Introduction
Gut microbiota composition and function have a vast influence on the developing immune system (Hansen et al., 2012b, 2013b; Weng and Walker, 2013). The immunological homeostasis is established right after birth in parallel with the initial colonization of the gastrointestinal tract by microorganisms from the birth canal, milk, mother, and the surrounding environment (Dominguez-Bello et al., 2010; Makino et al., 2013; Pannaraj et al., 2017). Oral tolerance toward commensal bacteria and food antigens is mediated through regulatory T-cells (Tregs) and their anti-inflammatory signals such as transforming growth factor-beta (TGF-β) and interleukin 10 (Il-10) (Chen et al., 2003; Round and Mazmanian, 2010; Geuking et al., 2011). Failure to establish a proper gut microbiota may lead to the skewing of the immune-microbial homeostasis such as suggested by Strachan (1989) and also observed in babies born by cesarean section (Dominguez-Bello et al., 2010; Hansen C. H. F. A. et al., 2014). This results in an imbalance between the effector and regulatory T-cells favoring a more reactive T-cell phenotype and increased risk of inflammatory disorders (Cani et al., 2008; Hansen et al., 2012b; Hansen A. K. et al., 2014). Colonization by the same bacteria before or after establishment of the immune-microbial homeostasis yields very different immunological reactions (Hansen et al., 2012b). Consequently, early life gut microbiome development may have long-lasting immunological effects beyond those of the immediate stimulation, and therefore it is possible to permanently influence the immune system through gut microbiota manipulation (Hansen et al., 2012b; El Aidy et al., 2013).
We have previously shown that especially members of the Clostridiales order and Verrucomicrobia phylum correlate positively with increased expression of the Forkhead box P3 (Foxp3) transcription factor, TGF-β, and IL-10, all of which are markers of Treg activity (Lindenberg et al., 2019a) indicating that the gut microbiota in horses can modulate immunity. In humans and mice both the Clostridiales spp. Faecalibacterium prausnitzii and the Verrucomicrobia spp. Akkermansia muciniphila are known to be very potent inducers of regulatory immunity and have been linked to prevention or amelioration of several inflammatory disorders (Atarashi et al., 2011; Hansen et al., 2012a; Qiu et al., 2013; Shin et al., 2014). The oligosachharides galacto- and fructo-oligosaccharides (GOS and FOS) have been shown to increase the abundance of Clostridiales spp. in the gut of newly weaned infants (Scholtens et al., 2006). Additionally, in a mouse model of human obesity and diabetes, mice fed fructooligosaccharides (FOS) had increased proportion of A. muciniphila and reduced obesity and inflammation (Everard et al., 2013). It is, therefore, reasonable to assume that horses would be able to increase their regulatory immunity, through oligosaccharide feeding increasing the presence of members of Clostridiales and Verrucomicrobia.
In recent years, various oligosaccharides have been used as pre-biotic food supplements, for both humans and animals, to promote the growth of anti-inflammatory bacteria species in the gut microbiota. Their therapeutic applications in relation to metabolic and inflammatory diseases such as type 2 diabetes (T2D) and obesity have been shown in mice and rats (Gobinath et al., 2010; Hansen et al., 2013a), and humans (Parnell and Reimer, 2009). Mannanoligosaccharides (MOS) has been shown to increase the colostrum levels of the immunoglobulins (Ig) IgA and IgG in mares (Spearman and Ott, 2004) and act as alternative binding sites for gastrointestinal pathogens in chickens (Spring et al., 2000; Baurhoo et al., 2007). High doses FOS have traditionally been used to induce a model of hindgut acidosis and laminitis in horses (van Eps and Pollitt, 2006; Bailey et al., 2009; Moreau et al., 2014). Recent studies showed that low doses of FOS supplementation had a stabilizing effect on the equine microbiota during sudden diet change and significantly decreased the basal plasma insulin level in insulin resistant horses (Respondek et al., 2008, 2011), suggesting a beneficial effect of low doses of FOS. It has been shown that neonatal IgA production in mice is regulated by the maternal microbiota (Mu et al., 2021).
In spite of the convincing evidence from other species, very little research has been conducted into the effects of prebiotics on the development of the gut microbiota and the immune system in horses. The aim of this study was to determine whether it was possible to stimulate the development and colonization of a gut microbiota by feeding a mixture of potentially beneficial oligosaccharides to mares and foals. This would in turn promote regulatory immunity in newborn foals. We hypothesized the following: a mixture of MOS and FOS will promote the growth of anti-inflammatory bacteria in the mare GIT which will be passed on to colonize the foal. This will affect the immune responses as assessed by the production of antibodies in the foal serum and by a mouse model of the local immune-microbial interaction in the horse gut. Finally, there will be an additive and more direct effect when the foals start eating the oligosaccharides themselves. Of particular interest was the effect of this diet regiment on the abundance of bacteria that we have previously correlated to regulatory immunity.
Discussion
As hypothesized a mixture of MOS and FOS promoted the growth of anti-inflammatory bacteria in the equine GIT. As significant increases were observed in the abundance of Akkermansia spp. The anti-inflammatory bacteria were not, as hypothesized passed on to colonize the foals, as no differences in microbial composition were detected in the foals before Day 49, at which time they were eating the mares’ food in addition to drinking milk. This indicates that ingestion of oligosaccharides from the diet was the primary and not as hypothesized an additive inducer of differences between test and control foals. It was noted that all foals ingested the mares feed but the amounts ingested was not measured thus it is not possible to say exactly how much of the supplement each foal had. In the early life foals are also coprophagic, thus microbiota may also have been passed on from mare to foal this way. However, as we did not detect any changes in the mares’ gut microbiota composition, we do not attribute this mode of microbiota transfer as the sole source of the differences in gut microbiota detected in the two groups of foals. In addition, no differences were observed between the gut microbiota composition of the test and control mares, suggesting that vertical transfer of bacteria from mares to foals have not contributed to the increase in Akkermansia spp. As hypothesized immune responses were influenced in the foals.
One Akkermansia spp., A. muciniphila which we have previously shown to be abundant in the equine gut (Lindenberg et al., 2019a) is known to inhabit the mucin layer and to help uphold the gut balance by conversion of mucin to the beneficial SCFAs acetate and propionate (Derrien et al., 2004, 2008). We have previously shown a positive correlation between Verrucomicrobia to which Akkermansia belongs and expression of the regulatory T-cell transcription factor FOXP3 (Lindenberg et al., 2019a). In that study, Akkermansia was far the most abundant member of the phylum suggesting that it was responsible for the majority of the immunological interaction, which may suggest that the unclassified Akkermansia may also have anti-inflammatory effects. Reduced abundances of A. muciniphila have been associated with inflammatory diseases such as inflammatory bowel disease (IBD) in humans (Png et al., 2010), colitis, T2D, and obesity in mice (Kang et al., 2013; Ellekilde et al., 2014a). Increasing the abundance of A. muciniphila significantly improved glucose metabolism, decreased pro-inflammatory cytokines IL-1b and IL-16 and increased levels of Tregs in diet induced obese mice (Shin et al., 2014). Vancomycin treatment of mice early in life, lead to higher abundances of A. muciniphila and significantly lower type 1 diabetes incidence (Hansen et al., 2012a) suggesting that increases in A. muciniphila in the early life may reduce the incidence of certain diseases later on. Of particular interest for this study are the findings by Everard et al. (2013), who found a direct link between FOS feeding, increased the abundance of A. muciniphila, and significant improvements in glucose- and insulin metabolism, mucosal barrier function, adipose tissue inflammation, reduced fat gain, and endotoxemia in mice. Their results support our results and hypothesis that oligosaccharides induce anti-inflammatory bacteria, which could positively impact the immune system and disease development.
In addition to Akkermansia spp. significant increases were observed within Campylobacter, and Helicobacter spp. in the test group compared to the control group. Certain species of Campylobacter are known pathogens and the cause of gastrointestinal disease (Van Vliet and Ketley, 2001). We and others have previously found this taxa in healthy horses suggesting that this might be a non-pathogenic species or that low abundances are part of the commensal equine microbiota (Moriarty et al., 2015; Lindenberg et al., 2019a), however, this should be investigated further, before non-pathogenicity can be assumed. We and others have also previously found Helicobacter spp. in the gut of healthy foals (Moyaert et al., 2009; Lindenberg et al., 2019b). Its presence has not been associated with any macro- or micro-pathologies in horses (Moyaert et al., 2007) suggesting it is also a part of the normal transient microbiota that initially colonizes the equine gut, however, pathogenic Helicobacter species exists, and further studies should determine the species of the Helicobacter in the equine GIT. Further, the betaproteobacteria Tremblayales, which so far has not been associated with either pathological or beneficial properties in healthy foals was detected (Lindenberg et al., 2019b). Proteobacteria contain LPS in the membrane, which in adults has been shown to induce endotoxemia and inflammation (Carvalho et al., 2012), but studies have shown that a lipopolysaccharide (LPS) boost early in life may lead to improved regulatory immunity (Lotz et al., 2006; Bendtsen et al., 2017). It may be speculated that by consuming mucin. Akkermansia may have reduced the mucosal layer. Such a mucosal layer reduction may facilitate interaction between bacterial LPS and the mucosal immune system thereby leading to an improved stimulation and subsequently improved the potential for inducing regulatory immunity (Bendtsen et al., 2018). Low abundances of both Campylobacter and Tremblayales have also been found in the colonic mucosa of healthy weanling piglets (Grosu et al., 2019), indicating that their presence may be normal in the microbiota of young animals. All foals who received the test supplement were healthy and did not show any gastro-intestinal symptoms. However, further studies determining the subspecies of both Campylobacter, Helicobacter and Tremblayales would be beneficial in determining whether increased abundance might have undesired effects on the gastro-intestinal system.
Mannanoligosaccharides fed to mares before parturition has previously been found to increase the levels of IgG and IgA in colostrum (Spearman and Ott, 2004). It may be speculated whether the small differences in IgA and IgG levels detected between the test and control foals indicate that the oligosaccharides may affect immunoglobulin levels, but further studies would be needed to warrant this. Future studies should investigate whether this effect is mediated through oligosaccharides increasing the colostrum immunoglobulin levels, which would then give reason to feed the mares pre-partum. Oligosaccharides may through their ability to stimulate the growth of specific species work as a “stabilizing agent” in the same way as it stabilizes the microbiota during diet changes (Respondek et al., 2008) and hence lead to the reduced alpha diversity observed in the test foals in this study. A similar effect was observed in a study comparing the gut microbiota of human infants fed an iso-caloric formula with or without galactooligosaccharides (GOS) (Matsuki et al., 2016). Oligosaccharides are routinely used in equine research to induce hindgut acidosis and subsequently laminitis, but we observed no detectable difference in pH between the test and control mares indicating that the test supplement in the given dose does not induce hindgut acidosis and as such is safe to use.
As it was not possible to obtain intestinal biopsies from the foals, we aimed to assess the local immunological effects in inoculated mice, which is also done in studies of human diseases (Zachariassen et al., 2017). Significant differences in the microbial composition were observed between the mice inoculated with microbiota from the test and control foals respectively, indicating that it is possible to transfer gut microbial differences from horses to mice. The microbial composition was not identical between test foals and inoculation test mice and control foals and inoculation control mice respectively, which is most likely due to the general loss of diversity during inoculation procedures (Ellekilde et al., 2014b), a reduced colonization rate (Lundberg et al., 2017) and differences observed in xeno-inoculations (Chung et al., 2012). The increased level of Akkermansia spp. was not transferred to the mice, but six genera, which we have previously correlated to regulatory immunity in horses, were found in increased abundances in the inoculation test group compared to the inoculation control group (Lindenberg et al., 2019a). This indicates that the microbial composition from the test foals had a greater potential of inducing regulatory immunity. The significantly increased expression of Il10 encoding the regulatory IL-10 cytokine in the ileum of the inoculation test group supports this hypothesis. Inoculation procedures and pre-treatment with antibiotics are associated with loss of species richness (Ellekilde et al., 2014b), and this is most likely the cause of the generally lower alpha-diversity in both the inoculation test and inoculation control mice. The lower alpha diversity may also explain the significantly lower Il12 expression in the inoculation control group compared to the test supplement control group, Il12 encode the pro-inflammatory interleukin 12 (IL-12) secreted by dendritic cells to stimulate the formation of T-helper 1 (Th1) cells that are stimulated by microbiota, and a reduced species richness in the inoculation control group may also have reduced the overall immune stimulation. The differences between the correlation analysis in the mice and the previous results from horses may be explained by the age of the mice at the time of inoculation. The mice were inoculated at 6 weeks of age at which time the window of opportunity for development of oral tolerance has closed (Hansen et al., 2012b). In the previous study, the horses were assumed to have acquired the same species as foals, and thereby have developed oral tolerance toward them. Optimally the inoculation should, therefore, have been performed on animals under 3 weeks of age, which can be done by inoculation of the mothers to have the pups born with the microbiota (Zachariassen et al., 2017). This should eventually be pursued in future horse to mouse inoculations. Additionally, to the potential regulatory immunity inducing taxa we also observed increased abundances of Candidatus Arthromitus also known as segmented filamentous bacteria (SFB) in the inoculation-test mice. We did not observe any SFBs in the foals, so the increase is not likely to be a direct effect of inoculation with SFB. However, the altered microbial composition may have changed the living conditions for certain resident bacteria and caused a bloom in SFBs. After ampicillin treatment a few species of resident bacteria may propagate in mice (Castro-Mejia et al., 2018). SFBs selectively stimulate the differentiation of T-helper 17 (Th17) cells, which are required for a normal immune balance, and SFB status has been shown to reduce the incidence of type 1 diabetes (Kriegel et al., 2011) and increase the severity of IBD (Ivanov et al., 2008, 2009) in mice. A further investigation into this increased abundance should, therefore, be conducted, preferably by also assessing the level of IL-17 between the two groups.
It should be noted that the applied oligosaccharides were extracted from yeast cell walls, and these also contain the polysaccharides known as β-Glucans (Fesel and Zuccaro, 2016), and therefore traces of these have been present in our preparation. β-Glucans from fungi exhibit a broad spectrum of biological activities including anti-tumor, immune-modulating and anti-inflammatory properties (Du et al., 2015), which may also have added to the effects observed in this study.
In conclusion, our oligosaccharide containing test supplement increased the abundances of Akkermansia spp. the presence of which is known to be associated with improved mucosal health, lower inflammatory level, and higher levels of regulatory immunity, which may reduce incidence or severity of inflammatory diseases later on. The test group also had increased abundance of Campylobacter, Helicobacter, and Tremblayales, which we believe to be part of the transient microbiota of young foals, but further studier should be carried out to determine this, as spp. are also known to be pathogenic. The effects of the test supplement seemed to be caused by intake of the diet by the foals and direct microbial stimulation by the oligosaccharides. Mouse inoculations showed an anti-inflammatory potential of the gut microbiota from the oligosaccharide fed test foals.