Human milk antibodies possess several advantages to protect against COVID-19, including a high concentration of secretory IgA (SIgA) and the presence of secretory IgM (SIgM) [1
]. SIgA is the most abundant antibody that protects mucosal surfaces against infectious microbes [4
]. SIgM also possesses a superior capacity to bind antigens than IgM [6
]. The secretory component is a critical constituent of SIgA and SIgM to perform immune exclusion and prevent pathogens’ invasion by blocking their attachment to the epithelial cells [7
]. Human milk SIgA and SIgM are resistant to digestive proteases and low pH [2
], allowing them to survive during infant digestion [1
]. Most importantly, human milk secretory antibodies specific to SARS-CoV-2 could help to protect infants against COVID-19 infection. Clinical manifestations of pediatric COVID-19 have been related to the respiratory and gastrointestinal (GI) systems [11
]. It is well-known that breastfeeding reduces the risk of viral infections in exclusively breastfed infants, but the mechanism of how human milk antibodies protect the respiratory tract against viral pathogens is still unclear [12
The spike protein (S) is present on coronaviruses’ surface and is composed of two subunits, S1 and S2, that are essential for both host specificity and viral infectivity [13
]. S1 subunit contains the receptor-binding domain (RBD) that recognizes the cell surface receptor [14
]. S2 subunit contains the required components for membrane fusion [15
]. Antibodies specific to S1 could disable the receptor interactions and block the viral infection process [13
]. S1- and S2-specific antibodies could also destabilize the prefusion structure and block coronaviruses’ entry [13
A recent study demonstrated that mothers with confirmed PCR SARS-CoV-2 had higher levels of antibodies specific to RBD SARS-CoV-2 than in unexposed mothers [16
]. Dong et al. [17
] showed that nucleocapsid-specific IgG and IgA levels were higher after five weeks of post-infection in women with confirmed PCR test than after two weeks. The difference of levels in human milk antibodies specific to S1 or S2 subunit SARS-CoV-2 in mothers with confirmed COVID-19 PCR test remains unknown.
Our recent study demonstrated that S1 + S2-reactive SARS-CoV-2 SIgA/IgA and SIgM/IgM were detected in most human milk samples collected during COVID-19, but mothers did not have a confirmed PCR test [18
]. These mothers had viral symptoms associated with COVID-19 infection. We speculated that human milk secretory antibodies might recognize a broad range of human coronaviruses (HCoVs) due to their polyreactive and cross-reactive properties, but no study has evaluated this hypothesis.
HCoV-OC43 and HCoV-229E cause respiratory infections, such as acute respiratory tract disease and pneumonia. The incidence of HCoV-OC43 infection in hospitalized infants (from birth to six months) is more often than HCoV-HK1, HCoV-229E, and HCoV-NL63 infections (i.e., HCoV-OC43 > HCoV-NL63 > HCoV-HKU1 > HCoV-229E) [19
]. Infants infected by HCoV-OC43 had immunity against HCoV-HKU1 infection [19
]. This result suggests that cross-reactive antibodies to HCoV-OC43 may protect against infection by other HCoVs. The positive correlation between human milk antibodies reactive to S1 + S2 subunits from HCoV-OC43, HCoV-229E, and SARS-CoV-2 is still unexplored.
This study aimed to evaluate the levels of SIgA/IgA, SIgM/IgM, and IgG specific to SARS-CoV-2 S1 or S2 subunit and reactive to HCoV-OC43 and HCoV-229E S1 + S2 subunits in milk from seven women with positive COVID-19 PCR test and six unexposed women (Ctl1-2018 pre-pandemic) as well as between 20 mothers that had viral symptoms during COVID-19 pandemic (no PCR test) and 16 unexposed mothers (Ctl2-2018 pre-pandemic). These results could help identify the critical factors influencing the maternal antibody response. This investigation’s clinical relevance is that women with COVID-19 PCR test and those with previous viral symptoms (but no PCR test) could secrete significant levels of antibodies against SARS-CoV-2 in human milk and provide passive immunity to their infants against coronavirus infection.
Infants and children (<2 years old) cannot receive the SARS-CoV-2 vaccine due to their immature antibody response [20
]. While respiratory symptoms are the main clinical manifestations of COVID-19, gastrointestinal symptoms are also reported in patients (including infants and children) infected with SARS-CoV-2 [21
]. Therefore, human milk antibodies specific to SARS-CoV-2 could protect the infant’s gut against COVID-19 infection. The potential protective effects of human milk antibodies may also decrease viral infection risk by other HCoVs. Antibodies that bind to S1 or S2 subunit could block the attachment or fusion of SARS-CoV-2. Our recent study demonstrated human milk antibodies’ presence reactive to SARS-CoV-2 S1 + S2 subunits in milk from mothers during the COVID-19 pandemic [18
]. However, the levels in antibodies specific to S1 and S2 subunits from SARS-CoV-2 in mothers with confirmed COVID-19 PCR remains to be evaluated. This present study compared the levels of antibodies against S1 or S2 subunit SARS-CoV-2, S1 + S2 subunits HCoV-OC43, and HCoV-229E between human milk collected from mothers diagnosed COVID-19 via PCR, from mothers with previous viral symptoms during COVID-19 pandemic, and from unexposed mothers in 2018. The study’s clinical relevance was to determine whether COVID-19 PCR mothers and mothers with previous viral symptoms had preexisting human milk antibodies against SARS-CoV-2 partially due to cross-reactive antibody reactive to HCoV-OC43 and HCoV-229E.
For the first time, we demonstrated that S2 subunit SARS-CoV-2-specific IgG levels were higher in the COVID-19 PCR and viral symptom groups than in the Ctl-2018 group, but the other antibodies did not differ. While S2 is less exposed than the S1 subunit on SARS-CoV-2, IgG may recognize S2 during the viral entry’s complex conformation change [14
]. Recovered COVID-19 individuals with cross-reactive B cell responses against the S2 subunit may enhance broad coronavirus protection [23
]. Antibodies targeting the S2 subunit had neutralizing activity, suggesting that the presence of S2-reactive IgG provided some protection against SARS-CoV-2 [24
]. Another study [25
] found that the epitope S2-78 exhibited potent neutralizing activity by interfering with the formation of 6-HB (helical bundle), an essential structure for cell membrane fusion. Most studies focus on the RBD region, which may induce potential mutations in this region and reduce the effectiveness of the RBD-specific therapeutic antibodies and vaccines [26
]. The identification of other targets that can confer neutralizing antibodies is critical to overcoming future SARS-CoV-2 mutations.
The lack of difference between COVID-19 PCR, viral symptom, and Ctl-2018 groups for SARS-CoV-2-reactive SIgA/IgA and SIgM/IgM and the high numbers of positive correlations between antigen and secretory antibodies in the COVID-19 PCR group could be related to their polyreactive capacity to bind different epitopes [27
]. These findings are in accordance with our recent study [18
], where S1 + S2 subunits SARS-CoV-2 IgG was higher in milk from women with viral symptoms (but no PCR test) during the COVID-19 pandemic 2020 than in the control group 2018. Moreover, the correlation matrix generated with the Ctl-2018 group had a significantly smaller number of positive correlations than those generated with the COVID-19 group. This observation suggests that human milk antibodies were more specific to SARS-CoV-2 in the COVID-19 PCR group than in the Ctl-2018 group.
The highest levels of antibodies specific to S1 or S2 subunit SARS-CoV-2 was observed in mother-1 with a confirmed + PCR test (M1). This observation could be related to the highest level of long-lived specific T cells that enhance B cell antibody production among COVID-19 mothers. M1 could also have antibodies that were disappearing less quickly (longer half-life) than the other mothers due to their immune status or viral exposure route. The time between the infection and collection did not affect the levels of antibodies specific to S1 or S2 SARS-CoV-2. Individual mothers have different levels of human milk antibody probably due to the differences of previous infections, preexisting immunity, age, genetic factors, and other factors affecting the immune response [29
S1 + S2 subunits HCoV-OC43-reactive IgG was higher in the COVID-19 group than in the control group but did not differ for S1 + S2 subunits HCoV-229E-reactive IgG or other antibodies. These findings are consistent with a recent study [23
]. IgG levels against S protein of HCoV-OC43 (but not against HCoV-229E) were higher in convalescent subjects than in non-SARS-CoC-2-exposed subjects and correlated strongly with anti-S2 IgG levels [23
]. B cells may have a stronger cross-reactivity between the S2 subunits of SARS-CoV-2 and human β-coronaviruses than α-coronaviruses [31
]. Moreover, we observed high SIgM/IgM levels reactive to SARS-CoV-2 S1 subunit and S1 + S2 subunits HCoV-OC43 and HCoV-229E in two unexposed mothers (Ctl1-2018) C4) and C5. These observations suggest the presence of cross-reactive antibodies against human coronaviruses. SARS-CoV-2, HCoV-OC43, and HCoV-229E could share similar epitopes on S1 and S2 subunits that are recognized by human milk antibodies. Our findings are in agreement with recent articles suggesting immunity to “common cold” coronaviruses [32
] may produce cross-reactive antibodies to SARS-CoV-2. Grifoni et al. [33
] detected SARS-CoV-2-reactive CD4+
T cells in ~40–60% of unexposed individuals (blood), suggesting cross-reactive T cell recognition between circulating “common cold” coronaviruses and SARS-CoV-2. SARS-CoV-2 RBD-reactive IgG, IgM, and IgA in serums from non-exposed individuals were detected, but the serum antibody titer was higher in individuals with COVID-19 than in serum from unexposed individuals [33
]. Mateus et al. [32
] demonstrated a range of preexisting memory CD4+
T cells that were cross-reactive to SARS-CoV-2, HCoV-OC43, HCoV-229E, HCoV-NL63, and HCoV-HKU1. S-reactive T cells in patients with COVID-19 responded similarly to HCoV-229E S and SARS-CoV-2 S, suggesting S-cross-reactive T cells’ presence was probably generated during past encounters with endemic coronaviruses [31
]. Preexisting SARS-CoV-2-reactive antibodies could be relevant because mothers with a high-level of preexisting antibodies recognizing SARS-CoV-2 could provide stronger passive immunity to their infants and reduce the risk of COVID-19 and common cold coronaviruses infections.
During COVID-19 infection in lactating women, SARS-CoV-2-specific IgA and IgM could be produced by plasma cells in the interstitial fluid of the mammary gland tissues (MEC) and transported by polymeric immunoglobulin receptor (PIgR) across the MEC (Figure 6
]. SIgA and SIgM are in the highest concentration than IgA or IgM due to this diffusion mechanism in the alveolar lumen. However, there is some proportion of IgA and IgM (without SC) in human milk, likely due to the production of antibodies by B cells present in human milk [1
]. SARS-CoV-2-reactive IgG could be produced in the maternal blood after contact with SARS-CoV-2 or other coronaviruses. To access the alveolar lumen, IgG needs to bind to the neonatal Fc receptor (FcRn) on the basolateral membrane of the MEC (Figure 6
]. Deactivated or intact viruses may activate the adaptive immunity to recognize the viral proteins from coronaviruses, including S1 subunit, S2 subunit, RBD, and nucleocapsid protein (Figure 6
L). As spike proteins from SARS-CoV-2, HCoV-OC43, and HCoV-229E share structural similarities, antibodies produced by B cells after antigen recognition could be cross-reactive to these human coronaviruses. Human milk antibodies against SARS-CoV-2 may provide additional immune defense to infants and reduce the risk of COVID-19 infection, but their neutralizing capacity remains to be investigated.
There are a few limitations in this study. First, the neutralizing capacity of HM antibodies against SARS-CoV-2 was not determined. Second, this study had a small sample size, but our sample size proved to be adequately powered based on the results.