Use of mitochondrial sequencing to detect gene doping in horses via gene editing and somatic cell nuclear transfer.

Gene editing and subsequent cloning techniques offer great potential not only in genetic disease correction in domestic animals but also in livestock production by enhancement of desirable traits. The existence of the technology, however, leaves it open to potential misuse in performance-led sports such as horseracing and other equestrian events. Recent advances in equine gene editing, regarding the generation of gene-edited embryos using CRISPR/Cas9 technology and somatic cell nuclear transfer, have highlighted the need to develop tools to detect potential prohibited use of the technology. One possible method involves the characterisation of the mitochondrial genome (which is not routinely preserved during cloning) and comparing it with the sequence of the registered dam. We present here our approach to whole-mitochondrial sequencing using tiled long-range PCR and next-generation sequencing. To determine whether the background mutation rate in the mitochondrial genome could potentially confound results, we sequenced 10 sets of dam and foal duos. We found variation between duos but none within duos, indicating that this method is feasible for future screening systems. Analysis of WGS data from over 100 Thoroughbred horses revealed wide variation in the mitochondria sequence within the breed, further displaying the utility of this approach.

Direct sequence confirmation of qPCR products for gene doping assay validation in horses.

The misuse of gene therapy by the introduction of transgenes via plasmid or viral vectors as a doping agent is an increasing concern in human and animal sports, not only in consideration to fair competition but also in potential detrimental effects to welfare. Doping events can be detected by polymerase chain reaction (PCR) amplification of a transgene-specific region of DNA. Quantitative real-time PCR (qPCR) is particularly suited to confirmatory investigations where precise limits of detection can be calculated. To fully validate a qPCR experiment, it is highly desirable to confirm the identity of the amplicon. Although post-PCR techniques such as melt curve and fragment size analysis can provide strong evidence that the amplicon is as expected, sequence identity confirmation may be beneficial as part of regulatory proceedings. We present here our investigation into two alternative processes for the direct assessment of qPCR products for five genes using next-generation sequencing: ligation of sequence-ready adapters to qPCR products and qPCR assays performed with primers tailed with Illumina flow cell binding sites. To fully test the robustness of the techniques at concentrations required for gene doping detection, we also calculated a putative limit of detection for the assays. Both ligated adapters and tailed primers were successful in producing sequence data for the qPCR products without further amplification. Ligated adapters are preferred, however, as they do not require re-optimisation of existing qPCR assays.

Screening for gene doping transgenes in horses via the use of massively parallel sequencing.

Throughout the history of horse racing, doping techniques to suppress or enhance performance have expanded to match the technology available. The next frontier in doping, both in the equine and human sports areas, is predicted to be genetic manipulation; either by prohibited use of genome editing, or gene therapy via transgenes. By using massively-parallel sequencing via a two-step PCR method we can screen for multiple doping targets at once in pooled primer sets. This method has the advantages of high scalability through combinational indexing, and the use of reference standards with altered sequences as controls. Custom software produces transgene-specific amplicons from any Ensembl-annotated genome to facilitate rapid assay design. Additional scripts batch-process FASTQ data from experiments, automatically quality-filtering sequences and assigning hits based on discriminatory motifs. We report here our experiences in establishing the workflow with an initial 31 transgene and vector feature targets. To evaluate the sensitivity of parallel sequencing in a real-world setting, we performed an intramuscular (IM) administration of a control rAAV vector into two horses and compared the detection sensitivity between parallel sequencing and real-time qPCR. Vector was detected by all assays on both methods up to 79 h post-administration, becoming sporadic after 96 h.