Coronavirus infections are common in humans and other mammals. Coronaviruses belong to the family Coronaviridae and possess a single strand, positive-sense RNA genome ranging from 26 to 32 kilobases in length. In December, 2019, a cluster of cases of pneumonia were reported in people associated with the Huanan Seafood Wholesale Market in Wuhan, Hubei Province. The subsequently research from the bronchoalveolar lavage fluid (BALF) of a hospitalized patient allowed the isolation of a novel coronavirus, which was named severe acute respiratory syndrome coronavirus 2 (SARS-CoV–2) and considered as the pathogen of the current outbreak of coronavirus disease (COVID–19)1. The common clinical features for COVID–19 include fever, cough, myalgia, fatigue, dyspnea, lymphopenia, and pneumonia, less common symptoms including sputum production, headache and haemoptysis and diarrhea2. As of February 19, 2020, a total of 75,204 cases had been reported in at least 25 countries with more than 10,000 in severe condition and 2,009 deaths3, whereas a sizable portion of infected but non-symptomatic people with potential of transmissibility was also reported4,5. Due to the urgency of the situation, World Health Organization declared that the outbreak of SARS-CoV–2 in China constitutes a Public Health Emergency of International Concern (PHEIC).
Animal models are essential for the study of pathogenesis of the viral infection, the evaluation of potential antiviral treatments or vaccine development. Here we report a nonhuman primate disease model for SARS-CoV–2. We experimentally infected Rhesus macaques (RM) using SARS-CoV–2 isolation from clinical bronchoalveolar lavage fluid6 and we evaluated the dynamics of SARS-CoV–2 in the blood, swabs and respiratory tract tissues of Rhesus macaques.
Six RM (1–6) between the ages of 6 and 12 years, three males and 3 females, were inoculated with 7×106 50% tissue-culture infectious doses (TCID50) of SARS-CoV–2 （IVCAS 6.7512）through intratracheal route. The blood, oropharyngeal swab, nasal swab and anal swab were evaluated post infection (Fig. 1a). No obvious clinical signs were observed during the study course except one animal showed reduced appetite. The body weight (Fig. 1b) did not show any change from 1 to 6 day post infection (d.p.i.) for all animals investigated while 7% and 8% weight loss was noticed on 14 d.p.i. in RM1 and RM4, respectively. The body temperature (Fig. 1c) was monitored from day 1 to day 14 and no obvious changes were found.
To examine the viral replication dynamics in the RM, the blood, oropharyngeal swab, nasal swab and anal swab were collected on 1–6, 9, 11 and 14 d.p.i. from the RMs available for study. No viral RNA could be detected from the blood through out day 1 to day 14 by quantitative reverse transcriptase-polymerase chain reaction (RT-qPCR) (Fig. 1d). Two peaks were observed on 1 and 5 d.p.i. in most animals from throat samples (Fig 1e). The first peak viral load on day 1 was examined in all six animals (RM1-RM6) from oropharyngeal swab and ranged 2.08×104–2.85×107 copies/ml. Due to the euthanization of two animals on day 3 and another two on day 6 for pathological examination, the animal sample size for the following investigation on day 4 were reduced to 4 whereas after day 6 were only two (see below). Notably, a second peak of viral load in oropharyngeal swab appeared on day 5 after infection with viral load of 1.03×107–3.60×107 copies/ml in three out of the four animals investigated (RM1, RM3 and RM4), whereas one monkey (RM5) that exhibited a very low viral RNA load (<10 copies/ml) on day 5 showed an increased viral load on day 6 after infection (around 200 copies/ml). The viral RNA levels in throat fell rapidly after day 6 in the two remaining RMs (RM1 and RM4) and reached undetectable level in these two animals by day 9. The nasal swabs also started to show positive viral RNA in 3 (RM1, RM2 and RM4) out of 6 RMs (Fig. 1f) on 1, 2 and 3 d.p.i. and maintained positive for 2–3 days. Furthermore, we evaluated the anal viral levels from 1 to 14 d.p.i. (Fig. 1g). Three out of six RMs exhibited detectable viral RNA on day 2 with RM1 showing the longest viral shedding till day 11. Overall, the viral RNA was detected from upper respiratory tract after infection. The two peaks in throat could represent the original input and viral replication in the throat, respectively. The fact that viral RNA was positive from anal swab suggested a possible occurrence of the infection in digestive tract.
The chest X-ray examination was conducted in animals on 0, 1, 3 and 6 d.p.i.. As shown in Fig. 2a (left panel), normal lungs were presented before infection, while patchy glass-ground opacity was observed in the lower parts of both left and right lungs on 1 d.p.i. (Fig. 2a middle left panel). On 3 d.p.i., multiple glass-ground opacity in left and right lungs was found (Fig. 2a, middle right panel) and the density decreased slightly on day 6 (Fig. 2a, right panel) while the scope of lung lesions extended from day 3 to day 6. Two animals (RM2 and RM6) were euthanized on day 3 while another two (RM3 and RM5) were euthanized on 6 d.p.i.. Necropsy was then performed in all the four animals. Postmortem examinations showed a variable degree of consolidation, edema, hemorrhage and congestion in bright red lesions throughout the lower respiratory tract which indicated the diffuse interstitial pneumonia (Fig.2b). After necropsy, the tissues and organs were harvested for quantification of viral RNA. RT-qPCR analysis was performed in respiratory tract tissues and other organs. The data revealed the widespread presence of SARS-CoV–2 in the respiratory tract (Fig. 2c-d). The amount of the viral RNA ranged from 3.0×104 to1.5×107 copies/g in trachea and bronchus tract on both 3 and 6 d.p.i.. In the lungs, viral RNA was detected with titre up to 2.0×107 copies/g and mainly localized in the lower lobes of lungs. Upon necropsy, the area of lung lobes affected by lesions was estimated by pathologists. Histopathological analysis showed damaged areas surrounding small bronchus, with the most serious injury in the inferior lobe of the left and right lungs. Thickened alveolar walls with fibroblast proliferation were observed in the majority of alveoli. Pulmonary hyaline-membrane formation, hemorrhage and edema could be seen in the alveoli (Fig. 2 e-g). Confocal microscopy analysis of lesion sites using specific antibodies (anti-RP3-CoV N protein) indicated the positive viral N protein (Fig. 2h). These results suggested that SARS-CoV–2 could infect the respiratory tract resulting in diffuse interstitial pneumonia in RMs.
To confirm the disease was caused by SARS-CoV–2, the virus was also re-isolated from bronchus, lung tissues and oropharyngeal swab. The whole genome sequence alignment analysis showed isolated viruses were >99.99% identical to the original input viral samples (Table 1). To examine whether neutralizing antibody was generated after infection, the sera from RM1 and RM4 were used for plaque reduction neutralization test (PRNT). As shown in table 2, the titre for RM1 reached to 1:1,350 on 14 d.p.i. and increased 1:4050 on 21 d.p.i., while RM4 maintained the same titre as 1:12,150 on both 14 and 21 d.p.i..
Collectively, SARS-CoV–2 caused acute localized-to-widespread pneumonia as proved by pathological studies in all animals studied, although without obvious clinical symptoms of respiratory disease. This animal model has confirmed the causal relationship between SARS-CoV–2 and respiratory disease in RM reminiscent of the mild respiratory disease or non-symptomatic cases in COVID–19 already reported in humans4,5, thus fulfilling Koch’s postulates. The model enables detailed studies of the pathogenesis of this illness and may play a critical role in the evaluation of therapeutic drugs and vaccines.