2.2.1. After infection with SARS-CoV-2 or other human coronaviruses
An early study in 35 convalescent COVID-19 patients identified distinct differences in functional antibody responses to SARS-CoV-2 in serum and nasal washings [
23]. IgA, IgG and IgM antibodies to SARS-CoV-2 were detected in nasal washings and in sera of patients with symptoms, regardless of whether they required hospitalization or not. SARS-CoV-2 infections were found to elevate levels of antibodies cross-reacting with other types of endemic human coronaviruses in nasal washings and sera [
23]. Importantly, virus neutralization assays showed that neutralization titres with IgA in nasal washings against SARS-CoV-2 were significantly higher in patients who did not require hospitalization compared with those with severe disease that did [
23]. This finding was consistent with the protection afforded by mucosal IgA antibodies in other human and animal coronavirus infections [
23]. IgG antibody-dependent phagocytosis by monocytes and virus neutralization with IgM antibodies were also observed in nasal washings from the same COVID-19 patients [
23]. More recent findings in COVID-19 patients showed that levels of IgA antibodies to the S1 subunit in nasal epithelial lining fluid correlated with virus neutralization titres and reduced viral load in the URT [
24].
Another study on COVID-19 patients admitted to hospital showed that (i) IgA antibodies to SARS-CoV-2 RBD were detected earlier in serum than IgG antibodies; (ii) a large proportion of plasmablasts synthesizing IgA in blood express chemokine receptor 10 (CCR10) which is a marker for their homing to mucosal sites; (iii) Neutralizing IgA anti-RBD antibodies were more abundant in saliva than serum, and were formed early during the onset of symptoms; (iv) neutralizing IgA antibodies remained detectable in saliva for 49 to 73 days after symptom onset [
25]. Nasal SARS-CoV-2 S-specific IgA antibody levels resulting from an infection were boosted by a subsequent infection and displayed significant strain-transcending neutralization capability against SARS-CoV-2 variants [
26]. Furthermore, IgA antibodies in nasal epithelial fluid, produced after infection with early strains of SARS-CoV-2 in 2020, (i) inhibited RBD-ACE 2 binding, (ii) were significantly higher than after two intramuscular vaccinations with the mRNA1273 vaccine expressing S, (iii) recognized RBD from the Delta and Omicron BA.1 variants, and (iv) remained elevated for several months [
19,
20]. These observations also suggest that higher levels antibodies in the URT, particularly antibodies of the secretory IgA type, are a likely correlate of protection in the URT.
CD4+ helper T cells (T
H) and CD8+ cytotoxic lymphocytes (T
C) have been implicated in protecting a group of healthcare workers lacking antibodies to SARS-CoV-2 against infection, despite their likely repeated exposure to the virus in hospitals [
27]. The CD4+ T
H cells obtained from their peripheral blood in this study were shown to recognize cross-reactive epitopes present in replication- transcription complex proteins of SARS-CoV-2 and other common human-infective endemic coronaviruses [
27]. While this [
27] and other indirect data [
14,
18] suggest that T
H and T
C cells are able to protect against SARS-CoV-2 infection, it has been technically difficult to obtain evidence for URT-located T cells in protection against infection. Sampling of nasal mucosal cells by nasal curettage after recovery from COVID-19 identified SARS-CoV-2-specific CD8+ T
C cells with a resident memory phenotype (CD8+ T
RM) persisting in the nasal mucosa for several months after clearance of the virus [
28]. A study of nasopharynx-associated lymphoid tissue (NALT) removed from children who underwent tonsillectomy and adenoidectomy, after a prior SARS-CoV-2 infection, provided important evidence for SARS-CoV-2-specific URT-resident B and T cells in the NALT [
29]. This study found SARS-CoV-2-specific germinal centre and memory B cells that had class-switched to IgA and IgG, and undergone somatic hypermutation in the variable region genes of immunoglobulins, in NALT [
29]. The B cell antigen receptor sequences in the cells were specific for S and matched known S-specific sequences identified in other studies. Tissue resident CD4+ T
H cells (CD4+ T
RM) with a memory phenotype and activated CD8+ T
C with T cell receptor sequences known to be specific for SARS-CoV-2 epitopes were also identified in the NALT. The T
H and T
C cells from the NALT synthesized appropriate cytokines and proteins on in vitro activation [
29]. This study [
29] therefore provides evidence for the presence of tissue resident memory B (B
RM) and T
RM cells recognizing SARS-CoV-2 antigens in the URT NALT after a SARS-CoV-2 infection.
Another recent study found a significant association between asymptomatic infection with SARS-CoV-2 and a class 1 HLA allele
HLA-B*15:01 in unvaccinated persons supporting a role for CD8+ T
C cells in early protection against infection, which is most likely to have occurred in the URT [
30]. CD8+ T
C cells collected in the pre-pandemic area from
HLA-B*15:01 persons reacted with an immunodominant SARS-CoV-2 S-derived peptide epitope with the amino acid (aa) sequence NQKLIANQF. Homologous aa sequences NQKLIANQF and NQKLIANAF were present in the common cold-causing endemic coronaviruses OC43-CoV and HKU1-CoV respectively. NQKLIANAF was shown to be presented by HLA-B*15:01 to CD8+ T cells and cross-react with the SARS-CoV-2 S epitope, thereby providing a likely explanation for pre-existing CD8+ T
C cell-mediated immunity against SARS-CoV-2 infection in the URT [
30].
These recent findings suggest that memory B and T cells are found in URT lymphoid tissue after recovery from COVID-19 or infections with other human coronaviruses, that generate a sufficiently early protective adaptive immune response to eliminate a subsequent infection with SARS-CoV-2 without the development of symptoms.
Figure 1 illustrates the main features relevant to SARS-CoV-2 infection and immunity in the URT.
2.2.2. After intramuscular vaccination of infection-naive persons
The widely used, intra-muscularly administered COVID-19 vaccines based on mRNA, replication-deficient adenovirus vectors, whole inactivated virus and adjuvanted SARS-CoV-2 spike protein (S), have been shown in clinical trials to elicit virus-specific T cells and antibodies in blood, and reduce severe disease and mortality [
5,
10,
14,
18]. The protection from severe disease raises the question whether intramuscularly administered vaccines also generate protective adaptive immunity in the URT which then limit SARS-CoV-2 infections in the URT, thereby minimising infection of the LRT and the development of symptomatic COVID-19. Consistent with this assertion, preclinical studies in animal models with intramuscular COVID-19 vaccines have demonstrated SARS-CoV-2-specific URT immunity [
10]. Experimental findings of potentially protective antibody responses in the URT elicited by the widely-used intramuscularly-delivered adenoviral vector and mRNA COVID-19 vaccines expressing S in persons who had never been infected with SARS-CoV-2 (
Table 1) are consistent with findings from preclinical studies.
Intramuscularly-administered mRNA S-based vaccinations were found to elicit S-specific antibodies in bronchoalveolar fluid of infection-naïve vaccinees, but at lower concentrations than in COVID-19 convalescent persons [
36]. URT responses were not investigated in this study [
36]. Reports of T
H and T
C cell responses in the URT to intramuscularly administered adenovirus-vectored and mRNA S-based vaccines in infection-naïve persons have been sparse [
10]. A recent analysis of cells obtained with nasopharyngeal swabs in persons vaccinated with the Pfizer/BioNTech S-based mRNA vaccine demonstrated the expansion of CD8+ T
RM cells as well as CD4+ T
H cells in persons with no known prior COVID-19 infection [
37]. Another study, however, could not detect S-specific nasal CD4+ T
H and CD8+ T
C cells in persons receiving S-based mRNA intramuscular vaccines unless the vaccinees subsequently became infected with SARS-CoV-2 [
38]. Reconciling these divergent findings is important because tissue resident T
RM and B
RM potentially induced in the URT by intramuscular vaccination can be expected to generate the rapid anamnestic immune response required to limit SARS-CoV-2 infections to the URT, a process which commonly occurs in persons who have recovered from COVID-19 (discussed
Section 2.2.1). Studies on donated organs show that B
RM and T
RM elicited by SARS-CoV-2 infection are found in the lungs, bone marrow, multiple lymph nodes, and spleen six months after infection [
39], raising the importance of investigating their presence also in the NALT and elsewhere in the URT after different types of COVID-19 vaccination, including mucosal vaccination discussed in section 3 below. B
RM and T
RM help generate robust anamnestic immune responses in many tissues [
40,
41,
42] and therefore their potential presence in the URT after COVID-19 vaccination may help limit SARS-CoV-2 infections in the URT.
The kinetics of antibody and immune cell responses in the blood of persons who had been vaccinated three times with S-based mRNA vaccines, and then become infected with the Omicron variant, are also pertinent in this context [
43]. Activation of S-specific CD4+ and CD8+ cells were detected during the first week of infection together with an expansion of S-specific plasmablasts. However, a substantial increase in S-specific neutralizing antibodies occurred mainly in the second week. Antibodies to S generated after Omicron infection were predominantly directed towards epitopes shared with the parent SARS-CoV-2 strain used in the vaccines suggestive of antigen imprinting [
43]. This study [
43] did not investigate URT responses but the findings suggest that eliminating virus early in the URT will require similar recall adaptive immune responses and adequate basal concentrations of antibodies to SARS-CoV-2, as discussed in section 2.2.1.
Many vaccinated persons have experienced breakthrough infections from SARS-CoV-2 variants carrying multiple mutations in S that have been selected to evade neutralization with vaccine-elicited antibodies [
16,
18]. However, S-based booster mRNA vaccines incorporating S from the recently widespread SARS-CoV-2 variants have yielded better protection in Nordic countries [
44], suggesting that this approach is also relevant for inducing URT immunity.