Evaluation of a real-time RT-PCR panel for detection of SARS-CoV-2 in bat guano

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of coronavirus disease 2019 (COVID-19), which is an ongoing global health concern. The exact source of the virus has not been identified, but it is believed that this novel coronavirus originated in animals; bats in particular have been implicated as the primary reservoir of the virus. SARS-CoV-2 can also be transmitted from humans to other animals, including tigers, cats, and mink. Consequently, infected people who work directly with bats could transfer the virus to a wild North American bat, resulting in a new natural reservoir for the virus, and lead to new outbreaks of human disease. We evaluated a reverse-transcription real-time PCR panel for detection of SARS-CoV-2 in bat guano. We found the panel to be highly specific for SARS-CoV-2, and able to detect the virus in bat guano samples spiked with SARS-CoV-2 viral RNA. Our panel could be utilized by wildlife agencies to test bats in rehabilitation facilities prior to their release to the wild, minimizing the risk of spreading this virus to wild bat populations.

Coronaviruses (CoVs) are found in a wide variety of animals in which they can cause disease.^2,9 Coronaviruses that infect animals, although rarely, can sometimes spread to humans. Severe acute respiratory syndrome (SARS), Middle Eastern respiratory syndrome (MERS), and coronavirus disease 2019 (COVID-19), are examples of diseases caused by CoVs that originated in animals and spread to humans to cause disease.^7,11,20 It is believed that the source of SARS-COV-2 was bats. SARS-CoV-2 spreads mostly person-to-person through respiratory droplets created by coughing and sneezing.¹⁹ However, some animals, including tigers, mink, cats, and dogs, have contracted the virus from either symptomatic or asymptomatic people.^4,12,14

Although little is known about the likelihood of reverse zoonotic exposure and infection, many federal agencies in the United States have taken steps to restrict wildlife rehabilitation activities involving bats. The current guidance in response to the COVID-19 outbreak is that wildlife agency biologists and researchers postpone activities that involve direct handling of bats, and that rehabilitators suspend the release of any bats currently held in their facilities.¹ Therefore, it is important to have a rapid test to detect SARS-CoV-2 in North American bats to mitigate the chance of creating a new natural reservoir for the virus.

Reverse-transcription real-time PCR (RT-rtPCR) is, to date, the standard assay for detection of SARS-CoV-2 in people.¹⁷ We evaluated a RT-rtPCR panel for detection of SARS-CoV-2 in bat guano. Our panel was adapted from the Centers for Disease Control and Prevention (CDC) SARS-CoV-2 panel.¹⁷ The panel targets 2 different regions of the SARS-CoV-2 nucleocapsid gene (N1 and N2) to avoid false-negative results as a result of potential genetic mutation. Each target primers–probe set was run in a separate RT-rtPCR assay (N1 and N2 assays). Additionally, an internal positive control (IPC) primer–probe set (VetMax Xeno IPC-VIC assay; Thermo Fisher) was included to detect an exogenous RNA control (VetMax Xeno IPC RNA; Thermo Fisher) that was spiked into all specimens during the extraction process to monitor for PCR inhibitors and to confirm the efficiency of specimen preparation and extraction.

To evaluate the analytical performance of each assay, 2 different RT-rtPCR kits (TaqPath and AgPath ID one step RT-PCR kits; Thermo Fisher) were used. For the AgPath ID kit, 25-µL total volume reactions included 2× RT-PCR buffer, 25× enzyme mix, primers–probes (final concentration of 500 nM of each primer and 125 nM of each probe; IDT),¹⁷ and 5 μL of extracted nucleic acid (NA). The RT-rtPCR was performed (ABI 7500 real-time PCR system, or the QuanStudio 5 PCR system; Thermo Fisher) with the following cycle parameters: 10 min at 48°C for reverse transcription and 10 min of inactivation at 95°C, followed by 40 cycles of 15 s at 95°C and 45 s at 60°C. For the TaqPath kit, 20-µL total volume reactions included 4× master mix, primers–probes (the same final concentration as mentioned above), and 5 μL of extracted NA. The RT-rtPCR thermocycler conditions matched the CDC protocol.¹⁷

The analytical sensitivity was determined for each target (assay) by testing 10-fold serial dilutions of SARS-CoV-2 RNA (kindly provided by Dr. Hon Ip, USGS National Wildlife Health Center, Madison, WI). This viral RNA was extracted from a SARS-CoV-2 isolate of known titer (10⁷ TCID₅₀/mL). The viral RNA dilutions (3 replicates per dilution) were tested by the panel using both RT-rtPCR kits to produce standard curves, and the efficiency of each assay was determined. Also, the analytical sensitivity of the 2 assays was evaluated using 10-fold serial dilutions of plasmid DNA containing the complete nucleotide gene of SARS-CoV-2 (2019-nCoV plasmid, 200,000 copies/µL; IDT).

To assess the repeatability of both assays, the intra- and inter-assay coefficients of variation (CVs) were calculated. The intra-assay variability was calculated using 5 replicates each of 3 different dilutions of plasmid. Inter-assay variability was determined with these dilutions tested in 6 different runs.

Specificity of both assays was evaluated by testing control plasmids (IDT) that contain the complete NC gene from SARS (Betacoronavirus group B, Bat SARS-like coronavirus isolate bat-SL-CoVZC45, GenBank MG772933.1) and MERS (Betacoronavirus group C, MERS-related coronavirus isolate KNIH/002_05_2015, GenBank MK796425.1). Confirmed PCR-positive specimens available in the Molecular Diagnostics section of the Pennsylvania Animal Diagnostic Laboratory System New Bolton Center Laboratory (PADLS-NBC; Kennett Square, PA) for infectious bronchitis virus (Gammacoronavirus), bovine coronavirus (Betacoronavirus group A), and equine coronavirus (Betacoronavirus group A) were tested via the panel as well. Further, the primers and probes were aligned against various bat CoV nucleotide sequences (Suppl. Table 1) available in GenBank (representative of bat alphacoronavirus and bat betacoronavirus groups B–D) using MAFT (Geneious software v.9.1.2; Biomatters)

To assess analytical performance with a prospective specimen matrix, 20 bat guano specimens were collected from clinically healthy bats from a rehabilitation facility that applies strict biosecurity measures, to decrease the chance that the bats used in our study could have been exposed to SARS-CoV-2 following intake. The biosecurity measures included wearing disposable gloves during handling of bats and wearing face masks and clothing that remained within the facility. New bat arrivals were quarantined for at least 1 wk in a separate room before being housed within the main room with the other bats. Guano pellets were placed in sterile labeled containers and shipped overnight on ice packs to PADLS-NBC; specimens were stored at −80°C until tested.

To confirm the absence of SARS-CoV-2 in the collected bat guano, the specimens were processed and tested with the SARS-CoV-2 panel. Nucleic acid was extracted from the specimens (QIAamp viral RNA mini kit; Qiagen) according to the manufacturer’s instructions with some modifications. Briefly, bat guano was suspended in 500–1,000 μL of phosphate-buffered saline (maximum dilution 1:10). The guano suspension was centrifuged at 4,000 × g for 5 min. Then, the manufacturer’s protocol was followed for purification of the NA through the column. To assess the limit of detection (LOD) of each assay using the specimen matrix, a pool of guano suspension was prepared from 2 guano specimens. Then, 10-fold and 2-fold serial dilutions of quantitative synthetic SARS-CoV-2 RNA (ORF, E, N [ATCC] of known concentration [10⁵ copies/μL]) was used to spike the aliquoted guano suspensions (5 replicates for each dilution). Then, NA was extracted from each suspension (pre-extraction spiked samples) and tested via the SARS-CoV-2 panel as mentioned above. Additionally, extracted bat guano NAs (post-extraction spiked samples) were spiked with 10-fold serial dilutions of the synthetic RNA and tested by the panel. The LOD was determined as the lowest concentration in which ≥95% (5 of 5) of the replicates were positive. Given that no positive guano specimens are available, the 20 bat guano suspensions were spiked with different concentrations of SARS-CoV-2 RNA (2–4 μL of viral RNA [dilutions 1 or 2] purified from SARS-CoV-2 isolate of known titer of 10⁷ TCID₅₀/mL) to mimic clinically positive samples. Then, the guano suspensions were processed and tested by the panel described above.

Our panel was highly sensitive, with detection of ~10 and ~100 copies/reaction of N1 and N2 targets, respectively, when tested using the TaqPath kit, and 100 copies/reaction of N1 and N2 targets when tested using the AgPath kit (Table 1). The analytical sensitivity of both N1 and N2 assays was 1 TCID₅₀/mL of the titrated virus when tested using the TaqPath kit, and 0.1 TCID₅₀/mL of the titrated virus when tested using the AgPath kit (Table 1). The standard curves produced with dilutions of the viral RNA were linear over a wide range of dilutions for both assays. N1 and N2 assays yielded R² values of 0.997 and 0.99 when tested using TaqPath, and 0.989 and 0.984 when tested using the AgPath kit, respectively. The efficiencies of the N1 and N2 assays were ≥90.5%. The assays were reproducible, with intra-assay CVs of 0.07–0.5% and inter-assay CVs of 0.47–1.6%. There was no amplification from non–SARS-COV-2 agents.

Table 1.

Analytical sensitivity of N1 and N2 assays using TaqPath and AgPath kits.

SARS-CoV-2 target	RNA from virus isolate with known titer (TCID50/mL)	TaqPath kit (mean Ct)	AgPath kit (mean Ct)	SARS-CoV-2 plasmid (copies/reaction)	TaqPath kit (mean Ct)	AgPath kit (mean Ct)
N1	100,000	17.5	NT	100,000	25.0	14.8
	10,000	21.3	21.6	10,000	28.4	28.2
	1,000	25.0	25.1	1,000	31.7	31.8
	100	28.3	28.5	100	34.4	34.9
	10	32.1	32.6	10	35.5	U
	1	35.6	36.6	1	U	U
	0.1	U	38.9	0.1	NT	NT
	0.01	U	U	0.01	NT	NT
N2	100,000	17.3	NT	100,000	26.1	25.1
	10,000	21.2	22.4	10,000	30.2	29.1
	1,000	25.0	25.9	1,000	33.8	32.8
	100	28.4	29.4	100	37.7	36.5
	10	32.5	33.6	10	U	U
	1	36.1	35.2	1	U	U
	0.1	U	38.9	0.1	NT	NT
	0.01	U	U	NT	NT	NT

When the cycle threshold (Ct) value of one or more of the replicates per dilution was undetermined, the mean Ct value of this dilution was considered undetermined.

NT = not tested; U = undetermined.

Alignment of the panel primer and probe sequences with various bat betacoronavirus, group B, nucleotide sequences showed that the forward primer sequence of the N1 assay has at least 2 mismatches in the 3'-end and 1 internal mismatch (Fig. 1A). Further, the reverse primer has at least 2 internal mismatches, and the probe has at least 1 internal mismatch and 1 mismatch in the 5'-end. The impacts of mismatches between primers–probes and template on the PCR amplification depend upon the position and number of mismatches. During PCR amplification, the DNA polymerases catalyze the addition of nucleotides to the primer 3'-OH, as specified by complementarity to the template DNA. Therefore, mismatches located in the 3'-end of a primer are known to be exceptionally detrimental to PCR priming.^8,18 Combinations of multiple mismatches in the 3'-end of a primer (either 2 or 3) have been documented to lead to abolishment of amplification; purine–purine and C–C mismatches have the most detrimental effect.¹⁵ The probe of the N2 assay has at least 2 mismatches in the 3'-end and 2 internal mismatches (Fig. 1B). PCR amplification is more prone to mismatches in the probe region, where even a single mismatch may lead to negative results.^5,10 These results revealed that the primer and probe combinations of both assays are designed for the specific detection of SARS-CoV-2.

Figure 1.

Alignment of the primers–probes and representatives of bat betacoronavirus group B. A. Alignment of N1 primers–probe shows that the probe and reverse primer sequences of the N1 assay has high sequence homology with bat SARS-like CoV sequences (2 nucleotide mismatches in the probe at nts 1 and 13, and 2–4 nucleotide mismatches in the reverse primer) and bat CoV RaTG13. The forward primer sequence showed no significant homology with any of the evaluated bat CoV sequences except bat CoV RaTG13, GenBank MN996532 (2 nucleotide mismatches at nts 12 and 14). B. Alignment of N2 primers–probe shows that the forward and reverse primer sequences showed high sequence homology to bat SARS-like CoV sequences. However, the probe sequences showed no significant homology with any of the evaluated bat CoV sequences (at least 5 nucleotide mismatches) except bat CoV RaTG13, GenBank MN996532 (2 nucleotide mismatches at nts 7 and 10).

It is important to note that the sequences of the primers and probe of both assays showed high sequence homology to bat CoV RaTG13. Phylogenetic analysis showed that bat CoV RaTG13 is the closest relative of SARS-CoV-2.²⁰ Bat CoV RaTG13 was only detected in Rhinolophus affinis from Yunnan Province, China; R. affinis is a bat species of the Rhinolophidae family that is very widespread throughout much of South Asia, southern and central China, and Southeast Asia. Also, there is much more CoV diversity in the old-world bat species than the new-world bat species. However, given that bats are known to harbor various species of CoV and the high mutation rate of RNA viruses,^6,20 it is recommended that any positive specimens be further tested using pancoronavirus primers and then sequenced⁶ to determine the genotype of the detected CoV.

All non-spiked bat guano specimens tested with the panel were initially negative for SARS-CoV-2; the IPC was detected in all samples. These results demonstrate that there is no cross-reactivity with the bat genome or bat microflora that would produce potential false-positive RT-rtPCR results. We used bat guano because research has shown that this specimen type is highly associated with CoV positivity.¹⁶ For the pre-extraction spiked bat guano suspensions, the LOD (Table 2) was 100 and 250 copies/μL for N1 and N2 assays using both RT-rtPCR kits, respectively. For the post-extraction spiked bat guano suspensions, the LOD was 1 copy/μL for both N1 and N2. The LOD was generally higher in pre-extraction spiked guano suspensions; this is very likely the result of the loss of synthetic RNA during the extraction step and not a result of RT-rtPCR inhibitors in the sample matrix. Both RT-rtPCR kits were able to identify SARS-CoV-2 RNA in the 20 spiked specimens (100%).

Table 2.

Limit of detection of N1 and N2 assays using specimen matrix.

SARS-CoV-2 target	Synthetic RNA spiked into bat guano samples prior to extraction (copies/µL)	TaqPath kit (mean Ct)	AgPath kit (mean Ct)	Synthetic RNA spiked into bat guano samples post-extraction (copies/µL)	TaqPath kit (mean Ct)	AgPath kit (mean Ct)
N1	1,000	35.6	35.6	1,000	NT	NT
	100	38.3	38.4	100	28.0	29.0
	10	U	U	10	31.8	31.9
	1	U	U	1	34.4	35.3
	500	35.2	35.6	500	NT	NT
	250	36.8	36.6	250	NT	NT
N2	1,000	37.5	36.9	1,000	NT	NT
	100	U	U	100	28.8	29.5
	10	U	U	10	31.8	32.8
	1	U	U	1	35.7	36.7
	500	36.2	35.6	500	NT	NT
	250	36.8	36.6	250	NT	NT

When the cycle threshold (Ct) of one or more of the replicates per dilution was undetermined, the mean Ct value of this dilution was considered undetermined.

NT = not tested; U = undetermined.

The overall performance of the N1 and N2 assays using both RT-rtPCR kits was well above the pre-defined required level;^3,13 therefore, the assays can be used for routine detection of SARS-CoV-2 in bat guano. As demand for SARS-CoV-2 testing continues to rise worldwide, some laboratories are facing shortages of critical reagents for virus detection; hence, the availability of multiple kit types is advantageous. To date, there is limited information about the susceptibility of North American bats to infection by SARS-COV-2, and it is difficult to determine the risk of transmission from humans to bats when handling them. Our RT-rtPCR panel could be utilized by wildlife agencies to test bats and permit rehabilitators to release bats currently under their care with minimal risk of spreading the virus among North American bats and other wild animal populations. Our panel could also be used to conduct passive surveillance to determine if SARS-CoV-2 is already present in wild North American bats.