Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is an enveloped, positive-stranded RNA virus and belongs to sarbecovirus in the genus Betacoronavirus, subfamily Orthocoronavirinae, family Coronaviridae1. SARS-CoV-2 has caused the pandemic of coronavirus disease 2019 (COVID-19), which began at the end of 2019, leading to global economic disruption and extensive changes to society and personal lifestyle.

Until now, its origin has not been fully understood. Bats are known to be a reservoir for SARS-CoV and SARS-related CoVs (SARSr-CoVs), thus making them a likely source of SARS-CoV-22,3,4. In previous reports, strains related to SARS-CoV-2 were identified in bats in China5. In Japan, Cambodia, and Thailand, SARSr-CoVs were also detected from insectivorous bats, Rhinolophus sp.6. In addition, one lineage of pangolin coronaviruses had 97.4% amino acid identity to the receptor binding domain (RBD) with SARS-CoV-27. Therefore, it is possible that bats are the natural host of SARS-CoV-2 and pangolins are one of the important animals in the transmission cycle of SARS-CoV-2 or SARSr-CoVs5,8.

In the current outbreak of COVID-19, natural SARS-CoV-2 infection of animals from SARS-CoV-2-positive pet owners and/or zookeepers has been reported, including cats, dogs, tigers, lions, minks, and gorillas9,10,11,12,13,14,15,16. Moreover, animal models of SARS-CoV-2 infection for preventive or therapeutic agents have been investigated, such as non-human primates, ferrets, cats, raccoon dogs, white-tailed deer, hamsters, mice, bats, and shrews17,18,19,20,21,22,23,24,25,26,27,28. These studies demonstrated that in addition to cats, rhesus macaques, hamsters, ferrets, and fruit bats have mild to high susceptibility to SARS-CoV-2 infection and thus could be useful for research as animal models.

To date, natural SARS-CoV-2 infections in dogs and cats have been reported in Japan29,30,31. In addition, pet owners were reported to have been infected with SARS-CoV-2 from their pet hamsters and cats32,33,34,35. These cases imply the circulation of the virus among humans and companion animals. In experimental infection, SARS-CoV-2 could replicate efficiently in cat tissues, with infected animals being asymptomatic or showing mild symptoms24,25. However, these studies, which used cats infected with a high dose of 105 TCID50, do not seem to reflect the conditions of natural infection. It is important to establish an experimental SARS-CoV-2 infection model in cats to imitate natural infection. This model could also contribute to confirming the possible transmission routes of SARS-CoV-2. A suitable animal model should be used for the evaluation of newly developed vaccines and therapeutic methods.

Since 2020, several variants with one or more mutations in their genome have newly emerged. The World Health Organization (WHO) designated a set of variations, as variants of interest (VOIs) and variants of concern (VOCs)36. To date, many variants have appeared and disappeared36. The alpha variant (B.1.1.7) was initially detected in the United Kingdom, and it has 9 mutations, including the well-known N501Y and D614G mutations, in the viral spike glycoprotein (S protein)37. The beta variant (B.1.351), which was detected in South Africa, contains three mutations, including E484K, in the receptor binding domain (RBD) of S protein38. The gamma variant (P.1) was discovered and caused an outbreak in Brazil39. The first reported cases of Gamma variant infection were from someone who had visited Brazil and were reported in Japan40. The delta variant (B.1.617.2) is a well-known variant, which caused a worldwide outbreak, and it was first reported in India41. It contains important mutations which cause the S protein to have strong affinity for the human angiotensin-converting enzyme 2 (ACE2) receptor and increase infectivity by facilitating cleavage of the S precursor protein to the active S1/S2 configuration42,43. These mutations were responsible for the faster spread of these variants in comparison to previous variants. The omicron variant (B.1.1.529) was first reported in Africa and it was the dominant VOC in 202244. This omicron variant includes several subvariants45. They have huge mutations, including three mutations at the furin cleavage site that facilitate transmission of the virus46,47. Recent studies indicate that the most important concern of the omicron variant is the ability to escape the immune response, which can cause re-infection and reduce vaccine effectiveness46,48,49.

Cats are considered to be an appropriate animal model with high susceptibility to SARS-CoV-224,25,50,51,52. In the present study, cats were infected with low, middle, and high doses of SARS-CoV-2, including wild type and some variants, to mimic natural infection. The pathogenicity of the SARS-CoV-2 strains was compared. Moreover, to confirm the immune evasiveness, the neutralizing activity was analyzed using convalescent sera from cats inoculated with wild type and variants.

Results

Experimental infection to cats with wild-type of SARS-CoV-2

Cats (n = 6) were experimentally inoculated with SARS-CoV-2 at several virus doses by intranasal (i.n.) and intraconjunctival (i.c.) routes and then monitored for 28 days (Table 1). Swab specimens from the eye, nose, oropharynx and rectum were subjected to virus isolation using Vero/TMPRSS2 cells53. SARS-CoV-2 was isolated from nasal and oropharyngeal swabs of all cats at 1 or 4 dpi to 7 dpi (Fig. 1). The highest titer was obtained from nasal swabs at 4 dpi or 7 dpi (Fig. 1A). At 11 dpi, SARS-CoV-2 was not isolated from nasal swabs of any cats. The titer from oropharyngeal swabs was highest at 1 dpi or 4 dpi, and then decreased (Fig. 1B). Similar to nasal swabs, SARS-CoV-2 was not isolated from 11 dpi. The virus titers of two cats of the low dose (102 TCID50) group and one cat of the middle dose (104 TCID50) group increased from 1 dpi, reaching a peak at 4 dpi, and then decreased. The titers at 1 dpi were high and showed a dose-dependent relationship with the amount of inoculated virus. In addition, the titers at 4 dpi of nasal and oropharyngeal swabs were higher in the low or middle dose groups than in the high dose group. Interestingly, SARS-CoV-2 was isolated from rectal swabs of high dose (106 TCID50)-inoculated cat No.1 at 7, 14, 18, 25, and 28 dpi, indicating persistent infection (Fig. 1C). This cat was depressed during the experiment and showed mild weight loss (Supplementary Figs. S1, S2). Furthermore, it was demonstrated that the low dose of SARS-CoV-2, just 100 TCID50, could infect cats and that the virus titers from this low dose group were similar to those of the middle or high dose groups.

Table 1 Information on viruses.
Full size table
Fig. 1
figure 1

Experimental infection of cats with SARS-CoV-2 WK-521, wild type. Virus titers (A–C) and virus-neutralizing (VN) antibody titers against SARS-CoV-2 (D) were measured from swab specimens and sera, respectively. (A) Infectious virus titers were determined using nasal swabs. (B) Infectious virus titers were determined using oropharyngeal swabs. (C) Infectious virus titers were determined using rectal swabs. (D) VN titers against wild type virus.

Full size image

Virus-neutralizing (VN) antibodies were detected using serum samples collected from cats. VN antibodies were detected from 11 to 28 dpi in all cats (Fig. 1D). Several cats showed a rapid increase in the VN antibody titer at 11 dpi, but then experienced a decrease from 14 dpi. At 11 dpi, the virus could not be isolated from nasal and oropharyngeal swabs, indicating that the VN antibody might neutralize the virus and inhibit the isolation of the virus. All cats maintained similar VN titers, 1:32–1:80 at 28 dpi.

Quantification of viral genome by quantitative RT-PCR (qRT-PCR)

The copy numbers of SARS-CoV-2 RNA from oropharyngeal, nasal and rectal swabs were measured using qRT-PCR. The copy numbers of viral RNA were the highest in the nasal swab, followed by oropharyngeal and rectal swabs (Supplementary Fig. S3). Viral RNA reached its peak at 4 dpi and then gradually declined. However, viral RNA was detected in some cats throughout the experimental period. The viral RNA in the low dose group increased from 1 to 4 dpi. Although this result was in line with the virus isolation result, viral RNA was still detected after 11 dpi.

Experimental infection to cats with SARS-CoV-2 variants

Cats (n = 12) were experimentally inoculated with the gamma (TY7–501, P.1 lineage), delta (TY11–927, AY.122 lineage), and omicron (TY38–873, BA.1.18 lineage) variants of SARS-CoV-2 and monitored for 28 days to compare the pathogenicity of VOCs in cats (Table 2). As a mimic of natural SARS-CoV-2 infection, cats were inoculated with a low dose of VOC, 102 TCID50/mL (TY7–501, TY11–927, and TY38–873), and a middle dose of TY38–873, 104 TCID50/mL. One of three cats inoculated with TY7–501 showed slight body weight loss at 1 and 4 dpi, and then recovered (Supplementary Fig. S1). Three cats inoculated with TY11–927 showed no clinical signs (Supplementary Fig. S2). Three cats inoculated with a middle dose of TY38–873 showed a loss of body weight from 1 or 4 dpi, while two showed body weight loss throughout the period of the experiment (Supplementary Fig. S1).

Table 2 Information on cats.
Full size table

Swab specimens were subjected to virus isolation. The virus was isolated from nasal and oropharyngeal swabs in two cats inoculated with TY7–501 (Fig. 2A,D) and 3 cats inoculated with TY11–927 (Fig. 2B,E). The highest titer was observed from nasal swab specimens. The virus titer was the highest at 4 dpi (TY11–927, Fig. 2B) and 7 dpi (TY7–501, Fig. 2A), and it was not isolated after 10 dpi. These results were similar to those of the WK-521 challenge experiment. Viral RNA was detected from 1 dpi to 14 or 21 dpi in nasal and oropharyngeal swabs (Supplementary Fig. S3). No virus was isolated or detected from the rectal swab in any of the cats, which is different from the result in WK-521-infected cats. The VN antibodies rapidly elevated from 7 dpi and remained elevated until 28 dpi in two cats inoculated with TY7–501 (Fig. 2G) and 3 cats inoculated with TY11–927 (Fig. 2H). These cats showed similar VN titers, 1:640–1:2560 at 28 dpi, which were higher than those of WK-521-inoculated cats.

Fig. 2
figure 2

Experimental infection of cats with SARS-CoV-2 variants. Three cats in each group were inoculated with 102 TCID50 of TY7–501, gamma variant (A,D,G), and TY11–927, delta variant (B,E,H), and 102 and 104 TCID50 of TY38–873, omicron variant (C,F,I). Virus titers from nasal swabs (A–C) and oropharyngeal swabs (D–F) and VN antibody titers against homologous virus (G–I)) were measured from swab specimens and sera, respectively.

Full size image

In contrast, the infectious virus was only isolated from a nasal swab in one of the three cats inoculated with the middle dose of TY38–873 (Fig. 2C,F). No virus was isolated from any specimens of the three cats that were inoculated with a low dose of TY38–873 (Fig. 2C,F, Supplementary Fig. S4A,B). Indeed, a significant difference was observed in the oropharyngeal swab of 4 dpi between TY11–927 and TY38–873 (Supplementary Fig. S4B). These results were quite different from those of WK-521 and other variant challenge experiments (Supplementary Fig. S4A,B). The VN antibody titer of cats inoculated with a low dose of TY38–873 was below the limit of detection, showing a significant difference between TY11–927 and TY38–873 (Fig. 2I, Supplementary Fig. S4C). The VN antibody titers of three cats inoculated with a middle dose increased from 10 or 14 dpi and were maintained at 1:10 to 1:320 at 28 dpi (Fig. 2I). These titers were lower than the titers of cats inoculated with TY7–501 and TY11–927.

Comparison of the antibody response among the wild-type strain and variants

The cross-reactivities of VN antibodies against several SARS-CoV-2 strains (Table 2) were compared among convalescent sera of cats inoculated with WK-521 and VOCs (Fig. 3). The VN titers of cats inoculated with WK-521 seemed to be higher against the alpha variant (QK002) and lower against the omicron variant (TY38–873), although the titers against other variants were not so different (Fig. 3A). The VN titers of 2 cats inoculated with TY7–501 were high against alpha (QK002), beta (TY8–612) and gamma variants (TY7–501) (Fig. 3B) but were low against the wild-type (WK-521), delta (TY11–912 and TY11–927) and omicron variants (TY38–873). The VN titers of cats inoculated with TY11–927 were the highest against delta variants (TY11–912 and TY11–927), followed by alpha (QK002), gamma (TY7–501), and beta variants (TY8–612) (Fig. 3C), showing a significant difference (Supplementary Fig. S5A). The VN titers of delta variant-inoculated cats seemed to be low against the wild type (WK-521) and omicron variant (TY38–873) (Fig. 3C), showing a significant difference (Supplementary Fig. S5A). The VN titers of the cats inoculated with the omicron variant (TY38–873) was significantly low against the other variants used in this experiment (Fig. 3D, Supplementary Fig. S5B), but the VN titers against omicron (TY38–873) and delta (TY11–927) variants were higher than those against the other viruses. Interestingly, the VN titer against WK-521 was below the limit of detection, thus indicating the absence of VN activity against early-phase SARS-CoV-2 strains.

Fig. 3
figure 3

VN titers of convalescent sera from cats inoculated with wild type (A), gamma (B), delta (C), and omicron variants (D). Wild type WK-521 and several variants including homologous viruses were used for the VN test.

Full size image

Discussion

In the present study, 9-month-old to 11-year-old cats (n = 18) were experimentally infected with SARS-CoV-2 at different doses (102, 104, and 106 TCID50) of wild type strain and VOCs isolated from COVID patients53. Two WK-521-inoculated cats, one gamma variant (TY7–501)-inoculated cat and two omicron variant (TY38–873)-inoculated cats showed mild to moderate depression, and appetite loss, which resulted in weight loss. Two cats inoculated with the omicron variant (TY38–873) showed transparent nasal discharge during the collection of nasal swab specimens, while the other cats showed no respiratory symptoms. Cats have been reported to demonstrate high susceptibility to SARS-CoV-2 infection, being asymptomatic with no clinical signs. However, some naturally infected cats exhibited respiratory or digestive symptoms, such as cough, nasal discharge, breathing difficulty, vomiting, and diarrhea35,54,55. In the previous study, cats inoculated with D614G and delta variants became lethargic, while the symptoms of cats inoculated with the omicron variant remained subclinical52. A number of studies52,56,57,58 reported that the omicron variant exhibits reduced pathogenicity in COVID-19 patients, hamsters, and domestic cats. In these animal experiments, cats were inoculated with a high dose of SARS-CoV-2. In our study, cats inoculated with a low dose (102 TCID50) of the gamma variant and a middle dose (104 TCID50) of the omicron variant showed body weight loss. Although the low dose of delta variant-inoculated cats did not show any symptoms, the viruses had extensively proliferated in the nasal and oropharyngeal cavity. This discrepancy might have been caused by different inoculation doses or the ages of the inoculated cats. We successfully established a challenge model of cats with a low dose of SARS-CoV-2 in this study.

The infectious virus titer was the highest in nasal swabs, followed by oropharyngeal and rectal swabs of cats inoculated with WK-521, delta (TY11–927) and gamma (TY7–501). Notably, one cat inoculated with WK-521 had persistently shed the virus via the rectal route. In human patients, the viral RNA positive rate in stool specimens was 29%, and in some cases viral RNA was persistently present in stool59,60. SARS-CoV-2 was also isolated in stool specimens from COVID-19 patients who did not have gastrointestinal symptoms59. In addition, SARS-CoV-1 survived for 4 days in stool samples61,62. These results support the potential fecal transmission of this virus as a source of continued spillover infection. Until now, there have been no reports on animal models of persistent infection. These findings indicate that some cats could persistently secrete infectious SARS-CoV-2.

SARS-CoV-2 in WK-521-inoculated cats was isolated from nasal and oropharyngeal swabs from 1 to 7 dpi. Since the titers of these specimens among different dose-inoculated cats were similar, the isolated virus from specimens at 1 dpi might not be the remnants of the inoculated virus. Since SARS-CoV-2 seems to replicate more rapidly in the upper respiratory tract63, this phenomenon is plausible. Due to this rapid replication, the virus could achieve rapid transmission from asymptomatic cats. These results indicate that SARS-CoV-2-infected cats could secrete high titers of viruses from these routes and could transmit infectious viruses to other animals, including humans. Interestingly, the viral secretion among cats inoculated with different doses of WK-521 was not dose-dependent. Cats inoculated with 104 TCID50 secreted more SARS-CoV-2 than cats inoculated with 106 TCID50. Blood cell counts showed that neutrophils increased and lymphocytes decreased faster from 1 dpi in cats inoculated with 106 TCID50 than in cats inoculated with 104 and 102 TCID50 (data not shown). The measurement of cytokine levels in sera revealed that IL-12p40 increased faster from 1 dpi in 106 TCID50-inoculated cats than in 104 and 102 TCID50-inoculated cats (data not shown). The high dose of virus inoculation could be thought to induce early innate and adaptive immune responses in comparison to the middle and low dose groups. In six cats, VN antibodies against SARS-CoV-2 increased from 11 dpi and reached a peak at 14 dpi, and thereafter persisted during the experimental period. This period of VN antibody elevation was consistent with the period of a reduced viral load. The cats inoculated with low doses of gamma variant, TY7–501 and delta variant, TY11–927 secreted a high titer of infectious virus through nasal and oropharyngeal routes. The fluctuation of VN titers of the cats inoculated with these variants was similar to that of the cats inoculated with a low dose of WK-521. As a result, it was confirmed that a low dose of SARS-CoV-2 (102 TCID50), especially WK-521, TY7–501, and TY11–927, could infect cats and proliferate well in cat tissues, such as the nose and oropharynx, in the acute phase, although there were no severe symptoms. In contrast, at 7 dpi, the infectious virus was only isolated in the nasal swab of one cat that was inoculated with a middle dose (104 TCID50) of the omicron variant TY38–873. The virus titer was very low in comparison to the other strains. The VN antibodies against TY38–873 elevated from 14 dpi and persisted until 28 dpi. This response of VN antibodies seemed to be slightly low and slow in comparison to other cats. Studies have shown that the omicron virus has attenuated pathogenicity in mice, hamsters, and cats52,56. Our results were consistent with those of previous studies52,56. Three distinct mutations in the furin cleave site region of S protein (P681H, H655Y and N679K) have been shown to decrease S1/S2 cleavage, fusogenicity, and syncytia formation, which are associated with attenuated pathogenicity57. Our findings demonstrated that the omicron variant could not replicate effectively in cat tissues, thus leading to the attenuated pathogenicity in cats.

Antigenicity differed between the wild type and VOCs. Sera from convalescent cats inoculated with WK-521, TY7–501 and TY11–927 showed similar levels of VN activity and the VN titers were the highest against the homologous strains. In contrast, the VN titers of cats inoculated with the omicron variant TY38–873 were low against several strains, including the homologous virus. In particular, they were unable to neutralize the wild type, WK-521. Among the unique mutations of the omicron variant, Q493R, G496S and Q498R have been linked to an increase in binding affinity to ACE264. G446S, S477N, E484A and Q493R have been reported to reduce the neutralizing activity of monoclonal or polyclonal antibodies48,65. These mutations of the omicron variant might contribute to the ability to evade the immune response that was acquired by the wild type and other variants. Since the phase was replaced with omicron variant, it has been reported that the omicron variant has a highly distinct antigenicity. It has also been reported that sera from convalescent and vaccinated individuals exhibit significantly reduced neutralizing activity against the omicron variant49. This phenomenon is consistent with the results in cats inoculated with the omicron variant.

The present study compared the pathogenicity and antigenicity of wild type and several VOCs in cats. One cat that was inoculated with WK-521 shed a low titer of infectious virus during the late phase when the protective immune response was established. SARS-CoV-2 at low doses, except for TY38–873, induced asymptomatic or mild changes and extensive proliferation of infectious virus in the acute phase in cats. This suggests that cats naturally infected with low doses of SARS-CoV-2 could transmit the virus to the environment, animals, and humans. The TY38–873 omicron variant was not able to proliferate effectively in cat tissues, but it did cause a loss of body weight. The convalescent sera from cats inoculated with WK-521, TY7–501, and TY11–927 exhibited low neutralizing activity against TY38–873. In addition, sera from convalescent cats inoculated with TY38–873 were unable to neutralize WK-521. This characteristic escape from VN antibodies against the other variants could cause reinfection with the omicron variant66,67,68,69. In conclusion, in many aspects cats are considered to be one of the best models of human COVID-19 infection. Additionally, this challenge model that uses cats could play a crucial role, not only as a model for human pathogenesis, but also in the development of preventative and therapeutic methods to control the infection and disease in humans and cats.

Materials and methods

Viruses and animals

Several strains of SARS-CoV-2 were used in this study (Table 1). Viruses were propagated in Vero/TMPRSS2 cells (JCRB1819)53. SPF cats were obtained from NAS Laboratory Co., Ltd. (Narita, Japan; from 9-month-old to 1-year-7-month-old-cats), Oriental Yeast Co., Ltd. (Tokyo, Japan; 1-year-2-month-old to 1-year-3-month-old cats), and School of Veterinary Medicine (Kitasato University, Aomori, Japan; 8- to 11-year-old cats). All cats were healthy. Each animal was housed in separate cages and isolators at the Biosafety Level 3 laboratory of the National Institute of Infectious Diseases (NIID, Musashimurayama, Japan).

Ethical statement

The animal experiments were performed in strict accordance with the Animal Experiment Guidelines of the NIID, and in compliance with the ARRIVE guidelines (https://arriveguidelines.org/about). The protocol was approved under Permission Number 120012-III and 121074-IV by the Institutional Animal Care and Use Committee of the NIID.

Animal infection and sample collection

Four groups of cats (n = 18, Table 2) were used in this experiment. The first group (n = 6) was i.n. and i.c. inoculated with 102, 104, and 106 TCID50/mL (WK-521, wild type), respectively. The second and third groups (n = 6) were i.n. inoculated with 102 TCID50/mL of TY7–501 (gamma) and TY11–927 (delta). The fourth group (n = 6) was i.n. inoculated with 102 and 104 TCID50/mL of TY38–873 (omicron variant). After inoculation, the cats were monitored for 4 weeks. The humane endpoints were set as follows: weight loss of 20%, loss of appetite (more than 4 days), clinical symptoms such as diarrhea, and fluctuation of blood cells demonstrated by a whole blood cell analysis. When cats met more than two points, they were considered to have reached the humane endpoint. At 0, 1, 4, 7, 10, 11, 14, 21, and 28 dpi, body weight, and body temperature were measured under anesthesia. Blood was collected from the veins and centrifuged for serum collection. After centrifugation, the supernatant was collected and stored at − 80 °C. Conjunctival, oropharyngeal, nasal and rectal swabs were obtained using cotton swabs70. Swab samples were suspended in 1 mL of Dulbecco’s Modified Eagle Medium (DMEM) with 2% fetal bovine serum (FBS). The samples were then centrifuged at 12,000×g for 1 min, and the supernatants were stored at − 80 °C.

Virus isolation

Tissue specimens were subjected to preparation of 10% homogenates in DMEM with 2% FBS. Then, the supernatants were harvested by centrifugation at 12,000×g for 3 min and stored at − 80 °C. Sera, swabs and tissue specimens were subjected to virus isolation using Vero/TMPRSS2 cells and the titers were measured as TCID50/mL53. Briefly, cells were prepared in the 96-well microplates to form a subconfluent monolayer. Then, 100 μL serially diluted specimens (1:101–108) were applied in duplicate to 4 wells on the following day, followed by cell culture at 37 °C in 5% CO2 for 3 days. At 3 dpi, wells with cytopathic effect (CPE) were counted, and the TCID50/mL was calculated. At the same time, cells with CPE were harvested and centrifuged at 12,000×g for 2 min. The supernatants were stored at – 80 °C.

VN test

The VN test was performed according to our institute protocol for seroepidemiological study71. Briefly, VeroE6/TMPRSS2 cells were prepared in a T75 flask using DMEM with 10% FBS, Geneticin (G418, Roche, Mannheim, Germany), and penicillin and streptomycin (Gibco, NY, USA; 10F-DMEM-G418-Penstrep). On the next day, SARS-CoV-2 was diluted to 100 TCID50/50 μL using DMEM with 10% FBS, penicillin, and streptomycin (10F-DMEM-Penstrep). Then, the test serum was subjected to twofold serial dilution, from 5- to 160-fold, with 10F-DMEM-Penstrep. Equal volumes of prepared virus and diluted test serum were mixed and incubated at 37 °C for 1 h. Vero/TMPRSS2 cells were treated with trypsin–EDTA (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan) and suspended in 10F-DMEM-G418-Penstrep. After the reaction, 100 μL of mixture were inoculated into each well of 4 wells in the 96-well microplate. Then, 100 μL of suspended cells were added to each well and incubated at 37 °C for 5 days. At 5 days after incubation, cells were fixed using 10% neutral formalin at room temperature for 1 h and then stained with crystal violet solution. The neutralization titers were determined as the reciprocals of the antiserum dilutions corresponding to a 50% reduction in cytopathic effect and were calculated by the geometric mean. VN tests were performed in duplicate.

Detection of SARS-CoV-2 RNA by quantitative one-step reverse transcription polymerase chain reaction (qRT-PCR)

A total of 100 μL of sera and swabs were subjected to RNA extraction using the QIAamp Viral RNA Mini kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. RNA from tissue specimens was extracted using Trizol (Invitrogen, CA, USA), PureLink® RNA Mini Kit (Invitrogen), and PureLink® DNase Set (Invitrogen) according to the manufacturer’s instructions. A qRT-PCR assay was performed with the QuantiTect Probe RT-PCR kit (QIAGEN) according to the manufacturer’s instructions. In brief, 5 μL of RNA, NIID_2019-nCOV_N_F2 (forward primer; AAATTTTGGGGACCAGGAAC), NIID_2019-nCOV_N_R2 (reverse primer; TGGCAGCTGTGTAGGTCAAC) and NIID_2019-nCOV_N_P2 (probe; FAM-ATGTCGCGCATTGGCATGGA-TAMRA) were added to the master mix72. This primer and probe set targets the nucleocapsid (N) protein gene of SARS-CoV-273. The thermal cycling conditions were as follows: reverse transcription at 50 °C for 30 min, denaturation at 95 °C for 15 min, and 45 cycles of 95 °C for 15 s and 60 °C for 1 min. Amplification was performed on a LightCycler 480 (Roche) instrument, software version 1.5.

Statistical analysis

Data were analyzed using an unpaired t-test with Welch’s correction and a two-way ANOVA with Tukey’s multiple comparison tests or the two-stage procedure of Benjamini, Krieger, and Yekutieli. All statistical analyses were performed using Graph Pad Prism 9.2 for Windows (GraphPad Software, San Diego, California, USA).