miR-514a regulates the tumour suppressor NF1 and modulates BRAFi sensitivity in melanoma.

To identify 'melanoma-specific' microRNAs (miRNAs) we used an unbiased microRNA profiling approach to comprehensively study cutaneous melanoma in relation to other solid malignancies, which revealed 233 differentially expressed (≥ 2 fold, p < 0.05) miRNAs. Among the top 20 most significantly different miRNAs was hsa-miR-514a-3p. miR-514a is a member of a cluster of miRNAs (miR-506-514) involved in initiating melanocyte transformation and promotion of melanoma growth. We found miR-514a was expressed in 38/55 (69%) melanoma cell lines but in only 1/34 (3%) other solid cancers. To identify miR-514a regulated targets we conducted a miR-514a-mRNA 'pull-down' experiment, which revealed hundreds of genes, including: CTNNB1, CDK2, MC1R, and NF1, previously associated with melanoma. NF1 was selected for functional validation because of its recent implication inacquired resistance to BRAFV600E-targeted therapy. Luciferase-reporter assays confirmed NF1 as a direct target of miR-514a and over-expression of miR-514a in melanoma cell lines inhibited NF1 expression, which correlated with increased survival of BRAFV600E cells treated with PLX4032. These data provide another mechanism for the dysregulation of the MAPK pathway which may contribute to the profound resistance associated with current RAF-targeted therapies.


Serum/plasma collection and Total RNA extraction
Serum and plasma were processed using standard methodologies and Total RNA was extracted using the Plasma/Serum Circulating RNA Purification Kit (#30000; Norgen Biotek, Ontario, Canada) according to manufacturer's instructions.
Advanced data analyses were performed in Genespring GX12.5 (Agilent Technologies, Santa Clara, USA) using the LOWESS normalized signal intensity values. All values <30 were considered 'background expression' (personal communication with LC Sciences) and changed to 0.01 prior to log2 transformation. So as to identify 'melanocyte-specific' miRNAs that were potentially more relevant to melanoma, samples were classified as either 'melanoma' or 'other cancers' (melanocytes, melanoblasts, nevocyte, and serum derived samples were excluded from these categories). To identify differentially expressed miRNAs, a Mann-Whitney U-test (unpaired) was applied to a 'volcano-plot' analysis with thresholds set at p<0.05 and ≥2 fold. The gene list derived from these analyses was then applied to all samples in an unsupervised manner using hierarchical clustering (Euclidean similarity with average linkage).

miScript quantitative RT-PCR validation
Briefly, all samples included on the microarray along with an extended cohort of melanoma cell lines (as described in (Boyle et al., 2011)) were reverse transcribed using the miScript II RT Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Real-time PCR was subsequently performed with a miScript SYBR Green PCR Kit (QIAGEN, Hilden, Germany); with the miRNA primer assays (QIAGEN, Hilden, Germany) hsa-miR-211-5p (#MS00003808), hsa-miR-514-3p (#MS00031948), and RNU-6 (#MS00033740), using the 7900HT Fast Real Time PCR System (Life Technologies, Foster City, CA, USA). Data were analyzed in Microsoft Excel using the ΔCT method compared to RNU6 which was assessed in every sample.

MITF inducible melanoma cell lines and lineage-specific miRNA Taqman Assays
RNA from the MITF inducible cells was used as templates for Taqman miRNA-specific cDNA synthesis as per manufacturer's instructions (Life Technologies, Foster City, USA).

Biotin pull-downs and microarray step-by-step methods
Biotin pull-downs and microarray hybridizations were modified to work with existing transfection conditions for melanoma cell lines. The protocol follows a step-by-step guide which was originally published by the Cloonan lab (Martin et al., 2014;Wani and Cloonan, 2014).

MATERIALS & REAGENTS
• Cell line of choice ■ PAUSE POINT Samples can now be stored at -80ºC. Alternatively, continue with quantification of RNA.
Step 60 Target mRNA and control RNA Quantification • TIMING 45mins, 20 mins hands-on 60. The target mRNA and the control lysate RNA can now be quantified on the Nanodrop 1000 spectrophotometer and on an Agilent 2100 Bioanalyser using the Agilent RNA 6000 Pico kit using manufacturer's protocol.
■ PAUSE POINT Samples can now be stored at -80ºC or continue on with RNA amplification.
Step 61-62 Target mRNA Labeling and Amplification

14
• TIMING 23 h, 2 h hands-on 61. Amplify and label 160-500ng of captured target mRNA using Illumina® TotalPrep RNA Amplification kit according to manufacturer's instructions. Also amplify and label 160-500 ng of control RNA.
▲ CRITICAL Carry out the 16 hour incubation for the IVT step.
Step 63 65. Extract the expression measurements using the GenomeStudio software.

Site-directed mutagenesis and dual-luciferase reporter assays
Putative binding sites were identified using the miRanda target prediction software (version August 2010) (Enright et al., 2003). A partial sequence of the cDNA of NF1 Melanoma cell lines were selected based upon transfection ability along with having detectable (by Western blot) endogenous NF1 protein levels (data not shown). NF1 mRNA expression was determined (relative to RNU6) using a QuantiTect Primer Assay (#QT00065016; QIAGEN) as described previously. Cell viability assays were performed and determined using a modified sulforhodamine B (SRB; Sigma, St Louis, USA) assay (Vichai and Kirtikara, 2006). Briefly, siRNAs, miRNA mimic, Negative control or LNA's were reverse-transfected with melanoma cell lines and seeded into a 96-well plate then incubated at 37ºC with 5% CO2 for 24hr. A serial 10-fold dilution series (100 nM-0.01 nM) of PLX4032 (Selleckchem, Houston, USA) in DMSO was added across each plate.
Plates were fixed on day 6 with methylated spirits prior to performing the SRB assay and read at 564 nm using a plate reader (Molecular Devices, Sunnyvale, USA). NF1 mRNA expression analysis RNA was reverse transcribed using the miScript II RT Kit (QIAGEN, Hilden, Germany) using the 'HiFlex' Buffer thus enabling both mRNA and miRNA to be analyzed in the same sample. The primer assays were used as described NF1 (#QT00065016) along with (RNU-6). Real-time PCR was performed and analyzed as previously stated.

20
(Benjamini-Hochberg). Normalised expression data of 233 miRNAs genes in all samples present on the microarray. Samples are grouped based upon tissue type. Normalised expression data is conditionally formatted (in Microsoft Excel) to assist in data visualisation.

Supplementary Table 2
All miRNAs genes that were ≥10 fold up and downregulated from those listed in Supplementary Table 1 are summarized here along with their relevance to melanoma.

Supplementary Table 3
Summary of all cell line names and associated tissue type that were used in the discovery (microarray) and validation cohorts (qRT-PCR).

Supplementary Table 4
List of all genes that we upregulated in 2/2 melanoma cell lines following a biotinlabelled miR-514a duplex pulldown of mRNA transcripts.

Supplementary Table 5
List of all possible binding sites of miR-514 in the NF1 transcript (5'UTR, coding, and 3'UTR) using the miRanda prediction algorithm. Binding threshold is set to 100 (default=140).

Supplementary Table 6
Table of P values associated with BRAFi sensitivity for the melanoma cell lines MM96L and MM253 ( Figure 5A-B). T-tests were performed with comparisons were made between siNF1 vs miR-514a; siNF1 vs siNF1+miR-514a; siNF1 vs miR-Neg-scr; miR-