Responding to the COVID-19 Pandemic: The Mouth Is a Site of SARS-CoV-2 Infection

The mouth is the largest gateway to inside the body and expectedly, is a vulnerable site for viral infections. A team of investigators present the first direct evidence that SARS-CoV-2 can infect and replicate in the cells of the mouth in humans, and that saliva is potentially infectious.

Like Comment
Read the paper

The world continues to pay a pernicious toll due to the SARS-CoV-2 pandemic, including more than 2.6 million deaths. One silver lining is the speed with which the scientific community has come together to elucidate the biology and pathogenesis of the virus. As a testament to the power of this ‘team science’ approach, we engaged an international, multidisciplinary team to understand the role of the oral cavity in SARS-CoV-2 infection. Our study presents the first direct evidence that SARS-CoV-2 can infect, and replicate in, cells of the mouth in humans. Though it might seem obvious that SARS-CoV-2 can be found in the mouth because saliva tests are now routinely used for diagnostics, the actual sources of the virus in saliva have been unclear. Our study identifies the oral cavity, and in particular the salivary glands, as one likely source of infectious SARS-CoV-2 in saliva (Figure 1).

Illustration representing the diversity of tissue sites represented in the mouth including both hard and soft tissues, and fluids (i.e., saliva). It is expected that the mouth presents an array of vulnerable sites for viral infection. (Illustration Credit: Alan Hoofring, MA (National Institutes of Health). 

The roots of this project can be traced to our respective research programs, who as early-career dentist-scientists and friends, were separately using a relatively new tool called single cell RNA sequencing (scRNAseq) to understand diseases that affect the mouth (i.e., Sjögren’s syndrome [Warner/NIH] and periodontal disease [Byrd/UNC-Chapel Hill]). As part of this work, we each had used scRNAseq to generate unpublished catalogs of the cell types making up the majority of tissues of the oral cavity in healthy humans.

        As much of the US was shutting down due to the pandemic in March 2020, we saw an opportunity to combine our efforts and contribute to the global response to understand the virus. With the help of US and international partners—as well as multiple institutes and centers across the National Institutes of Health (NIH)— we integrated our scRNAseq datasets and used them to map RNA the ‘entry factors’ used by SARS-CoV-2 to infect host cells. Among the common targets were angiotensin converting enzyme 2 (ACE2) and transmembrane protease, serine 2 (TMPRSS2), and we observed that these entry factors were expressed in the epithelia of the oral cavity, indicating likely vulnerability to SARS-CoV-2 infection at these barrier tissues.

        Armed with this data, we set out to determine whether oral tissues from individuals with COVID-19 were infected with SARS-CoV-2. Through a collaboration with the NIH COVID-19 Autopsy Consortium, and also an acutely infected NIH COVID-19 participant, we obtained oral tissues samples from people with active COVID-19. Using in situ hybridization, we found evidence of SARS-CoV-2 infection and replication in the ducts and acini of the salivary glands (Figure 2)

SARS-CoV-2 infects ACE2-expressing ductal and acinar cells of the human salivary glands. Credit: Paola Perez, PhD (Salivary Disorders Unit, National Institute of Dental and Craniofacial Research)

, and to a lesser degree, the oral mucosa. This information generated two new hypotheses: first, we hypothesized that de novo virus made by the salivary glands is potentially infectious; and second, that the lining epithelia of the oral could shed infected SARS-CoV-2 cells into the saliva.     

To test these hypotheses, we collected high viral load saliva samples from asymptomatic and mildly symptomatic individuals who had tested positive for COVID-19. We incubated cellular and acellular saliva fractions with healthy cultured cells using a standard infectivity assay and discovered that some high viral load acellular saliva, and to a lesser degree the cellular fraction, elicited viral cytopathic effects characteristic of SARS-CoV-2 infection ex vivo, which was later confirmed by sequencing (Figure 3).

Acellular and cellular saliva from asymptomatic high viral load subjects can infect and replicate in Vero cells, ex vivo. Samples demonstrating cytopathic effect characteristic of coronavirus infection were sequenced and confirmed to be SARS-CoV-2. (Credit: Bernard Lafont, PhD, NIAID)

This saliva was from some asymptomatic subjects and our results indicated that saliva could facilitate the spread of infectious SARS-CoV-2 to other organs, such as the gut or lungs, or be transmitted to other individuals. In analyses of saliva samples from individuals with COVID-19 from two different sites (UNC-Chapel Hill and NIH), we used in situ hybridization to show that oral epithelial cells shed into saliva not only expressed viral entry factors but also harbored SARS-CoV-2 RNA. In total, these results indicate that saliva is a medium not only for de novo virus but also a source of potentially infectious shed oral cells.

We next explored the concordance between nasopharyngeal and saliva swabs for virus detection, the relationship between reported COVID-19 symptoms and presence of the virus in saliva, and the presence of antibodies in saliva to SARS-CoV-2 antigens. Warner, with help from NIH leadership, designed a clinical study aimed at collecting paired saliva and nasopharyngeal swabs in mildly symptomatic and asymptomatic subjects with COVID-19. Immediately we recognized that, although there was general concordance between NP and salivary swabs for viral detection, this was not the case for every individual. Nasal-only and oral-only viral load was detected in some participants and could not be explained by sampling error. This observation may help explain reports of false-negative tests in the community. We then examined the relationship between self-reported COVID-19 symptoms and the presence of the virus in saliva. What we found was that the report of loss of taste and/or smell directly correlated with the presence of SARS-CoV-2 in saliva; however, the opposite was true for gastrointestinal symptoms. We also found that the presence of IgG antibodies for SARS-CoV-2 antigens—namely the spike and nucleocapsid proteins—was generally similar between blood and saliva samples.

The implications of this data extend beyond the primary findings of the manuscript. We believe that scRNAseq catalogs provide a simple roadmap of the cell type-specific viral vulnerabilities to both known and novel viruses— not just SARS-CoV-2— in the oral cavity. Ultimately, these catalogues may enable better preparedness for the next pandemic and offer targets for prevention or therapeutic intervention. Our results could also shed additional light on the “silent spreader” phenomenon where an appreciable proportion of participants were asymptomatic at collection, despite some with very high viral loads, and that some asymptomatic individuals had infectious virus in their saliva. Finally, we speculate—in addition to individuals experiencing “long COVID” symptoms— that the long-term impact of SARS-CoV-2 infection on health and wellness will not be fully realized for years to come.

In summary, this study provides the first extensive evidence in humans that SARS-CoV-2, the virus that causes COVID-19, can infect cells in the mouth, including the salivary glands. The results suggest that infected oral cells harbor replicating virus and that oral cavity tissues are one likely source of infectious virus in saliva. Moreover, the findings may have implications for understanding viral spread within the body—from one body region (the mouth) to another (the lungs or gut)—or from one person to another. Our conclusions underscore the importance of public health measures to reduce transmission of the virus (e.g., mask wearing and social distancing). These data also reveal the complexity and heterogeneity of SARS-CoV-2 infection and pathogenesis. Importantly, the presence of an oral reservoir of infection could underlie reported oral signs and symptoms of COVID-19 including taste loss, dry mouth, and salivary gland swelling, as well as mucosal ulcers, sores, or possibly tongue swelling or bumps (popularly known as “COVID tongue”); however the pathogenesis of these reported symptoms remains unknown.

"SARS-CoV-2 infection of the oral cavity and saliva" was published in Nature Medicine on March 25th, 2021 and can be found here:

Blake M. Warner

Principal Investigator, National Institutes of Health, National Institute of Dental and Craniofacial Research