Overall microRNA expression landscape in SCLC identifies miR-1 as a tumor suppressor gene in SCLC
MicroRNAs or non-coding RNAs are important molecules that regulate the expression pattern of multiple genes, including tumor suppressors and tumor promoters [27, 28]. Therefore, apart from identifying protein targets, microRNAs also provide a means to identify molecules modulating the various aspects of cancer and putting forward the potential of identifying tumor suppressor microRNAs as a useful strategy to develop or identify novel therapeutic agents. To understand the expression profile of microRNAs in SCLC, we performed miRNA sequencing from serum samples of eight SCLC patients and six healthy donors. The advantage of using serum samples is the secretory nature of miRNAs, which exist in exosomes and argonaute (AGO) proteins in circulation/serum that can be easily detected using noninvasive techniques [22, 29, 30]. We observed a differential expression pattern of miRs (65 up/13 down) in SCLC/donor serum samples, and interestingly, we observed a significant downregulation of miR-1 in the SCLC patients compared to healthy donors (Fig. 1A-C, supplementary Fig. S1A-C). To further understand miR-1 expression in SCLC and to validate the observation of miRNA-Seq, we also performed a bulk RNA-Seq analysis in the tumor tissues of SCLC patients and normal lung tissues (GSE19945). We observed a substantial downregulation of miR-1 in the SCLC tumors compared to normal lung tissues (Fig. 1D).
Further, to analyze the miR-1 expression in SCLC, we developed an advanced and highly sensitive nanoprobe-based biosensing assay that performs the absolute quantification of miR-1 in serum or cell line samples (Fig. 1E). For this, we synthesized and characterized a DNA-probe-functionalized Au-nanoprobe (DNA-AuNPs) that can be used to detect miR-1 from cell-derived total RNA or directly from serum or plasma. The synthesis and characterization are provided in supplementary information and supplementary figures S2-S3. Given that most of the miRNAs exist with AGO2 protein forming a RNA-induced silencing complex (RISC) and exosomes [22, 30], we performed a gold nanoprobe-mediated in situ (RNA isolation free) detection of miR-1 in the serum samples of human SCLC patients and healthy donors (Fig. 1F). Notably, to enhance the accuracy of serum miRNAs assay, a protein denaturation step was performed using a short treatment of the denaturation buffer on the serum followed by heat inactivation and removal of excess proteins (for details, see supplementary information) to facilitate the availability of miRNAs for biosensing purpose. Interestingly, we observed that miR-1 expression was drastically decreased in the serum samples of SCLC compared to healthy donors (Fig. 1F).
In addition, to determine the clinical significance of miR-1 expression, we performed in-situ hybridization on SCLC tissue microarray (Cat#, BS04116a, US Biomax, Inc. Derwood, MD), representing the pathology grade tumor tissues of 55 cases/100 cores for SCLC and ten cores for normal lung tissues. Using a miR-1 specific probe, the in-situ hybridization results validated the observations of miRNA-Seq and miR-1 serum profiling data that showed decreased expression of miR-1 in SCLC tumor tissues compared to normal lung tissues (Fig. 1G). Further, to determine miR-1 expression in SCLC cell lines, we performed a Taq-man-based miR-1 expression assay using a panel of SCLC cell lines (SBC3, SBC5, DMS273, NCI-H526, NCI-H82, DMS53, NCI-H69, and Colo668) and non-tumorigenic lung epithelial cells (BEAS-2B). Consistent with the bulk gene expression data, serum miRNAs sequencing, and nanoprobe-mediated expression profile of serum miR-1, a decrease to the null expression of miR-1 was observed in SCLC cell lines compared to non-tumorigenic lung epithelial cells (Fig. 1H). Together, our results suggest that downregulation of miR-1 accompanying SCLC, and consistent with the previous studies in other cancers [20, 31], establishes miR-1 as a tumor suppressor in SCLC.
miR-1 overexpression decreases the growth and metastasis of SCLC cells, whereas miR-1 inhibition potentiates cell growth and migration of SCLC cells
To investigate the role of miR-1 in the growth and metastasis of SCLC cells, we infected miR-1 expressing cell line (SBC3) with lentivirus containing miR-1Zip-GFP that was used to develop stable knockdown in the SBC3 cell line. For making a stable miR-1 overexpression model, we infected two miR-1 null SCLC cell lines (SBC5 and NCI-H69) with lentivirus that constitutively expresses mCherry and doxycycline (DOX) inducible miR-1 and luciferase expression. We have selected the positive clones using puromycin selection and single-cell sorting using GFP (for miR-1Zip) or mCherry (for DOX-On-miR-1) as markers. The DOX-On-miR-1 cells were further infected with lentivirus expressing rtTA having hygromycin as a selectable marker. In all the experiments, lentiviral-infected cells were maintained under the selection of the drug (puromycin/hygromycin) suitable for the corresponding lentiviral vector.
Before starting the functional studies, we first determined the miR-1 expression using miR-1 specific Taq-man qRT assay in SBC5-DOX-On miR-1 cells and observed that these cells showed significant upregulation in miR-1 expression following the 24–48 h induction with DOX compared to non-induced (-DOX-off) SBC5 cells (Fig. 2A). To monitor the effect of miR-1 on SCLC cells, SBC5-DOX-On miR-1 cells were plated in a low attachment cell culture plate and allowed to form spheres. We monitored and analyzed the growth of spheres formed by SBC5 (-DOX-off) Vs. SBC5-DOX-On miR-1 cells, and as expected [31, 32], once we induced miR-1 expression in SBC5 cells using DOX-On miR-1, it suppressed the sphere formation ability of SBC5 cells (Fig. 2B). In a complementary set of experiments on SBC3 and SBC3-miR-1Zip cells, it was found that miR-1 sponging (via miR-1Zip) supported the sphere-forming ability of SBC3 cells compared to parental SBC3 cells (Fig. 2C). In another variation of cell growth experiments, we also examined the role of miR-1 in colony formation by SCLC cells and found that miR-1 overexpression significantly decreased the colony formation in SBC5 cells, whereas SBC3-miR-1Zip cells had a higher colony-forming ability compared to parental cells (Fig. 2D-E).
Next, to see the impact of miR-1 on the migration properties of SCLC cells, we employed a live cell monitoring approach using IncuCyte coupled live-cell imaging. We seeded the SBC5(-DOX-off)/SBC5-DOX-On-miR-1 and SBC3/SBC3-miR-1Zip cells in a 96-well plate and simultaneously created a homogenous wound using the wound creating apparatus. We followed wound closure ability of SBC5(-DOX-off)/SBC5-DOX-On-miR-1 (without or with miR-1) and SBC3/SBC3-miR-1Zip cells in real-time for 72 h and found that miR-1 decreased the wound closure ability of SBC5 cells, and miR-1 sponging enhanced the wound closure or migratory properties of SBC3 cells as revealed by live-cell monitoring, scratch wound assay, and transwell migration studies (Fig. 2F-G & supplementary Fig. S4 A-D). We were then interested in discerning the apoptotic potential of miR-1 in SCLC cells (SBC5 and NCI-H69). For this, the DOX-On-miR-1 SBC5 and NCI-H69 cells were plated, and miR-1 expression was induced (+ DOX) for the different time periods (24–72 h), and apoptosis quantification was performed through flow cytometry-based annexin-V/PI staining. We observed that miR-1 overexpression induced apoptosis-like morphological changes in the SCLC cells (supplementary Fig. S5A-B). Further, we found a time-dependent significant increase in the percent apoptosis of miR-1 induced SBC5 and NCI-H69 cells compared to their -DOX-off controls (Fig. 2H-J). Thus, overall results showed that miR-1 decreased the sphere formation and cell migration properties of SCLC cells and induced apoptosis; on the other hand, miR-1 sponging enhanced sphere-forming and cell migration abilities of SCLC cells.
miR-1 modulates tumor growth and metastasis in SCLC xenografts
To probe deeper into the role of miR-1 in SCLC growth and metastasis, we performed tumorigenesis and metastasis studies with luciferase labeled SBC3 and SBC3-miR-1Zip cells using subcutaneous and intracardiac xenograft models in immunodeficient NOD/SCID gamma (NSG) mice (Fig. 3A). For tumorigenesis, the SBC3-luciferase and SBC3-miR-1Zip-luciferase cells were subcutaneously injected into the right flank of NSG mice (n = 6 per group with equal number of male and female mice) and monitored for tumor growth using IVIS-imaging (Fig. 3B). We found that the growth of SBC3-miR-1Zip xenografts was significantly higher compared to SBC3-parental cells suggesting that miR-1 sponging enhanced the tumorigenesis of SBC3 cells (Fig. 3C-E).
To determine the effect of miR-1 sponging on the metastatic potential of SCLC cells, we injected SBC3-luciferase and SBC3-miR-1Zip-luciferase cells intracardially in NSG mice, and we observed higher metastases in animals injected with SBC3-miR-1Zip cells (Fig. 3F). Remarkably, significant metastasis was observed in the liver, lung, intestine, ovary (female mice), and adrenal glands of mice injected with SBC3-miR-1Zip-luciferase cells (Fig. 3F-G). Lung, liver, ovary, and adrenal glands are the most common metastatic sites for SCLC [33, 34]. Further, IHC staining analysis of SBC3/SBC3-miR-1Zip-xenografts tumors with the proliferative and angiogenesis marker (Ki67 and CD31, respectively) revealed high Ki67 and CD31 staining in SBC3-miR-1Zip tumor tissues compared to SBC3 parental (Fig. 3H). These results suggest that miR-1 inhibition drives a highly metastatic SCLC phenotype in an intracardiac xenograft mouse model.
Next, to elucidate the impact of miR-1 overexpression on SCLC metastasis, we performed intracardiac injection of -DOX-off/DOX-On-miR-1 SBC5 cells (highly metastatic and relatively low miR-1 expression) in NSG mice (both male and female). Following two weeks of intracardiac injections, mice were randomized into (+)miR-1/DOX-On and (-)miR-1/-DOX-off groups (Fig. 4A). The metastatic lesions were monitored using IVIS-imaging (Fig. 4B &C). Consistent with the metastatic phenotype of SBC5 cells, intracardiac xenografts showed substantial metastasis, with a high metastasis to most common SCLC metastatic sites, including, lung, liver, brain, bone, ovary, adrenal, stomach, intestine, pancreas, and spleen (Fig. 4C-F). On the contrary, exogenous miR-1 overexpression markedly decreased the metastatic potential and growth of intracardially injected SBC5 cells (Fig. 4C-F). In context to most frequent metastatic sites of SCLC including lung, liver, brain, and bone [33, 34], it was observed that the intracardiac injections of SBC5 cells mimicked the metastatic properties of SCLC and validated our metastatic model to study SCLC metastasis. The key observations from the metastasis studies, clearly showed that miR-1 has the potential to decrease SCLC metastasis at the most frequent metastatic sites of SCLC, such as lung, liver, brain, and bone (Fig. 4F). In addition to the primary metastatic sites of SCLC, the endocrine organs such as the adrenal and ovary (in the case of female mice) were also reported as recurrent metastatic sites of SCLC [34–36]. The observation from metastasis studies suggests that the miR-1 overexpression decreased the adrenal and ovarian metastasis of SBC5 cells (Fig. 4C & F). Interestingly, tissue histology studies using H&E staining further suggest that miR-1 overexpression decreased the tumor burden in the most frequent metastatic sites of SCLC such as lung, liver, and brain (Fig. 4G, supplementary Fig S6-S7). Together, these findings suggest that miR-1 overexpression significantly inhibits SCLC growth and metastasis.
miR-1 modulates the CXCR4-FOXM1-RRM2 axis in SCLC
Having established the antitumorigenic potential of miR-1 in SCLC, we next examined the molecular mechanism or molecules involved in the accomplishment of the tumor-suppressive role of miR-1. We first performed a deep RNA sequencing (RNA-Seq) of SBC3, SBC3-miR-1Zip, SBC5(-DOX-off)/SBC5-DOX-On-miR-1 cells. Since SBC3 cells have a low or basal expression of miR-1 and other cells have no miR-1 expression (Fig. 1H), we performed a differentially expressed gene (DEG) analysis in SBC3, SBC3-miR-1Zip, and SBC5 cells (supplementary Fig. S8). Upon comparative analysis of the top 50 DEG in SBC3, SBC3-miR-1Zip, and SBC5 cells, we found that CXCR4 was the most differentially upregulated gene in SBC3-miR-1Zip and SBC5 cells (supplementary Fig. S8 & S9A).
CXCR4 has been previously implicated in the growth and metastasis of various cancers, including SCLC [37, 38]. Previous studies have shown that CXCR4 mediates cancer cell adhesion, migration, chemoresistance, and metastasis, and small molecule or peptide inhibitors of CXCR4 such as AMD3100, TF14016, and LY2510924 decreased tumor growth and metastasis [39–41]. In the second set of RNA-seq analysis, we performed a comparative analysis of SBC5(-DOX-off)/SBC5-DOX-On-miR-1 and SBC3/SBC3-miR-1Zip cells to assess the DEG in the presence of high miR-1 (SBC5-DOX-On-miR-1) and sponged or low miR-1 (miR-1Zip). We observed a marked downregulation of SCLC-specific gene clusters (CXCR4, RRM2, FOXM1, CCNB2, CEP55, PLK1, AURKA, AURKB, and CCNB1) in SBC5 cells overexpressing miR-1 (Fig. 5A). Interestingly, we observed upregulation or enrichment of similar gene sets (CXCR4, RRM2, FOXM1, CCNB2, CEP55, PLK1, AURKA, AURKB, CCNB1) in SBC3-miR-1Zip cells compared to SBC3 (Fig. 5B). In addition, we observed marked enrichment of CDH1 (gene encoding E-cadherin) in miR-1 overexpressing SBC5 cells, and consistent downregulation in SBC3-miR-1Zip cells (Fig. 5A-B).
Next, to investigate the tumor-suppressive mechanism of miR-1, we also performed clustering and interaction analysis of DEG (Fig. 5C-D). The clustering of DEG among miR-1 overexpressing and sponging groups suggested the existence of two to three clusters (Fig. 5C-D), solid lines connecting two members showed direct interaction within a cluster, whereas dotted line showed the interaction between the members of two clusters). Clustering analysis showed that in the first cluster, CXCR4 is shown to interact with AKT or CDH1, and in the other cluster, FOXM1 is represented as a central node for interaction with other DEG (Fig. 5C-D). FOXM1 has been previously implicated in SCLC tumorigenesis . Thus, in addition to CXCR4 (the candidate gene identified in the first set of RNA-seq analysis), the clustering analysis identified FOXM1 as a second candidate gene co-operating with CXCR4, and with RRM2 a prominent downstream target of FOXM1 (Fig. 5C-D, supplementary Fig. S10-S12A). RRM2 is known to regulate DNA-damage response, cancer aggressiveness, and drug resistance, and recent studies have validated FOXM1 as a RRM2-directing transcription factor [43, 44].
We further assessed the expression status of CXCR4, FOXM1, and RRM2 in the SCLC cell line panel (SCLC-NCI, SCLC-UTSW, SCLC CCLE-Broad-MIT, SCLC GDSC-MGH-Sanger, SCLC CTRP-Broad-MIT, and SCLC Global) from the CellMiner-SCLC (https://discover.nci.nih.gov/SclcCellMinerCDB/) and observed a positive correlation in CXCR4, FOXM1, and RRM2 expression (Fig. 5E, supplementary Fig. S10-S11). In most of the SCLC cell lines, a high expression of CXCR4, FOXM1, and RRM2 showing a positive correlation, was observed (Fig. 5E, supplementary Fig. S10-S12).
To better understand the transcriptional changes caused by miR-1 inhibition or overexpression, we performed protein expression studies by immunoblotting and IHC analyses of excised tumor tissues and the results further validated the overexpression of key targets observed in the RNA-seq/clustering analysis, including (1) high CXCR4 in SBC3-miR-1Zip tumor tissues compared to SBC3 parental tumor tissues (supplementary Fig. S13B), suggesting higher metastatic signaling that was confirmed in in-vitro CXCR4 surface expression analysis and oncogenic signaling (Fig. 5F, and 6); (2) high CXCR4 mediated signaling was also validated through CXCL12 treatment in SBC3 and SBC3-miR-1Zip cells, where a comparatively high colony formation, cell migration, and AKT activation was observed in CXCL12 treated SBC3-miR-1Zip cells compared to CXCL12 untreated or parental SBC3 cells (supplementary Fig. S4 C-F); (3) high levels of proliferation, EMT, and angiogenesis-related markers (Ki67, Zeb-1, snail, and CD31) in SBC3-miR-1Zip cells and tumor tissues (Fig. 3H, Fig. 6J); (4) increased expression of FOXM1 and RRM2 in SBC3-miR-1Zip cell lines and xenografts, which was validated in protein expression studies, and direct association of FOXM1 and RRM2 was validated through a ChIP-assay (Fig. 6, supplementary Fig. S13B, S14A). To explore the impact of miR-1 on the cell surface expression of CXCR4, we performed a flow cytometry-based cell surface expression analysis of CXCR4 in SCLC cells using APC-Cy-7-anti-CXCR4 (Fig. 5F). Remarkably, it was observed that miR-1Zip increased the cell surface expression of CXCR4 in SBC3 cells (Fig. 5F). On the other hand, exogenous expression of miR-1 using the DOX-ON-miR-1 system in SBC5 and NCI-H69 cells decreased the cell surface expression of CXCR4 (Fig. 5F). These experiments demonstrate that miR-1 decreases cell surface expression of CXCR4 in SCLC cell lines. Overall, our findings suggest that miR-1-mediated modulation of CXCR4 regulates oncogenic signaling and may drive the FOXM1-RRM2 axis in SCLC.
miR-1 directly target CXCR4 and alters FOXM1 accessibility to RRM2 promoter
We next sought to identify and characterize the miR-1 targeting site in the 3'-untranslated region (3'-UTR) of CXCR4. To this end, we retrieved and cloned the 3'-UTR of CXCR4 that contained a miR-1 binding site (position 265–272 of 3'-UTR, identified using TargetScan) in the pGL3-luciferase vector (for details, see supplementary information). In another set of 3'-UTR primers, we mutated the residues of the miR-1 binding site and cloned them into the pGL3-luciferase vector. A dual-luciferase assay was performed using miR-1 mimic with wild type or mutant 3'-UTR-pGL3 constructs in SBC5 cells (as these cells have no miR-1) and found a decreased luciferase activity in SBC5 cells transfected with wild-type-3'-UTR-pGL3/miR-1 mimic compared to wild-type-3'-UTR-pGL3/scramble miR control (Fig. 6A-B). In contrast, no change in luciferase activity was observed in SBC5 cells transfected with mutant-3'-UTR-pGL3 /miR-1 mimic and mutant-3'-UTR-pGL3/scramble miR control (Fig. 6B).
To further analyze whether miR-1 directly modulates the RRM2-promotor targeting of FOXM1 we used a ChIP-coupled qPCR assay (for details see supplementary information). We found that miR-1 overexpression in SBC5 cells (DOX-On-miR-1 SBC5) decreased the enrichment of FOXM1 with RRM2-promoter (Fig. 6C), whereas an enhanced binding or enrichment of FOXM1 to the RRM2-promoter has been observed in SBC3-miR-1Zip cells compared to SBC3 parental cells (Fig. 6D). Altogether, the CXCR4 cell surface expression analysis (Fig. 5F), 3'-UTR targeting, and FOXM1-RRM2 ChIP-qPCR experiments demonstrate that miR-1 targets CXCR4 and significantly decreases the accessibility of FOXM1 to RRM2-promoter, implying that miR-1 targets the CXCR4/FOXM1-RRM2 axis in SCLC.
miR-1 targets CXCR4/FOXM1/RRM2 axis, and pharmacological inhibition of CXCR4 and FOXM1 phenocopy miR-1 efficacies
Next, we aimed to validate our transcriptional findings in SCLC cell line models and xenograft tumor tissues through protein expression studies. Consistent with the RNA-Seq data, it was observed that miR-1 sponging increased the expression of FOXM1 and RRM2 in SBC3 cells compared to parental cells (Fig. 6E). Interestingly, the ectopic expression of miR-1 in DOX-On-miR-1 -SBC5 and/or -NCI-H69 cells consistently decreased the expression of FOXM1 and RRM2 (Fig. 6F). In concordance with CXCR4 cell surface expression studies (Fig. 5F), these observations suggested that miR-1 targets CXCR4/FOXM1/RRM2 axis in SCLC.
To further establish whether CXCR4/FOXM1/RRM2 axis is a viable target of miR-1, we took advantage of AMD3100 and FDI-6, the well-characterized and highly specific small molecule inhibitors of CXCR4 and FOXM1, respectively [45, 46]. The proposed model of utilizing AMD3100 and FDI-6 has been shown in Fig. 6G. As expected, it was found that AMD3100 treatment resulted in the decreased expression of FOXM1 and RRM2 in SBC3 and SBC3-miR-1Zip cells, and the treatment of cells with FDI-6 also decreased the expression of FOXM1 and RRM2 (Fig. 6H, supplementary Fig. S14B). We followed the treatment of AMD3100 and FDI-6 in SBC5 cells and compared the expression profile of FOXM1 and RRM2 in these cells with that of SBC5 cells that ectopically overexpressed miR-1 (DOX-On-miR-1 -SBC5). Remarkably, it was observed that the treatment of SBC5 cells with either AMD3100 or FDI-6 decreased the expression of FOXM1 and RRM2, and phenocopy the consequence of miR-1 overexpression concerning FOXM1 and RRM2 expression (Fig. 6I). Additionally, to validate the relationship of FOXM1/RRM2 axis in SCLC cells, we performed genetic perturbation of FOXM1 using small interfering RNA (siRNA)-mediated knockdown in SBC5 cells and found that knockdown of FOXM1 decreased the expression of RRM2 (supplementary Fig. S14C). Furthermore, we have analyzed the expression or activation of survival and metastasis-related markers (p-AKT/AKT, p-ERK/ERK, Snail, and Zeb-1), and observed that miR-1 inhibition enhances the activation of AKT and ERK and increases the expression of snail and Zeb-1 in SBC3 cells (Fig. 6J). The overexpression of miR-1 in SBC5 and NCI-H69 cells decreases the expression of snail, and the activation of AKT and ERK (Fig. 6K). Collectively, these results suggest that miR-1 modulates CXCR4/FOXM1/RRM2 axis in SCLC, and ectopic expression of miR-1 inhibited the expression of metastasis-associated proteins (snail, Zeb-1) and activation of AKT and ERK.
To determine whether CXCR4/FOXM1/RRM2 axis or the expression pattern that we observed in RNA-Seq and in vitro SCLC models follow a similar pattern in intracardiac xenografts in vivo models or not, we performed expression analysis of CXCR4, FOXM1, and RRM2 using IHC on metastatic liver tissues (as the liver is the most frequent metastatic site of SCLC). Metastatic liver tumors revealed a high expression of CXCR4, FOXM1, and RRM2 in the tissues excised from the no-miR-1 group (SBC5-DOX-off) of mice, whereas the majority of liver tissue sections from the + miR-1 group (SBC5-DOX-On-miR-1) of mice showed very low to no expression of CXCR4, FOXM1, and RRM2 (Fig. 7A). The pathological quantification of liver tissues stained with CXCR4, FOXM1, and RRM2 further confirmed that the expression of these markers significantly decreased in the + miR-1 group (Fig. 7B). We repeated the IHC staining of CXCR4, FOXM1, and RRM2 in the liver tissues from SBC3 and SBC3-miR-1Zip intracardiac xenografts and observed intense staining of these proteins at the liver metastatic sites developed from SBC3-miR-1Zip cells, and in contrast, no staining was observed in SBC3 injected group (Fig. 7C-D). Taken together, the IHC-analysis supports that miR-1 directs the CXCR4/FOXM1/RRM2 axis in SCLC.
SCLC xenografts derive greater survival benefits from high miR-1
Given that miR-1 loss aggravates SCLC growth and metastasis, whereas exogenous overexpression of miR-1 decreased the metastatic proficiencies of SCLC cells (Fig. 3–4), we next asked whether miR-1 has any effect on the overall survival of mice bearing SCLC tumors. Therefore, to explore the clinical implications of miR-1 in terms of survival, we again performed intracardiac injections of DOX-On-miR-1 SBC5 cells in NSG mice and looked for the establishment of metastasis using IVIS-imaging. The animals were randomized into (+)miR-1/+DOX and (-)miR-1/-DOX groups. Strikingly, miR-1 expression significantly prolonged the survival of mice standing intracardiac SCLC xenografts compared to the no miR-1 group (Fig. 7E). The estimated median survival was nearly 65 days for the + miR-1 group, whereas it was only 34 days for the without miR-1 group (Fig. 7E). Next, we explored the possibility of detecting miR-1 levels in the serum of mice intracardially injected with SCLC cells using the DNA-AuNPs nanoprobe-mediated detection assay we developed (Fig. 1E). Interestingly, a significantly high miR-1 level was observed in the serum samples of mice showing prolonged survival (Fig. 7F). Collectively, this data shows that miR-1 provided survival benefits in vivo as it promotes the survival of mice injected with SCLC cells and decreased tumorigenesis and metastasis.