Comparison of five methods for isolation of DNA from Mycoplasma cynos

Development of a long-read next generation sequencing workflow for improved characterization of fastidious respiratory mycoplasmas

Mycoplasma cynos and Mycoplasma felis are often associated with canine and feline infectious respiratory disease in dogs and cats, respectively. Mycoplasmas have a reduced genome and dearth of many biosynthetic pathways, making them dependent on rich medium for growth. Due to this fastidious nature, mycoplasmas have been historically underdiagnosed. The aim of this study was to develop a cost-effective and accurate sequencing workflow for genotypic characterization of clinical isolates of M. cynos and M. felis using a rapid long-read sequencing platform. We explored the following critical aspects of bacterial whole genome sequencing, including: (i) five solid and liquid-based culture approaches based on a specialized media formulation for Mycoplasma culture, (ii) three DNA extraction methods modified for long-read sequencing purposes, and (iii) two de novo assembly platforms, Flye and Canu, as key components of a bioinformatics pipeline. DNA extraction method 1, a solid-phase and column-based kit with enzymatic lysis, provided the best DNA quality and concentration followed by high coverage and sequencing contiguity. This was obtained with a culture volume of 45 ml in modified Hayflick's broth incubated for 48 h. DNA extracted directly from colonies on agar or from small broth volumes (6 ml) did not meet the criteria required for long-read sequencing. Overall, Flye generated more contiguous assemblies than the Canu assembler and was more time efficient. This 4-5 day sample-to-sequence workflow provides the scientific and clinical communities with a more comprehensive tool than laborious conventional methods for complete genomic characterization of M. cynos and M. felis clinical isolates.

Analysis of the lung microbiota in dogs with Bordetella bronchiseptica infection and correlation with culture and quantitative polymerase chain reaction

Infection with Bordetella bronchiseptica (Bb), a pathogen involved in canine infectious respiratory disease complex, can be confirmed using culture or qPCR. Studies about the canine lung microbiota (LM) are recent, sparse, and only one paper has been published in canine lung infection. In this study, we aimed to compare the LM between Bb infected and healthy dogs, and to correlate sequencing with culture and qPCR results. Twenty Bb infected dogs diagnosed either by qPCR and/or culture and 4 healthy dogs were included. qPCR for Mycoplasma cynos (Mc) were also available in 18 diseased and all healthy dogs. Sequencing results, obtained from bronchoalveolar lavage fluid after DNA extraction, PCR targeting the V1–V3 region of the 16S rDNA and sequencing, showed the presence of Bb in all diseased dogs, about half being co-infected with Mc. In diseased compared with healthy dogs, the β-diversity changed (P = 0.0024); bacterial richness and α-diversity were lower (P = 0.012 and 0.0061), and bacterial load higher (P = 0.004). Bb qPCR classes and culture results correlated with the abundance of Bb (r = 0.71, P < 0.001 and r = 0.70, P = 0.0022). Mc qPCR classes also correlated with the abundance of Mc (r = 0.73, P < 0.001). Bb infection induced lung dysbiosis, characterized by high bacterial load, low richness and diversity and increased abundance of Bb, compared with healthy dogs. Sequencing results highly correlate with qPCR and culture results showing that sequencing can be reliable to identify microorganisms involved in lung infectious diseases.

Evaluation of random amplified polymorphic DNA and multiple-locus variable number tandem repeat analyses for Mycoplasma cynos

Mycoplasma spp. can cause diseases of the respiratory system as well as urogenital infections, infertility, and anemia. The members of this genus have a low G + C content compared to other bacteria. Because primers used in the random amplified polymorphic DNA (RAPD) technique are only 10 bp long and have high GC content, this method can be inadequate for genotyping Mycoplasma spp. isolates. The aim of this study was to develop and evaluate multiple-locus variable number tandem repeat analysis (MLVA) and two-primer RAPD (TP-RAPD) procedures for subtyping Mycoplasma cynos isolates. A total of 55 M. cynos isolates obtained from 162 bronchoalveolar lavage fluid samples from shelter and pet dogs were used in this study. Seventy-four tandem repeat regions were detected in the M. cynos genome, and two of these loci were determined to be suitable and used for development of the MLVA scheme. The results of variable number tandem repeat (VNTR) analysis and TP-RAPD-PCR were compared with RAPD-PCR. The discriminatory power of TP-RAPD-PCR (Hunter-Gaston diversity index [HGDI] = 0.84) was higher than those of RAPD-PCR (HGDI = 0.727), VNTR1 (HGDI = 0.8), and VNTR3 (HGDI = 0.757). We observed that the TP-RAPD-PCR and MLVA methods provide clearer data and are more successful in determining genetic diversity, in contrast to the RAPD-PCR method for this species.