Biologically active peptides have attracted increasing attention in research on the development of new drugs. Mastoparans, a group of wasp venom linear cationic α-helical peptides, have a variety of biological effects, including mast cell degranulation, activation of protein G, and antimicrobial and anticancer activities. However, the potential hemolytic activity of cationic α-helical peptides greatly limits the clinical applications of mastoparans. Here, we systematically and comprehensively studied the hemolytic activity of mastoparans based on our wasp venom mastoparan family peptide library. The results showed that among 55 mastoparans, 18 had strong hemolytic activity (EC₅₀ ≤ 100 μM), 14 had modest hemolytic activity (100 μM < EC₅₀ ≤ 400 μM) and 23 had little hemolytic activity (EC₅₀ > 400 μM), suggesting functional variation in the molecular diversity of mastoparan family peptides from wasp venom. Based on these data, structure–function relationships were further explored, and, hydrophobicity, but not net charge and amphiphilicity, was found to play a critical role in the hemolytic activity of mastoparans. Combining the reported antimicrobial activity with the present hemolytic activity data, we found that four mastoparan peptides, Parapolybia-MP, Mastoparan-like peptide 12b, Dominulin A and Dominulin B, have promise for applications because of their high antimicrobial activity (MIC ≤ 10 μM) and low hemolytic activity (EC₅₀ ≥ 400 μM). Our research not only identified new leads for the antimicrobial application of mastoparans but also provided a large chemical space to support the molecular design and optimization of mastoparan family peptides with low hemolytic activity regardless of net charge or amphiphilicity.

Coagulation factor XIa is a new serine-protease family drug target for next-generation anticoagulants. With the snake venom Kunitz-type peptide BF9 as the scaffold, we obtained a highly active XIa inhibitor BF9-N17K in our previous work, but it also inhibited the hemostatic target plasmin. Here, in order to enhance the selectivity of BF9-N17K toward XIa, four mutants, BF9-N17K-L19A, BF9-N17K-L19S, BF9-N17K-L19D, and BF9-N17K-L19K, were further designed using the P2′ amino acid classification scanning strategy. The anticoagulation assay showed that the four P2′ single-point mutants still had apparent inhibitory anticoagulation activity that selectively inhibited the human intrinsic coagulation pathway and had no influence on the extrinsic coagulation pathway or common coagulation pathway, which indicated that the single-point mutants had minimal effects on the anticoagulation activity of BF9-N17K. Interestingly, the enzyme inhibitor assay experiments showed that the XIa and plasmin inhibitory activities were significantly changed by the P2′ amino acid replacements. The XIa inhibitory activity of BF9-N17K-L19D was apparently enhanced, with an IC₅₀ of 19.28 ± 2.53 nM, and its plasmin inhibitory was significantly weakened, with an IC₅₀ of 459.33 ± 337.40 nM. BF9-N17K-L19K was the opposite to BF9-N17K-L19D, which had enhanced plasmin inhibitory activity and reduced XIa inhibitory activity. For BF9-N17K-L19A and BF9-N17K-L19S, no apparent changes were found in the serine protease inhibitory activity, and they had similar XIa and plasmin inhibitory activities to the template peptide BF9-N17K. These results suggested that the characteristics of the charge of the P2′ site might be associated with the drug selectivity between the anticoagulant target XIa and hemostatic target plasmin. In addition, according to the molecular diversity and sequence conservation, a common motif GR/PCR/KA/SXIP-XYGGC is proposed in the XIa-inhibitory Kunitz-type peptides, which might provide a new clue for further peptide engineering. In conclusion, through P2′ amino acid classification scanning with the snake venom Kunitz-type peptide scaffold, a new potent and selective XIa inhibitor, BF9-N17K-L19D, was discovered, which provides a new XIa-targeting lead drug template for the treatment of thrombotic-related diseases.

Recent studies have revealed that the coagulation system plays a role in mammalian innate defense by entrapping bacteria in clots and generating antibacterial peptides. So, it is very important for the survival of bacteria to defend against the host coagulation system, which suggests that bacterial exotoxins might be a new source of anticoagulants. In this study, we analyzed the genomic sequences of Acinetobacter baumannii and a new bacterial exotoxin protein, F6W77, with five Kunitz-domains, KABP1-5, was identified. Each Kunitz-type domain features a classical six-cysteine framework reticulated by three conserved disulfide bridges, which was obviously similar to animal Kunitz-domain peptides but different from plant Kunitz-domain peptides. Anticoagulation function evaluation showed that towards the intrinsic coagulation pathway, KABP1 and KABP5 had apparently inhibitory activity, KABP4 had weak inhibitory activity, and KBAP2 and KABP3 had no effect even at a high concentration of 20 μg/mL. All five Kunitz-domain peptides, KABP1-5, had no inhibitory activity towards the extrinsic coagulation pathway. Enzyme-inhibitor experiments showed that the high-activity anticoagulant peptide KABP1 had apparently inhibitory activity towards two key coagulation factors, Xa and XIa, which was further confirmed by pull-down experiments that showed that KABP1 can bind to coagulation factors Xa and XIa directly. Structure-function relationship analyses of five Kunitz-type domain peptides showed that the arginine of the P1 site of three new bacterial anticoagulants, KABP1, KABP4 and KABP5, might be the key residue for their anticoagulation activity. In conclusion, with bioinformatics analyses, peptide recombination, and functional evaluation, we firstly found bacterial-exotoxin-derived Kunitz-type serine protease inhibitors with selectively inhibiting activity towards intrinsic coagulation pathways, and highlighted a new interaction between pathogenic bacteria and the human coagulation system.

Kunitz-type peptides from venomous and blood-sucking animals represent an important natural molecular resource for anticoagulant development towards XIa, a new target for antithrombotic drug discovery. However, the plasmin-inhibitory activity of these Kunitz-type XIa inhibitory anticoagulants limits their application. Here, using the snake Kunitz-type toxin analogue BF9-N17H, an XIa inhibitor with a unique P1-site histidine, as a template, P2′ scanning was performed to enhance its selectivity towards XIa. Ten BF9-N17H analogues were designed using different amino acid classification replacement at the P2′ site, including non-polar amino acids A and P, acidic amino acids D and E, polar amino acids S, N, and Y, and basic amino acids K, R, and H. Coagulation factor inhibitory experiments showed that three BF9-N17H analogues, BF9N17HP2, BF9N17HK8 and BF9N17HR9, lost their XIa inhibitory activity, but seven other analogues, BF9N17HA1, BF9N17HD2, BF9N17HS4, BF9N17HN5, BF9N17HY6, BF9N17HE7 and BF9N17HH10, effectively maintained their inhibitory activity towards XIa, with IC₅₀ values from 61.19 ± 7.04 nM to 4.54 ± 3.65 nM. One analogue, BF9N17HD2, inhibited XIa with an IC₅₀ of 6.19 ± 1.38 nM and had a selectivity index of 142.79 towards plasmin. It only had weak inhibitory activity toward plasmin, with an IC₅₀ of 883.88 ± 148.35 nM. Excitingly, five analogues, BF9N17HS4, BF9N17HN5, BF9N17HY6, BF9N17HE7 and BF9N17HH10, almost lost their plasmin-inhibitory activity completely, indicating their high selectivity towards XIa. To the best of our knowledge, this is the first report of Kunitz-type XIa inhibitors without plasmin-inhibitory activity, and it highlights that the Kunitz-type peptide scaffold might be used as a potential candidate for the development of potent and selective XIa-inhibitory antithrombotic agents.

Scorpion toxin peptides are useful bioactive substance resources, but no anticomplement peptides have been reported from scorpions. Here, we attempt to explore scorpion-derived anticomplement peptides by scanning a small and specific scorpion toxin peptide library. Given that complement activation constitutes a protease-driven proteolytic cascade reaction, we focused on two kinds of serine protease inhibitor-like peptides from scorpion Kunitz-type and Ascaris-type toxin subfamilies. Assays of complement-mediated erythrocyte lysis showed that one scorpion peptide SjAPI-2, with an Ascaris-type structural fold inhibited classical complement activation in a concentration-dependent manner. Complement-deficient experiments revealed that the anti-complement effect of SjAPI-2 was markedly enhanced in sera reconstituted with deficient complement proteins C3 or C4, and suggested that C3 and C4 are associated with the anticomplement effect of SjAPI-2. Combined ELISA and Co-IP assays further revealed that SjAPI-2 can directly bind to C4 and C4b. These results indicated that SjAPI-2 can inhibit classical complement activation via binding to C4 and C4b, which blocks the upstream proteolytic cleavage of C4 as well as the downstream cleavage and activation of C3. Alanine-scanning mutagenesis showed that SjAPI-2 suppressed complement activation via its functional loop from K30 to V38. Myocardial ischemia–reperfusion injury experiment showed that SjAPI-2 limited the extent of myocardial infarction and exerted a cardioprotective effect in vivo. To the best of our knowledge, SjAPI-2 is the first anticomplement toxin peptide from a scorpion, suggesting that scorpion venom peptides might be a new resource for anticomplement peptide discovery. The unique ten-cysteine peptide scaffold of SjAPI-2 also provided a novel lead drug template for C4/C4b associated complement diseases.