Genome-scale metabolic model predicts mechanistic differences in metabolic gradient of spatially resolved small intestine villi

Studying the spatial metabolic gradient provides significant opportunities for understanding the spatial division of labor within the tissue microenvironment. Enterocytes have the capacity to perform serial oxidation and conjugation reactions for detoxification, making the small intestine important as one of the first-pass metabolic organs. Recently, the enterocyte metabolic gradient was found to exhibit differential metabolic preferences depending on its location in the villus. However, it remains unclear how metabolism mechanistically differs in enterocyte microenvironments. To bridge this knowledge gap, we leveraged spatial transcriptomics data to (1) reconstruct genome-scale metabolic networks (GSMMs) that are location-specific, and (2) identify metabolic genes that may explain the differential susceptibility of different villus sections to diseases, diet, and other factors. We found that enterocytes at the bottom of the villus are enriched in genes related to intermediary metabolism, phase-I and -II metabolism, bile acid metabolism, and transporters. Comparing enterocyte GSMMs between the top and bottom of villi, we show that enterocytes at the top produce NAD and threonine more robustly compared to bottom enterocytes. Conversely, bottom enterocytes produce guanosine monophosphate (GMP) more readily than enterocytes at the top of the villus. These metabolic differences may have implications for differential villi susceptibility to diseases such as neuroendocrine tumors, acute graft-versus-host disease, and nutritional perturbations such as high-fat diets. Taken together, our findings demonstrate in a mechanistic manner the metabolic differences of enterocytes in the small intestine, providing information that can be applied to additional disease states and inform therapeutics development. AUTHOR SUMMARYSignificant metabolic gradients exist across different locations in a cell. Currently, bulk expression experiments, used in most animal studies and clinical trials, fail to describe this spatial variation due to the averaging effect in the tissue of interest. Recently, single-cell RNA sequencing technologies revealed heterogeneity within tissues. However, the underlying differences and mechanisms for the spatial metabolic variations remain understudied. In this research, we focus on the small intestine, which plays a critical role in nutrient absorption, distribution, and drug metabolism. The walls of the small intestine are composed of finger-like projections, called villi, that are composed of enterocytes. To investigate the potential mechanism underlying metabolic differences of enterocytes along the villi, we built and compared networks formed by gene-protein reaction relationships from spatially resolved bulk and scRNA-seq data. We found that enterocytes at the villus top more robustly produced key components in cell metabolism, such as nicotinamide adenine dinucleotide (nad) and threonine, whereas enterocytes at the villus bottom more robustly produced the nucleotide guanosine monophosphate (GMP). Our approach can be extended to the study of metabolic differences in other organs and diseases and to research the metabolism of specific compounds.