Identification of the C. sordellii lethal toxin receptor elucidates principles of receptor specificity in clostridial toxins

Clostridium sordellii lethal toxin (TcsL) is responsible for an almost invariably lethal toxic shock syndrome associated with gynecological C. sordellii infections. Here, using CRISPR/Cas9 screening, we identify semaphorins SEMA6A and SEMA6B as the cellular receptors for TcsL and demonstrate that soluble extracellular SEMA6A can protect mice from TcsL-induced edema. A 3.3 Å cryo-EM structure shows that TcsL binds SEMA6A with the same region that the highly related C. difficile TcdB toxin uses to bind structurally unrelated Frizzled receptors. Remarkably, reciprocal mutations in this evolutionarily divergent surface are sufficient to switch receptor specificity between the toxins. Our findings establish semaphorins as physiologically relevant receptors for TcsL, and reveal the molecular basis for the difference in tissue targeting and disease pathogenesis between highly related toxins.

glucose, the sugar donor for the glucosylation activity of all LCTs (6). The other was SEMA6A, encoding a transmembrane axon guidance molecule not previously linked to toxin function.
We then expressed and purified the soluble recombinant extracellular domain (rECD) of SEMA6A and tested its effect on TcsL toxicity on Vero cells, a commonly used cell line for studying toxin function. SEMA6A rECD counteracted TcsL toxicity in a dose-dependent manner (Fig. 1D).
Notably, SEMA6A rECD had a protective effect only when it was added to the cells before or simultaneously with TcsL ( fig. S1G). When cells were pretreated for one hour with TcsL, SEMA6A rECD had no effect on toxicity, suggesting that SEMA6A rECD must act before TcsL binds the cell membrane. We also repeated the competition assay with soluble ectodomains of SEMA6B, SEMA6C, and SEMA6D. Consistent with previous experiments, SEMA6B alleviated TcsL toxicity whereas SEMA6C and SEMA6D had no effect ( Fig. 1E and Fig. 1F). Together, these results strongly suggest that TcsL binds SEMA6A and SEMA6B on the cell surface and this interaction is required for TcsL entry into the cell.
One of the primary targets of TcsL during C. sordellii infection is the vascular endothelium of the lung (15). Therefore, we used immortalized human lung microvascular cells (HULECs) as a physiologically relevant cell line to study the role of SEMA6A and SEMA6B in TcsL intoxication.
HULECs express ~4-fold higher levels of SEMA6A and ~11-fold higher levels of SEMA6B than Hap1 cells ( Fig. 2A). We assayed the sensitivity of HULECs to four related clostridial toxins: TcdA and TcdB from C. difficile, TpeL from C. perfringens, and TcsL (Fig. 2B). TpeL showed low toxicity and TcdA and TcdB were several orders of magnitude less toxic to HULECs than what is reported for other cell types (16,17). Remarkably, the GI50 of TcsL was ~50 fM, suggesting that only ~200 toxin molecules/cell are lethal to HULECs. Furthermore, recombinant SEMA6A ectodomain but not SEMA6C ectodomain could block TcsL-induced rounding of HULECs in a dose-dependent manner ( Fig. 2C and Fig. 2D). These results strongly suggest that SEMA6A/SEMA6B are the physiologically relevant TcsL receptors in endothelial cells.
We then addressed the role of semaphorins in a mouse model of TcsL intoxication. We first examined the expression of SEMA6A and SEMA6B in mouse lung tissue by immunohistochemistry. Both proteins were highly expressed in lung endothelium and pneumocytes (Fig. 2E). We injected mice intraperitoneally with a lethal dose of TcsL together with mouse Sema6a ectodomain-Fc fusion, mouse Sema6c ectodomain-Fc fusion, or BSA (n = 3 for each group). Mice co-injected with BSA or Sema6c-Fc rapidly developed symptoms of TcsL intoxication, including decreased mobility and signs of ataxia and dehydration. Within four hours, all symptomatic mice had a buildup of fluid in the lungs (Fig. 2F). Histopathologically, symptomatic mice showed edema surrounding pulmonary vessels (Fig. 2G). In contrast, SEMA6A-Fc protected the mice from TcsL-induced symptoms. Mice co-injected with SEMA6A-Fc had no pleural effusion and did not show signs of edema after four hours ( Fig. 2E and 2F).
Taken together, these data support that SEMA6A, likely acting together with SEMA6B, is the physiologically relevant receptor for TcsL in vivo. We next structurally characterized the interaction between TcsL and SEMA6A by cryo-EM using a shortened TcsL1285-1804 construct, which bound recombinant SEMA6A ectodomain with nanomolar apparent affinity ( Fig. 3A and fig. S2A). Glutaraldehyde cross-linking was used to prevent dissociation of the TcsL1285-1804-SEMA6A complex during cryo-EM grid preparation, as described previously (18)(19)(20). Initial analysis of the dataset revealed that approximately 50% of the SEMA6A dimers were bound to TcsL using this sample preparation approach. Consequently, 2D and 3D classification resulted in two homogenous datasets corresponding to the TcsL-    Figure 3 The architecture of TcsL is similar to previously-determined structures of other clostridial toxins (8,(22)(23)(24) (25,26). Strikingly, the site on TcsL mediating the SEMA6A interaction is in the same location as the previously reported TcdB-FZD2 binding interface (8) ( Fig. 3B and fig. S8). In fact, the TcsL hydrophobic pocket burying M109 of SEMA6A is analogous to the TcdB hydrophobic pocket burying the FZD2 palmitoleic acid moiety (Fig. 3B).
Thus, TcsL and TcdB have evolved to bind different host receptors through the same interaction surface. This surface is highly diverged between the two toxins: only three of the 23 SEMA6A-interacting residues in TcsL are conserved in TcdB, and conversely, four of the 22 residues in TcdB that interact with FZD2 are conserved in TcsL. In stark contrast, 76% of non-interacting surface residues in TcdB are conserved in TcsL (p < 0.0001, Fisher's exact test), indicating that the divergence is specific to the receptor-binding surface. This surface may represent a more general receptor-binding site in clostridial toxins, as the divergence extends to all members of the family ( fig. S9). In particular, the region surrounding the hydrophobic pocket for SEMA6A M109 in TcsL and for the palmitoleic acid moiety in the FZD2/TcdB complex is highly variable between LCTs ( Fig. 4A and fig. S9).
We directly tested the role of the evolutionarily divergent surface in receptor specificity with recombinant proteins and size exclusion chromatography. As expected, TcsL1285-1804 interacted with SEMA6A, and TcdB1285-1804 formed a stable complex with FZD7 (Fig. 4B). However, TcsL did not interact with FZD7, consistent with the extensive divergence of the receptor-binding interface (Fig. 4B). We then generated a TcsL(FBD)1285-1804 variant that had a TcdB-like interface by changing 15 TcsL residues in the receptor-binding interface to those in TcdB ( Fig. 4C and fig.   S8 and S9). Remarkably, these mutations were sufficient to switch TcsL binding specificity: the TcsL1285-1804(FBD) hybrid protein robustly interacted with FZD7 but no longer bound SEMA6A (Fig. 4D).
These results establish that C. sordellii TcsL and C. difficile TcdB, two highly similar virulence factors, use the same surface to interact with distinct cognate receptors. The striking evolutionary divergence of this surface among all clostridial toxins suggests that it provides a common, malleable interface that can be modulated to bind structurally unrelated host receptors, allowing the pathogens to adapt to and attack novel host organisms and tissues. Moreover, extending the similarities between toxins, both TcdB and TcsL are likely to interfere with the host receptor function. TcdB, which binds Frizzled, can inhibit Wnt signaling in colonic organoids (7).
SEMA6A binds TcsL and its endogenous ligand Plexin A2 with overlapping interfaces, suggesting that TcsL could interfere with the semaphorin-plexin signaling axis in vascular endothelium.
Further work will address the intriguing possibility that TcsL and other LCTs use their cognate receptor both for gaining entry into the cell and for modulating host cell responses to their advantage.
We cannot exclude the possibility that, like other clostridial toxins (7,(10)(11)(12), TcsL also binds additional receptors on host cells. However, we did not identify additional factors in CRISPR/Cas9 screens in cells that do not express SEMA6A or SEMA6B. Moreover, if they exist, these receptors must be distinct from those of other LCTs. Our results with physiologically relevant models indicate that SEMA6A and SEMA6B are the major TcsL receptors in vivo. Consequently, targeting this interaction with antibodies or small molecules could provide a highly needed therapeutic intervention against rapidly lethal and currently untreatable C. sordellii infections.  TcsL/SEMA6A complex will be deposited to PDB and EMDB upon acceptance.

Materials and Methods
Table S1-S4 Movie S1

Cell lines
Vero

Next-generation sequencing library preparation
The gRNA sequences were PCR amplified from the extracted genomic DNA. Each amplified sample was then barcoded and processed on Illumina Next-seq high-output mode at a read depth of at least 5 million reads per sample. MAGeCK software (27) was used to generate rankings for positively enriched genes.

CRISPR screen validation
The gRNAs targeting SEMA6A, SEMA6B, SEMA6C, SEMA6D and UGP2 were chosen from TKOv3 library and were cloned into pX459 (Addgene plasmid #62988; gift from Feng Zhang) and lentiCRISPRv2 (Addgene plasmid #52961; gift from Feng Zhang). Plasmids were transfected into HAP1 cells seeded on a 6-well plate with Turbofectin (OriGene) following the manufacturer's protocol. One day post-transfection, the medium was changed to medium containing 2 µg/ml puromycin and was further selected for 3 days. Cells were washed and moved to a 10 cm plate with fresh growth medium with no antibiotics.
Wild-type and knock-out cells were seeded on a 96-well plate a day before toxin application at <40% confluency. Toxins were serially diluted in 1xPBS with 10% glycerol before applying to cells. Cells were incubated with toxins for 24 to 48 hours. Cell viability was measured either using AlamarBlue dye (Invitrogen) or CellTiter-Glo reagent (Promega) following the manufacturer's protocol.

Site-directed mutagenesis
Mutations were generated with Quikchange site-directed mutagenesis (Agilent) following manufacturer's protocol. immediately before TcsL at the indicated concentrations and the assay proceeded as above.  Recombinant FZD7 was expressed and purified as previously described (28). Briefly, a human Subsequently, the complex was concentrated to 0.9 mg/ml, spun down for 30 min at 14,500 x g and directly used for cryo-EM grid preparation.

Cryo-EM data collection and image processing
Homemade holey gold grids (29) were glow discharged in air for 15 s before use. TcsL-SEMA6A (3 µl, 0.9 mg/mL) was applied to grids, blotted for 12 s, and frozen in a mixture of liquid ethane and propane (30)  A total of 4581 raw movies were obtained for the TcsL-SEMA6A sample. Image processing was carried out in cryoSPARC v2 (18). Motion correction was performed with Patch Motion algorithm and CTF parameters were estimated from the average of aligned movie frames with Patch CTF.
4,659,284 particle images were selected by template matching and individual particle images were corrected for beam-induced motion using local motion algorithm (31) within cryoSPARC v2. Ab initio structure determination and classification revealed that ~50% particle images corresponded to a SEMA6A dimer and the remaining particles to a SEMA6 dimer-TcsL complex. The overwhelming majority of particles for the SEMA6A-TcsL complex had one TcsL molecule bound to the SEMA6A dimer, with no evidence of two TcsL molecules bound to the dimer that could be clearly identified for this cross-linked sample. Multiple rounds of heterogeneous refinement followed by non-uniform refinement resulted in a 3.3 Å resolution map of the SEMA6A-TcsL complex (155,353 particle images) and 3.1 Å resolution map of unliganded SEMA6A (229,275 particle images)

Model building
An initial model for TcsL was created using the Phyre2 server (32) with the TcdB crystal structure (PDB: 6C0B) as a reference. The atomic coordinates of SEMA6A dimer (PDB: 3OKW) were manually fitted into the density map using UCSF Chimera (33) to generate a starting model, followed by manual rebuilding using Coot (34). All models were refined using the phenix.real_space_refine (35) with secondary structure and geometry restraints. The final models were evaluated by MolProbity (36). Statistics of the map reconstruction and model refinement are presented in table S2.

Mouse studies
Female            Movie S1. 3D variability analysis of TcsL.
A Figure S1 [