Synthesis of fucosyllactose using α-L-fucosidases GH29 from infant gut microbial metagenome.

Fucosyl-oligosaccharides (FUS) provide many health benefits to breastfed infants, but they are almost completely absent from bovine milk, which is the basis of infant formula. Therefore, there is a growing interest in the development of enzymatic transfucosylation strategies for the production of FUS. In this work, the α-L-fucosidases Fuc2358 and Fuc5372, previously isolated from the intestinal bacterial metagenome of breastfed infants, were used to synthesize fucosyllactose (FL) by transfucosylation reactions using p-nitrophenyl-α-L-fucopyranoside (pNP-Fuc) as donor and lactose as acceptor. Fuc2358 efficiently synthesized the major fucosylated human milk oligosaccharide (HMO) 2'-fucosyllactose (2'FL) with a 35% yield. Fuc2358 also produced the non-HMO FL isomer 3'-fucosyllactose (3'FL) and traces of non-reducing 1-fucosyllactose (1FL). Fuc5372 showed a lower transfucosylation activity compared to Fuc2358, producing several FL isomers, including 2'FL, 3'FL, and 1FL, with a higher proportion of 3'FL. Site-directed mutagenesis using rational design was performed to increase FUS yields in both α-L-fucosidases, based on structural models and sequence identity analysis. Mutants Fuc2358-F184H, Fuc2358-K286R, and Fuc5372-R230K showed a significantly higher ratio between 2'FL yields and hydrolyzed pNP-Fuc than their respective wild-type enzymes after 4 h of transfucosylation. The results with the Fuc2358-F184W and Fuc5372-W151F mutants showed that the residues F184 of Fuc2358 and W151 of Fuc5372 could have an effect on transfucosylation regioselectivity. Interestingly, phenylalanine increases the selectivity for α-1,2 linkages and tryptophan for α-1,3 linkages. These results give insight into the functionality of the active site amino acids in the transfucosylation activity of the GH29 α-L-fucosidases Fuc2358 and Fuc5372. KEY POINTS: Two α-L-fucosidases from infant gut bacterial microbiomes can fucosylate glycans Transfucosylation efficacy improved by tailored point-mutations in the active site F184 of Fuc2358 and W151 of Fuc5372 seem to steer transglycosylation regioselectivity.

Infant Gut Microbial Metagenome Mining of α-l-Fucosidases with Activity on Fucosylated Human Milk Oligosaccharides and Glycoconjugates.

The gastrointestinal microbiota members produce α-l-fucosidases that play key roles in mucosal, human milk, and dietary oligosaccharide assimilation. Here, 36 open reading frames (ORFs) coding for putative α-l-fucosidases belonging to glycosyl hydrolase family 29 (GH29) were identified through metagenome analysis of breast-fed infant fecal microbiome. Twenty-two of those ORFs showed a complete coding sequence with deduced amino acid sequences displaying the highest degree of identity with α-l-fucosidases from Bacteroides thetaiotaomicron, Bacteroides caccae, Phocaeicola vulgatus, Phocaeicola dorei, Ruminococcus gnavus, and Streptococcus parasanguinis. Based on sequence homology, 10 α-l-fucosidase genes were selected for substrate specificity characterization. The α-l-fucosidases Fuc18, Fuc19A, Fuc35B, Fuc39, and Fuc1584 showed hydrolytic activity on α1,3/4-linked fucose present in Lewis blood antigens and the human milk oligosaccharide (HMO) 3-fucosyllactose. In addition, Fuc1584 also hydrolyzed fucosyl-α-1,6-N-acetylglucosamine (6FN), a component of the core fucosylation of N-glycans. Fuc35A and Fuc193 showed activity on α1,2/3/4/6 linkages from H type-2, Lewis blood antigens, HMOs and 6FN. Fuc30 displayed activity only on α1,6-linked l-fucose, and Fuc5372 showed a preference for α1,2 linkages. Fuc2358 exhibited a broad substrate specificity releasing l-fucose from all the tested free histo-blood group antigens, HMOs, and 6FN. This latest enzyme also displayed activity in glycoconjugates carrying lacto-N-fucopentaose II (Lea) and lacto-N-fucopentaose III (Lex) and in the glycoprotein mucin. Fuc18, Fuc19A, and Fuc39 also removed l-fucose from neoglycoproteins and human α-1 acid glycoprotein. These results give insight into the great diversity of α-l-fucosidases from the infant gut microbiota, thus supporting the hypothesis that fucosylated glycans are crucial for shaping the newborn microbiota composition. IMPORTANCE α-l-Fucosyl residues are frequently present in many relevant glycans, such as human milk oligosaccharides (HMOs), histo-blood group antigens (HBGAs), and epitopes on cell surface glycoconjugate receptors. These fucosylated glycans are involved in a number of mammalian physiological processes, including adhesion of pathogens and immune responses. The modulation of l-fucose content in such processes may provide new insights and knowledge regarding molecular interactions and may help to devise new therapeutic strategies. Microbial α-l-fucosidases are exoglycosidases that remove α-l-fucosyl residues from free oligosaccharides and glycoconjugates and can be also used in transglycosylation reactions to synthesize oligosaccharides. In this work, α-l-fucosidases from the GH29 family were identified and characterized from the metagenome of fecal samples of breastfed infants. These enzymes showed different substrate specificities toward HMOs, HBGAs, naturally occurring glycoproteins, and neoglycoproteins. These novel glycosidase enzymes from the breast-fed infant gut microbiota, which resulted in a good source of α-l-fucosidases, have great biotechnological potential.

Infant-gut associated Bifidobacterium dentium strains utilize the galactose moiety and release lacto-N-triose from the human milk oligosaccharides lacto-N-tetraose and lacto-N-neotetraose.

Much evidence suggests a role for human milk oligosaccharides (HMOs) in establishing the infant microbiota in the large intestine, but the response of particular bacteria to individual HMOs is not well known. Here twelve bacterial strains belonging to the genera Bifidobacterium, Enterococcus, Limosilactobacillus, Lactobacillus, Lacticaseibacillus, Staphylococcus and Streptococcus were isolated from infant faeces and their growth was analyzed in the presence of the major HMOs, 2'-fucosyllactose (2'FL), 3-fucosyllactose (3FL), 2',3-difucosyllactose (DFL), lacto-N-tetraose (LNT) and lacto-N-neo-tetraose (LNnT), present in human milk. Only the isolated Bifidobacterium strains demonstrated the capability to utilize these HMOs as carbon sources. Bifidobacterium infantis Y538 efficiently consumed all tested HMOs. Contrarily, Bifidobacterium dentium strains Y510 and Y521 just metabolized LNT and LNnT. Both tetra-saccharides are hydrolyzed into galactose and lacto-N-triose (LNTII) by B. dentium. Interestingly, this species consumed only the galactose moiety during growth on LNT or LNnT, and excreted the LNTII moiety. Two ß-galactosidases were characterized from B. dentium Y510, Bdg42A showed the highest activity towards LNT, hydrolyzing it into galactose and LNTII, and Bdg2A towards lactose, degrading efficiently also 6'-galactopyranosyl-N-acetylglucosamine, N-acetyl-lactosamine and LNnT. The work presented here supports the hypothesis that HMOs are mainly metabolized by Bifidobacterium species in the infant gut.

Infant gut microbiota modulation by human milk disaccharides in humanized microbiome mice.

Human milk glycans present a unique diversity of structures that suggest different mechanisms by which they may affect the infant microbiome development. A humanized mouse model generated by infant fecal transplantation was utilized here to evaluate the impact of fucosyl-α1,3-GlcNAc (3FN), fucosyl-α1,6-GlcNAc, lacto-N-biose (LNB) and galacto-N-biose on the fecal microbiota and host-microbiota interactions. 16S rRNA amplicon sequencing showed that certain bacterial genera significantly increased (Ruminococcus and Oscillospira) or decreased (Eubacterium and Clostridium) in all disaccharide-supplemented groups. Interestingly, cluster analysis differentiates the consumption of fucosyl-oligosaccharides from galactosyl-oligosaccharides, highlighting the disappearance of Akkermansia genus in both fucosyl-oligosaccharides. An increment of the relative abundance of Coprococcus genus was only observed with 3FN. As well, LNB significantly increased the relative abundance of Bifidobacterium, whereas the absolute levels of this genus, as measured by quantitative real-time PCR, did not significantly increase. OTUs corresponding to the species Bifidobacterium longum, Bifidobacterium adolescentis and Ruminococcus gnavus were not present in the control after the 3-week intervention, but were shared among the donor and specific disaccharide groups, indicating that their survival is dependent on disaccharide supplementation. The 3FN-feeding group showed increased levels of butyrate and acetate in the colon, and decreased levels of serum HDL-cholesterol. 3FN also down-regulated the pro-inflammatory cytokine TNF-α and up-regulated the anti-inflammatory cytokines IL-10 and IL-13, and the Toll-like receptor 2 in the large intestine tissue. The present study revealed that the four disaccharides show efficacy in producing beneficial compositional shifts of the gut microbiota and in addition, the 3FN demonstrated physiological and immunomodulatory roles.

Isolation and Characterization of Nitrate-Reducing Bacteria as Potential Probiotics for Oral and Systemic Health.

Recent evidence indicates that the reduction of salivary nitrate by oral bacteria can contribute to prevent oral diseases, as well as increase systemic nitric oxide levels that can improve conditions such as hypertension and diabetes. The objective of the current manuscript was to isolate nitrate-reducing bacteria from the oral cavity of healthy donors and test their in vitro probiotic potential to increase the nitrate-reduction capacity (NRC) of oral communities. Sixty-two isolates were obtained from five different donors of which 53 were confirmed to be nitrate-reducers. Ten isolates were selected based on high NRC as well as high growth rates and low acidogenicity, all being Rothia species. The genomes of these ten isolates confirmed the presence of nitrate- and nitrite reductase genes, as well as lactate utilization genes, and the absence of antimicrobial resistance, mobile genetic elements and virulence genes. The pH at which most nitrate was reduced differed between strains. However, acidic pH 6 always stimulated the reduction of nitrite compared to neutral pH 7 or slightly alkaline pH 7.5 (p < 0.01). We tested the effect of six out of 10 isolates on in vitro oral biofilm development in the presence or absence of 6.5 mM nitrate. The integration of the isolates into in vitro communities was confirmed by Illumina sequencing. The NRC of the bacterial communities increased when adding the isolates compared to controls without isolates (p < 0.05). When adding nitrate (prebiotic treatment) or isolates in combination with nitrate (symbiotic treatment), a smaller decrease in pH derived from sugar metabolism was observed (p < 0.05), which for some symbiotic combinations appeared to be due to lactate consumption. Interestingly, there was a strong correlation between the NRC of oral communities and ammonia production even in the absence of nitrate (R = 0.814, p < 0.01), which indicates that bacteria involved in these processes are related. As observed in our study, individuals differ in their NRC. Thus, some may have direct benefits from nitrate as a prebiotic as their microbiota naturally reduces significant amounts, while others may benefit more from a symbiotic combination (nitrate + nitrate-reducing probiotic). Future clinical studies should test the effects of these treatments on oral and systemic health.