Engineered botulinum neurotoxin B with improved efficacy for targeting human receptors

Abstract

Botulinum neurotoxin B is a Food and Drug Administration-approved therapeutic toxin. However, it has lower binding affinity toward the human version of its major receptor, synaptotagmin II (h-Syt II), compared to mouse Syt II, because of a residue difference. Increasing the binding affinity to h-Syt II may improve botulinum neurotoxin B's therapeutic efficacy and reduce adverse effects. Here we utilized the bacterial adenylate cyclase two-hybrid method and carried out a saturation mutagenesis screen in the Syt II-binding pocket of botulinum neurotoxin B. The screen identifies E1191 as a key residue: replacing it with M/C/V/Q enhances botulinum neurotoxin B binding to human synaptotagmin II. Adding S1199Y/W or W1178Q as a secondary mutation further increases binding affinity. Mutant botulinum neurotoxin B containing E1191M/S1199Y exhibits ~11-fold higher efficacy in blocking neurotransmission than wild-type botulinum neurotoxin B in neurons expressing human synaptotagmin II, demonstrating that enhancing receptor binding increases the overall efficacy at functional levels. The engineered botulinum neurotoxin B provides a platform to develop therapeutic toxins with improved efficacy.Humans are less sensitive to the therapeutic effects of botulinum neurotoxin B (BoNT/B) than the animal models it is tested on due to differences between the human and the mouse receptors. Here, the authors engineer BoNT/B to improve its affinity to human receptors and enhance its therapeutic efficacy.

Fig. 1

BACTH screen identifies mutations that enhance BoNT/B binding to h-Syt II. a Sequence alignment of the BoNT/B-binding site in mouse Syt I (m-Syt I), h-Syt I, h-Syt II, and mouse Syt II (m-Syt II), and a close-up view of the interface between H_CB and Syt II. Targeted residues in BoNT/B for mutagenesis studies are labeled in green. Syt II residues are marked in purple, with F54, F55, and I58 being highlighted. b A schematic illustration for construction of H_CB mutant libraries and the BACTH assay. c H_CB mutants identified in the BACTH assay were further analyzed by β-galactosidase activity assay, which reflects the levels of reconstituted adenylate cyclase (Student’s t-test, n = 6, *P < 0.05). Error bars represent SEM. d Point mutations at E1191, W1178, and S1199 were examined in pull-down assays, using immobilized m-Syt II (residues 1–87) or a mouse Syt II (1–87) containing F54L mutation that mimics h-Syt II sequence (designated as h-Syt II). Bound H_CB variants were detected by immunoblot analysis detecting the HA-tag fused to H_CB. Additional H_CB mutations at other sites are shown in Supplementary Fig. 2. One of two independent experiments is shown

Fig. 2

Combinational mutations in H_CB enhance its binding to h-Syt II and Syt I. a H_CB double mutations were examined in pull-down assays for their abilities to bind m-Syt II vs. h-Syt II. Three double mutations that showed robust binding to h-Syt II are marked in red. One of two independent experiments is shown. b Binding of H_CB to h-Syt II was quantified by the BLI assay. Representative association and dissociation curves are shown for WT H_CB and three H_CB mutants: E1191M/S1199Y, E1191V/S1199W, and E1191C/S1199W. Binding parameters for all 12 double mutations are listed in Table 1, and their representative binding traces are shown in Supplementary Fig. 3. c Binding of WT H_CB, H_CB (E1191M), and H_CB (E1191M/S1199Y) to immobilized GST-tagged h-Syt I (residues 1–80) was examined in pull-down assays, with (+) or without (−) gangliosides (Gangl.). Binding of WT H_CB to h-Syt I requires the presence of gangliosides, while H_CB (E1191M) and H_CB (E1191M/S1199Y) bind to h-Syt I in the absence of gangliosides. One of two independent experiments is shown. Further analysis of H_CB binding to h-Syt I by the BLI assay is shown in Supplementary Fig. 4 and Table 1

Fig. 3

H_CB_MY shows robust binding to neurons that express h-Syt II. a Humanized neurons were created by KD of endogenous Syt I and by expressing full-length h-Syt II in cultured rat cortical neurons via lentiviral transduction. KD efficiency was validated by immunoblot (Supplementary Fig. 5). Neurons that express full-length m-Syt II or m-Syt II (F54L) served as controls. These neurons were exposed to WT H_CB (100 nM) followed by immunostaining analysis. Bound H_CB was detected via an HA-tag fused to H_CB. Synapsin was labeled as a marker for presynaptic terminals. WT H_CB bound to WT neurons but not to Syt I KD neurons. Expression of full-length m-Syt II restored binding of H_CB. Expression of full-length m-Syt II (F54L) or full-length h-Syt II resulted in only low levels of binding of WT H_CB compared to m-Syt II. b Syt I KD abolished binding of H_CB_MY (100 nM) to neurons. The binding was rescued by expression of full-length m-Syt II, m-Syt II (F54L), or h-Syt II. Scale bars represent 20 µm. One of two independent experiments is shown

Fig. 4

BoNT/B_MY shows enhanced functional efficacy in neurons that express h-Syt II. a Humanized neurons were created as described in Fig. 3a. Neurons were exposed to a gradient of full-length WT BoNT/B or BoNT/B_MY for 24 h. Cell lysates were harvested and subjected to immunoblot analysis. β-tubulin served as an internal loading control. BoNT/B_MY cleaved more VAMP2 than WT BoNT/B at all concentrations tested, indicating that BoNT/B_MY entered neurons more efficiently than WT BoNT/B. One of three independent experiments is shown. b,c Humanized neurons were exposed to a gradient of WT BoNT/B or BoNT/B_MY for 24 h. The mIPSC was recorded by a whole-cell patch-clamp approach. b Representative mIPSC recordings at 30 pM toxins. c The mIPSC activities vs. toxin concentrations, normalized to neurons that were not exposed to toxins. The numbers of recorded neurons for each data point are noted within parentheses. The same number of neurons was recorded for WT BoNT/B and BoNT/B_MY. The half maximum inhibitory concentration (IC₅₀) was determined to be 89 pM for WT BoNT/B and 7.8 pM for BoNT/B_MY, demonstrating that enhancing binding to receptors increased functional efficacy of toxins in neurons. Statistical analysis was performed with Student’s t-test (*P < 0.01). All data shown are means ± SEM

Fig. 5

Variations in BoNT/B subtypes influence their binding to h-Syt II and h-Syt I. a Sequence alignment of BoNT/B subtypes in the region around residues E1191/S1199. b H_CB4 showed robust binding to m-Syt II, but no detectable binding to h-Syt II without gangliosides, which is similar to H_CB. Both showed low levels of binding to h-Syt II in the presence of gangliosides. c There are four residues that differ between H_CB4 and H_CB within the 19 key residues in the Syt-binding pocket besides E1191Q/S1199Y. Replacing all four residues in H_CB4 with the corresponding residues in H_CB created a mutant H_CB4 (V1113K/S1117P/S1196A/I1197P) that can bind to h-Syt II. The sequence alignment between H_CB and H_CB4 is shown in Supplementary Fig. 7. d Binding of H_CB and H_CB4 to immobilized GST-tagged h-Syt I was examined in pull-down assays. Binding of H_CB to h-Syt I requires the presence of gangliosides, while H_CB4 is capable of binding to h-Syt I without gangliosides in pull-down assays