TopBP1 and Claspin contribute to the radioresistance of lung cancer brain metastases

We assessed radiosensitivity of lung cancer cell lines (PC14PE6 and H460) with various doses of radiation. The clonogenic survival assays revealed that both cell lines showed clear radiation sensitivity in a dose-dependent manner, and the survival fraction of PC14PE6 showed 0.4 at 2.5 Gy whereas the survival fraction of H460 cells was more severely reduced down to 0.04 at 2.5 Gy (Figure 1A). These reduced radiation sensitivity of PC14PE6 indicated that PC14PE6 cells were relatively more resistant to radiation than H460 cells. We made lung cancer brain metastases xenograft models using PC14PE6 and H460 cells and tested whether tumor cells in xenograft models acquired radioresistance after radiotherapy. As shown in Figure 1B, xenografted mice models were exposed to radiation once, and the tumor cell mass was excised and cultured in vitro again. To validate whether radiosensitivity of tumor cells were changed, the tumor cells were irradiated with a range of radiation doses (5–20 Gy). Survival representing resistance to radiation was examined by cell proliferation assay and shown by percent of viability to non-irradiated cells (Figure 1C & D). Initially H460 cells have been less resistant to radiation, but cells from irradiated tumor cells (H460-5Gy and H460-10Gy) acquired radioresistance of over 90% of the survival rate (Figure 1C). H460-10Gy cells were even more radioresistant than H460 and H460-5Gy. These results show that radiosensitive H460 cells acquired radioresistance in a dose dependent manner. However, PC14PE6, originally more radioresistant than H460 did not show additional increase of resistance regardless of the radiation dose (Figure 1D).

After establishing radioresistant cells from the xenograft model, we explored the significant transcriptional changes by performing DNA microarray profiling. Using a filter criterion of at least 2-fold change with p < 0.05, the numbers of genes with changes in expression in H460-10Gy and PC14PE6-20Gy cells were determined: 385 genes increased and 329 genes decreased in H460 while 76 genes increased and 98 genes decreased in PC14PE6. Between the two cell lines, 64 of the increased genes and 63 of the decreased genes overlapped. These results are presented in a Venn diagram (Figure 2A). In particular, the number of affected genes in H460-10Gy cells was 4.1 fold higher than that in PC14PE6-20Gy. These results showed that transcriptional changes induced by radiation therapy are more extensive in radiosensitive cells (H460) than in resistant cells (PC14PE6). Functional annotation of the genes was carried out using gene ontology-based biological property analysis. The genes were categorized into 15 functional groups (Figure 2B & C). Most of the changed genes in the two cell lines were associated with cell differentiation, proliferation and cell cycle. The genes with significant changes in expression are listed in Additional file 1: Table S1-S4.

Genes involved in cellular response to DSBs are the major targets for cancer cell death [12]. Specifically, functional defects in DNA damage checkpoint pathways show an increased sensitivity to radiation [2, 10, 13], which led us to focus on the DNA Damage response (DDR) signaling pathway for subsequent analyses. The genes with expression change in DNA damage checkpoint pathway are listed in Additional file 1: Table S5. Notably, genes from H460-10Gy outnumber those from PC14E6-20Gy. We chose TopBP1 and Claspin genes as our targets for functional analysis as they play important roles for activating Chk1 in DDR pathway [6, 8, 14]. In particular, since Chk1 inhibition using inhibitors showed an enhancement of radiation sensitivity [13, 15], it is worth analyzing the role of genes upstream to Chk1 with regards to radiosensitivity. Chk1 is a well-established ATR target involved in DDR and normal cell cycle progression and thus represents a potential pharmacological target in the development of anti-cancer therapeutic agents [16, 17].

To confirm and validate the findings obtained from our microarray analysis, we used the radioresistant PC14PE6 cells and established stable lines in which TopBP1 and Claspin were regulated by shRNA via doxycycline treatment. It was confirmed that the mRNA levels of the two genes decreased by doxycycline treatment compared to untreated PC14PE6 cells (Additional file 1: Figure S1A & B). Induction of shRNA virtually completely abolished the expression of TopBP1 and Claspin (Figure 3A). It has been shown that TopBP1 and Claspin play an essential role in Chk1 activation/phosphorylation in response to DNA damage [6, 14]. We next examined the phosphorylation of Chk1 in response to irradiation after knockdown of TopBP1 and Claspin. As shown in Figure 3A, depletion of TopBP1 or Claspin abolished Chk1 phosphorylation as previously reported [6, 7, 14]. Immunocytochemistry of γ-H2AX and TopBP1 demonstrated that γ-H2AX foci were significantly decreased and the signals for TopBP1 were scantly detected after the depletion of TopBP1 (Additional file 1: Figure S1C).

We performed clonogenic survival assay to measure the change in sensitivity to radiation in these stable cell lines. The PC14PE6/shLuc cells showed a similar sensitivity to irradiation in comparison to original PC14PE6 cells (Figure 3Ba). TopBP1 and Claspin knockdown cells showed a higher sensitivity than control cells. The decrease in survival was particularly dramatic at higher dose of IR (e.g. 4 Gy) than at lower dose of IR (< 2 Gy) for both PC14PE6/shTopBP1 and PC14PE6/shClaspin cells (Figure 3Bb & c). Our results are consistent with a previous report describing an enhanced susceptibility to radiation in cells with down-regulated Chk1 [13].

The expression levels of TopBP1 and Claspin increased in induced radioresistant H460-10Gy cells. To clarify and further evaluate their roles with radioresistance, TopBP1, Claspin, and Chk1 were overexpressed in three independent cell lines, H460, H23, and U2OS which are all highly radiosensitive compared to PC14PE6 (Figure 3C & Additional file 1: Figure S2). The overexpression of each gene was examined by immunoblot (Figure 3C). In clonogenic assay, GFP expression vector was used as a control. In H460 and H23 cells, overexpressing TopBP1 showed higher survival (over 50%) than control cells after 2 Gy IR exposure (Figure 3D). In U2OS, the number of surviving colonies increased by about 20% compared to the control. These results indicate that overexpression of TopBP1, Claspin, and Chk1 induces radioresistance, consistent with the microarray data.

Based on results from in vitro functional assays, we hypothesized that TopBP1 or Claspin would be potential target genes in lung cancer brain metastasis models (Figure 4). We thus sought to estimate the effects of TopBP1 and Claspin knockdown based on the survival of lung cancer brain metastasis xenograft model animals in response to irradiation. We produced orthotopic xenograft animals with injection of PC14PE6/shTopBP1 or PC14PE6/shClaspin cells into the brain of athymic nude mice. Tumor-bearing mice were treated with doxycycline and/or radiation therapy as illustrated (Figure 4A). As shown in Figure 4B, without irradiation, there was no difference in survival rates between TopBP1-inhibited (i.e. doxycycline-treated) group and TopBP1-expressing group. The median survival time was 19 days. In contrast, when irradiated, xenografted mice with inhibited expression of TopBP1 showed a higher survival rate than the control mice with the mean life span of 42 days post injection (Figure 4C). In case of Claspin knockdown mouse, the survival also extended with irradiation and doxycycline treatment although the effect was not as dramatic as in the case of TopBP1 knockdown. Lung cancer brain metastasis models in this study do not represent patients completely. Radiotherapy did not mimic patient dosing and fractionation schedules for limitation of animal models. In sum, xenotransplant analysis further confirmed the proposed role of TopBP1 or Claspin in DDR response and acquisition of radioresistance.

Human non-small cell lung cancer (NSCLC) cell lines H23, H460, and human osteosarcoma cell (U2OS) (American Type Culture Collection) were maintained in RPMI 1640 medium containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin, and PC14PE6, kindly provided by MD Anderson Cancer Center was maintained in Eagle’s minimal essential medium (MEM) containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37 °C and 5% CO₂. Cell lines were regularly tested for mycoplasma infection using the MycoAlert mycoplasma detection kit (Lonza). Isolated PC14PE6 and H460 cells from xenograft mouse brains were immediately used to assay for the cell viability after dissociation procedure.

For overexpression of TopBP1, Claspin, and Chk1 in H23, H460 or U2OS cell lines, each cell lines were transfected with expression vectors by electroporation using Cell Line Nucleofector® Kit V with the Amaxa Nucleofector I system (Lonza). Stable cell lines of tetracycline-regulated expression of a short hairpin RNA were generated as previously described [13]. For shRNA induction, cells were seeded to achieve 30%-50% confluency on the day of induction and doxycycline was added into the culture medium at 1 μg/ml.

Cells were lysed and prepared for Western blotting analysis as previously described [19]. Anti-Chk1 and anti-α-tubulin antibodies were purchased from Santa Cruz Biotechnology. Anti-Claspin, anti-Chk1 phospho-Ser317, anti-Chk2 phospho-Thr68 and anti-γ-H2AX antibodies were obtained from Cell Signaling Technology. Anti-TopBP1 was obtained from Abcam. The secondary antibodies Cy3-conjugated anti-mouse IgG and FITC-conjugated anti-rabbit IgG for immunofluorescence were purchased from Jackson Immunoresearch.

siRNAs were obtained from Genolution Pharma, Korea. The sequences of the siRNAs targeting TopBP1 were: #1, CAGAAUUGUUGGUCCUCAAdTdT and #2, CGAUAGAGGAGACUCAUGAdTdT. The sequences of Claspin siRNA 1 and siRNA 2 were GGAAAGAAAGGCAGCCAGAdTdT and CCUUGCUUAGAGCUGAGUCdTdT, respectively. The control luciferase siRNA sequence was GTTACGCTGAGTACTTCGAAA. After the test of mRNA knockdown efficacy, plasmids expressing shRNA were constructed by shRNA cloning service from Genolution.

Cell proliferation was analyzed with Cell Counting Kit 8 (Dojindo Molecular Technology) and clonogenic survival assay was performed as previously described [13]. Cell proliferation was expressed as a percentage of the absorbance obtained in irradiated cells relative to that in untreated control cells. Each experiment in triplicates was repeated three times. Error bars represent the standard deviation.

All experiments were conducted in accordance with the Institute Laboratory Animal Research Guide for the Care and Use of Laboratory Animals and within the protocols approved by the appropriate Institutional Review Boards at the Samsung Medical Center (Seoul, Korea). Female athymic nude mice, 7 weeks of age were used for this study. Production of orthotopic lung cancer brain metastases models were performed as previously described [13]. Mock-irradiated or irradiated mice were anesthetized with 100 mg/kg ketamine and 10 mg/kg xylazine and restrained in custom-designed mice jig. Whole brain irradiation (15 Gy) was performed after 2 weeks. To induce knockdown of TopBP1 and Claspin, doxycycline, which was formulated in 5% sucrose, was administered for 5 days in drinking water from 12 days after tumor cells implantation at a dose of 1 mg/ml.

RNA expression array was performed as previously described [18]. RNA was transcribed into cDNA and hybridized onto Affymetrix GeneChip HG-U133Plus2.0 according to the Affymetrix GeneChip Expression Analysis Technical Manual. The gene expression CEL files were normalized using the Robust Multichip averaging procedure. The dataset was normalized using Central Tendency Median Normalization. Expression array data were deposited at http://​tbi.​skku.​edu/​microarray/​smc/​yoo-20140106.