Integration of metabolomic and transcriptomic analyses reveals novel regulatory functions of the ChREBP transcription factor in energy metabolism

Carbohydrate Response Element-Binding Protein (ChREBP) is a transcription factor that activates key genes involved in glucose, fructose, and lipid metabolism in response to carbohydrate feeding, but its other potential roles in metabolic homeostasis have not been as well studied. We used liver-selective GalNAc-siRNA technology to suppress expression of ChREBP in rats fed a high fat/high sucrose diet and characterized hepatic and systemic responses by integrating transcriptomic and metabolomic analyses. GalNAc-siChREBP-treated rats had lower levels of multiple short-chain acyl CoA metabolites compared to rats treated with GalNAc-siCtrl containing a non-targeting siRNA sequence. These changes were related to a sharp decrease in free CoA levels in GalNAc-siChREBP treated-rats, accompanied by lower expression of transcripts encoding enzymes and transporters involved in CoA biosynthesis. These activities of ChREBP likely contribute to its complex effects on hepatic lipid and energy metabolism. While core enzymes of fatty acid (FA) oxidation are induced by ChREBP knockdown, accumulation of liver acylcarnitines and circulating ketones indicate diversion of acetyl CoA to ketone production rather than complete oxidation in the TCA cycle. Despite strong suppression of pyruvate kinase and activation of pyruvate dehydrogenase, pyruvate levels were maintained, likely via increased expression of pyruvate transporters, and decreased expression of lactate dehydrogenase and alanine transaminase. GalNAc-siChREBP treatment increased hepatic citrate and isocitrate levels while decreasing levels of distal TCA cycle intermediates. The drop in free CoA levels, needed for the 2-ketoglutarate dehydrogenase reaction, as well as a decrease in transcripts encoding the anaplerotic enzymes pyruvate carboxylase, glutamate dehydrogenase, and aspartate transaminase likely contributed to these effects. GalNAc-siChREBP treatment caused striking increases in PRPP and ZMP/AICAR levels, and decreases in GMP, IMP, AMP, NaNM, NAD(P), and NAD(P)H levels, accompanied by reduced expression of enzymes that catalyze late steps in purine and NAD synthesis. ChREBP suppression also increased expression of a set of plasma membrane amino acid transporters, possibly as an attempt to replenish TCA cycle intermediates. In sum, combining transcriptomic and metabolomic analyses has revealed regulatory functions of ChREBP that go well beyond its canonical roles in control of carbohydrate and lipid metabolism to now include mitochondrial metabolism and cellular energy balance. Overall design: All procedures were approved by Duke University Institutional Animal Care and Use Committee and performed according to the regulations of the committee. Breeding pairs of Obese Prone CD (OP/CD) Sprague Dawley rats were gifts from Dr. Warren Grill and Dr. Eric Gonzalez, Duke University, and a colony was established and maintained by Duke Laboratory Animal Resources (DLAR). Starting at 4 weeks of age, male OP/CD rats were single-housed with a light cycle of 7 AM on/7 PM off, and fed ad libitum with a high-fat/high-sucrose (HF/HS) diet (D12451i, Research Diets) containing 47% fat (kcal) and 17% sucrose (kcal). Body weight and food intake were monitored weekly. After 9 weeks of feeding of the HF/HS diet, plasma samples were collected via saphenous vein bleeding. One week later, animals received an initial subcutaneous injection of one of two GalNAc-siRNA constructs at a dose of 9 mg/kg body weight, or an equal volume of the diluent (PBS), (see below for description of the two GalNAc-siRNA reagents). Additional doses of each GalNAc-siRNA construct were injected at 10, 18 and 25 days after the first injection. Animals were fasted overnight one day after the third injection (day 19), and subjected to an intraperitoneal glucose tolerance test (IPGTT) on the following day. Animals were weighed and a glucose solution (1g/kg body weight) was administered via intraperitoneal injection. Tail blood samples were obtained and glucose levels measured with a blood glucose meter (CVSHealth) immediately before and at 30,60, 90, 120, and 180 minutes after bolus injection of glucose. One day after the fourth GalNAc-siRNA or saline injection on day 25, plasma samples were collected via saphenous vein bleeding. A bolus of deuterium oxide (D2O, 10 ml/kg body weight, Sigma Aldrich) was then given by intraperitoneal injection and followed by free access to drinking water supplemented with 4% D2O for the rest of the experimental period. Saphenous plasma samples were collected again one day after the bolus delivery of D2O (day 27). On day 28 between 8 AM-noon, animals were anesthetized with Nembutal (250 mg/kg) by intraperitoneal injection and sacrificed for collection of plasma and liver samples that were rapidly frozen in liquid nitrogen and stored at -80oC for further analyses.