Discrete Correlation Summation Clustering Reveals Differential Regulation of Liver Metabolism by Thrombospondin-1 in Low-Fat and High-Fat Diet-Fed Mice

Abstract

Thrombospondin-1 (TSP1) is a matricellular protein with many important roles in mediating carcinogenesis, fibrosis, leukocyte recruitment, and metabolism. We have previously shown a role of diet in the absence of TSP1 in liver metabolism in the context of a colorectal cancer model. However, the metabolic implications of TSP1 regulation by diet in the liver metabolism are currently understudied. Therefore Discrete correlation summation (DCS) was used to re-interrogate data and determine the metabolic alterations of TSP1 deficiency in the liver, providing new insights into the role of TSP1 in liver injury and the progression of liver pathologies such as nonalcoholic fatty liver disease (NAFLD). DCS analysis provides a straightforward approach to rank covariance and data clustering when analyzing complex data sets. Using this approach, our previous liver metabolite data was re-analyzed by comparing wild-type (WT) and Thrombospondin-1 null (Thbs1−/−) mice, identifying changes driven by genotype and diet. Principal component analysis showed clustering of animals by genotype regardless of diet, indicating that TSP1 deficiency alters metabolite handling in the liver. High-fat diet consumption significantly altered over 150 metabolites in the Thbs1−/− livers versus approximately 90 in the wild-type livers, most involved in amino acid metabolism. The absence of Thbs1 differentially regulated tryptophan and tricarboxylic acid cycle metabolites implicated in the progression of NAFLD. Overall, the lack of Thbs1 caused a significant shift in liver metabolism with potential implications for liver injury and the progression of NAFLD.