Epoxygenase inactivation exacerbates diet and aging-associated metabolic dysfunction resulting from impaired adipogenesis

Abstract

Objective: When molecular drivers of healthy adipogenesis are perturbed, this can cause hepatic steatosis. The role of arachidonic acid (AA) and its downstream enzymatic cascades, such as cyclooxygenase, in adipogenesis is well established. The exact contribution of the P450 epoxygenase pathway, however, remains to be established. Enzymes belonging to this pathway are mainly encoded by the CYP2J locus which shows extensive allelic expansion in mice. Here we aimed to establish the role of endogenous epoxygenase during adipogenesis under homeostatic and metabolic stress conditions.

Methods: We took advantage of the simpler genetic architecture of the Cyp2j locus in the rat and used a Cyp2j4 (orthologue of human CYP2J2) knockout rat in two models of metabolic dysfunction: physiological aging and cafeteria diet (CAF). The phenotyping of Cyp2j4^-/- rats under CAF was integrated with proteomics (LC-MS/MS) and lipidomics (LC-MS) analyses in the liver and the adipose tissue.

Results: We report that Cyp2j4 deletion causes adipocyte dysfunction under metabolic challenges. This is characterized by (i) down-regulation of white adipose tissue (WAT) PPARγ and C/EBPα, (ii) adipocyte hypertrophy, (iii) extracellular matrix remodeling, and (iv) alternative usage of AA pathway. Specifically, in Cyp2j4^-/- rats treated with a cafeteria diet, the dysfunctional adipogenesis is accompanied by exacerbated weight gain, hepatic lipid accumulation, and dysregulated gluconeogenesis.

Conclusion: These results suggest that AA epoxygenases are essential regulators of healthy adipogenesis. Our results uncover their synergistic role in fine-tuning AA pathway in obesity-mediated hepatic steatosis.

Keywords: Adipogenesis; Aging; Arachidonic acid; Cafeteria diet; Cytochrome P450 2j4; Steatosis.

Figure 1

Cyp2j4^−/− mesenchymal stromal cells (MSCs) undergo spontaneous adipogenesis, and aging Cyp2j4^−/− rats show increased body weight and larger adipocytes. (A) The genomic synteny between human CYP2J2 (reference) locus and the corresponding rat and mice loci. [ ] denote the interruption in the genomic distance scale; positions are in bp. (B) Oil-red-o (ORO) staining in MSCs from WT and Cyp2j4^−/− before (left panel, pre-adipocytes) and after (right panel, adipocytes) adipocyte differentiation with the addition of adipogenic differentiation medium (DM). (C) ORO staining quantification as well as Fabp4 and Adipoq gene expression in MSC-derived pre-adipocytes and adipocytes. (D) Western blot analysis of PPARγ (pre-adipocytes) and C/EBPα (adipocytes) in Cyp2j4^−/− and WT cells. Blots are representative of 2 independent experiments. (E) Western blot analysis of PPARγ and C/EBPα in the adipose tissue from WT and Cyp2j4^−/− rats. (F) Glucose, insulin levels, and HOMA-IR index in 15-month old WT (n = 4) and Cyp2j4^−/− (n = 6) rats' plasma (G) Body weight evolution in aging WT (n = 4) and Cyp2j4^−/− (n = 6) rats under standard chow diet. (H) Representative Haematoxylin and Eosin (H&E) white adipose tissue (WAT) staining (left panel) and adipocyte area quantification in 15-month old WT and Cyp2j4^−/− rats. Error bars are s.e.m. Scale bars; 250 μm (B); 100 μm (H).

Figure 2

Cyp2j4^−/− rats show relatively increased metabolic dysfunction under CAF. (A) Percentage of body weight gain in WT (n = 5) Cyp2j4^−/− (n = 4) rats during 12 weeks of CAF. (B) Glucose, insulin, HOMA-IR, TAG, and NEFA levels measured in WT and Cyp2j4^−/− rats' plasma upon standard (STD) or cafeteria diet (CAF). At least n = 3 rats were used in each group. (C) Representative H&E staining in epididymal (left panel) and subcutaneous fat (right panel) sections. The arrows indicate the thickness of the subcutaneous adipose tissue layer and its larger magnification (×40) is shown at bottom left. (D) Adipocyte area distributions in WT CAF (open bars) and Cyp2j4^−/− CAF (black bars) where adipocytes are grouped into ascending sizes of 250 μm² (size range 250 μm²) in epididymal (left panel) and subcutaneous fat (right panel). For clarity, 9 group sizes are shown in the x-axis. Mean adipocyte area is shown for all groups (epididymal AT, top left). Subcutaneous (SBC) adipocyte layer and mean adipocyte layer are shown for WT (CAF) and Cyp2j4^−/− rats (subcutaneous AT, top right). (E) WAT cell density in WT and Cyp2j4^−/− rats in STD diet and CAF. At least n = 3 rats were used in each group. ns, non-significant. Scale bars, 100 μm (epididymal AT) and 500 μm (subcutaneous AT).

Figure 3

Macrophage infiltration and early fibrosis in Cyp2j4^−/− rats WAT upon CAF. (A) CD68 (rat ED-1; CLS denotes Crown-like structures) staining in WT (CAF) and Cyp2j4^−/− (CAF) and qRT-PCR for Cd68 and Adipoq (B) in STD or CAF-trated WT and Cyp2j4^−/− rats. (C) WAT hydroxyproline levels in STD or CAF-treated WT and Cyp2j4^−/− rats. (D) Representative immunofluorescence images for type I (left panel) and type VI (right panel) collagens and their quantification (bottom). (E) LC-MS/MS quantification of FUCA1 and IGF2R, two HIFa targets (TF-binding enrichment) in the stromal vascular fraction of WT and Cyp2j4^−/− rats under CAF. Error bars are s.e.m. At least n = 3 rats were used in each group. ns, non-significant. Scale bars, 100 μm (A) and 50 μm (D).

Figure 4

De novo lipogenesis and increased gluconeogenesis in Cyp2j4^−/− livers under CAF. (A) ORO staining and triglyceride (TAG) quantification in STD and CAF-treated WT and Cyp2j4^−/− rats' livers. (B) ORO staining (left) and TAG quantification in 15-month old WT and Cyp2j4^−/− livers. (C) C/EBPα Western blot analysis in STD and CAF-treated WT and Cyp2j4^−/− rats' livers. Blots are representative of 2 independent experiments. (D) qRT-PCR analysis of Fasn in STD and CAF-treated WT and Cyp2j4^−/− rats' livers. (E) LC-MS/MS heatmap displaying the proteins with significant differential protein abundance between Cyp2j4^−/− (CAF) when compared with Cyp2j4^−/− (STD) (150 and 164 up- and down-regulated proteins respectively, false discovery rate (FDR) < 0.05). In the heatmap, z-scores of the log-transformed intensities are displayed. Relevant functionally enriched pathways in these two protein sets are shown together with proteins contributing to these enrichments (red and blue bars). (F) LC-MS/MS heatmap (zoomed from E) and protein quantification profiles between WT (CAF) and Cyp2j4^−/− (CAF) for Pgm1, Fbp1, Aldob, Ldha, Gapdh, Gpi, Pklr, Eno1, and Dlat. (G) Phospho-AKT (Ser473 and Thr308) and total Akt Western blot in CAF-treated WT (n = 3) and Cyp2j4^−/−(n = 3) rats' livers. Numbers denote biological replicates. Error bars are s.e.m. At least n = 3 rats were used in each group. ns, non-significant. Scale bars, 100 μm.

Figure 5

Cyp2j4 deletion causes WAT dysfunction and a shunt in the AA pathway. (A) PPARγ and C/EBPα Western blot analyses in WT and Cyp2j4^−/− WAT (STD vs. CAF) or aging (4-month vs.15-month old) conditions. (B) Schematic representation of AA and LA-derived eicosanoids. The Cyp450 pathway is shown in grey to illustrate the inhibition of Cyp2j4-derived EET production. All eicosanoids in green are up-regulated either in aging or CAF conditions in WAT from Cyp2j4^−/− rats. For the quantitative data, see Supplementary Figure A.6A. (C) Schematic illustration of the quantitative lipidomics, proteomics, and RNA-seq datasets obtained from different tissues. The protein and mRNA levels of enzymes responsible for the generation of the eicosanoids detected in (B) were investigated in SVF LC-MS/MS, bone marrow-derived macrophage (BMDM) LC-MS/MS and RNA-seq. (D) Ptges3 and Cbr1 protein levels in WT and Cyp2j4^−/− BMDMs by LC-MS/MS (n = 3 rats per group). (E) Ptgis protein levels in SVF by LC-MS/MS (n = 4 rats per group). (F) Schematic illustration of the AA COX pathway showing the enzymes that catalyze the synthesis of different prostaglandin species. (G) Graphical summary showing WAT homeostasis under aging and CAF in WT Cyp2j4^−/− rats. Error bars are s.e.m.