We evaluated 20 pregnant individuals who received COVID-19 mRNA vaccines during pregnancy and their infants. Participants were vaccinated between December 2020 and April 2021. Gestational age at first dose ranged from 13 weeks to 40 weeks (mean 31.2, SD 5.9 weeks). Nineteen participants delivered live, singleton infants between January 2021 through April 2021 at gestational ages ranging from 37.4 to 41.1 weeks (mean 39.2, SD 1.1 weeks). One participant who was vaccinated at 13 weeks had a termination of pregnancy due to a lethal skeletal dysplasia of genetic etiology at 20.4 weeks. Eight participants received BNT-162b2 (Pfizer-BioNTech) and twelve received mRNA-1273 (Moderna) vaccines. Eighteen participants received both vaccine doses prior to delivery, and two participants received the second dose after delivery. The time from the first mRNA vaccine dose ranged from 6 to 97 (mean 51, SD 24.3) days prior to delivery, time from the second dose ranged from 2 to 75 (mean 32, SD 21.3) days prior to delivery, and in two participants 15 and 21 days after delivery (Table S1). No participants received a 3rd dose of vaccine prior to delivery. Infants born to vaccinated mothers were followed up at convenience timepoints ranging from age 3 weeks to 15 weeks of life (mean 8.3, SD 3.2). All participants were negative for prior COVID-19 infection by self-reported survey and by maternal anti-nucleocapsid IgG screening at delivery. Further demographic data is detailed in Table S1.
Vaccine mRNA products do not cross the placenta at readily detectable levels
To determine the transplacental transfer of mRNA vaccine-derived products, we examined available maternal blood at delivery, placenta tissue, and cord blood for Spike protein by western blot and vaccine mRNA by PCR. All available delivery samples (maternal blood, placental tissue, and cord blood) were negative for Spike protein by western blot (Supp Fig. 1, Supp Table 3) and did not have detectable levels of vaccine mRNA by PCR (Suppl Table 3). Together, this indicates that products of mRNA vaccination do not reach the fetus after vaccination during pregnancy at readily detectable levels.
mRNA vaccination in pregnancy leads to a robust antibody response
Similar to prior studies14,15,26, we found that mRNA vaccination during pregnancy led to an increase in anti-SARS-Cov-2 IgG following dose 1 (n = 7, mean 388.6, SD 224.8 RFU) and an even further robust increase after vaccination dose 2 (n = 12, mean 3214, SD 1383 RFU). Anti-SARS-CoV-2 IgM (n = 7, mean 53.3, SD 50.2 RFU) was detected in two maternal participants following dose 1, but only 1 participant following dose 2 (n = 12, mean 23.8, SD 17 RFU, Fig. 1).
Vaccine-induced anti-SARS-CoV-2 IgG and neutralizing antibodies are transplacentally transferred
We then evaluated the transplacental transfer of maternally-derived anti-SARS-CoV-2 IgG antibodies to their infants. Maternal blood at delivery was available in 19/20 participants and cord blood was available in 17/20 participants. Anti-SARS-CoV-2 IgG was detectable in 94.7% (18/19) of maternal blood samples at delivery (mean 3235, range [10, 7811] RFU). Anti-SARS-CoV-2 IgG was detectable in 88.2% (15/17) cord blood samples (mean 2243, range [2, 4959] RFU). One participant received one mRNA vaccine dose 9 days prior to delivery, and both the maternal and cord blood were negative for IgG at the time of delivery. Another participant received two doses of mRNA vaccine (23 and 2 days) prior to delivery and maternal blood was positive at 55 RFU (positive cutoff >50 RFU), however, cord blood IgG was negative (Fig. 2A). Maternal and cord blood anti-SARS-CoV-2 IgG levels were moderately correlated, but not statistically significant (p = 0.074, Rs = 0.446, Fig. 2A). All cord blood samples were anti-SARS-CoV-2 IgM negative.
We next evaluated the transplacental transfer of neutralizing antibody titers by a label-free surrogate neutralization assay (sVNT) from mother to cord blood. Maternal and cord blood at delivery had robust neutralizing responses (maternal n = 17, mean 220.2, range [0, 422]. Cord blood n = 16, mean 296.6, range [0, 485], Fig. 2B). All mother-infant dyads with positive IgG serology at delivery had detectable transplacental transfer of neutralizing antibodies with the exception of one pair in which the mother was borderline IgG positive at delivery and cord blood was negative, for which both maternal and cord blood were negative for neutralizing titers (Fig. 2B). However, maternal and cord blood neutralizing titers were not significantly correlated (p = 0.361, Rs = −0.244, Fig. 2B). Taken together, this indicates that maternal mRNA vaccination induces functional neutralizing antibodies that are transferred to the infant.
Maternally-derived vaccine-induced anti-SARS-CoV-2 IgG and neutralizing antibodies persist through early infancy
A subset of infants was sampled at convenience timepoints during follow-up (infant n = 11, weeks of life range [3,15] mean 8.3 weeks). Anti-SARS-CoV-2 IgG levels were positive in 81.8% of infants at follow-up (9/11 infants, mean 1290, range [1, 3225] RFU, Fig. 2A), with one infant still IgG positive at 12 weeks of age (Fig. 2C). The two infants that were IgG negative at follow-up were both born to mothers who received only one vaccine dose prior to delivery (6 and 9 days prior to delivery, respectively). One of these infants did not have paired maternal or cord blood available at the time of delivery for comparison, and the other was IgG negative in cord blood. Maternal and infant follow-up anti-SARS-CoV-2 IgG levels were not significantly correlated; however, cord blood and infant follow-up IgG levels were significantly associated (p = 0.492, Rs = 0.249 and p = 0.021, Rs = 0.76, respectively, Fig. 2A). All infants were IgM negative at the time of follow-up.
All infants with available IgG positive samples at follow-up had detectable neutralizing titers (n = 8, mean 154, range [41–256], Fig. 2B). Maternal and infant follow-up neutralizing titers were not significantly correlated, as well as cord and infant follow-up neutralizing titers (p = 0.665, Rs = −0.191 and p = 0.662, Rs = 0.214, respectively, Fig. 2B).
To compare the rate of decay of IgG antibody levels in mothers and their infants, we evaluated 5 dyads with paired maternal and infant blood samples on the same day at the time of follow-up (range 3–9 weeks post-delivery). Maternal antibody IgG levels decreased faster in mothers than infants (mean delta −974 RFU and −670 RFU, respectively. Figure 2E) at the follow-up timepoint. Taken together this indicates, maternally-derived functional vaccine-induced antibodies persist at high levels in newborns through early infancy during a critical time of immune vulnerability and may decay slower than maternal IgG antibodies.
Vaccine-induced antibody timing and transplacental facilitated transfer
We assessed the relationship of anti-SARS-CoV-2 IgG levels to neutralizing antibody levels. We found a strong correlation between IgG and neutralizing titers in maternal plasma at delivery (Rs = 0.744, p = 0.0012) and infant follow-up (Rs = 0.738, p = 0.046) timepoints, but no significant association between IgG and neutralizing titers in cord blood (Rs = 0.121, p = 0.656, Fig. 3).
We then evaluated the impact of timing of vaccination on maternal antibody levels at delivery. Consistent with the known kinetics of SARS-CoV-2 antibody responses16,27, two participants received their first dose of vaccine <30 days prior to delivery had low or absent levels antibody levels at delivery and were excluded from analysis. We found maternal IgM levels at delivery were statistically significantly correlated with time since vaccine dose 1 (Rs = −0.846, p < 0.0001, Fig. 4A). In addition, maternal IgG levels at delivery were significantly correlated with time since dose 1 (Rs = −0.866, p < 0.0001) and gestational age at delivery (Rs = 0.689, p = 0.002, Fig. 4B, C). We then evaluated neutralizing titers in those participants with known detectable IgG levels at delivery and found that maternal neutralizing titers was significantly correlated with days since vaccine dose 1 (Rs = −0.706, p = 0.002). However, maternal neutralizing titers at delivery was not associated with gestational age at dose 1 (Rs = 0.288, p = 0.279, Fig. 4D, E).
To assess facilitated antibody transfer, we evaluated cord-to-maternal antibody IgG and neutralization titer ratios by time since vaccination and gestational age in all participants with available samples. We found that cord-to-maternal IgG and neutralization titer transfer ratios were not significantly correlated (Rs = 0.257, p = 0.354, Fig. 4F). However, IgG ratios were highly correlated with both time since first maternal vaccination dose and gestational age at first dose (Rs = 0.917, p < 0.0001 and Rs = −0.739, p = 0.002, respectively, Fig. 4G, H). In contrast, neutralization titer cord-to-maternal ratios by time since first vaccination dose and gestational age at first dose were not significantly associated (Rs = 0.366, p = 0.179 and Rs = −0.032, p = 0.913, respectively, Fig. 4I, J). Together, this may indicate that timing of vaccination in pregnancy is critical for maternal-fetal antibody transfer, and functional neutralizing antibodies are differentially transferred to the fetus as compared to total anti-SARS-CoV-2 IgG during gestation.
mRNA vaccination in pregnancy leads to a unique SARS-CoV2 Spike protein antibody epitope binding signature
We next investigated antibody linear epitope binding and transplacental transfer using the PhIP-seq/VirScan SARS-CoV-2 Spike protein phage display array in mother-infant dyads at the time of birth (Fig. 5). We found that timing of vaccination was important for the transplacental transfer of Spike protein epitope binding antibodies. Two mother-infant dyads had minimal Spike protein-specific epitope binding. The first dyad only received one dose of mRNA vaccine 9 days prior to delivery, and the other dyad received the second vaccine dose 2 days prior to delivery.
We found high levels of SARS-CoV-2 Spike protein epitope binding in 4 major peaks we designate as regions 1–4 (Fig. 5A). Region 1 overlays the carboxy terminal of the N-terminal domain. Region 2 overlaps with key residues for the S1/S2 cleavage site. Regions 3 and 4 are within S2, flanking the fusion loop and the transmembrane portion of the Spike protein, respectively. However, we found minimal binding in the receptor binding domain (RBD) of Spike protein. Prior evaluation using the PhIP-seq/VirScan SARS-CoV-2 epitope phage array during SARS-CoV-2 infection demonstrated similar binding in regions 3 and 4, however, in SARS-CoV-2 infection there was minimal binding in regions 1 and 2 demonstrating that antibody epitope binding in these regions may be unique to vaccination during pregnancy28. In addition, there is proportional representation of linear epitope binding across the SARS-CoV-2 Spike protein proteome between mothers and infants (Fig. 5B). Taken together, SARS-CoV-2 antibody linear epitope binding after vaccination during pregnancy shows similar patterns, with multiple immunodominant regions found in the majority of mothers and infants. Some of these regions are unique to vaccination during pregnancy and not observed during natural infection in non-pregnant adults28,29,30.
#Evaluation #transplacental #transfer #mRNA #vaccine #products #functional #antibodies #pregnancy #infancy #Nature #Communications