| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Centre for Molecular Microbiology and Infection, Division of Cell and Molecular Biology, Imperial College London, London SW7 2AZ, UK
2 Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Correspondence
Neil F. Fairweather
n.fairweather{at}imperial.ac.uk
Received 27 September 2007
Accepted 25 January 2008
Abbreviations: CDAD, Clostridium difficile-associated disease; CWP, cell wall protein.
Fully annotated microarray data are available in µG@Sbase (http://bugs.sgul.ac.uk) and also ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) with accession numbers E-BUGS-55 (environmental shock) and E-BUGS-56 (growth in the presence of antibiotics).
Tables of gene expression data are available as supplementary data with the online version of this paper.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
To understand how C. difficile interacts with its host, it is necessary to elucidate the survival mechanisms necessary for colonization and pathogenesis, and for persistence outside the host. During infection of the host C. difficile is exposed to the low pH environment of the stomach and high pH bile salts, in addition to various oxidative stresses. Analysis of the responses of C. difficile to general stresses will prove useful in revealing possible adaptation and resistance mechanisms in vivo.
The general mechanism of the heat-shock response is conserved in all organisms (Boorstein et al., 1994; Gupta, 1995), and in Gram-positive bacteria it consists of three major regulons: HrcA, CtsR and SigB (reviewed by Rosen & Ron, 2002; Schumann, 2003). These responses are not solely regulated by heat shock, and may be induced by any stress producing damaged or denatured proteins, for example ethanol, acid or salt shocks. This can be seen in Clostridium acetobutylicum, where butanol stress induces several heat-shock genes (Tomas et al., 2004). The heat-shock-induced sigma B (SigB, 
B) regulon varies between different organisms, and can incorporate as many as 200 genes (Hoper et al., 2005; Pane-Farre et al., 2006; van Schaik et al., 2007). Some genes (e.g. clpP and clpC) are regulated by multiple mechanisms (Petersohn et al., 2001).
Previous studies in C. difficile of the heat-shock protein GroEL have shown both the mRNA and protein to be strongly induced after heat shock from 30 to 43 °C; and GroEL was also secreted into the supernatant. However, no change in expression is observed after heat shock from 30 to 48 °C (Hennequin et al., 2001). Acidic conditions and high osmolarity also strongly induced groEL mRNA but this was not accompanied by a large relative increase in protein levels (Hennequin et al., 2001). In this study, we have used microarray analysis to determine the proportional changes in the entire transcriptome of C. difficile 630 in response to heat, oxidative, acid or alkaline shocks in addition to growth in the presence of antibiotics.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Stress induction. All experiments were carried out in biological triplicate. For growth in the presence of antibiotics, an overnight liquid culture was inoculated into BHI containing either 1 µg amoxicillin ml–1, 50 µg clindamycin ml–1 or 0.15 µg metronidazole ml–1. The cells were harvested between OD600 0.5 and 0.7.
For heat-shock experiments, cultures were grown in BHI broth at 30 °C to OD600 0.5–0.7. Aliquots (15 ml) were sealed into 50 ml Falcon tubes and incubated in a water bath at either 30 °C (control) or 42 °C (heat) for 15 min before harvest.
For pH-shock experiments, it was determined that C. difficile could withstand pH shock between pH 4.8 and pH 8.3. Aliquots (15 ml) of exponential phase culture were added to 50 ml Falcon tubes containing 0.5 ml BHI with either 63 µl 10 M sodium hydroxide (designed to alter the final pH to pH 8.5), 24 µl concentrated hydrochloric acid plus 39 µl water (to pH 4.5), or 63 µl water, and then incubated anaerobically for 15 min before harvest.
For oxygen exposure experiments, 15 ml aliquots were sealed into 50 ml Falcon tubes and removed from the anaerobic cabinet. Half of the aliquots were exposed to the atmosphere and then resealed and shaken, three times. Control aliquots remained sealed with Parafilm and were shaken but not exposed to the atmosphere. All aliquots were then Parafilm sealed and returned to the 37 °C anaerobic cabinet for 10 min – it was determined that at least 15 % of cells survived this treatment.
After treatment, in each case two volumes of RNAprotect bacteria reagent (Qiagen) were immediately added to cells, mixed, and incubated for 5 min. Cells were harvested by centrifugation and the cell pellet was stored at –80 °C.
RNA extraction and real-time PCR. RNA was extracted using a TRIzol Max bacterial RNA isolation kit (Invitrogen) according to the manufacturer's instructions, and then DNase treated. A 16S rRNA PCR amplification was carried out using 0.5 µl RNA as template to ensure there was no residual DNA.
For real-time PCR, RNA was reverse transcribed using a high-capacity cDNA archive kit (Applied Biosystems) according to the manufacturer's instructions. Custom TaqMan gene expression assays were designed by Applied Biosystems and carried out according to the manufacturer's instructions.
Microarray analysis. In 2004, the Bacterial Microarray Group at St George's, University of London, London, UK ( µG@S) designed a microarray based on the preliminary C. difficile 630 genome sequence and annotation available at the time. The microarray is described by Stabler et al. (2006). The µG@S CDv1.0.0 array design is available in µG@Sbase (accession number A-BUGS-20; http://bugs.sgul.ac.uk/A-BUGS-20) and also ArrayExpress (accession number A-BUGS-20). To aid in interpretation of the results, the predicted proteome was subjected to large batch Pfam analysis (http://pfam.sanger.ac.uk); the detected domains are presented in Supplementary Tables S1–S14 (available with the online journal).
RNA was concentrated using an RNeasy MinElute cleanup kit (Qiagen) according to the manufacturer's instructions and then adjusted to 1.25 µg µl–1. Duplicate competitive genomic DNA/ RNA microarrays were then carried out according to the standard µG@S hybridization protocol. The slides were scanned using the standard µG@S Affymetrix 428 scanning protocol, and fluorescence was quantified and flagged using ImaGene v5.5 (BioDiscovery). For antibiotic stress data, datasets from images at multiple intensities were combined using MAVI Pro v2.6.0 (MWG Biotech) to create a single dataset with larger dynamic range than is possible to determine from a single image.
Paired datasets for Cy5 (RNA) and Cy3 (DNA) microarray element fluorescence intensity quantifications were then uploaded to GeneSpring GX v6.1 (Agilent Technologies). Values below 0.01 were set to 0.01, to allow logarithmic interpretation of the data. The log2 ratio of the intensity in the RNA channel (Cy5) to that in the control DNA channel (Cy3) was calculated. Data were then normalized to each slide's median intensity ratio (of present and marginal microarray elements). Each gene in each experiment was then normalized to the mean of the ratios obtained under its ‘control’ condition. The default interpretation mode was set to ‘log of ratio’ (of present and marginal microarray elements) to allow correct parametric statistical analysis. Each stress was then tested separately for statistical differences between stress and control conditions using ANOVA. A 1 or 5 % confidence level was used, in conjunction with Benjamini-Hochberg, Bonferroni or no multiple testing correction.
Fully annotated microarray data for environmental-shock response has been deposited in µG@Sbase (http://bugs.sgul.ac.uk/E-BUGS-55) and also in ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) with accession number E-BUGS-55; and for growth in the presence of antibiotics with accession number E-BUGS-56.
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
1.5 pH units), or 10 min of atmospheric exposure. RNA was extracted and analysed by competitive RNA/DNA microarray using the µG@S C. difficile 630 microarray. ANOVA statistical analysis of the behaviour of each microarray spot reveals that distinct (but overlapping) gene sets are significantly regulated after each stress, as shown in Table 1
. This paper describes a selection of the results obtained from the study, concentrating on several groups of genes whose expression is altered by environmental shock or antibiotic treatment. Complete lists of statistically regulated genes are given in Supplementary Tables S1–S14 (available with the online journal). A summary of the behaviour of every gene after every stress, including each gene's annotation data and our Pfam domain analysis, are given in Supplementary Tables S15–S16 (available with the online journal).
|
). clpB is located in a cluster with genes encoding two hypothetical proteins, which are also significantly upregulated. htpG, encoding a heat-shock chaperone protein, is upregulated by heat shock, and cspD, encoding a cold-shock protein, is downregulated. Analysis by real-time PCR of groEL expression over a time-course of heat shock reveals that it is induced most strongly at 15 min (13-fold, P<10–5), and is also significantly upregulated at all other time points examined (5, 30 and 60 min; >3-fold induction, P<10–3) (Fig. 1b). The heat-shock gene clusters are also upregulated after acid shock (Fig. 1a). The characterization of the response of the C. difficile vegetative cell to acid shock may have direct relevance to CDAD precipitated by gastric acid suppressants; in these patients gastric pH can rise above pH 5 (Jump et al., 2007), and we have shown vegetative C. difficile to be able to tolerate this.
|
Genes encoding four enzymes of the spermidine synthesis pathway – speA (encoding arginine decarboxylase), speD (encoding S-adenosylmethionine decarboxylase proenzyme), speE (encoding spermidine synthase), and speB (encoding putative agmatinase) – are situated in a cluster within the genome. All four genes are upregulated on average 3.1-fold in response to heat shock (Fig. 2a). Additionally, genes encoding the four components of the spermidine/putrescine ABC transporter – potA (encoding ATP-binding protein), potB (encoding permease protein), potC (encoding putative permease protein) and potD (encoding substrate-binding lipoprotein) – are located in a cluster with a gene encoding a putative transcriptional regulator, and all five genes are co-ordinately upregulated by heat shock by a mean of 3.3-fold (Fig. 2a). These spermidine synthesis and spermidine/putrescine ABC transporter gene clusters are also co-ordinately upregulated in response to alkali shock by a mean of 2.5- and 3.0-fold, respectively. Analysis by real-time PCR of the expression of speE (Fig. 2b) and potC (data not shown) over a time-course of heat shock reveals that significant induction is observed after 5 and 15 min: speE 4.1-fold (P<10–3) and 5.0-fold (P<10–4); and potC 7.2-fold (P<10–5) and 5.6-fold (P<10–3), respectively. No significant induction of either gene is seen at 30 or 60 min of heat shock. Spermidine is a polyamine involved in various cellular processes, and is linked to abiotic stress resistance in apple and pear fruit trees (Wen et al., 2007), and peroxide resistance in mouse fibroblasts (Rider et al., 2007). The Pot ABC transporter is known to be surface associated in Streptococcus pneumoniae (Shah et al., 2006) and may contribute to virulence in a murine sepsis model (Ware et al., 2006).
|
W) (Wiegert et al., 2001). However, BLASTP searching of the C. difficile 630 and R20352 (partial) published genome sequences reveals that there is no close homologue of W in C. difficile. Similarities between the alkali responses of C. difficile and B. subtilis can still be noted. The histidine synthesis pathway has previously been detected as being upregulated after alkaline shock (Wiegert et al., 2001), and this was also observed in this study. Tellurite and tellurium resistance proteins are upregulated in B. subtilis (Petersohn et al., 2001; Wiegert et al., 2001), and these genes were upregulated by a mean of 1.4-fold by alkali shock in this study, although none significantly so. However, induction of Na+/H+ antiporters and multidrug efflux protein encoding genes was not observed, and neither was any evidence of the phosphate starvation observed in Bacillus licheniformis by Hornbaek et al. (2004).
Despite being regarded as strictly anaerobic, C. difficile does appear to respond specifically to the oxidative stress of atmospheric exposure. In total, 7.6 % (18/238) of genes annotated as encoding proteins involved in electron transport are significantly upregulated. These include a gene cluster encoding a putative oxidoreductase (NAD/FAD binding subunit, electron transfer subunit and acetyl-CoA synthase subunit) that is co-ordinately upregulated (Fig. 3). Genes encoding a putative nitric oxide reductase flavoprotein and NorV (a putative anaerobic nitric oxide reductase flavorubredoxin), which may be involved in protection against nitric oxide, are also upregulated (Fig. 3), as are other redox proteins including thioredoxins, thiol peroxidase and rubrerythrin.
|
The observed response of C. difficile to atmospheric exposure is very interesting. The induction of a large number of presumably protective oxidoreductases and electron transport components implies that C. difficile does in fact have some mechanisms for tolerance of limited oxidizing conditions. The low viability of our atmospherically exposed cultures suggests that C. difficile 630 is substantially damaged by the level of oxygen exposure used in this study. However, the observed responses may promote survival, if not growth, in microaerophilic environments. The gut is not strictly anaerobic, and C. difficile either must grow and survive in strictly anaerobic niches or must be capable of tolerating microaerophilic conditions. It is interesting to note that sporulation does not appear to be induced by the environmental stresses tested, but this may be explained by the time point tested – no observed sporulation would be expected after 15 min of stress.
Growth in the presence of all three antibiotics tested resulted in the general upregulation of transcription and translation machinery. The genome region CD0059–CD0099 contains genes encoding many ribosomal proteins and other growth-related proteins. This region appears to be co-ordinately upregulated by amoxicillin (with the exception of CD0065, which is unchanged) by a mean of 4.4-fold (range 2.2- to 6.0-fold) (Fig. 4a). This includes significant upregulation of rpoB, rpoC and rpoA (encoding RNA polymerase β, β' and 
chains) and 25 ribosomal protein-encoding genes (plus 4 situated elsewhere). Overall, 30S and 50S ribosomal protein genes are induced by 3.7- and 3.8-fold, respectively, after growth in the presence of amoxicillin. Five tRNA synthetases are significantly upregulated after amoxicillin treatment, and genes encoding tRNA-associated proteins are on average upregulated 1.8-fold. The CD0059–CD0099 region is also co-ordinately upregulated after both clindamycin and metronidazole exposure, although no genes significantly so. Taken together these results imply that a general increase in transcription and translation may be a response to the presence of antibiotics, particularly amoxicillin.
|
The primary effect of growth in the presence of antibiotics appears to be on the surface composition of C. difficile. The region of the genome from CD2767–CD2801 encodes many surface-associated proteins, including cell wall biosynthesis and modification proteins (Calabi et al., 2001; Sebaihia et al., 2006). Also present are 13 genes encoding cell wall proteins (CWPs) containing the putative bacterial cell wall-binding Pfam domain PF04122; these are all paralogues of the slpA gene (Calabi et al., 2001; Karjalainen et al., 2001). In total, 27 of the 35 genes in this region (34 of 47 reporters) are over 1.3-fold upregulated by amoxicillin – 6 significantly so – and the cluster as a whole is on average 2.3-fold upregulated. In addition to the 13 CWP-encoding genes present in this cluster, there are a further 16 localized throughout the genome. Each of these slpA paralogues is represented on the microarray by two separate reporters each printed in duplicate. On average, all microarray reporters representing CWP-encoding genes are upregulated in the presence of ampicillin (2.3-fold), clindamycin (1.5-fold) and metronidazole (1.3-fold). At least one reporter for cwp2, cwp84, cwp6 and cwp7 is significantly upregulated by amoxicillin (Fig. 4b
) and for cwp21 is upregulated by clindamycin, whereas cwp23 is downregulated by clindamycin (data not shown). Antibiotic upregulation of slpA, cwp66 and cwp84 transcription has been previously demonstrated in C. difficile, although the level of regulation varied between strains and antibiotics (Anne Collignon, personal communication).
Expression of genes encoding several cell wall enzymes was affected by treatment with amoxicillin, which mimics D-alanyl-D-alanine and disrupts cell wall synthesis (reviewed by Macheboeuf et al., 2006). dltA and dacF were upregulated, as was a gene cluster incorporating murD, mraY and murF. All of the mur genes, which are involved in the peptidoglycan synthesis pathway, were upregulated at least 1.2-fold (mean value 1.9-fold). Two further genes were significantly induced by amoxicillin, encoding a penicillin-binding protein and a putative β-lactamase. In addition, induction was observed of a gene cluster encoding a penicillin-binding protein and a penicillin-binding protein repressor. Overall, all annotated penicillin-binding protein genes are either upregulated or downregulated by a mean of 1.7-fold; none remain unaffected by amoxicillin. These results suggest that C. difficile downregulates the targets with highest affinity for amoxicillin, and upregulates all alternative pathways. A cell wall stimulon has recently been described for Staphylococcus aureus (Utaida et al., 2003), which may be upregulated in vancomycin intermediate-resistance strains (McAleese et al., 2006). Comparison between this regulon and those genes regulated in this study reveals some overlap; however, some genes have disparate regulation and many S. aureus genes are not identified in C. difficile.
Three principle components of the secretion apparatus, secE, secY and secA2, were also upregulated by antibiotics, particularly amoxicillin (data not shown). Taken together with the upregulation of cell wall biosynthetic component transcripts, it is tempting to speculate that perhaps C. difficile alters the composition of its cell wall in response to antibiotics, perhaps producing a thicker or denser cell wall as a physical barrier to antibiotic penetration, similar to the mechanism that confers intermediate vancomycin resistance in S. aureus (reviewed by Appelbaum & Bozdogan, 2004). However, this phenomenon has not been previously reported for C. difficile in response to antibiotics.
The region of the genome CD0226–CD0272 contains primarily genes encoding flagellar proteins. These genes appear to be co-ordinately upregulated after growth in the presence of amoxicillin and clindamycin by at least 1.6-fold in each case (mean values 3.0- and 2.3-fold, respectively), of which 12 (out of 40) are significantly upregulated by amoxicillin (Fig. 4c). The 11 other chemotaxis-related protein-encoding genes, present in two clusters within the genome, are also upregulated by amoxicillin by a mean of 1.4-fold, and by clindamycin by a mean of 1.6-fold (although only 2 genes are significantly regulated). The induction of flagellar protein-encoding transcripts and other chemotaxis-related proteins by both amoxicillin and clindamycin implies that cellular motility is increased in the presence of these antibiotics. The gut mucus of animals and humans serves as a chemoattractant for C. difficile, and chemotactic motility is positively correlated with the relative virulence of several C. difficile strains in a hamster model (Borriello et al., 1988). Increased motility could perhaps allow the bacterium to move down a concentration gradient of antibiotic, or to move towards higher nutrient concentrations.
Concluding remarks
Our data show that some components of the stress responses of C. difficile are similar to those observed in other organisms, whilst other components appear to be unique. The transcriptional profiles of C. difficile 630 are similar after growth in the presence of amoxicillin and clindamycin. In contrast, metronidazole treatment resulted in a very small change in transcription patterns. The effect of these antibiotics on the synthesis of CWPs merits further study. Further analysis of other genes identified in this study will enable targeted investigation into disruption of the genes, proteins and pathways shown to be necessary for stress survival, and therefore adaptation to the host environment and colonization.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
B regulon in Staphylococcus aureus and its regulation. Int J Med Microbiol 296, 237–258.[CrossRef][Medline]B regulon of Bacillus cereus and conservation of B -regulated genes in low-GC-content gram-positive bacteria. J Bacteriol 189, 4384–4390.This article has been cited by other articles:
|
|
 |
|
  S. D. Baines, R. O'Connor, G. Huscroft, K. Saxton, J. Freeman, and M. H. Wilcox Mecillinam: a low-risk antimicrobial agent for induction of Clostridium difficile infection in an in vitro human gut model J. Antimicrob. Chemother., April 1, 2009; 63(4): 838 - 839. [Full Text] [PDF]
|
|
|
|
 |
|
 I. R. Poxton Proceedings from the 2nd International Clostridium difficile Symposium, Maribor, Slovenia, June 2007. J. Med. Microbiol., June 1, 2008; 57(Pt 6): 683 - 794. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
, 42 °C; 
, 30 °C.
