Detection of Group 1 Coronaviruses in Bats in North America

Sample Collection

Bats were sampled at 4 sites in the Rocky Mountain region in August 2006. At sites 1 and 2, bats of 2 species were sampled in colonies inhabiting 2 buildings 480 km apart on opposite sides of the continental divide of the Rocky Mountains. Eight occult myotis (Myotis occultus) and 1 Brazilian free-tailed bat (Tadarida brasiliensis) were captured in mist nets as they emerged from a roost in a building in Mancos in Montezuma County in southwestern Colorado (site 1) at dusk on August 19. This species was previously thought to be conspecific with the little brown bat (M. lucifugus) that is common throughout North America (16). Big brown bats (Eptesicus fuscus) were sampled at a roost in a building in Fort Collins in Larimer County in north-central Colorado (site 2) on August 7. Other bats (n = 27) were sampled at sites 3 and 4 incidental to ongoing, unrelated bat faunal surveys. One western small-footed myotis (M. ciliolabrum) and 1 long-eared myotis (M. evotis) were captured in mist nets over water on August 8 at Soapstone Prairie Natural Area in Larimer County (site 3). Four big brown bats, 3 long-eared myotis, 8 occult myotis, 1 Brazilian free-tailed bat, 7 long-legged myotis (M. volans), and 2 silver-haired bats (Lasionycteris noctivagans) were trapped in mist nets during the nights of August 14–20 as they drank or foraged near open water at 2 sewage treatment lagoons (9 km apart) (site 4) in Montezuma County, Colorado. Bats were captured under authority of a Colorado Division of Wildlife Scientific Collection License following procedures approved by the Institutional Animal Care and Use Committee of the US Geological Survey, Fort Collins Science Center. Typically, each bat was sampled within 5–10 minutes of capture and then released.

Whenever possible, 3 sample types were taken from each bat (Table). Sterile calcium alginate swabs were used for oral or anal area samples that were immediately placed into 2 mL of RNAlater (Ambion, Austin, TX, USA). Fecal samples were collected if the bat produced a fresh bolus during handling. Disposable latex gloves were changed between samples, and multiple forceps used to collect fecal boluses were rinsed, wiped in ethanol, and air-dried between samples. Samples were numbered, kept in a cooler in the field, stored at 4°C, and delivered to the laboratory on August 28.

RNA Extraction and Reverse Transcription (RT)

RNA from 140 μL of each of the 79 samples was extracted by using the QIamp viral RNA mini kit (QIAGEN, Valencia, CA, USA) following the manufacturer’s instructions. Extracted RNA was eluted in 50 μL of RNase-free water and stored at –80°C. We used Moloney murine leukemia virus reverse transcriptase (Invitrogen, Carlsbad, CA, USA) with random hexamers in a 20-μL reaction to generate cDNAs by using 10 μL of RNA as a template according to the manufacturer’s instructions. All samples were extracted and analyzed in triplicate. RT products were stored at –20°C.

PCR and Sequencing

All samples were screened by PCR and nested PCR. On the basis of previous reports, PCR with a pair of consensus primers that target a highly conserved region of coronavirus gene 1b was used to screen the cDNA samples (8). Three microliters of cDNA was amplified in a 50-μL reaction containing 1.5 mmol/L MgCl₂, 0.2 mmol/L deoxynucleoside triphosophates, 2.5 U of HotStarTaq (QIAGEN), and 0.2 μmol/L of primers 1 and 2: 5′-GGTTGGGACTATCCTAAGTGTGA-3′ (primer 1) and 5′-CCATCATCAGATAGAATCATCATA-3′ (primer 2) by using the following PCR program: 15 min at 95°C; 45 cycles for 1 min at 95°C, 1 min at 48°C, and 1 min at 72°C; and 10 min at 72°C.

For nested PCR, 5 μL from each PCR was amplified in a 50-μL reaction with primer 2 and primer 3 (5′-GTTGTACTGCTAGTGACAGG-3′), an internal primer based on nucleotide sequences of the PCR amplicons by using 40 cycles of the same PCR program. All RT-PCRs were conducted in an enclosed nucleic acid workstation equipped with a UV light (Clone Zone; USA Scientific, Ocala, FL, USA) in a room separate from the main laboratory. Water controls in all RT-PCRs did not show false-positive results. To overcome possible PCR inhibitors in fecal samples, PCR was performed both on the cDNA and on a 1:10 dilution of the cDNA. Amplicons were analyzed by agarose gel electrophoresis. For each positive specimen, amplicons from 2 independent RT-PCRs were sequenced on an ABI 3730 DNA sequencer (Applied Biosystems, Foster City, CA, USA) at the University of Colorado Health Science Center Cancer Center DNA Sequencing and Analysis Core. Numbered specimens were then correlated with lists of bat samples.

Data Analysis

Viral sequences were analyzed and aligned by using ClustalW (http://workbench.sdsc.edu). Phylogenetic trees were constructed by using the neighbor-joining method in the program PAUP* version 4.0 (Sinauer Associates, Inc., Sunderland, MA, USA) rooted with porcine respiratory and reproductive syndrome virus (GenBank accession no. NC_001961). Sequences used for alignment were AF304460 (HCoV-229E), AY567487 (HCoV-NL63), DQ648858 (BtCoV 512), AY594268 (BtCoV HKU2), DQ249224 (BtCoV HKU6), DQ249226 (BtCoV HKU7), and DQ249228 (BtCoV HKU8). The deduced sequences from this study were deposited in GenBank under accession nos. EF544563–EF544568.

Identification of Rocky Mountain Bat Coronaviruses (RM-Bt-CoVs)

A total of 79 samples (28 fecal samples, 29 anal swab specimens, and 22 oral swab specimens) were collected from 57 bats of 7 species in 4 locations in the Rocky Mountain region during a 2-week period in August 2006 (Table). PCR amplicons that target a conserved region in gene 1b containing the RNA-dependent RNA polymerase common to all coronaviruses were detected in reversed-transcribed RNA from 6 of the 79 samples. All samples positive for coronavirus RNA were from the 28 fecal samples tested (Table). None of the anal region or oral swab specimens were positive for coronavirus RNA.

Despite the small number of bats sampled, there was a high prevalence of coronavirus RNA shedding in fecal samples of 2 species of bats. Five (50%) of 10 fecal samples from occult myotis and 1 (17%) of 6 fecal samples from big brown bats were positive for coronavirus in screening tests. The 1 coronavirus-positive sample from big brown bat (bat sample 65) was from feces of 1 (33%) of 3 big brown bats sampled at site 2 in north-central Colorado, whereas the positive samples from the occult myotis (bat samples 3, 6, 11, 27, and 48) were from sites 1 and 4 in southwestern Colorado, ≈480 km from site 2 (Table). Most of the fecal samples were only positive in the PCRs with cDNA diluted 1:10, which suggested that PCR inhibitors were present in feces. In addition, most of the samples were positive only in the nested PCRs, which indicated that either the RNA was present in small amounts or that the primers used were not an optimal match for these viruses.

Phylogenetic Analysis of RM-Bt-CoVs

A 440-nt sequence in the RNA-dependent RNA polymerase region of gene 1b was amplified by RT-PCR from the 6 positive samples. Analysis of nucleotide sequences of these amplicons showed that all 6 RM-Bt-CoVs are members of coronavirus group 1 (Figure 1). Although these sequences were similar to those published for Asian bat group 1 coronaviruses, there was enough dissimilarity in this highly conserved region to suggest that the Rocky Mountain specimens represent unique coronaviruses (8,9). Phylogenetic analysis of this region of gene 1b suggests that the RM-Bt-CoVs cluster in 3 subgroups within group 1. Three of the 5 specimens from the occult myotis (samples 6, 11, and 48) were in 1 cluster and the other 2 (samples 3 and 27) formed a second cluster within group 1 coronaviruses. The 1 specimen from the big brown bat (sample 65) was a more distantly related group 1 coronavirus (Figure 2).