key: cord-0291684-u999bx7a
authors: Hawks, Seth A.; Prussin, Aaron J.; Kuchinsky, Sarah C.; Pan, Jin; Marr, Linsey C.; Duggal, Nisha K.
title: Infectious SARS-CoV-2 is emitted in aerosols
date: 2021-08-10
journal: bioRxiv
DOI: 10.1101/2021.08.10.455702
sha: 13a0c2219ee32a5ab9bc2a5235b706895bf26db2
doc_id: 291684
cord_uid: u999bx7a

Respiratory viruses such as SARS-CoV-2 are transmitted in respiratory droplets and aerosols, which are released during talking, breathing, coughing, and sneezing. Non-contact transmission of SARS-CoV-2 has been demonstrated, suggesting transmission in aerosols. Here we demonstrate that golden Syrian hamsters emit infectious SARS-CoV-2 in aerosols, prior to and concurrent with the onset of mild clinical signs of disease. The average emission rate is 25 infectious virions/hour on days 1 and 2 post-inoculation, with average viral RNA levels 200-fold higher than infectious virus in aerosols. Female hamsters have delayed kinetics of viral shedding in aerosols compared to male hamsters, with peak viral emission for females on dpi 2 and for males on dpi 1. The majority of virus is contained within aerosols <8 µm in size. Thus, we provide direct evidence that, in hamsters, SARS-CoV-2 is an airborne virus.

SARS-CoV-2 is a respiratory virus that has caused more than 190 million cases and 4.1 million 26 deaths, as of July 2021 [1] . Respiratory droplets and aerosol particles (aerosols), which may 27 contain virus, are expelled during coughing, sneezing, talking, and breathing and can vary 28 widely in size from less than 1 µm to greater than 100 µm [2, 3] . Their size significantly impacts 29 transmission risk and mode due to differences in the way that droplets and aerosols travel 30 through the air. There is a continuum of maximum distances that particles can reach. Those 31 smaller than 10 µm remain suspended in air for many minutes to hours, during which they can 32 travel long distances; this does not rule out their potential to transmit at close range, too. 33

Infectious SARS-CoV-2 has been cultured from aerosols sampled near . SARS-CoV-2 has also been isolated from aerosols <1 µm within a car driven by a 35 COVID-19 patient with mild illness [8] . The collection of exhaled breath condensate (EBC) is a 36 non-invasive sampling method of respiratory droplets and aerosols that has been used to 37 assess the airborne transmission potential of respiratory viruses, including seasonal human 38 coronaviruses, influenza viruses, and rhinoviruses [9] [10] [11] [12] . For COVID-19 patients, the emission 39 rate of SARS-CoV-2 RNA in EBC was estimated to be >1 million viral RNA copies/hour by 40 breathing [13, 14] . In non-human primates, SARS-CoV-2 RNA has also been detected in EBC 41 collected from inoculated animals [15, 16] . 42

Hamsters are a naturally-susceptible animal model for SARS-CoV-2 transmission that 43 develop few clinical signs of disease [17] [18] [19] [20] [21] [22] . Unlike mice, hamsters model asymptomatic 44 infection, which has been suggested to be the most important component of community 45

transmission of 24] . Oral swabs from inoculated hamsters contain high levels 46 of infectious virus, with levels similar to saliva collected from COVID-19 patients [25, 26] . 47 Importantly, inoculated hamsters have been shown to transmit SARS-CoV-2 to naïve hamsters 48 via non-contact transmission [17, 18, 22] , suggesting SARS-CoV-2 may be transmitted between 49

Chamber and nosecone. Aerosols generated by hamsters were sampled using two different 75 approaches. In the first, infected hamsters were placed in a 2L sealed chamber, in which they 76 were allowed to move freely. Air was supplied through an inlet to the chamber, and aerosols 77 were sampled through an outlet. This approach captured total aerosols produced in exhaled 78 breath, released from the fur, and resuspended from the chamber floor by the hamster's activity. 79

The second approach captured aerosols produced in exhaled breath only. Infected hamsters 80 were anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg) administered 81 by intraperitoneal (i.p.) injection. After the hamster was fully immobile, a non-rebreather 82 nosecone (Kent Scientific, VetFlo-0802) was placed on the hamster. Air was provided through 83 the nosecone's inlet tube, and aerosols were sampled via the outlet port. 84 85 Infectious virus collection. A condensation sampler (Aerosol Devices Inc. Series 110A) was 86 connected to the outlet of the chamber or nosecone. The sampler was operated at 1.5 L/min for 87 a total of 1 hour, collecting aerosols directly into a vial containing 400μl BA-1 medium. To 88 separate aerosols by size, a cyclone (URG 2000-30E-5-2.5-S) that removes aerosols larger 89 than 8 µm at the sampling flow rate used here was installed upstream of the condensation 90 sampler. Samples were collected for 30 minutes with the cyclone and 30 minutes without the 91 cyclone. For 4 hamsters, the cyclone was used during the first 30-minute period, and for 4 92 hamsters, the cyclone was used during the second 30-minute period. 93 94 Aerosol size distribution. An aerodynamic particle sizer (TSI 3321) was connected to the outlet 95 of the chamber or nosecone to measure the aerosol size distribution for 15 minutes in 1 minute 96 intervals at a flow rate of 1 L/min, with makeup air provided at the same flow rate. This 97 instrument detects aerosols over the size range 0.5-20 µm. Background concentrations 98 measured in an empty chamber and nosecone were subtracted to estimate the contribution 99 from the hamsters alone. 100 101 Viral quantification. 102

Plaque assays. Infectious virus was quantified via Vero cell plaque assay. Briefly, samples were 103 serially diluted, plated onto confluent Vero cells in a 6-well plate, and incubated for 1 hour at 104 37°C. After a 1-hour adsorption, 2 mL of an 0.8% agarose overlay medium was added to the 105 wells. Plates were incubated for 1 day at 37°C, after which a second 2 mL overlay containing 106 3% neutral red was added to the wells. One day later, plaques were counted. The limit of 107 detection was 1.4 log10 PFU/swab (oral/fur/rectal) and BAL fluid, 0.3 log10 PFU/air sample, and 108 0.7 log10 PFU/nasal wash. Aerosol generation by hamsters. To test whether inoculated animals shed SARS-CoV-2 in 126 aerosols, we established two aerosol sampling methods for hamsters. In the first method, 127 animals were allowed to move freely within an empty 2L chamber ( Figure 1A ). In the second 128 method, animals were anesthetized, and a nosecone was placed over the nose and mouth 129 ( Figure 1B ). Aerosols were sampled via an outlet port from the chamber or nosecone. 130

We measured aerosols produced by uninfected hamsters and calculated the emission 131 rate. Using the chamber, we found an average of 700 aerosol particles emitted per minute per 132 hamster ( Figure 1C ), with 99.9% of them <10 µm in size (Supplementary Figure 1) . Using the 133 nosecone, we found an average of 1 aerosol particle emitted per minute per hamster, which 134 was significantly fewer than the chamber approach (p<0.05). The size distribution was similar in 135 both cases, with very small particles being the most abundant. 136 137 Infectious SARS-CoV-2 is emitted in aerosols. Hamsters were inoculated intranasally with 138 SARS-CoV-2 strain USA-WA1/2020. Mild weight loss occurred from days post-inoculation (dpi) 139 2 through 5 ( Figure 2A ). Oral swabs and nasal washes were collected daily. Virus peaked on 140 dpi 1 in the oral swabs at 3.5 log10 PFU/swab and on dpi 2 in the nasal washes at 3.9 log10 141 PFU/wash ( Figure 2B and C). Significantly higher viral titers were observed in nasal washes 142 from males compared to females on dpi 1, with a 5,000-fold difference (p<0.001). Viral titers for 143 females peaked one day later than males. Viral titers decreased over time, with infectious virus 144 below the limit of detection by dpi 5 for most animals. Samples were tested for RNA, which was 145 detectable through dpi 10 ( Figure 2E and F). Sex-specific differences were not observed for 146 viral RNA levels. 147

Air samples were collected daily for 1 hour using a condensation sampler, which 148 maintains viral infectivity [27] . Infectious virus was detected in aerosols collected on dpi 1 and 2 149 from males, with a mean emission rate of 1.5 log10 PFU/hour, and infectious virus was detected 150 on dpi 2 from females, with a mean emission rate of 1.0 log10 PFU/hour ( Figure 2D ). 151

Significantly greater infectious viral titers were detected in air samples from males compared to 152 females on dpi 1 (p<0.01). The overall mean emission rate across dpi 1 and 2 was 1.4 log10 153 PFU/hour. A majority (75%) of inoculated animals released detectable levels of virus in the air 154 on dpi 2, and emission rates ranged from 0.9 to 1.8 log10 PFU/hour. Infectious virus was below 155 the limit of detection in air samples collected after dpi 2. Air samples were also tested for viral 156 RNA; viral RNA was detected through dpi 5, with levels below the limit of detection by dpi 10 157 ( Figure 2G ). Sex-specific differences were not observed for viral RNA levels in the air. For 158 samples with detectable infectious virus, the RNA levels were approximately 200-fold higher 159 than PFU levels on dpi 1 and 2 for males and approximately 300-fold higher than PFU levels for 160 females on dpi 2. Together, these data show that infectious SARS-CoV-2 is emitted in aerosols 161 early in infection, prior to and concurrent with the onset of mild clinical signs of disease. 162 163

for airborne virus, we inoculated additional hamsters in order to test whether small aerosols 165 contain infectious SARS-CoV-2. Here, we shortened the air sampling time to 30 minutes. Male 166 hamsters were used, as virus was more readily detected in their air samples compared to 167 females' air samples. To test whether the virus detected in the air was residual inoculum, we 168 collected an air sample at 4 hours post-inoculation using the chamber; no virus was detected 169 ( Figure 3A ). Then, we collected air samples on dpi 1 and 2 in the chamber, with and without a 170 cyclone separator, which removed aerosols >8 µm, placed upstream of the condensation 171 sampler. We detected similar titers in samples collected with and without the cyclone separator, 172 with no statistical significance in the difference in mean titers on dpi 1 (p=0.34) or dpi 2 (p=0.37), 173

indicating that size restriction did not alter the amount of virus detected ( Figure 3A ). To test 174 whether airborne SARS-CoV-2 was detectable in the breath, we anesthetized the hamsters and 175 collected their breath from the nosecone for one hour. Infectious virus was not detectable. 176 However, the breath rate was very low during anesthesia (13 ± 2 breaths/minute), and very few 177 total respiratory aerosols were collected with this method compared to the chamber method 178 ( Figure 1C) . 179

High levels of virus were detectable in oral swabs, nasal washes, and BAL fluid collected 180 from the animals ( Figure 3B-D) . A low level of virus was detected in fur and rectal swabs taken 181 from the animals, which indicates that some of the airborne virus detected in the chamber could 182 be resuspended from the body ( Figure 3E and F) . Together, these results indicate that SARS-183

CoV-2 is emitted primarily in small aerosols <8 µm. 184 185

In this study, we found that hamsters emitted infectious SARS-CoV-2 in aerosols primarily <8 187 µm in size in the absence of severe disease, with an emission rate of 1.4 log10 PFU/hour 188 (Figures 2 and 3) . Peak emission of infectious virus was delayed by 1 day for females compared 189 to males. SARS-CoV-2 viral RNA was detected in aerosols and the upper respiratory tract for a 190 longer duration than infectious virus. The upper and lower respiratory tract contained high titers 191 of virus, suggesting that the virus isolated from the air was primarily derived from the breath. 192

However, we found that the fur was contaminated with low levels of infectious virus, indicating 193 that virus resuspended from the fur may also be a viable mechanism of SARS-CoV-2 emission 194 into the air. 195 Aerosols <8 µm in size can remain airborne and be inhaled. Thus, our results suggest 196 that airborne transmission is likely a major driver of SARS-CoV-2 transmission. Our studies 197 support reports of infectious SARS-CoV-2 collected from aerosols near , as well as studies showing non-contact transmission of SARS-CoV-2 between ferrets and 199 hamsters [17, 18, 22] , including one study that demonstrated transmission over a 1-meter 200 distance [28] . Some previous studies have been unable to culture virus from air samples from 201 COVID-19 patients or inoculated non-human primates due to unknown collection times post-202 onset of disease or the use of air sampling equipment or buffers that do not maintain viral 203 infectivity [16, 29] . Our results support the idea that transmission is likely to occur prior to or 204 concurrent with symptom onset or in the absence of clinical disease, supporting studies that 205 have found that asymptomatic infection is a major driver of community transmission [23, 24] . 206

Identifying the modes of SARS-CoV-2 transmission is critical to designing interventions to 207 effectively prevent transmission, and our results support the use of masks and ventilation to 208 reduce SARS-CoV-2 transmission. 209

This study was limited by our inability to collect EBC from the hamsters. We were unable 210 to detect infectious virus directly from the breath when anesthetized; however, anesthesia 211 decreased the breath rate of hamsters, and we detected 700-fold fewer aerosol particles with 212 this method ( Figure 1 ). Thus, given the detection of infectious virus on the fur, we cannot 213 exclude that virus resuspended from the fur during movement may have contributed to the 214 particles that we detected using the chamber method. Movement in guinea pigs has been 215 shown to increase influenza particles in the air by increasing the resuspension of dust particles 216 from the body [30] . The contribution of resuspended aerosols to SARS-CoV-2 airborne 217 transmission has not been studied. 218

Sex-specific differences in COVID-19 disease severity are widely reported, with more 219 severe disease in men [31, 32] . Here, we detected a delay in transmission potential in female 220 hamsters compared to male hamsters. It is unclear whether these sex-specific differences in 221 infectious viral shedding detected are relevant to human transmission. Interestingly, viral RNA 222 levels in air samples were not significantly different between male and female hamsters, 223

suggesting that a female-specific factor may enhance the release of defective virions. Most 224 studies detect viral RNA, which may lead to the underestimation of differences in viral 225 transmission potential between sexes. Sex-specific differences in infectious viral titers have not 226 been observed by other groups, but differences in timing of samples collected may be relevant, 227 as the largest difference was observed on dpi 1 in this study, which was not tested in other 228 studies [33, 34] . 229

Public health agencies have recently begun describing SARS-CoV-2 as an airborne 230

virus. Here, we show that infectious virus is indeed culturable from the air early after infection, 231

with the majority of aerosols containing infectious virus <8 µm in size. This suggests that SARS-232

CoV-2 may be maintained in the air for hours and over larger distances that previously 233 recognized, with ventilation being an important tool for preventing transmission. Future studies 234 will be critical for establishing the transmission potential of small aerosols containing SARS-235

CoV-2. 

COVID-19) Dashboard

Mechanistic insights into the effect 252 of humidity on airborne influenza virus survival, transmission and incidence

Size distribution and sites of origin of droplets expelled from the 257 human respiratory tract during expiratory activities

Viable 260 SARS-CoV-2 in the air of a hospital room with COVID-19 patients

Aerosol and surface contamination of SARS-CoV-2 observed in quarantine and isolation care

Aerosol and Surface Distribution of 268

Severe Acute Respiratory Syndrome Coronavirus 2 in Hospital Wards

Aerodynamic analysis of SARS-272

CoV-2 in two Wuhan hospitals

275 Isolation of SARS-CoV-2 from the air in a car driven by a COVID patient with mild illness

Epub 2021/04/27

Respiratory 279 virus shedding in exhaled breath and efficacy of face masks

Infectious 282 virus in exhaled breath of symptomatic seasonal influenza cases from a college community

Influenza virus aerosols 287 in human exhaled breath: particle size, culturability, and effect of surgical masks

Rhinoviruses significantly affect day-to-day respiratory symptoms of children with asthma

Detection of SARS-CoV-2 in 295 Exhaled Breath from COVID-19 Patients Ready for Hospital Discharge

Earlier Stages Exhaled Millions of Severe Acute Respiratory Syndrome Coronavirus 2 Per Hour. 299 Clinical Infectious Diseases

Exhaled 301 aerosol increases with COVID-19 infection, age, and obesity

SARS-CoV-2 Aerosol Exhaled by 305 Experimentally Infected Cynomolgus Monkeys

Simulation of the clinical 312 and pathological manifestations of Coronavirus Disease 2019 (COVID-19) in golden Syrian 313 hamster model: implications for disease pathogenesis and transmissibility

Syrian 317 hamsters as a small animal model for SARS-CoV-2 infection and countermeasure development

Oral SARS-CoV-2 322 inoculation establishes subclinical respiratory infection with virus shedding in golden Syrian 323 hamsters

Defining the 326

Syrian hamster as a highly susceptible preclinical model for SARS-CoV-2 infection. bioRxiv

Surgical mask partition 330 reduces the risk of non-contact transmission in a golden Syrian hamster model for Coronavirus 331 Disease 2019 (COVID-19)

CoV-2 Transmission From People Without COVID-19 Symptoms

Quantifying asymptomatic infection and transmission 338 of COVID-19 in New York City using observed cases, serology, and testing capacity

PubMed 340 PMID: 33571106; PubMed Central PMCID: PMCPMC7936345

Comparison of SARS-CoV-2 detection in nasopharyngeal swab and saliva

Viral dynamics of SARS-CoV-2 in saliva from infected 346 patients

CoV-2 in the air of a hospital room with COVID-19 patients

-2 are transmitted through the air between ferrets over more than 354 one meter distance

Lack of viable 357 severe acute respiratory coronavirus virus 2 (SARS-CoV-2) among PCR-positive air samples 358 from hospital rooms and community isolation facilities

Influenza A virus is transmissible via aerosolized fomites

Male sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU 367 admission

32. GlobalHealth5050. The Sex, Gender and COVID-19 Project

Sex 372 Differences in Lung Imaging and SARS-CoV-2 Antibody Responses in a

2021:e0097421

Defining the 376

Syrian hamster as a highly susceptible preclinical model for SARS-CoV-2 infection