Human milk is a complex physiological fluid that provides nutrients as well as bioactive factors for the infant⁽Reference Donovan¹⁾. A distinctive property of human milk among all other species is its remarkable content and structural diversity of oligosaccharides⁽Reference Bode^2–Reference German, Freeman and Lebrilla⁴⁾. A broad range of functions has been attributed to human milk oligosaccharides (HMO), including serving as a component of the innate immunity of human milk by preventing attachment of potential pathogens to the intestinal lining⁽Reference Newburg, Ruiz-Palacios and Morrow^5, Reference Morrow, Ruiz-Palacios and Jiang⁶⁾ and by serving as a prebiotic to promote colonization by a healthy gut microbiota⁽Reference Bode^2–Reference German, Freeman and Lebrilla⁴⁾. HMO are resistant to digestion⁽Reference Chaturvedi, Warren and Buescher⁷⁾, thus little focus had been placed on the potential impact of HMO on the host epithelial cells. However, two recent articles published in the British Journal of Nutrition by Kuntz et al. ⁽Reference Kuntz, Rudloff and Kunz^8, Reference Kuntz, Kunz and Rudloff⁹⁾ provide compelling evidence that HMO modulate intestinal cell proliferation, differentiation and apoptosis via receptor binding and MAP-kinase signalling. These reports provide novel insight into a new contribution of HMO to neonatal intestinal development.

HMO are the third most predominant component in human milk after lactose and lipid, averaging about 20 g/l in colostrum and 5–10 g/l in mature milk⁽Reference Chaturvedi, Warren and Altaye^10, Reference Coppa, Pierani and Zampini¹¹⁾. HMO differ in molecular weight and structure, constituent sugar or sugar derivatives and pH⁽Reference Kunz, Rudloff and Baier³⁾. They are composed of both neutral and anionic components. HMO are synthesized in the mammary gland with a lactose core at the reducing end and elongation by N-acetyllactosamine units, with greater structural diversity provided by extensive fucosylation and/or sialylation wherein fucose and sialic acid residues are added at the terminal positions⁽Reference Bode^2–Reference German, Freeman and Lebrilla⁴⁾. HMO vary in size from three to thirty-two sugars⁽Reference German, Freeman and Lebrilla⁴⁾ and consist of linear and branched polymers linked together by at least twelve different types of glycosidic bonds⁽Reference Kunz, Rudloff and Baier³⁾. Current analytical methods have detected approximately 200 distinct molecular species of HMO consisting of mostly neutral and fucosylated oligosaccharides⁽Reference German, Freeman and Lebrilla^4, Reference Niñonuevo, Park and Yin¹²⁾.

HMO differ among women and four phenotypic groups, consistent with the Lewis blood group system, have been recognized based on the expression and activity of two fucosyltransferases⁽Reference Stahl, Thurl and Henker¹³⁾. The HMO components were recently identified in five donors by microfluidic chips and MS⁽Reference Niñonuevo, Perkins and Francis¹⁴⁾. The dominant oligosaccharides components were lacto-N-tetraose (LNT), lacto-N-neotetraose and lacto-N-fucopentaose I/V. A neutral oligosaccharide with neutral mass 709.3 Da (3Hex, 1HexNAc–LNT) was the most prominent oligosaccharide that was present in all human milk samples⁽Reference Niñonuevo, Perkins and Francis¹⁴⁾. Consistent with the bifidogenic activity of HMO, this HMO has been previously shown to be preferentially fermented by Bifidobacterium longum biovar infantis, an isolate from the infant gut⁽Reference LoCascio, Niñonuevo and Freeman^15, Reference Ward, Niñonuevo and Mills¹⁶⁾. The next most common species with relatively strong abundances consisted of neutral fucosylated oligosaccharides and a fucosylated species with a sialic acid residue.

HMO resist digestion within the stomach and intestine and a significant proportion of HMO passes into the lower gastrointestinal tract⁽Reference Chaturvedi, Warren and Buescher⁷⁾, where they are selectively metabolized by beneficial micro-organisms. The bifidogenic potential of HMO is frequently cited as an important reason why the gastrointestinal microbiota of breast-fed infants contains proportionally more bifidobacteria than that of formula-fed infants⁽Reference Harmsen, Wildeboer-Veloo and Raangs¹⁷⁾. To gain insight into the mechanisms underlying the bifidogenic properties of human milk, the transcriptome of B. longum LMG 13 197 grown in human milk or formula containing galacto-oligosaccharides and long-chain fructo-oligosaccharides was compared with each other and with bacteria grown on glucose-containing media⁽Reference Gonzalez, Klaassens and Malinen¹⁸⁾. Common genes that were highly up-regulated by both human milk and formula included putative genes for cell surface type 2 glycoprotein-binding fimbriae, which are implicated in attachment and colonization in the intestine⁽Reference Gonzalez, Klaassens and Malinen¹⁸⁾. Genes involved in carbohydrate metabolism formed the dominant group specifically up-regulated in breast milk and included putative genes for N-acetylglucosamine degradation and for metabolism of mucin and HMO via the galactose/lacto-N-biose gene cluster⁽Reference Gonzalez, Klaassens and Malinen¹⁸⁾. Thus, HMO exert bifidogenic properties by serving as a fermentable substrate⁽Reference LoCascio, Niñonuevo and Freeman^15, Reference Ward, Niñonuevo and Mills¹⁶⁾ and, in turn, modulate the bacterium's transcriptome to support its catalytic activity. Indeed, the genomic sequence of bifidobacterium revealed 700 genes that are unique to B. longum infantis, relative to other bifidobacterium including a variety of co-regulated glycosidases, suggesting a co-evolution of this strain of bifidobacterium to be uniquely suited for colonization of the human milk-fed infant⁽Reference German, Freeman and Lebrilla⁴⁾.

Certain HMO share common epitopes to those present on the infant's intestinal epithelia and known receptors for pathogens and, thus, act as decoys to prevent binding of pathogens to the epithelial cells⁽Reference Newburg, Ruiz-Palacios and Morrow^5, Reference Morrow, Ruiz-Palacios and Jiang^6, Reference Kunz and Rudloff¹⁹⁾. The significant immunological protection afforded by HMO has been partly attributed to the presence of α(1–2)-linked fucosylated oligosaccharides⁽Reference Morrow, Ruiz-Palacios and Jiang⁶⁾. Fucosylated oligosaccharides are capable of preventing diarrhoeal illness through diverse mechanisms, including inhibition of Escherichia coli activity by binding and blocking access to target receptors, prevention of Campylobacter adhesion to intestinal cells and competitive inhibition of the binding of Norovirus to the intestinal epithelium⁽Reference Newburg, Ruiz-Palacios and Morrow⁵⁾.

Last, HMO also can be absorbed into the circulatory system and are excreted in the urine of breast-fed infants⁽Reference Rudloff, Pohlentz and Diekmann²⁰⁾. HMO are transported via the circulation to other sites, such as the urinary tract, where they are thought to block pathogen adhesion⁽Reference Chaturvedi, Warren and Buescher^7, Reference Rudloff, Pohlentz and Diekmann²⁰⁾. HMO have structural similarities to selectin ligands, which mediate important cell–cell interactions in the immune system. Leucocyte infiltration into tissues is associated with many inflammatory conditions. Acidic HMO inhibited rolling and adhesion of leucocytes isolated from human blood to cultured human umbilical vein endothelial cells in a concentration-dependent manner⁽Reference Bode, Kunz and Muhly-Reinholz²¹⁾. Furthermore, Bode et al. ⁽Reference Bode, Rudloff and Kunz²²⁾ demonstrated that acidic (sialic-containing) HMO, but not neutral HMO, serve as anti-inflammatory components by inhibiting the formation of platelet–neutrophil complexes (PNC) and neutrophil activation. These findings support a role for HMO anti-inflammatory factors which may contribute to the lower incidence and severity of inflammatory diseases in breast-fed infants⁽Reference Kunz and Rudloff^19, Reference Bode, Kunz and Muhly-Reinholz^21, Reference Bode, Rudloff and Kunz²²⁾.

In a study published earlier this year in the British Journal of Nutrition, Kuntz et al. ⁽Reference Kuntz, Rudloff and Kunz⁸⁾ tested the hypothesis that HMO would affect intestinal epithelial cell dynamics. Effects of isolated acidic HMO and neutral HMO fractions or individual oligosaccharides on proliferation, differentiation and apoptosis were assessed in preconfluent transformed human intestinal cells (HT-29 and Caco-2) and non-transformed small-intestinal epithelial crypt cells of fetal origin (HIEC). Growth inhibition was induced by neutral and acidic HMO fractions in all cell lines in a dose-dependent manner. However, transformed cell lines (HT-29 and Caco-2 cells) were more sensitive than HIEC cells. In contrast, HMO induced differentiation (alkaline phosphatase activity) in HT-29 and HIEC cells, but not Caco-2 cells. Among individual oligosaccharides, only sialyllactoses induced differentiation. Induction of apoptosis was also detected in HT-29 and HIEC cells, but for neutral oligosaccharides, not for acidic fractions. Thus, for the first time, acidic and neutral HMO were shown to induce growth inhibition in intestinal cells through at least two different mechanisms, by suppressing cell cycle progression through the induction of differentiation and/or by influencing apoptosis⁽Reference Kuntz, Rudloff and Kunz⁸⁾.

To further elucidate the underlying molecular mechanisms of action of HMO on cell cycle dynamics, in a study published in this issue of the British Journal of Nutrition, Kuntz et al. ⁽Reference Kuntz, Kunz and Rudloff⁹⁾ exposed HT-29, HIEC and Caco-2 cells to neutral or acidic HMO and investigated cell cycle events via flow cytometry and expression levels of cell cycle regulators by using quantitative real-time RT-PCR. Consistent with their previous study⁽Reference Kuntz, Kunz and Rudloff⁹⁾, both acidic and neutral fractions induced a concentration-dependent decrease in proliferation, which was evidenced by G₂/M-arrest and associated changes in cyclin A and B mRNA expression. In addition, they observed that the expression of the cyclin-dependent kinase inhibitors p21^cip1 and p27^kip1p21^cip1 was p53-independent and necessary for arresting cells in the G2/M phase, while p27^kip1 was associated with differentiation effects. Lastly, both neutral and acidic HMO induced phosphorylation of the epidermal growth factor receptor and stimulated the MAP-kinase signalling pathway (ERK1/2 and p38 phosphorylation)⁽Reference Kuntz, Kunz and Rudloff⁹⁾.

Further studies are needed to determine the mechanism(s) whereby HMO are able to stimulate EGFR signalling and whether these effects can be blocked by specific inhibitors of EGFR- and p38-signalling pathways. In addition, the potential physiological relevance of these in vitro observations should be established by translation to preclinical animal models to determine whether HMO affect intestinal cell dynamics within the complex milieu of the gastrointestinal tract.