HtrA2 Independently Predicts Poor Prognosis and Correlates with Immune Cell Infiltration in Hepatocellular Carcinoma

High-temperature requirement protein A2 (HtrA2), a mitochondrial protein, is related to apoptosis regulation. However, the role of HtrA2 in hepatocellular carcinoma (HCC) remains unclear. In the present study, we explored the prognostic value and expression pattern of HtrA2 in HCC and confirmed its independent value for predicting outcomes via Cox analyses. LinkedOmics and GEPIA2 were used to construct the coexpression and functional networks of HtrA2. Additionally, the data obtained from TCGA was analyzed to investigate the relationship between the infiltration of immune cells and HtrA2 mRNA expression. Finally, the expression pattern of HtrA2 in HCC was confirmed by wet-lab experiments. The results showed high HtrA2 expression (P < 0.001) presented in tumor tissues in TCGA-HCC. Moreover, high HtrA2 expression was confirmed to be associated with poor HCC patient survival (P < 0.05). HtrA2 has also been recognized as an essential risk factor for overall survival (P=0.01, HR = 1.654, 95% CI 1.128–2.425), disease-specific survival (P=0.004, HR = 2.204, 95% CI 1.294–3.753), and progression-free interval (P=0.007, HR = 1.637, 95% CI 1.145–2.341) of HCC. HCC patients with low HtrA2 methylation had worse overall survival than patients with high methylation (P=0.0019). Functional network analysis suggests that HtrA2 regulates mitochondrial homeostasis through pathways involving multiple microRNAs and transcription factors in HCC. In addition, HtrA2 expression correlated with infiltrating levels of multiple immune cell populations. At last, increased expression of HtrA2 in HCC was confirmed using wet-lab experiments. Our study provides evidence that the upregulation of HtrA2 in HCC is an independent predictor of prognosis. Our results provide the foundation for further study on the roles of HtrA2 in HCC tumorigenesis.


Introduction
Liver cancer represents the second leading cause of cancerrelated deaths, with one out of every ten cancer-related deaths globally resulting from liver cancer [1]. Hepatocellular carcinoma (HCC) is the most frequent histological type of liver cancer (90% of all liver cancers), and its prognosis is poor [2,3]. Te survival rate of HCC is primarily determined by the stage of diagnosis [4]. Usually, the reported 5-year overall survival rate is only about 18% [5]. When HCC is diagnosed in an early phase, complete tumor resection and liver transplantation have led to improved 5-year survival rates of 47.9% and 59.3%, respectively [6]. Sadly, only a minority of patients with HCC received an early diagnosis. At the same time, a large proportion was diagnosed at an advanced stage, presenting symptoms of local tumor progression and distant metastasis [7][8][9].
Despite being the primary modality for treating tumor progression and distant metastasis, chemotherapy has shown limited antitumor efcacy in treating HCC due to poor targeting and low drug concentrations in tumor tissue [10][11][12][13]. Whereas, due to severe side efects and drug resistance, existing targeted drugs cannot provide satisfactory efcacy in treating HCC [14]. Targeting anti-HCC drugs with precise pharmacokinetics and reliable efcacy is an urgent need. In HCC, relatively little is known regarding the impact of host and genomic factors on the progression of preinvasive lesions [15]. Investigating these efects can improve our understanding of the molecular biology of HCC and favor the development of better screening strategies, improving patient prognosis. However, early diagnosis and targeted therapies for HCC are complex and challenging due to the lack of specifc markers associated with the stage or type.
High-temperature requirement protein A2 (HtrA2), a member of the high-temperature requirement family, is a serine protease localized to mitochondria [16]. HtrA2 was originally considered a heat shock-induced serine protease in Escherichia coli but is now identifed as a mitochondrial proapoptotic protein and participates in the maintenance of mitochondrial homeostasis [17][18][19]. Multiple reports show that HtrA2 is involved in the development of various cancers, including colorectal, breast, ovarian, and prostate cancer [20][21][22][23]. Recent studies have shown that plasma HtrA2 can be used as a clinical diagnostic marker for gastric carcinoma [24]. In addition, the occurrence of hepatocellular carcinoma cell apoptosis is highly dependent on HtrA2 expression [25]. Besides, HtrA2 has also been reported to participate in the suppression of hepatocellular carcinoma cell growth by staurosporine [26]. Previous studies on the relationship between HtrA2 and HCC were too specifc and needed a comprehensive view. Moreover, prior studies cannot provide a clear answer regarding whether HtrA2 serves as a predictive biomarker for HCC.
In the present study, we explored the expression, methylation level, and prognostic signifcance of HtrA2 in HCC patients in Te Cancer Genome Atlas (TCGA) and verifed our conclusions in multiple independent clinical cohorts. Furthermore, we used a multidimensional analysis to identify the functional enrichment of HtrA2-related genes in HCC and studied the relationship between HtrA2 and tumor immunity. Te pertinent fndings of this study may serve as the basis for developing new diagnostic and therapeutic strategies to treat HCC.

Materials and Methods
2.1. Data Acquisition. HCC patients' gene expression data, along with their clinical profles such as gender, age, TNM classifcation, tumor stage, and survival status, were downloaded from the TCGA database (https://portal.gdc. cancer.gov/). HCC and normal tissue gene expression profles of GSE121248, GSE101685, GSE76427, and GSE76297 were obtained from the Gene Expression Omnibus (GEO) database in the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih. gov/geo/).

HtrA2 Diferential Expression.
In the TCGA-HCC cohort, the receiver operating characteristic (ROC) curve was generated to evaluate the diagnostic accuracy of HtrA2 for HCC, and diferential expression analysis of mRNA levels of HtrA2 was performed by the Wilcoxon test, including unpaired and paired tests. Te study used R to analyze the diferential expression of HtrA2 in the GEO datasets (GSE121248, GSE101685, GSE76427, and GSE76297). We input HtrA2 in the "Gene_DE" module of TIMER2 (Tumor Immune Estimation Resource, Version 2 Web (https:// timer.comp-genomics.org/) and evaluated the expression diference of HtrA2 between tumor and normal tissues for the diferent tumors or specifc tumor subtypes of the TCGA project. Te associations of HtrA2 expression levels with clinical characteristics were examined using the nonparametric Wilcoxon rank sum test or Kruskal-Wallis test. R software was used for data visualization.

Survival Analyses of HtrA2 and Filtered Immune Cells.
Survival analysis in the TCGA-HCC cohorts was performed between high and low HtrA2 expression groups using the Survival package for R, and Kaplan-Meier survival curves were plotted using the survminer package. Kaplan-Meier Plotter (https://kmplot.com/) is a meta-analysis-based tumor biomarker assessment website. Te relation between HtrA2 expression and survival in HCC was also analyzed in Kaplan-Meier Plotter. Survival analyses of the fltered immune cells in the TCGA-HCC cohorts were performed by the Kaplan-Meier survival analysis, with a cut-of level set at the median value.

Te Independent Predictive
Value of HtrA2. Te Cox regression analyses were performed to confrm the independent predictive value of HtrA2 in HCC and explore the relationships between important clinical features and prognosis. First, 13 diferent variables were included in the univariate analysis, and their correlations with prognosis were checked. Ten, factors with P < 0.1 in univariate analyses were selected for multivariate analyses to evaluate whether each variable has an independent predictive value in HCC.

HtrA2 Methylation Level and Its Prognosis Analysis.
Te methylation level and copy number variation (CNV) data of HtrA2 were acquired through the cBioPortal web platform (https://www.cbioportal.org/). Furthermore, the correlation between HtrA2 methylation level and HtrA2 gene expression was analyzed using Pearson's correlation analysis. Te HtrA2 expression in diferent copy number variation groups was compared using the Kruskal-Wallis test. Te changes in the promoter methylation level of HtrA2in HCC and normal tissues were analyzed using the UALCAN online tool (https://ualcan.path.uab.edu/). Te predictive value of the HtrA2 methylation level in HCC was investigated using the MethSurv online tool (https://biit.cs. ut.ee/methsurv/).

Analyses of LinkedOmics and GEPIA2 Databases.
LinkedOmics (https://www.linkedomics.org) is a freely accessible website with multiomics data derived from 32 TCGA cancer types. HtrA2 coexpression analysis was performed using the Pearson test in the "LinkFinder" module of LinkedOmics, and its results were plotted as volcano plots and heatmaps. Analyses of Gene Ontology, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, miRNA-target enrichment, and transcription factor-target enrichment were performed using Gene Set Enrichment Analysis (GSEA) in "LinkInterpreter" module of LinkedOmics. Te rank criterion was false discovery rate (FDR) < 0.05, and simulations were 1000. Survival heatmaps of top coexpression genes were plotted using the GEPIA2 database (a web tool for analyzing the RNA sequencing expression data, https://gepia2.cancer-pku.cn/#correlation).

Analysis of the Relationship between Immune Infltration and HtrA2
Expression. Te ssGSEA (single-sample Gene Set Enrichment Analysis) was applied to quantify the relative tumor infltration levels of immune cell types in HCC using the R package "GSVA." Te relationship between the infltration of immune cells and the HtrA2 mRNA expression was evaluated by using the Spearman correlation.

Hepatocellular Carcinoma Tissue Samples and Cell
Culture. Tissue samples were obtained from 12 HCC patients without previous HCC treatment who underwent surgery. Tese patients were recruited from the Suining Central Hospital between 2014 and 2018. Patient consent and institutional ethics approval were obtained. Te normal human liver cell line HL7702 and human HCC cell lines HepG2, SMMC7721, and Huh7 were purchased from the ATCC (Manassas, VA, USA). All cells were cultured at 37°C in a 5% CO 2 atmosphere with 95% air in Dulbecco's Modifed Eagle Medium (DMEM, Lonza, NJ, USA) supplemented with 50 U/ml penicillin/streptomycin, 1% nonessential amino acids, and 10% fetal bovine serum (FBS, Lonza, NJ, USA).

Western Blot Analysis.
Total protein from cell lines was extracted using a whole-cell extraction bufer. Protein samples were separated by SDS-PAGE and transferred to polyvinylidene fuoride membranes (Millipore Corporation, Bedford, MA, USA). After blocking, the membranes were incubated overnight at 4°C with the following primary antibodies: HtrA2 (1 : 1000) and GAPDH (1 : 2500). Primary antibody against HtrA2 was purchased from ZEN BIO, China. Te primary antibody against GAPDH was purchased from Abcam, Cambridge, UK. Subsequently, the membranes were developed with secondary peroxidasecoupled anti-rabbit antibodies, followed by chemiluminescence imaging. Band intensity was quantifed using ImageJ software.

Statistical Analysis.
All experiments were repeated at least three times. Rank-based nonparametric Kruskal-Wallis test was used for non-normally distributed data. Statistical analysis of experimental data was performed using SPSS statistical software version 26.0. Te statistical signifcance was considered for P < 0.05.

High HtrA2 Expression in HCC.
Using the TCGA sequencing data, an analysis of HtrA2 mRNA expression in multiple human tumor tissues was performed. Te results showed that HtrA2 mRNA expression was signifcantly increased in various human tumor tissue in the TCGA database compared with that in the normal tissue ( Figure 1 ROC curve analysis of HtrA2 mRNA expression for the prediction of HCC revealed an area under the ROC curve (AUC) of 0.917 (95% CI: 0.885-0.949) (Figure 1(b)). In the TCGA-HCC cohort, the mRNA expression of HtrA2 in tumor tissue was compared with that in normal or paired adjacent tissues. Te paired and unpaired test results indicated that HtrA2 mRNA expression levels in HCC were elevated (all P < 0.001) (Figures 1(c) and 1(d)). In HCC, HtrA2 expression signifcantly correlated with the pathological stage (according to Edmondson-Steiner grade, Figure 1(f )) and histological grade (Figure 1(g)) but not with the T stage ( Figure 1(e)). HCC patients with an AFP level higher than 400 ng/mL had signifcantly higher HtrA2 expression levels than HCC patients with an AFP level less than or equal to 400 ng/mL (Figure 1(h)). Te high

High HtrA2 Expression Is Considered an Independent Risk
Factor for HCC. Te prognostic value of HtrA2 expression in HCC patients was further confrmed using the Cox proportional hazards model. Potential prognostic factors in the univariable analysis (Figures 3(a)-3(c)) with P < 0.1 were included in the multivariable analysis. Te multivariate Cox analyses ( Figure 3(d)) showed that only high HtrA2 expression is an important risk factor for OS and DSS. Consistently, in the PFI Cox analysis, HtrA2 and race were acting as potential risk roles in HCC patients.

Te Methylation of the HtrA2 Gene Is Associated with HtrA2 mRNA Expression and Predicts Patient Prognosis in HCC.
Te data sets on the HtrA2 gene, including mRNA expression data, copy number variation (CNV) data, and methylation data in HCC, were obtained from the cBio-Portal for Cancer Genomics. Te copy number for each HtrA2 was categorized according to copy number level per gene as shallow deletion, diploid, and gain. Patients with the HtrA2 gene copy number gain had a higher level of HtrA2 expression in HCC, but only 40 of 360 patients (11.1%) had the HtrA2 gene copy number gain (Figure 4(a)). Te results suggested that CNV may not be the leading cause of the high expression of HtrA2 in HCC. We further explored the relationship between HtrA2 methylation and gene expression, and these results proved that HtrA2 gene methylation was negatively correlated with gene expression (r � −0.18, P � 0.001) (Figure 4(b)). Based on the UALCAN database, we found that promoter regions of HtrA2 were shown to be signifcantly less methylated in tumor tissues of HCC than in the normal tissues adjacent to cancer (P < 0.001, Figure 4(c)). In addition, methylation of HtrA2 was strongly associated with survival in HCC patients, and hypomethylation was associated with a negative prognosis (HR � 0.566, P � 0.0019) (Figure 4(d)).

HtrA2 Coexpression Networks in HCC.
To obtain a deep understanding of the biological meaning of HtrA2 in HCC, the "LinkFinder" module in LinkedOmics was used to detect the coexpression pattern of HtrA2. Te results identifed 4843 genes (red dots) positively correlated with HtrA2 and 6070 genes (green dots) negatively correlated (P value <0.05)  c)). Gene ontology enrichment was determined using the GSEA. Te results showed that HtrA2 coexpressed genes involved mainly in protein localization to endoplasmic reticulum, mitochondrial respiratory chain complex assembly, ribonucleoprotein complex biogenesis, translational elongation, tRNA metabolic process, DNA damage response, protein folding, metallo-sulfur cluster assembly, and mRNA processing. However, positive regulation of cell motility, platelet-derived growth factor receptor signaling pathway, peptidyl-serine modifcation, respiratory tube development, and small GTPase-mediated signal transduction regulation were inhibited ( Figure 5(d)). KEGG enrichment analysis showed that HtrA2 coexpressed genes were primarily enriched in the ribosome, spliceosome, proteasome, RNA transport, ribosome biogenesis in eukaryotes, and nonalcoholic fatty liver disease (NAFLD) ( Figure 5(e)). Interestingly, the top 50 genes positively coexpressed with HtrA2, except for HERC3, were all independent prognostic indicators in HCC, and each of these genes had a high hazard ratio (HR, P value <0.05) ( Figure 5(f)). In addition, the HtrA2 coexpressed genes in HCC were analyzed using the GSEA method. Results demonstrated that the top 5 microRNAs were miR-186, miR-527, miR-26A, miR-519E, and miR-30A-5P (Table 1). Transcription factor enrichment analysis showed that the HtrA2 coexpressed genes were mainly related to V$IPF1_Q4, V$HFH4_01, PAX4_02, V$E4BP4_01, and V$ETF_Q6 (Table 1).

Te Correlation between HtrA2
Expression and the Levels of Infltrating Immune Cells. Because the immune system might play a key role in developing HCC, we further analyzed the relationship between HtrA2 mRNA expression and the levels of infltrating immune cells. Te correlation between HtrA2 mRNA expression and the infltration of immune cells is shown in Figure 6(a). Te results showed that the HtrA2 mRNA expression was signifcantly and negatively correlated with the infltration of B cells (Figure 6(b)), CD8 T cells (Figure 6(c)), T17 cells ( Figure 6(d)), and dendritic cells (DC) (Figure 6(e)). In addition, HtrA2 mRNA expression also positively correlated with T helper cells (Figure 6(f )) and T2 cells (Figure 6(g)).
In addition, we evaluated the prognostic impact of each of

Validation of HtrA2 Expression in HCC.
According to the results of IHC, HtrA2 protein expression in HCC was divided into weakly (Figure 8(a)), moderately (Figure 8(b)), and strongly positive (Figure 8(c)) as shown in Figure 8. Te results also showed that HtrA2 mRNA and protein expression in HCC cell lines (HepG2, SMMC7721, and Huh7) was signifcantly higher than that of HL7702 (all P < 0.001, Figures 8(d) and 8(e)).

Discussion
In the present study, HtrA2 was found to have high expression in HCC tumor tissue, and its high expression was associated with hypomethylation. In addition, HtrA2 mRNA expression and hypomethylation of HtrA2 were both associated with a poor prognosis in HCC. Univariate and multivariate analyses confrmed that the high expression of HtrA2 might be an independent prognostic factor in patients with HCC. We next explored the potential coexpression and regulatory networks of HtrA2. Furthermore, we examined the relationships between HtrA2 and immune cell infltration, fnding that overexpression of HtrA2 in HCC was associated with the infltration of various immune cells. At last, we experimentally confrmed the expression level of HtrA2 in HCC. Tis study sheds new light on the understanding of the role of HtrA2 in HCC development and provides directions for further research on the management of HCC. According to the results of previous studies, higher expression of HtrA2 had been found in malignant thyroid tumors, gastric cancer, and prostate cancer, and it was predictive of poor patient outcomes [27][28][29][30]. However, the expression of HtrA2 was found to be very low, which predicted poor outcomes in non-small-cell lung cancer and ovarian cancer [31,32]. Tese fndings refect the diferential roles of HtrA2 in diferent tumor tissues and stages of tumor progression.  Based on the TCGA database, GEO data sets, and our experimental data, this study found that the expression of HtrA2 was signifcantly higher in HCC than in normal liver tissues. We also found that HtrA2 was signifcantly overexpressed in multiple cancers in the TCGA data, which is consistent with previous studies. Terefore, HtrA2 has the potential to be a diagnostic marker for multiple tumor types, including HCC. According to our results, the diagnostic performance of HtrA2 in HCC detection was promising, with an AUC value of 0.917. Furthermore, the study demonstrated that a high HtrA2 was correlated with a more advanced pathologic stage, higher AFP level, and histologic grade in HCC. Tese results indicated that HtrA2 might be related to the malignant degree and growth of HCC. Survival analysis results also indicated that higher expression of HtrA2 portended a poor prognosis for patients with HCC. However, the predictive value of HtrA2 seemed to depend on the patient's sex. HtrA2 had a better predictive value for OS in male patients than in female patients. In addition, Cox regression analysis further proved that HtrA2 was an independent predictor of adverse clinical outcomes in HCC patients. Te present study also explored the mechanism of HtrA2 mRNA overexpression in HCC, and the results of our study showed that DNA hypomethylation was a possible mechanism contributing to HtrA2 mRNA overexpression in HCC.   XIAP  TBC1D8B  UHMK1  KLHL20  LMBRD2  ZNF281  HERC3  KLHL8  PROX1  SEC24A  C10orf118  HIP1  MTM1  MLL3  MAN1A2  IL6ST  ARHGEF12  HIPK3  MAP3K2  FTSJD1  ANKRD36BP1  ERN1  CLOCK  XRN1  WDR44  ZNF800  ERBB2IP  PHLPP2  CCNT1  CPEB4  R3HDM2  FNDC3A  CHD9  ATP13A3  KIAA0494  SECISBP2L  ETV3  DDI2  SNX13  IDS  REST  USP12  ZNF791  HIPK1  C9orf41  RALGAPA2  HEATR5A  RBL2  EXOC8  PIK3C2A Z-Score Group

Journal of Oncology
Interestingly, HtrA2 methylation states were also related to the prognosis of HCC, as HtrA2 methylation levels were negatively correlated with overall survival times in HCC patients. Tese fndings were consistent with our previous conclusion that high HtrA2 expression was associated with a poor prognosis in HCC patients. Although many mechanisms are likely to contribute to the overexpression of HtrA2 in HCC, hypomethylation may be one of the major regulatory mechanisms.
Previous studies support the conclusion that HtrA2 not only acts as an apoptotic-inducing protein but also contributes to removing denatured proteins from the mitochondria following exposure to heat shock or other stresses [33]. Functional enrichment analysis of coexpressed genes of HtrA2 found that HtrA2 might participate in multiple biological processes (such as protein localization to the endoplasmic reticulum, DNA damage response, and protein folding) and important protein synthesis pathways (including ribosome and ribosome biogenesis). Tese results are generally consistent with previous publications. As with HtrA2, many genes positively coexpressed with HtrA2 have a strong and independent prognostic role for HCC patients. Tis is further evidence that HtrA2 plays an important role in HCC prognosis.
We further explored the regulators responsible for HtrA2 dysregulation and found that HtrA2 was related to the network of microRNAs, such as miR-186, miR-527, miR-26A, miR-519E, and miR-30A-5P. miR-186 shows low expression in liver cancer stem cells, and its overexpression inhibits liver cancer stem cell self-renewal and tumorigenesis [34]. Upregulation of miR-186 may inhibit the nuclear β-catenin accumulation and the activation of Wnt/β-catenin signaling in HCC cells, reducing tumor cell proliferation and metastasis [35]. Furthermore, the highly upregulated long noncoding RNA can block the inhibitory efects of miR-186 on high mobility group A2, a validated oncogene in HCC [36]. Nonetheless, the exact pathogenetic mechanisms by which decreased miR-186 expression promotes hepatocarcinogenesis remain unclear. Several studies have confrmed the importance of miR-527-mediated gene regulation in HCC development. For example, miR-527 inhibits HCC tumorigenesis by regulating its targets, including UBE2A, FBXW7, and glypican-3 [37]. It has been well established in biomedical research that miR-26A suppresses tumor progression in multiple cancers, such as colorectal cancer, gastric cancer, and HCC [38][39][40]. Yang et al. show that miR-26A regulates Janus kinase 1 to inhibit HCC cell proliferation, invasion, and metastasis [41]. Mahati et al. found that nanosystem-mediated miR-26a delivery can inhibit the proliferation and migration of glypican-3-positive HCC cells in vitro and show an improved therapeutic efect in liver cancer xenografted mouse models [42]. Likewise, miR-30A-5P also has an important role in the development of HCC. Zhang et al. discovered that miR-30a-    5p/CLCF1 could regulate sorafenib resistance in HCC, and treatment with cholesterol-modifed agomiR-30a-5p signifcantly reduced tumor growth in mice harboring sorafenib-resistant HCC tumors [43]. Moreover, miR-30a-5p upregulation can block the enhanced migration and invasion of HCC cells induced by lncRNA loc339803 overexpression [44]. Pan et al. found that vimentin is a target of miR-30a-5p, and the miR-30a-5p-Vimentin axis is a potential molecular biomarker and therapeutic target in HCC [45]. Previous studies showed that miR-519E regulates tumor cell apoptosis, proliferation, and migration [46,47]. Since HtrA2 can regulate tumor cell apoptosis, there is a possible connection between miR-519E and HtrA2. Te above evidence may provide clues for further exploration of the relationship between HtrA2 and the network of microRNAs in HCC. Tis study found that HtrA2 dysregulation was also associated with multiple transcription factors, including IPF1, HFH4, PAX4, E4BP4, and ETF. All the major transcription factors regulate the cell apoptosis and proliferation of diferent types of human cells [48][49][50][51][52].
Our results indicate that HtrA2 may regulate the cell apoptosis and proliferation of HCC through these factors. We found that the expression of HtrA2 was signifcantly associated with the immune infltrate, and CD8 T cell infltration might be one of the critical factors of HtrA2 with prognostic value in HCC. Te results were consistent with those from studies in other tumor types. Hu et al. reported that HtrA2 regulates CCR2-mediated breast cancer cell growth and cellular invasion in a CCL2/CCR2 contextdependent manner [53]. While the importance of CCL2/ CCR2 signaling in macrophages during cancer progression is well documented, HtrA2 may regulate macrophage recruitment. In addition, previous studies found a strong relationship between the expression of HtrA2 and the expression of immune inhibitors (CD274, IDO1, TIGIT, etc.), immunostimulators (CD80, CD86, ICOS, etc.), and chemokines (CCL2, CCL7, CXCL9, etc.). Not only so, HtrA2 tends to express mainly at CD8 T, DC, plasma, fbroblasts, mast, and malignant cell clusters, possibly indicating that HtrA2 also functions in immune cells or stromal cells other than cancer cells [24]. However, studies on the correlation between HtrA2 and tumor immune infltration are currently lacking. More progress will be made in this research feld for the foreseeable future.
Nevertheless, the current study also has some shortcomings, which are described as follows. (1) Our study is limited by its retrospective data, and the fndings need to be confrmed by prospective studies. (2) Due to incomplete data, we were unable to assess the association of HtrA2 expression with the clinical staging of liver cancer, including the Barcelona Clinic Liver Cancer (BCLC) stage and China Liver Cancer staging (CNLC). (3) Te lack of appropriate in vivo and in vitro experimental data hampered the understanding of the role of HtrA2 in the HCC. Tis is, of course, a problem faced by other similar studies [54,55]. We have cultivated cell lines ready for further wet-lab experimentation to validate our results and investigate additional critical signaling pathways associated with HtrA2 in the HCC.
In summary, HtrA2 mRNA expression was overexpressed in HCC, while methylation of HtrA2 was decreased in HCC. Poor prognosis in HCC is associated with high HtrA2 mRNA expression and low HtrA2 methylation. Our results also showed that HtrA2 might play a key role in modulating the proliferation and apoptosis of HCC cells, possibly by interacting with microRNAs and multiple transcription factors. Furthermore, the infltration of most immune cells is signifcantly associated with HtrA2 expression, suggesting a possible involvement of HtrA2 in regulating the tumor immune response. Tis study demonstrated that HtrA2 has the potential to be a prognostic and diagnostic HCC biomarker, highlighting the potential therapeutic value of HtrA2 as an anticancer drug target.

Data Availability
Raw data are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that there are no conficts of interest.