Deregulation of hepatic insulin sensitivity induced by central lipid infusion in rats is mediated by nitric oxide

Abstract

Background: Deregulation of hypothalamic fatty acid sensing lead to hepatic insulin-resistance which may partly contribute to further impairment of glucose homeostasis.

Methodology: We investigated here whether hypothalamic nitric oxide (NO) could mediate deleterious peripheral effect of central lipid overload. Thus we infused rats for 24 hours into carotid artery towards brain, either with heparinized triglyceride emulsion (Intralipid, IL) or heparinized saline (control rats).

Principal findings: Lipids infusion led to hepatic insulin-resistance partly related to a decreased parasympathetic activity in the liver assessed by an increased acetylcholinesterase activity. Hypothalamic nitric oxide synthases (NOS) activities were significantly increased in IL rats, as the catalytically active neuronal NOS (nNOS) dimers compared to controls. This was related to a decrease in expression of protein inhibitor of nNOS (PIN). Effect of IL infusion on deregulated hepatic insulin-sensitivity was reversed by carotid injection of non selective NOS inhibitor NG-monomethyl-L-arginine (L-NMMA) and also by a selective inhibitor of the nNOS isoform, 7-Nitro-Indazole (7-Ni). In addition, NO donor injection (L-arginine and SNP) within carotid in control rats mimicked lipid effects onto impaired hepatic insulin sensitivity. In parallel we showed that cultured VMH neurons produce NO in response to fatty acid (oleic acid).

Conclusions/significance: We conclude that cerebral fatty acid overload induces an enhancement of nNOS activity within hypothalamus which is, at least in part, responsible fatty acid increased hepatic glucose production.

Figure 1. NO production by VMH neurons in response to oleic acid.

A and B: Bright field (A) and DAF-FM fluorescence intensity images of cultured (24 H) VMH neurons from rats. Cells (solid arrow) are plated with fluorescent beads (broken arrow) which are used to calibrate changes in cell fluorescence intensity. C: Example of VMH neurons which increase (solid line, 1 neuron) or do not change (dashed line, 8 neurons) their DAF-FM fluorescence intensity in response to oleic acid (2 µM). Neurons increasing their DAF-FM fluorescence intensity above 5% after 15–25 min post oleic acid addition are considered to produce NO in response to oleic acid. D: Percentage of VMH neurons that increased their DAF-FM fluorescence intensity [NO production] in response to oleic acid (2 µM) in the presence or absence (Control) of the non-selective NOS inhibitor N-Methyl-L-arginine (L-NMMA, 1 mM). Data are presented as mean±sem. *: p<0.05. Total number of cells – number of dishes used are given on the top of each bar.

Figure 2. Effect of lipid infusion on central NOS activities.

A: Hypothalamic and cortical NOS activities after 24 h of Intralipid/heparin infusion (IL 24 h) or saline/heparin infusion control group (C). Results are expressed as c.p.m per mg of tissue. B: Hypothalamic NOS activities of IL and C rats treated or not with acute injection of non selective NOS inhibitor, L-NMMA. The last right bar displays NOS activities in C rats injected with NOS substrate, L-Arginine. Results are expressed as percentage of C group activities. * p<0.05 vs. C.

Figure 3. Hepatic parameters after 24 h of infusion in C rats (NaCl/heparin) and in IL rats (IL/hep).

A: liver triglycerides content, B: liver PEPCK activity, C: liver glycogen, D: liver acetylcholinesterase activity. Values are means±SEM in each group. * p<0.05 vs. controls.

Figure 4. Assessment of glucose induced insulin secretion.

Time course of plasma glucose and insulin (A) concentrations in response to glucose load in C rats (open squares), C+L-NMMA rats (open triangles, dotted line), IL rats (filled squares) and IL+L-NMMA rats (filled triangles, dotted line). B, C, D, E, F: insulinogenic index (ΔI/ΔG), i.e. the area under curve of insulinemia time course divided by the area under curve of glycemia time course, after carotid injections of respectively with L-NMMA, 7-Ni, W1400, L-Arginine and SNP. Drugs or saline are injected 5 minutes before glucose loading into saphenous vein. Values are means±SEM in each group. * p<0.01 vs C rats. L-NMMA: non selective NOS inhibitor. 7-Ni: selective neuronal NOS inhibitor. W1400: selective inflammatory NOS inhibitor. SNP: chemical NO donor. IL 24 h: rats infused with of Intralipid/heparin through the carotid artery towards the brain for 24 h. C: rats infused for 24 h with saline/heparin.

Figure 5. Glucose turnover rate during euglycemic-hyperinsulinemic clamps.

Glucose Utilization and Hepatic Glucose Production during euglycemic-hyperinsulinemic clamp (0.4 U·kg⁻¹·h⁻¹) in Control and IL rats before and after L-NMMA (A), 7-Ni (B). L-Arginine (C) or SNP (E) carotid injection. L-NMMA: non selective NOS inhibitor. 7-Ni: selective neuronal NOS inhibitor. W1400: selective inflammatory NOS inhibitor. SNP: chemical NO donor. Values are means±SEM in each group. * p<0.01 vs C rats.

Figure 6. Hepatic parameters during euglycemic-hyperinsulinemic clamp procedure.

A: Glycogen content, B: Glucose-6-Phosphate content, C: PEPCK activity in the liver of control and IL rats (injected or not with non selective NOS inhibitor, L-NMMA) at the end of the clamp procedure. Values are means±SEM in each group. * p<0.01 vs. C rats; § p<0.01 vs. IL rats.

Figure 7. Hypothalamic gene expression after 24 h IL infusion.

A: mRNA expression of NO pathway proteins, inflammatory cytokines and oxidative stress enzymes in the hypothalamus of C rats and IL rats. Values are means±SEM in each group (arbitrary units). B: Western blot of nNOS monomers and dimers in C and IL rats, and D: expression of the ratio dimers/monomers (D/M) after image analysis (quantification of the intensity of the blots). Values are means±SEM in each group. * p<0.01 vs C rats.