Stability and heterogeneity in the antimicrobiota reactivity of human milk-derived immunoglobulin A

Abstract

Immunoglobulin A (IgA) is secreted into breast milk and is critical for both protecting against enteric pathogens and shaping the infant intestinal microbiota. The efficacy of breast milk-derived maternal IgA (BrmIgA) is dependent upon its specificity; however, heterogeneity in BrmIgA binding ability to the infant microbiota is not known. Using a flow cytometric array, we analyzed the reactivity of BrmIgA against bacteria common to the infant microbiota and discovered substantial heterogeneity between all donors, independent of preterm or term delivery. Surprisingly, we also observed intradonor variability in the BrmIgA response to closely related bacterial isolates. Conversely, longitudinal analysis showed that the antibacterial BrmIgA reactivity was relatively stable through time, even between sequential infants, indicating that mammary gland IgA responses are durable. Together, our study demonstrates that the antibacterial BrmIgA reactivity displays interindividual heterogeneity but intraindividual stability. These findings have important implications for how breast milk shapes the development of the preterm infant microbiota and protects against necrotizing enterocolitis.

Figure S1.

Isolation of IgA from breast milk by Peptide M columns substantially enriches for secretory IgA. (A) The soluble fraction of a breast milk sample was run over a Peptide M column and analyzed on an LDS-PAGE gel. Shown are the Peptide M bound fraction, the flow through, and the aqueous phase prior to separation. MW, molecular weight. (B) LDS-PAGE gels were prepared as in A, transferred to nitrocellulose, and blotted with antibodies to different proteins as described above the blot. L = ladder; M+ = Peptide M bound fraction and M− = flow through of Peptide M column. (C) Concordance of IgA estimations made my 280 nm light absorbance and IgA ELISA. (D) 10-fold dilutions of Peptide M purified IgA fractions run over the flow cytometric array as described in Fig. 1 A. Shown are two example bacteria. A and B are one example of two separate experiments. C is the aggregate of all of the donor samples used in the manuscript. D was performed once. Source data are available for this figure: SourceData FS1.

Figure 1.

A flow cytometric array for measuring the antibacterial reactivity of breast milk–derived IgA. (A) Design of the flow cytometric array. Made with BioRender.com. (B) Examples of Syto BC+/SSCDim staining (2 of 3,636) used to discriminate bacteria from debris/bubbles in the flow cytometer (control is empty well stained with Syto BC). Numbers represent the percentage of events inside the gate. (C) Two examples (of 101) of the magnitude of antibacterial IgA binding detected in our array comparing two donors (9 and 10) that differ in their antibacterial IgA responses. The bottom row shows the reactivity of an anti-HIV IgA antibody against the bacterial isolates (one experiment). Numbers in red represent the gMFI of that sample. (D) Breast milk–derived IgA reactivity, from 10 donors against the environmental bacteria B. japonicum. One experiment.

Figure 2.

Heterogeneity in the antibacterial reactivity of breast milk–derived IgA. Donor milk samples (term infants; >37 wk gestational age) were analyzed with our flow cytometric array (Fig. 1 A). (A) Heat map of normalized antibacterial IgA binding affinity of different donors. Hierarchical clustering (Spearman). The range of the normalized values across each row is indicated in the left-hand column. Donor numbers indicated on top of each column. (B) Scatter graph showing the normalized antibacterial IgA binding values for each donor (each color represents a different donor). (C) Scatter graph of the normalized BrmIgA binding to different isolates of E. coli separated according to donors selected from the analysis in A. Donor numbers indicated on bottom of each column. (D and E) A correlation network analysis was performed to describe which antibacterial IgA responses were predictive. (D)  Heat map indicating the level of correlation between different bacteria in our array. Black box drawn around Enterobacteriaceae family taxa. This figure is aggregate of experiments on 33 donor samples. (E) Network diagram indicating significantly correlated antibacterial IgA responses.

Figure 3.

Heterogeneity in breast milk–derived antibacterial IgA reactivity from donors who delivered preterm infants. Donor milk samples (preterm infants; 24–35 wk gestational age) were analyzed with our flow cytometric array (Fig. 1 A). (A) Bar graph showing the concentration of IgA purified from donor milk samples from mothers of term and preterm infants (ELISA). (B) Heat map of normalized antibacterial binding affinity of different preterm donors (Spearman). Samples where no data were collected due to insufficient bacteria in the well are colored gray. Donor numbers indicated on top of each column. (C) Scatter graph showing the normalized antibacterial IgA binding values for each preterm donor (each color represents a different donor). (D) PCA comparing aggregate antibacterial IgA binding between preterm and term samples. This figure is an aggregate of experiments on 15 donor samples.

Figure 4.

Temporal stability of antibacterial maternal IgA reactivity within one childbirth/infant. Multiple milk samples were collected from different donors over time and analyzed with our flow cytometric array (Fig. 1 A). (A) Heat map of normalized antibacterial binding affinity of different donors. Hierarchical clustering (Spearman) of various donors is indicated by colored bars above and below the heatmap. Date of collection indicated on heatmap: D## = number of days after delivery of sample collection. Donor numbers indicated on top of each column. (B) Scatter graph showing the normalized antibacterial IgA binding values for each sample from longitudinally collected donors (each color represents a different donor; from A). (C and D) PCA of the aggregate antibacterial IgA binding of longitudinally collected samples. Each donor colored as in A. (C) PCA of individual longitudinally collected samples where symbols indicate the time of collection (week after delivery). (D) PCA from C where ellipses indicate the maximum variance for each donor cluster along each axis. No ellipses are drawn for samples where fewer than four samples were available. This figure is an aggregate of experiments on 33 samples from eight donors.

Figure S2.

Longitudinal analysis of the antibaterial IgA response from sequentially collected breast milk samples. Graph of normalized IgA binding to all bacterial types from longitudinal milk samples. Samples from each donor are in order left to right. Each donor is represented by an individual graph and each bacteria by a different colored line. This experiment was performed once.

Figure 5.

Stability of breast milk–derived antibacterial IgA reactivity over the course of sibling infants. Breast milk samples were collected from consecutive siblings and analyzed with our flow cytometric array (Fig. 1 A). (A) Heat map of normalized antibacterial binding affinity of different donors. Hierarchical clustering (Spearman) of various donors is indicated by colored bars above and below the heatmap that correspond to each donor. Donor numbers indicated on top of each column. (B) PCA of aggregate antibacterial samples where each donor is displayed in a different color (from A). The first sibling is indicated by a circle and the second sibling a triangle. Samples colored as in A. (C) Paired Student’s t tests were calculated comparing the IgA binding of each donor between infant one and infant two for each bacterial taxon. The mean change ((Infant 2 − Infant 1; taxa 1) + (Infant 2 − Infant 1; taxa x))/36 (# of taxa) for each paired test was calculated and graphed. Significant increase in second infant = “up” triangle; significant decrease in second infant = “down” triangle; no statistical significance = circle. Colors are according to A. See Fig. S3 for each paired student’s t test. This figure is an aggregate of experiments on 20 samples from 10 donors.

Figure S3.

Detailed comparisons of the antibacterial IgA reactivity from donors collected over two separate infants. Paired Student’s t tests comparing the antibacterial IgA binding between breast milk collected from the first and second infant. Each dot represents a different bacterial taxon and each graph is a different donor dyad (as indicated). The right-hand graph is the difference (norm. IgA binding infant 2 − norm. IgA binding infant 1) for each taxon. This experiment was performed once.

Figure 6.

Holder pasteurization reduces the bacterial binding properties of breast milk–derived IgA. Breast milk samples from four donors were split into two where one half was pasteurized (62.5°C for 30 min) while the other was untreated as a control. IgA was then isolated from both halves and analyzed on our flow cytometric array (Fig. 1 A). (A) Paired Student’s t test (***P < 0.001) of the IgA concentration (mg/ml) of control (blue) and pasteurized samples (red), as measured by ELISA. (B) Paired Student’s t tests comparing control (blue) and Holder pasteurized (red) milk samples from the same donor. Each dot represents a different bacterial taxon. ****P < 0.0001. This figure is an aggregate of experiments on four donor samples.

Figure S4.

Effect of pasteurization on different antibody components. Two breast milk samples (A and D) were thawed and IgA isolated by passage over a Peptide M column. Each sample was split in two then treated by Holder pasteurization (62.5°C for 30 min) or left on ice for 30 min. Control samples were used just after Peptide M isolation. (A) Samples were boiled in Laemmli buffer and run on an LDS-PAGE gel (4–15% Gradient Acrylamide gel). Samples were normalized to protein content after pasteurization. (B) “D” samples were transferred onto nitrocellulose and blotted for heavy chain (right) or light chain (kappa; left). Samples were normalized to protein content after pasteurization. HP, Holder pasteurization; Ctrl, no treatment; SecF, SeF; IgAHC, IgA heavy chain; LC, light chain; MW, molecular weight. Source data are available for this figure: SourceData FS4