Decoding the HMO‒microbiome axis: bridging maternal milk to infant health outcomes

Abstract

Human milk oligosaccharides (HMO) are unique glycans in breast milk that critically mediate interactions between the microbiome and infant health. This review synthesizes recent advances in understanding HMOs as key connectors of maternal and neonatal microbiomes with health outcomes. HMO composition, shaped by genetic factors (e.g., Secretor/Lewis status), lactation stage, maternal diet, and health, exhibits significant structural diversity. Critically, a bidirectional relationship exists between HMOs and the maternal microbiota: the maternal microbiome influences HMO biosynthesis, while HMOs selectively shape the microbial community within breast milk. In neonates, indigestible HMOs function as prebiotics, driving the assembly of a beneficial gut microbiota dominated by Bifidobacterium. This HMO-guided microbial establishment is fundamental to infant health, conferring protection against pathogenic infections (respiratory, gastrointestinal, and urinary), reducing the risk of necrotizing enterocolitis and allergies, and promoting healthy weight regulation, neurodevelopment, and bone mineralization. The modulation of host‒microbe interactions by HMOs underpins these systemic benefits, highlighting their central role as microbial and immunological regulators. Understanding the HMO-microbiome axis provides a holistic framework for elucidating how breast milk components foster infant development and disease resilience.

1. Introduction

Human breast milk is a complex biological fluid that serves as the primary source of nutrition for infants, comprising carbohydrates, lipids, milk proteins and water, as well as numerous micronutrients essential for infant growth and development.¹ Beyond its nutritional value, breast milk contains a myriad of functional components including antibodies, antimicrobial peptides, lactoferrin, and mucins, which collectively contribute to infant health and immunity.² Among these bioactive constituents, human milk oligosaccharides (HMOs) stand out as unique and structurally diverse glycans that are exclusive to human milk.³ The third most abundant solid constituent in human milk following lactose and lipids,⁴ HMOs are complex oligosaccharides built from monosaccharides, including sialic acid and other units, linked via glycosidic bonds.⁵ As well as being a defining feature of human breast milk, HMOs also play a critical role in shaping the infant gut microbiota, significantly influencing the development of the infant gut microbiome and overall health.⁶

Infant health is modulated by HMOs through various mechanisms. Research has demonstrated that HMOs are integral to immune regulation,⁷ gut barrier integrity,⁸ pathogenic defense,⁹ and neurodevelopment.¹⁰ By acting as decoy receptors, HMOs prevent pathogenic bacteria from adhering to intestinal cells, thereby reducing the risk of infections and decreasing reliance on antibiotics.¹¹ Additionally, HMOs modulate the expression of mucins and polysaccharides, increase nutrient absorption, and exert immunomodulatory effects through Toll-like receptors, which can reduce gut permeability and prevent gastrointestinal disorders such as diarrhea.^12,13 Furthermore, indigestible HMOs that reach the colon can function as prebiotics, fostering the proliferation of beneficial bacteria, particularly genera, including Bifidobacterium ¹⁴, thus establishing a healthy gut microbiome in infants.¹⁵

HMOs are coevolved mediators of mother‒infant communication. This review advances the field by first integrating how lactation stage, maternal metabolic health, and environmental factors collectively modulate HMO synthesis with a comprehensive perspective not fully addressed elsewhere. A key novel insight we present is how HMOs function as microbiota-derived epigenetic regulators, translating signals from the maternal gut microbiome to actively shape infant metabolic programming. Collectively, these findings highlight the translational potential of HMO profiling for developing targeted microbial therapeutics and personalized nutrition strategies to mitigate early-life disease risk.

2. Factors affecting human milk oligosaccharide production

HMO profiles exhibit significant variation between individuals and over time. This variation is not random but is systematically influenced by a complex interplay of biological, maternal, and external factors. These determinants can be categorized into six primary groups (Figure 1). Genetic factors, including hereditary markers, fundamentally shape HMOs production, but maternal health, medication use, and physical activity modulate HMOs production. A diet that encompasses specific micronutrients, alongside the lactation stage and practical breastfeeding patterns, further refines the maternal HMO composition. Finally, broader environmental factors contribute to population-level differences. This integrated visual framework showcases the ways in which collective and interacting influences from multiple factors can lead to variable HMO production.

Figure 1.

Key factors influencing human milk oligosaccharide composition. This diagram illustrates the six interconnected categories impacting HMOs variation in breast milk: genetics, maternal factors, the environment, diet, lactation stage, and breastfeeding pattern.

2.1. Genetic factors

Genetic polymorphisms in the Secretor (Se) and Lewis (Le) blood groups constitute the primary biological mechanism governing HMO diversity and concentration. The expression of fucosyl transferase enzymes encoded by FUT2 (secretor gene) and FUT3 (Lewis gene) directly determines biosynthesis in fucosylated HMOs.^16-18 While FUT2 affects 2ʹ-fucosyllactose (2ʹ-FL) through α-1, 2-fucosylation pattern synthesis, FUT3 regulates α-1, 3/4-fucosylation processes,¹⁹ collectively generating four distinct phenotypic profiles: Se⁺/Le⁺, Se⁻/Le⁺, Se⁺/Le⁻, and Se⁻/Le⁻.²⁰ The functional dominance of FUT2 polymorphisms manifests in pronounced nutritional disparities between secretor phenotypes. Secretor-positive (Se⁺/Le⁺) mothers produce breast milk containing 50%–80% fucosylated HMOs with enhanced structural complexity, whereas non-secretor (Se⁻/Le⁻) milk exhibits <10% fucosylated HMOs.^21,22 This arises from rs601338 single nucleotide polymorphisms (SNPs) in FUT2 that abolish α1, 2-fucosyltransferase activity, creating a bimodal distribution where 72% of global populations are secretors versus 28% non-secretors, with notable ethnic variations (74% secretors in Caucasians vs. 60% in Asians).^23,24 Low level of the fucosylated HMOs can lead to reduced bacterial anti-adhesion efficacy due to diminished 3-fucosyllactose (3-FL) and lacto-N-fucopentaose I (LNFP-I) content.^21,22 Understanding these genetically determined HMO profiles is therefore essential for comprehending inter-individual variations in infant nutrition, gut microbiome development, and immune protection provided by breast milk.

The differences in the secretion characteristics of the population caused by specific SNP-HMO interactions not only exist at the ethnic level but also play a key role in the gene-dose dependent regulation of HMO biosynthesis. Specifically, FUT2 rs601338 directly governs 2ʹ-FL and LNFP-I production, while FUT3 variants rs28362459 and rs812936 regulate lacto-N-fucopentaose II (LNFP-II) synthesis.²⁵ Quantitative data show that secretor status mothers produce higher HMOs concentrations in milk—11.3 g/L compared to 5.8 g/L in non-secretors.^26-30 Additionally, 2ʹ-FL constitutes 38.6 ± 14.6% of the HMOs (0.12–6.4 g/L) in secretor milk, whereas 2ʹ-FL is almost absent (<0.1 g/L) in non-secretor milk.^26-30 This stark difference highlights the genetic underpinning of HMO production, while the significant considerable in 2ʹ-FL concentration (±14.6%) observed among secretors highlights the potential influence of non-genetic modifying factors.

2.2. Lactation stage

HMOs exhibit dynamic compositional and quantitative variations across lactation stages, directly affecting the nutritional and immunological properties of breast milk.³¹ Temporal shifts in HMO profiles influence microbial colonization by serving as selective substrates for beneficial microbiota while suppressing pathogenic species through competitive exclusion and receptor mimicry mechanisms.

During the colostrum phase (days 1–5 postpartum), human breast milk contains the highest HMOs concentration of HMOs (9–22 g/L).³² Short-chain oligosaccharides with type I structural motifs which are critical for neonatal immunity also dominate during the colostrum phase.³³ Key components include 2ʹ-FL, lacto-N-tetraose (LNT), LNFP-I, and lacto-N-difucohexaose I (LNDFH I), which collectively enhance intestinal barrier function and selectively stimulate bifidobacterial growth via preferential utilization of type I structures.^34,35 These early phase HMOs help to establish infant intestinal antimicrobial environments by competitively inhibiting pathogen adhesion to epithelial receptors while simultaneously promoting symbiotic microbial networks.^36,37 The transition phase (days 6–14 postpartum) exhibits reduced HMOs diversity compared to the colostrum phase, but retains high concentrations (8–19 g/L) containing increased long-chain glycans like lacto-N-neotetraose (LNnT).³⁸ These complex structures, enriched in type II motifs, support neurodevelopmental processes through sialic acid-mediated synaptic formation and enhance systemic immune priming.³⁹ At two weeks postpartum, milk production switches to the mature lactation stage where by HMO concentrations declines progressively (6–15 g/L at 1 month; 6–8 g/L post 4 months), and becomes dominated by sialylated structures including 6ʹ-sialyllactose (6ʹ-SL) and 3ʹ-sialyllactose (3ʹ-SL).^40,41 These, sialylated structures enhance molecule stability in the infant gastrointestinal tract,⁴² ensuring prebiotic activity.⁴³ In general, structures including 2ʹ-FL, 3ʹ-SL, 6ʹ-SL, LNnT, LNT, lacto-N-sialotetraose a (LST a), lacto-N-sialotetraose c (LST c), and LNFP-I decrease gradually as the lactation period progresses, whereas production of 3-FL, difucosyllactose (DFL) and lacto-N-fucopentaose III (LNFP-III) become increased as lactation period progresses.^44-47

2.3. Breast feeding pattern

Breast feeding practices significantly influence the concentration and composition of HMOs in breast milk. The evidence demonstrates that during direct breastfeeding, infant suckling stimulates a neuroendocrine cascade that enhances HMO synthesis.⁴⁸ Specifically, the mechanical stimulation from suckling triggers oxytocin release from the mother's posterior pituitary gland via the activation of sensory nerves innervating the nipple and areola.^49,50 This oxytocin surge induces areolar smooth muscle contraction, facilitating milk ejection while simultaneously upregulating HMO synthesis in mammary gland secretory cells.⁵¹ The interplay between suckling mechanics and endocrine regulation not only promotes copious milk production but also increases HMO concentrations.^52,53 Pumping breast milk for bottle-feeding can disrupt this physiological process, with a reduction in mechanical stimulation of the breast leading to decreased oxytocin release, resulting in reduced milk volume and HMOs concentration.⁵⁴ Emerging evidence suggests that the feeding frequency further modulates HMO production. Regular, frequent suckling may promote sustained milk synthesis and is correlated with enhanced HMOs production.^55,56 We speculate that prolonged, consistent lactation stimulation has the potential to optimize breast milk composition, potentially mitigating adverse effects on HMO synthesis resulting from mother‒infant separation or disrupted suckling patterns, although these mechanisms require further validation.^57,58 The complex interplay of mechanical stimulation, hormonal regulation, and feeding frequency highlights the critical role of direct breastfeeding in promoting optimal breast milk composition, particularly regarding HMOs concentrations, and maximizing the nutritional benefits for infants.

2.4. Diet

The maternal diet is now firmly established as a key modulator of breast milk HMO composition.^59,60 Emerging evidence demonstrates that lactogenic diets enriched with fiber, probiotics, and prebiotics significantly enhance HMOs biosynthesis.⁶¹ The gut‒mammary axis may be an important mechanistic pathway linking maternal nutrition to HMOs composition. Specifically, maternal consumption of fruits, vegetables, and whole grains may promote microbial fermentation of dietary fiber in the gut, generating metabolites that stimulate HMO production through cross-talk between the maternal microbiota and mammary gland metabolism.⁶¹ However, while the gut‒mammary axis represents a compelling framework, its precise function in humans, particularly regarding the role of specific microbial metabolites and host signaling pathways, remains only partially characterized and requires further broad-scale investigations.

The role of micronutrients as essential cofactors in HMO enzymatic synthesis is strongly supported. Evidence confirms positive correlations between maternal Vitamin A/C intake and 3-FL concentrations, while B-complex vitamin levels (B₁ and B₂) are predictive of elevated 2ʹ-FL or total 2ʹ-FL and 3-FL levels.⁶¹ Biochemical studies indicate that these vitamins facilitate glycosyltransferase activity through NAD⁺/FAD-dependent redox reactions and provision of nucleotide-sugar substrates.^62,63 Notably, zinc and potassium exhibit dose-dependent associations with 2ʹ-FL abundance, potentially stabilizing enzyme conformations or modulating glycosylation kinetics through metal-ion coordination in catalytic domains.⁶⁴ While vitamin C is well-known as an antioxidant, it may also support mammary epithelial metabolic function conducive to HMO synthesis.^65,66 The potential involvement of metal ions (such as K⁺, Mg²⁺, Ca²⁺, Zn²⁺) as enzymatic cofactors ⁶⁷ and indirect regulators via cellular redox state modulation remains plausible but requires further mechanistic characterization.⁶⁸ Importantly, while these correlations and biochemical roles are increasingly clear, translating this knowledge into precise dietary recommendations is challenging. Factors such as bioavailability, interactions between micronutrients, baseline maternal nutrient status, and genetic polymorphisms in nutrient transporters or HMO-synthetic enzymes significantly complicate predictive models of HMO outcomes based solely on intake. Clinical studies indicate that lactose intake from dairy products can predict LNnT and LNT concentrations in breast milk, as lactose serves as both a structural backbone and energy substrate for HMOs assembly in mammary epithelial cells.⁶¹ This nutrient‒HMO relationship is further modulated by dietary fat‒protein‒carbohydrate ratios, with balanced macronutrient profiles increasing HMOs diversity.^64,69 This underscores the complexity of dietary influence: single nutrients or food groups interact within the broader context of the overall diet, gut ecology, and maternal metabolism, making simplistic “silver bullet” dietary interventions unlikely to be effective. Preliminary evidence suggests that consuming fucoidan-rich foods (e.g., seaweed, legumes) may upregulate mammary FUT2/3 gene expression, potentially enhancing fucosylated HMO production, but further validation is required.⁷⁰ Furthermore, nicotinic acid supplementation shows a striking dose-response relationship, with each unit increase in niacin intake increasing 2ʹ-FL concentrations by 31.355 nmol/mL.⁶⁴ While this quantitative effect size is compelling, critical questions remain regarding its physiological significance for infant health outcomes, long-term safety of high-dose supplementation, and whether similar effects occur with dietary niacin sources compared to supplements.

2.5. Environmental factors

The composition of breast milk HMOs exhibits marked geographical heterogeneity, shaped by region-specific dietary patterns, environmental exposures, and population genetics.⁴⁷ Distinct HMO profiles have been observed across populations, with North American and Scandinavian cohorts containing high proportions of secretors (up to 95% in the US, 79%–87% in Sweden and Finland),^51,71 contrasting sharply with relatively low proportions of secretors in African (50%–80%),^71,72 Central Asian (40%–50%), and Pacific regions (some island ethnic groups as low as 30% - 40%).¹⁶ Even within secretor groups, production of HMOs can differ; Chinese and Dutch populations demonstrate secretor subgroup stratifications differentiated by signature HMO biomarkers including 2ʹ-FL, DFL, and LNFP-I in Chinese cohorts and 2ʹ-FL, LNT, and fucosylated lacto-N-octaose (F-LNO) in Dutch populations.⁷³ Fine scale differences in HMO production have also been observed, whereby Brazilian maternal HMO profiles exhibit individualized patterns independent of Se/Le phenotypes, correlating with allergic history, postpartum timing, and infant anthropometric parameters.⁷⁴ Such regional differences may arise from complex interactions between localized dietary customs and epigenetic adaptations. The environmental factors affecting HMOs may also be related to the degree of greening of the residential area, which is positively correlated with HMOs diversity. Studies have found that higher vegetation coverage is associated with increased concentrations of α-1-2-fucosylated HMOs and sialylated HMOs, which may be related to reducing maternal stress or improving air quality in a green environment.⁷⁵ However, defining the specific contribution of individual environmental variables remains a major challenge because of the inherent collinearity between factors such as diet, microbial exposure, pollution levels, and sociocultural habits in distinct geographical settings.

2.6. Maternal factors

2.6.1. Maternal pathophysiological state

Clinical conditions and pharmacological interventions significantly perturb the homeostatic regulation of HMO production. Emerging evidence implicates maternal glycemic control as a novel modulator, with hyperglycemia potentially interfering with lactose synthase complex formation.⁷⁶ Diabetic mothers show attenuated HMO synthesis, accompanied by altered fucosylation and sialylation profiles.⁷⁷ While these associations have been observed, a critical research gap exists in understanding the specific molecular mechanisms by which hyperglycemia disrupts HMOs synthesis (e.g., through altered nucleotide‒sugar availability, redox imbalance, or direct enzyme inhibition) and whether optimal glycemic control postdiagnosis can restore HMO profiles. Pharmacometabolomic studies further demonstrated that perinatal antibiotic exposure reduces nonfucosylated HMO production by 38%–42% compared to controls, suggesting microbiota-mediated regulation of glycosyltransferase activity.⁷⁸ These findings underscore the “gut‒mammary axis” concept, but crucially, it is unknown whether the effect is driven by direct microbial depletion, shifts in microbial metabolite profiles, or systemic inflammation. Therefore, there is a delicate equilibrium between maternal physiology and HMOs assembly, where systemic metabolic disturbances lead to alterations in milk composition. However, disentangling the direct effects of the pathophysiology itself from the confounding impacts of medications often used to manage these conditions (e.g., metformin, insulin, and antibiotics) remains a significant methodological challenge in human studies.

2.6.2. Psychological factors

Maternal psychological status exerts neuroendocrine regulation on HMO composition, and recent evidence suggests a direct correlation between maternal mental health and HMO biosynthesis.⁷⁹ Negative affective states and chronic stress activate the hypothalamic‒pituitary‒adrenal axis, triggering glucocorticoid-mediated interference with prolactin pulsatility and oxytocin secretion,⁸⁰ as well as disruption of oxytocinergic and prolactinergic pathways,^81,82 speculation may impair both the quantitative output and qualitative diversity of HMOs. This perturbation originates from stress-induced suppression of oxytocin receptors in mesocorticolimbic reward circuits and dopaminergic inhibition—critical mechanisms governing lactation physiology.⁸³ The pathophysiological cascade involves estrogen withdrawal-induced neurotransmitter imbalance, dysregulation of neuropeptide Y signaling, and altered vagal nerve tone, which all converge to disrupt mammary gland function.^84,85 While this proposed neuroendocrine cascade linking stress to mammary dysfunction is biologically plausible, demonstrating that direct causal links from a specific maternal psychological state through this cascade to quantifiable changes in specific HMO structures in humans remain complex. In a clinical study, significant variations in HMOs concentrations were observed in mothers experiencing psychological distress.⁷⁹ Differential levels of LNFP-III and lacto-N-hexaose (LNH) were found among secretors exhibiting depressive symptoms, while concentrations of DFL, LNFP-III, and disialyllacto-N-hexaose (DSLNH) were altered in Se mothers under stress conditions.⁷⁹ Notably, postpartum depression manifests as a clinical syndrome of this dysregulation, characterized by attenuated breastfeeding self-efficacy and disrupted maternal–infant neurohormonal synchrony.⁸⁶ We speculate that the observed correlation between depression and altered HMOs could be bidirectional: depression impacts HMOs via neuroendocrine pathways, while suboptimal HMO profiles might influence infant behavior and gut–brain axis signaling, potentially exacerbating maternal stress and depressive symptoms, creating a reinforcing cycle.

Psychosocial dimensions also modulate HMO profiles through behavioral–neurobiological interfaces. Maternal–infant attachment quality, body image perception, and developmental life-stage conflicts constitute key psychosocial determinants.⁸⁷ These factors influence breastfeeding frequency and intensity through dual pathways: (1) neuroendocrine modulation of milk ejection reflexes and (2) behavioral patterns affecting mammary stimulation. The objectification of lactating bodies in sociocultural contexts further exacerbates psychological barriers to optimal breastfeeding practices, indirectly altering HMO trajectories.^88,89 This highlights the importance of sociocultural context; however, quantifying its specific contribution versus individual psychological factors is extremely challenging, and cultural variations may significantly influence these dynamics.

Psychologically supportive interventions may directly and specifically modify HMOs composition to a clinically significant degree for infants, but outcomes are still emerging. Future research must prioritize rigorous intervention trials, such as mindfulness-based stress reduction and targeted psychological support programs with longitudinal HMO profiling, to test causality and effect sizes directly. Furthermore, translating biopsychosocial models into scalable, culturally adapted public health interventions poses significant practical and resource-related challenges that need systematic evaluation.

3. The maternal microbiome: as a key determinant of HMO profiles

Alongside a range of genetic and external factors involved in modulating HMO production, recent evidence suggests that internal factors can affect HMO production – one key factor being the maternal microbiome. Maternal microbial communities, particularly within breast milk and the gut, act as active modulators of the HMO composition secreted in breast milk. We examine how the maternal microbiome profoundly shapes the quantity, diversity, and structural characteristics of HMOs, highlighting its role as a core determinant of infant nutritional exposure.

3.1. Breast milk microbiome modulates HMO composition

The structural specificity of HMOs governs microbial selection and metabolic crosstalk in breast milk. Fucosylated HMOs (e.g., 2ʹ-FL and 3-FL) serve as selective substrates for Bifidobacterium longum subsp. infantis and B. bifidum, while remaining indigestible to most Lactobacillus strains and opportunistic pathogens.⁹⁰ This substrate specificity creates specific proliferation of Bifidobacterium, with elevated HMOs concentrations directly correlated with Bifidobacterium dominance.⁹¹ Sialylated HMOs (e.g., 3ʹ-SL, 6ʹ-SL) exhibit distinct metabolic outcomes, generating reduced lactic acid yields compared to neutral HMOs, thereby modulating microbial community dynamics.⁹² Mechanistically, fucosyltransferase and sialidase gene expression in Akkermansia muciniphila CSUN-19 and Bifidobacterium sp. align with their capacity to catabolize complex fucosylated/sialylated HMOs, a trait linked to α-fucosidase and β-galactosidase copy number variations.^93,94 However, this mechanistic understanding is largely derived from in vitro models and targeted genomic analyses of specific isolates. A lack of knowledge exists regarding the characterization of the in vivo functional meta-transcriptome and metabolome within the milk microbiome in response to HMO gradients. In the complex microbial ecosystem of human breast milk, interactions between multiple microbial taxa may modify individual strain responses, leading to conclusions that are completely different from those of in vitro experiments.

HMO‒microbiota interactions exhibit temporal specificity across lactation stages. Dynamic shifts in the breast milk microbiota occur as the lactation stage develops, with colostrum and early transitional milk (≤8 d postpartum) exhibiting reduced Chao1 and Shannon indices dominated by Blautia and mature milk (≥14 d) containing increased Ruminococcus, Dorea, and Escherichia‒Shigella relative abundances.⁹⁵ Notably, Bifidobacterium pseudocatenulatum and Bifidobacterium​​​​​ pseudolongum dominate the transitional phases, while Bifidobacterium ruminantium and Lactobacillus mucosae persist across all lactation periods.⁹⁵ These temporal patterns highlight how HMO composition can be a deterministic factor in microbial succession, selectively enriching taxa equipped with specialized HMO utilization systems while suppressing pathobionts such as Klebsiella pneumoniae through substrate exclusion.⁹⁶ It is worth noting that interpersonal variation in breast milk microbial ecology, potentially driven by maternal genetics, diet, health status, and geographic factors, complicates universal generalizations about HMOs–microbiome temporal dynamics and their functional outcomes for infants.

The synergistic interplay between HMOs and the breast milk-derived microbiota establishes a biochemical foundation for infant immunity and metabolic programming. Through competitive exclusion, vitamin synthesis, and mucosal barrier reinforcement, bifidobacteria and lactobacilli convert HMOs into bioactive metabolites that regulate host‒microbe interactions.⁹⁷ These coevolved mechanisms not only optimize nutrient bioavailability but also epitomize the functional integration of biochemical and microbiological components. Nevertheless, translating this foundational knowledge into actionable interventions, such as precise probiotics and prebiotics tailored to maternal HMO profiles, requires a deeper understanding of how interpersonal variations in the milk microbiome impact infant health outcomes.

3.2. Maternal gut microbiomes regulate HMO biosynthesis and secretion

The maternal gut microbiota exerts regulatory effects on HMO composition through microbial-derived signaling molecules. Gut microbiota-generated signaling mediators, including cytokines, microbial metabolites, and immunomodulatory compounds,⁹⁸ enter systemic circulation and translocate to mammary tissue where they modulate HMO biosynthesis and secretion.⁹⁹ Current evidence suggests that sialic acid and probiotic intervention may regulate the gut microbiota and 6ʹ-SL content in the breast milk of maternal mice through the GPR41-PI3K-Akt-PPAR pathway.¹⁰⁰ However, the precise mechanisms governing the translocation, stability, and mammary tissue targeting of these complex mediators remain largely hypothetical. Determining whether specific microbial metabolites reach the mammary epithelium at biologically relevant concentrations and how they cross cellular barriers to engage intracellular signaling networks regulating glycosylation requires sophisticated pharmacokinetic studies and tissue-specific metabolite tracing.

The metabolic competence of the maternal gut microbiome fundamentally underpins HMO biosynthesis through substrate provision regulation. Recent evidence suggests that intestinal dysbiosis can disrupt nutrient catabolism, potentially compromising maternal nutritional status and indirectly impairing HMO synthesis.^98,101 Beyond maintaining intestinal homeostasis and energizing epithelial cells, these microbial metabolites circulate systemically to mammary tissue, where they not only interact with lipid mediators to modify HMO structural diversity but also reprogram mammary cell metabolism through epigenetic modulation of glycosylation enzymes.^99,102-104 Although current evidence suggests that the systemic effects of these microbial metabolites are plausible, quantifying their mammary-specific bioavailability and establishing dose-responsive epigenetic reprogramming of specific glycosyltransferases in humans are critical next steps.

The maternal gut microbiome also affects immune-endocrine crosstalk that directly modulates HMO biosynthesis in mammary tissue. During lactation, leukocyte trafficking to mammary tissue dynamically adjusts HMO profiles.¹⁰⁵ Macrophages and lymphocytes not only provide pathogen surveillance but also secrete immunoregulatory molecules that fine-tune glycosyltransferase activity and adjust HMO production.¹⁰⁶ Notably, HMOs exhibit bidirectional immunomodulatory properties, while pro-inflammatory cytokines (TNF-α, IL-1β) suppress HMO synthase expression, specific HMOs structures mitigate inflammation via TNFR1 modulation and TLR4 antagonism.⁷⁶ Pathological dysregulation, as observed in mastitis, creates a feed-forward loop leading to microbial imbalance that increases Staphylococcus/Streptococcus dominance and elevates inflammatory cytokine levels (IFN-γ, IL-12).^107-109 HMOs can also mitigate TNF-α induced inflammation by interacting with TNFR1, thus modulating their own synthesis by regulating the inflammatory response.¹¹⁰ This bidirectional interplay presents a methodologically complex challenge, disentangling whether immune cell activity during periods of inflammation directly drives changes in HMO profiles, or whether altered HMO composition is an adaptive response to act locally as an anti-inflammatory agent.

Gut microbial-endocrine interactions represent a pivotal axis influencing HMOs profiles through systemic metabolic reprogramming. Specific gut bacteria, including Lactobacillus sp., Bifidobacterium sp., Escherichia coli sp., and Streptococcus sp., produce neuroendocrine analogs such as estrogen-mimetics and progesterone-like compounds.¹¹¹ These compounds can directly modulate signaling cascades in mammary epithelial cells.^112,113 Furthermore, the gut microbiome composition can impact mammary substrate availability for HMO biosynthesis. For example, Bacteroides facilitates calcium and magnesium absorption, while members of Bacillota drive carbohydrate metabolism.¹¹⁴ The FUT2-dependent secretory status establishes a distinct ecological niche within mammary tissue. Mothers who are nonsecretors exhibit reduced Bifidobacterium abundance and altered HMO-microbe coevolution patterns.¹¹⁵ However, it remains unresolved whether FUT2 status primarily shapes the maternal gut microbiota (thereby affecting HMO precursors/signals) or directly alters the mammary niche (influencing microbial colonization and local HMO utilization). While specific mechanisms are still fully characterized, it is evident that the microbiome can be a key regulator of mammary metabolic networks that govern HMO diversity.

4. Shaping the infant gut ecosystem: HMOs as microbial architects

The HMO profiles produced under maternal influence are not metabolic endpoints but rather bioactive molecules strategically deployed into the infant gut lumen and can therefore act as architects shaping the nascent neonatal gut microbiome. While research is still developing, recent evidence suggests that HMOs can create selective niches that favor beneficial symbionts, inhibit pathogens, and drive the establishment and coevolution of gut microbial communities critical for infant health and development.

4.1. HMOs utilization drives infant gut bacteria colonization

HMOs are key drivers of species-specific colonization in the infant gut, shaping distinct microbial community structures. A prime example is that Bifidobacterium, especially B. longum subsp. infantis dominates the microbiota of breastfed infants (Table 1), which is directly linked to its enhanced HMO utilization capacity.^116,117 This subspecies possesses a unique genomic arsenal encoding dedicated glycoside hydrolases and ABC transporters, enabling highly efficient metabolism of HMOs.¹¹⁸ Notably, comparative genomic analyses reveal specific HMO utilization gene clusters—including a pivotal 43 kb cluster in B. infantis—which dictate niche specialization by facilitating the near-complete catabolism of major HMO structures.¹¹⁹ The process of microbial utilization of oligosaccharides involves coordinated steps initiated by specialized membrane transport systems.¹²⁰ These systems selectively import intact HMOs or large fragments into the cell. Following uptake, complex HMO structures are degraded intracellularly. Some species first cleave HMOs extracellularly via membrane-bound glycoside hydrolases (e.g., sialidases, fucosidases, galactosidases) into smaller components. Subsequent intracellular hydrolysis releases constituent monosaccharides, including glucose, galactose, fucose, N-acetylglucosamine, and sialic acid.¹²¹ These monomers feed into central catabolic pathways: Bifidobacterium efficiently generates acetate, while glycolysis and specific pathways for fucose and sialic acid lead to pyruvate formation. Pyruvate fermentation then yields key short-chain fatty acids (SCFAs),¹²¹ including acetate (a primary product from Bifidobacterium and Bacteroides) and propionate (primarily from Bacteroides via succinate or fucose-derived propanediol pathways), alongside lactate, ethanol, and gases (Figure 2). Ultimately, the cascade of HMO metabolism promotes microbial growth and enriches the infant gut environment with bioactive SCFAs. This in turn increases the thickness of the intestinal mucus layer, promotes immune development, and influences systemic hormone metabolism through the bloodstream.¹²¹

Table 1.

Changes of gut microbiota in infants with different feeding methods at different months of age.

Infants' age in months	Main species of exclusively breastfed infants	Relative abundance in exclusively breastfed infants	Main species of exclusively formula fed infants	Relative abundance in exclusively formula fed infants	Reference
0–1 month	Bifidobacterium Bacteroides Staphylococcus Enterobacteriaceae Lactobacillus	35% 15% 15% 10% 5%	Bacteroides Enterobacteriaceae Bifidobacterium Lactobacillus Staphylococcus Clostridium difficile Veillonella	25% 15% 15% 15% 10% 5% 5%	122-127
2–3 months	Bifidobacterium Enterobacteriaceae Streptococcus Bacteroides Lactobacillus	45% 10% 5% 5% 5%	Bifidobacterium Streptococcus Enterobacteriaceae Bacteroides Lactobacillus Veillonella	30% 15% 15% 15% 10% 5%	123,127-130
4–6 months	Bifidobacterium Enterobacteriaceae Bacteroides Lactobacillus Streptococcus	60% 15% 5% 5% 5%	Enterobacteriaceae Bifidobacterium Streptococcus Bacteroides Lactobacillus Prevotella Veillonella	30% 20% 15% 15% 10% 5% 5%	124,128,129,131-137
7–12 months	Bifidobacterium Enterobacteriaceae Bacteroides Lactobacillus Streptococcus Clostridium	65% 15% 5% 5% 5% 3%	Enterobacteriaceae Bifidobacterium Bacteroides Streptococcus Clostridium Lactobacillus Veillonella	25% 20% 15% 15% 10% 5% 5%
>12 months	Bifidobacterium Enterobacteriaceae Bacteroidetes Lactobacillus Streptococcus Ruminococcaceae	70% 15% 5% 5% <5% <5%	Bifidobacterium Enterobacteriaceae Bacteroidetes Clostridium Streptococcus Veillonella Ruminococcaceae	20% 20% 15% 10% 10% 5% 5%	130,138-141

Figure 2.

Fate of human milk oligosaccharides in the infant gut. Following infant uptake of HMOs via human milk, complex HMO structures reach the infant gut and undergo microbial processing. Two major routes are shown: (A) certain species (e.g., B. longum subsp. infantis) import intact HMOs via bacterial uptake (transport) across the membrane and metabolize them through an intracellular digestion strategy; (B) other species (e.g., B. bifidum) initially performed an extracellular digestion strategy using membrane-bound glycoside hydrolases to cleave HMOs into smaller oligosaccharides or monosaccharides, after which these breakdown products are imported via bacterial transport and undergo further intracellular bacterial metabolism. In both routes, HMOs are ultimately hydrolyzed to monosaccharides that feed into central catabolic pathways. Bifidobacterium can efficiently generate acetate as a primary end-product, while glycolysis and dedicated pathways for fucose and sialic acid yield pyruvate. Subsequent fermentation of pyruvate produces short-chain fatty acids (SCFAs). This metabolic cascade fuels microbial growth and enriches the infant gut environment with bioactive SCFAs that support mucosal health and immune development.

The structural heterogeneity of HMO composition drives ecological succession in the developing gut microbiota through structure-specific growth modulation. In vitro coculture systems demonstrated that 6ʹ-SL selectively promotes B. longum subsp. infantis proliferation, whereas 2ʹ-FL and LNnT enhance cross-feeding symbiosis between bifidobacterial species.¹¹⁹ Recent metabolomic profiling revealed differential growth effects: Monofucosyl-lacto-N-hexaose-III (MFLNH-III) and monofucosyl-lacto-N-neohexaose (MFLNnH) stimulate B. dentium and B. bifidum growth, while LNnT, 6ʹ-SL and lacto-N-sialotetraose b (LST b) inhibited B. breve expansion in the gut.¹⁴² Sialylated oligosaccharides (3ʹ-SL/6ʹ-SL) distinctively enrich late-colonizing B. pseudocatenulatum ¹⁴³, demonstrating how HMOs heterogeneity actively choreographs microbial succession.

HMO‒mucin interactions alter glycocalyx viscosity and permeability, thereby modulating bacterial‒epithelial contact dynamics and spatial distribution patterns.^144,145 However, direct experimental evidence demonstrating altered mucus viscosity/permeability and consequent shifts in microbial biogeography specifically due to HMOs within the complex flow dynamics of the infant intestine is currently limited. Advanced imaging techniques or sophisticated gut-on-a-chip models incorporating infant mucus and microbiota are needed to validate this mechanism. Such dual-phase environmental engineering establishes biogeographical gradients that influence microbial colonization patterns and metabolic crosstalk within the gut ecosystem.

4.2. HMOs- mediate ecological conditioning via the breast milk microbial pool

HMOs shape the infant gut microbiota by modulating both the composition and function of maternal milk bacteria. Following birth, the previously sterile infant gastrointestinal tract undergoes rapid microbial colonization from maternal (feces, vagina, skin, oral cavity) and environmental sources.¹⁴⁶ Current metagenomic analyses identify Staphylococcus, Streptococcus, Enterococcus, Lactobacillus, Bifidobacterium and Leuconostoc as the dominant genera in the human milk microbiota, with their relative abundance directly influenced by HMO profiles.¹⁴⁷ This microbial transmission establishes a dynamic ecological network in which breast milk bacteria serve as a primary inoculum for the neonatal gut. The maternal gut microbiota contributes substantially to milk microbial composition through the gut‒mammary translocation pathway.¹⁴⁸ Bacterial translocation mechanisms involve dendritic cell-mediated transport of viable commensals from the intestinal lamina propria to the mammary glands during late gestation and lactation.^149,150 Consequently, HMOs not only influence the seeding microbes supplied to the infant but also likely do so by modulating a microbiome that already carries an imprint of the mother's own gut community, transferred through this sophisticated physiological route. This finding underscores the complex interdependence between maternal diet/metabolism, maternal gut microbiota, breast milk microbiota composition/function, and ultimately, infant gut colonization. Further clinical research is needed to fully elucidate how the immune imprinting and functional programming imparted by this HMO-guided microbial inoculum contributes to short-term colonization and long-term infant health.

Once consumed, the breast milk microbiota induces critical immune programming through direct interactions with infant immune cells. Bacterial stimulation triggers cytokine secretion by neonatal dendritic cells, facilitating the transition from intrauterine Th2-polarized immunity to balanced Th1/Th2 responses while expanding regulatory T-cell populations, such immunomodulatory processes increase intestinal barrier integrity and establish systemic immune tolerance, significantly reducing the risks of atopic and autoimmune disorders.¹⁵¹ Moreover, the specific bacterial composition of breast milk, particularly the occurrence of Lactobacillus and Bifidobacterium, is associated with a reduced risk of severe infant diseases, including a lowered incidence of Grade II or higher necrotizing enterocolitis (NEC) incidence, decreased mortality, reduced instance of late-onset sepsis, shorter hospital stays, and quicker achievement of full enteral nutrition.¹⁵² These clinical associations highlight the potential importance of the milk microbiota in infant health. However, establishing causality that specific milk bacteria directly confer these benefits is complex. These associations could reflect confounding factors, as mothers providing milk rich in beneficial bacteria may also have healthier lifestyles or diets. Furthermore, the observed benefits might also be mediated indirectly through the infant gut microbiota, which can be shaped by the consumed microbes and HMOs, rather than by the milk bacteria acting directly within the infant gut. Rigorous intervention studies (e.g., supplementing expressed milk with defined probiotics) are needed to test direct causality for specific outcomes such as NEC reduction.

5. Relationship between HMOs and infant health

While clinical links between HMOs and the establishment of basic microbial ecosystems are becoming evident, of greater significance is the role that this connection may play in determining infant health outcomes. Therefore, it is required to bridge the mechanistic insights of HMO‒microbiota interactions by examining the established and emerging causal links between HMO-driven microbial dynamics and identifying tangible benefits or protective effects for infants (Figure 3).

Figure 3.

Benefits of human milk oligosaccharides for infant health. HMOs play multifaceted roles in infant development. They can promote cognitive and neurological growth, strengthen defense mechanisms against pathogens and infections, reduce the risk of critical intestinal diseases, prevent allergic reactions, ensure appropriate weight regulation, and support robust bone development. HMOs are vital components of breast milk, acting through the gut microbes to holistically enhance infant health and well-being across multiple physiological systems.

5.1. HMOs are associated with infant weight gain

A significant correlation exists between HMO consumption and infant weight regulation, particularly in terms of early-life obesity and low birth weight associated with prematurity or malnutrition. Emerging evidence suggests that HMOs can mitigate obesity risks through infant gut microbiota regulation, a mechanism critical for metabolic programming.^153-156 Consumption of HMO-enriched breast milk by infants promotes gut colonization by beneficial gut microbes such as Bifidobacterium, which are diminished in obese infants.^154,157-159 Although establishing causal protection in humans remains challenging, the observed benefits could reflect broader advantages of breastfeeding (e.g., feeding patterns, bioactive factors beyond HMOs) or shared maternal/infant factors influencing both HMO profiles and metabolic health. Furthermore, the specific Bifidobacterium species and strains promoted through breastfeeding (e.g., B. infantis and B. breve), along with their functional metabolic outputs, may be crucial determinants of such protective effects, requiring deeper strain-level characterization.

HMOs are utilized by gut microbes to produce SCFAs,¹⁶⁰ which can regulate energy homeostasis, glucose metabolism and lipid storage. For instance, 2ʹ-FL and LNT selectively enhance Bifidobacterium proliferation in this way, which metabolizes HMOs into SCFAs such as acetate and butyrate.¹⁶¹ These SCFAs bind to G-protein-coupled receptors (GPR41/GPR43), activating pathways that regulate glucose tolerance, lipid oxidation, and adipocyte differentiation, thereby reducing fat accumulation.^162,163 Notably, preterm infants exposed to higher levels of LNnT and disialyllacto-N-tetraose (DSLNT) demonstrate accelerated weight gain, whereas elevated LNFP-II concentrations are negatively associated with growth rates.¹⁶⁴ This emphasizes the structure-dependent bioactivity of HMOs in the modulation of host‒microbe crosstalk and highlights a critical challenge in predicting the net metabolic outcomes of consuming a complex mixture of HMOs, which may exert antagonistic effects. Understanding how the overall HMO profile integrates these signals is therefore essential.

Preterm infants exhibit distinct gut microbiota colonization patterns, where HMOs serve as pivotal modulators of microbial succession and metabolic resilience. Maternal secretor status critically influences HMO profiles, with nonsecretor mothers linked to elevated intestinal Proteobacteria and reduced Firmicutes in their preterm offspring.¹⁶⁵ Such dysbiosis disrupts nutrient absorption and energy harvest, exacerbating faltering growth.^166-168 However, it raises the question of whether the observed dysbiosis is directly caused by the lack of specific fucosylated HMOs (e.g., 2ʹ-FL) or if non-secretor status is a marker for other genetic or environmental factors influencing both the maternal gut/milk environment and infant outcomes. Targeted supplementation studies in which specific fucosylated HMOs are added to infants of nonsecretor mothers are needed to test causality. Intervention studies have demonstrated that HMO supplementation, particularly when combined with galactooligosaccharides (GOSs), stabilizes B. longum dominance, enhances Lactobacillus colonization, improves intestinal barrier function and increases body weight in preterm infants.^169,170 These intervention results are promising, demonstrating the potential to enhance infant outcomes via supplements. However, achieving a consistent B. longum dominance and Lactobacillus enhancement across diverse preterm cohorts can be variable, suggesting that host factors (e.g., gestational age, antibiotic exposure, and feeding methods) significantly modulate the response. Other beneficial taxa promoted by HMOs, such as Faecalibaculum, enhance dietary fiber fermentation, amplifying SCFA-driven energy provision to intestinal epithelia and peripheral tissues.^171-173 Concurrently, HMO-regulated microbiota upregulate calcium, iron, and zinc absorption while modulating insulin and glucagon secretion, thereby regulating glucose homeostasis and lipogenesis.^174-177 Disruption of this regulatory axis through HMO deficiency or microbial imbalance–compromises nutrient utilization, leading to metabolic inefficiency and suboptimal growth trajectories.¹⁷⁸ Overall, the concept of a disrupted HMO-microbiota-metabolic axis provides a useful framework. However, future research must move beyond associations to establish causal links within this axis and identify thresholds or biomarkers for “deficiency” levels to guide targeted interventions for infants at risk of growth impairment.

5.2. Allergies

HMOs exhibit multifaceted protective effects against allergic disorders, including asthma, atopic dermatitis, and food allergies,¹⁷⁹ which have surged in prevalence among children in recent decades.^180,181 By modulating the gut microbiota composition, enhancing immune tolerance and fortifying intestinal barrier integrity, HMOs serve as critical bioactive agents in mitigating allergic pathogenesis. Experimental models have demonstrated that HMO-induced microbial shifts correlate with reduced IgE synthesis and eosinophil infiltration, which are hallmarks of allergic sensitization.¹⁸² In murine asthma models, oral administration of 2ʹ-FL and 6ʹ-SL reduces lung IL-4, IL-5, and IL-13 levels by 40%–60%, which is concomitant with elevated Clostridium abundance, illustrating a gut–lung axis modulated by HMOs.¹⁸¹ This evidence demonstrates the ability of HMOs to attenuate inflammation through microbiota-dependent pathways, providing a mechanistic basis for epidemiological associations between breastfeeding and reduced asthma risk. Clinical cohort analyses further revealed a threefold higher asthma risk in children who received antibiotics without breastfeeding, whereas no such association was observed in infants administered antibiotics during breastfeeding—a protective effect attributable to concurrent intake of fucosylated HMOs and B. infantis.¹⁸³ These clinical data reveal that the clinical value of HMOs lies not in isolated actions but in orchestrating host‒microbe symbiosis to mitigate environmental perturbations. The microbiota‒immune axis therefore emphasizes the role of HMOs in priming neonatal immunity toward tolerance rather than hypersensitivity. Beyond microbial modulation, HMOs directly obstruct allergic triggers by competitively inhibiting pathogen‒epithelial interactions.¹⁸⁴ As decoy receptors, fucosylated and sialylated HMOs, such as 2ʹ-FL and 6ʹ-SL, block the adhesion of Escherichia coli, Salmonella, and other enteropathogens to intestinal mucins, thereby reducing infection-driven inflammation that predisposes individuals to allergic responses.¹⁸⁵ This antiadhesive property not only limits microbial translocation but also minimizes antigenic overload, a critical factor in food allergy development. The well-established antiadhesive mechanism provides a direct pathway through which HMOs can reduce infection-driven inflammation—a known risk factor for allergic sensitization—thereby supporting their protective role against allergy development.

The maturation of the intestinal epithelial barrier represents a pivotal mechanism through which HMOs attenuate allergy risk. By upregulating tight junction proteins (e.g., occludin and zonula occludens-1) and stimulating goblet cell differentiation, HMOs reduce intestinal permeability, preventing allergen translocation across the epithelium.¹⁸⁶ These direct barrier-enhancing effects provide a fundamental physiological explanation for the allergy-protective properties of HMOs. 2ʹ-FL strengthens intestinal epithelial tight junction proteins and reduces allergen penetration.¹⁸⁷ Other HMOs, such as LNnT, can activate the metabolism of Bacteroides to produce indole-3-lactic acid, promote the differentiation of regulatory Treg cells, inhibit Th2 immune response, and ultimately alleviate milk protein allergy.¹⁸⁸ Targeted immunomodulation through defined microbial–metabolite interactions exemplifies the precision of HMO-mediated tolerance programming. Further, structural specialization clearly emerges: fucosylated and neutral nonfucosylated HMOs primarily support intestinal barrier integrity, whereas sialylated HMOs enhance mucosal and systemic immune responsiveness by promoting DC- and MO-mediated Th1 and Th17 responses and reducing Th2 responses.⁷ This structural diversification reflects evolutionary refinement, allowing HMO mixtures to provide comprehensive immune protection.

5.3. Mental development

HMOs exert multifaceted regulatory effects on infant mental development through interconnected biological pathways. HMOs demonstrate neurodevelopmental benefits by modulating gut–brain axis interactions, enhancing synaptic plasticity, and regulating neurochemical signaling systems.^186,189-191 The capacity to stimulate beneficial gut microbiota proliferation, followed by the generation of microbial metabolites that permeate systemic circulation and influence central nervous system maturation.¹⁹² This microbial-endocrine-neural network establishes a critical foundation for optimizing cognitive development during the early life window of neuroplasticity.

The neuroprotective properties of HMOs manifest through direct molecular interactions with developing neural circuits. Experimental models have demonstrated that sialylated HMOs components significantly enhance hippocampal long-term potentiation (LTP), the neurophysiological basis for memory consolidation, by strengthening synaptic efficacy and stabilizing neuronal connections.^193,194 Such enhancement of synaptic plasticity provides a fundamental explanation for the benefits of HMOs in cognitive function. In further support, administration of fucosylated and sialylated HMOs to different mammals, such as rats, mice, and piglets, has been observed to significantly enhances spatial memory and learning speed in behavioral tasks.¹⁹³ The reproducibility of these cognitive enhancements across species strengthens the biological plausibility of the direct neurodevelopmental role of HMOs. Crucially, rodent studies employing gene-edited models deficient in 6ʹ-SL biosynthesis reveal that early-life deprivation of sialylated HMOs leads to persistent cognitive deficits, including impaired working memory in T-maze tests and diminished spatial learning in the Barnes maze.¹⁹⁵ Complementary research in porcine models confirms that dietary supplementation with sialylated HMOs during critical developmental periods enhances hippocampal-dependent reference memory by 32% in spatial hole-board reversal tasks.¹⁹⁶ The magnitude of this improvement highlights the functional significance of adequate sialylated HMOs exposure for complex learning processes. Clinical analysis of breast milk samples from 76 mothers of extremely preterm infants at two weeks postpartum revealed significant associations between HMOs composition and infant neurodevelopment. Specifically, higher 3-FL levels were correlated with a reduced risk of neurodevelopmental impairment, while elevated concentrations of LST b were associated with improved cognitive and language composite scores.¹⁹⁷ This evidence strengthens the clinical evidence for the beneficial roles of sialylated HMOs in mediating neurodevelopmental trajectories, synaptic formation, and cognitive performance, thereby advancing translational medical insights in this field.

Molecular mechanisms underlying HMO-mediated neurodevelopment involve epigenetic regulation of neural gene networks. Cross-fostering experiments using 6ʹ-SL knockout murine models have demonstrated that HMO deprivation during lactation alters the expression of neurodevelopment genes in the prefrontal cortex, particularly those regulating axonal guidance and synaptic transmission.¹⁹⁸ These findings directly link HMO exposure to the epigenetic programming essential for proper neural circuit formation. Such genomic modifications correlate with measurable electrophysiological changes and metabolomic perturbations in circulating neuroactive compounds. Notably, sialylated HMOs increase cholinergic neurotransmission by increasing acetylcholine release and promoting survival of basal forebrain cholinergic neurons.^195,196 This direct biochemical pathway enhances a neurotransmitter system vital for attention, learning, memory, and offers a clear mechanism for the cognitive benefit provided by HMOs. Moreover, the improvement of memory function associated with HMO exposure may also involve its influence on brain structure and function,¹⁹⁹ promoting neuronal growth and synaptic connectivity,²⁰⁰ thereby enhancing the capacity for accurate information storage and retrieval.²⁰¹ Supporting gut‒brain axis involvement, rodent experiments have shown that oral HMO administration enhances conditioned reflexes and LTP via vagus nerve signaling.²⁰² The vagus nerve pathway provides a well-established neural conduit for gut-derived signals to rapidly modulate brain function and plasticity. In terms of cognitive ability, a lack of 6ʹ-SL in rodents has been linked with detrimental effects on attention, executive function and memory in adulthood.¹⁹⁸ The deficiency of 6ʹ-SL is associated with changes in the gut microbiota and serotonin system regulation.^198,203 Alterations in the serotonin system, which is crucial for mood regulation and cognitive flexibility, represent another plausible pathway linking HMO deficiency to impaired neurodevelopment. Further, a lack of 6ʹ-SL has also been linked to decreased expression of genes related to the development of the central nervous system in specific parts of the prefrontal cortex during specific developmental periods.¹⁹⁸ This spatiotemporal specificity in gene expression alterations reinforces the concept of critical windows during which HMOs exert their most profound neurodevelopmental effects.

5.4. Pathogenic infection

By selectively promoting colonization of beneficial Bifidobacterium species through nutrient competition and adhesion site occupation, HMOs establish a microbial equilibrium that suppresses pathogenic proliferation, thus effectively mitigating infection risks in three critical systems—gastrointestinal, respiratory, and urinary tracts—primarily.^36,204 This microbiota-centric defense forms a foundational barrier against enteric pathogens by leveraging ecological principles. Critically, the structural similarity of HMOs with epithelial cell receptors enables competitive inhibition of pathogen attachment to epithelial cells, with demonstrated efficacy against influenza viruses, rotavirus, HIV, norovirus, and multiple bacterial strains, including Salmonella and Escherichia coli ^205,206. Such receptor mimicry represents an evolutionarily refined strategy for broad-spectrum pathogen blockade at mucosal surfaces. Beyond competitive exclusion, HMOs exhibit direct antimicrobial activity through novel molecular interactions. Evidence suggests interference with pathogenic bacterial metabolic pathways and compromised membrane integrity, particularly against gram-positive species.²⁰⁶ This dual-action mechanism, which combines ecological niche occupation with direct antimicrobial effects, provides a robust defense against bacterial invasion. By extending protection beyond the gut, HMOs can significantly reduce uropathogenic E. coli invasion through urothelial cell stabilization and inflammatory pathway suppression.²⁰⁷ This demonstrated efficacy in the urinary tract emphasizes the systemic nature of HMO-mediated protection. Clinically, the reduction in respiratory tract infection risk observed in exclusively breastfed infants stems from this integrated defense strategy.²⁰⁸ Replication of this protective benefit in formulas supplemented with 2ʹ-FL and LNnT provides compelling clinical translational validation of the functional significance of these bioactive components.²⁰⁹

The role of the gut microbiome as a reservoir for pathogens further expands the significance of HMOs as protectants of infant health, as maintaining microbial equilibrium prevents pathogenic translocation across interconnected mucosal systems.²¹⁰ Colostrum-derived HMOs (C-HMOs) possess unique immunopriming properties critical for neonatal defense.²¹¹ Through precise modulation of fetal intestinal mucosa signaling pathways, these early-life HMOs establish immune networks that govern cellular communication and mucosal differentiation.^212-214 The capacity of HMOs to downregulate acute-phase inflammatory mediators (IL-1β, MCP-1/2) while upregulating tissue repair cytokines creates a balanced immunological microenvironment.^215-217 Such calibrated immune modulation optimizes host defense while minimizing collateral tissue damage. In particular, the 3ʹ-, 4ʹ-, and 6ʹ-galactosyllactose isomers demonstrate specific attenuation of virus-induced IL-8 responses, effectively calibrating mucosal immunity during critical developmental windows,²¹⁸ allowing the optimization of infection resistance while preventing inflammatory overactivation.²¹⁹ 3ʹ-SL and 6ʹ-SL can reduce the internalization of Pseudomonas aeruginosa by alveolar epithelial cells in a dose-dependent manner, suggesting that HMOs may have strong potential for preventing pulmonary infections.²²⁰ This structural specificity exemplifies the precision with which HMOs can regulate mucosal immunity during critical developmental windows. Collectively, HMOs–particularly C-HMOs–safeguard infant health by orchestrating microbiota regulation, immunomodulation, and mucosal barrier optimization, thereby enhancing resistance to infections while preventing detrimental inflammatory cascades.

5.5. Intestinal diseases

The capacity of HMOs to modulate the gut microbiota demonstrates their therapeutic potential to protect infants against multiple intestinal diseases, including NEC. Differences in the composition and concentration of maternal HMOs are closely related to the occurrence of neonatal NEC. Clinical studies have shown HMOs, particularly LNDFH-I, to be significantly lower in the breast milk of mothers whose infants developed NEC compared to those whose infants did not develop NEC.²²¹ Metagenomic analyses further revealed critical dysbiosis preceding NEC onset, marked by significantly diminished B. longum abundance alongside increased Enterobacter cloacae relative abundance.^222-224 Such microbial imbalance highlights the failure of early colonization resistance mechanisms in NEC pathogenesis. Further associations have been made between low DSLNT levels and disrupted microbiome maturation—specifically, hindering the transition to a Bifidobacterium-dominant community characteristic of healthy older infants.²²⁵ The role of DSLNT in guiding microbial succession positions it as a key regulator of gut ecosystem resilience. Beyond correlation, HMOs such as DSLNT exhibit substantial clinical utility as noninvasive biomarkers for identifying infants at elevated risk of NEC and for screening donated breast milk.^225,226 Furthermore, DSLNT and other HMOs may serve as natural templates for the development of novel NEC prevention strategies.^227,228

HMOs also confer direct preventative properties against intestinal diseases by enhancing intestinal barrier integrity and orchestrating immune responses in infants. HMOs, including 2ʹ-FL and 6ʹ-SL can prevent NEC in mouse and piglet models, and reduce NEC inflammation in human ileum by inhibiting toll like receptor 4 signaling and accelerating the regeneration of crypt cells.²²⁹ Observations of HMO combinations containing 2ʹ-FL, DFL, LNT, LNnT, 3ʹ-SL and 6ʹ-SL can have a positive effect on epithelial barrier function in vitro.^145,230 Such blends of HMOs can limit the cytokine induced transport of fluorescein isothiocyanate labeled dextran and/or the reduction of transepithelial resistance to increase the Teer value before the challenge of inflammation, indicating protection of the intestinal epithelial barrier to maintain intestinal health.¹²¹

5.6. Bone development

HMOs can enhance bone growth and mineralization by modulating osteoblast-osteoclast dynamics, gut microbiota composition, and immune-mediated pathways. A pivotal study in C57BL/6N mice demonstrated that high-dose supplementation (6361 mg/kg body weight) of 2ʹ-FL significantly increased femoral length, improved bone microarchitecture, and elevated serum calcium/phosphorus levels while upregulating osteogenic markers (e.g., osteocalcin) and suppressing osteoclast activity markers.²³¹ These structural and biochemical improvements confirm 2ʹ-FL's potent osteoanabolic properties. Mechanistically, HMOs exert dual regulatory control through Wnt/β-catenin pathway activation in osteoblasts coupled with cytokine-mediated inhibition of osteoclastogenesis, particularly via downregulation of RANKL and TNF-α expression.^232,233 These findings provide further evidence that mice inoculated with bacterial strains from stunted infants and fed sialylated bovine milk oligosaccharides (S-BMO) exhibited increased femoral trabecular bone volume and cortical thickness and reduced osteoclastogenesis alongside an altered Th2 response.²³⁴ These findings establish a novel gut‒bone axis in which HMOs can recalibrate bone remodeling through gut microbiome-mediated signaling. The discovery of this microbiota-dependent pathway significantly expands our understanding of nutritional skeletal regulation. Clinical evidence indicates that total HMOs intake is positively correlated with body length growth in preterm infants.²³⁵ Specifically, the concentrations of DFL, LNDFH-I, lacto-N-neodifucohexaose (LNnDFH), 2ʹ-FL, and 6ʹ-galactosyllactose (6ʹ-GL) are positively associated with infant growth parameters, whereas MFLNH III and 3-FL are negatively associated with these metrics.²³⁶ Importantly, while suggestive of cause-and-effect, such correlations do not definitively establish causality. Whether these specific HMOs directly regulate bone growth dynamics or merely serve as biomarkers reflecting other underlying processes requires further mechanistic investigation. Furthermore, the ultimate impact on bone development likely depends critically on the absolute concentration, the precise ratios between growth-promoting and inhibiting HMOs, and complex synergistic or antagonistic interactions with other bioactive components in human milk. This complexity necessitates caution when considering clinical translation. That is, indiscriminate enrichment of total HMOs intended to promote growth could inadvertently introduce significant quantities of 3-FL or MFLNH III, potentially leading to adverse effects. Consequently, these findings argue strongly against viewing HMO supplementation simplistically, and instead support a precise, individualized approach.

The gut microbiome serves as a fundamental mediator of HMOs-induced skeletal benefits. Through microbiota-derived metabolites, HMOs indirectly regulate bone metabolism via endocrine and paracrine signaling mechanisms.^237-239 This microbe‒endocrine axis provides a direct physiological link between gut ecology and skeletal homeostasis. Experimental evidence shows that propionate and butyrate induce osteoclast metabolic reprogramming by shifting energy production from oxidative phosphorylation to glycolysis, thereby suppressing TRAF6 and NFATc1 expression, which is essential for osteoclast differentiation.^240,241 Such metabolic intervention not only preserves bone mass but also demonstrates therapeutic potential against pathological bone loss, as evidenced by SCFA-mediated protection against postmenopausal and inflammation-induced osteoporosis in murine models.²⁴² The interaction between HMO-modified microbiota and skeletal health highlights the critical role of nutritional programming in early-life bone development.

6. Challenges and opportunities: Insights, crosstalk, and future directions

Despite advances in the understanding of HMO biology, significant challenges persist, chief among which is the difficulty in proving definitive causal links between specific HMO-mediated mechanisms and infant health outcomes in humans. Much of the evidence derives from animal models, preclinical models or correlative epidemiological studies, which cannot fully replicate the complexity of human development, especially regarding the myriads of confounding factors, such as genetics, maternal health, and environmental exposures. This is compounded by substantial methodological heterogeneity in HMO profiling and microbiome analysis across studies. Furthermore, the broad structural diversity of HMOs presents a fundamental challenge. Attributing biological effects of single oligosaccharides often oversimplifies their interactive effects within the complex environment of human breast milk. The interpersonal and temporal variability in HMO profiles, driven by factors beyond secretor status, such as diet, lactation stage, and geography, complicates generalization and necessitates personalized approaches. Proposed mechanisms involving intricate axes (e.g., the gut–mammary, gut–brain, gut-bone) frequently rely on plausible but incompletely characterized pathways in humans, such as the precise translocation of microbial metabolites to mammary tissue or the specificity of vagus nerve signaling in neurodevelopment. Crucially, evidence for long-term health impacts of early HMO exposure is sparse, requiring extended longitudinal cohorts.

The translational prospects for HMOs in infant nutrition hold great promise, with exciting pathways forward emerging. The supplementation of infant formula with specific HMOs, such as 2ʹ-FL and LNnT, shows potential for replicating certain benefits, including microbiota modulation or reduced infection risk. Long-term safety and efficacy data for supplemented containing HMOs formulas are still being gathered and analyzed, and optimal dosing strategies for diverse infant populations (e.g., preterm, nonsecretor infants) continue to be refined. While significant cost and scalability constraints remain challenges for large-scale complex HMO synthesis, promising advances in synthetic biology are steadily driving down costs and overcoming these barriers. Efforts to optimize maternal factors (e.g., diet and the microbiome) to enhance HMO profiles face additional challenges, including the development of reliable predictive models and the demonstration of clinically meaningful benefits in infants. Across diverse cohorts, cost-effective detection methods and clinical trials have been developed to actively validate the promising applications of specific HMOs, such as DSLNT, as both biomarkers for conditions such as NEC risk and novel therapeutic agents. Overcoming these limitations and continued progress through deeper mechanistic insights and robust clinical validation will unlock the immense potential of HMOs to revolutionize infant health and nutrition.

7. Conclusion

HMOs constitute a functionally vital and evolutionarily conserved component of breast milk, serving as key mediators between maternal physiology, the developing infant gut microbiome, and infant health outcomes. Research has established their fundamental roles as selective prebiotics that shape microbial colonization, modulators of intestinal barrier integrity and mucosal immunity, and contributors to defense against pathogens. The composition of HMOs is intricately governed by a myriad of factors, including maternal genetics, stage of lactation, maternal diet and metabolic health, environmental exposures, psychological state, and, crucially, the maternal microbiota itself via complex gut-mammary and potential endocrine axes.

However, realizing the full translational potential of this knowledge requires a precise assessment of current limitations and a strategic focus on unresolved questions. The field must go beyond the observational and correlative associations prevalent in human studies, as well as mechanistic insights derived from preclinical models, and move towards establishing robust causal links for specific HMO functions within the complex context of infant development. Future research should prioritize elucidating the precise molecular mechanisms underlying HMO actions, particularly concerning complex biological axes and structure-specific effects, utilizing advanced human-relevant models. Conducting well-designed and large-scale human intervention trials are paramount to confirm causal links between HMO profiles and infant health, evaluate the long-term safety and efficacy of HMO-supplemented formula (including optimal formulations for diverse populations), and test targeted strategies for modulating maternal HMO production.

Funding Statement

This work was supported by the earmarked funds from the Fundamental Research Funds for the Central Universities (SWU-KQ25008).

Disclosure of potential conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability statement

Data will be made available on request.

Abbreviations

2ʹ-FL

2ʹ-fucosyllactose

3ʹ-SL

3ʹ-sialyllactose

3-FL

3-fucosyllactose

6ʹ-GL

6ʹ-galactosyllactose

6ʹ-SL

6ʹ-sialyllactose

ABC

ATP-binding cassette

C-HMOs

colostrum-derived HMOs

DFL

difucosyllactose

DSLNH

disialyllacto-N-hexaose

DSLNT

disialyllacto-N-tetraose

F-LNO

fucosylated lacto-N-octaose

GOS

galactooligosaccharides

HMO

human milk oligosaccharides

LNDFH I

lacto-N-difucohexaose I

LNFP-I

lacto-N-fucopentaose I

LNFP-II

lacto-N-fucopentaose II

LNFP-III

lacto-N-fucopentaose III

LNnDFH

lacto-N-neodifucohexaose

LNnT

lacto-N-neotetraose

LST a

lacto-N-sialotetraose a

LST b

lacto-N-sialotetraose b

LST c

lacto-N-sialotetraose c

LNT

lacto-N-tetraose

Le

Lewis

LTP

long-term potentiation

MFLNH-III

monofucosyl-lacto-N-hexaose-III

MFLNnH

monofucosyl-lacto-N-neohexaose

NEC

necrotizing enterocolitis

Se

Secretor

SCFAs

short-chain fatty acids

S-BMO

sialylated bovine milk oligosaccharides

SNPs

single nucleotide polymorphisms