Ekaterina Chernyaeva

Genome-Wide Mycobacterium tuberculosis Variation (GMTV) Database: A New Tool for Integrating Sequence Variations and Epidemiology

BackgroundTuberculosis (TB) poses a worldwide threat due to advancing multidrug-resistant strains and deadly co-infections with Human immunodeficiency virus. Today large amounts of Mycobacterium tuberculosis whole genome sequencing data are being assessed broadly and yet there exists no comprehensive online resource that connects M. tuberculosis genome variants with geographic origin, with drug resistance or with clinical outcome.DescriptionHere we describe a broadly inclusive unifying Genome-wide Mycobacterium tuberculosis Variation (GMTV) database, (http://mtb.dobzhanskycenter.org) that catalogues genome variations of M. tuberculosis strains collected across Russia. GMTV contains a broad spectrum of data derived from different sources and related to M. tuberculosis molecular biology, epidemiology, TB clinical outcome, year and place of isolation, drug resistance profiles and displays the variants across the genome using a dedicated genome browser. GMTV database, which includes 1084 genomes and over 69,000 SNP or Indel variants, can be queried about M. tuberculosis genome variation and putative associations with drug resistance, geographical origin, and clinical stages and outcomes.ConclusionsImplementation of GMTV tracks the pattern of changes of M. tuberculosis strains in different geographical areas, facilitates disease gene discoveries associated with drug resistance or different clinical sequelae, and automates comparative genomic analyses among M. tuberculosis strains.

Contrasting origin of B chromosomes in two cervids (Siberian roe deer and grey brocket deer) unravelled by chromosome-specific DNA sequencing

BackgroundB chromosomes are dispensable and variable karyotypic elements found in some species of animals, plants and fungi. They often originate from duplications and translocations of host genomic regions or result from hybridization. In most species, little is known about their DNA content. Here we perform high-throughput sequencing and analysis of B chromosomes of roe deer and brocket deer, the only representatives of Cetartiodactyla known to have B chromosomes.ResultsIn this study we developed an approach to identify genomic regions present on chromosomes by high-throughput sequencing of DNA generated from flow-sorted chromosomes using degenerate-oligonucleotide-primed PCR. Application of this method on small cattle autosomes revealed a previously described KIT gene region translocation associated with colour sidedness. Implementing this approach to B chromosomes from two cervid species, Siberian roe deer (Capreolus pygargus) and grey brocket deer (Mazama gouazoubira), revealed dramatically different genetic content: roe deer B chromosomes consisted of two duplicated genomic regions (a total of 1.42-1.98 Mbp) involving three genes, while grey brocket deer B chromosomes contained 26 duplicated regions (a total of 8.28-9.31 Mbp) with 34 complete and 21 partial genes, including KIT and RET protooncogenes, previously found on supernumerary chromosomes in canids. Sequence variation analysis of roe deer B chromosomes revealed a high frequency of mutations and increased heterozygosity due to either amplification within B chromosomes or divergence between different Bs. In contrast, grey brocket deer B chromosomes were found to be more homogeneous and resembled autosomes in patterns of sequence variation. Similar tendencies were observed in repetitive DNA composition.ConclusionsOur data demonstrate independent origins of B chromosomes in the grey brocket and roe deer. We hypothesize that the B chromosomes of these two cervid species represent different stages of B chromosome sequences evolution: probably nascent and similar to autosomal copies in brocket deer, highly derived in roe deer. Based on the presence of the same orthologous protooncogenes in canids and brocket deer Bs we argue that genomic regions involved in B chromosome formation are not random. In addition, our approach is also applicable to the characterization of other evolutionary and clinical rearrangements.

Evaluation of a low-density hydrogel microarray technique for mycobacterial species identification

ABSTRACT In addition to the obligatory pathogenic species of the Mycobacterium tuberculosis complex and Mycobacterium leprae, the genus Mycobacterium also includes conditionally pathogenic species that in rare cases can lead to the development of nontuberculous mycobacterial diseases. Because tuberculosis and mycobacteriosis have similar clinical signs, the accurate identification of the causative agent in a clinical microbiology laboratory is important for diagnostic verification and appropriate treatment. This report describes a low-density hydrogel-based microarray containing oligonucleotide probes based on the species-specific sequences of the gyrB gene fragment for mycobacterial species identification. The procedure included the amplification of a 352-nucleotide fragment of the gene and its hybridization on a microarray. The triple-species-specific probe design and the algorithm for hybridization profile recognition based on the calculation of Pearson correlation coefficients, followed by the construction of a profile database, allowed for the reliable and accurate identification of mycobacterial species, including mixed-DNA samples. The assay was used to evaluate 543 clinical isolates from two regions of Russia, demonstrating its ability to detect 35 mycobacterial species, with 99.8% sensitivity and 100% specificity when using gyrB, 16S, and internal transcribed spacer (ITS) fragment sequencing as the standard. The testing of clinical samples showed that the sensitivity of the assay was 89% to 95% for smear-positive samples and 36% for smear-negative samples. The large number of identified species, the high level of sensitivity, the ability to detect mycobacteria in clinical samples, and the up-to-date profile database make the assay suitable for use in routine laboratory practice.

Characterization of multiple and extensively drug resistant Mycobacterium tuberculosis isolates with different ofloxacin-resistance levels

The goal of the study was to determine different mutation types in gyrA and gyrB genes in Mycobacterium tuberculosis strains with low-level (2 μg/ml) and high-level (10 μg/ml) ofloxacin (OFL) resistance and to compare genetic diversity of ofloxacin-resistant and susceptible M. tuberculosis isolates. M. tuberculosis isolates were collected in Leningrad Region in 2011. DNA sequencing showed that 54.3% of isolates with low-level and 76.9% of isolates with high-level OFL-resistance had mutations in gyrA gene. Few isolates carried mutations in gyrB gene - five among isolates with low-level resistance and two among high-level resistant isolates. Altogether, detection of point mutations in both DNA gyrase genes allows to identify 66.9% of mycobacterial isolates with low-level and 84.5% of isolates with high-level of OFL-resistance. Novel mutations S91L in gyrA gene and S512L in gyrB gene described in this study were detected in OFL-resistant isolates and may play role in M. tuberculosis fluoroquinolone resistance. M. tuberculosis Beijing family spoligotypes were identified among 70.8% of isolates with low-level resistance, 84.6% of isolates with high-level resistance and 50% of strains susceptible to all tuberculosis drugs. Fishers exact test revealed significant difference between Beijing prevalence in groups of drug-susceptible or high-level OFL-resistant M. tuberculosis strains (p-value = 0.032).

Molecular genetic analysis of Mycobacterium tuberculosis strains spread in different patient groups in St. Petersburg, Russia.

Molecular epidemiological features of Mycobacterium tuberculosis strains among different patient groups in Russia have not been studied well. The aim of our study was to compare the genotypes of M. tuberculosis strains circulating among tuberculosis (TB) patients from different groups: homeless, human immunodeficiency virus (HIV)-infected, prisoners, and the general population of St. Petersburg citizens. One hundred and forty-two M. tuberculosis complex isolates from different TB patient groups were studied using the spacer oligonucleotide typing (spoligotyping) method. The majority of the studied M. tuberculosis isolates in all groups belonged to the Beijing family (55% among homeless; 77% among HIV-infected; 60% among the general population; 81% among prisoners). There were no significant differences in the Beijing family prevalence among homeless patients, HIV/TB co-infected patients, and the general population of TB patients. The lowest genetic diversity of the pathogen was detected among imprisoned patients. The results of our study demonstrate that M. tuberculosis strains circulating among patients from high-risk groups are also spread among the general population of St. Petersburg citizens.

Next-Generation Sequencing of Mycobacterium tuberculosis.

To the Editor: Next-generation sequencing (NGS) technology is becoming more affordable and is increasingly being widely used for high-resolution molecular epidemiology of tuberculosis. Using an example of the emerging multidrug-resistant strain of Mycobacterium tuberculosis, we showed the value of informed understanding when in silico prediction from NGS data achieved with available bioinformatics tools is placed within the context of the existing genotyping framework. Spoligotyping is a classical method of M. tuberculosis genotyping, and the SITVIT_WEB database contains data on 7,105 spoligotype patterns of 58,180 isolates from 153 countries (http://www.pasteur-guadeloupe.fr:8081/SITVIT_ONLINE). Spoligotyping targets a variation of the DR/CRISPR locus, whose evolution in M. tuberculosis occurs through deletion of single or multiple spacers. By virtue of the orientation of the associated cas genes, the locus is situated on the minus strand, whereas its spacers are numbered within the locus, not the genome. Spoligotype international type (SIT) 266 (Figure, panel A) is an epidemiologically significant genotype. It constitutes a substantial proportion of the population structure of M. tuberculosis in Belarus, a post-Soviet state in Eastern Europe (1,2), and has been described sporadically in the neighboring provinces in northwestern and central Russia (2–5) and in Latvia (6). More important, it is multidrug resistant (MDR) (and most likely extensively drug resistant). In a recent Belarus study, SIT266 was found in 25 of 163 strains; all 25 were MDR (1). This situation contrasts clearly with its apparently parental type SIT264, which differs from SIT266 in a single spacer 8 (Figure, panel A). SIT264 is more widespread across Eastern Europe but at very low prevalence and is not associated with multidrug resistance (3,6,7). On the basis of 24 mycobacterial interspersed repetitive unit variable number tandem repeats clustering and robust phylogenetic single-nucleotide polymorphisms, SIT264 and SIT266 isolates are assigned to the Latin American–Mediterranean lineage of M. tuberculosis (2). Figure A) Spoligotyping hybridization profiles of H37Rv and BCG reference strains, Mycobacterium tuberculosis SIT266 (2 strains in this study) and SIT264 (previously published strain in [3]). SIT, spoligotype international type, designated according to SITVIT_WEB ... We recovered 2 MDR M. tuberculosis isolates of SIT266 from pulmonary tuberculosis patients from northwestern Russia in 2014. Bacterial DNA was subjected to macroarray-based spoligotyping (8) and whole-genome sequencing on the MiSeq platform (Illumina, San Diego, CA, USA). M. tuberculosis NGS data were deposited in the National Center for Biotechnology Information Sequence Read Archive (project no. PRJNA305488). The short sequencing reads were subjected to analysis by using the SpoTyping program (https://github.com/xiaeryu/SpoTyping) (9) and the TGS-TB online tool (https://gph.niid.go.jp/tgs-tb/) (10) to deduce their spoligoprofile. We used the TGS-TB tool to map the IS6110 insertion sites and detect drug resistance mutations. The reads also were mapped to the genome of reference strain H37Rv (GenBank accession no. NC_00962.3) by using the Geneious 9.0 package (Biomatters Ltd, Auckland, New Zealand). We obtained 1,294,895 and 816,693 paired reads for strains 4542 and 8279, respectively, and mapped them to the reference. Mean read length was 300 bp, and the average genome coverage was 72. Strains 4542 and 8279 were phenotypically MDR and harbored mutations associated with resistance to all 5 first line-drugs. The macroarray hybridization spoligotyping assigned both strains to spoligotype SIT266. However, by in silico typing, their spoligotype was predicted to be SIT264 (Figure, panel A). To reconcile these findings, we hypothesized that this discrepancy resulted from an IS6110 asymmetrically inserted in the direct repeat unit adjacent to the spacer 8 in a SIT266 isolate. This insertion would disrupt a target sequence for biotin-labeled DRa primer, thus preventing spacer 8 from amplification. Indeed, in both SIT266 isolates, the in silico analysis identified a forward IS6110 insertion that was mapped to position 3122916 in H37Rv genome (Technical Appendix Figure). This location correlates with the location of spacer 8 in this same genome from positions 3122954 to 3122918. Thus, IS6110 precedes spacer 8 in the genome of isolate with spoligotype SIT266, or follows it, within the DR locus (Figure, panel B; Technical Appendix Figure). An immediate excellent contribution of NGS with regard to tuberculosis treatment and control is its capacity to rapidly screen for multiple gene targets linked to the development of drug resistance. However, knowledge of strain genotype is no less clinically and epidemiologically relevant. A superspreading strain might be marked with other pathobiologically important features. In the case presented here (indeed emerging and MDR), the NGS-based in silico spoligotyping would confuse the MDR/extensively drug resistant SIT266 with “less dangerous” SIT264. To be precise, the revealed discrepancy is not inherent to the NGS technology itself. Although the general limitation of the use of short sequencing reads to infer repetitive genome regions is known, it did not pose a problem in our study. However, both bioinformatics tools predicted the spoligoprofile solely from the presence or absence of spacer sequences and did not take into account a “hiding” effect exerted by a putative IS6110 insertion on adjacent spacer under classical spoligotyping. In conclusion, we suggest that an accurate NGS-based prediction requires an integrative approach to all relevant information obtained by in silico analysis of a given genome locus. In particular, not only presence of CRISPR spacers but also presence and location of potentially interfering IS6110 insertion(s) should be considered for correct NGS-based assignment to internationally recognized spoligotypes. Technical Appendix: Whole-genome, short-sequencing reads of Mycobacterium tuberculosis strain 4542 mapped to complete genome of reference strain H37Rv (spacers 7–9 in CRISPR locus). Click here to view.(3.1M, pdf)

Emerging peak on the phylogeographic landscape of Mycobacterium tuberculosis in West Asia: Definitely smoke, likely fire

To date, a major attention was justly given to the global lineages of Mycobacterium tuberculosis. Here, we demonstrated an importance of the minor ones, on an example of intriguing and underestimated NEW-1 family that belongs to Euro-American lineage (lineage 4). Analysis of the global WGS/NGS datasets (5715 strains) identified 2235 strains of Lineage 4 and 66 strains of sublineage L4.5. This latter is marked with RD122 genomic deletion and includes NEW-1 family. Phylogenomic analysis confirmed a separate position of the NEW-1 family that we tentatively designate L4.5.1/Iran. We propose an evolution/migration scenario starting with origin of L4.5 1000-1300 ya in China, subsequent origin of the pre-NEW-1 intermediate genotype in Tibet, further migration to Xinjiang/Uyghur, and finally to Iran since 800 ya (origin of NEW-1), possibly, via expansion of the Mongol Yuan empire. Analysis of longitudinal phylogeographic datasets revealed a sharp increase in prevalence of NEW-1 strains in Iran and its eastwards neighbors in the last 20years; most alarmingly, it is accompanied with significant association with multidrug resistance (MDR). Ongoing migration, especially, Afghan refugees flows to developed countries emphasize a risk of the wider spread of the epidemic MDR subtype within NEW-1 family that we coin as emerging resistant clone of M. tuberculosis in West Asia.

Sequencing of Supernumerary Chromosomes of Red Fox and Raccoon Dog Confirms a Non-Random Gene Acquisition by B Chromosomes

B chromosomes (Bs) represent a variable addition to the main karyotype in some lineages of animals and plants. Bs accumulate through non-Mendelian inheritance and become widespread in populations. Despite the presence of multiple genes, most Bs lack specific phenotypic effects, although their influence on host genome epigenetic status and gene expression are recorded. Previously, using sequencing of isolated Bs of ruminants and rodents, we demonstrated that Bs originate as segmental duplications of specific genomic regions, and subsequently experience pseudogenization and repeat accumulation. Here, we used a similar approach to characterize Bs of the red fox (Vulpes vulpes L.) and the Chinese raccoon dog (Nyctereutes procyonoides procyonoides Gray). We confirm the previous findings of the KIT gene on Bs of both species, but demostrate an independent origin of Bs in these species, with two reused regions. Comparison of gene ensembles in Bs of canids, ruminants, and rodents once again indicates enrichment with cell-cycle genes, development-related genes, and genes functioning in the neuron synapse. The presence of B-chromosomal copies of genes involved in cell-cycle regulation and tissue differentiation may indicate importance of these genes for B chromosome establishment.

Whole-Genome Analysis of Mycobacterium tuberculosis from Patients with Tuberculous Spondylitis, Russia

Whole-genome analysis of Mycobacterium tuberculosis isolates collected in Russia (N = 71) from patients with tuberculous spondylitis supports a detailed characterization of pathogen strain distributions and drug resistance phenotype, plus distinguished occurrence and association of known resistance mutations. We identify known and novel genome determinants related to bacterial virulence, pathogenicity, and drug resistance.