Genomic Determinants Potentially Associated with Clinical Manifestations of Human-Pathogenic Tick-Borne Flaviviruses.

The tick-borne flavivirus group contains at least five species that are pathogenic to humans, three of which induce encephalitis (tick-borne encephalitis virus, louping-ill virus, Powassan virus) and another two species induce hemorrhagic fever (Omsk hemorrhagic fever virus, Kyasanur Forest disease virus). To date, the molecular mechanisms responsible for these strikingly different clinical forms are not completely understood. Using a bioinformatic approach, we performed the analysis of each amino acid (aa) position in the alignment of 323 polyprotein sequences to calculate the fixation index (Fst) per site and find the regions (determinants) where sequences belonging to two designated groups were most different. Our algorithm revealed 36 potential determinants (Fst ranges from 0.91 to 1.0) located in all viral proteins except a capsid protein. In an envelope (E) protein, most of the determinants were located on the virion surface regions (domains II and III) and one (absolutely specific site 457) was located in the transmembrane region. Another 100% specific determinant site (E63D) with Fst = 1.0 was located in the central hydrophilic domain of the NS2b, which mediates NS3 protease activity. The NS5 protein contains the largest number of determinants (14) and two of them are absolutely specific (T226S, E290D) and are located near the RNA binding site 219 (methyltransferase domain) and the extension structure. We assume that even if not absolutely, highly specific sites, together with absolutely specific ones (Fst = 1.0) can play a supporting role in cell and tissue tropism determination.

The relationship between the structure of the tick-borne encephalitis virus strains and their pathogenic properties.

Tick-borne encephalitis virus (TBEV) is transmitted to vertebrates by taiga or forest ticks through bites, inducing disease of variable severity. The reasons underlying these differences in the severity of the disease are unknown. In order to identify genetic factors affecting the pathogenicity of virus strains, we have sequenced and compared the complete genomes of 34 Far-Eastern subtype (FE) TBEV strains isolated from patients with different disease severity (Primorye, the Russian Far East). We analyzed the complete genomes of 11 human pathogenic strains isolated from the brains of dead patients with the encephalitic form of the disease (Efd), 4 strains from the blood of patients with the febrile form of TBE (Ffd), and 19 strains from patients with the subclinical form of TBE (Sfd). On the phylogenetic tree, pathogenic Efd strains formed two clusters containing the prototype strains, Senzhang and Sofjin, respectively. Sfd strains formed a third separate cluster, including the Oshima strain. The strains that caused the febrile form of the disease did not form a separate cluster. In the viral proteins, we found 198 positions with at least one amino acid residue substitution, of which only 17 amino acid residue substitutions were correlated with the variable pathogenicity of these strains in humans and they authentically differed between the groups. We considered the role of each amino acid substitution and assumed that the deletion of 111 amino acids in the capsid protein in combination with the amino acid substitutions R16K and S45F in the NS3 protease may affect the budding process of viral particles. These changes may be the major reason for the diminished pathogenicity of TBEV strains. We recommend Sfd strains for testing as attenuation vaccine candidates.

NS2B/NS3 protease: allosteric effect of mutations associated with the pathogenicity of tick-borne encephalitis virus.

The sequences of the protease domain of the tick-borne encephalitis (TBE) virus NS3 protein have two amino acid substitutions, 16 R→K and 45 S→F, in the highly pathogenic and poorly pathogenic strains of the virus, respectively. Two models of the NS2B-NS3 protease complex for the highly pathogenic and poorly pathogenic strains of the virus were constructed by homology modeling using the crystal structure of West Nile virus NS2B-NS3 protease as a template; 20 ns molecular dynamic simulations were performed for both models, the trajectories of the dynamic simulations were compared, and the averaged distance between the two models was calculated for each residue. Conformational differences between two models were revealed in the identified pocket. The different conformations of the pocket resulted in different orientations of the NS2B segment located near the catalytic triad. In the model of the highly pathogenic TBE virus the identified pocket had a more open conformation compared to the poorly pathogenic model. We propose that conformational changes in the active protease center, caused by two amino acid substitutions, can influence enzyme functioning and the virulence of the virus.