Search

Modification dependent restriction endonucleases (MDREs) often have separate catalytic and modification dependent domains. We systematically looked for previously uncharacterized fusion proteins featuring a PUA or DUF3427 domain and HNH or PD-(D/E)XK catalytic domain. The enzymes were clustered by similarity of their putative modification sensing domains into several groups. The TspA15I (VcaM4I, CmeDI), ScoA3IV (MsiJI, VcaCI) and YenY4I groups, all featuring a PUA superfamily domain, preferentially cleaved DNA containing 5-methylcytosine or 5-hydroxymethylcytosine. ScoA3V, also featuring a PUA superfamily domain, but of a different clade, exhibited 6-methyladenine stimulated nicking activity. With few exceptions, ORFs for PUA-superfamily domain containing endonucleases were not close to DNA methyltransferase ORFs, strongly supporting modification dependent activity of the endonucleases. DUF3427 domain containing fusion proteins had very little or no endonuclease activity, despite the presence of a putative PD-(D/E)XK catalytic domain. However, their expression potently restricted phage T4gt in Escherichia coli cells. In contrast to the ORFs for PUA domain containing endonucleases, the ORFs for DUF3427 fusion proteins were frequently found in defense islands, often also featuring DNA methyltransferases.

Nucleic acids research / Special publication 49(3), 1708 – 1723 (2021). doi:10.1093/nar/gkaa1218, Many modification-dependent restriction endonucleases (MDREs) are fusions of a PUA superfamily modification sensor domain and a nuclease catalytic domain. EVE domains belong to the PUA superfamily, and are present in MDREs in combination with HNH nuclease domains. Here, we present a biochemical characterization of the EVE-HNH endonuclease VcaM4I and crystal structures of the protein alone, with EVE domain bound to either 5mC modified dsDNA or to 5mC/5hmC containing ssDNA. The EVE domain is moderately specific for 5mC/5hmC containing DNA according to EMSA experiments. It flips the modified nucleotide, to accommodate it in a hydrophobic pocket of the enzyme, primarily formed by P24, W82 and Y130 residues. In the crystallized conformation, the EVE domain and linker helix between the two domains block DNA binding to the catalytic domain. Removal of the EVE domain and inter-domain linker, but not of the EVE domain alone converts VcaM4I into a non-specific toxic nuclease. The role of the key residues in the EVE and HNH domains of VcaM4I is confirmed by digestion and restriction assays with the enzyme variants that differ from the wild-type by changes to the base binding pocket or to the catalytic residues., Published by London

Modification dependent restriction endonucleases (MDREs) restrict modified DNA, typically with limited sequence specificity (∼2–4 bp). Here, we focus on MDREs that have an SRA and/or SBD (sulfur binding domain) fused to an HNH endonuclease domain, cleaving cytosine modified or phosphorothioated (PT) DNA. We independently characterized the SBD-SRA-HNH endonuclease ScoMcrA, which preferentially cleaves 5hmC modified DNA. We report five SBD-HNH endonucleases, all recognizing GpsAAC/GpsTTC sequence and cleaving outside with a single nucleotide 3′ stagger: EcoWI (N7/N6), Ksp11411I (N5/N4), Bsp305I (N6/N4-5), Mae9806I [N(8-10)/N(8-9)], and Sau43800I [N(8-9)/N(7-8)]. EcoWI and Bsp305I are more specific for PT modified DNA in Mg2+ buffer, and promiscuous with Mn2+. Ksp11411I is more PT specific with Ni2+. EcoWI and Ksp11411I cleave fully- and hemi-PT modified oligos, while Bsp305I cleaves only fully modified ones. EcoWI forms a dimer in solution and cleaves more efficiently in the presence of two modified sites. In addition, we demonstrate that EcoWI PT-dependent activity has biological function: EcoWI expressing cells restrict dnd+ GpsAAC modified plasmid strongly, and GpsGCC DNA weakly. This work establishes a framework for biotechnology applications of PT-dependent restriction endonucleases (PTDRs).

Journal of molecular biology 431(11), 2082 - 2094 (2019). doi:10.1016/j.jmb.2019.04.010, Specificity engineering is challenging and particularly difficult for enzymes that have the catalytic machinery and specificity determinants in close proximity. Restriction endonucleases have been used as a paradigm for protein engineering, but successful cases are rare. Here, we present the results of a directed evolution approach to the engineering of a dimeric, blunt end cutting restriction enzyme NlaIV (GGN/NCC). Based on the remote similarity to EcoRV endonuclease, regions for random mutagenesis and in vitro evolution were chosen. The obtained variants cleaved target sites with an up to 100-fold kcat/KM preference for AT or TA (GGW/WCC) over GC or CG (GGS/SCC) in the central dinucleotide step, compared to the only ~ 17-fold preference of the wild-type enzyme. To understand the basis of the increased specificity, we determined the crystal structure of NlaIV. Despite the presence of DNA in the crystallization mix, the enzyme crystallized in the free form. We therefore constructed a computational model of the NlaIV–DNA complex. According to the model, the mutagenesis of the regions that were in the proximity of DNA did not lead to the desired specificity change, which was instead conveyed in an indirect manner by substitutions in the more distant regions, Published by Elsevier, Amsterdam [u.a.]

Nucleic acids symposium series 47(22), 11943–11955 (2019). doi:10.1093/nar/gkz1017, EcoKMcrA from Escherichia coli restricts CpG methylated or hydroxymethylated DNA, and may act as a barrier against host DNA. The enzyme consists of a novel N-terminal specificity domain that we term NEco, and a C-terminal catalytic HNH domain. Here, we report that NEco and full-length EcoKMcrA specificities are consistent. NEco affinity to DNA increases more from hemi- to full-methylation than from non- to hemi-methylation, indicating cooperative binding of the methyl groups. We determined the crystal structures of NEco in complex with fully modified DNA containing three variants of the Y5mCGR EcoKMcrA target sequence: C5mCGG, T5mCGA and T5hmCGA. The structures explain the specificity for the two central base pairs and one of the flanking pairs. As predicted based on earlier biochemical experiments, NEco does not flip any DNA bases. The proximal and distal methyl groups are accommodated in separate pockets. Changes to either pocket reduce DNA binding by NEco and restriction by EcoKMcrA, confirming the relevance of the crystallographically observed binding mode in solution., Published by Oxford Univ. Press8619, Oxford

Restriction-modification systems digest non-methylated invading DNA, while protecting host DNA against the endonuclease activity by methylation. It is widely believed that the methylated DNA would not 'fit' into the binding site of the endonuclease in the productive orientation, and thus steric clashes should account for most of the protection. We test this concept statistically by grafting methyl groups in silico onto non-methylated DNA in co-crystal structures with restriction endonucleases. Clash scores are significantly higher for protective than non-protective methylation (P < 0.05% according to the Wilcoxon rank sum test). Structural data alone are sufficient to distinguish between protective and non-protective DNA methylation with 90% confidence and decision thresholds of 1.1 A and 48 A(3) for the most severe distance-based and cumulative volume-based clash with the protein, respectively (0.1 A was deducted from each interatomic distance to allow for coordinate errors). The most severe clashes are more pronounced for protective methyl groups attached to the nitrogen atoms (N6-methyladenines and N4-methylcytosines) than for C5-methyl groups on cytosines. Cumulative clashes are comparable for all three types of protective methylation.

Nucleic acids symposium series 46(18), 9829 - 9841 (2018). doi:10.1093/nar/gky731, Escherichia coli McrA (EcoKMcrA) acts as a methylcytosine and hydroxymethylcytosine dependent restriction endonuclease. We present a biochemical characterization of EcoKMcrA that includes the first demonstration of its endonuclease activity, small angle X-ray scattering (SAXS) data, and a crystal structure of the enzyme in the absence of DNA. Our data indicate that EcoKMcrA dimerizes via the anticipated C-terminal HNH domains, which together form a single DNA binding site. The N-terminal domains are not homologous to SRA domains, do not interact with each other, and have separate DNA binding sites. Electrophoretic mobility shift assay (EMSA) and footprinting experiments suggest that the N-terminal domains can sense the presence and sequence context of modified cytosines. Pyrrolocytosine fluorescence data indicate no base flipping. In vitro, EcoKMcrA DNA endonuclease activity requires Mn2+ ions, is not strictly methyl dependent, and is not observed when active site variants of the enzyme are used. In cells, EcoKMcrA specifically restricts DNA that is modified in the correct sequence context. This activity is impaired by mutations of the nuclease active site, unless the enzyme is highly overexpressed., Published by Oxford Univ. Press, Oxford

R.DpnI consists of N-terminal catalytic and C-terminal winged helix domains that are separately specific for the Gm6ATC sequences in Dam-methylated DNA. Here we present a crystal structure of R.DpnI with oligoduplexes bound to the catalytic and winged helix domains and identify the catalytic domain residues that are involved in interactions with the substrate methyl groups. We show that these methyl groups in the Gm6ATC target sequence are positioned very close to each other. We further show that the presence of the two methyl groups requires a deviation from B-DNA conformation to avoid steric conflict. The methylation compatible DNA conformation is complementary with binding sites of both R.DpnI domains. This indirect readout of methylation adds to the specificity mediated by direct favorable interactions with the methyl groups and solvation/desolvation effects. We also present hydrogen/deuterium exchange data that support ‘crosstalk’ between the two domains in the identification of methylated DNA, which should further enhance R.DpnI methylation specificity.

PvuRts1I is a prototype for a larger family of restriction endonucleases that cleave DNA containing 5-hydroxymethylcytosine (5hmC) or 5-glucosylhydroxymethylcytosine (5ghmC), but not 5-methylcytosine (5mC) or cytosine. Here, we report a crystal structure of the enzyme at 2.35 A resolution. Although the protein has been crystallized in the absence of DNA, the structure is very informative. It shows that PvuRts1I consists of an N-terminal, atypical PD-(D/E)XK catalytic domain and a C-terminal SRA domain that might accommodate a flipped 5hmC or 5ghmC base. Changes to predicted catalytic residues of the PD-(D/E)XK domain or to the putative pocket for a flipped base abolish catalytic activity. Surprisingly, fluorescence changes indicative of base flipping are not observed when PvuRts1I is added to DNA substrates containing pyrrolocytosine in place of 5hmC (5ghmC). Despite this caveat, the structure suggests a model for PvuRts1I activity and presents opportunities for protein engineering to alter the enzyme properties for biotechnological applications.

The PD-(D/E)XK type II restriction endonuclease ThaI cuts the target sequence CG/CG with blunt ends. Here, we report the 1.3 Å resolution structure of the enzyme in complex with substrate DNA and a sodium or calcium ion taking the place of a catalytic magnesium ion. The structure identifies Glu54, Asp82 and Lys93 as the active site residues. This agrees with earlier bioinformatic predictions and implies that the PD and (D/E)XK motifs in the sequence are incidental. DNA recognition is very unusual: the two Met47 residues of the ThaI dimer intercalate symmetrically into the CG steps of the target sequence. They approach the DNA from the minor groove side and penetrate the base stack entirely. The DNA accommodates the intercalating residues without nucleotide flipping by a doubling of the CG step rise to twice its usual value, which is accompanied by drastic unwinding. Displacement of the Met47 side chains from the base pair midlines toward the downstream CG steps leads to large and compensating tilts of the first and second CG steps. DNA intercalation by ThaI is unlike intercalation by HincII, HinP1I or proteins that bend or repair DNA.

The beta beta alpha-Me restriction endonuclease (REase) Hpy99I recognizes the CGWCG target sequence and cleaves it with unusual stagger (five nucleotide 5'-recessed ends). Here we present the crystal structure of the specific complex of the dimeric enzyme with DNA. The Hpy99I protomer consists of an antiparallel beta-barrel and two beta 4 alpha 2 repeats. Each repeat coordinates a structural zinc ion with four cysteine thiolates in two CXXC motifs. The beta beta alpha-Me region of the second beta 4 alpha 2 repeat holds the catalytic metal ion (or its sodium surrogate) via Asp148 and Asn165 and activates a water molecule with the general base His149. In the specific complex, Hpy99I forms a ring-like structure around the DNA that contacts DNA bases on the major and minor groove sides via the first and second beta 4 alpha 2 repeats, respectively. Hpy99I interacts with the central base pair of the recognition sequence only on the minor groove side, where A:T resembles T:A and G:C is similar to C:G. The Hpy99I-DNA co-crystal structure provides the first detailed illustration of the beta beta alpha-Me site in REases and complements structural information on the use of this active site motif in other groups of endonucleases such as homing endonucleases (e.g. I-PpoI) and Holliday junction resolvases (e.g. T4 endonuclease VII).

It has long been known that most Type II restriction endonucleases share a conserved core fold and similar active-sites. The same core folding motif is also present in the MutH protein, a component of the bacterial DNA mismatch repair machinery. In contrast to most Type II restriction endonucleases, which assemble into functional dimers and catalyze double-strand breaks, MutH is a monomer and nicks hemimethylated DNA. Recent biochemical and crystallographic studies demonstrate that the restriction enzymes BcnI and MvaI share many additional features with MutH-like proteins, but not with most other restriction endonucleases. The structurally similar monomers all recognize approximately symmetric target sequences asymmetrically. Differential sensitivities to slight substrate asymmetries, which could be altered by protein engineering, determine whether the enzymes catalyze only single-strand nicks or double-strand breaks.

LytM is a Staphylococcus aureus autolysin and a homologue of the S. simulans lysostaphin. Both enzymes are members of M23 metallopeptidase family (MEROPS) comprising primarily bacterial peptidoglycan hydrolases. LytM occurs naturally in a latent form, but can be activated by cleavage of an inhibitory N-terminal proregion. Here, we present a 1.45 Å crystal structure of LytM catalytic domain with a transition state analogue, tetraglycine phosphinate, bound in the active site. In the electron density, the active site of the peptidase, the phosphinate and the “diglycine” fragment on the P1′ side of the transition state analogue are very well defined. The density is much poorer or even absent for the P1 side of the ligand. The structure is consistent with the involvement of His260 and/or His291 in the activation of the water nucleophile and suggests a possible catalytic role for Tyr204, which we confirmed by mutagenesis. Possible mechanisms of catalysis and the structural basis of substrate specificity are discussed based on the structure analysis.

A series of tetrapeptide p -nitroanilide substrates of the general formula: suc-Ala-Ala-Pro-Aaa- p -nitroanilide was used to map the S 1 binding pocket of human cathepsin G. Based on the k cat / K m parameter, the following order of preference was found: Lys=Phe>Arg=Leu>Met>Nle=Nva>Ala>Asp. Thus, the enzyme exhibits clear dual and equal trypsin- and chymotrypsin-like specificities. Particularly deleterious were β-branched side chains of Ile and Val. The P 1 substrate preferences found for cathepsin G are distinctly different from many other serine proteinases, including fiddler crab collagenase and chymotrypsin. The k cat / K m values obtained for P 1 Lys, Phe, Arg and Leu substrates correlate well with those determined for analogous P 1 mutants of basic pancreatic trypsin inhibitor (BPTI) obtained through recombinant techniques. To characterise the subsite specificity of the enzyme, a series of Cucurbita maxima trypsin inhibitor I (CMTI I) mutants were used comprising P 2 –P 3 ′ and P 12 ′ positions. All the mutants obtained were inhibitors of cathepsin G with association constants in the range: 10 5 –10 9 M −1 . Some of the mutations destabilised complex formation. In particular, Met 8 →Arg substitution at P 3 ′, which increased association constant for chymotrypsin 46-fold, led to a 7-fold decrease of binding with cathepsin G. In addition, mutation of Ile 6 at position P 1 ′ either to Val or Asp was deleterious for cathepsin G. In two cases (Ala 18 →Gly (P 12 ′) and Pro 4 →Thr (P 2 )), about a 10-fold increase in association constants was observed.

The site-selective encoding of noncanonical amino acids (NAAs) is a powerful technique for the installation of novel chemical functional groups in proteins. This is often achieved by recoding a stop codon and requires two additional components: an evolved aminoacyl tRNA synthetase (AARS) and a cognate tRNA. Analysis of the most successful AARSs reveals common characteristics. The highest fidelity NAA systems derived from the Methanocaldococcus jannaschii tyrosyl AARS feature specific mutations to two residues reported to interact with the hydroxyl group of the substrate tyrosine. We demonstrate that the restoration of just one of these determinants for amino acid specificity results in the loss of fidelity as the evolved AARSs become noticeably promiscuous. These results offer a partial explanation of a recently retracted strategy for the synthesis of glycoproteins. Similarly, we reinvestigated a tryptophanyl AARS reported to allow the site-selective incorporation of 5-hydroxy tryptophan within mammalian cells. In multiple experiments, the enzyme displayed elements of promiscuity despite its previous characterization as a high fidelity enzyme. Given the many similarities of the TyrRSs and TrpRSs reevaluated here, our findings can be largely combined, and in doing so they reinforce the long-established central dogma regarding the molecular basis by which these enzymes contribute to the fidelity of translation. Thus, our view is that the central claims of fidelity reported in several NAA systems remain unproven and unprecedented.

The GIY-YIG nuclease domain is present in all kingdoms of life and has diverse functions. It is found in the eukaryotic flap endonuclease and Holliday junction resolvase Slx1-Slx4, the prokaryotic nucleotide excision repair proteins UvrC and Cho, and in proteins of 'selfish' genetic elements. Here we present the structures of the ternary pre- and post-cleavage complexes of the type II GIY-YIG restriction endonuclease Hpy188I with DNA and a surrogate or catalytic metal ion, respectively. Our structures suggest that GIY-YIG nucleases catalyze DNA hydrolysis by a single substitution reaction. They are consistent with a previous proposal that a tyrosine residue (which we expect to occur in its phenolate form) acts as a general base for the attacking water molecule. In contrast to the earlier proposal, our data identify the general base with the GIY and not the YIG tyrosine. A conserved glutamate residue (Glu149 provided in trans in Hpy188I) anchors a single metal cation in the active site. This metal ion contacts the phosphate proS oxygen atom and the leaving group 3'-oxygen atom, presumably to facilitate its departure. Taken together, our data reveal striking analogy in the absence of homology between GIY-YIG and ββα-Me nucleases.

Low molecular mass purine nucleoside phosphorylases (PNPs, E.C. 2.4.2.1) are homotrimeric enzymes that are tightly inhibited by immucillins. Due to the positive charge on the ribose like part (iminoribitol moiety) and protonation of the N7 atom of the purine ring, immucillins are believed to act as transition state analogues. Over a wide range of concentrations, immucillins bind with strong negative cooperativity to PNPs, so that only every third binding site of the enzyme is occupied (third-of-the-sites binding). 9-(5',5'-difluoro-5'-phosphonopentyl)-9-deazaguanine (DFPP-DG) shares with immucillins the protonation of the N7, but not the positive charge on the ribose like part of the molecule. We have previously shown that DFPP-DG interacts with PNPs with subnanomolar inhibition constant. Here, we report additional biochemical experiments to demonstrate that the inhibitor can be bound with the same K(d) ( approximately 190pM) to all three substrate binding sites of the trimeric PNP, and a crystal structure of PNP in complex with DFPP-DG at 1.45A resolution, the highest resolution published for PNPs so far. The crystals contain the full PNP homotrimer in the asymmetric unit. DFPP-DG molecules are bound in superimposable manner and with full occupancies to all three PNP subunits. Thus the postulated third-of-the-sites binding of immucillins should be rather attribute to the second feature of the transition state, ribooxocarbenium ion character of the ligand or to the coexistence of both features characteristic for the transition state. The DFPP-DG/PNP complex structure confirms the earlier observations, that the loop from Pro57 to Gly66 covering the phosphate-binding site cannot be stabilized by phosphonate analogues. The loop from Glu250 to Gln266 covering the base-binding site is organized by the interactions of Asn243 with the Hoogsteen edge of the purine base of analogues bearing one feature of the postulated transition state (protonated N7 position).

PspGI is a representative of a group of restriction endonucleases that recognize a pentameric sequence related to CCNGG. Unlike the previously investigated Ecl18kI, which does not have any specificity for the central base pair, PspGI prefers A/T over G/C in its target site. Here, we present a structure of PspGI with target DNA at 1.7 A resolution. In this structure, the bases at the center of the recognition sequence are extruded from the DNA and flipped into pockets of PspGI. The flipped thymine is in the usual anti conformation, but the flipped adenine takes the normally unfavorable syn conformation. The results of this and the accompanying manuscript attribute the preference for A/T pairs over G/C pairs in the flipping position to the intrinsically lower penalty for flipping A/T pairs and to selection of the PspGI pockets against guanine and cytosine. Our data show that flipping can contribute to the discrimination between normal bases. This adds a new role to base flipping in addition to its well-known function in base modification and DNA damage repair.

Restriction endonuclease BcnI cleaves duplex DNA containing the sequence CC/SGG (S stands for C or G, / designates a cleavage position) to generate staggered products with single nucleotide 5'-overhangs. Here, we show that BcnI functions as a monomer that interacts with its target DNA in 1:1 molar ratio and report crystal structures of BcnI in the absence and in the presence of DNA. In the complex with DNA, BcnI makes specific contacts with all five bases of the target sequence and not just with a half-site, as the protomer of a typical dimeric restriction endonuclease. Our data are inconsistent with BcnI dimerization and suggest that the enzyme introduces double-strand breaks by sequentially nicking individual DNA strands, although this remains to be confirmed by kinetic experiments. BcnI is remotely similar to the DNA repair protein MutH and shares approximately 20% sequence identity with the restriction endonuclease MvaI, which is specific for the related sequence CC/WGG (W stands for A or T). As expected, BcnI is structurally similar to MvaI and recognizes conserved bases in the target sequence similarly but not identically. BcnI has a unique machinery for the recognition of the central base-pair.

Restriction endonuclease MvaI recognizes the sequence CC/WGG (W stands for A or T, '/' designates the cleavage site) and generates products with single nucleotide 5'-overhangs. The enzyme has been noted for its tolerance towards DNA modifications. Here, we report a biochemical characterization and crystal structures of MvaI in an apo-form and in a complex with target DNA at 1.5 A resolution. Our results show that MvaI is a monomer and recognizes its pseudosymmetric target sequence asymmetrically. The enzyme consists of two lobes. The catalytic lobe anchors the active site residues Glu36, Asp50, Glu55 and Lys57 and contacts the bases from the minor grove side. The recognition lobe mediates all major grove interactions with the bases. The enzyme in the crystal is bound to the strand with T at the center of the recognition sequence. The crystal structure with calcium ions and DNA mimics the prereactive state. MvaI shows structural similarities to BcnI, which cleaves the related sequence CC/SGG and to MutH enzyme, which is a component of the DNA repair machinery, and nicks one DNA strand instead of making a double-strand break.

Restricion endonuclease Ecl18kI is specific for the sequence /CCNGG and cleaves it before the outer C to generate 5 nt 5′-overhangs. It has been suggested that Ecl18kI is evolutionarily related to NgoMIV, a 6-bp cutter that cleaves the sequence G/CCGGC and leaves 4 nt 5′-overhangs. Here, we report the crystal structure of the Ecl18kI–DNA complex at 1.7 A resolution and compare it with the known structure of the NgoMIV–DNA complex. We find that Ecl18kI flips both central nucleotides within the CCNGG sequence and buries the extruded bases in pockets within the protein. Nucleotide flipping disrupts Watson–Crick base pairing, induces a kink in the DNA and shifts the DNA register by 1 bp, making the distances between scissile phosphates in the Ecl18kI and NgoMIV cocrystal structures nearly identical. Therefore, the two enzymes can use a conserved DNA recognition module, yet recognize different sequences, and form superimposable dimers, yet generate different cleavage patterns. Hence, Ecl18kI is the first example of a restriction endonuclease that flips nucleotides to achieve specificity for its recognition site.

Structural investigations into quadruplexes, formed from guanine rich sequences found in the human chromosome, and development of small molecule ligands that can stabilize these structures have been the main focus of our research. The targeting of the telomere, a G-rich hexanucleotide repeat sequence TAGGGT, found at the end of all mammalian chromosomes has lead to the development of a series of ligands that can effectively inhibit the enzyme telomerase. Critically a 3’ single stranded overhang, located at the end of the chromosome, is the substrate for the ribonucleoprotein telomerase. Folding of these telomeric repeats into higher order DNA structures inhibits access of telomerase and prevents telomere maintenance. Additionally, the binding of other associated proteins that form part of the telosome, are also disrupted leading rapidly to chromosomal damage, such as chromosomal fusions and apoptosis.Wewill discuss our research into the folding topologies that are available to these telomeric sequences and other target sequence found in the chromosome, and the design of selective ligands that target these novel topologies. This is of particular importance with the identification of quanine rich sequences that have been shown to form stable quadruplex structures within promotor regions of onogenes. The structural diversity of the quadruplex structures formed from these G-rich DNA sequences will also be reviewed. m04.o03

We report our progress in understanding the structure-function relationship of the interaction between protein inhibitors and several serine proteases. Recently, we have determined high resolution solution structures of two inhibitors Apis mellifera chymotrypsin inhibitor-1 (AMCI-I) and Linum usitatissimum trypsin inhibitor (LUTI) in the free state and an ultra high resolution X-ray structure of BPTI. All three inhibitors, despite totally different scaffolds, contain a solvent exposed loop of similar conformation which is highly complementary to the enzyme active site. Isothermal calo- rimetry data show that the interaction between wild type BPTI and chymotrypsin is entropy driven and that the enthalpy component opposes complex formation. Our research is focused on extensive mutagenesis of the four positions from the protease binding loop of BPTI: P1, P1', P3, and P4. We mutated these residues to different amino acids and the variants were characterized by determination of the association constants, stability parameters and crystal structures of protease-inhibitor complexes. Accommodation of the P1 residue in the S1 pocket of four proteases: chymotrypsin, trypsin, neutrophil elastase and cathepsin G was probed with 18 P1 variants. High resolution X-ray structures of ten complexes between bovine trypsin and P1 variants of BPTI have been determined and compared with the cognate P1 Lys side chain. Mutations of the wild type Ala16 (P1') to larger side chains always caused a drop of the association constant. According to the crystal structure of the Leu16 BPTI-trypsin complex, introduction of the larger residue at the P1' position leads to steric conflicts in the vicinity of the mutation. Finally, mutations at the P4 site allowed an improvement of the association with several serine proteases involved in blood clotting. Conversely, introduction of Ser, Val, and Phe in place of Gly12 (P4) had invariably a destabilizing effect on the complex with these proteases.