Comparative Analysis of the Effect of the BRAF Inhibitor Dabrafenib in 2D and 3D Cell Culture Models of Human Metastatic Melanoma Cells

Abstract

Background/Aim

Melanoma, a variant of skin cancer, presents the highest mortality rates among all skin cancers. Despite advancements in targeted therapies, immunotherapies, and tissue culture techniques, the absence of an effective early treatment model remains a challenge. This study investigated the impact of dabrafenib on both 2D and 3D cell culture models with distinct molecular profiles.

Materials and Methods

We developed a high-throughput workflow enabling drug screening on spheroids. Our approach involved cultivating 2D and 3D cultures derived from normal melanocytes and metastatic melanoma cells, treating them with dabrafenib and conducting viability, aggregation, migration, cell cycle, and apoptosis assays.

Results

Dabrafenib exerted multifaceted influences, particularly on migration at concentrations of 10 and 25 μM. It induced a decrease in cell viability, impeded cellular adhesion to the matrix, inhibited cellular aggregation and spheroid formation, arrested the cell cycle in the G₁ phase, and induced apoptosis.

Conclusion

These results confirm the therapeutic potential of dabrafenib in treating melanoma with the BRAF V600E mutation and that 3D models are validated models to study the potential of new molecules for therapeutic purposes. Furthermore, our study underscores the relevance of 3D models in simulating physiological in vivo microenvironments, providing insights into varied treatment responses between normal and tumor cells.

Metastatic cutaneous melanoma skin cancer derived from an uncontrollable melanocyte proliferation located in the deep layer of the epidermis, with additional invasive and metastatic features, represents an important part of an invasive malignant melanoma (1,2). Invasive and metastatic cutaneous melanoma constitute the third most common type of skin cancer compared with non-melanoma skin cancer, such as squamous cell carcinoma and basal cell carcinoma (3). According to GLOBOCAN projections, it is estimated that by 2025, around 367,000 new cases of cutaneous melanoma will be diagnosed. The number of these new cases is estimated to increase by more than 50% by 2040, with 68% of expected deaths increasing from 57,000 to 96,000 in 2040 according to data produced by the International Agency for Research on Cancer (IARC) (4). Despite this, the five-year survival rate for metastatic cutaneous melanoma skin cancer remains at 15 to 20% (5).

Standard therapies for advanced melanoma include immunotherapies targeting PD-L1 and/or PD-1 antibodies, either alone or in combination with anti-CTLA as well as BRAF or Ras-Raf-MEK-ERK pathway inhibitors (6,7). In the context of personalized medicine for metastatic cutaneous melanoma, tumors are classified based on clinical, histological, site-based, and molecular biology subtypes, which include superficial spreading melanoma (SSM), nodular melanoma (NM), acral lentiginous melanoma (ALM) and lentigo malignant melanoma (LMM). Additionally, genetic subtypes and clinical/histopathological classifications are considered (8). According to the WHO classification, melanoma is categorized by frequency as follows: SSM (41-60% of cases), NM (16-20% of cases), ALM (5-10% of cases), and LMM (2-16% of cases) (9-11). Alternatively, genetic classifications using databases like The Skin Cutaneous Melanoma catalogue reveal prevalent mutations in the BRAF gene’s kinase domain (45-70% of cases), followed by NRAS (15-30% of cases) and KIT gene mutations (5-15% of cases) (12-14).

Molecular alterations in key genes such as BRAF, can hyperactivate proliferation, survival and carcinogenesis pathways like MAPK and/or Pi3K signaling (15,16). RAF mutation is a pivotal diagnostic marker for cutaneous melanoma and a notable therapeutic target (17-19). Current target therapies, including MEK/MAPK inhibitors (trametinib, cobimetinib, binimetinib) and BRAF inhibitors (vemurafenib, dabrafenib, encorafenib) have significantly improved patient outcomes, shifting from a median of six months with chemotherapy to a five-year overall survival (OS) and recurrence-free survival (RFS) with combination target therapies (20-22).

Despite these advancements, resistance development remains a challenge, with a median progression-free survival (PFS) of <12 months for combination therapy and <7 months for monotherapy in published trials (21,23). The development of resistance to BRAF inhibitor target therapies involves various mechanisms, caused by dividing persistent cells, reactivation of the MAPK signaling pathway through BRAF V600E gene amplification, NRAS or MEK secondary mutations, or COT (MAP3K8) over-expression (24). Other related adaptive mechanisms in the extracellular environment include growth factors involved in autocrine and paracrine signaling, metabolic reprogramming and dysregulation of the PI3K/AKT pathway (25,26). Preclinical models have demonstrated that both monotherapy and combined BRAF and MEK inhibition can suppress MAPK signaling and delay the development of acquired resistance (27,28).

In vitro cellular aggregates in the form of 3D- spheroids are increasingly recognized for their value in offering insight into cell growth characteristics, cell-cell, cell-extracellular matrix interactions and the complexity of the tumor environment (29,30). Gaseous exchange, metabolites and necrotic and hypoxic centers appear to reflect the characteristics of tumors more than cells from 2D cultures (31). Consequently, different studies have been using spheroid models to monitor drug penetration and resistance mechanisms with the overall target of reaching personalized therapies (32-34). However, the schema to assess multiple assays, such as proliferation, migration, cell cycle dynamics, apoptosis, aggregation, and drug efficacy adhesion has not been implemented as a validated model (33,35).

In our study, we conducted an analysis of public datasets to compare BRAF mutation prevalence in malignant cutaneous melanoma and healthy tissue. Next, we compared the presence of the BRAF mutation, clinical subtypes, and invasive features of these melanoma types according to the Clark classification and reported the phenotypic characteristics of the melanoma. We then investigated the anti-tumoral effect of dabrafenib against the malignant melanoma cell line BRAF(V600E) C32 and a normal primary cell line melanocyte HEMa in 2D and 3D models. The proliferation, migration, invasion, cell cycle dynamics, apoptosis, aggregation, and adhesion characteristics of these cell lines in the presence of the leading drug treatment dabrafenib are presented. This study therefore offers a detailed characterization of a melanoma spheroid model alongside normal melanocytes delineating their distinct behaviors, cell growth kinetics, and cycle dynamics. These baseline data are important if spheroids are to be established as rapid, relatively inexpensive, novel drug screening tools. Finally, we explored the association between the effect of dabrafenib and patients’ outcomes, considering BRAF status and melanoma clinical subtypes.

Materials and Methods

Analyses of public datasets. TNMplot public dataset was used to obtain BRAF expression data across metastatic, tumoral melanoma, and normal skin (36). We used the GENT2 public dataset and UCSC Xena visualization tool to analyze BRAF expression in different skin melanoma cancer subtypes (37). The UCSC Xena visualization tool was used to conduct co-expression analyses on two datasets: TCGA melanoma (SKCM) tumoral dataset (n=481), and GDC TCGA Melanoma (SKCM) tumoral dataset (n=477) (38). The ROCplot.org database (ROCplot) facilitated the exploration of responses to chemotherapy and cutaneous melanoma molecular subtype responses in connection with BRAF expression in solid tumors (39). KMplot public dataset was used to predict the response in overall survival in solid tumors with BRAF mutations (39).

Cell lines. The amelanotic malignant melanoma C32 cell line (ATCC – CRL-1586) and primary epidermal melanocyte HEMa cell line (ATCC-PCS-200-013) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). The C32 cell line was grown in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% fetal bovine serum (FBS; Gibco, Thermo Fisher Scientific, Waltham, MA, USA), antibiotics [(Ant) penicillin (100 U/ml), and streptomycin (100 μg/ml) (Gibco, Thermo Fisher Scientific)]. The HEMa cell line was grown in Dermal Cell Basal Medium (ATCC – PCS-200-030, Manassas, VA, USA) supplemented with Adult Melanocyte Growth Kit (ATCC – PCS-200-042), penicillin (100 U/ml), and streptomycin (100 μg/ml).

The C32 cell line with the BRAF mutation signature (V600E) was used to investigate the anti-tumoral effect of dabrafenib (Medchem – HY-12057) as a target inhibitor of BRAF (V600E) on malignant melanoma (40,41). It was selected on the basis that it possesses mutations in the BRAF and CDKN2A genes (missense mutations and copy number deletions, respectively) (42), that constitute the most common mutations in sporadic and familial melanoma (43,44). The non-tumor HEMa cell line was used to investigate the impact of dabrafenib in a non-BRAF (V600E) mutation cell line. For all cell lines, the culture medium was changed every three days and maintained at 37ºC under a 5% CO₂ humidified atmosphere. Cells were passaged by adding between 1-3 ml of trypsin 0.05% once they reached 70-80% confluency. Dabrafenib was dissolved in ethanol and stored long-term at -20ºC (1 mg/ml) or stored at 4˚C for no longer than two days when in use.

Culture and characterization of 3D Spheroids. To cultivate spheroids in a three-dimensional configuration, 50 μl of 2% low-gelling temperature agarose (SLCM8156) was plated per well in a 96-well plate and allowed to gel. Low-attachment culture plates (Thermo Fisher Scientific – 174927) were employed for spheroid maintenance. For each cell line, 10,000 or 20,000 cells were plated in 100 μl of fresh medium supplemented with 10% FBS and Ant per well and incubated using standard conditions. The media were changed every three days. Spheroids were imaged with EVOS™ M7000 Imaging Systems-2 (Advanced Microscopy Group, Paisley, Scotland, UK) every 24 h, and area and diameter analyzed using ImageJ software version 1.54 (National Institute of Health, Bethesda, MD, USA).

Spheroid aggregation assay. Cell culture multiwell plates were coated with 50 μl of agarose low gelling temperature (SIGMA SLCM8156) solution at 2% to prepare a non-adhesive surface as aforementioned. A total of 10,000 cells per well were plated on agarose-coated wells with 150 μl fresh medium supplemented with 10% FBS, Ant and different concentrations of dabrafenib (1-200 μM) or EtOH. These were followed up daily for seven days and imaged with EVOS™ M7000 Imaging System. Each experiment was independently conducted three times.

Drug sensitivity assay on 2D/3D cell cultures. Drug sensitivity in 2D and 3D cell cultures was assessed using the CellTiter-Glo 3D cell viability assay (Promega – G9682) following the manufacturer’s guidelines. For 2D culture, 10,000 cells were plated in 96 well plates with 100 μl of fresh media. After 24 h of incubation, the medium was substituted by fresh medium supplemented with 10% FBS, Ant and either dabrafenib (1-200 μM), ethanol (EtOH) as a vehicle control or fresh medium as a negative control. After incubation for 72 h, medium was replaced with 100 μl of fresh medium supplemented and 100 μl of CellTiter-Glo 3D reagent, mixed, and transferred to an opaque-walled multiwell plate, then incubated for 25 min at room temperature (RT). For 3D culture, spheroids were cultured for five days and then 100 μl of fresh medium supplemented with 10% FBS, Ant, and dabrafenib, EtOH or medium, was added and incubated for 72 h. Spheroids were transferred to opaque-walled multiwell plates and 100 μl of medium and CellTiter-Glo 3D reagent were added, shaken for 5 min, and then left at RT for 25 min. To measure luminescence readings, the TECAN Spark 20M (Mannedorf, Switzerland) plate reader was used. Viability for each group of spheroids was calculated using Prism GraphPad (GraphPad Software Inc., La Jolla, CA, USA). Each experiment was independently conducted three times.

Cellular migration and inhibition in 2D culture. For the migration assay in a monolayer culture, 1×10⁵ C32 melanoma and HEMa normal melanocyte cells were plated onto 24-well plates to reach 80-95% confluency. A wound was then created using a 200 μl tip. Detached cells were rinsed with 1X PBS, and then medium supplemented with 3% FBS, Ant, and either dabrafenib (10-25 μM), EtOH, or fresh medium was added to the plates. Wound healing was observed under an EVOS FL auto 2 microscope (Thermo Scientific™) at 0, 24, 48, and 72-h post-wound. The quantitative analysis of gap distance was evaluated using ImageJ. Each experiment was independently conducted three times.

Spheroid migration assay. The migration assay encompassed two distinct methods. The first method involved the utilization of a gel thin layer coating technique that coated the well with Matrigel. In this method, culture plates were coated with a mixture of Matrigel and medium at a 1:4 ratio, followed by a 30-min incubation period. Then, four spheroids were seeded on top of the coated plates in a fresh medium supplemented with 10% FBS, Ant, and dabrafenib at concentrations of 10 and 25 μM for the C32 cell line and 25 and 50 μM for the HEMa cell line.

The second method, known as the thick coating method, involved the preparation of a stock solution containing spheroids embedded in Matrigel at a 1:500 ratio, with a medium devoid of supplements (FBS and Ant). Four spheroids were then introduced into the stock solution, which was subsequently applied to the pre-coated plates. After 40 min of incubation, a fresh medium containing dabrafenib, 10% FBS, and Ant at the aforementioned concentrations was added. Spheroids were imaged at intervals of 0, 12, 48, and 72 h using EVOSTM M7000 Imaging Systems. The area invaded by the spheroid after these times was quantified using ImageJ software version 1.54. Each experiment was independently conducted three times.

Cell cycle assays. For cell cycle analysis in both 2D and 3D cell culture experiments, C32 and HEMa cells were serum-starved (0% FBS) for 24 h. Subsequently, 1×10⁵ cells were seeded onto 24-well plates with fresh medium supplemented with 10% FBS, Ant, and dabrafenib (10-25 μM for C32 and HEMa). After incubation for 24 and 72 h, cells were detached using 0.05% trypsin for 5 min and centrifuged. In 3D culture, spheroids were cultured for five days, exposed to fresh medium with 10% FBS, Ant, and either dabrafenib (10-25 μM for C32, 25-50 μM for HEMa), EtOH or fresh medium and maintained for 24 and 72 h. After incubation, spheroids were dissociated using 0.05% trypsin. For both experiments, cell pellets were fixed with cold ethanol (4˚C for 2 h). Following fixation, cells were stained with 200 μl of a 50 μg/ml propidium iodide solution, and 100 μg/ml Ribonuclease A in PBS for 1 h at 37˚C. Post-staining, the cells were analyzed using flow cytometry (Attune™ NxT Acoustic Focusing Cytometer, Life Technologies, Carlsbad, CA, USA), and data analysis was carried out using FlowJo VX software. Each experiment was independently conducted three times.

Apoptosis assays. For the 2D models, 1×10⁵ C32 and HEMa cells were seeded in 6-well plates overnight. Fresh medium with dabrafenib (10-25 μM) was added and incubated for 24, 48, and 72 h. Cells were detached using 0.05% trypsin for 5 min, and the resulting cell suspension was centrifuged. For the 3D cell culture, spheroids were cultured for five days, then exposed to fresh medium with 10% FBS, Ant, and dabrafenib (10-25 μM for C32, 25-50 μM for HEMa), EtOH or fresh medium with 10% FBS and Ant. This culture was maintained for 24, 48, and 72 h, and then spheroids were dissociated using 0.05% trypsin for 5 min. Fresh medium was added, and the cell suspension was centrifuged.

For both the 2D and 3D experiments, cell pellets were resuspended in assay buffer, and Apopxin Green Indicator (for apoptotic cells), 7-AAD 200X (for necrotic cells), and CytoCalcein 450 (for healthy cells) were added following the manufacturer’s instructions (apoptosis/necrosis assay kit ab176749; Abcam, Cambridge, UK). After a 60-min incubation period, 300 μl of assay buffer was added. The cell suspension was analyzed using an Attune NxT flow cytometer and data analysis was carried out using FlowJo VX software. Each experiment was independently conducted three times.

Statistical analysis. Experiments were performed in triplicate and independently replicated three times. Measurement values were expressed as the mean accompanied by the standard deviation (+SD) and statistical analyses were performed using ANOVA or t-test when parametric assumptions were met. If parametric assumptions were not met, Kruskal-Wallis test and Mann-Whitney Test were used. Statistically significant differences were denoted using the following p-values *p<0.05, **p<0.01, ***p<0.001, and “ns” indicated the absence of a significant difference. Statistical analysis and graphical representation were performed using GraphPad Prism software, version 9.0. Inkscape software was used for the organization and presentation of the graphical elements.

Results

BRAF expression across melanoma skin cancer. The demographic and clinical features of this study have been published in previous articles (45-48). We focused on the analysis of data from TNMplot.com to explore BRAF expression in metastatic melanoma compared to normal tissue. Our study revealed an increased BRAF expression in metastatic melanoma compared to normal tissue [fold change (FC)=1.15], with higher expression in non-invasive or non-metastatic melanoma (FC=0.59) compared to normal tissue (Kruskal-Wallis p-value<0.0001) (Figure 1A) (36). Using the GENT2 platform, we observed significantly higher BRAF expression in invasive stage according to the Clark level F-IIIB tumors compared to F-IIIC tumors (p-value=0.006) and higher than that of F-IV tumors (p-value=0.040). Additionally, elevated BRAF expression was observed in metastatic tumors of grade F-IIIC compared to M-IIIB and M-IIIC (p-value=0.001 and 0.013). No significant differences were found in BRAF expression between F-IIIC and F-IV tumors and metastatic M-IV tumors (p-value=0.261 and 0.390, respectively). Nevertheless, significant differences in BRAF expression were identified between metastatic melanomas of grade M-IIIB compared to M-IIIC and M-IV (p-value=0.007 and 0.005, respectively) (Figure 1B) (37).

Figure 1. BRAF expression patterns in melanoma skin cancer. (A) BRAF RNA expression profiles in primary tissue (n=103, Fc=0.59, p-value <0.001) compared to normal tissue (n=474), and in metastatic tumor tissue (n=368, Fc=1.15, p-value <0.01) analyzed using the Dunn`s test method. (B) BRAF expression patterns in different subtypes based on the Clark level (invasion). (C) Co-expression profile of BRAF, NRAS, KIT, TERT, and PTEN in primary and metastatic melanoma tumors subcategorized according to their invasion grade (Clark level), using the TCGA melanoma public dataset (n=481). (D) Co-expression profile of BRAF, NRAS, KIT, TERT, and PTEN in primary and metastatic melanoma tumors subcategorized according to their invasion grade (Clark level), using the GDC melanoma public dataset (n=477).

Analysis of TCGA Melanoma (SKCM) data (n=481) revealed that 38% of the population were women and 62% were male. Among women, 22% of melanomas were classified as primary tumors, while 78% were metastatic melanomas. Similar data were found in the male population where 20% of the melanomas were classified as primary and 80% as metastatic. Primary tumors with Clark levels greater than III exhibited increased BRAF expression, whereas primary tumors with low invasive capacity (Clark levels less than III or null) were associated with high KIT expression. Metastatic melanoma tumors with Clark levels greater than III showed increased BRAF expression. Co-expression analyses demonstrated a significant positive correlation between BRAF and NRAS (r=0.3817, p-value <0.001) and PTEN (r=0.1513, p-value=0.00095), according to Pearson`s correlation coefficient (Figure 1C) (38).

Additionally, the data obtained from GDC TCGA Melanoma (SKCM, n=477), revealed that 38% of the population were women and 62% were male. Among women, 24% of the melanomas were classified as primary tumors, and 76% metastatic. Similar data were found in the male population were 23% of the melanomas were classified as primary and 77% as metastatic. Primary tumors showed an increase in BRAF expression, irrespective of their invasive capacity (Clark level ranging from null to V). Conversely, metastatic malignant tumors displayed a significant increase in BRAF expression, irrespective of the Clark level. In the co-expression analyses of BRAF with genes, such as NRAS, KIT, PTEN, and TERT, it was observed that BRAF expression was positively correlated with NRAS (r=0.3207, p-value <0.001) and there was also a positive correlation between BRAF and PTEN (r=0.1560, p-value=0.000672). In contrast, BRAF expression was negatively correlated with KIT (r=-0.1198, p-value=0.0091). No correlation was found between BRAF expression and TERT (r=0.00038, p-value=0.933) (Figure 1D) (38).

Spheroid formation and characterization. The process began with the standardization and characterization of spheroids with varying cell concentrations, followed by continuous monitoring until day 23. Spheroid formation was initiated using two different initial cell numbers (10,000 to 20,000 cells per spheroid). Observations and imaging of the spheroids were conducted throughout this period to gain a comprehensive understanding of their behavior. Both C32 and HEMa cell lines readily formed spheroids in culture using different methodologies, such as hanging drops, matrix, and low-attached plates. Initially, within the first five days, the spheroids aggregated, establishing a connection between each other, and resulting in a decrease in spheroid diameter. However, after five days, the spheroids began expanding in both diameter and area, indicating an increase in cell population. Subsequent analyses were conducted from day 5 onwards due to this growth phase. C32 spheroids exhibited robust grape-like growth (700-1,200 μm), whereas HEMa spheroids showed a compact mass (500-900 μm).

Post-baseline characterization, spheroids were exposed to dabrafenib (1-100 μM) until day 7. HEMa spheroids at 10 μM formed amorphous structures, while concentrations above 10 μM did not induce aggregation. C32 spheroids formed mini spheroids (<50% diameter compared to control) below 25 μM. Concentrations over 25 μM did not induce aggregation, reflecting potential in vivo behavior with dabrafenib exposure as a BRAF inhibitor. The results of the behavior of spheroids in response to dabrafenib treatment inform the design of future experiments.

Drug sensitivity in 2D and 3D cultures. Organoids and spheroids in 3D cultures closely emulate tissue organization, providing an in vitro model for studying sensitivity to drugs (such as dabrafenib) in cutaneous melanoma cells (49). Viability assays were conducted to assess cell sensitivity to dabrafenib under various culture conditions as depicted in the timeline of Figure 2A. For cytotoxicity assays in 2D cell cultures, cells were exposed to varying concentrations of dabrafenib for three days, while in spheroids, treatment commenced five days after initiation of culture, with spheroids treated for three days thereafter.

Figure 2. Effect of the BRAF V600E inhibitor dabrafenib on the cell growth of 2D and 3D-culture skin cell lines. A) A detailed timeline for the cell treatment. B) Two cell lines were seeded under the same conditions in both 2D and 3D cultures and treated for three days with dabrafenib. Images of cells in 2D were captured on day 3, with a scale bar of 250 μm, while spheroids were imaged on day 5, with a scale bar of 500 μm. On the right-hand side, cellular viability data post-treatment with dabrafenib are displayed for both HEMa and C32 cell lines.

In 2D cultures, both HEMa and C32 cells exhibited a mesenchymal cell phenotype, characterized by elongated shape and disordered growth. HEMa-derived spheroids showed a spherical structure with a darker necrotic core, averaging a size between 500 to 700 μm. In contrast, C32-derived spheroids exhibited a grape-like structure with an approximate size ranging between 800 to 1,000 μm. Regarding dabrafenib sensitivity, HEMa cells in 2D cultures displayed an inhibitory concentration of 50 (IC₅₀) of 24.26 μM, whereas in spheroids it was 47.25 μM, representing double the concentration required to inhibit 50% of the treated cells’ viability. In contrast, C32 cells in 2D cultures presented an IC₅₀ of 16.36 μM, whereas in spheroids, it increased to 21.05 μM, indicating an approximately 30% rise compared to cells in 2D cultures, as presented in Figure 2B. Our finding showed a significant difference in dabrafenib sensitivity between 2D and 3D cell cultures, with notable variation in IC₅₀ values for both HEMa and C32 cells.

Migration capacity on 2D cell culture. Melanoma metastasis reduces survival rates and is challenging to detect early on, with cell migration being a key characteristic of metastatic behavior. To assess the potential impact of dabrafenib on cell migration, we employed the classical 2D wound-healing assay and evaluated migration within a 3D matrix. These experiments were conducted using serum-deprived media to limit cell proliferation, enabling us to focus specifically on migration capacities in both 2D and 3D conditions and evaluate the actual effects of dabrafenib on migration. After generating the wound, HEMa and C32 cell cultures were treated with 10 and 25 μM concentrations based on the IC₅₀ results. Area measurements were obtained using microscopy, and to ensure consistency, a coordinate pattern was generated and utilized for tracking.

Under standard conditions, normal HEMa melanocyte cells exhibited a migratory capacity of 30% at 72 h compared to the negative control. However, exposure to 10 and 25 μM concentrations resulted in a decreased migratory capacity of 20%, with a significant difference observed at 25 μM (p-value=0.0105) (Figure 3A).

Figure 3. Migration activity of dabrafenib in HEMa and C32 cells. (A) Normal HEMa melanocyte cells. (B) C32 melanoma cells where growth is shown as monolayer or 2D up to 72 hours. (C) The wound was created and imaged every 24 hours following the wound scratch. The scale bars represent 650 μm. Data were represented as mean±SD, which were derived from three independent experiments, *p<0.05, **p<0.01, ***p<0.001.

Conversely, C32 tumor cells demonstrated a migratory capacity of approximately 50% at 72 h under standard conditions, indicating wound closure. Exposure to 10 and 25 μM concentrations led to a significant reduction in wound closure, ranging from 1 to 8% compared to the negative control, with p-values of 0.044 and 0.034 for cells treated with dabrafenib at 10 and 25 μM, respectively (Figure 3B).

Migration capacity on 3D cell culture. Three-dimensional tumor models enable personalized therapy based on the patient’s molecular profile, enhancing treatment responses, survival rates, and disease-free periods (23). To investigate dabrafenib’s anti-migratory effects in the spheroid model, migration assays were conducted in two stages.

In the first stage, spheroids were directly exposed to 10 and 25 μM dabrafenib for C32 and 25 and 50 μM for HEMa spheroids, seeded in Matrigel-coated wells. HEMa cells exhibited increased cell expansion on Matrigel after 48 h, while spheroids treated with dabrafenib (25 and 50 μM) showed non-migratory behavior and significant size reduction (p-value=0.033 and 0.0004, respectively). Untreated C32 spheroids displayed adhesion and expansion, whereas those treated with 10 and 25 μM dabrafenib exhibited reduced size and detachment (p-value=0.010 and 0.0031 compared to the control) (Figure 4A).

Figure 4. 3D migration of HEMa normal melanocyte and C32 melanoma tumor spheroids on the top and embedded culture. (A) Graphical representation of cell migration models. (B) Spheroids were seeded on top of the Matrigel matrix exposed to dabrafenib, while other spheroids were embedded in the Matrigel matrix and exposed indirectly to dabrafenib. (C) The total covered area was calculated and then normalized against the initial spheroid size (mean±SD). The scale bars represent 650 μm. *p<0.05, **p<0.01, ***p<0.001. Data are representative of at least three independent experiments.

In the second stage, spheroids were embedded in Matrigel layers and indirectly exposed to dabrafenib. HEMa spheroids treated with dabrafenib showed no drastic size reduction but lacked migration through the Matrigel matrix. In the negative control, cells migrated, and spheroid size increased at 120 h (p-value=0.0036 and 0.00032 at 25 and 50 μM). C32 spheroids exhibited initial migration, but cells began to die by 72 h with indirect dabrafenib exposure, leading to significant disintegration at 120 h (p-value=0.018 at 25 μM) (Figure 4B).

Cell cycle arrest in 2D and 3D cell cultures. Cancer treatments aim to limit tumor expansion and induce selective cell death in rapidly proliferating cells, disrupting tumor growth. To assess the effects of dabrafenib on the cell cycle, we examined its impact on cell populations. Cell cycle analysis revealed significant differences in treatment responses between HEMa and C32 cell cultures in both 2D and 3D models. To synchronize cells, cells underwent FBS deprivation for 24 h across all conditions, including negative control, vehicle control, and dabrafenib treatment, in both 2D and 3D cultures. No differences were observed between measurements of the negative control and vehicle control in the assay.

In the 2D models, dabrafenib influenced cell cycle progression in the G₁ phase, with an increase of 10% on day 1 (70.2%) and day 3 (72.5%) compared to the control (60%) in HEMa cells treated with 25 μM. Conversely, in C32 cells, a cell cycle arrest in the G₁ phase was observed, with levels at 58.3% for the negative control increasing to 69% on day 1 after exposure to 25 μM. However, there was a decrease to 39% in the G₁ phase on day 3, indicating an increase in cell death (Figure 5A-C).

Figure 5. Cell cycle analysis in HEMa and C32 cells cultured in 2D and 3D models. A) Behavior of HEMa and C32 cells cultured in monolayer/2D treated with dabrafenib at a concentration of 25 μM. B) Behavior of spheroids treated with dabrafenib at concentrations of 25 μM. C) Cell cycle phase distribution of HEMa and C32 cells cultured in 2D model. D) Cell cycle phase distribution of HEMa and C32 cells cultured in 3D models, represented in bar graphs based on the percentage of cells in the G₁, S, G₂, and sub G₁ phases at 1 and 3 days of treatment with dabrafenib at 25 μM. Data are representative of at least three independent experiments.

In the 3D models, dabrafenib affected cell viability and cell cycle progression in HEMa spheroids treated at a concentration of 25 μM. The cells in the G₁ phase decreased from 56.6% in the negative control to 35.7% on day 1 and 45.6% on day 3, with an increase in cells in the death phase from 27.25% for the negative control to 48.5% on day 3 at 25 μM. For C32 spheroids, a similar pattern was observed at concentrations of 10 and 25 μM, transitioning from 54.4% of cells in the G₁ phase to 39.9% on day 3. Cells in the death phase increased by 21% from day 1 to day 3 compared to the negative control (Figure 5B-D). Our findings indicate that dabrafenib treatment influences cell cycle progression and induces cell death in both 2D and 3D melanoma cell models.

Cell death in 2D and 3D cell cultures. Building upon data from the previous experiment, our observations suggest that dabrafenib may induce senescence of cells, ultimately leading to apoptosis. To investigate differential sensitivity mechanisms of dabrafenib in 2D and 3D models, we studied two cell lines: HEMa and C32 with wild-type and V600E mutation in the BRAF gene. Cell death analysis revealed notable differences in behavior between HEMa and C32 cell cultures in both 2D and 3D models concerning live, apoptotic, and necrotic cells.

In the 2D model, HEMa cells exposed to 10 and 25 μM dabrafenib exhibited a gradual increase in apoptotic cells, from 3.97% to 5.73% on day 1, significantly rising to 39.6% on day 3 at 25 μM. Conversely, C32 cells at 25 μM showed a marginal increase in apoptotic cells from 9.37% on day 1 to 12.2% on day 3, compared to the 6.75% control (Figure 6A-C). In the 3D model, normal melanocytes exposed to 25-50 μM dabrafenib displayed a 20-30% increase in apoptotic cells on day 3. For C32 cell spheroids, concentrations of 10 and 25 μM induced apoptotic cell death, increasing from 12% to 15% after three days (Figure 6B-D). Our findings suggest that dabrafenib treatment elicits differential responses in cell lines HEMa and C32, with distinct effects observed in 2D and 3D models.

Figure 6. Detection of apoptotic cells in HEMa and C32 cell cultures in 2D and 3D models. A) Cells in 2D or monolayer models for both HEMa and C32 cells treated with concentrations of 25 and 10 μM, respectively. Each condition is represented in the dot plot on the flow cytometer diagram using a combination of Apopxin Green-FITC for cells in early apoptosis and 7-AAD for cells in necrosis. B) Spheroids treated with the same concentrations mentioned earlier for each cell line, focusing on apoptotic and necrotic cells up to day 3 of exposure. C) Different expression values as percentages of both apoptotic and necrotic cells in day 1. D) Different expression values as percentages of both apoptotic and necrotic cells in day 3. Data are representative of at least three independent experiments.

Association between BRAF expression and melanoma skin cancer outcomes. To validate our findings, we conducted in vitro experiments using C32 melanoma skin cancer cell lines. Also, we performed analyses based on publicly available datasets to investigate the association between BRAF expression and outcomes in patients with melanoma skin cancer. High BRAF expression was not significantly associated with overall patient survival when compared to low BRAF expression (p-value=0.6927; Figure 7A). Conversely, high BRAF expression was significantly associated with a lower progression-free interval (p-value=0.02680; Figure 7B) in patients with melanoma skin cancer. Kaplan-Meier curves were generated online using the Xena visualization tool. This highlights the importance of BRAF expression in disease-free survival.

Figure 7. Association between BRAF expression and melanoma skin cancer outcomes. A) Correlation between BRAF expression and melanoma overall survival. B) Correlation between BRAF expression and melanoma progression-free interval. Kaplan-Meier curves were generated online using the Xena visualization tool (https://xenabrowser.net, accessed on October 9, 2023). C) Association between BRAF expression and overall survival in melanoma skin cancer patients; Kaplan-Meier curves were created online at (https://tnmplot.com/, accessed on October 9, 2023). D) BRAF expression in melanoma samples from patients treated with dabrafenib who relapsed (non-responder) compared to those who did not relapse (responder) in the 5 years following diagnosis. Boxplots were generated online using ROCplot (http://www.rocplot.org/treatment/, accessed on October 10, 2023).

Furthermore, survival analysis data indicated that high BRAF expression was significantly associated with lower overall survival (OS; HR=1.41; log-rank p<0.0001; Figure 7C) in patients with melanoma skin cancer, as per the Kmplot.com gene expression public dataset. ROCplot analyses also confirmed that after stratification based on molecular subtype according to BRAF expression, patients treated with dabrafenib who relapsed within five years following diagnosis (n=1,267), showed a significant difference between non-responders with relapse (n=809) and responders without relapse (n=458) [Mann-Whitney p-value=0.022; (Figure 7D)].

Discussion

In this study, we conducted a validation of previous studies examining BRAF expression in patients with primary melanoma tumors and metastatic melanoma, comparing them with non-tumoral skin tissue, as analyzed by public datasets (36). Within the analyzed public datasets, elevated BRAF expression levels were confirmed in metastatic melanoma tumors compared to non-tumoral tissue and primary tumors. This correlation could be linked to increased proliferation, invasion, migration, and evasion of apoptosis in tumor cells harboring the BRAF V600E mutation (49). Additionally, BRAF expression was found to correlate with clinical characteristics of malignant tumors with invasive features (Clark III or IV) (50).

Correlation analyses allowed us to infer the interaction between BRAF and NRAS, KIT, PTEN, and TERT, showing both positive and negative correlations, primarily with KIT. A strong positive correlation between BRAF and KIT was evident in metastatic tumors, invasive tumors, and invasive primary tumors. This finding is consistent with a 2021 study suggesting that the interaction between BRAF and KIT could influence the prognosis of melanoma patients (51). To confirm dabrafenib’s response to the BRAF V600E mutation, we assessed in vitro models (2D and 3D) using HEMa and C32 cell lines. For cells without the mutation, the IC₅₀ doubled from 24.26 μM (2D) to 47.25 μM (3D), whereas malignant cells maintained similar IC₅₀ values (16.38-21.05 μM), consistent with earlier studies. These findings are also consistent with studies on A375 cells with BRAF V600E mutations treated with dabrafenib, with an IC₅₀ of 9.5 μM, or in resistant cells an IC₅₀ of 110.5 μM (52,53).

The formation of in vitro 3D spheroids models is emerging as an essential tool for cancer research. It allows a closer approximation to in vivo tumors during their formation and disease progression (54). Over 23 days, we observed grape-like and mass-like spheroids. Treatment of cells with dabrafenib (10 μM, 25 μM) influenced spheroid generation, aggregation, compaction, and growth of spheroids derived from normal and tumor cells. Similar results were obtained in A375 cells, both resistant and sensitive to dabrafenib, when exposed to compounds such as onconase (ONC), which affected colony formation, spheroid formation, migration, and invasive capacity (55).

Despite complex 2D to 3D transitions, comprehending compound effects on cell migration is vital. Normal HEMa cells demonstrated a wound closure process at 72 h. Additionally, 25 μM dabrafenib significantly reduced normal cell migration. In contrast, tumor cells showed high migratory capacity with approximately 50% wound closure at 72 h. However, dabrafenib (10 μM and 25 μM) induced a complete inhibition of BRAF-mutated cell migration.

In metastatic and invasive processes, the mesenchymal-epithelial transition plays a crucial role in tumor migration (56). To replicate in vivo conditions and to simulate the normal adhesion of circulating cells in colonization processes, spheroids were seeded on a thin layer of Matrigel. Tumor cells aggregated into spheroids and initiated colonization of the matrix within 24 h, while normal cells took 48-72 h, showing reduced colonization. Dabrafenib significantly hindered both tumor and normal cell colonization and adhesion, impacting migratory, adhesive, and tissue colonization capacities. Our study revealed reduced invasive capacity in tumor and normal cells in Matrigel-embedded spheroids after 48 h, after indirect exposure to dabrafenib. To compare the effects of dabrafenib on proliferation, cell division, and tumor progression both in vivo and in vitro, we conducted cell cycle assays using 2D and 3D models of normal and tumor cells. Dabrafenib induced G₁-S arrest in normal cells in both models, leading to an increase in cells in the sub-G₁ phase, which indicates senescence in the 3D model. In tumor cells, both 2D and 3D models exhibited G₁ phase arrest and an increase in sub-G₁ cells, indicative of senescence. Previous studies on 1205Lu and WN1366 melanoma cell lines have demonstrated that exposure to dabrafenib inhibited their growth and division, by targeting CDK4, phosphor-CHK1, p21 expression and NEK9 and CDK16 proteins. This was observed in both sensitive and resistant cells (57,58) and holds true for other BRAF mutations, such as V600K, in various types of mucosal melanoma cells (59,60).

Upon inducing cell cycle arrest, the cell death assay revealed that less than 5% of normal and tumor cells exhibited necrotic cell death. Between 20% and 40% of cells in both 2D and 3D cultures underwent apoptosis when exposed to dabrafenib (10 μM to 25 μM for tumor cells and 25 μM to 50 μM for normal cells). Dabrafenib induced endoplasmic reticulum stress and autophagy in A375 and MEL624 tumor cells (61). Additionally, in 23 mutant BRAF melanoma cell lines, concentrations above 5 nM led to increased apoptosis in 40% of the cell lines (53). This apoptotic induction involved MAPK kinase pathway inhibition, caspase activation, loss of mitochondrial membrane potential, and cytochrome C release, as observed in A375, MEWO, and SK-MEL-23 cell lines (62).

The data from diverse cellular models (2D and 3D) mark a significant advance in cancer research, replicating in vivo events. Spheroid studies revealed that cell-matrix interactions influence tumor cell behavior (54). 3D models enable anticipating patient tumor behavior before targeted therapy, offering insights into resistance model generation as cells adapt to microenvironmental abnormalities around tumor tissue post-exposure to these compounds (54). Our study indicates that these spheroids hold promise as effective tools for rapid preclinical drug screening, facilitating the evaluation of the behaviour of novel therapeutic compounds. Future directions that stem from this study should focus on the characterization of the underlying molecular mechanisms underlying the responses to dabrafenib treatment within this and other 3D spheroid models. Investigations into combination treatments involving dabrafenib and other therapeutic compounds should be conducted using this and similar models.

Conclusion

Our investigation into melanoma progression and therapeutic responses reveals a crucial interplay between drug sensitivity and cellular context. The use of C32 spheroid models provided a more physiologically relevant setting, showcasing differential responses in both normal and BRAF V600E positive tumor cells with respect to dabrafenib’s impact on migration and adhesion. Notably, our study highlights the significance of employing both traditional monolayer and advanced 3D culture models in melanoma research, as this duality enriches our understanding of changes triggered by drug treatments.

Funding

This research was financed by the Malta Council for Science and Technology through the Technology Development Programme (TDP-Lite) 2022 (R & I-2022-008L).

Conflicts of Interest

The Authors declare no conflicts of interest in relation to this study.

Authors’ Contributions

D.T.P., data collection or analysis and interpretation of data; experiments in cells, statistical analysis; writing of the manuscript or critical review of important intellectual content; critical review of the dataset, technical, and material support. M.Z.M, the study concept and design; funding, administrative, interpretation of data; writing of the manuscript or critical review of important intellectual content; interpretation; final approval of the final version of the manuscript.