GENOMIC CHARACTERIZATION OF BRUCELLA STRAINS ISOLATED FROM HUMAN BRUCELLOSIS PATIENTS IN KUWAIT

Abu Salim Mustafa

Department of Microbiology, Faculty of Medicine, Kuwait University, Kuwait

Abstract

Human brucellosis is the most common zoonotic infectious disease of word-wide distribution with an estimated 500,000 cases occurring annually. Brucellosis is present on all inhabited continents and it is endemic in many countries and regions of the world, including Kuwait and the Middle East. There are at least four species (Brucella abortus, Brucella melitensis, Brucella suis and Brucella canis,) and more than one dozen strains of Brucellae that cause human disease in various parts of the world. However, Brucella genomes are highly conserved among the various species and strains, and, therefore, their genotypic identification has been hindered using classical molecular methods, e.g. PCR and 16S rDNA sequencing, etc. Among the various species of Brucella causing infections in humans, B. melitensis is the most pathogenic species causing severe disease. The whole genome sequencing (WGS), by using next generation sequencing (NGS) technologies, is emerging as a method of choice for genomic characterization and genotypic identification of bacteria, including Brucella. This study was aimed to characterize 25 clinical isolates of B. melitensis isolated from patients attending various hospitals in Kuwait by NGS using Illumina MiSeq. Furthermore, bioinformatics and comparative genomics analyses were performed to identify the genotypes existing in Kuwait and determine the genetic differences among the isolates. All clinical isolates of B. melitensis were grown on culture plates and single colonies were suspended in saline. The bacterial suspensions were heated at 95°C for 10 minutes and DNA was purified using the QIAamp DNA mini kit (Qiagen, Hilden, Germany). DNA libraries were prepared using the Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA). The genomic DNA were sequenced using Illumina MiSeq paired-end (2 × 150 bp) sequencing technology. The data quality was checked with FastQC. Reads were trimmed and quality filtered using FastXtool (http://hannonlab.cshl.edu/fastx_toolkit/). The de novo assembly was performed using VelvetOptimiser, Velvet 1.2.10. The best velvet assembly was merged with spades assembly by using SPAdes 3.8. Quast was used to check the assembly quality. After ordering the obtained contigs against B. melitensis bv.1 str. 16M using Mauve contig aligner, each draft assembly was submitted to NCBI for annotation with PGAP 3.3. SNPs were detected relative to the genome of reference strain B. melitensis bv.1 str. 16M (GeneBank assembly accession: GCA_000007125.1) using BioNumerics 7.6 (Applied Maths, Belgium). All the assemblies had 57.2% G + C content and 3 rRNAs. Other assembly/genome characteristics, i.e. mean coverage, number of contigs, N50 contig length, assembly size, and number of genes, pseudogenes, tRNAs, protein coding sequences, genotypes and private as well as public SNPs were also identified. The most common genotypes were the genotype IIb and a new genotype IIg with some strains of genotype III and I. In addition to public and private SNPs, the genes with higher possibility of mutations were also identified. In conclusion, the WGS, followed by bioinformatics and comparative genomics, are useful to identify genotypes and find variations across a specific strain of B. melitensis. This information will be useful in contact-tracing and epidemiological studies. Supported by Kuwait University Research Sector grants MI04/15 and SRUL02/13.

Keywords: Brucella, Genomics, Whole genome sequencing, SNP analysis, Genotyping.