Gianni Panagiotou

The non-synonymous SNPs, the so-called non-silent SNPs, which are single-nucleotide variations in the coding regions that give “birth” to amino acid mutations, are often involved in the modulation of protein function. Understanding the effect of individual amino acid mutations on a protein/enzyme function or stability is useful for altering its properties for a wide variety of engineering studies. Since measuring the effects of amino acid mutations experimentally is a laborious process, a variety of computational methods have been discussed here that aid to extract direct genotype to phenotype information.

Molecular docking is the most commonly used technique in the modern drug discovery process where computational approaches involving docking algorithms are used to dock small molecules into macromolecular target structures. Over the recent years several evaluation studies have been reported by independent scientists comparing the performance of the docking programs by using default ‘black box’ protocols supplied by the software companies. Such studies have to be considered carefully as the docking programs can be tweaked towards optimum performance by selecting the parameters suitable for the target of interest. In this study we address the problem of selecting an appropriate docking and scoring function combination (88 docking algorithm-scoring functions) for substrate specificity predictions for feruloyl esterases, an industrially relevant enzyme family. We also propose the ‘Key Interaction Score System’ (KISS), a more biochemically meaningful measure for evaluation of docking programs based on pose prediction accuracy.

The parasite Plasmodium falciparum is the main agent responsible for malaria. In this study, we exploited a recently published chemical library from GlaxoSmithKline (GSK) that had previously been confirmed to inhibit parasite growth of the wild type (3D7) and the multi-drug resistance (D2d) strains, in order to uncover the weak links in the proteome of the parasite. We predicted 293 proteins of P. falciparum, including the six out of the seven verified targets for P. falciparum malaria treatment, as targets of 4645 GSK active compounds. Furthermore, we prioritized druggable targets, based on a number of factors, such as essentiality for growth, lack of homology with human proteins, and availability of experimental data on ligand activity with a non-human homologue of a parasite protein. We have additionally prioritized predicted ligands based on their polypharmacology profile, with focus on validated essential proteins and the effect of their perturbations on the metabolic network of P. falciparum, as well as indication of drug resistance emergence. Finally, we predict potential off-target effects on the human host with associations to cancer, neurological and dermatological disorders, based on integration of available chemical–protein and protein–protein interaction data. Our work suggests that a large number of the P. falciparum proteome is potentially druggable and could therefore serve as novel drug targets in the fight against malaria. At the same time, prioritized compounds from the GSK library could serve as lead compounds to medicinal chemists for further optimization.

Metabolic engineering is an attractive approach in order to improve the microbial production of drugs. Triterpenes is a chemically diverse class of compounds and many among them are of interest from a human health perspective. A systematic experimental or computational survey of all feasible gene modifications to determine the genotype yielding the optimal triterpene production phenotype is a laborious and time-consuming process. Based on the recent genome-wide sequencing of Saccharomyces cerevisiae CEN.PK 113-7D and its phenotypic differences with the S288C strain, we implemented a strategy for the construction of a β-amyrin production platform. The genes Erg8, Erg9 and HFA1 contained non-silent SNPs that were computationally analyzed to evaluate the changes that cause in the respective protein structures. Subsequently, Erg8, Erg9 and HFA1 were correlated with the increased levels of ergosterol and fatty acids in CEN.PK 113-7D and single, double, and triple gene over-expression strains were constructed. The six out of seven gene over-expression constructs had a considerable impact on both ergosterol and β-amyrin production. In the case of β-amyrin formation the triple over-expression construct exhibited a nearly 500% increase over the control strain making our metabolic engineering strategy the most successful design of triterpene microbial producers.

Identifying causative biological networks associated with relevant phenotypes is essential in the field of systems biology. We used ferulic acid (FA) as a model antioxidant to characterize the global expression programs triggered by this small molecule and decipher the transcriptional network controlling the phenotypic adaptation of the yeast Saccharomyces cerevisiae. By employing a strict cut off value during gene expression data analysis, 106 genes were found to be involved in the cell response to FA, independent of aerobic or anaerobic conditions. Network analysis of the system guided us to a key target node, the FMP43 protein, that when deleted resulted in marked acceleration of cellular growth (∼15% in both minimal and rich media). To extend our findings to human cells and identify proteins that could serve as drug targets, we replaced the yeast FMP43 protein with its human ortholog BRP44 in the genetic background of the yeast strain Δfmp43. The conservation of the two proteins was phenotypically evident, with BRP44 restoring the normal specific growth rate of the wild type. We also applied homology modeling to predict the 3D structure of the FMP43 and BRP44 proteins. The binding sites in the homology models of FMP43 and BRP44 were computationally predicted, and further docking studies were performed using FA as the ligand. The docking studies demonstrated the affinity of FA towards both FMP43 and BRP44. This study proposes a hypothesis on the mechanisms yeast employs to respond to antioxidant molecules, while demonstrating how phenome and metabolome yeast data can serve as biomarkers for nutraceutical discovery and development. Additionally, we provide evidence for a putative therapeutic target, revealed by replacing the FMP43 protein with its human ortholog BRP44, a brain protein, and functionally characterizing the relevant mutant strain.

Metabolomics is a rapidly evolving discipline that involves the systematic study of endogenous small molecules that characterize the metabolic pathways of biological systems. The study of metabolism at a global level has the potential to contribute significantly to biomedical research, clinical medical practice, as well as drug discovery. In this paper, we present the most up‐to‐date metabolite and metabolic pathway resources, and we summarize the statistical, and machine‐learning tools used for the analysis of data from clinical metabolomics. Through specific applications on cancer, diabetes, neurological and other diseases, we demonstrate how these tools can facilitate diagnosis and identification of potential biomarkers for use within disease diagnosis. Additionally, we discuss the increasing importance of the integration of metabolomics data in drug discovery. On a case‐study based on the Human Metabolome Database (HMDB) and the Chinese Natural Product Database (CNPD), we demonstrate the close relatedness of the two data sets of compounds, and we further illustrate how structural similarity with human metabolites could assist in the design of novel pharmaceuticals and the elucidation of the molecular mechanisms of medicinal plants.

Our understanding of how fungi evolved to develop a variety of ecological niches, is limited but of fundamental biological importance. Specifically, the evolution of enzymes affects how well species can adapt to new environmental conditions. Feruloyl esterases (FAEs) are enzymes able to hydrolyze the ester bonds linking ferulic acid to plant cell wall polysaccharides. The diversity of substrate specificities found in the FAE family shows that this family is old enough to have experienced the emergence and loss of many activities. In this study we evaluate the relative activity of FAEs against a variety of model substrates as a novel predictive tool for Ascomycota taxonomic classification. Our approach consists of two analytical steps; (1) an initial unsupervised analysis to cluster the FAEs substrate specificity data which were generated by cultivation of 34 Ascomycota strains and then an analysis of the produced enzyme cocktail against 10 substituted cinnamate and phenylalkanoate methyl esters, (2) a second, supervised analysis for training a predictor built on these substrate activities. By applying both linear and non-linear models we were able to correctly predict the taxonomic Class (∼86% correct classification), Order (∼88% correct classification) and Family (∼88% correct classification) that the 34 Ascomycota belong to, using the activity profiles of the FAEs. The good correlation with the FAEs substrate specificities that we have defined via our phylogenetic analysis not only suggests that FAEs are phylogenetically informative proteins but it is also a considerable step towards improved FAEs functional prediction.

One of the most intriguing groups of enzymes, the feruloyl esterases (FAEs), is ubiquitous in both simple and complex organisms. FAEs have gained importance in biofuel, medicine and food industries due to their capability of acting on a large range of substrates for cleaving ester bonds and synthesizing high-added value molecules through esterification and transesterification reactions. During the past two decades extensive studies have been carried out on the production and partial characterization of FAEs from fungi, while much less is known about FAEs of bacterial or plant origin. Initial classification studies on FAEs were restricted on sequence similarity and substrate specificity on just four model substrates and considered only a handful of FAEs belonging to the fungal kingdom. This study centers on the descriptor-based classification and structural analysis of experimentally verified and putative FAEs; nevertheless, the framework presented here is applicable to every poorly characterized enzyme family. 365 FAE-related sequences of fungal, bacterial and plantae origin were collected and they were clustered using Self Organizing Maps followed by k-means clustering into distinct groups based on amino acid composition and physico-chemical composition descriptors derived from the respective amino acid sequence. A Support Vector Machine model was subsequently constructed for the classification of new FAEs into the pre-assigned clusters. The model successfully recognized 98.2% of the training sequences and all the sequences of the blind test. The underlying functionality of the 12 proposed FAE families was validated against a combination of prediction tools and published experimental data. Another important aspect of the present work involves the development of pharmacophore models for the new FAE families, for which sufficient information on known substrates existed. Knowing the pharmacophoric features of a small molecule that are essential for binding to the members of a certain family opens a window of opportunities for tailored applications of FAEs.