Receptor-interacting protein 1 and 3 kinase activity are required for high-fat diet induced liver injury in mice

Abstract

Background: The RIP1-RIP3-MLKL-mediated cell death pathway is associated with progression of non-alcohol-associated fatty liver/steatohepatitis (NAFL/NASH). Previous work identified a critical role for MLKL, the key effector regulating necroptosis, but not RIP3, in mediating high fat diet-induced liver injury in mice. RIP1 and RIP3 have active N-terminus kinase domains essential for activation of MLKL and subsequent necroptosis. However, little is known regarding domain-specific roles of RIP1/RIP3 kinase in liver diseases. Here, we hypothesized that RIP1/RIP3 kinase activity are required for the development of high fat diet-induced liver injury.

Methods: Rip1^K45A/K45A and Rip3^K51A/K51A kinase-dead mice on a C57BL/6J background and their littermate controls (WT) were allowed free access to a diet high in fat, fructose and cholesterol (FFC diet) or chow diet.

Results: Both Rip1^K45A/K45A and Rip3^K51A/K51A mice were protected against FFC diet-induced steatosis, hepatocyte injury and expression of hepatic inflammatory cytokines and chemokines. FFC diet increased phosphorylation and oligomerization of MLKL and hepatocyte death in livers of WT, but not in Rip3^K51A/K51A, mice. Consistent with in vivo data, RIP3 kinase deficiency in primary hepatocytes prevented palmitic acid-induced translocation of MLKL to the cell surface and cytotoxicity. Additionally, loss of Rip1 or Rip3 kinase suppressed FFC diet-mediated formation of crown-like structures (indicators of dead adipocytes) and expression of mRNA for inflammatory response genes in epididymal adipose tissue. Moreover, FFC diet increased expression of multiple adipokines, including leptin and plasminogen activator inhibitor 1, in WT mice, which was abrogated by Rip3 kinase deficiency.

Discussion: The current data indicate that both RIP1 and RIP3 kinase activity contribute to FFC diet-induced liver injury. This effect of RIP1 and RIP3 kinase deficiency on injury is consistent with the protection of Mlkl^-/- mice from high fat diet-induced liver injury, but not the reported lack of protection in Rip3^-/- mice. Taken together with previous reports, our data suggest that other domains of RIP3 likely counteract the effect of RIP3 kinase in response to high fat diets.

Keywords: FFC diet; NAFLD; RIP1 kinase; RIP3 kinase; cell death; obesity.

Figure 1

RIP1 kinase activity was required for FFC-induced phosphorylation of MLKL and liver injury. Rip1^+/+ and Rip1^K45A/K45A mice were allowed free access to chow or FFC diet for 12 weeks. (A) Immunohistochemistry staining for pMLKL in paraffin-embedded liver sections. Images were acquired using a 10X objective. Representative images are shown from n=5-6 per group. (B) Hematoxylin and eosin (H&E) staining of liver sections. Images were acquired at 10X magnification. (C) ALT/AST concentration in plasma and hepatic triglyceride content in liver homogenates. (D) mRNA expression of genes of interest in livers was detected by qRT-PCR. Values with different alphabetical superscripts are significantly different from each other, n = 5-6 per group. p <0.05, assessed by ANOVA.

Figure 2

RIP3 kinase activity was required for FFC-induced phosphorylation and oligomerization of MLKL in mouse liver and liver injury. Rip3^+/+ and Rip3^K51A/K51A mice were allowed free access to chow or FFC diet for 12 weeks. (A) Immunohistochemistry staining for pMLKL in paraffin-embedded liver sections. Images were acquired using a 10X objective. Representative images are shown from n=5-6 per group. (B) Subcellular fractions (cytosol, P10 (10,000 g pellet) and plasma membrane (PM) were isolated from liver, proteins separated by SDS-PAGE under non-reducing condition, and immunoblotted (IB) with anti-MLKL antibody. Enrichment of sub-cellular fractions was verified by probing for Na⁺/K⁺-ATPase (plasma and intracellular membrane marker) and β-actin (cytosolic marker). Blots are representative from n = 3 for PM and 6 for cytosol and P10 independent experiments. Western blots were semi-quantified on Image J. Values for MLKL oligomer in the P10 fraction were 0.05 ± 0.02^a for chow-fed and 0.32 ± 0.14^b for FFC-fed in Rip3^+/+ and 0.05 ± 0.03^a for chow-fed and 0.12 ± 0.04^a for FFC-fed in Rip3^K51A/K51A (p<0.05). Values for MLKL oligomer in the PM fraction were 0.07 ± 0.03^a for chow-fed and 0.24 ± 0.06^b for FFC-fed in Rip3^+/+ and 0.05 ± 0.02^a for chow-fed and 0.07 ± 0.02^a for FFC-fed in Rip3^K51A/K51A (p<0.05). (C) Hematoxylin and eosin (H&E) staining of liver sections. Images were acquired at 10X magnification. (D) ALT/AST concentration in plasma and hepatic triglyceride content in liver homogenates. Values represent means ± SEM. Values with different alphabetical superscripts are significantly different from each other, n = 5-6 per group. p <0.05, assessed by ANOVA.

Figure 3

RIP3 kinase activity contributed to FFC-induced liver inflammatory responses. Rip3^+/+ and Rip3^K51A/K51A mice were allowed free access to chow or FFC diet for 12 weeks. (A) Expression of chemokine (Mcp-1, Cxcl1 and Cxcl5) and cytokine (Tnf-α, Il-1β and Il-10) mRNA in livers. (B) Expression of immune cell marker Cd11c, acute-phase protein Crp-1, and phagocytic gene (Havcr2, Abca1) mRNA in livers. Expression was normalized to 18S rRNA and expressed as the percent increase over chow-fed controls. Values represent means ± SEM. Values with different superscripts are significantly different from each other, n = 5-6 per group. p <0.05, assessed by ANOVA.

Figure 4

RIP3 kinase activity was involved in FFC-induced metabolic disturbances and regulation of genes involved in lipogenesis. Rip3^+/+ and Rip3^K51A/K51A mice were allowed free access to chow or FFC diet for 12 weeks. (A) Plasma concentrations of blood glucose, insulin and calculated HOMA-IR, as well as (B) cholesterol and triglycerides in plasma from Rip3^+/+ and Rip3^K51A/K51A mice. (C) Expression of lipogenesis-related mRNA including Srebp-1c, Pparγ, Fabp4, Fas and Mgat, in the liver was determined by qRT-PCR. Expression was normalized to 18S rRNA and expressed as the percent increase over chow-fed controls. Values represent means ± SEM. Values with different superscripts are significantly different from each other, n = 5-6 per group. p <0.05, assessed by ANOVA.

Figure 5

Effect of RIP3 kinase activity in FFC-induced cell death in mouse liver and PA-mediated cytotoxicity in primary hepatocytes. Rip3^+/+ and Rip3^K51A/K51A mice were allowed free access to chow or FFC diet for 12 weeks. (A, B) Paraffin-embedded livers were de-paraffinized followed by quantitative analyses of (A) TUNEL or (B) M30 staining. Images were acquired using 20X or 10X objective. The blue arrows indicate TUNEL-positive hepatocytes and red arrows indicate TUNEL-positive NPCs. TUNEL-positive hepatocytes and NPCs were counted and expressed as the total number of cells per 20X frame and M30 positive cells were quantified per 10X frame, n = 5-6 per group. (C, D) Primary hepatocytes were isolated from chow-fed WT and Rip3^K51A/K51A mice and then exposed to 500 µM palmitic acid (PA) for 24h. (C) MTS assay was performed to determine PA-induced hepatotoxicity and (D) location of MLKL at the cell surface (phalloidin: F-actin) in response to PA visualized by confocal microscopy. n=4 independent isolations. Values represent means ± SEM. Values with different superscripts are significantly different from each other, n = 3 per group. p <0.05, assessed by ANOVA. N.D., not detectable. TUNEL, Terminal deoxynucleotidyl transferase dUTP nick end labeling.

Figure 6

Absence of Rip1 or Rip3 kinase activity prevented FFC diet-induced inflammatory responses in the adipose. Mice were allowed free access to chow or FFC diet for 12 weeks and epididymal white adipose tissue collected and analyzed. (A/C) H&E staining of epididymal white adipose tissues and quantification of crown-like structures per 10X frame in (A) Rip1^+/+ and Rip1^K45A/K45A mice and (C) Rip3^+/+ and Rip3^K51A/K51A mice. (B/D) Expression of genes of interest were measured by qRT-PCR in adipose tissue from (B) Rip1^+/+ and Rip1^K45A/K45A mice and (D) Rip3^+/+ and Rip3^K51A/K51A mice. Values represent means ± SEM. Values with different superscripts are significantly different from each other, n = 5-6 per group. p <0.05, assessed by ANOVA.

Figure 7

Plasma adipokine profiles in epididymal adipose from Rip3 kinase deficient mice. (A) Mice were allowed free access to chow or FFC diet for 12 weeks and epididymal white adipose tissue collected and analyzed. (A) Plasma adipokine profiles were assess using an adipokine array. (B) Relative density was calculated using Image J; data were normalized to reference spots, and each experimental group was expressed as fold change of chow-fed Rip3^+/+ group. (C) Expression of mRNA for Leptin and Serpine-1 (PAI-1 gene) was normalized to 18S rRNA. Values represent means ± SEM. Values with different superscripts are significantly different from each other, n = 6 per group. p <0.05, assessed by ANOVA.