Maris Lapinsh

Classification of G-protein coupled receptors by alignment-independent extraction of principal chemical properties of primary amino acid sequences

We have developed an alignment‐independent method for classification of G‐protein coupled receptors (GPCRs) according to the principal chemical properties of their amino acid sequences. The method relies on a multivariate approach where the primary amino acid sequences are translated into vectors based on the principal physicochemical properties of the amino acids and transformation of the data into a uniform matrix by applying a modified autocross‐covariance transform. The application of principal component analysis to a data set of 929 class A GPCRs showed a clear separation of the major classes of GPCRs. The application of partial least squares projection to latent structures created a highly valid model (cross‐validated correlation coefficient, Q2 = 0.895) that gave unambiguous classification of the GPCRs in the training set according to their ligand binding class. The model was further validated by external prediction of 535 novel GPCRs not included in the training set. Of the latter, only 14 sequences, confined in rapidly expanding GPCR classes, were mispredicted. Moreover, 90 orphan GPCRs out of 165 were tentatively identified to GPCR ligand binding class. The alignment‐independent method could be used to assess the importance of the principal chemical properties of every single amino acid in the protein sequences for their contributions in explaining GPCR family membership. It was then revealed that all amino acids in the unaligned sequences contributed to the classifications, albeit to varying extent; the most important amino acids being those that could also be determined to be conserved by using traditional alignment‐based methods.

Development of proteo-chemometrics: a novel technology for the analysis of drug-receptor interactions

A novel method for the analysis of drug receptor interactions has been developed and used to explore mechanisms involved in the binding of 4-piperidyl oxazole antagonists to alpha1a-, alpha1b- and alpha1d-adrenoceptors. The method exploits affinity data for a series of organic chemical compounds binding to wild-type and artificially mutated receptors. The receptor sequences and compounds are assigned predictor variables that are correlated to the measured pharmacological activities using partial least-squares projections to latent structures. The predictor variables consist of one descriptor block derived from the chemical properties of the receptors primary amino acid sequences and another descriptor block derived from the chemical properties of the organic compounds. The cross-terms generated from the two descriptor blocks are also derived. Using this approach, very sturdy models were generated describing the interactions of the chemical compounds with the receptors. Models are useful to predict binding affinity and receptor subtype selectivity of compounds prior to their synthesis, and may find use in rational drug design. Moreover, models also give quantitative information about the interactions of the amino acids of the receptors with the ligands, thereby giving an insight into the molecular mechanisms involved in ligand binding.

Improved approach for proteochemometrics modeling: application to organic compound---amine G protein-coupled receptor interactions

MOTIVATIONnProteochemometrics is a novel technology for the analysis of interactions of series of proteins with series of ligands. We have here customized it for analysis of large datasets and evaluated it for the modeling of the interaction of psychoactive organic amines with all the five known families of amine G protein-coupled receptors (GPCRs).nnnRESULTSnThe model exploited data for the binding of 22 compounds to 31 amine GPCRs, correlating chemical descriptions and cross-descriptions of compounds and receptors to binding affinity using a novel strategy. A highly valid model (q2 = 0.76) was obtained which was further validated by external predictions using data for 10 other entirely independent compounds, yielding the high q2ext = 0.67. Interpretation of the model reveals molecular interactions that govern psychoactive organic amines overall affinity for amine GPCRs, as well as their selectivity for particular amine GPCRs. The new modeling procedure allows us to obtain fully interpretable proteochemometrics models using essentially unlimited number of ligand and protein descriptors.

Prediction of indirect interactions in proteins

BackgroundBoth direct and indirect interactions determine molecular recognition of ligands by proteins. Indirect interactions can be defined as effects on recognition controlled from distant sites in the proteins, e.g. by changes in protein conformation and mobility, whereas direct interactions occur in close proximity of the proteins amino acids and the ligand. Molecular recognition is traditionally studied using three-dimensional methods, but with such techniques it is difficult to predict the effects caused by mutational changes of amino acids located far away from the ligand-binding site. We recently developed an approach, proteochemometrics, to the study of molecular recognition that models the chemical effects involved in the recognition of ligands by proteins using statistical sampling and mathematical modelling.ResultsA proteochemometric model was built, based on a statistically designed protein librarys (melanocortin receptors) interaction with three peptides and used to predict which amino acids and sequence fragments that are involved in direct and indirect ligand interactions. The model predictions were confirmed by directed mutagenesis. The predicted presumed direct interactions were in good agreement with previous three-dimensional studies of ligand recognition. However, in addition the model could also correctly predict the location of indirect effects on ligand recognition arising from distant sites in the receptors, something that three-dimensional modelling could not afford.ConclusionWe demonstrate experimentally that proteochemometric modelling can be used with high accuracy to predict the site of origin of direct and indirect effects on ligand recognitions by proteins.

Unbiased descriptor and parameter selection confirms the potential of proteochemometric modelling

BackgroundProteochemometrics is a new methodology that allows prediction of protein function directly from real interaction measurement data without the need of 3D structure information. Several reported proteochemometric models of ligand-receptor interactions have already yielded significant insights into various forms of bio-molecular interactions. The proteochemometric models are multivariate regression models that predict binding affinity for a particular combination of features of the ligand and protein. Although proteochemometric models have already offered interesting results in various studies, no detailed statistical evaluation of their average predictive power has been performed. In particular, variable subset selection performed to date has always relied on using all available examples, a situation also encountered in microarray gene expression data analysis.ResultsA methodology for an unbiased evaluation of the predictive power of proteochemometric models was implemented and results from applying it to two of the largest proteochemometric data sets yet reported are presented. A double cross-validation loop procedure is used to estimate the expected performance of a given design method. The unbiased performance estimates (P2) obtained for the data sets that we consider confirm that properly designed single proteochemometric models have useful predictive power, but that a standard design based on cross validation may yield models with quite limited performance. The results also show that different commercial software packages employed for the design of proteochemometric models may yield very different and therefore misleading performance estimates. In addition, the differences in the models obtained in the double CV loop indicate that detailed chemical interpretation of a single proteochemometric model is uncertain when data sets are small.ConclusionThe double CV loop employed offer unbiased performance estimates about a given proteochemometric modelling procedure, making it possible to identify cases where the proteochemometric design does not result in useful predictive models. Chemical interpretations of single proteochemometric models are uncertain and should instead be based on all the models selected in the double CV loop employed here.

Rough set‐based proteochemometrics modeling of G‐protein‐coupled receptor‐ligand interactions

G‐Protein‐coupled receptors (GPCRs) are among the most important drug targets. Because of a shortage of 3D crystal structures, most of the drug design for GPCRs has been ligand‐based. We propose a novel, rough set‐based proteochemometric approach to the study of receptor and ligand recognition. The approach is validated on three datasets containing GPCRs. In proteochemometrics, properties of receptors and ligands are used in conjunction and modeled to predict binding affinity. The rough set (RS) rule‐based models presented herein consist of minimal decision rules that associate properties of receptors and ligands with high or low binding affinity. The information provided by the rules is then used to develop a mechanistic interpretation of interactions between the ligands and receptors included in the datasets. The first two datasets contained descriptors of melanocortin receptors and peptide ligands. The third set contained descriptors of adrenergic receptors and ligands. All the rule models induced from these datasets have a high predictive quality. An example of a decision rule is “If R1_ligand(Ethyl) and TM helix 2 position 27(Methionine) then Binding(High).” The easily interpretable rule sets are able to identify determinative receptor and ligand parts. For instance, all three models suggest that transmembrane helix 2 is determinative for high and low binding affinity. RS models show that it is possible to use rule‐based models to predict ligand‐binding affinities. The models may be used to gain a deeper biological understanding of the combinatorial nature of receptor‐ligand interactions. Proteins 2006.

Melanocortin Receptors: Ligands and Proteochemometrics Modeling

Abstract: The melanocortin receptors exist in five subtypes, MC1–5R. These receptors participate in important regulations of the immune system, central behavior, and endocrine and exocrine glands. Here we provide a short review on MCR subtype selective peptides and organic compounds with activity on the MCRs, developed in our laboratory. Also provided is an overview of our new proteochemometric modeling technology, which has been applied to model the interaction of MSH peptides with the MCRs.

Proteochemometric modeling reveals the interaction site for Trp9 modified α-MSH peptides in melanocortin receptors

The interactions of α‐MSH peptides with melanocortin receptors (MCRs) were located by proteochemometric modeling. Nine α‐MSH peptide analogues were constructed by exchanging the Trp9 residue in the α‐MSH core with the natural or artificial amino acids Arg, Asp, Cys, Gly, Leu, Nal, d‐Nal, Pro, or d‐Trp. The nine peptides created, and α‐MSH itself, were evaluated for their interactions with the 4 wild‐type MC1,3–5Rs and 15 multichimeric MCRs, each of the latter being constructed from three sequence segments, each taken from a different wild‐type MC1,3–5R. The segments of the chimeric MCRs were selected according to the principles of statistical molecular design and were arranged so as to divide the receptors into five parts. By this approach, a set of 19 maximally diverse MC receptor proteins was obtained for which the interaction activity with the 10 peptides were measured by radioligand binding thus creating data for 190 ligand–protein pairs, which were subsequently analyzed by use of proteochemometric modeling. In proteochemometrics, the structural or physicochemical properties of both interaction partners, which represent the complementarity of the interacting entities, are used to create multivariate mathematical descriptions. (Here, physicochemical property descriptors of the receptors and peptides amino acids were used). A valid, highly predictive (Q2 = 0.74) and easily interpretable model was then obtained. The model was further validated by its ability to correctly predicting the affinity of α‐MSH for new point and cassette‐mutated MC4/MC1Rs, and it was then used to identify the receptor residues that are important for affording the high affinity and selectivity of α‐MSH for the MC1R. It was revealed that these residues are located in several quite distant parts of the receptors transmembrane cavity and must therefore cause their influence at various stages of the dynamic ligand‐binding process, such as by affecting the conformation of the ligand at the vicinity of the receptor and taking part in the path of the ligands entry into its binding pocket. Our study can be used as a template how to create high resolution proteochemometric models when there are a limited number of natural proteins and ligands available. Proteins 2007.

N-alkylated dipeptide amides and related structures as imitations of the melanocortins’ active core

Thirty-three low molecular mass structures combining both peptide and peptoid features were prepared and tested on human melanocortin receptors MC1,3-5R. Most of them displayed low micromolar activity with preference for diamines, guanidino and 2-naphthyl derivatives compared to monoacetylated, amino and 3-indolyl counterparts. Some contained L- or D-histidine residues, but the change did not influence affinity. QSAR modelling yielded excellent models for the MC3-5 receptors explaining R2Y=0.89-0.91 and predicting Q2=0.77-0.80 of the affinity variations. One compound displayed MC1R selectivity (13-fold and more). An NMR study of showed that it exists as a mixture of four rotamers at its tertiary amide bonds. Comparisons with earlier data for melanocortin core tetrapeptide analogues indicate that the novel peptide-peptoids interact with the melanocortin receptors in a different way.