Final Report Summary - DIFFISORT (Characterization of a major family of cell wall proteins of Clostridium difficile – the sortase family.)
Clostridium difficile, a Gram-positive spore-forming bacterium, is the major cause of intestinal diseases associated with antibiotic therapy. C. difficile infection (CDI) can result in a spectrum of symptoms, ranging from mild diarrhea to life-threatening clinical conditions such as severe pseudomembranous colitis and colon perforation. C. difficile presents a serious risk of nosocomial infection (hospital acquired) but it is increasingly recognized also as a community-associated disease. C. difficile takes advantage of the disturbance of the normal colonic microbiota, following antibiotic treatment, to colonize the gastrointestinal tract. Therefore C.difficile must be implanted in the gut and attached to epithelial cells, in order to cause disease. C. difficile, like most bacterial pathogens, displays proteins on its surface (adhesins) that may interact with host proteins and are thought to be essential for the subsequent process of colonization. Surface proteins can be non-covalently associated to outermost cell wall components or covalently anchored to the peptidoglycan. Covalent attachment of the surface proteins to the peptidoglycan also occurs and is promoted by sortases, which are enzymes conserved in Gram-positive bacteria. Many clinically significant pathogens (e.g. S. aureus) require a functional sortase to be fully virulent, and sortase enzymes are promising therapeutic targets for the development of novel antibiotics.
At the beginning of this project, our preliminary bioinformatics analysis had identified a gene encoding a putative sortase in the genome of C. difficile and several putative sortase substrates had also been identified. However, there was nothing known about covalent linkage of surface proteins to peptidoglycan or the role of these proteins in C. difficile. Within this fellowship, the overall aim was to characterize a major family of cell wall proteins of C. difficile–the sortase family. It is important to characterize the function of cell wall proteins because they represent potential targets for novel therapeutic strategies, e.g. as antigens of novel vaccines or as targets for antibiotics. Sortase could constitute a new target to inhibit the colonization process and therefore block the development of infections.
In this project, the following tasks have been achieved:
(1) A sortase deletion mutant was successfully generated by allele exchange. The same method was also used to replace the native promoter of the chromosomal sortase encoding gene with an inducible promoter, leading to conditional expression of the sortase. Antibodies raised against the purified sortase protein were used to detect the sortase in the different stains. Whereas sortase could not be detected in the wild-type strain, the expression of the induced sortase was confirmed by Western-blot analysis and subcellular fractionation revealed that the sortase is a membrane protein.
(2) In-silico analysis led to the identification of 4 putative sortase substrates found in all sequenced strains of C. difficile. Three of these putative substrates encoding genes (CD0183, CD2831 and CD3392) have been expressed and purified in E. coli. Purified proteins were then used to raise specific antibodies. The level of the putative substrates has been examined by western-blot in subcellular extracts containing proteins present in cytosol, membrane, cell wall (peptidoglycan) or culture supernatant. Cell-wall extract was generated by digestion of the peptidoglycan by an amidase overexpressed and purified in our laboratory. CD0183 was well detected using specific antibodies but surprisingly, it was mainly found in the supernatant fraction in the wild-type strain, the sortase-induced strain or the sortase mutant strain, indicating that CD0183 is a secreted protein. Moreover, by addition of a 20 kDa tag to the C-terminus extremity of CD0183, we demonstrated using western blot that CD0183 is not cleaved by the sortase and therefore that it is not a sortase substrate.
(3) CD3392 could not be detected in the wild-type strain, so it was artificially overexpressed. Whereas overexpressed CD3392 was only weakly detected in all subcellular fractions of the wild-type strain, a massive level of CD3392 was present in the membrane fraction of the sortase mutant strain. This result suggested that CD3392 requires the activity of the sortase to be correctly translocated across the membrane and also that CD3392 is probably degraded after translocation as it is present only at low level in the wild-type strain. The importance of the conserved sorting motif, present at the C-terminal extremity of CD3392, was then investigated. Modification of this sorting motif through genetic manipulations revealed it is required for the correct translocation of CD3392 through the membrane. Taken together, these results clearly demonstrate that CD3392 is a substrate for the sortase SrtB.
(4-1) CD2831 could not be detected in the wild-type strain, so it was artificially overexpressed. Overexpressed CD2831 in the wild-type, the sortase induced or the sortase mutant strain led to the detection of this protein in the supernatant fraction mainly, indicating that CD2831 is a secreted protein. However, addition of a 20 kDa tag to the C-terminus extremity of CD2831 revealed that this protein could be cleaved by the sortase. Therefore, CD2831 represents the first example of a sortase substrate which is not found attached to the cell wall.
(4-2) A recent paper reported that CD2831 is efficiently cleaved in vitro by the secreted metalloprotease CD2830. To investigate whether the secretion of CD2831 into the supernatant might result of its cleavage by the CD2830 protease, a CD2830 in-frame deletion mutant was generated. Western blot analysis and immunofluorescence microscopy revealed that CD2831 is associated to the cell surface in the CD2830 deletion mutant. In addition, CD2831 returns in the supernatant in a srtB-CD2830 double mutant, indicating that CD2831 is actually a sortase substrate.
(4-3) The anchor structure of CD2831 has been investigated by LC-MS and MS-MS and the covalent anchoring of this protein to the meso-diaminopimelic acid residue of the peptidoglycan stem chain has been unambiguously shown.
(4-4) Cyclic-di-GMP-mediated regulation of both the expression and the CD2831 surface display has also been demonstrated. While expression of the protease CD2830 is repressed in presence of high level of c-di-GMP, the expression of the sortase substrate CD2831 is activated in this condition. By overexpressing the diguanylate cyclase DccA in the wild-type strain, we created a strain able to produce artificial high level of c-di-GMP. CD2831 is clearly expressed at higher level in this strain than in the wild-type strain and is found associated to the cell wall fraction, indicating that the protease CD2830 is no longer produced in these conditions.
(5) A role for the sortase enzyme in the adhesion to type V collagen and to the fibrinogen was also identified, indicating that the sortase is important to the physiology of C. difficile and might be involved in the virulence of C. difficile.
In conclusion, we showed for the first time in this project that C. difficile has a functional sortase enzyme able to covalently anchor some identified substrates to the peptidoglycan. However, the sortase system is unusual and even unique in C. difficile as the anchoring of its substrates to the cell wall is regulated by the activity of proteases. While the protease regulating the surface display of CD3392 remains unknown, we demonstrated that CD2831 is secreted into the supernatant upon cleavage by the protease CD2830. We also deciphered the c-di-GMP-mediated mechanism involved in the tight regulation of CD2831 surface anchoring. Phenotypic analysis revealed that the sortase is important for collagen or fibrinogen binding and therefore for the physiology of C. difficile. Thus, this project identified new virulence factors of C. difficile and contributed to expand our knowledge in the pathogenesis process of C. difficile. This project contributed to strengthen European expertise in the field of cell wall proteins of C. difficile and constitutes a first step toward the development of new therapeutic strategies.
At the beginning of this project, our preliminary bioinformatics analysis had identified a gene encoding a putative sortase in the genome of C. difficile and several putative sortase substrates had also been identified. However, there was nothing known about covalent linkage of surface proteins to peptidoglycan or the role of these proteins in C. difficile. Within this fellowship, the overall aim was to characterize a major family of cell wall proteins of C. difficile–the sortase family. It is important to characterize the function of cell wall proteins because they represent potential targets for novel therapeutic strategies, e.g. as antigens of novel vaccines or as targets for antibiotics. Sortase could constitute a new target to inhibit the colonization process and therefore block the development of infections.
In this project, the following tasks have been achieved:
(1) A sortase deletion mutant was successfully generated by allele exchange. The same method was also used to replace the native promoter of the chromosomal sortase encoding gene with an inducible promoter, leading to conditional expression of the sortase. Antibodies raised against the purified sortase protein were used to detect the sortase in the different stains. Whereas sortase could not be detected in the wild-type strain, the expression of the induced sortase was confirmed by Western-blot analysis and subcellular fractionation revealed that the sortase is a membrane protein.
(2) In-silico analysis led to the identification of 4 putative sortase substrates found in all sequenced strains of C. difficile. Three of these putative substrates encoding genes (CD0183, CD2831 and CD3392) have been expressed and purified in E. coli. Purified proteins were then used to raise specific antibodies. The level of the putative substrates has been examined by western-blot in subcellular extracts containing proteins present in cytosol, membrane, cell wall (peptidoglycan) or culture supernatant. Cell-wall extract was generated by digestion of the peptidoglycan by an amidase overexpressed and purified in our laboratory. CD0183 was well detected using specific antibodies but surprisingly, it was mainly found in the supernatant fraction in the wild-type strain, the sortase-induced strain or the sortase mutant strain, indicating that CD0183 is a secreted protein. Moreover, by addition of a 20 kDa tag to the C-terminus extremity of CD0183, we demonstrated using western blot that CD0183 is not cleaved by the sortase and therefore that it is not a sortase substrate.
(3) CD3392 could not be detected in the wild-type strain, so it was artificially overexpressed. Whereas overexpressed CD3392 was only weakly detected in all subcellular fractions of the wild-type strain, a massive level of CD3392 was present in the membrane fraction of the sortase mutant strain. This result suggested that CD3392 requires the activity of the sortase to be correctly translocated across the membrane and also that CD3392 is probably degraded after translocation as it is present only at low level in the wild-type strain. The importance of the conserved sorting motif, present at the C-terminal extremity of CD3392, was then investigated. Modification of this sorting motif through genetic manipulations revealed it is required for the correct translocation of CD3392 through the membrane. Taken together, these results clearly demonstrate that CD3392 is a substrate for the sortase SrtB.
(4-1) CD2831 could not be detected in the wild-type strain, so it was artificially overexpressed. Overexpressed CD2831 in the wild-type, the sortase induced or the sortase mutant strain led to the detection of this protein in the supernatant fraction mainly, indicating that CD2831 is a secreted protein. However, addition of a 20 kDa tag to the C-terminus extremity of CD2831 revealed that this protein could be cleaved by the sortase. Therefore, CD2831 represents the first example of a sortase substrate which is not found attached to the cell wall.
(4-2) A recent paper reported that CD2831 is efficiently cleaved in vitro by the secreted metalloprotease CD2830. To investigate whether the secretion of CD2831 into the supernatant might result of its cleavage by the CD2830 protease, a CD2830 in-frame deletion mutant was generated. Western blot analysis and immunofluorescence microscopy revealed that CD2831 is associated to the cell surface in the CD2830 deletion mutant. In addition, CD2831 returns in the supernatant in a srtB-CD2830 double mutant, indicating that CD2831 is actually a sortase substrate.
(4-3) The anchor structure of CD2831 has been investigated by LC-MS and MS-MS and the covalent anchoring of this protein to the meso-diaminopimelic acid residue of the peptidoglycan stem chain has been unambiguously shown.
(4-4) Cyclic-di-GMP-mediated regulation of both the expression and the CD2831 surface display has also been demonstrated. While expression of the protease CD2830 is repressed in presence of high level of c-di-GMP, the expression of the sortase substrate CD2831 is activated in this condition. By overexpressing the diguanylate cyclase DccA in the wild-type strain, we created a strain able to produce artificial high level of c-di-GMP. CD2831 is clearly expressed at higher level in this strain than in the wild-type strain and is found associated to the cell wall fraction, indicating that the protease CD2830 is no longer produced in these conditions.
(5) A role for the sortase enzyme in the adhesion to type V collagen and to the fibrinogen was also identified, indicating that the sortase is important to the physiology of C. difficile and might be involved in the virulence of C. difficile.
In conclusion, we showed for the first time in this project that C. difficile has a functional sortase enzyme able to covalently anchor some identified substrates to the peptidoglycan. However, the sortase system is unusual and even unique in C. difficile as the anchoring of its substrates to the cell wall is regulated by the activity of proteases. While the protease regulating the surface display of CD3392 remains unknown, we demonstrated that CD2831 is secreted into the supernatant upon cleavage by the protease CD2830. We also deciphered the c-di-GMP-mediated mechanism involved in the tight regulation of CD2831 surface anchoring. Phenotypic analysis revealed that the sortase is important for collagen or fibrinogen binding and therefore for the physiology of C. difficile. Thus, this project identified new virulence factors of C. difficile and contributed to expand our knowledge in the pathogenesis process of C. difficile. This project contributed to strengthen European expertise in the field of cell wall proteins of C. difficile and constitutes a first step toward the development of new therapeutic strategies.