Comparative protein engineering redirects the specificity of Clostridium botulinum proteases.

AI-redesigned starting points and outcomes enhance protein evolution

Engineered or laboratory-evolved proteins often have suboptimal stability, activity, or specificity. We applied AI-based protein sequence design to address challenges in experimental enzyme evolution. Using the model ProteinMPNN, we redesigned three distinct botulinum neurotoxin (BoNT) proteases, generating variants with improved stability and full catalytic efficiency1. We hypothesized that redesigned enzymes may be more mutationally robust than their wild-type counterparts, and therefore may serve as better starting points to evolve new function. We performed side-by-side phage-assisted continuous evolution (PACE) campaigns initiated with AI-redesigned proteases or with the corresponding wild-type proteases2. Evolving three distinct redesigned enzymes as starting points always yielded proteases with higher activity than evolving wild-type proteases in the same selection. Across four evolution campaigns, redesign conferred robustness that unlocked access to otherwise inaccessible highly functional sequences, confirmed by the inability of redesign-evolved mutations to function in wild-type enzyme backgrounds. When redesign raises fitness in sequence space local to the starting point, redesigned starting points adapt at a faster rate. Finally, we evolved both wild-type and AI-redesigned BoNT/E protease to selectively cleave the therapeutically relevant protein ataxin-2. Proteases evolved from the redesigned starting point reached higher catalytic efficiency and stability while minimizing native substrate cleavage, achieving >79-fold greater selected specificity for ataxin-2 than the best-performing variant evolved from wild-type BoNT/E. This study establishes a practical workflow using AI-redesigned starting points to evolve enzymes with improved properties over those evolved from natural proteins, with broad implications for protein science.

A modular platform for scalable recombinant production of highly toxic bacterial proteins

Bacterial protein toxins constitute a vast and largely untapped reservoir of antimicrobial activities with substantial therapeutic and biotechnological potential. However, their intrinsic toxicity frequently prevents stable recombinant expression in bacterial hosts, creating a major bottleneck for biochemical characterization, structural analysis, and development as antimicrobial agents. Here, we present a modular platform for the scalable recombinant production of highly toxic bacterial proteins based on transient intramolecular toxin neutralization. The strategy covalently links each toxin to its cognate immunity protein, promoting neutralization during biosynthesis while permitting recovery of the native toxin through site-specific proteolytic cleavage. Using this approach, we produced multiple previously intractable polymorphic toxin domains that could not be obtained using conventional inducible expression, toxin–immunity co-expression, or bacterial cell-free systems. We further streamlined the production workflow through intracellular protease-mediated cleavage, reducing the purification process from four steps to two and increasing protein recovery. To address cases in which native immunity proteins were insufficient, we incorporated computational protein design to engineer improved toxin-binding partners, enabling production of an additional toxin that remained refractory to the original platform. Purified toxins retained enzymatic activity following denaturation and refolding, confirming recovery of functional proteins and enabling identification of a previously uncharacterized nuclease activity. Together, these findings establish a scalable and adaptable microbial biotechnology platform for the production of intrinsically toxic proteins. The integration of transient intramolecular neutralization with computational engineering provides a route toward systematic production and characterization of toxic proteins for antimicrobial discovery, structural biology, protein engineering, and future biotechnological applications.

PROTEASE AND ITS MODULATION BY SMALL MOLECULES”

DNA Origami-Based Multivalent Nanobody Display Platform for Potent Neutralization of Botulinum Neurotoxin Type A.

Botulinum neurotoxins (BoNTs) are the most potent biological toxins discovered to date. Rapid capture and clearance of BoNTs from the bloodstream is a critical strategy for poisoning treatment. Nanobodies, with their high affinity, represent promising molecules for neutralizing BoNTs. Due to its small molecular weight, it has the limitation of a short half-life in the bloodstream. The BoNT/A-neutralizing nanobody, ciA-C2, binds to the receptor-binding domain of BoNT/A, blocking its interaction with the neuronal receptor SV2 and thereby preventing toxin entry into cells to achieve detoxification. This study leverages the nanoscale addressability of DNA origami technology. Through complementary base pairing, the BoNT/A-neutralizing nanobody ciA-C2 is displayed on a DNA origami nanostructure (DON), constructing a DON-ciA-C2 anti-BoNT/A system with prolonged half-life and spatially cooperative multivalency. The DON, measuring approximately 120 nm in size-about 30 times larger than ciA-C2-extends the half-life of the system due to its larger dimensions. The multivalent assembly of ciA-C2 on the DNA origami creates a "high-affinity" nanoscale surface. This design leverages dense regional proximity effects to significantly enhance toxin-binding efficiency and stability. Additionally, a physical barrier effect is formed, effectively shielding the receptor-binding domain of BoNT/A and preventing the toxin from accessing receptors on the cell membrane. In a mouse model of BoNT/A poisoning at 10 times the lethal dose (10 LD50), DON-ciA-C2 significantly improved survival rates to 100% in both prophylactic and therapeutic administration regimens. Moreover, it extended the effective half-life of ciA-C2 from 4 h to approximately 32 h. By orderly assembling nanobodies on DNA origami, this study creates a biomimetic nanoscale trap with triple functions-"capture, enrichment, and sequestration"-against BoNT/A. This strategy not only provides a highly efficient neutralization paradigm for anti-BoNT/A therapy but also highlights the transformative potential of DNA origami nanostructures in reshaping protein interaction interfaces for precision medicine.

Insights into the enzyme kinetics, structural features, and in silico inhibition of Leishmania donovani prolidase.

Overall, this study presents the first comprehensive report on the enzyme kinetics, structural characteristics, and in silico inhibition of metal-dependent prolidase from trypanosomatid parasites.

Structural and Functional Interrogation of Active Streptococcus pneumoniae Sortase A

Sortase A (SrtA) enzymes covalently anchor surface proteins to Gram-positive bacterial cell walls, promoting colonization and virulence. In Streptococcus pneumoniae, previous studies identified both a domain-swapped dimer and an active refolded monomer, but the active enzyme has not been characterized at the structural and residue-specific level. Here, we performed quantitative proteomic comparisons of wild-type and SrtA knockout strains that confirmed the loss of multiple LPxTG-containing virulence factors, including ZmpB, NanA, and IgA1 protease, consistent with an essential role for SrtA in surface protein anchoring. To enable mechanistic studies, we established a biochemical framework to produce monomeric Streptococcus pneumoniae SrtA by refolding and developed a gel-based assay using recombinant substrates to monitor catalytic activity. The refolded monomer, but not the swapped dimer, catalyzed cleavage and transpeptidation of a canonical LPxTG substrate in a metal-independent manner under the conditions examined. We further report high-resolution NMR backbone assignments for the active monomer and identify substrate-induced chemical shift perturbations that localize to the active site. Together, these findings provide an integrated proteomic, biochemical, and NMR characterization of monomeric, catalytically active Streptococcus pneumoniae SrtA and reveal residue-specific interactions with a canonical LPNTG recognition peptide.