A Highly Specific Holin-Mediated Mechanism Facilitates the Secretion of Lethal Toxin TcsL in Paeniclostridium sordellii

Abstract

Protein secretion is generally mediated by a series of distinct pathways in bacteria. Recently, evidence of a novel bacterial secretion pathway involving a bacteriophage-related protein has emerged. TcdE, a holin-like protein encoded by toxigenic isolates of Clostridioides difficile, mediates the release of the large clostridial glucosylating toxins (LCGTs), TcdA and TcdB, and TpeL from C. perfringens uses another holin-like protein, TpeE, for its secretion; however, it is not yet known if TcdE or TpeE secretion is specific to these proteins. It is also unknown if other members of the LCGT-producing clostridia, including Paeniclostridium sordellii (previously Clostridium sordellii), use a similar toxin-release mechanism. Here, we confirm that each of the LCGT-producing clostridia encode functional holin-like proteins in close proximity to the toxin genes. To characterise the respective roles of these holin-like proteins in the release of the LCGTs, P. sordellii and its lethal toxin, TcsL, were used as a model. Construction and analysis of mutants of the P. sordellii tcsE (holin-like) gene demonstrated that TcsE plays a significant role in TcsL release. Proteomic analysis of the secretome from the tcsE mutant confirmed that TcsE is required for efficient TcsL secretion. Unexpectedly, comparative sample analysis showed that TcsL was the only protein significantly altered in its release, suggesting that this holin-like protein has specifically evolved to function in the release of this important virulence factor. This specificity has, to our knowledge, not been previously shown and suggests that this protein may function as part of a specific mechanism for the release of all LCGTs.

Specialized pathogenic cells release Tc toxins using a type 10 secretion system

“…The T10SS-mediated release of Tc toxins also fits well into the emerging paradigm of anti-eukaryotic toxin export by phage-derived proteins ( 48 ). Notable examples include release of the Clostridium difficile toxins A and B via a holin and endolysin ( 56-58 ), the holin-dependent export of large clostridial glucosylating toxins ( 59, 60 ) and the endolysin-dependent secretion of typhoid toxin ( 61 ). Notably, in all the cases described above, the current consensus is that the holin and/or endolysin-mediated export occurs via a non-lytic mechanism ( 48, 56, 59-61 ), although in light of the data presented here the potential use of suicidal soldier cell subpopulations also by these pathogens is a fascinating possibility that is worth investigating.…”

“…Notable examples include release of the Clostridium difficile toxins A and B via a holin and endolysin ( 56-58 ), the holin-dependent export of large clostridial glucosylating toxins ( 59, 60 ) and the endolysin-dependent secretion of typhoid toxin ( 61 ). Notably, in all the cases described above, the current consensus is that the holin and/or endolysin-mediated export occurs via a non-lytic mechanism ( 48, 56, 59-61 ), although in light of the data presented here the potential use of suicidal soldier cell subpopulations also by these pathogens is a fascinating possibility that is worth investigating. Finally, combining the observation that YenTc deletion causes full loss of Y. entomophaga virulence ( 22 ) with our discovery that Tc toxins are produced by only few cells in a population leads to the startling insight that pathogen virulence can be determined by a small number of specialized soldier cells.…”

Occurrence and potential mechanism of holin-mediated non-lytic protein translocation in bacteria

“…All holins that are genetically associated with the known large clostridial toxins are able to transport endolysins and are therefore bona fide holins [ 40 ][ 37 ][ 41 ][ 42 ]. In case of the holin TcdE, its phage origin could be traced back, and remnants of the associated endolysin gene have been found in direct vicinity of tcdE in the PaLoc of C. difficile [ 40 ].…”

“…The holin specifically is required for the toxin transport, and as recombinant TpeE alone sufficed for TpeL secretion, it has been suggested that the holin may transport the toxin directly [ 42 ]. Also the holin of P. sordellii has been found to specifically promote toxin release in that organism [ 41 ]. It has not been shown directly that the toxins themselves are transported by the respective holins, which is why the authors carefully claimed only a function in the release of the toxin, but, like in the case of C. difficile , the toxins lack signal peptides for Sec transport, are genetically linked to holins on pathogenicity loci, and there are no known alternative pathways for the non-lytic release of folded proteins in these clostridia.…”

“…As the same endolysins have been shown to be translocated by holins of distinct families (e.g. [ 41 ]), and as similar holins likely can transport highly distinct protein substrates (such as toxins or bacteriocins), holins most likely use a conserved, yet to be identified general recognition principle. The “membrane-weakening and flipping” mechanism directly links recognition to translocation, and therefore similar mechanisms may be implemented by distinct holins to carry out non-lytic translocation, possibly involving a flipping of C-terminal domains.…”