Salivary Glands and Viral Pathogenesis

Hepatitis C virus

Hepatitis C virus (HCV) is a blood-borne pathogen that infects the liver, causing chronic inflammation and damage that can lead to cirrhosis, liver failure, and hepatocellular carcinoma. Since its discovery, several extrahepatic manifestations of HCV infection have been described. The early observation of a SjD-like clinical presentation in HCV-infected patients prompted further investigation into the role of HCV infection in the SGs, which revealed high rates of focal lymphocytic sialadenitis and reports of xerostomia (the sensation of oral dryness) as compared to healthy controls (Haddad et al. 1992). Several groups have since established key differences between classical SjD and HCV-related SG disease. The sialadenitis observed in HCV-infected patients is generally milder than in SjD, and patients present with an absence of SjD-specific autoantibodies (e.g., anti-Ro/SSA and anti-La/SSB) (Pirisi et al. 1994; Jorgensen et al. 1996). Further, a retrospective analysis of 177 HCV-infected patients found no association between anti-HCV treatment and the eventual development of SjD (Tung and Chen 2022). Despite the variation from primary SjD, HCV infection is commonly associated with xerostomia and/or salivary hypofunction (Grossmann et al. 2010; Maldonado et al. 2022). Chemical analysis of saliva from HCV-infected patients demonstrated an increase in sodium and a decrease in mucin 5b and 7 compared to saliva from healthy volunteers (Maldonado et al. 2022), indicating potential changes to SG composition and absorption/secretion function in the context of HCV infection. Overall, these findings suggest a distinct HCV-associated SG disease.

The exact mechanisms by which HCV infection leads to sialadenitis or salivary hypofunction are not fully understood, but HCV RNA has been detected in SG epithelial cells by polymerase chain reaction (PCR) and in situ hybridization (ISH) (Takamatsu et al. 1992; Arrieta et al. 2001; Maldonado et al. 2022). Both the viral genomic (positive) and replicative (negative) strands were detected in SG biopsies, indicating active replication in the SGs (Takamatsu et al. 1992; Arrieta et al. 2001). Whether this infection can trigger an inflammatory response, SG destruction, and resulting deleterious effects on SG function is still unknown. Studies in transgenic mouse models expressing the HCV envelope proteins revealed a focal inflammatory infiltrate in the parenchyma of the SGs (Koike et al. 1997), but the direct effect on saliva flow was not measured. Further research is needed to fully elucidate the mechanisms underlying HCV-associated sialadenitis and SG dysfunction to determine if the hyperinflammatory host response in other body sites, such as the liver, is also observed in the SGs.

Hepatitis D virus

Hepatitis delta virus (HDV) is a satellite RNA virus that requires a helper virus, such as hepatitis B virus, for viral particle assembly. HDV has primarily been studied in the context of liver dysfunction, where with hepatitis B virus coinfection it is associated with hepatitis and hepatocellular carcinoma. However, recent detection of HDV in patient minor SG samples indicates an extrahepatic tropism. The application of next-generation sequencing and high-density microarray technology in minor SGs from a cohort of patients with SjD led to identification of HDV in approximately 50% of patients with SjD (Weller et al. 2016). Immunohistochemistry revealed HDV antigen in the mitochondria and nuclei of both acinar and ductal cells (Hesterman et al. 2023). While viral genomes were detected in SG epithelial cells, viral replicative strands were not detected, possibly due to a low rate of HDV replication (Hesterman et al. 2023). Exploratory studies in mice confirmed that expression of HDV antigen in the SG could induce a SjD-like phenotype, including the development of anti-SSA/Ro and anti-SSB/La autoantibodies and decreased saliva flow (Weller et al. 2016). This finding supports further investigation into an association between HDV infection and SjD in the SG.

Enteric viruses

Enteric viruses are a diverse group of mostly RNA viruses that replicate in the gastrointestinal (GI) tract. Many, including human norovirus, rotavirus, and astrovirus, are well-known agents of acute gastroenteritis. Measures to limit the transmission of enteric viruses have primarily focused on transmission via the fecal–oral route and contact with contaminated surfaces. Genomic RNA or antiviral antibodies for norovirus (Zhuo et al. 2018; Anfruns-Estrada et al. 2020), rotavirus (Aiyar et al. 1990; Zhuo et al. 2018), and astrovirus (Zhuo et al. 2018) have also been detected in saliva samples or oral swabs from infected patients.

The source of virus in saliva has long been assumed to be a product of gut contaminants. However, a recent study by Ghosh et al. demonstrated that murine norovirus 1 (MNV-1) and murine rotavirus (epizootic diarrhea of infant mice [EDIM]) replicate in murine SGs (Ghosh et al. 2022). When uninfected dams were suckled by pups infected with MNV-1 or EDIM, virus was subsequently detected in the mammary glands, suggesting direct infection of mammary glands through saliva backflow from infected pups. MNV-1 and EDIM replicated in SG immune and epithelial cells and in ex vivo in murine salispheres. Finally, human norovirus replicated in human SG cell lines, providing a novel human cell culture system for studying norovirus infections (Ghosh et al. 2022). This study was the first to demonstrate the replication of acute gastroenteritis viruses in the SG, meriting future study of the role of the SGs in human acute gastroenteritis virus transmission and the potential chronic impact of infection on SG function.

Unlike viral agents of acute gastroenteritis, coxsackieviruses are enteric viruses that replicate in the GI tract but cause a broad range of diseases, including hand, foot, and mouth disease and myocarditis. Coxsackievirus infection has been linked to the SGs based on studies relating infection to SjD. A comparison of differentially expressed transcripts from minor SG biopsies revealed expression of a coxsackievirus B4 gene in samples from patients with primary SjD but not in healthy controls. Further analysis revealed presence of the 5′ noncoding region of the coxsackievirus genome and expression of the viral VP1 protein in primary SjD minor SG biopsies (Triantafyllopoulou et al. 2004). A follow-up study demonstrated that a portion of the coxsackie A21 protein 2B shared homology with an epitope of the host Ro60 protein and that both are targeted by anti-Ro antibodies in serum from patients with primary SjD, suggesting a form of molecular mimicry (Stathopoulou et al. 2005). Whether coxsackievirus can directly infect and replicate in the SGs remains an open question in need of further investigation.

Severe acute respiratory syndrome coronavirus 2

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the etiologic agent of coronavirus disease 2019 (COVID-19). Though the upper respiratory tract is recognized as the major site of SARS-CoV-2 replication, viral genomic RNA is readily detected in saliva of both symptomatic and asymptomatic individuals, allowing for widespread viral surveillance measures. Recent single-cell RNA sequencing (scRNA-seq) analysis of minor SG biopsies revealed SARS-CoV-2 in the ductal and acinar epithelial cells of the SGs, which also express the viral entry factors ACE2 and TMPRSS (Huang et al. 2021). In addition, postmortem biopsies of the major and minor SGs of patients who succumbed to COVID-19 revealed SARS-CoV-2 genomes in 75% of cases, with accompanying histopathologic degenerative changes in both the acinar and ductal epithelial cells (Matuck et al. 2021).

Initial studies in patients who have recovered from COVID-19 suggest that SARS-CoV-2–induced oral manifestations, including xerostomia and SG swelling, may persist following clearance of infection (Gherlone et al. 2021). Further, emerging evidence from postacute COVID-19 subjects suggests both symptomatic and histopathological overlap with SjD in the SG (Shen et al. 2023). The long-term effects of SARS-CoV-2 infection in the SGs and development of autoimmunity will likely continue to be recognized with a growing cohort of patients who have recovered from COVID-19.

Human T-cell leukemia virus type I

Human T-cell leukemia virus type I (HTLV-I) is a single-stranded RNA virus that primarily infects CD4⁺ T cells. HTLV-1 is the etiologic agent of adult T-cell leukemia/lymphoma (ATL) and HTLV-I–associated myelopathy/tropical spastic paraparesis (HAM/TSP). The link between HTLV-1 and SG disease began with the observation that patients with HAM/TSP exhibit higher rates of SjD-like SG inflammation (Vernant et al. 1988; Nakamura et al. 1997). Shortly after, a study in transgenic mice expressing the HTLV-1 tax gene revealed rapid proliferation of the SG epithelial duct and subsequent mononuclear cell infiltration and destruction of acinar architecture (Green et al. 1989). Several groups have since detected high levels of HTLV-1 tax in the SGs of patients with SjD (Sumida et al. 1994; Mariette et al. 2000). The cellular localization of tax is unclear, as some studies found HTLV-1 gene expression in the ductal lumen and inflammatory infiltrate (Tangy et al. 1999) while another only found HTLV-1 gene expression in the surrounding lymphocytic cells (Ohyama et al. 1998). One group detected HTLV-1 protein p19 in addition to tax in the acinar and ductal cells of the minor SGs (Lee et al. 2012). However, other groups failed to detect HTLV-1 tax in minor SG samples (Rigby et al. 1996). In vitro studies in primary SG epithelial cells cocultured with HTLV-1–infected T cells showed that SG epithelial cells can be infected with HTLV-1, albeit at low rates, and that coincubation with infected cells leads to the production of proinflammatory cytokines (Nakamura et al. 2015). Further, HTLV-I–specific T cells are present at high levels in the SGs of infected individuals, suggesting a potential role in SG dysfunction (Ohyama et al. 1998).

While these studies suggest an association between SjD and HTLV-1 infection, the detection of SjD-specific autoantibodies with the concomitant detection of the HTLV-1 tax gene is rare. HTLV-1 infection may instead induce an inflammatory phenotype in the SGs that is not autoimmune in nature. In a study of 272 patients infected with HTLV-1, 20% presented with sicca syndrome, but none presented with SjD-specific autoantibodies (Lima et al. 2016). Xerostomia was also one of the most common oral manifestations reported in patients infected with HTLV-1 and/or HTLV-2 (Martins et al. 2010). While the literature surrounding the HTLV-1/SjD association spans decades, there is still a significant knowledge gap in our understanding of whether HTLV-1 establishes a persistent infection in the SGs or if detection of viral gene expression is a remnant of a previous infection.

HIV

Human immunodeficiency virus (HIV) is a single-stranded RNA retrovirus that infects and depletes CD4⁺ T cells, which can lead to acquired immunodeficiency syndrome. While there are several manifestations of HIV infection, one of the most well-documented presentations in the head and neck region is HIV-associated salivary gland disease (HIV-SGD), which consists of swelling of the SGs, particularly the parotid glands, with or without xerostomia and/or decreased salivary flow (Schiødt et al. 1989). Several HIV-associated SG lesions have also been described. These lymphoepithelial lesions are the result of lymphocytic proliferation in the SGs as well as the presence of hyperplastic lymph nodes in the glands. HIV-SGD has been associated with diffuse infiltrative lymphocytosis syndrome and lymphoepithelial parotid cysts. Several reviews have been published detailing the clinical manifestations of HIV-SGD (see Jeffers et al. 2009; Islam et al. 2012).

The histopathology of HIV-SGD is similar to SjD, with lymphocytic infiltration of the SG tissues and increased fibrosis and collagen deposits in the glands, but differs in the localization of inflammation to the periductal region and absence of autoantibodies (Schiødt et al. 1989; McArthur et al. 2003). Therefore, although similar in presentation, HIV-SGD and SjD are generally considered separate entities. Interestingly, despite the introduction of highly active antiretroviral therapy (HAART), there is no conclusive evidence that HAART decreases the incidence of HIV-SGD, with some studies even showing an increase in HIV-SGD following the onset of HAART (Patton et al. 2000).

The etiology of HIV-SGD is still unknown. One hypothesis is that the phenotype in the SG is a direct result of HIV antigens. However, only a few studies have detected HIV in the salivary glands, with HIV p24 protein detected by immunohistochemistry in the periductal regions and dendritic cells of SGs from patients with HIV-SGD (Rivera et al. 2003). Another possibility is that HIV-associated opportunistic infections promote HIV-SGD. For example, human cytomegalovirus (HCMV) detection in the saliva of HIV coinfected patients was linked to xerostomia and hyposalivation (Greenberg et al. 1997) while Epstein–Barr virus (EBV) coinfection of the SGs was linked to benign lymphoepithelial cysts (Yen et al. 2004). Given the persistence of HIV-SGD prevalence despite HAART, further work is needed to identify the pathophysiological mechanism of SG disease associated with HIV infection.

Human cytomegalovirus

The human herpesvirus family consists of 8 viruses that establish chronic latent infections in different sites throughout the body and can be reactivated to enter a productive lytic cycle. While cellular tropism between the herpesviruses varies, several studies have investigated herpesvirus infections in the SG. For example, EBV genomes have been detected in lymphocytes in SG tissue biopsies from patients with chronic sialadenitis, and the link between EBV infection and SjD has been widely investigated (reviewed in Maslinska 2019). In addition, human herpesvirus 6 (HHV-6) and 7 (HHV-7) have been detected in SG biopsies from healthy individuals (Sada et al. 1996), supporting the notion that the SGs serve as a herpesvirus reservoir. Among the human herpesviruses, human cytomegalovirus (HCMV) has been the most comprehensively characterized for its role in SG infection. Children infected with HCMV exhibit high levels of viral shedding in saliva for weeks (Cannon et al. 2014), with higher levels of shedding in saliva from younger children as compared to seropositive adults (Stowell et al. 2014).

Murine cytomegalovirus (MCMV) recapitulates many features of HCMV pathogenesis and has contributed significantly to the understanding of HCMV immune response and tissue tropism. Early studies revealed persistent viral infection of the SGs, with viral replication and release into duct lumen, infiltration of inflammatory cells, and subsequent necrosis of SG acinar cells (Henson et al. 1972). Since then, several groups have described unique patterns of immune regulation in the SG that support persistent MCMV infection, including a decreased role of CD8⁺ T cells in clearing MCMV infection (reviewed in Campbell et al. 2008). MCMV encodes several genes to establish a chronic infection in the SG, including the SG growth gene, sgg1 (Manning et al. 1992), and the viral chemokine homolog MCK-2 (Fleming et al. 1999). Taken together, mouse models demonstrate the SG as a unique site of host immune evasion and viral persistence for MCMV infection.

Although HCMV DNA has been detected in minor SG biopsies from patients with SjD and nonspecific sialadenitis (Maitland et al. 1995), a causal link between HCMV infection and SG disease has not been established. Several groups have used murine models to show that MCMV infection coupled with specific genetic backgrounds can lead to SG dysfunction. MCMV infection of mice deficient in apoptosis signaling showed enhanced SG inflammatory T-cell infiltrates and histopathology (Fleck et al. 1998). Similarly, MCMV infection of mice prone to development of autoimmune disease led to progressive focal lymphocytic infiltration in the SGs, decline in salivary function, and elevated autoantibody levels (Ohyama et al. 2006; Carroll et al. 2012).

In addition to the effects of MCMV infection on SG function, HCMV may cause developmental defects in utero. Analysis of submandibular gland samples from fetuses from HCMV-positive mothers revealed inflammatory infiltrates and reduction in terminal branching of SG acini as compared to glands from fetuses of uninfected mothers (Gabrielli et al. 2017). These observations in human SGs correspond to earlier observations of abnormal development in MCMV-infected mouse embryonic submandibular gland explants. MCMV infection led to a decline in epithelial branching, appearance of pseudostratified ducts, and predominant hallmarks of mesenchymal, rather than epithelial, tissue (Melnick et al. 2006). Future experiments in in vitro models, such as SG explants, will be essential in understanding the embryopathology of HCMV infection.

BK polyomavirus

BK polyomavirus (BKV) is a small DNA tumor virus that ubiquitously infects the human population and establishes chronic infection in the uroepithelium. Although the urogenital tract is thought to be the primary viral reservoir, BKV has been detected in the saliva of healthy individuals, with salivary titers peaking in the third decade of life and remaining high during adulthood (Robaina et al. 2013), suggesting that the SG may also serve as a site of BK virus infection.

In addition to its detection in saliva, BKV has been described in relation to HIV-associated SG disease (HIV-SGD), which is characterized by lymphocytic infiltration of the SG and xerostomia (Schiødt 1992). BKV is shed at higher levels in HIV-positive individuals with SG disease as compared to healthy controls (Jeffers et al. 2009). Further, BKV isolated from throat wash samples from patients with HIV-SGD had a unique signature in the viral noncoding control region (NCCR), which dictates host cell permissiveness and replication efficiency. Most NCCRs of viruses isolated from throat wash samples were rearranged, and this NCCR architecture was distinct from strains derived from other sources (Burger-Calderon et al. 2014, 2016). These findings suggest that BK virus derived from HIV-SGD patients may represent a unique strain subset with a SG tropism.

BKV has been shown to infect and replicate in human SG cells in vitro. Studies in human SG cell lines demonstrated that lab strains of BK virus productively infect submandibular and parotid cell lines leading to the expression of viral genes and detection of virions in the cytoplasm by 5 days postinfection (Jeffers et al. 2009). BK virus isolated from throat wash samples from individuals with HIV-SGD was also able to replicate in human SG cell lines at levels comparable to laboratory strains and urine-derived virus (Burger-Calderon et al. 2014, 2016). Thus, the SG may serve as a novel site for BK virus replication, and further studies are needed to identify whether BK virus SG tropism can elicit an immune response in the glandular tissues.