E2F1 and TFDP1 Regulate PITX1 Expression in Normal and Osteoarthritic Articular Chondrocytes

Abstract

We previously reported a loss-of-PITX1 expression in patients suffering of knee/hip osteoarthritis (OA). Search for the mechanism underlying this event led us to discover that PITX1 repression was triggered by the aberrant nuclear accumulation of Prohibitin (PHB1), an E2F1 co-repressor, in OA articular chondrocytes. In the current study, we assessed in details the involvement of E2F transcription factors in regulating PITX1 expression. We also analyzed other genes that are similarly regulated by E2F in regard to osteoarthritis. The transcriptional regulation of the PITX1 promoter by E2F1 was analyzed with the luciferase reporter assay, and chromatin immunoprecipitation assays, which confirmed direct E2F1-PITX1 interactions. The probable binding sites for E2F1 in the PITX1 promoter were identified by DNA pulldown experiments. In silico and in vitro analyses show that the PITX1 proximal promoter region contains 2 specific sequences that are bound by E2F1. Overexpression of E2F1 enhances PITX1 promoter activity and mRNA transcription. In primary control and osteoarthritis chondrocytes, real time RT-PCR was used to measure the mRNA expression levels of candidate genes under E2F1 transcriptional control. Transcription Factor Dp-1 (TFDP1) knockdown experiments confirmed that the E2F1-TFDP1 complex regulates PITX1. Knockdown of TFDP1, an E2F1 dimerization partner, inhibits the activating effect of E2F1 and reduces both PITX1 promoter activity and mRNA transcription. Real time RT-PCR results reveal reduced expression of TFDP1 and a similar downregulation of their targets PITX1, BRCA1, CDKN1A, and RAD51 in mid-stage OA chondrocytes. Collectively, our data define a previously uncharacterized role for E2F1 and TFDP1 in the transcriptional regulation of PITX1 in articular chondrocytes. Additional E2F1 targets may be affected in OA pathogenesis.

Fig 1. Critical regulatory regions in the PITX1 promoter govern its expression in chondrocytes.

Fig 2. E2F1 directly regulates PITX1 gene expression in the C28/I2 cell line and in primary chondrocytes.

(A) Real time RT-PCR performed on the PITX1 gene in C28/I2 chondrocytes stably expressing pBabe-ER-E2F1 fusion plasmids or the empty control vector, treated with OHT or vehicle (100% ethanol) 24 hours prior to RNA extraction. Data represent mean ± standard deviation of 3 independent experiments. Asterisks represent a significant increase in OHT-treated cells compared with vehicle-treated cells (Two-way ANOVA; Bonferroni post hoc: *p < 0.0001). (2B-2C) Chromatin immunoprecipitation (ChIP) assays of E2F1 on PITX1 and TK promoters in (B) C28/I2 chondrocytes and (C) primary chondrocytes. Data are presented as relative quantification of DNA following immunoprecipitation when compared with the negative control. Error bars represent standard deviation of 3 independent experiments (Two-way ANOVA; Bonferroni post hoc: *p < 0.05, **p < 0.0001).

Fig 3. Point mutation analyses in the C28/I2 chondrocyte PITX1 promoter confirm specific E2F1 binding sites.

Biotinylated double-stranded probes representing different wild type or mutated regions of the PITX1 promoter were used in DNA pulldown assays. A nonspecific 30 bp biotinylated double-stranded probe was used as a negative control. In the representative immunoblots, lamin A/C was used as a loading control for the nuclear extract. (A) Different 30 bp probes corresponding to distinct PITX1 proximal promoter regions were used as baits to detect specific binding with E2F1, E2F2, or E2F3. (B) Fine mapping of E2F1 binding the -180 bp probe: m1, m2, m3 (point mutations). (C) Fine mapping of E2F1 binding the -115 bp probe: m4, m5, m6, m7 (point mutations). (D) Sequence of the -225/+75 bp PITX1 promoter region according to the transcription start site (+1). Predicted E2F1 response elements by MatInspector 8.0 software from the Genomatix Software Suite are highlighted in grey. The -180 bp and -115 bp probes are underlined with a dark and a dashed line, respectively. Asterisks represent the nucleotides that are essential for the binding of E2F1 to the PITX1 proximal promoter.

Fig 4. TFDP1 knockdown affects PITX1 gene regulation.

(A) Immunoblots of TFDP1 and PITX1 were performed in C28/I2 cells transfected with a nonspecific siRNA (siCTRL) or increasing amounts of TFDP1 siRNA (siTFDP1). α-TUBULIN was used as a loading control for immunoblotting. The percentage values of the relative expression of TFDP1 and PITX1 are indicated. (B) Luciferase assays were performed in C28/I2 cells transiently transfected with different constructs of the PITX1 promoter. The cells were co-transfected with either the pBabe plasmid expressing ER fused to E2F1 or the empty control vector along with either control siRNA or TFDP1 siRNA, and were induced with 4OH-tamoxifen (OHT) for 24 hours. The data represent mean ± standard deviation of 3 independent experiments. Asterisks represent a significant decrease in the luciferase activity (Two-way ANOVA; Bonferroni post hoc: *p < 0.01, **p < 0.001, ***p < 0.0001) compared with control cells. (C) Real time RT-PCR analysis of PITX1 mRNA in chondrocytes from healthy subjects transfected with either control siRNA or TFDP1 siRNA. Data represent mean ± standard deviation of 3 independent experiments. Asterisks represent a significant decrease in PITX1 levels in siTFDP1-transfected cells compared with control (Student's t-test: *p < 0.02).

Fig 5. Gene expression levels of E2F family members and selected targets in primary articular chondrocytes.

RNA was extracted from primary cultures of chondrocytes derived from the knees of 4 controls and 18 OA patients. The control subjects (CTRL) were attributed Kellgren-Lawrence (KL) scores of 0 (n = 3) and 1 (n = 1) while OA patients accounted for KL scores of 2 (n = 6), 3 (n = 5), and 4 (n = 7). Since chondrocyte proliferation may influence E2F-related gene expression levels, samples with KL 2–3 scores (mid-stage OA characterized with increased proliferation) were separated from KL 4 (end-stage OA). (A) Real time RT-PCR analysis of the E2F family members E2F1, E2F2, E2F3, and TFDP1. (B) Real time RT-PCR analysis of the E2F targets PITX1, BRCA1, CDKN1A, and RAD51. (A-B) Y-axis: relative expression (fold changes) compared with the lowest score of 1. (One-way ANOVA followed by Bonferroni post hoc: only significant p < 0.05 indicated on graphs).