DISCUSSION
The influence of early life nutrition on the development of the immune response has not previously been studied in the first few weeks of life. This is an important question as epidemiological data suggest that breastfeeding is associated with long-term health benefits, such as a lower incidence of allergy, asthma, diabetes, celiac disease and multiple sclerosis19-22. This mechanism may be imprinted early after birth when the immune system faces the dual challenge of establishing inflammatory capacity against pathogens whilst developing tolerance towards harmless antigens.
A striking finding was that Tregs expand substantially in the first 3 weeks of life and this expansion was more profound in breastfed babies in contrast to those receiving formula feed. Tregs of breastfed neonates also display an activated phenotype with increased expression of HLA-DR, a marker of increased suppressive activity23. On the other hand, changes in the profile of pro-inflammatory cytokine production were also observed at this early stage of life. IL-8 production, a major phenotypic attribute within cord blood, was largely maintained but a striking feature was the increase in intensity of IL-17 production by both CD4+ and CD8+ cells by 3 weeks of age. This is likely to reflect recognition of bacterial antigen during establishment of the microbiome and may be balanced by the coordinated Treg expansion during this period. An increase in serum IL-17 has been shown at 4 weeks and is likely explained by these observations5. Most of these features were independent of nutrition although IFN-γ production by T cells was lower in exclusively breastfed neonates. Our observation that CD8+ cells are less abundant and have reduced cytotoxic capacity in neonates compared to adults confirms previous results24,25.
One of the most interesting findings was that T cells from breastfed neonates display reduced proliferative responses and produce substantially lower levels of Th1 cytokines when challenged with maternal cells. This was specific to maternal antigens and was not present against unrelated PBMC and did not result from an intrinsic reduction of cytokine producing capacity. As such this reflects the development of immunological tolerance against non-inherited maternal antigens (NIMA) in exclusively breastfed neonates. Importantly, we were also able to show that NIMA-specific tolerance was mediated by Tregs, and is therefore linked to the expansion of this population in breastfed neonates.
An additional observation was that neonatal PBMC at 3 weeks of age triggered stronger immune responses from maternal PBMC compared to cord blood cells. This may reflect maturation of antigen presentation function by 3 weeks of age, although this was not assessed in this study. This also indicates that the fetal immune system may contribute to suppression of maternal immune recognition during pregnancy by maintaining a tolerogenic phenotype prior to parturition.
It is interesting to speculate on potential mechanisms by which breastfeeding can promote NIMA-specific tolerance in neonates. Our observations likely reflect immune tolerance to gastrointestinal presentation of maternal cells within breastmilk26. Transplacental passage of cells during pregnancy leads to reciprocal microchimerism that can persist for many years. Furthermore, this ’microchiome’ of maternal cells supports fertility in female offspring by promoting immune tolerance to NIMA during next-generation pregnancies27. Beneficial effects are also seen when NIMA are shared between donors and recipients of allogeneic renal or hemopoietic stem cell transplantation. Importantly, the establishment of NIMA-specific tolerance has been shown to be dependent on breastfeeding and nutritional history is also a determinant of NIMA-associated transplant outcome 28,29. Our findings show that breastfeeding directly promotes the development of Tregs that suppress recognition of maternal tissue, thus likely supporting maternal microchimerism and potentially conferring lifelong benefits in relation to fertility and immune protection against infectious agents and cancer27,30.
A further recent observation is that breastfeeding, through the transfer of human milk oligosaccharides, exerts important prebiotic and immunomodulatory effects including the development of tolerogenic dendritic cells which prime Tregs31,32.
We were also interested to assess how nutrition could impact on the formation of the early microbiome and how this might correlate with immune function. Microbial composition was broadly comparable in breastfed and formula-fed neonates and this is likely to reflect the fact that all babies in our cohort were delivered by caesarean section. Dysbiosis of the microbiota has been found to occur following delivery by caesarean section and in infants who are not breastfed33,34. Nevertheless, although it may take several months for nutrition to markedly influence microbiome composition35 subtle differences in microbial diversity were already apparent at 3 weeks of life. In line with previously published results36, we observed that the gut microbiome of breastfed neonates is more abundant in short chain fatty acid (SCFA) producing bacterial genera, such as Gemella andVeillonella . SCFAs, in particular propionate and butyrate, play an important role in promoting Treg differentiation and proliferation via the inhibition of histone deacetylases11. This notion is supported by a link between the presence of Veillonellaand the proportion of Tregs at 3 weeks of age in our network modelling analysis.
We selected elective caesarean deliveries for our study as labor is known to promote pro-inflammatory changes37. Further studies will therefore be needed to establish the relative contribution of mode of delivery and nutritional history on the development of NIMA-specific tolerance in the neonate. A further limitation of this work is the relatively low number of neonates in the examined feeding groups. Nevertheless, we established a unique and homogenous cohort of healthy neonates who were sampled at 3 weeks of age exclusively for the purposes of this study.
In summary, we demonstrate that the neonatal immune system undergoes substantial maturation in the first 3 weeks of life with an increase in IL-17 production in T cells and a balanced increase in the Treg population. Moreover, breastfed neonates show a specific and Treg dependent reduction in proliferative T cell responses to NIMA, associated with a reduction in inflammatory cytokine production. These findings add to our understanding of mechanisms by which early life nutrition can determine long term health outcomes38.