Fig 2: Contribution of disease severity and time since symptom onset on breastmilk antibody transfer(A) The scores plot of a partial least-squares regression (PLSR) model built on serum-derived antibody features with maternal disease severity, based on NIH criteria, as the outcome variable.(B) PLSR model built using breastmilk-derived antibody features with disease severity as the outcome variable. The dot plot (left) shows the scores of each sample, with each sample indicated by a dot and the color representing disease severity. The bar plot (right) illustrates the loadings of the features selected via Elastic Net on latent variable 1 (LV1), with the color indicating whether the feature was enriched in patients with milder (pink) or more severe (purple) COVID-19 symptoms.(C) Linear regression models fitting the relationship between days since symptom (Sx) onset and IgA1, IgG1, and IgM antibody titers in breastmilk (purple) and serum (teal). p values are reported below each plot.

Fig 3: HIV-specific responses and immune response capability of the case at 9.5 years of age. a Detection of HIV-specific antibodies at 9.5 years of age by western blot. The case antibody profile is compared with controls that are a high positive, low positive and HIV-negative. HIV proteins corresponding to bands in the blots are shown in the left grey-shaded block; the case profile was positive for the core proteins indicated in pink. b Quantitation of HIV-specific antibodies by multiplex bead array for all isotypes and subclasses (indicated on the left side—IgG1, IgG2, IgG3, IgG4, IgM, IgA1, IgA2) against gp41, Gag, RT, gp120, Nef, Tat, Vpu, peptide Vpu9 and V1V2 scaffold antigens (indicated at the top). Results are expressed as mean fluorescence intensities (MFI)), the colour key shows ranges of MFI according to colour intensity (the darker the more HIV-specific antibody detected). A result is considered positive if above the cut-off (mean ± 3 SD) determined from eight adult uninfected controls. Vpu9 amino acid sequence: STMVDMGHLRLLDVNDL. c Proportions of natural killer (NK) cells that respond to anti-CD16 antibody, and CD4+ and CD8+ T cells that respond to staphylococcal enterotoxin B (SEB) in a whole blood intracellular cytokine (ICC) assay that measures induction of interferon-? (IFN-?) and interleukin-2 (IL-2). HIV-uninfected adult reference values for comparison (n = 21; median % and range)—natural killer (NK) anti-CD16%: 37.92 (12–67.6), CD4 SEB%: 6.04 (0.25–11.91), CD8 SEB%: 5.82 (0.18–18.94). d A weak positive CD4+ T cell response to Gag (0.116%) in the absence of a detectable CD8+ T cell response to Gag (<0.1%; 0.023%). UN: addition of costimulatory antibodies anti-CD28 and anti-CD49d, no stimulation with peptides

Fig 4: mRNA-1273 vaccination elicits humoral responses to SARS-CoV-2 VOCs.(A and B) The line graphs show the vaccine-induced IgM, IgA1 and IgG1 recognition to D614G (wild-type; blue), alpha (B.1.1.7; yellow), beta (B.1.351; purple), delta (B.1.617.2; orange), and omicron (B.1.1.529; red) variants of concern RBDs (A) or full spike protein (B). (C) neutralizing titers are shown as pNT50 values. All plots in (A to C) show results for samples from children (n50 µg=6, n100 µg=9) and adults (n=14) at V2, where each individual’s response is linked across VOC antigens. (D) The line graphs show the Fc?R (Fc?R2a, Fc?R2b, Fc?R3a, Fc?R3b) binding profiles of vaccine-induced antibodies to RBD or spike protein VOC antigens across children (n50µg=6, n100µg=9) or adults (n=14) at V2, where each individual’s response is linked across VOC antigens. (E) Breadth score was calculated by summing up the number of VOC RBD- or spike protein-specific features above the median value for each individual (n50 µg=6, n100 µg=9) or adults (n=14) at V2. Horizontal bars in (E) indicate median. Background corrected data are shown and negative values were set to 100 for graphing purposes in (A to D). A Kruskal-Wallis test with a Benjamini-Hochberg post-test correction for multiple comparisons was used to test for statistical differences between wild-type and VOC titers within groups in (A to D) or between the groups in (E). P-values for significantly different features are shown above and fold change reductions of omicron titers compared to wild-type are shown below each dataset. Data for adult samples has been partly published previously ( 55 )).

Fig 5: SARS-CoV-2-infected women induce an antibody response in serum and breastmilk(A and B) The heatmaps summarize the antibody isotypes and FcR-binding signatures against SARS-CoV-2 in serum (A) and breastmilk (B) for SARS-CoV-2 uninfected (top) and infected (bottom) women. The color scale corresponds with the Z score for each antibody titer measured, with lighter coloration representing a positive Z score and darker coloration representing a negative Z score. The data represent the average of two replicates.(C) The dot plots show the IgG1, IgA1, and IgM titers and FcgR2a, FcgR3a, and FcaR binding against SARS-CoV-2 spike in the serum of SARS-CoV-2-infected (left, dark teal) and uninfected (right, light teal) mothers. Significance was determined by Mann-Whitney test; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. The data represent the average of two replicates.(D) The dot plots show the IgG1, IgA1, and IgM titers and FcgR2a, FcgR3a, and FcaR binding against SARS-CoV-2 spike in the breastmilk from SARS-CoV-2-infected (left, dark purple) and uninfected (right, light purple) mothers. Significance was determined by Mann-Whitney test; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. The data represent the average of two replicates.(E) The dot plots show the ratio of titers and FcR-binding against SARS-CoV-2 spike in breastmilk to serum. Significance was determined by a one-way ANOVA; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.