PCB 126 and other dioxin-like PCBs specifically suppress hepatic PEPCK expression via the aryl hydrocarbon receptor

Abstract

Dioxins and dioxin-like compounds encompass a group of structurally related heterocyclic compounds that bind to and activate the aryl hydrocarbon receptor (AhR). The prototypical dioxin is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), a highly toxic industrial byproduct that incites numerous adverse physiological effects. Global commercial production of the structurally similar polychlorinated biphenyls (PCBs), however, commenced early in the 20(th) century and continued for decades; dioxin-like PCBs therefore contribute significantly to total dioxin-associated toxicity. In this study, PCB 126, the most potent dioxin-like PCB, was evaluated with respect to its direct effects on hepatic glucose metabolism using primary mouse hepatocytes. Overnight treatment with PCB 126 reduced hepatic glycogen stores in a dose-dependent manner. Additionally, PCB 126 suppressed forskolin-stimulated gluconeogenesis from lactate. These effects were independent of acute toxicity, as PCB 126 did not increase lactate dehydrogenase release nor affect lipid metabolism or total intracellular ATP. Interestingly, provision of cells with glycerol instead of lactate as the carbon source completely restored hepatic glucose production, indicating specific impairment in the distal arm of gluconeogenesis. In concordance with this finding, PCB 126 blunted the forskolin-stimulated increase in phosphoenolpyruvate carboxykinase (PEPCK) mRNA levels without affecting glucose-6-phosphatase expression. Myricetin, a putative competitive AhR antagonist, reversed the suppression of PEPCK induction by PCB 126. Furthermore, other dioxin-like PCBs demonstrated similar effects on PEPCK expression in parallel with their ability to activate AhR. It therefore appears that AhR activation mediates the suppression of PEPCK expression by dioxin-like PCBs, suggesting a role for these pollutants as disruptors of energy metabolism.

Figure 1. Preliminary validation of PCB effects on metabolic functions and toxicity in primary mouse hepatocytes.

(A) Suppression of total intracellular glycogen by increasing doses of PCB 126; *p<0.05 vs. DMSO control. (B) Maximum suppression of total intracellular glycogen at 100 nM PCB 126 after 16 h of incubation; *p<0.05 vs. DMSO control. (C) PCB 126 is not cytotoxic at up to 1 mM, and also does not affect cellular expression of LDH. (D) PCB 126 has no effect on mitochondrial beta oxidation of palmitic acid; *p<0.05 vs. DMSO for each respective beta oxidation condition. (E) PCB 126 has no effect on basal or insulin-stimulated lipogenesis; *p<0.05 for insulin effect vs. respective no-insulin control. Unless otherwise noted, all PCB treatments were 16 h in length. Data are average of 3 experiments; error bars are −/+ SD.

Figure 2. Differential impact of PCB 126 forskolin stimulated hepatic GNG with lactate vs. glycerol.

(A) With 10 mM lactate as the carbon source, PCB 126 (100 nM) has no effect on basal GNG, but suppresses forskolin-stimulated glucose output. (B) With 10 mM glycerol as the carbon source, PCB 126 has no effect on either basal or forskolin-stimulated GNG, although total GNG is approximately 2-fold higher for all conditions. *p<0.05 for forskolin effect relative to respective −/+ PCB condition, #p<0.05 for forskolin effect in DMSO vs. PCB condition. All PCB treatments were 16 h in length and forskolin (25 µM) was added at the start of the GNG assay as described in Materials and Methods. Data are average of 3 experiments; error bars are −/+ SD.

Figure 3. Differential impact of PCB 126 on key forskolin-responsive gluconeogenic genes.

(A) PCB 126 has no effect on basal PEPCK or G6Pc expression. (B) PCB 126 selectively suppresses forskolin induction of PEPCK gene expression, but not that of G6pc. (C) Insulin strongly suppresses forskolin induction of PEPCK, and the effect of PCB 126 is additive with that of insulin; *p<0.05 for insulin effect within each condition and gene, #p<0.05 for PCB 126 effect on PEPCK induction independent of insulin. All PCB 126 treatments were 16 h in length while forskolin (25 µM) and insulin (10 nM) were added at the start of the assay. Data are average of 3 experiments for panels A–B, and 2 experiments for panel C; error bars are −/+ SD.

Figure 4. Interaction of CYP1A1, CYP1A2, EROD, and G6Pc versus PEPCK gene expression with PCB 126 dose-response.

(A) EC₅₀ of PCB 126 for CYP1A1gene expression and suppression of forskolin induction of PEPCK gene expression closely correlate. (B) EC₅₀ of PCB 126 for CYP1A2 gene expression and suppression of forskolin induction of PEPCK gene expression are modestly correlated but demonstrate the expected directionality. (C) EC₅₀ of PCB 126 for EROD activity induction and suppression of forskolin induction of PEPCK gene expression closely correlate. (D) Forskolin-induced G6pc gene expression fails to demonstrate a clear relationship with dose-dependent effect of PCB 126 on forskolin-induced PEPCK gene expression. Tested concentrations of PCB 126 in base 10 format are: 0.05, 0.20 1.0, 5.0, 10.0, 20.0, 100.0, and 500.0. Alphabetical letters separate doses of PCB 126 that elicited significantly (p<0.05) different degrees of response for the parameter designated on the left-hand y-axis (lowercase) or right-hand y-axis (uppercase). Data are average of 3 experiments; error bars are −/+ SD.

Figure 5. Validation and evaluation of myricetin, a putative competitive AhR inhibitor.

Doses of PCB 126 were selected based on the concentrations near the middle (5 nM) or top (10–20 nM) of the respective dose-response curves (established in figure 4). (A) Maximally effective myricetin doses (25 µM) significantly antagonize PCB 126 –induced CYP1A1 gene expression. (B) Similar effect of 25 µM myricetin on PCB 126 induction of CYP1A2 gene expression. (C) 25 µM myricetin antagonizes EROD activity induction by PCB 126. (D) Myricetin homogeneously suppresses forskolin-induced PEPCK expression and also restores PEPCK gene induction by forskolin in the presence of PCB 126. (E) Myricetin homogeneously suppresses forskolin-induced G6pc gene induction by PCB 126, which has no impact on forskolin induction of G6pc expression. (Panels A–C) *p<0.05 for myricetin effect with respective PCB 126 concentration. (Panels D–E) #p<0.05 for myricetin effect within each [PCB 126]. (All panels) Data are average of 2–3 experiments; error bars are −/+ SD.

Figure 6. Other dioxin-like PCBs that activate AhR suppress PEPCK gene induction.

(A) Dioxin-like PCBs 126, 169, 81, and 77 induce EROD activity, with dose-dependency for all dioxin-like PCBs with the exception of PCB 126. (B) Similar trends observed for when CYP1A1 was used as the readout instead of EROD. (C) CYP1A2 gene expression demonstrates a pattern in line with observations made in panels A and B. (D) Forskolin-induced PEPCK gene expression is suppressed by other dioxin-like PCBs in a dose-dependent fashion, with the exception of PCB 126. (E) G6pc gene induction is not affected by any tested PCBs at any tested concentration. *p<0.05 vs. DMSO, significant differences between doses of PCBs are designated by arrows. Note: for clarity, the PCB prefix has been eliminated for the tested congeners on the x-axis label. Data are average of 2–3 experiments; error bars are −/+ SD.

Figure 7. Simplified visual model of the impact of PCB 126 on hepatic glucose metabolism.

Glycolysis, glycogen synthesis, and gluconeogenesis in a mouse hepatocyte are presented. Forward metabolism of glucose (glycolysis and glycogen synthesis) is depicted by black arrows, while reverse metabolism of glucose (GNG) is denoted by gray arrows. Enzymes involved in forward glucose metabolism are in standard bold text, while enzymes involved in reverse glucose metabolism are in italicized bold text. In this model, PCB 126 inhibits GNG from lactate (and pyruvate) by suppression of PEPCK gene transcription, and also inhibits glycogen by an unknown mechanism. GNG from glycerol is not affected by PCB 126 because glycerol is under the control of Fbp1 and G6pc under conditions that promote GNG, and thus bypasses PEPCK.