Parvovirus

The fresh tissue samples including heart, lung, liver, spleen, and kidney of all fishing cats plus additional intestinal and mesenteric lymph node tissues of fishing cat no. 3, were subjected to viral nucleic acid extraction by individually homogenizing them with 1% (v/v) phosphate-buffered saline (PBS). Subsequently, the total viral nucleic acid was extracted using a commercial viral nucleic acid extraction II kit (Geneaid, Taipei, Taiwan) following the manufacturer’s suggestion. The extracted nucleic acids were then qualified and quantified by their A₂₆₀/A₂₈₀ absorbance ratio using the spectrophotometry method (NanoDrop, Thermo Scientific™, Waltham, MA, USA). All extracted nucleic acids were further subjected to viral molecular investigation using polymerase chain reaction (PCR) analysis with several pan-virologic-family PCR testing panels including the detection of the herpesvirus [23 (link)], paramyxovirus [24 (link)], pneumovirus [25 (link)], calicivirus [26 (link)], influenza virus [27 (link)], bocavirus [26 (link), 28 (link)], parvovirus [29 (link)] and coronavirus [30 (link), 31 (link)]. The PCR protocols, reagents, cycling conditions and positive/negative controls used in the reactions are described in the S1 File. PCR detection specific of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene of feline was used as an internal control as described previously [32 (link)]. Subsequently, the positive PCR amplicons were visualized using a QIAxcel capillary electrophoresis platform (Qiagen, Hilden, Germany) as previously described [33 (link)]. The positive amplicons of each PCR assay were purified and subjected to bidirectional Sanger sequencing (Macrogen Inc, Seoul, South Korea). Due to moderate autolysis of the tissues, it was not possible to perform bacteriological investigation on these fishing cats.

Samples 712 and 808 containing LuJo virus were prepared from human isolates [10 (link)]. RNA was extracted from the cerebrospinal fluid and serum of a liver transplant recipient. After digestion with DNase I to eliminate human chromosomal DNA, RNA preparations were amplified by means of reverse-transcriptase PCR (RT-PCR) with the use of random primers [11 (link), 12 (link)]. Amplification products were pooled and sequenced with the use of the 454 Genome Sequencer FLX platform (Roche, Branford, CT), but DNA fragmentation was omitted. The Zaria bat coronavirus samples 819 and 820 (and the negative control 806) were obtained from the GI tract of bats that tested positive (and negative for the control) for coronavirus by PCR [13 (link)]. Sample 28 containing GBV-D was obtained from bat serum [14 (link)] and prepared as detailed previously. The isolated RNA for both coronavirus and GBV-D samples was converted to cDNA and the library was prepared similarly to the LuJo virus isolates detailed above. The bat parvovirus sample, 1164, was obtained from the spleen of parvovirus PCR-positive bats (like those discovered in [15 (link), 16 (link)]), and DNA was isolated and the prepared libraries were sequenced on the 454 FLX (Roche, Branford, CT). Samples containing MERS-CoV (1500, 1501) [17 ] were prepared as previously described [18 (link)]. Viral cDNA was made using random primer RT-PCR from nasal swabs of camels. Further PCR amplifications were made using overlapping PCR primers spanning 2.0–2.5 kb fragments of MERS-CoV [19 ]. These amplification products were pooled and sequenced on the Ion Torrent PGM platform. The human serum spiked samples containing Y. pestis, F. tularensis, and B. anthracis, ​B. mallei, and B. psuedomallei were prepared for sequencing as described previously [20 (link), 21 (link)] and sequenced on 454 FLX (Roche, Branford, CT), Ion Torrent PGM (Life Technologies, Grand Island, NY), and Illumina MiSeq platforms (Illumina, San Diego, CA). SRA information for each sample analyzed here are available through the NBCI BioProject # PRJNA276557.

Age- and sex-matched non-obese diabetic (NOD) mice born and bred within our animal facility were used for all experiments. Mice were housed in OptiMouse cages with corn cob bedding and up to 5 cage-mates. Mice had ad libitum access to chow and reverse osmosis chlorinated (2-3ppm)-purified water. Housing rooms were kept on a 14.5/9.5-hour light/dark cycle with temperature maintained at 22-25°C and 50-70% humidity. Sentinel mice were placed on dirty bedding and nesting material in experimental rooms and subsequently tested on a quarterly basis for presence of parasites (pinworms and fur/follicular mites, Pneumocystis spp. (carinii, murina)), bacteria (Mycoplasma pulmonis), and viruses (RADIL Comprehensive Panel) including Mouse Hepatitis Virus (MHV), Mouse Minute Virus (MMV), Mouse Parvovirus (NS1 – Generic Parvovirus & MPV 1-5), Theiler’s Murine Encephalitis Virus (TMEV), Epizootic Diarrhea of Infant Mice (EDIM), Sendai Virus, Pneumonia Virus of Mice (PVM), Reo3 virus (REO3), Lymphocytic Choriomeningitis Virus (LCMV), Ectromelia virus, Murine Adenovirus I/II (MAV1/MAV11), and Polyomavirus. All experiments were performed in compliance with protocols approved by the Animal Care Committee of The University of British Columbia.
Virus stocks for coxsackievirus B4 Edwards strain 2 (CVB4) were prepared as previously described (15 (link)). Normoglycemic NOD mice at 11-12 weeks old were injected intraperitoneally with either 400 plaque-forming units (pfu) CVB4 or DMEM vehicle. Non-fasting blood glucose was monitored using a OneTouch LifeScan monitor. Mice were considered diabetic with two consecutive readings above 16.2 mg/dL separated by 24 hours.

Blood samples were collected at enrollment and stored at −80 °C. Within 24 h, the blood samples were transferred to BGI Medical Laboratory Institute (Wuhan, China) for mNGS detection. The mNGS detection was conducted according to the manufacturer’s protocol, which is reported in detail in Supplemental File S2.
With reference to previous literature, we have formulated the following criteria for positive mNGS results [16 (link)].
For bacteria (mycobacteria excluded), fungi, viruses, and parasites: mNGS identified a microbe (genus level) as positive whose reads were 3-fold greater than that of any other microbes. If only one pathogen was detected, it would be regarded as positive.
For mycobacteria: Mycobacterium tuberculosis was considered positive as long as there was more than 1 read mapped to the species or genus level.
Under the condition that the above criteria were met, the patient’s clinical characteristics were taken into consideration. Parvovirus, human herpes virus 5, etc., were appropriately excluded by clinicians and laboratory staff due to the consideration of laboratory contamination. Other results were treated as positive.