Autoantibodies to αS1-casein are induced by breast-feeding

Abstract

Background: The generation of antibodies is impaired in newborns due to an immature immune system and reduced exposure to pathogens due to maternally derived antibodies and placental functions. During nursing, the immune system of newborns is challenged with multiple milk-derived proteins. Amongst them, caseins are the main constituent. In particular, human αS1-casein (CSN1S1) was recently shown to possess immunomodulatory properties. We were thus interested to determine if auto-antibodies to CSN1S1 are induced by breast-feeding and may be sustained into adulthood.

Methods: 62 sera of healthy adult individuals who were (n = 37) or were not (n = 25) breast-fed against human CSN1S1 were investigated by a new SD (surface display)-ELISA. For cross-checking, these sera were tested for anti Epstein-Barr virus (EBV) antibodies by a commercial ELISA.

Results: IgG-antibodies were predominantly detected in individuals who had been nursed. At a cut-off value of 0.4, the SD-ELISA identified individuals with a history of having been breast-fed with a sensitivity of 80% and a specificity of 92%. Under these conditions, 35 out of 37 sera from healthy donors, who where breast-fed, reacted positively but only 5 sera of the 25 donors who were not breast-fed. The duration of breast-feeding was of no consequence to the antibody reaction as some healthy donors were only short term breast-fed (5 days minimum until 6 weeks maximum), but exhibited significant serum reaction against human CSN1S1 nonetheless.

Conclusion: We postulate that human CSN1S1 is an autoantigen. The antigenicity is orally determined, caused by breast-feeding, and sustained into adulthood.

Figure 1. Structure of the CSN1S1 fusion protein for SD ELISA development.

The environment of the fusion sites is given as sequences. The eight amino acids at the N-terminus and the two amino acids at the C-terminus that were added to CSN1S1 due to the cloning procedure are shown in italics. Restriction sites used for cloning are underlined. SP, signal peptide.

Figure 2. Surface accessibility of human CSN1S1 on E. coli.

2.1) SDS-PAGE of outer membrane protein preparations from E. coli UT5600(DE3) pKP010 which were incubated without trypsin (lane 1) and cells which were treated with trypsin (lane 2). OmpF/C and OmpA are natural outer membrane proteins of E. coli, and can be used as reporters for successful outer membrane protein preparations. 2.2) Flow cytometer analysis of E. coli cells displaying αS1-casein. (A) E. coli UT5600(DE3) pKP010 were incubated with a polyclonal rabbit antiserum against CSN1S1 casein and subsequently with a secondary antibody conjugated with FITC. (B) The mean fluorescence of E. coli UT5600(DE3) pKP010 was 10,071. The value of the fluorescence for E. coli UT5600(DE3) cells without a plasmid, used as a negative control, was 464. 2.3) Fluorescence microscopy of both cell types. The cells were treated identically as described for flow cytometer analysis (2.2). (A) E. coli UT5600(DE3) pKP010 ; (B) E. coli UT5600(DE3) as negative control; (C) E. coli UT5600(DE3) as negative control, transmission light control.

Figure 3. Antibody reaction against human CSN1S1 in human sera.

62 sera of healthy volunteers were tested for antibodies against CSN1S1 with the SD-ELISA. Every serum was measured three times independently, the average was calculated and is shown as a column. SD of each mean is given as a line.

Figure 4. Analyzing the serum reaction against human CSN1S1 with respect to formula-fed and breast-fed test persons.

4.1) For each cut-off value, sensitivity and specificity were calculated to choose the optimal cut-off value for the assay which was 0.4. sensitivity; ▪ specificity ▴ 4.2) The 62 probands were divided into two collectives (formula-fed and breast-fed). 25 of the 62 volunteers were formula-fed and 37 were breast-fed. Only 5 sera of the formula-fed collective have an absorption value over 0.4 and 2 sera of the breast-fed collective have an absorption value under 0.4. (dotted line = cut-off value).

Figure 5. Serum reaction against human CSN1S1 in probands who were breast-fed for a short period of time.

Eight probands who were breast fed for a short period of time (5 d–42 d), showed an increased antibody reaction against CSN1S1 in comparison to control persons, who were not breast-fed. (dotted line = cut-off value from Fig. 4).

Figure 6. Serum reaction against EBV-VCA in formula-fed and breast-fed probands analyzed as a control.

A commercially available ELISA for antibodies against EBV-VCA was used as described in material and methods.

Figure 7. Antibody reaction against bovine αS1-casein.

(A) Antibody reaction against E. coli UT5600(DE3) pSH3 (control, lane1), E. coli UT5600(DE3) displaying human CSN1S1 (lane 2) and E. coli UT5600(DE3) displaying bovine CSN1S1 (lane3) measured by SD-ELISA with a rabbit polyclonal anti- bovine CSN1S1 antibody. For detection a goat HRP conjugated anti rabbit antibody was used according to the conditions described for Fig. 3. (B) The antibody reaction against E. coli UT5600(DE3) pSH3 (control, black columns) and E. coli UT5600(DE3) displaying bovine CSN1S1 (white columns) were analyzed at different concentrations of the rabbit polyclonal anti-bovine CSN1S1 antiserum or in PBS with 3% FCS as negative control. (C) The SD ELISA against bovine CSN1S1 was repeated three times independently in triplicates at the highest antibody concentration applied (200 ng/ml) with E. coli UT5600(DE3)SH3 (control, lanes 1–3) and E. coli UT5600(DE3) displaying bovine CSN1S1 (lane4–6) in order to test the reproducibility.

Figure 8. IgM reaction against human CSN1S1 in human sera.

61 of the 62 sera shown in Fig. 3 with the IgG reaction on human CSN1S1 were analyzed for an IgM mediated serum response on the same antigen by SD-ELISA. Each serum was tested three times, the mean was calculated and shown as columns. SD of each mean is given as a line.

Figure 9. Analyzing the IgM serum reaction against human CSN1S1 with respect to formula-fed and breast-fed test persons.

61 of the 62 sera shown in Fig. 3 were divided into two consortia, a formula-fed and breast-fed, as it is shown for the IgG mediated response in Fig. 4. The formula-fed consortium comprises the sera of 25 volunteers, whereas within the breast-fed consortium 36 sera were analyzed.