Discovery at the VICB
Secret of Attack by a Deadly Bacterial Toxin
By: Carol A. Rouzer, VICB Communications
Published: December 22, 2016
Insight into the mode of entry of Clostridium difficile toxin TcdA into cells provides possible new pathways for fighting C. difficile colitis.
Clostridium difficile is a Gram-positive anaerobic bacterium that is a major cause of gastroenteritis, particularly among people who have been treated with broad-spectrum antibiotics. The single most common hospital-acquired infection, C. difficile colitis produces symptoms ranging from mild diarrhea to severe illness and death. In fact, C. difficile is the major cause of gastroenteritis-associated fatalities in the United States. Tissue damage resulting from C. difficile infection is primarily attributable to the production of two toxins, TcdA and TcdB, both of which cause epithelial cell death, fluid secretion, and inflammation. Despite their similar structures, TcdA and TcdB enter cells via different interactions and are taken up by distinct mechanisms. Recent work has demonstrated that uptake of TcdB occurs through a clathrin-dependent process. However, this does not appear to be the case for TcdA, leading Vanderbilt Institute of Chemical Biology investigator Borden Lacy, her graduate student Ramya Chandrasekaran, and their collaborator Anne Kenworthy to solve the puzzle [R. Chandrasekaran, et al., PLoS Pathog., (2016) 12, e1006070].
Much of what is known about the three-dimensional structure of TcdA comes from prior work in the Lacy laboratory (N. M. Chumbler, et al. (2016) Nat. Microbiol., 1, 15002, Figure 1). A large protein, TcdA comprises four domains that work together to gain entry to and then damage the cell. The N-terminal glucosyltransferase domain (GTD) is an enzyme that inactivates small intracellular GTPases, leading to cytoskeletal derangements, cell rounding, and ultimately apoptosis. Delivery of the GTD into the target cell occurs through the action of the other three domains. Evidence indicates that the C-terminal combined repetitive oligopeptides (CROPS) domain is at least partially responsible for the interaction of TcdA with receptor molecules on the target cell surface. Subsequent uptake into acidic endosomes results in activation of the delivery domain that then forms a pore in the endosomal membrane through which the GTD is transported along with the autoprotease domain (APD). Binding of cellular inositol hexakisphosphate then activates the APD, which cleaves off the GTD, allowing it to move freely into the cytosol.
FIGURE 1. Structure of TcdA (a) Primary structure of TcdA showing the four domains: GTD, glucosyltransferase domain; APD autoprotease domain; delivery domain; CROPS, combined repetitive oligopeptides domain. (b) Diagrammatic representation of the crystal structure of TcdA lacking the CROPS domain. The other three domains are colored as in (a). (c) same as (b) but rotated 90o. (d) Crystal structure of TcdA superimposed onto a map of the structure obtained by electron microscopy. The uncolored areas are presumed to correspond to the CROPS domain. Figure reprinted by permission from Macmillan Publishers Ltd from N. M. Chumbler, et al. (2016) Nat. Microbiol., 1, 15002. Copyright 2016.
The investigators began their work by challenging a prior conclusion that TcdA uses a clathrin-mediated process for cell entry. Choosing the human colorectal adenocarcinoma (Caco-2) cell line for these studies, they first used shRNA to knock down expression of the clathrin heavy chain (CHC) in order to inactivate clathrin-dependent uptake pathways. They found that cells expressing the shRNA directed against CHC were resistant to intoxication by TcdB but not TcdA. Furthermore, cells exposed to TcdA bearing a fluorescent tag (TcdA-546) demonstrated no colocalization of the toxin with CHC detected by immunofluorescent staining (Figure 2A). In contrast, when labeled transferrin (Tf-647) was added to the cells, it colocalized with CHC, consistent with its known clathrin-dependent uptake mechanism (Figure 2B). These results confirmed that TcdB but not TcdA is taken up by cells using a clathrin-mediated process.
FIGURE 2. TcdA does not colocalize with clathrin heavy chain CHC. (A) Caco-2 cells were exposed to TcdA bearing a fluorescent label (Tcd-546) at cold temperatures to allow toxin binding. Then the cells were warmed to 37 oC and incubated for the indicated times. CHC, stained using a specific antibody (red) did not colocalize with Tcd-546 (green) at any of the time points evaluated. (B) A control experiment using transferrin bearing a fluorescent label (Tf-647) reveals colocalization with CHC, as indicated by the yellow color (combined red and green) in the merged image. Figure reproduced under the Creative Commons CCO 1.0 Universal License from R. Chandrasekaran, et al., PLoS Pathog., (2016) 12, e1006070.
Dynamin is a large GTPase that plays a role in endocytosis by facilitating the scission and release of newly formed vesicles. To see if dynamin is involved in TcdA uptake, the researchers used siRNA to knock down dynamin-1 expression in Caco-2 cells. They found that this treatment improved the cells' survival upon exposure to TcdA. Furthermore, treatment of the cells with dynasore, an inhibitor of the dynamin GTPase suppressed TcdA-dependent Rac1 glycosylation. These results supported the hypothesis that dynamin is required for TcdA internalization in Caco-2 cells.
Caveolae-mediated endocytosis requires dynamin but not clathrin, leading the investigators to hypothesize that TcdA uptake might occur by a process involving caveolae. They tested this hypothesis by using RNAi to knock down the expression of caveolae-associated proteins. The results showed that knockdown of Cav1, cavin1, or PACSIN2 conveyed TcdA resistance to Caco-2 cells. These results were somewhat surprising, however, because Caco-2 cells have few caveolae, calling into question the importance of this pathway in TcdA endocytosis. Further studies using mouse embryonic fibroblasts (MEFs) that contain caveolae showed no colocalization of TcdA-546 with Cav1 or cavin1. Furthermore, TcdA was able to intoxicate MEFs lacking the gene for Cav1 or treated with RNAi to knockdown expression of Cav1. These results ruled out a role for caveolae in TcdA uptake by MEFs.
Despite the apparent lack of involvement of caveolae in TcdA endocytosis, the investigators found that TcdA-546 colocalized with PACSIN2 in MEF cells (Figure 3), and siRNA-mediated knockdown of PACSIN2 led to decreased uptake of TcdA. PACSIN2 plays a role in caveolae-dependent endocytosis, but the investigators could show that pools of PACSIN2 that associated with TcdA-546 did not also associate with Cav1. Colocalization of TcdA-546 and PACSIN2 also occurred in Caco-2 cells, though at a somewhat later time point (Figure 4). Use of shRNA to knockdown expression of PACSIN2 in these cells did not affect binding of TcdA to the cell surface, but it did inhibit uptake of the toxin.
The investigators concluded that TcdA enters cells by a novel mechanism that does not involve caveolin or clathrin, but does require dynamin and PACSIN2. Although these findings still leave the exact mechanism of TcdA-mediated cell entry in question, they provide considerable new insight into the process, insight that may be exploited for new approaches to treat C. difficile infections in the future.
FIGURE 3. Colocalization of TcdA-546 with PACSIN2. (A) MEFs were exposed to TcdA-546, and PACSIN2 was detected by immunofluorescence. Colocalization of TcdA-546 (green) and PACSIN2 (red) is indicated by the yellow color in the merged image. (B) Same as in (A) except that Cav1 (blue) was also detected by immunorluorescence, and the pink color indicates colocalizatin of Cav1 with PACSIN2. Figure reproduced under the Creative Commons CCO 1.0 Universal License from R. Chandrasekaran, et al., PLoS Pathog., (2016) 12, e1006070.
FIGURE 4. Colocalization of TcdA-546 with PACSIN2. Experimental conditions are the same as described for Figure 2. Colocalization of TcdA-546 and PACSIN2 was noted at the 10 min time point. Figure reproduced under the Creative Commons CCO 1.0 Universal License from R. Chandrasekaran, et al., PLoS Pathog., (2016) 12, e1006070.
View PLOS Pathogens article: Clostridium difficile Toxin A Undergoes Clathrin-Independent, PACSIN2-Dependent Endocytosis