MicroRNA-150 enhances radiosensitivity by inhibiting the AKT pathway in NK/T cell lymphoma

We used 5 cell lines as NK/T-cell lymphoma leukemia cell lines, which are commonly showing CD2⁺, sCD3⁻, CD3ε⁺, CD5⁻, CD56⁺, TCRαβ⁻ and TCRγδ⁻ phenotypes, including KHYG-1, NK-92, HANK-1, SNK-1, SNK-6. Of these, HANK-1, SNK-1 and SNK-6 are EBV⁺. Those cell lines were maintained in alpha-MEM (Minimum Essential Medium) supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA) and 100 U/ml recombinant IL-2. NK cells showing the sCD3⁻ and CD56⁺ phenotype were collected from 6 healthy donors as a control using a magnetic cell sorting system and flow cytometry (Additional file 1: Figure S1). All cell lines were cultured at 37 °C in a humidified atmosphere of 5% CO_2.

Paraffin-embedded tissues of 36 primary NKTL were collected from 22 nasal, 6 pharyngeal, 5 gastrointestinal tract and 3 lung, and obtained from the Department of Pathology of Zhujiang Hospital, Southern Medical University, between January 2010 and January 2016. These were assessed according to Ann Arbor Staging [16]. None of the patients received radiotherapy before biopsy sampling. The response of radiotherapy was assessed clinically for primary lesion based on CT (computed tomography) and PET (positron emission tomography)/CT 1 month after treatment according to the following criteria: complete response (CR) was defined as the complete disappearance of all assessable lesions, partial response (PR) was defined as a reduction of the sum of the lesions by 50% or more and no progression of assessable lesions, stable disease (SD) was indicated by a reduction of < 50% or an increase of < 25% in tumor size. All these conditions were required to last for at least 4 weeks with no appearance of new lesions. Progressive disease (PD) was defined as an increase of 25% in tumor size or the appearance of new lesions. The PR, SD and PD define as Non-CR.

Total RNA was extracted from the cells and tissues using TRIzol reagent according to the manufacturer’s protocol. The reverse transcription and polymerase chain reaction (PCR) primers for miR-150 and U6 were obtained from RiBoBio. Reverse transcription primersfor miR-150 were 5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGCACTGGTA-3′. The PCR primers for miR-150 were 5′-ACACTCCAGCTGGGTCTCCCAACCCTTGTACCA-3′ and 5′-CTCAACTGGTGTCGTGGA-3′. The PCR primers for U6 were 5′-CTCGCTTCGGCAGCACA-3′ and 5′-AACGCTTCACGAATTTGCGT-3′. The cDNA library was synthesized using the PrimeScript RT reagent kit. For quantification of mature miRNA, cDNA was generated using specific stem-loop universal primers. qRT-PCR for miRNA and mRNA was performed using SYBR Premix Ex Taq II and was measured using an ABI 7500 Sequence Detection System. U6 was used as the internal control.

The hsa-miR-150-5p gene fragment was synthesized chemically and cloned into lentiviral vector pLVX-ZsGreen-miRNA-Puro. The constructed recombinant plasmid pLVX-ZsGreen-miR-150-5p-Puro/pLVX-ZsGreen-miRNA-Puro, pspax2 and pMD2G was transfected to packaging cell line 293 T, and the packaged virus was determined for titer. Recombinant lentivirus particles harvested and concentrated, then 1 × 10⁴/L NK-92/Hank-1 cells were seeded into 24 well plates, we added the virus suspension into plates by Polybrene concentration gradient dilution method, the infection rate was low (about 30–40%) after 72 h by fluorescence microscope, then GFP-positive cells were selected transfer to 96 well plates for further culture by flow cytometry cell sorting, GFP-positive clones were expanded to construct stable cell line, and tested for expression level of miR-150 by qRT-PCR. The experiment was divided into three groups, including mock group, miR-control group, and miR-150 group.

1 × 10⁶ NK-92 or Hank-1cells of above three groups were irradiated for each well with 0, 2, 4, 6, 8, and 10 Gy, respectively. Cells were subjected to radiation exposure by a 6-Mv X ray (isocentric) linear accelerator. The dose rate was fixed as 300 cGy/min for three groups. The SSD (source to surface distance) was a 100 cm radiation field for 10 × 10 cm, Three days later, cell viability was measured by MTT assay. At least 3 independent experiments were performed.

1 × 10⁶ NK-92 or Hank-1cells were treated with 0, 2, 4, 6, 8, and 10 Gy of irradiation, respectively. Twelve days later, the cells were fixed with methanol and stained using 1% crystal violet in 70% ethanol. The colonies containing 50 or more cells were counted according to our previous study [17] . The survival curve was derived from a multi-target single-hit model, y = 1–1-exp.(−D/D0)n [18]. The radiation sensitivity enhancement ratio (SER) was measured according to the multi-target single-hit model. In some experiments, cells were treated with LY294002 starting 1 h before radiation for a duration of 24 h. At least 3 independent experiments were performed.

1 × 10⁶ NK-92 or Hank-1 cells of above miR-control and miR-150 group for each well were irradiated with 6 Gy, the cells were collected at 48 h after radiation. Annexin V fluorescein isothiocyanate and propidium iodide stains were used to determine the percentage of cells undergoing apoptosis. The apoptosis assay was conducted using the protocol supplied by the manufacturer. Each sample was then subjected to analyses by flow cytometry. At least 3 independent experiments were performed.

The 3′ untranslated region (UTR) of AKT2 and AKT3 containing the putative miR-150-binding site was amplified from genomic DNA via PCR using the following primers: wt-AKT2 F, 5′-GCAGTCTGCCCACGCAGAGG-3′, and Rv, 5′-TACAG ATGGATAGCTAGTTTA-3′, wt-AKT3 F, 5′-GTCTCTTTCATTCTGC TACTTCAC -3′, and Rv, 5′-GTAAAATGCCCTTTAACCCCCGT-3′. Mutation in 3′-UTR of both genes with miR-150 target binding site deleted was generated with the Quick Change Site-Directed Mutagenesis kit. Both the wild and mutant types of AKT2 and AKT3 genes were cloned into the psiCHECK-2 vector immediately downstream of the Renilla luciferase gene. A luciferase reporter construct containing the miR-150 consensus target sequence served as the positive control, and the psiCHECK-2 vector was used as the internal control. Cells were co-transfected with psiCHECK-2 firefly luciferase reporter (50 ng), psiCHECK-2 Renilla luciferase reporter (10 ng), and miR-150 (50 nM) or miR-control (50 nM) with Lipofectamine 2000 reagent. Cell lysates were prepared using passive lysis buffer for 48 h upon transfection, and luciferase activity was measured using the Dual-Luciferase Assay kit and normalized to firefly luciferase activity. Three independent experiments were performed in triplicate.

Cells were collected in ice-cold PBS 48 h after transfection and lysed on ice in cold modified RIPA buffer supplemented with protease inhibitors. The concentration of protein was determined by the BCA Protein Assay Kit. Then equal amount of protein were analyzed by SDS-PAGE. Gels were electroblotted onto nitrocellulose membranes, blocked in PBS/dry milk/0.05% Tween-20, membranes were incubated at 4 °C over night with first antibodies (Monoclonal rabbit anti-human anti-Akt (ab126433), anti-p-Akt (ab8805), anti-caspase 3 (ab136812), anti-AKT2 (ab38513), anti-AKT3 (ab152157), anti-mTOR (ab2732) and anti-p-mTOR (ab84400) were all purchased from Cell Signaling Technology (Beverly, MA, USA), and monoclonal mouse anti-human anti-Alpha-Tubulin (ab7750) and anti-GAPDH (ab8245) were purchased from Kang Chen Bio-tech (shanghai, china), according to the manufacturer’s instructions.

Four-week-old male severe combined immunodeficient mice were injected subcutaneously with 2 × 10⁵ Hank-1-miR-150 or Hank-1-miR-control cells, separately. When the tumor mass became palpable (about 110 mm³), mice in each group were randomly divided into subgroups, including control group, and IR-treated group. Each group contained six mice, and there was no difference in tumor size among the groups. Using this model, we monitored the efficacy of IR (2 Gy/2 days, total 40 Gy) in the miR-150 and miR-control groups, respectively. For the control group, dimethyl sulfoxide alone (100% DMSO, 40 uL, intraperitoneal injection every 2 days) was delivered to the mice. Tumor volume (V) was monitored by measuring the length (L) and width (W) of the tumor with calipers and was calculated with the formula V = (L*W²) *0.5. All the procedures were in accordance with the guidelines of the Animal Ethics Committee of Southern Medical University.

In this study, we detected miR-150 expression levels in NKTL tissues from 36 patients. The expression levels of miR-150 were obviously lower in the tumor tissues than those in normal sCD3⁻ and CD56⁺ NK cells (9.10 ± 0.19 vs 4.01 ± 0.22, P < 0.05). In addition, among five NKTL cell lines examined, four of them (NK-92, SNK-6, HANK-1, SNK-1) had significantly lower expression levels of miR-150 than NK cells obtained from normal peripheral blood. (P < 0.05, Fig. 1).

As shown in Table 1, low expression of miR-150 was closely correlated with the viral load of Epstein-Barr virus (EBV) (EBV ≧ 50 copy/ml severse as positive) (P = 0.007). No significant association was found between miR-150 expression and other clinico-pathological features, such as the patient’s gender, age, clinical stage, and IPI stage (P > 0.05). In our enrolled cases, CR, PR, NC, and PD were achieved in 12, 4, 5, and 15 patients at the evaluation time, respectively. Moreover, miR-150 expression was also the factor that showed a positive correlation with treatment response in our enrolled cases, in which low expression of miR-150 was observed more frequently in patients from the non-CR subset than those in the CR subset (P = 0.000).

To explore the underlying mechanism involved in the role of miR-150 in the treatment response of NKTL cells, we induced miR-150 expression upon lentiviral infection in NK-92 and Hank-1 cells, which had the lowest levels of miR-150 among our available NKTL cell lines. Increased expression of miR-150 was confirmed by qRT-PCR (Fig. 2a). In the MTT cell proliferation assay, there were no difference between three groups from two cell lines (P > 0.05). we tested whether miR-150 increased NKTL cell sensitivity to radiotherapy in vitro. NK-92 and Hank-1 cells were treated with various clinical doses of IR. As shown in Fig. 2c, IR was readily cytotoxic to NK-92 and Hank-1 cells. However, IR resulted in a higher level of cell death in miR-150 group compared to the miR-control and mock group (P < 0.05), however, there were no differences between the miR-control and mock group (P > 0.05). In addition, we applied colony formation assay by GraphPad Prism 5 software to further confirm that miR-150 increased the sensitivity of NKTL cells to IR in vitro. NKTL cells formed fewer colonies in miR-150 groups compared to other group (Fig. 2d), according to the linear quadratic model-fitted cell survival curve. The corresponding biological parameters of the survival curve are shown in Table 2. SERD0, SERDq, and SERSF₂ were 1.36, 2.050, and 1.064, respectively. These results provided evidence that miR-150 could increase the sensitivity of NKTL cells to radiotherapy in vitro.

To test whether the miR-150-induced sensitivity to IR was due to activation of apoptosis, we examined the effects of miR-150 on apoptosis in NKTL cell lines using flow cytometry. As shown in Fig. 3a, overexpressions of miR-150 significantly promoted the radiation-induced apoptosis of NK-92 cells [(39.8 ± 1.04)% vs (54.2 ± 1.12)%, P < 0.05] and Hank-1 cells [(39.8 ± 1.24)% vs (56.2 ± 1.04)%, P < 0.05] for Annexin V/PI analysis when treated with IR (10 Gy). As illustrated in Fig. 3b, the levels of cleaved caspase-3 [(0.16 ± 0.02 vs 0.40 ± 0.01,P < 0.05)] and cleaved PARP [(0.19 ± 0.03 vs 0.42 ± 0.01, P < 0.05)] in Hank-1 cells and the cleaved caspase-3 [(0.10 ± 0.02 vs 0.19 ± 0.01,P < 0.05)] and cleaved PARP [(0.14 ± 0.03 vs 0.28 ± 0.01, P < 0.05)] in NK-92 cells were increased in miR-150 group compared to the miR-control group after treated with IR (10 Gy) by western blot analysis. These data revealed that miR-150 promotes IR-induced apoptosis in NKTL cells.

We searched for miR-150 targets using three publicly available databases (TargetScan, Pictar and miRBase). We then used WB analysis to assess expression in various candidate targets, including c-Myb, FOXP1, AKT2, AKT3, BCAP, PRL1, c-Raf, MSK1, MIB1 and E2F3 in NK-92 and Hank-1 cells transduced with miR-150 or empty vector. The PI3K signaling pathway is usually activated in lymphoma, and inactivation of this pathway impairs DNA repair following radiation, then, we found that levels of both AKT2 and AKT3 were reduced in all miR-150 transductants (Fig. 4d). To demonstrate the direct inhibition by miR-150, luciferase reporter containing wild-type AKT2 and AKT3 3’UTR sequences or their mutant-derivates with deletion of putative miR-150-binding sites were cotransfected into both NKTL cell lines with miR-150 (Fig. 4a). As shown in Fig. 4b, miR-150 suppressed luciferase activities depending on the presence of miR-150 binding in the wild-type AKT2 and AKT3 3’UTR. Upon 3’UTR mutation of AKT2 and AKT3, miR-150 can not bind to its target site, resulting in increased luciferase activity. The ratio of wild-type AKT2 and AKT3 were decreased upon miR-150 overexpression in Hank-1 and NK-92 cell lines (Fig. 4c). Morever, overexpression of miR-150 significantly suppressed the protein expression of AKT2 and AKT3 in NK-92 and Hank-1 cells using WB analysis (Fig. 4d), All these results together suggested that AKT2 and AKT3 direct targets of miR-150 in NKTL cells.

mTOR is the downstream molecular of AKT signaling pathways. We found that overexpression of miR-150 reduced the expression of AKT meanwhile decreased the expression of p-mTOR by WB analysis. Moreover, we found that exposure to the tyrosinekinase inhibitor LY294002 activating the effect of miR-150 resulted in decrease of p-AKT/p-mTOR expresssion and formed colonies after different dose of radiation. There were statistical difference between the four groups in NK-92 cell (Χ² = 11.26 P = 0.021) and Hank-1 cell (Χ² = 10.38 P = 0.016) after 10 Gy radiation, and the total clone formation rate was minimum in the miR-150 + LY294002 group according to mean rank (Fig. 5a and b). Those function experiment confirmed that AKT pathway was the mechanism by which miR-150 sensitizes cells toward radiotherapy.

Freshly isolated Hank-1-miR-control and Hank-1-miR-150 cells were transplanted into NODSCID-12 IL-2 RG−/− (NSI) mice(purchased from Guangzhou Institutes of Biomedicine and Health (GIBH)) to establish P1 NKTL mice according to protocols [19]. We next examined the anti-tumor effect of miR-150 on NKTL xenografts treated with IR. The tumors were irradiated with 2 Gy every 2 day (total dose = 40 Gy) after tumor mass became palpable, followed by detection of NKTL xenograft growth (Fig. 6). It can be seen that the tumor volume was significantly decreased in the radiotherapy group (from day 10 to 30, P < 0.01) and combined therapy group (from day 10 to 30, P < 0.01) compared with control group. The tumor volume of control group showed no significant change compared with miR-150 group (from day 5 to 20, P > 0.05). The average tumor volume reached 1587.96 ± 83.56 mm³ in control group on day 30, while only 449.58 ± 29.20 mm³ in combined therapy group, 1448.02 ± 96.99 mm3 in miR-150 therapy group, and 1012.95 ± 78.77 mm³ in radiotherapy group. In addition, the tumor volume was significantly decreased in mice of combined therapy group compared with miR-150 therapy (from day 10 to 30, P < 0.01) or radiotherapy group (from day 15 to 30, P < 0.01).