Glutamine synthetase facilitates cancer cells to recover from irradiation-induced G2/M arrest

ABSTRACT

Resistance to radiation of cancer cells can be either intrinsic or acquired, leading to treatment failure. In response to DNA damage caused by IR, cancer cells are arrested in cell cycle showing limited proliferation and increased apoptosis. However, radiation-resistant cells are able to overcome the cell cycle block and proceed to proliferation, for which the detailed mechanism remains to be elucidated. In the present study, we showed that radioresistant cells exhibited a recoverable G2/M phase during prolonged cell cycle and manifested lower apoptosis rate and more colony formation. RNA-seq analysis revealed that glutamine synthetase (GS, GLUL) gene was highly expressed in radioresistant cancer cells in comparison with the parental cells, which was in accordance with the G2/M arrest after ionizing radiation. Knocking out of GS in radioresistant cells resulted in a delayed G2/M recovery and lowered proliferation rate after ionizing radiation treatment, which was accompanied with increased inhibitory phosphorylation of CDK1 at Y15 and downregulated Cdc25B, a dual specific phosphatase of CDK1. Moreover, there was an enhanced complex formation of CDK1 and Cyclin B1 when the cells were rescued by re-introducing GS. In vivo, knocking down of GS significantly sensitized CNE2-R xenografts to RT in mice. In this study, we demonstrate a novel role of glutamine synthetase independent of metabolic function in promoting recovery from G2/M arrest caused by ionizing radiation, thus, causing cancer cell resistance to radiotherapy.

Introduction

Nasopharyngeal carcinoma (NPC) is a type of head and neck cancers, commonly occurring in people in southern Asia and northern Africa.¹ Radiotherapy, delivered using ionizing radiations (IRs), is a frequently used approach to treatment of NPC.^2–4 IR relies on the generation of reactive oxygen species and DNA double-strand breaks in rapidly replicating cancer cells, thereby exhibiting a tumor-killing effect.^5,6 Alternatively, cancer cells initiate a DNA damage repair pathway to counteract the formation of detrimental lesions which requires the activation of cell cycle checkpoints.^6,7 However, dysregulation of cell cycle checkpoints, which also accounts for severe malignance of cancer cells, presents an obstacle in radiotherapy.^8–11

DNA damage induced by either radiotherapy or genotoxic agents could temporally initiate cell cycle arrest, ^12–14 giving the cells sufficient time for repair.⁶ The G2/M checkpoint is crucial in preventing premature mitosis in cells with damaged DNA, especially in cells with p53 functional deficiency.¹⁵ The G2/M checkpoint is initiated by an inactivated CDK1/Cyclin B1 complex.⁷ The activity of this complex is inhibited by phosphorylated site of CDK1 at Y15.^16,17 When damage is being repaired or tolerated, CDK1 can be dephosphorylated at Y15 and T14 by Cdc25¹⁸ and phosphorylated at T161 by CAKs (Cyclin-Dependent Activating Kinases), ¹⁹ resulting in activation of CDK1 and cell cycle recovery. However, it remains to be elucidated how the irradiated cells make the choice between overcoming arrest or dying.²⁰

Glutamine synthetase (GS) is a metabolic enzyme that catalyzes ligation of glutamate and ammonia to form glutamine.²¹ Glutamine is a non-essential amino acid but is required for synthesis of nucleotides.²² Moreover, GS contributes to catabolism fueling the tricarboxylic acid cycle through coupling with GLS (Glutaminase). Tumor cells employ glutamine as an essential metabolic fuel to meet the increased energy demands during fast proliferation or in genomic stress.^23–25 Recently, it was reported that GS might have a non-metabolic function in cell cycle regulation; however, the mechanism remains unknown.^26,27

In the present study, we found that IR-induced permanent arrest in the G2/M phase in IR-sensitive NPC cells (CNE2), whereas IR-resistant cells (CNE2-R) exhibited a tolerance to this arrest. Among the upregulated genes in IR-resistant cells, GS was found to help the cells escape from the G2/M arrest under irradiation, implying a non-metabolic function of GS in the cell cycle. Mechanistically, we found that GS promoted reactivation and assembling of CDK1/Cyclin B1 complex through up-regulated Cdc25B. Our study suggests a novel mechanism whereby cancer cells overcome G2/M arrest induced by IR, leading to radioresistance.

Results

Radioresistant NPC cells are resilient to IR-induced G2/M arrest

Cells under radiation stress tend to arrest at the cell cycle to facilitate DNA damage repair. To examine the features of cell cycle in radioresistant cancer cells, CNE2 and IR-resistant CNE2 (CNE2-R) were used as models. To confirm the IR resistance of CNE2-R, cell survival and apoptosis assays were performed. CNE2-R exhibited stronger colony formation ability but significantly lower apoptosis under irradiation than that of parental CNE2 cells (Figure 1(a,b)). Without irradiation, CNE2-R showed a comparatively lower proliferation rate compared to CNE2 (Figure 1(c)). However, the proliferation rates in CNE2 were greatly reduced by ionizing radiation in comparison to CNE2-R (Figure 1(d)).

Figure 1.

Resistance to IR-induced G2/M arrest in radioresistant NPC cells. (a), CNE2 and CNE2-R cells were exposed to 0, 2, 4 and 6 Gy IR. Formed colonies were counted after 2 weeks for survival fraction plotting. The radiobiological parameters D0, Dq, N, and sensitizing enhancement ratio (SER) were calculated and radiosensitivity was assessed by using a single-hit multitarget model of cell survival. (b), Four days post exposure to 8 Gy X-ray irradiation, apoptosis ratio of CNE2 and CNE2-R were analyzed by flow cytometry. (c), CNE2 and CNE2-R with or without 8 Gy X-ray radiation treatment were measured for cell proliferation rate in time course by cell counting. (d), CNE2 and CNE2-R cells were measured for cell proliferation rate with or without different doses of IR by cell counting. (e), Cells were synchronized with 2mmol/L double thymidine. The cell cycle profile was analyzed with or without 8 Gy X-ray radiation treatment by Flow cytometry. The arrows indicate the portion of the cells in G2/M phase. (f), The percentage of cells in G2/M was analyzed at 20 and 28 h after 8 Gy X-ray radiation treatment by Flow cytometry.

To analyze the cell cycle profiles in CNE2 and CNE2-R cells, the cells were synchronized with double thymidine and subjected to 8 Gy IR treatment. The cells were released and collected for FACS analysis at indicated times. Prolonged cell cycle, extending from 16 h to 40 h, was observed in both CNE2 and CNE2-R after IR treatment. Interestingly, CNE2 cells remained arrested at the G2/M even after 40 h, whereas CNE2-R cells started to bypass the G2/M arrest and returned to the cell cycle 16 h after IR (Figure 1(e)). Subsequently, the percentage of G2/M phase in CNE2 and CNE2-R was analyzed at 20 h and 28 h after IR, further confirming that CNE2-R could overcome G2/M arrest at approximately 28 h post IR treatment (Figure 1(f)). The absence of notable G2/M blockage and apoptosis in CNE2-R during the IR treatment suggests that the radiation-resistant cancer cells underwent some changes that endow it with resistance to cell cycle arrest.

GS promotes IR-resistant NPC cell recovery from IR-induced G2/M arrest

To examine the molecular events that caused unusual G2/M arrest transition in radioresistant cells, we performed RNAseq analysis. Glutamine synthetase was found to be one of the most up-regulated genes (Supplementary Figure 1). As shown in Figure 2(a), the GS expression in CNE2-R was significantly higher than that in CNE2. We found that the dynamic pattern of GS expression was well fitted to the cell cycle profile, where GS protein-level was gradually increased and reached the peak at 30 h in the phase of G2/M transition after IR (Figure 1(e), 2(a)). This suggested that GS expression inhibited irradiation-induced G2/M cell cycle arrest.

Figure 2.

Promotion of recovery from IR-induced G2/M arrest in radioresistant NPC cells. (a), Western blot analysis of the expression of GS in CNE2-R and CNE2 cells after exposure to 8 Gy X-ray irradiation. (b), Two GS knockout cells were verified by western blot. (c), Cell proliferation assay was performed for CNE2-R and two KO GS cells without IR treatment. (d), Cell proliferation assay was performed for CNE2-R and two KO GS cells with 8 Gy IR treatment. (e), Cell proliferation assay was performed for CNE2-R and two KO GS cells with different doses of IR (0, 4, 6, 8, 10 Gy) at day 4 post irradiation. (f), The cell cycle distribution and G2/M ratio was analyzed by Flow cytometry at 20 h and 28 h after 8 Gy IR treatment.

To further verify this observation, we performed a GS loss of function analyses. Two independent clones with GS knockout by CRISPR was established in CNE2-R cells (GS KO) (Figure 2(b)). Loss of GS did not change the proliferation rate without IR treatment compared with control cells (Figure 2(c)). However, the proliferation rate of GS cells became slower than that of CNE2-R after IR (Figure 2(d)). Under different doses of irradiation, GS KO cells grew slowly compared with CNE2-R cells (Figure 2(e)). Cell cycle profile analyses showed that depletion of GS in CNE2-R cells resulted in a tardy recovery from G2/M arrest referring to the CNE2 phenotype (Figure 2(f)), where majority of GS KO cells were blocked at G2/M phase after the IR treatment. To further validate the roles of GS, we over-expressed GS in CNE2 cells and found that GS overexpression led to a decrease in cells in G2/M phase, marginal increase in proliferation and enhanced resistance to irradiation (Supplementary Figure 2). These data imply that high expression of GS leads to resistance to IR in CNE2-R cells by promoting G2/M recovery.

GS accelerates G2/M transition via increased activity of CDK1

G2/M transition requires the activation of critical cell cycle kinase, CDK1, which involves in an interaction of it with Cyclin B1. To prevent premature mitosis in DNA-damaged cells, the activity of CDK1/Cyclin B1 complex is inhibited by phosphorylation of CDK1 on residues of Y15 that was mediated by Wee1.⁷ To examine whether inhibitory CDK1 phosphorylation was involved in the radioresistance, immunostaining of the Y15 phosphorylation was performed for the proteins from total cell lysates of CNE2 and CNE2-R cells at the indicated time following IR treatment. Consistent with cell cycle profiles, the inhibitory phosphorylation in parental CNE2 cells was higher under IR than that in radioresistant cells (Figure 3(a)). When GS was knocked out, the inhibitory phosphorylation level significantly increased compared with CNE2-R cells after 8 Gy IR treatment (Figure 3(b)). GS depletion mirrored the phenotype of CNE2, resulting in enhanced and prolonged Y15 phosphorylation of CDK1 in KO GS cells (Figure 3(b)). Moreover, when GS was re-expressed in GS KO cells, the inhibitory phosphorylation level was found to be reduced (Figure 3(c)). Together, these data showed that the decreased CDK1 inhibitory phosphorylation was associated with a high level of GS in radioresistant cells, which may suggest an involvement of GS in cell cycle regulation under IR stress.

Figure 3.

GS acceleration of G2/M transition via increased activity of CDK1. (a), The levels of phosphorylated CDK1 pY15 were compared between CNE2 and CNE2-R cells by western blots. (b), The levels of phosphorylated CDK1 at Y15 were compared between CNE2-R and two KO GS cells by western blots. (c), GS KO cells (KO1) were rescued by re-introduction of GS and assayed for Y15 CDK1 levels by western blots after 8 Gy IR treatment.

GS increased Cdc25B and enhanced the assembly of CDK1/cyclin B1 GS

Given that GS decreased Y15 phosphorylation of CDK1, we aimed to investigate whether Cdc25, a dual specific phosphatase, was involved in the process as Cdc25 plays a crucial role in initiating G2/M recovery specifically by dephosphorylating Y15 of CDK1.¹⁸ We found that expression of Cdc25B was elevated in CNE2-R cells in comparison with GS KO cells (Figure 4(a and 4b)), which was consistent with the decreased levels of Y15 of CDK1 (Figure 3(b and c)). When we increased the Glutamine concentration from 2mM (in normal RMPI medium) to 4mM in KO cells or CNE2 cells, IR treatment did not change the cdc25B level (Supplementary Figure 3), which further supported our notion that GS may play some nonmetabolic roles in radiation response of the cells. Previous report indicated that active CDK1, with the dephosphorylated T14 and Y15 mediated by Cdc25, is more likely to form a complex with Cyclin B1.^28,29 Thus, we further tested if GS affected the formation of CDK1/Cyclin B1 complex. As shown in Figure 4(c), an enhanced interaction between CDK1 and Cyclin B1 was detected by co-immunoprecipitation (co-IP) when GS was re-introduced into GS KO cells. Next, we examined whether there was a direct interaction between GS and CDK1. Co-IP demonstrated that GS could bind to CDK1 and the interactive domain with CDK1 was possibly within the N-terminal aa 1–24 of GS (Figure 4(d,e,f)). The findings indicate that GS promotes Cdc25B-mediated dephosphorylation at Y15 of CDK1 and efficient formation of CDK1/Cyclin B1 complex, which accelerates G2/M transition.

Figure 4.

GS involvement in increased level of Cdc25B and enhanced assembly of CDK1/Cyclin B1. (a) and (b), The levels of Cdc25B were compared between CNE2-R and two GS KO cells (a, KO1; b, KO2) by western blot after 8 Gy IR treatment. (c), GS KO cells (KO1) transfected with SBP-Flag-CDK1 and HA-Cyclin B1, and Flag-GS. Immunoprecipitants obtained with anti-SBP beads from lysates of cells were used for western blot to determine the interaction between CDK1 and Cyclin B1 affected by GS. (d), GS KO cells (KO1) transfected with HA-CDK1and Flag-GS Immunoprecipitants obtained with anti-Flag beads from lysates of cells were used for western blot to determine the interaction between GS and CDK1. (e), The diagram of constructs containing truncated version of GS. (f), Truncated GS constructs tagged by Flag were co-transfected with HA-CDK1 respectively into KO1 cells and the immunoprecipitants from anti-Flag were used for immunoblots.

GS enhances radioresistance of NPC cells

To investigate the biological roles of GS in radiation response, apoptosis ratio in CNE2-R and GS KO cells was determined under irradiation, showing that irradiation-induced apoptosis was increased in 2 GS KO NPC cells in comparison with CNE2-R cells (Figure 5(a)). Furthermore, colony formation assays demonstrated that GS knockout could sensitize the resistant CNE2-R cells to IR (Figure 5(b)). To confirm these observations, GS expression was rescued in 2 GS KO cells by re-introducing GS expression vector (Figure 5(c,d)). As expected, the rescued GS expression in KO cells enhanced resistance to IR (Figure 5(e,f)). Therefore, increased expression of GS from IR treatment promotes NPC cell resistance to IR via preventing IR-induced G2/M arrest.

Figure 5.

Sensitization of NPC cells to IR by GS depletion. (a), CNE2-R and two KO GS cells were exposed to 8 Gy X-ray radiation and apoptosis ratio of were analyzed by flow cytometry 4 days post IR. (b), CNE2-R and two KO GS cells were seeded in 35 mm dishes for 24 h and then exposed to 0, 2, 4, 6 Gy IR. The colonies were counted after 2 weeks. (c) and (d), GS expression in KO1 (c) and KO2 (e) cells after re-expression of GS. (e) and (f), KO1-ctr and KO1-reGS cells (d) and KO2-ctr and KO2-reGS cells (f) were seeded in 35 mm dishes for 24 h and then exposed to 0, 2, 4, 6 and 8 Gy IR. The colonies were counted after 2 weeks. The radiobiological parameters D0, Dq, N, and sensitizing enhancement ratio (SER) were calculated and radiosensitivity was assessed by using a single-hit multitarget model of cell survival.

GS knockdown sensitized tumors to IR in CNE2-R xenografts

To validate the results from cells, we further performed in vivo radiosensitizaion in mouse model by knocking down GS. CNE2-R cells stably expressing GS shRNA or shcon (CNE2-R shGS and CNE2-R shNT) were generated and used to form tumor xenografts in the nude mice. The mice were treated with 2 Gy IR twice in successive two days when volume of transplanted tumor xenografts attained 200 mm³. As expected, knockdown GS in CNE2-R cells significantly sensitized the tumors to IR treatment (Figure 6(a,b)). To confirm that GS regulated CDK1 in vivo, tumor samples generated from IR-treated tumor xenografts were stained with GS, CDK1 Y15 antibody. We observed that GS expression was remarkably decreased in CNE2-R shGS and the level of CDK1 CDK1 Y15 were elevated at the same time (Figure 6(c,d,e)). These data strongly support the notion that high level of GS contributes to radioresistance. Together, our findings indicate that GS promotes Cdc25B-mediated dephosphorylation at Y15 of CDK, leading to efficient formation of CDK1/Cyclin B1 complex that accelerates G2/M transition and promotes radioresistance (Figure 6(f)).

Figure 6.

Knocking down of GS radiosensitized NPC tumor in mice. (a), representative photos from the xenografts at the end of the experiments. (b), Tumor volumes derived from CNE2-R cells with or without knocking GS by shRNA, which received RT or sham irradiation. (c), Representative images of immunohistochemical staining of the xenografts derived from animal experiments. (d) and (e), Summary of immunohistochemical staining with antibodies against GS and CDK1-Y15 (n = 4). (f), The illustration of proposed mechanism whereby GS activates and assembles complex of CDK1 and Cyclin B1 via Cdc25B-mediated CDK1 dephosphorylation at Y15.

Discussion

It have been well established that cell cycle of the cells exposed to IR would halt to coordinate DNA damage repair.^11,14,30 The tumor cells from cancer patients that lack the ability to overcome cell cycle arrest will respond well to IR.^13,14,31,32 However, a portion of patients suffer from resistance to radiotherapy, which characterized by continuous proliferation and tumor relapse after standard radiotherapy.^20,33 Although there has been a wealth of studies that attempt to unveil the underlying mechanisms of radioresistance, none of them can fully accommodate the issues faced by oncologist possibly due to lack of in-depth understanding of radioresistance. In the present study, we focused on radio-response of the cells from NPC, to which radiotherapy has been a first-line treatment in clinical settings. We found that radioresistant NPC cells could overcome the cell cycle arrest, especially at G2/M checkpoint, after a prolonged cell cycle following irradiation, while radiosensitive NPC cells exhibited a persistent cell cycle arrest and higher percentage of apoptosis following IR. It is generally recognized that the therapeutic resistant cells tend to proliferate slower than the sensitive cells regardless the presence of therapeutic stress. This observation is indeed consistent with our finding, in which the CNE2 cells took less time than CNE2-R to go through cell cycle and were more sensitive to IR. These findings indicate that IR-resistant cells might have acquired the ability to more efficiently repair the damaged DNA for G2/M transition than the sensitive cells or to directly proceed to cell cycle in the stage of even incomplete DNA damage repair.

Considering the complexed regulatory network for cell cycle and DNA damage response, we raised questions: what are the changes in the regulatory network of cell cycle checkpoints in radioresistant cells? How can the radioresistant cells “tolerate” high dose of irradiation? To answer these questions, we conducted transcriptional profile analyses by RNAseq and found that GS was the most expressed gene among the key genes in regulation of metabolism and cell cycle. Although glutamine is a secondary essential growth-supporting nutrient besides glucose, the source of glutamine depends more on the extracellular uptake via transporters, such as SLC1A5/3 and SLC7A5.^34–37 Thus, high level of GS in radioresistant cells might indicate that GS plays some non-metabolic roles in mediating radioresistance. In this study, we showed a possible interaction between GS and CDK1. Although our data is not sufficient to demonstrate the necessary of GS-CDK1 interaction for GS function in cell cycle regulation, we at least provided a clue of potential mechanism by which radioresistant cells efficiently overcome G2/M checkpoint to promote activation and assembly of CDK1/Cyclin B1 complex for G2/M transition. This highlights a novel function of GS in radioresistance of NPC and underscores the importance of the finely tuned cell cycle network in radio-response that, once dysregulated, determines cancer response to radiotherapy.

Methods

Cell and cell culture

Human nasopharyngeal carcinoma cell line CNE2 cells were obtained from the Cancer Research Institute of Central South University and cultured in RPMI 1640 medium (Gibco, China) supplemented with 10% fetal bovine serum (Gibco, USA) and 1% penicillin/streptomycin (Invitrogen, USA). Cell cultures were grown in a humidified incubator at 37°C and 5% CO2.

Generation of radioresistant NPC cells

CNE2 cells were exposed to sublethal dose of irradiation (10 Gy), from which the survived cells were cultured, and were exposed to additional sublethal dose of irradiation to produce the next generation of subclone cells. The fifth-generation cells was defined as the radioresistant subclone cell line and named for CNE-2-R.³⁸

Plasmid constructions

GS, CDK1, and Cyclin B1 cDNA were reverse transcribed from CNE2 cells using a SuperScript™ IV First-Strand Synthesis System. GS cDNA was inserted into pRK5-Flag and pLVX-IRES (SBP-Flag); CDK1 cDNA was inserted into pRK5-HA and pLVX-IRES (SBP-Flag); Cyclin B1 cDNA was inserted into pRK5-HA. GS lentiCRISPR vector was conducted by ligation of sgRNA and lentiCRISPR V2 vector from Addgene. All sgRNAs were designed using the website (http://crispr.mit.edu) and produced by Sangon Biotech. All vectors were validated by sequencing or/and immunoblot.

Transfection, lentiviral package and stable cell constructions

Vectors were transfected to cells using a Lipofectamin 2000 Transfection Kit (Life Technology), followed the instructions provided by the manufacturer. For GS knockout cell construction, GS-lenti-CRISPR V2 vector was transfected to CNE2-R by method described above, and selected in 600 μg/mL puromycin after 48 h transfection for 7 days. Single cells were isolated by dilution and validated by genotyping and immunoblot. For GS rescued cell construction, GS-expressing lentivirus or control lentivirus was produced by co-transfecting the plasmids of GS-pLVX-IRES-Puro or empty vector, the packaging plasmids (Gag-Pol and Rev) and the envelope plasmid (VSV-G) to HEK-293T cells following the standard manuals of lentivirus production. Two GS KO cells were infected with high titer GS-expressing lentivirus or control lentivirus. The infected cells were selected by 600µg/ml puromycin for 7 days. Immunoblot analysis was used to assess the protein level of GS in the rescued cells.

Immunoblot, co-immunoprecipitation, and antibodies

For Co-IP test, cells were washed with PBS twice and then lysed in non-denatured IP buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 1 mM EDTA; 0.5% NP-40; 20 mM NaF; 10 mM β-glycerophosphate; 10 mM pyrophosphate; 0.5mM Na3VO4; and 1 mM PMSF proteasome inhibitor cocktail). After incubation on ice for 30 min with occasionally vortex, then the cell lysates were centrifuged at 14,000 g for 15 min at 4°C and supernatants were collected for immunoprecipitation. ANTI-FLAG M2 Affinity Gel (sigma) or SBP beads (NEB) were added into the lysate and incubated on roller shaker at 4°C for 4 h. The precipitated proteins were washed four times with IP wash buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 1 mM EDTA; 0.2% NP-40;) and then subjected to immunoblot analysis. For immunoblotting, cells were lysed directly in the RIPA lysis buffer (Beyotime). Equal amounts of proteins were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes (Bio-Rad Laboratories). The membranes were blocked in 5% milk with 0.1% Tween 20 TBS at room temperature for 1 h, and then incubated with primary antibodies, respectively, overnight at 4°C and then the membranes were incubated with HRP-conjugated secondary antibodies and detected using the ECL Plus kit (Thermo Fisher Scientific). The following antibodies were used in the study: anti-GS antibody (Abcam, 1:1000), anti-CDK1antibody (Cst, 1:1000), anti-CDK1 antibody (phosphor-Tyr15) (cst, 1:1000), anti-Cdc25B (Sangon biotech, 1:1000), anti-FLAG (Sigma, 1:1000), anti-HA (Sigma, 1:1000), anti-α-Tubulin (Santa, 1:1000), anti-Phosphothreonine antibody (Abcam, 1:1000), anti-Phosphoserine antibody (Abcam, 1:1000).

Cell cycle analysis

The cells were fixed in 70% anhydrous ethanol at 4°C overnight and then washed twice with PBS, suspended in PBS and incubated with RNase (50 μg/ml) (Sigma) at 37°C for 1 h, followed by incubation with 50 μg/ml propidium iodide (Sigma) at 37°C for 30 min. The cell cycle profiles were analyzed by flow cytometry (Guava easyCyte), and the data were analyzed using FlowJo software.

Apoptosis detection assay

Cells were harvested after treatment and gently washed twice with PBS. The cells were incubated with Annexin V-FITC and propidium iodide (PI) from the Apoptosis Detection Kit (KeyGEN BioTECH) at room temperature for 15 min. The percentage of positive cells was assessed by the flow cytometry (Guava easyCyte).

Cell counting

Cells were seeded in 24-well plates at a density of 1 × 10⁴ well, and then treated with particular dose of IR. Cells were re-suspended with the method of digestion with 100 ul trypsin and neutralization with 100ul culture medium. 10 ul cell suspension was used to take count of the number of cells. And the proliferation was calculated by the following formula: Proliferation rate = (mean Number _present – mean Number _{first day})/mean Number _{first day}.

CCK8 assay

Cells were seeded in 96-well plates at a density of 5 × 10³/well, and then treated with particular dose of IR. After 2 h incubation with 10 μl CCK8 solution (Bimake), the degree of cell proliferation was detected at 450 nm using a microplate reader (BioTek). And the proliferation rate was calculated by the following formula: Proliferation rate = (mean OD_present – mean OD_{first day})/mean OD_{first day}. The experiments were in triplicates with at least two independent repeats.

Colony formation assay

Cells were seeded into 35 mm petri dishes (350 cells/dish) for 24 h, then were treated with a gradually increasing IR doses, respectively (0 Gy, 2 Gy, 4 Gy, 6 Gy, 8 Gy, and 10 Gy). After 14 days in culture, cells were fixed and stained with 0.3% crystal violet in methanol and the colony number was counted.

Animal experiments

The animal experiment protocol was approved by the Animal Ethics Committee of Xiangya Hospital Central South University. CNE2-R shNT and CNE-R shGS stable cell lines were injected into the subcutaneous tissue of the left flank region of 6-week-old immune-deficient nude mice (BALB/C-nu/nu, SLAC Laboratory). Mice were randomly divided into four groups (n = 4/group): CNE2-R shNT IR-, CNE2-R shNT IR+, CNE2-R shGS IR- and CNE2-R shGS IR+. CNE2-R shNT IR+ and CNE2-R shGS IR+ were treated with 2 Gy IR twice at day 1 and day 2 when tumor size reached about 200 mm³. The mice were sacrificed humanely at day 7 post irradiation and tumor weight was measured, respectively.

Immunohistochemistry

The sections underwent routine dewaxing, dehydration with gradient ethanol, antigen-repair under high pressure for 1.5 min, and cooling under tap water for 10 min. After the remaining tap water on the sections was removed under running water, the sections were added one drop of endogenous peroxidase blocking solution, incubated at room temperature for 10 min, and rinsed with phosphate buffer solution (PBS) (3 min, 3 times). Then, the sections were added with suitable amount of primary antibody for incubation overnight at 4°C, followed by rinsing with PBS after being taken out (3 min, 3 times). The sections were successively incubated with regent 2 and regent 3 (horseradish peroxidase (HRP)-labeled secondary goat anti rabbit monoclonal antibody IgG/(ZSGH-BIO)) at 37°C for 30 min. The sections were rinsed with PBS (3 min, 3 times), colored with 3,3′-diaminobenzidine (DAB), and then rinsed under tap water to terminate the whole reaction. Later, the sections were re-stained with hematoxylin, dehydrated, cleared, and mounted. PBS was used as the negative control instead of the primary antibody. Finally, each section was randomly photographed under a light microscope (at 10× & 40× magnification) to get 3 non-overlapping visual fields. Staining index was determined by multiplying the score for staining intensity with score for positive area. The intensity was scored as follows: 0, negative; 1 weak; 2, moderate; and 3, strong. The frequency of positive cells was defined as follows: 0, less than 5%; 1, 5% to 25%; 2, 25% to 50%; and 4, greater than 75%.

Statistical analysis

All experiments were performed in at least triplicate with mean ± SD subjected analyzed by using a two-tailed unpaired t test between two groups or two-way ANOVA for multivariate analysis. Data analysis was performed using GraphPad Prism 5 software. ∗p < .05, ∗∗p < .01, and ∗∗∗p < .001 were considered to be significant for all of the tests.

Funding Statement

This work was supported by National Natural Science Foundation of China under Grant 81530084, Hunan Key Research and Development Program under Grant 2018SK2123, the Hunan Postgraduate Research Innovation Project under Grant CX2016B059 and China Hunan Provincial Science & Technology Department under Grant 2019JJ40482 and 2019JJ50947.

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Xiangya Hospital Central South University. All animals disposed in the experiment were in accordance with the ethical standards.

Disclosure of potential conflicts of interest

The authors declare that they have no competing interests.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website..