The refined crystal structure of a fully active semisynthetic ribonuclease at 1.8-A resolution

A new formulation of protein evolutionary models that account for structural constraints

Despite the importance of a thermodynamically stable structure with a conserved fold for protein function, almost all evolutionary models neglect site-site correlations that arise from physical interactions between neighboring amino acid sites. This is mainly due to the difficulty in formulating a computationally tractable model since rate matrices can no longer be used. Here, we introduce a general framework, based on factor graphs, for constructing probabilistic models of protein evolution with site interdependence. Conveniently, efficient approximate inference algorithms, such as Belief Propagation, can be used to calculate likelihoods for these models. We fit an amino acid substitution model of this type that accounts for both solvent accessibility and site-site correlations. Comparisons of the new model with rate matrix models and alternative structure-dependent models demonstrate that it better fits the sequence data. We also examine evolution within a family of homohexameric enzymes and find that site-site correlations between most contacting subunits contribute to a higher likelihood. In addition, we show that the new substitution model has a similar mathematical form to the one introduced in Rodrigue et al. (Rodrigue N, Lartillot N, Bryant D, Philippe H. 2005. Site interdependence attributed to tertiary structure in amino acid sequence evolution. Gene 347:207-217), although with different parameter interpretations and values. We also perform a statistical analysis of the effects of amino acids at neighboring sites on substitution probabilities and find a significant perturbation of most probabilities, further supporting the significant role of site-site interactions in protein evolution and motivating the development of new evolutionary models similar to the one described here. Finally, we discuss possible extensions and applications of the new substitution model.

Crystal structures of murine angiogenin-2 and -3-probing 'structure--function' relationships amongst angiogenin homologues

Angiogenin (Ang) is a potent inducer of neovascularization. Point mutations in human Ang have been linked to cancer progression and two neurodegenerative diseases: amyotrophic lateral sclerosis and Parkinson's disease. Intensive structural and functional analyses of Ang have been paramount in assigning functions to this novel homologue of bovine pancreatic RNase A. However, inhibitor-binding studies with crystalline Ang (for designing potential anti-cancer drugs) have been hampered as a result of the inaccessibility of the active site. Experiments with the murine homologues of Ang have not only overcome the obvious practical limitations encountered when studying the role of a human protein in healthy individuals, but also the crystal structures of murine angiogenins (mAng and mAng-4) have revealed themselves to have greater potential for the visualization of small-molecule inhibitor binding at the active site. In the present study, we report the crystal structures of two more murine Ang paralogues, mAng-2 and mAng-3, at 1.6 and 1.8 Å resolution, respectively. These constitute the first crystal structures of an Ang with a zinc ion bound at the active site and provide some insight into the possible mode of inhibition of the ribonucleolytic activity of the enzyme by these divalent cations. Both structures show that the residues forming the putative P₁, B₁ and B₂ subsites occupy positions similar to their counterparts in human Ang and are likely to have conserved roles. However, a less obtrusive conformation of the C-terminal segment in mAng-3 and the presence of a sulfate ion in the B₁ subsite of mAng-2 suggest that these proteins have the potential to be used for inhibitor-binding studies. We also discuss the biological relevance of the structural similarities and differences between the different Ang homologues.

Crystal structure of RNase A tandem enzymes and their interaction with the cytosolic ribonuclease inhibitor

Because of their ability to degrade RNA, RNases are potent cytotoxins. The cytotoxic activity of most members of the RNase A superfamily, however, is abolished by the cytosolic ribonuclease inhibitor (RI). RNase A tandem enzymes, in which two RNase A molecules are artificially connected by a peptide linker, and thus have a pseudodimeric structure, exhibit remarkable cytotoxic activity. In vitro, however, these enzymes are still inhibited by RI. Here, we present the crystal structures of three tandem enzymes with the linker sequences GPPG, SGSGSG, and SGRSGRSG, which allowed us to analyze the mode of binding of RI to the RNase A tandem enzymes. Modeling studies with the crystal structures of the RI-RNase A complex and the SGRSGRSG-RNase A tandem enzyme as templates suggested a 1 : 1 binding stoichiometry for the RI-RNase A tandem enzyme complex, with binding of the RI molecule to the N-terminal RNase A entity. These results were experimentally verified by analytical ultracentrifugation, quantitative electrophoresis, and proteolysis studies with trypsin. As other dimeric RNases, which are comparably cytotoxic, either evade RI binding or potentially even bind two RI molecules, inactivation by RI cannot be the crucial limitation to the cytotoxicity of dimeric RNases.

Conformational strictness required for maximum activity and stability of bovine pancreatic ribonuclease A as revealed by crystallographic study of three Phe120 mutants at 1.4 A resolution

The replacement of Phe120 with other hydrophobic residues causes a decrease in the activity and thermal stability in ribonuclease A (RNase A). To explain this, the crystal structures of wild-type RNase A and three mutants--F120A, F120G, and F120W--were analyzed up to a 1.4 A resolution. Although the overall backbone structures of all mutant samples were nearly the same as that of wild-type RNase A, except for the C-terminal region of F120G with a high B-factor, two local conformational changes were observed at His119 in the mutants. First, His119 of the wild-type and F120W RNase A adopted an A position, whereas those of F120A and F120G adopted a B position, but the static crystallographic position did not reflect either the efficiency of transphosphorylation or the hydrolysis reaction. Second, His119 imidazole rings of all mutant enzymes were deviated from that of wild-type RNase A, and those of F120W and F120G appeared to be "inside out" compared with that of wild-type RNase A. Only approximately 1 A change in the distance between N(epsilon2) of His12 and N(delta1) of His119 causes a drastic decrease in k(cat), indicating that the active site requires the strict positioning of the catalytic residues. A good correlation between the change in total accessible surface area of the pockets on the surface of the mutant enzymes and enthalpy change in their thermal denaturation also indicates that the effects caused by the replacements are not localized but extend to remote regions of the protein molecule.

The ultrahigh resolution crystal structure of ribonuclease A containing an isoaspartyl residue: hydration and sterochemical analysis

Crystals of the deamidated form of bovine pancreatic ribonuclease which contains an isoaspartyl residue in position 67 diffract to 0. 87 A at 100 K. We have refined the crystallographic model using anisotropic displacement parameters for all atoms to a conventional crystallographic residual R=0.101 for all observed reflections in the resolution range 61.0-0.87 A. The ratio observations/parameters is 7.2 for the final model. This structure represents one of the highest resolution protein structures to date and interestingly, it is the only example containing more than one molecule in the asymmetric unit with a resolution better than 1.0 A. The non-crystallographic symmetry has been used as a validation check of the geometrical parameters and it has allowed an estimate for an upper limit of errors associated with this high resolution model. In the present structure it was possible to obtain a more accurate picture of the active site whose electron density was not clearly interpretable in the previous 1.9 A resolution structure. In particular, the P1 site is alternatively occupied either by a sulphate anion or by a water molecule network. Most of hydrogen atoms were visible in the electron density maps, including those involved in C(alpha)-H(alpha).O interactions. Analysis of protein-solvent interactions has revealed the occurrence of an extensive cluster of water molecules, predominantly arranged in pentagonal fused rings and surrounding hydrophobic moiety of side-chains. Finally, in spite of the limited sample of residues, we have detected a clear dependence of backbone N-C(alpha)-C angle on residue conformation. This correlation can be fruitfully used as a valuable tool in protein structure validation.

Protein titration in the crystal state

Proteins are complex structures whose overall stability critically depends on a delicate balance of numerous interactions of similar strength, which are markedly influenced by their environment. Here, we present an analysis of the effect of pH on a protein structure in the crystalline state using RNase A as a model system. By altering only one physico-chemical parameter in a controlled manner, we are able to quantify the structural changes induced in the protein. Atomic resolution X-ray diffraction data were collected for crystals at six pH* values ranging from 5.2 to 8.8, and the six independently refined structures reveal subtle, albeit well-defined variations directly related to the pH titration of the protein. The deprotonation of the catalytic His12 residue is clearly evident in the electron density maps, confirming the reaction mechanism proposed by earlier enzymatic and structural studies. The concerted structural changes observed in the regions remote from the active-site point to an adaptation of the protein structure to the changes in the physico-chemical environment. Analysis of the stereochemistry of the six structures provided accurate estimates of p Kavalues of most of the histidine residues. This study gives further evidence for the advantage of atomic resolution X-ray crystallographic analyses for revealing small but significant structural changes which provide clues to the function of a biological macromolecule.

Crystal structures of two mutants that have implications for the folding of bovine pancreatic ribonuclease A

The Tyr92-Pro93 peptide group of bovine pancreatic ribonuclease A (RNase A) exists in the cis conformation in the native state. From unfolding/refolding kinetic studies of the disulfide-intact wild-type protein and of a variant in which Pro93 had been replaced by Ala, it had been suggested that the Tyr92-Ala93 peptide group also exists in the cis conformation in the native state. Here, we report the crystal structure of the P93A variant. Although there is disorder in the region of residues 92 and 93, the best structural model contains a cis peptide at this position, lending support to the results of the kinetics experiments. We also report the crystal structure of the C[40, 95]A variant, which is an analog of the major rate-determining three-disulfide intermediate in the oxidative folding of RNase A, missing the 40-95 disulfide bond. As had been detected by NMR spectroscopy, the crystal structure of this analog shows disorder in the region surrounding the missing disulfide. However, the global chain fold of the remainder of the protein, including the disulfide bond between Cys65 and Cys72, appears to be unaffected by the mutation.

The structures of RNase A complexed with 3'-CMP and d(CpA): active site conformation and conserved water molecules

The interactions of RNase A with cytidine 3'-monophosphate (3'-CMP) and deoxycytidyl-3',5'-deoxyadenosine (d(CpA)) were analyzed by X-ray crystallography. The 3'-CMP complex and the native structure were determined from trigonal crystals, and the d(CpA) complex from monoclinic crystals. The differences between the overall structures are concentrated in loop regions and are relatively small. The protein-inhibitor contacts are interpreted in terms of the catalytic mechanism. The general base His 12 interacts with the 2' oxygen, as does the electrostatic catalyst Lys 41. The general acid His 119 has 2 conformations (A and B) in the native structure and is found in, respectively, the A and the B conformation in the d(CpA) and the 3'-CMP complex. From the present structures and from a comparison with RNase T1, we propose that His 119 is active in the A conformation. The structure of the d(CpA) complex permits a detailed analysis of the downstream binding site, which includes His 119 and Asn 71. The comparison of the present RNase A structures with an inhibitor complex of RNase T1 shows that there are important similarities in the active sites of these 2 enzymes, despite the absence of any sequence homology. The water molecules were analyzed in order to identify conserved water sites. Seventeen water sites were found to be conserved in RNase A structures from 5 different space groups. It is proposed that 7 of those water molecules play a role in the binding of the N-terminal helix to the rest of the protein and in the stabilization of the active site.

The occupancy of two distinct conformations by active-site histidine-119 in crystals of ribonuclease is modulated by pH

Structures of a semisynthetic RNase have been obtained to a resolution of 2.0 A at pH values of 5.2, 6.5, 7.5, and 8.8, respectively. The principle structural transformation occurring over this pH range is the conversion of the side chain of active site residue His-119 from one conformation (chi 1 = -43 degrees to -57 degrees) at low pH to another (chi 1 = +159 degrees to +168 degrees) at higher pH values. On the basis of this observation, a model is proposed that reconciles the disparate pK values for His-119 in the enzyme-substrate complex that have been deduced from kinetic studies and from proton NMR measurements in the presence of pseudosubstrates.

Structural investigation of catalytically modified F120L and F120Y semisynthetic ribonucleases

The structures of two catalytically modified semisynthetic RNases obtained by replacing phenylalanine 120 with leucine and tyrosine have been determined and refined at a resolution of 2.0 A (R = 0.161 and 0.184, respectively). These structures have been compared with the refined 1.8-A structure (R = 0.204) of the fully active phenylalanine-containing enzyme (Martin PD, Doscher MS, Edwards BFP, 1987, J Biol Chem 262:15930-15938) and with the catalytically defective D121A (2.0 A, R = 0.172) and D121N (2.0 A, R = 0.186) analogs (deMel VSJ, Martin PD, Doscher MS, Edwards BFP, 1992, J Biol Chem 267:247-256). The movement away from the active site of the loop containing residues 65-72 is seen in all three catalytically defective analogs--F120L, D121A, and D121N--but not in the fully active (or hyperactive) F120Y. The insertion of the phenolic hydroxyl of Tyr 120 into a hydrogen-bonding network involving the hydroxyl group of Ser 123 and a water molecule in F120Y is the likely basis for the hyperactivity toward uridine 2',3'-cyclic phosphate previously found for this analog (Hodges RS, Merrifield RB, 1974, Int J Pept Protein Res 6:397-405) as well as the threefold increase in KM for cytidine 2',3'-cyclic phosphate found for this analog by ourselves.

Crystal packing in six crystal forms of pancreatic ribonuclease

We compare the molecular packing of bovine pancreatic ribonuclease A (RNase A) in six crystal forms, two grown with alcohol, three with high salt and one with polyethylene glycol as a precipitant. The six packings differ in the number of molecules in contact and in the extent of the contacts, which bury 1570 A2 to 2790 A2 of the RNase surface. Regions of the protein surface involved in the six packings cover almost the whole RNase molecule. The abundance of polar interactions, about one per 200 A2, is the same in all types of precipitants. All molecule-to-molecule contacts are different in the six crystal forms, except for the one that forms a RNase dimer. The dimer has a large interface covering 1800 A2 and eight to ten polar interactions. Its presence in the three salt-grown crystal forms suggests that it is an intermediate in salt induced crystallization. In contrast, the two alcohol-grown forms contain only small interfaces, implying a different mechanism of nucleation.