Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Essential <i>Bacillus</i> <i>subtilis</i> genes

Essential Bacillus subtilis genes Essential Bacillus subtilis genes a b,c d d e f g h i K. Kobayashi , S. D. Ehrlich , A. Albertini , G. Amati , K. K. Andersen , M. Arnaud , K. Asai , S. Ashikaga , S. Aymerich , j k l m n b o p P. Bessieres , F. Boland , S. C. Brignell , S. Bron , K. Bunai , J. Chapuis , L. C. Christiansen , A. Danchin , f b q e e p r i ´ ´ M. Debarbouille , E. Dervyn , E. Deuerling , K. Devine , S. K. Devine , O. Dreesen , J. Errington , S. Fillinger , k s d f m t u l v S. J. Foster , Y. Fujita , A. Galizzi , R. Gardan , C. Eschevins , T. Fukushima , K. Haga , C. R. Harwood , M. Hecker , w p n x a h h y z D. Hosoya , M. F. Hullo , H. Kakeshita , D. Karamata , Y. Kasahara , F. Kawamura , K. Koga , P. Koski , R. Kuwana , w w t s i aa x m bb D. Imamura , M. Ishimaru , S. Ishikawa , I. Ishio , D.LeCoq , A. Masson , C. Mauel , R. Meima , R. P. Mellado , k a s h a o cc q e A. Moir , S. Moriya , E. Nagakawa , H. Nanamiya , S. Nakai , P. Nygaard , M. Ogura , T. Ohanan , M. O’Reilly , k l x f r bb x u g M. O’Rourke , Z. Pragai , H. M. Pooley , G. Rapoport , J. P. Rawlins , L. A. Rivas , C. Rivolta , A. Sadaie , Y. Sadaie , y w o e q aa t p M. Sarvas , T. Sato , H. H. Saxild , E. Scanlan , W. Schumann , J. F. M. L. Seegers , J. Sekiguchi , A. Sekowska , aa dd dd x z cc w r S. J. Seror , M. Simon , P. Stragier , R. Studer , H. Takamatsu , T. Tanaka , M. Takeuchi , H. B. Thomaides , b m z l t s s n ee V. Vagner , J. M. van Dijl , K. Watabe , A. Wipat , H. Yamamoto , M. Yamamoto , Y. Yamamoto , K. Yamane , K. Yata , s u v a K. Yoshida , H. Yoshikawa , U. Zuber , and N. Ogasawara a b Graduate School of Information Science, Nara Institute of Science and Technology, Nara 630-0101, Japan; Ge´ne ´ tique Microbienne, Institut National de la d e Recherche Agronomique, 78530 Jouy en Josas, France; Genetica e Microbiologia, Universita ` di Pavia, 1 via Ferrata, 27100 Pavia, Italy; Genetics, Smurfit f g Institute, Trinity College, Dublin 2, Ireland; Biochimie Microbienne, Institut Pasteur, 25 Rue du Dr. Roux, 75015 Paris, France; Faculty of Science, Saitama h i University, Saitama 338-8570, Japan; College of Science, Rikkyo (St. Paul’s) University, Tokyo 171-8501, Japan; Ge´ne ´ tique Mole ´ culaire et Cellulaire, Institut National de la Recherche Agronomique–Centre National de la Recherche Scientifique–Institut National Agronomique Paris-Grignon, 78850 Thiverval-Grignon, France; Mathe ´ matiques Informatique Ge ´ nomes, Institut National de la Recherche Agronomique, 78530 Jouy en Josas, France; k l Molecular Biology and Biotechnology, University of Sheffield, Sheffield S10 2TN, United Kingdom; Cell and Molecular Bioscience, Newcastle University Medical School, Framlington Place, Newcastle upon Tyne NE2 4HH, United Kingdom; Genetics, Groningen Biomolecular Sciences and Biotechnology n o Institute, 9750 AA, Haren, The Netherlands; Institute of Biological Sciences, University of Tsukuba, Ibaraki 305-8572, Japan; Biological Chemistry, Institute of Molecular Biology, Solvgade 83, 1307 K, Copenhagen, Denmark; Genetique des Genomes Bacteriens, Institut Pasteur, Unite ´ de Recherche Associe´e, q r Centre National de la Recherche Scientifique 2171, 75015 Paris, France; Institute of Genetics, Bayreuth University, D-95440 Bayreuth, Germany; Sir William Dunn School of Pathology, Oxford University, Oxford OX1 3RE, United Kingdom; Faculty of Life Science and Biotechnology, Fukuyama University, t u Hiroshima 729-0292, Japan; Faculty of Textile Science and Technology, Shinshu University, Nagano 386-8564, Japan; Department of Bioscience, Tokyo v w University of Agriculture, Tokyo 156-8502, Japan; Institute for Microbiology, Ernst-Moritz-Arndt-University, D-17487 Greifswald, Germany; Department of International Environmental and Agricultural Science, Tokyo University of Agriculture and Technology, Tokyo 183-8509, Japan; Institut de Genetique y z et de Biologie Microbiennes, CH-1005 Lausanne, Switzerland; National Public Health Institute, 00300, Helsinki, Finland; Faculty of Pharmaceutical Sciences, aa Setsunan University, Osaka 573-0101, Japan; Institut de Ge´ne ´ tique et Microbiologie, Centre National de la Recherche Scientifique Unite ´ Mixte de bb Recherche 8621, Universite ´ Paris-Sud, 91405 Orsay Cedex, France; Centro Nacional de Biotecnologı´a, Campus de la Universidad Auto ´ noma, cc dd Cantoblanco, 28049 Madrid, Spain; School of Marine Science and Technology, University of Tokai, Shizuoka 424-8610, Japan; Institut de ee Biologie Physico-Chimique, 75005 Paris, France; and Radioisotope Center, National Institute of Genetics, Shizuoka 411-8540, Japan Communicated by Richard M. Losick, Harvard University, Cambridge, MA, January 27, 2003 (received for review November 10, 2002) To estimate the minimal gene set required to sustain bacterial life M. genitalium and Haemophilus influenzae, led to a description of in nutritious conditions, we carried out a systematic inactivation of a smaller set of some 260 genes (2). More recently, an experi- Bacillus subtilis genes. Among 4,100 genes of the organism, only mental approach involving high-density transposon mutagenesis 192 were shown to be indispensable by this or previous work. of the H. influenzae genome led to a much higher estimate of Another 79 genes were predicted to be essential. The vast majority 670 putative essential genes (3), whereas transposon mutagen- of essential genes were categorized in relatively few domains of esis of two mycoplasma species led to an estimate of 265–360 cell metabolism, with about half involved in information process- essential genes (4). Another experimental approach using anti- sense RNA to inhibit gene expression led to the identification of ing, one-fifth involved in the synthesis of cell envelope and the some 150 essential genes in Staphylococcus aureus (5). However, determination of cell shape and division, and one-tenth related to these approaches have limitations. Computation is likely to cell energetics. Only 4% of essential genes encode unknown underestimate the minimal gene set because it takes into account functions. Most essential genes are present throughout a wide only those genes that have remained similar enough during the range of Bacteria, and almost 70% can also be found in Archaea course of evolution to be recognized as true orthologues. and Eucarya. However, essential genes related to cell envelope, Transposon mutagenesis might overestimate the set by misclas- shape, division, and respiration tend to be lost from bacteria with sification of nonessential genes that slow down the growth small genomes. Unexpectedly, most genes involved in the Emb- without arresting it but can also miss essential genes that tolerate den–Meyerhof–Parnas pathway are essential. Identification of un- transposon insertions (3, 6). Finally, the use of antisense RNA known and unexpected essential genes opens research avenues to is limited to the genes for which an adequate expression of the better understanding of processes that sustain bacterial life. inhibitor y RNA can be obtained in the organism under study. To obtain an independent and possibly more reliable estimate he definition of the minimal gene set required to sustain a of a minimal protein-encoding gene set for bacteria, we system- T living cell is of considerable interest. The functions specified atically inactivated Bacillus subtilis genes. B. subtilis was chosen by such a set are likely to provide a view of a ‘‘minimal’’ bacterial because it is one of the best studied bacteria (7) and is a model cell. Many functions should be essential in all cells and could be for low-GC Gram-positive bacteria, which include both deadly considered as a foundation of life itself. The determination of the pathogens, such as Bacillus anthracis, and bacteria widely used in range of essential functions in different cells should reveal food and industr y, such as lactococci and bacilli. Because the possible solutions for sustaining life. Computational and exper- essentiality of a gene depends on the conditions under which the imental research has previously been carried out to define a organism is propagated, we used an environment likely to be minimal protein-encoding gene set. An upper-limit estimate of optimal for B. subtilis and thus carried out inactivation on a a minimal bacterial gene set was obtained from the sequence of the entire Mycoplasma genitalium genome, which contains only 480 genes (1). A computational approach, based on the To whom correspondence should be addressed. E-mail: ehrlich@jouy.inra.fr and assumption that essential genes are conser ved in the genomes of ehrlich@is.aist-nara.ac.jp. 4678 – 4683  PNAS  April 15, 2003  vol. 100  no. 8 www.pnas.orgcgidoi10.1073pnas.0730515100 Table 1. Essential and nonessential B. subtilis genes Table 2. B. subtilis essential genes Essential Nonessential Total DNA metabolism 27 Basic replication machinery 16 This study* 150 2,807 2,957 Packaging and segregation 9 Previous studies 42 614 656 Methylation 2 Prediction 79 106 185 RNA metabolism 14 Phage genes 0 303 303 Basic transcription machinery 4 Total 271 (6.6%) 3,830 (94.4%) 4,101 RNA modification 6 A list of the genes and their classifications can be accessed at http: Regulation 4 bacillus.genome.ad.jp. Protein synthesis 95 *We included 18 essential genes here that were inactivated in the course of Ribosomal proteins 52 this study and also studied previously. Aminoacyl-tRNA synthetases 24 Carried out in B. subtilis. Translation factors 10 Full list is presented as Table 3. Protein folding and modification 3 Excluded are four genes that were not studied because of technical reasons Protein translocation 6 (too short for insertional inactivation and too inconveniently placed for Cell envelope 44 chloramphenicol replacement). Membrane lipids 16 Cell wall 28 standard laborator y rich medium at 37°C. This choice also Cell shape and division 10 allowed for a comparison of the results obtained in many Glycolysis 8 laboratories and many previous studies, nevertheless leaving Respiratory pathways 22 open the possibility that a different gene set is essential under Isoprenoids 8 Menaquinone 8 different growth conditions. Analysis of the mutants, in con- Cytochrome biogenesis 3 junction with the literature data, leads us to conclude that there Thioredoxin 3 are only 271 genes indispensable for growth in LB when inac- Nucleotides 10 tivated singly. These fall into a relatively few large domains of cell Cofactors 15 physiology and are ver y broadly conser ved in microorganisms. CoA 1 Methods Folate 3 NAD 4 The approach used for gene inactivation has been described (8). S-Adenosylmethionine 1 Brief ly, it involved insertion of a nonreplicating plasmid into the Iron–sulfur cluster 6 target gene via a single crossover recombination. The expression Other 15 of the downstream genes from the same operon was controlled Unknown 11 by an isopropyl -D-thiogalactoside (IP TG)-regulated promoter Total 271 present on the inserted plasmid. A gene was deemed essential if it could not be inactivated by insertion (i.e., no transformants A complete list of genes and the evidence used to ascertain their essential were obtained when competent recipient cells were mixed with nature are presented in Table 4. the insertional plasmid) and if the strain became IP TG depen- dent when an intact copy of the gene was placed under control 106 are not (Table 3). We inactivated all but 4 of the remaining of the regulated promoter (8). IP TG-dependent strains could genes and found that 150 are essential. This analysis leads us to not be constructed for six essential genes, possibly because the conclude that there are 271 genes indispensable for growth when regulated promoter was either not strong enough or not suffi- ciently tuned to provide appropriate gene expression levels. An inactivated singly (Table 1). For 96% of these, we propose alternative strategy was followed for 160 genes shorter than assignment to various domains of cell metabolism (Table 2; the 300 bp, where insertional inactivation was limited by the insuf- complete list of genes is given in Table 4, which is published as ficient gene length. These genes were replaced by a chloram- supporting information on the PNAS web site). phenicol resistance marker, and if replacement failed they were Functional Assignment of Essential Genes. Information processing. rendered IP TG-dependent. All mutations were made in the About half of the essential genes are involved in DNA and RNA standard laborator y strain 168. Inactivation was not attempted metabolism and protein synthesis. Sixteen genes encode the for 656 genes studied previously in B. subtilis, and 185 genes basic DNA replication machiner y. They comprise five genes having a high degree of similarity with genes well characterized in other bacteria or involved in well characterized processes, for involved in the initiation of replication (dnaA, B, D, and I, and which we could predict essentiality with confidence (Table 3, priA), eight genes encoding components of the replisome (dnaC, which is published as supporting information on the PNAS web E, G, N, and X, holA and B, and polC), DNA ligase, and the Ssb site, www.pnas.org). Complete microbial genomes included in protein. One gene, pcrA, has no clearly identified role, but could the Microbial Genome Database for Comparative Analysis be involved in the progression of the replication fork (10). Among genes involved in DNA packaging and segregation, five (http:mbgd.genome.ad.jp), comprising 54 bacteria, 16 ar- encode topoisomerases (topA, gyrA and B, and parD and E), one chaea, and 2 yeasts, were analyzed for the presence of the B. subtilis essential gene homologs by using the default parameters, encodes the general DNA-binding protein Hbsu, and three with 10 as a cut-off value. encode the proteins that act in the condensation of the nucleoid (smc, and scpA and B; ref. 11). The remaining two genes encode Results modification methylases, expected to be essential unless the There are 4,100 annotated genes in the B. subtilis genome (9). cognate nucleases are inactivated. Some 303 are encoded on prophages that can be eliminated from Among 14 essential genes involved in RNA metabolism, four the genome and are not essential. Previous studies on 656 B. (rpoA, B, and C, and sigA) encode components of the basic transcription machiner y, whereas six are involved in RNA subtilis genes identified 42 that are essential (Table 1). Through predictions we propose that 79 other genes are essential, whereas modification. rnc and rnpA encode RNases, cspR and trmD and Kobayashi et al. PNAS  April 15, 2003  vol. 100  no. 8  4679 GENETICS U encode methylases, and cca encodes tRNA nucleotidyl trans- appears to be required for both fatty acid and phospholipid ferase. Only four genes are involved in regulation of RNA biosynthesis in a way that is not well understood (19). synthesis: a two-component system yycF and G (12), a gene Synthesis of peptidoglycan, the main component of the cell involved in the coupling between translation and termination of wall, comprises two stages, the synthesis of the precursor mol- RNA synthesis, nusA (13), and an anti-sigma factor, YhdL (14). ecules and the polymerization of peptidoglycan (20). All of the The largest categor y, comprising 95 essential genes, is that essential genes are involved in the first stage, which encompasses involved in protein synthesis. Over half of the genes encode a variety of biosynthetic pathways: (i) Synthesis of aminosugars ribosomal proteins. Although there is no experimental evidence (Fig. 6, which is published as supporting information on the that they are essential in B. subtilis, we suggest that they belong PNAS web site) by conversion of fructose-6-phosphate to UDP- to the essential set, because the ribosome itself is essential. This N-acetyl-glucosamine and UDP-N-acetyl-manosamine. The first suggestion is supported by the obser vation that the inhibition of two steps, leading to glucosamine-1-phosphate, are catalyzed by synthesis of 21 different ribosomal proteins is lethal in S. aureus the products of glmS and ybbT genes. The last two steps are (5). Among these are proteins such as L24, which was not carried out by the products of the gcaD and yvyH. More than one absolutely essential in E. coli, but cells that lacked it grew ver y gene product seems to be able to acetylate glucosamine-1- slowly and were thermosensitive (15). We suggest that there are phosphate, because there is no single essential gene for this step. 20 essential genes that encode aminoacyl-tRNA synthetases, (ii) Diaminopimelate (Fig. 7, which is published as supporting corresponding to 18 amino acids. All but two are present in information on the PNAS web site) is synthesized from L- unique copies. We showed that one of the unique copy genes, aspartate by eight successive reactions, six of which are carried lysS, is essential and assumed that others are too, without seeking out by products of essential genes asd, dapA, B, and F, and ykuQ further experimental evidence. There are two genes encoding and R. The first and the fifth step can be catalyzed by products tRNA-Tyr and tRNA-Thr synthetases. Only t yrS was essential of three (dapG, lysC, and yclM) and two genes (mtnV and ywfG), when inactivated singly whereas either thrS or thrZ could assure respectively; thus, none of the five is essential if inactivated the viability. We grouped with the synthetases three genes that singly. (iii) Two essential genes, racE and alr, encode racemases are required for the conversion of the tRNA-Glu to tRNA-Gln that convert L-glutamate and L-alanine into the corresponding D (gatC, B, and A) and one gene that is required for the formylation isomers. racE cannot be replaced by a homologue, yrpC. The of methionyl tRNA ( fmt). Of the10 essential genes involved in essential ddl gene is required for synthesis of the dipeptide mRNA translation, 3 are required for initiation (infA, B, and C), D-Ala-D-Ala. (iv) Eight essential genes, murAA, murB, C, D, E, 3 are required for elongation (tufA, tsf, and fusA), and 4 are F, and G, and mraY, are required for synthesis of the lipid-linked required for termination and ribosome recycling (prfA and B, disaccharide-pentapeptide peptidoglycan precursor (Fig. 8, pth, and frr). There is one essential gene involved in posttrans- which is published as supporting information on the PNAS web lation modification, map, that encodes methionine aminopepti- site) from UDP-N-acetyl-glucosamine, phosphoenolpyruvate, dase. Deformylation is also required, but can be carried out by D-glutamine, diaminopimelate, D-ala dipeptide, and an isopre- products of two genes, def and ykrB, neither of which is essential nylphosphate. Polymerization of peptidoglycan is carried out by when inactivated singly (16). Two essential genes, groEL and ES, the products of functionally redundant genes in B. subtilis. The are involved in protein folding. Finally, there are six essential cell wall of B. subtilis contains teichoic acid (21), and there are genes that encode key components of the machiner y for protein seven essential genes involved in its synthesis. Four, tagA, B, D, insertion into the membrane and secretion. These include the and O, are required for the synthesis of linkage units and three, targeting factors Ffh and FtsY, the translocation motor SecA, tagF, G, and H, are required for chain polymerization, translo- two components of the translocation channel, SecY and E, and cation, and linkage to peptidoglycan (Fig. 9, which is published the folding catalyst PrsA. The essential DNA-binding protein as supporting information on the PNAS web site). Hbsu is also a part of the signal recognition particle (17). Ten essential genes are involved in cell shape and division. Cell envelope, shape, and division. About one-fifth of the essential Septum formation requires seven ( ftsA, L, W, and Z, divIB and genes are required for these processes (Table 2). The synthesis C, and pbpB; ref. 21), whereas cell shape requires three (rodA, of the cell envelope involves 44 essential genes, all required for and mreB and C). membrane and cell wall formation. Membrane lipids, phospho- Embden–Meyerhof–Parnas (EMP) pathway and respiration. About 10% lipids, and glycolipids are synthesized from fatty acids. Fatty acid of essential genes, which have in common the provision of energy synthesis (Fig. 4, which is published as supporting information on for the cell, are required for these processes. A majority of genes the PNAS web site) is initiated by products of four genes, accA, composing the ubiquitous EMP pathway are essential (Fig. 10, B, C, and D, together with acpA and fabD gene products. acpS which is published as supporting information on the PNAS web is required for the conversion of AcpA from the apo to the holo site). The process can be viewed as consisting of two parts: the form, whereas birA is required for the addition of a biotinyl top, which converts hexose sugars to trioses, and the bottom, group to carboxylase. The fatty acid chains are elongated by the which converts these compounds to pyruvate, funneled into products of two essential genes, fabFG. The elongation cycle pyruvate dehydrogenase. The top part comprises four steps when involves two additional steps that are catalyzed by pairs of genes glucose is the carbon source, the last two of which are catalyzed with overlapping functions (ycsD and ywpB, and fabI and L), by products of essential genes pfkA and fbaA, whereas the bottom none of which is essential when inactivated singly (18). Two of part comprises six steps, four of which are encoded by essential the essential genes required for phospholipid synthesis (Fig. 5, genes tpiA, pgk, pgm, and eno. The two remaining essential genes which is published as supporting information on the PNAS web related to glycolysis are tkt and prs. The first encodes a transke- site), gpsA and yhdO, are involved in the conversion of dihy- tolase, involved in the pentose pathway, whereas the second gene droxyacetone phosphate to phosphatidic acid, which is a pre- codes for a pyrophosphokinase that converts ribose-5-phosphate cursor of complex lipids. Interestingly, yerQ, which encodes an to 5-phospho-ribose-1-diphosphate, a common precursor of enzyme with a diacylglycerol kinase catalytic domain found in nucleotides and cofactors, such as NAD, which likely accounts eukar yotes and presumably catalyses synthesis of phosphatidic for its essential role. Taken together, these results are rather acid from another precursor (diacylglycerol), is also essential, unexpected. First, our experiments were carried out on a rich whereas a homologue, dgkA, is not. Two essential genes, cdsA medium, which contains numerous compounds that could pro- and pgsA, are required for synthesis of phosphatidylglycerol vide the energy and building blocks for cell life, the two known phosphate, which might be converted into phosphoglycerol by a functions of the EMP pathway. Addition of glucose to LB did not nonspecific phosphatase. The remaining essential gene, plsX, restore growth of any of the nonviable EMP mutants. Second, in B. 4680  www.pnas.orgcgidoi10.1073pnas.0730515100 Kobayashi et al. subtilis a part of the EMP pathway can be bypassed via the pentose is published as supporting information on the PNAS web site), shunt, and it is surprising that both are simultaneously required for and only the four genes involved in the salvage pathway (yueK, viability. Possibly, the enzymes revealed as essential have novel and yqeJ, nadE, and yjbN) were essential. We speculate that the unexpected functions in the cell. It should be noted that pgm and accumulation of nicotinate might repress de novo synthesis of eno mutants have been isolated previously and had very slow nicotine mononucleotide in the absence of yueK, rendering this growth (22), suggesting that the difference between lethal and gene essential. There are three essential genes involved in folate almost-lethal mutation can be due to subtle differences in the metabolism (Fig. 16, which is published as supporting information experimental conditions and the strain background. on the PNAS web site). One, dfrA, codes for dihydrofolate reduc- Respiration can provide energy for the cell, in the absence of tase, which converts folate, presumably imported from the medium, glycolysis. We identified 22 essential genes involved in this to tetrahydrofolate. Two other genes, glyA and folD, are required for process. Under the aerobic condition used in our experiments, conversion of the latter compound to 10-formyl tetrahydrofolate, a respiration involves the transfer of electrons by various dehy- one-carbon donor molecule for a number of reactions. S- drogenases to menaquinone and then to cytochromes (23). adenosylmethionine (SAM) is another one-carbon donor, synthe- Menaquinone is synthesized from chorismate in seven steps, the sized from ATP and methionine by SAM synthetase, encoded by last six of which are catalyzed by products of essential genes, the essential metK gene. There is only one essential gene involved menA, B, C, D, E, and H (Fig. 11, which is published as in the biosynthesis of CoA, ytaG, that is required for the last step supporting information on the PNAS web site). Two genes, menF in the pathway (Fig. 17, which is published as supporting informa- and dhbC, appear to be able to catalyze the first step, and neither tion on the PNAS web site), suggesting that the precursor, dephos- is essential if inactivated singly. The penultimate step involves pho-CoA, is transported from the medium into the cell. The condensation of dihydroxynaphthoic acid with an isoprenoid remaining cofactor is an iron–sulfur cluster, which forms part of biphosphate. Isoprenoids (Fig. 12, which is published as sup- proteins that participate in many aspects of the cell physiology, porting information on the PNAS web site) are synthesized from including redox and nonredox catalysis, as well as sensing for pyruvate and glyceraldehyde-3-phosphate by a nonmevalonate regulatory processes. There are five essential genes, yurU, V, W, X, pathway in B. subtilis. The first six steps, leading to isopentenyl and Z, involved in the synthesis of this cluster. We included here diphosphate, involve seven essential genes, dxs, dxr, ispE, yacM yrvO, a homologue of yurV. and N, and yqfP and Y. Three other essential genes, hepS and T Other processes. Only 15 essential genes that have a clear bio- and yqiD, are required for the synthesis of farnesyl diphosphate chemical function were not associated with any of the large and more complex compounds that are used for menaquinone domains of cellular physiology discussed above. Among these are synthesis. Altogether, of 22 essential genes involved in respira- six GT P-binding proteins of the EraObg family. Only one, obg, tion, 16 are required for menaquinone synthesis. There are only has been studied previously in B. subtilis and been shown to three essential genes involved in cytochrome biogenesis, resA, B, affect the stress response mediated by  . Five other genes, and C. No cytochrome structural genes are essential, possibly mrpA, B, C, D and F, encode a sodium– hydrogen antiporter, ref lecting overlapping functions of their products (24). We have which is required to maintain pH homeostasis in the presence of included trxA and B, which encode thioredoxin and thioredoxin sodium chloride concentrations similar to those found in LB reductase with the respiration genes, because of the role of TrxA (27). ppaC encodes the inorganic pyrophosphatase, which drives in electron transport, although this protein is involved in many the anabolic f luxes by pyrophosphate hydrolysis in various other oxido-reduction reactions. We also included here a puta- biochemical reactions, whereas gcp encodes a sialopeptidase of tive thioredoxin reductase gene, yumC. Nucleotides and cofactors. Metabolism of these compounds requires 10% of the essential genes (Table 2). The metabolism of nucleotides is quite complex, comprising complementar y de novo synthesis and salvage pathways (25). Nevertheless, we found 10 essential genes involved in this process. Among the four that participate in purine metabolism (Fig. 13, which is published as supporting information on the PNAS web site), two (adk and gmk) specify kinases, which phosphor ylate A MP or GMP to the respective diphosphates. Absence of guanine from the medium accounts for the essential nature of guaB. Surprisingly, hprT,a gene from the purine salvage, is also essential, raising a possi- bility that its product has a second, unsuspected role in the cell. Two essential genes involved in pyrimidine metabolism (Fig. 14, which is published as supporting information on the PNAS web site), cmk and tmk, also encode kinases that phosphor ylate CMP and TMP to corresponding diphosphates. The remaining essen- tial gene, pyrG, encodes cytidylate synthetase, which converts UT P into CT P. This might ref lect the paucity of cytidine in the rich medium. Interestingly, two B. subtilis essential genes encode enzymes present in the E. coli degradosome [yjbN (ppnK) and eno, a member of the EMP pathway], which provides CDP for DNA synthesis and further nucleotide metabolism, while con- trolling mRNA turnover (26). Finally, there are three essential Fig. 1. B. subtilis essential gene homologues are widely conserved. (Upper) genes involved simultaneously in purine and pyrimidine metab- Genes are ordered by their relative abundance among 54 Bacteria (blue) and 18 olism, nrdE and F and ymaA, that encode subunits of nucleoside- Archaea and Eucarya (red). The position (rank) of a gene is shown on abscissa and diphosphate reductase, which converts the ribose into deoxyri- the fraction of organisms in which a gene is present is shown on the ordinate. bose derivatives. (Lower) Fraction of genes present in different kingdoms of life (a gene counted Synthesis of only five cofactors, involving 16 genes, was as ‘‘all kingdoms’’ is present in at least one archaeon and one eukaryote, in required under our experimental conditions. NAD synthesis can addition to bacteria, whereas a gene counted as ‘‘bacteria’’ is not present in any take place de novo or by salvaging of precursors (Fig. 15, which archae or eukaryote). The list of genes and organisms is presented in Table 4. Kobayashi et al. PNAS  April 15, 2003  vol. 100  no. 8  4681 GENETICS Fig. 2. The number of B. subtilis essential gene homologues depends on genome size. (Top) All genes. Bacilli and close relatives denote Bacillus species and other low-GC Gram-positive bacteria, but not clostridia, mycoplasma, and ureaplasma. (Middle and Bottom) Different bacterial gene categories. Empty red circles in Bottom refer to Bacilli and close relatives, whereas filled red circles refer Fig. 3. Phylogenetic profiling of essential genes. The 271 B. subtilis genes to other bacteria. Interpolated lines throughout the figure correspond to the best were grouped in 266 clusters. Only one gene, yhdL, which encodes a possible fitting polynomial of the second or the fourth order. The number of genes is: anti-sigma protein, had no orthologues in the database and is not presented information processing, 136; envelope, respiration, cell shape, and division, 76; here. Each line and column corresponds to individual gene and organism, cofactors, other, and unknown, 41; and nucleotides and glycolysis, 18. respectively. Presence and absence of a gene is indicated by a black and white square, respectively. The list of genes and organisms is given in Table 5, which is published as supporting information on the PNAS web site and the ordering unknown role. The last two genes, pdhA and odh, encode is described in the text. subunits of pyruvate and 2-oxoglutarate dehydrogenase, respec- tively; growth of the mutants could be restored by addition to LB of the metabolites (acetate and succinate, respectively) related 3 Mb (highlighted in red). Other bacteria with genomes of a to the activity of the proteins they encode. similar size have, on average, slightly 80% of the B. subtilis Unknown. The last categor y groups 11 essential genes for which we essential gene homologues. This proportion drops to 57% with were unable to suggest a role in cell physiology. Biochemical decreasing bacterial genome size, indicating progressive loss of functions, a protease and a hydrolase of the metallo--lactamase essential genes. Archaea and Eucar ya maintain, on average, 36% superfamily, can be suggested for products of two gene, ydiC and of the essential gene homologues, with the proportion var ying ykqC. One gene, yneS, encodes a putative membrane protein, between 33% and 44% almost linearly with genome size. In and another, ymdA, encodes a protein with an HD domain of bacteria, gene loss occurs mainly in three categories (cell enve- metal-dependent phosphohydrolases, whereas three, yloQ, yqjK, lope, shape and division, and respirator y pathways) and to a and ywlC, encode proteins with recognizable signatures, an lower extent in three other categories (cofactor synthesis, other AT PGT P-binding site, a metallo--lactamase motif, and a processes, and unknown functions). In contrast, information putative RNA-binding motif, respectively. Four genes, yacA, processing, glycolysis, and nucleotide synthesis genes are largely ydiB, ylaN, and yqeI, have no easily recognizable features. retained (Fig. 2 Middle and Bottom). Phylogenetic profiling of essential B. subtilis genes is summarized Conservation of Essential Genes. The average level at which ho- in Fig. 3. Organisms were grouped into four classes and ordered mologues of essential B. subtilis genes are present in bacteria is within each class on the basis of the number of essential gene rather high (approaching 80%), one-fourth being found in all homologues they share with B. subtilis, placing the organisms with bacteria and three-fourths in at least 75% (Fig. 1 Upper). The fewest conserved genes at the right of each class. Genes were average is 36% in Eucar ya and Archaea, but some 20% of the grouped in categories and ordered by abundance among all bac- genes are nevertheless present in all 18 organisms we analyzed teria, which placed the less abundant genes at the bottom of each (Fig. 1 Upper). About one-third of the genes are found in all three category. A number of general features are easily discernible from kingdoms of life, and a further one-third are shared between this analysis. (i) The five top categories are composed of genes Bacteria and either Archaea or Eucar ya (Fig. 1 Lower). present in 80% of Bacteria and at least 40% of Eucarya and The number of B. subtilis essential gene homologues present Archaea, with the exception of RNA synthesis, which is less well in an organism depends on at least two parameters: phylogenetic conserved in the last two kingdoms. (ii) The next two categories, proximity to B. subtilis and genome size (Fig. 2 Top). The highest number is found in bacilli and close relatives, having genomes of DNA metabolism and cell shape and division, contain genes 4682  www.pnas.orgcgidoi10.1073pnas.0730515100 Kobayashi et al. present in most bacteria and genes specific for Gram-positive bacteria with smaller genomes. Concomitantly, genes involved in organisms. This can most easily be seen from the appearance of the the determination of cell shape, division, and respiration are also lost. This suggests that it may be possible to build, maintain, and relatively broad horizontal white bars at the bottom of the two reproduce the cell compartment in a simpler way than that used classes. (iii) The categories that contain genes missing from bacteria by bacteria with larger genomes, and that glycolysis can be with small genomes are easily identified by the presence of the sufficient to generate energy for the cell. A minimal essential vertical white band at the right of the low-GC Gram-positive gene set could thus be significantly smaller than the one present bacteria class, corresponding to Mycoplasma and Ureaplasma urea- in bacteria with genomes larger than 3 Mb. lyticum. In addition, there is an enlargement of the white zone at the right end of the ‘‘Other bacteria’’ class, noticeable for cell envelope, Unexpected Essential Genes. Notwithstanding the grouping of respiration, and unknown functions. (iv) Genes in the last two most essential functions in a few large categories, our study has categories, unknown and other, although often found only in the revealed genes that were not expected to have an essential closest relatives of B. subtilis, are nevertheless present in over a half function under the experimental conditions used, such as eight of other bacteria. EMP pathway genes and a gene involved in purine biosynthesis. These obser vations suggest previously unsuspected links be- Discussion tween different domains of cell physiology. A Simple Bacterial Cell. Of some 4,100 genes of B. subtilis, only 271 are essential for growth under our experimental conditions when Redundant Genes for Essential Functions. Our analysis does not inactivated singly. About 80% of the functions they encode fall detect essential functions encoded by redundant genes, because in a few large categories; namely, information processing, cell only a single gene was inactivated in each mutant strain. The list envelope, shape, division, and energetics. These obser vations of the essential genes given here is thus likely to be underesti- lead to a view of a rather simple bacterial cell, consisting of a mated, because synthetic lethal mutants are well known. A rigorous detection of the missing functions would require the compartment, formed by a membrane and a wall, enclosing the systematic combination of all of the mutations in a single strain, elements necessar y to synthesize proteins that carr y out reac- which is beyond the present genetic technology. However, it is tions required for (i) the duplication and inheritance of the remarkable that single gene inactivation did reveal large cate- genetic information; (ii) the division of the compartment; and gories of essential functions, suggesting that most of the vital cell (iii) the provision of energy. These processes do not appear to be processes are encoded by nonredundant genes. The presence of coordinated by modulation of gene expression, because the expres- paralogues for 50% of B. subtilis genes (9) might thus allow the sion regulators are by and large not essential. We suggest that the cell to respond to changing environmental conditions rather than coordination might be carried out, at least in part, by the essential provide back-up for vital processes. GTP-binding proteins, as appears to be the case in eukaryotes. Isogenic Mutant Collection. Finally, it should be noted that the Broad Distribution of Essential Genes and Functions. Over 80% of isogenic set of 3,000 mutants that we have generated can be essential B. subtilis gene homologues are present in all bacteria used to identify genes, and thus functions, that are essential with genomes above 3 Mb, and 57% are found even in bacteria under conditions different from those used here. Furthermore, with the smallest genomes (mycoplasma). Almost 70% of genes the mutant set is a unique bacterial resource for studying various are present in at least one kingdom other than Bacteria. Many phenotypes and may thus lead to deeper insight into the me- organisms thus appear to rely on a similar set of essential tabolism of the bacterial cell. functions, supporting the simple microbial cell view outlined above. The similarity might be even higher, because some of the This work was supported, in part, by European Union Grant BIO4- genes might have diverged beyond recognition and some func- CT95-0278 and a Grant-in-Aid for Scientific Research on Priority Areas tions can be encoded by unrelated genes (28). However, genes (C) ‘‘Genome Biology’’ from the Ministr y of Education, Culture, Sports, involved in the synthesis of the cell envelope tend to be lost from Science and Technology of Japan. 1. Fraser, C. M., Gocayne, J. D., White, O., Adams, M. D., Clayton, R. A., 16. Haas, M., Beyer, D., Gahlmann, R. & Freiberg, C. (2001) Microbiology 147, Fleischmann, R. D., Bult, C. J., Kerlavage, A. R., Sutton, G., Kelley, J. M., et 1783–1791. al. (1995) Science 270, 397– 403. 17. Nakamura, K., Yahagi, S., Yamazaki, T. & Yamane, K. (1999) J. Biol. Chem. 2. Mushegian, A. R. & Koonin, E. V. (1996) Proc. Natl. Acad. Sci. USA 93, 274, 13569 –13576. 10268 –10273. 18. Heath, R. J., Su, N., Murphy, C. K. & Rock, C. O. (2000) J. Biol. Chem. 275, 3. Akerley, B. J., Rubin, E. J., Novick, V. L., Amaya, K., Judson, N. & Mekalanos, 40128 – 40133. J. J. (2002) Proc. Natl. Acad. Sci. USA 99, 966 –971. 19. Morbidoni, H. R., de Mendoza, D. & Cronan, J. E., Jr. (1996) J. Bacteriol. 178, 4. Hutchison, C. A., Peterson, S. N., Gill, S. R., Cline, R. T., White, O., Fraser, 4794 – 4800. C. M., Smith, H. O. & Venter, J. C. (1999) Science 286, 2165–2169. 20. Foster, S. J. & Popham, D. L. (2002) in Bacillus subtilis and Its Closest Relatives: 5. Ji, Y., Zhang, B., Van Horn, S. F., Warren, P., Woodnutt, G., Burnham, M. K. From Genes to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, R. (Am. Soc. & Rosenberg, M. (2001) Science 293, 2266 –2269. Microbiol, Washington DC), pp. 21– 41. 6. Gerdes, S. Y., Scholle, M. D., D’Souza, M., Bernal, A., Baev, M. V., Farrell, 21. Errington, J. & Daniel, R. A. (2002) in Bacillus subtilis and Its Closest Relatives: M., Kurnasov, O. V., Daugherty, M. D., Mseeh, F., Polanuyer, B. M., et al. From Genes to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, R. (Am. Soc. (2002) J. Bacteriol. 184, 4555– 4572. Microbiol, Washington, DC), pp. 97–109. 7. Sonenshein, A. L., Hoch, J. A. & Losick, R., eds. (2002) Bacillus subtilis 22. Leyva-Vazquez, M. A. & Setlow, P. (1994) J. Bacteriol. 176, 2788 –2795. and Its Closest Relatives: From Genes to Cells (Am. Soc. Microbiol., Wash- 23. von Wachenfeldt, C. & Hederstadt, L. (2002) in Bacillus subtilis and Its Closest ington, DC). Relatives: From Genes to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, 8. Vagner, V., Der vyn, E. & Ehrlich, S. D. (1998) Microbiology 144, 3097–3104. R. (Am. Soc. Microbiol, Washington, DC), pp. 163–179. 9. Kunst, F., Ogasawara, N., Moszer, I., Albertini, A. M., Alloni, G., Azevedo, V., Bertero, 24. Winstedt, L. & von Wachenfeldt, C. (2000) J. Bacteriol. 182, 6557– 6564. M. G., Bessieres, P., Bolotin, A., Borchert, S., et al. (1997) Nature 390, 249–256. 25. Switzer, R. L. (2002) in Bacillus subtilis and Its Closest Relatives: From Genes 10. Petit, M. A. & Ehrlich, S. D. (2002) EMBO J. 21, 3137–3147. to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, R. (Am. Soc. Microbiol, 11. Soppa, J., Kobayashi, K., Noirot-Gros, M. F., Oesterhelt, D., Ehrlich, S. D., Washington, DC), pp. 255–269. Der vyn, E., Ogasawara, N. & Moriya, S. (2002) Mol. Microbiol. 45, 59 –71. 26. Nitschke ´, P., Guerdoux-Jamet, P., Chiapello, H., Faroux, G., Henaut, C., 12. Fabret, C. & Hoch, J. A. (1998) J. Bacteriol. 180, 6375– 6383. Henaut, A. & Danchin, A. (1998) FEMS Microbiol. Rev. 22, 207–227. 13. Ingham, C. J., Dennis, J. & Furneaux, P. A. (1999) Mol. Microbiol. 31, 651– 663. 27. Ito, M., Guffanti, A. A., Oudega, B. & Krulwich, T. A. (1999) J. Bacteriol. 181, 14. Horsburgh, M. J. & Moir, A. (1999) Mol. Microbiol. 32, 41–50. 2394 –2402. 15. Nishi, K., Dabbs, E. R. & Schnier, J. (1985) J. Bacteriol. 163, 890 – 894. 28. Koonin, E. V. (2000) Annu. Rev. Genomics Hum. Genet. 1, 99 –116. Kobayashi et al. PNAS  April 15, 2003  vol. 100  no. 8  4683 GENETICS http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Proceedings of the National Academy of Sciences Unpaywall

Essential <i>Bacillus</i> <i>subtilis</i> genes

Download PDF
6 pages

Loading...

Page 2

Loading...

Page 3

Loading...

Page 4

Loading...

Page 5

Loading...

Page 6

 
/lp/unpaywall/essential-i-bacillus-i-i-subtilis-i-genes-IV2bGZGZln

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0027-8424
DOI
10.1073/pnas.0730515100
Publisher site
See Article on Publisher Site

Abstract

Essential Bacillus subtilis genes a b,c d d e f g h i K. Kobayashi , S. D. Ehrlich , A. Albertini , G. Amati , K. K. Andersen , M. Arnaud , K. Asai , S. Ashikaga , S. Aymerich , j k l m n b o p P. Bessieres , F. Boland , S. C. Brignell , S. Bron , K. Bunai , J. Chapuis , L. C. Christiansen , A. Danchin , f b q e e p r i ´ ´ M. Debarbouille , E. Dervyn , E. Deuerling , K. Devine , S. K. Devine , O. Dreesen , J. Errington , S. Fillinger , k s d f m t u l v S. J. Foster , Y. Fujita , A. Galizzi , R. Gardan , C. Eschevins , T. Fukushima , K. Haga , C. R. Harwood , M. Hecker , w p n x a h h y z D. Hosoya , M. F. Hullo , H. Kakeshita , D. Karamata , Y. Kasahara , F. Kawamura , K. Koga , P. Koski , R. Kuwana , w w t s i aa x m bb D. Imamura , M. Ishimaru , S. Ishikawa , I. Ishio , D.LeCoq , A. Masson , C. Mauel , R. Meima , R. P. Mellado , k a s h a o cc q e A. Moir , S. Moriya , E. Nagakawa , H. Nanamiya , S. Nakai , P. Nygaard , M. Ogura , T. Ohanan , M. O’Reilly , k l x f r bb x u g M. O’Rourke , Z. Pragai , H. M. Pooley , G. Rapoport , J. P. Rawlins , L. A. Rivas , C. Rivolta , A. Sadaie , Y. Sadaie , y w o e q aa t p M. Sarvas , T. Sato , H. H. Saxild , E. Scanlan , W. Schumann , J. F. M. L. Seegers , J. Sekiguchi , A. Sekowska , aa dd dd x z cc w r S. J. Seror , M. Simon , P. Stragier , R. Studer , H. Takamatsu , T. Tanaka , M. Takeuchi , H. B. Thomaides , b m z l t s s n ee V. Vagner , J. M. van Dijl , K. Watabe , A. Wipat , H. Yamamoto , M. Yamamoto , Y. Yamamoto , K. Yamane , K. Yata , s u v a K. Yoshida , H. Yoshikawa , U. Zuber , and N. Ogasawara a b Graduate School of Information Science, Nara Institute of Science and Technology, Nara 630-0101, Japan; Ge´ne ´ tique Microbienne, Institut National de la d e Recherche Agronomique, 78530 Jouy en Josas, France; Genetica e Microbiologia, Universita ` di Pavia, 1 via Ferrata, 27100 Pavia, Italy; Genetics, Smurfit f g Institute, Trinity College, Dublin 2, Ireland; Biochimie Microbienne, Institut Pasteur, 25 Rue du Dr. Roux, 75015 Paris, France; Faculty of Science, Saitama h i University, Saitama 338-8570, Japan; College of Science, Rikkyo (St. Paul’s) University, Tokyo 171-8501, Japan; Ge´ne ´ tique Mole ´ culaire et Cellulaire, Institut National de la Recherche Agronomique–Centre National de la Recherche Scientifique–Institut National Agronomique Paris-Grignon, 78850 Thiverval-Grignon, France; Mathe ´ matiques Informatique Ge ´ nomes, Institut National de la Recherche Agronomique, 78530 Jouy en Josas, France; k l Molecular Biology and Biotechnology, University of Sheffield, Sheffield S10 2TN, United Kingdom; Cell and Molecular Bioscience, Newcastle University Medical School, Framlington Place, Newcastle upon Tyne NE2 4HH, United Kingdom; Genetics, Groningen Biomolecular Sciences and Biotechnology n o Institute, 9750 AA, Haren, The Netherlands; Institute of Biological Sciences, University of Tsukuba, Ibaraki 305-8572, Japan; Biological Chemistry, Institute of Molecular Biology, Solvgade 83, 1307 K, Copenhagen, Denmark; Genetique des Genomes Bacteriens, Institut Pasteur, Unite ´ de Recherche Associe´e, q r Centre National de la Recherche Scientifique 2171, 75015 Paris, France; Institute of Genetics, Bayreuth University, D-95440 Bayreuth, Germany; Sir William Dunn School of Pathology, Oxford University, Oxford OX1 3RE, United Kingdom; Faculty of Life Science and Biotechnology, Fukuyama University, t u Hiroshima 729-0292, Japan; Faculty of Textile Science and Technology, Shinshu University, Nagano 386-8564, Japan; Department of Bioscience, Tokyo v w University of Agriculture, Tokyo 156-8502, Japan; Institute for Microbiology, Ernst-Moritz-Arndt-University, D-17487 Greifswald, Germany; Department of International Environmental and Agricultural Science, Tokyo University of Agriculture and Technology, Tokyo 183-8509, Japan; Institut de Genetique y z et de Biologie Microbiennes, CH-1005 Lausanne, Switzerland; National Public Health Institute, 00300, Helsinki, Finland; Faculty of Pharmaceutical Sciences, aa Setsunan University, Osaka 573-0101, Japan; Institut de Ge´ne ´ tique et Microbiologie, Centre National de la Recherche Scientifique Unite ´ Mixte de bb Recherche 8621, Universite ´ Paris-Sud, 91405 Orsay Cedex, France; Centro Nacional de Biotecnologı´a, Campus de la Universidad Auto ´ noma, cc dd Cantoblanco, 28049 Madrid, Spain; School of Marine Science and Technology, University of Tokai, Shizuoka 424-8610, Japan; Institut de ee Biologie Physico-Chimique, 75005 Paris, France; and Radioisotope Center, National Institute of Genetics, Shizuoka 411-8540, Japan Communicated by Richard M. Losick, Harvard University, Cambridge, MA, January 27, 2003 (received for review November 10, 2002) To estimate the minimal gene set required to sustain bacterial life M. genitalium and Haemophilus influenzae, led to a description of in nutritious conditions, we carried out a systematic inactivation of a smaller set of some 260 genes (2). More recently, an experi- Bacillus subtilis genes. Among 4,100 genes of the organism, only mental approach involving high-density transposon mutagenesis 192 were shown to be indispensable by this or previous work. of the H. influenzae genome led to a much higher estimate of Another 79 genes were predicted to be essential. The vast majority 670 putative essential genes (3), whereas transposon mutagen- of essential genes were categorized in relatively few domains of esis of two mycoplasma species led to an estimate of 265–360 cell metabolism, with about half involved in information process- essential genes (4). Another experimental approach using anti- sense RNA to inhibit gene expression led to the identification of ing, one-fifth involved in the synthesis of cell envelope and the some 150 essential genes in Staphylococcus aureus (5). However, determination of cell shape and division, and one-tenth related to these approaches have limitations. Computation is likely to cell energetics. Only 4% of essential genes encode unknown underestimate the minimal gene set because it takes into account functions. Most essential genes are present throughout a wide only those genes that have remained similar enough during the range of Bacteria, and almost 70% can also be found in Archaea course of evolution to be recognized as true orthologues. and Eucarya. However, essential genes related to cell envelope, Transposon mutagenesis might overestimate the set by misclas- shape, division, and respiration tend to be lost from bacteria with sification of nonessential genes that slow down the growth small genomes. Unexpectedly, most genes involved in the Emb- without arresting it but can also miss essential genes that tolerate den–Meyerhof–Parnas pathway are essential. Identification of un- transposon insertions (3, 6). Finally, the use of antisense RNA known and unexpected essential genes opens research avenues to is limited to the genes for which an adequate expression of the better understanding of processes that sustain bacterial life. inhibitor y RNA can be obtained in the organism under study. To obtain an independent and possibly more reliable estimate he definition of the minimal gene set required to sustain a of a minimal protein-encoding gene set for bacteria, we system- T living cell is of considerable interest. The functions specified atically inactivated Bacillus subtilis genes. B. subtilis was chosen by such a set are likely to provide a view of a ‘‘minimal’’ bacterial because it is one of the best studied bacteria (7) and is a model cell. Many functions should be essential in all cells and could be for low-GC Gram-positive bacteria, which include both deadly considered as a foundation of life itself. The determination of the pathogens, such as Bacillus anthracis, and bacteria widely used in range of essential functions in different cells should reveal food and industr y, such as lactococci and bacilli. Because the possible solutions for sustaining life. Computational and exper- essentiality of a gene depends on the conditions under which the imental research has previously been carried out to define a organism is propagated, we used an environment likely to be minimal protein-encoding gene set. An upper-limit estimate of optimal for B. subtilis and thus carried out inactivation on a a minimal bacterial gene set was obtained from the sequence of the entire Mycoplasma genitalium genome, which contains only 480 genes (1). A computational approach, based on the To whom correspondence should be addressed. E-mail: ehrlich@jouy.inra.fr and assumption that essential genes are conser ved in the genomes of ehrlich@is.aist-nara.ac.jp. 4678 – 4683  PNAS  April 15, 2003  vol. 100  no. 8 www.pnas.orgcgidoi10.1073pnas.0730515100 Table 1. Essential and nonessential B. subtilis genes Table 2. B. subtilis essential genes Essential Nonessential Total DNA metabolism 27 Basic replication machinery 16 This study* 150 2,807 2,957 Packaging and segregation 9 Previous studies 42 614 656 Methylation 2 Prediction 79 106 185 RNA metabolism 14 Phage genes 0 303 303 Basic transcription machinery 4 Total 271 (6.6%) 3,830 (94.4%) 4,101 RNA modification 6 A list of the genes and their classifications can be accessed at http: Regulation 4 bacillus.genome.ad.jp. Protein synthesis 95 *We included 18 essential genes here that were inactivated in the course of Ribosomal proteins 52 this study and also studied previously. Aminoacyl-tRNA synthetases 24 Carried out in B. subtilis. Translation factors 10 Full list is presented as Table 3. Protein folding and modification 3 Excluded are four genes that were not studied because of technical reasons Protein translocation 6 (too short for insertional inactivation and too inconveniently placed for Cell envelope 44 chloramphenicol replacement). Membrane lipids 16 Cell wall 28 standard laborator y rich medium at 37°C. This choice also Cell shape and division 10 allowed for a comparison of the results obtained in many Glycolysis 8 laboratories and many previous studies, nevertheless leaving Respiratory pathways 22 open the possibility that a different gene set is essential under Isoprenoids 8 Menaquinone 8 different growth conditions. Analysis of the mutants, in con- Cytochrome biogenesis 3 junction with the literature data, leads us to conclude that there Thioredoxin 3 are only 271 genes indispensable for growth in LB when inac- Nucleotides 10 tivated singly. These fall into a relatively few large domains of cell Cofactors 15 physiology and are ver y broadly conser ved in microorganisms. CoA 1 Methods Folate 3 NAD 4 The approach used for gene inactivation has been described (8). S-Adenosylmethionine 1 Brief ly, it involved insertion of a nonreplicating plasmid into the Iron–sulfur cluster 6 target gene via a single crossover recombination. The expression Other 15 of the downstream genes from the same operon was controlled Unknown 11 by an isopropyl -D-thiogalactoside (IP TG)-regulated promoter Total 271 present on the inserted plasmid. A gene was deemed essential if it could not be inactivated by insertion (i.e., no transformants A complete list of genes and the evidence used to ascertain their essential were obtained when competent recipient cells were mixed with nature are presented in Table 4. the insertional plasmid) and if the strain became IP TG depen- dent when an intact copy of the gene was placed under control 106 are not (Table 3). We inactivated all but 4 of the remaining of the regulated promoter (8). IP TG-dependent strains could genes and found that 150 are essential. This analysis leads us to not be constructed for six essential genes, possibly because the conclude that there are 271 genes indispensable for growth when regulated promoter was either not strong enough or not suffi- ciently tuned to provide appropriate gene expression levels. An inactivated singly (Table 1). For 96% of these, we propose alternative strategy was followed for 160 genes shorter than assignment to various domains of cell metabolism (Table 2; the 300 bp, where insertional inactivation was limited by the insuf- complete list of genes is given in Table 4, which is published as ficient gene length. These genes were replaced by a chloram- supporting information on the PNAS web site). phenicol resistance marker, and if replacement failed they were Functional Assignment of Essential Genes. Information processing. rendered IP TG-dependent. All mutations were made in the About half of the essential genes are involved in DNA and RNA standard laborator y strain 168. Inactivation was not attempted metabolism and protein synthesis. Sixteen genes encode the for 656 genes studied previously in B. subtilis, and 185 genes basic DNA replication machiner y. They comprise five genes having a high degree of similarity with genes well characterized in other bacteria or involved in well characterized processes, for involved in the initiation of replication (dnaA, B, D, and I, and which we could predict essentiality with confidence (Table 3, priA), eight genes encoding components of the replisome (dnaC, which is published as supporting information on the PNAS web E, G, N, and X, holA and B, and polC), DNA ligase, and the Ssb site, www.pnas.org). Complete microbial genomes included in protein. One gene, pcrA, has no clearly identified role, but could the Microbial Genome Database for Comparative Analysis be involved in the progression of the replication fork (10). Among genes involved in DNA packaging and segregation, five (http:mbgd.genome.ad.jp), comprising 54 bacteria, 16 ar- encode topoisomerases (topA, gyrA and B, and parD and E), one chaea, and 2 yeasts, were analyzed for the presence of the B. subtilis essential gene homologs by using the default parameters, encodes the general DNA-binding protein Hbsu, and three with 10 as a cut-off value. encode the proteins that act in the condensation of the nucleoid (smc, and scpA and B; ref. 11). The remaining two genes encode Results modification methylases, expected to be essential unless the There are 4,100 annotated genes in the B. subtilis genome (9). cognate nucleases are inactivated. Some 303 are encoded on prophages that can be eliminated from Among 14 essential genes involved in RNA metabolism, four the genome and are not essential. Previous studies on 656 B. (rpoA, B, and C, and sigA) encode components of the basic transcription machiner y, whereas six are involved in RNA subtilis genes identified 42 that are essential (Table 1). Through predictions we propose that 79 other genes are essential, whereas modification. rnc and rnpA encode RNases, cspR and trmD and Kobayashi et al. PNAS  April 15, 2003  vol. 100  no. 8  4679 GENETICS U encode methylases, and cca encodes tRNA nucleotidyl trans- appears to be required for both fatty acid and phospholipid ferase. Only four genes are involved in regulation of RNA biosynthesis in a way that is not well understood (19). synthesis: a two-component system yycF and G (12), a gene Synthesis of peptidoglycan, the main component of the cell involved in the coupling between translation and termination of wall, comprises two stages, the synthesis of the precursor mol- RNA synthesis, nusA (13), and an anti-sigma factor, YhdL (14). ecules and the polymerization of peptidoglycan (20). All of the The largest categor y, comprising 95 essential genes, is that essential genes are involved in the first stage, which encompasses involved in protein synthesis. Over half of the genes encode a variety of biosynthetic pathways: (i) Synthesis of aminosugars ribosomal proteins. Although there is no experimental evidence (Fig. 6, which is published as supporting information on the that they are essential in B. subtilis, we suggest that they belong PNAS web site) by conversion of fructose-6-phosphate to UDP- to the essential set, because the ribosome itself is essential. This N-acetyl-glucosamine and UDP-N-acetyl-manosamine. The first suggestion is supported by the obser vation that the inhibition of two steps, leading to glucosamine-1-phosphate, are catalyzed by synthesis of 21 different ribosomal proteins is lethal in S. aureus the products of glmS and ybbT genes. The last two steps are (5). Among these are proteins such as L24, which was not carried out by the products of the gcaD and yvyH. More than one absolutely essential in E. coli, but cells that lacked it grew ver y gene product seems to be able to acetylate glucosamine-1- slowly and were thermosensitive (15). We suggest that there are phosphate, because there is no single essential gene for this step. 20 essential genes that encode aminoacyl-tRNA synthetases, (ii) Diaminopimelate (Fig. 7, which is published as supporting corresponding to 18 amino acids. All but two are present in information on the PNAS web site) is synthesized from L- unique copies. We showed that one of the unique copy genes, aspartate by eight successive reactions, six of which are carried lysS, is essential and assumed that others are too, without seeking out by products of essential genes asd, dapA, B, and F, and ykuQ further experimental evidence. There are two genes encoding and R. The first and the fifth step can be catalyzed by products tRNA-Tyr and tRNA-Thr synthetases. Only t yrS was essential of three (dapG, lysC, and yclM) and two genes (mtnV and ywfG), when inactivated singly whereas either thrS or thrZ could assure respectively; thus, none of the five is essential if inactivated the viability. We grouped with the synthetases three genes that singly. (iii) Two essential genes, racE and alr, encode racemases are required for the conversion of the tRNA-Glu to tRNA-Gln that convert L-glutamate and L-alanine into the corresponding D (gatC, B, and A) and one gene that is required for the formylation isomers. racE cannot be replaced by a homologue, yrpC. The of methionyl tRNA ( fmt). Of the10 essential genes involved in essential ddl gene is required for synthesis of the dipeptide mRNA translation, 3 are required for initiation (infA, B, and C), D-Ala-D-Ala. (iv) Eight essential genes, murAA, murB, C, D, E, 3 are required for elongation (tufA, tsf, and fusA), and 4 are F, and G, and mraY, are required for synthesis of the lipid-linked required for termination and ribosome recycling (prfA and B, disaccharide-pentapeptide peptidoglycan precursor (Fig. 8, pth, and frr). There is one essential gene involved in posttrans- which is published as supporting information on the PNAS web lation modification, map, that encodes methionine aminopepti- site) from UDP-N-acetyl-glucosamine, phosphoenolpyruvate, dase. Deformylation is also required, but can be carried out by D-glutamine, diaminopimelate, D-ala dipeptide, and an isopre- products of two genes, def and ykrB, neither of which is essential nylphosphate. Polymerization of peptidoglycan is carried out by when inactivated singly (16). Two essential genes, groEL and ES, the products of functionally redundant genes in B. subtilis. The are involved in protein folding. Finally, there are six essential cell wall of B. subtilis contains teichoic acid (21), and there are genes that encode key components of the machiner y for protein seven essential genes involved in its synthesis. Four, tagA, B, D, insertion into the membrane and secretion. These include the and O, are required for the synthesis of linkage units and three, targeting factors Ffh and FtsY, the translocation motor SecA, tagF, G, and H, are required for chain polymerization, translo- two components of the translocation channel, SecY and E, and cation, and linkage to peptidoglycan (Fig. 9, which is published the folding catalyst PrsA. The essential DNA-binding protein as supporting information on the PNAS web site). Hbsu is also a part of the signal recognition particle (17). Ten essential genes are involved in cell shape and division. Cell envelope, shape, and division. About one-fifth of the essential Septum formation requires seven ( ftsA, L, W, and Z, divIB and genes are required for these processes (Table 2). The synthesis C, and pbpB; ref. 21), whereas cell shape requires three (rodA, of the cell envelope involves 44 essential genes, all required for and mreB and C). membrane and cell wall formation. Membrane lipids, phospho- Embden–Meyerhof–Parnas (EMP) pathway and respiration. About 10% lipids, and glycolipids are synthesized from fatty acids. Fatty acid of essential genes, which have in common the provision of energy synthesis (Fig. 4, which is published as supporting information on for the cell, are required for these processes. A majority of genes the PNAS web site) is initiated by products of four genes, accA, composing the ubiquitous EMP pathway are essential (Fig. 10, B, C, and D, together with acpA and fabD gene products. acpS which is published as supporting information on the PNAS web is required for the conversion of AcpA from the apo to the holo site). The process can be viewed as consisting of two parts: the form, whereas birA is required for the addition of a biotinyl top, which converts hexose sugars to trioses, and the bottom, group to carboxylase. The fatty acid chains are elongated by the which converts these compounds to pyruvate, funneled into products of two essential genes, fabFG. The elongation cycle pyruvate dehydrogenase. The top part comprises four steps when involves two additional steps that are catalyzed by pairs of genes glucose is the carbon source, the last two of which are catalyzed with overlapping functions (ycsD and ywpB, and fabI and L), by products of essential genes pfkA and fbaA, whereas the bottom none of which is essential when inactivated singly (18). Two of part comprises six steps, four of which are encoded by essential the essential genes required for phospholipid synthesis (Fig. 5, genes tpiA, pgk, pgm, and eno. The two remaining essential genes which is published as supporting information on the PNAS web related to glycolysis are tkt and prs. The first encodes a transke- site), gpsA and yhdO, are involved in the conversion of dihy- tolase, involved in the pentose pathway, whereas the second gene droxyacetone phosphate to phosphatidic acid, which is a pre- codes for a pyrophosphokinase that converts ribose-5-phosphate cursor of complex lipids. Interestingly, yerQ, which encodes an to 5-phospho-ribose-1-diphosphate, a common precursor of enzyme with a diacylglycerol kinase catalytic domain found in nucleotides and cofactors, such as NAD, which likely accounts eukar yotes and presumably catalyses synthesis of phosphatidic for its essential role. Taken together, these results are rather acid from another precursor (diacylglycerol), is also essential, unexpected. First, our experiments were carried out on a rich whereas a homologue, dgkA, is not. Two essential genes, cdsA medium, which contains numerous compounds that could pro- and pgsA, are required for synthesis of phosphatidylglycerol vide the energy and building blocks for cell life, the two known phosphate, which might be converted into phosphoglycerol by a functions of the EMP pathway. Addition of glucose to LB did not nonspecific phosphatase. The remaining essential gene, plsX, restore growth of any of the nonviable EMP mutants. Second, in B. 4680  www.pnas.orgcgidoi10.1073pnas.0730515100 Kobayashi et al. subtilis a part of the EMP pathway can be bypassed via the pentose is published as supporting information on the PNAS web site), shunt, and it is surprising that both are simultaneously required for and only the four genes involved in the salvage pathway (yueK, viability. Possibly, the enzymes revealed as essential have novel and yqeJ, nadE, and yjbN) were essential. We speculate that the unexpected functions in the cell. It should be noted that pgm and accumulation of nicotinate might repress de novo synthesis of eno mutants have been isolated previously and had very slow nicotine mononucleotide in the absence of yueK, rendering this growth (22), suggesting that the difference between lethal and gene essential. There are three essential genes involved in folate almost-lethal mutation can be due to subtle differences in the metabolism (Fig. 16, which is published as supporting information experimental conditions and the strain background. on the PNAS web site). One, dfrA, codes for dihydrofolate reduc- Respiration can provide energy for the cell, in the absence of tase, which converts folate, presumably imported from the medium, glycolysis. We identified 22 essential genes involved in this to tetrahydrofolate. Two other genes, glyA and folD, are required for process. Under the aerobic condition used in our experiments, conversion of the latter compound to 10-formyl tetrahydrofolate, a respiration involves the transfer of electrons by various dehy- one-carbon donor molecule for a number of reactions. S- drogenases to menaquinone and then to cytochromes (23). adenosylmethionine (SAM) is another one-carbon donor, synthe- Menaquinone is synthesized from chorismate in seven steps, the sized from ATP and methionine by SAM synthetase, encoded by last six of which are catalyzed by products of essential genes, the essential metK gene. There is only one essential gene involved menA, B, C, D, E, and H (Fig. 11, which is published as in the biosynthesis of CoA, ytaG, that is required for the last step supporting information on the PNAS web site). Two genes, menF in the pathway (Fig. 17, which is published as supporting informa- and dhbC, appear to be able to catalyze the first step, and neither tion on the PNAS web site), suggesting that the precursor, dephos- is essential if inactivated singly. The penultimate step involves pho-CoA, is transported from the medium into the cell. The condensation of dihydroxynaphthoic acid with an isoprenoid remaining cofactor is an iron–sulfur cluster, which forms part of biphosphate. Isoprenoids (Fig. 12, which is published as sup- proteins that participate in many aspects of the cell physiology, porting information on the PNAS web site) are synthesized from including redox and nonredox catalysis, as well as sensing for pyruvate and glyceraldehyde-3-phosphate by a nonmevalonate regulatory processes. There are five essential genes, yurU, V, W, X, pathway in B. subtilis. The first six steps, leading to isopentenyl and Z, involved in the synthesis of this cluster. We included here diphosphate, involve seven essential genes, dxs, dxr, ispE, yacM yrvO, a homologue of yurV. and N, and yqfP and Y. Three other essential genes, hepS and T Other processes. Only 15 essential genes that have a clear bio- and yqiD, are required for the synthesis of farnesyl diphosphate chemical function were not associated with any of the large and more complex compounds that are used for menaquinone domains of cellular physiology discussed above. Among these are synthesis. Altogether, of 22 essential genes involved in respira- six GT P-binding proteins of the EraObg family. Only one, obg, tion, 16 are required for menaquinone synthesis. There are only has been studied previously in B. subtilis and been shown to three essential genes involved in cytochrome biogenesis, resA, B, affect the stress response mediated by  . Five other genes, and C. No cytochrome structural genes are essential, possibly mrpA, B, C, D and F, encode a sodium– hydrogen antiporter, ref lecting overlapping functions of their products (24). We have which is required to maintain pH homeostasis in the presence of included trxA and B, which encode thioredoxin and thioredoxin sodium chloride concentrations similar to those found in LB reductase with the respiration genes, because of the role of TrxA (27). ppaC encodes the inorganic pyrophosphatase, which drives in electron transport, although this protein is involved in many the anabolic f luxes by pyrophosphate hydrolysis in various other oxido-reduction reactions. We also included here a puta- biochemical reactions, whereas gcp encodes a sialopeptidase of tive thioredoxin reductase gene, yumC. Nucleotides and cofactors. Metabolism of these compounds requires 10% of the essential genes (Table 2). The metabolism of nucleotides is quite complex, comprising complementar y de novo synthesis and salvage pathways (25). Nevertheless, we found 10 essential genes involved in this process. Among the four that participate in purine metabolism (Fig. 13, which is published as supporting information on the PNAS web site), two (adk and gmk) specify kinases, which phosphor ylate A MP or GMP to the respective diphosphates. Absence of guanine from the medium accounts for the essential nature of guaB. Surprisingly, hprT,a gene from the purine salvage, is also essential, raising a possi- bility that its product has a second, unsuspected role in the cell. Two essential genes involved in pyrimidine metabolism (Fig. 14, which is published as supporting information on the PNAS web site), cmk and tmk, also encode kinases that phosphor ylate CMP and TMP to corresponding diphosphates. The remaining essen- tial gene, pyrG, encodes cytidylate synthetase, which converts UT P into CT P. This might ref lect the paucity of cytidine in the rich medium. Interestingly, two B. subtilis essential genes encode enzymes present in the E. coli degradosome [yjbN (ppnK) and eno, a member of the EMP pathway], which provides CDP for DNA synthesis and further nucleotide metabolism, while con- trolling mRNA turnover (26). Finally, there are three essential Fig. 1. B. subtilis essential gene homologues are widely conserved. (Upper) genes involved simultaneously in purine and pyrimidine metab- Genes are ordered by their relative abundance among 54 Bacteria (blue) and 18 olism, nrdE and F and ymaA, that encode subunits of nucleoside- Archaea and Eucarya (red). The position (rank) of a gene is shown on abscissa and diphosphate reductase, which converts the ribose into deoxyri- the fraction of organisms in which a gene is present is shown on the ordinate. bose derivatives. (Lower) Fraction of genes present in different kingdoms of life (a gene counted Synthesis of only five cofactors, involving 16 genes, was as ‘‘all kingdoms’’ is present in at least one archaeon and one eukaryote, in required under our experimental conditions. NAD synthesis can addition to bacteria, whereas a gene counted as ‘‘bacteria’’ is not present in any take place de novo or by salvaging of precursors (Fig. 15, which archae or eukaryote). The list of genes and organisms is presented in Table 4. Kobayashi et al. PNAS  April 15, 2003  vol. 100  no. 8  4681 GENETICS Fig. 2. The number of B. subtilis essential gene homologues depends on genome size. (Top) All genes. Bacilli and close relatives denote Bacillus species and other low-GC Gram-positive bacteria, but not clostridia, mycoplasma, and ureaplasma. (Middle and Bottom) Different bacterial gene categories. Empty red circles in Bottom refer to Bacilli and close relatives, whereas filled red circles refer Fig. 3. Phylogenetic profiling of essential genes. The 271 B. subtilis genes to other bacteria. Interpolated lines throughout the figure correspond to the best were grouped in 266 clusters. Only one gene, yhdL, which encodes a possible fitting polynomial of the second or the fourth order. The number of genes is: anti-sigma protein, had no orthologues in the database and is not presented information processing, 136; envelope, respiration, cell shape, and division, 76; here. Each line and column corresponds to individual gene and organism, cofactors, other, and unknown, 41; and nucleotides and glycolysis, 18. respectively. Presence and absence of a gene is indicated by a black and white square, respectively. The list of genes and organisms is given in Table 5, which is published as supporting information on the PNAS web site and the ordering unknown role. The last two genes, pdhA and odh, encode is described in the text. subunits of pyruvate and 2-oxoglutarate dehydrogenase, respec- tively; growth of the mutants could be restored by addition to LB of the metabolites (acetate and succinate, respectively) related 3 Mb (highlighted in red). Other bacteria with genomes of a to the activity of the proteins they encode. similar size have, on average, slightly 80% of the B. subtilis Unknown. The last categor y groups 11 essential genes for which we essential gene homologues. This proportion drops to 57% with were unable to suggest a role in cell physiology. Biochemical decreasing bacterial genome size, indicating progressive loss of functions, a protease and a hydrolase of the metallo--lactamase essential genes. Archaea and Eucar ya maintain, on average, 36% superfamily, can be suggested for products of two gene, ydiC and of the essential gene homologues, with the proportion var ying ykqC. One gene, yneS, encodes a putative membrane protein, between 33% and 44% almost linearly with genome size. In and another, ymdA, encodes a protein with an HD domain of bacteria, gene loss occurs mainly in three categories (cell enve- metal-dependent phosphohydrolases, whereas three, yloQ, yqjK, lope, shape and division, and respirator y pathways) and to a and ywlC, encode proteins with recognizable signatures, an lower extent in three other categories (cofactor synthesis, other AT PGT P-binding site, a metallo--lactamase motif, and a processes, and unknown functions). In contrast, information putative RNA-binding motif, respectively. Four genes, yacA, processing, glycolysis, and nucleotide synthesis genes are largely ydiB, ylaN, and yqeI, have no easily recognizable features. retained (Fig. 2 Middle and Bottom). Phylogenetic profiling of essential B. subtilis genes is summarized Conservation of Essential Genes. The average level at which ho- in Fig. 3. Organisms were grouped into four classes and ordered mologues of essential B. subtilis genes are present in bacteria is within each class on the basis of the number of essential gene rather high (approaching 80%), one-fourth being found in all homologues they share with B. subtilis, placing the organisms with bacteria and three-fourths in at least 75% (Fig. 1 Upper). The fewest conserved genes at the right of each class. Genes were average is 36% in Eucar ya and Archaea, but some 20% of the grouped in categories and ordered by abundance among all bac- genes are nevertheless present in all 18 organisms we analyzed teria, which placed the less abundant genes at the bottom of each (Fig. 1 Upper). About one-third of the genes are found in all three category. A number of general features are easily discernible from kingdoms of life, and a further one-third are shared between this analysis. (i) The five top categories are composed of genes Bacteria and either Archaea or Eucar ya (Fig. 1 Lower). present in 80% of Bacteria and at least 40% of Eucarya and The number of B. subtilis essential gene homologues present Archaea, with the exception of RNA synthesis, which is less well in an organism depends on at least two parameters: phylogenetic conserved in the last two kingdoms. (ii) The next two categories, proximity to B. subtilis and genome size (Fig. 2 Top). The highest number is found in bacilli and close relatives, having genomes of DNA metabolism and cell shape and division, contain genes 4682  www.pnas.orgcgidoi10.1073pnas.0730515100 Kobayashi et al. present in most bacteria and genes specific for Gram-positive bacteria with smaller genomes. Concomitantly, genes involved in organisms. This can most easily be seen from the appearance of the the determination of cell shape, division, and respiration are also lost. This suggests that it may be possible to build, maintain, and relatively broad horizontal white bars at the bottom of the two reproduce the cell compartment in a simpler way than that used classes. (iii) The categories that contain genes missing from bacteria by bacteria with larger genomes, and that glycolysis can be with small genomes are easily identified by the presence of the sufficient to generate energy for the cell. A minimal essential vertical white band at the right of the low-GC Gram-positive gene set could thus be significantly smaller than the one present bacteria class, corresponding to Mycoplasma and Ureaplasma urea- in bacteria with genomes larger than 3 Mb. lyticum. In addition, there is an enlargement of the white zone at the right end of the ‘‘Other bacteria’’ class, noticeable for cell envelope, Unexpected Essential Genes. Notwithstanding the grouping of respiration, and unknown functions. (iv) Genes in the last two most essential functions in a few large categories, our study has categories, unknown and other, although often found only in the revealed genes that were not expected to have an essential closest relatives of B. subtilis, are nevertheless present in over a half function under the experimental conditions used, such as eight of other bacteria. EMP pathway genes and a gene involved in purine biosynthesis. These obser vations suggest previously unsuspected links be- Discussion tween different domains of cell physiology. A Simple Bacterial Cell. Of some 4,100 genes of B. subtilis, only 271 are essential for growth under our experimental conditions when Redundant Genes for Essential Functions. Our analysis does not inactivated singly. About 80% of the functions they encode fall detect essential functions encoded by redundant genes, because in a few large categories; namely, information processing, cell only a single gene was inactivated in each mutant strain. The list envelope, shape, division, and energetics. These obser vations of the essential genes given here is thus likely to be underesti- lead to a view of a rather simple bacterial cell, consisting of a mated, because synthetic lethal mutants are well known. A rigorous detection of the missing functions would require the compartment, formed by a membrane and a wall, enclosing the systematic combination of all of the mutations in a single strain, elements necessar y to synthesize proteins that carr y out reac- which is beyond the present genetic technology. However, it is tions required for (i) the duplication and inheritance of the remarkable that single gene inactivation did reveal large cate- genetic information; (ii) the division of the compartment; and gories of essential functions, suggesting that most of the vital cell (iii) the provision of energy. These processes do not appear to be processes are encoded by nonredundant genes. The presence of coordinated by modulation of gene expression, because the expres- paralogues for 50% of B. subtilis genes (9) might thus allow the sion regulators are by and large not essential. We suggest that the cell to respond to changing environmental conditions rather than coordination might be carried out, at least in part, by the essential provide back-up for vital processes. GTP-binding proteins, as appears to be the case in eukaryotes. Isogenic Mutant Collection. Finally, it should be noted that the Broad Distribution of Essential Genes and Functions. Over 80% of isogenic set of 3,000 mutants that we have generated can be essential B. subtilis gene homologues are present in all bacteria used to identify genes, and thus functions, that are essential with genomes above 3 Mb, and 57% are found even in bacteria under conditions different from those used here. Furthermore, with the smallest genomes (mycoplasma). Almost 70% of genes the mutant set is a unique bacterial resource for studying various are present in at least one kingdom other than Bacteria. Many phenotypes and may thus lead to deeper insight into the me- organisms thus appear to rely on a similar set of essential tabolism of the bacterial cell. functions, supporting the simple microbial cell view outlined above. The similarity might be even higher, because some of the This work was supported, in part, by European Union Grant BIO4- genes might have diverged beyond recognition and some func- CT95-0278 and a Grant-in-Aid for Scientific Research on Priority Areas tions can be encoded by unrelated genes (28). However, genes (C) ‘‘Genome Biology’’ from the Ministr y of Education, Culture, Sports, involved in the synthesis of the cell envelope tend to be lost from Science and Technology of Japan. 1. Fraser, C. M., Gocayne, J. D., White, O., Adams, M. D., Clayton, R. A., 16. Haas, M., Beyer, D., Gahlmann, R. & Freiberg, C. (2001) Microbiology 147, Fleischmann, R. D., Bult, C. J., Kerlavage, A. R., Sutton, G., Kelley, J. M., et 1783–1791. al. (1995) Science 270, 397– 403. 17. Nakamura, K., Yahagi, S., Yamazaki, T. & Yamane, K. (1999) J. Biol. Chem. 2. Mushegian, A. R. & Koonin, E. V. (1996) Proc. Natl. Acad. Sci. USA 93, 274, 13569 –13576. 10268 –10273. 18. Heath, R. J., Su, N., Murphy, C. K. & Rock, C. O. (2000) J. Biol. Chem. 275, 3. Akerley, B. J., Rubin, E. J., Novick, V. L., Amaya, K., Judson, N. & Mekalanos, 40128 – 40133. J. J. (2002) Proc. Natl. Acad. Sci. USA 99, 966 –971. 19. Morbidoni, H. R., de Mendoza, D. & Cronan, J. E., Jr. (1996) J. Bacteriol. 178, 4. Hutchison, C. A., Peterson, S. N., Gill, S. R., Cline, R. T., White, O., Fraser, 4794 – 4800. C. M., Smith, H. O. & Venter, J. C. (1999) Science 286, 2165–2169. 20. Foster, S. J. & Popham, D. L. (2002) in Bacillus subtilis and Its Closest Relatives: 5. Ji, Y., Zhang, B., Van Horn, S. F., Warren, P., Woodnutt, G., Burnham, M. K. From Genes to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, R. (Am. Soc. & Rosenberg, M. (2001) Science 293, 2266 –2269. Microbiol, Washington DC), pp. 21– 41. 6. Gerdes, S. Y., Scholle, M. D., D’Souza, M., Bernal, A., Baev, M. V., Farrell, 21. Errington, J. & Daniel, R. A. (2002) in Bacillus subtilis and Its Closest Relatives: M., Kurnasov, O. V., Daugherty, M. D., Mseeh, F., Polanuyer, B. M., et al. From Genes to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, R. (Am. Soc. (2002) J. Bacteriol. 184, 4555– 4572. Microbiol, Washington, DC), pp. 97–109. 7. Sonenshein, A. L., Hoch, J. A. & Losick, R., eds. (2002) Bacillus subtilis 22. Leyva-Vazquez, M. A. & Setlow, P. (1994) J. Bacteriol. 176, 2788 –2795. and Its Closest Relatives: From Genes to Cells (Am. Soc. Microbiol., Wash- 23. von Wachenfeldt, C. & Hederstadt, L. (2002) in Bacillus subtilis and Its Closest ington, DC). Relatives: From Genes to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, 8. Vagner, V., Der vyn, E. & Ehrlich, S. D. (1998) Microbiology 144, 3097–3104. R. (Am. Soc. Microbiol, Washington, DC), pp. 163–179. 9. Kunst, F., Ogasawara, N., Moszer, I., Albertini, A. M., Alloni, G., Azevedo, V., Bertero, 24. Winstedt, L. & von Wachenfeldt, C. (2000) J. Bacteriol. 182, 6557– 6564. M. G., Bessieres, P., Bolotin, A., Borchert, S., et al. (1997) Nature 390, 249–256. 25. Switzer, R. L. (2002) in Bacillus subtilis and Its Closest Relatives: From Genes 10. Petit, M. A. & Ehrlich, S. D. (2002) EMBO J. 21, 3137–3147. to Cells, eds. Sonenshein, A. L., Hoch, J. A. & Losick, R. (Am. Soc. Microbiol, 11. Soppa, J., Kobayashi, K., Noirot-Gros, M. F., Oesterhelt, D., Ehrlich, S. D., Washington, DC), pp. 255–269. Der vyn, E., Ogasawara, N. & Moriya, S. (2002) Mol. Microbiol. 45, 59 –71. 26. Nitschke ´, P., Guerdoux-Jamet, P., Chiapello, H., Faroux, G., Henaut, C., 12. Fabret, C. & Hoch, J. A. (1998) J. Bacteriol. 180, 6375– 6383. Henaut, A. & Danchin, A. (1998) FEMS Microbiol. Rev. 22, 207–227. 13. Ingham, C. J., Dennis, J. & Furneaux, P. A. (1999) Mol. Microbiol. 31, 651– 663. 27. Ito, M., Guffanti, A. A., Oudega, B. & Krulwich, T. A. (1999) J. Bacteriol. 181, 14. Horsburgh, M. J. & Moir, A. (1999) Mol. Microbiol. 32, 41–50. 2394 –2402. 15. Nishi, K., Dabbs, E. R. & Schnier, J. (1985) J. Bacteriol. 163, 890 – 894. 28. Koonin, E. V. (2000) Annu. Rev. Genomics Hum. Genet. 1, 99 –116. Kobayashi et al. PNAS  April 15, 2003  vol. 100  no. 8  4683 GENETICS

Journal

Proceedings of the National Academy of Sciences – Unpaywall

Published: Apr 7, 2003

There are no references for this article.