HCoV-NL63 and SARS-CoV-2 Share Recognized Epitopes by the Humoral Response in Sera of People Collected Pre- and during CoV-2 Pandemic

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) can cause serious illness in older adults and people with chronic underlying medical conditions; however, children and young people are often asymptomatic or with mild symptoms. We evaluated the presence of specific antibodies (Abs) response against Human coronavirus NL63 (HCoV-NL63) S protein epitopes (NL63-RBM1, NL63-RBM2_1, NL63-RBM2_2, NL63-RBM3, NL63-SPIKE541–554, and NL63-DISC-like) and SARS-CoV-2 epitopes (COV2-SPIKE421–434 and COV2-SPIKE742–759) in plasma samples of pre-pandemic, mid-pandemic, and COVID-19 cohorts by indirect ELISA. Moreover, a competitive assay was performed to check for cross reactivity response between COV2-SPIKE421–434 and NL63-RBM3 among patients with a definitive diagnosis of SARS-CoV-2. Immune reaction against all SARS-CoV-2 and HCoV-NL63 epitopes showed a significantly higher response in pre-pandemic patients compared to mid-pandemic patients. The results indicate that probably antibodies against HCoV-NL63 may be able to cross react with SARS-CoV-2 epitopes and the higher incidence in pre-pandemic was probably due to the timing of collection when a high incidence of HCoV-NL63 is reported. In addition, the competitive assay showed cross-reactivity between antibodies directed against COV2-SPIKE421–434 and NL63-RBM3 peptides. Pre-existing HCoV-NL63 antibody response cross reacting with SARS-CoV-2 has been detected in both pre- and mid-pandemic individual, suggesting that previous exposure to HCoV-NL63 epitopes may produce antibodies which could confer a protective immunity against SARS-CoV-2 and probably reduce the severity of the disease.

The novel coronavirus disease 2019 (COVID19) was first reported in Wuhan, China, in December 2019, and declared a worldwide pandemic on 11 March 2020 (https://www.ecdc.europa.eu/en/novelcoronavirus/event-background-2019). As of 12 December 2020, Johns Hopkins, University of Medicine, has reported 71,081,574 confirmed cases and 1,594,777 death cases worldwide [4]. Older adults and people with chronic underlying medical conditions are at higher risk for serious illness from SARS-CoV-2; however, children represent a small portion of patients whose clinical symptoms are usually milder than adults [5]. There has been multiple evidence about reduced susceptibility in children, probably due to cross-protection from previous CoVs infections [6,7].
SARS-CoV-2, as well as HCoV-NL63, interact with host cell by the envelope protein spike (S) required for the binding with angiotensin converting enzyme 2 (ACE2) receptor through Receptor Binding Domain (RBD) and for membrane fusion [8][9][10], these first steps point out to be crucial for the beginning of the infection [11].
RBD contains Receptor-Binding Motifs (RBMs) capable of interacting with ACE2 [12] and prompting the synthesis of neutralizing antibodies thanks to the presence of epitopes with immunogenic properties [13]. In regard to what mentioned above, the infusion of convalescent plasma (CP) from recovered COVID19 subjects into ongoing diseased patients seems to protect from the severity of the infection [14].
Different studies evidenced a certain degree of antigenic cross-protection for human CoVs belonging to the same genetic group, such as the protection offered against HcoV-229E by HcoV-NL63 seroconversion (both alpha-coronaviruses) or the protection against HcoV-HKU1 reinfection related to HcoV-OC43 seroconversion [15]. Seroprevalence surveys pinpoint that children are seropositive to HcoV-NL63 by six years old [16] and that reinfection by the same virus occurs throughout life, keeping in mind that nearly 20% of clinical cold can be accounted to four different genotypes of CoVs. Given the lack of symptoms in the majority of pediatric patients during SARS-CoV-2 infection and the capability of both SARS-CoV-2 and HcoV-NL63 S-protein to bind ACE2, we pondered the possibility that antibodies against HcoV-NL63 may cross react and offer protection against SARS-CoV-2 infection. The Receptor Binding Domain (RBD), within S, is the main target for neutralizing antibodies and was selected as the ideal vaccination subunit against SARS-CoV-2 and previous SARSs. The presence of antibodies against these specific epitopes may correlate with disease outcome and eventually with vaccine efficacy. Different monoclonal antibodies (mAbs) against S protein have been reported [17]. None of the mAbs that were reported bound to prefusion ectodomain trimers of the HcoV-OC43 or MERS-CoV S glycoproteins, which indicated a lack of cross-reactivity outside the Sarbecovirus subgenus.
With the aim to identify cross-reacting antibodies recognizing different coronaviruses, we evaluated the antibody response against different epitopes of SARS-CoV-2 and HcoV-NL63 S-protein in a pre-pandemic, a mid-pandemic, and a SARS-CoV-2 positive cohort.

Subjects and Blood Collection
Study population was formed by a pre-pandemic COVID19 cohort of 160 patients (109 females and 51 males; mean age ± SD, 39.63 ± 22.85 years), a mid-pandemic COVID19 cohort of 141 patients (105 females and 36 males; mean age ± SD, 55.55 ± 16.99). Both cohorts were recruited from the Endocrinology Department of University Hospital, Sassari; the pre-pandemic group was enrolled during the winter and early spring of 2017 and 2018, while the mid-pandemic one from May to August 2020. Peripheral venous blood samples were collected in K2-EDTA tube test and separated plasma was tested for the presence of antibodies against SARS-CoV-2 and HCoV-NL63 Spike protein-derived peptides. A group of 46 patients with a definitive diagnosis of SARS-CoV-2 determined by swab test viral PCR was tested against SARS-CoV2 epitopes; four of them (1 female and 3 males; mean age ± SD, 72.5 ± 4.36 years) were admitted to the intensive care unit (ICU) and forty-two (25 females and 17 males; mean age ± SD, 56.95 ± 17.46 years) were hospitalized in the SARS-CoV-2 wards, or other clinical departments. The investigation of plasma was approved by the ethical committee of ASL 1 Sassari (2149/CE). (Table 1)
The obtained absorbance values (at 405 nm) were normalized to a highly positive control serum with absorbance reactivity set at 1.0 arbitrary units (AU/mL).

Rapid Test
nCoV-K003: SARS-CoV-2-IgM-IgG combined rapid test kit (Bio Bench) was used to confirm the presence of IgG and IgM antibodies against SARS-CoV-2 in selected pre-pandemic, co pandemic sample and COVID-19 patients. 10 uL of plasma sample diluted in 60 µL of PBS (pH 7.2) was incubated for 15 min before the reading the result.

Competitive ELISA
Plasma from 11 COVID19 patients, with strong and medium antibody response against the COV2-SPIKE 421-434 peptide, were incubated on PBS-T (1:100) over night at 4 • C with NL63-RBM3 at concentrations of 10 µM to perform the competitive assay (cELISA).
Treated COVID19 plasma was then subjected to ELISA on plates coated with COV2-SPIKE 421-434 . In the same assay, peptide without the plasma was used as a negative control, COV2-SPIKE 421-434 positive plasma was used as positive control.

Statistical Analysis
All data were analyzed using GraphPad Prism 8.2.0 software (GraphPad Software, San Diego, CA, USA). Mann-Whitney U and Kruskal-Wallis tests were used to analyze non-parametric data and compare differences between two or three groups, respectively. Differences with a p-value < 0.05 were considered statistically significant. Receiver-operating characteristic (ROC) was used to choose the cut-off value to assess the sample positivity, which was consequently tested through Fisher's exact test. Spearman Correlation tests were performed between HCoV-NL63 and SARS-CoV2-derived peptides.

Results
Humoral response against five selected epitopes of HCoV-NL63 and two epitopes of SARS-CoV-2 were evaluated in plasma of pre-pandemic, mid-pandemic, and SARS-CoV-2 positive groups.
Antibody response against NL63-RBM1, showed a statistical difference between pre-pandemic and mid-pandemic populations (p = 0.0055) analyzed by Mann-Whitney U test. A cut-off values were chosen by ROC analysis and percentage of positive and negative samples were determined by Fisher's exact test. 63.1% of pre-pandemic population (101 patients out of 160) and 48.9% in mid-pandemic population (69 patients out of 141) were seropositive (p = 0.01, Fisher's exact test), with a cut-off value of 0.100 and AUC = 0.593 ( Figure 1A).

Microorganisms 2020, 8, x FOR PEER REVIEW 6 of 15
Correlations analysis were performed on the basis of OD values obtained from evaluation of humoral response against HCoV-NL63 and SARS-CoV-2 epitopes. Strikingly, in pre-pandemic COVID19 group, the antibody response against SARS-CoV-2 epitopes revealed a positive correlation with the antibody response to NL63-derived peptides.
In order to study the age-variability of the pre-pandemic population, we performed a correlation analysis between HCoV-NL63 and SARS-CoV-2 epitopes, calculated using Spearman correlation test. The heatmaps (Figure 4) show r's value between pairs of epitopes of HCoV-NL63 and SARS-CoV-2 in pre-pandemic population divided by age groups, as shown in Table 3.
In order to study the age-variability of the pre-pandemic population, we performed a correlation analysis between HCoV-NL63 and SARS-CoV-2 epitopes, calculated using Spearman correlation test. The heatmaps (Figure 4) show r's value between pairs of epitopes of HCoV-NL63 and SARS-CoV-2 in pre-pandemic population divided by age groups, as shown in Table 3.  Table 3.   Table 3. Table 3. Details of pre-pandemic population divided by age groups (years old, y/o).

Age Groups
Number In the 11-20 years group ( Figure 4B) a non-significant p-value has been found for the correlation between NL63-RBM2_1 and NL63-RBM2_2, NL63-SPIKE 541-554 and the epitopes NL63-RBM1, NL63-RBM2_1, NL63-RBM2_2 and NL63-RBM3; the correlations were no statistically significant also between NL63-DISC-like and NL63- The serological presence of antibodies against COV2-SPIKE 742-759 and COV2-SPIKE 421-434 was examined in a group of 46 patients with a diagnosis of SARS-CoV-2, a Kruskal-Wallis test was performed to compare COVID19 group with the pre-pandemic and the mid-pandemic cohorts. A statistically significant difference has been observed between pre-pandemic and mid-pandemic populations (p < 0.0001) and between mid-pandemic and COVID19 populations (p < 0.0001). No significant difference has been found in COVID19 group compared with the pre-pandemic population ( Figure 5). The serological presence of antibodies against COV2-SPIKE742-759 and COV2-SPIKE421-434 was examined in a group of 46 patients with a diagnosis of SARS-CoV-2, a Kruskal-Wallis test was performed to compare COVID19 group with the pre-pandemic and the mid-pandemic cohorts. A statistically significant difference has been observed between pre-pandemic and mid-pandemic populations (p < 0.0001) and between mid-pandemic and COVID19 populations (p < 0.0001). No significant difference has been found in COVID19 group compared with the pre-pandemic population ( Figure 5).  From these results, 11 COVID19 patients were selected based on the plasma antibody response against the COV2-SPIKE 421-434 peptide (five strong and six medium responders). All the strongly responsive patients showed outstanding antibody cross-reactivity with a noticeable reduction of the antibody reaction after the incubation with NL63-RBM3 peptide.
From these results, 11 COVID19 patients were selected based on the plasma antibody response against the COV2-SPIKE421-434 peptide (five strong and six medium responders). All the strongly responsive patients showed outstanding antibody cross-reactivity with a noticeable reduction of the antibody reaction after the incubation with NL63-RBM3 peptide.

Discussion
Cross-reactivity between HCoVs has long been hypothesized to offer transient protection against infection with different CoVs [7,18]. In particular, Sagar et al. [19] have reported less severe coronavirus disease 2019 illness in patients with a previously detected endemic human CoVs (eCoV) infections; they suggested that pre-existing immune responses against eCoVs can alleviate disease manifestations from SARS-CoV-2 infection.
Three known antigenic groups of CoVs are associated with diseases in animals and humans. The known human CoVs (HCoVs) are HCoV-229E and HCoV-NL63 belonging to the alpha group and HCoV-OC43 within the beta group, are generally recognized to cause mild upper respiratory tract diseases and, rarely, lower respiratory tract diseases. However, HCoV-NL63 is the only CoV responsible of mild infections that use the ACE2 receptor as target for its S protein [20].
HCoV-NL63 S-protein has been well characterized. Although HCoV-NL63 and SARS-CoV have no structural homology in RBD cores or RBM, the two viruses recognize common ACE2 regions, largely because of a "virus-binding hotspot" on ACE2. Using the IEDB site we tried to identify possible common "binding hotspot" on ACE2 by analyzing the S-proteins on both CoVs. Moreover, S-protein structural and phylogenetic analysis revealed low homology in the aminoacidic sequence in SARS-CoV and SARS-CoV-2, on the other hand both S-protein RBDs manifested the same ACE2binding mode [21].
With regard to what expressed above, we wondered if a serologic cross-reactivity may be found between HCoV-NL63 and SARS-CoV-2. To date, few studies investigated antibody cross-reaction between alpha and beta-coronaviruses.
Recently, Loos et al. [22] reported some positive relationships between the IgGs response to the common CoVs and SARS-CoV-2 RBD-specific immunity, most relationships among the SARS-CoV-

Discussion
Cross-reactivity between HCoVs has long been hypothesized to offer transient protection against infection with different CoVs [7,18]. In particular, Sagar et al. [19] have reported less severe coronavirus disease 2019 illness in patients with a previously detected endemic human CoVs (eCoV) infections; they suggested that pre-existing immune responses against eCoVs can alleviate disease manifestations from SARS-CoV-2 infection.
Three known antigenic groups of CoVs are associated with diseases in animals and humans. The known human CoVs (HCoVs) are HCoV-229E and HCoV-NL63 belonging to the alpha group and HCoV-OC43 within the beta group, are generally recognized to cause mild upper respiratory tract diseases and, rarely, lower respiratory tract diseases. However, HCoV-NL63 is the only CoV responsible of mild infections that use the ACE2 receptor as target for its S protein [20].
HCoV-NL63 S-protein has been well characterized. Although HCoV-NL63 and SARS-CoV have no structural homology in RBD cores or RBM, the two viruses recognize common ACE2 regions, largely because of a "virus-binding hotspot" on ACE2. Using the IEDB site we tried to identify possible common "binding hotspot" on ACE2 by analyzing the S-proteins on both CoVs. Moreover, S-protein structural and phylogenetic analysis revealed low homology in the aminoacidic sequence in SARS-CoV and SARS-CoV-2, on the other hand both S-protein RBDs manifested the same ACE2-binding mode [21].
With regard to what expressed above, we wondered if a serologic cross-reactivity may be found between HCoV-NL63 and SARS-CoV-2. To date, few studies investigated antibody cross-reaction between alpha and beta-coronaviruses.
Recently, Loos et al. [22] reported some positive relationships between the IgGs response to the common CoVs and SARS-CoV-2 RBD-specific immunity, most relationships among the SARS-CoV-2 response and common CoVs, influenza virus, and Respiratory Syncytial Virus (RSV) were largely driven by individuals exhibiting low to undetectable titers. They suggest that preexisting CoV RBD-specific and other pathogen immunity plays a limited role in shaping SARS-CoV-2 humoral immune responses. However, they used the HCoV-NL63 RBD (accession no. AKT07952, aa residues 481 to 616) without any immunogenic analysis of selected highly immunogenic B cells epitopes as we performed.
Mateus et al. [25] showed three cases in which HCoV analogs were better antigens than the SARS-CoV-2 peptide, suggesting that they may be the cognate immunogen, one of them belonging to HCoV-NL63.
These findings of cross-reactive HCoV T-cell specificities are in stark contrast to HCoV-neutralizing antibodies, which are HCoV species specific and did not show cross-reactivity against SARS-CoV-2 RBD as reported by Premkumar et al. [26]. Here, archived human samples collected before SARS-CoV-2 (20 American adults and samples from individuals in South Asia, the Caribbean, and Central America) were tested for binding against RBD spike antigens from HCoVα (NL63) and HCoVβ (HKU1) with no reaction [27]. Whereas Sotgia F. and Lisanti MP. argued that mild pathogenic coronaviruses epitopes can be used for a vaccine [26].
Of note, Ng et al. (Science, 2020) have identified several epitopes that were recognized with cross-reactive antibodies by uninfected patients proving the presence of preexisting antibodies recognizing SARS-CoV-2 [28]. These results considering also the finding of preexisting T cell memory against seasonal HCoVs and SARS-CoV-2 (21) may shed light on the SARS-CoV-2 natural infection having in mind that this protective cross immunity does not last for long time and probably is not sterilizing as well.
In our study, we first assessed the antibody response against five peptides of HCoV-NL63 and for two epitopes of SARS-CoV-2 from S-protein in a pre-pandemic and a mid-pandemic group. The pre-pandemic population shows good chances to have been recent infected with eCoVs, including NL63, as it is composed for about one-third of patients under 18 years old, that are the most exposed people [29]. The decision to use short peptides, instead of expressed protein, represents a limitation of the study, but it can also a resource showing epitopes that could be previously hidden.
The pre-pandemic population exhibited a strong humoral response against HCoV-NL63 compared to that observed in the mid-pandemic population for NL63-RBM1 and NL63-DISC-like.
At the same time, we observed an unexpected and strong humoral response in the pre-pandemic population towards SARS-CoV-2 S-protein epitopes. Of course, this response cannot be attributed to SARS-CoV-2 since the samples from the pre-pandemic population were collected before 2019. These data were confirmed by the rapid test we performed on one of the most reactive pre-pandemic sample for SARS-CoV-2 epitopes, obtaining a positive result. Considering that 72.5% and 68.7% of the pre-pandemic population are seropositive against COV2-SPIKE 421-434 and COV2-SPIKE 742-759 , respectively, we think a heterologous cross-reactivity between HCoV-NL63 antibodies against SARS-CoV-2 epitopes.
The correlations found between HCoV-NL63 and SARS-CoV-2 peptides in both cohorts corroborate the hypothesis of an antibody cross-reactivity between the two virus species and the possibility that HCoV-NL63 may offer a certain degree of protection against SARS-CoV-2. Further, we have looked at the correlations among the epitopes used in this study in pre-pandemic population divided by age classes (Figure 4); the finest correlation can be observed in 5-10 years group and in 40-60 years group. Of course, to better understand this correlation more in depth-studies are necessary but one of the assumptions can be that the antibodies against NL63 may help for a better manage of the SARS-CoV-2 infection.
To confirm these data, the results from a cELISA, performed on patients with a positive diagnosis to SARS-CoV-2, highlight a significant difference in antibody response against SARS-CoV-2 epitopes between pre-pandemic and mid-pandemic cohorts and between COVID19 and mid-pandemic groups. Noteworthy, HCoV-NL63 infections were observed primarily in the winter season, period when our pre-pandemic samples were collected with a 30.6% of children under 18 years old, that represent the most HCoV-NL63 affected people [29]; this can explain the similar antibody response in pre-pandemic and COVID-19 groups. Related to this, Canducci et al. demonstrated that on 322 infants suffering from acute respiratory disease, the 8.7% of the cases examined were caused by coronaviruses, with HCoV-NL63 accounting for 21.4% of the latter [30].
This support our initial hypothesis that antibodies directed against HCoV-NL63 S-protein may recognize SARS-CoV-2 S-protein epitopes.
These results are particularly relevant in both strongly and medium responsive patients (Figure 6), we observed a significant reduction of the antibodies against COV2-SPIKE 421-434, except for two patients (COVID19#8 and COVID19#10) who showed a reduction inferior to 40%. Considered the increased seroprevalence of HCoV-NL63 infections and the low incidence of SARS-CoV-2 in children, it is possible that antibodies directed against HCoV-NL63 may have a protective effect in some individuals affected by SARS-CoV-2.
A richer interaction energies of the human ACE2 molecular recognition by SARS-CoV-2 compared to SARS-CoV and HCoV-NL63 was recently reported [31], explaining why SARS-CoV-2 has an enhanced ability for pathogenicity. An other study [32] identified the conserved residues in the binding interface of S proteins in all three strains. The systematic point mutations show that these conserved residues in the respective strains are evolutionarily favored at their respective positions. Notwithstanding, further experiments, larger cohorts and different time collection, are necessary to establish the length of the protection and to assess the nature of the antibodies and their possible neutralizing effect against SARS-CoV-2, but also to investigate the possible contribution of pre-existing immunity offered from other coronaviruses, such as HCoV-OC43, mediated by different proteins than Spike.