GDF15 deficiency exacerbates chronic alcohol- and carbon tetrachloride-induced liver injury

Growth differentiation factor 15 (GDF15) has recently been shown to have an important role in the regulation of mitochondrial function and in the pathogenesis of complex human diseases. Nevertheless, the role of GDF15 in alcohol-induced or fibrotic liver diseases has yet to be determined. In this study, we demonstrate that alcohol- or carbon tetrachloride (CCl4)-mediated hepatic GDF15 production ameliorates liver inflammation and fibrosis. Alcohol directly enhanced GDF15 expression in primary hepatocytes, which led to increased oxygen consumption. Moreover, GDF15 reduced the expression of pro-inflammatory cytokines in liver-resident macrophages, leading to an improvement in inflammation and fibrosis in the liver. GDF15 knockout (KO) mice had more TNF-α-producing T cells and more activated CD4+ and CD8+ T cells in the liver than wild-type mice. Liver-infiltrating monocytes and neutrophils were also increased in the GDF15 KO mice during liver fibrogenesis. These changes in hepatic immune cells were associated with increased tissue inflammation and fibrosis. Finally, recombinant GDF15 decreased the expression of pro-inflammatory cytokines and fibrotic mediators and prevented the activation of T cells in the livers of mice with CCl4-induced liver fibrosis. These results suggest that GDF15 could be a potential therapeutic target for the treatment of alcohol-induced and fibrotic liver diseases.

the Jackson Laboratory (Bar Harbor, ME, USA). GDF15 KO mice derived from the inbred C57BL/6 strain were provided by Dr. S. Lee (Johns Hopkins University School of Medicine, Baltimore, MD, USA). All mice were maintained in a specific pathogen-free animal facility (Chungnam National University Hospital Preclinical Research Center) in a controlled environment (12 h light/12 h dark cycle; humidity, 50-60%; ambient temperature, 22 ± 2 °C). Mice were placed on a Lieber-DeCarli low-fat liquid diet (Dyets, Dyets, Inc., Bethlehem, PA, USA) containing 1 kcal/ml, of which 18% was derived from protein, 12% from fat, and either 70% from carbohydrate (control diet) or 43% from carbohydrate and 27% from ethanol (alcohol diet). Alcohol was administered gradually by escalating the content by 1% (v/v) each day until the mice were consuming a diet containing 5% (v/v) ethanol. This was continued for six more weeks. At the end of this period, mice were sacrificed, and liver tissues and whole blood were collected for further analysis. All animal experiments were approved by the Institutional Review Board on Animal Experimentation of Chungnam National University School of Medicine (CNUH-A0001) and performed in accordance with the guidelines and regulations of Chungnam National University.

Carbon tetrachloride (CCl 4 )-induced liver fibrosis in mice.
In age-matched (8-week-old) WT and GDF15 KO male mice, liver fibrosis was induced by intraperitoneal injection of CCl 4 (Sigma-Aldrich, MO; 2 mL/ kg in olive oil, 20% v/v) three times per week for 3 weeks. Mice were sacrificed 12 hours after the final CCl 4 injection.
Serum biochemical measurements. Serum was collected from whole blood by centrifugation at 10,000 rpm for 5 minutes at room temperature and was assayed for aspartate aminotransferase (AST) and alanine aminotransferase (ALT) using a Fuji Dri-Chem 4000i analyzer (Fujifilm, Tokyo, Japan). Serum or hepatic levels of tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6) were measured using an enzyme linked immunosorbent assay (ELISA) kit (R&D Systems, Minneapolis, MN, USA).

Histological analyses.
Sections of the left and medial lobes of the liver were fixed with 10% neutral buffered formalin (BBC Biochemical, Mt. Vernon, WA, USA) or frozen in Tissue-Tek OCT compound (Sakura, Tokyo, Japan). For formalin-fixed samples, after deparaffinization in xylene and rehydration with ethanol, 4 μm-thick sections were stained with hematoxylin and eosin (H&E) and Sirius Red (Sigma-Aldrich, MO, USA) for detection of collagen deposition in the liver. Frozen sections (10 μm thick) were stained with Oil Red O Kit (ab150678; Abcam, Cambridge, UK) for visualization of lipid accumulation in the liver tissue.
Measurement of hepatic triglyceride levels. Hepatic lipids were extracted from 100 mg liver tissue using a mixture of chloroform and methanol (2:1 ratio), as previously described 22 . Hepatic lipid extracts were lyophilized with nitrogen gas, and the triglyceride (TG) levels in resuspended lipids were analyzed on a Fuji Dri-Chem 4000i analyzer according to the manufacturer's instructions (Fujifilm, Tokyo, Japan).
Isolation of liver mononuclear cells. Liver mononuclear cells (MNCs) were isolated as described previously 5,23 . Briefly, liver tissues were cut into small pieces and incubated with pre-warmed media containing dissociation enzymes (Miltenyi Biotec, Bergisch Gladbach, Germany) for 30 minutes at 37 °C. After enzymatic digestion, the liver cell suspensions were quickly homogenized in C-Tubes using the GentleMACS Dissociator (Miltenyi Biotec) on the m_liver_03 program. After removal of debris, the cells were resuspended in phosphate-buffered saline and centrifuged at 1,000 × g for 5 minutes for the elimination of hepatocytes. The supernatant was removed with mechanical suction and filtered through a 70 μm nylon cell strainer (BD Falcon, Millville, NJ). Hepatic MNCs were isolated by centrifugation at 1,200 × g for 10 minutes at 4 °C and resuspended in RPMI-1640 medium (Welgene, Daegu, South Korea).
SCIeNtIfIC RepoRts | 7: 17238 | DOI:10.1038/s41598-017-17574-w Isolation of hepatic stellate cells and Kupffer cells. As described previously 5,23 , hepatic stellate cells (HSCs) were isolated by in situ collagenase perfusion of the liver. After anesthesia by intraperitoneal injection with 100-200 mg/kg ketamine/xylazine cocktails, the liver was perfused in situ with collagenase type I through the portal vein. Non-parenchymal cells were separated by centrifugation at 500 rpm and 4 °C for 5 minutes. Then, the HSCs and Kupffer cells were isolated by differential centrifugation on a density gradient of 11.5% and 20% OptiPrep (Sigma-Aldrich, MO, USA) at 3,000 rpm and 4 °C for 17 minutes. HSCs and Kupffer cells were located in the upper and lower layers, respectively. Kupffer cells were positively purified by labeling with anti-mouse F4/80 microbeads (Miltenyi Biotec) and subjected to real-time PCR analysis. HSCs and Kupffer cells were treated with alcohol (50-200 mM), lipopolysaccharide (10 ng/mL; Sigma-Aldrich, MO, USA), or recombinant GDF15 (100 ng/mL; R&D Systems, Minneapolis, MN, USA).
Co-culture hepatocyte with Kupffer cells or HSCs. Mouse Kupffer cells were cultured in the lower chambers of 12-well plates with lipopolysaccharide (10 ng/mL) overnight. Isolated HSCs were cultured in the lower chambers of 12-well plates with lipopolysaccharide (10 ng/mL) for 4 days. The cells were then rinsed two times with PBS, and WT or GDF15 KO hepatocytes were seeded into the Transwell inserts (Corning Inc., Corning, NY). The two cell lines were co-cultured for 3 hours and the co-cultured Kupffer cells and HSCs were subjected to real-time PCR analysis.

Real-time polymerase chain reaction.
Total RNA was extracted from cells or tissues using TRIzol reagent (ThermoFisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. Complementary DNA (cDNA) was prepared from the same quantity of total RNA using M-MLV reverse transcriptase and oligo-dT primers (Invitrogen/ThermoFisher Scientific, Waltham, MA, USA). The resulting cDNA was amplified on a 7500 Fast Real-Time PCR System (Applied Biosystems, Carlsbad, CA) using 2X SYBR Green mix (Applied Biosystems). All reactions were performed in triplicate with the primers listed in Supplementary  Table 1. Relative gene expression was calculated using the ΔΔCT method, with normalization to 18 s expression. Values are expressed as the fold change versus the control group.
Measurement of the oxygen consumption rate. The mitochondrial oxygen consumption rate (OCR) was measured using a Seahorse XF-24 analyzer (Seahorse Bioscience Inc., North Billerica, MA, USA) in 24-well plates. Hepatocytes were seeded at 2 × 10 4 cells per well 24 h before the analysis. On the day before the OCR measurement, the sensor cartridge was placed into calibration buffer (Seahorse Bioscience) and incubated in a non-CO 2 incubator at 37 °C. Hepatocytes were washed and incubated in DMEM without sodium bicarbonate. The medium and mitochondrial OXPHOS inhibitors were adjusted to pH 7.4 on the day of the OCR assay. The basal OCR was measured three times, and three readings were taken after the addition of each mitochondrial OXPHOS inhibitor [oligomycin (2 µg/mL) and rotenone (1 µM)]. The basal and post-oligomycin OCRs were calculated by averaging the last three measurements after maintaining a steady state. Coupled respiration was expressed as the percent decrease from basal respiration. Additionally, carbonyl cyanide m-chlorophenyl hydrazone (CCCP; 5 µM) was used to measure maximal mitochondrial respiration of the cells. OCR was automatically calculated and recorded by the sensor cartridge and the Seahorse XF-24 software. Statistical analysis. Statistical analyses were performed using Stat Graph Prism 6 (GraphPad, San Diego, CA, USA). Data are reported as the mean ± SEM. All data from animal studies were analyzed by two-way repeated-measures ANOVA followed by Bonferroni's multiple comparisons, one-way ANOVA followed by Tukey's post-hoc test, or a two-tailed Student's t-test. P-values < 0.05 were considered statistically significant.

Alcohol and mitochondrial inhibitors increase hepatic expression of GDF15 in vitro and in vivo.
Alcohol decreases the activity of the oxidative phosphorylation (OXPHOS) complexes 26 and increases ROS production 27 , leading to mitochondrial stress or dysfunction. Recently, mitochondrial stress has been shown to promote the induction of GDF15, a mitokine affecting systemic energy metabolism 16 . Based on previous observations that GDF15 expression is associated with the mitochondrial stress response, we examined the effect of alcohol on GDF15 expression in hepatocytes and non-parenchymal cells. As shown in Fig. 1a, Gdf15 was exclusively expressed in hepatocytes and not HSCs or liver-resident macrophages. Gdf15 expression was increased in hepatocytes treated with alcohol in a concentration-dependent manner (Fig. 1b). Treatment with alcohol also increased the levels of secreted GDF15 in supernatants of cultured hepatocytes (Fig. 1c). Additionally, we found HSC activation was not related to Gdf15 expression (Fig. 1d), but that treatment with alcohol increased Gdf15 expression in D1 HSCs (Fig. 1e). We next investigated the effect of mitochondrial dysfunction on GDF15 expression in hepatocytes. Inhibition of the mitochondrial OXPHOS complex using oligomycin or rotenone significantly increased Gdf15 expression and GDF15 secretion by cultured hepatocytes (Fig. 1f and g). To confirm this observation in vivo, WT mice were pair-fed an alcohol or control liquid diet for 6 weeks to investigate the effects of alcohol on GDF15 expression. No significant difference was detected in body weight gain or food intake between the two groups ( Fig. 1h and i). Hepatic Gdf15 expression and serum levels of GDF15 were remarkably increased in alcohol-fed mice ( Fig. 1j-l). Taken together, these data demonstrate that treatment with alcohol induces hepatic GDF15 expression in vitro and in vivo.

Recombinant GDF15 enhances oxygen consumption by hepatocytes and reduces inflammatory cytokine production by liver-resident macrophages.
Although GDF15 is mainly expressed and secreted by hepatocytes, several types of liver-resident cells, including HSCs, macrophages, and liver sinusoidal endothelial cells as well as hepatocytes can be affected by GDF15. Therefore, hepatocytes were treated with recombinant GDF15 (rGDF15) and the OCR was measured to determine the effect of GDF15 on hepatocyte mitochondrial function. Treatment with rGDF15 increased the basal OCR and the maximal respiration in hepatocytes ( Fig. 2a and b). Mitochondrial oxidative metabolism is linked to anti-inflammatory macrophage activation 28 . Therefore, Kupffer cells were treated with lipopolysaccharide (LPS) in the presence or absence of rGDF15 to determine the anti-inflammatory effect of GDF15 on hepatic resident macrophages. GDF15 reduced the expression of pro-inflammatory cytokines in Kupffer cells treated with LPS (Fig. 2c). The anti-inflammatory effect of rGDF15 on LPS-treated Kupffer cells was dose-dependent (Fig. 2c). To further evaluate the paracrine effect of GDF15 released from hepatocytes on Kupffer cells, WT or GDF15 KO hepatocytes were seeded into Transwell inserts, and lipopolysaccharide-treated Kupffer cells were placed in the lower chambers (Fig. 2d). The expression of pro-inflammatory cytokines by the LPS-treated Kupffer cells was significantly increased by co-culture with GDF15 KO hepatocytes as compared to co-culture with WT hepatocytes (Fig. 2e). However, co-culturing intermediate activated D4 HSCs with WT or GDF15 KO hepatocytes in the presence of LPS (10 ng/mL) did not result in any differences in the expression of fibrotic mediators in co-cultured D4 HSCs (Fig. 2f). These data suggest that GDF15 may regulate the hepatic immune microenvironment by inhibiting pro-inflammatory cytokine production from liver-resident macrophages.

GDF15 deficiency exacerbates alcoholic liver injury in mice.
To investigate the effect of GDF15 depletion on alcoholic liver injury, WT and GDF15 KO mice were fed an alcohol liquid for 6 weeks. Blood chemistry analysis revealed that serum levels of AST and ALT were increased in alcohol-fed GDF15 KO mice compared to controls (Fig. 3a). Hepatic triglyceride content was also enhanced in the alcohol-fed GDF15 KO mice (Fig. 3b). As shown in Fig. 3c, H&E staining and Oil Red O staining showed that hepatic fat accumulation and immune cell infiltration were increased in alcohol-fed GDF15 KO mice compared to controls (Fig. 3c). Serum levels of TNF-α and IL-6 ( Fig. 3d) and hepatic expression of pro-inflammatory cytokines were also increased in alcohol-fed GDF15 KO mice (Fig. 3e). Taken together, these data demonstrate that GDF15 is essential for the regulation of alcohol-mediated liver injury and steatosis.

GDF15 is required for tissue homeostasis in CCl 4 -induced liver injury and fibrosis.
Although several animal models for alcoholic liver disease have been developed, they mainly lead to steatosis and low-grade inflammation but not significant fibrosis, even after prolonged administration of alcohol [29][30][31][32] . Therefore, to determine the effect of GDF15 on advanced liver disease, we used the CCl 4 -induced model of liver injury. Interestingly, serum levels of AST and ALT were significantly increased in GDF15 KO mice compared to WT mice (Fig. 4a). In addition, serum levels of TNF-α were enhanced in GDF15 KO mice treated with CCl 4 compared to controls (Fig. 4b).
Although there was no significant difference between WT and GDF15 KO mice treated with vehicle, GDF15 KO mice showed more severe liver injury than WT mice treated with CCl 4 (Fig. 4a and c). In addition, collagen deposition was significantly increased in the livers of GDF15 KO mice treated with CCl 4 (Fig. 4c). Furthermore, Tnf-α and fibrotic mediators such as type 1 collagen 1 alpha 1 (Col1a1) and alpha smooth muscle actin (Acta2) were upregulated in the livers of GDF15 KO mice treated with CCl 4 (Fig. 4d). These data demonstrate that GDF15 deficiency induces hepatic inflammation and fibrosis in the CCl 4 -induced model of liver fibrosis.

GDF15 suppresses activation of hepatic immune cells in CCL 4 -induced liver injury.
To further evaluate the role of GDF15 on hepatic inflammation induced by CCl 4 , the phenotype and function of hepatic immune cells in WT and GDF15 KO mice treated with CCl 4 were analyzed by FACS. There were no significant differences in the numbers of liver-infiltrating CD4 + or CD8 + T cells between the two groups ( Fig. 5a and b). However, the percentages of activated (CD44 + ) and CD4 + and CD8 + T cells were significantly increased in the GDF15 KO mice compared to WT mice (Fig. 5c-e). Moreover, the neutrophil (CD11b + Ly6G + ) population was increased in GDF15 KO mice. However, the percentage of monocytes (CD11b + Ly6C high ) in CCl 4 -injured livers Relative quantification of each protein was done by densitometry. (l) Serum levels of GDF15 in WT or GKO mice fed vehicle or an alcohol liquid diet for 6 weeks. All data are representative of three independent experiments and are expressed as the mean ± SEM. *P < 0.05 and **P < 0.01, versus the corresponding controls.
was not changed by GDF15 deletion (Fig. 5f-h). Furthermore, the percentage of TNF-α + CD8 + T cells in the livers of GDF15 KO mice was higher than in WT mice, but the percentages of TNF-α + CD4 + T cells and Th17 cells were not significantly different between the two groups ( Fig. 5i-l). Collectively, these data indicate that GDF15 is required for the regulation of the hepatic inflammatory response during CCl 4 -mediated liver fibrogenesis.
Recombinant GDF15 attenuates CCl 4 -induced liver fibrosis by suppressing the hepatic inflammatory response. Based on the data described above, GDF15 deficiency exacerbates liver inflammation and fibrosis. Thus, to investigate the beneficial effect of GDF15 in liver fibrosis, mice were co-injected with CCl 4 and rGDF15 for 3 weeks. rGDF15 administration reduced collagen accumulation in GDF15 KO mice during liver fibrogenesis (Fig. 6a). As shown in Fig. 6b, serum levels of AST and ALT were lower in GDF15 KO mice after treatment with rGDF15 than GDF15 KO mice treated with CCl 4 alone. Treatment with rGDF15 also reduced the serum levels of pro-inflammatory cytokines such as TNF-α and IL-6 in the GDF15 KO mice (Fig. 6c). Consistent with these findings, hepatic expression of Tnf-α, Acta2, and Col1a1 was significantly downregulated in GDF15 KO by treatment with rGDF15 (Fig. 6d). To decipher the molecular basis by which rGDF15 reduces CCl 4 -induced liver injury and fibrosis, phosphorylation of NF-kB, JNK, and p38 was evaluated in the livers of WT and GDF15 KO mice treated with CCl 4 and rGDF15. NF-kB phosphorylation was significantly increased by injection of CCl 4 in GDF15 KO mice, while treatment with rGDF15 reduced CCl 4 -induced phosphorylation of NF-kB and JNK in the livers of GDF15 KO mice (Fig. 6e). CCl 4 -induced p38 phosphorylation was also significantly decreased in the livers of GDF15 KO mice treated with rGDF15 (Fig. 6e). Additionally, CYP2E1 expression was downregulated in the livers of GDF15 KO mice treated with rGDF15 (Fig. 6e). This pharmacologic inhibition of NF-kB, JNK and HSCs co-cultured with WT or GDF15 KO hepatocytes in medium containing lipopolysaccharide (10 ng/mL) for 3 hours. All data are representative of three independent experiments and are expressed as the mean ± SEM. *P < 0.05 and **P < 0.01, versus the corresponding controls.
SCIeNtIfIC RepoRts | 7: 17238 | DOI:10.1038/s41598-017-17574-w CYP2E1 expression and p38 phosphorylation may contribute to the suppression of α-SMA expression in the liver (Fig. 6e), leading to attenuation of CCl 4 -induced liver fibrosis. To analyze the effect of rGDF15 on hepatic resident immune cells, immune infiltrates in the livers of mice with CCl 4 -induced liver injury treated with rGDF15 were analyzed by FACS. Treatment with rGDF15 reduced the percentage of CD4 + and CD8 + CD44 + T cells and increased the percentage of CD4 + and CD8 + CD62L + T cells in the liver (Fig. 6f). The numbers of monocytes (CD11b + Ly6C high ) and neutrophils in the liver were also significantly decreased in the GDF15 KO mice treated with rGDF15 compared to GDF15 KO mice treated with vehicle (Fig. 6g). Moreover, treatment with rGDF15 reduced the numbers of CD8 + TNF-α + T cells in the livers of GDF15 KO mice (Fig. 6h). These data demonstrate that rGDF15 regulates hepatic immune homeostasis, thereby ameliorating CCl 4 -induced liver injury and fibrosis.

Discussion
The link between serum levels of GDF15 and liver disease is well described in mice and in humans, but it remains to be determined whether GDF15 has direct effects on hepatic inflammation and fibrosis. In the present study, we clearly show that alcohol and CCl 4 promote hepatic GDF15 induction, which is required for protection from alcohol-and CCl 4 -induced liver injury in mice. The findings of this study are summarized in Fig. 7.
Although the GDF15 receptor and the signaling pathways that drive its biological action have not been unequivocally elucidated, GDF15 is currently considered a reliable biomarker for liver disease, cardiovascular disease, and cancer [12][13][14]33 . Moreover, GDF15 is induced by the cellular adaptive response to tissue injury or exposure to carcinogens in a number of solid organs 34,35 . However, most data on the relationship between GDF15 and disease come from observational studies, and there is no conclusive evidence on causality. In the current study, we show that alcohol and CCl 4 -induced liver inflammation and fibrosis are more severe in GDF15 KO mice, which suggests a causal relationship between GDF15 and liver disease. These data demonstrate that GDF15 is a negative regulator of hepatic inflammation and fibrosis.
Recent emerging evidence has suggested that cellular GDF15 is involved in mitochondrial dysfunction in humans and rodents. Specifically, serum levels of GDF15 are increased in patients with mitochondrial disease 36 . Moreover, mitochondrial OXPHOS dysfunction in skeletal muscle promotes GDF15 expression in mice 16 . Meanwhile, mitochondria are abundant in liver cells and are essential in hepatic metabolism, cell survival, and ROS homeostasis 37 . Specifically, mitochondria are more involved in glucose, lipid, and protein metabolism in the liver than in other organs 37 . Therefore, hepatic mitochondrial dysfunction is implicated in the pathogenesis of liver diseases such as alcoholic and non-alcoholic steatohepatitis, insulin resistance, and fibrosis. Chronic alcohol feeding alters the expression of respiratory chain complexes in the liver mitochondria 38 , and patients with non-alcoholic steatohepatitis show morphological changes in the liver mitochondria 39 . Heterozygosity for the mitochondrial trifunctional protein promotes hepatic fat accumulation and insulin resistance in mice 40 . Moreover, mitochondrial dysfunction in hepatocytes has a critical role in the progression of CCl 4 -induced liver fibrosis in mice 41 . However, recently, mild mitochondrial perturbation in the muscle increased the lifespan and delayed age-dependent locomotory impairment in Drosophila 42 . The protective effects of muscle mitohormesis may be associated with the mitochondrial unfolded protein response, which is caused by dysfunction of the mitochondrial respiratory chain complex I 42 . Additionally, reduced mitochondrial OXPHOS function in muscle improved systemic energy homeostasis and glucose intolerance via GDF15 secretion from muscle in mice 16 . In the present study, we also showed that treatment with mitochondrial inhibitors such as oligomycin and rotenone induced GDF15 production in primary murine hepatocytes (Fig. 1), which may drive the protective effect against alcohol-or CCl 4 -induced liver injury. This evidence suggests that mitochondrial dysfunction may be a double-edged sword, as it may affect disease progression as well as protection.
The liver can be regarded as an immunologic organ, with diverse resident and infiltrating immune cells 2 . These immune cells contribute to the maintenance of hepatic immune homeostasis, but dysregulation of immune cells is involved in the development of inflammation and fibrosis in the liver. We have investigated the effect of GDF15 on the number and function of hepatic immune cells in WT and GDF15 KO mice with CCl 4 -induced liver injury. As shown in Fig. 5, the numbers of CD4 + CD44 + T cells, CD8 + CD44 + T cells, and TNF-α-producing CD8 + T cells were increased in GDF15 KO mice compared to WT mice during CCl 4 -induced liver fibrogenesis. Additionally, treatment with rGDF15 reduced the numbers of CD8 + TNF-α + T cells and the numbers of activated CD4 + and CD8 + T cells, but increased the numbers of naïve CD4 + and CD8 + T cells in the livers of GDF15 KO mice (Fig. 6). These findings suggest that GDF15 has an anti-inflammatory role in CCl 4 -induced liver injury, resulting in reduced collagen deposition in the liver. In agreement with our results, a recent study also showed that GDF15 inhibited the inflammatory response and septic heart and renal injury induced by lipopolysaccharide 43 . Furthermore, NF-kB, JNK, and p38 signaling pathways, which are closely associated with hepatic inflammation and fibrosis 44 , were inhibited by treatment with rGDF15 in a mouse model of liver fibrosis (Fig. 6). This result is also consistent with the anti-inflammatory effect of GDF15 in hepatic immune cells. However, further evaluation (c-e) Percentages of hepatic CD4 + CD44 + and CD8 + CD44 + T cells in WT and GDF15 KO mice with liver fibrosis. (f-h) Percentages of hepatic monocytes (CD11b + Ly6C high ) and neutrophils (CD11b + Ly6G + ) in WT and GDF15 KO mice with liver fibrosis. (i-l) Intracellular staining for TNF-α and IL-17A in hepatic CD4 + and CD8 + T cells. All data are expressed as the mean ± SEM. *P < 0.05 and **P < 0.01, versus the corresponding controls.
SCIeNtIfIC RepoRts | 7: 17238 | DOI:10.1038/s41598-017-17574-w is required to clarify whether GDF15 directly reduces the numbers and functional activity of immune cells in the liver, and to identify the molecular mechanism responsible for the GDF15 effect on signaling pathways.
In a previous investigation, GDF15 did not affect the regeneration capacity of the liver after acute liver injury with a single injection of CCl 4 45 . Liver regeneration, as assessed by BrdU labeling, was also not different between WT and GDF15 KO mice after partial hepatectomy and a single CCl 4 injection 45 . By contrast, our study showed that GDF15 has anti-inflammatory and anti-fibrotic roles in CCl 4 -induced chronic (3 week) liver fibrosis, which is more generally used to induce liver fibrosis than a single CCl 4 injection. Additionally, we suggested that liver fibrosis can be attenuated by the anti-inflammatory action of GDF15, based on the observation that the numbers of neutrophils and activated T cells were lower in the inflamed liver of mice treated with rGDF15. However, the strain (C57BL/6) of GDF15 KO mice used in our study was different from the mouse strain (129/SvJ) used in their study, indicating that the differences between these two studies could be due to differences in experimental methodology and the genetic background of the mice.
Many previous investigations have revealed a remarkable induction of GDF15 in a number of diseases, such as inflammation, cancer, and various metabolic diseases, suggesting that it is induced as a compensatory measure. Alcohol-and CCl4-induced liver injury also induces a remarkable increase in GDF15 levels in WT mice. Because the existing high levels of GDF15 in various injury models make it difficult to pinpoint the role of rGDF15 in WT mice, we administered rGDF15 to GDF15 KO mice to obtain a clearer picture of the role played by GDF15 after hepatic cell injury. However, to determine the therapeutic role of GDF15, rGDF15 should be administered to WT mice rather than to GDF15 KO mice. Therefore, further study is required to establish a therapeutic role for alcoholic steatosis and fibrotic liver diseases. (f) Percentages of hepatic CD4 + CD44 + and CD8 + CD44 + T cells in GDF15 KO mice treated with vehicle or rGDF15. (g) Percentages of hepatic monocytes (CD11b + Ly6C high ) and neutrophils (CD11b + Ly6G + ) in GDF15 KO mice treated with vehicle or rGDF15. (h) Intracellular staining for TNF-α in CD8 + T cells in the livers of GDF15 KO mice treated with vehicle or rGDF15. All data are expressed as the mean ± SEM. **P < 0.01, versus the corresponding controls.
In summary, we demonstrate that GDF15 has an anti-inflammatory role in alcohol-and CCl 4 -induced liver injury in mice. Mitochondrial dysfunction caused by alcohol and CCl 4 promotes GDF15 production in hepatocytes, which inhibits alcohol-mediated fat accumulation and CCl 4 -induced fibrosis in the liver. Although future studies should be performed to elucidate the mechanism of cellular GDF15 induction and secretion in damaged hepatocytes, GDF15 could be a novel therapeutic target for the treatment of liver diseases. Figure 7. Graphical summary. Alcohol and CCl 4 increase GDF15 production in hepatocytes. During alcoholinduced and fibrotic liver injury, GDF15 decreases hepatic inflammation and fibrosis via regulation of liver immune cells.