Probing the structure and function of the protease domain of botulinum neurotoxins using single-domain antibodies

Botulinum neurotoxins (BoNTs) are among the deadliest of bacterial toxins. BoNT serotype A and B in particular pose the most serious threat to humans because of their high potency and persistence. To date, there is no effective treatment for late post-exposure therapy of botulism patients. Here, we aim to develop single-domain variable heavy-chain (VHH) antibodies targeting the protease domains (also known as the light chain, LC) of BoNT/A and BoNT/B as antidotes for post-intoxication treatments. Using a combination of X-ray crystallography and biochemical assays, we investigated the structures and inhibition mechanisms of a dozen unique VHHs that recognize four and three non-overlapping epitopes on the LC of BoNT/A and BoNT/B, respectively. We show that the VHHs that inhibit the LC activity occupy the extended substrate-recognition exosites or the cleavage pocket of LC/A or LC/B and thus block substrate binding. Notably, we identified several VHHs that recognize highly conserved epitopes across BoNT/A or BoNT/B subtypes, suggesting that these VHHs exhibit broad subtype efficacy. Further, we identify two novel conformations of the full-length LC/A, that could aid future development of inhibitors against BoNT/A. Our studies lay the foundation for structure-based engineering of protein- or peptide-based BoNT inhibitors with enhanced potencies and cross-subtypes properties.

Proteolytic activation of the epithelial sodium channel (ENaC) by factor VII activating protease (FSAP) and its relevance for sodium retention in nephrotic mice

Proteolytic activation of the epithelial sodium channel (ENaC) by aberrantly filtered serine proteases is thought to contribute to renal sodium retention in nephrotic syndrome. However, the identity of the responsible proteases remains elusive. This study evaluated factor VII activating protease (FSAP) as a candidate in this context. We analyzed FSAP in the urine of patients with nephrotic syndrome and nephrotic mice and investigated its ability to activate human ENaC expressed in Xenopus laevis oocytes. Moreover, we studied sodium retention in FSAP-deficient mice (Habp2−/−) with experimental nephrotic syndrome induced by doxorubicin. In urine samples from nephrotic humans, high concentrations of FSAP were detected both as zymogen and in its active state. Recombinant serine protease domain of FSAP stimulated ENaC-mediated whole-cell currents in a time- and concentration-dependent manner. Mutating the putative prostasin cleavage site in γ-ENaC (γRKRK178AAAA) prevented channel stimulation by the serine protease domain of FSAP. In a mouse model for nephrotic syndrome, active FSAP was present in nephrotic urine of Habp2+/+ but not of Habp2−/− mice. However, Habp2−/− mice were not protected from sodium retention compared to nephrotic Habp2+/+ mice. Western blot analysis revealed that in nephrotic Habp2−/− mice, proteolytic cleavage of α- and γ-ENaC was similar to that in nephrotic Habp2+/+ animals. In conclusion, active FSAP is excreted in the urine of nephrotic patients and mice and activates ENaC in vitro involving the putative prostasin cleavage site of γ-ENaC. However, endogenous FSAP is not essential for sodium retention in nephrotic mice.

Cryo-EM structure of hexameric yeast Lon (PIM1) highlights importance of conserved structural elements

Lon protease is a conserved ATP-dependent serine protease composed of an AAA+ domain that mechanically unfolds substrates and a serine protease domain that degrades unfolded substrates. In yeast, dysregulation of Lon protease (PIM1) attenuates lifespan and leads to gross mitochondrial morphologic perturbations. Although structures of bacterial and human Lon protease reveal a hexameric assembly, PIM1 was speculated to form a heptameric assembly, and is uniquely characterized by a $\sim$50 residue insertion between the ATPase and protease domains. To understand the yeast-specific properties of PIM1, we determined a high-resolution cryo-EM structure of PIM1 in a substrate-translocating state. Here, we reveal that PIM1 forms a hexamer, conserved with that of bacterial and human Lon proteases, wherein the ATPase domains form a canonical closed spiral that enables pore loop residues to translocate substrate to the protease chamber. In the substrate-translocating state, PIM1 protease domains form a planar protease chamber in an active conformation and are uniquely characterized by a $\sim$15 residue C-terminal extension. These additional C-terminal residues form an alpha-helix that is located along the base of the protease domain. Finally, we did not observe density for the yeast-specific insertion between the ATPase and protease domains, likely due to high conformational flexibility. Biochemical studies to investigate the insertion using constructs that truncated or replaced the insertion with a glycine-serine linker suggest that the yeast-specific insertion is dispensable for PIM1 enzymatic function. Altogether, our structural and biochemical studies highlight unique components of PIM1 machinery and demonstrate evolutionary conservation of Lon protease function.

Human Prothrombin Variants with Modifications in the Active Site S4 Subpocket Demonstrate Resistance to Dabigatran

Chymotrypsin-like serine proteases are hallmarked by a protease domain comprising the catalytic triad residues His57, Asp102, and Ser195 (chymotrypsinogen numbering) situated in the active site cleft. While the catalytic triad in conjunction with the oxyanion hole residues regulate substrate cleavage, the active site subpockets (S1-4) control substrate recognition and binding. The high structural homology of the serine protease domains allows for analogous strategies in drug design, which is underscored by the direct oral anticoagulants (DOACs) for the prophylactic management of stroke in atrial fibrillation and prevention and treatment of venous thrombosis. DOACs inhibit coagulation serine proteases by reversibly engaging the active site with high affinity. To expand the repertoire of DOAC-specific reversal agents we have previously successfully modified the S4 active site subpocket of human factor Xa to prevent DOAC binding while preserving catalytic activity [Verhoef 2017 Nature Commun.]. To explore whether an analogous strategy can be applied to create DOAC resistance in the serine protease thrombin, specific substitutions or sequences in or around the dabigatran-binding S4 subsite derived from naturally occurring serine proteases or plasma proteins were introduced in prothrombin. A panel of prothrombin variants was generated and transfected into HEK293 cells to allow for stable protein expression. In some of the generated prothrombin variants comprising insertions in amino acid sequence 91-99 that is directly adjacent to the S4 subsite protein expression was severely impaired. This indicates that exchange of any surface-exposed serine protease or plasma protein region into the prothrombin protease domain is not necessarily compatible with protein expression. In contrast, exchange of the human prothrombin 91-99 sequence for that of human kallikrein 3 or targeted amino acid replacement of S4 subsite residue Ile174 resulted in prothrombin protein expression levels similar to wild-type prothrombin. Following expression, prothrombin variants were purified to homogeneity using the CaptureSelect tm affinity matrix that selects for fully gamma-carboxylated prothrombin. The specific prothrombin clotting activity analyses of the purified prothrombin variants KL3 (0.7 ± 0.2 U/mg), I174A (0.8 ± 0.2 U/mg), and I174F (0.8 ± 0.3 U/mg) demonstrated an overall ~10-fold reduced specific activity relative to wild-type prothrombin (7.5 ± 0.1 U/mg). As such, modification of the S4 subsite likely interferes with the binding and subsequent conversion of fibrinogen by thrombin. To determine whether the prothrombin variants supported tissue factor-initiated thrombin formation in human plasma, prothrombin-deficient plasma was supplemented with increasing plasma concentrations of prothrombin variant (90-180 ug/mL). Consistent with their reduced specific clotting activity, 180 ug/mL prothrombin variant was required to obtain substantial thrombin generation but with reduced thrombin generation parameters (peak thrombin, ETP) relative to supplementation with plasma concentrations of wild-type prothrombin (90 ug/mL). This calibrated automated thrombin generation assay set-up was used to assess the DOAC-resistance of the prothrombin variants. While thrombin formation reached half-maximum inhibition at 532 ± 58 nM dabigatran in wild-type prothrombin-supplemented plasma, addition of the prothrombin variants displayed a ~2-fold reduced sensitivity to dabigatran inhibition (IC50: 1186 ± 136 nM prothrombin-KL3; 851 ± 97 nM prothrombin-I174F; 772 ± 80 nM prothrombin-I174A). This demonstrates that the S4 subsite-modified prothrombin variants are able to support thrombin generation in the presence of physiological plasma concentrations of inhibitor. Collectively, our findings indicate that human prothrombin variants comprising a single point mutation at position Ile174 in the S4 subsite or at a region directly adjacent to the S4 subsite are able to generate thrombin in plasma inhibited by dabigatran. Hence, serine proteases with S4 subpocket modifications have the potential to bypass DOAC therapy and could provide a generic strategy in the development of novel DOAC-bypassing agents. Figure 1 Figure 1. Disclosures Reitsma: VarmX. B.V.: Current Employment, Current holder of individual stocks in a privately-held company, Current holder of stock options in a privately-held company, Patents & Royalties. Verhoef: VarmX. B.V.: Current Employment, Current holder of individual stocks in a privately-held company. Bos: VarmX B.V.: Research Funding; uniQure Biopharma B.V.: Research Funding.

Precursors of Viral Proteases as Distinct Drug Targets

Viral proteases are indispensable for successful virion maturation, thus making them a prominent drug target. Their enzyme activity is tightly spatiotemporally regulated by expression in the precursor form with little or no activity, followed by activation via autoprocessing. These cleavage events are frequently triggered upon transportation to a specific compartment inside the host cell. Typically, precursor oligomerization or the presence of a co-factor is needed for activation. A detailed understanding of these mechanisms will allow ligands with non-canonical mechanisms of action to be designed, which would specifically modulate the initial irreversible steps of viral protease autoactivation. Binding sites exclusive to the precursor, including binding sites beyond the protease domain, can be exploited. Both inhibition and up-regulation of the proteolytic activity of viral proteases can be detrimental for the virus. All these possibilities are discussed using examples of medically relevant viruses including herpesviruses, adenoviruses, retroviruses, picornaviruses, caliciviruses, togaviruses, flaviviruses, and coronaviruses.

Biochemical characterization of protease activity of Nsp3 from SARS-CoV-2 and its inhibition by nanobodies

Of the 16 non-structural proteins (Nsps) encoded by SARS CoV-2, Nsp3 is the largest and plays important roles in the viral life cycle. Being a large, multidomain, transmembrane protein, Nsp3 has been the most challenging Nsp to characterize. Encoded within Nsp3 is the papain-like protease domain (PLpro) that cleaves not only the viral polypeptide but also K48-linked polyubiquitin and the ubiquitin-like modifier, ISG15, from host cell proteins. We here compare the interactors of PLpro and Nsp3 and find a largely overlapping interactome. Intriguingly, we find that near full length Nsp3 is a more active protease compared to the minimal catalytic domain of PLpro. Using a MALDI-TOF based assay, we screen 1971 approved clinical compounds and identify five compounds that inhibit PLpro with IC50s in the low micromolar range but showed cross reactivity with other human deubiquitinases and had no significant antiviral activity in cellular SARS-CoV-2 infection assays. We therefore looked for alternative methods to block PLpro activity and engineered competitive nanobodies that bind to PLpro at the substrate binding site with nanomolar affinity thus inhibiting the enzyme. Our work highlights the importance of studying Nsp3 and provides tools and valuable insights to investigate Nsp3 biology during the viral infection cycle.

Structure of human factor VIIa–soluble tissue factor with calcium, magnesium and rubidium

Coagulation factor VIIa (FVIIa) consists of a γ-carboxyglutamic acid (GLA) domain, two epidermal growth factor-like (EGF) domains and a protease domain. FVIIa binds three Mg2+ ions and four Ca2+ ions in the GLA domain, one Ca2+ ion in the EGF1 domain and one Ca2+ ion in the protease domain. Further, FVIIa contains an Na+ site in the protease domain. Since Na+ and water share the same number of electrons, Na+ sites in proteins are difficult to distinguish from waters in X-ray structures. Here, to verify the Na+ site in FVIIa, the structure of the FVIIa–soluble tissue factor (TF) complex was solved at 1.8 Å resolution containing Mg2+, Ca2+ and Rb+ ions. In this structure, Rb+ replaced two Ca2+ sites in the GLA domain and occupied three non-metal sites in the protease domain. However, Rb+ was not detected at the expected Na+ site. In kinetic experiments, Na+ increased the amidolytic activity of FVIIa towards the synthetic substrate S-2288 (H-D-Ile-Pro-Arg-p-nitroanilide) by ∼20-fold; however, in the presence of Ca2+, Na+ had a negligible effect. Ca2+ increased the hydrolytic activity of FVIIa towards S-2288 by ∼60-fold in the absence of Na+ and by ∼82-fold in the presence of Na+. In molecular-dynamics simulations, Na+ stabilized the two Na+-binding loops (the 184-loop and 220-loop) and the TF-binding region spanning residues 163–180. Ca2+ stabilized the Ca2+-binding loop (the 70-loop) and Na+-binding loops but not the TF-binding region. Na+ and Ca2+ together stabilized both the Na+-binding and Ca2+-binding loops and the TF-binding region. Previously, Rb+ has been used to define the Na+ site in thrombin; however, it was unsuccessful in detecting the Na+ site in FVIIa. A conceivable explanation for this observation is provided.

SARS-CoV-2 Entry Protein TMPRSS2 and Its Homologue, TMPRSS4 Adopts Structural Fold Similar to Blood Coagulation and Complement Pathway Related Proteins

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) utilizes TMPRSS2 receptor to enter target human cells and subsequently causes coronavirus disease 19 (COVID-19). TMPRSS2 belongs to the type II serine proteases of subfamily TMPRSS, which is characterized by the presence of the serine-protease domain. TMPRSS4 is another TMPRSS member, which has a domain architecture similar to TMPRSS2. TMPRSS2 and TMPRSS4 have been shown to be involved in SARS-CoV-2 infection. However, their normal physiological roles have not been explored in detail. In this study, we analyzed the amino acid sequences and predicted 3D structures of TMPRSS2 and TMPRSS4 to understand their functional aspects at the protein domain level. Our results suggest that these proteins are likely to have common functions based on their conserved domain organization. Furthermore, we show that the predicted 3D structure of their serine protease domain has significant similarity to that of plasminogen which dissolves blood clot, and of other blood coagulation related proteins. Additionally, molecular docking analyses of inhibitors of four blood coagulation and anticoagulation factors show the same high specificity to TMPRSS2 and TMPRSS4 3D structures. Hence, our observations are consistent with the blood coagulopathy observed in COVID-19 patients and their predicted functions based on the sequence and structural analyses offer avenues to understand better and explore therapeutic approaches for this disease.

Conformational Plasticity-Rigidity Axis of the Coagulation Factor VII Zymogen Elucidated by Atomistic Simulations of the N-Terminally Truncated Factor VIIa Protease Domain

The vast majority of coagulation factor VII (FVII), a trypsin-like protease, circulates as the inactive zymogen. Activated FVII (FVIIa) is formed upon proteolytic activation of FVII, where it remains in a zymogen-like state and it is fully activated only when bound to tissue factor (TF). The catalytic domains of trypsin-like proteases adopt strikingly similar structures in their fully active forms. However, the dynamics and structures of the available corresponding zymogens reveal remarkable conformational plasticity of the protease domain prior to activation in many cases. Exactly how ligands and cofactors modulate the conformational dynamics and function of these proteases is not entirely understood. Here, we employ atomistic simulations of FVIIa (and variants hereof, including a TF-independent variant and N-terminally truncated variants) to provide fundamental insights with atomistic resolution into the plasticity-rigidity interplay of the protease domain conformations that appears to govern the functional response to proteolytic and allosteric activation. We argue that these findings are relevant to the FVII zymogen, whose structure has remained elusive despite substantial efforts. Our results shed light on the nature of FVII and demonstrate how conformational dynamics has played a crucial role in the evolutionary adaptation of regulatory mechanisms that were not present in the ancestral trypsin. Exploiting this knowledge could lead to engineering of protease variants for use as next-generation hemostatic therapeutics.