Effects of early nutrition on the wnt and bone morphogenic protein pathways in the developing intestine

Infant formula is consumed by the majority of infants in the United States for at least part of the first year of life. Infant formula lacks many of the bioactive compounds that are naturally occurring in breast milk. Because of this, there has been an increased interest by the companies that manufacture infant formula to include additives that would potentially allow formula to more closely mimic breast milk activity. One such ingredient currently being added to infant formula is prebiotics. Prebiotics are non-digestible food ingredients that beneficially affect the host by selectively stimulating the growth of specific healthful bacteria in the colon. It is speculated that prebiotics replicate the activity of breast milk oligosaccharides, which through the production of butyrate by intestinal microbiota, may interact with the Wnt/BMP pathways. The Wnt/BMP pathways regulate intestinal stem cells, which determine the growth, development and maintenance of the intestine. Therefore, the objective of this study was to explore the effects that the addition of prebiotics to formula have on the regulation of the Wnt/BMP pathways when fed to neonatal piglets, a model commonly used in the study of infant nutrition. Piglets (n=5) were randomized into sow-reared (SR), fed control formula (F), or fed formula with added prebiotics (F+P). Fructooligosaccharides (FOS) (2 g/L) and polydextrose (PDX) (2 g/L) were chosen as the prebiotics for this study, because this combination had been less studied than other combinations. Ileum and ascending colon were collected at 7 and 14 days-of-age. Dry matter content, pH, and short chain fatty acid (SCFA) content was measured. The mRNA expression of β-catenin, sFRP3, sFRP4, frizzled 6, DKK1 (Wnt pathway), gremlin (BMP pathway), TNF-a, HNF-4α and osteopontin (OPN) was measured by RT-qPCR. Piglets fed the F+P diet had greater acetate concentration and lower pH in the ileum at day 14 and in the colon at day 7 and day 14 than F piglets. Butyrate concentrations were highest in SR with F+P not differing from F in ileum at day 14 and colon at day 7 and day 14. Effects of age were seen in all genes, with the exception of OPN, sFRP-3 and sFRP-4. On day 7, no effect of diet was observed in the ileum, however, mRNA expression of DKK1 and frizzled 6 were greater in F+P than SR (p≤0.05). On day 14, gremlin expression was lower and OPN was greater in the ileum of SR piglets compared to F and F+P. Also on day 14, HNF-4α mRNA expression was greater in both ileum and colon of F+P piglets and sFRP3 mRNA expression was greater in the colon than F or SR . In summary, differences were observed between gene expression of F+P and SR piglet intestines, but the supplementation of 2 g/L scFOS and 2 g/L PDX to formula did not shift expression of genes in the Wnt/BMP pathways to be more similar to SR than F. As the Wnt/BMP pathway is known to exist in a gradient along the crypt-villus axis, with Wnt expression dominating in the crypt region and BMP expression dominating in the villi, it was possible that pooling whole tissue reduced our ability to detect treatment effects that would be concentrated in either region. A method was therefore developed to remove intestinal epithelial cells along the villus-to-crypt axis. Twenty-five-day-old F and SR piglets were euthanized and ileal tissue was collected and placed in a dissociation buffer in a shaking water bath. Exfoliated cells were removed at increasing time points from 5 to 100 minutes in order to remove cells along the villus-to-crypt axis. After the final incubation, remaining mucosal tissue was removed using a sterile glass microscope slide and pooled with the final exfoliated cell isolation. After each cell collection, a section of tissue was fixed in formalin for histomorphological examination. Expression of genes in the Wnt/BMP pathways, along with crypt marker genes (CDK5 and v-myb), were measured in both whole ileal tissue, pooled epithelial cells, and separate epithelial cell isolations from the same piglet. The expression of β-catenin, HNF-4α, TNF-α, TGF-β and the crypt marker v-myb matched the expected villus-to-crypt pattern in cells collected after 10 (incubation 1), 30 (incubation 2) and 60 (incubation 3) minutes. However, expression of expression in cells collected after 100 minutes (incubation 4) was variable, which may be due to the fact that crypt cells were not efficiently removed and the presence of unwanted non-epithelial tissue. Gremlin, OPN, DKK1, sFRP3 and sFRP4 expression was not statistically different along the villus-to-crypt axis. Frizzled 6 and CDK5 did not express as we had predicted, with expression highest towards the villi. In summary, the epithelial cell collection method used was not entirely successful. While much of the gene data suggests that cells were removed along the villus-to-crypt axis through the first three incubations, the last incubation, which involved scraping the tissue, removed non-epithelial components of the mucosa, while leaving the crypts intact. In conclusion, the addition of 2 g/L PDX and 2 g/L scFOS did not cause gene expression of the Wnt/BMP pathways to mirror either F or SR expression. New isolation methods to extract cells along the crypt-villus axis should be considered, including the use of a laser capture microdissection. While this combination of prebiotics did not yield the intended effects, future research should be done on other combinations, such as the inclusion of galactooligosaccharides (GOS), which is commonly added to food products including infant formula.