Short-term tamoxifen administration improves hepatic steatosis and glucose intolerance through JNK/MAPK in mice

Abstract

Nonalcoholic fatty liver disease (NAFLD) which is a leading cause of chronic liver diseases lacks effective treatment. Tamoxifen has been proven to be the first-line chemotherapy for several solid tumors in clinics, however, its therapeutic role in NAFLD has never been elucidated before. In vitro experiments, tamoxifen protected hepatocytes against sodium palmitate-induced lipotoxicity. In male and female mice fed with normal diets, continuous tamoxifen administration inhibited lipid accumulation in liver, and improved glucose and insulin intolerance. Short-term tamoxifen administration largely improved hepatic steatosis and insulin resistance, however, the phenotypes manifesting inflammation and fibrosis remained unchanged in abovementioned models. In addition, mRNA expressions of genes related to lipogenesis, inflammation, and fibrosis were downregulated by tamoxifen treatment. Moreover, the therapeutic effect of tamoxifen on NAFLD was not gender or ER dependent, as male and female mice with metabolic disorders shared no difference in response to tamoxifen and ER antagonist (fulvestrant) did not abolish its therapeutic effect as well. Mechanistically, RNA sequence of hepatocytes isolated from fatty liver revealed that JNK/MAPK signaling pathway was inactivated by tamoxifen. Pharmacological JNK activator (anisomycin) partially deprived the therapeutic role of tamoxifen in treating hepatic steatosis, proving tamoxifen improved NAFLD in a JNK/MAPK signaling-dependent manner.

Fig. 1

Tamoxifen decreased hepatocyte lipotoxicity in vitro. a Huh7, HepG2, AML12 cells, and primary hepatocytes were seeded in six-well plates. After 12 h, 0.3 mM sodium palmitate was added to the medium and after 36 h, DMSO/tamoxifen (10, 20, 40 μM) was added. After 36 h, cells were stained with ORO, pictured using an inverted phase contrast microscope, and quantified by ORO staining areas. Scale bar: 100 μm. b Statistical analysis of ORO staining. For cellular TC and TG test, cells were digested by trypsin, washed with PBS and collected using centrifuge. The following steps were performed according to the manufacturer’s protocols. All experiments were repeated at least three biological times. Bars = means ± SD; n = 3–10; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 2

Tamoxifen inhibited lipid accumulation in mice fed with normal diet. a Dosing scheme of tamoxifen on male and female C57BL/6 mice fed with normal diets. Dose of tamoxifen: 100 mg/kg. b Since initiating tamoxifen administration, mice were pair-fed and body weight was recorded weekly until mice were sacrificed. c Food consumption weight per cage was recorded when mice were administrated with tamoxifen. d Since initiating tamoxifen administration, mice were free-fed and body weight was recorded every other day for 4 weeks. e Food consumption weight per cage was recorded weekly when free-fed mice were administrated with tamoxifen. f Serum TC, HDL-C and LDL-C analysis of male and female mice fed with normal diets and administrated with vehicle/tamoxifen for 1, 2, 4, 8 weeks, respectively. g GTT test was performed on male and female mice administrated with tamoxifen or vehicle for 8 weeks and area under curve (AUC) was calculated and compared. h ITT test was performed on male and female mice administrated with tamoxifen or vehicle for 8 weeks and area under curve (AUC) was calculated and compared. i Frozen liver sections from male and female mice administrated with tamoxifen or vehicle for 1 and 4 weeks was performed ORO staining and ORO staining area was quantitatively compared. Scale bar:100 μm. j Frozen liver sections from male and female mice administrated with tamoxifen or vehicle for 8 weeks were performed ORO staining and ORO staining area was quantitatively compared. Scale bar:100 μm. Bars = means ± SD; n = 3– 5; ns, no significance; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 3

Tamoxifen alleviated hepatic steatosis in MCD and CDAA-induced models. a Dosing scheme of tamoxifen on male C57BL/6 mice fed with MCD diets for 6 weeks. Dose of tamoxifen: 100 mg/kg. b Liver sections from tamoxifen group and vehicle group in mice fed with MCD diets were performed H&E, ORO, F4/80 IF, and PSR staining. H&E fat cavitation area, ORO staining area, F4/80 positive cells percentage and PSR staining area were quantitatively compared. Scale bar: 100 μm. c Dosing scheme of tamoxifen on male C57BL/6 mice fed with CDAA diets for 10 weeks. Dose of tamoxifen: 100 mg/kg. d Liver sections from tamoxifen group and vehicle group in mice fed with CDAA diets were performed H&E, ORO, F4/80 IF, and PSR staining. H&E fat cavitation area, ORO staining area, F4/80 positive cells percentage and PSR staining area were quantitatively compared. Scale bar:100 μm. e RNA was extracted from liver tissues of MCD diets-induced NASH mice administrated with tamoxifen or vehicle and expression of lipogenesis, inflammation, and fibrosis-related genes was determined by RT-qPCR with β-actin as an internal control. f RNA was extracted from liver tissues of CDAA diets-induced NASH mice administrated with tamoxifen or vehicle and expression of lipogenesis, inflammation, and fibrosis-related genes were determined by RT-qPCR with β-actin as an internal control. g Serum analysis of TC, TG, HDL and LDL in MCD diets-induced mice administrated with tamoxifen or vehicle. h Liver tissues from MCD and CDAA-induced NASH mice administrated with tamoxifen or vehicle were harvested and TG concentrations were measured using commercial kits. Hepatic TG contents were normalized by hepatic protein levels. Bars = means ± SD; n = 3–6; ns, no significance; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 4

Tamoxifen ameliorated NAFLD in HFD-induced model. a Dosing scheme of tamoxifen on male C57BL/6 mice fed with high-fat diets for 20 weeks. Dose of tamoxifen: 100 mg/kg. b Liver sections from tamoxifen group and vehicle group in mice fed with high fat diets were performed H&E, ORO, F4/80 IF, and PSR staining. Mice fed with normal diets and administrated with vehicle were as controls. H&E fat cavitation area, ORO staining area, F4/80 positive cells percentage and PSR staining area were quantitatively compared. Scale bar:100 μm. c Serum analysis of ALT, AST, TC, TG, and LDL in high fat diets-induced mice administrated with tamoxifen or vehicle. Mice fed with normal diets and administrated with vehicle were as controls. d GTT experiment was performed on control mice or high fat diets-induced mice administrated with tamoxifen or vehicle and area under curve was quantitatively compared. e ITT experiment was performed on control mice or high fat diets-induced mice administrated with tamoxifen or vehicle and area under curve was quantitatively compared. f Hepatic TG levels were examined and normalized by protein levels. g RNA was extracted from isolated hepatocytes of control mice or high fat diets-induced mice administrated with tamoxifen or vehicle and expression of lipogenic genes was determined by RT-qPCR with β-actin as an internal control. h Dosing scheme of oral tamoxifen administration on male C57BL/6 mice fed with high-fat diets for 20 weeks. Dose of tamoxifen: 100 mg/kg. i Male C57BL/6 mice fed with high-fat diets for 20 weeks were administrated with 100 mg/kg tamoxifen intraperitoneally for 2 weeks and ceased treatment for 2 or 4 weeks. j Liver sections from male HFD-induced mice administrated with tamoxifen or vehicle orally for 2 weeks were performed H&E and ORO staining and H&E fat cavitation area, and ORO staining area was quantitatively compared. Scale bar:100 μm. k Hepatic TG levels were examined and normalized by protein levels. l GTT and ITT experiments were performed on male HFD-induced mice administrated with tamoxifen or vehicle orally for 2 weeks and area under curve was quantitatively compared. m Liver sections from male HFD-induced mice administrated with tamoxifen or vehicle for 2 weeks and ceased tamoxifen treatment for 2 or 4 weeks were performed H&E and ORO staining and H&E fat cavitation area and ORO staining area was quantitatively compared. Scale bar: 100 μm. n Hepatic TG levels were examined and normalized by protein levels. Bars = means ± SD; n = 3 to 6; ns, no significance; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 5

The therapeutic role of tamoxifen was independent of sex disparity and estrogen receptor. a Dosing scheme of tamoxifen on female C57BL/6 mice fed with MCD diets for 6 weeks. Dose of tamoxifen: 100 mg/kg. b Dosing scheme of tamoxifen combined with ER antagonist-fulvestrant on C57BL/6 mice fed with MCD diets for 6 weeks. Dose of tamoxifen: 100 mg/kg. c Serum analysis of ALT, AST, TC, TG, HDL, and LDL in MCD diets-induced male and female mice administrated with tamoxifen or vehicle. d Comparison of the decreased serum TC, TG, HDL, and LDL concentrations caused by tamoxifen between male and female mice. e Liver sections from male or female MCD diets-induced mice administrated with tamoxifen or vehicle were performed H&E and ORO staining. H&E fat cavitation area and ORO staining area were quantitatively compared. Scale bar: 100 μm. f Hepatic TG measurement in female MCD-induced NASH mice administrated with tamoxifen or vehicle. g Comparison of the decreased H&E fat cavitation area and ORO staining area caused by tamoxifen between male and female mice. h Serum analysis of ALT, AST, TC, TG, HDL, and LDL in male MCD diets-induced mice administrated with tamoxifen or tamoxifen + fulvestrant. i Protein levels of estrogen receptor in liver tissues from NASH mice administrated with tamoxifen w/o fulvestrant were determined by western blotting and quantitatively compared. j Liver sections from male MCD diets-induced mice administrated with vehicle, fulvestrant, tamoxifen or tamoxifen + fulvestrant were performed H&E and ORO staining. H&E fat cavitation area and ORO staining area were quantitatively compared. Scale bar: 100 μm. k Hepatic TG levels were examined on male MCD diets-induced mice administrated with vehicle, fulvestrant, tamoxifen, or tamoxifen + fulvestrant. l Liver sections from female MCD diets-induced mice administrated with vehicle, fulvestrant, tamoxifen or tamoxifen + fulvestrant were performed H&E and ORO staining. H&E fat cavitation area and ORO staining area were quantitatively compared. Scale bar: 100 μm. m Hepatic TG levels were examined on female MCD diets-induced mice administrated with vehicle, fulvestrant, tamoxifen, or tamoxifen + fulvestrant. n WT and ER knockdown AML12 cells were treated with PA for 36 h and then treated with 40 μM tamoxifen for another 36 h. ORO staining and cellular TG measurement were performed. Scale bar:100 μm. Bars = means ± SD; n = 3–6; ns, no significance; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 6

Tamoxifen inhibited JNK/MAPK signaling pathway. RNA was extracted from isolated hepatocytes of male MCD diets-induced NASH mice administrated with tamoxifen or vehicle and RNA-seq was performed. Fold change > 1.5, P < 0.05 or FDR < 0.25 were considered significant. a Gene expression heatmap of lipid metabolism, inflammation, fibrosis and apoptosis-related genes. b KEGG analysis of DEGs. c Heatmap of MAPK signaling pathway-related DEGs. d GSEA analysis of MAPK and Ras signaling pathways. e GSEA analysis of MAPK signaling pathway and inflammation, fibrosis and apoptosis-related signaling pathways. Fig. a–e were generated on Omicsmart platform (https://www.omicsmart.com/). f Protein levels of three key MAPK molecules-P38, JNK, and ERK were determined by western blotting and quantitatively compared. g Total protein was extracted from liver tissues in normal diets-fed mice administrated with tamoxifen or vehicle for 8 weeks. Protein levels of JNK and p-JNK were determined by western blotting and quantitatively compared. Bars = means ± SD; n = 3 to 4; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 7

Pharmacological activation of JNK/MAPK signaling pathway partly abolished the therapeutic effect of tamoxifen. a Dosing scheme of tamoxifen coupled with JNK activator-anisomycin (ANI) on male C57BL/6 mice fed with MCD diets for 6 weeks. Dose of tamoxifen: 100 mg/kg. Dose of anisomycin: 50 mg/kg. b Total protein was extracted from liver tissues in mice administrated with vehicle, ANI, tamoxifen or tamoxifen + ANI. Protein levels of JNK and p-JNK were determined by western blotting and quantitatively compared. c Serum analysis of ALT, AST, TC, TG, and HDL in mice administrated with vehicle, tamoxifen or tamoxifen + ANI. d Liver sections from MCD diets-induced mice administrated with vehicle, tamoxifen, or tamoxifen + ANI were performed H&E and ORO staining. Scale bar:100 μm. e H&E fat cavitation area and ORO staining area were quantitatively compared. f Hepatic TG levels were measured and normalized by protein levels. g AML12 cells were stimulated by 0.3 mM PA and treated with 40 μM tamoxifen. 10 μM anisomycin or DMSO was added 2 h before tamoxifen treatment. ORO staining was performed. Scale bar: 100 μm. h Cellular TG contents were measured and quantitatively compared. i Total proteins were extracted from AML12 cells treated with DMSO/ ANI/ tamoxifen/ tamoxifen+ANI. Protein levels of JNK and p-JNK were determined by western blotting. j The ratio of p-JNK/JNK and p-JNK/GAPDH were quantitatively compared. k AML12 cells were stimulated by 0.3 mM PA and treated with 40 μM tamoxifen. 30 μM CC930 or DMSO were added 2 h before tamoxifen treatment. ORO staining was performed. Scale bar: 100 μm. l Cellular TG contents were measured and quantitatively compared. m Total proteins were extracted from AML12 cells treated with DMSO/tamoxifen/CC930/tamoxifen+CC930. Protein levels of JNK and p-JNK were determined by western blotting. n The ratio of p-JNK/JNK and p-JNK/GAPDH were quantitatively compared. Bars = means ± SD; n = 3–6; ns, no significance; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001