Regulation of 4-HNE via SMARCA4 Is Associated with Worse Clinical Outcomes in Hepatocellular Carcinoma

Accumulation of 4-hydroxynonenal (4-HNE), a marker of lipid peroxidation, has various favorable and unfavorable effects on cancer cells; however, the clinicopathological significance of its accumulation in hepatocellular carcinoma (HCC) and its metabolic pathway remain unknown. This study analyzed 4-HNE accumulation and its clinicopathological significance in HCC. Of the 221 cases, 160 showed relatively low accumulation of 4-HNE in HCC tissues, which was an independent prognostic predictor. No correlation was found between 4-HNE accumulation and the expression of the antioxidant enzymes glutathione peroxidase 4, ferroptosis suppressor protein 1, and guanosine triphosphate cyclohydrolase 1. Therefore, we hypothesized that 4-HNE metabolism is up-regulated in HCC. A database search was focused on the transcriptional regulation of aldo-keto reductases, alcohol dehydrogenases, and glutathione-S-transferases, which are the metabolic enzymes of 4-HNE, and seven candidate transcription factor genes were selected. Among the candidate genes, the knockdown of SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily a, member 4 (SMARCA4) increased 4-HNE accumulation. Immunohistochemical analysis revealed an inverse correlation between 4-HNE accumulation and SMARCA4 expression. These results suggest that SMARCA4 regulates 4-HNE metabolism in HCC. Therefore, targeting SMARCA4 provides a basis for a new therapeutic strategy for HCC via 4-HNE accumulation and increased cytotoxicity.


Introduction
Hepatocellular carcinoma (HCC) is a major malignancy, accounting for approximately 90% of primary liver cancer cases, and its carcinogenesis has been reported to be associated with hepatitis C virus (HCV) and hepatitis B virus (HBV) infection, alcohol, cirrhosis, and nonalcoholic steatohepatitis.It is the fourth leading cause of cancer-related death worldwide, and, therefore, new cancer therapies are critical to improve the prognosis of patients with HCC [1].
Metabolic changes in cancer cells or metabolic reprogramming occurs during carcinogenesis and cancer progression [2,3], resulting in various abnormalities in metabolites that accumulate in cancer cells [4].4-hydroxynonenal (4-HNE) is a lipid peroxide, a marker reflecting the lipid oxidation state, as well as a metabolite with various effects on cells [5].
avidin and biotinylated enzyme (ABC) method (VECTASTAIN ABC kit; Vector Laboratories, Newark, CA, USA) or the polymer method (ImmPRESS Reagent kit; Vector Laboratories), and the staining was visualized using Vector 3,3 -Diaminobenzidine Substrate (Vector Laboratories).The details of the staining conditions are presented in Table S1.
The cytoplasmic expression levels of 4-HNE, GPX4, FSP1, and GST1 were evaluated, compared with those in non-tumor tissues, and scored as follows: 0, weak; 1, comparable; and 2, strong.For SMARCA4, nuclear expression was scored as follows: 1, comparable; 2, partially strong; and 3, strongly positive in whole tumor tissue.All tumor and non-tumor boundaries were identified, and viable tumor cells and non-tumor hepatocyte components of the borderline were compared.The most dominant staining attitude of both areas was used as the result, and the necrotic area was excluded from the evaluation.
For univariate and multivariate analyses, 4-HNE accumulation and the expression of GPX4, FSP1, and GCH1 were analyzed.

Cell Line and Culture
The HCC cell line HepG2 was purchased from the Japanese Collection of Research Bioresources Cell Bank and incubated in Dulbecco's Modified Eagle's Medium (Wako Chemicals, Richmond, VA, USA) with 10% fetal bovine serum (Gibco, Waltham, MA, USA) and 1% penicillin and streptomycin (Gibco).The cells were passaged at a ratio of 1:5 every 2-3 days.

Quantitative Reverse Transcription Polymerase Chain Reaction
RNA was extracted from siRNA-transfected cells using the RNeasy Mini Kit (Qiagen, Hilden, Germany) and was reverse transcribed to complementary DNA (cDNA) using Taq-Man reverse transcription reagents (Applied Biosystems, Woburn, MA, USA).Messenger RNA (mRNA) expression was quantified on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems) using the THUNDERBIRD SYBR qPCR Mix (Toyobo, Tokyo, Japan).In this study, β-actin, FOXP1, MCRS1, PROX1, RUVBL1, SMARCA4, SP1, and ZMYM3 mRNA levels were measured, with β-actin as a normalization control.The primer sequences are shown in Table S3.

ELISA
For sample preparation, 2 × 10 5 siRNA-transfected cells were suspended in 300 µL radioimmunoprecipitation assay buffer (50 mmol/L Tris-HCl [pH 8.0], 150 mmol/L NaCl, 0.1% w/v sodium dodecyl sulfate, 1.0% w/v NP40, and 0.5% w/v deoxycholic acid) and allowed to rest on ice for 30 min, after which they were sonicated and centrifuged at 9000× g for 15 min.The supernatant was collected, and 4-HNE concentration was quantified using the Lipid Peroxidation (4-HNE) Assay Kit (Abcam, Cambridge, UK) according to the manufacturer's instructions.This experiment was repeated two times.

Western Blotting
The efficiency of SMARCA4 knockdown by siRNA in HepG2 cells was confirmed using Western blotting.Methods previously described [37] were applied to HepG2 in this study.The primary antibodies used in this study were SMARCA4/BRG1 (21634-1-AP) (Proteintech, Chicago, IL, USA) and b-Actin (13E5) (Cell Signaling Technologies, Danvers, MA, USA) at the dilution of 1:1000.Anti-Rabbit IgG, HRP-Linked Whole Ab Donkey (Cytiva, Tokyo, Japan), was used as a secondary antibody at the dilution rate 1:5000.

Statistical Analysis
Correlations between two groups were determined using Fisher's exact test.Overall survival duration was calculated from the date of diagnosis to the date of death or last follow-up.Kaplan-Meier survival curves were used to estimate overall survival rates, and log-rank tests were used to assess differences in survival between groups.Univariate and multivariate analyses were performed using the log-rank test and the Cox proportionalhazard regression model, respectively.One-way analysis of variance with Dunnett's test was used for ELISA analysis.All statistical analyses were performed using EZR software (version 4.0.3;Saitama Medical Center, Jichi Medical University, Saitama, Japan).The error bars represent the standard deviation in this study.

Survival and Clinicopathological Analyses with 4-HNE Accumulation and Antioxidant Enzymes
Patients were stratified according to 4-HNE accumulation and the expression levels of GPX4, FSP1, and GCH1, and the correlation between the accumulation or expression levels and overall survival was analyzed (Figure 2a-h).Results showed that 4-HNE accumulation and GPX4 protein expression were correlated with overall survival.In the comparison among the three groups (scores 0, 1, and 2), the prognosis tended to be worse in cases of cancer tissues with less 4-HNE accumulation than in noncancerous tissues (Figure 2a).In contrast, a worse prognosis was observed in cases with higher or less GPX4 expression in cancer tissues than in noncancerous tissues (Figure 2c).The prognosis based on 4-HNE was significantly worse in the score 0 group (p = 0.034; Figure 2b), while that based on GPX4 was worse in the score 2 group (p = 0.005, Figure 2d).There were no significant differences in the prognosis classified according to the expression of FSP1 or GCH1 (Figure 2e-h).

Survival and Clinicopathological Analyses with 4-HNE Accumulation and Antioxidant Enzymes
Patients were stratified according to 4-HNE accumulation and the expression levels of GPX4, FSP1, and GCH1, and the correlation between the accumulation or expression levels and overall survival was analyzed (Figure 2a-h).Results showed that 4-HNE accumulation and GPX4 protein expression were correlated with overall survival.In the comparison among the three groups (scores 0, 1, and 2), the prognosis tended to be worse in cases of cancer tissues with less 4-HNE accumulation than in noncancerous tissues (Figure 2a).In contrast, a worse prognosis was observed in cases with higher or less GPX4 expression in cancer tissues than in noncancerous tissues (Figure 2c).The prognosis based on 4-HNE was significantly worse in the score 0 group (p = 0.034; Figure 2b), while that based on GPX4 was worse in the score 2 group (p = 0.005, Figure 2d).There were no significant differences in the prognosis classified according to the expression of FSP1 or GCH1 (Figure 2e-h).Next, we analyzed the correlations of 4-HNE accumulation or the expression levels of GPX4, FSP1, and GCH1 with clinicopathological factors in HCC and found that 4-HNE accumulation significantly correlated with the number of lesions (p = 0.049), vascular Next, we analyzed the correlations of 4-HNE accumulation or the expression levels of GPX4, FSP1, and GCH1 with clinicopathological factors in HCC and found that 4-HNE accumulation significantly correlated with the number of lesions (p = 0.049), vascular invasion (p = 0.002), differentiation (p = 0.002), stage (p = 0.004), and Ki-67 score (p = 0.001), whereas GPX4 expression levels significantly correlated with the presence of cirrhosis (p = 0.004) and HCV infection (p = 0.016).Meanwhile, FSP1 and GCH1 expression did not show a marked correlation with clinicopathological factors (Tables S5-S9).In addition, univariate analysis of clinicopathological factors, 4-HNE accumulation, and GPX4, FSP1, and GCH1 expression, was performed, and 4-HNE accumulation (p = 0.003), GPX4 expression (p = 0.005), tumor size (p = 0.001), number of lesions (p = 0.006), and vascular invasion (p = 0.004) were significantly associated with prognosis (Table 1).Multivariate analysis of these factors revealed that tumor size (hazard ratio (HR): 2.386, 95% confidence interval (CI): 1.328-4.286,p = 0.004), 4-HNE score 0 (HR: 2.726, 95% CI: 1.127-6.596,p = 0.026), and relatively high GPX4 expression (HR: 2.418, 95% CI: 1.379-4.242,p = 0.002) were independent predictive factors for poor prognosis.Decreased 4-HNE accumulation may be advantageous for tumor cells to survive; therefore, we conducted a detailed analysis of the factors that contribute to the decrease in 4-HNE.

Correlation between 4-HNE Accumulation and Antioxidant Enzyme Expression
The estimated mechanism for the reduction in 4-HNE in HCC is that antioxidant enzymes are overexpressed and cancel lipid peroxidation, thereby avoiding the consequent generation of 4-HNE and its associated cytotoxicity.Therefore, we examined the relationship between 4-HNE accumulation and the expression of antioxidant enzymes; however, no significant correlation was found between them (Table 2).Next, the relationship between the expression of GPX4, FSP1, and GCH1 with score 2 in each case and the amount of 4-HNE accumulated was analyzed.The percentage of cases with score 2 for any one of the antioxidant enzymes was 60.0% for 4-HNE with score 0 and 60.7% for 4-HNE with score 1 and score 2, with no significant difference (Table 3).These findings suggest that 4-HNE accumulation is reduced in HCC through mechanisms other than the elimination of lipid peroxidation by GPX, FSP1, and GCH1.

In Vitro Validation of Genes Selected via Database Analysis
Seven candidate genes identified by database analysis were examined for their association with 4-HNE accumulation in vitro.Each gene was knocked down using siRNA (Figure S1), and the intracellular accumulation of 4-HNE was evaluated using ELISA.The results showed a significant increase in 4-HNE accumulation following SMARCA4 knockdown (2.12-fold, p = 0.049, Figure 3b).Western blot analysis confirmed SMARCA4 protein expression was down regulated in cells transfected with si SMARCA4 (Figure 3c).

In Vitro Validation of Genes Selected via Database Analysis
Seven candidate genes identified by database analysis were examined for their association with 4-HNE accumulation in vitro.Each gene was knocked down using siRNA (Figure S1), and the intracellular accumulation of 4-HNE was evaluated using ELISA.The results showed a significant increase in 4-HNE accumulation following SMARCA4 knockdown (2.12-fold, p = 0.049, Figure 3b).Western blot analysis confirmed SMARCA4 protein expression was down regulated in cells transfected with si SMARCA4 (Figure 3c).

IHC of SMARCA4 and the Relation between Its Expression and 4-HNE Accumulation
Considering ELISA results, SMARCA4 is a candidate transcription factor that regulates the expression of a group of metabolic enzymes; therefore, immunohistochemical analysis of the association between SMARCA4 expression and 4-HNE accumulation in HCC tissues was performed (Figure 4a-c and Table 4).The result showed a significant correlation between the increased expression of SMARCA4 and decreased accumulation of 4-HNE (p = 0.041).In addition, patients were stratified according to SMARCA4 expression levels, and the result of their association with the outcome is shown in Figure 4d.When classified into three groups for each score, a trend toward worse prognosis was observed in the group with score 3. The results were divided into two groups: (a) group with score 1 and score 2 and (b) group with score 3. The group with score 3, which showed higher expression of SMARCA4 in whole tumor tissue than in normal hepatocytes, had a significantly poorer prognosis than the group with score 1 and score 2 (p < 0.001).

IHC of SMARCA4 and the Relation between Its Expression and 4-HNE Accumulation
Considering ELISA results, SMARCA4 is a candidate transcription factor that regulates the expression of a group of metabolic enzymes; therefore, immunohistochemical analysis of the association between SMARCA4 expression and 4-HNE accumulation in HCC tissues was performed (Figure 4a-c and Table 4).The result showed a significant correlation between the increased expression of SMARCA4 and decreased accumulation of 4-HNE (p = 0.041).In addition, patients were stratified according to SMARCA4 expression levels, and the result of their association with the outcome is shown in Figure 4d.When classified into three groups for each score, a trend toward worse prognosis was observed in the group with score 3. The results were divided into two groups: (a) group with score 1 and score 2 and (b) group with score 3. The group with score 3, which showed higher expression of SMARCA4 in whole tumor tissue than in normal hepatocytes, had a significantly poorer prognosis than the group with score 1 and score 2 (p < 0.001).

Discussion
4-HNE is the final metabolite of lipid peroxidation, resulting from the peroxidation of ω6 polyunsaturated fatty acids such as linolenic and arachidonic acids.It is associated

Discussion
4-HNE is the final metabolite of lipid peroxidation, resulting from the peroxidation of ω6 polyunsaturated fatty acids such as linolenic and arachidonic acids.It is associated with carcinogenesis and cancer progression, has highly reactive C=C and C=O, and forms adducts with intracellular substances such as proteins, nucleic acids, and phospholipids, leading to impaired intracellular signal transduction, genetic mutations, and cytotoxicity [5,41].This study showed that the accumulation of 4-HNE is relatively low in cancer tissues compared with that in noncancerous tissues, which is associated with a poor prognosis.It has been reported that exposure to low concentrations of 4-HNE promotes cell proliferation, whereas exposure to high concentrations of 4-HNE inhibits proliferation.Moreover, higher concentrations of 4-HNE induce cell death, such as apoptosis and necrosis [6,42].In cases with lower 4-HNE accumulation in HCC, 4-HNE may be less cytotoxic and favor proliferation and survival.On the other hand, a high accumulation of 4-HNE also has a poor prognosis, but in these cases, it may have acquired some form of viability that exceeds the cytotoxicity of 4-HNE itself, which may include the influence of other factors not examined in this study.
An inverse correlation between GPX4 expression and accumulation of 8-hydroxy-2deoxyguanosine, the end-product of oxidation, has been observed in diffuse large B-cell lymphoma (DLBCL) [43], whereas GPX4 expression is upregulated, and oxidative stress is reduced in colon cancer [44].However, no significant inverse correlation was observed between 4-HNE accumulation and antioxidant enzyme expression in HCC, which indicates that 4-HNE accumulation in HCC may be reduced by means other than the high expression of antioxidant enzymes and elimination of lipid peroxidation in cases with relatively low 4-HNE accumulation.Among the candidate factors regulating the 4-HNE-metabolizing enzymes AKRs, ALDHs, and GSTs, SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily a, member 4 (SMARCA4) is suggested to regulate the metabolism of 4-HNE in HCC.SMARCA4 is involved in chromatin remodeling and promotes or inhibits the expression of several genes by forming the BRM-associated factor complex [45].Since SMARCA4-deficient tumors in thoracic malignancies were reported in 2015 [46], SMARCA4 expression and its genetic mutations have received attention in various cancers [47][48][49][50].SMARCA4 acts as a tumor suppressor in hypercalcemic small cell carcinoma of the ovary, and its deficiency has been reported to be involved in carcinogenesis [47].SMARCA4 not only acts as a tumor suppressor but also cooperates with p53 loss and Kras activation in mice, and SMARCA4 enhances oxidative phosphorylation [48].In contrast, upregulation of SMARCA4 is observed in HCC, in which it promotes cell proliferation and progression, and a worse clinical outcome was observed in the group with higher SMARCA4 expression [49,50].Furthermore, SMARCA4 expression, in concert with c-MET and NRAS v12 mutations, is involved in HCC carcinogenesis [49], whereas its inhibition reduces cell migration [51].A database search (http://chip-atlas.org/,accessed on 31 September 2021) showed that SMARCA4 regulates GSTA4, a 4-HNE metabolic enzyme.Of the GSTs, GSTA4 is known to have a particularly high metabolic capacity for 4-HNE [52].GSTs contribute to the metabolism of 4-HNE either by forming a complex with reduced glutathione and being excreted through efflux transporters (MRP and RLIP76) or via further metabolism of this complex by AKR and ALDH [10].
In this study, SMARCA4 negatively regulated 4-HNE accumulation in HCC, suggesting that it positively regulates the expression of 4-HNE metabolic enzymes and contributes to 4-HNE metabolism and its decrease in cancer tissues.This reduces the cytotoxicity of 4-HNE and may favor the survival of HCC cells.Furthermore, SMARCA4 upregulated the expression of HO-1, an antioxidant enzyme [53], which may reduce intracellular oxidative stress.Therefore, targeting SMARCA4 and its downstream metabolic enzymes is expected to lead to the development of novel therapeutic strategies due to 4-HNE accumulation and its associated cytotoxicity and oxidative stress accumulation.
In addition, a poor prognosis was observed in the group with high expression of the antioxidant enzyme GPX4, suggesting that lipid peroxidation was suppressed in this group.
In turn, this suggests that low 4-HNE expression may be partly due to the overexpression of GPX4, which is a key negative regulator of ferroptosis (nonapoptotic cell death related to lipid peroxidation) that has recently attracted attention [23].The overexpression of GPX4 may suppress ferroptosis, thus favoring cancer cell survival.It is advantageous to establish the molecular basis for a novel antitumor strategy by inducing ferroptosis in HCC with high GPX4 expression.
Despite these findings, this study has several limitations.The expression levels of GPX4, FSP1, and GCH1 in HCC differed from case to case; however, we were unable to investigate the molecular mechanisms that regulate this in the present study.Overexpression of GPX4 in DLBCL has been shown to upregulate SECISBP2, which regulates GPX4 expression, and its expression may be controlled by a similar mechanism in HCC [54].Moreover, we found that 4-HNE reduction serves as a poor prognostic factor in HCC tissues and is detected in SMARCA4 knockdown cells in vitro; however, the current study lacks in vivo support.Therefore, additional in vivo experiments are necessary to support our claim.In addition, the mechanism of SMARCA4 overexpression, which is a poor prognostic predictor of HCC, remains unclear.Furthermore, the metabolic enzymes related to 4-HNE that are regulated by SMARCA4 have not been fully investigated, and further studies are needed to determine whether SMARCA4 regulates GSTA4 and other metabolic enzymes.In addition to SMARCA4, the possibility that FOXP1 also regulates 4-HNE is undeniable, although ELISA experiments showed no statistical differences.The standard deviation was large due to the small number of repeated experiments in this study.Therefore, further studies are needed to obtain more reliable results.

Conclusions
Collectively, our findings elucidate the clinicopathological significance of 4-HNE in HCC and the molecular mechanisms underlying the 4-HNE metabolic pathway.Future studies on the details of SMARCA4-mediated metabolism of 4-HNE and examination of preclinical models may help establish evidence for SMARCA4 as a therapeutic target and improve the prognosis of patients with HCC.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the ethics committee of Tokyo Medical and Dental University (approval number M2015-548, (date of approval: 25 March 2016)).
Informed Consent Statement: Patient consent was waived as the used material was historical pathological material.

Figure 2 .
Figure 2. Overall survival analysis of hepatocellular carcinoma stratified according to immunohistochemical staining of 4-hydroxynonenal (4-HNE), glutathione peroxidase 4 (GPX4), ferroptosis suppressor protein 1 (FSP1), and guanosine triphosphate cyclohydrolase 1 (GCH1).(a) Classification into three groups based on 4-HNE staining intensity showed no significant association with overall survival.(b) 4-HNE score 0 group showed significantly worse prognosis than 4-HNE score 1 and score 2 groups (p = 0.003).(c) Classification into three groups based on GPX4 staining intensity showed no significant association with overall survival.(d) GPX4 score 2 group showed significantly worse prognosis than the GPX4 score 0 and score 1 group (p = 0.005).(e) Classification into three groups based on FSP1 staining intensity.(f) None of the groups (FSP1 score 0, 1, and 2) were significantly associated with overall survival.(g) Classification into three groups based on GCH1 staining intensity.(h) None of the groups (GCH1 score 0, 1, and 2) were significantly associated with overall survival.

Figure 2 .
Figure 2. Overall survival analysis of hepatocellular carcinoma stratified according to immunohistochemical staining of 4-hydroxynonenal (4-HNE), glutathione peroxidase 4 (GPX4), ferroptosis suppressor protein 1 (FSP1), and guanosine triphosphate cyclohydrolase 1 (GCH1).(a) Classification into three groups based on 4-HNE staining intensity showed no significant association with overall survival.(b) 4-HNE score 0 group showed significantly worse prognosis than 4-HNE score 1 and score 2 groups (p = 0.003).(c) Classification into three groups based on GPX4 staining intensity showed no significant association with overall survival.(d) GPX4 score 2 group showed significantly worse prognosis than the GPX4 score 0 and score 1 group (p = 0.005).(e) Classification into three groups based on FSP1 staining intensity.(f) None of the groups (FSP1 score 0, 1, and 2) were significantly associated with overall survival.(g) Classification into three groups based on GCH1 staining intensity.(h) None of the groups (GCH1 score 0, 1, and 2) were significantly associated with overall survival.

Figure 4 .
Figure 4. Representative images of immunohistochemical staining of SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily a, member 4 (SMARCA4) in tissues from patients with hepatocellular carcinoma (HCC).(a) Score 1, (b) score 2, and (c) score 3. Overall survival analysis of HCC based on SMARCA4 staining intensity.(d) Classification into three groups based on SMARCA4 staining intensity showed no significant association with overall survival.(e) SMARCA4 score 3 group showed significantly worse prognosis than SMARCA4 score 1 and score 2 groups (p < 0.001).

Figure 4 .
Figure 4. Representative images of immunohistochemical staining of SWI/SNF-related, matrixassociated, actin-dependent regulator of chromatin, subfamily a, member 4 (SMARCA4) in tissues from patients with hepatocellular carcinoma (HCC).(a) Score 1, (b) score 2, and (c) score 3. Overall survival analysis of HCC based on SMARCA4 staining intensity.(d) Classification into three groups based on SMARCA4 staining intensity showed no significant association with overall survival.(e) SMARCA4 score 3 group showed significantly worse prognosis than SMARCA4 score 1 and score 2 groups (p < 0.001).
Abbreviations: a HR: Hazard ratio, b CI: Confidence interval.

Table 2 .
Correlation between 4-hydroxynonenal accumulation and the expression of antioxidants against lipid peroxidation.

Table 3 .
Correlation between 4-hydroxynonenal accumulation and the number of the high expression of antioxidants for lipid peroxidation.