Excipient substitution in botulinum toxin type A formulations: Insights from multi-scale in silico modelling

Botulinum neurotoxin type A (BoNT/A) formulations differ mainly in excipients, which determine stability, aggregation, and diffusion. Using multi-scale in silico modelling (AesthetiSIM™), we evaluated excipient substitution across six marketed products. The 150 kDa neurotoxin backbone showed lower RMSD (3.6 ± 0.4 Å) than the 900 kDa complex (4.8 ± 0.6 Å, p = 0.019). Among excipients, trehalose formed 23.1 ± 2.7 hydrogen bonds and -106 ± 6 kcal/mol interaction energy, outperforming sucrose (19.8 ± 2.5 bonds, -93 ± 5 kcal/mol) and lactose (9.6 ± 1.8 bonds, -74 ± 5 kcal/mol). Trehalose combined with HSA yielded synergistic stabilization with 30.1 ± 3.2 bonds, interaction energy of -112 ± 7 kcal/mol, and reduced RMSD to 3.1 ± 0.3 Å. Coarse-grained simulations showed the lowest aggregation with trehalose-HSA (mean cluster size 2.4 ± 0.4 molecules), compared with sucrose (3.7 ± 0.7) and lactose (5.4 ± 1.0, p < 0.001). Finite element modelling predicted diffusion radii of 1.25 ± 0.05 mm for trehalose-HSA versus 1.36 ± 0.06 mm with sucrose (p = 0.014). Reformulated PRABO and LETYBO with trehalose-HSA achieved the longest durations (186 ± 8 and 184 ± 9 days), followed by INCO (176 ± 7), ONA (170 ± 8), DAXI (167 ± 10), and ABO (160 ± 7). Reconstitution with 0.5 mL per 100 units minimized spread (1.22 ± 0.04 mm) and maximized persistence (188 ± 9 days, p < 0.01 vs 2.5 mL). Across molecular, mesoscopic, and tissue levels, trehalose-HSA consistently provided superior stability, aggregation resistance, and duration, offering a rational basis for next-generation BoNT/A formulations.