Structural and functional analogy between pneumolysin and proaerolysin

Pneumolysin and proaerolysin are bacterial toxins that form pores in host cells by oligomerization. We propose that they may have similar structures despite a poor sequence identity. The crystal structure of proaerolysin reveals a protein composed of four domains, arranged in the shape of an elongated comma. Electron microscopy of the pneumolysin monomer shows a similar arrangement of domains. The sequence of pneumolysin recognizes the template of proaerolysin from a library of protein folds. A three-dimensional model of pneumolysin has been constructed by the comparative approach using the structure of proaerolysin. This model, together with results on the activity of site- specific mutants and the positions of antigenic sites, has been used to propose functional roles of individual domains.      

Stereochemical modeling of disulfide bridges. Criteria for introduction into proteins by site-directed mutagenesis

A computer modeling procedure for assessing the stereochemicalsuitability of pairs of residues in proteins as potential sitesfor introduction of cystine disulfide crosslinks has been developed.Residue pairs with C – C distances of 6.5 Å andC^beta;–C^ß distances of 4.5 Å are chosenfor geometrical fixation of S atoms using the program MODIP.The stereochemistry of the modeled disulfides is evaluated usinglimits for the structural parameters of the various torsionangles and S–S bond length in the disulfide bridge. Theability of the procedure to correctly model disulfides has beenchecked with examples of cystine peptides of known crystal structuresand 103 disulfide bridges from 25 available protein crystalstructures determined at 2 Å resolution. An analysis ofresults on three proteins with engineered disulfides, T4 lysozyme,dihydrofolate reductase and subtilisin, is presented. Two positionsfor the introduction of ‘stereochemically optimal’disulfides are identified in subtilisin.     

Modelling multiple disulphide loop containing polypeptides by random conformation generation. The test cases of {alpha}-conotoxin GI and edothelin I

A general procedure for arriving at 3-D models of disulphidericholypeptide systems based on the covalent cross-link constraintshas been developed. The procedure, which has been coded as acomputer program, RANMOD, assigns a large number of random,permitted backbone conformations to the polypeptide and identifiesstereochemically acceptable structures as plausible models basedon strainless disulphide bridge modelling. Disulphide bond modellingis performed using the procedure MODIP developed earlier, inconnection with the choice of suitable sites where disulphidebonds could be engineered in proteins (Sowdhamini,R., Srinivasan,N.,Shoichet,B., Santi,D.V., Ramakrishnan,C. and Balaram,P. (1989)Protein Engng, 3, 95-103). The method RANMOD has been testedon small disulphide loops and the structures compared againstpreferred backbone conformations derived from an analysis ofputative disulphide subdatabase and model calculations. RANMODhas been applied to disulphiderich peptides and found to giverise to several stereochemically acceptable structures. Theresults obtained on the modelling of two test cases, a-conotoxinGI and endothelin I, are presented. Available NMR data suggestthat such small systems exhibit conformational heterogeneityin solution. Hence, this approach for obtaining several distinctmodels is particularly attractive for the study of conformationalexcursions.     

Functional sites and evolutionary connections of acylhomoserine lactone synthases

Acylhomoserine lactone (AHL) synthases act as chemical communicationsignals or pheromones in Gram-negative bacteria and regulatediverse physiological events in a cell density-dependent manner.The recent crystal structure determination of EsaI, a key enzymein this pathway, shows that the AHL synthase superfamily membersadopt the fold of the N-acetyltransferase superfamily. We suggest,by the identification of intermediate sequences, that the twosuperfamilies are evolutionarily related. Evolutionary traceanalyses of aligned sequences and docking studies have beenused to discuss functionally important residues of EsaI homologues.     

Insights into the structure of hepatocyte growth factor/scatter factor (HGF/SF) and implications for receptor activation

The modular structure of HGF/SF offers a reductionist or 'divide and rule' approach to the analysis of structure and function. Domain deletion experiments have established that the N domain, kringle 1 and kringle 2 are essential for HGF/SF activity and that truncated variants containing the N domain and kringle 1 (NK1) or kringles 1 and 2 (NK2) can exhibit partial agonistic or antagonistic activity depending on target cells. Comparative modelling has been used to predict the 3D structures of the six domains of HGF/SF. More recently, NMR methods have shown that the N domain has a novel fold, the charge distribution of which suggests a heparin binding site. Crystals of NK1 indicate the relationship of this domain to the kringle 1, offering further insights into the mechanism of domain interactions and receptor activation.     

SMoS: a database of structural motifs of protein superfamilies

The Structural Motifs of Superfamilies (SMoS) database providesinformation about the structural motifs of aligned protein domainsuperfamilies. Such motifs among structurally aligned multiplemembers of protein superfamilies are recognized by the conservationof amino acid preference and solvent inaccessibility and areexamined for the conservation of other features like secondarystructural content, hydrogen bonding, non-polar interactionand residue packing. These motifs, along with their sequenceand spatial orientation, represent the conserved core structureof each superfamily and also provide the minimal requirementof sequence and structural information to retain each superfamilyfold. Received April 25, 2003; revised September 9, 2003; accepted September 24, 2003.     

The recognition of protein structure and function from sequence: adding value to genome data

The explosion of DNA sequence data from genome projects presents many challenges. For instance, we must extend our current knowledge of protein structure and function so that it can be applied to these new sequences. The derivation of rules for the relationships between sequence and structure allow us to recognize a common fold by the use of tertiary templates. New techniques enable us to begin to meet the challenge of rule-based modelling of distantly related proteins. This paper describes an integrated and knowledge-based approach to the prediction of protein structure and function which can maximize the value of sequence information.