Histograms of log₂ of mean expression ratios of miRNA levels in WM1552C to that in the untransformed melanocyte cell line HEM-l (control) are plotted as histograms. Asterisks indicate the respective levels of statistical significance, indicated below the diagram.

A) Levels of three individual miRNAs, as measured by qRT-PCR in eight different melanoma cell lines relative to their respective levels in the melanocyte cell line HEM-l, are plotted as histograms. RQ =  relative quantification index. B) Northern blot analysis of miR-211 and miR-let-7g in five melanoma cell lines and melanocytes. miR-let-7g is consistently over-expressed in melanoma cells. miR-211 is expressed at high level in melanocytes but is not detectable in any of the melanoma cell lines. Error bars are standard errors of mean of six independent measurements.

Histograms show normalized ratios of miR-211 levels in clinical samples relative to its level in the melanocyte cell line HEM-l (normalized to 1.0), as determined by real time quantitative RT-PCR analysis. Normal skin shows low levels of miR-211 because melanocytes constitute a small fraction of normal skin cells (see text). By two-tailed t-test, the mean RQ of miR-211 in the four groups (primary melanoma, regional, distant, and nodal metastatic melanoma) compared to the mean RQ of miR-211 in HEM-l were all statistically significant at P<10⁻⁶. MC = Melanocytes, NS = Normal Skin, PM = Primary Melanoma, RM = Regional Metastases, DM = Distant Metastases, and NM = Nodal Metastases.

Histograms of log₂ transformed mean expression ratios (fold change) of mRNAs in WM1552C to those in the melanocyte line HEM-l are plotted. The computationally predicted target genes were selected according to criteria described in Methods. Only genes with statistically significant fold change in expression were plotted.

WM1552C and A375 melanoma cell lines were transfected with expression cassettes containing the pre-miR-211 sequences, and stable transfectants were selected (see Methods). Expression levels of miR-211 target genes in HEM-l, A375, WM1552C, A375/211 and WM1552C/211 were measured by hybridization of cDNA (made from total RNA) to Affymetrix microarrrays. Histograms represent the log₂ ratios of expression in different cell lines as indicated in the figure. MC = HEM-l.

A) Western blot analysis of KCNMA1 protein expression in melanocytes and melanoma cell lines. Lysates were prepared from cultures of cells complementing those analyzed by northern blot in Figure 2B, including HEM-l, WM1552C, RPMI-7951, SK-MEL2, HT-144, and A375 and probed by Western blotting with antibodies against KCNMA1 or β-tubulin. B) Relative expression of KCNMA1 mRNA in WM1552C compared to WM1552C expressing miR-211 [WM1552C/211(400)]. Histograms represent relative quantification ratio (RQ) as measured by qRT-PCR analysis. Assays were performed in triplicate. C) Western blot analysis of KCNMA1 protein expression in WM1552C stable cell lines. Lysates were prepared from cultures of untransfected WM1552C and WM1552C stably transfected with expression vectors containing: no miR (WM1552C/VO), miR-211 [both WM1552C/211(400) and WM1552C/211(800)], and an shRNA against the KCNMA1 transcript (WM1552C/KC KO), respectively, and probed by Western blotting with antibodies against KCNMA1 or β-tubulin. D) Inverse correlation between miR-211 expression and KCNMA1 protein levels. miR-211 mean RQ was measured by quantitative RT-PCR in three different strains: WM1552C/VO (normalization standard), WM1552C/211(400), and WM1552C/211(800); KCNMA1 protein levels were measured from relative fluorescence in western blots normalized against fluorescence intensity in WM1552C/VO and β-tubulin load controls. Error bars are standard errors of mean for mean RQ, and standard deviations of relative fluorescence intensity. E) Anti-miR-211 inhibitor reverses KCNMA1 protein levels in melanocytes. Melanocytes were transfected with anti-miR-211 inhibitors, and KCNMA1 protein expression was measured in transfected cells by western blot analysis using a KCNMA1 antibody (β-tubulin was used as a load control). Derepression of KCNMA1 protein in the transfected cells is shown in the lane marked as MC+Inh. MC and MC+Scr are melanocyte controls. F) Inhibitory effect of miR-211 on mRNA containing the KCNMA1 3′-UTR sequences. An expression plasmid containing the KCNMA1 3′-UTR seed sequence for miR-211 was fused to a lacZ reporter gene (labelled, KCNMA1) such that the lacZ mRNA would contain the KCNMA1 3′-UTR sequences (harbouring the miR-211 target site) and was co-transfected into the melanoma cell line A375 with one of the following synthetic miRNAs: miR-211, miR-16-1, miR-34b, miR-let-7a, miR-CE (cel-miR-67), and no miRNA. Histograms are measurements of β-galactosidase activity at OD₄₂₀. To directly confirm the importance of the miR-211 seed sequence, a plasmid containing the LacZ gene was fused to a mutant KCNMA1 3′UTR seed sequence (labelled, KCNMA1 Mutant), and the expression vector itself without any 3′-UTR fusion to LacZ (labelled, positive control) were also included. The assays were performed in triplicate. The only sample with statistically significant difference is indicated by an asterisk (Kruskal Wallis test, χ² = 24.142, P<0.001).

Relative gene expression levels of MITF, TRPM1, and miR-211 in MITF knock-down cells. Three different doses of MITF siRNA (5 nM, 10 nM and 15 nM) was used to knock-down MITF gene and expression values are normalized to scramble siRNA control. Histograms represent the Ratio of RNA Concentration, as measured by qRT-PCR analysis.

A and B: Relative mean cell titers of (A) WM1552C/211(400), WM1552C/211(800), and WM1552C/VO (vector only) cells to that of WM1552C cells and of (B) A375/211(400) and A375/VO (vector only) cells to that of A375 cells. (C) and (D): Cell invasion assays comparing WM1552C to WM1552C/VO, WM1552C/211(400), WM1552C/211(800), and WM1552C/KC KO stable derivatives, as well as to WM1552C/211(800) transfected with the Anti-miR miRNA Inhibiter for hsa-miR-211 [labelled “WM1552C/211(800) + miR-211”] or Negative Control #1 (labelled “WM1552C/211(800) + miR-Scramble”) over 48 hours. Each assay was performed in triplicate. Statistical significance is indicated by an asterisk in the figure pertaining to the experimental group delimited by a bar over the histograms (P-value<0.001). E) The artificial expression of KCNMA1 protein in WM1552C/211(800) cells increases melanoma cell invasiveness. Western blot results show that KCNMA1 protein levels are elevated in transfected cells [“WM1552C/211(800) + KCNMA1 Vector” relative to control cells without KCNMA1 expression vector] (bottom). β-tubulin was used as a load control. Results from the invasion assay illustrate that the KCNMA1 protein expression increased melanoma cell invasiveness (top).

MITF, a transcription factor with tumor-suppressor activity, is active in melanocytes, where it is required for the expression of TRPM1, the structural gene for melatonin-1. miR-211 gene is located within the sixth intron of TRPM1, which is co-transcribed with the TRPM1 transcript, is processed to active miR-211 and subsequently the latter acts on the 3′-UTR of KCNMA1 transcript. We propose that inhibition of KCNMA1 translation is needed for preventing the development of invasive melanoma. MITF activity is low in melanoma cells, which is expected to reduce TRPM1 as well as miR-211 transcription, therefore would induce the expression of KCNMA1. While it is widely held on the basis of expression pattern alone that TRPM1 is a putative tumor suppressor, we show that miR-211, contained within the pre-mRNA of TRPM1 transcript, also has a tumor suppressor activity through its negative regulation on KCNMA1.