1 INTRODUCTION

Since December 2019, an outbreak of a novel coronavirus (SARS-CoV-2), which started in Wuhan, China, has been spreading rapidly, resulting in a potentially life-threatening viral respiratory disease named as coronavirus disease 19 (COVID-19) by the World Health Organization (WHO).¹ By 23 August 2020, more than 20 million cases of COVID-19 and more than 800 000 deaths have been confirmed in over 200 countries, which has been classified as a pandemic by the WHO on 11 March 2020.^{2, 3} This in turn had a severely detrimental impact on the public health and economy of several countries worldwide. As one of three diagnostic criteria according to guidelines set by health authorities in China, the positive detection rate of SARS-CoV-2 nucleic acids in oropharyngeal swabs of infected patients is relatively high.⁴ At the same time, the presence of SARS-CoV-2 in various other human secreta and excreta such as sputum, faeces and urine has also been reported. Nevertheless, as one of the most readily accessible and easily collected bodily fluids, it has not been confirmed whether there are detectable levels of SARS-CoV-2 within the saliva of COVID-19 patients.

Like blood, saliva is rich in multiple biological markers such as DNA, RNA, proteins, with readily detectable levels of microorganism, as the two biofluids share many similarities in molecular components.^{5, 6} Hence, there is a high probability of antibodies and viruses of the human body being present in saliva. It has been reported that some viruses of large-scale infectious diseases, particularly the respiratory diseases such as severe acute respiratory syndrome (SARS) and middle east respiratory syndrome (MERS), can be detected in saliva.^{7, 8} Moreover, saliva has already been utilized for the detection of human immunodeficiency virus (HIV), hepatitis B virus (HBV) and various drugs like cocaine and alcohol. ^{9, 10} As a result, it is imperative to determine whether SARS-CoV-2 can be detected in saliva, as well as characterize the initial oral symptoms manifested in COVID-19 patients.

Single-cell RNA-seq data analysis of angiotensin-converting enzyme II (ACE2) expression and serological investigation of patient samples have indicated that ACE2 may be the cell receptor of SARS-CoV-2,^{11, 12} thus suggesting that ACE2-expressing cells are likely to be the major target cell type which are vulnerable to SARS-CoV-2 infection. There is typically high expression of ACE2 r on the epithelial cells of oral mucosa, which is particularly enriched in the epithelial cells of the tongue.¹³ Prior to this study, there have been few reports that specifically investigated whether the salivary gland epithelial cells express ACE2 receptors. To confirm this, bulk RNA-seq profiles from four public datasets (https://www.proteinatlas.org/) including GTEx dataset, HPA dataset, FANTOM5 dataset and Consensus dataset were analysed. Furthermore, to verify that there are detectable levels of SARS-CoV-2 nucleic acids in saliva, we have screened the saliva of COVID-19 patients (who displayed positive results for SARS-CoV-2 nucleic acids detection before or on the day of sample collection). The results have been compared with oropharyngeal swabs, which is one of three diagnostic criteria.⁴ Additionally, we also carried out a survey on the oral health status of COVID-19 patients, to evaluate the health status of the salivary glands. Our study thus provides a more comprehensive understanding of the detection of SARS-CoV-2 in saliva, and initial oral symptoms upon COVID-19 infection of oral tissues.

2 MATERIALS AND METHODS

This study was approved by the Medical Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China (20200062).

3 RESULTS

The data from GTEx dataset, HPA dataset, FANTOM5 dataset, and Consensus dataset in the Human Protein Atlas (Human Protein Atlas available from https://www.proteinatlas.org/, images/data available from v19.3.proteinatlas.org) showed the expression and distribution of ACE2 within normal human tissues, particularly in salivary glands. The sample ID in GTEx data set, HPA data set and FANTOM5 data set is shown in Tables S1-S3. The results showed that ACE2 is highly expressed in the gastrointestinal system, testes, kidney and heart muscles. Although ACE2 is also expressed in the salivary glands, its expression is at a relatively low level. The results of RNA-seq showed that in normal salivary gland tissues, ACE2 is mainly expressed by glandular cells (Figure 1).

In 31 COVID-19 patients, based on the inclusion criteria, there were 15 male and 16 female patients. The median age was 60.6 years, ranging from 18 to 86 years. There were 5 critically ill patients who needed ventilator support to breathe in these 31 cases. Saliva and oropharyngeal swabs samples were collected at the same time. In our study, 13 cases were tested positive for oropharyngeal swab nucleic acids detection. Among these 13 patients, there were 4 cases with positive nucleic acids detection in saliva, of which 3 cases were critically ill patients on ventilator support, and 1 case being ordinary or heavy without ventilator support (Table 2). The information and test results of the 4 cases are shown in Table 3.

Of the 108 valid questionnaires, 52 respondents were male and 56 respondents were female. The age data of one female patient were missing. After excluding this case, the overall average age was 52.0, 51.1 years for males and 52.9 years for females.

Among the 14 oral-related symptoms listed, amblygeustia (47.2% overall, 36.5% in males, 57.1% in females) and dry mouth (46.3% overall, 46.2% in males, 46.4% in females) had the highest incidence. 11.1% (13.5% in males, 8.9% in females) of the patients exhibited dryness and inflammation of mouth. One female patient (0.9% overall, 1.8% of female patients) had enlargement of lymph nodes in the submandibular regions. (Table 4 and Figure 2 present the results of the questionnaire).

4 DISCUSSION

In our study, we found 4 cases with positive detection of SARS-CoV-2 nucleic acids in saliva. This result thus confirmed the possibility of SARS-CoV-2 being present in saliva. Hence, the potential transmission risk of salivary fluids should not be ignored. Our discovery is consistent with the tissue distribution of ACE2 expression reported in the public database (https://www.proteinatlas.org/). The ACE2 expression was present at detectable levels in the salivary glands, but was lower than other tissues like the gastrointestinal system, testes, kidney and heart muscles (Figure 1). This might explain why there was just one positive salivary detection in ordinary and heavy patients. A previous study reported the detection rate of SARS-CoV-2 in self-collected saliva.¹⁶ However, in that study, patients were asked to cough out saliva from their throat into sterile containers, and hence the saliva samples were mainly sputum from the lower respiratory tract. In our study, the saliva samples were the pure fluid from the canal of salivary glands.

It is noteworthy that the positive saliva detection rate was as high as 75% (3/4) in critically ill patients on ventilator support in our study. There might be two possible explanations. First, viral replication in critically ill patients is unchecked by a weakened immune response, so there is a very large amount of virus in the blood or tissue fluids. Hence, the salivary glands are attacked due to the high viral load. Second, in critically ill patients, there is usually a weakened immune system, electrolyte disorders and multiple organ dysfunction, particularly at the late stage of the disease which could destroy the salivary glands and lead to virus invasion of the salivary glands. The results indicated that as SARS-CoV-2 infection progressed, the chances of viral particles appearing in the saliva increase. The emergence of the virus in saliva may be an indication that the disease condition of the patient has deteriorated and that the disease has entered the terminal stage. Furthermore, it needs to be emphasized that healthcare workers (HCWs) should pay more attention to increased risks of transmission of COVID-19 virus through saliva, and that patients with a positive saliva detection may indicate a poor prognostic outcome.

Specifically, for critically ill patients, saliva has a higher potential for detection of SARS-CoV-2. Although positive nucleic acids detection of oropharyngeal swab is one of the diagnostic criteria for COVID-19 testing,⁴ HCWs are exposed to relatively high infection risks while collecting samples. Moreover, collecting oropharyngeal samples may cause discomfort to patients, such as pain, nausea and even bleeding. Therefore, saliva may be a new source of diagnostic specimens for critically ill patients, since it can be easily collected without any invasive procedures, which is advantageous for both HCWs and patients, especially for multiple sampling and continuous monitoring of viral loads.

Another result worth noting in this study is that according to the analysis of the questionnaire, the two major oral-related symptoms, dry mouth and amblygeustia, were manifested by a relatively high proportion of 108 COVID-19 patients. A previous study identified fever, fatigue and dry cough as the most common symptoms of COVID-19 infection.¹⁷ However, a number of recent studies have found that COVID-19 patients generally have oral symptoms such as dry mouth and amblygeustia, which are described in a review by Pellegrino et al.¹⁸ An international multicentre study collected information from 394 patients diagnosed with COVID-19 from multiple countries, 161 of whom had olfactory or taste disorders, accounting for 41%.¹⁹ A cross-sectional study reported that 20% of COVID-19 patients would feel taste abnormalities.²⁰ There have been news reports that even children recovering from mild COVID-19 are still unable to taste food.²¹ The researchers found a correlation between dry mouth and bitter taste in COVID-19 patients in a recently published study.²² Ren et al also suggest that COVID-19 patients may develop early oral symptoms (including taste loss and dry mouth), even before fever and dry cough.²³ Therefore, some experts have suggested that olfactory or taste disorders should be included in COVID-19 screening criteria.²⁴ Oral health researchers may play a more active role in the early diagnosis and treatment of COVID-19 by exploring the mechanism of dry mouth and amblygeustia.

Besides viral invasion, we speculate that oral symptoms may also be due to the patient's change in psychological status, poor oral hygiene or microbiota imbalance caused by therapeutic drugs. In particular, high ACE2 expression can be found on the epithelial cells of oral mucosa, which is enriched in epithelial cells of the tongue, thus providing possible routes of entry for SARS-CoV-2.¹³ Therefore, oral mucosa might be at potential risk of infection by SARS-CoV-2, suggesting that oral symptoms could also be considered as initial symptoms of COVID-19 infection, which might be new diagnostic criteria. At the same time, doctors should pay more attention to the oral condition and oral hygiene of patients, and give proper oral treatment during the clinical treatment process.

This study thus confirms that ACE2 is also expressed in the salivary glands, SARS-CoV-2 can be detected in saliva, and oral symptoms may be frequently manifested by COVID-19 patients. The findings of this study suggest that saliva may carry a risk of SARS-CoV-2 transmission, particularly in critically ill patients, and that SARS-CoV-2 could cause partial impairment of oral tissues, thus providing a new insight to the clinical prevention, diagnosis and treatment of COVID-19. More in vitro and in vivo evidence and in-depth histological data are needed to further confirm and reinforce these findings.

CONFLICT OF INTEREST

We declare no competing interests.

AUTHOR CONTRIBUTIONS

S. W., L. W. and Y. J. designed and supervised this study, and revised the manuscript; L. C. carried out the experiments, analysed the data, wrote and revised the article; J. Z. and J. P. searched the databases, analysed the data, wrote and revised the article; X. D. analysed the data and revised the article; X. L., Z. S. and F. G. collected clinical samples and case data; Z. G. and Q. Z. performed the laboratory testing.