In this study lactobacilli were detected more frequently in breastfed than formula-fed 4 month-old infants in saliva and mucosal swab samples as we previously observed in a different population of infants . L. gasseri was the dominant Lactobacillus species detected, which was identified from 16S RNA gene sequences of isolates. Probiotic potential of L. gasseri was found to include growth inhibition of F. nucleatum, A. naeslundii, A. oris, S. sobrinus and C. albicans in addition to the previously reported S. mutans and S. sanguinis. Other characteristics of L. gasseri were inhibition of adhesion to hydroxyapatite in the presence of saliva, salivary gp40 and MUC7 suggesting possible mechanisms for probiotic activity.
The infants sampled were recruited from a randomized clinical trial of MFGM supplemented infant formula compared with a standard formula and breastfeeding. Compliance to the feeding regimens was acceptable according to diet records obtained from the parent study. Infants recruited into the parent study were between 0 and 2 months of age. The estimated intake of breast milk at study enrollment was similar in the standard formula and the MFGM formula groups. When infants were sampled at 4 months of age, they had been exposed to either formula or breast milk for two months [40, 41]. The lack of difference between the formula-fed groups suggests that this period might not have been long enough or that the different formulations do not induce changes in the oral microbiota. Previous studies, however, have observed that feeding mode, method of delivery, use of antibiotics and probiotic products may influence the oral and intestinal microbiota [2, 13, 40, 42]. We accounted for these possible confounders in the PLS analysis, and found they had only marginally influential for feeding group allocations and total lactobacilli counts.
L. gasseri was identified as the dominant Lactobacillus species in the oral cavities of the 4 month-old infants. This is consistent with previous studies on Lactobacillus detection in the oral cavity [13, 16] and the infant gut [43, 44]. L. gasseri is a member of the L. acidophilus complex, which includes L. acidophilus, Lactobacillus amylovorus, Lactobacillus crispatus, Lactobacillus gallinarum and Lactobacillus johnsonii. Strains belonging to the L. gasseri complex have been extensively studied for “probiotic” traits, including attachment to epithelial cells, growth inhibition, replacement or binding inhibition of pathogens and immunomodulation [46, 47]. L. gasseri strains from feces and human milk have been observed to (i) adhere to intestinal epithelial cells and intestinal mucus (mainly MUC2) [48, 49], (ii) produce bacteriocins [50, 51], (iii) reduce mutagenic enzymes in feces , (iv) stimulate macrophages and lymphocytes, (v) modulate the immune systems through the toll receptors  and (vi) show resistance to gastric and small intestine fluids . In the current report, salivary L. gasseri demonstrated several probiotic traits including: attachment to the human gingival epithelial cells HGEPp.05 and saliva, growth inhibition of several oral species and reduced attachment of the cariogenic S. mutans to saliva. Potential in vivo effects on the microbiota as well as short and long term biological processes remain to be demonstrated, but in vivo effects might be anticipated as we observed growth inhibition at L. gasseri concentrations as low as 103 CFU/mL, which are the levels reported for human milk [6, 16].
Studies have reported that breast milk contains L. gasseri, L. salivarius and L. fermentum, of which L. gasseri was the most prevalent species [15, 16], but the prevalence of L. gasseri detection has not been reported. We cultured Lactobacillus species, predominantly L. gasseri, from approximately one third of breastfed infants with lower to non-detectable levels from formula-fed infants. This is consistent with our previous rapport . Breast milk was not collected from the mothers, so we do not know whether detection of L. gasseri in infants reflects its presence in the mother’s milk. Other possible reasons for variability of L. gasseri detection in infants saliva include: individuality in adhesion site blocking on L. gasseri (presumably by saliva because L. gasseri aggregated in saliva but not in milk), and phenotypic host receptor variation. Few studies have examined host receptors for, and adhesion properties of, L. gasseri and lactobacilli in general . Binding of various lactobacilli species to saliva gp340 , peroxidase  and gastric and intestinal mucus [46, 48], blood group antigens and histone H3  has been reported. Most of these host receptors are heavily glycosylated and several carry blood group antigens [55, 56], which is consistent with the present findings of more avid binding of L. gasseri to submandibular/sublingual saliva, gp340, MUC7 and MFGM. Interestingly, it was reported recently  that the innate immunity peptide LL37, which has been detected in the mouth on epithelial cells and in submandibular/sublingual saliva , alters the surface of L. crispatus with a possible influence on its adhesive traits . Since gp340 and MUC7 (here identified as host receptors for L. gasseri binding) exist as polymorphic variants [34, 35], and phenotypic variation in gp340 relates to S. mutans adhesion avidity (gp340 here shown as shared host receptor for L. gasseri and S. mutans), it seems possible that phenotypic host receptor variation can influence L. gasseri colonization in breastfed infants. This would suggest that bacterial acquisition in infancy, and potential beneficial effects from probiotic products, may vary among individuals.
Pre-incubation of L. gasseri with saliva reduced detectable salivary gp340, and thus the observed S. mutans binding to gp340, suggesting that L. gasseri and S. mutans share a binding epitope in saliva. Competitive binding has previously been observed between S. mutans and other lactobacilli species with gp340 . L. gasseri strains have also been shown to compete with, displace, and inhibit the adhesion of the enteric pathogens Cronobacter sakazakii and Clostridium difficile to intestinal mucus . This suggests that L. gasseri may play a similar role in the oral cavity as has been observed in the gut. Although saliva from adults was used in the present study, gp340 has been detected in saliva in infants . Saliva has been shown to have a stable pattern of salivary proteins and glycoproteins from early infancy, with the exception of albumin and the mucins, with early dominance of MUC7 later followed by MUC5B .
Infants fed the MFGM supplemented formula tended to have higher oral levels of total lactobacilli and L. gasseri than infants fed a standard formula. This could reflect that MFGM provides a wide range of potential carbohydrate binding epitopes on glycoproteins and glycolipids, and that L. gasseri bound to purified MFGM coated on hydroxyapatite (present study). An increased content of MFGM supplementation could potentially foster acquisition of L. gasseri and/or other Lactobacillus species in the gastro-intestinal tract, but this concept needs further study.