A novel peptide PDHK1-241aa encoded by circPDHK1 promotes ccRCC progression via interacting with PPP1CA to inhibit AKT dephosphorylation and activate the AKT-mTOR signaling pathway

Abstract

Background: Clear cell renal cell carcinoma (ccRCC) is the most prevalent kidney cancer with high aggressive phenotype and poor prognosis. Accumulating evidence suggests that circRNAs have been identified as pivotal mediators in cancers. However, the role of circRNAs in ccRCC progression remains elusive.

Methods: The differentially expressed circRNAs in 4 paired human ccRCC and adjacent noncancerous tissues ccRCC were screened using circRNA microarrays and the candidate target was selected based on circRNA expression level using weighted gene correlation network analysis (WGCNA) and the gene expression omnibus (GEO) database. CircPDHK1 expression in ccRCC and adjacent noncancerous tissues (n = 148) were evaluated along with clinically relevant information. RT-qPCR, RNase R digestion, and actinomycin D (ActD) stability test were conducted to identify the characteristics of circPDHK1. The subcellular distribution of circPDHK1 was analyzed by subcellular fractionation assay and fluorescence in situ hybridization (FISH). Immunoprecipitation-mass spectrometry (IP-MS) and immunofluorescence (IF) were employed to evaluate the protein-coding ability of circPDHK1. ccRCC cells were transfected with siRNAs, plasmids or lentivirus approach, and cell proliferation, migration and invasion, as well as tumorigenesis and metastasis in nude mice were assessed to clarify the functional roles of circPDHK1 and its encoded peptide PDHK1-241aa. RNA-sequencing, western blot analysis, immunoprecipitation (IP) and chromatin immunoprecipitation (ChIP) assays were further employed to identify the underlying mechanisms regulated by PDHK1-241aa.

Results: CircPDHK1 was upregulated in ccRCC tissues and closely related to WHO/ISUP stage, T stage, distant metastasis, VHL mutation and Ki-67 levels. CircPDHK1 had a functional internal ribosome entry site (IRES) and encoded a novel peptide PDHK1-241aa. Functionally, we confirmed that PDHK1-241aa and not the circPDHK1 promoted the proliferation, migration and invasion of ccRCC. Mechanistically, circPDHK1 was activated by HIF-2A at the transcriptional level. PDHK1-241aa was upregulated and interacted with PPP1CA, causing the relocation of PPP1CA to the nucleus. This thereby inhibited AKT dephosphorylation and activated the AKT-mTOR signaling pathway.

Conclusions: Our data indicated that circPDHK1-encoded PDHK1-241aa promotes ccRCC progression by interacting with PPP1CA to inhibit AKT dephosphorylation. This study provides novel insights into the multiplicity of circRNAs and highlights the potential use of circPDHK1 or PDHK1-241aa as a therapeutic target for ccRCC.

Keywords: CircPDHK1; Clear cell renal cell carcinoma; Dephosphorylation; Novel peptide; PPP1CA.

Fig. 1

Identification and characteristics of upregulated circPDHK1 in ccRCC. A The workflow for circRNA screening in 4 paired ccRCC tissues as indicated. B Cluster heatmap showed the differentially expressed 233 circRNAs in the ME blue module with WGCNA generated from 4 ccRCC tissues and paired adjacent normal tissues. The red and blue strips indicate up-regulated and down-regulated circRNAs, respectively. C Relative normalized signal detected by circRNA probe in microarray analysis of the 9 candidate circRNAs that intersected with the 233 circRNAs in the GSE100186 data. D RT-qPCR analysis of circPDHK1 expression in 148 paired ccRCC samples and normal adjacent tissues. E Comparison of circPDHK1 expression between patients with WHO/ISUP stage III–IV (n = 51) and those with WHO/ISUP stage I–II (n = 97), detected by RT-qPCR. F Comparison of circPDHK1 expression between patients with tumor T stage T1a-T1b (n = 103) and those with tumor T stage T2a-T4 (n = 45), detected by RT-qPCR. G Association analyses of circPDHK1 expression and distant tumor metastasis. H Relationships between circPDHK1 expression and VHL mutations in ccRCC paired tissues with genetic test reports (n = 106). I Relationships between circPDHK1 expression and Ki-67 in ccRCC paired tissues (n = 112). J Illustration of circPDHK1 formation by splicing exons 2 and 8 from the PDHK1 parental gene as identified in CircBase (http://circrna.org). Specific divergent and convergent PCR primers to detect circPDHK1 and linear PDHK1 are indicated, respectively. K PCR products of circPDHK1 in cDNA and gDNA amplified using convergent or divergent primers in Caki-1 and 786-O. L Schematic illustration of circPDHK1 conformation. The exon 2–8 of PDHK1 mRNA formed PDHK1 through back splicing. As shown, Sanger sequencing was used to verify the back-splice junction site of circPDHK1. M Reverse transcription using oligo dT and random priming to identify circPDHK1 loop characteristics by RT-qPCR. N CircPDHK1 and linearPDHK1 mRNA stability was assessed using RNase R treatment. O RT-qPCR analysis of the abundance of circPDHK1 and PDHK1 linear mRNA in ccRCC cells treated with actinomycin D at the indicated time points. P RT-qPCR analysis of circPDHK1 location in the nucleus or cytoplasm in Caki-1 and 786-O cells. GADPH served as a marker of cytoplasmic location, while U6 served as a marker of nuclear location. Q Fluorescence in situ hybridization (FISH) was utilized to detect circPDHK1 localization in Caki-1 and 786-O cell lines. Bars = 20 μm. *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance

Fig. 2

Inhibition of circPDHK1 suppresses the proliferation, migration and invasion of ccRCC cells in vitro and in vivo. A Colony formation assays, (B) CCK8 assays and (C, D) EdU assays were performed to detect the cell viability and proliferation activity of Caki-1 and 786-O cells transfected with two circPDHK1 siRNA. Bars = 200 μm. E–G Representative images and quantification of the transwell migration and invasion assays in circPDHK1 knockdown, and control ccRCC cells. Bars = 100 μm. H-I Representative images and quantification of the transwell migration and invasion assays in circPDHK1 knockdown, and control ccRCC cells. Bars = 100 μm. J Subcutaneous 786-O tumors in NSG mice injected with cholesterol-modified circPDHK1 siRNA, and tumor volumes were calculated every 2 d. At Day 37 after treatment, all mice were sacrificed, and (K) subcutaneous tumors were dissected and recorded. L Subcutaneous tumor was also photographed. (n = 6, each group). M Representative photomicrographs of Ki-67 immunohistochemical (IHC) staining of subcutaneous tumors. N We also established a lung metastasis model by injecting 6 × 10⁶ Caki-1 cells into the tail vein to evaluate the ability of cells to metastasize in vivo. Representative in vivo imaging of mice, photographs of whole lung tissues and hematoxylin–eosin (H&E) staining of lung metastatic nodules. (n = 3, each group) Bars = 200 μm. *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance

Fig. 3

CircPDHK1 encodes a novel peptide, PDHK1-241aa. A The predicted 815 nt circPDHK1 sequence encodes 241 amino acids (PDHK1-241aa). Potential m6A sites and IRES (Internal Ribosome Entry Site) are also indicated. B IRES constructs in luciferase expression vectors. Full-length IRES, IRES del-1, IRES del-2 and mutated IRES types inserted in Rluc and hLuc reporter vectors containing independent start and stop codons. C Relative luciferase activity detected using dual-luciferase reporter assays to tested circPDHK1 translation ability in 293 T cells. D Illustration of the synthetic circRNA expression plasmid pLC5-ciR, a commercial circRNA blank expression vector; We also established circPDHK1-3 × flag plasmid for pLC5-circPDHK1 fused with 3 × flag tag and circPDHK1-3 × flag ATG mut, the start codon ATG of circPDHK1-3 × flag was mutated to CTG with a commercial kit to validate the coding potential of circPDHK1. E Sanger sequencing of the circPDHK1-3 × flag ATG mut vector to verify its construction. F-G RT‒qPCR and western blot were used to analyze the efficiency of overexpressing circPDHK1-3 × flag and circPDHK1-3 × flag ATG mut vectors to verified circPDHK1 translation ability in 293 T cells. H Immunofluorescence staining using a Flag antibody to detect the intracellular localization of circPDHK1-3 × flag in Caki-1 and 786-O cells. The red (anti-flag) indicated the circPDHK1-3 × flag, the blue (Hochest) indicated the nucleus. Bars = 20 μm. I PDHK1-241aa expression detected by western blot following SDS‒PAGE of IP pull-downs from cell lysates as indicated. Gel bands corresponding to 31 kDa were manually removed and submitted to LC‒MS for identification. J Identification of the amino acid sequence of PDHK1-241 by LC‒MS. K Western blot analysis of PDHK1-241aa expression in 16 paired ccRCC samples and normal adjacent tissues. **P < 0.01; ***P < 0.001

Fig. 4

PDHK1-241aa, but not circPDHK1, promotes the proliferation, migration and invasion of ccRCC cells in vitro and in vivo. A Colony formation assays, B CCK-8 assays and (C, D) EdU assays were performed to detect the cell viability and proliferation activity of Caki-1 and 786-O cells transfected with pLC5-ciR, circPDHK1 and circPDHK1 ATG mut vectors. Bars = 200 μm. E–G Transwell assay to detect cell migration and invasion of Caki-1 and 786-O cells transfected with pLC5-ciR, circPDHK1 and circPDHK1 ATG mut vectors. Bars = 100 μm. H, I wound healing assays were conducted to evaluate cell migratory abilities in Caki-1 and 786-O cells tansfected with pLC5-ciR, circPDHK1 and circPDHK1 ATG mut vectors. Bars = 100 μm. J Xenograft 786-O tumors in NSG mice were injected with circPDHK1 and circPDHK1 ATG mut overexpression lentiviruses. Tumor volumes were calculated every 2 d. At Day 18 after treatment, all mice were sacrificed, and (K) subcutaneous tumors were dissected and recorded. L Subcutaneous tumor was also photographed. (n = 5, each group). M Representative photographs of Ki-67 IHC staining in subcutaneous tumors. N We also established a lung metastasis model by injecting 6 × 10⁶ Caki-1 cells into the tail vein to evaluate the ability of cells to metastasize in vivo. Representative in vivo imaging of mice, photographs of whole lung tissues and hematoxylin–eosin (H&E) staining of lung metastatic nodules. (n = 3, each group) Bars = 200 μm. *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance

Fig. 5

PDHK1-241aa interacts with PPP1CA to inhibit AKT dephosphorylation and activate the AKT-mTOR signaling pathway. A Volcano plot of differentially expressed protein-coding genes following circPDHK1 interference and overexpression. B The RNA-seq analysis showed that relevant KEGG enrichment pathways following circPDHK1 interference and overexpression in Caki-1 cells. C GSEA results plotted to visualize correlations between circPDHK1 and AKT-mTOR pathway expression in Caki-1 cells. Western blot detection of proteins related to the AKT-mTOR pathway in (D) 8 pairs of ccRCC tissues and (E) Caki-1 and 786-O cells following circPDHK1 interference and overexpression as indicated. F Coimmunoprecipitation (co-IP) assays were used to verify the interaction between PDHK1-241aa and AKT or p-AKT (Thr308). G Total proteins of PDHK1-241aa-flag-transfected HEK293T cells separated by SDS‒PAGE. 9 interacting proteins of circPDHK1 were identified by LC‒MS and listed with their corresponding scores. The LC‒MS analysis showed that identification of PPP1CA in the PDHK1-241aa protein complex. H Co-IP used to verify binding between PPP1CA and PDHK1-241aa. I Immunofluorescence localization of PDHK1-241aa-flag and PPP1CA from overexpression vectors in Caki-1 and 786-O cells. The green (anti-flag) indicated the circPDHK1-3 × flag; The red(anti-PPP1CA) indicated the PPP1CA; The blue (hochest) indicated the nucleus. Bars = 20 μm. J Variations in PPP1CA and PDHK1-241aa localization verified by nuclear and cytoplasmic separation assays. K NCBI database (https://www.ncbi.nlm.nih.gov/cdd/) predicted that the PDHK1 domain is composed of BCDHK domain and HATPase domain, respectively. L Schematic illustrations of PDHK1 expression plasmids. M We also established truncation mutants BCDHK domain 3 × flag and HATPase domain 3 × flag plasmid. Immunoblot analysis following cotransfection with PDHK1-241aa and the indicated truncation mutants. N The specific binding between PDHK1-241aa and PPP1CA was predicted by PyMOL (https://pymol.org/2/) and HDOCK (http://hdock.phys.hust.edu.cn/). *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance

Fig. 6

Phenotypes of Caki-1 and 786-O cells cotransfected with siRNAs for circPDHK1 and PPP1CA. A Colony formation assays, (B) CCK-8 assays and (C, D) EdU assays were performed to detect the cell viability and proliferation activity of Caki-1 and 786-O cells transfected with circPDHK1 siRNA and/or PPP1CA siRNA. Bars = 200 μm. E–G Transwell migration and invasion assays to detect cell migration ability of Caki-1 and 786-O cells transfected with the control siRNA, a circPDHK1 siRNA, a PPP1CA knockdown siRNA and cotransfected with the circPDHK1 siRNA and the PPP1CA knockdown siRNA. Bars = 100 μm. H Western blot detection of the phosphorylation levels of the AKT-mTOR signaling pathway in Caki-1 and 786 cells. I Analysis of subcutaneous tumors in mice injected with cholesterol-modified the control siRNA, a circPDHK1 siRNA and cotransfected with the circPDHK1 siRNA and the PPP1CA knockdown siRNA. Tumor volumes calculated every 2 d. J At Day 37 after treatment, all mice were sacrificed, and subcutaneous tumors were dissected and recorded. Tumor weights in each group as indicated. K Images of subcutaneous tumors. (n = 6, each group). L, M Representative photographs and quantification of Ki-67 IHC staining in subcutaneous tumors. *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance

Fig. 7

Biogenesis of circPDHK1 is activated by HIF-2A. A HIF-2A expression data in KIRC samples and normal adjacent tissues from TGCA. B, C RT-qPCR to detected relative HIF-2A expression in 106 paired ccRCC tissues and normal adjacent tissues. D Comparison of HIF-2A expression between patients with WHO/ISUP stage III–IV (n = 38) and those with WHO/ISUP stage I–II (n = 68), detected by RT-qPCR. E Comparison of HIF-2A expression between patients with tumor T stage T1a-T1b (n = 71) and those with tumor T stage T2a-T4 (n = 35), detected by RT-qPCR. F Association analyses of HIF-2A expression and distant tumor metastasis. G, H Relationships between HIF-2A expression and VHL mutation (n = 106) and Ki-67 expression (n = 93). I RT-qPCR validation of correlation between circPDHK1 and HIF-2A expression in ccRCC paired tissues (n = 106). J, K Regulatory relationships between circPDHK1 and HIF-2A using RT‒qPCR detection and dual-luciferase assays in Caki-1 and 786-O cells. L Schematic of the HIF-2A binding motif and the HREs in the PDHK1 promoter obtained from the Contra V3 database (http://bioit2.irc.ugent.be/contra/v3) as indicated. We also designed three primers for the different binding sites in PDHK1 promoter. M, N ChIP assay to assess HIF-2A binding site interactions in the PDHK1 promoter region. *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance

Fig. 8

Cotransfection analysis of circPDHK1 siRNA and HIF-2A overexpression vector in Caki-1 and 786-O cells. A Colony formation assays, (B) CCK-8 assays and (C, D) EdU assays were performed to detect the cell viability and proliferation activity of Caki-1 and 786-O cells transfected with circPDHK1 siRNA and/or HIF-2A overexpression vector. Bars = 200 μm (E–G) Transwell migration and invasion assays to detect cell migration ability of Caki-1 and 786-O cells cotransfected with the control siRNA and pCDNA3.1 vector, a circPDHK1 siRNA and pCDNA3.1 vector, the control siRNA and HIF-2A overexpressed vector, the circPDHK1 siRNA and the HIF-2A overexpressed vector. Bars = 100 μm. H Graphical depiction of circPDHK1 encoding PDHK1-241aa, which binds to PPP1CA to inhibit AKT dephosphorylation and activate the AKT-mTOR signaling pathway, thereby promoting the proliferation and metastasis of ccRCC. The schematic was drawn by Figdraw (https://www.figdraw.com). *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance