A Multiplex Targeted NGS Panel for Identifying Pathogens in Canine Neurological and Reproductive Diseases

Jobin Jose Kattoor^1,2Eman Anis³Nelly Elshafie²Rebecca P. Wilkes^2*

• ¹Department of Infectious Diseases and Microbiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania, United States
• ²Animal Disease Diagnostic Laboratory, College of Veterinary Medicine, Purdue University, West Lafayette, Indiana, United States
• ³Pennsylvania Animal Diagnostic Laboratory System, New Bolton Center, School of Veterinary Medicine, University of Pennsylvania, Kenneth Square, United States

Multiple pathogens can infect the canine reproductive and central nervous systems, including organisms that are zoonotic, such as Brucella canis, pathogenic Leptospira spp., Anaplasma phagocytophilum, Histoplasma capsulatum, and Blastomyces dermatitidis. In this study, we developed a targeted next-generation sequencing (tNGS) panel to identify common infectious agents related to neurologic and reproductive disease in canines while incorporating less common zoonotic agents into a single test.Primer pools were developed to detect 34 pathogens and used in two multiplex PCR assays, which were combined prior to library preparation and sequencing using Ion Torrent technology. A feasibility study was performed with known positive clinical samples, bacterial isolates, or synthetic DNA (gBlocks) spiked into nucleic acids from negative canine clinical samples. Of the 34 organisms included in the panel, 33 were detectable. Analytical specificity was determined through in-silico evaluation of the primer sets as well as feasibility testing. Some primer sets were not specific for the intended target organism, based on BLAST analysis (NCBI) of the obtained sequences. Analytical sensitivity was evaluated with dilutions of gBlocks spiked into negative canine nucleic acid samples (for 10 organisms) or via testing dilutions of positive clinical samples and compared to Ct values from real-time PCR assays (for 13 organisms). Compared to real-time PCR assays, pathogens could be detected at Ct values from 33-38, depending on the pathogen, and at approximately 100-1000 copies based on gBlock testing. Diagnostic sensitivity and specificity testing were performed based on comparisons to those results from bacterial culture and real-time PCR assays. A total of 76 samples (39 positive and 40 negative) were tested, representing neurological and reproductive samples. Diagnostic sensitivity and Formatted: Superscript specificity for the assay were calculated as 89% and 98%, respectively. The tNGS assay had the added benefit of strain-typing Canine distemper virus and Canine parvovirus-2 in the positive samples. For reproducibility, a blinded panel was tested by our laboratory and another laboratory using the same tNGS protocol where the assay had an agreement for 16 out of 18 samples with a Cohen's kappa value of 0.77, indicating high reproducibility.