The Forespore Line of Gene Expression in Bacillus subtilis

doi:10.1016/j.jmb.2006.01.059

J. Mol. Biol. (2006) 358, 16–37

The Forespore Line of Gene Expression in Bacillus subtilis Stephanie T. Wang1, Barbara Setlow2, Erin M. Conlon3 Jessica L. Lyon4, Daisuke Imamura5, Tsutomu Sato5 Peter Setlow2, Richard Losick1* and Patrick Eichenberger1,4 1

Department of Molecular & Cellular Biology, Harvard University, Cambridge MA 02138, USA 2 Department of Molecular Microbial and Structural Biology, University of Connecticut Health Center Farmington, CT 06032, USA 3

Department of Mathematics and Statistics, University of Massachusetts, Amherst MA 01003, USA 4

Department of Biology and Center for Comparative Functional Genomics New York University New York, NY 10003, USA 5

International Agricultural Science, Tokyo University of Agriculture and Technology Fuchu, Tokyo 183-8509, Japan *Corresponding author

Endospore formation by Bacillus subtilis involves three differentiating cell types, the predivisional cell, the mother cell, and the forespore. Here we report the program of gene expression in the forespore, which is governed by the RNA polymerase sigma factors sF and sG and the DNA-binding proteins RsfA and SpoVT. The sF factor turns on about 48 genes, including the gene for RsfA, which represses a gene in the sF regulon, and the gene for sG. The sG factor newly activates 81 genes, including the gene for SpoVT, which turns on (in nine cases) or stimulates (in 11 cases) the expression of 20 genes that had been turned on by sG and represses the expression of 27 others. The forespore line of gene expression consists of many genes that contribute to morphogenesis and to the resistance and germination properties of the spore but few that have metabolic functions. Comparative genomics reveals a core of genes in the sF and sG regulons that are widely conserved among endospore-forming species but are absent from closely related, but non-spore-forming Listeria spp. Two such partially conserved genes (ykoU and ykoV), which are members of the sG regulon, are shown to confer dry-heat resistance to dormant spores. The ykoV gene product, a homolog of the non-homologous end-joining protein Ku, is shown to associate with the nucleoid during germination. Extending earlier work on gene expression in the predivisional cell and the mother cell, we present an integrated overview of the entire program of sporulation gene expression. q 2006 Elsevier Ltd. All rights reserved.

Keywords: sporulation; forespore; sigma factor; transcriptional profiling; promoters

Introduction The formation of a multicellular organism involves the generation of multiple types of specialized cells by cellular differentiation. The differentiation of each cell type is governed by its own program of gene expression and is coordinated with the differentiation of other cells by intercellular signaling pathways. A major challenge in the field of developmental biology is to comprehensively describe the programs of gene expression for all cell

Abbreviations used: SASP, small, acid-soluble protein; FFL, feed-forward loop; GFP, green fluorescence protein. E-mail address of the corresponding author: [email protected]

types in a developing organism and to elucidate the regulatory circuits to which they are subject. An attractive model system in which to undertake this challenge is endospore formation in Bacillus subtilis because of the relative simplicity of the organism and its amenability to the tools of traditional and molecular genetics.1,2 Endospore (henceforth simply spore) formation is a seven to eight hour process that is triggered by conditions of nutrient limitation.1,2 Spore formation involves three cell types known as the predivisional cell, the forespore (or prespore) and the mother cell. Cells enter the pathway to sporulate in response to conditions of nutrient limitation, which results in the formation of the predivisional cell. Next, the predivisional cell undergoes a process of asymmetric division in which a septum is formed near

0022-2836/$ - see front matter q 2006 Elsevier Ltd. All rights reserved.

Forespore Line of Gene Expression in B. subtilis

one pole, thereby creating dissimilar-sized progeny cells. These are the forespore (the smaller cell) and the mother cell. Initially, the forespore and the mother cell lie side-by-side but later in development the forespore is engulfed by the mother cell. This phagocytic-like process results in a cell-within-acell in which the forespore and its membrane are wholly surrounded by an outer layer of membrane derived from the engulfing mother cell. During subsequent morphogenesis, the forespore, which will become the core of the mature spore, undergoes dehydration driven in part by the replacement of water by calcium dipicolinate, and its chromosome becomes packaged by a family of small, acidsoluble proteins (SASPs) into a toroid-like structure in which it is protected against many types of DNA damage. Meanwhile, a thick layer of cell wall material known as the cortex is produced in the space between the membranes that separate the two cells, and an outer shell of protein termed the coat is deposited around the developing spore from within the mother cell. Thus, when morphogenesis is complete the spore core is encased by protective outer layers of cortex and coat material. Eventually, the spore is released by lysis of the mother cell. Upon release, the mature spore can remain dormant for long periods of time, but under favorable conditions can germinate and rapidly resume vegetative growth. The master regulator for entry into sporulation is Spo0A, a member of the response regulator family of DNA-binding proteins.3,4 Spo0A, which is activated by phosphorylation, orchestrates gene expression in the predivisional cell, acting both as an activator and a repressor. Some gene expression in the predivisional cell is activated indirectly by Spo0A in a pathway involving Spo0A-mediated repression of a repressor gene called abrB.5–8 The product of this gene, AbrB, prevents the expression during growth of certain genes that are turned on at the start of sporulation when AbrB is depleted from cells in which Spo0A has been activated. Approximately 121 genes, which are organized as 30 single-gene units and 24 operons, are under the direct control of Spo0A.9 Forty of these genes are up-regulated by Spo0A and 81 are down-regulated. Interestingly, the levels of Spo0A rise gradually in the early stages of sporulation with different genes being turned on or off at different levels of phosphorylated Spo0A.10 Thus, some genes in the regulon are low-threshold genes (requiring a low level of Spo0A to be turned on or off) and others high-threshold genes. The abrB gene, for example, is a low-threshold gene that is repressed at low levels of the master regulator. Asymmetric division sets in motion two parallel but interconnected programs of gene expression.11 The earliest-acting, cell-specific regulatory protein in the mother cell line of gene expression is the RNA polymerase sigma factor sE. The sE factor turns on the expression of about 262 genes, including the genes for GerR and SpoIIID.12,13 These DNAbinding proteins repress the expression of many

17 of the genes in the sE regulon. In addition, SpoIIID, acting in conjunction with sE, turns on about ten genes including genes involved in the appearance of the next regulatory protein in the mother cell line of gene expression, sK. The sK factor, in turn, switches on about 75 genes, including the gene for GerE. Finally, GerE, acting both as a repressor and an activator, switches off many of the genes in the sK regulon while turning on the expression of a terminal gene set of about 36 genes. Thus, the mother cell program of gene expression is governed by a hierarchical regulatory cascade consisting of a linked series of feed-forward loops (FFLs).14–16 FFLs are wide spread regulatory motifs in which a primary regulator (such as sE) directs the synthesis of a secondary regulator (such as SpoIIID) and both regulatory proteins control the expression of a set of target genes. Meanwhile, a parallel program of gene expression is played out in the forespore compartment of the sporangium. The earliest-acting regulatory protein in the forespore is sF, which turns on the synthesis of RsfA17 and sG. RsfA is a DNAbinding protein that represses one of the genes in the sF regulon, spoIIR, a signaling gene that is involved in triggering the appearance of sE in the mother cell.18,19 The sG factor is an activator that turns on the next set of genes in the cascade, including genes for the SASP family of proteins (see above), and the gene for the DNA-binding protein SpoVT.20 SpoVT (as we shall see) is both a repressor of genes in the sG regulon as well as an activator of the terminal gene set in the forespore line of gene expression. Finally, the forespore and mother cell lines of gene expression are linked to each other by a series of intercellular signaling pathways in which sF, by turning on the synthesis of the secreted signaling protein SpoIIR (above), triggers the appearance of sE in the mother cell and sG, by turning on the synthesis of the secreted signaling protein SpoIVB, triggers the appearance of sK in the mother cell.11 Meanwhile, sE sets in motion a still poorly understood chain of events that triggers the activation of sG in the forespore. Thus, gene expression in the forespore and the mother cell is linked in criss-cross fashion by a successive series of signals that go back and forth from one cell to the other. As enumerated above, gene expression in the predivisional cell and in the mother cell were elucidated previously in a comprehensive manner involving multiple complementary approaches, including transcriptional profiling during sporulation, transcriptional profiling in cells engineered to produce regulatory proteins during growth, biochemical analysis with purified regulatory proteins, chromatin immunoprecipitation, promoter identification by transcriptional start site mapping, bioinformatics and the use of gene reporters, and systematic gene inactivation. Recently, and while this work was in progress, an analysis of genes under the control of sF and sG was reported largely based on transcriptional profiling during

18 sporulation.21 Here we report a comprehensive analysis of the forespore line of gene expression based on transcriptional profiling during sporulation and a robust system for producing the forespore sigma factors during growth, genome wide analysis of the roles of RsfA and SpoVT, transcriptional start site mapping, bioinformatics, comparative genomics, and systematic gene inactivation. Our work elucidates the overall program of gene activation and repression in the forespore, the conservation of forespore-expressed genes among other endospore-forming species, and the function of these genes in spore maturation. In particular, we report on the discovery of an operon that markedly influences the resistance properties of the spore. Our analysis also provides insights into the logic of the regulatory circuit governing the differentiation of the forespore. Finally, as this work essentially completes the genome-wide identification of genes under developmental control during sporulation, we present an integrated overview of the entire program of gene expression in the three differentiating cell types of this primitive developing organism.

Results and Discussion Transcriptional profiling The forespore line of gene expression was elucidated by transcriptional profiling with microarrays imprinted with oligonucleotides (65-mers) corresponding to essentially all of the annotated open-reading frames in the genome (4106). Two complementary sets of transcription profiling experiments were carried out. In one, RNA was collected from the following mutants at times after suspension in sporulation medium chosen to correspond to the peak of the activity of the transcription factor in question: at 2 h after suspension (hour two of sporulation) RNA from cells wild-type for sF (strain SW3) was compared against RNA from cells mutant for sF (strain SW7). Because sF triggers the activation of sE and directs transcription of the gene for sG, strains SW3 and SW7 were designed to be mutant for these two regulatory proteins so that the transcriptional profile would be confined to genes under the direct control of sF. At hour three, RNA from cells wild-type for RsfA (strain RL560) was compared against RNA from cells mutant for RsfA (strain SW211). To eliminate indirect effects of RsfA on downstream gene expression, both strains contained a mutation in the gene for the next regulator in the hierarchy, sG. At hour four, RNA from cells wild-type for sG (strain RL3230) was compared against RNA from cells mutant for sG (strain SW64). Since sG triggers the appearance of sK, a null mutation in the gene for sK was included in strains RL3230 and SW64. Finally, at hour five, RNA from cells wild-type for SpoVT (strain RL3230) was compared against RNA from cells

Forespore Line of Gene Expression in B. subtilis

mutant for SpoVT (strain SW68), with both strains being mutant for sK. Examples of the microarray results are shown as logarithmic scale plots of spot intensities in Supplementary Data, Figure S1. The scatter plots compare the fluorescence intensities for each gene in the genome between strains wildtype and mutant for sF (Figure S1A) or sG (Figure S1B). The majority of spots fall on or near the line with a slope of 1, which represents genes whose RNA levels were not significantly affected by the presence of sF or sG. In the second set of experiments, transcriptional profiling was carried out with strains engineered to produce sF or sG during growth. RNA was isolated from cells induced for the synthesis of the transcription factor and compared against uninduced cells of the same strain. The sF regulon Candidates for members of the sF regulon were identified, as described above, both by comparing RNA from a strain wild-type for sF against RNA from a strain mutant for sF, and by comparing RNA from cells artificially induced for the synthesis of sF during the exponential phase of growth against RNA from uninduced cells. The transcription profiling experiments revealed 48 genes, organized in 36 transcription units that were under the control of sF. Twenty of the genes were previously known to be under sF control, whereas the remaining 28 genes were newly identified members of the regulon (Table 1). Of the 20 previously identified regulon members, nine gave a signal above the cut-off for significance in both sets of transcription profiling experiments, and ten missed the cut-off (which was a ratio of 2.0 and score of 0.95) in one of the two sets of experiments. One gene, tlp, did not give a significant signal in either experiment, but is located downstream, and in the same operon as, a known sF- (and sG-) controlled gene (sspN).22 Twenty-eight genes were identified as new members of the sF regulon. Of the 28 genes, eight genes (arsB, pbpI, tuaG, ylbB, ylbC, yqhQ, yqhH, ytfI) gave a signal that missed the cut-off for significance in one or the other of the two sets of transcriptional profiling experiments and a ninth (ywnJ) missed the cut-off in both sets of experiments. These nine genes were nonetheless included in the regulon because in most cases the signal was close to the cut-off for significance. Also, 5 0 RACE experiments (below) revealed that in each case transcription originated from a promoter sequence that significantly matched that for known sF-controlled promoters. A list of the members of the sF regulon is presented in Table 1. One member of the sF regulon, rsfA, is known to encode a DNA-binding protein that inhibits the transcription of the sF-controlled gene spoIIR and is reported to stimulate the transcription of several other regulon members.17 A transcriptional profiling experiment in which RNA extracted from cells

19

Forespore Line of Gene Expression in B. subtilis Table 1. sF-regulated genes Possible operon

Gene

Ratioa sporulation

Ratiob induction

Functionc

F

A. Previously identified s -regulated genes bofC 1.9d csfB 15.1 dacF 4.2 gerAA –e gerAA-AB-AC gerAB 0.5d gerAC –e gpr 3.0 gpr-spoIIP spoIIP 3.1 katX 5.9 lonB 5.8 mutTA 1.1d rsfA 7.6

33.9 1.3d 132.4 8.8 13.5 185.0 71.8 55.6 10.8 159.4 2.7 4.8

sigG

1.2d

3.3

spoIIQ spoIIR spoIVB

11.9 3.6 2.1

158.3 1.1d 905.9

sspN 3.7 tlp 1.2d yhcN 1.1d ywhE 2.5 B. Newly identified sF-regulated genes sF promoter mapped by 5 0 -RACE-PCR pbpI 1.9d

–e 0.2d 4.2 32.3

sspN-tlp

4.1

yhfM 4.7 145.2 yhfW 4.5 11.9 ylbB 1.7d 38.7 ylbBC d 16.6 ylbC 1.7 yqhQ 3.1 1.3d yqzG 2.5 19.4 ytfI 1.4d 1414.0 ytfIJ ytfJ 7.2 42.4 yuiC 4.0 26.5 1.4d ywnJ 1.8d yyaC 8.7 19.2 Up-regulated in both sporulation and induction experiments 8.3 arsB-arsC arsB 1.4d arsC 2.2 4.7 cydD 3.0 89.0 ksgA 2.7 4.4 mcsA/ 2.0 3.1 yacH pgk-tpiA-pgm pgk 2.2 2.5 tpiA/tpi 2.3 3.1 pgm 2.5 4.2 tuaFGH tuaF 2.3 15.6 d tuaG 1.5 84.9 tuaH 3.4 13.4 yabT 2.8 29.1 yjbA 2.8 45.4 ypfB 3.8 3.4 yqhH 1.8d 5.1 yqhHG yqhG 2.5 16.8

Forespore regulator of the sigma-K checkpoint Sigma-F-transcribed gene Penicillin-binding protein Germination response to L-alanine germination response to L-alanine Germination response to L-alanine Spore protease (degradation of SASPs) Required for dissolution of the septal cell wall Major catalase in spores Lon-like ATP-dependent protease Antimutator 8-oxo-dGTPase Probable regulator of transcription of sigma-F-dependent genes RNA polymerase sporulation forespore-specific (late) sigma factor Required for completion of engulfment Required for processing of pro-sigma-E Intercompartmental signalling of pro-sigma-K Processing/activation in the mother-cell Small acid-soluble spore protein (minor) Small acid-soluble spore protein Similar to unknown proteins Similar to penicillin-binding protein

Similar to penicillin-binding protein, under sigma E-control (microarray) Unknown Similar to Rieske [2Fe-2S] iron-sulfur protein Similar to IMP dehydrogenase Similar to unknown proteins from B. subtilis Similar to unknown proteins Unknown Similar to unknown proteins Unknown Similar to unknown proteins Unknown Similar to gpr Probable arsenic resistance operon Reduction of arsenate to arsenite ABC transporter required for expression of cytochrome bd Dimethyladenosine transferase Modulator of CtsR repression Phosphoglycerate kinase Triose phosphate isomerase Phosphoglycerate mutase Biosynthesis of teichuronic acid Biosynthesis of teichuronic acid Biosynthesis of teichuronic acid Similar to serine/threonine-protein kinase Similar to unknown proteins Unknown Similar to SNF2 helicase Similar to unknown proteins

The underlined genes are conserved in at least five of the following species: B. licheniformis, B. anthracis, B. cereus, B. halodurans, O. iheyensis, C. acetobutylicum, C. perfringens, and absent in L. monocytogenes and L. innocua. a Ratios of relative RNA levels in sigFC versus sigF mutant. b Ratios of relative RNA levels from sigF overexpressed strain versus uninduced strain. c Function according to SubtiList (http://genolist.pasteur.fr/SubtiList). d Missed ratio cutoff for up-regulated gene. e Spot excluded from data analysis.

wild-type for RsfA was compared against RNA from cells mutant for RsfA (in a sG null mutant background) revealed a significant increase in the level of spoIIR expression in the absence of the

DNA-binding protein (but no other effects that met the cut-off for significance). The significance of the repressive effect of RsfA on spoIIR is considered below.

20 Genome-wide distribution of sF-controlled genes A physical map of the location of the members of the sF regulon reveals two noteworthy features (Figure 1). The first is that regulon members are predominantly located in the left-hand side of the genome. Indeed, of the 48 members of the regulon, 35 are located to the left of the axis of symmetry extending from the origin to 1808. We have no explanation for this enigmatic left-hand bias. The second feature is a conspicuous cluster of sFcontrolled genes that straddles, but is biased to the left of, the origin of replication, extending from position K563 kb (tuaF) relative to the origin to C102 kb (mcsA). This distribution is reminiscent of that previously seen for the binding sites for RacA, a sporulation-specific, DNA-binding protein that is responsible for anchoring the origin region to the cell pole. Binding sites for RacA are known to be distributed over a 612 kb stretch of the chromosome extending from position K412 kb to position C200 kb.23 Thus, at the time of asymmetric division genes in the cluster could be expected to be trapped in the nascent forespore chamber of the sporangium, an inference in accord with the results

Forespore Line of Gene Expression in B. subtilis

of previous studies. Given that the 612 kb centromere (also known as the “polar localization element”) is the portion of the chromosome closest to the cell pole, it might be expected that the portion of the chromosome that is 1808 opposite to it (that is, the 1708 region) would be the last chromosomal segment to enter the forespore during DNA translocation. If so, then it is interesting to note this distal portion of the chromosome is strikingly deficient in genes under the control of sF. Clustering of genes that are under sF control in and near binding sites for RacA makes good sense, as these genes would be appropriately situated to be switched on immediately after asymmetric division. Likewise, the under representation of sF-controlled genes in the portion of the chromosome distal to the centromere is consistent with the view that such genes would be the last to reach the chamber in which the forespore transcription factor is active. The sG regulon As in the case of sF and as described above, genes under sG control were identified from complementary sets of transcriptional profiling experiments

Figure 1. Map of the sF regulon. For operons only the first gene for each sF-controlled transcription unit is shown. The principal gene (spoIIR) repressed by RsfA is shown inside a circle. The dotted line divides the chromosome symmetrically from the origin to 1808. RacA binding sites are distributed across the region indicated by the arc.

21

Forespore Line of Gene Expression in B. subtilis

involving RNA from cells undergoing sporulation and RNA from cells engineered to produce the sporulation regulatory protein during growth. In this way, 95 genes were identified as members of the s G regulon, including 56 previously known members and 39 genes newly identified members (Table 2). Of the 39 newly identified members, nine were not activated by sG alone but additionally required the DNA-binding protein SpoVT as we explain below. Of the remaining 30 newly identified genes, 16 gave a signal that exceeded the cut-off in both sets of transcriptional profiling experiments, whereas the remaining 14 exhibited a significant signal only in profiling experiments involving expression during sporulation. Nonetheless, in most cases the signal under these conditions was high and in all cases 5 0 RACE experiments (see below) showed that transcription originated from a promoter that significantly conformed to the consensus for sG-controlled promoters. One of the members of the sG regulon is spoVT, which encodes a DNA-binding protein that is known to influence the expression of other sGcontrolled genes.20 To investigate the role of SpoVT in modulating the expression of other genes in the sG regulon and to do so on a genome-wide scale, a transcriptional profiling experiment was performed in which we compared RNA from sporulating cells of a strain that was wild-type for the DNA-binding protein against RNA from cells that were mutant for the protein. Of the 95 genes in the sG regulon, the level of expression of 47 genes was significantly affected by SpoVT (Table 2). In nine cases, as mentioned above, transcription was almost completely dependent upon SpoVT, whereas in 11 cases, transcription was significantly enhanced by the DNA-binding protein but not completely dependent on it. In 27 other cases transcription was inhibited by SpoVT. Noteworthy among the targets of SpoVT are sigG, whose expression it inhibits, and spoIVB, whose expression it stimulates. sigG is the gene for sG and spoIVB is the gene for an intercellular signaling protein that triggers the appearance of sK in the mother cell. Thus, SpoVT would appear to be a switch protein that shunts transcription from the forespore to the mother cell at an intermediate-tolate stage of sporulation. Finally, Figure 2 presents a map of the distribution of genes in the sG regulon, with genes under the negative control of SpoVT shown on the inside and genes whose expression is stimulated by SpoVT highlighted in bold on the outside. In striking contrast to the distribution seen for sF-controlled genes, members of the sG regulon are distributed throughout the chromosome, although sometimes in clusters. Computational predictions of sF and sG-controlled promoters The algorithms BioProspector and BioOptimizer were used to identify common motifs among

regions extending 200 bp upstream of each of the transcription units in the sF and sG regulons. For members of the sF regulon, a two-block-motif was stipulated with one block of 5 bp corresponding to the K35 region and a second block of 10 bp corresponding to the K10 region, and with a spacing of 14(G1) bp. For members of the sG regulon, a two-block motif was searched with 5 bp for the K35 element and 7 bp for the K10 region, and a spacing of 17 or 18 bp. Among the 36 upstream regions for sF-controlled transcription units that were analyzed, BioProspector/BioOptimizer recognized 11 of the 14 known sF-controlled promoters and identified potential promoter sequences in 20 of the remaining 22 cases. The sG dataset consisted of 61 upstream sequences (excluding genes whose transcription was strongly dependent on SpoVT). The algorithms identified 22 of the 33 known sG-promoters and found putative promoters for 22 of the remaining 28 cases. The consensus sequence logos derived from the BioProspector/BioOptimizer analysis of upstream regions are shown in Figure 3(a) for the sFcontrolled transcription units and Figure 3(c) for the sG-controlled units. Transcriptional mapping of sF and sG-controlled promoters To evaluate the validity of the BioProspector/ BioOptimizer predictions for promoters recognized by sF and sG, 5 0 -RACE-PCR was performed to map the transcription start site for ten of the newly identified sF transcription units and 19 of the newly identified s G units (Figure 4). A comparison of the mapped promoters against the computer analysis revealed that the success rate of BioProspector/BioOptimizer in predicting newly identified sF-controlled promoters was 80% (eight of ten mapped promoters), and for newly identified sG-controlled promoters 63% (12 of 19 mapped promoters). (The overall success rate for both newly and previously mapped promoters was 79% (19/24) in the case of sF and 65% (34/52) in the case of sG.) The consensus sequence logos obtained from previously known sF and sG promoters and the promoters mapped in this study are compared in Figure 3(a) to (d) from which we draw the following conclusions. First, the results derived from the computational analysis are in excellent accord with the consensus sequences obtained from promoters whose start sites were mapped. Second, and as noted previously, sF and sG-controlled promoters are strikingly similar to each other in their K10 and K35 sequence elements but differ from each other by the presence of a highly conserved G on the upstream side of the K10 element (position 20 in Figure 3(a) and (b)) and a somewhat conserved G at the adjacent position, whose functional significance had been demonstrated previously.24

22

Forespore Line of Gene Expression in B. subtilis

Table 2. sG-regulated genes Possible operon

Gene

Ratioa SigG

Ratiob induction

Ratioc SpoVT

1.0e 0.9e –g 2.2 –g 4.0 2.2 1.3e 1.3e 1.2e 2.3 1.1e 3.9 0.8e 1.1e 1.5e 0.8e 9.6 1.9e 1.8e 9.0

1.2 3.9f 0.4h 0.4h 1.2 0.4h 0.3h 0.3h 0.4h 0.3h 0.4h 0.3h 0.2h 0.9 0.4h 0.4h 0.5h 0.6 0.7 1.0 1.4

Functiond

G

A. Previously identified s -regulated genes bofC 1.1e coxA 3.5 csgA 5.2 dacF 0.7e exoA 0.8e glcU-gdh glcU 13.8 gdh 9.6 gerAA 2.4 gerAA-AB-AC gerAB 2.2 gerAC 2.5 gerBA 4.1 gerBA-BB-BC gerBB 2.9 gerBC 2.7 gerD 1.6e gerKA 2.5 gerKA-KC-KB gerKC 2.5 gerKB 1.8e gpr 1.1e gpr-spoIIP spoIIP 1.1e pdaA 4.1 rsfA 0.9e

sleB-ypeB splA-splB

spoVAA-VAB-VACVAD-VAE-VAF

sigG

10.0

8.9

0.4h

sleB ypeB splA

5.6 9.4 4.5

7.7 6.1 2.1

0.4h 0.3h –g

splB spoIVB

2.6 3.5

1.5e 3.3

1.1 2.0f

spoIVH spoVAA

1.2e 5.7

0.6e 4.8

1.1 1.9

spoVAB spoVAC spoVAD spoVAE spoVAF spoVT

6.1 7.3 7.6 6.1 5.5 1.8e

3.3 4.1 7.7 2.4 3.8 1.4e

1.6 1.6 1.5 1.3 1.3 1.0

3.5 9.0 6.4 2.5 4.2 3.1 2.0 9.0 8.7 9.8 7.0 7.4 1.4e 1.2e 2.1 2.9 5.0 1.6e 6.8 0.8e 6.2

2.7 2.5 –g 2.1 1.1e 1.0e 0.9e 2.1 1.8e 2.7 1.3e 1.4e 3.9 1.0e 1.0e 1.6e 0.9e –g 1.9e 1.4e 1.3e

0.7 2.3f –g 1.9 0.9 0.9 0.9 0.8 1.4 1.4 0.8 1.9 1.6 1.3 2.6f 2.3f 3.9f 2.9f 0.6 0.7 1.0

3.4

4.4

0.3h

2.3 2.8 3.1 9.7

3.5 –g 3.6 5.1

0.4h 0.4h 0.2h 0.9

sspA sspB sspC sspD sspE sspF sspH sspI sspJ sspK sspL sspM sspN-tlp sspN tlp sspO-sspP sspO sspP ybaK-cwlD ybaK cwlD yhcN yqfS ywhE B. Newly identified sG-regulated genes sG promoter mapped by 5 0 -RACE-PCR ctpB yckD ydfR ydfS yhcQ

Forespore regulator of the sigma-K checkpoint Spore cortex protein Sporulation-specific SASP protein Penicillin-binding protein Multifunctional DNA-repair enzyme Probable glucose uptake protein Glucose 1-dehydrogenase Germination response to L-alanine Germination response to L-alanine Germination response to L-alanine Germination response to AGFK Germination response to AGFK Germination response to AGFK Germination response to L-alanine and to AGFK Germination response to AGFK Germination response to AGFK Germination response to AGFK Spore protease (degradation of SASPs) Required for dissolution of the septal cell wall Production of muramic delta-lactam residues Probable regulator of transcription of sigma-Fdependent genes RNA polymerase sporulation forespore-specific (late) sigma factor Spore cortex-lytic enzyme Unknown; similar to unknown proteins Transcriptional negative regulator of the spore photoproduct lyase operon Spore photoproduct lyase Intercompartmental signalling of pro-sigma-K processing/activation in the mother-cell Similar to thioredoxin Mutants lead to the production of immature spores Mutants lead to the production of immature spores Mutants lead to the production of immature spores Mutants lead to the production of immature spores Mutants lead to the production of immature spores Mutants lead to the production of immature spores Transcriptional positive and negative regulator of Sigma-G-dependent genes Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Small acid-soluble spore protein Unknown N-acetylmuramoyl-L-alanine amidase (germination) Similar to unknown proteins Similar to endonuclease IV Similar to penicillin-binding protein Similar to PDZ-containing serine protease (sigK activation) Unknown Unknown; controlled by sigma-K (microarray) Unknown; controlled by sigma-K (microarray) Unknown (continued on next page)

23

Forespore Line of Gene Expression in B. subtilis

Table 2 (continued) Possible operon yitG-yitF yndD-yndE-yndF

yraG-yraF-adhByraE-yraD

Gene

Ratioa SigG

Ratiob induction

Ratioc SpoVT

yhcV yitG yitF yndD yndE yndF yoaR yozQ ypzA yraG

2.6 4.8 3.3 2.5 2.0 1.9e 2.0 17.8 3.4 9.5

1.3e 0.6e 0.5e 3.0 3.3 4.1 5.9 16.6 1.5e 1.1e

1.9 0.7 0.6 0.4h 0.4h 0.5h 0.6 0.4h 0.9 0.6

Similar to IMP dehydrogenase Similar to multidrug resistance protein Similar to mandelate racemase Similar to spore germination protein Similar to spore germination protein Similar to spore germination protein Unknown Unknown Unknown Similar to spore coat protein

0.5 0.6 0.5 0.7 2.6f 1.6 0.6 0.4h 0.6 0.4h 8.4f

Similar to spore coat protein Alcohol dehydrogenase Similar to spore coat protein Similar to spore coat protein Similar to unknown proteins Unknown Unknown Similar to NAD(P)H dehydrogenase Unknown Similar to permease Unknown

2.1f 0.6 7.8f 0.3h

Unknown Unknown Unknown Unknown

2.9f 2.8f 2.7f 4.4f 2.5f 3.0f 2.4f 2.8f 2.7f

Probable aminomethyltransferase Probable glycine decarboxylase Probable glycine decarboxylase Flavohemoglobin Similar to epoxide hydrolase Similar to unknown proteins Similar to ATP-dependent DNA ligase Unknown Unknown

yraF 9.0 0.9e adhE 8.7 1.2e yraE 5.1 1.9e yraD 4.7 1.0e yrrD 3.5 2.8 yteA 3.9 2.4 yuzA 3.5 1.5e yvaB 2.1 3.5 yvdQ 3.7 1.1e yveA 6.7 0.3e yxeD 6.9 4.3 Up-regulated in both sporulation and induction experiments yfhD 2.3 5.4 yfkD 3.5 2.5 yqfX 6.4 5.6 yusN 3.3 6.1 Up-regulated by SpoVT onlyi gcvT-gcvPA-gcvPB gcvT 0.9 1.1 gcvPA 0.9 0.8 gcvPB 0.9 1.0 hmp 2.9 0.9 yfhM 1.2 1.7 ykoV 3.8 1.2 ykoV-ykoU ykoU 1.7 2.0 ykzE 1.9 1.3 ylyA 1.3 1.2

Functiond

The underlined genes are conserved in at least five of the following species: B. licheniformis, B. anthracis, B. cereus, B. halodurans, O. iheyensis, C. acetobutylicum, C. perfringens, and absent in L. monocytogenes and L. innocua. a Ratios of relative RNA levels in sigGC versus sigG mutant. b Ratios of relative RNA levels from sigG overexpressed strain versus uninduced strain. c Ratios of relative RNA levels from spoVTC versus spoVTK mutant. d Function according to SubtiList (http://genolist.pasteur.fr/SubtiList). e Missed ratio cutoff for up-regulated gene. f Activated by sG and SpoVT. g Spot excluded from data analysis. h Activated by sG and repressed by SpoVT. i Genes in this category did not meet the criteria for sG regulation, but their expressions were significantly activated by SpoVT. These genes were included as members of the sG regulon.

Regulatory logic of the forespore line of gene transcription In summary, the forespore line of gene transcription involves the activation of 129 genes, which are grouped into 93 transcription units. The program of transcription is orchestrated by two RNA polymerase sigma factors, sF and sG, and two DNA-binding proteins, RsfA and SpoVT, which appear successively in the order: sF RsfA sG SpoVT. Fourteen genes are members of both the sF and sG regulons. The regulatory circuit is characterized by a series of FFLs. FFLs are circuits involving two regulatory proteins in which the primary one turns on the

synthesis of a secondary regulatory protein, and both then act together to control the expression of target genes. Two common kinds of FFLs are coherent type 1 FFLs, in which both regulatory proteins switch on the transcription of target genes, and incoherent type 1 FFLs, in which the primary protein acts as an activator and the secondary protein acts as a repressor.14,15 The forespore circuit consists of at least two incoherent FFLs and one coherent FFL (Figure 5). Thus, sF turns on the expression of RsfA, which then acts to repress spoIIR, representing an incoherent FFL. Regulatory circuits of this kind are thought to be responsible for causing gene expression to occur in a burst. The product of spoIIR

24

Forespore Line of Gene Expression in B. subtilis

Figure 2. Map of the sG regulon. For operons only the first gene for each sG-controlled transcription unit is shown. Genes repressed by SpoVT are inside the circle. Genes dependent on SpoVT for expression are indicated in bold.

is an intercellular signaling protein that is known to trigger the appearance of sE in the mother cell (through a pathway involving the processing of an inactive proprotein precursor to the sigma factor) almost immediately after sF is activated in the forespore. Thus, a principal function of the RsfA FFL may be to restrict SpoIIR synthesis to a brief period just after the activation of sF. Meanwhile, SpoVT, whose synthesis is turned on by sG, is the basis for both kinds of FFLs. Thus, some genes in the sG regulon are subject to a coherent, type 1, FFL in which their transcription is switched on or augmented by the synthesis of SpoVT. Conversely, other regulon members are subject to an incoherent, type 1, FFL in which SpoVT curtails their expression. In these regards, the role of SpoVT parallels, and is analogous to, that of the mother-cell-specific, DNA-binding protein GerE. Like SpoVT, whose synthesis is turned on at the terminal step of the forespore line of gene expression GerE synthesis is turned on by sK at the terminal step of the mother-cell-line of gene expression. Also, like SpoVT, GerE both activates the expression of certain members of its regulon while inhibiting the transcription of other regulon members. Finally, we note that the forespore line of gene expression is not a free-running program but is

linked to gene expression in the mother cell through an as yet poorly understood intercellular pathway that influences the activation of sG.11 Functional categories of genes in the forespore line of gene expression Members of the sF regulon fall into two principal categories: regulatory and those whose products directly contribute to morphogenesis or the resistance and germination properties of the spore, such as structural proteins and enzymes. Regulatory genes include sigG, which encodes sG;25 spoIIR, which signals activation of the mother cell-specific sigma factor sE;18,19 rsfA, which, as considered here, encodes a DNA-binding protein that represses spoIIR, and bofC, which is involved in the appearance of sK in the mother cell. A newly identified member of the sF regulon, mcsA, can be grouped in the regulation category. McsA is a modulator of CtsR repression.26 In non-stressed cells, CtsR is stabilized by McsA and acts as a repressor of heat shock genes, including clpC and clpP. The ClpCP protease mediates the degradation of SpoIIAB, which is the anti-sigma factor of sF. This raises the possibility that the activation of sF is inhibited by McsA. Genes whose products directly contribute to morphogenesis or spore resistance and germination

Forespore Line of Gene Expression in B. subtilis

25

Figure 3. Consensus promoter sequences for sF and sG-controlled promoters. Consensus sequences are displayed as sequence logos, in which the height of the letters indicates the information content at each position.86 (a) Consensus sequence logo for sF obtained from the compilation of 34 common motifs in regions upstream of the sF-regulated genes as identified by the BioProspector/BioOptimizer algorithm. (b) Consensus sequence logo for sF based on promoters for which the transcription start site was mapped previously (14 cases) or in the present work (ten cases) by 5 0 -RACE-PCR. (c) Consensus sequence logo for sG obtained from the compilation of 55 common motifs in upstream regions of sGregulated genes found by BioProspector/BioOptimizer. (d) Consensus sequence logo for sG based on promoters for which the transcription start site was mapped previously (33 cases) or in the present work (19 cases) by 5 0 -RACE-PCR.

properties include spoIIQ, which plays multiple roles in spore formation, including tethering certain proteins produced in the mother cell to the outer membrane surrounding the forespore;27–30 katX, which encodes a catalase that helps protect the spore from hydrogen peroxide;31 dacF32 and ywhE,33 which encode penicillin-binding proteins; the germination operon gerA;34 and gpr,35,36 which encodes a protease responsible for the degradation of SASPs during germination. Seven newly identified sF-controlled genes may fall in this category: pbpI (yrrR), whose product has similarity to a penicillin-binding protein and may function together with dacF and ywhE in the cross-linking of the cortical peptidoglycan; ywnJ, whose product contains a VanZ domain found in enzymes involved in vancomycin resistance and therefore may be important for peptidoglycan remodeling; yyaC, which exhibits sequence similarity to gpr and may play a role in SASP degradation; tuaF, tuaG and

tuaH members of the tua operon, which is involved in the biosynthesis of teichuronic acid; yqhH, which shows extensive similarity to a class of SNF helicases and whose product may be involved in changing DNA conformation and/or in DNA repair. In addition, it is interesting to note that two genes involved in arsenic resistance, arsB and arsC are expressed under the control of sF. These two genes are normally part of the larger ars operon (arsR-yqcK-arsB-arsC) expressed under the control of the DNA binding protein ArsR in the presence of arsenate or arsenite.37 Here, it seems that the requirement for ArsR is bypassed given that the putative promoter for sF is found within the yqcK gene, i.e. 625 bp downstream of the Pars promoter. The ars operon is located within the skin element, a sequence of 48 kb that interrupts the sigK gene. In the mother cell, skin is excised under the control of the recombinase SpoIVCA to allow expression of

26

Forespore Line of Gene Expression in B. subtilis

Figure 4. Positions of 5 0 termini of transcripts for sF and sG-controlled genes as mapped by 5 0 -RACE-PCR. The K35 and K10 sequence elements (uppercase letters in bold), the 5 0 ends of mRNAs (underlined uppercase), ribosomebinding sites (underlined lowercase), and transcription start sites (uppercase letters) are indicated. H is A or C or T, M is A or C, Y is C or T, W is A or T, R is A or G.

Figure 5. Regulatory circuits governing sporulation. Shown are the regulatory circuits for the predivisional cell, the forespore and the mother cell, with the forespore circuit highlighted in bold. See the text for a detailed description. AND Gates are so indicated. Arrowheads represent positive inputs and short, horizontal lines negative inputs. Note that the forespore and mother cell circuits are linked to each other at three points as indicated by the horizontal lines emanating from the signaling proteins SpoIIR and SpoIVB and via an unknown pathway from sE.

27

Forespore Line of Gene Expression in B. subtilis

an intact sK protein.38 In the forespore, however, skin remains integrated in the forespore chromosome because expression of the recombinase is dependent on the mother-cell-specific s factor sE. Having forespore-specific genes on skin may provide a selective advantage during sporulation (e.g. arsenic resistance) and help to ensure that skin is maintained in the germ cell. Members of the sG regulon can also be grouped into regulatory and genes whose products directly contribute to the proper formation of the spore. In the first category are spoIVB, which is involved in triggering the appearance of sK, and spoVT, which, as discussed above, encodes a DNAbinding protein that modulates the expression of other sG-controlled genes. Also present in this category are splA,39 which encodes a transcriptional regulator of the spl operon (which is involved in DNA repair) and ctpB, a gene previously shown to be under sE control.40 Like SpoIVB, CtpB is a PDZ domain-containing serine protease that contributes to activation of sK. Thus, ctpB belongs to a small subset of genes (discussed below) that are under the dual control of mother cell and forespore s factors. The second functional category includes cwlD41 and pdaA,42,43 which are involved in the production of muramic d-lactam, a chemically unique feature of the layer of peptidoglycan (the cortex) that surrounds the core of the spore; coxA, encoding a protein found in the cortex;44 sleB,45 which encodes a cortex lytic enzyme; dacF32 and ywhE,33 which encode penicillin-binding proteins (and which are also under the control of sF); yoaR, a newly identified member of the regulon, may be involved in cortex formation given that its protein product exhibits a VanW domain also found in enzymes that confer vancomycin resistance; the spoVA operon, whose products are responsible for the import of dipicolinic acid from the mother cell into the forespore;46 and germination genes such as the gerA,34 gerB47 and gerK operons and gerD.48 A newly identified member of the sG regulon, yndDEF, encodes proteins with similarity to the products of other germination operons, but a mutant of the operon is known not to affect germination.49 Another newly identified member of the sG regulon, the ykoVU operon, encodes DNA break repair proteins, whose functional significance is considered in detail below. Interestingly, and in contrast to the mother cell regulons, very few genes involved in metabolic reactions were identified in the forespore regulons. In the sG regulon, genes involved in metabolism include the glcU-gdh operon (for glucose uptake), the gcvT-gvcPA-gcvPB operon (for glycine degradation), yveA (previously identified as the major 50 L-Asp transporter in B. subtilis ) and the pgk-tpiApgm operon (which encode enzymes of the glycolytic pathway) in the sF regulon. Thus, it seems that the mother cell provides most of the energy and building blocks required for spore morphogenesis.

Systematic gene inactivation Because of the long history of genetic screens for mutants defective in sporulation and germination, it seemed unlikely that mutants of any of the newly identified members of the sF and sG regulons would represent new spo or ger genes with conspicuous phenotypes. Nevertheless, and with one exception, we systematically generated null mutants of the 23 newly identified genes (representing 19 operons) under the control of sF, and the 34 new members (representing 29 operons) of the sG regulon. In the case of operons, only the first gene in the transcription unit was deleted. The exceptional case was the pgk-tpiA-pgm operon in the sF regulon because it is essential for viability. Thus, a total of 47 (18C29) null mutants were created. The mutants were tested for their capacity to sporulate as measured by the production of heatresistant spores, and for the resulting spores to germinate using the tetrazolium overlay assay. None of the mutants exhibited an observable sporulation or germination defect. We also considered the possibility that functions for these genes were masked by functional redundancy with other members of the regulons.51 Accordingly, 56 double mutants were generated for the sF-controlled genes and 94 double mutants and six triple mutants were created for sG-controlled genes. All of the double and triple mutants were assayed for sporulation and germination efficiency, but no abnormal phenotypes were identified. Nevertheless, it is likely that at least some of these genes play a role in spore formation because they are widely conserved among other endospore-forming bacteria as we now consider. Comparative genomics Sporulation genes in B. subtilis are often conserved among other members of the genus Bacillus, such as Bacillus anthracis and Bacillus cereus and members of the genus Clostridium, but not among members of the genus Listeria, which is otherwise closely related to B. subtilis but does not form spores and lacks almost all sporulation genes.52 We therefore investigated the conservation of members of the sF and sG regulons in the genomes of the endospore-forming species Bacillus licheniformis, B. anthracis, B. cereus, Bacillus halodurans, Oceanobacillus iheyensis, Clostridium acetobutylicum, and Clostridium perfringens, and the non-endosporeforming species Listeria monocytogenes, and Listeria innocua (Supplementary Data, Tables 2 and 3). Among previously known members of the sF regulon, 15 orthologs were detected that were present in at least five of the seven endosporeforming species examined but were absent in the two Listeria species. Among the newly identified members of the regulon, 11 genes met the same criteria of being conserved in endospore-forming bacteria. Therefore, even though a conspicuous phenotype was not detected for mutants under

28 laboratory conditions, we infer that these 11 genes play a significant role in spore formation. Among previously known members of the sG regulon, 35 orthologs were detected that were present in at least five of the endospore-forming species but were absent in Listeria. Among the newly identified members of the regulon, 16 genes were conserved in endospore-forming bacteria but not in Listeria. In particular, the ykoVU operon, which we consider next, was present in some of the endospore-formers but not in Listeria. DNA strand break repair proteins YkoV and YkoU A striking finding from our transcriptional profiling analysis was the discovery that an operon, ykoVU, that governs the repair of DNA breaks is under the control of sG. The ykoV and ykoU gene products are homologous to repair proteins that mediate “non-homologous, end-joining” in eukaryotes (Ku and an ATP-dependent ligase, respectively), and mutants of the operon were previously reported to cause hypersensitivity to ionizing radiation in stationary phase cells.53 Given that ykoVU is under sG control and is widely conserved among endospore formers but is absent in Listeria, it seemed appealing to imagine that the principal function of the operon is in DNA repair during spore germination. Accordingly, spores were prepared from otherwise isogenic wild-type (PS832) and ykoV and ykoVU mutant strains and subjected to dry heat, a procedure that is known to cause breaks in spore DNA (possibly by generating apurinic/apyrimidinic sites).54 ykoV and ykoVU mutant spores were found to be significantly more sensitive to dry heat than were wild-type spores (Figure 6(a)). In contrast, wild-type spores and ykoV and ykoVU mutant spores were all equally sensitive to wet heat at 90 8C (data not shown), a treatment that does not cause DNA damage in spores containing a/b-type SASP.55 The a/b-type SASP

Forespore Line of Gene Expression in B. subtilis

are abundant, small proteins produced in the forespore under sG control that bind to the chromosome and protect it from many types of DNA damage.56 Hydrogen peroxide is also known to induce DNA breaks in spore DNA but only in mutant spores that lack the majority of their a/b-type SASP (aKbK spores).57 The results of Figure 6(b) show that the ykoV or ykoVU mutations did not cause sensitivity to hydrogen peroxide treatment by otherwise wildtype spores, but their presence in aKbK spores caused a striking decrease in the capacity of the spores to survive treatment with this oxidizing agent (Figure 6(c)). The ykoV and ykoVU mutations also sensitized aKbK spores significantly to wet heat at 84 8C (data not shown), a treatment that kills aKbK spores by causing DNA strand breaks, most likely by generating abasic sites.55 A time-course experiment during spore germination visualizing the sub-cellular localization of YkoV using a fusion to the green fluorescence protein (GFP) is consistent with the idea that these proteins function during germination to repair breaks in DNA. YkoV-GFP-harboring spores were germinated in Luria Broth at 37 8C (Figure 7). The results show that at zero time the spores were phase bright and the GFP signal was faint and diffuse in the spore core, but at 30 min, when the spores were phase dark (germination had commenced), the GFP signal was bright and located in the center of the spore core, presumably with the nucleoid. By 90 min, when the germinating spores had swollen, the GFP had disappeared (although autofluorescence from the spore shell could be seen). We note that the original report indicating that ykoVU operon mutations cause sensitivity to ionizing radiation was carried out with an overnight culture of cells that were well into stationary phase53 (Aidan Doherty, personal communication). Thus, it seems likely that the effect of the mutations was due to a sub-population of spores, rather than cells. This makes sense in that cells can repair

Figure 6. (a)–(c) Dry heat and hydrogen peroxide sensitivity of spores with and without YkoV and YkoU. Spores of various strains were prepared and cleaned as described.87 Spores were subjected to either (a) dry heat (90 8C), or (b) and (c) 5% hydrogen peroxide in 50 mM KPO4 buffer (pH 7) at 23 8C, and spore survival determined as described.49,52 The symbols used for the various strains in (a) and (b) are: ,, PS832 (wild-type); B, PS3721 (ykoV); and 6, PS3722 (ykoVU). The symbols used in (c) are: ,, PS355 (aKbK); B, PS3750 (aKbK ykoV); and 6, PS3751 (aKbK ykoVU).

29

Forespore Line of Gene Expression in B. subtilis

Table 3. Genes activated during sporulation Trancriptional regulator Spo0A Activated by Spo0A Repressed by Spo0A sF Repressed by RsfA sG Further activated by SpoVT Repressed by SpoVT Unaffected by SpoVT Activated by SpoVT only Total in the forespore sE Repressed by GerR Activated by SpoIIID Repressed by SpoIIID Unaffected by GerR or SpoIIIR sK Repressed by GerE Activated by GerE Unaffected by GerE Total in the mother cell

Figure 7. Localization of YkoV fusion to GFP. Spores from strain JL1 were germinated in LB medium at 37 8C and samples were taken at (a) 0 min, (b) 30 min, and (c) 90 min. The left panel shows the intensity and localization of the GFP signal, and the right panel represents the phase contrast picture of the spores.

double-strand breaks by an alternative pathway involving recA-mediated recombination. Since spores have only one chromosome, the recombination pathway cannot operate during spore germination.58 The overall program of sporulation gene expression The developmental process of spore formation involves three differentiating cell types: the predivisional cell, the mother cell and the forespore. Having now elucidated the forespore-line of gene expression, we revisit the overall program of gene expression involved in the conversion of a growing cell into a spore, beginning with the predivisional cell. The master regulator for the earliest phase of sporulation gene expression is the transcription factor Spo0A, which directly or indirectly controls at least 525 genes, representing about 12% of the 4106 annotated protein-coding genes in the genome.59 Of these 525 genes, 283 are up-regulated and 242 are down-regulated (Table 3), indicating that an important role of Spo0A is to shut down the expression of many genes that are normally expressed during vegetative growth. A ChIP-onchip analysis combined with gel electrophoretic mobility shift assays revealed that a total of 121 genes are direct targets of Spo0A.9 Evidence indicates that the levels of Spo0A rise gradually over the course of the first 2 h of sporulation and that some genes are turned on or off at a low level of

Genes (transcription units) 121 (54) 40 81 48 (36) 1 (1) 95 (67) 11 (9) 27 (17) 48 (35) 9 (6) 129 (93) 272 (171) 14 (10) 10 (8) 112 (62) 126 (90) 111 (71) 55 (36) 36 (27) 17 (12) 383 (242)

the master regulator whereas other genes require a high threshold concentration to be activated or repressed.10,60 Other transcriptional regulators that play a role at this early stage of sporulation are the stationary-phase sigma factor, sH,61 the transition state regulator, AbrB,62 and the master regulator for biofilm formation, SinR.63 Thus, gene expression in the predivisional cell is subject to complex regulation involving progressive increases in the level of the master regulator and the action of other, auxiliary regulatory proteins. Not only is Spo0A the master regulator for the earliest phase of sporulation gene expression but it comes to be a cell-specific transcription factor that accumulates to high levels selectively in the mother cell after the asymmetric division stage of development. Indeed, this cell-specific accumulation of Spo0A is believed to contribute to the cell-specific appearance of sE, the earliest-acting regulatory protein in the mother-cell line of gene expression.64 The mother cell and forespore programs of gene expression consist of about 504 genes (corresponding to about 12% of the protein-coding genes in the genome), with 383 mother-cell-specific genes and 129 forespore-specific genes (Table 3). Thus, the number of mother-cell-specific genes is about threefold higher than the number of forespore-specific genes. The overlap between the programs is limited to only eight genes (discussed below). Genes expressed after asymmetric division belong to four regulons, the mother cell regulons under the control of sE and sK and the forespore regulons under the control of sF and sG. By far the largest regulon is that of sE (253 genes), which has more members than the other three regulons combined, whereas the smallest is the sF regulon (48 genes). Several additional transcriptional regulators modulate gene expression within each regulon: RsfA in the sF regulon, SpoIIID and GerR

30 in the sE regulon, SpoVT in the sG regulon and GerE in the sK regulon. Of the approximately 504 genes that belong to the mother cell and forespore lines of gene expression, a total of 209 genes are repressed (by RsfA, GerR, SpoIIID, SpoVT or GerE) shortly after having been turned on, and thus are expressed only in a short pulse. By definition, the repressor and its corresponding sigma factor form an incoherent type 1 FFL. In contrast, 109 genes are activated further (by SpoIIID, SpoVT or GerE); as a result, full expression of these genes is delayed until significant levels of the activator protein have been produced. In such cases, the activator and its corresponding sigma factor form a coherent type 1 FFL. SpoIIID, SpoVT or GerE, because they act both as repressors and activators of gene expression, function as molecular switches between temporal classes of genes by ensuring that significant numbers of genes are turned off before the next class of genes is activated. Programs of gene expression within each compartment can be seen as hierarchical regulatory cascades of the form sF/ RsfA/sG/SpoVT in the forespore and sE/GerR/ SpoIIID/sK/GerE in the mother cell. The two lines of gene expression run in parallel and are connected at the post-transcriptional level by intercellular signaling pathways (Figure 5): sE activation is dependent on the secreted, signaling protein SpoIIR, whose expression is controlled by sF and RsfA; sG activation requires sE-dependent gene expression and sK activation requires the secreted, signal protein SpoIVB, whose expression is dependent on sG and SpoVT. As mentioned above, members of the forespore line of gene expression largely fall into two broad functional categories: regulatory genes and genes whose products directly contribute to the formation of the spore and the acquisition of its resistance and germination properties. Members of the sE regulon were previously assigned to regulation and to those involved in formation of the spore (such as genes involved in engulfment and the synthesis of spore protective structures) but in addition included numerous genes involved in metabolism.12 Similarly, members of the sK regulon were also assigned to regulation, spore morphogenesis and metabolism.13 Thus, a striking difference between the mother cell and the forespore lines of gene expression is the relative absence in the latter of genes performing metabolic functions. Given that the core of the spore is preparing for dormancy, it may not come as a surprise that metabolism is reduced there. Even several enzymes produced in the forespore under the control of sF and sG, such as the GPR protease and enzymes involved in DNA repair, become active only upon germination. As a further indication that the mother cell carries out most of the energy consuming tasks is that dipicolinic acid, which represents 10% of the dry weight of the spore,65 is synthesized in the mother cell and later imported into the forespore. Dipicolinic acid is produced by enzymes encoded by the sK-controlled spoVF operon and is imported into

Forespore Line of Gene Expression in B. subtilis

the spore core by the products of the sG-controlled spoVA operon.66 Spore morphogenesis requires the contribution of hundreds of genes and while some aspects of the morphogenetic process are specific to one compartment, others necessitate the participation of genes expressed from both compartments. For instance, shortly after asymmetric division, engulfment of the forespore by the mother cell is controlled by two mother-cell-specific genes, spoIID and spoIIM, one forespore-specific gene, spoIIQ, and one gene expressed in both compartments, spoIIP.30,67 In order to achieve maximal protection of the spore genetic material, at least three different macromolecular structures need to be assembled during sporulation: the spore coat, the spore cortex and the toroid-shaped spore chromosome. The spore coat is the outermost layer of the spore and consists of a shell of more than 50 proteins68 that are all expressed in the mother cell line of gene expression. Thus, more than 10% of the mother-cellspecific genes (and more than 1% of all protein coding genes in B. subtilis) encode components of the spore coat. A subset of forespore-specific genes, (yraD, yraE, yraF, yraG and yhcQ) display extensive sequence similarity to cotF, which codes for a wellcharacterized spore coat protein. However, it is unlikely that these forespore genes code for components of the spore coat. Tertiary structure predictions† using the protein sequences of YraG, YraF and CotF indicate that the CotF-similarity domain is a predicted four-helix bundle also found in iron-binding (ferritin-like) proteins. The spore cortex is a modified form of peptidoglycan synthesized between the inner and outer forespore membranes. It is essential for maintaining the spore core in a dehydrated state, which is the basis for heat resistance. Consistent with its intermediate location between the mother cell and the forespore compartments, synthesis of the cortex requires gene expression from both compartments. A unique feature of the cortex is the presence of muramic d-lactam, whose synthesis is catalyzed by the muramoyl-L-alanine amidase CwlD and the deacetylase PdaA.42,43 cwlD is expressed under the dual control of sG and sE and pdaA is expressed under the control of sG, whereas spoVIE (ybaN), a gene that encodes a putative deacetylase with similarity to PdaA,51 is under sE control. Another characteristic of the cortical peptidoglycan is its low degree of cross-linking, which is regulated by two D, 69 D-carboxypeptidases. The first one, DacB, is expressed from the mother cell compartment under the control of sE, whereas the second one, DacF,32 is expressed from the forespore compartment under the control of both sF and sG. The spore chromosome constitutes the only copy of the B. subtilis genome during dormancy. Therefore, it is critical to prevent damage to the spore † http://protevo.eb.tuebingen.mpg.de/toolkit/index. php?viewZhhpred

31

Forespore Line of Gene Expression in B. subtilis

DNA. If damage happens nevertheless, it has to be repaired efficiently during germination before the first round of replication. Whereas gene expression in the mother cell indirectly contributes to the protection of the genetic material by formation of the spore cortex and the spore coat structures, the most direct contributions to spore DNA protection and repair are achieved by the products of genes expressed in the forespore. At least 24 members of the forespore regulons (about 20% of the foresporespecific genes) are involved in this task. SASPs, which consist of small DNA binding proteins that bind to the spore chromosome, are the main players of this group of genes. In addition, several DNA repair enzymes, including the products of the sG-dependent ykoVU operon, complete the arsenal of spore-specific proteins involved in maintaining the integrity of the genetic material. As we have demonstrated, the sub-cellular localization of the Ku-like protein YkoV-GFP is consistent with a role in DNA repair during the early stages of germination. Our analysis has shown that although the number of genes expressed after asymmetric division is quite large, compartmentalization is extensive. The overlap between the forespore and mother cell regulons is limited to eight genes: ctpB (yvjB), coxA (yrbB), cwlD, pbpI (yrrR), spoIIP, spoIVH (ykvV, stoA), ydfR and ydfS. The coxA, cwlD, spoIIP and spoIVH genes form operons with the next gene upstream, whose expression is also compartmentalized during sporulation. Thus, a first compartment-specific promoter is found upstream of the operon and drives the expression of both genes, whereas in the intergenic region, a second promoter drives the expression in the other compartment of the downstream gene only. The majority of the genes expressed in both compartments appear to play a role in cortex formation (cwlD, pbpI, spoIVH) and/or their product has been located between the inner and outer forespore membrane (coxA, spoIIP). If they are secreted proteins, the side from which they reach this intermembrane space may not matter. However, if they remain associated with the membrane, it may be advantageous to have them on both sides of the nascent cortex. In the case of cwlD,70 it has been shown that expression from the mother cell side is critical, whereas expression from the forespore side is dispensable. By contrast, in the case of spoIVH,71 expression from the mother cell side is dispensable. In the case of spoIIP,72 it is likely that expression from both sides is required. Undoubtedly, our efforts and those of others to elucidate the complete program of sporulation gene expression are not fully complete. Nonetheless, to a substantial extent it is now possible to describe comprehensively the programs of gene expression in the three differentiating cell types of the developing sporangium. As a consequence, we can now see the overall logic of the regulatory circuits that operate during sporulation and ascribe functions to many of the hundreds of genes that are involved in transforming a cell into a spore.

Materials and Methods Strains All strains (Table 4) were derivatives of the wild-type strain PY79, with the exception of the sigF-inducible strain and strains PS355, PS832, PS3721, PS3722, PS3750 and PS3751, which are all isogenic and also derivatives of 168. SW3, SW7, SW171, SW211, RL3230, SW64, SW68, SW388, and TSBF1 were used for transcriptional profiling experiments. SW7 was obtained by transformation of strain RL1265 (sigFTkan)59 to spectinomycin resistance with chromosomal DNA from strain SW3 (sigE-sigGTspc). SW211 was obtained by transformation of strain RL560 (sigGTcat)73 to tetracycline resistance with chromosomal DNA from strain SW171 (rsfATtet). SW64 (sigGTcat, sigKTerm) was generated by transformation of strain RL560 with RL3230 (sigKTerm). SW68 (spoVTTspc, sigKTerm) was obtained by transformation of RL3230 to spectinomycin resistance with chromosomal DNA from SW21 (spoVTTspc). The strain used for the overexpression of sF, TSBF1 (amyETPxyl-sigF cm), was generated by double cross-over recombination of XhoI-lineared plasmid pMFsigF at the amyE locus. Strain SW372 (Pspank-hysigG), which was used for the overproduction of sG, was created by single cross-over integration of plasmid pSW1 at the sigG locus. Strains PS3721 and PS3750 were obtained by transformation of strains PS832 and PS355, respectively, to Macrolide, Lincosamide and Streptogramin B (MLS) resistance with chromosomal DNA from strain SW530 (ykoVTerm). Strains PS3722 and PS3751 were obtained by transformation of strains PS832 and PS355, respectively, to MLS resistance with chromosomal DNA from strain SW533 (ykoVUTerm). Strain JL1 (ykoVUpJL1 (ykoV-gfp) spcr) was obtained by single cross-over integration of plasmid pJL1 into PY79 and selection for spectinomycin resistance. Deletion mutants were generated by long-flanking homology PCR.74 Chromosomal DNA isolated from Table 4. B. subtilis strains used in this study Strain

Genotype

Source

RL3 RL24 RL322 RL560 RL1265 RL3217 RL3230 SW3 SW7 SW21 SW64 SW68 SW171 SW211 SW372 SW388 SW530 SW533 TSBF1 PS355 PS832 PS3721 PS3722 PS3750 PS3751 JL1

PY79 sspB-lacZUcat cotETcat sigGTcat sigFTkan gerETcat sigKTerm sigE-sigGTspc sigE-sigGTspc, sigFTkan spoVTTspc sigGTcat, sigKTerm spoVTTspc, sigKTerm rsfATtet rsfATtet, sigGTcat Pspank-hy-sigG Pspank-hy-sigG, sspB-lacZ ykoVTerm ykoVUTerm amyETPxyl-sigF aKbK in PS832 Wild-type ykoVTerm in PS832 ykoVUTerm in PS832 ykoVTerm, aKbK in PS355 ykoVUTerm, aKbK in PS355 ykoVUpJL1 (ykoVKgfp) spcr

Laboratory stock Laboratory stock Laboratory stock Laboratory stock Laboratory stock Laboratory stock Laboratory stock This study This study This study This study This study This study This study This study This study This study This study This study JBact 167:174 Laboratory stock This study This study This study This study This study

32 these strains was analyzed by PCR to confirm the insertion of the resistance cassette. Double mutants were generated by transformation of chromosomal DNA from one deletion strain into competent cells of a second mutant strain. Triple mutants were obtained by transformation of chromosomal DNA from one deletion mutant into a double mutant strain. Plasmids pMFsigF (Pxyl-sigF) was constructed by placing the sigF gene under the control of a xylose-inducible promoter Pxyl. The HindIII-BamHI digested sigF gene fragment, which was amplified using primers SigF-F (5 0 CCCAAGCTTTGGGAACAACGATTCGC-3 0 ) and SigF-R (5 0 -CGCGGATCCTCTCATTCATTCATCCGCTCG-3 0 ), was cloned into pMF20. The resultant clone was used to transform Escherichia coli JM105 to ampicillin resistance in order to generate plasmid pMFsigF. To induce sigG expression during vegetative growth, sigG was placed under the control of a LacI-repressible, isopropyl-ß-Dthiogalactopyranoside (IPTG)-inducible promoter Pspank-hy to generate plasmid pSW1 (Pspank-hy-sigG Nter). A 536 bp fragment containing an optimal ribosome binding site,75 the ATG start codon, and the N-terminal coding sequence of sigG, was PCR-amplified with primers swp373 (5 0 -GCCGTCGACACATAAGGAG GAACTACTATGTCGAGAAATAAAGTCGAAATCTG3 0 ) and swp375 (5 0 -GACATGCATGCCCGCCGTCGTTA TAGATCGGTTCAAATAGAG-3 0 ). The SalI-SphI digested PCR product was cloned into pPE3061 and transformed into E. coli DH5a cells to generate plasmid pSW1. Plasmid pJL1 (ykoV-gfp) was obtained by PCR amplification of the 3 0 end of ykoV with primers JLL1 ACTGACGGATCCT GAAAATTGTATCGTCATGGAGTC and JLL2 ACT GACCTCGAGTGATGTGCCGGAGGCTTTT from PY79 chromosomal DNA, followed by digestion with XhoI and BamHI (restriction sites underlined) and cloning into pCVO119.76 Growth and sporulation conditions Strains SW3, SW7, SW171, SW211, RL3230, SW64, and SW68 were grown in a 37 8C shaking water bath in hydrolyzed casein medium to A600nm of 0.6. After centrifugation at 4000g for 5 min, pellets were resuspended in Sterlini-Mandelstam medium.77,78 Sporulating cultures were returned to growth at 37 8C, and samples were collected at the indicated times after resuspension. A culture (0.5 ml) of the sF overexpression strain TSBF1 at A600nm of 0.8 was inoculated into 100 ml of LB medium with or without 1 mM xylose. Cells were grown at 37 8C and harvested 2 h after xylose induction (A600nm of 0.8). SW388 was grown in 50 ml of LB medium at 37 8C to A600nm of 0.5, at which time the culture was split into two and IPTG was added to one of the cultures to a final concentration of 1 mM. At 2 h after IPTG induction, samples were harvested from both the induced and the uninduced cultures for RNA extraction. Transcriptional profiling with oligonucleotide microarrays Construction of oligonucleotide microarrays All transcriptional profiling experiments were carried out with oligonucleotide microarrays except those involving engineering synthesis of sF during growth,

Forespore Line of Gene Expression in B. subtilis

which is described below. B. Subtilis OligoLibrarye (96 well format, 43 plates) was purchased from Sigma Genosys. The oligonucleotides in this collection were 65mers with a 5 0 -C6 amino modification. One oligo was designed and sequence-optimized to minimize crosshybridization of homologous genes for each gene in the B. subtilis genome. The library consisted of 4128 oligos representing 4106 B. subtilis genes, ten controls from E. coli and the Brome mosaic virus, and 12 random oligonucleotides. Each 1 nmol/well oligonucleotide was resuspended in 25 ml of distilled water to yield an initial concentration of 40 mM. 15 ml of the 40 mM oligonucleotide was then transferred to 385 well plates and 15 ml of 2X QMT Spotting Solution I (Quantifoil) was added to generate a final concentration of 20 mM. The oligonucleotides were spotted on aldehyde slides (Quantifoil) with the GeneMachines OmniGrid arrayer. Within one to two days after printing the oligos on aldehyde slides, the array was immobilized. A humid chamber (Sigma) was prepared by filling 1/3 with distilled water and placed at 37 8C for 10 min. The slides were placed in the humid chamber at 37 8C for 15 min, and transferred to a 120 8C oven for 60 min. The immobilized slides were stored in a dark and desiccated box until use. Immediately before hybridization of labeled cDNA, the slide was prepared by washing twice in 0.2% (w/v) sodium dodecyl sulfate, twice in milliQ water, blocking in aldehyde blocking solution (2.5% (w/v) NaBH4, 75% (v/v) PBS, 20% (v/v) ethanol), followed by washing twice in 0.2% SDS, twice in milliQ water, and centrifuged at 600 rpm for 5 min. Sample preparation and RNA isolation Samples of sporulating cells were immediately mixed with equal volume of methanol at K20 8C and centrifuged for 5 min at 4000g, 4 8C. Cell pellets were stored at K80 8C until RNA extraction. RNA was isolated by a hot acid-phenol procedure as described.59 The quantity of RNA was measured by A260nm, and the quality was determined by the ratio of A260nm/A280nm and examined by visualizing the 23 S and 16 S rRNA bands on an agarose gel. Labeling and hybridization Labeled cDNA was obtained from RNA by direct incorporation of Cy3 or Cy5 labeled dUTP. 25 mg of RNA was incubated with 1 mg of random hexamers at 70 8C for 10 min and then placed on ice for 2 min. Either Cy3-dUTP or Cy5-dUTP (Perkin Elmer Life Sciences) was added to the RNA. A reaction mix containing reverse transcription buffer, 10 mM dithiothreitol, deoxynucleoside triphosphate (1 mM dATP, 1mM dCTP, 1 mM dGTP, and 0.4 mM dTTP), and SuperScript II Reverse Transcriptase (Invitrogen) was added to the RNA mixture. The reaction was incubated at room temperature for 10 min and 42 8C for 2 h. The reaction was inactivated by heating at 70 8C for 10 min. RNA was digested by incubation with RNase H for 20 min at 37 8C. The differentially labeled cDNAs were mixed prior to the removal of unincorporated nucleotides by QiaQuick purification spin columns (Qiagen). The volume of labeled cDNA was reduced to 5 ml. A hybridization mixture (5 ml of labeled probe, 1 ml of 20 mg/ml yeast tRNA, 3.75 ml of 20! SSC, 0.75 ml of 10% SDS, and 4.5 ml of 100% formamide) was spotted onto the prepared microarray slide (as described above) and

33

Forespore Line of Gene Expression in B. subtilis

covered with a coverslip in a CMT-hybridization chamber (Corning). Hybridization was performed overnight in a 42 8C water bath. The slide was washed for 1 min in each of the following buffers: 2! SSC supplemented with 0.2% SDS, 2! SSC, and 0.2! SSC (SSC is 0.15 NaCl, 0.015 trisodium citrate (pH7.0)). After centrifugation at 800 rpm for 2 min, the array was scanned on a GenePix 4000B scanner (Axon Instruments).

a second gene-specific primer located at about 150 bp downstream from the predicated promoter. The PCR products were sequenced using the second gene-specific primers. The nucleotide immediately preceding the poly(A) residues was defined as the mapped transcription start site.

Transcriptional profiling with DNA microarrays

B. subtilis genome sequence was obtained from SubtiList†, L. monocytogenes (strain EGD-e) and L. innocua (strain CLIP 11262) genome sequences were obtained from ListiList‡, and B. licheniformis (strain ATCC14580), B. anthracis (strain Ames Ancestor), B. cereus (strain ATCC14579), B. halodurans (strain C-125), O. iheyensis (strain HTE831), C. acetobutylicum (strain ATCC 824), and C. perfringens (strain 13) genome sequences were obtained from the NCBI website§. Paralogues were searched by using Blast on the SubtiList website. An E value O1!10K20 was considered of high similarity to the protein of interest and an E value between 1!10K20 and 1!10K4 was considered low similarity. Orthologues were identified by blasting the protein sequence corresponding to the gene of interest on NCBI and by using the SmithWaterman report (Nrprot) on SubtiList.

DNA microarray based on PCR-amplified open-reading frames were used for the transcriptional profiling experiments involving sF production during growth. The procedures were as described.12 Statistical data analysis Statistical analysis of the microarray data was carried out as described by Eichenberger et al.12 Briefly, each slide was first normalized with an iterative rank-variant method. The normalized replicate slides were then combined using a Bayesian hierarchical model79 that incorporated experimental variations across slides. The posterior median of the log-ratio of expression for each gene and the corresponding Bayesian confidence interval were produced by a Markov chain Monte Carlo implementation of the model using 4000 iterations. The score of the gene is defined as the posterior probability for a positive log-expression ratio. Promoter search BioProspector and BioOptimizer The programs BioProspector80 and BioOptimizer81 were employed to predict the sF and sG-controlled promoters for the newly identified genes. BioProspector is a stochastic algorithm for finding conserved motifs across a set of DNA sequences. The program is capable of identifying dimeric motifs with variable gaps, such as in the cases of sF and sG promoters. The 200 bp long regions upstream of every gene in the sF and sG regulons, including previously reported genes, were analyzed by BioProspector. BioOptimizer is an optimization algorithm that improves upon the BioProspector motifs. The scoring function used by BioOptimizer is the exact log-posterior distribution of the Bayesian motif-discovery model. The BioProspector motifs for sF and sG were subjected to BioOptimizer analysis in order to select the best motifs.

Comparative genomics

Germination assays Germination efficiency was assayed using 2,3,5-triphenyltetrazolium chloride (Tzm) overlay as described.13,78 Strains PE454 (gerETcat)13 and RL322 (cotETcat)84 were utilized as negative controls and wild-type PY79 as a positive control. Sporulation was induced in Difco sporulation medium (DSM) and allowed for completion by incubation at 37 8C for three days. Heat activation of spores and killing of vegetative cells was performed in a 60 8C oven for 3 h. Measuring spore resistance Heat in water Sporulation efficiency as determined by heat resistance was performed as described.76,78 Briefly, cells were grown to exhaustion in DSM at 37 8C for 30 h. Serial tenfold dilution of the culture was carried out six times in T base supplemented with 1 mM MgSO4. Then 100 ml of the 10K4 and 10K5 dilutions were plated on LB agar plates. The dilutions were incubated at 80 8C for 15 min, and 100 ml of heat-treated cells were plated on LB. After incubation at 37 8C overnight, the number of colonies was determined.

5 0 -RACE-PCR The transcription start sites of ten newly identified sFcontrolled and 18 new sG -controlled genes were determined using rapid amplification of cDNA ends (RACE)-PCR.12,82,83 A total of 25 mg of RNA extracted from SW3 (sigFC) and RL3230 (sigGC) was reverse transcribed using the SuperScript II system (Invitrogen) in the presence of up to 16 gene-specific primers (sequence available upon request) located at about 400 bp downstream from the predicated promoter. The cDNA generated was purified with QiaQuick purification spin column (Qiagen). A homopolymeric T-tail was added to the 3 0 end of cDNA in a terminal transferase reaction (Roche). The sample was purified using a QiaQuick column. PCR was performed with poly(T) tailed cDNA as the template, a poly(dA) primer and

Dry heat A total of 100 ml of purified spores in dH2O at A600nm of 1.0 was dried in sterile glass test tubes. Dry spores were heated in an oil bath at 120 8C for 0, 30, 60, 90, and 120 min. After cooling on ice, 1 ml of dH2O was added to the dry spores. Following sonication for 30 s, 10 ml aliquots of serial dilutions were spotted on LB plates with appropriate antibiotics. The plates were incubated at 30 8C overnight before shifting to 37 8C. All time point experiments were performed in duplicate. † http://genolist.pasteur.fr/SubtiList/ ‡ http://genolist.pasteur.fr/ListiList/ § http://www.ncbi.nlm.nih.gov/

34 Ultraviolet radiation Resistance to ultraviolet radiation was measured as described.85 UV resistance was determined as the dose in Joules/m2 necessary to result in a tenfold reduction in spore viability. Hydrogen peroxide Measurement of spore resistance to hydrogen peroxide was performed as described.57 Spore preparation, germination and microscopy Cells of strain JL1 were grown to exhaustion in 20 ml of DSM supplemented with 20 ml of 0.01 M MnCl2, 20 ml 1 M Ca(NO3)4 and 20 ml 1 mM FeSO4 for 48 h at 37 8C in a shaking water bath. The culture was then removed from the water bath and pelleted by tabletop centrifugation (5000 rpm), and the pellet resuspended in 5 ml cold dH2O. After five sequential washes, the pellet was resuspended in 4 ml of cold dH2O, pelleted, and finally resuspended in 2 ml of cold dH20. A 2 ml volume of purified spores in dH2O was pelleted by centrifugation, and the pellet resuspended in 1.2 ml of Luria-Bertani broth and placed in a 37 8C water bath. A 3 ml volume was removed at 10 min intervals, and placed on a microscope slide. The sample was covered with a poly L-lysine treated coverslip. Images were obtained using a Nikon 90i motorized fluorescent microscope and Roper 1K monochrome digital camera. Image acquisition preformed with Metamorph software version 6.3, and processed with Adobe Photoshop.

Acknowledgements We thank P. Piggot for advice on the manuscript and A. Doherty for discussions and for communicating unpublished results. This work was supported by a Grant-in-aid for Scientific Research on Priority Areas (genome biology) from the Ministry of Education, Science, Sports and Culture of Japan (to T.S.), by the Department of the Army award number W81XWH-04-01-0307 (to P.E.), by the Army Research Office (to P.S.) and by NIH grant GM18568 (to R.L.). The content of this material does not necessarily reflect the position or the policy of the Government and no official endorsement should be inferred.

Supplementary Data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/ j.jmb.2006.01.059

References 1. Piggot, P. J. & Losick, R. (2002). Sporulation genes and inter-compartmental regulation. In Bacillus subtilis and its Closest Relatives: From Genes to Cells (Sonenshein, A. L., Hoch, J. A. & Losick, R., eds), pp. 483–518, American Society for Microbiology, Washington, DC.

Forespore Line of Gene Expression in B. subtilis

2. Hilbert, D. W. & Piggot, P. J. (2004). Compartmentalization of gene expression during Bacillus subtilis spore formation. Microbiol. Mol. Biol. Rev. 68, 234–262. 3. Hoch, J. A. (1993). Regulation of the phosphorelay and the initiation of sporulation in Bacillus subtilis. Annu. Rev. Microbiol. 47, 441–465. 4. Stephenson, K. & Hoch, J. A. (2002). Evolution of signalling in the sporulation phosphorelay. Mol. Microbiol. 46, 297–304. 5. Strauch, M. A., Spiegelman, G. B., Perego, M., Johnson, W. C., Burbulys, D. & Hoch, J. A. (1989). The transition state transcription regulator abrB of Bacillus subtilis is a DNA binding protein. EMBO J. 8, 1615–1621. 6. Fujita, M. & Sadaie, Y. (1998). Feedback loops involving Spo0A and AbrB in in vitro transcription of the genes involved in the initiation of sporulation in Bacillus subtilis. J. Biochem. (Tokyo), 124, 98–104. 7. Perego, M., Spiegelman, G. B. & Hoch, J. A. (1988). Structure of the gene for the transition state regulator, abrB: regulator synthesis is controlled by the spo0A sporulation gene in Bacillus subtilis. Mol. Microbiol. 2, 689–699. 8. Zuber, P. & Losick, R. (1987). Role of AbrB in Spo0Aand Spo0B-dependent utilization of a sporulation promoter in Bacillus subtilis. J. Bacteriol. 169, 2223–2230. 9. Molle, V., Fujita, M., Jensen, S. T., Eichenberger, P., Gonzalez-Pastor, J. E., Liu, J. S. & Losick, R. (2003). The Spo0A regulon of Bacillus subtilis. Mol. Microbiol. 50, 1683–1701. 10. Fujita, M. & Losick, R. (2005). Evidence that entry into sporulation in Bacillus subtilis is governed by a gradual increase in the level and activity of the master regulator Spo0A. Genes Dev. 19, 2236–2244. 11. Losick, R. & Stragier, P. (1992). Crisscross regulation of cell-type-specific gene expression during development in B. subtilis. Nature, 355, 601–604. 12. Eichenberger, P., Jensen, S. T., Conlon, E. M., van Ooij, C., Silvaggi, J., Gonzalez-Pastor, J. E. et al. (2003). The sigmaE regulon and the identification of additional sporulation genes in Bacillus subtilis. J. Mol. Biol. 327, 945–972. 13. Eichenberger, P., Fujita, M., Jensen, S. T., Conlon, E. M., Rudner, D. Z., Wang, S. T. et al. (2004). The program of gene transcription for a single differentiating cell type during sporulation in Bacillus subtilis. PLoS Biol. 2, e328. 14. Milo, R., Shen-Orr, S., Itzkovitz, S., Kashtan, N., Chklovskii, D. & Alon, U. (2002). Network motifs: simple building blocks of complex networks. Science, 298, 824–827. 15. Shen-Orr, S. S., Milo, R., Mangan, S. & Alon, U. (2002). Network motifs in the transcriptional regulation network of Escherichia coli. Nature Genet. 31, 64–68. 16. Mangan, S. & Alon, U. (2003). Structure and function of the feed-forward loop network motif. Proc. Natl Acad. Sci. USA, 100, 11980–11985. 17. Juan Wu, L. & Errington, J. (2000). Identification and characterization of a new prespore-specific regulatory gene, rsfA, of Bacillus subtilis. J. Bacteriol. 182, 418–424. 18. Londono-Vallejo, J. A. & Stragier, P. (1995). Cell-cell signaling pathway activating a developmental transcription factor in Bacillus subtilis. Genes Dev. 9, 503–508. 19. Karow, M. L., Glaser, P. & Piggot, P. J. (1995). Identification of a gene, spoIIR, that links the activation of sigma E to the transcriptional activity of sigma F during sporulation in Bacillus subtilis. Proc. Natl Acad. Sci. USA, 92, 2012–2016.

35

Forespore Line of Gene Expression in B. subtilis

20. Bagyan, I., Hobot, J. & Cutting, S. (1996). A compartmentalized regulator of developmental gene expression in Bacillus subtilis. J. Bacteriol. 178, 4500–4507. 21. Steil, L., Serrano, M., Henriques, A. O. & Volker, U. (2005). Genome-wide analysis of temporally regulated and compartment-specific gene expression in sporulating cells of Bacillus subtilis. Microbiology, 151, 399–420. 22. Cabrera-Hernandez, A., Sanchez-Salas, J. L., Paidhungat, M. & Setlow, P. (1999). Regulation of four genes encoding small, acid-soluble spore proteins in Bacillus subtilis. Gene, 232, 1–10. 23. Ben-Yehuda, S., Fujita, M., Liu, X. S., Gorbatyuk, B., Skoko, D., Yan, J. et al. (2005). Defining a centromerelike element in Bacillus subtilis by Identifying the binding sites for the chromosome-anchoring protein RacA. Mol. Cell, 17, 773–782. 24. Sun, D., Fajardo-Cavazos, P., Sussman, M. D., Tovar-Rojo, F., Cabrera-Martinez, R. M. & Setlow, P. (1991). Effect of chromosome location of Bacillus subtilis forespore genes on their spo gene dependence and transcription by E sigma F: identification of features of good E sigma F-dependent promoters. J. Bacteriol. 173, 7867–7874. 25. Sun, D. X., Cabrera-Martinez, R. M. & Setlow, P. (1991). Control of transcription of the Bacillus subtilis spoIIIG gene, which codes for the forespore-specific transcription factor sigma G. J. Bacteriol. 173, 2977–2984. 26. Kruger, E., Zuhlke, D., Witt, E., Ludwig, H. & Hecker, M. (2001). Clp-mediated proteolysis in Gram-positive bacteria is autoregulated by the stability of a repressor. EMBO J. 20, 852–863. 27. Sun, Y. L., Sharp, M. D. & Pogliano, K. (2000). A dispensable role for forespore-specific gene expression in engulfment of the forespore during sporulation of Bacillus subtilis. J. Bacteriol. 182, 2919–2927. 28. Doan, T., Marquis, K. A. & Rudner, D. Z. (2005). Subcellular localization of a sporulation membrane protein is achieved through a network of interactions along and across the septum. Mol. Microbiol. 55, 1767–1781. 29. Blaylock, B., Jiang, X., Rubio, A., Moran, C. P., Jr & Pogliano, K. (2004). Zipper-like interaction between proteins in adjacent daughter cells mediates protein localization. Genes Dev. 18, 2916–2928. 30. Londono-Vallejo, J. A., Frehel, C. & Stragier, P. (1997). SpoIIQ, a forespore-expressed gene required for engulfment in Bacillus subtilis. Mol. Microbiol. 24, 29–39. 31. Bagyan, I., Casillas-Martinez, L. & Setlow, P. (1998). The katX gene, which codes for the catalase in spores of Bacillus subtilis, is a forespore-specific gene controlled by sigmaF, and KatX is essential for hydrogen peroxide resistance of the germinating spore. J. Bacteriol. 180, 2057–2062. 32. Schuch, R. & Piggot, P. J. (1994). The dacF-spoIIA operon of Bacillus subtilis, encoding sigma F, is autoregulated. J. Bacteriol. 176, 4104–4110. 33. Pedersen, L. B., Ragkousi, K., Cammett, T. J., Melly, E., Sekowska, A., Schopick, E. et al. (2000). Characterization of ywhE, which encodes a putative highmolecular-weight class A penicillin-binding protein in Bacillus subtilis. Gene, 246, 187–196. 34. Feavers, I. M., Foulkes, J., Setlow, B., Sun, D., Nicholson, W., Setlow, P. & Moir, A. (1990).

35.

36.

37. 38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

The regulation of transcription of the gerA spore germination operon of Bacillus subtilis. Mol. Microbiol. 4, 275–282. Sussman, M. D. & Setlow, P. (1991). Cloning, nucleotide sequence, and regulation of the Bacillus subtilis gpr gene, which codes for the protease that initiates degradation of small, acid-soluble proteins during spore germination. J. Bacteriol. 173, 291–300. Sanchez-Salas, J. L. & Setlow, P. (1993). Proteolytic processing of the protease which initiates degradation of small, acid-soluble proteins during germination of Bacillus subtilis spores. J. Bacteriol. 175, 2568–2577. Sato, T. & Kobayashi, Y. (1998). The ars operon in the skin element of Bacillus subtilis confers resistance to arsenate and arsenite. J. Bacteriol. 180, 1655–1661. Stragier, P., Kunkel, B., Kroos, L. & Losick, R. (1989). Chromosomal rearrangement generating a composite gene for a developmental transcription factor. Science, 243, 507–512. Pedraza-Reyes, M., Gutierrez-Corona, F. & Nicholson, W. L. (1994). Temporal regulation and foresporespecific expression of the spore photoproduct lyase gene by sigma-G RNA polymerase during Bacillus subtilis sporulation. J. Bacteriol. 176, 3983–3991. Pan, Q., Losick, R. & Rudner, D. Z. (2003). A second PDZ-containing serine protease contributes to activation of the sporulation transcription factor sigmaK in Bacillus subtilis. J. Bacteriol. 185, 6051–6056. Sekiguchi, J., Akeo, K., Yamamoto, H., Khasanov, F. K., Alonso, J. C. & Kuroda, A. (1995). Nucleotide sequence and regulation of a new putative cell wall hydrolase gene, cwlD, which affects germination in Bacillus subtilis. J. Bacteriol. 177, 5582–5589. Fukushima, T., Kitajima, T. & Sekiguchi, J. (2005). A polysaccharide deacetylase homologue, PdaA, in Bacillus subtilis acts as an N-acetylmuramic acid deacetylase in vitro. J. Bacteriol. 187, 1287–1292. Fukushima, T., Yamamoto, H., Atrih, A., Foster, S. J. & Sekiguchi, J. (2002). A polysaccharide deacetylase gene (pdaA) is required for germination and for production of muramic delta-lactam residues in the spore cortex of Bacillus subtilis. J. Bacteriol. 184, 6007–6015. Takamatsu, H., Hiraoka, T., Kodama, T., Koide, H., Kozuka, S., Tochikubo, K. & Watabe, K. (1998). Cloning of a novel gene yrbB, encoding a protein located in the spore integument of Bacillus subtilis. FEMS Microbiol. Letters, 166, 361–367. Moriyama, R., Fukuoka, H., Miyata, S., Kudoh, S., Hattori, A., Kozuka, S. et al. (1999). Expression of a germination-specific amidase, SleB, of Bacilli in the forespore compartment of sporulating cells and its localization on the exterior side of the cortex in dormant spores. J. Bacteriol. 181, 2373–2378. Tovar-Rojo, F., Chander, M., Setlow, B. & Setlow, P. (2002). The products of the spoVA operon are involved in dipicolinic acid uptake into developing spores of Bacillus subtilis. J. Bacteriol. 184, 584–587. Corfe, B. M., Moir, A., Popham, D. & Setlow, P. (1994). Analysis of the expression and regulation of the gerB spore germination operon of Bacillus subtilis 168. Microbiology, 140, 3079–3083. Kemp, E. H., Sammons, R. L., Moir, A., Sun, D. & Setlow, P. (1991). Analysis of transcriptional control of the gerD spore germination gene of Bacillus subtilis 168. J. Bacteriol. 173, 4646–4652.

36 49. Paidhungat, M. & Setlow, P. (2000). Role of ger proteins in nutrient and nonnutrient triggering of spore germination in Bacillus subtilis. J. Bacteriol. 182, 2513–2519. 50. Lorca, G., Winnen, B. & Saier, M. H., Jr (2003). Identification of the L-aspartate transporter in Bacillus subtilis. J. Bacteriol. 185, 3218–3222. 51. Silvaggi, J. M., Popham, D. L., Driks, A., Eichenberger, P. & Losick, R. (2004). Unmasking novel sporulation genes in Bacillus subtilis. J. Bacteriol. 186, 80898095. 52. Stragier, P. (2002). A gene odyssey: exploring the genomes of endospore-forming bacteria. In Bacillus subtilis and its Closest Relatives: From Genes to Cells (Sonenshein, A. L., Hoch, J. A. & Losick, R., eds), pp. 519–526, American Society for Microbiology, Washington, DC. 53. Weller, G. R., Kysela, B., Roy, R., Tonkin, L. M., Scanlan, E., Della, M. et al. (2002). Identification of a DNA nonhomologous end-joining complex in bacteria. Science, 297, 1686–1689. 54. Setlow, B. & Setlow, P. (1995). Small, acid-soluble proteins bound to DNA protect Bacillus subtilis spores from killing by dry heat. Appl. Environ. Microbiol. 61, 2787–2790. 55. Fairhead, H., Setlow, B. & Setlow, P. (1993). Prevention of DNA damage in spores and in vitro by small, acidsoluble proteins from Bacillus species. J. Bacteriol. 175, 1367–1374. 56. Setlow, P. (1995). Mechanisms for the prevention of damage to DNA in spores of Bacillus species. Annu. Rev. Microbiol. 49, 29–54. 57. Setlow, B. & Setlow, P. (1993). Binding of small, acidsoluble spore proteins to DNA plays a significant role in the resistance of Bacillus subtilis spores to hydrogen peroxide. Appl. Environ. Microbiol. 59, 3418–3423. 58. Frenkiel-Krispin, D., Sack, R., Englander, J., Shimoni, E., Eisenstein, M., Bullitt, E. et al. (2004). Structure of the DNA-SspC complex: implications for DNA packaging, protection, and repair in bacterial spores. J. Bacteriol. 186, 3525–3530. 59. Fawcett, P., Eichenberger, P., Losick, R. & Youngman, P. (2000). The transcriptional profile of early to middle sporulation in Bacillus subtilis. Proc. Natl Acad. Sci. USA, 97, 8063–8068. 60. Fujita, M., Gonzalez-Pastor, J. E. & Losick, R. (2005). High- and low-threshold genes in the Spo0A regulon of Bacillus subtilis. J. Bacteriol. 187, 135713–135768. 61. Britton, R. A., Eichenberger, P., Gonzalez-Pastor, J. E., Fawcett, P., Monson, R., Losick, R. & Grossman, A. D. (2002). Genome-wide analysis of the stationary-phase sigma factor (sigma-H) regulon of Bacillus subtilis. J. Bacteriol. 184, 4881–4890. 62. Hamon, M. A., Stanley, N. R., Britton, R. A., Grossman, A. D. & Lazazzera, B. A. (2004). Identification of AbrB-regulated genes involved in biofilm formation by Bacillus subtilis. Mol. Microbiol. 52, 847–860. 63. Kearns, D. B., Chu, F., Branda, S. S., Kolter, R. & Losick, R. (2005). A master regulator for biofilm formation by Bacillus subtilis. Mol. Microbiol. 55, 739–749. 64. Fujita, M. & Losick, R. (2003). The master regulator for entry into sporulation in Bacillus subtilis becomes a cell-specific transcription factor after asymmetric division. Genes Dev. 17, 1166–1174. 65. Murrell, W. G. & Warth, A. D. (1965). Composition and heat resistance of bacterial spores. In Spores III (Campbell, L. L. & Halvorson, H. O., eds), pp. 1–24, American Society for Microbiology, Washington, DC.

Forespore Line of Gene Expression in B. subtilis

66. Nicholson, W. L., Sun, D. X., Setlow, B. & Setlow, P. (1989). Promoter specificity of sigma G-containing RNA polymerase from sporulating cells of Bacillus subtilis: identification of a group of forespore-specific promoters. J. Bacteriol. 171, 2708–2718. 67. Abanes-De Mello, A., Sun, Y. L., Aung, S. & Pogliano, K. (2002). A cytoskeleton-like role for the bacterial cell wall during engulfment of the Bacillus subtilis forespore. Genes Dev. 16, 3253–3264. 68. Kim, H., Hahn, M., Grabowski, P., McPherson, D. C., Otte, M. M., Wang, R. et al. (2006). The Bacillus subtilis spore coat protein interaction network. Mol. Microbiol. 59, 487–502. 69. Simpson, E. B., Hancock, T. W. & Buchanan, C. E. (1994). Transcriptional control of dacB, which encodes a major sporulation-specific penicillin-binding protein. J. Bacteriol. 176, 7767–7769. 70. Gilmore, M. E., Bandyopadhyay, D., Dean, A. M., Linnstaedt, S. D. & Popham, D. L. (2004). Production of muramic delta-lactam in Bacillus subtilis spore peptidoglycan. J. Bacteriol. 186, 80–89. 71. Imamura, D., Kobayashi, K., Sekiguchi, J., Ogasawara, N., Takeuchi, M. & Sato, T. (2004). spoIVH (ykvV), a requisite cortex formation gene, is expressed in both sporulating compartments of Bacillus subtilis. J. Bacteriol. 186, 5450–5459. 72. Dworkin, J. & Losick, R. (2005). Developmental commitment in a bacterium. Cell, 121, 401–409. 73. Karmazyn-Campelli, C., Bonamy, C., Savelli, B. & Stragier, P. (1989). Tandem genes encoding sigmafactors for consecutive steps of development in Bacillus subtilis. Genes Dev. 3, 150–157. 74. Wach, A. (1996). PCR-synthesis of marker cassettes with long flanking homology regions for gene disruptions in S. cerevisiae. Yeast, 12, 259–265. 75. Vellanoweth, R. L. & Rabinowitz, J. C. (1992). The influence of ribosome-binding-site elements on translational efficiency in Bacillus subtilis and Escherichia coli in vivo. Mol. Microbiol. 6, 1105–1114. 76. van Ooij, C., Eichenberger, P. & Losick, R. (2004). Dynamic patterns of subcellular protein localization during spore coat morphogenesis in Bacillus subtilis. J. Bacteriol. 186, 4441–4448. 77. Sterlini, J. M. & Mandelstam, J. (1969). Commitment to sporulation in Bacillus subtilis and its relationship to development of actinomycin resistance. Biochem. J. 113, 29–37. 78. Harwood, C. R. & Cutting, S. M. (1990). Molecular Biological Methods for Bacillus, Wiley, ChiChester,United Kingdom. 79. Tseng, G. C., Oh, M. K., Rohlin, L., Liao, J. C. & Wong, W. H. (2001). Issues in cDNA microarray analysis: quality filtering, channel normalization, models of variations and assessment of gene effects. Nucl. Acids Res. 29, 2549–2557. 80. Liu, X., Brutlag, D. L. & Liu, J. S. (2001). BioProspector: discovering conserved DNA motifs in upstream regulatory regions of co-expressed genes. Pac. Symp. Biocomput. 6, 127–138. 81. Jensen, S. T. & Liu, J. S. (2004). BioOptimizer: a Bayesian scoring function approach to motif discovery. Bioinformatics, 20, 1557–1564. 82. Frohman, M. A. (1994). On beyond classic RACE (rapid amplification of cDNA ends). PCR Methods Appl. 4, S40–S58. 83. Price, C. W., Fawcett, P., Ceremonie, H., Su, N., Murphy, C. K. & Youngman, P. (2001). Genome-wide analysis of the general stress response in Bacillus subtilis. Mol. Microbiol. 41, 757–774.

Forespore Line of Gene Expression in B. subtilis

84. Driks, A., Roels, S., Beall, B., Moran, C. P., Jr & Losick, R. (1994). Subcellular localization of proteins involved in the assembly of the spore coat of Bacillus subtilis. Genes Dev. 8, 234–244. 85. Mason, J. M. & Setlow, P. (1986). Essential role of small, acid-soluble spore proteins in resistance of Bacillus subtilis spores to UV light. J. Bacteriol. 167, 174–178.

37 86. Schneider, T. D. & Stephens, R. M. (1990). Sequence logos: a new way to display consensus sequences. Nucl. Acids Res. 18, 6097–6100. 87. Paidhungat, M., Setlow, B., Driks, A. & Setlow, P. (2000). Characterization of spores of Bacillus subtilis which lack dipicolinic acid. J. Bacteriol. 182, 5505–5512.

Edited by M. Gottesman (Received 16 December 2005; received in revised form 13 January 2006; accepted 17 January 2006) Available online 8 February 2006