Computer Program for Detection and Analyzing the Porin-Mediated Antibiotic Resistance of Bacteria.

The aim of this work was to develop a new software tool for identifying gene mutations that determine the porin-mediated resistance to antibiotics in gram-negative bacteria and to demonstrate the functionality of this program by detecting porin-mediated resistance to carbapenems in clinical isolates of Pseudomonas aeruginosa. Materials and Methods: The proposed algorithm is based on searching for a correspondence between the reference and the studied genes. When the sought nucleotide sequence is found in the analyzed genome, it is compared with the reference one and analyzed. The genomic analysis is then verified by comparing between the amino acid sequences encoded by the reference and studied genes. The genes of the susceptible P. aeruginosa ATCC 27853 strain were used as the reference nucleotide sequences encoding for porins (OprD, OpdD, and OpdP) involved in the transport of carbapenems into the bacterial cell. The complete genomes of clinical P. aeruginosa isolates from the PATRIC database 3.6.9 and our own collection were used to test the functionality of the proposed program. The analyzed isolates were phenotypically characterized according to the CLSI standard. The search for carbapenemase genes in the studied genomes of P. aeruginosa was carried out using the ResFinder 4.1. Results: The developed program for detecting the genetic determinants of non-plasmid antibiotic resistance made it possible to identify mutations of various types and significance in the porin genes of P. aeruginosa clinical isolates. These mutations led to modifications of the peptide structure of porin proteins. Single amino acid substitutions prevailed in the OpdD and OpdP porins of carbapenem-susceptible and carbapenem-resistant isolates. In the carbapenem-resistant strains, the gene encoding for OprD porin was found heavily modified, including insertions and/or deletions, which led to premature termination of porin synthesis. In several isolates resistant to meropenem, no mutations were detected in the gene encoding for OprD, which might be associated with alternative mechanisms of resistance to carbapenems. Conclusion: The proposed software product can become an effective tool for deciphering the molecular genetic mechanisms of bacterial chromosomal resistance to antibiotics. Testing the program revealed differences between the occurrences of mutations significant for carbapenem resistance in the oprD, opdD, and opdP genes.

In silico design of high-affinity ligands for the immobilization of inulinase.

Using computer modeling, virtual screening of high-affinity ligands for immobilization of inulinase - an enzyme that cleaves inulin and fructose-containing polymers to fructose - has been performed. The inulinase molecule from Aspergillus ficuum (pdb: 3SC7) taken from the database of protein structures was used as a protein model and the target for flexible docking. The set of ligands studied included simple sugars (activators, inhibitors, products of enzymatic catalysis), as well as high-molecular weight compounds (polycation and polyanion exchange resins, glycoproteins, phenylalanine-proline peptide, polylactate, and caffeine). Based on the comparative analysis of the values of the total energy and the localization of ligand binding sites, we made several assumptions concerning the mechanisms of interaction of the suggested matrices for the immobilization of enzyme molecules and the structural features of such complexes. It was also assumed that the candidates for immobilization agents meeting the industrial requirements may be glycoproteins, for which we propose an additional incorporation of cysteine residues into their structure, aimed to create disulfide «anchors¼ to the surface.

[Experience in simulating the structural and dynamic features of small proteins using table supercomputers].

The results of theoretical studies of the structural and dynamic features of peptides and small proteins have been presented that were carried out by quantum chemical and molecular dynamics methods in high-performance graphic stations, "table supercomputers", using distributed calculations by the CUDA technology.

[On possible influence of different pathways of binding of amides with water on the "pyramidalization" of nitrogen valence bonds of peptide groups].

With the aim to study solvation effects in peptide structure organization, the energy of different types of hydration in simple amines and amides has been analyzed. It was shown based on the quantum-chemical DFT and PM3 calculations of amino derivatives CH3-(CH2)3-NH2, (CH3)2-NH, CH3-NH2, NH3, CH2=CH-NH2, H-CC-NH2, O=C(CH)3-N(CH3)2, O=C(CH3)-NH(CH3), O=C(CH3)-NH2, O=CH-N(CH3)H, and O=CH-NH2 that: (1) in the given set of molecules, the proton acceptor N...H-O variant of hydrogen bonding of NH2-group with a water molecule is dominating only for the simplest amines. Being primordially weaker, the proton donor N-H...OH variant of water H-bonding gradually increases in energy in the given set as the basicity of the compound decreases, and for the case of amides of carbon acids it becomes already a significant channel of the hydration; (2) the intermolecular N-H...O=C bonding of trans-N-methylacetamides, which models the peptide hydrogen bonds in proteins, induces a "planarization" of its initially nonplanar O=C-NH fragments. However, the addition of water molecules to the complex through the proton acceptor N...H-O variant of binding of N atom not only "restores", but even strengthens the "pyramidalization" of valent bonds of peptide groups in this place.

[Torsion lability of the O=C-N-H fragment in single dipeptide molecules of L-amino acids].

On the basis of ab initio MP2 and semi-empirical PM3 quantum-chemical calculations the bistability of the nonplanar O=C-N-H fragment in the structure of simple amides and dipeptides is discussed. The influence of the nature of amino acids on the structural polymorphism of the peptide group in dipeptides is shown.