Clinical profiling of BCL-2 family members in the setting of BRAF inhibition offers a rationale for targeting de novo resistance using BH3 mimetics

Abstract

While response rates to BRAF inhibitiors (BRAFi) are high, disease progression emerges quickly. One strategy to delay the onset of resistance is to target anti-apoptotic proteins such as BCL-2, known to be associated with a poor prognosis. We analyzed BCL-2 family member expression levels of 34 samples from 17 patients collected before and 10 to 14 days after treatment initiation with either vemurafenib or dabrafenib/trametinib combination. The observed changes in mRNA and protein levels with BRAFi treatment led us to hypothesize that combining BRAFi with a BCL-2 inhibitor (the BH3-mimetic navitoclax) would improve outcome. We tested this hypothesis in cell lines and in mice. Pretreatment mRNA levels of BCL-2 negatively correlated with maximal tumor regression. Early increases in mRNA levels were seen in BIM, BCL-XL, BID and BCL2-W, as were decreases in MCL-1 and BCL2A. No significant changes were observed with BCL-2. Using reverse phase protein array (RPPA), significant increases in protein levels were found in BIM and BID. No changes in mRNA or protein correlated with response. Concurrent BRAF (PLX4720) and BCL2 (navitoclax) inhibition synergistically reduced viability in BRAF mutant cell lines and correlated with down-modulation of MCL-1 and BIM induction after PLX4720 treatment. In xenograft models, navitoclax enhanced the efficacy of PLX4720. The combination of a selective BRAF inhibitor with a BH3-mimetic promises to be an important therapeutic strategy capable of enhancing the clinical efficacy of BRAF inhibition in many patients that might otherwise succumb quickly to de novo resistance. Trial registrations: ClinicalTrials.gov NCT01006980; ClinicalTrials.gov NCT01107418; ClinicalTrials.gov NCT01264380; ClinicalTrials.gov NCT01248936; ClinicalTrials.gov NCT00949702; ClinicalTrials.gov NCT01072175.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Pre treatment BCL-2 expression in tumors of patients with metastatic melanoma negatively correlates with response to BRAF inhibition.

(A) Linear regression between duration of patient response to BRAF inhibition and BCL-2 expression levels relative to endogenous gene βActin prior to treatment (n = 17, R = −0.56, p = 0.0097, dotted red lines indicate 95% confidence intervals). (B) Patients who progress in less than 5.3 months have higher levels of BCL2 mRNA expression relative to βActin pre-treatment than patients with an above average duration of response to BRAF inhibition (P≤0.05).

Figure 2. BRAF inhibition is associated with changes in BCL2 family member expression in tumors of patients with metastatic melanoma.

mRNA levels of BIM, BCL2-XL, BCL2-W, BID increased in patients with metastatic melanoma undergoing treatment with a selective inhibitor of BRAFV600E while mRNA levels of MCL-1 decreased; BCL2 levels did not change significantly across patients. (A) mRNA expression levels of each gene from pre and on treatment biopsies from each patient were quantified by real-time PCR and are plotted as log fold change. Each number along the x-axis indicates an individual patient identifier, the y-axis indicates the mRNA level changes of BCL-2 family members for each patient. (B) Changes in mRNA expression levels across patients 10–14 days after initiation of BRAFi are plotted on a log scale as fold change from pre-treatment levels using box and whisker plots (* = P≤0.05).

Figure 3. Protein expression levels of BCL2 family members in patients undergoing treatment with a BRAF inhibitor.

RPPA analysis of tumors from patients with metastatic melanoma shows a significant increase of BID and BIM on BRAFi. (A) Protein expression levels of each gene from pre and on treatment biopsies for each patient are shown as log fold change on treatment. (B) Changes in protein expression levels across patients 10–14 days after initiation of BRAFi are plotted on a log scale as fold change from pre-treatment levels using box and whisker plots (* = P≤0.05).

Figure 4. Expression levels of BCL2 family members in a panel of BRAF mutant cell lines undergoing BRAF inhibition.

(A) mRNA expression levels of various BCL2 family members were quantified by real-time PCR changes and are plotted as log fold difference from vector control (DMSO). (B) Across our panel of cell lines, BCL2-W and BIM increased significantly from control in the context of BRAF inhibition. PLX4720 (1 µM) was used as BRAF inhibitor.

Figure 5. The effect of BRAF inhibition, BH3-mimetics or their combination on cell proliferation, apoptosis and protein expression levels of BCL2 family members in BRAFV600E melanoma cell lines.

(A) MTT assay demonstrating the effect of BCL-2 inhibition, BRAF inhibition, and their combination, on cell proliferation with their respective combination index (CI) value. (B) Corresponding isobolograms. (C) Fluorescence activated cell-sorting (FACS) for Annexin after indicated drug treatment in a BRAFV600 cell line, A375. Drug combinations used at a 1∶1 ratio. (D) Western blotting of BIM and MCL1 in a BRAFV600 cell line, A375 after 2, 6 and 24 hours treatment with a BRAFi, ABT and the combination of both BRAFi and ABT.

Figure 6. The effect of BH-3 mimetic treatment, BRAF inhibition, and their combination on tumor growth in BRAF V600E mutant xenografts.

(A) A375 xenograft. (B) A2058 xenograft. PLX4720 was used as BRAF inhibitor and ABT-263 was used as BH3-mimetic. Both inhibitors were given to mice PO daily at 100 mg/kg for 12 days according to treatment group. Mice were euthanized when tumors reached maximal allowed tumor volume. This occurred between days 9 and 12 for some but not all animals. Error bars represent standard error of the mean (SEM).