The potential of HMOs in neonatology

Editor(s): Lars Bode.

The potential of HMOs in neonatology

Lars Bode

Unlike the milk of many other mammals, human milk contains a high amount of diverse complex sugars called human milk oligosaccharides (HMOs). The composition varies between women and, to a certain extent, changes over the course of lactation. Current research focuses on the maternal factors that drive HMO variation and aims to understand how HMO composition impacts immediate and long-term infant health and development.

The milk of every woman has a distinct HMO composition; it’s almost like a thumb print (Bode, Glycobiology 2012). For example, data from the Canadian CHILD cohort with 1,200 analyzed milk samples reveals a wide spec- trum of different concentrations for individual HMOs (Azad et al., J Nutr 2018). Instead of comparing individual HMOs, principal component analysis of the entire HMO composition in each milk sample further highlights the wide variation in HMO composition between different women.

Maternal Drivers
The principal component plot (Fig. 1) includes HMO data from almost 10,000 milk samples collected from different sites around the world. Each dot in the three-dimensional space represents the HMO composition in one milk sample. The closer the dots are to each other, the more similar the HMO composition in those two samples. The further apart the dots, the more dissimilar or different the HMO composition. The plot highlights the milk samples from two different cohorts with the CHILD cohort samples in red and an India rotavirus cohort in green (ref). The samples cluster in slightly different locations, suggesting that HMO composition is different in different parts of the world. However, even more evident is the strong left-right separation that is independent of cohorts and geographic location, but can be explained by maternal gene- tics. In fact, single nucleotide polymorphisms (SNPs; the difference in one single base pair in a person’s DNA) is responsible for the two HMO “lactotypes”. The SNPs introduce a premature stop codon and inactivate an enzyme called fucosyltransferase 2 (FUT2), which participates in HMO synthesis. Secretor mom's (right cluster) have an active FUT2 and are able to synthesize a specific HMO subset. Nonsecretor moms (left cluster) have an inactive FUT2 and are unable to synthesize these specific HMOs, but they can still synthesize many other HMOs. This one base pair change in mom’s DNA dramatically changes her entire HMO composition. The frequency of Secretors and Nonsecretors varies around the world.

For example, in South America the abundance of Secretors is ~95%, in some parts of Africa it’s only ~65% (McGuire et al., Am J Clin Nutr 2017).

In addition to genetics, other maternal fixed and modifiable factors like diet and exercise, probiotics and other supplements, health status and drugs, may influence HMO composition and are currently under investigation.

Infant Health and Development
Understanding the maternal factors that drive the variation in HMO composition represents only one axis of the mother-milk-infant ‘triad’. The other axis links HMO composition with infant health and development, and the available data is a mixed bag of associations in observational cohort studies, mechanistic insights from tissue culture and animal models, and recently intervention studies with individual HMOs added to infant formula. To this extent, it is believed that HMOs serve as prebiotics, antimicrobials and antiadhesives, but also directly interact with epithelial and immune cells either locally in the gut or systemically after absorption in the circulation.

Prebiotic effects:
HMOs can be utilized by and provide a growth advantage to certain potentially beneficial bacteria in the infant gut. Other, potentially harmful bacteria, may not be able to utilize HMOs and are at a disadvantage. HMO utilization leads to the production of short chain fatty acids and other HMO biotransformation products that potentially benefit the infant, but also lower the pH in the gut lumen and help keep other bacteria in check. However, an overabundance of HMO-utilizers like specific Bifidobacteria may also create a disadvantage as HMOs, particular the higher molecular structures, are being used and are no longer available to exert other protective effects.

Antiadhesive effects:
Due to the constant flow and movement in the gut, many microbes need to be able to attach to the intestinal surface to avoid being washed out. Many microbes have specific anchor proteins that allow them to adhere to complex sugar molecules (glycocalyx) on the intestinal surface. HMOs, complex sugars themselves, resemble the structure of the intestinal surface sugars and serve as decoy receptors and block the adhesion of microbes, which are no longer able to attach, proliferate and in some cases invade and cause disease. For example, specific HMOs block the adhesion of enter- opathogenic E. coli (EPEC) both in vitro and in suckling mice. EPEC is one cause of diarrheal disease, which continues to takes the lives of more than 2,000 children under the age of 5 every single day.

Rotavirus infection and vaccination:
Rotavirus is another major cause of infectious diarrhea and death. One specific rotavirus strain called G10P[11] is prominent in certain areas in India and mainly affects neonates.

Surprisingly, specific HMOs increase this strain’s infectivity in tissue culture and the very same HMOs are associated with symptomatic rotavirus infections in a mother-infant cohort in India (Ramani et al., Nat  Commun 2018). Does that mean that first pathogens are starting to exploit HMOs to their advantage? Or can we turn this around and use the generated knowledge to develop new vaccination strategies that include HMOs?

In addition to protecting from infections, HMOs may also affect infant metabolism – either indirectly through shaping microbial communities in the infant gut or directly by affecting metabolically relevant host tissues like fat, liver, muscles – and influence body weight and composition in infancy and beyond. For example, an explorative study in the Danish SKOT III cohort with 30 mother-infant pairs showed that exclusively breastfed infants with extreme weight gain in the first six months of life received mother’s milk that contained more 2’FL and less LNnT than the normal weight group (Larsson et al., Frontiers Pediatrics 2018). A Finnish study within the STEPS cohort (N=802 mother-infant pairs) corroborated these results in healthy children: again, higher concentrations of 2’FL and lower concentrations of LNnT in the mother’s milk were associated with higher length and weight during the breastfeeding period and up to 5 years of age. (Lagstrom/Rautava, manuscript in revision). While there is currently no data showing causalities or mechanistic understanding, these and other cohort studies suggest that HMOs are associated with infant body weight and composition, and high- light that it’s not only individual HMOs, but the relative abundance of different HMOs to each other that drive outcomes.

Necrotizing Enterocolitis (NEC):
5% of all VLBW preterm infants develop NEC. Most intriguingly, infants who receive human milk are at six-to-ten-fold lower risk to develop this devastating and often deadly disorder.

HMOs improve survival and reduce NEC pathology in animal models, with disialyllacto- N-tetraose (DSLNT) being the most effective individual HMO. Human cohort studies with mothers and their VLBW infants corroborate these results, showing that infants who develop NEC receive less DSLNT with the mother’s milk than infants who do not develop the disorder. Data from animal studies and human cohorts together establish association and causalities, strongly indicating that DSLNT is protective against NEC.