Hepatic n-3 polyunsaturated fatty acid depletion promotes steatosis and insulin resistance in mice: genomic analysis of cellular targets

Abstract

Patients with non-alcoholic fatty liver disease are characterised by a decreased n-3/n-6 polyunsaturated fatty acid (PUFA) ratio in hepatic phospholipids. The metabolic consequences of n-3 PUFA depletion in the liver are poorly understood. We have reproduced a drastic drop in n-3 PUFA among hepatic phospholipids by feeding C57Bl/6J mice for 3 months with an n-3 PUFA depleted diet (DEF) versus a control diet (CT), which only differed in the PUFA content. DEF mice exhibited hepatic insulin resistance (assessed by euglycemic-hyperinsulinemic clamp) and steatosis that was associated with a decrease in fatty acid oxidation and occurred despite a higher capacity for triglyceride secretion. Microarray and qPCR analysis of the liver tissue revealed higher expression of all the enzymes involved in lipogenesis in DEF mice compared to CT mice, as well as increased expression and activation of sterol regulatory element binding protein-1c (SREBP-1c). Our data suggest that the activation of the liver X receptor pathway is involved in the overexpression of SREBP-1c, and this phenomenon cannot be attributed to insulin or to endoplasmic reticulum stress responses. In conclusion, n-3 PUFA depletion in liver phospholipids leads to activation of SREBP-1c and lipogenesis, which contributes to hepatic steatosis.

Figure 1. Accumulation of lipids in the livers of n-3 PUFA depleted mice.

Oil red staining performed on frozen liver sections of fasted mice fed a control (CT) or n-3 PUFA-depleted (DEF) diet for 3 months. Bar = 50 µm (A). TG, free cholesterol and esterified cholesterol content in the livers of mice fed a control (CT) or n-3 PUFA-depleted (DEF) diet for 3 months. Data are the mean ± SEM. ^*: mean values significantly different (P<0.05, Student's t-test) (B).

Figure 2. n-3 PUFA depletion leads to decreased hepatic fatty acid oxidation and increased TG secretion and synthesis.

Mice were fed a control (CT) or n-3 PUFA-depleted (DEF) diet for 3 months. To measure hepatic TG secretion, fasted CT mice (closed squares, n = 4) and DEF mice (closed circles, n = 6) were injected with tyloxapol (0.5 mg tyloxapol/g body weight). Data are the mean ± SEM. ^*: mean values significantly different (P<0.05, two-way ANOVA; for the area under the curve, P<0.05, Student's t-test) (A). Precision-cut liver slices (PCLSs) obtained from fasted CT (n = 8) and DEF (n = 6) mice were incubated with [¹⁴C]-palmitate for 3 hours to measure CO₂ produced from fatty acid oxidation (B). TG synthesis from [¹⁴C]-acetate (C) and fatty acid esterification into TG from [¹⁴C]-palmitate (D) were measured in PCLSs obtained from fed CT (n = 7) and DEF mice (n = 8). Data are the mean ± SEM. ^*: mean values significantly different (P<0.05, Student's t-test).

Figure 3. n-3 PUFA-depleted mice exhibited higher mRNA content of enzymes involved in fatty acid and cholesterol synthesis.

n-3 PUFA depletion induces increased hepatic expression of enzymes involved in lipogenesis (A) and cholesterol synthesis (B). Microarray analysis was performed on a pool of RNA obtained from fasted mice fed a control (CT) or n-3 PUFA-depleted (DEF) diet for 3 months. In red are depicted enzymes and factors for which mRNA content in the liver of DEF mice was higher and in green those for which mRNA content was lower, than the values measured in CT mice in the fasted state. The color bare refers to the log₂ fold change in DEF mice versus value obtained in CT mice. No threshold was applied on the fold change. We applied a statistical significance threshold calculated by the MAS5 algorythm as recommended by the manufacturer (Affymetrix). EC  =  esterified cholesterol. See Table S2 for raw data.

Figure 4. n-3 PUFA depletion leads to hepatic SREBP-1 and SREBP-2 activation.

Liver immunoblot analysis of the hepatic SREBP-1 precursor (pSREBP-1) and nuclear (nSREBP-1) form (A) and of the hepatic SREBP-2 precursor (pSREBP-2) and nuclear (nSREBP-2) form (B) from fasted mice fed a control (CT; n = 6) or n-3 PUFA depleted (DEF; n = 7) diet for 3 months. Quantification of immunoblots is shown at right. Data are the mean ± SEM. ^*: mean values significantly different (P<0.05, Student's t-test).

Figure 5. n-3 PUFA-depleted mice exhibited hepatic insulin resistance.

Insulinemia (A) and glycemia (B) in fasted mice fed a control (CT; n = 6) or n-3 PUFA-depleted (DEF; n = 7) diet for 3 months. Hepatic glucose production (C) and glucose infusion rate (D) were measured during steady-state euglycemic-hyperinsulinemic clamp performed on fasted mice fed a control (CT; n = 4) or n-3 PUFA depleted (DEF; n = 6) diet for 3 months. Data are the mean ± SEM. ^*: mean values significantly different (P<0.05, Student's t-test).

Figure 6. Absence of hepatic ER stress under n-3 PUFA depletion.

Mice were fed a control (CT; n = 6) or n-3 PUFA-depleted (DEF; n = 7) diet for 3 months. The hepatic mRNA content of glucose-regulated protein 78 (GRP78), glucose-regulated protein 94 (GRP94), protein disulfide isomerise (PDI), C/EBP homologous protein (CHOP), ER degradation enhancer, mannosidase alpha-like 1 (EDEM1), unspliced (uXBP-1) and spliced (sXBP-1) X-box binding protein-1 is shown from fasted CT and DEF mice. The qPCR results were analysed according to the 2^{−ΔΔ ct} method and were normalised to RPL-19 mRNA (A). Immunoblot analysis of GRP78, PDI, GRP94 and total and phosphorylated eukaryotic translation initiation factor 2α (eIF2α) (B) in the livers from fasted CT and DEF mice. Quantification of immunoblots is shown below. Data are the mean ± SEM. ^*: mean values significantly different (P<0.05, Student's t-test).

Figure 7. Proposed metabolic pathways involved in fatty acid and cholesterol accumulation in the livers of n-3 PUFA depleted mice.

The reduction of fatty acid oxidation and the induction of lipogenesis both contribute to the accumulation of lipids in the livers of n-3 PUFA-depleted mice. The reduced fatty acid degradation could result from the inhibition of the PPARα pathway. De novo lipogenesis is promoted through SREBP-1c activation. Data suggest that SREBP-1c activation could result from both increased hepatic 2-AG content and LXR activation. The lower n-3 PUFA content in hepatic phospholipids can lead to higher 2-AG production, which could therefore stimulate CB1 and then SREBP-1c expression. Both hepatic n-3 PUFA depletion and higher cholesterol content contribute to LXR activation. Increased SREBP-2 activation, occurring by an unknown mechanism, could play a role in the increased cholesterol synthesis observed in the n-3 PUFA-depleted livers.