Figure 1.
ER–associated polysome profiling of lean and obese liver tissues.
(A) ER-associated polysome profile of lean and obese mouse liver tissues without overnight fasting. Membrane bound polysomes were released from ER and centrifuged through sucrose gradients to separate ribosomes according to their density with light ribosomes (ribosome subunits and single ribosomes) at the top and heavy ones (polysomes) at the bottom of the separation columns. The sucrose gradients were then fractionated from top to bottom and the concentration of ribosomes in each fraction was continuously monitored by UV absorbance (A260 for ribosomal RNA). The left to right of the X-axis correlates the top (light ribosome subunits and single ribosome) to the bottom (heavy, multiple ribosomes assembled on a single transcript) of the sucrose gradient respectively. The Y-axis A260 measures the total amount of ribosomal RNA present in each fraction. Ribosomes from equivalent amounts of ER were fractionated. (B–C) ER associated polysome profile of lean and obese mouse liver tissues at 3 and 6 months of age without overnight fasting. (D) Area under the curve measurements of the polysome fraction of lean and obese mouse livers at 3 months of age (n = 3, p<0.05, Student's t-test). (E) Quantification of de novo protein synthesis in lean and obese primary hepatocytes as measured by 35S-Methionine pulse-tracing (t = 30 minutes, n = 4, p<0.01, Student's t-test). (F) ER associated polysome profile of lean and obese mouse liver tissues at 3 months of age after overnight fasting. Equal amount of ER isolated from lean and obese mouse livers were loaded onto the sucrose gradient for polysome fractionation.
Figure 2.
Genome-wide translational changes in the obese mouse liver.
(A) Scatterplot of logP-values versus log2FC (fold of change, obese/lean). Green denotes genes up-regulated in the obese liver ER translatome while red denotes downregulation. (B) Differential regulation of transcripts as measured by microarray (Y-axis) and quantitative RT-PCR (X-axis) for 22 genes that were calculated as differentially regulated (student t-test, P<0.05) by either one or both methods. (C) Gene Ontology analysis of differentially regulated genes in the ER translatome of nonfasted lean and obese liver. X-axis denotes the number of genes categorized into each differentially-regulated function on the Y-axis. Color scale corresponds to logP-values converted to positive values for functional categories comprised of upregulated (obese/lean) genes. (D) Measurement of RNA synthesis in lean and obese primary hepatocytes based on 32P-UTP incorporation (t = 2 hours, n = 4, p<0.01, Student's t-test). (E) Immunoblot measurement of mitochondria protein expressions in the lean and obse mouse liver samples. (F) Immunoblot measurement of overall mitochondrial protein acetylation (IB: α-Ac, top panel) and the acetylation of the succinate dehydrogenase (IP: α-Ac; IB: α-SDHA, middle panel). Acetylation of the pyruvate dehydrogenase enzyme 1 (PDH-E1α) is shown (bottom panel) as a positive control for α-Ac immunoprecipitation as it is fully charged with acetyl-CoA under both lean and obese conditions. (G) Measurement of mitochondria oxygen consumption rate (OCR) for lean and obese primary hepatocytes (n = 7, p = 0.01, Student's t-test).
Figure 3.
Dynamic regulation of the hepatic ER translatome by fasting.
(A) Illustration of the overlap between translatomic changes induced by the development of obesity (obese nonfasted, ONF versus lean nonfasted, LNF) and fasting in lean mice (lean fasted, LF versus lean nonfasted, LNF). (B) Correlation scatterplot of fold changes (logarithmically transformed) for commonly regulated genes induced by obesity (Y-axis) and fasting in the lean (X-axis). Each dot represents a gene that is differentially regulated in the translatome in both comparisons (ONF versus LNF and LF versus LNF). (C) Diagram illustrates the reversal of obesity-induced translatomic changes (ONF versus LNF) by overnight fasting in the obese mice (OF versus ONF). (D) Scatterplot of logarithmically transformed fold changes induced by obesity (Y-axis) and its reversal by overnight fasting (X-axis). Each dot represents a gene that is differentially regulated in both comparisons (ONF versus LNF and OF versus ONF). “R” denotes Pearson coefficiency. (E) Distribution of commonly regulated genes between ONF versus LNF and LF versus LNF or OF versus ONF. X-axis: P-values of LF versus LNF; Y-axis: P-values of OF versus ONF; Z-axis: number of genes differentially regulated in the ONF versus LNF comparison. Negative values correspond to downregulated genes. I, genes upregulated in both ONF versus LNF comparison and LF versus LNF comparison; II, genes commonly downregulated between these comparisons; III, genes upregulated in ONF versus LNF comparison but downregulated in OF versus ONF comparison; IV, genes downregulated in the ONF versus LNF comparison but upregulated in the OF versus ONF comparison.
Figure 4.
Dysregulation of bile acid metabolism in the obese ER translatome.
(A) Heatmap of 50 most differentially regulated genes in the obese ER-associated translatome (25 up, 25 down) that are also suppressed in the lean fasted liver. ONF: obese non-fasted, LNF: lean non-fasted, OF: obese overnight-fasted, LF: lean fasted. Fold changes are calculated based on the mean of all ten samples. (B) Validation of feeding-induced upregulation of Cyp7b1 and Slco1a1, but not CD36 and Igfbp2, in the translatome of lean mouse liver. (C) Western blot of Cyp7b1 from the liver of lean and obese mice with or without overnight fasting.
Figure 5.
Cyp7b1 and Slco1a1 overexpression regulates glucose homeostasis.
(A) Examination of Cyp7b1 expression in the liver of ob/ob mice transduced with control and Cyp7b1 over-expressing adenoviruses by quantitative RT-PCR and immunoblot analysis. (B) Changes of body weight in control and Cyp7b1 and Slco1a1 overexpressing mice in a 14 day period post-virus injection. (C) Glucose levels of control and experimental mice 5 days post-virus injection. (D–E) Insulin and glucose tolerance test of control and experimental mice with bolus injections of insulin (1.0 IU/kg) or glucose (0.75 g/kg) and measurement of plasma glucose levels at indicated times (n = 9). (F) Immunoblot analysis of insulin signaling stimulated by portal vein injection as indicated by insulin receptor (IR) and AKT phosphorylation. The graph on the right side shows the quantitation of the data after correction for total protein amount. (G) Expression of gluconeogenic genes in the liver of control and experimental mice as measured by with real time, quantitative RT-PCR (n = 4). “*” denotes p<0.05 comparing control versus Cyp7b1-overexpressing mice, and “#” denotes p<0.05 comparing control versus Slco1a1 overexpression, Student's t-test. PEPCK: phosphoenolpyruvate carboxykinase; G6Pc: Glucose-6-phosphatase catalytic subunit.
Figure 6.
Cyp7b1 overexpression causes cholestasis.
(A–C) Changes of hepatic steatosis in the liver of control and experimental mice as measured by liver triglyceride levels and H&E staining. White vesicles indicate fat infiltration. (D–E) Measurement of hepatic cholesterol and bile acid levels in the control and experimental mice. (F–G) Measurement of plasma ALT/AST levels and bile acid levels from the control and experimental mice. (H) Transcript levels of genes involved in inflammation and fibrosis in the liver of control and experimental mice as measured by qPCR. (I) Measurement of total bile collected from the gallbladder of control and experimental mice after 6 hours of food withdrawal. (J–K) Measurement of transcript levels of genes involved in cholesterol and bile synthesis and transport in the liver tissues of control and experimental mice with real time, quantitative RT-PCR. “*” denotes p<0.05 (Student's t-test, n = 6 except for qPCR, in which n = 4/group).