CGI-58 KD dissociates hepatic steatosis from insulin resistance. A: Photographs depicting hepatic steatosis and adiposity in C57BL/6N mice fed either a chow or HFD for 10 weeks in conjunction with biweekly injections (25 mg/kg) of either a nontargeting control ASO or ASO targeting KD of CGI-58 (CGI-58 ASO). B–G: Despite hepatic lipid insult, KD of CGI-58 enhances insulin signaling. Mice were fed either a chow or HFD and treated with ASOs for 8 weeks. Mice were fasted for 10 h before saline or insulin injection into the portal vein. Exactly 5 min later, tissues were excised and immediately snap-frozen in liquid nitrogen. Protein extracts from the liver (B), skeletal muscle (C), and adipose tissue (D) were analyzed by Western blotting for total Akt, phospho (p)-Akt (Ser473 and Thr308), and phospho (p)-FoxO1 (Ser256); three representative animals are shown for each group. E–G: Densitometric analyses of insulin signaling: Phospho-Akt protein levels were normalized to total Akt in liver (E), skeletal muscle (F), and adipose tissue (G). □, control ASO; ■, CGI-58 ASO. Data represent the mean ± SEM from three mice per group, and values not sharing a common superscript letter differ significantly (P < 0.05). AU, arbitrary unit. (A high-quality digital representation of this figure is available in the online issue.)

CGI-58–generated signaling lipids are necessary for maximal TNFα signaling in the liver. A–C: Mice were maintained on a chow diet for 4 weeks in conjunction with biweekly injections (25 mg/kg) of either a nontargeting control ASO (□) or ASO targeting knockdown of CGI-58 (CGI-58 ASO; ■). Mice were fasted for 10 h before injection of saline or TNFα (10 ng) into the portal vein. Exactly 5 min later, the liver was excised and immediately snap-frozen in liquid nitrogen for signaling analyses. A: Hepatic levels of PA and phosphatidylglycerol (PG) were analyzed by mass spectrometry. B: Total hepatic LPAAT activity. Data in A and B represent the mean ± SEM from four mice per group, and values not sharing a common superscript letter differ significantly (P < 0.05). C: Protein extracts from the liver were analyzed for total IκB α (IκBα) and phospho-IκBα (p-IκBα; Ser32); data from four representative animals are shown for each group. D–F: Acute stress kinase activation in primary hepatocytes. Following 4 weeks of ASO treatment, hepatocytes were isolated from control and CGI-58 ASO-treated mice by collagenase perfusion. Freshly isolated hepatocytes were stimulated for 15 min (15’) or 1 h with 100 ng/mL TNFα (D), 10 ng/mL IL-1β (E), or 10 ng/mL IL-6 (F). Downstream signaling was analyzed by immunoblotting for p-JNK (Thr183/Tyr185), phospho-S6 ribosomal protein (p-S6; Ser235/236), and β-actin. Data in D–F represent responses of hepatocytes isolated from three individual mice per condition.

Proposed model for CGI-58’s integrated role in inflammatory responses and insulin action in the liver. In response to inflammatory stimuli, such as an HFD or LPS, plasma levels of inflammatory cytokines, such TNFα, IL-1β, and IL-6, are increased. These inflammatory cytokines normally signal through their membrane-bound receptors (TNF-R, IL-6-R, and IL-1-R) to promote CGI-58–driven generation of signaling lipids either directly from LPAAT activity or indirectly by coactivating ATGL to generate lipid signals from TAG hydrolysis. CGI-58–generated PA, and likely other signaling lipids, can subsequently act as a critical second messenger to promote the activation of inflammatory stress kinases, such as IKK-β, S6K1, and mTOR. Collectively, these activated stress kinases (IKK-β, S6K1, and mTOR) can facilitate serine phosphorylation (pS) of critical serine residues (Ser307, Ser612, Ser632, and Ser1101) on IRS-1, thereby dampening hepatic insulin signaling. Knocking down CGI-58 diminishes this potent negative regulatory loop, thereby improving hepatic insulin action. IR, insulin receptor. TNF-R, TNF receptor.

CGI-58 KD alters HFD-induced inflammation: Evidence of hepatic cytokine resistance. C57BL/6N mice were fed either a standard chow or HFD in conjunction with biweekly injections of either a nontargeting control ASO (□) or ASO targeting KD of CGI-58 (CGI-58 ASO; ■) for 10 weeks. A: Plasma levels of proinflammatory cytokines, including IL-6 and IL-12p40; monocyte chemoattractant protein-1 (MCP-1); macrophage inflammatory protein-2 (MIP-2); CXCL1 (KC); and regulated upon activation, normal T-cell expressed, and RANTES. Data represent the mean ± SEM from five mice per group, and values not sharing a common superscript letter differ significantly (P < 0.05). ND, levels below limit of detection. B and C: HFD-induced stress kinase activation. Representative immunoblots from liver (B) or epididymal adipose tissue (C) are shown for phospho-IκB kinase α/β (p-IKKα/β; Ser176/180), phospho-mTOR (p-mTOR; Ser2448), and phospho-S6 ribosomal protein (p-S6; Ser235/236). Membranes were probed for β-actin and CGI-58 to serve as loading controls; data from four representative animals are shown for each group.

CGI-58 KD alters the systemic inflammatory and metabolic response to endotoxin. C57BL/6N mice were fed a standard chow diet in conjunction with biweekly injections of a nontargeting control ASO (□) or CGI-58 ASO (■) for 4 weeks. Thereafter, mice received a single intraperitoneal injection of either saline or LPS (5 μg/mouse) and were necropsied 6 h after injection. A: Plasma cytokine and acute-phase response protein levels were measured at 1 h (for TNFα only) and 6 h after injection (for IL-6 and IL-12p40 and for SAA and haptoglobin). ND, levels below limit of detection. B and C: qPCR analyses of hepatic (B) and epididymal adipose tissue (C) gene expression. SAP, serum amyloid P-component. WAT, white adipose tissue. D: Hepatic TAG levels. E: Plasma TAG levels at 6 h after injection. Data in A, D, and E represent the mean ± SEM (n = 6), and qPCR data in B and C represent the mean ± SEM (n = 5). Within each panel, values not sharing a common superscript letter differ significantly (P < 0.05). AU, arbitrary unit.