Deep mutational engineering of broadly-neutralizing nanobodies accommodating SARS-CoV-1 and 2 antigenic drift

Here, we report the molecular engineering of nanobodies that bind with picomolar affinity to both SARS-CoV-1 and SARS-CoV-2 receptor-binding domains (RBD) and are highly neutralizing. We applied deep mutational engineering to VHH72, a nanobody initially specific for SARS-CoV-1 RBD with little cross-reactivity to SARS-CoV-2 antigen. We first identified all the individual VHH substitutions that increase binding to SARS-CoV-2 RBD and then screened highly focused combinatorial libraries to isolate engineered nanobodies with improved properties. The corresponding VHH-Fc molecules show high affinities for SARS-CoV-2 antigens from various emerging variants and SARS-CoV-1, block the interaction between ACE2 and RBD, and neutralize the virus with high efficiency. Its rare specificity across sarbecovirus relies on its peculiar epitope outside the immunodominant regions. The engineered nanobodies share a common motif of three amino acids, which contribute to the broad specificity of recognition. Our results show that deep mutational engineering is a very powerful method, especially to rapidly adapt existing antibodies to new variants of pathogens.

Di-phosphorylated BAF shows altered structural dynamics and binding to DNA, but interacts with its nuclear envelope partners

Barrier-to-autointegration factor (BAF), encoded by the BANF1 gene, is an abundant and ubiquitously expressed metazoan protein that has multiple functions during the cell cycle. Through its ability to cross-bridge two double-stranded DNA (dsDNA), it favours chromosome compaction, participates in post-mitotic nuclear envelope reassembly and is essential for the repair of large nuclear ruptures. BAF forms a ternary complex with the nuclear envelope proteins lamin A/C and emerin, and its interaction with lamin A/C is defective in patients with recessive accelerated aging syndromes. Phosphorylation of BAF by the vaccinia-related kinase 1 (VRK1) is a key regulator of BAF localization and function. Here, we demonstrate that VRK1 successively phosphorylates BAF on Ser4 and Thr3. The crystal structures of BAF before and after phosphorylation are extremely similar. However, in solution, the extensive flexibility of the N-terminal helix α1 and loop α1α2 in BAF is strongly reduced in di-phosphorylated BAF, due to interactions between the phosphorylated residues and the positively charged C-terminal helix α6. These regions are involved in DNA and lamin A/C binding. Consistently, phosphorylation causes a 5000-fold loss of affinity for dsDNA. However, it does not impair binding to lamin A/C Igfold domain and emerin nucleoplasmic region, which leaves open the question of the regulation of these interactions.

Mechanism of MRX inhibition by Rif2 at telomeres

Specific proteins present at telomeres ensure chromosome end stability, in large part through unknown mechanisms. In this work, we address how the Saccharomyces cerevisiae ORC-related Rif2 protein protects telomere. We show that the small N-terminal Rif2 BAT motif (Blocks Addition of Telomeres) previously known to limit telomere elongation and Tel1 activity is also sufficient to block NHEJ and 5' end resection. The BAT motif inhibits the ability of the Mre11-Rad50-Xrs2 complex (MRX) to capture DNA ends. It acts through a direct contact with Rad50 ATP-binding Head domains. Through genetic approaches guided by structural predictions, we identify residues at the surface of Rad50 that are essential for the interaction with Rif2 and its inhibition. Finally, a docking model predicts how BAT binding could specifically destabilise the DNA-bound state of the MRX complex. From these results, we propose that when an MRX complex approaches a telomere, the Rif2 BAT motif binds MRX Head in its ATP-bound resting state. This antagonises MRX transition to its DNA-bound state, and favours a rapid return to the ATP-bound state. Unable to stably capture the telomere end, the MRX complex cannot proceed with the subsequent steps of NHEJ, Tel1-activation and 5' resection.

The C-terminal domain of HpDprA is a DNA-binding winged helix domain that does not bind double-stranded DNA

DNA-processing protein A, a ubiquitous multidomain DNA-binding protein, plays a crucial role during natural transformation in bacteria. Here, we carried out the structural analysis of DprA from the human pathogen Helicobacter pylori by combining data issued from the 1.8-Å resolution X-ray structure of the Pfam02481 domain dimer (RF), the NMR structure of the carboxy terminal domain (CTD), and the low-resolution structure of the full-length DprA dimer obtained in solution by SAXS. In particular, we sought a molecular function for the CTD, a domain that we show here is essential for transformation in H. pylori. Albeit its structural homology to winged helix DNA-binding motifs, we confirmed that the isolated CTD does not interact with ssDNA nor with dsDNA. The key R52 and K137 residues of RF are crucial for these two interactions. Search for sequences harboring homology to either HpDprA or Rhodopseudomonas palustris DprA CTDs led to the identification of conserved patches in the two CTD. Our structural study revealed the similarity of the structures adopted by these residues in RpDprA CTD and HpDprA CTD. This argues for a conserved, but yet to be defined, CTD function, distinct from DNA binding.

High affinity anchoring of the decoration protein pb10 onto the bacteriophage T5 capsid

Bacteriophage capsids constitute icosahedral shells of exceptional stability that protect the viral genome. Many capsids display on their surface decoration proteins whose structure and function remain largely unknown. The decoration protein pb10 of phage T5 binds at the centre of the 120 hexamers formed by the major capsid protein. Here we determined the 3D structure of pb10 and investigated its capsid-binding properties using NMR, SAXS, cryoEM and SPR. Pb10 consists of an α-helical capsid-binding domain and an Ig-like domain exposed to the solvent. It binds to the T5 capsid with a remarkably high affinity and its binding kinetics is characterized by a very slow dissociation rate. We propose that the conformational exchange events observed in the capsid-binding domain enable rearrangements upon binding that contribute to the quasi-irreversibility of the pb10-capsid interaction. Moreover we show that pb10 binding is a highly cooperative process, which favours immediate rebinding of newly dissociated pb10 to the 120 hexamers of the capsid protein. In extreme conditions, pb10 protects the phage from releasing its genome. We conclude that pb10 may function to reinforce the capsid thus favouring phage survival in harsh environments.

Crystal structure of full-length human collagenase 3 (MMP-13) with peptides in the active site defines exosites in the catalytic domain

Matrix metalloproteinase (MMP)-13 is one of the mammalian collagenases that play key roles in tissue remodelling and repair and in progression of diseases such as cancer, arthritis, atherosclerosis, and aneurysm. For collagenase to cleave triple helical collagens, the triple helical structure has to be locally unwound before hydrolysis, but this process is not well understood. We report crystal structures of catalytically inactive full-length human MMP-13(E223A) in complex with peptides of 14-26 aa derived from the cleaved prodomain during activation. Peptides are bound to the active site of the enzyme by forming an extended β-strand with Glu(40) or Tyr(46) inserted into the S1' specificity pocket. The structure of the N-terminal part of the peptides is variable and interacts with different parts of the catalytic domain. Those areas are designated substrate-dependent exosites, in that they accommodate different peptide structures, whereas the precise positioning of the substrate backbone is maintained in the active site. These modes of peptide-MMP-13 interactions have led us to propose how triple helical collagen strands fit into the active site cleft of the collagenase.

High throughput screening identifies disulfide isomerase DsbC as a very efficient partner for recombinant expression of small disulfide-rich proteins in E. coli

Background

Disulfide-rich proteins or DRPs are versatile bioactive compounds that encompass a wide variety of pharmacological, therapeutic, and/or biotechnological applications. Still, the production of DRPs in sufficient quantities is a major bottleneck for their complete structural or functional characterization. Recombinant expression of such small proteins containing multiple disulfide bonds in the bacteria E. coli is considered difficult and general methods and protocols, particularly on a high throughput scale, are limited.

Results

Here we report a high throughput screening approach that allowed the systematic investigation of the solubilizing and folding influence of twelve cytoplasmic partners on 28 DRPs in the strains BL21 (DE3) pLysS, Origami B (DE3) pLysS and SHuffle® T7 Express lysY (1008 conditions). The screening identified the conditions leading to the successful soluble expression of the 28 DRPs selected for the study. Amongst 336 conditions tested per bacterial strain, soluble expression was detected in 196 conditions using the strain BL21 (DE3) pLysS, whereas only 44 and 50 conditions for soluble expression were identified for the strains Origami B (DE3) pLysS and SHuffle® T7 Express lysY respectively. To assess the redox states of the DRPs, the solubility screen was coupled with mass spectrometry (MS) to determine the exact masses of the produced DRPs or fusion proteins. To validate the results obtained at analytical scale, several examples of proteins expressed and purified to a larger scale are presented along with their MS and functional characterization.

Conclusions

Our results show that the production of soluble and functional DRPs with cytoplasmic partners is possible in E. coli. In spite of its reducing cytoplasm, BL21 (DE3) pLysS is more efficient than the Origami B (DE3) pLysS and SHuffle® T7 Express lysY trxB^-/gor^- strains for the production of DRPs in fusion with solubilizing partners. However, our data suggest that oxidation of the proteins occurs ex vivo. Our protocols allow the production of a large diversity of DRPs using DsbC as a fusion partner, leading to pure active DRPs at milligram scale in many cases. These results open up new possibilities for the study and development of DRPs with therapeutic or biotechnological interest whose production was previously a limitation.

A multiscale coarse-grained polarizable solvent model for handling long tail bulk electrostatics

A multiscale coarse-grained approach able to handle efficiently the solvation of microscopic solutes in extended chemical environment is described. That approach is able to compute readily and efficiently very long-range solute/solvent electrostatic microscopic interactions, up to the 1-μm scale, by considering a reduced amount of computational resources. All the required parameters are assigned to reproduce available data concerning the solvation of single ions. Such a strategy makes it possible to reproduce with good accuracy the solvation properties concerning simple ion pairs in solution (in particular, the asymptotic behavior of the ion pair potentials of mean force). This new method represents an extension of the polarizable pseudoparticle solvent model, which has been recently improved to account for the main features of hydrophobic effects in liquid water (Masella et al., J. Comput. Chem. 2011, 32, 2664). This multiscale approach is well suited to be used for computing the impact of charge changes in free energy computations, in terms of both accuracy and efficiency.

Grafting of functional motifs onto protein scaffolds identified by PDB screening--an efficient route to design optimizable protein binders

Artificial miniproteins that are able to target catalytic sites of matrix metalloproteinases (MMPs) were designed using a functional motif-grafting approach. The motif corresponded to the four N-terminal residues of TIMP-2, a broad-spectrum protein inhibitor of MMPs. Scaffolds that are able to reproduce the functional topology of this motif were obtained by exhaustive screening of the Protein Data Bank (PDB) using STAMPS software (search for three-dimensional atom motifs in protein structures). Ten artificial protein binders were produced. The designed proteins bind catalytic sites of MMPs with affinities ranging from 450 nm to 450 μm prior to optimization. The crystal structure of one artificial binder in complex with the catalytic domain of MMP-12 showed that the inter-molecular interactions established by the functional motif in the artificial binder corresponded to those found in the MMP-14-TIMP-2 complex, albeit with some differences in geometry. Molecular dynamics simulations of the ten binders in complex with MMP-14 suggested that these scaffolds may allow partial reproduction of native inter-molecular interactions, but differences in geometry and stability may contribute to the lower affinity of the artificial protein binders compared to the natural protein binder. Nevertheless, these results show that the in silico design method used provides sets of protein binders that target a specific binding site with a good rate of success. This approach may constitute the first step of an efficient hybrid computational/experimental approach to protein binder design.

RASMOT-3D PRO: a 3D motif search webserver

Detection of structural motif of residues in protein structures allows identification of structural or functional similarity between proteins. In the field of protein engineering, structural motif identification is essential to select protein scaffolds on which a motif of residues can be transferred to design a new protein with a given function. We describe here the RASMOT-3D PRO webserver (http://biodev.extra.cea.fr/rasmot3d/) that performs a systematic search in 3D structures of protein for a set of residues exhibiting a particular topology. Comparison is based on Cα and Cβ atoms in two steps: inter-atomic distances and RMSD. RASMOT-3D PRO takes in input a PDB file containing the 3D coordinates of the searched motif and provides an interactive list of identified protein structures exhibiting residues of similar topology as the motif searched. Each solution can be graphically examined on the website. The topological search can be conducted in structures described in PDB files uploaded by the user or in those deposited in the PDB. This characteristic as well as the possibility to reject scaffolds sterically incompatible with the target, makes RASMOT-3D PRO a unique webtool in the field of protein engineering.

Combining a polarizable force-field and a coarse-grained polarizable solvent model: application to long dynamics simulations of bovine pancreatic trypsin inhibitor

The dynamic coupling between a polarizable protein force field and a particle-based implicit solvent model is described. The polarizable force field, TCPEp, developed recently to simulate protein systems, is characterized by a reduced number of polarizable sites, with a substantial gain in efficiency for an equal chemical accuracy. The Polarizable Pseudo-Particle (PPP) solvent model represents the macroscopic solvent polarization by induced dipoles placed on mobile Lennard-Jones pseudo-particles. The solvent-induced dipoles are sensitive to the solute electric field, but not to each other, so that the computational cost of solvent-solvent interactions is basically negligible. The solute and solvent induced dipoles are determined self-consistently and the equations of motion are solved using an efficient iterative multiple time step procedure. The solvation cost with respect to vacuum simulations is shown to decrease with solute size: the estimated multiplicative factor is 2.5 for a protein containing about 1000 atoms, and as low as 1.15 for 8000 atoms. The model is tested for six 20 ns molecular dynamics trajectories of a traditional benchmark system: the hydrated Bovine Pancreatic Trypsin Inhibitor (BPTI). Even though the TCPEp parameters have not been refined to be used with the solvent PPP model, we observe a good conservation of the BPTI structure along the trajectories. Moreover, our approach is able to provide a description of the protein solvation thermodynamic at the same accuracy as the standard Poisson-Boltzman continuum methods. It provides in addition a good description of the microscopic structural aspects concerning the solute/solvent interaction.

Development of potent and selective phosphinic peptide inhibitors of angiotensin-converting enzyme 2

Angiotensin-converting enzyme 2 (ACE2), a recently identified human homologue of angiotensin-converting enzyme, is a zinc metallocarboxypeptidase which may play a unique role in cardiovascular and renal function. Here we report the discovery of potent and selective inhibitors of ACE2, which have been identified by evaluating a series of phosphinic di- and tripeptides of the general formula: Z-Xaa(PO 2-CH 2)YaaOH and Ac-Zaa-Xaa(PO 2-CH 2)YaaOH. The most potent inhibitor in this series is a tripeptide that displays a K i value of 0.4 nM toward ACE2 and is 3 orders of magnitude less potent toward carboxypeptidase A. Phosphinic tripeptides exhibit high potency exclusively when the Xaa position is occupied by a pseudoproline. A model of interaction between one inhibitor of this series and ACE2 suggests that the critical role played by a proline in inhibitors, but also for substrates hydrolysis, may rely on the presence of Tyr (510) in the ACE2 active site.

Structure-based secondary structure-independent approach to design protein ligands: Application to the design of Kv1.2 potassium channel blockers

We have developed a structure-based approach to the design of protein ligands. This approach is based on the transfer of a functional binding motif of amino acids, often referred as to the "hot spot", on a host protein able to reproduce the functional topology of these residues. The scaffolds were identified by a systematic in silico search in the Protein Data Bank for proteins possessing a group of residues in a topology similar to that adopted by the functional motif in a reference ligand of known 3D structure. In contrast to previously reported studies, this search is independent of the particular secondary structure supporting the functional motif. To take into account the global properties of the host protein, two additional criteria were taken into account in the selection process: (1) Only those scaffolds sterically compatible with the positioning of the functional motif as observed in a reference complex model were retained. (2) Host proteins displaying electrostatic potentials, in the region of the transferred functional motif, similar to that of the reference ligand were selected. This approach was applied to the development of protein ligands of the Kv1.2 channel using BgK, a small protein isolated from the sea anemone Bunodosoma granulifera, as the reference ligand. Four proteins obtained by this approach were produced for experimental evaluation. The X-ray structure of one of these proteins was determined to check for similarity of the transferred functional motif with the structure it adopts in the reference ligand. Three of these protein ligands bind the Kv1.2 channel with inhibition constants of 0.5, 1.5, and 1.6 microM. Several mutants of these designed protein ligands gave binding results consistent with the presumed binding mode. These results show that protein ligands can be designed by transferring a binding motif on a protein host selected to reproduce the functional topology of this motif, irrespective to the secondary structure supporting the functional motif, if the host protein possesses steric and electrostatic properties compatible with the binding to the target. This result opens the way to the design of protein ligands by taking advantage of the considerable structural repertoire of the Protein Data Bank.

Development of Selective Inhibitors and Substrate of Matrix Metalloproteinase-12

Four phosphinic peptide libraries with compounds having the general formula p-Br-Ph-(PO2-CH2)-Xaa'-Yaa'-Zaa'-NH2 have been prepared and screened against 10 matrix metalloproteinases (MMPs). We identified two phosphinic peptides with Ki values of 0.19 and 4.4 nM toward MMP-12 (macrophage elastase) that are more than 2-3 orders of magnitude less potent toward the other MMPs tested. These highly selective MMP-12 inhibitors contain a Glu-Glu motif in their Yaa'-Zaa' positions. Incorporation of this Glu-Glu motif into the sequence of a nonspecific fluorogenic peptide cleaved by MMPs provides a highly selective substrate for MMP-12. A model of one of these inhibitors interacting with MMP-12 suggests that the selectivity observed might be due, in part, to the presence of two unique polar residues in MMP-12, Thr239 and Lys177. These MMP-12-selective inhibitors may have important therapeutic applications to diseases in which MMP-12 has been suggested to play a key role, such as in emphysema, atherosclerosis, and aortic abdominal aneurysm.

Combined use of selective inhibitors and fluorogenic substrates to study the specificity of somatic wild-type angiotensin-converting enzyme

Somatic angiotensin-converting enzyme (ACE) contains two homologous domains, each bearing a functional active site. Studies on the selectivity of these ACE domains towards either substrates or inhibitors have mostly relied on the use of mutants or isolated domains of ACE. To determine directly the selectivity properties of each ACE domain, working with wild-type enzyme, we developed an approach based on the combined use of N-domain-selective and C-domain-selective ACE inhibitors and fluorogenic substrates. With this approach, marked differences in substrate selectivity were revealed between rat, mouse and human somatic ACE. In particular, the fluorogenic substrate Mca-Ala-Ser-Asp-Lys-DpaOH was shown to be a strict N-domain-selective substrate of mouse ACE, whereas with rat ACE it displayed marked C-domain selectivity. Similar differences in selectivity between these ACE species were also observed with a new fluorogenic substrate of ACE, Mca-Arg-Pro-Pro-Gly-Phe-Ser-Pro-DpaOH. In support of these results, changes in amino-acid composition in the binding site of these three ACE species were pinpointed. Together these data demonstrate that the substrate selectivity of the N-domain and C-domain depends on the ACE species. These results raise concerns about the interpretation of functional studies performed in animals using N-domain and C-domain substrate selectivity data derived only from human ACE.

Future challenges facing the development of specific active-site-directed synthetic inhibitors of MMPs

Despite a deep knowledge on the 3D-structure of several catalytic domains of MMPs, the development of highly specific synthetic active-site-directed inhibitors of MMPs, able to differentiate the different members of this protease family, remains a strong challenge. Due to the flexible nature of MMP active-site, the development of specific MMP inhibitors will need to combine sophisticated theoretical and experimental approaches to decipher in each MMP the specific structural and dynamic features that can be exploited to obtain the desired selectivity.

Residues determining the binding specificity of uncompetitive inhibitors to tissue-nonspecific alkaline phosphatase

Recent data have pointed to TNALP as a therapeutic target for soft-tissue ossification abnormalities. Here, we used mutagenesis, kinetic analysis, and computer modeling to identify the residues important for the binding of known ALP inhibitors to the TNALP active site. These data will enable drug design efforts aimed at developing improved specific TNALP inhibitors for therapeutic use.

INTRODUCTION

We have shown previously that the genetic ablation of tissue-nonspecific alkaline phosphatase (TNALP) function leads to amelioration of soft-tissue ossification in mouse models of osteoarthritis and ankylosis (i.e., Enpp1-/- and ank/ank mutant mice). We surmise that the pharmacologic inhibition of TNALP activity represents a viable therapeutic approach for these diseases. As a first step toward developing suitable TNALP therapeutics, we have now clarified the residues involved in binding well-known uncompetitive inhibitors to the TNALP active site.

MATERIALS AND METHODS

We compared the modeled 3D structure of TNALP with the 3D structure of human placental alkaline phosphatase (PLALP) and identified the residues that differ between these isozymes within a 12 A radius of the active site, because these isozymes differ significantly in inhibitor specificity. We then used site-directed mutagenesis to substitute TNALP residues to their respective homolog in PLALP. In addition, we mutagenized most of these residues in TNALP to Ala and the corresponding residues in PLALP to their TNALP homolog. All mutants were characterized for their sensitivity toward the uncompetitive inhibitors l-homoarginine (L-hArg), levamisole, theophylline, and l-phenylalanine.

RESULTS AND CONCLUSIONS

We found that the identity of residue 108 in TNALP largely determines the specificity of inhibition by L-hArg. The conserved Tyr-371 is also necessary for binding of L-hArg. In contrast, the binding of levamisole to TNALP is mostly dependent on His-434 and Tyr-371, but not on residues 108 or 109. The main determinant of sensitivity to theophylline is His-434. Thus, we have clarified the location of the binding sites for all three TNALP inhibitors, and we have also been able to exchange inhibitor specificities between TNALP and PLALP. These data will enable drug design efforts aimed at developing improved, selective, and drug-like TNALP inhibitors for therapeutic use.

Phosphinic peptides as zinc metalloproteinase inhibitors

Solid-phase synthesis of phosphinic peptides was introduced 10 years ago. A major application of this chemistry has been the development of potent synthetic inhibitors of zinc metalloproteases. Specific properties of the inhibitors produced in recent years are reviewed, supporting the notion that phosphinic pseudo-peptides are useful tools for studying the structural and functional biology of zinc proteases.

Structural Determinants of RXPA380, a Potent and Highly Selective Inhibitor of the Angiotensin-Converting Enzyme C-Domain

RXPA380 (Cbz-PhePsi[PO(2)CH]Pro-Trp-OH) was reported recently as the first highly selective inhibitor of the C-domain of somatic angiotensin-converting enzyme (ACE), able to differentiate the two active sites of somatic ACE by a selectivity factor of more than 3 orders of magnitude. The contribution of each RXPA380 residue toward this remarkable selectivity was evaluated by studying several analogues of RXPA380. This analysis revealed that both pseudo-proline and tryptophan residues in the P(1)' and P(2)' positions of RXPA380 play a critical role in the selectivity of this inhibitor for the C-domain. This selectivity is not due to a preference of the C-domain for inhibitors bearing pseudo-proline and tryptophan residues, but rather reflects the poor accommodation of these inhibitor residues by the N-domain. A model of RXPA380 in complex with the ACE C-domain, based on the crystal structure of germinal ACE, highlights residues that may contribute to RXPA380 selectivity. From this model, striking differences between the N- and C-domains of ACE are observed for residues defining the S(2)' pocket. Of the twelve residues that surround the tryptophan side chain of RXPA380 in the C-domain, five are different in the N-domain. These differences in the S(2)' composition between the N- and C-domains are suggested to contribute to RXPA380 selectivity. The structural insights provided by this study should enhance understanding of the factors controlling the selectivity of the two domains of somatic ACE and allow the design of new selective ACE inhibitors.

Evaluation of P1'-diversified phosphinic peptides leads to the development of highly selective inhibitors of MMP-11

Phosphinic peptides were previously reported to be potent inhibitors of several matrixins (MMPs). To identify more selective inhibitors of MMP-11, a matrixin overexpressed in breast cancer, a series of phosphinic pseudopeptides bearing a variety of P(1)'-side chains has been synthesized, by parallel diversification of a phosphinic template. The potencies of these compounds were evaluated against a set of seven MMPs (MMP-2, MMP-7, MMP-8, MMP-9, MMP-11, MMP-13, and MMP-14). The chemical strategy applied led to the identification of several phosphinic inhibitors displaying high selectivity toward MMP-11. One of the most selective inhibitors of MMP-11 in this series, compound 22, exhibits a K(i) value of 0.23 microM toward MMP-11, while its potency toward the other MMPs tested is 2 orders of magnitude lower. This remarkable selectivity may rely on interactions of the P(1)'-side chain atoms of these inhibitors with residues located at the entrance of the S(1)'-cavity of MMP-11. The design of inhibitors able to interact with residues located at the entrance of MMPs' S(1)'-cavity might represent an alternative strategy to identify selective inhibitors that will fully differentiate one MMP among the others.

Selective inhibition of the C-domain of angiotensin I converting enzyme by bradykinin potentiating peptides

Somatic angiotensin I converting enzyme (ACE) contains two functional active sites. Up to now, most of the studies aimed at characterizing the selectivity of inhibitors toward the two ACE active sites relied on the use of ACE mutants containing a single functional active site. By developing new fluorogenic synthetic substrates of ACE, we demonstrated that inhibitor selectivity can be assessed directly by using somatic ACE. This useful screening approach led us to discover that some bradykinin potentiating peptides turned out to be selective inhibitors of the C-domain of ACE. The peptide pGlu-Gly-Leu-Pro-Pro-Arg-Pro-Lys-Ile-Pro-Pro, with K(i)(app) values of 30 nM and 8 microM, respectively, for the C- and N-domain of ACE, is to our knowledge the most highly selective C-domain inhibitor of ACE so far reported. Inhibitors able to block selectively either the N- or C-domain of ACE will represent unique tools to probe the function of each domain in the regulation of blood pressure or other physiopathological events involving ACE activity.

Potency and selectivity of RXP407 on human, rat, and mouse angiotensin-converting enzyme

By screening phosphinic peptide libraries, we recently reported the discovery of RXP407 (Ac-Asp-PheY(PO2-CH2)LAla-Ala-NH2), a potent N-domain-selective inhibitor of recombinant human angiotensin-converting enzyme (ACE). Preliminary studies to evaluate the in vivo activity of RXP407 in rat led us to suspect possible differences in the binding property of RXP407 between human and rat ACE. The aim of the present study was thus to determine the potency of RXP407 toward rat and mouse ACEs, as compared to non-recombinant human ACE, and to assess the efficacy of this inhibitor in discriminating between the N- and C-domains of these ACE enzymes. By comparing the ability of RXP407 to block purified somatic and germinal ACE from mice, RXP407 was shown to be a potent N-domain-selective inhibitor of mouse somatic ACE, a behavior similar to that observed with human somatic ACE. In contrast, RXP407 appeared less potent toward purified ACE from rat and furthermore was unable to block ACE activity present in crude rat plasma. This study demonstrated that for further evaluation of the in vivo efficacy of RXP407, mice rather than rats should be used as the animal model. Thus, following the change in the Ac-S-D-K-P plasmatic levels, after i.v. injection of RXP407 to mice, will permit the potency and selectivity of this novel ACE inhibitor to be assessed.

Crystal structure of the stromelysin-3 (MMP-11) catalytic domain complexed with a phosphinic inhibitor mimicking the transition-state

Stromelysin-3 (ST3) is a matrix metalloproteinase (MMP-11) whose proteolytic activity plays an important role in tumorigenicity enhancement. In breast cancer, ST3 is a bad prognosis marker: its expression is associated with a poor clinical outcome. This enzyme therefore represents an attractive therapeutic target. The topology of matrix metalloproteinases (MMPs) is remarkably well conserved, making the design of highly specific inhibitors difficult. The major difference between MMPs lies in the S(1)' subsite, a well-defined hydrophobic pocket of variable depth. The present crystal structure, the first 3D-structure of the ST3 catalytic domain in interaction with a phosphinic inhibitor mimicking a (d, l) peptide, clearly demonstrates that its S(1)' pocket corresponds to a tunnel running through the enzyme. This open channel is filled by the inhibitor P(1)' group which adopts a constrained conformation to fit this pocket, together with two water molecules interacting with the ST3-specific residue Gln215. These observations provide clues for the design of more specific inhibitors and show how ST3 can accommodate a phosphinic inhibitor mimicking a (d, l) peptide. The presence of a water molecule interacting with one oxygen atom of the inhibitor phosphinyl group and the proline residue of the Met-turn suggests how the intermediate formed during proteolysis may be stabilized. Furthermore, the hydrogen bond distance observed between the methyl of the phosphinic group and the carbonyl group of Ala182 mimics the interaction between this carbonyl group and the amide group of the cleaved peptidic bond. Our crystal structure provides a good model to study the MMPs mechanism of proteolysis.

Phosphinic peptide inhibitors as tools in the study of the function of zinc metallopeptidases

The development of the combinatorial chemistry of phosphinic peptides has led to the discovery of both highly potent and selective inhibitors of various zinc metalloproteinases. Several properties of these compounds are reviewed, supporting the view that this class of inhibitors should represent useful tools for probing several aspects of the function of this broad family of proteases in vivo.

Role of the S1' subsite glutamine 215 in activity and specificity of stromelysin-3 by site-directed mutagenesis

The influence of Gln215 in stromelysin-3 (MMP-11), a residue located in the S1' subsite, was determined by producing three single mutants of this position. As compared to wild-type stromelysin-3, the kinetic parameters K(M) and k(cat) for the degradation of the fluorogenic substrate Dns-Pro-Leu-Ala-Leu-Trp-Ala-Arg-NH(2) (Dns-Leu) by these mutants indicated that the Gln/Leu substitution led to a 4-fold decrease in catalytic efficiency, whereas the mutations Gln/Tyr and Gln/Arg increased this parameter by a factor 10. The cleavage of alpha1-protease inhibitor (alpha1-PI), a natural substrate of stromelysin-3, by these mutants was also determined. Their relative activities for the degradation of alpha1-PI correspond to those observed with the synthetic substrate Dns-Leu. The catalytic efficiency of wild-type stromelysin-3 and its mutants to cleave the P1' analogue of Dns-Leu, containing the unusual amino acid Cys(OMeBn) (Dns-Cys(OMeBn)), was also determined. The values of the specificity factor, calculated as the ratio (k(cat)/K(M))Dns-Cys(OMeBn))/(k(cat)/K(M))Dns-Leu, were observed to vary from 26 for the wild-type stromelysin-3 to 120 for the Gln/Leu mutant and 25 for the Gln/Arg mutant. The Gln/Tyr mutant did not cleave the substrate when its P1' position is substituted by the unusual amino acid Cys(OMeBn). Altogether these observations established that both the catalytic activity and the specificity of stromelysin-3 are dependent on the nature of the residue in position 215. Finally, the cleavage efficiency of the Dns substrates by three representative matrixins, namely, MMP-14 (215 = Leu), MMP-1 (215 = Arg), and MMP-7 (215 = Tyr), was determined. Interestingly, the trends observed for these enzymes were similar to those established for the three mutants of stromelysin-3, pointing out the influence of position 215 toward the selectivity in this family of enzymes.

Phosphinic pseudo-tripeptides as potent inhibitors of matrix metalloproteinases: a structure-activity study

Several phosphinic pseudo-tripeptides of general formula R-XaaPsi(PO(2)-CH(2))Xaa'-Yaa'-NH(2) were synthesized and evaluated for their in vitro activities to inhibit stromelysin-3, gelatinases A and B, membrane type-1 matrix metalloproteinase, collagenases 1 and 2, and matrilysin. With the exception of collagenase-1 and matrilysin, phosphinic pseudo-tripeptides behave as highly potent inhibitors of matrix metalloproteinases, provided they contain in P(1)' position an unusual long aryl-alkyl substituent. Study of structure-activity relationships regarding the influence of the R and Xaa' substituents in this series may contribute to the design of inhibitors able to block only a few members of the matrix metalloproteinase family.

RXP 407, a phosphinic peptide, is a potent inhibitor of angiotensin I converting enzyme able to differentiate between its two active sites

The human somatic angiotensin converting enzyme (ACE) contains two homologous domains, each bearing a zinc-dependent active site. All of the synthetic inhibitors of this enzyme used in clinical applications interact with these two active sites to a similar extent. Recently, several lines of evidence have suggested that the N-terminal active site of ACE might be involved in specific hydrolysis of some important physiological substrates, like Acetyl-Seryl-Aspartyl-Lysyl-Proline, a negative regulator of hematopoietic stem cell differentiation and proliferation. These findings have stimulated studies aimed at identifying new ACE inhibitors able to block only one of the two active sites of this enzyme. By screening phosphinic peptide libraries, we discovered a phosphinic peptide Ac-Asp-(L)Phepsi(PO2-CH2)(L)Ala-Ala-NH2, called RXP 407, which is able to differentiate the two ACE active sites, with a dissociation constant three orders of magnitude lower for the N-domain of the enzyme. The usefulness of a combinatorial chemistry approach to develop new lead structures is underscored by the unusual chemical structure of RXP 407, as compared with classical ACE inhibitors. As a highly potent and selective inhibitor of the N-terminal active site of wild ACE (Ki = 12 nM), RXP 407, which is metabolically stable in vivo, may lead to a new generation of ACE inhibitors able to block in vivo only a subset of the different functions regulated by ACE.

Membrane type-1 matrix metalloprotease and stromelysin-3 cleave more efficiently synthetic substrates containing unusual amino acids in their P1' positions

The influence of the substrate P1' position on the specificity of two zinc matrix metalloproteases, membrane type-1 matrix metalloprotease (MT1-MMP) and stromelysin-3 (ST3), was evaluated by synthesizing a series of fluorogenic substrates of general formula dansyl-Pro-Leu-Ala-Xaa-Trp-Ala-Arg-NH2, where Xaa in the P1' position represents unusual amino acids containing either long arylalkyl or alkyl side chains. Our data demonstrate that both MT1-MMP and ST3 cleave substrates containing in their P1' position unusual amino acids with extremely long side chains more efficiently than the corresponding substrates with natural phenylalanine or leucine amino acids. In this series of substrates, the replacement of leucine by S-para-methoxybenzyl cysteine increased the kcat/Km ratio by a factor of 37 for MT1-MMP and 9 for ST3. The substrate with a S-para-methoxybenzyl cysteine residue in the P1' position displayed a kcat/Km value of 1.59 10(6) M-1 s-1 and 1.67 10(4) M-1 s-1, when assayed with MT1-MMP and ST3, respectively. This substrate is thus one of the most rapidly hydrolyzed substrates so far reported for matrixins, and is the first synthetic peptide efficiently cleaved by ST3. These unexpected results for these two matrixins suggest that extracellular proteins may be cleaved by matrixins at sites containing amino acids with unusual long side chains, like those generated in vivo by some post-translational modifications.

Structural basis of antibody cross-reactivity: solution conformation of an immunogenic peptide fragment containing both T and B epitopes

A synthetic octadecapeptide with the amino acid sequence of residues 23-40 of toxin alpha from Naja nigricollis, cyclized with a disulfide bridge between residues 23 and 40, induces antibodies that cross-react with toxin alpha. We report a structural analysis of this peptide in aqueous solution using NMR spectroscopy and molecular modeling. Structures compatible with the 151 obtained NMR distance restraints were generated using a random simulated annealing protocol followed by restrained high-temperature dynamics and energy minimization. The generated structures are compared with that of the corresponding sequence in the native toxin. The two stretches 23-28 and 37-40 adopt a canonical beta-strand structure in the toxin but are disordered in the peptide. The region 28-36 is ordered in both the peptide and the toxin. Residues 28-30 and 34-36 adopt beta-strand structures in the toxin but loop structures in the peptide. Residues 30-33 form a reverse turn in both the peptide and the toxin. Residues Val-27, Trp-28, Ile-35, and Ile-36 form a hydrophobic cluster. The similar, reverse-turn fold of residues 30-33 in the peptide and the toxin may be associated with the immunogenic cross-reactivity.

Three-dimensional structure of cyclohexapeptides containing a phosphinic bond in aqueous solution: a template for zinc metalloprotease inhibitors. A NMR and restrained molecular dynamics study

The 3D structures of two phosphinic cyclic hexapeptide inhibitors of bacterial collagenase, cyclo-(Gly1-Pro2-Phe3 psi[PO2-CH2]Gly4-Pro5-Nle6) (compound I) and cyclo(Gly1-Pro2-D-Phe3 psi[PO2-CH2]-Gly4-Pro5-Nle6) (compound II), in aqueous solution, as derived from NMR spectroscopy and molecular dynamics simulations, are described. The general structures of these cyclic hexapeptides closely resemble the "canonic" two-reverse-turn structure, with the proline occupying the (i + 1) position of the turns and the glycine the connecting positions. The phosphinic bond is located between the (i + 2) and (i + 3) positions of one of these turns. However, a striking feature of the backbone structure of these peptides is the presence of double type VIII-turns in compound I, and in compound II of type VIII- and tentatively named type IX-turns. The comparison of the 3D structures of these two cyclic hexapeptides shows that the stereochemistry of the phenylalanylphosphinyl residue influences not only the local conformation but also the global topology of the peptide macrocycle. The differences in the 3D structure of these compounds are discussed in relation to their inhibitory potencies and with the view of using these constrained cyclic peptides as a scaffold for the development of rigid metalloproteases inhibitors.

Cyclic peptides with a phosphinic bond as potent inhibitors of a zinc bacterial collagenase

A series of cyclic peptides containing a phosphinic bond were synthesized and evaluated as inhibitors of a zinc bacterial collagenase from Corynebacterium rathaii. Among this series of pseudopeptides of different sizes of cycles, only two molecules Ia (cyclo[Gly-Pro-Phe psi(PO2CH2)-Gly-Pro-Ahx]) and Va (cyclo[beta Ala-Pro-Phe psi (PO2CH2)Gly-Pro-Ahx]) were found to be rather potent inhibitors of this protease, with Ki values of 120 and 90 nM, respectively. Besides the influence of the peptide ring size, this study suggests that both the stereochemical and the conformational properties of the pseudophenylalanine residue in these cyclic peptides may determine their potency. Interestingly, the kinetic analysis for the binding of the cyclic peptide inhibitors Ia and Va to the collagenase, as compared to a linear parent compound, reveals that the lower potency of the cyclic peptides is mostly the consequence of a lower rate constant for association to the enzyme. To our knowledge, this is the first report on cyclic phosphinic peptides and on their activities as inhibitors of a zinc protease.

The abasic site as a challenge to DNA polymerase. A nuclear magnetic resonance study of G, C and T opposite a model abasic site

An abasic site in DNA creates a strong block to DNA polymerase and is a mutagenic base lesion. In this study, we present structural and dynamic properties of duplex oligodeoxynucleotides containing G, C and T opposite a model abasic site studied by one and two-dimensional nuclear magnetic resonance spectroscopy. We have demonstrated that A opposite the abasic site was positioned within the helix as if paired with T, and that the A residue melted co-operatively with the surrounding helix. We report here that G opposite the abasic site is also observed to be predominantly intrahelical in a normal anti conformation at low temperature. With increasing temperature, the mobility of the G residue increases rapidly and apparently is in a "melted state" well before denaturation of the helix. At low temperature, two species are found for T opposite the abasic site; one, intrahelical, one extrahelical. These species are in slow exchange with one another on a proton nuclear magnetic resonance time-scale. The two species then move into fast exchange with increasing temperature and the proportion of the extra-helical form increases. When C is positioned opposite the abasic site, both the C residue and the abasic sugar are extrahelical, the helix collapses, and the adjacent G.C base-pairs stack over one another. On the basis of these observations, we propose a model that explains why the abasic site acts to block DNA replication. Further, we suggest an explanation for the observed polymerase preference for base selection at abasic sites.

Molecular mechanics and dynamics of an abasic frameshift in DNA and comparison to NMR data

In a previous publication (Ph. Cuniasse, L.C. Sowers, R. Eritja, B. Kaplan, M.F. Goodman, J.A.H. Cognet, M. Le Bret, W. Guschlbauer and G.V. Fazakerley, Biochemistry 28, 2018 (1989), we determined by two dimensional NMR studies and molecular mechanics calculations the three-dimensional structure of a non-selfcomplementary oligonucleotide: [sequence; see text] where dr, at the center of the first strand, is a model abasic site. In order to explain all the results arising from NMR measurements, we found that an equilibrium between two conformations was necessary. These conformations differ mainly by the sugar pucker of G5 which is C2' endo or C3' endo. The latter is stabilized by addition of counterions between phosphate residues P3 and P4. In this paper, we have constructed systematically, all possible structures as a function of torsion angles delta of dr4 and of G5 by molecular mechanics in the presence or absence of counterions. Since these conformations were not forced with NMR distance measurements, this method allows detailed comparisons between all possible conformations and NMR data. Maps of contour lines of the potential energy, of fits to NMR distance measurements, and of helical twist as a function of torsion angles delta of dr4 and of G5 unravel the difficulties associated with the study of the G5 sugar pucker conformation equilibrium. Sugar puckers and proton distances are very sensitive criteria to monitor molecular dynamics. Relying on these experimental criteria, we have tested many molecular dynamics preparation phases and we propose a new warm-up and equilibration procedure for molecular dynamics. Thus we show with a 290 ps molecular dynamic run that G5 is in conformational equilibrium and that all NMR data are well reproduced.

Abasic frameshift in DNA. Solution conformation determined by proton NMR and molecular mechanics calculations

We have determined the three-dimensional structure of a non-self-complementary oligodeoxynucleotide duplex that contains a model abasic site. The duplex contains six GC base pairs plus the abasic site at the center of one strand and corresponds to an abasic frameshift. Two-dimensional NMR studies on the nonexchangeable protons show that the guanine bases on either side of the abasic site are stacked over each other and that the abasic site is rotated out of the helix. Close proton-proton interactions are observed between the H4' proton of the abasic site and sugar protons of the guanosine in the 5' direction, which allows the position of the free sugar to be well-defined. NOE buildup curves from NOESY spectra recorded at very short mixing times were used to calculate a set of interproton distances. This data set was incorporated into the refinement of the oligonucleotide structure by molecular mechanics calculations. Two conformations that differ in the sugar conformation of the guanosine next to the abasic site in the 3' direction were necessary to fit all the NMR data. One of these two conformations could only be stabilized by addition of counterions at specific sites.

An abasic site in DNA. Solution conformation determined by proton NMR and molecular mechanics calculations

We have determined the three-dimensional structure of a non-selfcomplementary nonanucleotide duplex which contains an abasic (apyrimidinic) site in the centre, i.e. a deoxyribose residue opposite an adenosine. The majority of the base and sugar proton resonances were assigned by NOESY, COSY and 2DQF spectra in D2O and H2O. We have measured the initial slope of buildup of NOEs in NOESY spectra at very short mixing times (25 to 50 ms), and from these were able to establish interproton distances for the central part of the duplex. We propose a different strategy for proton-proton distance determinations which takes into account the observed variations in correlation times for particular proton-proton vectors. A set of 31 measured interproton distances was incorporated into the refinement of the oligonucleotide structure by molecular mechanics calculations. Two structures were obtained which retain all aspects of a classical B DNA in which the unpaired adenine and the abasic deoxyribose lie inside the helix. We observe that the non-hydrogen bonded adenine is held well in the helix, the Tm of this base being the same as that of the A.T base pairs in the same duplex.