Linagliptin Ameliorates Hepatic Steatosis via Non-Canonical Mechanisms in Mice Treated with a Dual Inhibitor of Insulin Receptor and IGF-1 Receptor

Abstract

Abnormal hepatic insulin signaling is a cause or consequence of hepatic steatosis. DPP-4 inhibitors might be protective against fatty liver. We previously reported that the systemic inhibition of insulin receptor (IR) and IGF-1 receptor (IGF1R) by the administration of OSI-906 (linsitinib), a dual IR/IGF1R inhibitor, induced glucose intolerance, hepatic steatosis, and lipoatrophy in mice. In the present study, we investigated the effects of a DPP-4 inhibitor, linagliptin, on hepatic steatosis in OSI-906-treated mice. Unlike high-fat diet-induced hepatic steatosis, OSI-906-induced hepatic steatosis is not characterized by elevations in inflammatory responses or oxidative stress levels. Linagliptin improved OSI-906-induced hepatic steatosis via an insulin-signaling-independent pathway, without altering glucose levels, free fatty acid levels, gluconeogenic gene expressions in the liver, or visceral fat atrophy. Hepatic quantitative proteomic and phosphoproteomic analyses revealed that perilipin-2 (PLIN2), major urinary protein 20 (MUP20), cytochrome P450 2b10 (CYP2B10), and nicotinamide N-methyltransferase (NNMT) are possibly involved in the process of the amelioration of hepatic steatosis by linagliptin. Thus, linagliptin improved hepatic steatosis induced by IR and IGF1R inhibition via a previously unknown mechanism that did not involve gluconeogenesis, lipogenesis, or inflammation, suggesting the non-canonical actions of DPP-4 inhibitors in the treatment of hepatic steatosis under insulin-resistant conditions.

Keywords: DPP-4 inhibitors; diabetes; hepatic steatosis; insulin resistance; insulin signaling; phosphoproteomics; proteomics.

Figure 1

Linagliptin (Lina) improved OSI-906-induced hypertriglyceridemia in mice after 7 days of OSI-906 administration. (a) Experimental protocol (n = 5–6). (b) Body weight during the experiment. Data represent the mean ± SEM. * p < 0.05, OSI-906 vs. vehicle; ^† p < 0.05, OSI-906 vs. Lina by repeated measures ANOVA followed by Bonferroni multiple comparison test (n = 5–6 per group). (c) Blood glucose levels determined just before and 4 h after the administration of OSI-906 or the vehicle during the experiment. ** p < 0.01, OSI-906 vs. vehicle; ^†† p < 0.01, OSI-906 vs. Lina, ^ǂ p < 0.01, OSI-906 + Lina vs. vehicle; ^§ p < 0.01, OSI-906 + Lina vs. Lina; ^ǁ p < 0.01, OSI-906 + Lina vs. OSI-906; ^¶ p < 0.01, Lina vs. vehicle by repeated measures of ANOVA followed by Bonferroni multiple comparison test (n = 5–6 per group). (d) Serum insulin, (e) serum triglyceride (TG), (f) serum free fatty acid (FFA), and (g) serum glutamic pyruvic transaminase (GPT) levels on day 7. Data represent the mean ± SEM. * p < 0.05, ** p < 0.01 (n = 5–6 per group) by ANOVA with an additional Tukey–Kramer post-hoc test.

Figure 2

Linagliptin improved hepatic steatosis evoked by OSI-906. (a) Hematoxylin and eosin-stained sections of liver on day 7. Scale bar = 200 μm. (b) Ratio of liver weight to body weight on day 7. (c,d) Triglyceride (TG) and glycogen content in the liver on day 7. Data represent the mean ± SEM. * p < 0.05, ** p < 0.01 (n = 5 per group) by ANOVA with an additional Tukey–Kramer post-hoc test. (e) Masson–Goldner-stained section of liver on day 7. Scale bar = 200 μm. (f) Non-alcoholic fatty liver disease (NAFLD) activity score (NAS), (g) fibrosis staging, and (h) the degree of steatosis, (i) hepatocyte ballooning, (j) lobular inflammation of liver sections according to the NAFLD activity score (NAS) score. Data represent the mean ± SEM. *p < 0.05, ** p < 0.01 (n = 4–6 per group) by Kruskal–Wallis test.

Figure 3

Hepatic gene expressions of the indicated molecules on day 7. Data represent the mean ± SEM. * p < 0.05, ** p < 0.01 (n = 4–5 per group) by ANOVA with an additional Tukey–Kramer post-hoc test.

Figure 4

Proteomic and phosphoproteomic analysis of the liver from OSI-906- and linagliptin-treated mice. (a) Protocol for proteomic and phosphoproteomic analyses using liver samples (n = 5 per group). (b) Volcano plot of hepatic proteomic data. Molecules identified in comparisons of vehicle vs. OSI-906 (left) and OSI-906 vs. OSI-906 + linagliptin (right) are shown (n = 5 per group). (c) Venn diagram of differentially expressed molecules in OSI-906 vs. vehicle and OSI-906 + Lina vs. OSI-906 in proteomic analyses of liver samples. The numbers of proteins that were significantly upregulated or downregulated (ANOVA p-value < 0.05, fold change > 1.2) compared with the respective controls are shown (n = 5 per group). The lists of overlapping proteins are shown in Supplementary Table S1.

Figure 5

Linagliptin reversed the increased expression of acetylated lysine, perilipin-2, and nicotinamide N-methyltransferase (NNMT) in OSI-906-treated liver. (a) Immunoblotting for pan acetyl-lysine, perilipin-2, and NNMT in the liver (n = 3 per group). Densitometry was plotted in lower graphs. (b) Hepatic gene expressions of Cyp2b10, Plin2, and Nnmt on day 7. Data represent the mean ± SEM. * p < 0.05, ** p < 0.01 (n = 4–5 per group) by ANOVA with an additional Tukey–Kramer post-hoc test.

Figure 6

Canonical pathway analysis of proteomics in the livers of OSI-906- and linagliptin-treated mice. A canonical pathway analysis of the hepatic proteomic data was performed using molecules that were differentially expressed (ANOVA p < 0.05) in comparisons of OSI-906 vs. vehicle and OSI-906 + Lina vs. OSI-906. The color of the bar indicates the z-score (orange, positive z-score; blue, negative z-score). The orange line depicts the statistical significance threshold (p = 0.05).

Figure 7

Canonical pathway analysis of phosphoproteomics in the livers of OSI-906- and linagliptin-treated mice. A canonical pathway analysis of the hepatic phosphoproteomic data was performed using molecules that were differentially expressed (ANOVA p < 0.05) in comparisons of OSI-906 vs. vehicle or OSI-906 + Lina vs. OSI-906. The color of the bar indicates the z-score (orange, positive z-score; blue, negative z-score). The orange line depicts the statistical significance threshold (p = 0.05).

Figure 8

Impact of linagliptin or liraglutide on gene expressions in OSI-906-treated AML-12 cells. The gene expressions of the indicated genes in AML-12 cells treated with 10 nM linagliptin or 100 nM liraglutide in the presence of 200 nM OSI-906 are shown. Cells were serum-starved overnight and then incubated with OSI-906 for 4 h before the treatment with linagliptin or liraglutide. Then, the cells were incubated for 24 h with the indicated drugs. Data represent the mean ± SEM. * p < 0.05, ** p < 0.01 (n = 4–5 per group) by ANOVA with an additional Tukey–Kramer post-hoc test.

Figure 9

Schematic regulatory pathway of linagliptin in OSI-906-induced hepatic steatosis. Linagliptin improved hepatic steatosis induced by acute IR/IGF1R signaling inhibition with OSI-906 through an insulin signaling-independent pathway. Pathways involving perilipin-2 and NNMT have been proposed as possible mechanisms for the amelioration of OSI-906-induced hepatic steatosis by linagliptin.