Seroprevalence of endemic and emergent coronaviruses among SARS-COV-2 patients and healthcare workers in Abidjan, Côte d’Ivoire

Abstract

Background

Understanding the serological landscape of endemic and emergent coronaviruses is critical to interpreting early-pandemic immune responses and evaluating hypotheses of cross-reactivity. It was proposed that prior exposure to endemic coronaviruses could affect susceptibility or shape symptom severity through cross-reactive antibody responses. However, little was known about baseline coronavirus seroprevalence in many global regions, including Abidjan, Côte d’Ivoire. Characterizing this landscape provides key insights into early pandemic immunity and the potential influence of prior coronavirus exposures on SARS-CoV-2 immune response.

Methods

Here, we probe this using data from syndromic surveillance in Abidjan, Côte d’Ivoire, collected between September 2020 and July 2021. We quantified IgG antibody levels to both spike and nucleocapsid proteins for emergent coronaviruses (SARS-CoV-1, SARS-CoV-2, and MERS-CoV) and endemic coronaviruses (HKU1, OC43, NL63 and 229E) using high-throughput multiplex bead assay. Samples were collected from SARS-CoV-2 negative healthcare workers (N = 202) and SARS-CoV-2 positive patients (N = 207). SARS-CoV-2 positive patients returned for repeat sampling at day 28 (N = 131).

Results

SARS-CoV-2 negative healthcare workers had higher SARS-CoV-1 seropositivity [0.27 (CI: 0.21–0.33) vs. 0.18 (CI: 0.13–0.24)] and SARS-CoV-2 seropositivity [0.52 (CI: 0.46–0.59) vs. 0.37 (CI: 0.31–0.44)] than SARS-CoV-2 positive patients. There were no significant differences among endemic coronaviruses between the healthcare workers and patients. Among the endemic coronaviruses, seropositivity was highest for 229E at 0.96 (95% CI: 0.94–0.98) and lowest for HKU1 at 0.56 (95% CI: 0.51–0.61) We found no significant sex difference in seropositivity to any coronavirus.

Conclusions

These findings provide a snapshot of endemic and emergent coronavirus seroprevalences during the beginning of the COVID-19 pandemic in Abidjan. We observed high seroprevalence to endemic alphacoronaviruses (229E and NL63), slightly lower levels for betacoronaviruses (HKU1 and OC43), and cross-reactive antibody signals to SARS-CoV-1. Among SARS-CoV-2 positive patients sampled again after 28 days, we did not observe evidence of boosting antibody levels to endemic coronaviruses, suggesting no cross-reactive responses. However, studies incorporating conserved S2 regions are needed to more fully assess cross-reactivity.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-13133-9.

Background

The emergence of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) prompted broad interest in characterizing the landscape of immunity around coronaviruses, both in considering whether previous exposure with a coronavirus was protective against SARS-CoV-2 infection or severity of symptoms, as well as the magnitude of individual heterogeneity in immune responses to SARS-CoV-2 [1]. Endemic coronaviruses, which include alpha coronaviruses (HCoV-229E and HCoV-NL63) and beta coronaviruses (HCoV-OC43 and HCoV-HKU1), are known to circulate seasonally with a majority of the world’s population estimated to have been exposed to at least one strain [2–4]. Prior to the SARS-CoV-2 pandemic, it was known that immunity to infection by endemic coronaviruses is short lived, with reinfection possible within 12 months [5, 6].

Partial genome sequence identity across viral lineages, especially in the immunogenic receptor binding domain of the spike protein, means that antibody responses generated to one coronavirus may also respond to others. Among endemic coronaviruses, cross-reactivity has been shown within alpha coronaviruses (229E and NL63) and beta coronaviruses (OC43 and HKU1) but not between them [5, 7–9]. The beta coronaviruses also include emergent and pathogenic coronaviruses (SARS-CoV-1, SARS-CoV-2 and Middle East Respiratory Syndrome Coronavirus (MERS-CoV)). Studies examining cross-reactivity with SARS-CoV-1 following endemic coronavirus infection have found mixed results, with some studies showing little cross-reactivity [10–14]. In addition, SARS-CoV-1 infection has been shown to induce cross-reactive antibodies with both endemic and emergent coronaviruses [10, 11, 15, 16], suggesting that these relationships might be asymmetric. Aligning with this, prior infection with SARS-CoV-2 is associated with lower incidence of endemic coronaviruses [17]. Due to this cross-reactivity, antibody detection for a given coronavirus may reflect past or ongoing infection with the target virus, or past or ongoing infection with antigenically related coronaviruses capable of eliciting cross-reactive immune responses. Consequently, observed seropositivity may represent a mixture of direct and cross-reactive antibody responses rather than a definitive history of infection with a single virus [18].

Despite a vast body of work on coronaviruses in the wake of the pandemic, the role of cross-protection from previously circulating coronavirus strains in preventing infection or hospitalizations during the SARS-CoV-2 pandemic remains unresolved. For example, pre-pandemic sera have been found to both lack [19] or contain [20–22] SARS-CoV-2 reactive antibodies. Such pre-existing cross-reactivity has been found to be protective against SARS-CoV-2 [23], have no benefit [20] or even be harmful [24–28].

In addition to questions surrounding the presence and effectiveness of cross-immunity, questions were also raised about the low reported mortality of SARS-CoV-2 infection in some areas, particularly Africa (though these observations are likely subject to bias in underestimating mortality in some African countries [29]). One of the proposed hypotheses for variation in SARS-CoV-2 outcomes was the existence of cross-immunity from coronaviruses circulating in Africa prior to the SARS-CoV-2 outbreak [30]. In Gabon and Senegal, for example, cross-reactive antibodies against SARS-CoV-2 were more prevalent than in samples from Canada, Brazil and Denmark but ultimately failed to neutralize the virus [31]. Another study found that SARS-CoV-2 cross-reactivity was significantly higher in samples from sub-Saharan Africa than samples from the United States [32]. However, few studies have characterized seroprevalence of both endemic and emerging coronaviruses in West African populations, particularly in the early stages of the pandemic.

To address this gap, we conducted a multiplex serological survey of coronavirus immunity in Abidjan, Côte d’Ivoire with samples collected in 2020 and 2021, a period where Alpha (B.1.1.7) and Eta (B.1.525) became prevalent in Côte d’Ivoire to describe antibody profiles to both endemic and emergent coronaviruses [33]. We used samples obtained through syndromic surveillance in Abidjan, Côte d’Ivoire, which we define as the recruitment of individuals based on clinical presentation and symptom profiles prior to laboratory confirmation. By analyzing samples collected early in the pandemic, prior to widespread vaccine rollout, our aim was to provide a baseline snapshot of coronavirus seroprevalence and explore how SARS-CoV-2 infection may shape antibody responses to related viruses. Our analysis focuses on SARS‑CoV‑2 positive patients and SARS‑CoV‑2 negative healthcare workers. We compared seropositivity across seven human coronaviruses using a multiplex assay that included S1 and N antigens, and did not capture antibodies targeting the conserved S2 subunit, which may limit the detection of broader cross-reactive responses.

Methods

Study population

Patients with SARS-CoV-2 like symptoms and healthcare workers were sampled across various locations in Abidjan from September 2020 to July 2021 (N = 207) (Fig. 1 and Table S1). Patients with SARS-CoV-2 symptoms were sampled at the Treichville Hospital, Yopougon Hospital, and the Institut Pasteur Infectious Diseases Department. Healthcare workers were sampled at the Treichville Hospital, Cocody Hospital, Institute Pasteur of Côte d’Ivoire and the National Institute of Public Health and Hygiene from September 2020 to July 2021 (N = 202). We included individuals who resided in Abidjan at least 6 months prior to the start of the pandemic and excluded individuals with a known autoimmune pathology, individuals who were pregnant and individuals who had a symptom onset date more than 6 days prior to sampling. All individuals provided consent to participate.

Fig. 1.

Sampling sites and period. (A) Map depicting Abidjan, Côte d’Ivoire with study sites labeled A-E. (B) Weekly confirmed COVID-19 cases per million people in Côte d’Ivoire during study period, with arrows indicating sampling periods for each of the two study populations: (1) SARS-COV-2 negative healthcare workers and (2) SARS-CoV-2 positive patients. For each population, the sampled sites and total sample sizes are indicated. See Table S1 for further breakdown. Data source for weekly confirmed COVID-19 cases: Our World in Data, World Health Organization (2025); Population based on various sources (2024)

Data

All individuals provided a blood sample at day 0 (D0) and a nasopharyngeal sample. Data recorded included age, sex, date of sampling, sampling site, and commune of residence. We conducted RT-PCR on the nasopharyngeal samples to detect current SARS-CoV-2 infection with the Qiagen QIAamp viral RNA kit (Qiagen, Hilden, Germany). Individuals with SARS-CoV-2 symptoms and a positive RT-PCR were asked to return for an additional blood sample at day 28 (D28) (Table S2). Individuals with SARS-CoV-2 symptoms and a negative RT-PCR were excluded from this study. All sampled healthcare workers had a negative RT-PCR. Thus, we established two groups, SARS-CoV-2 negative healthcare workers (N = 202) and SARS-CoV-2 positive patients (N = 207) (Fig. 1). All blood samples were tested for antibodies to both spike and nucleocapsid proteins for three recently emergent coronaviruses, SARS-CoV-1, SARS-CoV-2, and MERS-CoV, and four endemic coronaviruses, 229E, HKU1, OC43, and NL63. Samples were tested using a high throughput multiplex assay to detect IgG antibodies with sensitivity and specificity > 99% for SARS-CoV2 [18] which has been used previously in serosurveys conducted in Africa [34–36].

We used commercially available recombinant nucleocapsid and spike proteins (Sinobiologicals, Interchim, Montluçon, France). The proteins were purchased as lyophilized powders and resuspended in buffer following the manufacturer’s instructions. Recombinant nucleocapsid (2 µg/1.25 × 106 beads) and spike proteins (1 µg/1.25 × 106 beads) were covalently coupled on carboxyl functionalized fluorescent magnetic beads (Luminex Corp., Austin, TX) with the BioPlex amine coupling kit (Bio-Rad Laboratories, Marnes-la-Coquette, France) following the manufacturer’s instructions.

Diluted samples (1:200) were incubated with coupled beads for 16 h at 4 °C. Reactions were revealed following incubation with a biotin-labeled anti-human IgG and streptavidin-R-phycoerythrin conjugate. Antigen-antibody reactions were then read on BioPlex-200 equipment (Bio-Rad, Marnes-la-Coquette. France) with results expressed as median fluorescence intensity (MFI) per 100 beads.

Cut-offs for SARS-CoV-1, SARS-CoV-2, and MERS-CoV antigens were established using receiver operating characteristic (ROC) curve analysis performed in GraphPad Prism 8 (San Diego, CA, USA). The assay was also expanded to include eight additional antigens: nucleoprotein (N) and spike (S) proteins from the four endemic human coronaviruses (HCoV-NL63, HCoV-OC43, HCoV-HKU1, and HCoV-229E). These antigens were sourced from Sinobiological and Cusabio and processed identically to the pathogenic coronavirus proteins.

To validate these additional targets and determine seropositivity thresholds, we analyzed 2,568 samples from Guinea. This panel included 1,740 pandemic-era samples (collected in 2021 as part of the ARIACOV project) and 828 pre-pandemic samples (collected in 2018 as part of a demographic and health survey) [37]. Cut-offs were determined using a mixture distribution modeling and bootstrapping approach, as described previously [38]. The resulting MFI thresholds for each antigen (Spike and Nucleoprotein) were applied across the study and are listed in Table S3. It is important to note that the multiplex assay only included S1 subunit spike proteins and no S2 antigens. As a result, our analysis does not capture S2-mediated cross-reactivity which may underestimate the presence of cross-reactive antibodies.

Analysis

First, we classified individuals sampled at D0 based on MFI values to both spike and nucleocapsid proteins as either seropositive, seronegative, or inconclusive based on thresholds provided in Table S3. Antibody titers, levels, and MFI are used interchangeably in this study. For MERS-CoV, SARS-CoV-1, and SARS-CoV-2, the threshold of seropositivity was 1000 MFI for spike specific levels and 500 MFI for the nucleocapsid specific levels. For HKU1, OC43, NL63, and 229E the threshold of seropositivity was 1300 MFI for both spike protein specific levels and nucleocapsid specific levels. Individuals above the threshold of seropositivity for both spike and nucleocapsid specific levels were classified as seropositive. Individuals below the threshold for both spike and nucleocapsid specific levels were classified as seronegative while individuals above the threshold for one protein specific level and below the threshold for the other protein specific level were classified as inconclusive. Given known cross-reactivity among coronaviruses, seropositivity in this study may reflect a combination of antibodies induced by past or ongoing infection with the target virus, or by past or ongoing infection with antigenically related coronaviruses capable of generating cross-reactive responses. Vaccination against SARS-CoV-2 may also contribute to observed antibody titers. However, since vaccination in Côte d’Ivoire began in March 2021 and the first mass campaign was not conducted until December 2021, we expect very limited vaccine-related antibody responses among participants sampled during our study period (September 2020–July 2021).

We measured seroprevalence by dividing the total number of seropositive individuals by the total sample size across all coronaviruses, sampling periods and sampling groups. Inconclusive individuals were retained in the denominator for all seroprevalence estimates to avoid upward bias in seroprevalence estimates. Then we compared seroprevalence across all coronaviruses, groups and sex. To assess whether seroprevalence significantly differed between SARS-CoV-2–positive patients and SARS-CoV-2–negative healthcare workers for each virus, we conducted Fisher’s exact tests.

Next, we used Rsero, an R package designed to analyze serological survey data, to assess seroprevalence across age groups for all individuals with information on age at the time of sampling [39]. For this age-specific analysis, seroprevalence estimates for ages 0 to 9 were excluded due to low sample size. 95% confidence intervals were calculated for all seroprevalence estimates.

We then analyzed antibody levels quantitatively by looking at how both spike- and nucleocapsid-specific antibody levels varied by age at the time of sampling for all individuals sampled at D0 where age information was available. We used the mgcv package to fit a generalized additive model smoothed across age for each group [40]. Lastly, we assessed how antibody levels for all coronaviruses changed from D0 to D28 among SARS-CoV-2 positive patients by calculating the Log2 fold change of MFI levels for the 131 individuals who returned at D28 for a subsequent blood test.

Results

A total of 409 individuals were initially sampled across 5 sampling sites in Abidjan upon their first visit (D0) (Table S1). 207 individuals were patients with SARS-CoV-2 symptoms and positive RT-PCR results and 202 participants were healthcare workers with negative RT-PCR results. 131 patients with SARS-CoV-2 symptoms and positive RT-PCR results returned for subsequent sampling 28 days later (Table S2).

First, we classified each individual’s serostatus based on MFI levels to both spike and nucleocapsid (Figs. S1-S2) and measured seroprevalence to each coronavirus across sampling periods and groups (Figs. S3-S4). Among SARS-CoV-2 positive patients, no changes in seroprevalence were seen across the sampling period (Fig. S3). There was an increase in SARS-CoV-2 seroprevalence among sampled SARS-CoV-2 negative healthcare workers across the sampling period from 0.49 (95% CI: 0.37–0.61) post wave 1 to 0.76 (95% CI: 0.63–0.88) post wave 2 (Fig. S4).

When combining across all sampling periods and comparing among sampling groups, we found that SARS-CoV-1 and SARS-CoV-2 seroprevalence was highest among SARS-CoV-2 negative healthcare workers (Fig. 2). SARS-CoV-1 seroprevalence was 0.27 (95% CI: 0.21–0.33) among SARS-CoV-2 negative healthcare workers and 0.18 (95% CI: 0.13–0.24) among SARS-CoV-2 positive patients (Fig. 2). SARS-CoV-2 seroprevalence was 0.52 (95% CI: 0.46–0.59) among SARS-CoV-2 negative healthcare workers and 0.37 (95% CI: 0.31–0.44) among SARS-CoV-2 positive patients (Fig. 2). There were no significant differences among endemic coronaviruses between the healthcare workers and patients (Fig. 2). Endemic alphacoronavirus seroprevalence for SARS-CoV-2 negative healthcare workers was 0.84 (95% CI: 0.79–0.89) for NL63 and 0.95 (95% CI: 0.92–0.98) for 229E. Endemic alphacoronavirus seroprevalence for SARS-CoV-2 positive patients was 0.80 (95% CI: 0.74–0.85) for NL63 and 0.96 (95% CI: 0.94–0.99) for 229E. Endemic betacoronavirus seroprevalence for SARS-CoV-2 negative healthcare workers was 0.56 (95% CI: 0.49–0.63) for HKU1 and 0.75 (95% CI: 0.69–0.81) for OC43. Endemic betacoronavirus seroprevalence for SARS-CoV-2 positive patients was 0.57 (95% CI: 0.50–0.63) for HKU1 and 0.82 (95% CI: 0.76–0.87) for OC43. We observed no sex differences in seroprevalence across both sampling groups (Fig. S5).

Fig. 2.

Emergent and endemic coronavirus seroprevalence in Abidjan, Ivory Coast (N = 409). Bars display the proportion of individuals who were above the threshold of seropositivity for both spike and nucleocapsid protein specific antibodies. 95% confidence intervals are displayed by the red bars. Group classification on the x axis indicates whether an individual was a SARS-Cov-2 negative healthcare worker (N = 202) or whether they were a SARS-CoV-2 positive patient (N = 207). For each virus, differences in seroprevalence between groups were assessed using Fisher’s exact test. Asterisks indicate statistically significant differences between groups (p < 0.05 (*), p < 0.01 (**)); non‑significant comparisons are labeled “ns”

Next, we combined samples across all sampling periods and groups to assess seroprevalence for all sampled individuals (N = 409). As expected, we found higher seropositivity to endemic coronaviruses compared to the emergent coronaviruses. Among the endemic coronaviruses, seropositivity was highest for 229E at 0.96 (95% CI: 0.94–0.98) and lowest for HKU1 at 0.56 (95% CI: 0.51–0.61) (Fig. S6 and Table S4). This trend remained consistent when looking at age-specific seroprevalence, with 229E seroprevalence remaining high across sampled age groups (Fig. 3). Among the emergent coronaviruses, seropositivity was highest for SARS-CoV-2 at 0.45 (95% CI: 0.40–0.50) and lowest for MERS-CoV at 0.02 (95% CI: 0.005–0.03) (Fig. S6 and Table S4). SARS-CoV-2 seroprevalence remained at similar levels across age groups sampled (Fig. 3).

Fig. 3.

Age-specific seroprevalence of emergent and endemic coronavirus in Abidjan, Ivory Coast (N = 297) generated with Rsero package. Black points display mean seroprevalence while black bars display 95% confidence intervals. Age at sampling is displayed on the x axis. Individuals above the threshold of seropositivity for both spike and nucleocapsid proteins were classified as seropositive. Individuals without age information were excluded and all sampling groups (SARS-CoV-2 positive patients and SARS-CoV-2 negative healthcare workers) were combined for this analysis. Seroprevalence estimates for ages 0 to 9 were excluded due to low sample size

Lastly, we assessed quantitative antibody levels across age, by sampling group and among SARS-CoV-2 positive patients on D0 and D28. No notable trends emerged for MFI levels across age (Fig. S7). After 28 days, we saw an increase in antibodies to SARS-CoV-2 spike and nucleocapsid in addition to increases in SARS-CoV-1 nucleocapsid specific antibodies among SARS-CoV-2 positive patients (Figs. 4 and S8). There was a slight increase in NL63 nucleocapsid MFI levels seen among samples collected at the Treichville hospital (Figs. 4 and S8).

Fig. 4.

Changes in coronavirus antibody levels among SARS-CoV-2 patients (N = 131). SARS-CoV-2 positive patients provided a sample at day 0 and day 28. Changes in MFI levels, a measure of IgG antibodies, for each individual are shown by the black line connecting their day 0 and day 28 sample. Mean MFI levels are shown in red. The “S” following the coronavirus label indicates spike specific antibodies while the “N” indicates nucleocapsid specific antibodies

Discussion

We leveraged syndromic surveillance conducted in Abidjan, Côte d’Ivoire (September 2020-July 2021) among SARS-CoV-2 negative healthcare workers and SARS-CoV-2 positive patients to probe endemic and emergent coronavirus seroprevalence and how SARS-CoV-2 infection may shape antibody responses to other related coronaviruses. To our knowledge, this is the first study to map population-level seroprevalence to both endemic and emergent coronaviruses in Côte d’Ivoire. Seroprevalence for endemic coronaviruses among our participants was higher than emergent coronaviruses as expected since they are constantly circulating (Figs. 2 and S6). Since sampling was done early on during the SARS-CoV-2 pandemic, not everyone had been exposed to SARS-CoV-2 yet so we expect that present-day seroprevalence to SARS-CoV-2 has increased to similar seropositivity levels seen with the endemic coronaviruses.

Our estimates of SARS-CoV-2 seroprevalence in a sample from Abidjan, Côte d’Ivoire (37–52%) were coarsely similar to those observed from studies in other African localities with similar study design and timing (2020 to mid-2021). In Nigeria, a randomized serosurvey of the general population of two states in June 2021 found 40–43% of individuals were seropositive for SARS-CoV-2, while seroprevalence was already much higher in an urban area (Lagos) by early 2021 (72.4% of healthcare workers and 60.3% of the general population) [41, 42]. In urban residential areas of Ghana, seroprevalence estimates ranged from 43.9% in Kumasi, to 53.3% in Accra and 84.7% in Tamale [43]. Together these data indicate that our study locality, Abidjan, Côte d’Ivoire, was experiencing similar rates of SARS-CoV-2 infection in the early waves of the pandemic as other population centers in the region.

Among our sampled population, 229E seroprevalence was the highest at 96% (95% for healthcare workers and 97% for patients), while HKU1 was the lowest at 56% (56% for healthcare workers and 57% for patients) (Figs. 1 and S6 and Table S4). A 2013 study of children under five with respiratory illness in Côté d’Ivoire found that 229E was the most commonly detected human coronavirus responsible for 12.1% of positive cases [44]. Our findings differ from serological data from South Africa in 2020–2021, where 229E seroprevalence was lower at 30.6% with the most prevalent coronavirus seroprevalence being NL63 at 37.1% [45]. Among a sample of SARS-CoV-2 seropositive women in Dakar, Senegal, HKU1 seropositivity was most commonly observed (95.7%), followed by 229E (85.1%), and OC43 (70.2%) [46]. Prior to the SARS-CoV-2 pandemic, a majority of studies of the seroprevalence of endemic coronaviruses focused on children (≤ 5 years old) with acute respiratory infections, complicating comparison to studies of adult populations such as healthcare workers, but seroprevalence varied widely across the 10 African countries with published estimates [47]. These differences may reflect variation in circulation of endemic coronaviruses across Africa. Our study establishes a baseline serological map of coronavirus exposure in Abidjan, capturing early-pandemic patterns of immunity across both endemic and emergent coronaviruses.

We also detected antibody reactivity to SARS-CoV-1 (23%) and MERS-CoV (2%) among our participants. However, neither virus is known to have circulated in Côte d’Ivoire, and these findings are unlikely to reflect true past infections. MERS-CoV has never been identified in Côte d’Ivoire, and the primary reservoir host, camels, are absent in the study area, making local transmission implausible [48]. Similarly, SARS-CoV-1 circulation has not been documented in Côte d’Ivoire [49]. Therefore, these antibody signals are likely resulting from cross-reactive antibodies induced by other coronaviruses or background assay noise. For comparison to other settings, seroprevalence of SARS-CoV-1 was 0% in a review of human coronaviruses in Africa prior to the SARS-CoV-2 pandemic, while seroprevalence of MERS-CoV varied from low (0.18%) in livestock handlers in Kenya to high (> 50%) in some pilgrims returning to Sudan from Hajj [47]. The higher SARS-CoV-1 seropositivity observed in SARS-CoV-2 negative healthcare workers compared to SARS-CoV-2 positive patients likely reflects prior SARS-CoV-2 exposure via either natural infection or vaccination and resulting cross-reactivity. Overall, it is possible that some healthcare workers or patients had experienced prior SARS-CoV-2 infections earlier in the epidemic, which could have influenced observed antibody levels.

We were unable to use the age-specific seroprevalence to estimate force of infection with sero-catalytic models due to the age structure of our sample and the limited number of younger individuals under 10 years of age (Fig. 3). It is likely that we are missing important seroconversion dynamics which would only be visible in younger children.

It is important to note that the relationship between seropositivity and seroprotection is not clear due to factors like individual immune heterogeneities and cellular immunity. There were numerous individuals close to the seropositivity threshold who were seropositive to one antigen and not the other across all coronaviruses (Figs. S1-S2). Although these individuals might be classified as seronegative to that specific coronavirus, they might actually be protected from infection through other means e.g., cellular immunity. False positives and negatives are also possible due to assay noise. In addition, our assay did not include S2 targets and thus might have underestimated seroprevalence.

As expected because IgG antibodies to SARS-CoV-2 are detectable around two weeks after symptom onset, SARS-CoV-2 spike and nucleocapsid MFI levels for SARS-CoV-2 positive patients increased from day 0 to day 28 (Figs. 4, S8) [18, 50]. SARS-CoV-1 nucleocapsid MFI levels also increased from day 0 to day 28 as to be expected since there is higher homology in the SARS-CoV-1 nucleocapsid than the spike protein [51, 52]. A previous study using the same assay found that 100% of the fully converted COVID-19 individuals were 100% cross-reactive to SARS-CoV-1 nucleocapsid and 45.9% were cross-reactive with SARS-CoV-1 spike protein [18]. Surprisingly, we noticed a slight increase from day 0 to day 28 in NL63 nucleocapsid MFI levels among individuals sampled at the Treichville hospital (Fig. S8). We believe that this is a result of ongoing NL63 circulation in the area and not a result of cross-reactivity since a similar pattern was not observed in the samples collected at the sampling site in Yopougon (Fig. S8). There were no significant changes in spike or nucleocapsid MFI levels from day 0 to day 28, providing no strong evidence of cross-reactive antibodies.

Importantly, although follow‑up sampling at day 28 allowed us to capture later stages of the SARS‑CoV‑2 antibody response, including rising IgG levels, we did not observe increases in antibody levels to endemic coronaviruses across spike or nucleocapsid targets. This suggests limited evidence for boosting of endemic coronavirus antibody responses during the early convalescent period following SARS‑CoV‑2 infection within the antigen targets measured here.

However, antibody affinity and avidity continue to mature beyond day 28, and differences in maturation kinetics may influence cross‑reactive potential. Together with the lack of S2 antigens in our assays, these factors limit our ability to fully characterize cross‑reactivity and highlight the need for longitudinal studies incorporating later timepoints and broader antigenic targets.

Conclusion

To conclude, our analysis provides the first serological snapshot of endemic and emergent coronavirus exposure in Abidjan, Côte d’Ivoire during the early phase of the COVID-19 pandemic. We found high seroprevalence to endemic alphacoronaviruses and lower levels for betacoronaviruses, as well as detectable cross-reactive antibody signals to SARS-CoV-1 in the absence of known circulation. SARS-CoV-2 and SARS-CoV-1 seropositivity were higher among healthcare workers compared to PCR-confirmed SARS-CoV-2 patients, likely reflecting earlier exposure. Lastly, we found no strong evidence of cross-boosting from SARS-CoV-2 infection to other betacoronaviruses over a 28-day period. These findings contribute to our knowledge of coronavirus immunity in West Africa and underscore the need for continued serological surveillance to characterize antibody landscapes and cross-reactive responses.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

DK, MP, AOT, MD, AA, and BR contributed to study conceptualization, study design, sample collection, and laboratory testing. DK, AA, and BR also contributed to data analysis. AM, BLR, CJEM, ALG, and SC contributed to data analysis and manuscript writing. All authors reviewed and edited the manuscript. All authors read and approved the final manuscript.

Funding

Princeton Precision Health. IRD.

Data availability

The datasets used and analysed during the current study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

Authorization on August 4th 2020 from the “Comité National d’Ethique (CNESVS CI)” to implement the research protocol entitled “Projet Immunité croisée entre SARS-Co V2 et les autres coronavirus : impact épidémiologique sur la Covid − 19 en Côte d’Ivoire”.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.