We conducted a retrospective cross-sectional study firstly on archived plasma samples collected between 2009 and 2018 before the COVID-19 pandemic in Cameroon and secondly on samples collected during the COVID-19 pandemic period between 2021 and 2022.

This study utilized de-identified, anonymized plasma samples biobanked after malaria research studies (19, 20), and it was approved by the Centre Regional Ethics Committee for Human Health Research (CE-N⁰0146/CRERSHC/2021). Administrative authorizations were obtained from the Ministry of Public Health of Cameroon and from local health centers. For samples collected during the pandemic period (2021–2022), authorization was granted by the directors of the three selected health facilities. An informational note was provided to all eligible participants, who then gave their written informed consent prior to enrollment in the study. The confidentiality of study participants was maintained using identification codes.

Samples prior the outbreak of COVID-19 in Cameroon used in this study were collected in three geographic areas of the city of Yaoundé in Cameroon, a rural area in Ngali, a peri-urban area at the Marie-Reine d’Etoudi Medical Center and an urban area at the Yaoundé Central Hospital and the Efoulan District Hospital. Processing was carried out at the Immunology Laboratory of the Biotechnology Center of the University of Yaoundé I and at the Center for Research on Emerging and Reemerging Diseases (CREMER) of the Institute of Medical Research and Medicinal Plant Studies.

Overall, we included 1,329 plasma samples from pregnant women among which 1,290 were archival samples collected before the COVID-19 pandemic (pre-COVID-19) and 39 plasma samples from women at delivery collected during the pandemic (peri-COVID-19). Those samples were previously anonymized and codified as per institutional biobanking procedures. The archival samples were collected from 629 consenting pregnant women recruited during the second trimester between the 12^th and 27^th weeks of pregnancy and 661 at delivery. All women were aged between 15 to 45 years (Figure 1).

SARS-CoV-2 antibodies were tested using the Abbott Panbio™ COVID-19 IgG/IgM Rapid Diagnostic Test as per manufacturer’s instructions. Assay performance as reported by the manufacturer was as follows: sensitivity (97.8%); specificity (92.8%); precision (>99% both for intra-assay and inter-assay assessments) (21). Briefly, plasma samples were mixed by low-speed vortex after which 10µL of supernatant was applied to the specimen well (S) of the test device. Two drops (approximately 60µL) of buffer were added and incubated for 10 minutes before reading the test device.

For serology assays, recombinant SARS-CoV-2 Hexapro spike protein was used for this study (22). The SARS-CoV-2, OC43 and NL63 spike proteins were expressed in Human Embryonic Kidney (HEK) 293F suspension cells by transfecting the cells with the spike plasmid as previously described (23). After 6 days, proteins were purified using a nickel resin followed by size-exclusion chromatography. Relevant fractions were collected and frozen at -80°C until use.

Participants’ serostatus was assessed using an Enzyme-Linked Immunosorbent Assay (ELISA) for quantitative detection of anti-SARS-CoV-2 Spike IgG in plasma samples (24, 25). 96-well high-binding plates were coated separately with 50 µL of SARS-CoV-2, OC43 or NL63 spike protein diluted in 2 µg/ml of 1x PBS and incubated overnight at 4°C. After the overnight incubation of the plates coated by the antigens, the plates were washed 3 times with 250 μL of wash buffer per well using an automated plate washer. 5% skimmed milk powder diluted in wash buffer (1x PBS, 0.05% Tween 20 and distilled water) was used as blocking buffer. 200 µL of blocking buffer was then added to each well and incubate at room temperature (RT) for one hour. After this incubation, the plates were washed as previously described and 100 µL plasma samples pre-diluted to a 1:100 dilution in blocking buffer, were added to the plates and incubated for 75 minutes. Titrations were performed by doing a serial dilution of each positive sample in blocking buffer to determine the effective concentration of the antibodies present. Following incubation, plates were washed and the secondary antibody diluted at 1:3000 in blocking buffer was added to the plates and further incubated for 60 minutes then washed. 100 µL TMB substrate (Thermofisher Scientific) was then added to all the wells and incubated for 5 min at RT in the dark. Upon stopping the reaction with 1M H₂SO₄, absorbance was measured at a 450 nm wavelength and used to calculate the half maximal effective concentration (EC50) using GraphPad Prism v9.0.0.121. The mAb CR3022 was used as positive control and Palivizumab was used as a negative control. A Pregnant women was considered to be seropositive at anti-SARS-CoV-2 IgG antibodies when OD450nm ≥ 0.4.

Pseudoviruses were produced by co-transfection of HEK293T cells with the SARS-CoV-2 614G spike (D614G) plasmids in conjunction with a firefly luciferase encoding lentivirus backbone plasmid (pNL4plasmid) with PEIMAX (Polysciences). Culture supernatants were clarified of cells by a 0.45 μM filter and stored at −70°C (26, 27).

The 293T/ACE2.MF cells modified to overexpress human ACE2 were kindly provided by M. Farzan (Scripps Research). Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Life Technologies) containing 10% heat-inactivated serum (FBS) and 3 μg/mL puromycin at 37°C, 5% CO2. Plasma samples were heat-inactivated and clarified by centrifugation. Pseudovirus and serially diluted plasma were incubated for 1 h at 37°C, 5% CO₂. Cells were added at 1 × 10⁴ cells per well and after 72 h of incubation at 37°C, 5% CO₂ luminescence was measured as a reduction in luciferase gene expression after single-round infection of 293T/ACE2.MF cells with spike-pseudotyped viruses (26–28).

The threshold value used to determine the neutralization titer was given by the ID50, which represents the dilution at which 50% of viral activity is inhibited. Our starting dilution factor was 25. Then, for all pregnant women with an ID50 greater than 25 were considered positive and their antibody titers were taken into account.

Data collected were entered on Microsoft Excel 2013 and Graphpad Prism version 9.0.0.121 was used for statistical analysis. Association analyses were performed using Pearson ChiSquare Test. We used a linear regression in order to understand the relationship between anti-D614G and anti-OC43 and anti-NL63 antibodies. The Mann-Whitney test was also used in this study to determine whether there is a significant difference between the median hemoglobin level of women at the second trimester and for those of women at delivery. The paired t-test was used to compare the means of mother’s anti-SARS-CoV-2 antibodies profile against their newborn’s anti-SARS-CoV-2 antibodies profile. Statistical significance was considered at p<0.05.

In total, 1,329 samples were included in this study, among which 1,290 were archived plasmas while 78 were collected during the COVID-19 period. The socio-demographic parameters are summarized in Table 1.

This study included a total of 1,329 women with 629 (47.3%) in their second trimester and 700 (52.7%) at delivery. The participants’ average age was 26.1 years, with ages ranging from 15 to 44. There was a statistical difference in both gravidity and parity among this study population (p < 0.0001). Among the women in their second trimester, 103 (81.7%) resided in rural areas, and 222 (31.3%) lived in peri-urban areas, while 304 (66.7%) were from urban areas. Among those at delivery, 26 (2.1%) resided in rural areas, 486 (68.6%) lived in peri-urban areas, and 152 (33.3%) were from urban areas. However, there was no statistical variation (p=0.2) of age among women between those areas of residence. Samples were collected during two periods: pre-COVID-19 and peri-COVID-19. The pregnant women who gave birth during the COVID-19 pandemic were significantly (p=0.01) older (average of 28 years old) than those before the COVID-19 pandemic (average of 26 years old). During the pre-COVID-19 period, 629 women (48.8%) were sampled in their second trimester, and 661 women (51.2%) were sampled at delivery. During the peri-COVID-19 period, 39 women were sampled, all at delivery. A significant difference in hemoglobin (HGB) levels was observed between women in their second trimester and those at delivery (p < 0.0001) (Table 1).

The seroprevalence of pregnant women with cross-reactive anti-SARS-CoV-2 antibodies was 16.5% among 1,290 women tested using RDT (Figure 2A) with significantly higher prevalence (p=0.0086) at delivery (19.2%) compared to the prevalence obtained during the second trimester (13.7%) (Figure 2A). Since this study primarily focuses on seroprotection, which is mainly represented by neutralization, we laid emphasis on samples with IgG which is efficiently transported across the placenta to the fetus. A higher positivity rate (p<0.0001) to IgM (80.3%) compared to IgM/IgG (19.7%) was seen among the pregnant women during the second trimester (80.2% and 19.8% respectively) and at delivery (80.3% and 19.7% respectively) as shown (Figure 2B).

Additionally, the study showed a prevalence of pregnant women seropositive for cross-reactive anti-SARS-CoV-2 antibodies was 12.2% (20/90) (Figure 3A) with a prevalence of 14.3% for women at delivery, lower compared to those seropositive in the second trimester (7.4%) although not significant (p=0.4943, p=0.909) (Figure 3B).

After performing both a full antibody screening and titration for 90 samples using ELISA which enabled us to identify samples that can bind efficiently to the SARS CoV-2 antigens (antibody binding), 12.2% (11/90) of pregnant women met that threshold, however, none of the antibodies in these samples were neutralizing.

To investigate whether anti-SARS-CoV-2 cross-reactivity in pre-COVID-19 samples correlated with past exposure to other endemic coronaviruses [HCoVs (OC43 and NL63)] in the same samples, pre-COVID-19 plasma samples of pregnant women that cross-reacted with SARS-CoV-2 were tested to determine their anti-HCoVs (anti-OC43 and anti-NL63) responses (Figure 4). Among the 90 pregnant women assessed, 22.2% of them had cross-reactive anti-SARS-CoV-2 antibodies directed against the Spike. However, all those women were seropositive to the HCoVs tested in this study (Figure 4A). No correlation was found between cross-reactive anti-SARS-CoV-2 (anti-D614G) antibodies and anti-OC43 antibodies (Figure 4B) and between (anti-D614G) and anti-NL63 antibodies (Figure 4C).

To assess cross-neutralization in samples collected during the COVID-19 pandemic as a comparative group, we processed 39 plasma samples of pregnant women at delivery from during the peri-COVID-19 era by RDT. We found that 53.8% (21/39) of women at delivery were seropositive with anti-SARS-CoV-2 antibodies (Figure 5A) meanwhile, all these women (n=28) were seropositive to anti-SARS-CoV-2 antibody binding by ELISA (Figure 5C). By isotyping the antibodies, more pregnant women expressed higher levels of IgG (p=0.0003) than IgM and IgM+IgG during the pandemic (Figure 5B). Overall, 20/22-90.9% of plasma from pregnant women assessed using the spike-pseudotyped lentivirus neutralization assay had neutralizing antibodies.

In this study, transplacental transfer was evaluated in pregnant women who showed seropositivity for anti-SARS-CoV-2 antibodies in their peripheral blood. The transfer was assessed by detecting these antibodies both in the maternal peripheral blood and in the corresponding cord blood, which reflects the newborn’s blood. Among the 39 women included in the transplacental transfer analysis of this work, 30 women (76.9%) transferred antibodies by measuring antibodies level in cord blood, while 9 women (23.1%) did not. Specifically, during the pre-COVID-19 period, 4/11 women (13.3%) transferred antibodies and 7/11 women (77.8%) did not. During the peri-COVID-19 period, 26/28 women (86.7%) transferred antibodies and 2/28 women (22.2%) did not. The difference between these groups was statistically significant, with a p<0002. No significant difference (p= 0.6) was observed between the mean maternal age of women who transferred antibodies (28 years old) and women who did not transfer antibodies (28 years old). When looking at the area of residence, 24/31women (77.4%) from peri-urban areas and 6/8 women (75%) from urban areas transferred antibodies. In contrast, 7/31 women (22.6%) from peri-urban areas and 2/8 women (25%) from urban areas did not transfer antibodies. Also, no significant difference was found between primigravida (80%) and multigravida (85.7) and between primiparous (85%) and multiparous (73.3%) who transferred antibodies (p=0.7 and p=0.4 respectively) (Table 2).

The efficient transfer of IgG antibodies from seropositive pregnant women to their newborns was also demonstrated in this study (Figure 6). Cord blood was used as a surrogacy for newborn blood. Cross-reactive antibodies to SARS CoV-2 were transferred from mother to child prior to and during the pandemic, however, at a lower proportion prior to the pandemic (36.4%) compared to during the pandemic (92.9%). In fact, the transplacental transfer of anti-SAR-CoV-2 antibodies occurred both before and during the COVID-9 pandemic. However, that transfer was significantly represented during the pandemic (p<0.0001) than prior to COVID-19 (p=0.2) (Figure 6A). There is no significant difference in the antibody titers between mothers and cord samples that could signify an efficient transplacental transfer of those antibodies from the mother to the fetus (p=0.6347) (Figure 6B). There was a strong, statistically significant correlation (r² = 0.96; p<0.0001) between the levels of neutralizing antibodies in mothers and their cords, with the transfer of neutralizing antibodies from mother to cord (ratio=1) (Figure 6C).

To evaluate the efficiency of transplacental antibody transfer during the peri-COVID period, we compared the neutralizing capacity of anti-SARS-CoV-2 antibodies between maternal and cord blood samples (Figure 6D). Although, there was no significant difference in the neutralizing capacity of anti SAR-CoV-2 antibodies between maternal and their newborns peri-COVID-19 (p=0.6698).

Limitations include a relatively small sample size and potential biases in the selection process due to inadequate sample size, which may lead to certain groups being underrepresented. Indeed, before being subjected to ELISA testing, the plasma samples were first screened using the Abbot rapid diagnostic test (RDT), which detects the presence or absence of IgG antibodies. This initial screening served as an indicator of potential seropositivity, which is the primary focus of this study. All these procedures significantly reduced the sample size, to the extent that no samples were seropositive to cross-reactive anti-SARS-CoV-2 antibodies in certain areas as in rural area. Such comparisons using ELISA results would likely introduce bias, given the widely disproportionate number of samples tested by ELISA in certain locations. Another biggest shortcomings of this study was the lack of follow-up that could have helped us track exactly how antibody levels behaved between the second and third trimesters. Targeting the second trimester may reflect maternal seroconversion. This is insufficient to assess transplacental transfer of antibodies, as the peak occurs mainly during the third trimester due to increased expression of the neonatal Fc receptor (FcRn) and placental maturity. The lack of detailed investigation into other HCoVs and SARS-CoV-2 variants, which could contribute to cross-protection observed in pre-pandemic samples also limited our findings. To mitigate these issues, we employed rigorous statistical methods, ensured diverse sampling, and validated our findings with multiple approaches to enhance the reliability and generalizability of our results. We also analyzed peri-pandemic samples to assess long-term immunity and protection levels in mothers and children. Another key limitation of this study is the inability to include vaccinated pregnant which restricted our capacity to assess vaccine-related immune responses throughout pregnancy because current national data on SARS-CoV-2 vaccination in pregnant women remain limited. In this context, our study, although focused on pre-vaccination samples, highlights the critical need for strengthened surveillance systems capable of monitoring antibody responses and vaccine impact in pregnant women.