Rad51C-ATXN7 fusion gene expression in colorectal tumors

To identify the fusion gene in colorectal tumors, Rad51C exons 1–7 (accession number NM_058216) and exons 6–13 of ATXN7 (accession number NM_000333) mRNA sequences were joined. According to the sequence, RT-PCR specific primers were designed with forward primer spanning exon-5 of Rad51C and reverse primer located in exon-8 of ATXN7. Total RNA was isolated from colorectal tumors and non-tumors following the protocol from TRIzol as mentioned in materials and methods. The cDNA containing exons 5–7 of Rad51C and exons 6-8 of ATXN7 was amplified by one step RT-PCR using primers (LF-F and LF-R; Additional file 1: Table S1) that produced either two amplicons with size of 376 bp, and 316 bp as shown in Fig. 1a, or one product with size of 376 bp (Fig. 1b). The RT-PCR products were gel purified and subjected to sequencing using the primers (LF-F and LF-R). Sequencing analysis showed that the 376 bp product, which we named Variant 1 of the fusion gene, contained Rad51C (exon 5–7) and ATXN7 (exon- 6–8) (Fig. 1c). The 316 bp fragment, named Variant 2, contained exons 5–6 of Rad51C and exons 6–8 of ATXN7 (Fig. 1d). In silico analysis showed that the Variant 1 extends for only three amino acids (aa) after the fusion junction, and results in a truncated protein and the Variant 2 forms an inframe fusion transcript with predicted molecular weight of protein 110 KDa (Additional file 2: Table S2). Therefore, we used the term of fusion gene to represent the Variant 2 which produces a fusion protein in the subsequent text. The structure of the gene was depicted in Fig. 2.

To identify the Rad51C-ATXN7 fusion gene in colorectal tumors, RNA was isolated from 67 tumors following TRIzol protocol. To amplify the fusion mRNA, we designed Variant specific primers according to the junction sequences between Rad51C and ATXN7 obtained previously. The primers (SJF and SJR) were listed in the Additional file 1: Table S1. Using the fusion specific RT-PCR (Fig. 3), we analysed 67 tumor samples. Of the 67 tumor RNA samples, 24 (36 %) of tumors showed the fusion mRNA (Table 1). Overall, the fusion RNA is expressed mainly in tumors (Fig. 3, Additional file 3: Table S3). We found six of the 24 tumours expressed both variants. In addition, we Topo®TA cloned 45 RT-PCR products amplified using primers LF-F and LF-R, 12(27 %) clones contained this fusion mRNA (Variant 2). In addition, Rad51C-ATXN7 fusion DNA was detected in LS-174 T cell and lung tumor genomic DNA samples by PCR analysis (Additional file 4: Figure S1).

To assess for functional somatic deficiency in the FA repair pathway, a triple staining immunofluorescence method (FATSI) was previously developed in our lab to identify the presence of FANCD2 repair foci in the nucleus of proliferating cells [22]. Based on FATSI, 20 of 67 (30 %) colorectal tumors examined were defective for FANCD2 foci formation and 47 (70 %) were FANCD2 foci positive (Table 1). We investigated the expression of Rad51C-ATXN7 fusion gene according to FANCD2 foci status. Of the 20 FANCD2 foci negative tumors, 12 (60 %) showed expression of the fusion gene. Among the 47 FANCD2 foci positive tumors, 12 (26 %) showed expression of the fusion gene (Table 1). Thus, the fusion gene expression is more associated with foci negative tumors (p = 0.012, fisher exact two tailed test).

Microsatellite status was obtained from the surgical pathology report based on standard Immuno-histochemistry (IHC) analysis evaluating expression of MSH2, MSH6, MLH1 and PMS2 proteins, in the colorectal tumors [23]. Of the 67 colorectal tumors, we had 54 colorectal tumors known for microsatellite status, 43 of which were microsatellite instable (MSS) and 11 were microsatellite instable (MSI). Among the 43 MSS tumors, 13 (30 %) tumors expressed the fusion gene. Of the 11 MSI tumors, fusion gene was in 6 (55 %) tumors. The fusion gene is more expressed in the MSI tumors. However, there is no significant association statistically between the fusion gene expression and status of microsatellite (p = 0.166, fisher exact two tailed test).

We have shown several hypermethylated CpG sites within the promoter of Rad51C [24] and treatment with hypomethylating agents causes demethylation at the three methylated CpG sites in the promoter region. We evaluated if the expression of the Rad51C-ATXN7 fusion gene could be altered by treatment with DNA hypomethylating agents. LS-174 T and RKO cells were treated with 5-Azacytidine at 5 μM for 72 h and the total RNA isolated from the cells. To quantitatively analyze the effect of promoter methylation on expression of Rad51C-ATXN7 fusion gene, RNA from 5-Azacytidine treated cells as well as pre-treatment control cells was analyzed with Taqman real time PCR. The real time PCR primers and respective amplicon sizes for Variant 2 are described in Additional file 1: Table S1. 18S ribosomal rRNA was used as endogenous control. The real time PCR analysis showed a 3.51 fold increase in the expression of the fusion gene, respectively, following treatment with 5-Azacytidine as compared to untreated controls (Fig. 4). RKO cells showed 1.2 fold mild increases in the expression of the fusion gene post treatment. Immunoprecipitation of total Rad51C protein from the post 5-Azacytdine treated LS-174 T cells showed higher expression of protein (data not shown) in comparison to the untreated control.

The Rad51C-ATXN7 fusion transcripts identified from the colorectal tumors were in silico translated using EMBOSS program [25]. The fusion gene (Variant 2) produced an inframe fusion transcript with predicted molecular weight of protein 110 KDa (Additional file 2: Table S2).

To confirm the existence of the fusion protein between Rad51C and ATXN7, immunoprecipitation of Rad51C protein was carried out using Dyna MagTM -2 beads from the total cell extracts prepared from the colon cancer cell lines LS-174 T and RKO. The total protein was immunoprecipitated using anti Rad51C N-terminus antibody and the pull down proteins were analysed with Western immunoblot using anti-ATXN7 antibody. As shown in Fig. 5, a protein with molecular weight of around 110 KDa was observed in the total and Rad51C antibody immunoprecipitated samples from LS-174 T cells, which was absent in the RKO cell line.

To evaluate the influence of the fusion gene in regarding to cell viability following exposure to cisplatin, we evaluated survival fractions for the cells that are expressing the fusion gene (LS-174 T) as compared with non-expressing cells (RKO) post treatment of cisplatin (5 μg/ml). MTT assay was used for the cell viability analysis and an averaged absorbance was recorded 24, 48, and 72 h post treatment. MTT data showed that the LS-174 T cells had 60 and 27.5 % of viable cells compared to non-treatment controls 48 and 72 h post treatment with cisplatin. In contrast, there were 80 and 46 % of viable cells in the RKO population 48 and 72 h post cisplatin treatment (Fig. 6). These results indicate LS-174 T cells are more sensitive to the treatment of cisplatin, a DNA interstrand crosslinking agent.

After Institutional Review Board (IRB) approval, fresh human tumor and paired non-tumor tissue samples were obtained from tissue procurement at The Ohio State University, tissues used for RNA and protein analysis were frozen in liquid nitrogen and stored in a –80 °C freezer. Tissues used for histological analysis were placed in 10 % neutral buffered formalin for 12 h, and then stored in PBS. Total RNA and DNA from the colorectal tumor tissue samples were isolated by homogenization and sequential precipitation following the protocol from TRIzol reagent (Life Technologies). RNA and DNA obtained from the colorectal tumor tissue samples was quantified at absorbance 260 nm and 280 nm using Nanodrop 2000C (Thermoscientific).

The forward and reverse primers were designed using the Primer3 software [28]. The forward and reverse primers for the Variant 2 Rad51C-ATXN7 fusion gene are shown in Additional file 1: Table S1.

The total RNA obtained from the colorectal cancer patients were reverse transcribed to cDNA following the Superscript™ II Reverse transcriptase protocol (Invitrogen, CA).

The optimal temperature was determined for each pair of primers to be around 60 °C. PCR reactions were performed in 25 μl reaction volume with 1x PCR buffer, 10 mM dNTP mixture, 50 mM MgCl_2, 0.2 μM primer of each forward and reverse, 0.2 μl 1.0 unit Platinum ® Taq DNA polymerase following the protocol from (Invitrogen, CA). PCR conditions for each amplicon included 95 °C for 4 min, 95 °C for 30 s, annealing temperature at 60 °C for 30 s and extension at 72 °C for 30 s. Post PCR, the amplicons were analyzed on 1.5 % agarose gel.

RT-PCR was performed using one step RT-PCR Kit (Qiagen, CA). The junction specific primers for Variant 2 are SJF, SJR and the sequences are shown in the Additional file 1: Table S1, along with their respective amplicon sizes. The RT-PCR was performed in 25 μl reaction containing final concentration (1x) of 5 μl of 5x Qiagen one step RT-PCR buffer, 400 μM of each dNTPs, 0.6 μM each of forward and reverse primers, 1 μl of Qiagen One step RT-PCR enzyme mix, template RNA up to 2 μg and the volume was made up to 25 μl using PCR grade water. The PCR was carried out as reverse transcription for 34 min at 52 °C, initial PCR activation for 15 min at 95 °C, 32 cycles of initial denaturation for 30 s at 95 °C, annealing for 30 s at 60 °C, extension for 30 s at 72 °C and final extension for 4 min at 72 °C. The amplified products were gel extracted and purified using QIAquick Gel Extraction Kit (Qiagen, CA). The purified products were subjected to direct sequencing using the primers (Additional file 1: Table S1).

For rapid and reliable identification of Rad51C-ATXN7 fusion gene, Topo TA cloning was performed following the protocol from (Invitrogen, CA). The obtained clones were sent for direct sequencing. To identify the Variant 2, the sequences obtained from TA clones were compared to sequences of Rad51C and ATXN7 mRNA sequences in Genbank.

The FA triple staining immunofluorescence method has been previously described [22]. Briefly, FFPE tumor tissue was cut at 4 microns, placed on positively charged slides, and stained with hematoxylin and eosin. Additional sections for immunofluorescence staining were placed in a 60 °C oven for one hour, cooled, deparaffinized, and rehydrated through xylenes and graded ethanol solutions to water in standard fashion. Antigen retrieval was performed by placing slides in Dako’s TRS antigen retrieval solution (Dako) in a calibrated vegetable steamer (Black and Decker) at 94 °C for 25 min. Slides then were placed on a Dako Autostainer for automated staining. The tissue sections were incubated with a primary antibody cocktail of rabbit polyclonal FANC-D2 antibody at a dilution of 1:1,000 and a monoclonal anti-Ki67 mouse antibody (Dako) at a dilution of 1:150, for one hour at room temperature. Sections then were incubated with a secondary antibody (FITC conjugated to anti-rabbit IgG and Alexa fluor 594 donkey anti-mouse IgG, Invitrogen) at 1:1,000 for one hour at room temperature. The sections were mounted on glass slides using a 4′ 6-diamidino-2-phenylindole (DAPI)-containing embedding medium (Vysis Dapi 1, Abbott Laboratories). The slides were analyzed under a Nikon E-400 fluorescence microscope [22].

Human colon cancer cell lines LS-174 T and RKO were purchased from ATCC® and were grown in complete Eagle’s Minimum Essential Medium with 10 % fetal bovine serum and 1 % penicillin/streptomycin at 37 °C (5 %, CO₂). For the treatment of cells, 5-Azacytidine was filter-sterilized and added directly to the fresh cell culture media. Fresh 5-Azacytidine was added every 24 h at dose of 5 μM for 72 h. The post treatment cells were lysed, and the pellets were collected and stored at −80 °C.

The RNA expression levels of the target fusion gene were analyzed by Taqman real time PCR. The RNA was isolated from the 5-Azacytidine treated cells and reverse transcribed to cDNA using Superscript™ II Reverse Transcriptase from Invitrogen (CA, USA).

The primers for real time PCR amplification were designed using Primer3 software. The TaqMan MGB FAM labelled probes were purchased from Applied Biosystems (Foster City, CA). The 6-FAM labelled probes and the primers used for real time PCR amplification for Variant 2 are listed in Additional file 1: Table S1, along with respective amplicon sizes. The cDNA was used as template and the final reaction consisted of 1x Taqman® Fast Advanced Master Mix (Applied Biosystems, Carlsbad, CA), 3.0 μM FAM labelled probe, 10pmol/μl of forward and reverse primers, Human 18S rRNA VIC-MGB internal control, Applied Biosystems (Foster City, CA).

Human colon cancer cell lines LS-174 T and RKO were grown in EMEM medium with 10 % fetal bovine serum and penicillin/streptomycin at 37 °C (5 %CO₂). The cells were lysed in lysis buffer containing 1 mM PMSF, 0.5 mM DTT and 1 mM cocktail. The total cell lysates containing equal amounts of protein (500 μg) were immunoprecipitated using magnetic beads coupled protein G following Dynabeads® protocol (novex®, Life technologies). Briefly, total cell lysates were pre-incubated with mouse monoclonal Rad51C antibody (Santa Cruz Biotechnology, CA) overnight at 4 °C while rotating, washed and eluted the IP product in 20 μl volume using elution buffer by the supplier (novex®, Life technologies) and stored at −80 °C for detection using immunoblot.

Equal amount of proteins isolated from cells and tumor tissues were size fractionated on 4–12 % NuPAGE gel. Proteins were then transferred onto nitrocellulose membrane. The membrane was blocked with blocking buffer (5 % non fat milk). The blocked membrane was then incubated with target antibody rabbit polyclonal ATXN7 antibody (Bethyl Laboratories INC, TX) at 4 °C overnight. After washing the membrane with TBS-T (20 mM Tris, 0.9 %Nacl) for three times 20 min each, the membrane was then incubated with anti-rabbit secondary polyclonal antibody (GE Healthcare, UK) for 2 h at room temperature. The detection of specific protein binding was performed with a chemiluminescence kit (Amersham, GE Healthcare, UK).

Five thousand cells from each line were seeded in each well of a 96-well plate 24 h before treatment. Cells were treated with cisplatin (5 μg/ml). Dimethylthiazolyl-2-5-diphenyltetrazolium bromide (MTT) dye solution (Sigma, St. Louis, MO) was added into the 96-well plate. The plate was incubated at 37 °C for 4 h, and then the treatment was terminated by adding stop solution (isopropanol with 0.04 N HCl). MTT was cleaved by live cells to a colored formazan product. Absorbance at 560 nm wavelength was recorded using a Bio-Rad microplate reader 680 (Bio-Rad Laboratories, Hercules, CA). Each treatment was repeated in triplicate. An averaged absorbance of blank values (containing all reagents except cells) was subtracted from all absorbance to yield corrected absorbance. The relative absorbance of each sample was calculated by comparing the average of corrected absorbance with an average of corrected untreated control.