Polymorphism rs2395655 affects LEDGF/p75 binding activity and p21WAF1/CIP1 gene expression in esophageal squamous cell carcinoma

Abstract

p21WAF1/CIP1 (p21) plays critical roles in cell-cycle regulation and DNA repair and is transcriptionally regulated through p53-dependent or -independent pathways. Bioinformatic analysis predicated one stress-response element (STRE) implicated in single nucleotide polymorphism (SNP) rs2395655 of the p21 promoter. Here, we investigated the transcriptional regulatory function of rs2395655 variant genotype and analyzed its associations with the p21 expression and clinical outcomes in esophageal squamous cell carcinoma (ESCC) patients. Luciferase assay results showed significantly increased transcriptional activity of the rs2395655 G allele-containing p21 promoter compared with rs2395655 A allele-containing counterpart, especially in ESCC cells with ectopic LEDGF/p75 expression. Furthermore electrophoretic mobility shift assay using the rs2395655 G or A allele-containing probe and chromatin immunoprecipitation assay with specific anti-LEDGF/p75 antibody indicated the potential binding activity of LEDGF/p75 with the STRE element implicated in rs2395655 G allele of the p21 promoter. Subsequent specific RNA interference-mediated depletion or ectopic expression of LEDGF/p75 caused obviously down- or up-regulated expression of p21 mRNA in ESCC cells harboring rs2395655 GG genotype but not cells with rs2395655 AA genotype. Furthermore, rs2395655 GG genotype carriers showed significantly elevated p21 protein expression and conferred survival advantage in both univariate and multivariate analyses in total 218 ESCC patients. Our findings suggest that LEDGF/p75 regulates the p21 expression in ESCC cells through interacting with STRE element implicated in polymorphism rs2395655 and the elevated p21 protein expression and rs2395655 GG genotype may serve as positive prognostic factors for ESCC patients.

Keywords: esophageal squamous cell carcinoma; lens epithelium-derived growth factor; p21WAF1/CIP1; prognostic factor; single-nucleotide polymorphism.

Figure 1

Reporter gene assay with constructs containing the pivotal region of the p21 promoter. (A) Schematic of four reporter gene constructs encompassing nucleotides −2308 to +204 of the p21 promoter, with sequences identical except for rs3829963 (–2119C/A) and rs2395655 (–809G/A) polymorphic sites. Luciferase expression in the four constructs in ESCC cells EC‐109 (B) and KYSE150 (C) were standardized by cotransfection with pRL‐SV40. LEDGF/p75 expression vector was cotransfected to evaluate its potential effect on the activity of the p21 promoter, using pcDNA3.0 empty vector as control. Fold increase was measured by defining the activity of the pGL3‐Basic vector and pcDNA3.0 empty vector as 1. The value of each reported construct was calculated as the mean fold increase ± SD from three independent experiments each performed in duplicate (top panel). LEDGF/p75 (p75 in brief) protein levels in EC‐109 and KYSE150 cells were examined by Western blot analysis (bottom panel). *P < 0.05 and **P < 0.01 vs rs2395655 A allele‐containing counterparts, respectively

Figure 2

LEDGF/p75 interacted with human genomic DNA sequence containing rs2395655 G allele. (A) The rs2395655 A>G transition created one cis‐regulatory element, that is STRE element (AGGGGT, nucleotides −810 to −805), in the p21 promoter. Polymorphic sites of rs2395655 were italicized. (B) Nuclear extracts from KYSE150 cells were used for electrophoretic mobility shift assay (EMSA) with two biotin‐labeled probes carrying rs2395655 G and A allele, respectively. Anti‐LEDGF/p75 antibody (C‐16, 2 mg), IgG control antibody or a 400‐fold molar excess of cold probe was added to the binding reaction to examine the specificity of the protein‐DNA complex formation. The black and hollow arrows indicated the shift and supershift bands, respectively. (C) Chromatin immunoprecipitation was performed using anti‐LEDGF/p75 antibody (C‐16) and genomic DNA mixture of EC‐109 and KYSE410 cells, carrying rs2395655 GG and AA genotype, respectively. Human p21 region of interest was amplified and sequenced. Human GAPDH and FBXO10 loci were included as respective negative and positive LEDGF/p75 binding controls

Figure 3

LEDGF/p75 regulated the expression of p21 gene in ESCC cells with rs2395655 GG genotype. (A) Specific small interference RNAs (si‐p75‐1 and si‐p75‐2)‐mediated depletion of LEDGF/p75 (p75 in brief) induced obviously down‐regulated expression of p21 mRNA in EC‐109 and KYSE150 cells carrying rs2395655 GG genotype but not KYSE410 and KYSE450 cells with rs2395655 AA genotype. (B) Ectopic expression of p75 apparently increased the expression level of p21 mRNA in EC‐109 and KYSE150 cells, while no change of the p21 expression level was detected in KYSE410 and KYSE450 cells. For specific siRNA transfection and ectopic expression of p75, the mismatched siRNA (mock) and empty vector transfections were used as control, respectively. The relative signal intensity of each target gene was quantified and normalized to internal control, β‐actin. The graph indicates the relative mRNA expression levels of p75 and p21 and the values are mean ± SD for three independent experiments. M, DNA marker. Neg, negative control. *P < 0.01 and **P > 0.05 vs controls, respectively

Figure 4

Immunohistochemistry staining of the p21 protein in ESCC tissues. No‐anti p21, negative control with primary antibody replaced by PBS. Anti‐p21, the brown signals represented positive staining for p21 protein and the staining was scored on a scale as indicated in the Materials and Methods. Positive cases were defined as those with moderate to strong p21 nuclear staining in ≥ 10% of tumor cells. Magnification, × 100 (top panel) and × 400 (bottom panel), respectively

Figure 5

Kaplan–Meier survival estimation of 218 ESCC patients related to possible predictors. (A) Elevated p21 protein expression indicated better postoperative outcome for ESCC patients (32.0 and 19.0 months of median survival for p21‐positive and negative patients, respectively, P = 0.001). (B) The median survival time of ESCC patients with rs2395655 GG genotype was significantly increased compared with rs2395655 AA or AG genotype (30.0 vs 20.0 months, P = 0.003)