| INTRODUCTIONLung cancer is one of the leading causes for cancer death throughout the world at present. Small cell lung cancer (SCLC) accounted for around15%of all lung cancers1,2. The2-year survival rate is generally less than5%in patients with SCLC and90%of patients often die within5years after final diagnosis. Chemotherapy remains a major treatment for SCLC. Though it shows remarkable sensitivity to chemotherapeutic drugs, SCLC is characterized by high relapse rates and subsequent poor prognosis. The aggressiveness of SCLC could, in part, be due to their intrinsic and extrinsic chemoresistance3,4. Therefore, it is increasingly challenging to understand the molecular mechanism of chemoresistance and develop effective therapeutic strategies to overcome drug resistance of SCLC.Epithelial and endothelial tyrosine kinase (Etk), also known as bone marrow X kinase (Bmx), is one of the important Tec family members of non-receptor tyrosine kinase5"7. It has been reported that Etk/BMX is involved in various biological processes, including proliferation, differentiation, apoptosis and cell migration8-10. Besides, our previous studies found that Etk/BMX participated in the chemoresistance. We knockdown Etk/BMX expression in multi-drug resistant SCLC H69/AR cells by Etk/BMX specific small interfering RNA, leading to increased sensitivity of H69/AR cells to chemotherapeutic drug11. Thus far, the mechanism of chemoresistance mediated by Etk/BMX in SCLC remains to be elucidated.Accumulating studies have revealed that EMT is linked with drug resistance such as in bladder cancer and non-small cell lung caner12>13. The hallmark of EMT is loss of the epithelial markers, such as E-cadherin and P-catenin, and gain of mesenchymal molecules, such as N-cadherin and Vimentin (Vim). The factors containing Snail, Slug, Twist, and so on, which interact with E-box elements, are significant inducers of EMT by repressing E-cadherin expression14,15. Whether EMT plays an important role in drug resistance of SCLC has not been reported yet. We analyzed the data from cDNA microarray in cellular models of SCLC which were widely used as sensitive (H69) and resistant (H69/AR) cell lines to chemotherapy in previous study. The results showed that EMT related markers (Zeb-2, Twist, Vim and P-catenin) were significantly different (>3-fold)(Supplementary figure1). But it remains unclear whether aberrant expression of these EMT-associated genes is directly responsible for chemoresistance in SCLC and regulated by Etk/BMX.It is becoming clear that miRNAs play a pivotal role in the drug resistance16,17. For example, up-regulation of miR-214induces cell survival and cisplatin resistance primarily through targeting the PTEN/Akt pathway in human ovarian cancer’. Over-expression of miR-216a/217in hepatocellular carcinoma cells results in an acquired resistance to sorafenib by activating the PI3K/Akt and TGF-p pathways19. We also found miR-495was differentially expressed and closely associated with chemoresistance of SCLC. Furthermore, the sensitivity of H69/AR cells to anti-cancer drugs greatly increased following transfection with the miR-495mimics20. Besides, the bioinformation analysis implicated that Etk/BMX may be a targeted gene of miR-495. Whether miR-495directly influenced the expression of Etk/BMX should be further confirmed.As mentioned above, we proposed the hypothesis that Etk/BMX involving the chemoresistance of SCLC may be under the regulation of Etk/BMX and through EMT process. To prove this hypothesis, we performed two drug resistant cells H69/AR and H446/AR for a series of assays. Loss-and gain-of-function experiments showed Etk/BMX can regulate cancer cells proliferation, migration, invasion and tumor growth. Moreover, we found that up-expression of Etk/BMX in SCLC specimens correlated with poor pathologic stage and survival time.MATERIALS AND METHODSTissue preparation and Immunohistochemical stainingEighty-six SCLC samples were obtained by biopsy in our department since2005.1-2009.1. All patients had not received chemotherapy. This study was approved by the Ethics Committee of the Southern Medical University. Each case was independently reviewed by two pathologists according to the WHO criteria.Tissues sections were deparaffinized and rehydrated routinely. Before adding the primary antibody, antigen was retrieved by heating sections. After blocking with0.3%H2O2and goat serum, the slides were then incubated with a primary antibody (Etk,1:100; BD Biosciences, USA) at4℃overnight. Following three washings in PBS, an avidin biotin complex (Vector Labs, Burlingame, USA) and alkaline phosphatase mixture were applied. Reaction products were visualized by3’diaminobenzidine (DAB),5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT) and3-amino-9-ethylcarbazole (AEC)(MaximBiotech, Inc CA, USA). All slides were subsequently counterstained with hematoxylin. Brown-yellow granules in cytoplasm were considered positive staining for Etk/BMX. Negative controls were performed by replacing the primary antibodies stated above with PBS.Cell culture and transfectionThe human small lung cancer cell line NCI-H446, NCI-H69and the corresponding drug resistant subline H69/AR were purchased from the AmericanType Culture Collection (ATCC) and maintained in RPMI1640medium supplemented with10%new born bovine serum (GIBCO) at37℃with5%CO2. H446/AR cell had been developed as described previously28. The two resistant cell lines were tested regularly for maintained resistance to the selected drugs. Growth and morphology of all cell lines were monitored on a weekly basis. The plasmid encoding the full-length Etk was kindly provided by Dr Hsing-Jien Kung29. The short hairpin RNA (shRNA) targeted Etk, the small interfering RNA (siRNA) targeted for Zeb-2and Twist, and the mimics and antagomir of miR-495were all supplied by Shanghai GenePharma Co, Ltd. The transfection process was according to the instruction of Lipofectamine2000(Invitrogen). Sequences for RNAs oligo were shown as following:Name Sense sequence (5’-3’) Antisense sequence (5’-3’) miR-495AAACAAACAUGGUGCACUUC GAAGUGCACCAUGUUUGUU mimics UU UUU miR-495AAGAAGUGCACCAUGUUUG antagomir UUU Etk/BMX GGGAAATGGAATCTGGGAAC shl T Etk/BMX GCCGAAGTCAGTGGTTGAAA sh2G Zeb-2si1GGCAAGGCCUUCAAAUAUAT UAUAUUUGAAGGCCUUGCC T TT Zeb-2si2GACCACUCCAGGAGUAAUAT UAUUACUCCUGGAGUGGUC T TT Twist si1GAUGGCAAGCUGCAGCUAUT AUAGCUGCAGCUUGCCAUC T TT Twist si2GCAAGAUUCAGACCCUCAAT UUGAGGGUCUGAAUCUUGC T TTReal-time quantitative reverse transcriptase-PCR (qRT-PCR)Total RNA was extracted using Trizol reagent (Invitrogen), and quantified by NanoDrop2000. qRT-PCR analysis was performed to validate the mRNA expression levels. Reverse transcription reactions were performed for15min at37℃, followed by5s at85℃for complementary DNA synthesis. Real time reactions were performed using the SYBR(?) PrimeScriptTM RT-PCR Kit (Takara Biotechnology Co, Ltd. Dalian, China). Glyceraldehyde3-phosphate dehydrogenase (GAPDH) or U6snRNA was used as the housekeeping gene. The mRNA expression levels of all samples were normalized to the housekeeping gene and analyzed using the comparative expression level2"△△Ct method. The all reactions were performed at least three times. Sequences for primers were shown as following: Name Forward primer (5’-3’) Reverse primer (5’-3’) Etk/BMX CATAGTGGGTTCTTCGTGGAC TGCCCGAGGTATCTTCAGC Zeb-2GGTCCAGATCGAAGCAGCTCA GTGACTATGTTTGTTCACAT AT T Twist CAGTCTTTACGAGGAGCTG GCTTGAGGGTCTGAATCT Vim TGCTAACTACCAAGACACT GTAGGTGCAATCTCAATG β-catenin ACAAACTGTTTTGAAAATCCA CGAGTCATTGCATACTGTCC GAPDH AGAAGGCTGGGGCTCATTTG AGGGGCCATCCACAGTCTT CWestern blot analysisCell proteins were extracted with RIPA lysis buffer and determined by the standard BCA method (BCATM Protein Assay Kit, Pierce, USA). Each protein sample was homogenized in the loading buffer and boiled for5min, then separated on8%SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Etten-leur, Netherlands). The membranes were blocked with5%non-fat dried milk for2h, then treated with primary antibodies including Etk, Zeb-2, Twist, Vim, and β-catenin overnight at4℃. The membranes were washed again with TBST (10mM Tris, pH8.0,150mM NaCl, and0.1%Tween20), followed by incubation with horseradish peroxidase-labeled secondary antibody (Beijing Biosynthesis Biotechnology Co. Ltd.) for45min at room temperature. Finally, after treatment with super echochemiluminescence (ECL) plus detection reagents, the protein bands of membranes were visualized by exposure to X-ray film. The intensity of protein bands were quantified by Quantity One software.In vitro drug resistance assayThe ranges of drug concentrations were based on earlier studies and aimed at obtaining half maximal (50%) inhibitory concentration of a substance (IC50)11.30. A total of three anticancer drugs Cisplatin (DDP; Shangdong, China), Etoposide (VP-16; Jiangshu, China), and Adriamycin (ADM; Jiangshu, China) were obtained from commercial sources and were dissolved according to the manufacturer’s instructions. Anti-cancer drugs induced cell death was quantified using the Cell Counting Kit-8assay (CCK-8). Cells were seeded into96-well plates and then treated for24h in200 μl of medium with anti-cancer drugs. CCK-8reagent (Dojindo, Kumamoto, Japan) was then added and the cells were incubated at37℃for2h before reading the absorbency using a micro-plate reader at450nm. The assay was conducted in six replicate wells for each sample and three parallel experiments were performed.Cell proliferation assayThe cells were seeded in96-well plates at2×103cells/well. At the indicated times (0h,24h,48h,72h, and96h),10μl cck-8and100μl RPMI1640were added to each well. The cells were incubated for2h and absorbance at450nm was measured to calculate cell growth rates. Growth rate=(absorbance at450nm at x h-absorbance at450nm at0h)/(absorbance at450nm at0h).Cell scratch-wound healing assayThe cells were seeded on24-well plates and grown to monolayer. Wound areas were scraped using100μl plastic tips. At the indicated times (0h,12h, and24h), wound areas were photographed and the wound healing rate was calculated. Healing rate=(width of wound at x h-width of the wound at0h)/width of wound at0h.Cell invasion assayUsing24-well transwell units with8μm pore size polycarbonate insert, matrigel (50μl) as a basement membrane was spread on the polycarbonate membrane and allowed to solidify for1h at room temperature. Cells suspended in RPMI1640containing5g/L BSA were added to each upper compartment of the transwell units. After being cultured for48h, cells migrating through the matrigel-coated polycarbonate membrane were fixed by paraformaldehyde, stained with crystal violet and counted in five different fields randomly.Flow cytometry assay for cell apoptosisThe cells treated with or without chemotherapeutic drugs including DDP, VP-16and ADM were harvested, washed with phosphate buffered saline (PBS), and resuspended in binding buffer containing7-AAD for10minutes, followed by the addition of Annexin V-PE. Cell apoptosis analysis was carried out using a flow cytometer (BD Biosciences, Oxford, United Kingdom).In vivo tumorigenicity assayMale nude mice (six weeks old, weighing18-20g, from the Medical Experimental Animal Center of Guangdong Province, China) were used for in vivo assay. The mice were raised under pathogen-free conditions. All procedures were performed according to guidelines of Association for Assessment and Accreditation of Laboratory Animal Care International.Cells were harvested, washed with PBS and re-suspended in normal culture medium at a concentration of1x107/0.1ml. Cells in RPMI1640were subcutaneously inoculated into the legs of nude mice to establish the tumor model, respectively. The tumor volume was determined three times a week by direct measurement with sliding caliper and calculated by the following equation:V=(4/3)×71×R12×R2, where Rl is radius1and R2is radius2and R1<R2. Growth curves of the tumors were constructed. Twenty days later,5mice of each group were sacrificed and tumors were excised. The tumors were homogenized to extract protein for detecting the expression of Etk/BMX, Twist, Zeb-2, Vim and P-catenin. The other mice of each group were continually raised to record the survival time.Luciferase reporter assayThe wild and mutational3’UTR segments of Etk/BMX predicted to interact with miR-495were amplified by PCR from human genomic DNA and inserted into psiCHECK-2vector immediately downstream from the stop codon of luciferase (Promega) to develop psiCHECK2-Etk/BMX-3’UTR and psiCHECK2-Etk/BMX-mut-3’UTR. Cells in24-well plates were transfected with psiCHECK2-Etk/BMX-3’UTR, psiCHECK2-Etk/BMX-mut-3’UTR or psiCHECK-2, respectively. Besides, miR-495mimics and antagomir was also con-transfected into the cells, respectively. Luciferase activity was assayed at48h post-transfection using a dual-luciferase reporter assay system (Promega).Statistical analysisData were analyzed using SPSS20.0software. Results are presented using means±SD. Comparison of means between two samples was performed using Student’s t test. Statistical comparisons of more than two groups were performed using one-way analysis of variance (ANOVA). and then multiple comparisons were performed using least-significant difference (LSD). The association between Etk/BMX expression and clinical features were analyzed by Pearson Chi-Square test. Survival curves were obtained by Kaplain-Meier analysis. In all cases, P<0.05was considered statistically significant.RESULTSRelationship between ETK/BMX expression and clinicopathological characteristics in SCLC patientsTable1summarized the relationship between Etk/Bmx expression and clinical characteristics in SCLC patients. Of the66cases with limited stage small cell lung cancer,42(63.6%) were expressed for Etk/Bmx, while all20(100.0%) cases were positive in patients with extensive stage. The expression of Etk/Bmx revealed obvious correlation with the pathologic stage (P=0,001). Positive expression for Etk/Bmx was more frequent in cases with survival rate<1year (62/76,81.6%) than that in cases with survival rate>1year (0/10,0.0%)(P<0.001). The expression of Etk/Bmx exhibited no significant relationship with gender (P=0.448) and age (P=0.106). The Kaplan-Meier analysis revealed that Etk/Bmx expression was closely associated with overall survival rate of the SCLC patients (Figure1).Etk/Bmx was involved in chemoresistance of SCLC by inhibiting apoptosis induced by chemotherapeutic drugsThe drug resistant cells H69/AR and H446/AR showed greatly elevated mRNA and protein expression of Etk/Bmx compared to their parental cells by qRT-PCR and Western blot assay (Figure2A). To further study the role of Etk/Bmx in SCLC cells, we developed Etk/Bmx stable over-expression cells (H69/Etk and H446/Etk) and down-expression cells (H69/AR/sh and H446/AR/sh)(Figure2B). Then, we analyzed the viabilities of SCLC cells when exposed to chemotherapeutic drugs. The IC50values to chemotherapeutic drugs were greatly decreased in H69/AR/sh and H446/AR/sh cells, while notably increased in H69/Etk and H446/Etk cells (Table2). These data indicated that Etk/Bmx modulated SCLC cells to chemotherapeutic drugs.Subsequently, the number of apoptotic rate was analyzed following the cells treated with chemotherapeutic drugs including ADM, DDP and VP-16using flow cytometry assay. The cell apoptosis was significantly suppressed in H69/Etk and H446/Etk cells. However, it was obviously increased in H69/AR/sh and H446/AR/sh cells (Figure2C and Supplementary figure2). These data provided strong indication that Etk/Bmx was involved in chemoresistance of SCLC by inhibiting apoptosis induced by chemotherapeutic drugs.ETK/BMX promoted cancer cells proliferation, migration, invasion, tumor growth and shortened the survival timeTo investigate the biological roles of Etk/Bmx in SCLC cells, the cell proliferation was assessed by CCK-8assay. Comparing with the results from the control cells, H69/Etk and H446/Etk cells presented a significant increase of the cell growth rates, while H69/AR/sh and H446/AR/sh cells showed obviously decreased cell growth rates (Figure3A). We also detected the effects of Etk/Bmx on migration and invasion ability by scratch-wound healing and transwell-invasion assay. As shown in Figure3B and C, Etk/Bmx over-expression enhanced the healing rates of H446cells. However, knockdown of Etk/Bmx in H69/AR and H446/AR cells inhibited the healing rates (Figure3B). For transwell-invasion assay, more H446/Etk cells penetrated the matrigel-coated membrane than the control H446cells did. By contrast, H69/AR/sh and H446/AR/sh cells showed a notably reduced number of the cells penetrating the matrigel-coated membrane (Figure3C).The tumorigenic properties of Etk/Bmx in vivo were conducted in nude mice xenograft. The nude mice were subcutaneously inoculated with Etk/Bmx over-or down-expressing SCLC cells. The tumors of H69/Etk and H4467Etk cells grew significantly rapidly compared with that of H69and H446cells. But the tumors grew slowly in H69/AR/sh and H446/AR/sh cells. The null mice presented longer survival time accompanied by knockdown of Etk/Bmx, while shorter survival time with elevation of Etk/Bmx (Figure3D).Taken together, both loss-and gain-of-function experiments demonstrated that Etk/Bmx promoted the SCLC proliferation, migration, invasion and tumor growth, but shorten the survival time.Reversing EMT reduced the chemoresistance of SCLCOur previous studies had showed that the EMT related markers (epithelial molecule P-catenin, and the mesenchymal markers including Zeb-2, Twist, and Vim) expressed significantly between H69and H69/AR cells by cDNA microarray (Supplementary figure1). In the current study, we further examined the expression of related markers in H69/AR and H446/AR cells at mRNA and protein level. The results were consistent with the data from cDNA microarray (Figure2A). Zeb-2and Twist have been documented as the important transcriptional repressors of EMT12. Knockdown of Zeb-2and Twist can cause a reversal of EMT21,22. Then, the Zeb-2and Twist expression were inhibited with Zeb-2siRNA or Twist siRNA in H69/AR and H446/AR cells. The expression of Vim was subsequently suppressed while β-catenin was increased accompany with the decreased Zeb-2and Twist. The results indicated that down-regulation of Zeb-2and Twist reversed the EMT process (Figure4). We further investigated whether reversing EMT can influence the drug sensitivity of SCLC cells. The results showed the IC50values of the cells down-expression Zeb-2and Twist were notably reduced when cells were exposed to chemotherapeutic drugs including ADM, DDP and VP-16(Table3). It suggested that the drug resistance to chemotherapeutic drugs was inhibited with reversing EMT in SCLC. Etk/BMX regulated chemoresistance of SCLC through EMTAs mentioned above, the EMT process participated in the regulation of drug resistance of SCLC. We further studied whether Etk/BMX modulated the sensitivity of SCLC cells to chemotherapeutic drugs through EMT. The expression of Zeb-2, Twist, Vim and β-catenin was evaluated after up or down-regulation of Etk/BMX. The results revealed that the expression of Zeb-2, Twist and Vim at mRNA and protein level was remarkably increased while β-catenin was inhibited by over-expression of Etk/BMX (Figure5A). In contrast, suppression of Etk/BMX reduced the expression of Zeb-2, Twist and Vim, but promoted the expression of β-catenin (Figure5B). We also analyzed the expression of the EMT markers stated above in xenograft tumors. The results were consistent with the data from the cell lines (Figure5C).Our observations provided valid evidence that Etk/BMX may influence SCLC chemoresistance through EMT process.miR-495modulated chemoresistance by targeting Etk/BMXThe sensitivity of the H69/AR cells to anti-cancer drugs was greatly increased following transfection with the miR-495mimics as described previously20. To better understand the roles of miR-495in the drug sensitivity of SCLC cells, both H69/AR and H446/AR cells were used to quantify the expression of miR-495by qRT-PCR. The H69/AR and H446/AR cells presented lower level of miR-495than their corresponding control groups. Then, the miR-495expression was up-regulated in H69/AR and H446/AR cells by miR-495mimics and down-regulated in H69and H446cells by antagomir, respectively (Figure6A). Furthermore, the effect of miR-495on the drug sensitivity of SCLC cells was analyzed. We found that H69and H446cells treated with miR-495antagomir exhibited higher IC50values to chemotherapeutic drugs including ADM, DDP and VP-16. In contrast, the IC50values were obviously declined in H69/AR and H446/AR cells by miR-495mimics (Table4). These findings demonstrated that the miR-495was related to chemoresistance of SCLC cells. Considering on Etk/BMX as a potential targeted gene of miR-495by the bioinformation analysis (Figure6C), we examined the Etk/BMX expression after modulation of miR-495by qRT-PCR and Western blot. The result showed that the Etk/BMX mRNA and protein expression were elevated in H69and H446cells by inhibition of miR-495, while decreased in H69/AR and H446/AR cells as a result of up-regulation of miR-495(Figure6B). The correlation between miR-495and Etk/BMX was further detected by luciferase reporter assay. The Etk/BMX luciferase activity was significantly declined in H69/AR cells treated with miR-495mimics, but obviously elevated by treatment with miR-495antagomir (Figure6C).For the reasons mentioned above, all these results suggested that the influence of Etk/BMX on SCLC chemoresistance may be regulated by miR-495.DISCUSSIONIt has been reported that Etk/BMX is associated with various cellular processes including proliferation, differentiation, apoptosis and tumorigenicity8"10. In this study, we investigated the influence of Etk/BMX on the biological behaviors of SCLC by loss-and gain-of-function experiments. The proliferation, migration and invasion abilities were decreased after suppression of Etk/BMX, while they were all elevated following the up-regulation of Etk/BMX. The in vivo data revealed that over-expression of Etk/BMX can promote tumor growth and prolong survival time. By clinical samples, we observed that expression of Etk/BMX showed closely relationship with poor pathologic stage and survival time. In sum, our findings indicated that Etk/BMX played an important role in the biological behaviors of SCLC, suggesting that Etk/Bmx may serve as a predictor for poor prognosis in SCLC.Our previous studies had showed that suppression of Etk/BMX expression reduced the chemoresistance of H69/AR cells. In this study, we further confirmed the effect of Etk/BMX on chemoresistance of SCLC using two multi-drug resistant cells H69/AR and H446/AR cells. The H69/AR/sh and H446/AR/sh cells showed much more sensitivity to chemotherapeutic drugs. In contrast, over-expression of Etk/BMX in H69and H446cells reduced the drug sensitivity. Besides, our research revealed that the cell apoptosis induced by chemotherapeutic drugs decreased significantly after up-regulation of Etk/Bmx, but it increased obviously with knockdown of Etk/Bmx. There was a negative correlation between Etk/BMX and cell apoptosis. These results provided sufficient proof to demonstrate that Etk/Bmx affected the chemoresistance of SCLC by inhibiting apoptosis induced by chemotherapeutic drugs.Studies have shown that EMT process confers resistance to chemotherapy in ovarian carcinoma, colorectal cancer and so on23-25. Base on our previous study, the role of EMT in SCLC chemoresistance was firstly clarified in this study. Inhibition of Zeb-2and Twist in H69/AR and H446/AR cells led to declined Vim while elevated β-catenin. The results indicated that suppression of Zeb-2and Twist can reverse the EMT process. Furthermore, the IC50values to chemotherapeutic drugs were suppressed after knockdown of Zeb-2and Twist. It suggested that EMT was associated with chemoresistance of SCLC. Subsequently, the relationship between Etk/BMX and EMT was explored. The H69/AR/sh and H446/AR/sh cells exhibited a significant decrease in the expression of mesenchymal markers, while increased epithelial markers. However, the H69and H446cells presented an opposite results by elevation of Etk/BMX. The expression of the EMT markers stated above in xenograft tumors was consistent with the data from the cell lines. Taken together, Etk/BMX modulated the chemoresistance of SCLC possibly through EMT process.miR-495has been reported as a tumor suppressor by inhibiting gastric cancer cell migration and invasion26. Ectopic expression of miR-495in breast cancer cells promotes tumorigenesis in the mice27. Our previous study firstly reported that transfection of the H69/AR cells with the miR-495mimics significantly reduced the chemoresistance. In the present study, we used two drug resistant cells to further investigate the role of miR-495in drug sensitivity of SCLC. Up-or down-regulation of miR-495can modulate IC50values to chemotherapeutic drugs, indicating miR-495was involved in the chemoresistance of SCLC cells. Elevation of miR-495resulted in decreased expression of Etk/BMX. On the contrary, Etk/BMX was over expressed following the reduced expression of miR-495. The result from luciferase reporter assay confirmed that Etk/BMX was one of the directly targeted genes of miR-495. All these data presented that miR-495may participate in the chemoresistance and regulate the expression of Etk/BMX in SCLC.Taken together, our study supported for the first time that elevation of Etk/BMX can led to chemoresistance of SCLC by EMT process and was under the regulation of miR-495. Therefore, our research provided a new insight into the mechanism of SCLC chemoresistance. Etk/BMX can be a useful predictor for drug sensitivity, raising the possibility of Etk/BMX depletion as a promising strategy for interfering with chemoresistance in SCLC. |
PDF Full Text Request