Prospecting Human Milk Oligosaccharides as a Defense Against Viral Infections

Abstract

In addition to providing maximal nutritional value for neonatal growth and development, human milk functions as an early defense mechanism against invading pathogens. Human milk oligosaccharides (HMOs), which are abundant in human milk, are a diverse group of heterogeneous carbohydrates with wide ranging protective effects. In addition to promoting the colonization of beneficial intestinal flora, HMOs serve as decoy receptors, effectively blocking the attachment of pathogenic bacteria and viruses. HMOs also function as bacteriostatic agents, inhibiting the growth of gram-positive bacteria. Based on this precedence, an emerging area in the field has focused on characterizing the antiviral properties of HMOs. Indeed, HMOs have been evaluated for their potential as antiviral agents, with many shown to possess activity against life-threatening infections. This targeted review provides insight to the known glycan-binding interactions between select HMOs and influenza, rotavirus, respiratory syncytial virus (RSV), human immunodeficiency virus (HIV), and norovirus. Additionally, we review the role of HMOs in preventing necrotizing enterocolitis (NEC), an intestinal disease linked to viral infections. We close with a discussion of what is known, broadly regarding human milk oligosaccharides and their interactions with coronaviruses.

Introduction

The past twenty-five years have seen the development of antiviral agents that not only inhibit the viral growth cycle selectively but do so without causing collateral damage to the host. Antivirals function by inhibiting the viral infection cycle through a number of distinct mechanisms, including obstructing entry, uncoating, interfering with receptor recognition, and disrupting viral protein and nucleic acid synthesis.¹ Indeed, a myriad of clinically approved drugs exist that treat auto-immune disease syndrome (AIDS),² human immunodeficiency virus (HIV), herpes simplex virus (HSV), influenza virus, and hepatitis B and C viruses (HBV and HCV, respectively).³ Not surprisingly, these drugs are plagued by the increasing emergence of resistance, limited therapeutic efficacy, and severe side effects.⁴ For emerging and/or neglected life-threatening viral diseases (e.g., Dengue virus and Zika virus), the lack of approved medications for clinical use represents a largely unmet need.^{5, 6}

In the early stages of an infection, viruses recognize the host’s blood group carbohydrates and other sialylated (substituted variants of neuraminic acid) glycoproteins as receptors. In theory, any carbohydrate that shares structural homology with these cell surface glycans can function as a receptor decoy for viral adhesions, preventing an early stage of infection. Indeed, human milk contains secretary blood group carbohydrates (e.g. fucosylated Lewis antigens and sialylated glycoproteins) that prevent viral infections.⁷ Human milk contains complex oligosaccharides (HMOs) that possess related functionality. Not surprisingly, HMOs have demonstrated the ability to protect infants against a number of viruses by serving as receptor decoys. In this targeted review, we argue the case for studying the antiviral properties of HMOs, a group of complex carbohydrates that possess both prebiotic and antimicrobial activity. HMOs also possess novel structural features (in comparison to common anti-viral agents), high efficacy, and lack the possibility of collateral damage to the host.

Human Milk Oligosaccharides

Human milk has long been considered the gold standard for infant nutrition, as most health experts, including the American Academy of Pediatrics, recommend exclusive breast feeding for the first six months of life.⁸ Breast milk not only contains essential nutrients for growth and development, but also is dynamic in its composition—continuously changing to meet the unique needs of the infant. Colostrum, the earliest milk produced by mammals, is a thick, yellow fluid rich in antibodies, proteins, and oligosaccharides.⁹ The mother’s body starts producing colostrum around mid-pregnancy and continues its secretion up to five days after birth. At this time, the body transitions to producing mature milk over an approximately 14-day window. 95% of the total energy supplied in mature breast milk exists in the form of milk fats and lactose.^{10, 11} The average macromolecular profile of one liter of human milk contains 9–12 g of proteins, 32–36 g of fats, 67–78 g of lactose, and 5–20 g of human milk oligosaccharides (HMOs).^{10, 12–14}

HMOs are the third largest macromolecule found in human milk.¹³ While present in the milk of most mammals, complex oligosaccharides such as HMOs are significantly more complex and abundant in primate milk. Indeed, the concentration of HMOs produced by the primate mammary gland are highest in the colostrum (ca. 20 g/L) and averages from 5 to 15 g/L in mature milk.^{13, 15, 16} To date, over 200 HMOs have been identified, ranging from simple derivatives of lactose containing 3 monosaccharides to complex polymers that incorporate upward of 20 monosaccharides.^17–19,20 The isolation of HMOs is a well-established stepwise procedure beginning with the removal of fats via centrifugation, followed by protein precipitation with ethanol. The abundant lactose component in the remaining mixture is hydrolyzed into its glucose and galactose monomers using the enzyme β-galactosidase. The individual HMOs are then purified using size-exclusion chromatography and can be further characterized. These various separation and profiling techniques include: high pH anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD), nuclear magnetic resonance (NMR) spectroscopy, reverse-phase high-performance liquid chromatography (RP-HPLC), hydrophilic interaction chromatography HPLC (HILIC), matrix assisted laser desorption ionization-time-of-flight-mass spectrometry (MALDI-TOF-MS), and permethylation followed by liquid chromatography coupled with high-resolution tandem mass spectrometry (LC-MS/MS).^{17, 21–24} While these methods have significantly enhanced the availability of HMOs for study, there is still a limited supply of them, therefore chemoenzymatic synthesis and microbial fermentation have been employed to synthesize a library of HMOs for research and supplementation into formula.^{25, 26}

Structurally, HMOs are composed of 5 monosaccharides: L-fucose (Fuc), D-glucose (Glc), D-galactose (Gal), N-acetylglucosamine (GlcNAc), and N-acetylneuraminic acid (Neu5Ac) (Figure 1).¹³ Lactose (Lac, Gal-β1-4-Glc) forms the reducing end of all HMOs (Figure 1A). To synthesize an HMO, lactose is elongated with lacto-N-biose (LNB; Galβ1-3GlcNAc) or N-acetyllactosamine (LacNAc; Galβ1-4GlcNAc) to generate a type I or type II chain respectively. Generally, LNB is used to terminate a growing HMO chain, while LacNAc is used as an elongation unit to extend the chain at O-3 or O-6 of lactose. Branching occurs when β1-6 linkages are installed (iso-HMOs), while linear chains occur through installation of β1-3 linkages (para-HMOs). While most women generally produce the same “core” oligosaccharides, Lewis blood group and secretor status will determine how Fuc and Neu5Ac are installed. HMOs can be fucosylated via α 1-2, α 1-3 or α 1-4 linkages (Figure 1B) or sialylated via α 2-3 or α 2-6 linkages. These modifications can occur on the lactose core or the elongated chain (Figure 1C). In general, the abundance of fucosylated HMOs is significantly higher (35–50%) than sialylated HMOs (10–15%). Non-fucosylated, neutral HMOs (40–55%) make up the remainder of HMOs (Figure 1A).^{27, 28} While a neonate consumes several grams of HMOs per day, they are non-nutritive. Indeed, only 3% of HMOs reach systemic circulation. HMOs resist digestion by both enzymatic hydrolysis and intestinal acidity. Thus, ca. 97% of HMOs accumulate in the distal small intestine and colon.²⁹ Previous work in the field has demonstrated that HMOs possess a number of biological functions including modulation of gut microbiota, inhibition of pathogenic adhesion, promotion of brain development, protection of the epithelial barrier, and stimulation of the immune response.³⁰

Figure 1:

The most common human milk oligosaccharides (HMOs) found in human milk and their biosynthesis. All HMOs are composed from five monosaccharides (Legend): glucose (Glc), galactose (Gal), N-acetyl glucosamine (GlcNAc), fucose (Fuc), and N-acetylneuraminic acid (Neu5Ac). (A) The basic blueprint for HMO biosynthesis with lactose forming the reducing end for all oligosaccharides. Lactose (Lac) can be elongated to with either lacto-N-biose (LNB) to form type I chains, or with N-acetyl lactosamine (LacNAc) to form type II chains. B) Representative fucosylated HMOs synthesized through the addition of fucose. C) Representative sialylated HMOs characterized by the addition of N-acetylneuraminic acid (Neu5Ac).

A primary function of HMOs is to serve as prebiotics, facilitating the establishment of beneficial flora in the infant’s gut microbiome. Accordingly, HMOs select for the growth of a small number of dominant beneficial species which suppress the growth of pathogenic bacteria. As a consequence, the microbial composition of the infant gut varies significantly between breastfed and formula-fed babies. Since infant formula does not replicate the composition of breast milk, and lacks HMOs, the microbiome of formula fed infants possesses greater diversity. In 1954 Kuhn and György reported that a combination of HMOs, which they termed the “bifidus factor”, promotes the growth of Bifidobacterium bifidum (previously classified as Lactobacillus bifidus).^{31, 32} Bifidobacterium fermentation of HMOs leads to the production of short-chain fatty acids (SCFAs; acetate, propionate, and butyrate). SCFAs protect the commensal flora against invading pathogens and inflammation by regulating gut pH and enhancing the immune system.^33–35

In general, the gut flora of a breastfed infant is dominated by Bacteroides, Lactobacillus, and Bifidobacterium spp. as these strains can metabolize HMOs.^{36, 37} Most Bifidobacterium spp are able to metabolize the principal HMOs present in breast milk, including lacto-N-tetraose (LNT) and 2’-fucosylactose (2’-FL).^{36, 38} Additional Bifidobacterium and Bacteroides spp., specifically B. bifidum and B. fragilis, use β-galactosidase, α-fucosidase, and sialidase enzymes to hydrolyze HMOs. Next, the released monosaccharides are metabolized.³⁶ Indeed, abundance of the probiotic Bifidobacterium spp. (Bifidobacterium longum subsp. infantis, Bacteroides fragilis, and Bacteroides vulgatus) that function under this mechanism is prominent in breastfed children.³⁹ While Bifidobacterium spp. still dominate the microbiota in formula fed babies, in general there is an overall greater diversity with an increase in Staphylococcus, Streptococcus, Enterococcus, and Clostridium species.^40–42 Low microbial in breastfed babies is a result of an abundance of beneficial Bifidobacterium and Lactobacillus species.⁴³

In addition to serving as prebiotics, our group has shown that HMOs possess bacteriostatic properties and antibiofilm.⁴⁴ Initial studies revealed that both homogeneous and heterogeneous HMOs govern bacterial growth and biofilm assembly.^45–49 We discovered that HMO extracts possess antibiofilm and antimicrobial properties against Streptococcus agalactiae (GBS), antimicrobial properties against the Gram-negative aerobe Acinetobacter baumannii, and antibiofilm properties against methicillin-resistant Staphylococcus aureus (MRSA). In a second-generation study, we observed that HMOs potentiate the actions of aminoglycosides, anti-folates, macrolides, lincosamides, and tetracyclines against GBS, S. aureus and A. baumannii.^{46, 50}

Virus Prevention or Treatment with HMOs

HMOs prevent the colonization of viral pathogens.⁵¹ The virulence of enteric viruses is dependent, in part, on the pathogen’s ability to adhere to epithelial surfaces (Figure 2A). There are two proposed mechanisms for how HMOs modulate viral pathogenesis. As HMOs share structural homology with epithelial cell surface glycans, they serve as soluble decoy receptors to prevent early cellular attachment (Figure 2B). Additionally, HMOs bind epithelial cell surface receptors to block viral adhesions (Figure 2C).⁵² The interactions between these two classes of molecules include the well-known carbohydrate - lectin interactions that are critical to the viral infection process as the viral cell surface is decorated with oligosaccharides that recognize lectins on human cells.⁵³

Figure 2:

Proposed mechanism of actions of how HMOs can prevent viral adhesion to gut epithelial cells. A) In the absence of HMOs, viruses recognize surface glycans that are necessary for pathogenic adhesion, the first step in establishing infection. B) HMOs resemble surface glycans, acting as soluble decoy receptors and blocking the attaching of viral pathogens to the epithelial cells. C) HMOs additionally can indirectly prevent viral adhesion through binding to the epithelial surface causing a structural change in the receptor.

The innate immune system possesses pattern recognition receptors that identify pathogens.⁵⁴ In a related process, viral surface lectins recognize human epithelial cell-surface glycans to identify hosts during an infection. HMOs have been explored to tactically protect against virus invasions by mimicking epithelial cell surface glycans. HMOs also possess immunomodulatory activity by reducing viral infection in influenza, rotavirus, respiratory syncytial virus (RSV), human immunodeficiency virus (HIV), norovirus, and the disease necrotizing enterocolitis (NEC), which can accompany viral infection (Figure 3).

Figure 3:

HMOs with known immunomodulatory activity against influenza, rotavirus, respiratory syncytial virus (RSV), human immunodeficiency virus (HIV), norovirus and necrotizing enterocolitis (NEC). With the exception of NEC, all of the viruses discussed in this review have known protection with 2’-fucosyllactose (2’-FL). Influenza and rotavirus additionally have shown reduced infection rates with 3’-sialyllactose (3’-SL), 6’-siallylactose (6’-SL), and Lacto-N-neotetraose (LNnT). 3-fucosyllactose (3-FL) reduces viral load in RSV, HIV, and norovirus. lacto-N-difucohexaose LNFH I, lacto-N-fucopentaose I (LNFP I), and lacto-N-fucopentaose III (LNFP III) all also known to inhibit viral binding. Only one HMO, disialyllacto-N-tetraose (DSLNT) has been shown to reduce NEC infection.

Influenza

Influenza belongs to the Orthomyxoviridae family of negative-sense, single-stranded RNA viruses. Negative-sense RNA must first be converted into positive-sense RNA by an RNA-dependent RNA polymerase before translation.⁵⁵ The influenza viruses are classified into types A, B, and C. Influenza A and B are principle causes of seasonal epidemics and are responsible for three to five million cases of acute respiratory infections per year.^56–58 Influenza C infection is rare and causes mild symptoms.⁵⁹ Influenza virions are spherical or filamentous in shape with an outer lipid membrane covered in glycoprotein spikes of haemagglutinin (HA) and neuraminidase (NA).⁶⁰ Both avian and human influenza viruses recognize oligosaccharides containing the sialylated galactose structural motif, where sialic acid is bound to galactose in an α2,6 or α2,3 linkage.⁶¹ HA functions by binding sialic acid which enables cell penetrance.⁶⁰ NA, also known as the receptor-destroying enzyme, cleaves cell-surface sialic acid residues, promoting viral budding and release.⁶⁰ The complex of HA, NA, and matrix protein M2 is currently the target of a number of antiviral drugs.⁶⁰ Within the viral core, the nuclear export protein (NEP) and ribonucleoprotein (RNP) complex are required elements of viral transcription and replication.⁶² Due to the fact that a number of HMOs contain the sialylated galactose sub-structure, it is not surprising that 3’-siallylactose (3’-SL) and 6’-siallyllactose (6’-SL) reduce influenza infection by binding to HA glycoprotein spikes.^63–65 Through related binding mechanisms, lacto-N-neotetraose (LNnT), which is non-sialylated, decreases viral load in the epithelium. Interestingly, 2’-FL has been shown to trigger the immune response in an influenza vaccination model.^65–67 While the role of 2′-FL-mediated protection is likely due to modulation of the microbiota, the relationship to vaccine efficacy remains unknown.

Rotavirus

Rotaviruses are non-enveloped double-stranded RNA viruses that cause 138 million cases of severe diarrhea, vomiting, and dehydration each year.^68–70 The genome of the triple-layered virus is composed of eleven segments that code for six structural viral proteins (VP1-4, 6, 7) and six non-structural proteins (NSP1-6).⁷¹ The three major structural proteins include two surface proteins, VP7 and VP4 (middle layer). VP4 is a spike protein responsible for cell attachment and membrane penetrance.⁷² Additionally, VP4 is proteolytically cleaved by trypsin, producing VP5* and VP8*, which results in higher rates of infection.⁷³

Both fucosylated and sialylated HMOs have been shown to reduce rotavirus infection by acting as decoy receptors to inhibit binding of intestinal histo-blood group antigens (HBGAs) to the viral proteins. 3’-SL and 6’-SL, in particular, reduce viral infectivity; however, the combination of the two HMOs has proven to be more effective in binding to VP8* in a porcine rotavirus model.⁷⁴ The most abundant HMO found in human milk globally, 2’-FL, causes a significant decrease in infectivity when used as a therapy after appearance of viral symptoms.⁷⁴ An additional non-fucosylated HMO, LNnT, while failing to reduce viral load in vitro, surprisingly reduced rotavirus infection in an in vivo model by binding to VP8*.⁷⁵

Respiratory syncytial virus

Respiratory syncytial virus (RSV) is a leading cause of respiratory tract and lung infections in children.⁷⁶ In 2015, it was estimated that the virus resulted in 33.1 million cases and 59,600 deaths worldwide, with children under six months of age making up 1.4 million cases and 27,300 deaths.⁷⁷ RSV belongs to the Paramyxoviridae family of single-stranded negative-sense RNA viruses, which includes measles, mumps, Nipah virus, and Hendra virus.⁷⁸ Structurally, RSV is an enveloped sphere consisting of ten open reading frames and eleven viral proteins of which nine are structural and two are nonstructural.^{79, 80} The three proteins that are required for replication and to protect the ribonucleoprotein (RNP) core are the nucleoprotein (N), phosphoprotein (P), and the RNA-dependent RNA polymerase large protein (L).⁸⁰ The fusion protein (F), attachment glycoprotein (G), and short hydrophobic protein (SH) span the membrane, and the F and G proteins are integral for attachment and infection initiation.⁸¹ The two membrane-associated proteins (M2-1 and M2-2) are involved in improving the efficiency of transcription and replication.^{80, 82} Both 2’-FL and 3-FL bind to glycoprotein G, reducing the RSV viral load in airway epithelial cells.⁸³

Human Immunodeficiency Virus

Human Immunodeficiency Virus (HIV), which can lead to acquired immunodeficiency syndrome (AIDS), belongs to the Retroviridae family of pathogens that attack the immune system. The virus is classified as either HIV-1, which is more widespread and aggressive, or HIV-2, which is found mostly in western Africa and is associated with lower pathogenicity.⁸⁴ This two-stranded RNA virus contains 9 open reading frames and 15 viral proteins.⁸⁵ The lipid bilayer encloses the RNA core, which is protected by the protein capsid. Two glycoproteins that are embedded in the envelope, gp120 and gp41, are necessary for attachment to the cluster of differentiation 4 (CD4) receptor on the host cell and viral fusion to mediate cell penetration.⁸⁶ The matrix Gag protein p17, surrounds the capsid, plays a crucial role in replication, and anchors both gp120 and gp41 to the envelope.^{87, 88} Within the viral core, the major structural protein p24 has recently become a target for HIV vaccines as it is found in high abundance in the blood of infected patients.⁸⁹ The mechanism in which HIV gains viral entry across the infant’s mucosal barrier is through binding of the receptor, dendritic cell-specific ICAM-3 grabbing non-integrin (DC-SIGN) to the high mannose containing glycans on HIV envelope glycoprotein, gp120. This initiates infection in the CD4+ T lymphocytes of the host.^{90, 91} There is a correlation between high concentrations of HMOs present in breast milk to lower risk of transmission of HIV to the infant through breastfeeding.⁹² Carbohydrates in general, but specifically the Lewis blood group antigens found in HMOs, have been shown to inhibit binding of HIV to DC-SIGN.^{92, 93} DC-SIGN is highly reactive towards the fucose Lewis antigens Le^a,b,x,y, including 2’-FL and 3-FL.⁹³

Norovirus

Norovirus is a highly contagious virus that causes gastroenteritis along with some symptoms such as vomiting, diarrhea, and nausea. According to the CDC, norovirus is responsible for 685 million cases worldwide, with 200 million of those cases occurring in children under five.⁹⁴ Noroviruses belong to the Caliciviridae family which also include additional positive single-stranded RNA viruses associated with respiratory disease, hemorrhages, and gastroenteritis.⁹⁵ The norovirus genome encodes three open reading frames (ORFs). ORF1 encodes the six nonstructural proteins, while ORF2 and ORF3 encode the two structural proteins, VP1 and VP2, that form the major and minor capsid proteins.^{96, 97} Structurally, VP1 is composed of 90 dimers which are further separated into a shell domain (S domain) and a protruding domain (P domain). These domains are connected by a flexible hinge linker, 10 to 14 amino acids in length.^{98, 99} The β-barrel folded P domain is composed of a highly variable P2 subdomain extending out from the P1 subdomain.¹⁰⁰ Since P2 is located on the distal surface, it promotes immune recognition as well as contains the receptor binding site.^{100, 101} Noroviruses, like rotaviruses, recognize histo-blood group antigens (HGBAs) found in the saliva or protective mucosa of the digestive tract - leading to increased infection rates. Based on X-ray crystallography analysis, both 2’-FL and 3-FL bind to a pocket at the top of the P1 domain, acting as a decoy as they structurally mimic HBGAs, effectively inhibit the virus from binding to HGBA.¹⁰² In another notable study, it was found that two fucosylated HMOs, lacto-N-fucopentaose III (LNFP III) and 2’-FL bind to one norovirus strain VA387.¹⁰³ In the same study, another strain of norovirus (Norwalk) bound to lacto-N-fucopentaose I (LNFP I ) and lacto-N-difucohexaose (LNDFH I) to effectively inhibit viral infection.¹⁰³

Necrotizing enterocolitis (NEC)

The five viruses described above are the most common viral pathogens tested for antiviral activity, however, another disease, necrotizing enterocolitis (NEC) is of importance when discussing HMO treatment options. The exact cause of NEC infections in still unclear but it has been associated with specific bacteria, viruses, and fungi. Rotavirus, norovirus, echovirus, astrovirus, enterovirus, cytomegalovirus, and coronavirus have all been implicated as possible causes of NEC.³⁴ NEC is a devastating intestinal disease that is most common in premature and very low birth weight infants. NEC is associated with a mortality rate between 20 and 40%, with a heightened risk associated with surgical intervention.^104–106 The disease is characterized by ulceration, intestinal inflammation, abdominal distention, hemorrhages, and bacterial overgrowth.^{107, 108} While a low diversity of HMOs in breast milk is associated with NEC development in infants, one specific HMO, disialyllacto-N-tetraose (DSLNT), has been found most effective in preventing NEC.^{109, 110} It was determined that the removal of one or both sialic acid residues of DSLNT caused those HMOs to have no effect on treating NEC.¹⁰⁹ The underlying mechanism of how DSLNT reduces the incidences remains to be elucidated, however, several receptor-mediated hypotheses remain in play. It is unlikely that that DSLNT is selectin-mediated since these transmembrane glycoproteins only bind to fucosylated glycans.⁶⁵ Other classes of lectins such as siglecs which bind sialylated HMOs, or galectins which bind to sulfated, sialylated, fucosylated, and lactose-containing HMOs, are more likely candidates for DSLNT targeting.^{65, 111, 112}

Coronaviruses.

Coronaviruses are a large family of viruses that are named for the crown-like spikes on their surface. There are four main sub-groupings of coronaviruses, known as alpha, beta, gamma, and delta. Coronaviruses typically cause mild to moderate upper-respiratory tract infections such as the common cold. There are hundreds of coronaviruses, most of which circulate among mammals such as bats and porcine. Occasionally coronaviruses jump to humans, a so-called spillover event, where they cause serious disease. Four of the seven known coronaviruses that sicken people cause only mild to moderate disease. These are 229E (alpha coronavirus), NL63 (alpha coronavirus), OC43 (beta coronavirus), and HKU1 (beta coronavirus).

Three coronaviruses have emerged from animal reservoirs over the past several decades that cause serious and often fatal disease. The first is a virus that causes severe acute respiratory syndrome (SARS) known as SARS coronavirus (SARS-CoV). The virus emerged in 2002 and disappeared by the end of 2004. SARS-CoV is thought to be an animal virus from a thus far, unknown animal reservoir, most likely a bat. The second virus is known as Middle East respiratory syndrome (MERS) and is caused by the MERS coronavirus (MERS-CoV). Transmitted from an animal reservoir in camels, MERS was identified in 2012 and continues to cause irregular, albeit contained outbreaks. The third novel coronavirus to emerge this century is known as SARS-CoV-2. It causes coronavirus disease 2019 (COVID-19), which emerged from China in late 2019 and was declared a global pandemic by the World Health Organization (WHO) in early 2020.

In the context of human milk, it is currently unclear whether severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is shed into breastmilk and transmitted to a baby through breastfeeding. Indeed, recent investigations of appreciable viral load in human milk - but sample sizes are small. ¹¹³ It is currently believed that pregnant people are at an increased risk for severe illness from COVID-19 compared to non-pregnant people. Additionally, pregnant people with COVID-19 might be at increased risk for other adverse outcomes, such as preterm birth. However, there is no appreciable evidence that COVID-19 is transmitted through breast milk. Whether or not HMOs are able to serve as receptor decoys or are able to bind the pathogen is currently unknown.

Future Directions

In the last two decades, humankind face six major viral threats – SARS (2002, 2020), swine flu (2009), MERS (2012), avian influenza (2013), and Ebola (2014). While the 2020 SARS pandemic has been catastrophic, it will assuredly not be the final contagion we face. Modern society functions such that disease spillover into humans is far more facile than in the past. Humankind must discover new tools to meet this challenge.

The first step in controlling interspecies transmission of viruses and increasing our ability to treat those that adversely affect human health and wellness, we must characterize the molecular interactions between viruses and their glycoconjugate receptors on host cells. Competitive inhibitors that mimic cell surface receptor glycans may hold the key to preventing viral entry. For example, the earliest stages of norovirus and rotavirus infection feature attachment to HBGAs. As described above, HMOs that share structural homology with HBGAs protect against rotavirus and norovirus infections. Moreover, HMOs target a conserved region of the capsid glycoprotein and prevent viral invasion, which means they are less susceptible to resistance evolution. One would thus expect these molecules to be of interest as an alternative therapeutic.

Before the antimicrobial activity of HMOs can be translated to prophylactic or therapeutic application, the community must elucidate the structural basis of viral inhibition, characterize the potency of HMOs, and define strain specificity. Similar to other carbohydrate-protein interactions, HMOs generally bind to viral capsid glycoproteins with low affinity. By comparison, viral adhesion to host cells is of high affinity due to multivalent interactions with cell surface glycan receptors. Thus, from a structural perspective, synthetic multivalent HMO constructs (similar to the tumor associated carbohydrate antigen vaccine constructs produced by the Danishefsky lab) are of great interest to our group and may serve as highly effective entry inhibitors. In addition to novel multivalent HMO based constructs, we believe that a second subset of HMOs worth exploring are mixed or multi-sialylated / fucosylated HMOs which incorporate several virus-recognizing epitopes in a single molecule.

The greatest roadblock to exploiting HMOs to prevent norovirus, rotavirus, and even influenza infections is their accessibility and general structural characterization. HMOs are a highly diverse group of compounds. Sadly, their benefits (which generally appear to be structure-dependent) cannot be readily attributed to a specific structure as most HMOs have not been characterized. Indeed, fewer than 10% of proposed HMOs have been evaluated for antiviral activity. Thus, advances to the chemical and chemoenzymatic synthesis of HMOs is necessary to achieve broad structure-activity studies.

HMO research has undergone significant advances since Moro and Tissier observed that breastfeeding governs the infant gut microbiome nearly 125 years ago. In the present day, the community’s dedication to characterizing the biological activity of several “small” HMOs has been validated by the first clinical trials in infants using either 2′-FL or 2`-FL/LNnT. These studies have revealed that HMO supplementation of formula is safe and produces a similar growth pattern as breast-fed infants. Thus, a number of exciting opportunities exists for future generations of human milk scientists. Within this context, our view of the future of HMO-based antivirals is optimistic and the first clinical trial of an HMO antimicrobial agent is likely on the horizon.