MEK Inhibition, Alone or in Combination with BRAF Inhibition, Affects Multiple Functions of Isolated Normal Human Lymphocytes and Dendritic Cells

As a result of the identification of mutations in the BRAF kinase as an oncogenic driver in cutaneous melanomas (1, 2), BRAF kinase inhibitors (BRAFi) have now become the standard of care for treatment of patients with metastatic melanoma bearing V600 mutations (3). Unfortunately, clinical benefit is generally short-lived and BRAFi-resistant disease develops with a median of 6 to 7 months (4, 5). Multiple mechanisms have been described, but, importantly, each of the resistance mechanisms has the common theme of reactivating mitogen-activated protein kinase (MAPK) pathway signaling or activating parallel signaling pathways that drive cell proliferation and survival (6–14).

Trametinib (GSK1120212; GlaxoSmithKline Pharmaceuticals) is an orally available, small-molecule, selective inhibitor of MEK1 and MEK2 (15). Monotherapy with trametinib improved progression-free and overall survival when compared with chemotherapy administered to patients who had BRAF-mutant melanoma, validating trametinib as a therapeutic approach (16). In an attempt to extend the duration of clinical benefit, MEK inhibitors (MEKi) have been combined with BRAFi therapy (17), with the combination of trametinib with dabrafenib (GSK2118436; GlaxoSmithKline Pharmaceuticals) that resulted in a 3.5-month prolongation of progression-free survival when compared with dabrafenib monotherapy. The addition of trametinib with dabrafenib, therefore, represents an effective strategy for delaying the emergence of BRAFi resistance (17).

Immunotherapies have also shown considerable promise in the treatment of metastatic malignancy, and inhibitory antibodies that target the molecular immune checkpoints CTLA-4 and PD-1 have been reported to induce durable clinical remissions as well as prolong survival in responding patients (18–20). The anticancer effects of these agents are mediated by cellular effectors of the immune system. This has provided a rationale for combining MAPK pathway inhibitors with immunotherapy, as BRAFi can sensitize BRAF-mutant melanoma cells to immune recognition and relieve immune suppression (21–26). Amid the speculation as to the role for combined MAPK pathway inhibition and immunotherapy, the degree to which kinase inhibition, specifically dabrafenib and/or trametinib, may directly affect immune function remains poorly defined.

To optimally combine kinase inhibitors with immunotherapy, it is critical to understand how the combination of dabrafenib and trametinib influences immune cell subpopulations. We assessed the effects of dabrafenib and trametinib, alone and in combination, on healthy donor T lymphocytes and monocyte-derived dendritic cells (moDC).

BRAF inhibitor (GSK2118436, dabrafenib) and MEK1/2 inhibitor (GSK1120212, trametinib) were provided by GlaxoSmithKline (GSK) and solubilized in 100% dimethyl sulfoxide (DMSO) at 10 and 2.5 mmol/L, respectively. Stock solutions were stored at −20°C and physiologically relevant concentrations tested (27, 28). All phenotyping antibodies were from BD Biosciences, unless indicated otherwise. NY-ESO-1–specific CD8⁺ T-lymphocyte clones were generated as described (29, 30). Complete media contained RPMI-1640 with 20 mmol/L HEPES, 60 mg/L penicillin, 12.5 mg/L streptomycin, 2 mmol/L L-glutamine, 1% nonessential amino acids, and 10% heat-inactivated fetal calf serum (Invitrogen). All flow cytometry was performed on a BD FACSCanto II, and data analyses were performed using the FlowJo software (TreeStar Inc).

Peripheral blood mononuclear cells (PBMC) from buffy coat preparations from healthy donors (Red Cross Blood Bank) were prepared by Ficoll-Paque (GE Healthcare) density gradient centrifugation. T cells were isolated by positive selection using magnetic anti-CD4 or -CD8 microbeads (Miltenyi Biotech). Cells were counted and selection confirmed by flow cytometry (purity confirmed to be ≥90%).

Purified CD4⁺ and CD8⁺ T lymphocytes were labeled with 10 μmol/L carboxyfluorescein succinimidyl ester (CFSE) dye (Invitrogen) at 37°C, washed, and plated at 1 × 10⁵ cells per well in a 96-well U-bottom plate (BD Falcon). Cells were stimulated with beads coated with anti-CD3/CD28 antibodies (Invitrogen) at a 2:1 ratio for 5 days, with or without the indicated concentration of MAPK inhibitors. Media and inhibitors were replenished on day 3. After staining with a violet live/dead fluorescent viability dye and anti-CD3, anti-CD4, and anti-CD8 fluorescent antibodies, cells were analyzed on the BD FACSCanto II, gating for lymphocytes with FSC/SSC parameters. Data analysis was performed using the FlowJo proliferation tool (TreeStar). The “proliferation index” was used to define a proliferation value. This value, which represents the average number of cell divisions undergone by the responding cells, more faithfully reflects the biology of the system by considering only the fraction of proliferating cells (http://www.flowjo.com/v7/html/proliferation.html).

PBMCs were plated at 1 × 10⁵ cells per well in a 96-well U-bottom plate. Cells were cultured in the presence and absence of the mitogens phorbol myristate acetate (PMA, Sigma Aldrich; 5 ng/mL) and ionomycin (1 μmol/L; Sigma-Aldrich) for 24 hours, with or without the indicated concentrations of the BRAF and/or MEK inhibitors. Golgi plug (CD8⁺ culture) and Golgi STOP (CD4⁺ culture) was added for the final 12 hours to halt cytokine secretion. Cells were then stained with anti-CD3, -CD4, and -CD8 antibodies, permeabilized and stained for IFN-γ, TNF-α, and interleukin (IL)-17, and analyzed using a BD-FACSCanto II. Data analyses were performed using FlowJo, gating on viable CD3⁺CD4⁺ and CD3⁺CD8⁺ cytokine-producing T cells.

Immature moDCs were generated by culturing CD14⁺ monocytes with granulocyte macrophage colony-stimulating factor (20 ng/mL) and IL-4 (10 ng/mL) for 4 days. On day 5, moDCs were plated at 1 × 10⁵ cells per well in a 96-well U-bottom plate with or without BRAF and/or MEK inhibitors in the presence of 100 ng/mL lipopolysaccharide (LPS; Sigma Aldrich) for 24 hours. Phenotypic analysis was performed using the following antibodies: anti–HLA-DR, anti-CD86, and anti-CD83 in combination with a violet live/dead fluorescent dye (Invitrogen). Cells were analyzed using a BD FACSCanto II and data analyses were performed using FlowJo, gating on viable moDCs (live/dead dye exclusion).

In vitro antigen presentation experiments were performed with moDCs pulsed for 16 hours with (i) immune complexes formed with purified recombinant NY-ESO-1 protein produced in Escherichia coli (GSK Vaccines; refs. 30, 31) and anti-His₆ Ab (mouse monoclonal MAB050; R&D Systems); (ii) control recombinant NY-ESO-1 protein and anti-IgG1 isotype control (mouse monoclonal MAB002; R&D Systems); and (iii) the cognate peptide recognized by the respective T-cell clone in the absence or presence of different concentrations of BRAF and MEK inhibitors. After 16 hours, dendritic cells (DC) were washed and used as antigen-presenting cells (APC) for NY-ESO-1–specific CD8⁺ T-cell clones (HLA-A2/NY-ESO-1_157–165) at a 1:10 T cell-to-APC ratio with GolgiPlug for 4 hours. T-cell activation was measured by intracellular cytokine staining (ICS) for IFN-γ. Where indicated, before the ICS assessment, DCs were pulsed for 1 hour with 1 μg/mL of the NY-ESO-1 peptide followed by washing.

Cells were treated with media containing the indicated kinase inhibitors or equal amount of DMSO for the indicated time. Viability was assessed using violet live/dead fluorescent dye (Invitrogen) and analyzed using a BD FACSCanto II and data analysis performed using the FlowJo software (TreeStar Inc.).

HLA-A2⁺ healthy donor PBMCs were plated at 1 × 10⁶ cells per well in a 48-well plate in culture media to which 1 μg/mL (final concentration) of EBV BMLF1_280-288 (GLCTLVAML) peptide were added with 25 IU recombinant hIL-2, in the presence or absence of the BRAF and MEK inhibitors. This treatment was repeated every 3 days until day 10. On day 10, the inhibitors were removed, and the cells were washed thoroughly and then restimulated with or without 1 μg/mL BMLF1 peptide for 4 hours in the presence of Golgi-stop. T-cell activation was measured by ICS for IFN-γ. Flow cytometry was performed using the BD FACSCanto II and data were analyzed using FlowJo, gating on viable CD3⁺CD8⁺ cytokine-producing T cells.

The BRAFi vemurafenib and its derivative PLX4720 do not have a negative effect on healthy donor T lymphocytes (21, 32), whereas MEK inhibitors, U0126, PD0325901 (21), and PD98059 (33, 34), have been shown to have negative effects on T-lymphocyte proliferation and cytokine secretion. The effect of dabrafenib and trametinib, alone and in combination, on healthy human T lymphocytes has not yet been reported. Therefore, we examined the effects of trametinib and dabrafenib, alone and in combination, on stimulated CD8⁺ and CD4⁺ T lymphocytes by assessing cellular proliferation, viability, and cytokine secretion.

To investigate the effects of kinase inhibition on lymphocyte proliferation and viability, purified CD4⁺ and CD8⁺ T lymphocytes were isolated from normal donor blood, labeled with the membrane-incorporating dye CFSE, and activated with anti-CD3/CD28 beads in the presence of dabrafenib or trametinib alone or in combination. After 5 days, the proliferation index and viability for each treatment condition and population was determined by Flow cytometry. Proliferation plots for activated CD4⁺ and CD8⁺ T lymphocytes are shown in Fig. 1A and B. A statistically significant decrease in proliferation of CD4⁺ and CD8⁺ T lymphocytes was observed following trametinib or concurrent treatment, whereas following dabrafenib alone proliferation was comparable with that of the control. No effect on CD4⁺ or CD8⁺ T-lymphocyte viability was observed (Fig. 1C and D).

Cytokine-producing CD4⁺ and CD8⁺ T lymphocytes were enumerated by ICS 24 hours following nonspecific activation with PMA–ionomycin in the presence of kinase inhibitors (Fig. 2). Significant differences were seen for IFN-γ- and IL-17–producing CD4⁺ T lymphocytes (Fig. 2A and B), and IFN-γ– and TNF-α–producing CD8⁺ T lymphocytes (Fig. 2C and D). Trametinib alone or in combination reduced the frequency of cytokine-producing cells even at the lowest concentration (3 nmol/L; Fig. 2).

Taken together, these data demonstrate that treatment of human T lymphocytes with trametinib alone or in combination with dabrafenib both constrains proliferation and inhibits cytokine production by both CD4⁺ and CD8⁺ T lymphocytes.

To investigate the effect of dabrafenib, trametinib, and their combinations on lymphocyte activation and proliferation, in vitro peptide antigens were used to stimulate antigen-specific CD8⁺ T lymphocytes in whole PBMC cultures. An immunodominant epitope from the Epstein-Barr virus (EBV) BMLF1 protein, (peptide position 280–288) was selected to stimulate memory responses, as the majority of healthy HLA-A2⁺ donors generally have a preexisting immune responses to this epitope (35). PBMCs were stimulated in vitro with BMLF1 in the presence of titrating concentrations of the inhibitors. After 10 days in culture, cells were restimulated with the BMLF1 peptide and PBMCs (Fig. 3A), CD4⁺ and CD8⁺ (Fig. 3B) viability was determined, and reacting CD8⁺ T lymphocytes were enumerated by ICS for TNF-α (Fig. 3C) or IFN-γ (Fig. 3D). Figure 3C and D show a concentration-dependent decrease in the generation of antigen-specific T lymphocytes with trametinib alone and in combination (6 donors). No statistically significant change with any tested dose of dabrafenib was observed when compared with the DMSO control (Fig. 3C and D). This result suggests that the MEK inhibitor trametinib specifically inhibits the activation of antigen-specific T lymphocytes.

If dabrafenib and trametinib are to be employed in combination with immunotherapeutics, it is critical to understand their potential effect not only on T-cell activation and function but also on the capacity of the DCs to process and present antigen. Therefore, we studied the effects of trametinib and dabrafenib alone and in combination on the ability of human moDCs to process and present the cancer–testis antigen NY-ESO-1 (30, 31). MoDCs were isolated from HLA-A*0201⁺ donors and incubated with NY-ESO-1 as an immune complex (20 μg/mL anti-His₆ antibody), in the presence or absence of inhibitors (Fig. 4A). NY-ESO-1_157–165 peptide-pulsed DCs (Fig. 4B) and NY-ESO-1 protein with anti-IgG1 isotype control (Fig. 4C) was used as a control. The HLA-A*0201/NY-ESO-1_157–165 epitope was processed and presented efficiently as assessed by the response by an epitope-specific CD8⁺ T-cell clone in a 6-hour ICS assay for IFN-γ (Fig. 4A). There was significantly reduced, dose-dependent presentation of this epitope in the presence of trametinib when combined with dabrafenib (P < 0.05).

To confirm these findings using a second epitope, the ability of HLA-Cw3 moDCs to cross-present NY-ESO-1 epitope (92–100) was examined (31, 36). Similarly, but to a lesser extent, the presentation of this epitope by moDC was reduced by 30 nmol/L trametinib and the combinations containing 10 and 30 nmol/L trametinib (data not shown).

It has been shown previously that inhibition of ERK signaling potentiates DC maturation (37–40) and mature DCs are less effective at taking up antigen for processing than immature DCs. We therefore sought to investigate this further as a potential explanation for the negative impact of kinase inhibition on antigen cross-presentation. MoDCs were incubated with LPS in the presence of kinase inhibitors for 24 hours. LPS alone upregulated the expression of CD86, CD83, and HLA-DR that was further augmented by trametinib alone or in combination with dabrafenib (Fig. 5A). The effect of trametinib was reflected in both the percentage of positive cells and the median fluorescent intensity (MFI) of the whole population of cells, demonstrating that inhibition of the ERK signal transduction pathway enhances phenotypic maturation as has been shown previously by others (refs. 39, 41; Fig. 5A). No effect on the normal maturation pathway was observed when the MEKi was used alone or in combination (in the absence of LPS; data not shown). The toxicity exerted by the individual and combined inhibitors was assessed using a fluorescent live/dead dye. There was a significant increase in the number of dead moDCs following 24- and 48-hour combination treatment and a dose-dependent increase (trend, not significant) in cell death when incubated with trametinib for 48 hours (Fig. 5B).

Targeting oncogenic BRAF with kinase inhibitors has been one of the great success stories in the treatment of melanoma (3, 5, 42), and combination with MEK inhibition has further prolonged the survival of patients with metastatic disease (16, 17). Unfortunately, relapse is the rule rather than the exception, and the addition of effective agents will be required to build on these early successes (43). Among the most promising other therapies are the immunotherapeutics, and antibodies that target immune checkpoints such as CTLA-4 and PD-1/PD-L1/PD-L2 have attracted considerable attention (18–20). Clinical approaches combining kinase inhibitors with immunotherapeutics are a logical and evolving next step. To inform these approaches, it is helpful to better understand the effects of kinase inhibition on immune cell function.

There are limited data showing the effect of dabrafenib, trametinib alone, or their combination on immune cell function. Because the efficacy of immunotherapy is dependent on immune cell function, it is useful to ascertain the effects of these kinase inhibitors alone and in combination on these cell subsets before attempting any combination strategies. The current study sought to assess the effects of these drugs alone and in combination on isolated immune cells obtained from healthy donors. We show that while dabrafenib had no demonstrable inhibitory effects on human CD4⁺ and CD8⁺ T lymphocytes or moDCs, trametinib inhibited or modulated a variety of functions.

The idea of combining BRAFi with immunotherapy arose following the discovery that expression of melanoma differentiation antigens (MDA; refs. 21, 32) and other tumor antigens, including cancer–testis antigens (22), on melanoma cells increased in response to vemurafenib (or its derivative, PLX4720). The increase in MDA and tumor antigen expression was associated with enhanced antigen-specific T-lymphocyte recognition in vitro (21, 22) and in clinical trial patients (24), providing the first evidence for combining vemurafenib treatment with T-lymphocyte–directed immunotherapy. BRAFi also enhances the levels of MHC molecules on melanoma cells, reverses immune suppression of T lymphocytes by inhibiting IL-1α and IL-1β (25), and inhibits VEGF secretion from tumor cells (26), providing an additional explanation for why T-cell recognition is enhanced.

Regarding the direct effect of kinase inhibition on healthy T lymphocytes, previous in vitro studies have shown that vemurafenib and its derivative (PLX4720) do not compromise T-lymphocyte function (21, 32), whereas MEKi U0126 and PD0325901 are detrimental. We asked whether the function of normal healthy donor T lymphocytes was affected by direct dabrafenib or trametinib or combined treatment. Trametinib at a monotherapy dose of 10 nmol/L or in combination (2 μmol/L dabrafenib combined with 3 nmol/L trametinib) significantly decreased T-lymphocyte proliferation and resulted in the dose-dependent suppression of cytokine-producing CD4⁺ T lymphocytes and CD8⁺ T lymphocytes. In contrast, dabrafenib had no effect on T-lymphocyte proliferation or the percentage of cytokine-positive cells. This result parallels the in vitro findings with the other BRAFi (vemurafenib) and its derivative (PLX4720), which have no effect on PBMC viability or proliferation (32), T-lymphocyte proliferation (21), or cytokine secretion (21, 23).

MEK inhibitors are not selective and inhibit ERK phosphorylation regardless of genotype or cell type. The MEK inhibitors U0126 and PD0325901 have been shown to have deleterious effects on T-cell viability, proliferation, and IFN-γ production (21), whereas trametinib suppressed TNF-α and IL-6 production from PBMCs and reactivation of antigen-specific memory T lymphocytes (44). We demonstrate that trametinib has deleterious effects on T-lymphocyte proliferation and activation of antigen-specific T lymphocytes.

In response to maturation stimuli, inhibition of ERK activation can result in enhanced DC maturation with higher expression of MHC class II, costimulatory, and adhesion molecules (38, 39), indicating that ERK signaling helps in the maintenance of the immature state of DCs (37). The MEK inhibitors U0126 and PD98059 have been shown to promote the maturation of moDCs in the presence of inducing agents LPS or TNF-α (37, 39). Similarly, we found that in the presence of LPS, trametinib and the combined inhibitors enhanced phenotypic maturation. In a recent article, Ott and colleagues report on their finding that maturation is inhibited by U0126; however, they suggest that the differences in the maturation stimuli are one important factor that might account for some of the inconsistencies (45).

The fact that trametinib and combined inhibitors enhanced phenotypic maturation provides a possible explanation why the NY-ESO-1 cross-presentation was impaired when moDCs were pretreated with trametinib and the combined inhibitors. Previous studies with PD98059 illustrated that blockade of the MAPK–ERK pathway not only potentiates phenotypic maturation but also enhances the loss of endocytic activity. This suggests that moDCs matured in the presence of trametinib and combined inhibitors lose their ability to capture and/or internalize NY-ESO-1–antibody immune complexes resulting in reduced cross-presentation of both NY-ESO-1_92–100 and NY-ESO-1_157–165 epitopes.

In vivo, BRAF inhibition has no detectable negative impact on existing systemic immunity or the de novo generation of tumor-specific T lymphocytes in patients with BRAF-mutant melanoma (46). In vivo BRAF inhibitors increase both the number of tumor-infiltrating lymphocytes (24, 47) and the intratumoral CD8⁺ T:FoxP3⁺CD4⁺ T-cell ratio, which has been linked to the downregulation of tumor CCL2 gene expression and production (48), and they inhibit the generation of myeloid-derived suppressor cells. Together, these studies demonstrated that BRAF inhibition engages a host immune response and that future trials combining BRAF inhibitors with immunotherapy would not be expected to impair but rather prolong clinical response. Immunotherapeutics are inducing durable remissions in a significant proportion of patients (18–20), and clinical trials that combine kinase inhibition with immunotherapeutics are a potentially promising strategy; however, these approaches should incorporate a careful evaluation of any effects on host immunity, as interference with immune function has the potential to compromise long-term outcomes.

J.S. Cebon has received honoraria from the speakers' bureau and serves as a consultant/advisory board member for GlaxoSmithKline. No potential conflicts of interest were disclosed by the other authors.

Conception and design: A.-L. Puaux, J. Louahed, W. Chen, J.S. Cebon

Development of methodology: L.J. Vella, A. Pasam, A. Knights, W. Chen, K. Woods

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L.J. Vella, A. Pasam, N. Dimopoulos, A. Knights

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L.J. Vella, A. Pasam, N. Dimopoulos, M. Andrews, J. Louahed, W. Chen, J.S. Cebon

Writing, review, and/or revision of the manuscript: L.J. Vella, A. Pasam, M. Andrews, A. Knights, A.-L. Puaux, J. Louahed, W. Chen, K. Woods, J.S. Cebon

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L.J. Vella, A. Pasam, J.S. Cebon

Study supervision: W. Chen, J.S. Cebon

This study was supported by grants from GlaxoSmithKline Biologicals and S.A. Rue de I'Institut 89 (Rixensart, Belgium). Operational Infrastructure Support Program funding was also provided by the Victorian Government.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.