SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

A 14.6 kb prophage-like insertion, termed skinCd, was found to interrupt the sigK gene, which encodes an RNA polymerase sigma factor essential for sporulation, in six strains of Clostridium difficile. Until now, Bacillus subtilis was the only spore-former shown to carry such an insertion, and the presence of the insertion is not required for efficient sporulation in this organism. The B. subtilis and C. difficile skin elements proved to be divergent in sequence, inserted at different sites within the sigK gene and in opposite orientations. The skinCd element was excised from the chromosome specifically during sporulation, forming a circular molecule. Two natural isolates of C. difficile lacked the skinCd element and were defective in sporulation. When a merodiploid strain was created that carries both interrupted and uninterrupted versions of the sigK gene, the cells became Spo, showing that the uninterrupted gene is dominant and inhibits sporulation. C. difficile sigK genes, whether skinCd+ or skinCd–, lack the N-terminal pro-sequence found in all other sigK genes studied to date. Thus, regulated excision of skinCd appears to be a critical mechanism for achieving proper temporal activation of σK.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

The two major regulatory systems that govern spore formation in Bacillus subtilis, the transcription factor Spo0A (Hoch, 1993) and a series of alternative sigma factors of RNA polymerase (Errington, 1993; Stragier and Losick, 1996), appear to be conserved in all sequenced species of Bacillus and Clostridium (Brown et al., 1994; Sauer et al., 1994; Arcuri et al., 2000).

The sporulation-specific sigma factors of B. subtilis are controlled at several different levels. Each is expressed from a gene that is under developmental regulation, and each is synthesized under conditions that prevent immediate activity. For the late mother cell factor, σK, proteolytic activation of an inactive precursor is required (Lu et al., 1990). A 20-residue pro-sequence is thought to direct the pro-sigma factor to the mother cell side of the prespore septum (Zhang et al., 1998) where, upon appropriate signalling from the forespore compartment (Losick and Stragier, 1992), a protease in the membrane cleaves off the pro-sequence, releasing the active sigma factor into the mother cell cytoplasm (Cutting et al., 1990; Resnekov and Losick, 1998; Rudner et al., 1999; Yu and Kroos, 2000; Rudner and Losick, 2002).

In B. subtilis, an additional event is required for the expression of σK. The B. subtilis sigK gene is disrupted by a 48 kb, prophage-like element termed skin (sigK intervening sequence) (Stragier et al., 1989; Takemaru et al., 1995). A recombination event that occurs only at stage III of sporulation and only within the mother cell excises skin from the chromosome creating an intact sigK gene (Stragier et al., 1989; Kunkel et al., 1990). Although the presence of the skin element was expected to contribute to the temporal and spatial regulation of σK activity, deletion of skin does not result in any detectable sporulation defect (Kunkel et al., 1990). Apparently, the dependence of sigK transcription on the mother cell sigma factor σE and the regulated cleavage of pro-σK are sufficient. Moreover, no other Bacillus or Clostridium species examined to date contains an interrupted sigK gene (Adams et al., 1991; Sauer et al., 1994; Takami et al., 2000).

Clostridium difficile is the major causative agent of antibiotic-associated diarrhoea and the more severe condition, pseudomembranous colitis (Kelly and LaMont, 1998). Infection by C. difficile is thought to be initiated by the ingestion of spores present in the environment (Kelly and LaMont, 1998). Only individuals with a compromised colonic flora, primarily those who have received antibiotic therapy, are susceptible to colonization with the organism and proceed to symptomatic infection (Larson et al., 1980; Kelly and LaMont, 1998).

To begin to understand spore formation in C. difficile, we sought to identify the sporulation-specific sigma factors in this organism. We report here the unexpected presence of a phage-like element disrupting the sigK gene in several strains of C. difficile. The C. difficile element, called here skinCd, is excised, like B. subtilis skin, from the mother cell chromosome only during sporulation. Unlike the situation in B. subtilis, however, the presence of this element is required for efficient sporulation.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

The sigK gene of C. difficile strain CD630 is disrupted by an element similar to B. subtilis skin

Homologues of the sporulation-specific sigma factors of B. subtilis and Clostridium acetobutylicum were identified in the unfinished C. difficile genome database by BLAST analysis (Altschul et al., 1997), using the partial sequence of the C. difficile sigG gene (see Experimental procedures) as a probe. The identified genes were distinguishable as sporulation-specific sigma factors based on the sequences of regions 2.4 and 4.2, known to be involved in direct promoter interactions (Lonetto et al., 1992; Wosten, 1998). For σF, σE and σG, the identity of adjacent genes (Haldenwang, 1995; Sauer et al., 1995) was consistent with functional assignments. Surprisingly, the apparent N-terminal and C-terminal encoding portions of C. difficile sigK were contained on different contigs and proved to be separated by 14.660 kb (Fig. 1).

image

Figure 1. The disrupted sigK genes of C. difficile and B. subtilis. The B. subtilis spoIVCB (sigK 5′), spoIIIC (sigK 3′) and skin sequences were obtained from the SubtiList database (http:genolist.pasteur.frSubtiList). Shaded rectangles/arrows represent the 5′ and 3′ components of the disrupted sigK genes. Open arrows designate the site-specific recombinases present within the sigK intervening sequence, skin, from each organism. DNA sequences flanking sigK are represented by a thin line; skin sequences are represented by a thick line. The length of each intervening sequence is noted below the diagram.

Download figure to PowerPoint

The DNA immediately upstream of the C-terminal region of sigK contains an open reading frame (ORF) that would encode a potential site-specific recombinase of the large resolvase subfamily (Bannam et al., 1995; Crellin and Rood, 1997; Thorpe and Smith, 1998; Wang and Mullany, 2000). The predicted protein contains a resolvase/invertase domain at its amino-terminus but has an extended C-terminal domain making it much larger (505 amino acids) than most resolvases. The closest relative of this putative gene product is YokA, the site-specific recombinase of phage SPβ (30% identity, 50% similarity). It is also related to other members of the large resolvase family, TndX (22% identity, 42% similarity) and TnpX (21% identity, 39% similarity), encoded within the Clostridium conjugative transposons Tn5397 and Tn4451 respectively (Bannam et al., 1995; Wang and Mullany, 2000). At the amino acid level, the putative C. difficile recombinase shares 25% identity with the resolvase domain of SpoIVCA, the site-specific recombinase that is encoded within the B. subtilis skin element and catalyses its excision (Kunkel et al., 1990; Sato et al., 1990; Popham and Stragier, 1992). In B. subtilis, spoIVCA lies adjacent to the N-terminal region of sigK (spoIVCB) and is in convergent orientation. Thus, C. difficile has a skin-like element that is in inverted orientation with respect to the sigK gene relative to that of the B. subtilis skin element (Fig. 1).

Disruption of the sigK gene was confirmed, using a polymerase chain reaction (PCR) assay, for strains CD630, VPI10463, CD196, CD2001, CD65982 and CD79685 (Fig. 2). Using chromosomal DNA prepared from vegetatively grown cells as a template, primer pairs OJH39 + OJH40 and OJH41 + OJH42 amplified the N-terminal and C-terminal regions of sigK as expected (Fig. 3A). Primer pairs OJH39 + OJH43 and OJH44 + OJH42 directed amplification of products with sizes indicating that skin-specific sequences are indeed located adjacent to the N-terminal and C-terminal regions of sigK (Fig. 3A). Primers that would amplify the entire sigK ORF (OJH39 + OJH42) gave no product (Fig. 3A), consistent with the idea that there is a large intervening sequence separating the N- and C-terminal ORFs. We conclude that the sigK gene of six different C. difficile strains is disrupted by a DNA element, hereafter referred to as skinCd.

image

Figure 2. PCR strategy to determine sigK organization. At the top, the shaded rectangle and arrow represent the disrupted sigK gene. skinCd is represented by a thick line bounded by arrowheads between the two sigK sequences. The expected products of skinCd excision are shown below. The rearranged sigK gene is represented by a shaded arrow, and skinCd is represented as a ligated circle. Small numbered arrows indicate the sites of oligonucleotide annealing and the direction of priming. Primer sequences are given in Table 3. PCR with primers OJH39 + OJH40 would yield a product, N (519 bp), regardless of the organization (disrupted or undisrupted) of sigK on the chromosome; primers OJH41 + OJH42 would yield a product, C (251 bp), regardless of the organization (disrupted or undisrupted) of sigK on the chromosome; primers OJH39 + OJH43 or OJH44 + OJH42 would yield products, Nφ (796 bp) or Cφ (487 bp), respectively, containing both sigK sequence and skinCd sequence, if present; primer pair OJH39 + OJH42 would only yield a product, sigK (847 bp), if the undisrupted form of sigK is present; primer pair OJH44 + OJH43 would only yield a product, CF (424 bp), if skinCd is excised to form a circular molecule.

Download figure to PowerPoint

image

Figure 3. PCR analysis of C . difficile sigK from cells grown under different conditions. Chromosomal DNA was purified from (A) vegetative or (B) sporulating populations of C. difficile strain VPI10463. This DNA was used as a template in the PCR assay depicted in Fig. 2. Reaction products were separated on 1% agarose gels. Arrows to the right of the gels indicate the predicted sizes of the sigK (847 bp) and CF (424 bp) products. C. Chromosomal DNA was purified from a vegetatively grown culture of C. difficile strain CD37 (skinCd–) and subjected to the PCR assay as above. Lanes are labelled as described for (A) and (B).

Download figure to PowerPoint

skin Cd excises from the chromosome during sporulation to reconstitute intact sigK

When we prepared chromosomal DNA from a sporulating population of cells, we found that, in addition to the PCR products observed from a non-sporulating population of cells, a PCR product expected for an uninterrupted version of the sigK gene could be amplified with primers OJH39 and OJH42 (Fig. 3B). (Products that reflect the contiguity of skinCd with sigK were also amplified from sporulating cell DNA because not all cells in the population reach the stage of sporulation at which skinCd is excised from the chromosome.) Thus, uninterrupted sigK appears to be generated by excision of skinCd during sporulation of C. difficile.

The sequence of the PCR product generated by OJH39 and OJH42 confirmed that intact sigK is created by excision of skinCd and that a 12 bp repeat is the site of recombination (Fig. 4A). A 22 bp imperfect inverted repeat (Fig. 4A), which is found adjacent to the 12 bp repeat at both N-terminal and C-terminal sites, might provide binding sites for the skinCd recombinase (Popham and Stragier, 1992; Hallet and Sherratt, 1997). This predicted recombination site contains within it the 5 bp sequence, AATGA, that is identical to the site at which recombination occurs in the B. subtilis sigK gene. The C. difficile recombination site, however, lies 80 bp downstream of the recombination site in the B. subtilis gene. The location corresponding to the B. subtilis recombination site in the C. difficile gene also contains the sequence AATGA, but the pentanucleotide in this case is not contained within a larger 12 bp repeat (data not shown).

image

Figure 4. A. Alignment of sigK 5′ and 3′ recombination sites before chromosomal rearrangement. The 12 bp repeated sequence within which recombination takes place is highlighted in grey. The central GA dinucleotide in the recombination site is indicated with bold type. A 22 bp imperfect inverted repeat sequence is indicated by the arrows (the 20 perfectly matching bases are in bold). The predicted translation products of the 5′ and 3′ sequences are given above and below the sequences respectively. B. Sequence of the sigK recombination junction after the recombination event. The remaining copy of the 12 bp recombination site is highlighted in grey. The central GA dinucleotide is indicated with bold type. The predicted translation product of sigK after recombination is listed below the sequence.

Download figure to PowerPoint

Excision of skinCd results in the formation of a circular molecule

The B. subtilis skin element circularizes when it excises from the chromosome (Stragier et al., 1989; Kunkel et al., 1990). Other recombinases of the large resolvase subfamily have also been shown to catalyse the formation of circular DNA molecules (Bannam et al., 1995; Thorpe and Smith, 1998; Wang and Mullany, 2000). To test whether skinCd circularizes, we assayed the ability of the primer pair OJH43 + OJH44 to direct amplification of a product from DNA prepared from sporulating cells. These primers would only yield a PCR product if a circular form of skinCd were present, as it is only in this configuration that the primers would face each other. Indeed, a product representing the circular junction was observed (Fig. 3B). When the same PCR was performed with template DNA isolated from cells during vegetative growth, no product was obtained (Fig. 3A).

Absence of skinCd correlates with a sporulation defect

Although six strains of C. difficile carry skinCd within the sigK gene, two other strains, ATCC9689 and CD37, produced a different amplification profile in our assays. PCRs using primers specific for skinCd gave no product (Fig. 3C), and chromosomal DNA from vegetatively grown cells of these strains contained the uninterrupted version of sigK (Fig. 3C).

As σK is an essential factor for late sporulation gene expression, we questioned whether skinCd+ and skinCd– strains would have different sporulation efficiencies. As C. difficile colonies undergo sporulation on brain–heart infusion (BHI) plates, they change from pale yellow and translucent to creamy and opaque (J. D. Haraldsen, unpublished observation). The skinCd– strains, ATCC9689 and CD37, did not exhibit this characteristic change in morphology even after 48 h of incubation on BHI agar, whereas this change was apparent in skinCd+ strains after only 24 h of incubation. Using a heat resistance assay (see Experimental procedures) to determine the extent of sporulation, we found that < 1% of the population of skinCd– strains had formed spores after 24 h compared with 30–50% of skinCd+ cells (Table 1). The sequence of the sigK gene from the skinCd– strains was identical to that of the uninterrupted sigK gene created during sporulation of skinCd+ cells. The absence of skinCd therefore correlates with a sporulation defect.

Table 1. . Sporulation of various C. difficile strains.
StrainGenotypeHeatR spores ml−1Total viablecells ml−1aSporulationefficiencybRelative sporulationfrequencyc
  • a . The total number of viable cells was calculated by taking the sum of the number of heat-resistant colonies ml −1 (spore population) and the number of colonies ml−1 without heating (vegetative population) (see Experimental procedures).

  • b . [(No. of heat R spores ml−1)/(no. of total viable cells ml−1)] ×100.

  • c

    . The sporulation efficiency of a given strain was compared with that of strain CD196, normalized to 100%.

CD196 sigK::skinCd8.55 × 1062.70 × 10731.67%(100%)
ATCC9689 sigK 5.90 × 1041.24 × 107 0.48%1.52%
JH-C68 sigK::skinCd,Tn916::sigK1.55 × 1061.50 × 108 1.03%3.25%
JH-C72 sigK::skinCd,Tn916::sigK4.00 × 1059.54 × 107 0.42%1.33%

Clostridium difficile σK does not contain an N-terminal pro-sequence

Based on the B. subtilis model, the sporulation defect in C. difficile strains lacking skinCd was surprising. It would only be understandable if other levels of sigK regulation were defective. For instance, a skin-less mutant of B. subtilis that also lacks the 20-amino-acid pro-sequence sporulates poorly (Oke and Losick, 1993). Alignment of the deduced C. difficileσK coding sequence with σK sequences from nine other Bacillus and Clostridium species showed that the pro-sequence is uniquely absent in C. difficile CD630 (Fig. 5). The homology begins at residue 22 of the B. subtilis sequence, near the site of proteolytic cleavage.

image

Figure 5. Amino acid sequence alignment of the N-terminal regions of σK from various spore-forming bacteria. The B. stearothermophilus, Bacillus cereus and B. haloduransσK sequences were identified in the NCBI Microbial Genomes BLAST database. The C. perfringens sequence was provided by V. L. Stirewalt and S. B. Melville. Sequences were aligned by the CLUSTALW program at EBI (http:www2.ebi.ac.ukclustalW). Black boxes highlight identical residues at a given position in 50% or more of the sequences. Grey boxes highlight similar residues at a given position in> 50% of the sequences. Asterisks indicate nonsense codons. Similar residues were defined as I, L, V and M; A and G; F, Y and W; K and R; S and T; N and Q; D and E. The site of pro-sequence cleavage by B. subtilis SpoIVFB is indicated by an arrow. Species names are abbreviated as follows, Bsu, B. subtilis; Bth, B. thuringiensis; Bst, B. stearothermophilus; Bce, B. cereus; Ban, B. anthracis; Bha, B. halodurans; Cac, C. acetobutylicum; Cpe, C. perfringens; Cbo, C. botulinum; Cdi, C. difficile.

Download figure to PowerPoint

To verify this sequence analysis, we amplified the region upstream of the coding sequence from all eight C. difficile strains by PCR using primers OJH39 + OJH40 and sequenced the products. The resulting sequences were identical in all cases and confirmed the absence of the pro-sequence in both skinCd+ and skinCd– strains. The sigK start codon was defined based on the presence of a plausible ribosome binding site 7 bp upstream and the presence of a stop codon within the same reading frame located six codons upstream of the assigned start. Three potential σE-dependent promoter sequences were identified within the 150 bp upstream of the start codon.

Not surprisingly, no homologues of SpoIVFB, the protease that processes pro-σK in B. subtilis, or the other members of the pro-σK processing complex (BofA, SpoIVFA) can be found in the C. difficile CD630 genome sequence (Stragier, 2002; data not shown).

skinCd+/ skin Cd– merodiploid strains are Spo

The combined absence of the sigK pro-sequence and skinCd in Spo strains such as ATCC9689 and CD37 suggested that regulated excision of skinCd is required for the proper temporal expression of sigK. The sporulation defect of skinCd– strains would then be attributable to premature activity of σK.

To prove that regulated excision of the element is required for sporulation in C. difficile, we constructed merodiploid strains containing both interrupted (containing skinCd) and uninterrupted copies of the sigK gene. As there is no easy way of introducing DNA into or directing mutations in C. difficile, we took advantage of the observation that Tn916 can mediate its own transfer by conjugation from B. subtilis to C. difficile (Mullany et al., 1994). A recombinant version of Tn916 carrying the uninterrupted copy of sigK, associated with its native promoter (amplified from CD79685 after recombination) (Fig. 6), was used to drive transfer of sigK to a skinCd+C. difficile strain, CD196 (see Experimental procedures).

image

Figure 6. Integration of Tn916::sigK into the C. difficile chromosome. A. Schematic representations of Tn916::sigK (top) after excision from the B. subtilis chromosome and the sigK locus (bottom) of skinCd+C. difficile. The disrupted sigK gene is represented by the black arrow and rectangle. Genes encoded by Tn916 and pSMB47 are represented by grey arrows. The region in which homologous recombination occurred is indicated by an X. B. Schematic representation of the sigK locus in merodiploid strains JH-C68 and JH-C72. C. PCR products diagnostic of the mechanism of Tn916::sigK integration in JH-C72 separated on a 1.0% agarose gel. Lane 1, molecular weight markers; 2, sigK (OJH39 + OJH42, 847 bp); 3, skinCdsigK C-terminus (OJH44 + OJH42, 487 bp); 4, tetM (OJH59 + OJH60, 611 bp); 5, ermB (OJH64 + OJH65, 1428 bp); 6, ermBsigK C-terminus (OJH65 + OJH106, 2220 bp); 7, 5′ flanking region–sigK C-terminus (OJH73 + OJH42, 1612 bp); 8, sigK N-terminus−3′ flanking region (OJH76 + OJH104, 1382 bp).

Download figure to PowerPoint

Several transconjugants were selected for further analysis. We proved that these strains were merodiploid for sigK by using the primer pairs OJH39 + OJH42, which can only amplify the uninterrupted copy of the gene, and OJH39 + OJH43 (or OJH44 + OJH42), which can amplify the sigK/skinCd junction (Fig. 6). PCR assays also showed the presence of Tn916 and pSMB47 markers in the transconjugants. Integration of Tn916::sigK into the recipient chromosome could occur by either transposition or homologous recombination. We distinguished between these mechanisms by amplification of the regions of Tn916 flanking sigK and by amplification of the chromosomal DNA that normally flanks the sigK locus. In fact, all the transconjugants analysed yielded a product from the primer pair OJH76 + OJH104, establishing that the N-terminus of sigK is contiguous with the 3′ flanking chromosomal DNA (Fig. 6). This result indicates that Tn916::sigK entered the chromosome by homologous recombination within the C-terminus of sigK (Fig. 6).

All the transconjugants examined had a Spo colony morphology on BHI agar containing erythromycin after 24 h of incubation and produced very few or no phase-bright spores as assessed by phase-contrast microscopy. After 48 h of incubation, many of the transconjugants appeared to produce phase-grey prespores that were not released from the mother cell. This sporulation defect was quantified by heat resistance assays. The sigK merodiploid strains had a sporulation efficiency 100-fold lower than that of the isogenic parent strain (Table 1). In fact, the extent of sporulation observed for the merodiploid strains correlates well with the sporulation efficiency of the naturally skinCd– strain ATCC9689. We conclude that expression from the uninterrupted copy of sigK is dominant over expression from the native, skinCd+ copy of sigK and results in a Spo phenotype.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

Chromosomal rearrangements are associated with several important developmental processes in eukaryotic organisms. Examples include the rearrangements of T-cell receptor and immunoglobulin genes in the vertebrate immune system (Grawunder et al., 1998; Bassing et al., 2002), mating type switching in yeast (Haber, 1998) and the dramatic genome rearrangements that occur during macronucleus development in many protozoan parasites (Jahn and Klobutcher, 2002).

Discovery of the skin element in B. subtilis raised the possibility that regulated chromosomal rearrangement may be a common way of controlling development in prokaryotic organisms with terminally differentiated cell types (Stragier et al., 1989), as the chromosome of the terminally differentiated cell need not be maintained. Several examples of this type of regulated rearrangement had, in fact, already been identified in Anabaena (Golden et al., 1985; 1988), a cyanobacterium that develops terminally differentiated heterocysts for nitrogen fixation. However, later work showed that B. subtilis skin plays no essential role in the regulation of sporulation or sigK expression (Kunkel et al., 1990). Temporally regulated excision of skin in B. subtilis proved to be only one of three mechanisms regulating the synthesis and activity of σK. Moreover, as all other Bacillus and Clostridium species examined to date contain uninterrupted sigK genes, interest in skin declined. Our work has now identified at least one organism, C. difficile, in which interruption of the sigK gene is essential for efficient sporulation.

The absence of a second regulatory element, the N-terminal pro-sequence, in skinCd– (and skinCd+) strains appears to account for the sporulation defect observed. Perhaps transcriptional regulation alone is not sufficient to restrict σK activity to the appropriate time and location, making regulated rearrangement of sigK essential for efficient sporulation in C. difficile. There is, in fact, evidence suggesting that σK initiates a feedback loop that inhibits expression of sigE, the gene for the early mother cell sigma factor, as well as other genes in the σE regulon (Zhang and Kroos, 1997; Zhang et al., 1999). In normal cells, this feedback may facilitate the switch from σE-dependent to σK-dependent transcription. But if σK activity occurs prematurely, as in pro-less B. subtilis strains or (as we predict) in skinCd–C. difficile strains, this negative feedback may contribute to the sporulation defect observed.

The dominance of undisrupted sigK over sigK::skinCd argues that premature expression and subsequent activity of σK are detrimental to the developing forespore, perhaps causing improper assembly of the spore coat and cortex. As it is the assembly of these structures that confers refractility to spores (Losick et al., 1986; Catalano et al., 2001), such a defect would explain our inability to detect mature refractile spores in these strains by phase-contrast microscopy. The Spo phenotype that we observed is consistent with, but more pronounced than, the phenotype described for a double mutant of B. subtilis lacking skin and deleted for the pro-sequence of sigK (Oke and Losick, 1993). A skin-less, pro-less B. subtilis strain sporulates approximately four- to 50-fold less efficiently than a wild-type strain, whereas the C. difficile merodiploid strains sporulate up to 100-fold less efficiently. The B. subtilis spores that are produced have germination defects (Oke and Losick, 1993), consistent with abnormalities in the spore coat or cortex.

The apparent lack of pro-σK processing in C. difficile has another important implication. In B. subtilis, the cleavage of pro-σK is dependent on events occurring in the forespore (Cutting et al., 1990; 1991; Lu et al., 1990). A complex of proteins, composed of SpoIVFA and BofA, keeps the SpoIVFB protease inactive (Resnekov and Losick, 1998; Rudner et al., 1999; Rudner and Losick, 2002) until a signal is received from the forespore protein SpoIVB (Cutting et al., 1991), thereby co-ordinating mother cell gene expression with that of the forespore. Although there is a homologue of SpoIVB in C. difficile, the absence of all other proteins involved in pro-σK processing implies that there may be no coupling of forespore and mother cell gene expression late during sporulation in this organism, or that late sporulation gene expression may be co-ordinated through a completely different mechanism in C. difficile.

The interruptions of sigK by prophage-like elements in B. subtilis and C. difficile clearly represent independent events during evolution. First, the skin elements in these two organisms are not identical. Yet they are related elements, as they integrate into the identical 5 bp sequence (contained within a 12 bp repeat in C. difficile). Secondly, although the same gene is disrupted in both organisms and, in both cases, the critical DNA recognition elements of the sigma factor are separated, the exact location of skin integration within the two ORFs is different. Thirdly, skinCd is in the opposite orientation with respect to sigK compared with skin of B. subtilis. Thus, the two skin elements both integrated after the separation of the Clostridium and Bacillus genera and after speciation of each genus.

If we assume that skin elements were once temperate phage genomes, it may have been advantageous for them to have integrated into the coding sequence of a sporulation gene. The skin elements would remain integrated under conditions that support rapid cell growth but might become expressed under conditions that would induce sporulation. If so, it would be important for the skin element to excise from the mother cell genome only, as the mother cell is destined to lyse. This idea is supported by the fact that, although there are five potential sites for skinCd integration in the C. difficile chromosome, the element integrated stably only at the site that is within a mother cell-specific sporulation gene.

Whether the loss of the N-terminal pro-sequence in C. difficile preceded or followed integration of skinCd is not easily determinable. The small size of skinCd compared with skin suggests that skinCd has lost large portions of its original sequence and may therefore be more ancient than B. subtilis skin. skinCd also appears to have undergone extensive genetic deterioration. Only four predicted ORFs, including the site-specific recombinase, appear to have been maintained, as assessed by sequence comparisons with phage genomes. The remainder of skinCd sequences with homology to other genes appear to contain loss-of-function mutations. The acquisition of skin would allow a pro-sigK gene to lose its pro-sequence and its processing machinery without loss of regulation of σK activity. B. subtilis sigK, which currently maintains both the pro-sequence and skin, may be undergoing a similar evolutionary process and may in future aeons become more like C. difficile sigK.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

Bacterial strains

The C. difficile strains used in this study are listed in Table 2. The reference strain VPI10463 was obtained from the culture collection of the Department of Anaerobic Microbiology, Virginia Tech, Blacksburg, VA, USA. Strains CD37 and CD630 (the strain currently being sequenced by the Sanger Centre) were obtained from P. Mullany, University College London, UK. All other C. difficile strains used in this study were from the culture collection of M. R. Popoff (Institut Pasteur, Paris, France).

Table 2. . Strains and plasmids.
StrainsRelevant characteristics skin Cd Reference or source
  • a

    . PMC, pseudomembranous colitis.

C. difficile
 CD630Multiply antibiotic-resistant clinical isolate, Sanger Centre C. difficile genome project strain+ Hachler et al. (1987)
 VPI10463Human abdominal wound isolate+ Sullivan et al. (1982)
 CD196Human PMCa isolate from patient treated with amoxicillin+ Popoff et al. (1988)
 CD2001 +M. R. Popoff
 CD37  Hachler et al. (1987)
 CD65982 +M. R. Popoff
 CD79685Human PMCa isolate+ Eveillard et al. (1993)
 ATCC9689Type strainM. R. Popoff
 JH-C68CD196 sigK::skinCd, Tn916::sigK+, –This study
 JH-C72CD196 sigK::skinCd, Tn916::sigK+, –This study
B. subtilis
 BS49CU2189::Tn916 P. Mullany
 JH-B31CU2189::Tn916::pJH16 This study
Plasmids
 pSMB47Tn916 integrational vector, CmR, ErmR  Manganelli et al. (1998)
 pJH16 C. difficile sigK cloned in pSMB47 digested with HindIII–BamHI This study

Cell growth and sporulation

All C. difficile strains were grown in an atmosphere of 10% H2, 5% CO2 and 85% N2 in an anaerobic chamber (Coy Laboratory Products) at 37°C. BHI agar or broth (Difco) was used for vegetative growth of C. difficile. Sporulation was induced by overnight growth on BHI agar and was quantified by a modification of a published procedure (Kamiya et al., 1989). Briefly, a loopfull of C. difficile cells was resuspended in 1 ml of BHI and divided into two aliquots. One aliquot was untreated, serially diluted and plated on BHIS (BHI agar supplemented with 5 g l−1 yeast extract and 0.1%L-cysteine HCl) to assess the number of vegetative cells present in the population. We and others (Kamiya et al., 1989) have found that C. difficile spores germinate very poorly under these conditions. The second aliquot was removed from the anaerobic chamber and heated to 80°C for 10 min. After cooling briefly on ice, the sample was treated with sodium thioglycolate (165 µM) for 30 min at 50°C, pelleted by centrifugation, resuspended in 0.5 ml of lysozyme (4 mg ml−1) and incubated for 15 min at 37°C to stimulate germination. The sample was returned to the anaerobic chamber, and serial dilutions were plated on BHIS agar to determine the number of heat-resistant spores.

Isolation of an internal fragment of the C. difficile sigG gene

An internal portion of the C. difficile spoIIIG (sigG) gene was identified by PCR amplification using primers OJH3 + OJH4 (Table 3), designed to anneal to well-conserved regions of the σ70 family of sigma factors of RNA polymerase (Lonetto et al., 1992; Haldenwang, 1995). Chromosomal DNA isolated from VPI10463 was used as template. Amplification conditions are described below. The PCR product was sequenced, and the predicted translation product of the incomplete spoIIIG sequence was used to search the unfinished C. difficile genome database (Sanger Centre; http:www.sanger.ac.ukProjectsCdifficile) for the remaining sequence of the gene as well as for additional sigma factor genes. Candidate genes were analysed by the BLAST (Altschul et al., 1997) program.

Table 3. . PCR primers used in this study.
Primer5′ sequence 3′
  1. Bold type indicates homology to C. difficile genome sequence or to Tn916.

OJH3CGCGAATTCGGATCCGTTGATACATCTAAACTTCCAG
OJH4CGCGAATTCCATATATTGCATCTCCACCGTC
OJH39CCGAATTCAAGCTTCCCACTCTTTATGTCATATTCC
OJH40 CTTTATCTGTTCCGATAGGG
OJH41 CAAGTTGGAAAACTTTACG
OJH42CCCGAGCTCCTAATGATGTACGATATATCC
OJH43 GATACCAAGTATCTGTATGC
OJH44 CCATGATTCAGATTCCCTTGG
OJH59 CCGATTCTGAACAATGGGATACGG
OJH60 GCGGATCACTATCTGAGATTTCC
OJH64CCCCGGAATTCGCCGCATGCGCGAGATCTGAAAGTCCTACTTTGCCG
OJH65CCGCCGCTGCAGGTAGCACCTGAAGTCAGCCCC
OJH73CCCGCGGATCCGCCTGTCCTTTCTAGTTACATAGC
OJH76 GGGAATACTAATGAAATACTGCG
OJH104 CCCCTGAAGAAGAGATTGAG
OJH106 GTGTCTTATAGCCATAAGGAGTTAACCC

Isolation of chromosomal DNA from C. difficile

Vegetative cells were harvested by centrifugation from 50 ml broth cultures of C. difficile grown to an OD600 of 0.4–0.8. Sporulating populations of C. difficile cells were scraped from BHI agar plates and collected by centrifugation. Chromosomal DNA was isolated from each population of cells by a published method (Wren and Tabaqchali, 1987). Using this method, chromosomal DNA from the mother cell compartments of prespores is isolated preferentially.

Amplification and sequencing of C. difficile sigK

Primers used to amplify various regions of the sigK gene and flanking sequences from C. difficile are listed in Table 3. Amplification reactions consisted of 40 cycles of denaturation (95°C, 1 min), annealing (50°C, 1 min) and extension (72°C, 1 min). PCR products were purified using the QIAquick PCR purification kit (Qiagen) and submitted to the Tufts University core facility for sequencing. PCR products from the 5′ region of sigK were sequenced using OJH40 as primer, whereas both OJH39 and OJH42 were used as primers for sequencing the PCR product containing the uninterrupted sigK gene.

Engineering of Tn916 and interspecies conjugation

The uninterrupted form of sigK with its native promoter was amplified by PCR, using chromosomal DNA prepared from sporulating cells of strain CD79685, and cloned in pSMB47, a suicide vector that integrates into the conjugal transposon Tn916 by homologous recombination (Manganelli et al., 1998). The resultant plasmid, pJH16, was used to transform a B. subtilis strain, BS49, containing Tn916. Integrants were selected on plates containing erythromycin (0.5 µg ml−1). One strain, JH-B31 (BS49::Tn916::pJH16), was selected for conjugation with C. difficile using the protocol of D. Lyras and J. I. Rood (Mani et al., 2002) with modifications. The skinCd+C. difficile recipient strain CD196 was grown in BHIS broth for 6 h. Simultaneously, the B. subtilis donor strain, JH-B31, was grown to mid-exponential phase in BHI broth containing a subinhibitory concentration of tetracycline (2.5 µg ml−1) to induce the excision of Tn916::pJH16 from the chromosome and expression of tra genes (Celli and Trieu-Cuot, 1998). Samples (100 µl) of the donor and recipient strains were mixed and spread on BHIS agar plates supplemented with KNO3 (5 mM) to support anaerobic growth of the B. subtilis donor strain (Nakano and Zuber, 1998). The mating plates were incubated anaerobically at 37°C for 20–24 h. Growth from the BHIS plates was resuspended in sterile BHI broth. Samples of this mating mixture were plated on BHIS containing erythromycin (2 µg ml−1) to enrich for C. difficile cells that had participated in a conjugation event. After overnight incubation, growth was resuspended from the BHIS + Erm plates and spread on BHIS plates supplemented with both erythromycin (2 µg ml−1) and neomycin (10 µg ml−1) to select for transconjugants (C. difficile is naturally resistant to neomycin).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

We thank P. Stragier for helpful discussions and for communicating his genomic analysis before publication, T. Wilkins, P. Mullany and M. Popoff for providing C. difficile strains, and S. Melville for providing C. perfringens sigK sequence prior to publication. Preliminary sequence data from the Bacillus anthracis genome were obtained from The Institute for Genomic Research. Sequencing of B. anthracis was accomplished with support from the Office of Naval Research (ONR), the US Department of Energy (DOE), the National Institute of Allergy and Infectious Diseases (NIAID) and DERA. Preliminary sequence data from the Bacillus stearothermophilus genome were produced by the Advanced Center for Genome Technology at the University of Oklahoma. Preliminary sequence data from the C. difficile and Clostridium botulinum genomes were produced at the Sanger Centre. This work was supported by a predoctoral training grant (5 T32 AI07422) and a research grant (GM 42219) from the US Public Health Service.

Note added in proof

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References

We have learned from P. Stragier (pers. comm.) that the recently completed Clostridium tetani genome (Brüggemann, H., Baümer, S., Fricke, W. F., Wiezer, A., Liesegang, H., Decker, I., et al. (2003) The genome sequence of Clostridium tetani, the causative agent of tetanus disease. Proc Natl Acad Sci USA100: 1316–1321) reveals the presence of a third skin element, skinCt, disrupting the sigK gene of this organism. skinCt is more similar to B. subtilis skin than to skinCd in some respects; it is 47 kb in size and is inserted into the sigK gene in the same orientation as in B. subtilis. The site of insertion of skinCt, however, is the same as for skinCd. Thus, the sigK gene in different endospore-formers has sustained three independent insertions by similar elements.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. Note added in proof
  9. References
  • Adams, L.F., Brown, K.L., and Whiteley, H.R. (1991) Molecular cloning and characterization of two genes encoding sigma factors that direct transcription from a Bacillus thuringiensis crystal protein gene promoter. J Bacteriol 173: 38463854.
  • Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 33893402.
  • Arcuri, E.F., Wiedmann, M., and Boor, K.J. (2000) Phylogeny and functional conservation of σE in endospore-forming bacteria. Microbiology 146: 15931603.
  • Bannam, T.L., Crellin, P.K., and Rood, J.I. (1995) Molecular genetics of the chloramphenicol-resistance transposon Tn4451 from Clostridium perfringens: the TnpX site-specific recombinase excises a circular transposon molecule. Mol Microbiol 16: 535551.
  • Bassing, C.H., Swat, W., and Alt, F.W. (2002) The mechanism and regulation of chromosomal V (D) J recombination. Cell 109 (Suppl.): S45S55.
  • Brown, D.P., Ganova-Raeva, L., Green, B.D., Wilkinson, S.R., Young, M., and Youngman, P. (1994) Characterization of spo0A homologues in diverse Bacillus and Clostridium species identifies a probable DNA-binding domain. Mol Microbiol 14: 411426.
  • Catalano, F.A., Meador-Parton, J., Popham, D.L., and Driks, A. (2001) Amino acids in the Bacillus subtilis morphogenetic protein SpoIVA with roles in spore coat and cortex formation. J Bacteriol 183: 16451654.
  • Celli, J., and Trieu-Cuot, P. (1998) Circularization of Tn916 is required for expression of the transposon-encoded transfer functions: characterization of long tetracycline-inducible transcripts reading through the attachment site. Mol Microbiol 28: 103117.
  • Crellin, P.K., and Rood, J.I. (1997) The resolvase/invertase domain of the site-specific recombinase TnpX is functional and recognizes a target sequence that resembles the junction of the circular form of the Clostridium perfringens transposon Tn4451. J Bacteriol 179: 51485156.
  • Cutting, S., Oke, V., Driks, A., Losick, R., Lu, S., and Kroos, L. (1990) A forespore checkpoint for mother cell gene expression during development in B. subtilis. Cell 62: 239250.
  • Cutting, S., Driks, A., Schmidt, R., Kunkel, B., and Losick, R. (1991) Forespore-specific transcription of a gene in the signal transduction pathway that governs pro-σK processing in Bacillus subtilis. Genes Dev 5: 456466.
  • Errington, J. (1993) Bacillus subtilis sporulation: regulation of gene expression and control of morphogenesis. Microbiol Rev 57: 133.
  • Eveillard, M., Fourel, V., Barc, M.C., Kerneis, S., Coconnier, M.H., Karjalainen, T., et al. (1993) Identification and characterization of adhesive factors of Clostridium difficile involved in adhesion to human colonic enterocyte-like Caco-2 and mucus-secreting HT29 cells in culture. Mol Microbiol 7: 371381.
  • Golden, J.W., Robinson, S.J., and Haselkorn, R. (1985) Rearrangement of nitrogen fixation genes during heterocyst differentiation in the cyanobacterium Anabaena. Nature 314: 419423.
  • Golden, J.W., Carrasco, C.D., Mulligan, M.E., Schneider, G.J., and Haselkorn, R. (1988) Deletion of a 55-kilobase-pair DNA element from the chromosome during heterocyst differentiation of Anabaena sp. strain PCC 7120. J Bacteriol 170: 50345041.
  • Grawunder, U., West, R.B., and Lieber, M.R. (1998) Antigen receptor gene rearrangement. Curr Opin Immunol 10: 172180.
  • Haber, J.E. (1998) Mating-type gene switching in Saccharomyces cerevisiae. Annu Rev Genet 32: 561599.
  • Hachler, H., Berger-Bachi, B., and Kayser, F.H. (1987) Genetic characterization of a Clostridium difficile erythromycin-clindamycin resistance determinant that is transferable to Staphylococcus aureus. Antimicrob Agents Chemother 31: 10391045.
  • Haldenwang, W.G. (1995) The sigma factors of Bacillus subtilis. Microbiol Rev 59: 130.
  • Hallet, B., and Sherratt, D.J. (1997) Transposition and site-specific recombination: adapting DNA cut-and-paste mechanisms to a variety of genetic rearrangements. FEMS Microbiol Rev 21: 157178.
  • Hoch, J.A. (1993) Regulation of the phosphorelay and the initiation of sporulation in Bacillus subtilis. Annu Rev Microbiol 47: 441465.
  • Jahn, C.L., and Klobutcher, L.A. (2002) Gene remodeling in ciliated protozoa. Annu Rev Microbiol 10: 10.
  • Kamiya, S., Yamakawa, K., Ogura, H., and Nakamura, S. (1989) Recovery of spores of Clostridium difficile altered by heat or alkali. J Med Microbiol 28: 217221.
  • Kelly, C.P., and LaMont, J.T. (1998) Clostridium difficile infection. Annu Rev Med 49: 375390.
  • Kunkel, B., Losick, R., and Stragier, P. (1990) The Bacillus subtilis gene for the developmental transcription factor σK is generated by excision of a dispensable DNA element containing a sporulation recombinase gene. Genes Dev 4: 525535.
  • Larson, H.E., Price, A.B., and Borriello, S.P. (1980) Epidemiology of experimental enterocecitis due to Clostridium difficile. J Infect Dis 142: 408413.
  • Lonetto, M., Gribskov, M., and Gross, C.A. (1992) The σ70 family: sequence conservation and evolutionary relationships. J Bacteriol 174: 38433849.
  • Losick, R., and Stragier, P. (1992) Crisscross regulation of cell-type-specific gene expression during development in B. subtilis. Nature 355: 601604.
  • Losick, R., Youngman, P., and Piggot, P.J. (1986) Genetics of endospore formation in Bacillus subtilis. Annu Rev Genet 20: 625669.
  • Lu, S., Halberg, R., and Kroos, L. (1990) Processing of the mother-cell sigma factor, σK, may depend on events occurring in the forespore during Bacillus subtilis development. Proc Natl Acad Sci USA 87: 97229726.
  • Manganelli, R., Provvedi, R., Berneri, C., Oggioni, M.R., and Pozzi, G. (1998) Insertion vectors for construction of recombinant conjugative transposons in Bacillus subtilis and Enterococcus faecalis. FEMS Microbiol Lett 168: 259268.
  • Mani, N., Lyras, D., Barroso, L., Howarth, P., Wilkins, T., Rood, J.I., et al. (2002) Environmental response and autoregulation of Clostridium difficile TxeR, a sigma factor for toxin gene expression. J Bacteriol 184: 59715978.
  • Mullany, P., Wilks, M., Puckey, L., and Tabaqchali, S. (1994) Gene cloning in Clostridium difficile using Tn916 as a shuttle conjugative transposon. Plasmid 31: 320323.
  • Nakano, M.M., and Zuber, P. (1998) Anaerobic growth of a ‘strict aerobe’ (Bacillus subtilis). Annu Rev Microbiol 52: 165190.
  • Oke, V., and Losick, R. (1993) Multilevel regulation of the sporulation transcription factor σK in Bacillus subtilis. J Bacteriol 175: 73417347.
  • Popham, D.L., and Stragier, P. (1992) Binding of the Bacillus subtilis spoIVCA product to the recombination sites of the element interrupting the σK-encoding gene. Proc Natl Acad Sci USA 89: 59915995.
  • Popoff, M.R., Rubin, E.J., Gill, D.M., and Boquet, P. (1988) Actin-specific ADP-ribosyltransferase produced by a Clostridium difficile strain. Infect Immun 56: 22992306.
  • Resnekov, O., and Losick, R. (1998) Negative regulation of the proteolytic activation of a developmental transcription factor in Bacillus subtilis. Proc Natl Acad Sci USA 95: 31623167.
  • Rudner, D.Z., and Losick, R. (2002) A sporulation membrane protein tethers the pro-σK processing enzyme to its inhibitor and dictates its subcellular localization. Genes Dev 16: 10071018.
  • Rudner, D.Z., Fawcett, P., and Losick, R. (1999) A family of membrane-embedded metalloproteases involved in regulated proteolysis of membrane-associated transcription factors. Proc Natl Acad Sci USA 96: 1476514770.
  • Sato, T., Samori, Y., and Kobayashi, Y. (1990) The cisA cistron of Bacillus subtilis sporulation gene spoIVC encodes a protein homologous to a site-specific recombinase. J Bacteriol 172: 10921098.
  • Sauer, U., Treuner, A., Buchholz, M., Santangelo, J.D., and Durre, P. (1994) Sporulation and primary sigma factor homologous genes in Clostridium acetobutylicum. J Bacteriol 176: 65726582.
  • Sauer, U., Santangelo, J.D., Treuner, A., Buchholz, M., and Durre, P. (1995) Sigma factor and sporulation genes in Clostridium. FEMS Microbiol Rev 17: 331340.
  • 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., and Losick, R. (eds). Washington, DC: American Society for Microbiology Press, pp. 519525.
  • Stragier, P., and Losick, R. (1996) Molecular genetics of sporulation in Bacillus subtilis. Annu Rev Genet 30: 297241.
  • Stragier, P., Kunkel, B., Kroos, L., and Losick, R. (1989) Chromosomal rearrangement generating a composite gene for a developmental transcription factor. Science 243: 507512.
  • Sullivan, N.M., Pellett, S., and Wilkins, T.D. (1982) Purification and characterization of toxins A and B of Clostridium difficile. Infect Immun 35: 10321040.
  • Takami, H., Nakasone, K., Takaki, Y., Maeno, G., Sasaki, R., Masui, N., et al. (2000) Complete genome sequence of the alkaliphilic bacterium Bacillus halodurans and genomic sequence comparison with Bacillus subtilis. Nucleic Acids Res 28: 43174331.
  • Takemaru, K., Mizuno, M., Sato, T., Takeuchi, M., and Kobayashi, Y. (1995) Complete nucleotide sequence of a skin element excised by DNA rearrangement during sporulation in Bacillus subtilis. Microbiology 141: 323327.
  • Thorpe, H.M., and Smith, M.C. (1998) In vitro site-specific integration of bacteriophage DNA catalyzed by a recombinase of the resolvase/invertase family. Proc Natl Acad Sci USA 95: 55055510.
  • Wang, H., and Mullany, P. (2000) The large resolvase TndX is required and sufficient for integration and excision of derivatives of the novel conjugative transposon Tn5397 J Bacteriol 182: 65776583.
  • Wosten, M.M. (1998) Eubacterial sigma-factors. FEMS Microbiol Rev 22: 127150.
  • Wren, B.W., and Tabaqchali, S. (1987) Restriction endonuclease DNA analysis of Clostridium difficile. J Clin Microbiol 25: 24022404.
  • Yu, Y.T., and Kroos, L. (2000) Evidence that SpoIVFB is a novel type of membrane metalloprotease governing intercompartmental communication during Bacillus subtilis sporulation. J Bacteriol 182: 33053309.
  • Zhang, B., and Kroos, L. (1997) A feedback loop regulates the switch from one sigma factor to the next in the cascade controlling Bacillus subtilis mother cell gene expression. J Bacteriol 179: 61386144.
  • Zhang, B., Hofmeister, A., and Kroos, L. (1998) The prosequence of pro-σK promotes membrane association and inhibits RNA polymerase core binding. J Bacteriol 180: 24342441.
  • Zhang, B., Struffi, P., and Kroos, L. (1999)σK can negatively regulate sigE expression by two different mechanisms during sporulation of Bacillus subtilis. J Bacteriol 181: 4081–4088.