Integrated Genomics Identifies miR-181/TFAM Pathway as a Critical Driver of Drug Resistance in Melanoma

Abstract

MicroRNAs (miRNAs) are attractive therapeutic targets and promising candidates as molecular biomarkers for various therapy-resistant tumors. However, the association between miRNAs and drug resistance in melanoma remains to be elucidated. We used an integrative genomic analysis to comprehensively study the miRNA expression profiles of drug-resistant melanoma patients and cell lines. MicroRNA-181a and -181b (miR181a/b) were identified as the most significantly down-regulated miRNAs in resistant melanoma patients and cell lines. Re-establishment of miR-181a/b expression reverses the resistance of melanoma cells to the BRAF inhibitor dabrafenib. Introduction of miR-181 mimics markedly decreases the expression of TFAM in A375 melanoma cells resistant to BRAF inhibitors. Furthermore, melanoma growth was inhibited in A375 and M14 resistant melanoma cells transfected with miR-181a/b mimics, while miR-181a/b depletion enhanced resistance in sensitive cell lines. Collectively, our study demonstrated that miR-181a/b could reverse the resistance to BRAF inhibitors in dabrafenib resistant melanoma cell lines. In addition, miR-181a and -181b are strongly down-regulated in tumor samples from patients before and after the development of resistance to targeted therapies. Finally, melanoma tissues with high miR-181a and -181b expression presented favorable outcomes in terms of Progression Free Survival, suggesting that miR-181 is a clinically relevant candidate for therapeutic development or biomarker-based therapy selection.

Keywords: BRAF inhibitors; Dabrafenib; TFAM; biomarkers; cancer resistance; melanoma; miR-181; microRNA; mitochondria; target therapy.

Figure 1

(A) Hierarchical clustering analysis and heat map based on the expression profiles of 65 miRNAs in responding and non-responding melanoma cells to dabrafenib treatment. Cluster analysis grouped samples and miRNAs according to similarity in expression. miRNAs are in rows and samples in columns. The miRNA clustering tree is indicated on the left and the sample clustering tree is at the top. Red color represents up-regulated expression and blue marks downregulated genes. Yellow indicates resistant cells and blue indicates responsive cells. miRNA, microRNA. To create the heatmap we converted the read counts into log2-counts-per-million (logCPM) values. (B) Putative target genes of differentially expressed miRNAs were obtained from TargetScan and used for KEGG pathway enrichment analysis. Only pathways with an adjusted p value < 0.01 were considered and listed according to a decreasing value of the combined score. (C,D) The expression levels of miR-181a and b in A375, A375-BIR, M14 and M14-BIR melanoma cells, measured by quantitative Real-Time PCR (qRT-PCR). (E,F) The expression levels of miR-181a and b in A375, A375-BIR, M14 and M14-BIR melanoma cells, treated with 200 nM dabrafenib (DAB) and vehicle control (DMSO), measured by qRT-PCR. The miRNA expression levels were normalized to the internal control U6. **** p < 0.0001; ns, not significantly.

Figure 2

(A,B) A375 and A375-BIR cells were transiently transfected with 40 nM of the miR-181a and -181b mimic or inhibitor or the appropriate controls and assayed for proliferation by the MTT viability assay. Data are expressed in terms of percentage of cell viability as compared to untransfected cells. Each value represents the arithmetic mean of three independent experiments performed with triplicate samples. *** p < 0.001. (C,D) M14 and M14-BIR cells were transiently transfected with 40 nM of the miR-181a and -181b mimic or inhibitor or the appropriate controls and assayed for proliferation by the MTT assay. Data are expressed in terms of percentage of cell viability as compared to untransfected cells. Each value represents the arithmetic mean of three independent experiments performed with triplicate samples. *** p < 0.001. (E–G) Colony formation assays were performed to determine the colony formation ability of A375 and A375-BIR melanoma cells transfected with miR-181a and -b mimics or inhibitor or the appropriate controls. Representative images and quantification of visible colonies have been presented. The colony number in the DMSO-treated group was set as 100%. All the experiments were performed in triplicate and the relative colony formation rates are shown as the mean +/− SD. *** p < 0.001; **** p < 0.0001; ns, not significant. H–I) Effect of miRNA-181a and b forced expression or inhibition on CASP-9 activation in A375 and A375-BIR cells. The determination of CASP-9 activity was carried out by using Caspase-Glo^® 9 assay. Data are expressed in fold change relative to vehicle control +/− SD of three independent assays with 3 replicates for each one. *** p < 0.001 and ** p < 0.01 versus vehicle control; ns, not significant.

Figure 3

(A,B) A375 (A) and A375-BIR (B) cells transiently transfected with 40 nM of the miR-181a and -181b mimic or inhibitor or the appropriate controls were treated with 200 nM dabrafenib and assayed for proliferation by the MTT assay. Data are expressed in terms of percentage of cell viability as compared to untransfected cells. Each value represents the arithmetic mean of three independent experiments performed with triplicate samples. *** p < 0.001; * p < 0.01. (C) Effect of miR-181a/b knockdown on cell proliferation in A375 and A375-BIR melanoma cells treated with increasing doses of dabrafenib. Dose-response curves of cell viability according to the sensitivity to dabrafenib is showed for cells knockdown for pre-miR-181a/b (181KD) cluster and controls bearing empty vector (EV, pcDNA3.1-GFP). Data are represented as mean +/− SD, n = 3. (D,E) Colony formation assays were performed to determine the colony formation ability of A375, A375-BIR and A375-181KD cells transfected with miR-181a and -181b mimics or the appropriate controls. Representative images and quantification of visible colonies have been presented. The colony number in the DMSO group was set as 100%. All the experiments were performed in triplicate and the relative colony formation rates are shown as the mean +/− SD. *** p < 0.001; **** p < 0.0001; ns, not significative. (F,G) Effect of forced expression or inhibition of miRNA-181a and -181b on CASP-9 activation in A375, A375-BIR and A375-181KD cells. The determination of CASP-9 activity was carried out by using Caspase-Glo^® 9 assay. Data are expressed in fold change relative to vehicle control +/− SD of three independent assays with 3 replicates each one. *** p < 0.001; ** p = 0.001 and * p < 0.01 versus vehicle control; ns, not significant.

Figure 4

(A,B) Box-and-whisker diagrams of miRNA-181a (A) and -181b (B) levels in melanoma patients grouped according to best response to the treatment and in normal skin sample. The horizontal bar within each box indicates the median. Data were analysed by non-parametric Wilcoxon matched-pairs signed-rank test. ** p < 0.001; * p < 0.01; ns = not significant. (C,D) Relationship between the expression of miRNA-181a (C) and -181b (D) in melanoma tissues and the prognosis of patients. The PFS (progression free survival) rate of the high miRNA-181a (B) and -181b (C) expression group was higher than that of the low microRNAsexpression group (p = 0.0065; p = 0.0091). (E,F) Kaplan-Meier Log-rank survival analysis for overall survival (OS) of melanoma patients according to miR-181a (E) and -181b (F) expression (p = 0.0518; p = 0.1036).

Figure 5

(A) Scatter plot of the RNA-Seq expression data showing differentially expressed miR-181a and -181b-target genes (DEG) between A375 (responders) and A375-BIR (resistant) cell lines. Dots represent DEGs that are: significantly overexpressed (red) and down-regulated (green). The most critical upregulated and downregulated DEGs targets are indicated. (B) Volcano plot showing the DEGs up (red dots) and down-regulated (blue dots) in patients expressing high levels of miR-181a and -181b (responders group). FDR: False discovery rate. (C) Plot of functional gene set enrichment analysis (GSEA) indicating hallmark gene sets significantly modulated in melanoma patients expressing high levels of miR-181a and -181b (responders group). (D) Network of miRNA-181a- and -181b-negatively regulated predicted target genes in melanoma patients (responders group) visualized by Cytoscape. Green dots represent predicted target genes and the red dot represents hub miRNA. (E,F) GO term enrichment analysis of miRNA-181a- and -181b-negatively regulated predicted target genes in melanoma patients (responders group). (E) Scatter plot of enriched GO pathways according to numbers of genes and FDR. (F) GO network analysis plot shows the relationship between enriched pathways. Darker green nodes are more significantly enriched gene sets. Bigger nodes represent larger gene sets. Most critical networks are highlighted in yellow.

Figure 6

(A) TFAM 3′UTR contains miR-181a and -181b binding sites. Red-labelled nucleotide indicate the differences between the sequences of miR-181a and -181b. (B) The dual luciferase reporter assay was performed with HeLa cells as described in Materials and Methods. Briefly, HeLa cells were transfected with miR-181a or miR mimic negative control together with the plasmid encoding for the Firefly luciferase (FFL) carrying part of the TFAM 3′-UTR in its wild type form (TFAM-3UTR) or TFAM 3ʹ-UTR with the mutagenized form of the miR-181a and -181b binding site (TFAM-3UTR-MUT). FFL activities were internally normalized to Renilla luciferase activities yielding relative light units (RLU). Histograms show the mean +/− SE from 3 to 6 independent experiments upon normalization to the miR mimic control. ** p < 0.01; *** p < 0.001 in un-paired t-test. (C,D) TFAM mRNA expression analysis in A375 (C) and A375-BIR (D) cells transfected with miR-181a, -181b mimics or their inhibitors or relative controls. Histograms show the mean +/− SE from 3 independent experiments upon normalization to the miR mimic control. ** p < 0.01; *** p < 0.001; **** p < 0.0001 in un-paired t-test. (E,F) Treatment resistant and responder melanomas display miR-181a/b/TFAM axis deregulation. Lower miR-181a (E) and -181b (F) and higher TFAM expression levels in patients resistant during therapy compared to tumors from the responder group. p < 0.001 by Mann-Whitney U test.