Mucosal immunity to severe acute respiratory syndrome coronavirus 2 infection

Purpose of review Despite its crucial role in protection against viral infections, mucosal immunity has been largely understudied in the context of coronavirus disease 2019 (COVID-19). This review outlines the current evidence about the role of mucosal immune responses in the clearance of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, as well as potential mucosal mechanisms of protection against (re-)infection. Recent findings The angiotensin-converting enzyme 2 cellular entry receptor for SARS-CoV-2 is most highly expressed in the upper respiratory tract and most SARS-CoV-2 shedding occurs from the upper respiratory tract. Viral shedding peaks early during infection around the onset of symptoms, before dropping rapidly in most individuals within 7 days of symptom onset, suggesting mucosal inhibition of viral infection. Serum and mucosal immunoglobulin G and immunoglobulin M responses were found to be strongly correlated in infected patients, whereas correlations were much weaker for immunoglobulin A (IgA). Mucosal IgA responses have been detected in infected cases in the absence of serum antibody responses, with mucosal antibody levels correlating strongly with virus neutralization. Bulk and single-cell RNA sequencing analysis of nasopharyngeal swabs and bronchoalveolar lavage samples of COVID-19 patients revealed the induction of mucosal chemokine and cytokine genes, complement pathways, Janus Kinase/Signal Transducer and Activator of Transcription signaling and cytotoxic T cells. Summary Although most clinical studies focus on antibodies and cellular immunity in peripheral blood, mucosal immune responses in the respiratory tract play a key role in the early restriction of viral replication and the clearance of SARS-CoV-2. Identification of mucosal biomarkers associated with viral clearance will allow monitoring of infection-induced immunity. Further studies are needed to understand how the systemic immunological endpoints measured in vaccination studies translate to mucosal protection against SARS-CoV-2 infection.


INTRODUCTION
Since the beginning of 2020, the world has been severely impacted by the coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), a newly emerged coronavirus. Coronavirus disease is characterized by a large spectrum of clinical symptoms. Approximately 20% of infected cases develop severe or critical COVID- 19  ]. Severe and critical COVID-19 is associated with respiratory failure, acute respiratory distress syndrome (ARDS), sepsis and septic shock, thromboembolism, multiorgan failures and other complications [2 & ]. Several risk factors have been identified for severe morbidity of COVID-19 and/or mortality, including older age, obesity, high blood pressure and diabetes (reviewed elsewhere [3]). The majority of individuals infected with SARS-CoV-2 develop no or mild symptoms (40%) or moderate symptoms (40%) [2 & ]. Young age in particular is an important determinant for mild and asymptomatic infection. Mild or moderate COVID-19 is characterized by flu-like symptoms, respiratory symptoms, gastrointestinal symptoms and loss of smell [4][5][6][7][8]. Cognitive symptoms may also develop, including dizziness, confusion, and memory loss [8]. Around 10% of mild cases develop what is now known as 'long COVID-19', characterized by longer-term sequelae [9].

Viral replication and shedding
Viral infectious dose and the duration and frequency of exposure are key determinants in the establishment of SARS-CoV-2 infection [10]. It is likely that many -if not most -transmission events of SARS-CoV-2, that is, the transfer of virus particles between individuals, will not lead to the establishment of infection. For instance, virus particles may get trapped in the mucous layer that lines the respiratory epithelium, a complex gel-like matrix of mediators containing, for example, mucins and glycans, and subsequently removed by mucociliary clearance [11]. Having overcome the mucous barrier, virus particles attach to host cells through binding of the viral spike protein to the cellular entry receptor angiotensin-converting enzyme 2 (ACE-2) [12 & ]. Facilitated by host factors such as transmembrane protease, serine 2 (TMPRSS2) [12 & ], virus particles are then internalized and fusion occurs at the cellular or endosomal membrane. Viral genomic RNA is then released and viral replication initiates. The replication cycle of SARS-CoV-2 has been estimated at approximately 10 h [13], after which viral progeny is released by exocytosis [14]. Although the expression level and tissue distribution of ACE-2 and TMPRSS2 vary with genetic factors, sex, age, and comorbidities like obesity [15,16], they are mostly highly expressed in the upper respiratory tract (URT) [17,18]. The salivary glands also act as a viral repository [19 & ,20]. Thus, initial viral replication and shedding mostly takes place in the URT, and precedes viral replication in the lower respiratory tract (LRT) [1 && ]. Once the virus reaches the LRT, pulmonary cells are infected and patients may become viremic. The route of the virus and the timing of viral replication, first in the URT and later in the LRT, is reflected in the clinical presentation; URT symptoms including congestion, sore throat and loss of taste precede LRT symptoms such as dysp- 21]. Investigations on the transmission to close contacts indicate that shedding of SARS-CoV-2 viral particles is highest in the URT at the first stage of infection and can already occur during the incubation period, that is, the time between exposure and symptom onset, known as the presymptomatic period [2 & ]. This is underlined by research looking at the viral load dynamics, which found that viral load peaks approximately two days before symptom onset after which it declines [1 && , 22,23]. Predictive modeling suggests that up to 80% of transmission occurs prior to symptom onset [24].

Role of mucosal immunity in viral clearance and protection against re-infection
Given that viral loads decrease rapidly after the onset of (mostly) URT symptoms and that most infections with SARS-CoV-2 remain mild ora symptomatic, mucosal immune responses presumably play a central role in infection clearance. Despite this, surprisingly few studies have examined mucosal immunity in the context of SARS-CoV-2. The present review outlines the current evidence about the role of human mucosal immune responses in the clearance of SARS-CoV-2 infection, as well as the role of human infection-and vaccinationinduced mucosal immunity in protection against re-infection.

Mucosal immune response to infection
To sustain essential respiratory function, the mucosal immune system in the respiratory tract must strike a careful balance between minimizing inflammation-induced tissue damage and acting adequately on threats such as SARS-CoV-2 infection. Although there have been many publications on profiling immune response dynamics in COVID-19 patients, including advanced single-cell analysis methods, virtually all of these studies focused on peripheral blood. Nonetheless, there have been

KEY POINTS
The kinetics of viral replication and shedding of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) suggest a central role of the mucosal immune response in the protection against infection.
Studies analyzing mucosal antibody responses against SARS-CoV-2 are scarce, but point to a role for mucosal immunoglobulin A in early viral neutralization.
Gene expression studies identify mucosal induction of cytokine networks, antiviral type I interferon signaling, complement pathways and T-cell responses.
Current vaccination studies primarily focus on immune responses in peripheral blood, highlighting a knowledge gap in how these systemic responses translate to mucosal protection. some attempts to study local responses in the respiratory tract, albeit generally with low patient numbers. Single cell (sc) and bulk RNA sequencing of nasopharyngeal swabs and bronchoalveolar lavage samples -performed on a single timepointidentified various genes and cell types associated with COVID-19 and severe disease, and show that cellular responses in the lungs differ from those in the URT. In nasopharyngeal swab samples, COVID-19 patients expressed gene sets involved in cytokine -cytokine interaction, complement pathways, and Janus Kinase/Signal Transducer and Activator of Transcription signaling. These gene sets were found to be more highly expressed in severe/critical patients [25]. Cellular interactions between epithelial and immune cells were also linked to the development of severe symptoms, reflected by increased activation of inflammatory macrophages and cytotoxic T cells [26 & ]. In the bronchoalveolar lavage fluid, a high expression of cytokines, chemokines, and antiviral interferon-stimulated genes [27] was found, as well as chemokine receptors, suggestive of macrophage and neutrophil recruitment [28]. scRNA-seq found that severe COVID-19 patients had CD8 þ T-cell populations that were less expanded and phenotypically heterogeneous, whereas mild COVID-19 patients had higher CD8 þ T-cell counts, with features of tissue residence [29 & ]. These results suggest that during severe disease, inflammatory processes are dysregulated, leading to the recruitment of inflammatory monocytes and neutrophils. In contrast, in mild cases, virusspecific T cells are recruited and -presumably -able to more effectively control viral replication. It should be noted though that none of these studies examined temporal changes of the cellular responses in relation to viral infection kinetics.
Although mucosal B cells were not identified in the scRNA-seq analysis, the production of mucosal antibodies through B-cell activation remains a key component of the antiviral response. In serum, virus-specific immunoglobulin M (IgM) can be detected approximately 7 days post symptom onset, after which class switching occurs and specific immunoglobulin A (IgA) and immunoglobulin G (IgG) are detected. IgG shows the strongest response and is present in high concentrations in convalescent plasma, while the IgA response is more transient [ [35 && ]. Although this study requires further validation, it is tempting to speculate that (mild) infection may not always induce systemic responses but can still activate local immunity. Another preprint study found an inverse correlation between serum and nasal fluid antibody isotypes, with higher mucosal IgA levels correlated with lower IgG. Of note, viral neutralization of nasal fluid samples correlated most strongly with nasal IgA, while for serum strongest correlations with neutralization were observed for IgG [36].

Factors influencing mucosal immune responses
Both genetic and nongenetic factors may contribute to phenotypic variation in the mucosal immune response to SARS-CoV-2 infection. Nongenetic factors may include preexisting cross-reactive memory to seasonal coronaviruses, SARS-CoV-1 and Middle East respiratory syndrome-related coronavirus has been described [37 & ], in particular for the nucleocapsid (N)-protein. This applies to both cross-reactive T cells [38 & ] and antibodies [39]. Whether preexisting memory may promote or hinder clearance is not yet known. It has also been hypothesized that previous vaccinations may contribute to protection against COVID-19 due to shared epitopes. For instance, diphtheria, tetanus, and pertussis (DTP) vaccines are predicted to share CD4 þ and CD8 þ T-cell epitopes and B-cell epitopes to SARS-CoV-2 spike protein [40]. Similarly, the BCG vaccine has been described to contain similar 9-amino acid sequences with SARS-CoV-2, which were related to HLA class I molecules [41]. However, it remains to be seen whether these small changes translate into partial protection or not. Another contributing factor may be the microbiome. Although the direct role of the microbiome on the host response is not yet investigated, several studies suggest that disruption of the gut and nasal microbiome results in more severe COVID-19 disease [42,43].

Vaccine-induced mucosal immunity
At the moment of writing, two COVID-19 vaccines have been authorized for emergency use, that is, Pfizer-BioNTech BNT162b2 and Moderna mRNA-1273. Both vaccines contain mRNA molecules encoding the viral spike protein. The primary efficacy endpoint for all COVID-19 vaccines thus far has been protection against disease [44,45]. Although the phase 3 studies were not powered to detect differences in asymptomatic or subclinical infections, a key question is how well these vaccines provide herd immunity. Although larger studies are now ongoing that will hopefully soon be able to answer this question, protection against subclinical infection -even against homologous virus strains -will likely wane sooner than protection against disease, given that most nonreplicating vaccines do not offer life-long protection [46]. It is therefore crucial to define the immunological correlates of protection for disease as well as for infection. Whereas all clinical studies to date have primarily focused on serum neutralizing antibodies as immunogenicity endpoint [47,48], it is not yet known if and how the new SARS-CoV-2 vaccines elicit durable mucosal immunity. Both the mRNA vaccines as well as the adenovirus vector vaccines currently in phase 3 trials induce strong CD4 þ T-cell responses in blood [49][50][51]. A similar induction has been observed for CD8 T cells in the blood [49,50]. An important question is whether parenteral COVID-19 vaccines will be able to induce tissue resident memory T cells (Trm) in the human respiratory tract. The ability to do so may vary between vaccine formulations. For instance, vaccination with acellular subunit pertussis vaccines did not induce Trm cells in the lungs, whereas Trm cells were observed after vaccination with a whole-cell pertussis vaccine [52].

CONCLUSION
Although severe SARS-CoV-2 infection can have farreaching systemic effects, the infection begins at the mucosal surface and -in most cases -remains restricted to the respiratory tract. It is therefore essential to not only focus on immunological endpoints in blood, but also to establish mucosal immunological correlates of protection. Two distinct immunological processes can be identified that both contribute to protection against infection. Firstly, complete prevention of infection is likely solely mediated by sufficiently high concentrations of neutralizing mucosal antibodies that block the attachment of virus particles to host cells, inducing sterilizing immunity. Polymeric mucosal IgA and IgM may efficiently trap virus particles to facilitate their removal by cilial beating. Mucosal antibody responses are typified by the high production of dimeric secretory IgA (sIgA). sIgA differs from monomeric serum IgA in its molecular form and effector functions [53,54 & ]; with monomeric IgA not effectively being transported to the mucosal surface [54 & ]. sIgA is highly effective against various viruses, can neutralize virus particles within epithelial cells [54 & ], and has anti-inflammatory properties [53,55]. This is fundamentally different from the systemic antibody environment, which is dominated by IgG and is typically pro-inflammatory [56][57][58].
In case virus particles overcome this 'antibody barrier', either by waning of immunity, by exposure to high numbers of virus particles, or by suboptimal binding/neutralization due to e.g. mutations in the spike protein, antibodies may not fully block viral attachment to ACE-2, resulting in viral replication. In this case, additional processes will be required for clearance of viral infection. Antibodies may contribute to this process by binding to viral antigens on infected cells and induce antibody-dependent cellular cytotoxicity (ADCC) or other Fc-mediated effector functions [59]. Additionally, cytotoxic T cells may kill infected cells through the release of granzymes, which induce apoptosis, and perforin that punches holes in the cell membrane.
There are both challenges and advantages to measuring mucosal immunity. Challenges are related to sampling standardization and validation and to availability of robust assays to measure mucosal immunity. Historically, there have been difficulties in the standardization of mucosal sampling and dilution effects are frequently observed with for, example, saliva. Furthermore, mucosal antibody concentrations are significantly lower than in serum, necessitating sensitive detection assays. However, many of these challenges can be or have already been overcome. For instance, precision sampling of undiluted nasal mucosal lining fluid by nasosorption allows robust measurements of antibodies [60]. Highly sensitive and specific detection methods have been established to measure antibodies to SARS-CoV-2 [61]. Although mucosal antibodies are a more likely candidate for correlates of protection, mucosal cellular immunity may also be studied via sampling of mucosal tissue via swabs, curettage or brushing. Whilst cell numbers are typically low in these samples, advances in single-cell analysis methods allow unprecedented in-depth analysis. Importantly, all of these sampling techniques are noninvasive and can therefore be performed in people of all ages, including children. Expanding our knowledge on mucosal immunity to SARS-CoV-2 is essential to understand what causes protection against infection, and to properly monitor the immunological effects of vaccination. The ease of collecting mucosal samples provides the opportunity to analyze immune responses in a noninvasive manner at multiple time points. Mucosal findings will need to be bridged with existing data on systemic immunological endpoints, but recent advances in both sampling and analysis methods pave the way for mucosal correlates of protection.

None.
Financial support and sponsorship None.

Conflicts of interest
There are no conflicts of interest. 2.