PSMD4 drives progression of hepatocellular carcinoma via Akt/COX2 pathway and p53 inhibition

The ubiquitin-dependent proteolytic pathway is crucial for cellular regulation, including control of the cell cycle, differentiation, and apoptosis. Proteasome 26S Subunit Ubiquitin Receptor, Non-ATPase 4, (PSMD4) is a member of the ubiquitin proteasome family that is upregulated in multiple solid tumors, including hepatocellular carcinoma (HCC), and the existence of PSMD4 is associated with unfavorable prognosis. In this study, transcriptome sequencing of HCC tissues and non-tumor hepatic tissues from the public database Cancer Genome Atlas (TGCA) revealed a high expression of PSMD4. Additionally, PSMD4 loss in HCC cells suppressed the tumor development in mouse xenograft model. PSMD4, which is maintained by inflammatory factors secreted from tumor matrix cells, positively mediates cell growth and is associated with Akt/GSK-3β/ cyclooxygenase2 (COX2) pathway activation, inhibition of p53 promoter activity, and increased p53 degradation. However, the domain without the C-terminus (VWA+UIM1/2) sustained the activation of p53 transcription. Thus, our findings suggest that PSMD4 is involved in HCC tumor growth through COX2 expression and p53 downregulation. Therapeutic strategies targeting PSMD4 and its downstream effectors could be used for the treatment of PSMD4-abundant HCC patients.


Introduction
Hepatocellular carcinoma (HCC) is a highly aggressive form of cancer and has been the fourth leading cause of cancerrelated death worldwide [1]. Clinically, surgery with curative intent is feasible for only 15-25% of patients, and the longterm prognosis of HCC remains unsatisfactory due to high recurrence rates. Moreover, the overall therapeutic effect is unsatisfactory owing to the highly radiotherapy-resistant characteristic of the tumor and high cytotoxicity induced by chemotherapeutic drugs [2]. Currently, the increasing number of positive phase III studies on lenvatinib as first-line treatment and cabozantinib and ramucirumab as second-line treatments, respectively, appears as a promising foreground [3][4][5]. Meanwhile, it is important to appreciate the molecular pathogenesis of HCC to find new targets for diagnosis, chemoprevention, and treatment in the future.
The proteasome 26S subunit, also known as non-ATPase 4 (PSMD4), is a major receptor member of the ubiquitindependent proteolytic protein system. It is an intrinsic subunit of the 19S regulatory particle that involved in substrate recognition and bound to ubiquitin molecule [6]. The 26S 1 3 proteasome recognizing the poly ubiquitin chain is important in the selective degradation of target proteins, and the subunit of the 26S proteasome, PSMD4, preferentially binds to the poly ubiquitin chain in vitro [7]. Additionally, PSMD4 is an important regulator of specific biological functions, including apoptosis, modulation of inflammatory processes, and cancer pathogenesis [8][9][10]. PSMD4 is a well-known oncoprotein commonly overexpressed in HCC; it can induce tumor progression in HCC through the binding of hypoxia status-induced transcription factor HIF1α to the PSMD4 promoter [11].
The phosphoinositide-3-kinase (PI3K)/Akt pathway is involved in multiple cellular processes, including cell proliferation, survival and motility, which are critical adaptation in tumorigenesis. Increasing research suggests that activation of Akt signaling is crucial for the development of human hepatocarcinoma [12,13].
Cyclooxygenase-2 (COX2), an enzyme that plays a critical role in inflammation [14], is principally induced by the activation of phospholipase A2, which catalyzes the hydrolysis of cell membrane phospholipids by multiple factors [15]. Recent evidence suggests that COX2 signaling is involved in hepatocarcinogenesis [16]. COXs and their metabolic products, such as prostaglandins (PGs) and leukotrienes, are considered as novel preventive and therapeutic targets in cancer [17].
The major regulatory transcription factor, tumor suppressor-p53, plays an important role in the regulation of cell cycle, apoptosis, metabolism, and genetic stability of mammalian cells [18]. Thus, that is no surprise that p53 inactivation is one of the most frequent events during human carcinogenesis, making it critical to explore the molecular mechanisms behind p53 dysregulation in tumors.
In the study, we researched the possible crosstalk between PSMD4 and the Akt/cyclooxygenase2 (COX2) pathway in hepatocarcinogenesis, using in vitro and in vivo approaches. We found that PSMD4 expression in human HCC tissues correlated positively with COX2 and was dependent on the Akt pathway. Additionally, we found that the C-terminal of PSMD4 plays an essential role in p53 transcription activity and induces carcinogenesis in HCC. Thus, PSMD4 could be a valuable prognostic marker and potential therapeutic molecule target for human HCC patients.

Bioinformatics analysis
The public datasets (GSE114269, GSE14520, GSE54236) from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas Program (TCGA) containing HCC tissues and non-tumor tissues were used to analyze the gene expression. The Cox proportional-hazards model was introduced to analyzing survival data. Functional enrichment analysis of Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to predict the pathway activated by PSMD4 overexpression. All single-cell RNA-seq analyses were publicly available cell-by-gene expression matrices that were aligned to the HCC transcriptome. All bioinformatic analyses were conducted using the R language procedure and R package.

Cell lysis and immunoblotting
Schizolytic proteins (30 μg) separation were carried out by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride (PVDF) membrane. Then, membranes were blocked using 5% fatfree milk for 2 h at 26 °C and were incubated with primary antibodies overnight at 4 °C. Membranes with secondary antibodies were washed and developed using a commercial chemiluminescence detection kit (Cat. #BL520B).

Immunohistochemistry (IHC)
Tumor samples and non-tumor tissue microarrays derived from patients with HCC were purchased from Bioaitech (Shanxi, CHN). For IHC analysis, deparaffinized sections with xylene and gradient alcohol were heated to just below the boiling temperature in Tris-Ethylene Diamine Tetraacetic Acid buffer (Cat#. C1037, Solarbio) for antigen retrieval. To inhibit endogenous peroxidase, the 3% hydrogen peroxide was incubated with retrieval sections for 5 min. Then, the samples were blocked with 5% Bovine Serum Albumin (BSA) for 1 h at 25 °C. The PSMD4 protein in tumor tissue sections was incubated with PSMD4 primary (Cat. #14899-1-AP, Proteintech Group) and secondary antibodies and detected using a 3,3'-Diaminobenzidine (DAB) detection kit (Cat. #ZLI-9018, ZSGB-BIO). All specimens were evaluated using the tissue array method. The immunoreactive intensity for each case was scored as none, weak, moderate, or strong. Histological examinations were performed by an experienced pathologist who was blinded to clinical information.

Cell immunofluorescence staining
For immunofluorescence, the cells were seeded on sterile glass coverslips, fixed using 4% paraformaldehyde, and blocked with 3% BSA. Transient transfection of cells was performed with polyclonal antibodies against PSMD4 and tetramethylrhodamine isothiocyanate isomer R (TRITC)conjugated secondary antibodies. After washing, to visualize the cell nuclei, the cells were stained with 1 μg/mL Hoechst 33,258 for 15 min and mounted using an anti-fade mounting reagent. The cells were examined using confocal microscopy (LCM 510, Carl Zeiss, Jena, Germany).

Cell colony formation assay
1000 cells were seeded and cultured in 6-well plates in an incubator at 37 ℃ for 2 weeks for colony formation assay. The fixed cells in methanol were then stained with 0.05% crystal violet. Viable colony formation was observed using a camera.

Methylthiazol tetrazolium (MTT) assay
The MTT assay was used to evaluate the proliferation ability of cells according to the manufacturer's instructions. Cells (2 × 10 4 ) seeded in 24-well plates were cultured for 24, 48, 72, 96, 120, and 144 h, respectively. Cells incubated with 200 μL of 0.5 mg/mL MTT for additional 4 h, then 200 μL dimethyl sulfoxide were used to replace the cell medium for resolving the cell crystals. Absorbance was read at 540 nm and recorded using a 96-well plate reader.

Soft agar formation assay
To allow the bottom agar layer to solidify, 750 μL 2 × cell culture medium (containing 20% FBS and 2 × penicillin and streptomycin) and 750 μL 1.2% Seaplaque agarose solution mixed at 1:1 ratio and transferred into clear, flatbottom 6-well plates, which were then incubated at 25 °C for 30 min. 750 μL of cell suspension was mixed with an equal volume of 0.7% agar solution and placed on the bottom agar layer; this mixture was immediately solidified at 25 °C for 15 min. 1 mL of DMEM containing 10% FBS per well was used to incubate the plates for 20 days at 37 °C and 5% CO 2 condition. Change the medium every 3-4 days.

Mouse tumorigenicity assay
Four-week-old female athymic nude mice from the Beijing Vital River Laboratory Animal Technology were used in all experiments. The animals were housed under specific pathogen-free condition at 20-25 °C with 40-60% humidity and 12-h light/dark cycles; 5 × 10 6 shPSMD4 cells or-Luc-shPSMD4 cells were injected into the flank of nude mice. Monitoring mice daily, and survival was recorded for each mouse. The tumor volume was calculated according to the following equation: V (mm 3 ) = width 2 (mm 2 ) × length (mm)/2. Tumor growth was observed when the tumor reached a size of 50 mm 3 .

Construction of recombinant adenovirus
The DNA sequence encoding GFP-PSMD4 was synthesized and cloned into pAd/CMV/V5-dest by an LR reaction kit (Cat. # 11791020, Invitrogen). pAd/CMV/V5-dest containing PSMD4 was transfected into HEK293T cells to generate an adenovirus using Lipofectamine 3000. The adenovirus supernatant was harvested and purified using an adenovirus purification kit (Cat. #C2901S, Beyotime Biotechnology). The adenovirus titer was determined, and cells were treated with adenovirus encoding PSMD4 at the recommended MOI.

Phospho-proteome profiling
Transfected cells were lysed in a NP-40 lysis buffer (10 μg/ mL aprotinin, and 10 μg/mL leupeptin) by rocking at 4 °C for 30 min. The lysates were centrifuged at 14,000×g for 15 min at 4 °C, and collect the supernatants. BCA Protein Assay Kit (Cat. #P0012, Beyotime) was used to quantify total protein. Capture antibodies spotted in duplicates on nitrocellulose membranes binding to specific target proteins were presented in the sample lysates (500 μg) according to the manufacturer's instructions. Array data were visualized using chemiluminescence reagents.

Luciferase assay
Cells were transfected with p53-PTA-luc or PTA-luc vector construct using lipofectine3000 reagents for 48 h. The Dual-Luciferase Reporter Assay System (Cat. #E1910, Promega) was used to measure the activity of luciferase in firefly and luciferase in Renilla using a microplate reader.

Chromatin immunoprecipitation assay
HLK3 cells transfected with different p53 promoters were used to establish the binding of PSMD4 upstream of the p53 gene by a chromatin immunoprecipitation assay. Crosslinking protein-DNA complexes were tattered using an ultrasonic cell disruptor in chromatin immunoprecipitation (ChIP) lysis buffer. Then, ChIP-grade antibodies of PSMD4 and Protein A/G Agarose were used for immunoprecipitation of isolated chromatin linked to the PSMD4 protein. Both the input and IgG groups were used as control groups according to the protocol. The chromatin complex was eluted using Protein A-Sepharose beads (Cat. #CL-4B, Cytiva) in the elution buffer. DNA extraction by phenol/chloroform was precipitated with ethanol and performed polymerase chain reaction (PCR) using Taq DNA polymerase (Cat. #27-0798-04, Cytiva) according to the manufacturer's protocol. Primers (Table 1) were used to amplify the DNA products.

Statistical analysis
All experimental results are expressed as the mean ± standard deviation of at least three independent experiments. Statistical evaluations were conducted using two-tailed t tests, Mann-Whitney U tests, and chi-square tests. Survival rates and differences in survival curves were analyzed by the Kaplan-Meier method and log-rank test, respectively. All statistical analyses were performed using IBM SPSS Statistics 26 (IBM SW, Cambridge, MA, USA). P < 0.05 was considered as statistically significant (*P < 0.05, **P < 0.01, ***P < 0.001).

PSMD4 is enriched in HCC and related to patient survival
To assess the function of PSMD4 expression in HCC, we first analyzed its expression pattern using Gene Expression Profiling Interactive Analysis (GEPIA) online on TCGA public database between HCC tumors (T) and non-tumor samples (N) (Fig. 1A). PSMD4 was preferentially overexpressed in HCC tissues compared with non-tumor tissues, and the results were identified in Array Express-Functional Genomics Data (E_TABM_36) and Gene Expression Omnibus (GEO) databases (GSE144269, GSE14520, GSE54236) ( Supplementary Fig. 1A). Kaplan-Meier survival analysis demonstrated that the higher the levels of PSMD4, the shorter the overall survival (OS) time (P = 0.038; Fig. 1B). To explore the effect of PSMD4 expression on the survival of HCC patients, a Cox proportional-hazards model was introduced to calculate the hazard ratio (HR), and we concluded that the HR in different groups can exceed 1 (except GSE10141 database) ( Supplementary Fig. 1B). In addition to HCC cells, data retrieved from our findings also revealed a PSMD4 overexpression in multiple cell types, including renal epithelial cells (HEK293T), Chang liver cells, THLE2, HepG2, Hep3B, PLC/PRF/5, Huh7, SK-HEP-1, SH-J1, HLK2, HLK3, HKK2, HLK4, HLK1, HLK5, HKK1, HKK6, LM-3, SNU-182, SNU-387, and Li-7 cell lines (Fig. 1C). Thus, protein levels of PSMD4 were positively correlated with HCC development. We then determined protein levels of PSMD4 in HCC patients' tissue microarray samples and non-tumor hepatic tissues and found that compared to non-tumor tissues, the HCC tissues exhibited high levels of PSMD4 protein expression (Fig. 1D). In addition, the correlation analysis of PSMD4 protein expression and clinicopathological features in the HCC tissues was carried out Table 2). Chi-square analysis indicated a significant association between PSMD4 expression and gender (P = 0.0002) and grade (P = 0.0060).

PSMD4 overexpression increases cell proliferation
To assess the effect of PSMD4 on tumor cell growth, PSMD4 cell lines stably expressing HLK3 were established by G418 screening (P-3, P-19, P-23, and P-28) and cultured in the same medium as the parental cell line ( Fig. 2A). Specific bands corresponding to the myc-PSMD4 and GFP-PSMD4 proteins were detected by PSMD4 antibody (Supplementary Fig. 2A), and the specific immunofluorescence of PSMD4 transfected with the GFP-tagged PSMD4 expression plasmid was detected in HLK-3 and HepG2 cells by an immunofluorescence assay. The results revealed that endogenous PSMD4 and ectopic expression of PSMD4 were localized mainly in the nuclei of cells and PSMD4 was less abundant in the cytoplasm (Supplementary Fig. 2B). Colony generation ability analysis and cellular growth measurement showed that PSMD4 expression in HCC cells had conspicuously maximal intrinsic effectors on cell proliferation compared to the parental cells ( Fig. 2B-D). Additionally, the clonogenicity of PSMD4-knockdown cells was significantly reduced in the MTT assay (Fig. 2E). Furthermore, confirming the importance of PSMD4 in HCC cell proliferation, colony formation, and anchorage-independent cell growth, PSMD4 knockdown formed fewer and smaller colonies (Fig. 2F-H). To assess whether the effect of PSMD4 on HCC proliferation was mediated by advancing mitosis, immunofluorescence demonstrated that PSMD4 overexpression began with the pre-prophase and was arrested to its nadir during interphase ( Supplementary Fig. 3A). In G2/M phase, thymidine-aphidicolin double-arrest of mammalian cells revealed cell cycle-dependent expression of PSMD4, in which the cyclinB1/Cdc2 complex was essential (Supplementary Fig. 3B), and PSMD4 knockdown decreased the phosphorylation of Cdc25C or Cdc2, inhibiting the M phase ( Supplementary Fig. 3C).

Knockdown PSMD4 inhibits hepatocellular carcinoma cell migration and invasion
A previous study demonstrated that PSMD4 promotes tumor progression, and a wound healing assay showed that knockdown of PSMD4 impaired the migration and invasion abilities of HCC cells in both SK-HEP-1 and Huh7 cell lines (Fig. 3A, B). As an important biological process for HCC cell invasion, that PSMD4 is associated with Epithelial-mesenchymal transition (EMT) was examined in these HCC cells. PSMD4 knockdown caused increase in E-cadherin expression and decrease in vimentin expression. However, PSMD4 knockdown did not affect cytokeratin 8 or cytokeratin 18 expression (Fig. 3C). This observation was confirmed by bioinformatic analysis, in addition, we extracted single-cell transcriptomic sequencing data from metastatic HCC on the GEO platform (http:// omic. tech/ scrna-hcc/), and divided the malignant hepatocyte cells into PSMD4 high expression and low-expression groups. We then performed VISION algorithm analysis for both groups of data and found that in all the malignant hepatocyte cells, the value score of cells expressing high PSMD4 was close to 1 (Fig. 3D). Thus, PSMD4 ablation in HCC affects its migration and invasion abilities.

Knockdown of PSMD4 inhibits tumorigenicity in vivo
Stable PSMD4-knockdown cells were subcutaneously inoculated into nude mice to assess the effect of PSMD4 on HCC cell growth in vivo. Two weeks post-inoculation, in vivo studies showed that PSMD4 knockdown significantly suppressed the tumorigenicity of SK-HEP-1 and Huh7 cells (Fig. 4A, B). For further clarification, we established PSMD4-knockdown cells expressing luciferin in SH-J1 and confirmed that knockdown of PSMD4 inhibited tumorigenicity in vivo using a NOD/SCID mouse model (Fig. 4C). Through linear fitting analysis, the growth fitting line of the tumor tissue size in each group and the slope indicated the growth rate of the tumor (Fig. 4D). Histopathological analysis showed that compared to the parental cell line, PSMD4-knockdown cells had minimal heterogeneity (Fig. 4E).

Activation PI3K/Akt/GSK-3β signaling pathway by PSMD4
To further elucidate the molecular mechanism by which PSMD4 promotes cell proliferation in vivo and in vitro, we first confirmed the PSMD4 expression in HLK3 cells infected with PSMD4 recombinant adenovirus (Fig. 5A). Two antibody microarrays-the Proteome Profiler Human Phospho-Kinase Array Kit (Cat. #ARY003C, R&D Systems), and Human Phospho-MAPK Array kit (Cat. #ARY002B, R&D Systems)-were used to reveal the downstream phosphorylated molecule of the PSMD4 protein ( Fig. 5B and Supplementary Fig. 4A). PSMD4 expression induced phosphorylation of PI3K, Akt (S473 and S474), GSK-3β (S9), mTOR (S2488), and 4E-BP1 (T37/46) (Fig. 5C, D). However, PSMD4 overexpression did not affect the phosphorylation site, T308 of Akt or the phosphorylation status of PDK1. Furthermore, to verify this result, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was used in the high and low PSMD4 expression groups; the analysis revealed high PSMD4 expression contributed to extreme activation of the PI3K-Akt signaling pathway ( Supplementary  Fig. 4B). These results demonstrate that PSMD4 enhances the activation of the PI3K-Akt pathway in HCC.

Phosphorylated Akt/GSK-3β as mediator of PSMD4 modulating COX2 expression
HLK3 cells treated with Ad-PSMD4 plasmid were subjected to RNA transcriptome sequence analysis. Among the genes significantly upregulated following PSMD4 overexpression, COX2 was the first gene to be highly expressed, as shown in the heatmap (Fig. 6A). COX2 encodes the enzyme COX2, which assumes the responsibility for the formation of PGs from arachidonic acid. COX2 expression is a marker of poor prognosis, and PI3K/Akt modulates COX2 expression in colorectal, breast, and gastric cancer cells [20][21][22][23]. We confirmed that PSMD4 increases COX2 protein and mRNA levels (Fig. 6B). This result is consistent with our observations in Huh7 and 293A cell lines (Fig. 6C). Simultaneously, our immunoblot and immunofluorescence results showed that PSMD4 expression inducing COX2 protein was almost entirely located in the nucleus and not in the cytoplasm ( Fig. 6D and Supplementary Fig. 5). Further investigation showed that only the cells expressing endogenous phosphorylated Akt (S473) expressed COX2 (Fig. 6E). To further identify the correlation between Akt and COX2, HLK3 and Huh7 cells were transfected with wild-type Akt (Wt-Akt1), activated Akt (Act-Akt1), and dominant-negative Akt (dn-Akt1) constructs. The results showed that phosphorylation of Akt leads to significant upregulation of COX2. Similarly, dominant-negative Akt failed to trigger an increase in COX2 levels (Fig. 6F). We also verified that PSMD4 expression significantly increased the transcriptional activity of COX2 (Fig. 6G). Our western blot analysis revealed that the Akt inhibitor downregulated COX2 expression induced by the Act-Akt construct, and the GSK-3β inhibitor lithium chloride (LiCl) enhanced COX2 expression. However, the ERK inhibitor PD98059 failed to modulate COX2 expression (Fig. 6H). In summary, activated Akt can mediate the regulation of PSMD4 to COX2.

Negative regulation of p53 by PSMD4
PSMD4 inhibits the degradation of p53 protein work as a proteasomal ubiquitin receptor, leading to the accumulation of ubiquitinated p53 [24]. To explore the role of PSMD4 in the regulation of p53 expression in HCC cells, PSMD4 knockdown in cell lines (SK-HEP-1, Huh7, HepG2, and A549) significantly induced p53 protein and mRNA expression ( Fig. 7A-C). Furthermore, to elucidate the underlying mechanisms of PSMD4 suppressing p53 expression, a dual-luciferase reporter assay was performed, which showed that PSMD4 loss could prominently enhance the luciferase activity of the p53 promoter (Fig. 7D). PSMD4 reduced the transcriptional activity of p53 in a dose-dependent manner (0-2 µg) (Fig. 7E). Moreover, western blot analysis of HCT116 p53 (-/-) , p53 (-/+) , and p53 WT cells demonstrated that the expression of p53 was reduced in case of high PSMD4 expression (Fig. 7F). Meanwhile, cell proliferation was increased in COX2 transfectants and inhibited in p53 transfectants compared with control LacZ by MTT assay (Supplementary Fig. 6A). These findings indicate that PSMD4 negatively regulates the protein expression of p53 by reducing transcription viability and facilitating the expression of COX2. Furthermore, p53 suppression induced by PSMD4 was rescued following MG132 treatment ( Supplementary  Fig. 6B).

PSMD4 interacts with p53 dependent on the VWA + UIM1/2 domain
To define the functional domain of PSMD4 involved in HCC expansion, diverse truncated forms of PSMD4 that deplete different fragments were constructed. PSMD4 (amino acids 1-377) contains an N-terminal consensus domain comprising a VWA domain (amino acids 5-188) and two C-terminal ubiquitin-interacting motifs (UIM1, amino acids 211-230; and UIM2, amino acids 282-301) [24,25]. Import of substrates to the 20S subunit for proteolytic degradation is dependent on the interacting of the proteasome complex and ubiquitin-interacting motif (UIM) domains. VWA domain of S5a limits ubiquitin and Ubl binding only upon proteasomal association [26]. Diverse deletions of the PSMD4 fragment were also cloned into pcDNA3.1/myc-hisA and pEGFPN3 vectors in both HLK3 and HepG2 cells (   Supplementary Fig. 7). Next, a colony formation assay was performed, and the results showed that FL (PSMD4full length), VWA, VWA+UIM1, ∆N0, ∆N1, ∆VWA, ∆N2, ∆N3, ∆UIM1, ∆UIM2, and ∆UIM1/2 significantly upregulated colony formation in HLK3 cells; however, the VWA+UIM1/2 domain, which does not contain a C-terminal, negatively affected colony formation ability in HCC cells (Fig. 8B). This result suggested that VWA+UIM1/2 plays an important role in PSMD4-mediated cell proliferation. As a potential tumor-suppressive factor, p53 can cause proteasomal degradation via PSMD4-mediated ubiquitylation process. [24] However, to determine whether the VWA+UIM1/2 domain could regulate the transcription of p53 and inhibit colony formation of HCC cells, the p53 luciferase reporter plasmid (p53-Luc (− 3123− + 396)) was transfected into the HLK cell line with diverse PSMD4 truncation plasmids. The results showed that the VWA+UIM 1/2 domain successfully elicited a response (Fig. 8C). Furthermore, the 1704 kb upstream region (− 1998 base to − 255 base) from the transcription starting point of p53 was dissected into six fragments to predict the binding of PSMD4 by ChIP-PCR analysis, and segment 5 (P5) and segment 6 (P6) were predicted as binding sites (Fig. 8D). To further validate whether the VWA+UIM1/2 domain is correlated with p53 expression, we examined expression levels of p53 in SK-HEP-1 cells using western blot analysis. p53 expression was decreased in SK-HEP-1 cells transfected with full-length (FL) PSMD4, VWFA1-2, and ∆UIM1/2 segments compared with the VWA+UIM1/2 truncated form (Fig. 8E). Thus, the C-terminal domain of PSMD4 is essential for p53 transcription. Accumulating evidence suggests that chronic inflammation increases the susceptibility and risk of cancer development [27]. Furthermore, to confirm the effect of tumor microenvironmental inflammation

Discussion
Currently, clinical targeted therapy to cancer is gaining popularity; it is preferred over conventional (cytotoxic) chemotherapy that acts non-selectively and affects the proliferation of rapidly dividing cells in the body while directly targeting one or more molecules in tumor cells by antibodies and kinase inhibitors [28]. Recent advances in bispecific or monoclonal antibody therapeutic candidates targeting unique tumor antigens demonstrated effective and safe treatments in clinical trials [29]. Further research is needed on novel therapeutic agents and for understanding the pathophysiological mechanisms of HCC to improve the survival of patients [30]. In this study, we provided important insights regarding how antigen protein PSMD4 determines HCC progression and how it is affected by inflammatory factors in the tumor microenvironment, linking the poor prognosis of HCC to high PSMD4 expression. We established a concrete mechanism regarding PSMD4 regulation and HCC progression. The ubiquitin-proteasome system (UPS) is a tanglesome and tight quality control system, which is responsible for degradation of 80-90% proteins; it plays a core role in regulating cellular function and maintaining protein homeostasis to enable cells adapt to environmental conditions [31]. The UPS mediates cellular functions and is the preferred target of anticancer therapies [32]. Increased understanding of the roles of ubiquitylation in cell cycle control, DNA damage repair, and signal transduction regulation in the process of tumor development provides molecular insights into the crucial events of hepatoma cancer [33]. As a ubiquitin receptor in the proteasome 26S subunit, PSMD4 binds to ubiquitin via two independent ubiquitin-interacting motifs (UIMs) [34]; it is an important factor for targeting misfolded proteins to multiple quality control destinations, such as the proteasome, lysosomes, and aggresomes, as well as for triggering mitophagy to eliminate defective mitochondria [35]. Previous literature has shown the pathological effect of PSMD4 in the occurrence and development of HCC [11,36], but the mechanism and factors affecting PSMD4 expression in tumors remains not clear. Recent research elucidated that PSMD4 expression in HCC cells is induced by hypoxia and HIF-α is a key upstream regulator of PSMD4 that directly binds to the promoter of this gene, resulting in the development of HCC [11]; however, there is insufficient information regarding the downregulation of PSMD4 in HCC development. Furthermore, according to the epidemiology data of hepatocellular carcinoma, men have a higher risk of hepatocellular carcinoma (HCC) than women. One primary factor propelled this gender bias is potentially that female sex hormones play a protective role for bodies [37]. PSMD4, as an important proto-oncogene that promotes cancer cell growth, we conjecture that its relatively lower expression of humor tissues from female are affected by sex hormones. Meanwhile, the expression level of PSMD4 is higher among the patients with poorer pathological grading. An accumulating pile of evidence supports PSMD4 as a potent therapeutic target for multiple tumors [38,39].
We reported the tumor-supporting role of PSMD4 in hepatoma cancer cell growth. Analysis from the TCGA database and immunohistochemistry staining indicated that PSMD4 expression in HCC tissues was significantly higher than that in non-tumor liver tissues. The prognoses of patients in hepatoma cancer analyses also suggested that high PSMD4 expression is significantly associated with poor prognosis. For the hepatoma cancer cell growth, PSMD4mediated transcription of COX2 depended on Akt phosphorylation. PSMD4 loss in hepatoma cancer cell activates the transcription activity of p53. As a member of the UPS proteasome regulatory particle base subunit 1, PSMD4 may destabilize the tumor-suppressive factor, p53, and the process is dependent on its C-terminal motif.
COX, the limiting enzyme in the synthesis of PGs from arachidonic acid [40], is overexpressed in several types of malignancies, including breast, colorectal, and non-small cell lung cancer [41][42][43]. Its overexpression is intimately linked to poor prognosis in human cancers and portends carcinogenesis in experimental animal models [42]. Here, we identified PSMD4 as an indirect transcriptional regulator of COX2, the gene that encodes the limited enzyme COX2; the transcriptional process is dependent on the phosphorylation of Akt at the 473 tryptophan residue.
Inflammatory ligands secreted by CAFs are major components of the stroma; they promote tumor proliferation, therapy resistance, and immune exclusion [44]. Chronic, dysregulated, persistent, and unresolved inflammation is regulated to promote the occurrence of malignant diseases [45]. We found that the inflammatory factors prostaglandin E2, interleukin-1β, and interleukin-6 contributed in the induction of PSMD4 for regulating tumorigenesis.
In conclusion, we established that PSMD4 activates COX2 over-production via the PI3K/Akt pathway, inhibits p53 transcriptional activity, drives p53 degradation, and induces tumorigenesis. Our findings showed that inflammatory factors trigger over-activation of the PSMD4/Akt/COX2 axis and inhibit p53 expression (Fig. 8F).
Ethical approval All mouse experiments were authorized by the Laboratory Animal Ethical and Welfare Committee of Hebei Medical University (IACUC-Hebmu-2022008) and performed in accordance with Laboratory animal-Guideline for ethical review of animal welfare of People's Republic of China National Standard (GB/T 35892-2018).