Rad51 mRNA and protein levels are increased in cancer cells. (A) Diagram of the human Rad51 gene. Transcription start site is indicated by arrow. Coding exons are represented by solid black boxes. Upstream to the start of transcription is the Rad51 regulatory region. (B) Analysis of Rad51 transcript levels in normal and cancerous cells by quantitative RT-PCR. The top bands are RT-PCR products of Rad51 mRNA and the bottom bands are RT-PCR products of 18S subunit ribosomal RNA used as a reference. The histogram represents the relative intensity of the Rad51 band normalized to the 18S band. (C) Analysis of alternatively spliced Rad51 transcript levels by quantitative RT-PCR. Same PCR primers are used as in B but the number of the PCR cycles is increased and the gel is overexposed to visualize the less abundant alternative splice variant of Rad51. The top band is full length (not quantified because of saturation), and the band directly below it is the alternative slice variant. The histogram represents the relative intensity of the Rad51 splice variant band with the 18S band used as a reference. (D) Western blot analysis of Rad51 protein. Protein levels for each cell line were normalized using α-tubulin as a loading control and are displayed in the histogram above the gels. All of the experiments were repeated three times and error bars are SD.

The construct containing Rad51 promoter fused to diphtheria toxin A (DTA) gene selectively kills cancer cells. (A) Diagram of the pRad51-DTA construct. Transcription start site is indicated by arrow. (B) Decline in cell counts after transfection with pRad51-DTA. Cells were cotransfected with the indicated amounts of pRad51-DTA and promoterless plasmid pGL3 to bring the total amount of DNA in each transfection to 0.1 μg. Attached cells were harvested 72 h posttransfection and counted using a Beckman Coulter counter. For each cell line, the cell counts obtained after transfection with pRad51-DTA were divided by cell counts in the control transfections with 0 μg of pRad51-DTA and normalized for the efficiency of transfection. The experiments were repeated three times and error bars show SD. (C) Inhibition of protein synthesis by pRad51-DTA. In this experiment inhibition of SV40-luciferase expression is used as a model of reduction in de novo protein synthesis by DTA. Cells were transfected with pRad51-DTA and pGL3 as described above along with 1 μg of a plasmid encoding firefly luciferase under SV40 promoter. Cells were harvested 72 h posttransfection and subjected to luciferase assay. For each cell line, luciferase activity after transfection with pRad51-DTA was divided by luciferase activity in control transfections with 0 μg of pRad51-DTA. The experiments were repeated three times and error bars show SD.

Rad51 promoter fused to luciferase gene shows dramatic difference in promoter activity between normal and cancer cells. (A) Diagram of the pRad51-Luc construct with the firefly luciferase gene under control of Rad51 promoter. Transcription start site is indicated by arrow. (B) Luciferase assays measuring Rad51 promoter activity in 13 cell lines. Cells were transfected with 2 μg of pRad51-Luc and luciferase activity was analyzed in cell extracts 72 h posttransfection. Luciferase activity was normalized for the efficiency of transfection determined by transfection with GFP-expressing plasmid, shown in C. The values for luciferase activity for all of the cell lines are provided in Table S1. The experiments were repeated three times and error bars show SD. (C) Transfection efficiency in 13 cell lines. In parallel with pRad51 luciferase transfections shown in B, cells were transfected with 2 μg of the GFP-expressing plasmid pEGFP-N1 and analyzed by flow cytometry 72 h posttransfection. The parameters for FACS analysis were set so as to detect all cells with green fluorescence above the background. This ensures that all transfected cells are scored regardless of the differences in expression in different cell lines. Efficiency of transfection is expressed as the percentage of GFP+ cells. The experiments were repeated three times and error bars show SD.