|Home | About | Journals | Submit | Contact Us | Français|
Stereotactic ablative radiotherapy (SABR) has emerged as a highly promising treatment for medically inoperable early-stage non-small cell lung cancer patients. Treatment outcomes after SABR have been excellent compared to conventional fractionated radiotherapy (CFRT). However, the biological determinants of the response to ablative doses of radiation remain poorly characterized. Furthermore, there's little data on the cellular and molecular response of genetically distinct NSCLC subtypes to radiation. We assessed the response of 3 genetically distinct lung adenocarcinoma cell lines to ablative and fractionated ionizing radiation (AIR and FIR). We studied clonogenic survival, cell proliferation, migration, invasion, apoptosis and senescence. We also investigated the effect of AIR and FIR on the expression of pro-invasive proteins, epithelial-to-mesenchymal transition (EMT), extracellular signal-regulated kinases (ERK1/2) and the transmembrane receptor cMET. Our findings reveal that AIR significantly reduced cell proliferation and clonogenic survival compared to FIR in A549 cells only. This differential response was not observed in HCC827 or H1975 cells. AIR significantly enhanced the invasiveness of A549 cells, but not HCC827 or H1975 cells compared to FIR. Molecular analysis of pathways involved in cell proliferation and invasion revealed that AIR significantly reduced phosphorylation of ERK1/2 and upregulated cMET expression in A549 cells. Our results show a differential proliferative and invasive response to AIR that is dependent on genetic subtype and independent of intrinsic radioresistance. Further examination of these findings in a larger panel of NSCLC cell lines and in pre-clinical models is warranted for identification of biomarkers of tumor response to AIR.
NSCLC, non-small cell lung cancer; EGFR, epidermal growth factor receptor; KRAS, kirsten rat sarcoma viral oncogene homolog; cMET, mesenchymal epithelial transition factor; EMT, epithelial mesenchymal transition; SABR, stereotactic ablative radiotherapy; AIR, ablative ionizing radiation; FIR, fractionated ionizing radiation; IR, ionizing radiation; ERK, extracellular signal-regulated kinases; Gy, Gray
Non-small cell lung cancer (NSCLC) represents the most common type of lung cancer, affecting 75–80% of patients.1,2 Surgical resection is the treatment of choice for stage I NSCLC, resulting in 5-year survival rates of approximately 50–70%.3-6 However, up to a third of early-stage patients are medically inoperable. Historically, these patients were treated with conventional fractionated radiotherapy (CFRT) comprised of 60–66 Gy in 2–3 Gy fractions, which offered poor control with an in-field tumor progression rate of 80–90%.7-10 In recent years, stereotactic ablative radiation therapy (SABR), a technique allowing accurate delivery of very high doses of radiation in a small number of fractions, has evolved as the standard therapy for inoperable patients.2,8,11-14 Typically, a dose of 34–60 Gy is delivered in 1–8 fractions distributed over 1–2 weeks. Reported treatment outcomes with SABR have been excellent, with 3-year local control greater than 90%, and overall survival rates at 3 y that are in excess of 50%.12,13,15
Despite its widespread use, the biology underlying cellular responses and cell fate following SABR remain understudied. It is increasingly appreciated that NSCLC is not a uniform disease but consists of many genotype-defined subtypes. The most common subtypes are defined by oncogenic mutations in KRAS and EGFR, which may confer resistance and sensitivity to radiation, respectively.16-18 Importantly, it is still unknown how these genotypes can affect responses to SABR.
Controversially, a number of studies have demonstrated that SABR-like radiotherapy fractions can promote metastatic features such as migration and invasion in glioma,19-21 breast,22-24 liver,25 pancreas,26 melanoma27 and NSCLC cell lines.28-30 However, these observations were made without comparison to responses to fractionated radiation. Whether SABR has the capacity to enhance metastatic potential in NSCLC remains unknown. Furthermore, it is not known whether this effect is specific to certain genotype-defined tumor subtypes. This study was designed to assess the diversity in cellular and molecular responses to ablative ionizing radiation (AIR) and fractionated ionizing radiation (FIR) in a panel of oncogene driven lung adenocarcinoma cell lines. Our results highlight a differential response to AIR among genetically distinct NSCLC cells.
Human lung adenocarcinoma cells A549, H1975 and HCC827 were obtained from the American type culture collection (ATCC, VA). Cells were cultured at 37°C and 5% CO2 in RPMI 1640 media except for A549, which was cultured in DMEM F/12 media. Media was supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. Cell irradiation was performed using a Faxitron X-Ray machine (Faxitron X-ray Corporation, IL). X-ray tube voltage was set to 160 kVp, current of 6.3 mA and dose rate of 1450 R/min. Unless otherwise stated, cells were exposed to 12 Gy for AIR or 4 daily fractions of 3 Gy for FIR.
Inhibition of cell proliferation was assessed using the Vybrant® 3-[4,5- dimethylthiazol-2-yl]-2,5-diphenyltetrazoliumbromide (MTT) assay according to manufacturer instructions (Molecular Probes, OR). Cells were plated in 96-well plates at optimal densities 24 h after irradiation (8 Gy or 12 Gy for AIR and 4 daily fractions of 2 Gy or 3 Gy for FIR) and incubated for 3–5 d The number of fractions in FIR was limited to 4 to maintain optimal cell confluence and tissue culture conditions. Absorbance readings were performed at 560 nm using a standard microplate reader (Thermo Scientific, Multiskan Spectrum). Separate wells were kept for morphological analysis with the fluorescent stain, DAPI.
The colony-forming assay was performed as reported previously.31 Exponentially growing cells were irradiated with increasing doses of IR (0–10 Gy) and cultured until colonies appeared (10 days). Colonies with 50 cells or more were counted. The surviving fraction was determined by dividing the number of colonies by the number of cells plated with a correction for plating efficiency. Data was plotted on semi-log graphs and fitted using the linear-quadratic (LQ) model.32
Cell death was assessed in all cell lines after exposure to AIR using Trypan blue exclusion assay. Cells were plated at equal densities in 6-well plates, irradiated and harvested at 0h, 24 h, 48 h, 72 h and 96 h. Harvested cells were counted and analyzed using an automated cell counter (TC-10, Bio-Rad, CA). Cytochemical staining for senescence-associated b-galactosidase (SA-bgal) was performed 5 d after treatment using a commercial kit (Cell Signaling, MA).33 Images were acquired at 200x magnification. The number of stained and unstained cells was counted in at least 5 fields and averaged. Percent senescence was calculated as the fraction of positively staining cells over the total number of cells. Similarly, cell polyploidy was assessed 5 d after treatment as an indicator of cells undergoing mitotic catastrophe. Cells were fixed and stained with DAPI and analyzed under a fluorescent microscope (Zeiss Axio Vert.A1). Images from 5 20x fields were obtained and analyzed for the presence of multinucleated cells. Percent polyploidy was calculated as the fraction of multinucleated cells over the total number of cells.
Cell cycle analysis was performed on our panel of NSCLC cell lines 24-hours after exposure to 12 Gy. Briefly, cells from treated and untreated plates were collected and fixed in 70% ethanol for 30 minutes. Cells were centrifuged and resuspended in 1mg/mL RNase A (Sigma-Aldrich, St. Louis, MO) and incubated at 37°C for 30 minutes. Cells were stained with propidium iodide (Sigma-Aldrich St. Louis, MO) by incubating for 20 minutes at room temperature. Cell cycle analysis was performed using a BDFACS Canto II Flow cytometer (BD Biosciences, San Jose, USA). The percentage of subG1, G1, S and G2 populations was calculated with FlowJo software (version 9.6.2, FlowJo, LLC, Ashland, OR, USA).
Cell migration and invasion was assessed using modified Boyden chambers (8-μm pore, BD Biosciences, CA) according to previously established methods.34 Chambers were either non-coated for migration or pre-coated with BD Matrigel basement membrane for invasion studies. Assays were performed 5 d after exposure to AIR or FIR. Invading cells were counted under 200x magnification from at least 5 fields per chamber and averaged. Microscopic images of invading cells were taken using a Zeiss camera attached to a Zeiss Axio Vert.A1 microscope.
A549 cells were incubated for 24 h or 72 h after irradiation. After the incubation period, cells were plated in triplicates in 6-well plates and allowed to adhere overnight. Wounds were made using a sterile 200 uL pipette tip.35 Wells were washed with PBS and cells were replenished with starving media (0.5% FBS). Phase contrast images were acquired from the same region at 0 h, 24 h and 48 h at 100x magnification. Wound closure was measured using ImageJ software (http://imagej.nih.gov/ij) and normalized to the initial wound width (0% closure, 0h).
A549 cells were serum starved 16 hours prior to irradiation and cell lysates collected at 3 hr, 24 hr and 48 h time points. Protein concentration was quantified using a BCA protein kit (Thermo Fisher Scientific, MA), according to manufacturer instructions. Equal amounts of protein were separated by SDS-PAGE under reducing conditions and blotted onto polyvinylidene difluoride membranes.36 Membranes were blocked in 5% nonfat milk and probed with primary antibodies directed against p53 (Santa Cruz Biotechnology, TX), p21, poly(ADP-ribose) polymerase-1 (PARP-1), E-cadherin, vimentin, phospho-FAK-tyr397, FAK, c-MET, phospho-ERK1/2, ERK1/2, phospho-AKT, AKT, phospho-mTOR, mTOR or β-actin (Cell Signaling, MA). Densitometric measurements were performed using ImageJ software.
All experiments were performed in triplicates and repeated at least in 3 independent settings. Unless otherwise stated, statistical analysis was performed using 2-way ANOVA in Prism Software and all values are expressed as mean ± standard error of the mean (SEM). Statistical significance was set at *p < 0.05; **p < 0.01; ***p < 0.001.
To assess the response to AIR, 3 genetically distinct NSCLC adenocarcinoma cell lines were used. A549 cells harbor mutant KRAS (G12S) whereas HCC827 and H1975 cells harbor mutant EGFR (Del E746-A750 and L858R respectively). In addition, H1975 cells harbor the T790M mutation which has been shown to confer resistance to EGFR tyrosine kinase inhibitors.37 Cells were exposed to 8 Gy or 12 Gy and proliferation was assessed by MTT. Inhibition of proliferation was normalized to untreated control and compared to FIR comprised of 4 fractions of 2 Gy or 3 Gy. Significant differences in the inhibitory capacity of AIR compared to FIR were observed in A549 cells only. Cell proliferation was reduced by 60% and 71% following exposure to single doses of 8 Gy and 12 Gy, respectively. Equivalent fractionated total doses of 8 Gy and 12 Gy resulted in 38% and 36% reduction, respectively (Fig. 1A). In contrast, HCC827 and H1975 cells did not exhibit a significant proliferative difference after exposure to AIR compared to FIR. In HCC827 cells, AIR of 8 Gy and 12 Gy decreased proliferation by 43% and 65%, while fractionated doses reduced cell proliferation by 47% and 72%, respectively (Fig. 1B). Similar findings were also observed in H1975 cells (Fig. 1C).
To determine whether the difference in inhibition of proliferation between AIR and FIR in A549 cells is due to differences in the biologically equivalent dose (BED), we compared a single-dose of 8 Gy to 4 fractions of 3 Gy (BED of 14.4 Gy vs 15.6 Gy, respectively). Our findings show similar inhibition of proliferation between 8 Gy and 12 Gy single-doses. Therefore, for subsequent experiments a dose of 12 Gy was used and compared to an equivalent dose in 4 fractions.
To determine whether the differences in proliferation were related to intrinsic radiosensitivity of the individual cell lines, we performed clonogenic cell survival assays and assessed the survival fraction at 2 Gy (SF2). Consistent with literature values,38 our results show that A549 and HCC827 cells are relatively less radiosensitive (SF2 0.74 and 0.67 respectively) than H1975 cells (SF2 0.34) (Fig. 1D). Collectively, these results show a differential proliferative response to AIR based on genetic subtype and independent of intrinsic radioresistance as measured by SF2. AIR differentially regulated cell proliferation of A549 cells compared to FIR. In contrast, AIR and FIR were equally effective in inhibiting proliferation in HCC827 and H1975 cells.
We sought to examine the effects of AIR on cell fate in our panel of cell lines. Following exposure to AIR, A549 cells acquired a distinct phenotype with large cell cytoplasm and formation of numerous cell protrusions. In contrast, H1975 and HCC827 cells exposed to AIR became enlarged and developed fragmented and floating bodies. Time-course analysis of cell death, up to 96h after AIR, using Trypan blue exclusion revealed significant increase in cell death in HCC827 and H1975 cells, but not in A549 cells compared to untreated cells (Fig. 2A). Analysis of SA-bgal staining identified the presence of similar levels of senescent cells in all cell lines, 5 d after exposure to AIR (Fig. 2B). Further analysis, revealed that all senescent HCC827 and H1975 cells were highly multinucleated and contained micronuclei indicative of cells undergoing mitotic catastrophe.39 In contrast, less than 5% of senescent A549 cells were multinucleated (Fig. 2C). Since senescence did not dominate the entire population of AIR-treated A549 cells (less than 40%), in separate experiments we observed that A549 cells regained their full proliferative capacity after 28 d (data not shown).
Cell cycle analysis of our panel of lung adenocarcinoma cell lines 24 hours following exposure to 12 Gy using flow cytometry showed substantial increase in subG1 cells in HCC827 and H1975 cells, indicative of apoptosis (Fig. 3B, C). The reduction in S-phase population in HCC827 and H1975 cells was at the expense of an elevated G1-phase population and a drastic reduction in the G2-phase population. These results indicate that AIR of 12 Gy induces apoptosis and G1/S cell cycle arrest in HCC827 and H1975 cells. Conversely, A549 cells demonstrated a lack of G1 cell cycle arrest, a reduction in S-phase population and elevated G2-phase population indicating G2/M arrest with no evidence of apoptosis or cell death (Fig. 3A). Consistent with the observed G1 cell-cycle arrest in A549 cells, AIR induced significantly higher levels of p53 and downstream p21 compared to control as shown in Fig. 3.
We investigated the effect of AIR on cell migration and invasion in NSCLC cells using modified Boyden chambers. A549 cells exposed to AIR, but not FIR showed a statistically significant increase (fold2-) in cell migration and invasion (Fig. 4A, D) compared to untreated cells. HCC827 cells exposed to either AIR or FIR showed significantly lower migration and invasion compared to untreated cells (Fig. 4B,E). In H1975 cells, AIR did not significantly impact migration levels, but significantly reduced invasion, while FIR significantly reduced migration and invasion compared to untreated cells (Fig. 4C, F). These results clearly demonstrate that AIR-induced invasion is dependent on the cell's molecular subtype.
To gain better visualization of morphological alterations associated with cell migration of A549 cells after exposure to IR, we performed wound-healing assays. In A549 cells exposed to FIR, no significant difference in wound-healing capacity was observed compared to untreated cells. In contrast, A549 cells exposed to AIR showed a significant increase in motility and wound closure at 24h and 48h relative to untreated as well as FIR-treated cells (Fig. 5).
Exposure to IR can induce epithelial-to-mesenchymal transition (EMT) and activate pro-invasive pathways involving focal adhesion kinas (FAK), extracellular signal-regulated kinases (ERK1/2), mechanistic target of rapamycin (mTOR) and the transmembrane receptor c-MET.21,26,40-42 However, the effect of radiation regimen (AIR versus FIR) on radiation-induced invasion has not been previously investigated. We did not observe a significant difference in the expression of vimentin and E-cadherin at 24h following AIR compared to FIR (Fig. 6). However, AIR upregulated c-MET expression by 2.fold3- in A549 cells compared to FIR (Fig. 6). Phosphorylation of ERK1/2 was significantly reduced after AIR compared to FIR. This is consistent with the role of ERK1/2 in cell-cycle progression and cell proliferation.43 No significant difference was observed in total and phosphorylated levels of FAK, AKT and mTOR after AIR compared to FIR (Fig. 6).
NSCLC consists of many genotype-defined subtypes. The common subtypes are defined by oncogenic mutations in KRAS or EGFR.44 SABR is increasingly used in the clinic as a curative treatment for inoperable early-stage NSCLC patients. The molecular and cellular characterization of the response to ablative-doses of radiation compared to conventional fractionation in genetically distinct NSCLC cells remain poorly investigated. Our study is the first attempt to examine the heterogeneity of response to AIR across genetically distinct adenocarcinoma cell lines. Our results identify significant differences in the proliferative, clonogenic, invasive and cell death response to AIR across 3 lung adenocarcinoma cells lines harboring KRAS and EGFR mutations.
It is generally established that ablative-doses (above 8 Gy) of radiation provide superior tumor growth inhibition compared to equivalent fractionated doses.45 Our results confirm this finding in A549 cells, but we demonstrate that in HCC827 and H1975 cells, the response to AIR and FIR is similar, irrespective of intrinsic radioresistance. Radioresistance is traditionally defined by the survival fraction at 2 Gy (SF2) derived from cell survival curves of clonogenic assays.45 The similar response to AIR and FIR observed in HCC827 and H1975 cells can be explained partly by the presence of tyrosine kinase mutations in EGF receptors of both cell lines. HCC827 cells harbor a deletion at the E746-A750 of exon 19 while H1975 cells harbor 2 mutations resulting in amino acid change in the tyrosine kinase domain of EGFR (L858R and T790M). Previous studies have shown that EGFR can translocate to the nucleus upon activation with IR where it binds to DNA-dependent protein kinase catalytic subunits (DNA-PKcs), leading to activation of the nonhomologous end-joining DNA repair pathway.46,47 Mutations in EGFR block its translocation and binding to DNA-PKcs.48 This may result in comparable levels of DNA damage after AIR and FIR, resulting in similar inhibition of proliferation as observed in our study.
Cell fate is an important consideration when assessing the response to AIR. The decrease in cell proliferation observed in HCC827 and H1975 cells following AIR, was associated with increased apoptosis whereas in A549 cells, inhibition of proliferation was attributed to cell-cycle arrest leading to senescence. The exact mechanism controlling cell fate following AIR remains unknown. The p53 tumor suppressor protein is a key mediator of the DNA damage response that is activated following exposure to ionizing radiation.49,50 Wild type p53 is known to induce cell cycle arrest through the induction of cyclin-dependent kinase inhibitor, p21CIP1/WAF151. p53-dependent activation of p21 is also a key pathway in stress-induced senescence and promotes apoptosis resistance in several cancer cell lines.43,52 A549 cells harbor wild-type p53, which may have contributed to the induction of senescence and decreased proliferation in A549 cell exposed to AIR. However, stress-induced senescence is not the most desirable outcome of cancer treatment as senescent cells remain metabolically active and can secrete growth factors, cytokines and matrix-degrading enzymes which result in increased migration and invasion of nearby tumor cells.53-56 Converting senescence to apoptosis is an active area of research and represents the most desirable outcome of anti-neoplastic treatment.17,52
Radiation-induced invasion has been reported by a number of investigators. Canazza et al compared a single-dose of 8 Gy to 4 fractions of 2 Gy in a glioblastoma cell line and reported increased invasion after single-dose, but not fractionated doses.20 In Lewis lung tumor cells, irradiation with 7.5 Gy increased invasion in vitro while in tumor-bearing rats irradiation with 50 Gy in 5 fractions was shown to promote pulmonary metastasis.57 Our findings support these reports, as we demonstrate that A549 cells that survived AIR develop a more invasive phenotype compared to FIR. Studies have also shown that radiation can activate pro-invasive pathways and induce epithelial-to-mesenchymal transition in vitro.30,41 Our analysis did not reveal significant changes in EMT marker expression after AIR compared to FIR. However, we showed increased expression of cMET after exposure to AIR but not FIR, which may have contributed to the development of more invasive profile in AIR-treated A549 cells. Activity of the cMET tyrosine kinase receptor has been shown to promote invasion and metastasis in NSCLC.58,59 Furthermore, cMET overexpression was shown to promote survival and inhibit apoptosis in NSCLC cells.60 More studies are needed to define the potential role of c-MET in AIR-induced invasiveness. This is of clinical significance as c-MET inhibition was shown as an effective strategy to radiosensitize NSCLC tumors and is currently being tested in clinical trials of NSCLC.61,62
Current radiotherapy schedules do not account for a tumor's genetic composition. In particular, knowledge of EGFR and K-RAS status may not only impact patient prognosis, but also the choice of radiotherapy treatment regimen in NSCLC. Our findings underscore the need for genetic characterization of tumors and elucidation of radiation-induced pathways in order to provide better treatment with SABR. Combining SABR with targeted therapy may provide an effective regimen to circumvent the pro-invasive effects of radiation. Identification of biomarkers that predict tumor response to ablative radiotherapy will help identify patients who may derive the most benefit from SABR, and will allow tailoring of ablative radiation doses to genetically distinct NSCLC subtypes.
No potential conflicts of interest were disclosed.
This work was done with grant support from the Montreal General Hospital foundation. AO was funded by Fonds de recherche du Quebec - Santé (FRQS).