Mutant-selective degradation by BRAF-targeting PROTACs

Abstract

Over 300 BRAF missense mutations have been identified in patients, yet currently approved drugs target V600 mutants alone. Moreover, acquired resistance inevitably emerges, primarily due to RAF lesions that prevent inhibition of BRAF V600 with current treatments. Therefore, there is a need for new therapies that target other mechanisms of activated BRAF. In this study, we use the Proteolysis Targeting Chimera (PROTAC) technology, which promotes ubiquitination and degradation of neo-substrates, to address the limitations of BRAF inhibitor-based therapies. Using vemurafenib-based PROTACs, we achieve low nanomolar degradation of all classes of BRAF mutants, but spare degradation of WT RAF family members. Our lead PROTAC outperforms vemurafenib in inhibiting cancer cell growth and shows in vivo efficacy in a Class 2 BRAF xenograft model. Mechanistic studies reveal that BRAF^WT is spared due to weak ternary complex formation in cells owing to its quiescent inactivated conformation, and activation of BRAF^WT sensitizes it to degradation. This study highlights the degree of selectivity achievable with degradation-based approaches by targeting mutant BRAF-driven cancers while sparing BRAF^WT, providing an anti-tumor drug modality that expands the therapeutic window.

Fig. 1. Vemurafenib-based PROTAC SJF-0628 induces degradation of mutant BRAF.

a Chemical structure of vemurafenib and BRAF targeting PROTAC, SJF-0628, and its epimer, SJF-0661. SJF-0628 is composed of vemurafenib, a short piperazine-based linker, and a VHL recruiting ligand. SJF-0661 has an identical warhead and linker as SJF-0628 but contains an inverted hydroxyl group in the VHL ligand and is therefore unable to engage VHL to induce ubiquitination. b Inducible NIH3T3 cells expressing indicated V5-BRAF constructs (doxycycline 100–200 ng/mL, 24 h) treated with increasing amounts of SJF-0628. c SK-MEL-28 cells (homozygous BRAF^V600E) treated with indicated amounts of SJF-0628 induced BRAF degradation and suppression of MEK and ERK phosphorylation. d Quantitation of ERK inhibition in SK-MEL-28 cells treated with SJF-0628 or SJF-0661 (mean ± SD, n = 3 biologically independent samples) P value calculated by multiple unpaired t-tests. e Quantitation of SJF-0628 treatment time course (100 nM) at indicated times in SK-MEL-28 cells shows maximal degradation within 4 h (n = 2 biologically independent samples). f SJF-0628 induces selective degradation of p61-BRAF^V600E mutant and inhibits MEK and ERK phosphorylation but spares BRAF^WT and CRAF in SK-MEL-239-C4 cells. g H1666 (heterozygous BRAF^G466V) treated with SJF-0628 shows BRAF degradation, but incomplete suppression of ERK signaling. h BRAF^WT is spared by SJF-0628 in OVCAR-8 cells but induces slight activation of ERK phosphorylation. i Covalent inhibition of KRAS^G12C by MRTX849 in H23 cells hinders PROTAC induced BRAF^WT degradation (n = 3 biologically independent samples). j Quantification of 1i (mean ± SD, n = 3 biologically independent samples). P value calculated by one-way ANOVA. Source data are provided as a Source Data file.

Fig. 2. BRAF^WT is unable to form a PROTAC-induced ternary complex in cells and thus not degraded.

a IC₅₀ values of radiolabeled kinase assay for WT RAF and Class 1 and 2 BRAF mutants (mean, n = 2 biologically independent experiments). Plotted values shown in Table 1. b Purified protein ternary complex assay. GST-VBC (VHL, Elongin B, Elongin C) is immobilized on glutathione beads and incubated with DMSO, SJF-0661 (500 nM) or increasing concentrations of SJF-0628 and purified full length-BRAF to observe VBC:PROTAC:BRAF ternary complex. c Quantification of 2b with respect to 1% input (mean ± SD, n = 3 biologically independent samples). Replicates shown in source data; WT = black circles, V600E = blue circles. d Cell lysate based ternary complex assay (as described in b) but using NIH3T3 cell lysates (doxycycline 800 ng/mL) containing V5-BRAF^WT or V5-BRAF^V600E as input. e Quantification of 2d with respect to 1% input (mean ± SD, n = 3 biologically independent samples). Replicates shown in source data. WT = black circles, V600E = blue circles. f NIH3T3 cells expressing indicated V5-BRAF treated with DMSO or 1 µM SJF-0628 for 1-hour followed by immunoprecipitation of V5-BRAF. g NanoBRET ternary complex assay. HEK293T cells ectopically expressing NanoLuc-BRAF(donor) and HaloTag-VHL covalently labeled with a HaloTag 618 ligand (acceptor) were treated with DMSO, epimer SJF-0661(1 μM) or indicated concentration of SJF-0628 for 3 h. Data represented as BRET ratio (mean ± SD, n = 4 biologically independent experiments); WT = black circles, V600E = blue circles. h Tandem Ubiquitin Binding Entities 1 (TUBE1) pull down of tetra-ubiquitinated proteins in NIH3T3 cells expressing indicated V5-BRAF after 1-hour treatment with vehicle or SJF-0628. Immunoblotted for V5-BRAF. Source data are provided as a Source Data file.

Fig. 3. BRAF-activating MEK inhibitors also sensitize BRAF to PROTAC-induced ubiquitination and degradation.

a NIH3T3 cells with trametinib (1 µM, 5 h) or cobimetinib (500 nM, 3 h) pre-treatment subsequently treated with increasing amounts of SJF-0628 (20 h) promote degradation of BRAF^WT and show a marked increase in p-MEK (n = 2 biologically independent samples). b OVCAR8 cells pre-treated with cobimetinib and trametinib (1 µM, 2 h) promote MEK and CRAF phosphorylation as well as BRAF degradation in the presence of SJF-0628 (n = 2 biologically independent samples). c Cell lysate-based ternary complex assay shown in 2c but using NIH3T3 lysates expressing BRAF^WT and pre-treated with DMSO or 1 µM cobimetinib for 3 h. Cobimetinib pre-treatment promotes ternary complex formation. d V5-BRAF immunoprecipitation in NIH3T3 cells pre-treated with 1 µM of cobimetinib (2 h) followed by treatment of SJF-0628 for 2.5 h. e TUBE1 pulldown in 293 T-Rex cells stably expressing HA- BRAF^WT treated with cobimetinib (cobi) (2 h, 1 µM) and subsequently treated with SJF-0628 (2 h). f NIH3T3 cells pre-treated with 1 µM PD0325901 (MEK inhibitor) or SCH772984 (ERK inhibitor) for 3 h followed by treatment with indicated amount of SJF-0628 for 20 h. g A431 cells pre-treated with GDC-0623 and cobimetinib (500 nM for 3 h) then treated with SJF-0628 for 20 h. Source data are provided as a Source Data file.

Fig. 4. SJF-0628 outperforms vemurafenib in inhibiting growth of cell lines expressing mutant BRAF.

a Cell proliferation assay in SK-MEL-28 cells treated with increasing amounts of vemurafenib, SJF-0628, or SJF-0661 for 3 days (mean ± SD, n = 3 biologically independent samples). EC₅₀ = 215 ± 1.09 nM, 37 ± 1.2 nM, and 243 ± 1.09 nM, respectively; vemurafenib = blue, SJF-0628 = burgundy, SJF-0661 = purple. b Cell proliferation assay in vemurafenib resistant SK-MEL-239-C4 cells treated with increasing amounts vemurafenib, SJF-0628, or SJF-0661 for 5 days (mean ± SD, n = 3 biologically independent samples); vemurafenib = blue, SJF-0628 = burgundy, SJF-0661 = purple. c Cell proliferation assay in SK-MEL-246 (Class 2) cells treated with increasing amounts vemurafenib, SJF-0628, or SJF-0661 for 5 days (mean ± SD, n = 3 biologically independent samples); vemurafenib = blue, SJF-0628 = burgundy, SJF-0661 = purple. d SJF-0628 EC50 = 218 nM ± 1.06 c,H1666 cells treated with SJF-0628, vemurafenib, or SJF-0661 for 5 days (mean ± SD, n = 3 biologically independent samples); vemurafenib = blue, SJF-0628 = burgundy, SJF-0661 = purple. e Treatment of CAL-12-T cells with vemurafenib, SJF-0628, or SJF-0661 for 5 days shows minimal effect on cell viability (mean ± SD, n = 3 biologically independent samples). f Results of an efficacy study in SK-MEL-246 tumor xenografts implanted in female athymic mice showing tumor regression with 50 mg/kg IP twice daily (mean ± SD, n = 3 biologically independent animals). g Scatter plot result of final tumor volumes of SKMEL-246 xenografts treated with SK-MEL-246 (mean ± SD, n = 3 biologically independent animals). P value calculated by unpaired t-test. Source data are provided as a Source Data file.