Heat shock protein family A member 8 serving as a co-activator of transcriptional factor ETV4 up-regulates PHLDA2 to promote the growth of liver cancer

Abstract

Heat shock protein family A member 8 (HSPA8) participates in the folding or degradation of misfolded proteins under stress and plays critical roles in cancer. In this study, we investigated the function of HSPA8 in the development of liver cancer. By analyzing the TCGA transcriptome dataset, we found that HSPA8 was upregulated in 134 clinical liver cancer tissue samples, and positively correlated with poor prognosis. IHC staining showed the nuclear and cytoplasmic localization of HSPA8 in liver cancer cells. Knockdown of HSPA8 resulted in a decrease in the proliferation of HepG2 and Huh-7 cells. ChIP-seq and RNA-seq analysis revealed that HSPA8 bound to the promoter of pleckstrin homology-like domain family A member 2 (PHLDA2) and regulated its expression. The transcription factor ETV4 in HepG2 cells activated PHLDA2 transcription. HSPA8 and ETV4 could interact with each other in the cells and colocalize in the nucleus. From a functional perspective, we demonstrated that HSPA8 upregulated PHDLA2 through the coactivating transcription factor ETV4 to enhance the growth of liver cancer in vitro and in vivo. From a therapeutic perspective, we identified both HSPA8 and PHDLA2 as novel targets in the treatment of HCC. In conclusion, this study demonstrates that HSPA8 serves as a coactivator of ETV4 and upregulates PHLDA2, leading to the growth of HCC, and is a potential therapeutic target in HCC treatment.

Introduction

Hepatocellular carcinoma (HCC) is the second most common cause of cancer-related deaths. Although approximately 50% of HCC patients receive systemic therapies, the 2-year progression-free survival (PFS) and 5-year overall survival (OS) rates are far from satisfactory, and the detailed mechanisms are poorly understood [1]. Moreover, the underlying mechanism of hepatocarcinogenesis is poorly understood. Heat shock protein family A member 8 (HSPA8) belongs to the heat shock protein 70 (HSP70) family, and plays critical roles in cancer development and progression by participating in the folding or degradation of misfolded proteins [2]. HSP70 can maintain protein homeostasis during cellular stress conferred by the metabolic requirements of cancer cells, leading to misfolding of proteins. HSPA8 is abundant in the cytoplasm, plasma membrane, and extracellular space, and it is transiently present in the nucleus, where it shuttles between the cytoplasm and nucleus depending on the stress state [3, 4]. HSPA8 is constitutively expressed in cells and is involved in processes vital for cell fate and organism survival, including metabolism, glycolysis, and translation [5]. HSPA8 plays an oncogenic role by regulating the levels of client proteins. For example, HSPA8 prevents Rab1A degradation and apoptosis, resulting in colon cancer progression [6]. HSPA8 can be modified at the translational level by the nestin protein, and this event increases glioblastoma cell growth and invasion [7]. As a tumor-derived exosomal protein, HSPA8 might be an ideal diagnostic marker for cancers [8]. HSPA8 is essential for cancer cell survival, and drugs that inhibited tumor cell expression of HSPA8 showed a therapeutic effect in head and neck squamous cell carcinoma [9, 10]. Additionally, HSPA8 is the core component of the chaperone-mediated autophagy (CMA) machinery, mediating the delivery of cytoplasmic proteins to lysosomes, participating in cancer cell proliferation, and reducing drug sensitivity [11, 12]. HSPA8 significantly influences the development of HCC through its role in protein modification, but the effect of HSPA8 on gene transcription is unknown [13–15].

A cancer cell behaves differently from a normal cell because it expresses different housekeeper genes and regulated genes [16]. The regulated genes are controlled by many mechanisms, among which transcriptional regulation is very important [17, 18]. Transcription requires the coordination of transcription factor (TF) binding and other coactivators or repressors. TFs that bind to cis-regulatory DNA sequences or motifs, and interact with transcriptional cofactors (COFs) are responsible for either positively or negatively influencing the transcription of specific genes [19–21]. COFs are central effectors of transcriptional activation and gene expression, and have the ability to activate the core promoters of target genes [22]. Studies have shown that COFs, such as the peroxisome proliferator-activated receptor-gamma coactivator (PGC1) and transcriptional coactivator with PDZ-binding motifs (TAZ), play a critical role in HCC progression. The function of COFs is related to their subcellular localization [23, 24]. However, the influence of HSPA8 as a COF in HCC is unclear.

Pleckstrin homology like domain family A member 2 (PHLDA2) is an imprinted gene. Studies have shown that PHLDA2 plays a pleiotropic role in cancers. For example, PHLDA2 was found to be related to the pathogenesis and prognosis of endometrial cancer [25]. Overexpression of PHLDA2 was found to inhibit osteosarcoma progression via the PI3K/AKT pathway, while knockdown of PHLDA2 was found to inhibit the proliferation and invasion of tumor cells in breast cancer, colorectal cancer, and pancreatic ductal adenocarcinoma [26–29]. PHLDA2 knockdown was found to result in the downregulation of DPPA4 protein expression, thus playing an important role in tumor progression [30]. However, the significance of PHLDA2 in HCC has not been reported. ETV4 belongs to the ETS transcription factor family, and plays a critical role in cell proliferation and the cell cycle. Dysregulated expression of ETV4 and its downstream target genes is implicated in the development of liver cancer [31, 32]. However, the relationships among HSPA8, ETV4, and PHLDA2 in HCC are poorly understood.

In this study, we tried to identify the novel function of HSPA8 in HCC. Interestingly, our findings showed that HSPA8 is a novel transcriptional coactivator in cells. HSPA8 bound to the promoter of PHLDA2 and regulated PHLDA2 expression. PHLDA2 transcription was activated by the transcription factor ETV4. HSPA8 and ETV4 interacted with each other in cells, displaying colocalization in the nucleus. HSPA8 upregulated PHDLA2 by acting as a coactivator of ETV4, leading to the growth of liver cancer. Our findings provide new insight into the mechanism by which HSPA8 promotes the development of liver cancer.

Materials and methods

Public databases and bioinformatics analysis

RNA sequencing data and clinicopathological features of liver cancer were obtained from TCGA. The data were cleaned and transformed log2 using R version 4.0.5 (R Core Team, Vienna, Austria). In addition, the comparison between HCC and normal liver tissues was analyzed using GraphPad Prism7 software (La Jolla, CA, USA).

Patient specimens and clinical information

The study protocol conformed to the ethical guidelines and was approved by the Clinical Research Ethics Committee of Tianjin Medical University Cancer Hospital. HCC specimens from 55 patients were obtained from Tianjin Medical University Cancer Hospital. These specimens were used to measure the levels of mRNAs and proteins. Informed consent was obtained from each patient.

Animals

Male BALB/c nude mice (4 weeks of age) were purchased from Beijing Vital River Laboratory Animal Technology (Beijing, China). HepG2 cells were injected subcutaneously (1 × 10⁶ cells in 50 μL of DMEM). The mice were sacrificed 4 weeks after tumor cell injection. The tumor dimensions were measured with calipers, and the volume was calculated according to the following equation: tumor volume (mm³) = (length × width²) × 1/2. Then, tumors were homogenized for Western blot analysis and real-time PCR (RT-PCR). The study was authorized by the Ethics Committee of Tianjin Medical University and performed per the Declaration of Helsinki. All the animal experiments were approved by Tianjin Medical University Animal Care and Use Committee.

Cell lines and cell culture

Liver cancer cell lines (such as HepG2 and Huh-7) and 293 T cells were cultured with Dulbecco′s modified Eagle′s medium (DMEM; Gibco, Grand Island, NY, USA) which contained 10% fetal bovine serum (FBS, Gibco, NY, USA) and 1% Penicillin-Streptomycin (Biosharp, Beijing, China).

Transfection and infection

The lentiviral vectors expressing shRNAs against HSPA8, PHLDA2, and ETV4 were purchased from GENEWIZ (Suzhou, China). The PHLDA2 and ETV4 overexpression plasmids were constructed by GENEWIZ. Cells were infected with lentiviruses according to the manufacturer′s manual. The PHLDA2 and ETV4 overexpression plasmids and their control plasmids were transfected into cells using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). All shRNA sequences are listed in Supplementary Table S1.

RNA extraction and reverse-transcription quantitative real-time PCR (RT-qPCR)

Total RNA was extracted from tumor tissue of nude mice xenografts, clinical species of patients, and liver cancer cells and used for RT-qPCR using Hieff qPCR SYBR Green Master Mix from YEASEN (Shanghai, China). GAPDH was used as an internal control for normalization. All primers involved in the article are documented in Supplementary Table S2.

Tissue microarray and IHC

Liver cancer tissue microarrays (n = 79) with normal liver tissues were purchased from Shanghai Outdo Biotech (Shanghai Cohort, China). Moreover, we used the above 55 HCC specimens and matched normal liver tissues to generate the tissue microarrays. The microarray information was shown in Supplementary Table S3 and S4. All of the formalin-fixed paraffin-embedded samples were used to perform IHC staining with anti-HSPA8 antibody and anti-PHLDA2 antibody. IHC staining intensity was quantified by using the software of ImageJ (ImageJ, NIH, Bethesda, MA). The H score was defined by the HSPA8 staining intensity (0–3+) multiplied by the expression percentage (0–100) for each slide. All the antibodies were listed in Supplementary Table S5.

Western blot analysis

Liver cancer cells and tumor tissues from patients and nude mice were harvested by centrifugation and lysed in the Radio Immunoprecipitation Assay (RIPA) buffer with a proteinase inhibitor cocktail. The nuclear and cytoplasmic proteins in HepG2 cells were extracted using the Nuclear and Cytoplasmic Protein Extraction Kit according to the manufacturer’s protocol (Beyotime, China, P0028). Protein concentrations were determined by a bicinchoninic acid (BCA) assay. Proteins were subjected to electrophoresis in a 10% SDS-PAGE gel and were then transferred to a nitrocellulose membrane. The membrane was blocked with 5% nonfat milk and incubated primary antibodies as indicated in each experiment (anti-HSPA8, anti-PHLDA2, and anti-ETV4) for 1 h at room temperature. After incubation with the HRP-conjugated secondary antibody, the protein bands were visualized by ECL in a Genebox system from Gene Company Limited (Hong Kong, China). All the antibodies are listed in Supplementary Table S5.

Dual-Luciferase reporter assay (DLR)

According to GeneCopeia Company (Rockville, MD, USA), the firefly luciferase reporter backbone was inserted into the PHLDA2 promoter region 1408 bp from the transcription start site (−1408 bp ~ +1 bp). Cells were seeded in 96-well white tissue culture plates at 3000 cells per well. After 24 h, cells were transfected with 50 ng pGL3-PHLDA2 promoter or control luciferase plasmid (pRL-TK) using Lipofectamine 3000 in triplicate. Distinct lengths of the PHLDA2 promoter regions were constructed into the pGL3-basic vector to produce pGL3-1408, pGL3-908, and pGL3-404. These fractions were co-transfected with pRL-TK, shNC, or shHSPA8. After transfection for 48 h, the cells were lysed for luciferase activity measurement using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA). The sequences of specific primers are shown in Supplementary Table S2.

Confocal assays

The subcellular locations of HSPA8 and ETV4 were detected using confocal microscopy. HepG2 cells were planked and then fixed using a methanol-fix solution. The cells were then incubated with anti-HSPA8 and anti-ETV4 antibodies, FITC-conjugated mouse monoclonal secondary antibody directed to the HSPA8, and PE-conjugated rabbit polyclonal antibody directed to the ETV4, and the nucleus was stained with DAPI.

Co-immunoprecipitation (Co-IP)

Lysates harvested from hepatoma cells transfected with different constructs according to the manufacturer′s instructions were subjected to immunoprecipitation overnight at 4 °C using 2–5 g of the antibodies. Negative controls with IgG were included in each experiment. Immune complexes were incubated with protein A/G agarose beads at 4 °C. The precipitates were washed six times with ice-cold lysis buffer, resuspended in PBS, and resolved by SDS–PAGE prior to Western blot analysis.

Chromatin immunoprecipitation (ChIP)

ChIP assays were conducted with the ChIP A/G (Invitrogen, Carlsbad, CA, USA) according to the manufacturer′s protocol in HepG2 cells. Briefly, cells were harvested and washed twice in PBS, crosslinked with formaldehyde, lysed with sodium dodecyl sulfate buffer, and sonicated. Then, the fragmented chromatin was added to the ChIP dilution buffer and incubated overnight with an anti-HSPA8 antibody. A normal rabbit IgG was added as a negative control antibody. IP products were collected after incubation with Protein A/G-coated magnetic beads. The bound chromatin was eluted and digested with proteinase K, and then DNA was purified for RT-qPCR analysis. All primers are listed in Supporting Table S4.

RNA sequencing

Total RNA was extracted from HepG2 cells transfected with shHSPA8 and shCon, and subjected to RNA sequencing (RNA-seq) analysis performed by Shanghai Majorbio Bio-pharm Technology Co. Ltd. The data were analyzed on the free online platform of Majorbio Cloud Platform. The RNA sequencing results were shown in Supplementary Fig. S3a.

Cell proliferation and colony formation assays

Cell proliferation was conducted with an MTS reagent. In brief, the cells were plated in a 96-well-plate with 5000 cells (HepG2 and Huh-7) per well and cultured in DMEM with 10% FBS. MTS was added into each well and incubated for 1 h, and then absorbance was measured at 490 nm. For colony formation assays, cells were seeded in six-well plates. After two weeks, cell colonies were fixed with 4% paraformaldehyde for 10 min, stained with 0.01% crystal violet for 5 min, and washed twice with PBS.

Statistical analysis

All data analysis was performed using GraphPad Prism 7 software. All data analysis were performed using the SPSS statistical software package (version 22; Chicago, IL, USA). All data represent multiple independent experiments conducted in triplicate and were presented as mean ± SD. Student′s t-test was used for comparisons between groups. Overall survival (OS) and progression-free survival were calculated according to the Kaplan–Meier method and compared using the log-rank test.

Results

HSPA8 is overexpressed in HCC and positively correlates with poor prognosis

It has been reported that HSPA8 plays crucial roles in cancer progression by functioning as a molecular chaperone or in the CMA machinery [33, 34]. To better understand the significance of HSPA8 in HCC, we leveraged data from The Cancer Genome Atlas Liver Hepatocellular Carcinoma (TCGA-LIHC) dataset. Analysis of this TCGA transcriptome dataset revealed that HSPA8 was upregulated in HCC samples (n = 371) compared with normal tissues (n = 50) (P < 0.001, Fig. 1a) [35]. Patients with high HSPA8 expression exhibited a worse prognosis (P < 0.05) (Fig. 1b and Supplementary Fig. S1a). Moreover, the findings were supported by analysis with the public tool KMplotter (https://kmplot.com) (Supplementary Fig. S1b), collectively suggesting that high levels of HSPA8 are significantly correlated with poor progression-free survival (PFS) and overall survival (OS) in HCC. HSPA8 expression was correlated with aggressive tumor stage and poor differentiation (P < 0.001, Fig. 1c, d and Supplementary Fig. S1c, d). The expression of HSPA8 in the tissue of Stage1, Stage2, and Stage3 tumors was significantly higher than that in normal tissues. Stage 2 and Stage 3 of tumors had much higher expression levels of HSPA8 than that in Stage 1 tumors. Analysis based on pathological tumor grade also showed that the expression of HSPA8 in tumor tissues was higher than that in normal tissues. There was no statistically significant difference between normal tissue and tissues from Stage 4 tumors (n = 6). Moreover, the expression of HSPA8 was examined by IHC staining using a liver cancer tissue microarray (total, n = 134). The results showed that the percentage of HSPA8-positive cells was 27.6% (37 out of 134) in normal tissues and 78% (105 out of 134) in HCC tissues, suggesting that HSPA8 is overexpressed in liver cancer. Notably, HSPA8 was widely distributed in both the cytoplasm and nucleus of liver cancer cells (Fig. 1e and Supplementary Fig. S1e), implying that HSPA8 plays a role in the nucleus of liver cancer cells. The expression of HSPA8 was validated by RT-qPCR in 55 pairs of HCC tissues and Western blot analysis in 32 pairs of tissues (Fig. 1f, g and Supplementary Fig. S1f), with results consistent with those of IHC staining, supporting the idea that HSPA8 is highly expressed in HCC. In addition, the mRNA and protein levels of HSPA8 were significantly increased in liver cancer cell lines compared with LO-2 normal liver cells (Fig. 1h). Therefore, we conclude that HSPA8 is overexpressed in HCC and positively correlates with poor prognosis. HSPA8 potentially plays a role in the nucleus of liver cancer cells.

Fig. 1. HSPA8 is overexpressed in HCC and positively correlates with poor prognosis.

a The expression of HSPA8 was significantly higher in HCC tumor tissues than in adjacent normal tissues in the TCGA database. b Kaplan–Meier overall survival curves for all 370 patients with HCC stratified by the expression of HSPA8 (high and low). c, d The expression of HSPA8 was correlated with unfavorable clinicopathological characteristics in the TCGA HCC database. e Representative images of HSPA8 IHC staining intensity in 134 paired tissues (Tianjin cohort, n = 55; Shanghai Outdo Biotech, n = 79). Blue represents the nucleus stained with DAPI, and brown represents HSPA8 staining in the nucleus and cytoplasm of positive cells. The H score of HSPA8 staining was calculated (right). Scale bars, 50 μm. f The relative mRNA levels of HSPA8 were higher in tumor tissues than in matched adjacent normal tissues (n = 56). g Protein expression levels of HSPA8 in 8-paired HCC tissues were measured by Western blot analysis. h The mRNA and protein levels of HSPA8 were measured by RT-qPCR and Western blot analysis, respectively, in different HCC cell lines. ***P < 0.001, **P < 0.01. N normal tissues, T tumor tissues, P patient, WB Western blot.

HSPA8 is able to promote the growth of liver cancer in vitro and in vivo

To explore the role of HSPA8 in the modulation of HCC growth, we interfered with the expression of HSPA8 in HepG2 and Huh-7 cells by introducing an shRNA targeting the HSPA8 mRNA. The knockdown efficiency was verified by RT-qPCR and Western blot analysis in the system (Fig. 2a, b). The following experiments were performed with the shHSPA8-2 plasmid. Given the correlations between HSPA8 expression and clinicopathological characteristics in HCC, we assumed that HSPA8 might function to promote tumor cell proliferation. To verify the oncogenic activity of HSPA8, colony formation, and MTS assays were performed in HSPA8-knockdown HepG2 and Huh-7 cells. Our data showed that the colony formation ability and cell proliferation rate were markedly decreased after HSPA8 was knocked down compared with those in control cells (Fig. 2c–f), suggesting that HSPA8 is crucial for the proliferation of liver cancer cells in vitro. Next, we confirmed these findings in vivo in xenograft mouse models. Control and shHSPA8 HepG2 cells were subcutaneously injected into the right flanks of nude mice. The tumor volumes were measured every week, and the final tumor weights were recorded. The expression of HSPA8 in xenograft tumors was validated by Western blot analysis (Supplementary Fig. S2a). Our data showed that HSPA8-knockdown cells exhibited a significant decrease in the ability to form tumors in nude mice compared with negative control cells, based on the tumor volumes, and final tumor weights (Fig. 2g–i). The expression levels of HSPA8 and Ki-67 were significantly decreased in the HSPA8-knockdown xenograft tumor tissues relative to the control tissues (Fig. 2j and Supplementary Fig. S2b). Furthermore, the relationships among the expression levels of HSPA8, Ki-67, and PCNA in TCGA HCC specimens demonstrated that the expression level of HSPA8 was significantly correlated with those of Ki-67 and PCNA (Supplementary Fig. S2c, d). We didn′t find significant differences in the expression of HSPA8 among tumors of similar sizes, while HSPA8 expression tended to increase with tumor size (Supplementary Fig. S2e). Considering these results collectively, we concluded that HSPA8 is able to promote the growth of liver cancer in vitro and in vivo.

Fig. 2. HSPA8 is able to promote the growth of liver cancer in vitro and in vivo.

a, b The mRNA and protein levels of HSPA8 were measured by RT-qPCR and Western blot analysis, respectively, in shHSPA8-knockdown HepG2 and Huh-7 cells. c, d Knockdown of HSPA8 expression in HepG2 cells decreased the colony formation ability of HepG2 cells and Huh-7 cells. e, f Knockdown of HSPA8 expression in HepG2 cells inhibited the proliferation of HepG2 cells and Huh-7 cells, as determined by MTS assays. g–i The growth of tumors derived from shNC or shHSPA8 transduced HepG2 cells injected into nude mice was measured every 5 days. Images of tumors excised from five mice at 27 days and the average weight of the tumors are shown. j HSPA8 and Ki-67 staining showed that shHSPA8 transduced cells had a lower proliferation capacity (scale bar, 20 μm; magnification, ×200). **P < 0.01; Student′s t-test. The experiment was repeated at least three times.

HSPA8 binds to the PHLDA2 promoter region

According to our previous study, HSPA8 was confirmed to recruit HBc onto the HBV cccDNA minichromosome and confer transcription of cccDNA and replication of HBV [36]. In this study, we observed that HSPA8 could localize in the nucleus. Thus, we hypothesized that HSPA8 might bind to the promoters of some genes in mammals, serving as a transcriptional coactivator. To identify the underlying mechanism by which HSPA8 promotes the growth of liver cancer, we performed RNA-seq analysis in shHSPA8-treated HepG2 cells and shNC-control cells. The RNA-seq profiles showed that shHSPA8 introduction resulted in changes in the expression of 4079 genes at the transcriptional level, with 2092 upregulated genes and 1987 downregulated genes (log₂ fold change ≥1.5, P adjusted <0.05, Supplementary Fig. S3a). GO analysis was performed for three different categories, namely biological process (BP), cellular component (CC), and molecular function (MF). According to the enrichment analysis results, HSPA8 participates in transcriptional regulatory activity, cell proliferation, cellular process, biological regulation, and molecular function regulation (Fig. 3a), implying that HSPA8 is involved in the modulation of gene transcription. Moreover, KEGG pathway enrichment analysis also showed that HSPA8 was related to transcription (Fig. 3b). To validate the role of HSPA8 in modulating gene transcription, we conducted ChIP-Seq to map the genomic distribution of HSPA8 in HepG2 cells. The ChIP-Seq profiles revealed a total of 870 peaks relative to annotated genomic elements, and HSPA8 was found to bind to the promoters of 86 genes. The binding preferences of HSPA8 for genomic elements were calculated, and the results revealed that 34.8% of the peaks were mapped to intergenic regions; 39.7% were mapped to introns; 13% were mapped to exons, 5’UTR exons or 3’UTR exons; and 10.0% were mapped to promoter regions (Fig. 3c). To further identify the target gene regulated by HSPA8, we integrated transcriptome sequencing data with the ChIP-Seq data and found that a total of 6 genes might be directly regulated by HSPA8 (Fig. 3d). The ChIP-quantitative PCR assay showed that HSPA8 was enriched on all fragments of the PHLDA2 promoter region but on only some fragments of the RCOR2 and CAPN15 promoters. These results showed that PHLDA2 was the most affected gene (Fig. 3e and Supplementary Fig. S3b, c). According to the results of RNA-seq, the expression of PHLDA2 was significantly decreased in the shHSPA8 group (Fig. 3f). Furthermore, RT-qPCR analysis showed that HSPA8 and PHLDA2 expression had a strong correlation in 55 HCC samples (r = 0.603, P < 0.001) (Fig. 3g), suggesting that HSPA8 may transcriptionally regulate PHLDA2 in HCC. Then, we explored the expression of these three genes in the TCGA transcriptome database and found that the expression of only PHDLA2 is significantly different between the HSPA8 high expression group and the HSPA8 low expression group (Supplementary Fig. S3d–f). The expression levels of CAPN15 and RCOR2 had weaker correlations with that of HSPA8 than did that of PHLDA2 (Supplementary Fig. S3g–i). Additionally, in the TCGA database, patients with high CAPN15 expression exhibited a worse prognosis, while high expression of RCOR2 did not affect the prognosis of patients (Supplementary Fig. S3j, k). The cytoplasm-nucleus separation assay showed that HSPA8 was distributed in both the cytoplasm and nucleus of cells (Supplementary Fig. S3l). Moreover, RNA sequencing analysis identified 4172 DEGs, of which 2873 were upregulated and 1353 were downregulated (log2-fold change ≥2.0, P adjusted <0.01, Fig. 3h, i) in HepG2-shPHLDA2 cells relative to HepG2-shNC cells. GO analysis revealed that the terms most enriched with the downregulated DEGs included negative regulation of growth and regulation of the extrinsic apoptotic pathway (Supplementary Fig. S3m). Beyond the enriched GO terms, KEGG pathway enrichment analysis showed that the downregulated genes were more strongly involved in cell proliferation in HepG2-shPHLDA2 cells than in HepG2-shNC cells, as the enriched terms included the PI3K-Akt signaling pathway and insulin resistance (Fig. 3j), suggesting that the HSPA8 mediated increase in PHLDA2 expression plays crucial roles in hepatocarcinogenesis. Taken together, we concluded that HSPA8 can bind to the PHLDA2 promoter region, and PHLDA2 may be involved in the progression of HCC and regulated by HSPA8.

Fig. 3. HSPA8 was identified to bind to the PHLDA2 promoter region.

a A circular histogram of the results of Gene Ontology analysis of differentially expressed genes in HSPA8-knockdown HepG2 cells compared with control HepG2 cells. b KEGG enrichment analysis was performed to analyze the HSPA8 regulated pathways. c Peaks in the distribution of HSPA8 on genomic elements, as determined through ChIP assays. d Venn diagram indicating the overlap of genes downregulated in the HSPA8 knockdown group and genes identified to be occupied by HSPA8 by ChIP-seq. e The binding of HSPA8 to the PHLDA2 promoter was examined by ChIP assays in HepG2 cells. f The RNA-seq results for PHLDA2 in shHSPA8-treated HepG2 cells and shNC-control cells. g The correlation between HSPA8 and PHLDA2 expression was examined by RT-qPCR in 55 HCC tissues. (Pearson correlation coefficient, r = 0.603). h Volcano plot indicating the differentially expressed genes identified by RNA-seq analysis in HepG2 cells transduced with shPHLDA2 (log2-fold change ≥2, P adjusted <0.05). i The expression of PHLDA2 in RNA-seq of HepG2-shPHLDA2 cells. j KEGG enrichment analysis were performed with the downregulated genes in PHLDA2-knockdown cells. ***P < 0.001; Student′s t-test. The experiment was repeated at least three times.

HSPA8 activates PHLDA2 to enhance the growth of liver cancer in vitro and in vivo

Gene activation in cells requires the concerted action of transcription factors and coactivator proteins [37, 38]. As described above, we observed that HSPA8 could be expressed in the nucleus, and ChIP assays revealed that HSPA8 was able to bind to the promoter region of PHLDA2. Accordingly, we hypothesized that HSPA8 might transcriptionally activate PHLDA2 in liver cancer cells. RT-qPCR and Western blot analysis showed that the expression of PHLDA2 was significantly decreased when HSPA8 was downregulated in shHSPA8 HepG2 and Huh-7 cells relative to the control cells (Fig. 4a and Supplementary Fig. S4a, b), suggesting that HSPA8 might upregulate PHLDA2 expression at the transcriptional level. To test this hypothesis, three fragments of the PHLDA2 promoter region, pGL3-1408 (−1362 ~ +45), pGL3-908 (−862 ~ +45), and pGL3-404 (−358 ~ +45), were cloned and transiently transfected into HepG2 and Huh-7 cells. The dual-luciferase reporter assay revealed that pGL3-1408 exhibited the maximum luciferase activity, while the luciferase activity of pGL3-908 was sharply reduced (Fig. 4b). However, the luciferase activity of pGL3-404 was not influenced when the expression of HSPA8 was downregulated (Fig. 4c), suggesting that the −862/−358 fragment constitutes the core region of the PHLDA2 promoter. ChIP-qPCR assay with an antibody against HSPA8 showed that recruitment to the core region of the PHLDA2 promoter was significantly decreased in HSPA8-knockdown cells relative to control cells (Supplementary Fig. S4c). Functionally, the MTS and colony formation assays showed that overexpression of PHLDA2 significantly reversed the proliferation inhibition mediated by HSPA8 knockdown in the cells (Fig. 4d–f and Supplementary Fig. S4d–e). Moreover, we validated this finding in animal models (Fig. 4g–i), and these collective results suggest that HSPA8 enhances the growth of liver cancer by upregulating PHLDA2. Thus, we concluded that HSPA8 activates PHLDA2 to enhance the growth of liver cancer in vitro and in vivo.

Fig. 4. HSPA8 activates PHLDA2 transcription to enhance the growth of liver cancer in vitro and in vivo.

a The mRNA and protein levels of PHLDA2 were measured by RT-qPCR and Western blot analysis, respectively, in HSPA8-knockdown HepG2 cells. b PHLDA2 promoter activity was examined by using DLR assays. HepG2 and Huh-7 cells were transfected with pGL3-basic or reporter constructs containing various lengths of the PHLDA2 promoter region. c HSPA8 knockdown reduced PHLDA2 promoter activity in cells, as determined by using DLR assays. d–f Overexpression of PHLDA2 significantly reversed the inhibition of proliferation and colony formation caused by HSPA8 knockdown in HepG2 cells and Huh-7 cells, as shown by Western blot analysis, MTS assays, and colony formation assays. g–i The growth of tumors in nude mice injected with shNC, shHSPA8-transfected, and shHSPA8 combined with overexpression of PHLDA2 HepG2 cells was measured every 5 days. Images of tumors excised from five mice at 27 days and the average weight of the tumors are shown. **P < 0.01; Student′s t-test. The experiment was repeated at least three times.

HSPA8 coactivates the transcription factor ETV4 by binding to ETV4 at the promoter of PHLDA2 to upregulate PHLDA2

To better understand the underlying mechanism by which HSPA8 upregulates PHLDA2, we tried to identify the HSPA8 coactivated transcription factor targeting PHLDA2. The Cistrome DB Toolkit (http://dbtoolkit.cistrome.org/) is a resource of human cis-regulatory data derived from ChIP-seq profiling assays, through which the genome-wide locations of transcription factor binding sites can be mapped [39]. Therefore, we examined TFs that might regulate PHLDA2 gene transcription using the Cistrome DB Toolkit. Interestingly, we found that several transcription factors, such as ETV4, GABPA, HNF4A, MAFK, MAX, and ZBTB7A, might have regulatory effects on PHLDA2 expression in cells. Among those candidates, only ETV4 modulated PHLDA2 expression in HepG2 cells (Fig. 5a). Subsequent analysis using the TCGA database revealed that ETV4 mRNA expression showed the best correlation (r = 0.527) with PHLDA2 mRNA expression in HCC tissues (Supplementary Fig. S5a–f). Moreover, the expression level of the transcription factor ETV4 was positively associated with unfavorable clinicopathological features and poor prognosis in HCC patients (Supplementary Fig. S5g–j). Next, we predicted putative binding sites of ETV4 in the promoter of PHLDA2 by using JASPAR (https://jaspar.genereg.net/). We observed that the binding site was located in the region encompassing positions −362 ~ −353 (Fig. 5b and Supplementary Fig. S5k). Then, we used RT-qPCR to validate the effect of ETV4 on PHLDA2 in the cells. As expected, ETV4 knockdown decreased PHLDA2 expression in HepG2 and Huh-7 cells (Fig. 5c). From a clinical perspective, the expression levels of PHLDA2 and ETV4 were significantly correlated in HCC specimens (n = 55, r = 0.721, P < 0.001, Fig. 5d). Moreover, co-IP assays demonstrated that HSPA8 could interact with ETV4 in HepG2 cells, and immunofluorescence staining showed that HSPA8 and ETV4 were colocalized in the nucleus of HepG2 cells (Fig. 5e, f). ChIP assays revealed that the recruitment of ETV4 to the core region of the PHLDA2 promoter was significantly decreased when HSPA8 was knocked down in the cells and that overexpression of ETV4 restored PHLDA2 promoter activity and protein levels when HSPA8 was knocked down in the cells (Fig. 5g–i), suggesting that HSPA8 bound to ETV4 on the promoter of PHLDA2 to activate the transcription of PHLDA2. Therefore, we screened out ETV4 and confirmed that it was coactivated by HSPA8 and then promoted PHLDA2 transcription.

Fig. 5. HSPA8 coactivates the transcription factor ETV4 by binding to ETV4 in the promoter of PHLDA2 to upregulate PHLDA2.

a TFs with high regulatory potential in HepG2 cells (10 k distance to TSS). b ETV4 binding motif. c The mRNA levels of PHLDA2 were measured by RT-qPCR in ETV4-knockdown HepG2 and Huh-7 cells. d The expression correlation between ETV4 and PHLDA2 was measured by RT-qPCR analysis in 55 HCC tissues (Pearson correlation coefficient, r = 0.721). e The interaction between ETV4 and HSPA8 was examined by co-IP assays in HepG2 cells. f The co-localization of and interaction between HSPA8 and ETV4 were assessed by confocal microscopy in HepG2 cells. g A ChIP assay and agarose electrophoresis were performed to evaluate ETV4 binding to the PHLDA2 promoter in the shNC and shHSPA8 groups. h, i PHLDA2 promoter activity was examined by using DLR assays, and protein levels were measured by Western blot analysis. HepG2 cells were transfected with ETV4 or shHSPA8 plasmids. **P < 0.01; Student′s t-test. All experiments were repeated at least three times.

HSPA8–elevated PHLDA2 contributes to the growth of liver cancer in vitro and in vivo

PHLDA2 has been linked to cancer-related cellular processes, and no study has investigated the role of PHLDA2 in HCC [40]. To identify the importance of the HSPA8 mediated increase in PHLDA2 expression in HCC progression, we analyzed gene expression from transcriptomic data in TCGA and found that PHLDA2 (n = 371) was significantly upregulated in HCC tissues compared with adjacent normal tissues (n = 50) (P < 0.001, Fig. 6a). Moreover, the expression level of PHLDA2 was significantly correlated with tumor stage, pathological differentiation status, Ki-67 expression, and PCNA expression in HCC (P < 0.001, Fig. 6b, and Supplementary Fig. S6a–c). Overall survival analysis revealed that HCC patients with a high level of PHLDA2 expression had a worse prognosis than those with a low level of PHLDA2 expression (Fig. 6c). The silencing efficiency of the shRNA targeting HSPA8 mRNA was validated by RT-qPCR and Western blot analyses in HepG2 and Huh-7 cells (Fig. 6d, and Supplementary Fig. S6d). From a functional perspective, the MTS and xenograft assays confirmed that the growth rates were markedly decreased after PHLDA2 knockdown compared with negative control treatment in vitro and in vivo (Fig. 6e–i), suggesting that PHLDA2 greatly contributes to the growth of liver cancer. Thus, we concluded that HSPA8 coactivates the transcription factor ETV4 by binding to the promoter of PHLDA2, thereby upregulating PHLDA2 expression and contributing to the growth of liver cancer (Fig. 6j).

Fig. 6. The HSPA8 mediated increase in PHLDA2 expression contributes to the growth of liver cancer in vitro and in vivo.

a The mRNA levels of PHLDA2 in HCC tissues and adjacent nontumor liver tissues from the TCGA database. b The correlation between the mRNA levels of PHLDA2 and MKI67 in TCGA HCC tissues is shown (Pearson correlation coefficient, r = 0.41). c Kaplan–Meier overall survival curves for all 370 patients with HCC stratified by the expression of PHLDA2 (high and low). d The knockdown efficiency of PHLDA2 in HepG2 and Huh-7 cells was evaluated by Western blot analysis. e, f Knockdown of PHLDA2 in liver cancer cells inhibited the proliferation of HepG2 cells, as determined by an MTS assay. g–i The growth of tumors in nude mice injected with shNC or shPHLDA2-transfected HepG2 cells was measured every 5 days. Images of tumors excised from six mice at 25 days and the average weight of the tumors are shown. j A model showing the mechanism by which HSPA8 activates PHLDA2 transcription by promoting the binding of ETV4 to the PHLDA2 promoter region and then promoting tumor proliferation. **P < 0.01, ***P < 0.001; Student′s t-test. The experiment was repeated at least three times.

Discussion

HCC is one of the most challenging cancers to treat due to the poor understanding of HCC pathogenesis. HSPA8 plays critical role in the development and progression of cancer by ensuring the proper folding or degradation of misfolded proteins [41]. Moreover, as a core component of the chaperone-mediated autophagy (CMA) machinery, HSPA8 recognizes cytosolic proteins with a specific motif, and directly targets them for translocation to lysosomes for degradation [42]. However, the underlying mechanism by which HSPA8 promotes hepatocarcinogenesis is not well documented. In this study, we identified the novel function of HSPA8 in liver cancer.

We first examined the expression of HSPA8 and found that it was markedly upregulated in clinical liver cancer tissues. Notably, in patients with HCC, a high level of HSPA8 expression was significantly correlated with worse prognosis and unfavorable clinicopathological characteristics. However, the statistics were not available due to the small sample size of patients with Stage 4 disease.

Then, our evidence showed that HSPA8 could significantly promote the growth of liver cancer cells. Regarding its function in HCC tumorigenesis, decreased lipid oxidation via an HSPA8-mediated mechanism results in the progression of nonalcoholic steatohepatitis [43]. Baccarini and his colleagues revealed that YAP1 accumulated in lysosomes by binding to HSPA8 and that this event resulted in the promotion of tumor growth in HCC [14]. Although HSPA8 has been identified as a tumor-promoting factor in multiple cancers, the mechanism of HSPA8 in HCC progression has not been thoroughly explored [44, 45]. In this study, we observed that HSPA8 was expressed in the nucleus of liver cancer cells using IHC staining of tissue microarrays. This implies that HSPA8 may play a role in modulating gene transcription. Interestingly, RNA-seq analysis showed that HSPA8 participated in transcriptional regulatory activity, cell proliferation, cellular processes, biological regulation, molecular function regulation, and cell growth. Thus, we hypothesized that HSPA8 might be involved in gene regulation in HCC. Moreover, the ChIP-Seq profiles revealed that 10.0% of binding peak locations mapped to promoter regions. Then, integrated analysis of the transcriptome sequencing data and the ChIP-Seq data identified six genes that might be directly regulated by HSPA8. PHLDA2, RCOR2, and CAPN15 were demonstrated to be regulated by HSPA8, as shown by RT-qPCR. The ChIP-quantitative PCR assays showed that PHLDA2 was the gene most affected by HSPA8. Our findings suggest that HSPA8 is able to transcriptionally upregulate PHLDA2 in liver cancer cells. From a clinical perspective, the expression level of PHDLA2 showed a significant difference between the HSPA8 high expression group and the HSPA8 low expression group. Patients with high PHLDA2 expression exhibited a worse prognosis. The expression levels of HSPA8 and PHLDA2 showed a strong correlation in 55 HCC patients as determined by RT-qPCR (r = 0.603, P < 0.001). This strongly supports the hypothesis that HSPA8 upregulates PHLDA2 in HCC.

In seeking to better understand the underlying mechanism by which HSPA8 upregulates PHLDA2, we used the Cistrome DB Toolkit to predict that the transcription factor ETV4 might regulate PHLDA2 expression. It has been reported that ETV4 is overexpressed in liver cancer and promotes the proliferation and invasion of HCC cells [46]. However, the role of ETV4 in modulating HCC progression is not yet comprehensively understood. Interestingly, we found that HSPA8 and ETV4 could colocalize in the nucleus. Co-IP assays further demonstrated that HSPA8 could interact with ETV4 in cells. Dual-luciferase reporter assays revealed that HSPA8 upregulated the expression of PHLDA2 by enhancing the transcriptional activity of ETV4 toward PHLDA2. HSPA8 knockdown markedly decreased the regulatory effect of ETV4 on PHLDA2 transcription. ChIP assays revealed that ETV4 could be recruited to the core region of the PHLDA2 promoter in cells. This suggests that HSPA8 upregulates PHLDA2 by activating the transcription factor ETV4.

Considering that PHLDA2 promotes tumor progression via different mechanisms in breast cancer, pancreatic ductal carcinoma, and colorectal cancer, we tried to identify the importance of the HSPA8 mediated increase in PHLDA2 expression in liver cancer. We found that PHLDA2 was overexpressed in HCC tissues and correlated with a poor prognosis. From a mechanistic perspective, the functions of PHLDA2 in the promotion of tumor growth might be related to the extrinsic apoptosis pathway, cell proliferation, PI3K-Akt signaling pathway, and insulin resistance, according to the RNA-seq data. In this study, we first reported that PHLDA2 was a target gene regulated by HSPA8 in HepG2 cells. The DLR assays confirmed that HSPA8 could activate the PHLDA2 promoter. From a functional perspective, PHLDA2 was capable of promoting the growth of liver cancer cells in vitro and in vivo.

In summary, HSPA8 acts as a coactivator of the transcription factor ETV4 and upregulates PHLDA2 in liver cancer. The HSPA8/ETV4/PHLDA2 axis drives the growth of HCC cells. Therefore, we conclude that HSPA8 serves as a transcriptional coactivator of ETV4 and promotes the progression of liver cancer by upregulating PHLDA2. From a therapeutic perspective, HSPA8 and PHLDA2 are potential targets for HCC treatment.

Author contributions

SW, YFW, and GY performed most of the experiments; HHZ, HFY, CYH, LNZ, YHS, JS, LLS, and PL accomplished some of the in vitro and in vivo experiments; XDZ, WL, NNZ, and YS conceived and designed the project and wrote the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41401-023-01133-3.