Upregulated expression of miR-4443 and miR-4488 in drug resistant melanomas promotes migratory and invasive phenotypes through downregulation of intermediate filament nestin

Abstract

Background: BRAF-mutant melanoma patients benefit from the combinatorial treatments with BRAF and MEK inhibitors. However, acquired drug resistance strongly limits the efficacy of these targeted therapies in time. Recently, many findings have underscored the involvement of microRNAs as main drivers of drug resistance. In this context, we previously identified a subset of oncomiRs strongly up-regulated in drug-resistant melanomas. In this work, we shed light on the molecular role of two as yet poorly characterized oncomiRs, miR-4443 and miR-4488.

Methods: Invasion and migration have been determined by wound healing, transwell migration/invasion assays and Real Time Cell Analysis (RTCA) technology. miR-4488 and miR-4443 have been measured by qRT-PCR. Nestin levels have been tested by western blot, confocal immunofluorescence, immunohistochemical and flow cytometry analyses.

Results: We demonstrate that the two oncomiRs are responsible for the enhanced migratory and invasive phenotypes, that are a hallmark of drug resistant melanoma cells. Moreover, miR-4443 and miR-4488 promote an aberrant cytoskeletal reorganization witnessed by the increased number of stress fibers and cellular protrusions-like cancer cell invadopodia. Mechanistically, we identified the intermediate filament nestin as a molecular target of both oncomiRs. Finally, we have shown that nestin levels are able to predict response to treatments in melanoma patients.

Conclusions: Altogether these findings have profound translational implications in the attempt i) to develop miRNA-targeting therapies to mitigate the metastatic phenotypes of BRAF-mutant melanomas and ii) to identify novel biomarkers able to guide clinical decisions.

Keywords: Invasion; Metastatic melanoma; MicroRNA; Migration; Targeted therapy.

Fig. 1

A375 resistant melanoma cells have increased migratory/invasive phenotypes and altered cytoskeletal remodeling as compared to sensitive counterparts. a A375 sensitive cells (SEN) and their resistant counterparts (RES) were cultured in the presence of double well culture inserts to perform wound healing assays. After 24 h, inserts were removed and each well was photographed at 10 × magnification (scale bar 500 μm) immediately after insert removal (T-0) and after 24 h (T-24H) (left panel). The open residual area (right panel) was measured by using the ImageJ software in RES cells as compared to SEN ones. The values were calculated as “fold change” (± SD). b To perform migration assays, A375 SEN and RES cells were seeded in serum-free media into the upper chamber of a transwell, whereas the lower chamber was filled with 10% FBS RPMI. After 8 h, cells remaining on the top side of the membrane were counted and the average number ± SD of cells are reported as fold change respect to control considered as 100. c To perform invasion assays, A375 SEN and RES cells were seeded into the upper chamber of a transwell coated with Matrigel, whereas the lower chamber was filled with 10% FBS RPMI. After 24 h, invading cells were counted and the average number ± SD of cells are reported as fold change respect to control considered as 100. Three independent experiments were performed (a, b, c); each experiment was performed at least in quadruplicate. d, e xCELLigence RTCA technology was used to measure cell migration and invasion of A375 SEN and RES cells. To evaluate migration (d), cells were seeded in serum-free medium on filters in the upper chamber, whereas the lower chamber was filled with 10% FBS RPMI. Cell migration was monitored for 6 h. To evaluate invasion (e), cells were seeded in serum-free medium on filters coated with Matrigel in the upper chamber, whereas the lower chamber was filled with 10% FBS RPMI. Cell invasion was monitored for 25 h. Each experiment was performed at least twice in quadruplicate (d, e). The slope represents the change rate of cell index values generated in a 0–1 h time frame. f, g Representative images by confocal microscopy of F-Actin and DAPI immunostaining of A375 SEN and RES cells that were cultured into 8-well μ-slides for 48 h. Red arrows indicate filopodia and stress fibers. Magnification 63x. h Quantification analyses of the long cell axis measured from the aforementioned confocal images were performed by Zeiss Zen control software. All the experiments have been performed at least in triplicate ± standard deviation (SD) and p-value < 0.05 was considered as significant (Student’s t-test)

Fig. 2

miR-4443 and miR-4488 transient overexpression affects migration/invasive potential and cytoskeletal remodeling of melanoma cells. a Total RNAs were extracted from A375 sensitive cells (SEN) and their resistant counterparts (RES) to perform qRT-PCR analyses for miR-4443 and miR-4488 expression levels. Each miRNA transcript was quantified using the comparative Ct method for relative quantification (ΔΔCt) using U6 as normalizer. The values were calculated as “fold change” (± SD) compared to SEN considered as 1. b A375 SEN cells were transiently transfected with miR-4443, miR-4488 or the relative scrambled (SCR) for 48 h. Transfected cells have been then harvested and cultured in the presence of double well culture inserts to perform wound healing assays. After 24 h, inserts were removed and each well was photographed at 10 × magnification (scale bar 500 μm) immediately after insert removal (T-0) and after 24 h (T-24H) (left panel). The open residual area (right panel) was measured by using the ImageJ software in oncomiR-transfected cells as compared to SCR-transfected ones. The values were calculated as “fold change” (± SD). c To perform migration assays, A375 SEN cells, transfected as described above, were seeded in serum-free media into the upper chamber of a transwell, whereas the lower chamber was filled with 10% FBS RPMI. After 8 h, cells remaining on the top side of the membrane were counted and the average number ± SD of cells are reported as fold change respect to control (SCR) considered as 100. d To perform invasion assays, A375 SEN cells, transfected as described above, were seeded into the upper chamber of a transwell coated with Matrigel, whereas the lower chamber was filled with 10% FBS RPMI. After 24 h, invading cells were counted and the average number ± SD of cells are reported as fold change respect to control (SCR) considered as 100. e, f Migration and invasion assays were performed as described above using A375 RES cells transiently transfected with the anti-miR-4443, the anti-miR-4448 or the SCR for 48 h before performing the appropriate assays. For all migration and invasion assays, three independent experiments were performed; each experiment was performed at least in quadruplicate. g Representative images by confocal microscopy of F-Actin and DAPI immunostaining of A375 SEN cells transiently transfected with miR-4443, miR-4488 or the SCR for 48 h. Red arrows indicate filopodia and stress fibers. Magnification 63x. h Quantification analyses of the long cell axis measured from the aforementioned confocal images were performed by Zeiss Zen control software. All the experiments have been performed at least in triplicate ± standard deviation (SD) and p-value < 0.05 was considered as significant (Student’s t-test)

Fig. 3

The intermediate filament Nestin is a molecular target of miR-4443 and miR-4488 in melanoma. a Schematic representation of the entire 3’UTR of mRNA relative to nestin containing the two seed regions for miR-4488 and for miR-4443 located at positions 197–203 and 336–342, respectively (source miRWalk3.0). b, c To perform luciferase assays, HEK-293 cells were plated in a 6-wells plate and then co-transfected with a plasmid containing the 3’ UTR of nestin and Renilla luciferase plasmid together with miR-4443 (b), miR-4488 (c) or the relative scrambled (SCR). After 48 h of transfection, cells were harvested and luminescence was evaluated using Dual-Luciferase® Reporter Assay System. The values were calculated as fold change (± SD) compared to SCR considered as 1. d A375 SEN cells were transiently transfected with miR-4443, miR-4488 or with the SCR for 72 h. Cells have been then harvested and lysed to extract total proteins that have been subjected to Western blot analyses with the indicated antibodies (left panel). Tubulin levels have been estimated for the protein equal loading and densitometric analyses were performed using ImageJ software (right panel). Results were expressed as mean values from three independent experiments. The values were calculated as fold change (± SD) compared to SCR considered as 1. e A375 SEN cells transfected as described above have been stained with an antibody PE mouse anti-nestin for 45 min to perform flow cytometry analyses. Non-stained cells were used as negative control. Black arrows indicate the peaks relative to cells transfected with miR-4443 or miR-4488 mimics (left graphs). Quantitative analyses were calculated as “fold change” (± SD) compared to SCR considered as 1 (right graph). Data were analyzed using CytExpert version 2.2 software. f Representative images by confocal microscopy of nestin and DAPI immunostaining of A375 SEN cells transiently transfected as described above. Magnification 63x. All the experiments have been performed at least in triplicate ± standard deviation (SD) and p-value < 0.05 was considered as significant (Student’s t-test)

Fig. 4

Nestin levels are reduced in drug resistant melanoma cells and are associated with targeted therapy response in patients. a A375 SEN and RES cells were harvested and lysed to extract total proteins to perform Western blot analyses with the indicated antibodies. Tubulin levels have been estimated for the protein equal loading. b A375 SEN and RES cells have been stained with an antibody PE mouse anti-nestin for 45 min to perform flow cytometry analyses. Non-stained cells were used as negative control. Black arrow indicates the peaks relative to RES cells (left graph). Quantitative analyses were calculated as “fold change” ± standard deviation (SD) compared to SEN cells considered as 1. Data were analyzed using CytExpert version 2.2 software. c, d Migration and invasion assays were performed as described above using A375 RES cells transiently transfected with a plasmid encoding for nestin (NES) or the relative empty control (CTR) for 48 h. Results have been quantified by counting cells remaining on the top side of the membrane and the average number ± SD of cells are reported as fold change respect to control (SCR) considered as 100. e The online software GEPIA 2.0 (http://gepia2.cancer-pku.cn/#index) was interrogated to unveil nestin prognostic value based on Skin Cutaneous Melanoma (SKCM) data from The Cancer Genome Atlas (TCGA). Patients have been stratified according to the mutational status and the Kaplan Meier (KM) curves represent the data relative to the BRAF-mutant subtype (n = 149). f The mRNA levels relative to nestin (NES) have been extracted from bulk RNA-seq data deriving from 18 melanoma patients before starting MAPKi therapy (GSE65185) deposited in Gene Expression Omnibus (GEO) database. Data have been used to plot KM curves to assess the predictive value of NES for therapy response (Discovery cohort). g NES levels were evaluated by immunohistochemistry (IHC) in 14 melanoma patients’ biopsies collected before starting MAPKi therapy. Data have been used to plot KM curves to assess the predictive value of NES for therapy response (internal validation cohort). High and low levels were assessed by considering positive and negative z-scores, respectively. The hazard ratio, Cox models and the log-rank p values were evaluated to plot KM curves

Fig. 5

Nestin levels in representative melanoma patients. Melanoma biopsies collected before starting MAPKi therapy were subjected to immunohistochemistry (IHC) analyses to measure nestin levels. Results were quantified counting positive cells in at least ten fields and expressed as percentage of positivity. Representative images of four cases are reported (TT07, TT09, TT13, and TT17). 10X magnification (scale bar 500 μm). Progression Free Survival (PFS) is expressed in days. All the clinical data are available in Suppl. Table 1