Calculation of population size (n)

The population size (n) was calculated considering the number of women who received anti-tetanus and anti-pneumococcal vaccines or had infectious episodes of the intestinal tract during pregnancy, 95% confidence interval (Z= 1.96), 5% error.

Population selection

This study included colostrum samples from 113 pregnant, clinically healthy, women (Table 1) who delivered their babies at three large hospitals in Mexico City: Hospital Regional 1º de Octubre ISSSTE (HR 1º October), Hospital Regional de Alta Especialidad de Ixtapaluca 77 (HRAEI), or Hospital de la Mujer (HMuj). All procedures were explained to women, who signed assent and informed consent forms, following the Helsinki Declaration revised in 2013¹³. Clinical information related to the number of infectious episodes of the skin and respiratory, gastrointestinal, or urogenital tracts during pregnancy was obtained from each patient through an interview and clinical examination. Information depicted in the individual clinical files validated the data obtained from the interview at the time of sample collection.

Inclusion and exclusion criteria

Colostrum samples were obtained from women between 20 and 35 years of age, with a full-term pregnancy (38-42 weeks), with vaginal delivery (VD) or cesarean section (C-section), and, mainly, singleton births. Exclusion criteria were the presence of chronic maternal disease, previous abortions, preterm pregnancy, multiparous parturition, women with hormonal treatment, or acute disease within 3 weeks before delivery.

Sample processing

Colostrum samples were collected from clinically healthy women between 1 and 12 h post-delivery and frozen at −20°C until use. After defrosting at room temperature, all samples were centrifuged at 2000 g for 10 min at 4°C (Allegra™ X-22R Benchtop Centrifuges Beckman Coulter Life Sciences, Indianapolis, IN, USA) to prepare the corresponding dilutions (see below).

ELISA for quantification of specific Igs

A modification of the quantitative ELISA previously reported¹⁰ was used to quantify IgA1 and IgA2 concentrations against TT, PPS, or LPS in colostrum. Anti-TT antibodies quantification: ninety-six-well plates (Immuno Plate MaxiSorp, Thermo Scientific, NY, USA, Cat. Num. 1132249) were coated with 50 µL of 0.5 µg/mL of antigen (TT from C. tetani, Sigma-Aldrich®, Darmstadt, Germany, Cat. Num. T3194-25UG) in borate buffer. Plates were incubated overnight at 4°C. Plates were washed with 100 µL of 0.05% polyoxyethylene sorbitan monolaurate (Tween 20, Sigma-Aldrich Company®, USA, Cat. Num. 9005-64-5) in phosphate-buffered saline (PBS) (PBS-T) for 8 times. The non-specific binding free sites were blocked with 100 µL of 5% skim milk formula (Svelty® diluted in PBS) by incubation during 60 min at 37°C. Anti-PPS antibodies quantification: plates were coated with 50 µL of 5 µg/mL of total antigen (capsular polysaccharides from 23 pneumococcal serotypes, Pulmovax®, Schering-Plough®, West Point, PA, USA, Lot. Num. R033367) in PBS; plates were incubated during 5 h at 37°C and then overnight at 4°C. Plates were washed with 100 µL of PBS-T for 5 times. The non-specific binding free sites were blocked with 5% of bovine serum albumin (BSA, Sigma-Aldrich Company®, USA, Cat. Num. 9048-46-8) with serum fetal bovine (SFB, Biowest®, Riverside, MO, USA, Cat. Num. US1520) diluted in PBS-T (200 µL per well) by incubation during 60 min at 37°C. Anti-LPS antibodies quantification: plates were coated with 50 µL of 5 µg/mL of antigen (LPS from Escherichia coli, Sigma-Aldrich®, Darmstadt, Germany, Cat. Num. L2143) in PBS. Plates were incubated overnight at 4°C and washed with 100 µL of PBS-T for 5 times. The non-specific binding free sites were blocked with 5% BSA diluted in PBS-T (200 µL per well) by incubation for 30 min at 37°C.

From this point, all plates were processed under the same conditions: centrifuged samples were processed by 10-fold serial dilutions starting at ratios of 1:10, 1:100, 1:1000, and 1:10,000. For standard curves, 50 µL of standard serum dilutions from healthy individuals vaccinated against TT, PPC, or reference human serum to LPS were added and incubated for 2 h at 37°C. For the detection of specific IgA subclasses, 50 µL per well of biotinylated antibodies anti-IgA1 (monoclonal mouse antibody, Abcam®, Cat. Num. ab99796; 1:2500) or anti-IgA2 (monoclonal mouse antibody, Abcam®, Cat. Num. ab128731; 1:2000) were added and incubated under the same experimental conditions. Then, 50 µL of streptavidin-horseradish peroxidase (HRP) complex (Streptavidin HRP, Abcam®, Cat. Num. ab7403) dilution were added and incubated for 1 h at 37°C. The presence of labeled antibodies was revealed using 50 µL per well of chromogenic substrate 3’3’’-5-5-tetramethylbenzidine (TMB ELISA Substrate, Abcam®, Cat. Num. 171523) and the reaction was stopped by adding 50 µL of 0.2 M sulfuric acid per well. The absorbance was measured in a Microplate ELISA plate reader spectrophotometer (Sunrise, Tecan® TX, USA) at 450 nm. IgA1 and IgA2 standard curves were obtained for each plate to determine each sample concentration by absorbance interpolation and dilution factor. Results were finally expressed in nanograms of each IgA subclass per milliliter of colostrum (ng/mL).

Western blot

The ability of colostrum to recognize protein antigens from E. coli was evaluated by Western blot. For these assays, 10 representative colostrum samples were evaluated. E. coli O157:H7 from the microbial collection of the National School of Biological Sciences-IPN (Mexico) was grown overnight in LB medium without antibiotics. Bacteria were lysed for 30 min at 4°C with RIPA buffer (20 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM Na2 EDTA, 1 mM EGTA, 1% NP-40, 1% sodium deoxycholate, 2.5 mM sodium pyrophosphate, 1 mM b-glycerophosphate, 1 mM Na₃VO₄, and 1 µg/ml leupeptin). The lysate was then centrifuged at 12,000 ×g for 15 min at 4°C. The supernatant was separated and quantified by the Lowry method, 6X Laemmli buffer was added to the proteins, and samples were heated at 95°C for 5 min. Two-hundred micrograms of proteins were loaded and separated in 10% polyacrylamide preparative gel. The proteins were transferred to a nitrocellulose membrane (0.45 µm; BIO-RAD Cat. Num. 1620115) by a humid chamber (BIO-RAD Mini Trans-Blot® Cell Cat. Num. 1703930) at 100 volts for 1.5 h¹⁴. Non-specific binding free sites in the membrane were blocked with 5% skim milk formula (Svelty® diluted in PBS) by incubation during 60 min at 37°C. Colostrum dilutions (1:1000) in 0.05% PBS-T were used as the primary antibody, and the membrane was incubated overnight at 4°C. Biotinylated antibodies such as anti-total IgA (rabbit polyclonal antibody, Abcam®, Cat. Num. ab97218; 1:5000), anti-IgA1 (Abcam®, Cat. Num. ab99796; 1:2500), or anti-IgA2 (Abcam®, Cat. Num. ab128731; 1:2000) in 0.05% PBS-T were added and incubated 1 h at room temperature. Streptavidin-HRP was used at 1:8000 dilution and incubated 1 h at room temperature. Clarity Western ECL (luminol and peroxide) was used as a substrate. Membranes were documented using ChemiDoc™ MP Imaging System with Image Lab™ Software.

Statistical analysis

The IgA concentration distribution was analyzed by the Kolmogorov–Smirnov test. A comparison of Ig concentrations between samples was performed by non-parametric Mann–Whitney U-test or Kruskal-Wallis test. All statistical tests and graphics were developed using statistical program GraphPad Prism (GraphPad® Software version 5, La Jolla, CA, USA).

Population data

The mean age of mothers was 27.4 ± 5.2 years; 50% had a cesarean delivery (C-section) and the other half a VD. For 95% of women, this was their first child. Information on previous infectious episodes and vaccine protocols during pregnancy is shown in Table 1.

Levels of IgA subclasses in colostrum against studied antigens

The levels of IgA subclasses against TT, PPS, and LPS were quantified in colostrum. None of the colostrum Ig levels had a normal distribution. As in the previous reports⁷, we found that colostrum had higher amounts of IgA1 against TT and, in contrast, IgA2 was more abundant against PPS and LPS (Fig. 1).

Figure 1 Immunoglobulin (Ig)A subclasses in the function of the type of antigen. The bar graph shows the concentrations of IgA1 (blue) and IgA2 (red) against tetanus toxin, Escherichia coli lipopolysaccharide, and Pulmovax 23® (pneumococcal polysaccharides) (n = 113). All values are represented as mean ± standard deviation. All data are expressed in nanograms of Ig/milliliter of colostrum (ng/mL). Statistical analysis was performed using the Mann–Whitney U-test. ***p < 0.001. 

IgA subclasses against E. coli antigens by Western blot

The ability of colostrum to recognize protein antigens from E. coli was evaluated by Western blot. For these assays, 10 representative colostrum samples were evaluated. The pattern of proteins recognized by total IgA was similar among the different colostrum samples, showing variations in the intensity of the bands between the different samples (Fig. 2). When IgA1 recognition was evaluated, we found that the banding pattern changed considerably; although some samples varied in the number of proteins recognized, and in the intensity of the bands, the pattern was considerably different when compared with total IgA. In sharp contrast, IgA2 recognized fewer antigen proteins with reduced intensity. The pattern and the intensity of the bands recognized by IgA2 differed from that seen with IgA1 and total IgA.

Figure 2 The SDS-PAGE analysis of colostrum immunoglobulin (Ig)A subclasses against Escherichia coli antigens. Colostrum samples were analyzed to determine the recognition profile of E. coli O157:H7 proteins. The membranes were developed with anti-total IgA (A), anti-IgA1 (B), and anti-IgA2 (C). MW markers, for size comparison, are shown at the left. Lanes 1-10 contain 10 selected colostrum samples. 

IgA subclasses did not show variations among population descriptors

Many descriptors played a pivotal role in identifying the characteristics of the population analyzed. The following descriptors were included in the function of the size of the sampled population: the hospital where the mothers were recruited (Supplementary Fig. S1a) and the delivery method (Supplementary Fig. S1b). These comparisons did not show any statistical differences between groups. Hence, we concluded that Ig levels were independent of the descriptors used to stratify this population.

Figure S1 Stratification of colostrum immunoglobulin (Ig) levels in the function of population descriptors. The bar chart shows comparative amounts of IgA against specific antigens in colostrum in the function of (A) hospitals: HR 1° October, HRAEI, and HMuj; and (B) delivery methods: vaginal delivery and cesarean section. Results are shown as mean ± standard deviation. Statistical analysis was performed using the Kruskal-Wallis test for non-parametric independent data. 

The concentrations of colostrum IgA subclasses vary in women with previous antigenic stimulation during pregnancy

The amount of colostrum IgA subclasses was compared between women without vaccination or previous infection reports during pregnancy. Results showed that both IgA subclasses against TT, PPS, or LPS increased in women vaccinated with specific antigens (TT or PPS) or women that had infections at the gastrointestinal tract during pregnancy, in comparison with women without infections (Fig. 3). These results were compared with total IgA subclasses levels in the colostrum (Supplementary Fig. S2). This effect was analyzed by the Pearson’s correlation test to evaluate whether multiple interactions with the same antigen during pregnancy had an effect on IgA subclasses levels in colostrum. Although we did not find a statistical correlation, levels of IgA1 tended to increase in colostrum from women with more than one TT vaccination dose. In contrast, IgA2 increased in colostrum from women with multiple episodes of gastrointestinal tract infections during pregnancy when LPS was used as an antigen in the test (Supplementary Fig. S3). As it was already mentioned, 98 women vaccinated with Pulmovax® developed a strong response against polysaccharide, with predominant recognition by IgA2 (Fig. 3).

Figure 3 Distribution of immunoglobulin (Ig)A subclasses in the function of the type of antigen, comparing women with or without previous specific antigen stimulation. The amounts of Ig from women with previous infections at the gastrointestinal tract (lipopolysaccharide [LPS]) or vaccination during pregnancy (tetanus toxin [TT] or pneumococcal polysaccharides [PPS]) are compared with each IgA subclass. The bar graphs show IgA1 (blue) and IgA2 (red) between women without documented stimulation during pregnancy (bars with rods n = 63) and women with at least one antigenic stimulation during their pregnancy (bars without rods = 50): TT (n = 27) or PPS (n = 98) vaccination or LPS (n = 13). All values are represented as mean ± standard deviation. All data are expressed in nanograms of Ig per milliliter of colostrum (ng/mL). Statistical analysis was performed using the Mann–Whitney U-test. ***p<0.001. 

Figure S2 Quantification of the total immunoglobulin (Ig)A subclasses present in colostrum. Comparative amounts of IgA types in colostrum (IgA, IgA1, and IgA2), n = 113. Bars indicate mean ± standard deviation. All data are expressed in milligrams of Ig per milliliter of colostrum (mg/mL). 

Figure S3 Comparison of immunoglobulin (Ig)A1 and IgA2 levels in colostrum against (A) tetanus toxin (TT) or (B) lipopolysaccharide (LPS), in function of the number of vaccination shots (TT) or infections at the gastrointestinal tract (LPS) during pregnancy. All values are represented as mean ± standard deviation. All data are expressed in nanograms of Ig per milliliter of colostrum (ng/mL).