Nanobodies against Clostridioides difficile CDTb provide a toolkit for potent toxin neutralization and highly sensitive quantitation

Abstract

Clostridioides difficile is a pathogenic bacterium and a leading cause of antibiotic-associated diarrhea. Symptoms of the infection arise because of the production of large clostridial toxins that disrupt the intestinal barrier and cause an acute host inflammatory response. Epidemic C. difficile strains also produce the C. difficile transferase toxin (CDT), a binary toxin consisting of separate enzymatically active (CDTa) and cell-binding (CDTb) components. However, the role of CDT during C. difficile pathogenesis remains poorly understood. We created a CDTb nanobody (Nb) clone library and identified and purified five clones with promising CDTb-binding properties. Studies using the Carterra LSA^XT platform revealed high-affinity binding interactions between the Nbs and three distinct CDTb epitopes. Functionally, these Nbs potently neutralize cellular cytopathic effects of CDT at equimolar concentrations in vitro. We further identified two distinct neutralization mechanisms-inhibition of CDTb heptamer formation and inhibition of cell surface binding, both of which are crucial for CDTa delivery into the host cell. These Nbs were used in a sandwich ELISA assay to monitor CDTb levels between 1- and 7-day post R20291 infection in the cecal material of infected mice. Notably, levels of CDTb spiked during days 3 and 4, with monomers constituting the majority of CDTb. We anticipate that these reagents will allow researchers to further expand toxin intervention and monitoring strategies to obtain a deeper understanding of the CDT mechanism of action.

Keywords: C. difficile; epitope characterization; molecular mechanism; nanobody; protein–protein interactions; toxin.

Figure 1

The initial characterization of 94 lysates provided evidence of diverse sequences capable of recognizing three or more distinct epitopes on CDTb.A, color-coded schematic of CDTb domain organization and transition from the monomer (1) to prepore (2) to CDTa-bound pore (3) (PDB IDs: 6O2N and 6V1S). Pseudo-ADP ribosyltransferase (Padprt) and ADP ribosyltransferase (ADPRT) of CDTa are shown in light blue and yellow, respectively (PDB ID: 2WN6). B, after derivation of nanobody (Nb) clones, CDTb was expressed as three distinct truncation proteins for screening: D1–D3 (blue, residues 43–615), D3’ (red, residues 616–750), and D4 (green, residues 760–876). C, protein lysates of the 94-clone panel were tested for binding to CDTb_D1–D3 (blue bars), CDTb_D3’ (red bars), and CDTb_D4 (green bars) by ELISA. The threshold level for each protein construct was defined as the average value of the negative controls (clones 1 and 2) multiplied by two and shown as horizontal lines in respective colors on the Y-axis. Clones with values above the thresholds were colored on the cladogram. The experiment was independently performed one time with one technical replicate. D, a cladogram representing the amino acid sequence alignments of the 63 Nb sequences obtained. Clones with identical sequences (eight groups) are highlighted in yellow and circled in red. The five Nbs that were further characterized in this study are shown in bold font. “C1” in front of each Nb correlates to the internal plate numbering, and for conciseness purposes, the “C1” plate identifier will be omitted throughout the rest of the article. Domain specificity was assigned based on ELISA results shown in C (D1–D3, blue; D3′, red; and D4, green). CDTa, Clostridioides difficile transferase toxin, A subunit; CDTb, Clostridioides difficile transferase toxin, B subunit; PDB, Protein Data Bank.

Figure 2

Binding kinetics between CDTb constructs and nanobodies (Nbs) reveal tight protein interactions. Representative surface plasmon resonance (SPR) sensorgrams of (A) CDTb_D1–D4, (B) CDTb_D2–D3’, (C) CDTb_D3’, and (D) CDTb_D4 injected over immobilized A9, C6, G2, H8, and H12 Nbs. Data shown come from different ligand (Nb) densities: higher ligand density regions of interest (ROIs) are shown for low molecular weight (MW) constructs like CDTb_D3’ and CDTb_D4, whereas low ligand density ROIs are shown for CDTb_D1–D4 and CDTb_D2–D3’. Analyte injections are shown in a green–blue palette, with the lowest analyte concentration in green and the highest in blue. Fits to one-to-one kinetics model are shown in black. Noninteracting pairs are highlighted in gray. Summary of interactions is shown in Table 1, Table 2, Table 3, Table 4. The experiment was independently performed one time with three to four technical replicates. CDTb, Clostridioides difficile transferase toxin, B subunit; ROI, region of interest.

Figure 3

Epitope binning of nanobodies (Nbs) confirmed three distinct CDTb epitopes that can be utilized in quantitative ELISA.A, schematic of the epitope binning assay. The figure was created with BioRender.com using a premium institutional license. B, heat map showing blocking relationships of analyte–ligand Nb pairs. Brown squares represent competition between the corresponding analyte and ligand, and light blue squares indicate “sandwiching” or no competition. The competition threshold was set to 0.05 to 0.1. Shaded brown squares along the diagonal represent self–self interactions. These were confirmed to be blocking in both orientations. C, network plot showing that Nbs were grouped into three distinct epitope bins—bin 1 (red—A9 and G2), bin 2 (green—C6), and bin 3 (blue, H8 and H12). Nbs are represented as nodes, blocking relationships are represented as chords, and epitope bins are represented as envelopes. D, CDTb-binding Nb pairs (capture/detection) used to generate quantitative ELISA assays. Five Nb pairs show good agreement with the CDTb limit of detection (LOD) of 10 pM (shown for each pair in color-matched lines at the bottom of the Y-axis). G2–C6 pair (in bold) was used for downstream analyses. E, standard curves in uninfected mouse cecal and fecal materials diluted to 10 mg/ml in PBS-T–BSA with spiked in titration of purified CDTb. Neither background changed the LOD compared with the mock condition (PBS-T–BSA only). In both (D) and (E), standard curves were constructed by interpolating the data using a sigmoidal four-parameter logistic (4-PL) curve fit. Data points are connected by solid lines, and color-matched curve fits are shown in dashed lines. Each data point represents the mean ± SD of three independent biological experiments (n = 3). Each biological experiment consisted of two technical replicates, and the average value of both technical replicates was used. F, CDTb was quantified in cecal material (diluted to 10 mg/ml in PBS-T–BSA) of euthanized mice during the indicated timepoints of R20291 or R20291 ΔcdtB infections. Bars represent mean ± SEM of the group; dots represent an individual mouse within the group. Number of animals: day 1—n = 6 (R20291), n = 6 (R20291 ΔcdtB); day 2—n = 7 (R20291), n = 7 (R20291 ΔcdtB); day 3—n = 6 (R20291), n = 6 (R20291 ΔcdtB); day 4—n = 5 (R20291), n = 5 (R20291 ΔcdtB); and day 7—n = 7 (R20291), n = 5 (R20291 ΔcdtB). In vivo experiments were independently performed two times (with three to four animals per group). BSA, bovine serum albumin; CDTb, Clostridioides difficile transferase toxin, B subunit; PBS-T, PBS with Tween-20.

Figure 4

High neutralization potency of anti-CDTb nanobodies (Nbs).A, schematic of the neutralization assay in Vero-GFP cells. The figure was created with BioRender.com using a premium institutional license. The concentrations of purified CDTb and CDTa were kept constant (7 nM of activated monomer and 1 nM, respectively), and Nbs were titrated according to the graphs. Cell rounding curves depict the protective effects of (B) A9, (C) C6, (D) G2, (E) H8, and (F) H12 Nbs. G, neutralization data summarized by EC₅₀ values obtained via a nonlinear fit to the (agonist) versus response variable slope (four parameters) model (GraphPad Prism). Each data point represents the mean ± SD of three independent biological experiments at a 3-h postintoxication timepoint (n = 3). Each biological experiment consisted of two technical replicates, and the average value of both technical replicates was used. For H12 (F), the analysis did not deduce EC₅₀ value (most likely because of a sharp change in percent rounding between the two reported concentrations). Thus, an approximate value of 6.0 nM was manually assigned based on the curve fit. CDTa, Clostridioides difficile transferase toxin, A subunit; CDTb, Clostridioides difficile transferase toxin, B subunit.

Figure 5

Functional in vitro assays support predicted mechanisms of neutralization.A, oligomerization assay showing that C6 prevents whereas A9 and G2 promote CDTb oligomerization in vitro. Representative overlayed size-exclusion chromatography (SEC) profiles of full-length (pro) CDTb trypsinized with or without nanobodies (Nbs). The red box shows a double heptamer peak, which is the readout of CDTb oligomerization. About 1.5 mg of full-length (pro) CDTb at 7.9 mg/ml final concentration was trypsinized for each run. Each nanobody was added in 5X molar excess relative to CDTb. B, quantification of the area under the curve of the double heptamer peak. Bars represent mean ± SD of the group; dots represent an individual independent biological experiment. One-way ANOVA with Holm–Šídák’s multiple comparisons test was used to calculate statistical significance. The experiment was independently performed two times (n = 2). Each biological experiment consisted of one technical replicate. Quantification of Vero cells bound by (C) trypsinized CDTb-Alexa 647 or (D) full-length (pro) CDTb-Alexa 647 in the presence of Nbs using flow cytometry. Mean fluorescence intensity (MFI) of Alexa-647⁺ live cells is reported relative to CDTb treatment without the Nbs (treated as 100). About 25 nM of CDTb preincubated with 100 nM of Nbs (4X excess) were applied to the cells. Bars represent mean ± SD of the group; dots represent an individual independent biological experiment. One-way ANOVA with Holm–Šídák’s multiple comparisons test was used to calculate statistical significance. For trypsinized CDTb, the experiment was independently performed four times (n = 4). For full-length (pro) CDTb, the experiment was independently performed five times (n = 5). Each biological experiment consisted of one technical replicate. CDTb, Clostridioides difficile transferase toxin, B subunit.