If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Microbiology, Tumor and Cell Biology (MTC), Karolinska Institutet, Stockholm, SwedenWhitehead Institute for Biomedical Research, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
Department of Dermatology, University of Turku and Turku University Hospital, Turku, FinlandMediCity Research Laboratory, University of Turku, Turku, Finland
Department of Dermatology, University of Turku and Turku University Hospital, Turku, FinlandMediCity Research Laboratory, University of Turku, Turku, Finland
Unit of Dermatology and Venereology, Department of Medicine, Karolinska Institutet, Stockholm, SwedenDepartment of Medicine, Karolinska University Hospital, Stockholm, Sweden
Unit of Dermatology and Venereology, Department of Medicine, Karolinska Institutet, Stockholm, SwedenDepartment of Medicine, Karolinska University Hospital, Stockholm, Sweden
Cutaneous squamous cell carcinoma (cSCC) is the second most common cancer and a leading cause of cancer mortality among solid organ transplant recipients. MicroRNAs (miR) are short RNAs that regulate gene expression and cellular functions. Here, we show a negative correlation between miR-203 expression and the differentiation grade of cSCC. Functionally, miR-203 suppressed cell proliferation, cell motility, and the angiogenesis-inducing capacity of cSCC cells in vitro and reduced xenograft tumor volume and angiogenesis in vivo. Transcriptomic analysis of cSCC cells with ectopic overexpression of miR-203 showed dramatic changes in gene networks related to cell cycle and proliferation. Transcription factor enrichment analysis identified c-MYC as a hub of miR-203–induced transcriptomic changes in squamous cell carcinoma. We identified c-MYC as a direct target of miR-203. Overexpression of c-MYC in rescue experiments reversed miR-203–induced growth arrest in cSCC, which highlights the importance of c-MYC within the miR-203–regulated gene network. Together, miR-203 acts as a tumor suppressor in cSCC, and its low expression can be a marker for poorly differentiated tumors. Restoration of miR-203 expression may provide a therapeutic benefit, particularly in poorly differentiated cSCC.
). cSCC is a particular problem in organ-transplant recipients, in whom its frequency is increased 65- to 250-fold, and the tumors are more aggressive, with a 30-fold worse survival rate (
). cSCC arises primarily on sun-exposed skin because of accumulation of somatic mutations leading to the activation of oncogenic pathways and the inactivation of tumor suppressor genes (
). However, the exact molecular mechanisms remain to be explored.
Recent works have delineated a class of endogenous small noncoding RNAs called microRNAs (miRNAs), which can regulate gene expression by binding to the 3′-untranslated region (3′UTR) of their target mRNAs, resulting in the translational inhibition or target mRNA degradation (
). Depending on the set of targets that they regulate, miRNAs may act as oncogenes (oncomiRs) or tumor suppressors, and because of their ability to regulate multiple targets simultaneously, they are attractive therapeutic targets for cancer treatment (
). We recently showed that miR-203 acts as a tumor suppressor in the most common keratinocyte-derived human malignancy, basal cell carcinoma (BCC), and that local delivery of miR-203 mimics can inhibit tumor growth in vivo (
In this study we investigated the expression of miR-203 in cSCC and showed an inverse correlation between differentiation grade of cSCC and miR-203 expression level. Enforced miR-203 overexpression suppressed cell proliferation, scratch-wound healing, cell migration, cell invasion, colony formation, and the angiogenesis-inducing ability of cSCC cell lines. In vivo, subcutaneous injection of miR-203–overexpressing cSCC cells into NOD scid gamma mice resulted in decreased tumor size and vessel density. We identified the c-MYC proto-oncogene as a direct target of miR-203 in cSCC. Taken together, these results indicate a tumor suppressive effect of miR-203 in cSCC and imply that miR-203 mimics have potential as a therapeutic modality in miR-203–deficient cancers.
Results
MiR-203 is down-regulated in poorly differentiated cSCC
Recent reports have shown that miR-203 is involved in skin morphogenesis and promotes epidermal differentiation by repressing stemness of keratinocytes (
). This observation prompted us to investigate the expression of miR-203 in a keratinocyte-derived tumor, cSCC. To this end, we performed in situ hybridization analysis using locked nucleic acid probes to determine the expression levels and distribution of miR-203 in 10 healthy and 40 cSCC skin samples including tumors of World Health Organization grades I–III (Figure 1). The differentiation markers keratin-10 and involucrin showed a gradual decrease in tumors with increasing grade (less differentiated) (Figure 1a). In situ hybridization showed expression of miR-203 in suprabasal epidermal layers of healthy skin with comparable levels in well-differentiated cSCCs. In poorly differentiated cSCCs, however, the expression of miR-203 was lower. Semiquantitative scoring of in situ hybridization showed a gradual decrease in mean score with increasing grade: the mean score of miR-203 staining in grade III tumors (poorly differentiated) was significantly decreased compared with healthy skin samples, whereas tumors of grades I (well-differentiated) and II (moderately differentiated) displayed trends in miR-203 down-regulation (Figure 1a, and see Supplementary Figure S1 online). Quantitative real-time PCR (qPCR) validation on an independent set of cSCC samples (grade I, n = 20; grade II, n = 20; and grade III, n = 15) showed significant reduction of miR-203 expression level in grade III tumor samples compared with healthy skin and well-differentiated tumor samples (Figure 1b). miR-203 expression was also determined in a panel of cSCC cell lines, showing that cSCC cell lines have significantly decreased expression of miR-203 compared with primary normal human epidermal keratinocytes (Figure 1c). Together, the decreased expression of miR-203 correlates with the differentiation grade of tumors, suggesting a biomarker potential and functional role for miR-203 in cSCC.
Figure 1miR-203 is down-regulated in cSCCs. (a) Expression of epidermal differentiation markers (K10 and involucrin) and miR-203 was analyzed by immunofluorescence staining and in situ hybridization (ISH), respectively, in healthy human skin (n = 10) and squamous cell carcinomas (n = 40). Dashed lines demarcate tumor-stroma boundary. Scale bar = 100 μm. In situ hybridization score is shown as a box plot. (b) qPCR analysis of miR-203 in healthy skin (n = 13) and squamous cell carcinomas: grade I (n = 20), grade II (n = 20), and grade III (n = 15). (c) Expression level of miR-203 in cSCC cell lines (UT-SCC) compared with primary NHEKs. ∗P < 0.05, ∗∗∗P < 0.001, Student t test. cSCC, cutaneous squamous cell carcinoma; Epi, epidermis; K, keratin; miR, microRNA; NHEK, normal human epidermal keratinocytes; qPCR, quantitative PCR; Str, stroma; Tu, tumor.
), was found to be up-regulated in cSCC (see Supplementary Figure S2 online) and thus may contribute to reduced miR-203 expression.
MiR-203 regulates cancer-associated gene networks in cSCC and inhibits cell-cycle transition from G1 to S phase
To obtain further insights into the function of miR-203 in cSCC, we performed transcriptomic profiling of a metastatic cSCC cell line (UT-SCC-7) transfected with miR-203 mimics or scrambled oligonucleotides (ODNs). Gene set enrichment analysis of the profiling data showed significant enrichment of genes in several well-defined biological processes known to be involved in carcinogenesis (see Supplementary Figure S3 online). Among 33,297 annotated genes, 596 genes were significantly altered in miR-203–overexpressing cells (377 down-regulated and 219 up-regulated) (see Supplementary Table S1 online). Gene ontology term analysis showed significant alteration in the expression of genes related to cell cycle, cell proliferation, and cell motility in miR-203–overexpressing cells compared with scrambled ODN–transfected cells (Figure 2a, and see Supplementary Figure S4 online). Several of the down-regulated genes have previously been shown to promote skin carcinogenesis/angiogenesis, for example, proliferating cell nuclear antigen (PCNA), MYC, JUN, hepatocyte growth factor (HGF), EGFR, TP53, suppressor of cytokine signaling (SOCS3), and IL8 (
) (see Supplementary Figure S5 online). Moreover, gene set enrichment analysis showed enrichment within the profiling data for the reported cell proliferation gene ontology term (see Supplementary Figure S6a online). Down-regulation of selected cell cycle- and cell proliferation-related genes by miR-203 overexpression is summarized as a heat map in Figure 2b.
Figure 2MiR-203 suppresses proliferation in cSCC cell lines. (a) Gene ontology term analysis of deregulated genes detected in miR-203–overexpressing UT-SCC-7 cells by Affymetrix whole transcript expression analysis (Affymetrix, Santa Clara, CA). (b) Heat map showing differentially expressed genes related to cell cycle and cell proliferation. Predicted target genes of miR-203 are labeled in red. (c) UT-SCC-7 and A431 were transfected with a synthetic precursor molecule for miR-203 (miR-203) or scrambled ODNs as negative control (scramble). After 48 hours, cell proliferation and cell cycle progression were measured by EdU labeling. (d) Expression of Ki67 proliferation marker in UT-SCC-7 and A431 after miR-203 transfection was determined by qPCR. The inlet showed immunofluorescence staining for Ki67. Scale bar = 50 μm. (e and f) Two- and three-dimensional (soft agar) colony formation assay of UT-SCC-7 and A431 transfected with miR-203 mimic. Optical density of crystal violet-stained colony is shown as bar chart. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, Student t test. cSCC, cutaneous squamous cell carcinoma; h, hours; miR, microRNA; ODN, scrambled oligonucleotide; qPCR, quantitative PCR; Scr, scramble.
Simultaneous regulation of an unusually large number of cell cycle-/proliferation-associated genes in cSCC cells prompted us to test the hypothesis that miR-203 regulates cell proliferation in cSCC. miR-203 reduced the amount of newly synthesized DNA, as indicated by a significantly decreased percentage of 5-ethynyl-2 deoxyuridine (EdU)-incorporating cells in two human cSCC cell lines, UT-SCC-7 and A431. Analysis of cell cycle distribution in miR-203–overexpressing cSCC cells showed that miR-203 overexpression enforced arrest at the G1 phase of the cell cycle, as shown by an accumulation of cells in the G1 phase and a reduction of the number of cells entering S phase (Figure 2c, and see Supplementary Figure S6b). Furthermore, the expression level of Ki-67, a proliferation marker, in miR-203–overexpressing cells was diminished at both the mRNA and protein levels, as determined by qPCR and immunofluorescence staining (Figure 2d). Consistent with these results, Western blot analysis showed the suppression of another cell proliferation marker, PCNA, by miR-203 (see Supplementary Figure S6c). Suppression of proliferation in miR-203-overexpressing cells was also observed by the CyQuant cell proliferation assay (Life Technologies, Carlsbad, CA) (see Supplementary Figure S5d). The low baseline level of miR-203 in SCC cell lines rendered depletion experiments unfeasible in this model system; however, inhibition of miR-203 in primary human keratinocytes led to increased proliferation (see Supplementary Figure S7a online). Moreover, inhibition of miR-203 suppressed calcium-induced differentiation in keratinocytes, as judged by expression level of involucrin (see Supplementary Figure S7b and c).
Long-term cell proliferation and self-renewal ability was investigated by two-dimensional colony formation assay, and a significant reduction in the number of colonies was found in miR-203–overexpressing cells (Figure 2e). Anchorage-independent cell growth was assayed by soft agar colony formation assay and massively decreased colony-forming ability was detected in miR-203–overexpressing cells (Figure 2f). These results clearly show that miR-203 suppresses the proliferative and self-renewal abilities of cSCC cells.
MiR-203 suppresses cell motility and invasiveness in SCC
Next, we studied the effect of miR-203 overexpression on migration of cSCC cells in a scratch wound assay and found that miR-203 suppressed the rate of scratch wound closure in UT-SCC-7 and A431 cell cultures (Figure 3a). The effect of miR-203 on migration and invasion of UT-SCC-7 and A431 cells was additionally determined by Transwell (Corning Life Sciences, New York, NY) cell migration and invasion chamber, respectively. A significant reduction in the number of migrating and invading cells upon miR-203-overexpression in both UT-SCC-7 (Figure 3g) and A431 (Figure 3c) cells was observed. The average number of migrating cells per captured field was decreased from 352 ± 50 and 35 ± 5 cells in scrambled oligonucleotide-transfected cells to 225 ± 32 and 9 ± 1 cells after miR-203 mimic transfection (P < 0.05) in UT-SCC-7 and A431 cells, respectively. The number of invading cells in miR-203–overexpressing UT-SCC-7 and A431 cells was 80 ± 7 and 77 ±10 cells per captured field, approximately 47% and 62% reduced compared with scrambled controls. These results show that miR-203 suppresses the motility and invasiveness of cSCC cells.
Figure 3miR-203 suppresses migration and invasion of cSCC cells. (a) UT-SCC-7 and A431 transfected with miR-203 mimic or scrambled ODNs were subjected to scratch assay. Images were taken at 0, 7, 11, and 24 hours after the monolayer was scratched. Depicted are representative images and indications of the migration distances. ImageJ software (National Institutes of Health, Bethesda, MD) was used to determine the migration distance. Percentage of healing, which represents random cell migration, was calculated and is presented as a graph. (b and c) Transwell (Corning Life Sciences, New York, NY) migration and Matrigel (Corning Life Sciences) invasion assay of miR-203 mimic– or scrambled ODN–transfected UT-SCC-7 and A431. Data shown are means of cell counts obtained from three separate wells, and error bars represent standard deviations. ∗P < 0.05, ∗∗P < 0.01, Student t test. cSCC, cutaneous squamous cell carcinoma; h, hours; miR, microRNA; ODN, scrambled oligonucleotide.
), and our transcriptomic profiling of miR-203–overexpressing cSCC cells showed down-regulation of several genes encoding growth factors and cytokines, which have been reported to play a role in angiogenesis (including VEGF, FGF, PDGF, and FST) (Figure 4a) (
). Therefore, we next investigated the role of miR-203 in cSCC angiogenesis. Using TaqMan qPCR (Applied Biosystems, Carlsbad, CA), we found that miR-203 significantly suppressed the levels of IL8, VEGFA, and CCL2 mRNAs in both UT-SCC-7 and A431 cells (Figure 4b). Consistent with the qPCR data, the amount of secreted IL-8 was decreased in the supernatants of miR-203–overexpressing UT-SCC-7 and A431 cells (Figure 4c). To find out whether changes at the mRNA and protein levels translate into biological function, we investigated the effect of miR-203 on the angiogenesis-inducing capacity of cSCC cells by the human umbilical vein endothelial cells tube formation assay. The number of well-organized nodes formed by Matrigel (Corning Life Sciences, New York, NY)-overlaid human umbilical vein endothelial cells was substantially decreased upon exposure to conditioned medium collected from miR-203–overexpressing cSCC cells (Figure 4d). This suggets that miR-203 could affect tumor growth not merely by cell-intrinsic mechanisms but also by affecting tumor-driven angiogenesis by modulating the level of pro-angiogenic factors secreted by tumor cells to the microenvironment.
Figure 4Inhibition of angiogenesis-inducing ability of cSCC by miR-203. (a) Heat map showing angiogenesis-related genes deregulated in miR-203–overexpressing cells. Putative miR-203 targets predicted by TargetScan algorithm (http://www.targetscan.org/vert_71/) are labeled in red. (b) qPCR analysis of VEGF-A, IL-8, and CCL2 mRNA levels in UT-SCC-7 and A431 cells transfected with miR-203 mimic or scramble oligo. (c) IL-8 level of supernatant collected from miR-203 mimic or scrambled ODNs transfected UT-SCC-7 and A431 cells detected by ELISA. (d) HUVECs tube formation assay was performed by resuspending HUVEC cells in supernatant collected from UT-SCC-7 and A431 cells transfected with miR-203 mimic or scrambled ODNs at 48 hours after transfection. HUVEC cell suspension was seeded into a 48-well plate with Matrigel (Corning Life Sciences, New York, NY) layer at the density of 4.5 × 104 per well. After 16–20 hours of incubation, photos were taken, and the number of well-organized nodes was counted. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, Student t test. h, hours; HUVEC, human umbilical vein endothelial cell; miR, microRNA; ODN, scrambled oligonucleotide; qPCR, quantitative PCR.
We next sought to determine the molecular basis for the observed effects of miR-203 in cSCC cells. Enriched transcription factors whose targets are overrepresented in the differentially expressed gene in miR-203–overexpressing cSCC cells were identified. c-MYC, one of the most frequently activated oncogenes in human cancer, which has previously been implicated in cSCC pathogenesis (
), was among the top-listed transcription factors in addition to c-Jun, another miR-203 target (Figure 5a). Moreover, 3′UTR of c-MYC was found to contain an miR-203–binding site (Figure 5b). To experimentally validate this miRNA-to-mRNA interaction, we first performed luciferase assays using a pMIR-REPORT luciferase construct (Applied Biosystems, Carlsbad, CA) containing wild-type c-MYC 3’UTR. As shown in Figure 5b, miR-203 significantly decreased c-MYC 3′UTR reporter activity compared with scrambled ODNs. The inhibitory effect of miR-203 was abolished when the predicted miR-203–binding site was mutated, showing a sequence-specific interaction between miR-203 and MYC 3′UTR.
Figure 5MiR-203 directly targets c-MYC. (a) MetaCore (Thomson Reuters, New York, NY) analysis of genes regulated by miR-203 in miR-203–overexpressing cSCC cells. (b) c-MYC 3′UTR and MUT c-MYC 3′UTR reporter were constructed, and luciferase reporter assay was performed. (c) Western blot analysis of c-MYC and β-actin proteins in UT-SCC-7 cells transfected with miR-203 mimic or scrambled ODNs. (d) Expression level of c-MYC oncoprotein was analyzed in the section of healthy skin and cSCC samples by immunofluorescence staining. Scale bar = 200 μm. (e) c-MYC transcription factor activity of UT-SCC-7 cells transfected with miR-203 mimic was determined by TransAM c-MYC kit (Active Motif, Carlsbad, CA). (f) Expression level of proliferation-related MYC effectors in UT-SCC-7 cells transfected with miR-203 mimic or scrambled ODNs as determined by qPCR. (g) Restoration of c-MYC could override miR-203 effect on cell cycle. (h) Expression level of MYC and a well-known effector, CCND1, was measured in UT-SCC-7 double-transfected with miR-203/pcDNA or miR-203/pcDNA-MYC by qPCR. c-MYC protein level was determined by Western blot analysis. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, Student t test. A, adenine; C, cytosine; cSCC, cutaneous squamous cell carcinoma; Epi, epidermis; G, guanine; h, hours; miR, microRNA; MUT, mutated; ODN, scrambled oligonucleotide; PD, poorly differentiated squamous cell carcinoma; qPCR, quantitative PCR; Str, stroma; Tu, tumor; U, uracil; UTR, untranslated region; WD, well-differentiated squamous cell carcinoma; WT, wild type.
Moreover, overexpression of miR-203 led to markedly decreased c-MYC protein levels in cSCC cells (Figure 5c). A reverse correlation between miR-203 and c-MYC expression was shown in tumor tissues from cSCC with different differentiation grades, with higher expression of c-MYC in poorly differentiated cSCC compared with well-differentiated cSCC (Figure 5d).
Next, we aimed to determine the downstream processes by which miR-203–mediated c-MYC regulation can affect cell proliferation in cSCC. c-MYC transcription factor activity assay of nuclear fraction showed decreased c-MYC transcription factor activity in miR-203–overexpressing cells (Figure 5e). Not only c-MYC–dependent DNA-binding activity, but several proliferative signaling-related MYC effectors including cyclin D1 (CCND1), CAD, CDK1, and ODC1, were significantly suppressed by miR-203 in UT-SCC-7 cells as determined by qPCR (Figure 5f).
These results show that c-MYC is a direct target of miR-203 in cSCC and suggest that decreased expression of miR-203 in cSCC may contribute to the derepression of the c-MYC oncogene in poorly differentiated/high-grade cSCC, thereby contributing to the progression of tumorigenesis.
To define the relative importance of c-MYC on cSCC phenotypes mediated by miR-203, UT-SCC-7 cells were transfected with an miR-203 mimic together with a control expression plasmid (pcDNA) or a c-MYC expression plasmid (pcDNA-MYC). Using the EdU assay, we found that the G1 arrest induced by miR-203 was rescued in the cells co-transfected with pcDNA-MYC (Figure 5g). In c-MYC–restored cells, both the number of cells entering S phase and the number of EdU-incorporated cells was increased compared with control cells. Overexpression of c-MYC in transfected cells was confirmed by qPCR (Figure 5h). Cyclin D1, one of c-MYC effectors that drives cell cycle progression, was also induced in the c-MYC–restored cells (Figure 5h). Double transfection of primary human keratinocytes with miR-203 inhibitor and siRNAs targeting cMYC resulted in neutralization of the cell proliferation-promoting effect from miR-203 depletion (see Supplementary Figure S7a). These results suggest that miR-203–mediated suppression of cell proliferation, at least in part, is mediated by the suppression of the c-MYC gene network in cSCC.
MiR-203 suppresses cSCC growth and angiogenesis in vivo
Because miR-203 suppressed several cancer phenotypes of cSCC in vitro, we set out to study the role of this miRNA in cSCC tumor growth and neoangiogenesis in vivo. To this end, UT-SCC-7 cells with stable miR-203 overexpression were established by lentiviral transduction and were subcutaneously injected into the flank of immune-deficient NOD scid gamma mice. Overexpression of miR-203 in tumors from miR-203–overexpressing cells was confirmed by qPCR (see Supplementary Figure S8 online). MiR-203 overexpression led to a significant reduction in tumor volume (see Figure 6a and b). Notably, the volume of tumors formed by miR-203–overexpressing cells was significantly reduced from 2 weeks after injection, and they were significantly smaller even after 4 and 5 weeks. Average tumor weight was reduced from 0.92 ± 0.34 g in the scramble UT-SCC-7 injected group to 0.64 ± 0.12 g in the miR-203–overexpressing UT-SCC-7 injected group (Figure 6c). Western blot analysis of protein lysate from harvested tumors showed decreased c-MYC levels in tumors from miR-203–overexpressing cSCC cells (Figure 6d). Immunofluorescence staining showed decreased expression of miR-203 targets c-MYC and c-Jun, a previously identified miR-203 target (
), as well as proliferation marker Ki67, in the tumors derived from miR-203–overexpressing cSCC cells (Figure 6e). Comparison of the transcriptomic profiles of miR-203–overexpressing and control tumors showed that MYC-target genes were significantly enriched among the transcripts regulated in miR-203–overexpressing tumors (see Supplementary Figure S9 online). The significant reduction of CCND1 and ODC1 in miR-203–overexpressing tumors was shown by qPCR (Figure 6f). Our in vitro angiogenesis assay results showed that miR-203 can regulate the expression and production of several angiogenesis-related genes. In accordance with this, we observed a significant enrichment of angiogenesis- and inflammation-related gene sets among genes differentially expressed in miR-203–overexpressing tumors (see Supplementary Figure S9) and decreased vasculature, as shown by staining for CD36 (Figure 6g). Taken together, miR-203 is associated with loss of differentiation in cSCC, it regulates a proliferation- and angiogenesis-related gene network at least partly through c-MYC, and it can suppress tumor growth and neoangiogenesis in vivo.
Figure 6miR-203 suppresses tumor growth in xenograft model. (a) Stable miR-203 or scrambled ODNs overexpressing UT-SCC-7 cells was established. Ten million scramble or miR-203–overexpressing cells were injected into the left and right flank of each mouse, respectively (n = 6). (b) Tumor width and length were measured by caliper, and tumor volume was calculated as (width2 × length)/2. (c) When tumor volume reached 1,000 mm3 or mice conditions reached humane endpoint, tumors were harvested and weighed. Mean ± standard deviation for each group is indicated in the table. (d) Western blot analysis for the expression of MYC protein in tumors from scrambled ODNs or miR-203–overexpressing cells. (e) Immunofluorescent staining of c-MYC, c-Jun, and Ki67 in tumors. (f) Expression level of proliferation-related MYC effectors in the harvested tumors was determined by qPCR. (g) Immunohistochemical analysis of the endothelial marker CD36 (left) and quantification of the stained area (mean ± standard deviation of stained area) by ImagePro Plus software (Media Cybernetics, Rockville, MD) (right). Scale bar = 100 μm. ∗P < 0.05, ∗∗P < 0.01, Mann-Whitney U test. H&E, hematoxylin and eosin; miR, microRNA; ODN, scrambled oligonucleotide; qPCR, quantitative PCR; Scr, scramble; Wk, week.
The differentiation grade of cSCC is one of the most important determinants of clinical behavior of the tumor and marks an important biological difference between low- and high-risk tumors (
). Here we report a negative correlation between miR-203 expression and cSCC differentiation grade, with significant down-regulation of this miRNA in poorly differentiated cSCC. We show that miR-203 suppresses cSCC proliferation, migration, invasion, angiogenesis, and in vivo tumor growth, at least partially through targeting c-MYC. Inverse correlation of miR-203 expression with differentiation grade in human cSCC is in line with the role for miR-203 as a differentiation-promoting and stemness-suppressing miRNA in keratinocytes (
), but its role in cSCC has remained unclear, especially because it was not found to be differentially expressed when the miRNA expression profile of cSCC (from all grades) was compared with healthy skin.
Integration of the concept of miRNAs into the transcriptional network in cancer is more complex than one miRNA targeting a single gene; instead, dozens to hundreds of target genes in multiple pathways are regulated by each miRNA across the transcriptome (
). miR-203 is not an exception, and its virtue as a putative anticancer therapeutic is based on the fact that it can target multiple cancer genes simultaneously. Various genes have been reported to be directly targeted by miR-203, for example, SOCS-3 and p63 during human fetal skin development in keratinocytes (
). Gene set enrichment and transcription factors enrichment analyses could detect the fingerprint of several of its direct targets on cSCC transcriptome, such as JUN and p63; however, MYC was predicted with a higher score, suggesting its importance in the miR-203 gene network specifically in the context of SCC. In vitro, we showed that miR-203 can suppress proliferation in cSCC cells. A mouse xenograft experiment further showed suppression of the tumorigenic potential of cSCC by miR-203. These results conform to a number of previous studies that reported the inhibition of in vivo tumor formation by miR-203 in other types of cancers (
In addition to inhibiting proliferation, miR-203 also suppressed the motility, invasiveness, and angiogenesis-inducing ability of cSCC cells. Inhibition of cell migration and invasion by miR-203 has previously been reported in other types of cancers through several target genes such as SNAI2, ZEB2, LASP1, NUAK1, and SPARC (
). It could also suppress the scratch-wound healing rate in primary human keratinocytes (see Supplementary Figure S10 online). Tumor growth and angiogenesis suppressor roles for miR-203 by directly targeting vascular endothelial growth factor-A, a well-known angiogenesis regulator, has also been shown in cervical cancer (
We identify the c-MYC proto-oncogene as a direct target for miR-203 in cSCC. Our results suggest that miR-203 functions as the upstream regulator of c-MYC, evidenced by the suppression of c-MYC oncoprotein expression and transcriptional activity upon miR-203 overexpression. Suppression of c-MYC expression by miR-203 was also observed in primary human keratinocytes (see Supplementary Figure S11 online). Several downstream effectors of c-MYC, which have been reported to be associated with poor survival and highly metastatic cSCCs (e.g., CCND1 and CDK1) (
), were also suppressed in miR-203–overexpressing cSCC cells. In vivo, down-regulation of c-MYC gene signature and reduction of Ki67 level were also observed in miR-203–overexpressing cSCC xenografts.
c-MYC restoration in miR-203 mimic–transfected UT-SCC-7 cells reversed the growth inhibitory effect of miR-203, as indicated by the rebound of EdU-incorporated cell numbers and percentage of cells in S phase. This finding, together with our results showing decreased expression of c-MYC and its target genes in miR-203–overexpressing SCC xenografts in vivo, suggests that miR-203–mediated suppression of tumor growth is at least partially mediated through c-MYC, even though the effect of restoration of c-MYC or other targets in miR-203–overexpressing xenograft tumors should be assessed in future experiments. The cyclin D1 gene, an oncogene that acts as a growth factor sensor to integrate extracellular signals with cell cycle machinery, was also found to be up-regulated in c-MYC rescued cells. It is known that c-MYC collaborates with transforming growth factor-α, EGFR, Ras, phosphatidylinositol 3 kinase/protein kinase B, and NF-κB in part via coordination in regulation of cyclin D1 gene expression, because cyclin D1 is a common downstream effector of these growth pathways (
). Coordination of c-MYC with cyclin D1 or its upstream activators not only accelerates tumor formation but may also drive tumor progression to a more aggressive phenotype.
Proto-oncogene c-MYC is an attractive therapeutic target in cancer because of its high frequency of deregulation in a wide range of tumor types, affecting multiple cancer phenotypes (
). Preclinical data have proven the efficacy of antisense oligodeoxynucleotides targeting MYC and several small molecules attacking MYC targets such as ODC, CAD, and Bcl2 in the treatment of malignancies (
). MYC gains and amplifications are frequent cytogenetic abnormalities in cSCC, and activation of c-MYC in adult suprabasal epidermis rapidly triggers proliferation and disrupts differentiation of postmitotic keratinocytes in transgenic mice (
), was also observed in tumors formed from miR-203–overexpressing UT-SCC-7 cells. It is well known that c-MYC and c-Jun have important roles in cell growth and differentiation, not only during normal growth but also in the development of neoplasia (
In conclusion, we propose that miR-203 functions as a tumor suppressor miRNA in cSCC. We showed an inverse correlation between the miR-203 level and the differentiation grade of cSCC and showed that miR-203 suppresses in vitro cancer hallmarks and in vivo tumor growth of cSCC cells. Regulation of c-MYC gene network was identified as a key underlying mechanism for the tumor suppressive effect of miR-203 in cSCC. On the basis of strong evidence for the potential of miRNA therapeutics in cancer, MRX34, a double-stranded RNA mimic of tumor suppressor miR-34, encapsulated in a liposomal nanoparticle formulation is the first microRNA mimic to enter phase I clinical trials (
), and the miR-203/c-MYC pathway presents a potential candidate target for cSCC treatment. Our study identifies miR-203 as a regulator of multiple cancer hallmarks in cSCC and suggests that restoration of miR-203 is a potential approach for cSCC treatment.
Materials and Methods
Clinical samples
Paraffin-embedded skin samples from patients with cSCC were obtained from the Karolinska University Hospital Biobank, Stockholm, Sweden. The diagnosis was confirmed and differentiation grade was assessed by histopathological evaluation by a dermatopathologist. Control skin samples (4-mm punch biopsy samples) were taken from healthy volunteers after written informed consent. The study was approved by the Regional Committee of Ethics.
Further details of the cell culture conditions, reagents, transfection, in situ hybridization, qPCR, in vitro functional assay, luciferase reporter assay, Western blot, immunostaining, MYC restoration, and in vivo xenograft experiment are available in Supplementary Materials and Methods online.
We are grateful to Mari-Anne Hedblad for assistance with the clinical sample collection. We also would like to thank the core facility at Novum Bioinformatics and Expression Analysis, which is supported by the board of research at the Karolinska Institute and the research committee at the Karolinska hospital. The study was supported by the Swedish Research Council (VR 2015-02844), the Swedish Cancer Society (CAN 2012/730 and CAN 2015/694), the Swedish Society of Medicine (Svenska Läkaresällskapet), the European Skin Research Foundation, the Welander and Finsens Foundation, the Tore Nilssons Foundation, the Lars Hierta Memorial Foundation, the Sigurd and Elsa Golje Memorial Foundation, the Stockholm County Council, the Finnish Cancer Research Foundation, Sigrid Jusélius Foundation, and Turku University Hospital EVO grant (project 13336).
32121-2&id=gr1.jpg){kind=link}
32121-2&id=gr2.jpg){kind=link}
32121-2&id=gr3.jpg){kind=link}
32121-2&id=gr4.jpg){kind=link}
32121-2&id=gr5.jpg){kind=link}
32121-2&id=gr6.jpg){kind=link}