmpA [ Thome90 ]. A skp degP double mutant exhibits heat sensitivity and periplasmic buildup of denatured proteins [ Schafer99 ]. A skp surA double mutant exhibits a growth defect with formation of filaments [ Rizzitello01 ], accumulates unfolded OmpA, and prevents proper formation of LamB trimers in the outer membrane [ Sklar07a ]. A degP surA double mutation is lethal [ Rizzitello01 ]. A surA mutant forms LamB trimers more slowly than a skp degP mutant [ Sklar07a ]. Depletion of SurA reduced OMPs in the outer membrane, but a skp degP double mutant did not have significantly reduced amounts of OMPs in the outer membrane [ Sklar07a ]. Overproduction of Skp increases production of some recombinant proteins [ Bothmann98 , Strachan99 , Mavrangelo01 , Levy01 , Lin08c ] or proteins produced for phage display [ Bothmann98 , Hayhurst99 ]. Expression of skp also reduces extracytoplasmic stress associated with expression of recombinant outer membrane proteins [ Narayanan ].

Crystal structures of Skp have been solved [ Walton04 , Korndorfer04 , Schlapschy04 ]. Skp is periplasmic and soluble [ Thome91 , Chen96a ] and forms stable homo-trimers [ Schlapschy04 ]. Wild-type Skp translocation to the periplasm requires the SecA [ Thome91 , Ernst94 ] and SecY [ Thome91 ] proteins, but not SecB [ Ernst94 ]. Wild-type Skp translocation also requires ATP [ Thome91 ] and the proton gradient [ Thome91 , Ernst94 ]. Skp is subject to post-translational processing [ Holck88 , Thome91 ]. A 20-residue N-terminal leader is removed to generate the mature species [ Holck88 ]. Skp exhibits interactions with OmpA, OmpC, OmpF, LamB, PhoE OmpG, and YaeT outer membrane proteins at a ratio of one Skp trimer per OMP protein [ Chen96a , De99f , Harms01 , Qu07 ]. Proteomic analyses have identified 30 other interacting proteins, especially from the outer membrane, among these FadL and BtuB, and from the periplasm, MalE and OppA [ Jarchow08 ]. Skp shows a conformational change to protease resistance upon interaction with phospholipids in the presence of divalent cations by inserting into the membrane, and this change is inhibited in the presence of divalent cations and lipopolysaccharides [ De99f ]. Nuclear magnetic resonance of Skp with bound OmpA shows that the OmpA βbarrel is maintained in an unfolded state within the Skp cavity whereas the folded periplasmic domain of OmpA protrudes outside of the cavity [ Walton09 ]. The interaction of Skp with OmpA has been studied using site directed fluorescence spectroscopy [ Qu09 ].

Expression of skp is regulated by σE as well as by the CpxAR two-component response regulator [ Dartigalon01 , Rhodius05 ].

There was some confusion among the neighboring genes skp and lpdX/firA in early studies [ Dicker91 , Aasland88 , Thome90 ].

HlpA: "histone-like protein" / HLP-I: "histone-like protein I" / Skp: "seventeen kDa protein"

Front 200

Signal Recognition Particle Protein Translocation System

Back 200

The Ffh/4.5S RNA/FtsY complex is believed to function in Escherichia coli in the targeting and integration of inner membrane proteins [ Tian02 ]. Ffh is one of three E.coli translocation factors that are homologs of subunits of the eukaryote signal recognition particle (SRP), a soluble ribonucleoprotein complex comprised of six polypeptides and a 300 nucleotide RNA that work together to target proteins, using the energy of GTP hydrolysis, to the secretory pathway.

Based on sequence similarity, Ffh is homologous to SRP54 which is a subunit of SRP involved in preprotein signal-sequence binding. The 4.5S RNA (ffs gene product) is homologous to 7S RNA of SRP RNA which in eukaryotes targets the nascent peptide-ribosome complex to the ER membrane [ Poritz90 ]. One other SRP- associated E.coli protein, FtsY, is homologous to the Sr(alpha) subunit of the SRP receptor which in eukaryotes is the ER membrane receptor for SRP [ Bernstein89a ]. Several studies have supported the hypothesis that these three particles represent a diminished version of the eukaryote SRP system to target proteins to the membrane. Antiserum to Ffh also precipitates 4.5S RNA from E.coli extracts, suggesting that Ffh and 4.5S RNA are able to form a ribonucleoprotein particle similar to the mammalian SRP [ Luirink92 ]. Finally, the Ffh/4.5S RNA particle, in conjunction with FtsY, can effectively substitute for SRP and SR(alpha) in a mammalian in vitro translocation system [ Powers97 ] strongly suggesting that Ffh/4.5S RNA and FtsY can perform membrane targeting functions in E.coli.

The 4.5S RNA and Ffh together form the signal recognition particle (SRP) [ Ribes90 , Poritz90 , Luirink92 ], which binds to the signal sequence [ Luirink92 ] and targets the nascent protein to the SRP receptor, FtsY, [ Miller94 ]. The SRP is involved in integration of nascent proteins into the membrane [ Macfarlane95 , deGier96 , deGier98 , Koch99 , Tian00 ]. The interaction between 4.5S RNA and Ffh [ Wood92 , Lentzen94 , Lentzen96 , Zheng97 , Suzuma99 , Buskiewicz09 ], the GTP-dependent [ Miller94 ] interaction between the SRP and the SRP receptor [ Miller94 , Peluso00 , Jagath01 , Bradshaw09 ], and the GTP hydrolysis cycle [ Miller94 , Powers95 , Moser97 , Jagath98 , Peluso01 ] have been examined in detail. Incorrect cargos are discriminated against at each step of the SRP pathway ensuring fidelity of the system despite the degenerate nature of signal sequences [ Zhang10e ].

Deletion studies of ffh, ffs and ftsY show that deletion of any of the three genes causes defects in membrane insertion of the membrane proteins AcrB, MalF and FtsQ [ Tian02 ]. Other studies using a synthetic lethal upon overexpression ("Slo" phenotype method) have shown that Ffh/4.5S RNA and FtsY are involved in the integration of a subset of polytopic inner membrane proteins [ Ulbrandt97 ].

The structure of Ffh has been analyzed alone and in a complex with the 4.5 S RNA [ Buskiewicz05 , Buskiewicz05a ]. A crystal structure of the SRP in complex with FtsY has been resolved at 3.9 Å [ Ataide11 ].

Oxidized Ffh is unable to bind 4.5 S RNA and requires the activity of methionine sulfoxide reductases to return it to its reduced form [ Ezraty04 ].

FtsY and SecY have been shown to interact in vitro through chemical crosslinking/immunoprecipitation, coimmunoprecipitation, and co-affinity purification studies. This interaction couples signal recognition of membrane proteins to their translocation [ Angelini05 ]. FtsY contains two lipid-binding helices that may contribute to the stability of the FtsY-membrane interaction [ Braig09 ].

An experimental approach using alkaline phosphatase (PhoA) fusions to protein signal sequences has allowed discrimination between the major modes of transport, including the SRP translocation system, across the inner membrane [ Marrichi08 ]

Front 201

eno

Back 201

Enolase catalyzes the interconversion of 2-phosphoglycerate and phosphoenolpyruvate. It is also a component of the degradosome, a complex that degrades RNA.

Enolase dimerizes, with two active sites per dimer. It is dependent on Mg ion for its structure and denatures in the absence of Mg.[ Spring71 ] The crystal structure of enolase was initially determined to a resolution of 2.5 Å, revealing that its dimer interface is enriched in charged residues relative to typical protein-protein interfaces [ Kuhnel01 ]. A more recent 1.6 Å structure shows enolase bound to its recognition site on RNase E [ Chandran06 ].

Enolase is functionally similar to enolases in other organisms, notably in its dependence on Mg, inhibition by fluoride ion in the presence of phosphate and in its catalytic parameters. Its pH optimum is significantly higher than vertebrate enolases and is somewhat above those of yeast and plant enolases. [ Spring71 ]

Enolase is required for the rapid, degradosome-mediated degradation of ptsG mRNA in response to high levels of glucose 6-P or fructose 6-P [ Morita04 ]

Front 202

efeB = ycdB

Back 202

EfeB is a heme-containing component of the cryptic EfeUOB ferrous iron transporter.

Due to the lack of an outer membrane receptor for heme, E. coli K-12 is unable to utilize heme as a source of iron [ Sasarman68 , McConville79 ]. However, expression of the heme receptor protein HasR from Serratia marcescens enables utilization of heme as a source of iron. This requires the presence of either the Dpp dipeptide ABC transporter or of the EfeUO transporter, which is not functional in E. coli K-12 due to a frameshift mutation disrupting efeU. Similar to YfeX, the EfeB protein, although it is normally transported to the periplasm, is able to catalyze the cytoplasmic release of iron from hemin without destroying the tetrapyrrol ring [ Letoffe09 ].

EfeB is a dimeric, non-covalently bound heme-containing peroxidase enzyme of the DyP-type peroxidase family, and is a substrate of the twin arginine translocation (Tat) system [ Sturm06 ]. The Tat-system allows EfeB to assemble in the cytoplasm prior to transport [ Sturm06 ]. The heme cofactor has been identified as FeIII-protoporphyrin IX [ Sturm06 ].

EfeB may act to reduce Fe3+ to Fe2+ for transport or to oxidize Fe2+ during transport [ Cao07 ]. EfeB exhibits guaiacol peroxidase activity at an optimum pH of about 4.0 and may function under acid-stress conditions [ Sturm06 ].

The crystal structure of EfeB has been determined in the apo form to a resolution of 2.0 Å; crystals with bound haem were obtained and diffract to 2.9 Å [ Cartron07 ].

A yfeX efeB double mutant in a HasR-expressing strain is unable to use heme as an external source of iron [ Letoffe09 ].

Front 203

yggE

Back 203

Expression of yggE is upregulated in a superoxide dismutase-deficient strain that is exposed to oxidative stress [ Kim05 ] and by UVA irradiation and elevated temperature [ Ojima09 ]. In microarray experiments, yggE expression was found to be σS-dependent [ Lacour04 ], upregulated by AI-2 quorum signaling [ DeLisa01 ], and belonging to the Rcs regulon [ Hagiwara03 ].

Overexpression of yggE from a plasmid increases the maximum specific growth rate in paraquat-containing medium and decreases ROS (reactive oxygen species) levels [ Kim05 ].

Proteomic analysis of membrane preparations suggests that YggE forms a homo-oligomeric complex [ Maddalo11 ]. The C-terminus of YggE is in the periplasm and thus YggE may be a periplasmic protein that associates with the inner membrane [ Maddalo11 ].

Front 204

dppA

Back 204

DppA is the periplasmic binding component of the dipeptide ABC transporter.

Front 205

dipeptide ABC transporter

Back 205

The DppABCDF dipeptide transport system is a member of the ATP-Binding Cassette (ABC) Superfamily of transporters [ Wu95 ]. Based on sequence similarity, DppA is the substrate-binding component, while DppB and DppC are the membrane components, and DppD and DppF are the ATP-binding components of the ABC transporter. DppABCDF is similar in sequence and subunit composition to the oligopeptide uptake system OppABCDF, suggesting similar subunit functions.

DppA's unbound structure has been resolved by x-ray crystallography to resolutions of 3.2 Å [ Dunten93 ] and 2.0 Å, and shows two domains connected by two 'hinge' segments [ Nickitenko95 ]. The structure of DppA has also been determined with bound glycyl-L-leucine has been determined to a resolution of 3.2 Å [ Dunten95 ]. The structure reveals that the binding site recognizes the peptide backbone allowing for accommodation of various side chains [ Dunten95 ]. There is also a requirement for an unsubstituted α-amino group for transport of a peptide [ Gilvarg65 ].

Loss of DppA or DppBCDF resulted in pro mutants being unable to utilize Pro-Gly as a proline source [ Olson91 ]. Pro-Gly transport was inhibited by His-Glu, suggesting His-Glu is an additional substrate for DppABCDF [ Olson91 ]. Mutations in dpp displayed resistance to the toxic dipeptide Lys-aminoxyAla, the loss of ability to utilize Leu-Trp as a source of its required amino acids [ Payne84 ], resistance to Gly-Val, Leu-Val, and Val-Leu, and reduced uptake of Gly-Gly [ De73 ]. Substrate specificity of DppA was studied in a filter binding assay in which column fractions were monitored for binding activity towards radioactively labeled dipeptides and tripeptides. DppA was observed to mediate the ATP driven uptake of dipeptides and, to a lesser extent, tripeptides from the periplasm [ Smith99 ]. When an outer membrane heme receptor is expressed in E. coli, the dipeptide ABC transporter is also capable of transporting heme and the heme precursor, δ aminolevulinic acid, from the periplasm into the cytoplasm [ Letoffe06 ]. Binding of heme to purified DppA has been demonstrated [ Letoffe06 ].

DppA accumulates to high levels when grown in minimal media, but levels of DppA are reduced when the medium is supplemented with casamino acids [ Olson91 ]. DppA levels were decreased after 4 hours exposure to zinc stress [ Easton06 ] and in response to glucose limitation [ Wick01 ]. When grown in rich medium, gcvB deletion mutants had high constitutive expression of dppA compared to the parent strain [ Urbanowski00 ]. dppA expression is also repressed by PhoB during phosphate limitation [ Smith92 ].

Front 206

rbbA

Back 206

RbbA is associated with the 30S subunit of the ribosome and has ribosome-dependent ATPase activity [ Kiel99 , Kiel01 ]. Stimulation of protein synthesis by RbbA is dependent on ATP hydrolysis, and the ATPase activity is inhibited by hygromycin [ Kiel01 , Ganoza01 ]. The effect of hygromycin may be due to its ability to release RbbA from the ribosome [ Ganoza01 ].

RbbA specifically interacts with EF-Tu and crosslinks to the 30S ribosomal subunit protein S1 [ Kiel01 ]. The E-site base A937 of 16S rRNA is protected by RbbA in the intact 70S ribosome [ Xu06 ].

RbbA contains two ATP-binding domains in its N terminus [ Kiel99 ]. The ATP-binding domains together with five predicted transmembrane helices led to the initial functional prediction as the ATP-binding component of an ABC transporter [ Saurin99 ]. A truncated form of RbbA which does not contain the predicted transmembrane domains still stimulates protein synthesis and has ribosome-dependent ATPase activity [ Xu06 ].

RbbA cross-reacts with antibody raised against fungal elongation factor EF-3 [ Kiel99 ].

Front 207

dnaK

Back 207

chaperone Hsp70; autoregulated heat shock proteins
Hsc56 exhibits specificity toward Hsc62, as Hsc56 does not activate DnaK or Hsc66 ATPase activity [ Yoshimune02 ].

Oleanolic acid caused a slight increase in DnaK synthesis, suggesting a moderate ability for inducing the heat shock response [ Grudniak11 ].

Front 208

dnaJ

Back 208

chaperone with DnaK; heat shock protein

Front 209

ftsA

Back 209

essential cell division protein FtsA
FtsA is an essential cell division protein which colocalizes with FtsZ to the septal ring structure; localization is FtsZ-dependent [ Ma96a , Addinall96 ]. Both FtsA and ZipA are required for the recruitment of FtsEX [ Schmidt04 ] and FtsK [ Wang98f , Di03a ], and thus other cell division proteins downstream, to the Z ring.
The FtsA protein belongs to the actin family of ATPases [ Kabsch95 ] and is associated with the cytoplasmic side of the inner membrane [ Pla90a ]. It is well conserved in bacteria, but is conspicuously absent in certain groups such as the mycobacteria, streptomycetes, and cyanobacteria.
Overproduction of FtsA causes inhibition of cell septation [ Wang90b ]. To enable cell division, FtsA and FtsZ have to be present at a ratio of about 1:100 in the cell [ Dai92 , Dewar92 ], although this ratio may be closer to 1:10 [ Rueda03 ]. FtsA, together with ZipA, supports stabilization of the Z ring [ Pichoff02 ]. A point mutation in FtsA (named FtsA*) can bypass the requirement for the ZipA protein in cell division [ Geissler03 ].
FtsA interacts directly with the C-terminal core domain of FtsZ [ Wang97b , Ma99 ]. The extreme C terminus of FtsA is required for biological activity, localization to the Z ring, and for normal self-interaction [ Yim00 ]. A conserved C-terminal amphipathic helix is required for targetting of FtsA to the membrane and the Z ring [ Pichoff05 ]. Subdomain 1C is essential for interaction with other FtsA molecules [ Carettoni03 ] and sufficient for recruitment of FtsQ, FtsN and FtsI, but is not required for localization of FtsA to the Z ring [ Rico04 , Corbin04 ]. If this subdomain is independently targeted to the Z ring, it can partially suppress a ftsAts mutation; it may therefore constitute an independent domain involved in the recruitment of other septation proteins [ Corbin04 ].
FtsA: "filamentous temperature sensitive" [ Ricard73 ]

Front 210

mukB

Back 210

cell division protein involved in chromosome partitioning
The MukB protein is involved in chromosome condensation and partitioning. It belongs to the family of SMC (Structural Maintenance of Chromosomes) proteins [ Soppa01 ]. MukB acts as a macromolecular clamp that compacts DNA; condensation is cooperative, and ATP stimulates its initiation, but not propagation [ Cui08 ]. In single-molecule observations, both MukB alone and the MukBEF complex promote shrinkage of large DNA molecules in the presence of ATP [ Chen08d ].
GFP-marked MukB protein is observed to occupy the same location as the nucleoid in the presence of MukE and MukF [ Ohsumi01 ], and fluorescent antibodies also show localization to the nucleoid, but no colocalization with the FtsZ ring [ denBlaauwe01 ]. In vivo, MukEF is required for association of MukB with the nucleoid [ She07 ]. Mutations in the MukF linker region affect localization of MukB [ Shin09b ]. MukB generally colocalizes with the origin of replication (oriC) throughout the cell cylce [ Danilova07 ], although the number of MukB foci is larger than the number of oriC foci [ Adachi08a ]. The oriC region is aberrantly positioned at the cell pole in a mukB mutant [ Danilova07 ].
The MukB protein forms a homodimer [ Niki92 ]. Like other members of the SMC protein family, MukB is a large protein made up of five domains: N- and C-terminal globular "head domains", a central linker region or "hinge domain", and two coiled-coil rod domains that separate the head domains from the hinge domain. This structure can be visualized by EM; the coiled-coil regions are arranged in an antiparallel fashion in the homodimer [ Melby98 , Matoba05 ]. The relative alignment of the N- and C-terminal halves of the coiled-coil region has been investigated by site-directed crosslinking [ Li09c ].
The C-terminal region of MukB is essential for its DNA binding activity [ Niki92 , Saleh96 ], while the hinge domain does not appear to interact with DNA [ Ku10 ]. ATP is not required for DNA binding [ Petrushenk06 ]. A purified N-terminal domain of MukB binds to FtsZ in vitro [ Lockhart98 ]. A crystal structure of the 227 N-terminal amino acid residues of MukB has been solved at 2.2 Å resolution [ vandenEnt99 ]. Crystal structures of the central hinge domain responsible for dimerization of MukB, including a portion of the adjacent coiled-coil domain, have been solved at 3.1 and 2.3 Å resolution [ Li10b , Ku10 ]. The hinge domain interacts directly with the C-terminal domain of ParC and stimulates the DNA relaxation and decatenation activity of topoisomerase IV in vitro [ Hayama10 , Li10c ].
A mukB null mutant has a cell division defect and can not form colonies [ Niki91 ]. mukB mutants are temperature sensitive and produce anucleate cells. A mukB mutation causes unfolding of the nucleoid; this phenotype is supressed by a mutation in seqA [ Weitao99 , Onogi00 ]. Additional supressors of the mukB null mutant phenotype have been isolated in smbA [ Yamanaka92 ], msmA, msmB, msmC [ Yamanaka94 ], smbB [ Kido96 ], crcA, cspE, crcB [ Hu96a ], topA [ Sawitzke00 ], and gyrB [ Adachi03b ]. In a mukB mutant, the chromosome appears to be less supercoiled [ Weitao00 ]. Overproduction of MukB leads to condensation of chromosomes, even in the absence of MukE and MukF [ Wang06g ].

Front 211

appC

Back 211

cytochrome bd-II terminal oxidase subunit I
The appC encoded protein is subunit 1 of E. coli cytochrome bd-II terminal oxidase.

Front 212

torA

Back 212

trimethylamine N-oxide reductase, catalytic subunit
TorA contains the active site and molybdenum cofactor [ Zhang08e ].

The TorA protein is exported to the periplasm via the twin-arginine transport (TAT) system. A 39 amino acid signal peptide is cleaved [ Mejean94 ].

There is evidence that the catalytic TorA subunit can exist as both a dimer and a monomer. [ Silvestro88 , Silvestro89 ]

Front 213

fdnG

Back 213

formate dehydrogenase N, α subunit
FdnG is the catalytic subunit of formate dehydrogenase-N. It contains the bis-molybdopterin guanine dinucleotide (MGD) cofactor and selenocysteine [ Berg91a ].

FdnG is translocated to the periplasm via the Tat system; interaction with the FdnI subunit localizes the protein to the periplasmic face of the cytoplasmic membrane [ Sargent98a , Stanley02 ].

Production of FdnG is regulated at the translational level. A segment of the mRNA encoding the N terminus of FdnG is able to form a stable stem-loop structure; disruption of the structure by site-directed mutagenesis leads to overproduction of FdnG-β-galactosidase and FdnG-CAT translational fusion proteins; if such mutations are introduced at the chromosomal fdnG locus, formate dehydrogenase-N activity increases [ Punginelli04 ].

Front 214

eno

Back 214

Enolase catalyzes the interconversion of 2-phosphoglycerate and phosphoenolpyruvate. It is also a component of the degradosome, a complex that degrades RNA.

Enolase dimerizes, with two active sites per dimer. It is dependent on Mg ion for its structure and denatures in the absence of Mg.[ Spring71 ] The crystal structure of enolase was initially determined to a resolution of 2.5 Å, revealing that its dimer interface is enriched in charged residues relative to typical protein-protein interfaces [ Kuhnel01 ]. A more recent 1.6 Å structure shows enolase bound to its recognition site on RNase E [ Chandran06 ].

Enolase is functionally similar to enolases in other organisms, notably in its dependence on Mg, inhibition by fluoride ion in the presence of phosphate and in its catalytic parameters. Its pH optimum is significantly higher than vertebrate enolases and is somewhat above those of yeast and plant enolases. [ Spring71 ]

Enolase is required for the rapid, degradosome-mediated degradation of ptsG mRNA in response to high levels of glucose 6-P or fructose 6-P [ Morita04 ].

Front 215

mreB

Back 215

Immunofluorescence microscopy has shown that MreB, an actin homolog, forms helical filaments beneath the surface of the cell [ Kruse03 ]. Filament formation has been shown to be dependent upon the rod-shape of the cell [ Kruse05 ]. MreB is also incorporated into cytoskeletal rings that are located near the midcell during cell division [ Vats07 ].
MreB is responsible for proper chromosome segregation and movement. Overexpression of MreB inhibits cell division. Overexpression of dysfunctional MreB results in altered MreB filament morphology, inhibition of cell division, mislocalized origin and terminus regions of the chromosome, and perturbed DNA segregation [ Kruse03 ].
BolA represses mreB transcription [ Freire09 ].
Overexpression of ftsQAZ suppresses the lethality of MreBCD depletion by increasing the supply of monomers for the enlarged Z ring of round cells during division. Fractionation and GFP fusions studies have shown MreC and MreD associate with the inner membrane. Two-hybrid experiments have shown that MreBCD form a complex in which MreB interacts with itself and MreC, and MreC interacts with itself and MreD.

The coiled-coil domain of MreC is believed to allow it to dimerize while its alpha helices are embedded within the inner membrane. MreD is predicted to be membrane bound with five transmembrane α-helices. The cytoplasmic 13 -14 N-terminal amino acids of MreC are believed to interact with MreB lying just beneath the cell surface. MreC is also believed to interact with PBP2, which is responsible for lateral wall peptidoglycan synthesis, suggesting a role for the MreBCD complex in directing peptidoglycan formation through this interaction [ Kruse05 ]. Immunofluorescence microscopy has shown PBP2 localization in the cell periphery in band-like structures is similar to MreB localization and is dependent upon MreB in Caulobacter crescentus [ Figge04 ]. and E.coli [ Vats09 ]. Immunofluorescence microscopy has also shown that assembly of MreB, MreC and MreD into the cytoskeletal rings and coiled structures occurs independently. [ Vats09 ].
E. coli with a mutation in mreB have a decreased MIC of β-lactams for β-lactamase-producing cells [ Pradel09 ].
Component of: longitudinal peptidoglycan synthesis/chromosome segregation-directing complex

Front 216

secA

Back 216

SecA is an inner membrane component of the Sec Protein secretion system and is peripherally associated with the multi-subunit translocation apparatus SecYEG. Conditional lethal mutation analyses [ Oliver81 ] indicate that secA mutants accumulate several periplasmic proteins (MalE, PhoA, LamB and the ompF gene product ) in the cytosol. Studies [ Brundage90 ] of purified proteins reconstituted into proteoliposomes indicate that SecA, SecB, the translocation complex SecY/E, a precursor protein and ATP are sufficient for the in vitro reconstitution of translocation. In the presence of precursor protein, the energy of ATP binding and hydrolysis drives the membrane insertion of SecA and its bound precursor protein into the cytoplasmic membrane.
Biochemical studies have shown that SecA serves multiple and dynamic functions in the translocation process. Studies using cell fractionation and protease protection indicate that SecA cycles between three different cellular locations: the cytoplasm, the inner-membrane as a peripheral protein, and an integral inner-membrane protein [ Danese98 ]. SecA can function as a cytoplasmic chaperone, similar to SecB, targeting its precursor protein and delivering it to the inner membrane translocation apparatus [ Danese98 ]. SecA also functions as a peripheral membrane protein. It can either associate nonspecifically with the inner leaflet phospholipid heads or, alternatively, bind to the SecEGY tranlocase apparatus in the cytoplasmic membrane. Purification of SecA and reconstitution into inverted membrane vesicles in the presence of ATP and preprotein followed by proteolysis indicates that it is while bound to the SecEGY translocase that SecA becomes protected from proteolysis due to membrane insertion, and thus, it is at this point that translocation of the periplasmic protein precursor occurs [ Eichler97 ], [ Economou95 ].
Sites of interaction between SecA and SecY have been identified which are important in the catalytic activity of SecA [ vanderSlui06 ]. Crosslinking studies have demonstrated that the loop at the tip of the two-helix finger of SecA interacts with a polypeptide chain right at the entrance into the SecY pore. Mutagenesis has indicated that residues within this loop are particularly important for translocation of the protein substrate [ Erlandson08 ]. SecA exists in equilibrium between its monomeric and dimeric forms. Mutation, complementation and chemical cross-linking studies show that the dimeric form of SecA is required for protein translocation [ Jilaveanu05 , Jilaveanu06 , Das08 ]. Cryo-electron microscopy (cryo-EM) has been used to study ligand-free SecA in solution. This revealed that SecA adopts an antiparallel dimeric conformation in solution, with many electrostatic interactions occurring at the dimer interface [ Chen08e ]. When SecA interacts with SecB, a cytosolic chaperone, two protomers of SecA must be bound to SecB for the complex to be active in vitro [ Randall05 ]. The presence of two SecA protomers has been investigated in cytoplasmic membrane vesicles and shown to be necessary to achieve maximal coupling efficiency between ATP hydrolysis and translocation [ Mao09 ].
SecA differs from other helicases in that in addition to the helicase DEAD motor, it contains two flexible specificity domains. One of these domains is a preprotein binding domain (PBD), which is essential for Sec-dependent protein translocation [ Papanikou05 ] and thought to be involved in regulation of ATPase activity. To investigate this further, the structure and properties of SecA without PDB were characterized. This study indicated that removal of the PDB does not cause large structural changes in the DEAD motor and results in a slight elevation of the ATPase activity. In addition, structural analyses allowed construction of a predicted model of the transition state for ATP hydrolysis [ Nithianant08 ].
Binding of a signal sequence to soluble SecA reduces its nucleotide binding capability and induces its oligomerization, enhancing membrane insertion of the SecA-signal peptide complex [ Shin06 ]. Once inserted, SecA regains nucleotide binding capability [ Shin06 ]. Binding and hydrolysis of ATP induces disorder-order folding transitions of the nucleotide binding domain (NBD) which communicates allosterically with other protein domains to drive translocation [ Keramisano06 ]. Upon hydrolysis of ATP, SecA is released, at least partially, from the translocating protein, whose signal sequence has been cleaved by signal peptidase. At this point, in concert with additional translocase components SecD/F/YajC, ATP again binds to the SecA binding site and additional stepwise translocation of the polypeptide resumes [ Duong97 , Erlandson08a ]. Translocation occurs in a stepwise manner with each cycle of ATP binding and hydrolysis driving approximately 50 amino acids across the membrane [ Tomkiewicz06 ]. Preporoteins with stretches of positively charged or negatively charged amino acids inserted in their polypeptide chain are unable to stimulate secA ATPase activity and do not undergo translocation by the Sec translocase in vitro [ Nouwen09 ].
SecA contains distinct binding sites for the signal peptide and the mature domain of its preprotein substrates [ Gouridis09 ]. ProPhoA association to SecA is only marginally reduced if the signal peptide is impaired but substantially reduced if the mature region is truncated [ Prinz96 , Gouridis09 ]. Tight signal peptide binding to SecA promotes 'triggering' of the translocase (that is, lowering the ATPase activation energy) and drives 'trapping' of the mature protein in the translocase so that is irreversibly engaged in the channel [ Gouridis09 ].
The kinetics of the ATP hydrolysis cycle of SecA have been studied in vitro. Basal ATPase activity is low; release of ADP is stimulated by binding of SecA to SecYEG and ADP release is fastest when a substrate protein is being translocated [ Robson09a ].
Site-directed spin labelling in combination with electron paramagnetic resonance (EPR) has been utilized to study regions on the surface of SecA which interact with binding partners during export [ Cooper08 ].
SecA-dependence of protein translocation requires periplasmic loops of at least 30 residues for multiple spanning proteins but not for single spanning proteins suggesting different roles for SecA in translocation of the two protein types [ Deitermann05 ].
The crystal structures of homodimeric SecA in complex with ATP, ADP and adenosine 5'-[beta,gamma-imido]triphosphate (AMP-PNP) have been resolved to 2 Å [ Papanikola07 ].

Front 217

tatE

Back 217

TatE is a subunit of the TatABCE (twin-arginine translocation) complex for the export of folded proteins across the cytoplasmic membrane. TatE shares overlapping functions with TatA and deletion mutation studies [ Sargent98a ] show that deletion of tatE results in a decrease in the range of substrates that the complex can export.

Front 218

TatABCE protein export complex

Back 218

The twin-arginine translocation (Tat) system works in parallel with the E. coli Sec translocation system to transport folded proteins across the cytoplasmic membrane [ Weiner98 ]. While the Sec system transports unfolded proteins, Tat translocase functions to move structured macromolecular substrates, usually containing cofactors, across the cytoplasmic membrane. These proteins include those critical for bacterial respiratory and photosynthetic energy metabolism. Substrates utilizing the Tat pathway are characterized by essentially invariant amino-terminal sequences which contain consecutive arginine residues.

Cofactor-containing Tat substrates acquire their cofactors in the cytoplasm where they attain a folded conformation [ Berks00 ]. In studies using mutated tat strains [ Sargent98a ], precursor proteins that accumulate in the cytoplasm contain cofactors. In even more direct studies using folded green fluorescent protein (GFP) fused with the Tat signal peptide [ Santini01 ], the GFP was found to localize to the periplasm.

The Tat apparatus in E. coli is encoded by genes located in two genetic loci. The tatA operon encodes tatABCD; tatE is coded for in a separate locus [ Bogsch98 ]. TatA, TatB and TatE are similar in structure, predicted to comprise a membrane-spanning alpha helix at the amino terminus, followed by an amphipathic helix at the cytoplasmic side of the membrane and a variable-length carboxy terminus. TatA has been shown to be a fully integral membrane protein [ De01 ]. TatA and TatE share 50% sequence identity and share overlapping functions in Tat translocation. Deletion of either of these genes results in a decrease in the range of substrates, while deletions in both results in complete loss of Tat-dependent export [ Sargent98a ]. Although TatB has 20% sequence identity with TatA/TatE, it serves a distinct function in export. A deletion mutation of tatB alone is enough to completely abolish the translocation of some but not all endogenous Tat substrates [ Ize02 ]. TatC has similarly been shown to be essential for Tat-dependent protein export [ Bogsch98 ]. tatD encodes a soluble cytoplasmic protein with nuclease activity [ Wexler00 ]. Deletion studies of tatD and two of its homologues indicate that TatD family proteins are not essential for Tat-dependent protein translocation.

Functionally, protein translocation in the Tat system is energized exclusively by a transmembrane proton electrochemical gradient with no involvement of nucleotide hydrolysis [ Mould91 ] unlike the Sec translocation system which is powered by ATP hydrolysis.

Localization studies using fusion proteins with green fluorescent protein (GFP) demonstrated that TatA, TatB and TatC proteins all localize to the cellular poles, suggesting that active translocon poles are primarily located at polar positions in E. coli [ Berthelman04 ].

Within the purified Tat complex, TatB and TatC are present in a strict 1:1 ration and a TatBC fusion protein supports Tat dependent transport [ Bolhuis01 ]. Three-dimensional structures of TatBC-substrate complexes and unliganded TatBC have been obtained by single particle electron microscopy. The structures show substrate binding on the periphery of the TatBC complex and suggest that TatBC undergoes structural reorganisation upon substrate binding [ Tarry09 ].

An experimental approach using alkaline phosphatase (PhoA) fusions to protein signal sequences has allowed discrimination between the major modes of transport, including the Tat protein export system, across the inner membrane [ Marrichi08 ].

Front 219

ndh

Back 219

NADH:ubiquinone oxidoreductase II
NADH:ubiquinone oxidoreductase II (NDH-2) is a type IIA NADH dehydrogenase that catalyzes the transfer of electrons from NADH to the quinone pool in the cytoplasmic membrane. It is thus part of the aerobic respiratory chain of the cell, and its primary function may be maintenance of the [NADH]/[NAD+] balance of the cell. NDH-2 is one of two distinct NADH dehydrogenases in E. coli. In contrast to NDH-1 (encoded by the nuo genes), NDH-2 utilizes NADH exclusively, and electron flow from NADH to ubiquinone does not generate an electrochemical gradient [ Matsushita87 , Hayashi89 , Calhoun93 ].

NDH-2 links the major catabolic and energy-producing pathways in the cell [ Jaworowski81 ]. The comparative energy efficiency of utilization of the various components of the aerobic respiratory chain has been examined [ Calhoun93a , Helling02 , Noguchi04 ].

NDH-2 is a strongly membrane-associated protein; it contains an FAD cofactor and copurifies with phospholipids [ Young78 , Jaworowski81a , Campbell83 ]. Phosphatidylethanolamine appears to be required for full enzymatic activity in vivo [ Mileykovsk93 ]. The enzyme contains a thiolate-bound Cu(I) ion, and a putative copper binding site has been identified [ Rapisarda02 ]. Heterooligomers of NDH-1 and NDH-2 have been identified by electrophoresis and sucrose gradient centrifugation suggestive of a supramolecular organisation in the membrane [ Sousa10 ].

NDH-2 can generate superoxide radicals and hydrogen peroxide by autooxidation of the FAD cofactor [ Messner99 ] when the enzyme is overproduced or in the absence of quinones [ Seaver04 ]. However, under regular growth conditions, it is not the primary source of intracellular hydrogen peroxide [ Seaver04 ]. NDH-2 also has cupric reductase activity that is dependend on FAD or quinones. Under experimental conditions, the reductase activity was observed simultaneously with the dehydrogenase function linked to the respiratory chain [ Rapisarda99 ].

A topology model with two C-terminal transmembrane domains has been proposed [ Rapisarda02 ]. Based on the remote similarity of NDH-2 to the SCOP family of FAD/NAD-linked reductases, a structural model of NDH-2 has been proposed [ Schmid04 ].

The originally isolated ndh mutant [ Young76 ] was not a single-locus mutant; it also contained a disruption of the nuo locus, encoding NDH-1 [ Calhoun93 ]. Only strains containing mutations in both NADH dehydrogenases are unable to grow on mannitol as the sole source of carbon [ Calhoun93 ]. Metabolic effects of an ndh null mutation alone and in combination with various other mutations have been investigated [ Yun05 ].

Expression of ndh is repressed during anaerobic growth [ Spiro89 , Green94 ]. In an F1-ATPase-defective mutant, ndh transcription and NDH-2 activity is increased [ Noda06 ].
Ndh: "NADH dehydrogenase" [ Young76 ]

Front 220

hrpA

Back 220

HrpA has sequence similarity to DEAH-box RNA helicases [ Moriya95 ]. HrpA is involved in the post-transcriptional processing of the daa operon mRNA, which encodes proteins involved in fimbrial biogenesis of an enteropathogenic E. coli strain. Activity requires the predicted nucleotide triphosphate binding and hydrolysis functions [ Koo04a ]. The ATPase activity of His-tagged partially purified HrpA has been measured [ Jain06a ].

Front 221

rimL

Back 221

ribosomal-protein-L12-serine acetyltransferase
RimL is the acetyl transferase that acetylates the Nα-terminal serine residue of ribosomal protein L12, converting it into its modified form, L7 [ Isono81 , Tanaka89 , Miao07 ]. Acetylation of L12 does not appear to be essential for its function [ Isono81 ].

RimL can also acetylate a mutant ribosomal protein L12 which contains alanine instead of serine at the N-terminal position (after cleavage of the leading methionine); substitution of the adjacent amino acid with aspartate leads to lower catalytic efficiency [ Miao07 ].

A rimL mutant lacks ribosomal protein L7, the acetylated form of ribosomal protein L12 [ Isono81 ].
RimL shows similarity to RimJ [ Tanaka89 ].

Front 222

rhlA

Back 222

SrmB, DEAD-box RNA helicase
SrmB is a DEAD-box protein with RNA helicase activity that facilitates an early step in the assembly of the 50S subunit of the ribosome [ Charollais03 ]. The SrmB protein was found to be associated with a pre-50S ribosomal particle [ Charollais03 ] and forms a complex with ribosomal proteins L4, L24, and the region of 23S rRNA that interacts with L4 and L24 [ Trubetskoy09 ].

The ATPase and helicase activities of SrmB are stimulated by long single-stranded RNAs [ Bizebard04 , Worrall08a ]. SrmB can directly interact with poly(A) polymerase I [ Raynal99 ] and can stabilize overexpressed mRNAs [ Iost94 ].

Deletion of the srmB gene causes a growth defect at low temperature [ Charollais03 ], while overexpression supresses a temperature sensitive lethal mutation in ribosomal protein L24 [ Nishi88a ]. The C-terminal domain of SrmB is not required for function [ Trubetskoy09 ].

Front 223

secG

Back 223

SecG is an inner membrane protein involved in the Sec protein secretion pathway.

Front 224

degQ

Back 224

serine endoprotease, periplasmicDegQ is a serine endoprotease that may be involved in degrading transiently denatured proteins [ Waller96 , Kolmar96 ].
DegQ may function as a dodecamer (note, however, that the other protein identified as a dodecamer in this paper, DegP, appears to actually operate as a hexamer) [ Kolmar96 ].
DegQ expression increases during anaerobic growth at pH 8.5 [ Yohannes04 ].
DegQ is a multi-copy suppressor of prc deficiency [ Bass96 ].

Front 225

lldD

Back 225

L-lactate dehydrogenase is an FMN-dependent membrane-associated dehydrogenase [ Futai77 , Dong93a ]. It functions in aerobic respiration and also has a role in anaerobic nitrate respiration [ Nishimura83 , Iuchi94 ]. L-lactate dehydrogenase is associated with the inner membrane [ Kung72 , Futai77 , LopezCampi05 ].

LldD is one of three lactate dehydrogenase enzymes which interconvert pyruvate and lactate in E. coli. The other two enzymes are specific for D-lactate: the soluble LdhA , an NAD-linked fermentative enzyme, and Dld , a membrane-associated respiratory enzyme.

L-lactate dehydrogenase is induced by aerobic growth on L-lactate and L-fucose [ Cocks74 , Futai77 , Dong93a ] and in the presence of lactate under nitrate, fumarate and TMAO respiration conditions [ Nishimura83 ]. Repression of lldD expression under anaerobic conditions is mediated by ArcA [ Iuchi94 ]. An lldD insertion mutant has lost the ability to grow on L-lactose as the sole source of carbon and energy, but can still utilize D-lactate [ Dong93a ].

Front 226

tatA

Back 226

TatA is a subunit of the TatABCE (twin-arginine translocation) complex for the export of folded proteins across the cytoplasmic membrane. TatA is fully integrated into the inner membrane [ Gohlke05 ].

TatA shares overlapping functions in Tat translocation. Deletion mutations [ Sargent98a ] of tatA result in a decrease in the range of substrates that the complex is able to export.

TatA is capable of forming homomultimeric tubes in the cytoplasm which can interact together in the presence of TatC. E. coli TatA therefore has the previously unrecognized capacity to extend in three-dimensions [ Berthelman08 ].

Fluorescence microscopy indicates that TatA complexes have a heterogeneous size distribution [ Leake08 ]. Fourier analysis suggests that the complexes are assembled from tetramer units [ Leake08 ]. TatA complexes do not form in cells lacking TatBC [ Leake08 ].

Front 227

tatB

Back 227

TatB is a component of the TatABCE (twin-arginine translocation) complex for the export of folded proteins. TatB has been shown [ Ize02 ] through deletion mutations to be required for the transport of some, but not all, endogenous Tat substrates across the cytoplasmic membrane.

In vitro assays suggest that TatB forms an oligomeric binding site that contacts folded Tat precursors in an intermediate binding step that precedes translocation [ Maurer10 ].

Front 228

tatC

Back 228

TatC is a subunit of the TatABCE (twin-arginine translocation) complex for the export of folded proteins across the cytoplasmic membrane. Although its exact function is not known, it has been shown through deletion mutation studies [ Bogsch98 ] to be essential for Tat-dependent protein export. Membrane topology predictions using experimentally determined C terminus locations indicate that TatC has 6 transmembrane helices and the C-terminus is located in the periplasm [ Rapp04 ],[ Drew02 ]. TatC may form functional dimers [ Maldonado11 ].

Front 229

pepQ

Back 229

PepQ, a proline dipeptidase, hydrolyzes dipeptide substrates containing a proline residue at the carboxy-terminal end. In addition, the enzyme can catalyze the stereoselective hydrolysis of organophosphate di- and triesters [ Park04 ].

Front 230

rimI

Back 230

RimI is an alanine acetyltransferase that is specific for ribosomal protein S18 [ Isono80 , Yoshikawa87 ].
A rimI mutant exhibits a defect in acetylation of the N-terminal alanine of ribosomal protein S18, but shows no defect in acetylation of ribosomal proteins S5 or L12 [ Isono80 ].
RimI and RimJ share C-terminal similarity [ Yoshikawa87 ].

Front 231

rpsT

Back 231

The S20 protein is a component of the 30S subunit of the ribosome and appears to be involved in translation initiation and the association of the 30S and 50S subunits [ Gotz90 ]. S20 was also identified as antizyme 1, an inhibitor of the biosynthetic ornithine and arginine decarboxylases; these enzymes are involved in the biosynthesis of polyamine [ Panagiotid84 ]. S20 is identical to L26.

The synthesis of S20 is regulated at the post-transcriptional level [ Parsons83 , Wirth81 ]; in an in vitro system, S20 represses its own synthesis [ Wirth82 ]. However, S20 doesn't show affinity for its own mRNA [ Donly88 ]. The UUG initiation codon appears to be important for regulation [ Parsons88 ]. Processing and degradation of rpsT mRNA have been studied extensively; see for example [ Baker03a , Mackie00 ] and references therein. rpsT mRNA stability and translational efficiency are linked [ Parsons88 , Rapaport94 ].

S20 physically interacts with ornithine and arginine decarboxylase, and overexpression of S20 decreases the production of polyamine in vivo [ Panagiotid95 ]. Levels of S20 increase in response to polyamines; the effect is thought to be due to an increase in the level of transcription of rpsT [ Huang90 ].

S20 binds to the 5' domain and the 3' minor domain of 16S rRNA [ Marquardt79 , Stern88 , Weitzmann93 , Cormack91 ]. C-terminal residues of S20 are required for binding to 16S rRNA [ Donly88 ]. Hydroxyl radical footprinting provided evidence that S20 is located near the bottom of the 30S subunit [ Culver98 ]. Further footprinting studies provided insight into 30S subunit assembly [ Dutca08 ]. S20 has a local effect on the stability of the folded structure of 16S rRNA [ Ramaswamy09a ].

The sup(S20) mutation in rpsT suppresses defects of an alanyl-tRNA synthetase mutant [ Bock74 , Buckel76b ]. A spontaneous mutation in rpsT causes a temperature-sensitive phenotype and increases the ability of the ribosome to misread nonsense stop codons [ RydenAulin93 ]. An rpsT null mutant shows impaired growth, but is viable [ Bubunenko07 ]. A 30S ribosomal subunit lacking S20 shows a reduced rate of mRNA binding and 70S complex formation [ Tobin10 ].

Front 232

pepD

Back 232

Peptidase D is a dipeptidase capable of breaking down a number of dipeptides with unblocked N termini, including cysteinylglycine [ Schroeder94 , Suzuki01a ]. Peptidase D functions as a dimer of PepD monomers [ Klein86 ].
Transcription of pepD increases five fold during phosphate starvation [ Henrich92 ].
cytosol nonspecific dipeptidase (peptidase D)


a dipeptide + H2O <=> 2 a standard α amino acid

The reaction direction shown, that is, A + B ↔ C + D versus C + D ↔ A + B, is in accordance with the Enzyme Commission system.
The reaction is physiologically favored in the direction shown.

Front 233

pepN

Back 233

Aminopeptidase N is an aminoendopeptidase, capable of breaking down small peptides as well as whole, native proteins [ Lazdunski75 , Chandu03 ].

Aminopeptidase N catalyzes the hydrolysis of a peptide bond either at the end of a small peptide or in the middle of a protein. In its exopeptidase activity, it shows a cleavage preference for basic and small amino acids, in the order arginine > alanine > lysine > glycine [ Chandu03a ]. Other substrates include cysteinylglycine [ Suzuki01a ]. Phosphorylation of peptide side chains near the cut site can prevent hydrolysis of the peptide bond [ Fernandez96 ].

Some results suggest that aminopeptidase N synthesis is independent of growth conditions, though other experiments have demonstrated increased synthesis of this protein during phosphate starvation, anaerobiosis and growth on glycerol and succinate [ McCaman82 , Murgier82 , Gharbi85 ].

Crystal structures of solitary aminopeptidase N and the enzyme complexed with bestatin have been determined to 1.5 and 1.6 Å resolution, respectively [ Ito06 ].

Front 234

rne

Back 234

Ribonuclease E (RNase E) is a single-strand-specific endonuclease that is essential for viability. It processes rRNA, tRNA and other RNAs, is involved in plasmid and phage stability and is part of the degradosome, a multienzyme complex involved in mRNA degradation.
RNase E is involved in processing and cleavage of several rRNAs. It processes the 9S rRNA precursor to yield the mature 5S rRNA by cleaving quite near the 5' end and downstream from the 3' end of the final product [ Ghora78 , Roy83 , Apirion78 ]. RNase E also participates in the 5' maturation of 16S rRNA from its 17S precursor, as well as being able to cleave single-stranded regions within mature 16S and 23S rRNAs [ Li99a , Bessarab98 ].
RNase E initiates the processing of both poly- and monicistronic tRNA transcripts, including those within rRNA transcripts, by cleaving within a few nucleotides of the mature 3' CCA terminus, thus allowing RNase P and other 3' to 5' exonucleases to complete tRNA maturation [ Ow02 , Li02 ]. RNase E similarly cleaves at the 3' CCA terminus of the ssrA precursor to yield its final form [ LinChao99 ]. RNase E may also be involved in processing of the 5' leader of precursor tRNAs [ Soderbom05 ].
RNase E carries out the 3' processing of M1 mRNA, which codes for the catalytic subunit of RNase P [ Lundberg95 , Sim01 ]. Other mRNA processing substrates include the cell division inhibitor DicF, the RNA polymerase sigma70 activity modulator 6S RNA, the polycistronic histidine operon mRNA and the papAB primary transcript, which is cleaved to yield stable papA and unstable papB mRNA [ Faubladier90 , Kim04 , Alifano94 , Nilsson91 ].
The stability of plasmids R1 and Colicin E1 is influenced by RNase E. It initiates degradation of CopA, the R1 copy regulator RNA [ Soderbom98 ]. RNase E also cleaves near the 5' end of the sok component of the hok/sok sense/antisense RNA plasmid stabilization mechanism from R1, allowing subsequent degradation by another degradeosome component, PNPase [ Dam97 ]. The Colicin E1 DNA synthesis inhibitor RNA, RNAI, is also cleaved at its 5' end by RNase E [ Tomcsanyi85 , LinChao91 ]. Finally, RNase E cleaves FinP, which binds to the 5'-untranslated region of the positive F-plasmid transfer regulator traJ [ Jerome99 ].
RNase E processing maintains the balance between phage f1 proteins pII and PX by cleaving the mRNA coding for pII, thus maintaining a normal replication cycle [ Kokoska98 ]. RNase E is required more generally for production of certain phage f1 mRNAs as well [ Stump96 ]. RNase E processes T4 gene 32 mRNA, cleaves T4 soc mRNA and is involved generally in the destabilizing of T4 mRNA [ Mudd88 , Otsuka03 , Mudd90a ].
A number of cellular mRNAs are degraded by RNase E. mRNA decay slows 2-3 fold in an rne mutant, and RNase E is the rate-limiting enzyme in the degradation of many of its substrates [ Babitzke91 , Jain02 ]. RNase E cleaves both sodB mRNA and its antisense RNA RyhB, though cleavage of the latter can be blocked by Hfq binding to the cleavage site [ Afonyushki05 , Masse03 , Moll03 , Folichon03 ]. Hfq also overlaps cleavage sites in the dsrA and ompA mRNAs [ Moll03 ]. The rpsO-pnp transcript is cleaved near the beginning of the rpsO coding sequence and on both sides of the rspO 3' stem-loop terminator, after which it is rapidly degraded by PNPase [ Hajnsdorf99 , Regnier91 , Braun96 , Hajnsdorf94 ]. Ribosome binding blocks this cleavage [ Braun98a ]. RNase E is responsible for a number of cleavages within the unc transcripts, which code for subunits of the F1/F0-ATPase [ Patel92 , Patel95 ]. It destabilizes the secE, nusG, L11-L1, L10 and beta cistrons from transcripts from the secEnusG and rplKAJLrpoBC operons, though this is not reflected in a change in mRNA abundance [ Chow94a ]. Other transcripts that are degraded by RNase E include ftsA-ftsZ, thrS, pstG, pnp and rnb [ Cam96 , Nogueira01 , Kimata01 , Hajnsdorf94a , Zilhao95 ]. RNase E is also involved in limiting the abundance of mRNAs from rspT, dsbC, pth and tetR [ Le02b , Zhan04 , CruzVera02 , Baumeister91 ]. Finally, although overexpressed RNase G can partially complement a lack of RNase E, about a hundred RNAs are only degraded by RNase E, including many mRNAs coding for proteins involved in energy generation and macromolecule synthesis and degradation [ Lee02 ].
RNase E regulates its own abundance by cleaving within the 5' untranslated region of rne mRNA. As RNase E activity can be titrated by other substrates, this acts to modulate its expression to match cellular needs [ Mudd93 , Diwa02 , Sousa01 ]. Appending the 5' region of rne to heterologous RNA confers RNase E regulation [ Jain95 ].
In a pnp/rnb/rne triple mutant, RNA polyadenylation is longer and more abundant [ OHara95 ]. Conversely, RNase E indirectly increases polyadenylation by generating new 3' ends on which PAP I, which has a binding region for RNase E, can act [ Mohanty00 , Raynal99 ]. Increased polyadenylation stabilizes the rne transcript [ Mohanty99 , Mohanty02 ].
RNase E cleaves at regions that are single stranded and rich in A/U sequences [ Kim04h , Mackie92 , Bessarab98 , Babitzke91 ]. Though RNase E has no canonical target sequence, the effects of local sequence on cleavage placement and effectiveness have been thoroughly characterized [ Kaberdin03 , McDowall94 ]. Secondary structure in the form of adjacent stem-loops has been shown to be necessary for RNase E cleavage for a number of substrates, and it has been suggested that these structures maintain a stretch of single-stranded RNA for the enzyme to cleave [ Ehretsmann92 , Cormack92 , Diwa00 ]. In other cases, however, secondary structures play no definite role in susceptibility or actually impede RNase E cleavage [ Mackie93 , McDowall95 , Lopez96 ].
RNase E binds to the 5'-monophosphate end of its substrate but then cleaves farther in moving 3' to 5', suggesting a scanning mechanism [ Feng02 ]. In the absence of a 5'-monophosphate, cleavage is slowed [ Jiang04 ]. Blocking the 5' end, either by circularizing the RNA or by adding a 5'-triphosphate also inhibits cleavage [ Mackie00 , Mackie98 ].
The catalytic parameters of RNase E have been thoroughly evaluated [ Redko03 ].
RNase E's enzymatic and RNA-binding functions are split between its amino-terminal and carboxy-terminal portions, respectively [ McDowall96 , Tarasevici95 ]. The carboxy-terminal section of the protein and its arginine-rich RNA-binding domain (ARRBD) is required for mRNA degradation and enhances RNase E autoregulatory cleavage of rne mRNA, but is dispensible for rRNA processing [ Ow00 , Lopez99 , Jiang00 , Kaberdin00 ]. Contradicting this observation, it has been reported that the RNA-binding domain of RNase E is not required for feedback regulation [ Diwa02a ]. Mutations within the RNA-binding do lead to defective binding, but have no effect on RNA cleavage activity [ Shin08 ].
RNase E is catalytically active only as a tetramer, with its RNA-binding domains facing outward [ Callaghan03 ]. Crystallographic and NMR analysis of the isolated RNA-binding domain indicates that it forms a homodimer, possibly contributing to overall tetramer formation [ Schubert04 ]. A crystal structure of the amino-terminal catalytic domain to 2.9 Å resolution shows that the tetramer consists of a dimer of dimers and contains divalent magnesium ion [ Callaghan05 ]. The tetrameric structure is maintained by cysteine-zinc-cysteine linkages between adjacent Rne monomers [ Callaghan05a ].
Both RrnA and CspE bind and inhibit RNase E, and T7 gene 0.7 protein kinase phosphorylates its carboxy-terminal half, stabilizing T7 mRNAs against RNase E degradation [ Lee03d , Feng01 , Marchand01 ].
RNase E is required for cell division to occur [ Goldblum81 ]. Inviability of rne mutants may be due to reduced levels of the cell-division protein FtsZ [ Takada05 ].
The previously reported RNase K appears to be a proteolytic fragment of RNase E [ Mudd93 ].

Front 235

oppB

Back 235

OppB is an integral membrane component of the oligopeptide ABC transporter and the murein tripeptide ABC transporter.

Front 236

oppD

Back 236

OppD is an ATP-binding component of the oligopeptide ABC transporter and the murein tripeptide ABC transporter.

Front 237

hisS

Back 237

Histidyl-tRNA synthetase (HisRS) is a member of the family of aminoacyl-tRNA synthetases, which interpret the genetic code by covalently linking amino acids to their specific tRNA molecules. The reaction is driven by ATP hydrolysis. HisRS belongs to the Class II aminoacyl tRNA synthetases, which share three regions of homology [ Eriani90 , Cusack91 ].

HisRS is a dimer in solution [ Kalousek74 ]. The C-terminal domain of the protein is required for dimerization, while the N-terminal domain contains most of the catalytic activity. The two domains do not complement each other in trans [ Augustine97 ].

Specificity determinants within tRNAHis that are important for recognition by HisRS have been identified; the unique G-1:C73 base pair was found to play a crucial role [ Himeno89 , Yan94 , Fromant00 , Rosen04a , Guth07 ]. Specificity determinants and residues within HisRS that are important for catalytic activity have been investigated [ Yan95 , Ruhlmann97 , Bovee99 , Hawko01 , Connolly04 , Guth07 ], and a model for the catalytic cycle was proposed [ Guth07 ]. The C-terminal domain of HisRS was found to be largely responsible for recognition of the tRNAHis anticodon [ Yan96 ].

Crystal structures of HisRS have been determined, and a reaction mechanism was proposed [ Arnez95 , Arnez97 ]. Various types of experiments support a substrate-assisted concerted reaction mechanism [ Guth05 ]. Catalysis may occur at sites alternating between the two monomers; conformational changes may be rate-limiting for product formation [ Guth07 , Guth09 ]. The mechanism of substrate discrimination has been modeled [ Banik09 , Banik10 ].

Front 238

Enzymatic reaction of: histidyl-tRNA synthetase

Back 238

HisRS
tRNAhis + L-histidine + ATP + H+ <=> L-histidyl-tRNAhis + diphosphate + AMP
The reaction direction shown, that is, A + B ↔ C + D versus C + D ↔ A + B, is in accordance with the direction of enzyme catalysis.
The reaction is physiologically favored in the direction shown.
In Pathways: tRNA charging
Cofactors or Prosthetic Groups: Mg2+ [Airas96 ]
Inhibitors (Competitive): histidinol [ Kalousek74 ]
KM for ATP: 890 μM [ Augustine97 ]
KM for L-histidine: 30 μM [ Augustine97 ]
KM for tRNAhis: 3.5 μM [ Himeno89 ]
pH(opt): 7.4 [Kalousek74 ]

Front 239

pepB

Back 239

Aminopeptidase B (PepB) is one of four cysteinylglycinases in E. coli.

Aminopeptidase B cleaves Leu-Gly, Leu-Gly-Gly, Cys-Gly and Leu-Gly in vitro [ Miller78 , Hermsdorf79 , Suzuki01 ]. In vivo, aminopeptidase B is, along with aminopeptidases A and N, and dipeptidase D, one of four cysteinylglycinases [ Suzuki01a ].

Divalent cations, including some that are not effective stimulators of activity, stabilize aminopeptidase B against heat inactivation [ Suzuki01 ].

Front 240

trmE (also mnmE)

Back 240

MnmE is required for wild-type 5-methylaminomethyl-2-thiouridine modification of tRNA [Elseviers84]. Together with MnmG, MnmE is thus involved in maintenance of the correct reading frame [Brierley97, Urbonavici01, Bregeon01, Urbonavici03].

MnmE also appears to play a role in oxidation of thiophene and furan compounds [Alam91] and regulates glutamate-dependent acid resistance [Gong04].

MnmE is a GTP-binding protein that also exhibits GTPase activity, showing rapid GTP hydrolysis and low nucleotide affinity. The nucleotide binding and hydrolysis activities are localized within the central 17 kDa GTPase domain [Cabedo99]. The GTPase activity as well as the Cys451 residue in the C-terminal domain are required for the wild-type tRNA modification function [Yim03], but not sufficient [MartinezVi05]. Dimerization of the GTPase domain is potassium ion-dependent; subsequent GTP hydrolysis activity is dependent on dimerization [Scrima06]. Low pH inhibits the GTP hydrolysis activity [Monleon07]. Unlike other GTPases, MnmE does not appear to use an "arginine finger" for catalysis [Scrima06, Monleon07].

Solution and crystal structures of the G-domain of MnmE have been solved [Scrima06, Monleon07]. MnmE can homomultimerize and localizes to the cytoplasm, showing some association with the cytoplasmic membrane [Cabedo99]. MnmE interacts specifically with MnmG [Yim06].

Viability of an mnmE mutation is dependent on the strain background [Cabedo99]. mnmE mutants are defective in the tRNA modification 5-methylaminomethyl-2-thiouridine; tRNA anticodons that are modified with 5-methylaminomethyl-2-thiouridine in the wild type show 2-thiouridine modification in the mutant, and mutants exhibit a defect in UAG readthrough [Elseviers84]. Unexpectedly, the hypomodified tRNALys of an mnmE mutant leads to decreased misreading of the anticodon [Hagervall98].

Expression of mnmE is increased during stationary phase, but independent of the stationary phase sigma factor RpoS. Expression is also subject to catabolite repression and is decreased in the absence of oxygen [Zabel00].

Front 241

rpsU

Back 241

he S21 protein is a component of the 30S subunit of the ribosome. S21 was found to associate with the 50S subunit of the ribosome as well as the 30S subunit [ Odom84 ].

Interactions between S21 and 16S rRNA have been mapped [ Stern88a ]. S21 also crosslinks to the 4.5S RNA of the signal recognition particle [ Gu05 ]. Interactions between S21 and mRNA can be shown by fluorescence energy transfer [ Czworkowsk91 ] and crosslinking [ Brandt92 ].

S21 is required for full activity of translation initiation [ Held74a , Van81 ]. S21 was also shown to crosslink to IF3 [ MacKeen80 , Cooperman81 ].

Processing and degradation of the rpsU-dnaG-rpoD operon mRNA has been studied; see for example [ Yajnik93 , Lupski84a ] and references therein.

Front 242

dppF (also called dppE)

Back 242

DppF is an ATP-binding component of the dipeptide ABC transporter.

Front 243

dppD

Back 243

DppD is an ATP-binding component of the dipeptide ABC transporter.

Front 244

dppC

Back 244

DppC is an integral membrane component of the dipeptide ABC transporter.

Front 245

dppB

Back 245

DppB is an integral membrane component of the dipeptide ABC transporter.

Front 246

dppA

Back 246

DppA is the periplasmic binding component of the dipeptide ABC transporter.

Front 247

mnmG (or trmF)

Back 247

MnmG and MnmE both act in a tRNA modification pathway that reduces +2 frameshift errors in translation [Bregeon01]. Transcription of mnmG, which is adjacent to oriC, appears to affect DNA replication [Ogawa91, Theisen93].

The MnmG protein appears to be a dimer in solution, interacts specifically with MnmE, and binds FAD. Mutations in the conserved G13 and G15 residues of the proposed FAD binding site lead to loss of FAD binding and loss of methylaminomethyl modification of tRNAs [Yim06].

mnmG is allelic with trmF [Bregeon01, Yim06]. tRNA isolated from a trmF mutant carries 2-thiouridine instead of 5-methylaminomethyl-2-thiouridine modifications in the wobble position of the anticodon in certain tRNAs [Elseviers84, Yim06]. mnmG complements the temperature-sensitive growth defect of a mutant that requires the normally non-essential tRNAleuX for growth [Nakayashik98].

The effect of an mnmG null mutation is dependent on the strain background; an mnmG null mutation can not be introduced into the V5701 strain background, but only leads to a reduced growth rate in the MG1655 background [Yim06].

Regulation of transcription during the cell cycle has been described [Kolling88, Gielow88, Ogawa91, Theisen93, Bogan97, Zhou97b].

Front 248

fis

Back 248

Fis, "factor for inversion stimulation", is a small DNA-binding and bending protein whose main role appears to be the organization and maintenance of nucleoid structure through direct DNA binding and by modulating gyrase [Cho08a, Schneider01] and topoisomerase I production [WeinsteinF07], as well as regulation of other proteins that modulate the nucleoid structure, such as CRP, HNS, and HU. Fis directly modulates several cellular processes, such as transcription, chromosomal replication, DNA inversion, phage integration/excision, and DNA transposition [Finkel92, Travers01].

As a transcriptional regulator, Fis regulates the expression of many genes involved in translation (rRNA and tRNA genes), virulence, biofilm formation, energy metabolism, stress response, central intermediary metabolism, amino acid biosynthesis, transport, cell structure, carbon compound metabolism, amino acid metabolism, nucleotide metabolism, motility, and chemotaxis [Bradley07, Finkel92, Sheikh01]. A transcriptome analysis has shown that transcription of approximately 21% of genes is modulated directly or indirectly by Fis [Cho08a].

Fis, together with proteins such as HNS, HU, IHF, and Dps, is one of the largest components of the nucleoid. A ChIP-chip analysis has shown that Fis binds to 894 DNA regions in the genome, which results in two sites of Fis per supercoiling domain. These regions include both intergenic regions and regions within genes, and not all genes to which Fis binds are affected by this protein, a fact that is in agreement with the primary role assigned to Fis in maintenance of nucleoid structure [Cho08a].

Under optimal growth conditions, Fis is the dominant DNA-binding protein in the cell. Up to 60 000 copies of the protein can be found in a single cell under log phase, but it is nearly imperceptible in stationary phase (<100 molecules per cell) [Ali99]. Fis bends the DNA between 40 and 90° [Finkel92]. This bending stabilizes the DNA looping to regulate transcription and to promote DNA compaction [Skoko06, Travers98].

The crystal structure of Fis has been solved by several research groups [Cheng00, Yuan91, Kostrewa91, Safo97], and they have shown that this protein has an α-helical core of four helices (A to D). The C-terminal domain contains the C and D helices, forming a helix-turn-helix motif, characteristic of many DNA-binding proteins. The N-terminal domain has the A helix and a flexible β-hairpin arm, which are involved in DNA inversion; this region appears to make contact with the DNA invertase Hin in the invertasome structure [Safo97, Finkel92].

As expected for a gene involved in the modulation of many cellular processes, expression of fis is regulated by several systems and at different levels. At the transcription level, Fis is autoregulated, induced by high supercoiling levels [Schneider00], and regulated by both growth rate-dependent and stringent control systems that require the presence of a GC motif downstream of the -10 region [Ninnemann92].

Transcription of fis is also regulated by the availability of the nucleotide triphosphate CTP, which is the nucleotide with which transcription of fis is initiated and whose largest concentration is seen during log phase, a fact that correlates with the pattern of fis expression. If transcription of this gene is initiated with an A or a G instead of a C, transcription is induced at the same level in early stationary phase rather than in log phase [Walker04].

A promoter that does not transcribe any ORF is located upstream of the promoter that transcribes fis and in the opposite direction. Under RNA polymerase limitation, both promoters compete for the protein. This might be a type of fis regulation expression, but there is no evidence to support this claim [Nasser01].

DksA, a protein that is associated with RNA polymerase in regulating transcription, inhibits transcription of fis by increasing the inhibitory effects of ppGpp, decreasing the lifetime of the RNA polymerase-fis promoter complex, and increasing the sensitivity to the CTP nucleotide [Mallik06].

BipA is a protein required for fis translation; this protein appears to destabilize the strong interaction between the 5� end of the untranslated region of the fis mRNA and the ribosome [Owens04].

In a specific way and with tight binding affinity, Fis recognizes and binds as a dimer to sites that have poorly related sequences. However, a very degenerated consensus sequence has been proposed for Fis sites, with only a subset of common nucleotides. This consensus sequence shows a core binding site of 15 bp with partial dyad symmetry [Finkel92]. This sequence presents only four highly conserved nucleotides, a G and a C at the 1st and at the 15th nucleotide, respectively, and a pyrimidine (A or G) and a purine (T or C) at the 5th and 11th positions, respectively. The central region commonly presents an AT-rich sequence [Shao08, Hengen97]. Some of these sites have internal sequences that are potential methylation targets, 5�-GATC-3�, suggesting that under some circumstances Fis binding is controlled by methylation [Weinreich92]. The binding of Fis is destabilized when the tension of the double helix increases [Xiao11].

Front 249

obgE

Back 249

ObgE is an essential member of the Obg family of small GTP-binding proteins [Arigoni98, Kobayashi01c]. A variety of essential cellular processes, including chromosome segregation and ribosome assembly, are affected by ObgE.

Purified ObgE has moderate affinity and high guanine nucleotide exchange rate constants for both GTP and GDP and a relatively low GTP hydrolysis rate [Kobayashi01c, Wout04, Persky09]. The physiologically relevant ligand may be ppGpp. During amino acid starvation, ObgE is important for cell survival and appears to alter the ratio of pppGpp to ppGpp [Persky09].

ObgE is involved in late steps of the assembly of the 50S ribosomal subunit. In cell lysates, the ObgE protein is associated with the 50S ribosomal subunit [Wout04, Jiang06a], but not during amino acid starvation [Jiang07b]. Purified ObgE also cofractionates with the 30S ribosomal subunit [Sato05, Jiang06a]. ObgE copurifies with a number of ribosomal subunit proteins [Sato05]. Earlier experimental results that showed no association of ObgE with the ribosome [Kobayashi01c] are thought to be due to instability of the association under the conditions used [Wout04]. ObgE is involved in 16S and 23S rRNA processing [Sato05, Jiang06a]; in the presence of GTP, purified ObgE cosediments with both 16S and 23S rRNA [Sato05]. ObgE also copurifies and interacts with SpoT [Wout04]. Overproduction of ObgE suppresses the growth and ribosome assembly defects of a mutant lacking the RrmJ heat shock protein and rRNA methyltransferase [Tan02]. Depletion of ObgE does not appear to affect translation [Foti07].

ObgE may play a role in cell survival after replication fork arrest. A mutant containing a hypomorphic obgE allele (obgE::Tn5) shows increased sensitivity to replication inhibitors such as hydroxyurea and guanazole, and concomitant loss of recA or recB substantially enhances this phenotype. The obgE::Tn5 allele is synthetically lethal with seqA, suggesting that ObgE plays a role in the regulation of replication or in organization of the replication fork [Foti05]. Cells with depleted ObgE show perturbed SeqA foci [Foti07]. Levels of the replication-initiator DnaA depend directly on the abundance of ObgE [Ulanowska03, Sikora06], indicating that ObgE acts by regulating DnaA [Sikora06].

ObgE appers to be involved in cell cycle progression and chromosome partitioning [Dutkiewicz02, Kobayashi01c, Foti07]. A heat sensitive mutant exhibits a defect in chromosomal partitioning and elongated cell morphology, but has no defect in DNA replication [Kobayashi01c]. Overproduction causes a defect in chromosomal partitioning and leads to oversized cell morphology, but does not cause gross defects in DNA replication, RNA transcription, or protein synthesis [Kobayashi01c, Dutkiewicz02]. Depletion of ObgE delays chromosome segregation, but is not lethal; growth resumes after expression of ObgE is restored [Foti07]. Purified ObgE associates with DNA and shows some membrane association [Kobayashi01c], but it does not appear to be associated with the nucleoid in vivo [Sato05].

ObgE is an abundant protein [Kobayashi01c], and the growth rate correlates with the cellular level of ObgE [Sato05]. Levels of ObgE protein increase five-fold after exposure to UV irradiation, and moderate overexpression of ObgE leads to increased resistance to UV radiation [Zielke03]. ObgE was found to be thioredoxin-associated [Kumar04].

ObgE has been shown to enhance survival of UV-irradiated cells; this UV resistance increases with ObgE overexpression. Temperature-sensitive obgE mutants were found to have a lower basal level of the RecA repair protein compared with wild-type; recA expression was not stimulated by UV irradiation in these mutants. Taken together, these results strongly suggest that ObgE plays a role in the DNA repair process, most likely through stimulation of the RecA-dependent DNA repair pathway [Zielke03].

Front 250

rnhA

Back 250

RNase H cleaves RNA in RNA-DNA hybrids. Targets for RNase H include RNA primers for DNA synthesis, especially longer primers that are cleaved by RNase H prior to digestion by the exonuclease activity of DNA polymerase I [Ogawa84, Kitani85]. RNase H also acts to block the formation of R-loops and the initiation of DNA replication from sites other than the classical origin of replication, oriC.

RNase H carries out the endonucleolytic cleavage of RNA in RNA-DNA hybrids, cleaving near the 3' terminus of the RNA and then digesting the remainder of the RNA [Miller73, Crooke95]. Though RNase H only cleaves RNA in hybrid duplexes, it binds DNA-DNA, DNA-RNA, and RNA-RNA duplexes [Oda93]. This binding occurs in the minor groove of the duplex, and may involve hydrogen bonds with 2' oxygens in the RNA [Daniher97, Nakamura91]. A-form (RNA-RNA, RNA-DNA) duplexes are bound 60 times more tightly than B-form (DNA-DNA) duplexes, and 300 times more tightly than single-stranded oligonucleotides [Lima97]. Tests with boranophosphate DNA analogs demonstrate that weaker hybridization yields more rapid RNA degradation [Rait99]. Evaluation of duplexes containing arabinonucleic acids suggests that RNase H may recognize RNA-DNA hybrid duplexes by dint of a minor groove width intermediate between that seen in RNA-RNA and DNA-DNA duplexes [Noronha00, Minasov00]. Steric modification of the minor groove results in a substantial increase in kM, whereas modification of the major groove has no effect and altering the 2'-hydroxyl does not alter binding but lowers the cleavage rate [Uchiyama94]. Damaged DNA can disrupt cleavage of hybridized RNA, directing it to sites adjacent to the damaged nucleotides [Shiels01]. Cleavage can also be slowed by the presence of nearby RNA secondary structure [Lima97a]. However, RNase H appears to be able to induce hybrid formation and cleavage even with complementary DNA that exists in stable stem-loops or duplexes [Li98].

Mutants lacking RNase H function can undergo "constitutive stable DNA replication" (cSDR), replication in the absence of protein synthesis [Casaregola87]. This replication can also take place without functional DnaA and OriC; rnhA mutation suppresses the negative effects of loss of DnaA function [Kogoma83, Lindahl84]. Though RNase H is not required for initiation of replication, it serves as a specificity factor blocking replication from sites other than OriC [Ogawa84a, Hong93, Kogoma85]. Indeed, replication in an double mutant lacking RnhA and OriC starts from multiple origins [deMassy84]. Though cSDR normally requires RecA, a lexA mutant can suppress this requirement, as long as DNA Pol I polymerase activity is present [Cao93]. This mutation can also suppress the lethality of an rnhA, polA double mutant [Kogoma97]. Various other genotypes allowing cSDR have been examined [Torrey87]. Double mutants lacking rnhA and the helicase gene recG are inviable; mutants in recG alone also display cSDR and a requirement for DNA Pol I function [Hong95]. Double mutants in rep and rnha are also inviable [Sandler05]. Plasmid ColE1 can also undergo cSDR in the absence of RecA [Naito86]. cSDR requires PriA and PriB, but not PriC [Masai94, Sandler05].

RNase H mutants affect plasmid replication even in the presence of protein synthesis. pBR322 can replicate in the absence of DNA Pol I in an rnhA mutant [Kogoma84]. Both pBR322 and ColE1-type plasmids form concatamers in rnhA mutants [Subia86]. Loss of rnhA allows replication of replication-defective ColE1, even if the entire origin of replication is deleted [Naito84, Ohmori87].

RNase H exerts control over the formation of R-loops (single strands of DNA that loop out when RNA binds to their cognate strand), probably by degrading the RNA that induces looping. In the absence of DNA topoisomerase I, DNA gyrase can induce R-loop formation, impairing rRNA transcript elongation and limiting the induction of heat shock proteins. Overexpressed RNase H blocks these effects [Drolet95, Cheng03, Hraiky00]. Overexpressed RNase H is detrimental at 21 degrees C, though combined rnhA, nusG double mutants are lethal at this temperature [Masse99, Harinaraya03]. Overexpression of topoisomerase III blocks both the detrimental effects of a topoisomerase I mutant and those of RNase H overexpression at low temperature [Broccoli00]. RNase H overexpression is also detrimental following UV irradiation, which may explain why its abundance is unaltered even during UV-induced stable DNA replication [Bockrath87, Bialy86].

RNase H has been subject to extensive structural analysis. A crystal structure to 1.8 Å resolution showed that RNase H has two domains, and that its metal binding site is near a cluster of four acidic residues, a feature conserved in homologs in other organisms [Katayanagi90]. Based on a 1.48 Å resolution structure, RNase H comprises a five-stranded beta sheet and five α helices [Katayanagi92]. A subsequent cocrystal with divalent magnesium ion shared a similar backbone conformation with the pure protein with minor differences in the substrate binding region, and a single coordinated magnesium [Katayanagi93]. Crystal structures of several active site mutants reveal only localized conformational change around this magnesium-binding site [Katayanagi93a]. A cocrystal with divalent manganese shows two metal-binding sites, one needed for enzymatic activity, the other possibly involved in inhibition [Goedken01]. Indeed, RNase H is activated by low concentrations and inhibited by high concentrations of manganese ion [Keck98]. Crystal structures have also been determined for RNase H bound to RNA-DNA hybrid substrate [Nowotny05]. Alteration of the coordinating residues Asp10 and Glu48 to arginines yields a protein that is structurally similar to wild type by circular dichroism, binds nucleic acids well but does not complement a mutant lacking RNase H activity [Tsunaka01]. Alterations in a "handle" region from residues 84-99 does alter protein conformation and raises kM 3-5 fold [Kanaya91]. Deletion of the carboxy-terminal E helix does not affect structure but diminishes activity; adding a peptide corresponding to this region stimulates activity in this mutant variant [Goedken97]. Removal of a basic helix/loop region that can render homologous HIV Rnase H partially active also diminishes enzymatic activity [Keck96].

Molecular dynamics modeling of various divalent metal ions in RNase H has been carried out, and the stereochemistry of the RNase H reaction has been described [Babu03, Krakowiak02]. RNase H folding has been studied as an example of general protein folding principles [Spudich04, Cecconi05].

The SOS response is active in rnhA mutants [Kogoma93]. Two isoelectric points were found for this protein, 9.0 and 9.6 [Arendes82].

Front 251

rraA

Back 251

RraA inhibits ribonuclease E (RNase E, Rne) activity by binding to and masking the C-terminal RNA binding domain of RNase E. The interaction of RraA with the degradosome is facilitated by protein-RNA remodeling via the ATPase activity of RhlB [Gorna10].

RraA physically interacts with RNase E, but does not interact with the RNA substrates [Lee03a]. High-affinity binding of RraA to RNase E requires the C-terminal domain (CTD) of RNase E [Lee03a, Gao06]. RraA interacts with both RNA-binding sites of RNase E and interferes with their interaction with RNA [Gorna10]. RraA also interacts with the RhlB helicase component of the degradosome, and a ternary complex of RraA, RNase E and RhlB can be observed [Gorna10]. Binding of RraA or RraB, a second modulator of RNase E activity, differently affect the composition of the degradosome [Gao06].

A crystal structure of RraA is presented at 2.0 Å resolution. RraA forms a homotrimer; the complex is shaped like a ring with a hole of 12 Å across. RraA is structurally related to a family of aldolases [Monzingo03].

The regulatory interaction between RraA and RNase E and their orthologs appears to be evolutionarily conserved [Yeom08, Yeom08a, Lee09j].

Transcription of rraA is σS-dependent and increased upon entry into stationary phase. The stability of rraA mRNA itself is dependent on the activity of RNase E [Zhao06c]. Overproduction of RraA causes pleiotropic phenotypes due to increased abundance of RNAs that are usually substrates of RNase E [Lee03a]. Overexpression of RraA rescues cells overexpressing RNase E from growth arrest [Yeom06].

RraA was originally mis-annotated as a SAM-dependent methyltransferase predicted to act in menaquinone biosynthesis, and given the name MenG (Hudspeth et al., unpublished; GenBank record U56082); however, the protein was found to lack both structural or functional indications of methyltransferase activity [Lee03a, Monzingo03].

Front 252

rpsD

Back 252

The S4 protein, a component of the 30S subunit of the ribosome, functions in the assembly of the 30S ribosomal subunit, the mRNA helicase activity of the ribosome, the regulation of translation of a subset of ribosomal proteins, and transcription antitermination of rRNA operons.

S4 interacts directly with helical elements at the 5' domain of the 16S rRNA [Stern86, Powers95a, Vartikar89, Bellur09, Ramaswamy09]. The N-terminal domain of unbound S4 is dynamically disordered, which is exploited to initiate S4 binding to helix 16 in the unfolded five-way junction of 16S rRNA [Chen10c]. The ability of both S4 and S7 to bind 16S rRNA by themselves indicates that they function as initiator proteins for the assembly of the 30S subunit of the ribosome. The S20, S16, S15, S6, and S8 subunits appear to depend on S4 for assembly [Nowotny88].

The S4 protein is involved in the regulation of translation of the other ribosomal proteins encoded by the α operon, RpsM (S13), RpsK (S11), RplQ (L17) and S4 itself [Yates80, Thomas87, JinksRober82]. The α operon leader region is required for translational repression by S4 [Thomas87]; S4 specifically interacts with a double pseudoknot structure which overlaps with the ribosome binding site and initiation codon for RpsM [Deckman85, Deckman87, Deckman87a, Tang89, Tang90, Gluick95]. There may be a second binding site for S4 upstream of the RplQ open reading frame [Meek84]. The same protein domain appears to be responsible for both mRNA and rRNA binding [Baker95a, Conrad87].

S4 can also act as a general transcription antitermination factor similar to NusA; it associates with RNA polymerase and is involved in rRNA operon antitermination [Torres01a].

S4 influences translational fidelity [Topisirovi77]. Certain mutations in rpsD confer a "ribosomal ambiguity" (ram) phenotype, which is characterized by decreased growth rate, increased streptomycin sensitivity, and increased errors in translation [Zimmermann71, Andersson83a, Andersson82a, Olsson79]. Ribosomes containing the S4 rpsD12 allele display the ram phenotype; these ribosomes appear to exploit only the initial phase of tRNA selection, thus reducing discrimination against near-cognate aa-tRNAs [Zaher10]. Cells carrying the rpsD14 allele have a mutator phenotype [Balashov03]. The ribosome was found to have mRNA helicase activity, and mutations in the S3 and S4 subunits impair this activity [Takyar05].

Front 253

frr

Back 253

Rrf recycles the ribosome upon translation termination [Janosi94, Janosi98]. Rrf, release factor RF-3, and elongation factor EF-G are involved in this recycling process [Pavlov97, Pavlov97a]. At termination, Rrf and EF-G catalyze release of the 50S ribosomal subunit from the 70S complex, a GTPase-dependent process [Karimi99]. Release factor RF-1 inhibits Rrf activity [Pavlov97a]. Rrf with EF-2 and RF-3 increase dissociation of peptidyl-tRNA from the ribosome [HeurgueHam98]. Rrf stimulates translation [Ryoji81] and inhibits amber readthrough [Ogawa75, Ryoji81, Ryoji81a].

Rrf is essential for viability [Janosi94, Janosi98]. A mutant lacking 25 C-terminal residues is inviable [Fujiwara99b].

Rrf has been structurally characterized [Janosi00, Yun00a, Kim00e, Yoshida02c]. Crystal structures [Kim00e, Nakano02] and NMR studies [Yoshida03a, Yoshida02c] are presented. Rrf has structural similarity to tRNAs [Kim00e, Nakano02], and was thought to bind to the ribosome like tRNA [Fujiwara01a, Hirokawa02, Hirokawa02a]; however, protection assays indicate that the interaction of Rrf with the ribosome is physically distinct from the interaction of tRNA with the ribosome [Lancaster02].

Rrf shows physical interactions with 50S ribosomal subunits, and shows weaker interaction with 70S ribosomes and 30S ribosomal subunits [Ishino00, Todorova03]. Interaction with the 50S ribosome involves Rrf residue R132 [Ishino00] and is sensitive to some antibiotics [Ishino00, Todorova03].

The rrf mutant phenotype is functionally complemented by coproduction of Mycobacterium tuberculosis Rrf and Mycobacterium tuberculosis EF-G [Rao01], by coproduction of Thermus thermophilus Rrf with specific EF-G mutant proteins or an E. coli-Thermus thermophilus chimeric EF-G protein [Ito02a], or by production of a mutant Thermus thermophilus Rrf with a five-residue C-terminal truncation [Fujiwara99b].

Regulation has been described [Ichikawa89, Shimizu91].

Front 254

rpsO

Back 254

The S15 protein is a component of the 30S subunit of the ribosome and also functions in the post-transcriptional regulation of its own expression.

The S15 protein binds to 16S rRNA in the absence of other ribosomal proteins [Zimmermann72, Zimmermann75, Gregory84]. Nucleotides essential for the S15-16S rRNA interaction have been determined by mutagenesis [Stark84, Serganov01] and nuclease protection [Wiener88, Mougel88]. Binding of S8 to 16S rRNA influences the central domain organisation and affects the rRNA environment of S15 [Jagannatha03].

In addition to its function in the ribosome, the ribosomal protein S15 binds to its own mRNA, stabilizing a pseudoknot secondary structure and impeding translation initiation [Portier90, Philippe90, Portier90a, Philippe94, Benard94, Philippe95, Benard98]. S15 appears to prevent the formation of a functional ternary 30S-mRNA-tRNA(fMet) complex, trapping the ribosome in a preinitiation complex [Philippe93]. rpsO mRNA and 16S rRNA compete for binding to S15 [Philippe94]; common structural determinants between the mRNA and rRNA bindig sites have been investigated, showing that there is limited similarity between the two targets [Serganov02, Mathy04].

Ribosomes lacking S15 can suppress the rpoH11 mutation [Yano89]. A mutation in rpsO (secC) suppresses a secA(Ts) allele [FerroNovic84]. S15 appears to be required for the optimal synthesis of lipoprotein [Watanabe88].

Processing and degradation of rpsO mRNA have been studied extensively; see for example [Le02, Marujo03, Folichon03, Folichon05a] and references therein.

Front 255

pcnB (or PAP I)

Back 255

Poly(A) polymerase I is responsible for the polyadenylation of 3' ends of RNA molecules.

Poly(A) polymerase polyadenylates the vast majority of mRNA transcripts [Mohanty06]. Unlike in eukaryotes, increased polyadenylation of mRNAs leads to decreased mRNA half-life [Mohanty99, Mohanty06]. Rho-independent transcription terminators appear to serve as targeting signals for polyadenylation [Mohanty06]. The Hfq protein appears to be involved in the recognition of 3' termini of RNA by poly(A) polymerase I [Le03].

Intracellular levels of poly(A) polymerase I as well as the level of pcnB transcription vary inversely with growth rate [Jasiecki03]. Overexpression of poly(A) polymerase I is toxic and leads to slowed growth [Mohanty99, Mohanty06]. Use of AUU as the translational start codon results in InfC discrimination (as with production of the IF-3 translation initiation factor) and results in low levels of poly(A) polymerase I in the cells [Binns02].

A his-tagged version of poly(A) polymerase I showed reduced activity following phosphorylation of its his tag. In other proteins, this sometimes correlates with the protein itself being regulated via phosphorylation [Jasiecki06].

Front 256

acnB

Back 256

There are two aconitases in E. coli, both of which catalyze the reversible isomerization of citrate and iso-citrate via cis-aconitate. AcnB also plays a role in the methylcitrate cycle for degradation of propionate, where it is responsible for hydration of 2-methyl-cis-aconitate to (2R,3S)-2-methylisocitrate [Brock02]. The apo form of AcnB is able to bind mRNA and enhances translation of AcnB [Tang99].

AcnB appears to function as the main catabolic enzyme, while the main role of AcnA appears to be as a maintenance or survival enzyme during nutritional or oxidative stress [Cunningham97]. The AcnB enzyme is less stable, has a lower affinity for citrate and is active over a more narrow pH range than the AcnA enzyme [Jordan99, Varghese03]. Unlike AcnA, AcnB is sensitive to oxidation in vivo [Brock02, Varghese03]. AcnB rapidly loses catalytic activity when the iron concentration is low [Varghese03].

The N-terminal region of AcnB mediates the formation of AcnB homodimers in the presence of Fe2+; the 4Fe-4S cluster or catalytic activity is not required for dimer formation. In the absence of Fe2+, the same region is able to bind to mRNA [Tang05]. AcnB also interacts weakly with isocitrate dehydrogenase [Tsuchiya08]. The catalytically inactive AcnB apo-protein, lacking its iron-sulfur cluster, has a negative effect on SodA synthesis in vitro [Tang02].

A crystal structure of AcnB has been solved at 2.4 Å resolution [Williams02].

An acnB mutant does not grow on acetate as the sole source of carbon, grows poorly on other carbon sources such as glucose and pyruvate [Gruer97], contains high levels of citrate, and excretes substantial amounts of citrate into the medium [Varghese03]. An acnB mutant is more sensitive to peroxide stress than wild type and shows increased SodA synthesis [Tang02].

Expression of acnB increases early in exponential phase and decreases during entry into stationary phase [Gruer97, Cunningham97].

Front 257

rplA

Back 257

The L1 protein is a component of the 50S subunit of the ribosome and also functions in the post-transcriptional regulation of the ribosomal protein genes encoded in the L11 operon. Ribosomes lacking L1 show a lower translation activity than wild type [Subramania80] and are defective in binding of aminoacylated tRNA [Sander83]. L1 has been identified within a 3-D map of the 70S ribosome constructed by cryo-electron microscopy [Malhotra98].

L1 interacts with a region in the 5' end of 23S rRNA [Branlant76b, Branlant80, Egebjerg91]. It also can be crosslinked to a region near the central fold of aminoacylated tRNA in the P and E site [Podkowinsk89, Rosen93, Osswald95]. L1 is located within 21 Å of nucleotide C2475 of 23S rRNA, near the peptidyltransferase center [Muralikris95].

L1 is a translational repressor of the synthesis of L11 and L1, the proteins encoded by the L11 operon [Brot81, Thomas87, Stoffler81, Yates80, Dabbs81]. Synthesis of L1 is regulated by translational coupling to the synthesis of L11 [Yates81, Baughman83]. The target site for L1 binding to the mRNA is near the translation initiation site of L11 [Yates81, Baughman83, Thomas87a], and the presence of 23S rRNA relieves translational inhibition by L1 [Yates81]. The predicted secondary structure of the L1 binding region within 23S rRNA and rplKA mRNA is similar [Branlant80, Gourse81] and has been studied experimentally [Baughman84, Kearney87, Said88, Drygin00].

Both the growth rate control and stringent control of the synthesis of ribosomal proteins L11 and L1 are resulting from translational regulation by L1 [Cole86].

Front 258

rpsA

Back 258

The S1 protein is essential in E. coli; it is a component of the ribosome and is likely required for translation of most mRNAs [Sorensen98]. During translation initiation, mRNAs are selected and bound to the ribosome with the help of two components: the conserved 3' end of 16S rRNA is complementary to the Shine-Dalgarno region of the typical mRNA, while the S1 protein binds to the leader sequence of mRNAs, upstream of the Shine-Dalgarno region [Boni91, Komarova02]. S1 promotes RNA strand displacement in vitro [Rajkowitsc07]. The S1 protein is not required for the translation of leaderless mRNAs [Tedin97, Moll02]. Association of S1 with the ribosome is unstable [Subramania77], and the S2 protein is required for binding of S1 to the 30S subunit of the ribosome [Moll02].

The S1 protein is the largest of the ribosomal proteins and assumes a complex elongated shape; it is located at the junction of the head, platform, and main body of the 30S subunit [Sengupta01]. RNA binding is associated with structural modification of the S1 protein [Aliprandi08]. NMR structures of domains 4 and 6 have been determined [Salah09].

The translation initiation region of rpsA mRNA, which encodes the S1 protein, lacks a canonical Shine-Dalgarno region, but nevertheless supports high levels of translation [Boni00, Skorski06]. Expression of S1 is autoregulated; the S1 protein acts as a specific repressor of translation of its own mRNA [Christians81, Skouv90]. Supported by various types of evidence including site-directed mutagenesis, footprinting and phylogenetic studies, a model was proposed in which the rpsA leader region folds into three stem-loop structures which enable efficient translation; binding of free S1 protein disturbs the conformation of the rpsA mRNA and thus specifically represses translation [Boni01, Tchufistov03].

S1 proteins carrying a C-terminal truncation due to IS10R insertion in the ssyF29 mutant lack the R4 RNA binding domain and are able to stimulate translation of the rpsA mRNA compared to wild type [Boni00]. This effect was found to be due to destabilization of the rpsA mRNA [Skorski07]. The S1 protein represses its own translation. The N-terminus of S1 was reported to be sufficient for repressing the expression of rpsA [Skouv90]. Repression may have been due to full-length S1 protein that had been displaced from the ribosome [Boni00].

The S1 protein has been suggested to play a role in trans-translation by binding transfer-messenger RNA (tmRNA) and delivering it to stalled ribosomes [Bordeau02, Karzai01, Wower00, Wower04, Okada04]. An alternative model suggests that S1 binds to both mRNA and tmRNA molecules indiscriminately and may not play a direct role in tmRNA-mediated tagging of incompletely translated polypeptides [McGinness04]. Later studies disagree on whether S1 is [Saguy07] or is not [Qi07] required for trans-translation in vitro.

The S1 protein has been reported to bind to RNA polymerase and stimulate transcription of a number of promoters [Sukhodolet03]. S1 activates transcriptional cycling in vitro [Sukhodolet06]. S1 may also inhibit RNaseE-dependent mRNA decay [Delvillani11].

Front 259

rpoA

Back 259

RpoA is the α subunit of the RNA polymerase core enzyme. It consists of two domains connected by a flexible linker.

The RpoA amino-terminus is both necessary and sufficient for dimerization of RpoA and subsequent assembly of the RNA polymerase core complex [Zhang98d]. The amino-terminus has been analyzed both by NMR and via a 2.5 Å resolution cystral structure [Otomo00, Zhang98d].

The amino-terminal and carboxy-terminal domains of RpoA are connected by a flexible linker, which has been shown to affect transcription in a promoter-dependent fashion [Jeon97, Fujita00, Meng00].

The carboxy-terminal domain of RpoA is involved in antitermination, rho-dependent termination, and is a target for interactions with transcription termination/antitermination L factor that control termination and pausing [Schauer96, Kainz98]. Interaction with the RpoA carboxy-terminal domain activates RNA binding by transcription termination/antitermination L factor [Mah00]. The RpoA carboxy-terminal domain is also required for some kinds of transcriptional activation and plays a role in some transcriptional initiation [Zou97, Burns99].

Front 260

degP

Back 260

Protease Do, or DegP, is a periplasmic serine protease required for survival at high temperatures [Lipinska89, Strauch89, Seol91]. DegP degrades abnormal proteins in the periplasm, including mutant proteins, oxidatively damaged proteins and aggregated proteins [Strauch88, Strauch89, SkorkoGlon99, Laskowska96]. DegP has been specifically shown to degrade the mutant periplasmic protein MalS, as well as unassembled subunits from protein complexes, including HflK, LamB and PapA [Spiess99, Kihara98, Misra91, Jones02].

DegP also proteolyzes a range of other proteins that may not be quality control substrates, such as the DNA methyltransferare Ada, various forms of the colicin A lysis protein and the replication initiation inhibitor IciA [Lee90, Cavard89, Cavard95, Yoo93]. DegP also binds to the ssrA-encoded degradation tag, though this PDZ-domain-mediated interaction does not appear to allow DegP proteolysis of tagged proteins [Spiers02]. Finally, strains lacking DegP are more susceptible to the cationic antimicrobial peptide Lactoferricin B, indicating a possible role for DegP in degradation of that molecule [Ulvatne02].

DegP also has an independent chaperone activity that functions even in proteolytically inactive mutants of DegP [Spiess99]. This chaperone activity is required for survival in the case of disrupted outer membrane assembly, preventing buildup of toxic aggregates [Misra00]. There may be some redundancy between DegP and the chaperones Skp and SurA [Rizzitello01].

DegP is a six-membered ring-shaped structure with a central cavity which contains its proteolytic sites [Swamy83, Kim99]. The hexamer is built from a pair of staggered trimeric rings, with the proteolytic cavity accessible from the sides rather than the ends [Krojer02]. There are two PDZ domains in each monomer which are required for this assembly, and which may be involved in opening and closing the lateral openings [Sassoon99]. Binding of substrate to the PDZ1 domain induces oligomer conversion from a resting hexameric state to a higher order active complex [Krojer10, Merdanovic10]. The PDZ1 domain anchors substrate, facilitating its presentation to the proteolytic domain [Krojer08]. DegP is a processive protease - cleaving its substrate into peptides with a mean size of 13-15 residues [Krojer08]. The PDZ1 domain is required for protease activity and for binding of unfolded proteins, while the PDZ2 domain is primarily required for maintaining a hexameric configuration [Iwanczyk07]. The inner cavity also has several hydrophobic patches, which may be involved in its chaperone function [Krojer02].

Hexameric DegP assembles into large catalytically active spherical structures around its substrate [Krojer08a, Jiang08]. The spherical multimers exhibit proteolytic and chaperone-like activity [Shen09]. A model polypeptide substrate binds each DegP subunit at two sites in the crystal structure of a DegP dodecamer [Kim11]. Substrate binding drives the formation of proteolytically active dodecamers and larger cages of 18, 24 and 30 subunits while substrate cleavage promotes cage disassembly [Kim11].

DegP's proteolytic activity is increased at high temperatures but drops dramatically at low temperatures, leaving its chaperone function unaffected [SkorkoGlon95, Spiess99]. DegP interacts with phosphatidylglycerol on the periplasmic face of the inner membrane, undergoing a conformational change that correlates with the temperature dependence of its proteolytic capacity [SkorkoGlon97].

The mature form of DegP is derived by cleavage of its first twenty-six amino acids by leader peptidase [Lipinska90, Lipinska89]. Targeting of DegP to the Sec-translocase for transport across the inner membrane is SecB-dependent [Baars06].

DegP is a member of the HtrA (high temperature requirement) family of proteases which combine a protease domain with one or more PDZ domains and function as higher order oligomers [Kim05].

DegP is downregulated during low osmolarity [Forns05].

Front 261

lon

Back 261

Lon is an ATP-dependent protease responsible for degradation of misfolded proteins as well as a number of rapidly degraded regulatory proteins. Key regulatory proteins that are Lon substrates include the cell division regulator SulA [Schoemaker84, Higashitan97], the capsule synthesis regulator RcsA [TorresCaba87] and possibly TER components involved in blocking septation sites during the SOS response [Dopazo87]. Lon is required for degradation of misfolded proteins and the prevention of aggregate formation [Chung81, Ryzhavskai, Laskowska96]. In the absence of Lon function, aggregation triples [Rosen02]. At least some of this degradation of misfolded proteins depends on the chaperone DnaK [Sherman92].

Lon also degrades the lamba phage N and Xis proteins, with degradation of the latter promoting lysogeny over lysis [Maurizi87, Leffers98]. Other substrates included HU1 in the absence of its partner HU2, HemA and DAM methylase [Bonnefoy89, Wang99b, Calmann03].

Lon degrades the antitoxin protein in many toxin/antitoxin protein pairs, including both plasmid and chromosomal versions. Lon proteolysis of the antitoxin protein in plasmid-encoded pairs is required for plasmid maintenance, as the antitoxin has a shorter half life in lon+ cells than the toxin, thus requiring the continued presence of the plasmid for cell survival. Plasmid-encoded antitoxin substrates include CcdA from F plasmid, relBP307 and PasA [Van96a, Van94, Gronlund99, Smith98]. Lon proteolyzes chromosomal toxin/antoxin pairs as well, including RelB and YoeB [Christense01, Christense04]. This degradation of chromosomal pairs may regulate part of the starvation stress response, as the breakdown of RelB leaves RelE, which suppresses translation [Christense01]. Starvation-induced transcription of chpA also depends on Lon [Christense03].

Lon is an ATP-dependent protease with chymotrypsin-like specificity based on a Serine (679)-Lysine (722) dyad [Charette81, Waxman85, Botos04, Nishii05]. Lon has one proteolytic and four ATP-binding sites, two high affinity, the other two low affinity [Chin88, Menon87]. Detailed kinetic analysis shows that the two types of ATP sites use ATP at different rates as well [Vineyard06]. Lon's protease function depends on its ATPase activity; both require Mg2+, two ATP molecules are used per peptide bond hydrolyzed and loss of ATPase functions leads to concomitant loss of peptidase function [Menon87, Menon87a, Fischer94, Waxman82]. ATPase activity continues in mutants that are unable to proteolyze [Pohl76]. Though its ATPase activity is required for protein degradation, Lon is capable of breaking down small peptides in the absence of ATP or ATPase function [Goldberg85, Rasulova98]. The isolated ATPase domain undergoes conformational change in response to ADP and ATP binding [Vasilyeva02].

Lon has an independent protein-binding domain in addition to its proteolytic domain. This domain binds unfolded proteins [Chin88]. Protein binding substantially stimulates peptide degradation and ATPase activity, the latter even in mutants incapable of peptidase function [Waxman86, Pohl76].

Lon binds DNA via its DNA-binding domain [Charette81, Chin88]. Addition of DNA, especially denatured DNA, stimulates substrate proteolysis in vitro, as well as stimulating ATPase activity even in the absence of substrate [Chung82, Charette84]. Lon may have specificity for promoter regions, explaining how it targets regulatory proteins [Fu97].

Lon can form a complex with inorganic polyphosphate, allowing subsequent degradation of ribosomal proteins, including S2, L9 and L13 [Kuroda01]. Lon's DNA-binding domain binds polyphosphate with greater affinity than DNA [Nomura04]. Lon complexed with polyphosphate may be an octamer instead of a tetramer [Nishii05].