A novel method for the identification of correlated pairs in aligned homologous protein sequences is presented and evaluated against a model of simulated protein evolution incorporating covariation. Our method is shown to be capable of identifying all coevolutionary pairs of sites, with minimal interference by background correlations, in aligned sequence sets containing approximately 60 sequences with a tree depth of at least 30 accepted point mutations. This result is expected even in the presence of a large degree of neutral and non-correlated evolution. It is postulated that, since naturally occurring protein families may be subject to stronger selection pressures and a lesser degree of neutral evolution, this method of covariation analysis may be generally more robust than the model would indicate.

Correlations between amino-acid residues can be observed in sets of aligned protein sequences, and the analysis of their statistical and evolutionary significance and distribution has been thoroughly investigated. In this paper, we present a model based on such covariations in protein sequences in which the pairs of residues that have mutual influence combine to produce a system analogous to a Hopfield neural network. The emergent properties of such a network, such as soft failure and the connection between network architecture and stored memory, have close parallels in known proteins. This model suggests that an explanation for observed characters of proteins such as the diminution of function by substitutions distant from the active site, the existence of protein folds (superfolds) that can perform several functions based on one architecture, and structural and functional resilience to destabilizing substitutions might derive from their inherent network-like structure. This model may also provide a basis for mapping the relationship between structure, function and evolutionary history of a protein family, and thus be a powerful tool for rational engineering.

The structural and functional evolution of the Kunitz/bovine pancreatic trypsin inhibitor (BPTI) family of proteins, which includes serine proteinase inhibitors and potassium channel blockers, was analysed with the evolutionary trace method. This method highlights sites in aligned primary sequences whose side-chain variation can be strongly linked with the past development of different functional classes or subgroups within the family. A total of 16 such "class-specific" positions distributed throughout the molecular fold were identified. On average, the side-chain chemistry at these positions had been more conserved and made greater contribution to molecular stability than the side-chain chemistry at remaining sites of variation. It was possible to use these 16 positions to describe the division of the Kunitz/BPTI family into general functional classes. According to known complexes of inhibitor variants with serine proteinases, only two of the 16 class-specific positions appear to be directly involved in intermolecular recognition via the "antiproteinase site". Instead, from various critical locations in the fold, the remainder seem to have been associated with various degrees of intramolecular conformational adjustment to the underlying framework of the antiproteinase site. It is, therefore, implied that functional diversification in this family has been founded upon both sustained evolutionary selection and conformational adjustment. The findings are important for protein engineers wishing to alter the binding selectivity of these molecules, because it appears that the issue of target recognition is dependent on the conformation of the chain segment to which the interactive side-chains are attached. To retarget members of this family towards potentially novel peptide binding sites, substitutions at certain structurally significant class-specific positions could be a good starting point.

Dendrotoxins are a family of small proteins isolated from mamba (Dendroaspis) snake venoms. The toxins contain 57-60 amino acid residues cross-linked by three disulphide bridges. They are homologous to Kunitz-type serine proteinase inhibitors, such as aprotinin, (BPTI) although they have little anti-proteinase activity. The dendrotoxins block some subtypes of voltage-dependent potassium channels in neurones. Studies with cloned K+ channels indicate that alpha-dendrotoxin from green mamba Dendroaspis angusticeps blocks Kv1.1, Kv1.2 and Kv1.6 channels in the nanomolar range. Engineered and modified versions of dendrotoxin are being investigated to define the interactive site and account for differences in K+ channel selectivity.

Mechanisms for the ring conversion of GTP into cofactor precursors by GTP cyclohydrolases and molybdopterin synthase are proposed and discussed in the context of the crystal structure of the GTP cyclohydrolase I - 2′-deoxy-GTP complex. The mechanisms suggest that common features of acid-base catalysis may underly the reactions catalysed in all three cases.

The Kunitz BPTI proteinase inhibitor family is divisible into subgroups based on source and bioactivity. Variants from snake venoms are of special interest because some show only weak inhibitory activity against the common proteinases while others are neurotoxic. We analysed the sequences for each subgrouping in the context of the common chain fold to predict the 3D location of interactive sites. The method used was an enhanced from of the previously devised 'regiovariation analysis.' This revealed the foci in 3D of amino acid side chain conservation in each subgroup. Locally high levels of side-chain conservation extending substantially in three dimensions can be associated more with the preservation of function than conformation, hence the foci probably reveal the most functionally relevant sites. For the inhibitor variants that do not originate from snake venom, regiovariation analysis gave an exact prediction of the antiproteinase site revealed by X-ray crystallography of inhibitor-enzyme complexes. However, this site is not the principal focus of evolutionary conservation in the inhibitors from snake venom, and other areas of the molecular surface are more prominent. The neurotoxic variants from snake venom (the dendrotoxins) have the principal focus of conservation near their C-terminal region, so this may be the origin of their special properties.

The synonym quotas within the genetic code for the 20 common amino acids are examined in relation to the ways in which these amino acids can be marshalled into different sets on the basis of shared physico-chemical properties. This reveals which shared properties are encouraged or discouraged during the course of protein evolution by the arrangement of the code. A dominant theme is that the synonym quotas are allocated in favour of small and chemically uncomplicated residues, and to the disadvantage of large and chemically prominent ones. From amongst the various measurements that can be considered to quantitatively express aspects of amino acid residue “size and complexity” (e.g. side chain volume, bulkness and formula weight), formula weight has the highest correlation with the synonym quota for each amino acid. However, the correlation is weak. A specially derived “size/complexity” scale for the amino acids based on their relative atomic composition improved the correlation only marginally. The existence of another weak correlation between the synonym quotas and the general amino acid composition of proteins prompted an investigation of the correlations between this composition and the previously considered amino acid properties. Again, the highest correlations are with amino acid formula weight and “size/complexity”, but in this instance the correlations are high enough to be truly significant. It is suggested that the biased synonym quotas in the genetic code are intended to ensure that proteins as a whole maintain a certain amino acid composition, even to the extent that the quotasinclude compensatory biasesto counter opposing influences upon this composition caused by the processes of natural selection for protein function. It is the need for these compensatory biases that prevents a simple correlation between the quotas and measures of amino acid complexity. The final outcome, in which amino acidsaredeployed in functional proteins in approximate proportion to their chemical complexity may serve both as a means of minimising the negative consequences of random genetic mutation (by reducing the chance appearance of the more “disruptive” types of side chain in proteins) and as a means of ensuring the most economic use of biosynthetic resources. According to this reasoning, the code is not a “frozen accident”; it is universally appropriate because it provides the best compromise that can be achieved between biosynthetic cost and biological return in respect of the rate of protein evolution.

The pentapeptide methionine enkephalin is readily hydrolysed by the oligopeptidase activity contained in Taiwan cobra (Naja naja atra) venom. It is a significantly better substrate than the peptides previously used to identify the presence of this enzyme, but it retains many of the sequence characteristics shared by these other peptides. Analysis of the manner of hydrolysis by means of high-performance liquid chromatography and electrospray mass spectrometry revealed the simultaneous actions of at least two types of oligopeptidase on the neuropeptide, producing two routes of initial breakdown. By one route, an endopeptidase cleaved the Gly-Phe bond of enkephalin first to release Tyr-Gly-Gly and Phe-Met. By the other route, an aminopeptidase was able to release Tyr and Gly-Gly-Phe-Met by cleaving the Tyr-Gly bond first. From amongst the various peptide fragments produced, Tyr-Gly-Gly was subject to immediate aminopeptidase action to release Tyr and Gly-Gly. The free Tyr produced in these reactions was in turn quickly transformed by the L-amino acid oxidase in the venom. The kinetic qualities of the enkephalin hydrolysis, and the conversions of the fragments Tyr-Gly-Gly and Tyr, were measured. Methionine enkephalin has potential as a routine assay for venom oligopeptidases, either in testing the venoms from other species or in attempts to purify these enzymes. Moreover, the ease of hydrolysis of this bioactive peptide, coupled with the revelation of the other enzymic steps involving the fragments generated, may provide important clues as to the possible role of the oligopeptidases (and L-amino acid oxidase) in the venom.

A simple methodology is described to apply to aligned protein sequence sets for which at least one representative 3-D C alpha structure is known. The evolutionary variation observed at each residue position in the sequence alignment is qualified by taking into account the residue variation that has occurred at other positions located within 7 A (according to the probable chain fold). This expresses the evolutionary behaviour of any residue position in the more appropriate context of its immediate surroundings and distinguishes between invariant residues on the basis of the variation of their environment. The highest mechanistic significance is attached to conserved residues in conserved surroundings, but the quantitative nature of the analysis means that all residue vicinities can be ranked and merged according to the degree of conservation that they exhibit and the residue positions that comprise them. Therefore, with the aid of the chain fold, contour maps can be constructed that show graded foci of evolutionary conservation in the underlying superstructure of the protein type, and the irregular shapes and extents of large conserved areas. To test the methodology, it was applied to cytochromes c and the carboxypeptidases A and B.

Cobra venoms (Naja species) contain a little-understood peptidase activity which shows specificity towards small peptides containing glycine and non-polar aromatic/aliphatic residues. We have examined the ability of whole cobra venom to degrade several types of peptide with emphasis on the action of Taiwan cobra (Naja naja atra) venom on L-alanylglycylglycine and glycylglycyl-L-phenylalanine. These are competing substrates, and it proved possible to generate inhibitors of the degradation of glycylglycyl-L-phenylalanine by synthesizing L-alanylglycylglycine analogues in which the peptide bond between the second and third residues had been replaced by different linkages. These analogues were themselves resistant to hydrolysis. The peptidase activity can also be inhibited by bestatin, captopril and chloromethyl ketones. Kinetic analyses suggested that even the best substrate discovered was of poor efficacy, so the natural peptide substrate remains to be identified. In unsuccessful attempts to devise a reliable chromogenic assay, it was found that the venom had activity against N-blocked amino acid p-nitrophenol esters, but not against leucine p-nitroanilide or ester substrates for trypsin-like and chymotrypsin-like enzymes.

Small angle x-ray scattering and viscometric analyses of the alpha-zeins of maize in solution indicated that the molecules were asymmetric. Structure predictions of consensus sequences for the two classes of alpha-zeins, Z19 and Z22, were in good agreement with the alpha-helical contents determined by circular dichroism. Dimensions determined by small angle x-ray scattering and viscometry indicated a predominantly alpha-helical conformation. The data are discussed in relation to models for the solution conformation and to earlier models for alpha-zeins structure.

Synopsis:With the recent advances in gene technology, a substantial premium has been placed on the ability to predict protein structure and mechanism from sequence data alone. This is because the direct experimental approaches are more time and labour intensive, and progress is slower. One area of theoretical investigation which has a clear potential to assist in predictive exercises is the natural evolution of protein structure and sequence.Some aspects of protein evolution, such as the fact that homology of sequence often denotes common ancestry, are routinely considered in predicting three-dimensional structure and mechanism. However, there are other aspects to this natural process which are either not realised or are not appreciated as potential analytical tools.This article introduces the different ways protein evolution can be investigated and indicates developing techniques as well as those already well established.

It is not widely appreciated that mammals can be venomous in the manner of snakes and lizards. However, it was first demonstrated scientifically 50 years ago in the case of the American short-tailed shrew. Subsequently, similar evidence has been obtained from European shrews and the Haitian solenodon, but research in this area has been almost completely neglected for the last 20 years. In complete contrast to what has been learned about other animal venoms, the identity and mode of action of mammal venom toxins are still unknown. This review draws attention once more to the pioneering work undertaken in the 1940s and 1950s, exploring in more detail than hitherto why the implications of mammal venom are just as important as the chemistry and pharmacology of the phenomenon itself.

It is hypothesised that the characteristic twin domain structure of serine proteases permits important allosteric responses in the molecule when peptide and protein substrates bind. Such movement would be ideal for stressing the scissile bond in the substrate, thereby making the task of hydrolysis substantially easier. The control of the domain movement can be closely associated with substrate binding, via the N- and C-terminal regions of the enzyme. The hypothesis also suggests that certain inhibitory peptides exert their effect by binding without inducing the domain movement.

A novel chymotrypsin inhibitor identified in fat body and hemocyte cDNA libraries of Boophilus microplus was named BmCI (B. microplus Chymotrypsin Inhibitor) (Genbank EU636772). The putative BmCI amino acid sequence presented a 22-residue-signal peptide and 58-residue-mature protein. BmCI amino acid sequence analysis allowed its classification as a Kunitz–BPTI inhibitor with six cysteine residues, a theoretical pI of 7.8, and the presence of Tyr at P1 position in the putative reactive site, suggesting inhibitory activity toward chymotrypsin. In this work, we reported the biochemical characterization of BmCI. The recombinant BmCI expressed in yeast Pichia pastoris was purified by size exclusion and reverse phase chromatographies. rBmCI expression yield was of 1 mg L−1 of culture. Purified rBmCI confirmed its chymotrypsin inhibitory activity with a low Ki (6.2 pM). The BmCI gene expression analysis by semi-quantitative RT-PCR indicated its transcription in the hemocytes, salivary gland and ovary. The cytotoxic activity of purified rBmCI was demonstrated in BALB/c 3T3 mouse fibroblasts. As assessed by the MTT reduction assay, rBMCI induced a dose-dependent decrease in 3T3 fibroblasts viability (IC50 = 8 μM). Moreover, flow cytometry analysis revealed that rBmCI is able to induce apoptosis, whereas no effect was observed on cell cycle progression. In conclusion, we demonstrated that rBmCI is cytotoxic against mammalian cells and obtained evidence that this growth inhibition is caused by an apoptosis-inducing activity.

Snake envenomation is listed as Category A Neglected Tropical Diseases (NTD) by World Health Organization, indicates a severe public health problem. The global figures for envenomation cases are estimated to be more than 1.8 million annually. Even if the affected victims survive the envenomation, they might suffer from permanent morbidity due to local envenomation. One of the most prominent local envenomation is dermonecrosis. Dermonecrosis is a pathophysiological outcome of envenomation that often causes disability in the victims due to surgical amputations, deformities, contracture, and chronic ulceration. The key venom toxins associated with this local symptom are mainly attributed to substantial levels of enzymatic and non-enzymatic toxins as well as their possible synergistic actions. Despite so, the severity of the local tissue damage is based on macroscopic observation of the bite areas. Furthermore, limited knowledge is known about the key biomarkers involved in the pathogenesis of dermonecrosis. The current immunotherapy with antivenom is also ineffective against dermonecrosis. These local effects eventually end up as sequelae. There is also a global shortage of toxins-targeted therapeutics attributed to inadequate knowledge of the actual molecular mechanisms of cytotoxicity. This chapter discusses the characterization of secretory phenotypes of dermonecrosis as an advanced tool to indicate its severity and pathogenesis in envenomation. Altogether, the secretory phenotypes of envenomed cells and tissues represent the precise characteristics of dermonecrosis caused by venom toxins.

A simple method has been developed to detect protein microenvironments which are likely to be the focus of natural selection, and thereby important for function. It relies on distinguishing between the probability of an amino acid type arising by genetic mutation and the probability that it will be chosen by natural selection. When applied to proteins of known tertiary structure, the method revealed that major differences exist in the balance between neutral and selected change, and also that functional sites can be highlighted.

Circular dichroism spectra were obtained for ten scorpion neurotoxins (representing five species of scorpion) in order to provide an understanding of their relative conformations in solution. Despite a high degree of amino acid sequence homology, the toxins clearly differ from each other in terms of CD-detectable structure. When superimposed, the CD spectra suggest that the toxins form a series of related conformational variants. Since the resemblances amongst the individual CD spectra can be correlated to degrees of sequence resemblance and pharmacological specificity, conformational balance could be an important factor in both toxin evolution and target recognition.

The distinctive amino acid compositions of protein exteriors and interiors were compared to the composition bias imposed by genetic code redundancy. It transpired that the synonym allocation is biased more in favour of those residues which are preferred in interiors, and this leads to an average interior residue being more probable and less mutable compared to an exterior residue. The general implications for protein evolution are discussed in association with the known evolutionary behavior of particular protein families. It is suggested that some proteins may have their structural history "fossilised" in their interiors and that the "amino acid" code is in reality a "protein" code.