PHF5A facilitates the development and progression of gastric cancer through SKP2-mediated stabilization of FOS

Abstract

Background: Gastric cancer (GC) is the fifth most common cancer and the third most common cause of cancer death worldwide. Plant homeodomain (PHD)-finger domain protein PHF5A has been demonstrated to play a promoting role in a variety of cancers. This study aimed to clarify the role of PHF5A in the progression of GC and its potential mechanism of action.

Methods: Immunohistochemical staining experiments were performed based on tissues from clinical GC patients to reveal PHF5A expression. A series of functional experiments in vitro and in vivo were used to clarify the role of PHF5A in GC.

Results: Clinically, PHF5A was abundantly expressed in GC and existed clinical value indicating poor prognosis. In addition, GC cells with knockdown of PHF5A expression showed slowed proliferation, enhanced sensitivity to apoptosis and inhibition of migration. Mechanically, knockdown of PHF5A led to decreased protein stability of FOS, which was mediated ubiquitination of E3 ubiquitin ligase S-phase kinase-associated protein 2 (SKP2). Moreover, downregulation of FOS attenuated the promotion of PHF5A overexpression on GC cells. Consistently, Pladienolide B (PHF5A inhibitor) treatment reversed the induction of PHF5A overexpression on the malignant phenotypes and tumor formation of GC cells.

Conclusion: Knockdown of PHF5A inhibited the progression of GC through SKP2-mediated ubiquitination of FOS, which may be a promising candidate target with potential therapeutic value.

Keywords: FOS; Gastric cancer; PHF5A; Phenotype; Ubiquitination.

Fig. 1

The high expression of PHF5A in GC exerts clinical value for poor prognosis. A PHF5A was highly expressed in tumor tissues compared with adjacent normal tissues in GC patients. (The magnification is 200× and 400×). B The mRNA level of PHF5A in GC tumor tissues and normal tissues. C, D The mRNA and protein levels of PHF5A in human GC cell lines MGC-803, AGS, SGC-7901, BGC-823 and normal gastric mucosa cells GES1cells. The presented results were representative of experiments repeated at least three times. Data was represented as mean ± SD. ***P < 0.001. E, F Correlation between PHF5A expression level and overall survival/disease-free survival (months) in patients with GC

Fig. 2

Knockdown down of PHF5A acts as a tumor suppressor in GC cells. A The effect of reduced PHF5A expression on the proliferation of AGS and MGC-803 cells was estimated by Celigo cells counting assay. B Flow cytometry was used to estimate the effect of reduced PHF5A expression on the apoptosis ability of AGS and MGC-803 cells. C, D Transwell and wound-healing experiments were performed to evaluate the effect of reduced PHF5A expression on AGS and MGC-803 cell migration. E Effect of PHF5A knockdown on GC stem cells was evaluated by Sphere-forming assay. The presented results were representative of experiments repeated at least three times. Data was represented as mean ± SD. **P < 0.01, ***P < 0.001

Fig. 3

PHF5A regulates the protein stability of FOS through SKP2-mediated ubiquitination. A Prime View Human Gene Expression Array was performed to identify DEGs between shCtrl and shPHF5A groups of MGC-803 cells. In the heat map of cluster analysis, each column represents a sample and each row represents a differential gene. The red indicates that the gene expression is relatively up-regulated, the green indicates that the gene expression is relatively down-regulated, the black indicates that the gene expression is not significantly changed, and the gray indicates that the signal strength of the gene is not detected. B, C The most significant DEGs in MGC-803 cells with PHF5A knockdown were screened by qRT-PCR and western blot analysis. D The effect of FOS, PTEN, PRKCZ knockdown on proliferation was detected using Celigo cell counting assay. E The E3 ubiquitin ligase affecting FOS protein stability was predicted by bioinformatics analysis. F, G The protein stability of FOS in MGC-803 cells after PHF5A knockdown and SKP2 overexpression was examined. H, I After treatment with MG-132, the protein stability of FOS in MGC-803 cells after PHF5A knockdown and SKP2 overexpression was examined. J, K The lysates of MGC-803 cells after PHF5A knockdown and SKP2 overexpression were immunoprecipitated and WB was performed to examine the ubiquitination of FOS. L Co-IP analysis of interaction of PHF5A and SKP2 in MGC-803 cells. The presented results were representative of experiments repeated at least three times. Data was represented as mean ± SD. ***P < 0.001

Fig. 4

Downregulation of FOS attenuates the promotion of PHF5A overexpression on GC cells. A FOS was highly expressed in tumor tissues compared with adjacent normal tissues in GC patients. (The magnification is 200 × and 400 ×). B Cell proliferation in MGC-803 was measured using Celigo cell counting assay. Results were normalized to viability on day 1 and represented as fold change. C The clone number of MGC-803 cells was estimated by colony formation assay. D Apoptosis ability of MGC-803 cells was analyzed using Annexin V-APC staining, followed by flow cytometry. E, F Migration of MGC-803 cells was determined by Wound healing assay and Transwell assay. G Effect of PHF5A knockdown on MGC-803 cells stem was evaluated by Sphere-forming assay. The presented results were representative of experiments repeated at least three times. Data was represented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 5

Pladienolide B treatment reverses the induction of PHF5A overexpression on the progression of GC cells. A–D PHF5A overexpressed cells were treated with Pladienolide B to detect changes in cell viability, apoptosis, and migration. Notably, for Figure D, H is the migration time of the cell. The presented results were representative of experiments repeated at least three times. Data was represented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. E–G PHF5A overexpression and control MGC-803 cells were subcutaneous injected into nude mice followed by intravenous Pladienolide B treatment to observed tumor volume, tumor size and immunohistochemical staining. Data was represented as mean ± SD (n = 5). *P < 0.05, **P < 0.01, ***P < 0.001