Protein Phosphatase 4 Promotes Hepatocyte Lipoapoptosis by Regulating RAC1/MLK3/JNK Pathway

Lipotoxicity-induced apoptosis, also referred to as lipoapoptosis, is one of the important initial factors promoting the progression from hepatosteatosis to nonalcoholic steatohepatitis (NASH). Saturated free fatty acids (SFAs), which are increased significantly in NASH, are directly hepatotoxic which induce hepatocyte lipoapoptosis. Previously, we reported that protein phosphatase 4 (PP4) was a novel regulator of hepatic insulin resistance and lipid metabolism, but its role in hepatic lipoapoptosis remains unexplored. In this study, we found out that PP4 was upregulated in the livers of western diet-fed-induced NASH mice and SFA-treated murine primary hepatocytes and HepG2 cells. In addition, we found for the first time that suppression of PP4 decreased SFA-induced JNK activation and expression of key modulators of hepatocyte lipoapoptosis including p53-upregulated modulator of apoptosis (PUMA) and Bcl-2-interacting mediator (Bim) and reduced hepatocyte lipoapoptosis level as well both in vitro and in vivo. Further study revealed that PP4 induced JNK activation and lipoapoptosis-related protein expression by regulating the RAC1/MLK3 pathway instead of the PERK/CHOP pathway. The effects of palmitate-treated and PP4-induced lipoapoptosis pathway activation were largely abolished by RAC1 inhibition. Moreover, we identified that PP4 interacted with RAC1 and regulated GTPase activity of RAC1. In conclusion, these results demonstrated that PP4 was a novel regulator of hepatocyte lipoapoptosis and mediated hepatocyte lipoapoptosis by regulating the RAC1/MLK3/JNK signaling pathway. Our finding provided new insights into the mechanisms of this process.


Introduction
The occurrence of large amounts of hepatocytes death is one fundamental difference between benign steatosis and progressive nonalcoholic steatohepatitis (NASH) [1]. Lipotoxicityinduced apoptosis, also referred to as lipoapoptosis, is one of the main modes of hepatocyte death that contributes to NASH progress [2,3].
Protein phosphatase 4 (PP4) is an evolutionarily conserved protein serine/threonine phosphatase that belongs to the PP2A family, and mouse PP4 protein shares 100% amino acid sequence identity with human PP4 [14]. It has been shown to participate in multiple cellular processes, including the DNA damage response [15], cell proliferation [16], and multiple signaling pathways regulation [14,17]. Recently, our group verified the role of PP4 in hepatic insulin resistance and lipid metabolism. PP4 forms a complex with IRS-1 and participates in inflammatory-related hepatic insulin resistance [18]. PP4 promotes hepatic lipogenesis by dephosphorylating ACC-1 [19]. Additionally, PP4 interacted with hnRNPU and involved in proliferation of HepG2 cells [16]. However, the contribution of PP4 to SFA-induced hepatocyte lipoapoptosis has not been explored.
To determine the role of PP4 on hepatocyte lipoapoptosis, the expression level of PP4 in human HepG2 cells, murine primary hepatocytes, and western diet-fed-induced NASH mice was decreased either by PP4 siRNA or by AD-PP4shRNA transfection/injection and increased by AD-PP4 transfection/injection. We demonstrated that PP4 promoted PUMA and Bim upregulation through a RAC1/MLK3/JNKdependent signaling pathway, which led to hepatocyte lipoapoptosis. Moreover, PP4 interacted with RAC1 and regulated the GTPase activity of RAC1. Therefore, our findings suggested that PP4 was a positive regulator of RAC1/MLK3/JNK pathway and participated in hepatocyte lipoapoptosis.

Preparation of BSA-Bound
Palmitate. Palmitate (P5585) and BSA (A2058) were obtained from Sigma-Aldrich Co. (St Louis, MO, USA). According to a previously described protocol [20], palmitate was dissolved in EtOH at a stock concentration of 250 mM. Palmitate was conjugated with BSA at a 4 : 1 molar ratio before treatment. Specifically, the palmitate stock and 10% BSA was separately incubated in 60°C water bath for 10 min. 480 μl palmitate stock was drop wise added into 20 ml prewarmed 10% BSA to make 6 mM palmitate solution. The palmitate solution should be heated at 60°C water bath for 15 min before experiment. For 600 μM palmitate treatment, DMEM medium containing 1% bovine serum albumin was used for cell culture.
2.4. Cell Culture and Treatment. Mouse primary hepatocytes were isolated as previously described [18]. HepG2 cells purchased from ATCC were cultured in high-glucose DMEM supplemented with 10% fetal bovine serum (HyClone), 100 U/ml penicillin (Gibco), and 0.1 mg/ml streptomycin (Gibco). For palmitate treatment, cells were fasted in serum-free DMEM for 12 h before being treated with 600 μM palmitate conjugated with 1% bovine serum albumin for 48 h. For AD-PP4, AD-PP4shRNA, and RAC1shRNA lentiviral treatment, cells were transfected with adenovirus at a 30 multiplicity of infection (MOI) before treatment with palmitate. Each experiment was carried out at least 3 separate times.
2.5. Animals and Treatment. C57BL/6J male mice were purchased from the Peking University Health Science Center (Beijing, China) and housed in individual cages in a temperature-controlled environment with a 12-h light/dark cycle. Six-to eight-week-old C57BL/6J mice were fed a standard chow diet or a western diet (41% kcal from fat, Research Diet mD12079B, USA) for 16 weeks to establish a dietinduced NASH model [21].
Mice were intravenously injected through the tail vein with control adenovirus (control AD), PP4 (AD-PP4), or an inhibitor (AD-PP4shRNA) at a dose of 5 × 10 8 plaqueforming units (PFU) in 100 μl PBS (0.1 ml/25 g body weight). Mice were sacrificed, and tissues were harvested on day 9 after adenovirus injection. All animal experiments were performed in accordance with recommendations of the National Research Council Guide for Care and Use of Laboratory Animals, and the protocols were approved by the Animal Ethics Committee of the Beijing Hospital (2018BJYYEC-164-01).
2.9. GTPase Activation Assay. RAC1 activation was analyzed using a GTPase activation assay kit purchased from Upstate Biotechnology according to the manufacturer's instructions. Briefly, cell lysates were incubated with 10 μg of glutathione S-transferase-(GST-) agarose beads coupled to the GST-Pak-1-binding domain (PBD). Samples were subjected to SDS-PAGE and immunoblotting. Membranes were probed with a mouse monoclonal antibody against RAC1 (Abcam) followed by western blot analysis. The amount of PBDbound RAC1 was normalized to the amount of protein in cell lysates.
2.11. Statistical Analysis. All data are presented as the mean ± SD. The statistical significance of the differences between various treatments was determined by Student's t-test or one-way ANOVA (Tukey-Kramer honestly significant difference). A p value less than 0.05 was considered significant.

PP4 Expression Level
Was Upregulated by Palmitate-Treated Hepatocytes. Hepatocyte lipoapoptosis can be induced in vitro by incubating hepatocytes with SFAs. First, we examined PP4 expression in the hepatic lipoapoptosis cell model. Mouse primary hepatocytes treated with palmitate 600 μM for 24 h. Oil red staining showed significant lipid accumulation (Figure 1(a)). Apoptosis level elevated as indicated by TUNEL staining (Figures 1(a) and 1(b)) and DNA ladder (Figure 1(c)). Along with JNK activation and the upregulated expression of lipoapoptosis-related proteins such as PUMA, Bim, and c-caspase 3, the expression of PP4 significantly increased following palmitate treatment (Figures 1(d) and 1(e)). Moreover, we did the phenotype analysis in HepG2 cells treated with palmitate, and similar results were obtained from HepG2 cells (Figures 1(f)-1(j)). These data demonstrated that palmitate could induce hepatic lipoapoptosis and upregulate the expression of PP4. Together, these data suggested that knockdown of PP4 could restore palmitate-induced hepatic lipoapoptosis, and overexpression of PP4 promoted palmitate-induced lipoapoptosis.

PP4 Regulated Lipoapoptosis in the Liver of Western Diet-Fed Mice.
To determine whether PP4 is involved in hepatic lipoapoptosis in vivo, we analyzed its expression in the mouse liver after western diet treatment (Figure 3(a)). Feeding with western diet for 16 weeks induced significant increase of body weight and liver weight (Figures 3(e) and 3(f)) in C57BL/6J mice. In addition, serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were elevated in mice fed with western diet compared with control mice (Figure 3(g)). Oil red O and TUNEL staining revealed a significant increase of lipid deposition and hepatic apoptosis rate in the livers of western diet-fed mice ( Figure 3(b)). Compared with control mice, western dietfed mice exhibited elevated JNK phosphorylation and lipoapoptosis-related protein expressions in the livers, including PUMA, Bim, and cleaved-caspase 3 (Figures 3(c) and 3(d)). In addition, western diet-fed mice showed a significant increase in hepatic PP4 expression (Figures 3(c) and 3(d)).
To determine whether PP4 is involved in hepatic lipoapoptosis, we suppressed PP4 expression in the liver of western diet-fed mice with AD-PP4shRNA by administering tail vein injections. Ten days after administration, the expression level of hepatic PP4 was reduced to~40% by AD-PP4shRNA compared with control adenovirus vectors (Control AD) (Figures 3(m) and 3(n)). Compared with Control AD, AD-PP4shRNA treatment did not alter mouse body weight but      Oxidative Medicine and Cellular Longevity decrease liver weight and serum levels of ALT and AST (Figures 3(h)-3(k)). Moreover, the 60% reduction in PP4 expression decreases hepatic lipid deposition and hepatic lipoapoptosis level (Figure 3(l)). In addition, this effect was associated with decreased expression levels of PUMA, Bim, c-caspase 3 protein, and phosphorylation level of JNK in the liver (Figures 3(m) and 3(n)). These results suggest that inhibition of PP4 expression in the liver may prevent hepatic lipoapoptosis in western diet-fed mice. Additionally, we upregulated PP4 expression in the liver of western diet-fed mice with AD-PP4. Our data showed that the upregulation of PP4 did not alter body weight but elevated liver weight and aggravated liver injury (Figures 3(h)-3(k)) by promoting western diet-induced lipid deposition, lipoapoptosis (Figure 3(l)), and lipoapoptosis-related signaling pathway activation (Figures 3(o) and 3(p)), which is consistent with the phenotype of PP4 inhibition.
Together, PP4 regulated lipoapoptosis in the liver of western diet-fed mice.  Figure 3: PP4 regulated lipoapoptosis in the liver of western diet-fed mice. Six-to eight-week-old male C57BL/6J mice fed with western diet for 16 weeks were injected with control AD, AD-PP4shRNA, or AD-PP4 (n = 5 for each group) through the tail vein and sacrificed 10 days postinjection. Body weight and liver weight in western diet-fed mice (a, e, f) and western diet-fed mice injected with control AD, AD-PP4shRNA, or AD-PP4 (h-j). Oil red O staining, TUNEL staining, and/or H&E staining of frozen liver sections in western diet-fed mice (scale bar: 50 μM) (b) and western diet-fed mice injected with control AD, AD-PP4shRNA, or AD-PP4 (l) (scale bar: 20 μM). Serum ALT and AST levels were measured in western diet-fed mice (g) and western diet-fed mice injected with control AD, AD-PP4shRNA, or AD-PP4 (k). Western blots of hepatic PP4, PUMA, Bim, c-caspase 3, and phosphorylation levels of JNK in western diet-fed mice (c), western diet-fed mice injected with AD-PP4shRNA (m), and AD-PP4 (o). Relative protein levels were quantified by densitometry and normalized by GAPDH (d, n, p). Data are expressed as means ± SD, and statistical significance for protein levels was determined by one-way ANOVA (Tukey-Kramer honestly significant difference), ( * p < 0:05, * * p < 0:01).
3.6. RAC1 Mediates PA-Induced MLK3/JNK Activation and Lipoapoptosis. To verify the role of RAC1 in hepatic lipoapoptosis, the RAC1 expression was suppressed by RAC1-shRNA lentivirus (LV) transfection in HepG2. As shown in Figures 6(a) and 6(b), in palmitate-treated HepG2 cells, RAC1 suppression led to decreased MLK3/JNK activation and PUMA, Bim, and c-caspase 3 expression levels induced by palmitate. Moreover, RAC1 inhibition did not induce significant change in PP4 expression.

PP4 Regulated Lipoapoptosis through
Regulating the Activity of RAC1. Previous studies showed that PP4 was a positive regulator of Rho GTPases, including RAC1 in HEK293 cells by regulating their GTP-binding activity [22]. So, we speculated that PP4 may induce hepatic lipoapoptosis through regulating RAC1 activation. Next, we assessed the role of RAC1 in PP4-mediated lipoapoptosis. We found that in RAC1-silenced cells, PP4-induced MLK3/JNK activation was abolished by RAC1 inhibition (Figures 6(c) and 6(d)). More importantly, the knockdown of RAC1 reduced PP4 upregulation-induced PUMA, Bim, and c-caspase 3 levels. Taken together, these data suggested that RAC1 was a key mediator in PP4 induced lipoapoptosis.
Then, we analyzed the effect of PP4 on RAC1 activity. We found that PP4 upregulation by AD-PP4 transfection led to RAC1 activation (Figures 6(e) and 6(f)), and PP4 Data are expressed as means ± SD, and statistical significance for protein levels was determined by Student's t-test ( * p < 0:05, * * p < 0:01).

12
Oxidative Medicine and Cellular Longevity suppression by siRNA transfection decreased the levels of PBD-RAC1 (Figures 6(g) and 6(h)). We addressed whether PP4 was able to regulate RAC1 directly. Coimmunoprecipitation revealed that RAC1 was immunoprecipitated with PP4 by an anti-PP4 antibody ( Figure 6(i)) and vice versa ( Figure 6(j)). Taken together, we identified RAC1 as a novel interaction protein with PP4 and PP4-regulated lipoapoptosis through regulating the activity of RAC1.

Discussion
In this study, we provided new insights into the mechanisms of hepatocyte lipoapoptosis and described a positive function of PP4 in this model. We had demonstrated for the first time that (1) PP4 was a positive regulator of hepatocyte lipoapoptosis both in vitro and in vivo, (2) PP4 promoted JNKmediated lipoapoptosis by activating the RAC1/MLK3 instead of PERK/CHOP signaling pathway, and (3) PP4 interacted with RAC1 and regulated RAC1 GTPase activity. Each of these results is discussed in greater detail below.
Lipoapoptosis induced by excess toxic lipids is one of the main modes of hepatocyte death that contributes to NASH progress [2]. Accumulating evidence have shown that the SFAs can induce hepatic lipoapoptosis both in vivo and in vitro. In hepatocytes, the activation of JNK signaling pathways is central regulator of hepatic lipoapoptosis [6,7]. Two types of BH3-only proteins, PUMA and Bim, were reported to be key sensitizers of hepatocytes to lipoapoptosis, which further triggers mitochondrial outer membrane permeabilization [23], the activation of the executioner caspases 3 and 7, and cell death [3,5]. But the molecular mechanisms involved in hepatocyte lipoapoptosis are incompletely understood.
PP4 is a novel regulator of hepatic glucose and lipid metabolism. Previously, our group showed that PP4 formed a complex with IRS-1 and participated in inflammatoryrelated insulin resistance [18]. Additionally, PP4 promoted hepatic lipogenesis by dephosphorylating ACC-1 [19]. However, the contribution of PP4 to SFA-induced hepatocyte lipoapoptosis has not been explored. In this study, we found that along with elevated levels of hepatocyte apoptosis and the activation of lipoapoptosis signaling, the expression of PP4 significantly increased in the liver of western diet-fed mice, palmitate-treated HepG2 cells, and murine primary hepatocytes, suggesting that PP4 may be involved in hepatocyte lipoapoptosis.
In order to verify the important role of PP4 on lipoapoptosis, we assessed the regulatory effect of PP4 on these regulators both in vitro and in vivo. In murine primary hepatocytes and HepG2 cells, the inhibition of hepatic PP4 expression reduced hepatic lipoapoptosis and was accompanied by decreased expression levels of lipoapoptosis-related molecules. Additionally, the PP4 upregulation increased lipoapoptosis and activated related molecules. On the other hand, to examine the role of PP4 in hepatic lipoapoptosis in vivo, we suppressed PP4 expression in the liver of western diet-fed mice with AD-PP4shRNA and upregulated PP4  Figure 6: PP4 regulated lipoapoptosis through regulating the activity of RAC1. HepG2 cells were pretreated with control LV and RAC1 shRNA LV for 24 h followed by palmitate treatment or AD-PP4 transfection. The levels of RAC1, PUMA, Bim, PP4, and c-caspase 3 and phosphorylation levels of JNK and MLK3 were measured by western blot analysis in palmitate treatment cells (a) and AD-PP4-transfected cells (c). Relative protein levels were quantified by densitometry and normalized (b, d). RAC1 activity was determined by binding to GST-PAK PBD and subsequent immunoblot analysis in AD-PP4-(e, f) and PP4 siRNA-(g, h) transfected cells. HepG2 cells were transfected with control AD or AD-PP4 for 24 h. The interaction between PP4 and RAC1 was confirmed using coimmunoprecipitation with an anti-PP4 antibody (i) and RAC1 antibody (j). Data are expressed as means ± SD, and statistical significance for protein levels was determined by Student's t-test or one-way ANOVA ( * p < 0:05, * * p < 0:01). expression with AD-PP4 by administering tail vein injections. The knockdown of PP4 reversed western dietinduced lipoapoptosis signaling. It is noteworthy that liver weight and lipoapoptosis were ameliorated in the AD-PP4 shRNA-injected mice. Also, PP4 upregulation promoted western diet-induced lipid deposition, lipoapoptosis level, and lipoapoptosis-related signaling pathway activation. These data confirmed that PP4 contributed to hepatic lipoapoptosis.
How is PP4 involved in hepatic lipoapoptosis? In the liver, both RAC1/MLK3 [5,9,10] and ER stress [11][12][13]24] signaling contribute to SFA-induced JNK activation. In accordance with previous studies, we found that the RAC1/MLK3 signaling pathway and ER stress pathway were simultaneously activated when hepatocytes underwent lipoapoptosis. While inhibition of PP4 had significant effect on RAC1/MLK3 activation, it did not affect the activation of ER stress signaling both in vitro and in vivo. Similarly, the upregulation of PP4 expression also enhanced hepatocyte lipoapoptosis by activating the MLK3/JNK signaling pathway instead of the PERK/CHOP pathway both in vivo and in vitro. These data strongly suggested PP4 likely regulated hepatocyte lipoapoptosis via the MLK3/JNK signaling pathway.
What is the mechanism underlying PP4 involvement in the palmitate-induced RAC1/MLK3 pathway activation? In palmitate-treated hepatocytes, RAC1, a key type of GTPbinding protein, has been reported to be an important component of a SFA-stimulated signaling pathway that regulates MLK3-dependent activation of JNK in hepatocytes [9,25]. RAC1 is activated by palmitate treatment and participates in MLK3/JNK activation. The combination of RAC1 and the CRIB motif of MLK3 releases the self-inhibition of MLK3, followed by the activation of the MLK3/JNK signaling pathway [9,26]. In accordance with a previous study, palmitate stimulation increased the GTPase activity of RAC1 in hepatocytes, while silencing RAC1 expression in HepG2 cells significantly inhibited the activation of the MLK3/JNK  14 Oxidative Medicine and Cellular Longevity signaling pathway induced by palmitate. Our data confirmed that palmitate-induced JNK activation and lipoapoptosis were through a RAC1-dependent pathway. Moreover, PP4 has been reported to be a positive regulator of Rho GTPases, including RAC1, in HEK293 cells by regulating their GTPbinding activity [22]. So, we speculated that PP4 may induce hepatic lipoapoptosis through regulating RAC1 activation. As we expected, PP4 also regulates the GTPase activity of RAC1 in hepatocytes. The upregulation of PP4 enhances the GTPase activity of RAC1, while the suppression of PP4 significantly reduces its GTPase activity. Moreover, RAC1 inhibition significantly inhibited PP4-induced MLK3/JNK pathway activation. In addition, we found PP4 was able to bind RAC1 directly by coimmunoprecipitation. Our data indicated that PP4 regulated lipoapoptosis through regulating the activity of RAC1. However, how PP4-regulated RAC1 activity needed further verification? For many years, RAC1 activity is believed to be regulated only by the activation family guanine nucleotide exchange factors (GEFs), inhibition family GTPase-activating proteins (GAPs), and guanine nucleotide disassociation inhibitors (GDIs) [26]. Recent studies show that reversible phosphorylation is also involved in the regulation of RAC1 activity. Ser71 (64yrplsyp73) is a conserved Akt site on RAC1 that can be phosphorylated by Akt. The phosphorylation of Ser71 by AKT inhibits the GTP-binding activity of RAC1, which further inhibits the activation of downstream signals, including the JNK signaling pathway [27]. Another study shows that thr108 (106pntp109) is an ERK phosphorylation site of RAC1. It can be phosphorylated by ERK and induces the translocation of RAC1 from the cytoplasm to the nucleus, which prevents the transmission of extracellular signals to the cell [28]. In addition, there are more than 20 candidate phosphorylation sites in RAC1 by PhosphoSite analysis, but whether these sites participate in the regulation of RAC1 activity needs further verification. For many years, due to the lack of specific phosphatase motifs, research on protein dephosphorylation has been relatively lagging. To date, the phosphatase that dephosphorylates the ser71, thr108, and other phosphorylation sites of RAC1 were still unknown, and whether these sites were regulated by PP4 needed further verification.
In conclusion, increased expression of hepatic PP4 contributed to lipoapoptosis by enhancing the GTPase activity of RAC1 and activating the MLK3/JNK pathway, which led to elevated expression of Bcl-2 proapoptotic family members and promoted hepatocyte lipoapoptosis (Figure 7). Overall, we revealed the significant role of PP4 in hepatocyte lipoapoptosis, which may provide benefit in developing novel therapeutic approaches for NASH treatment in the future.

Data Availability
All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors.

Conflicts of Interest
The authors declare that they have no conflict of interest.