Published OnlineFirst August 15, 2019; DOI: 10.1158/0008-5472.CAN-19-0244 Cancer Translational Science Research Src-Dependent DBL Family Members Drive Resistance to Vemurafenib in Human Melanoma Charlotte R. Feddersen1, Jacob L. Schillo1, Afshin Varzavand2, Hayley R. Vaughn1, Lexy S. Wadsworth1, Andrew P. Voigt1, Eliot Y. Zhu1, Brooke M. Jennings2, Sarah A. Mullen2, Jeremy Bobera2, Jesse D. Riordan1, Christopher S. Stipp2,3, and Adam J. Dupuy1,3 Abstract The use of selective BRAF inhibitors (BRAFi) has pro- signaling axis capable of driving resistance to both current duced remarkable outcomes for patients with advanced and next-generation BRAFis. However, we show that the cutaneous melanoma harboring a BRAFV600E mutation. SRC inhibitor, saracatinib, canblocktheDBL-drivenresis- Unfortunately, the majority of patients eventually develop tance. Our work highlights the utility of our straightforward drug-resistant disease. We employed a genetic screening genetic screening method in identifying new drug combina- approach to identify gain-of-function mechanisms of BRAFi tions to combat acquired BRAFi resistance. resistance in two independent melanoma cell lines. Our screens identified both known and unappreciated drivers of Significance: A simple, rapid, and flexible genetic screen- BRAFi resistance, including multiple members of the DBL ing approach identifies genes that drive resistance to MAPK family. Mechanistic studies identified a DBL/RAC1/PAK inhibitors when overexpressed in human melanoma cells. Introduction characterized (7). Thus, unexplained cases of resistance to MAPK inhibition (MAPKi) in human melanoma represent an important Melanoma is the deadliest form of skin cancer, with around unmet clinical need. 90,000 diagnoses of invasive disease and approximately 10,000 Mechanisms of vemurafenib resistance have been studied in deaths per year (1). Patients had few treatment options until the BRAFV600E-mutant human melanoma cell lines using genome- development of vemurafenib, a highly selective kinase inhibitor wide shRNA and CRISPR loss-of-function screens (9–11). Over- that specifically targets the BRAFV600E-mutant protein present in all, these studies showed little overlap in candidate mechan- approximately 50% of all melanoma cases (2). Initially, vemur- isms. Two screens have been reported that attempted to identify afenib provided complete or partial response in over 50% of drivers of vemurafenib resistance by high-throughput overex- patients and increased progression-free survival (3). Unfortunate- pression of genes via lentiviral libraries (12, 13). Importantly, ly, most patients relapse once tumors acquire resistance to these screens failed to identify known mechanisms of vemur- vemurafenib. afenib resistance (e.g., BRAFV600E amplification or N-terminal Genetic analysis of progression samples has identified resis- truncation; refs. 4, 6). These observations led us to develop a tance mechanisms, including amplification of BRAFV600E, expres- simple insertional mutagenesis screening approach using the sion of truncated BRAFV600E, and RAS mutation (4–6). However, Sleeping Beauty (SB) transposon system to identify novel drivers these mechanisms explain only approximately 60% of cases of of vemurafenib resistance in an unbiased forward genetic screen. BRAF inhibitor (BRAFi) resistance (5, 7, 8). Drug resistance can be The SB system is a well-established tool for developing delayed by combining vemurafenib with cobimetinib, an MEK mouse models of spontaneous cancer in which transposon- inhibitor (MEKi), but most patients eventually develop progres- induced somatic mutations drive transformation (14). In this sive disease via resistance mechanisms that have not been well context, the SB system consists of two parts: a mutagenic transposon vector and the transposase enzyme. When intro- 1Department of Anatomy and Cell Biology, Carver College of Medicine, Univer- duced into the same cell, the transposase excises the transposon 2 sity of Iowa, Iowa City, Iowa. Department of Biology, College of Liberal Arts and from a donor vector and integrates it at a TA dinucleotide site in Sciences, University of Iowa, Iowa City, Iowa. 3Holden Comprehensive Cancer the host cell genome. In the context of some selective pressure Center, University of Iowa, Iowa City, Iowa. (e.g., proliferation, drug treatment), cells with specificmuta- Note: Supplementary data for this article are available at Cancer Research tions conferring the ability to outcompete neighboring cells Online (http://cancerres.aacrjournals.org/). undergo clonal expansion, facilitating subsequent identifica- Corresponding Authors: Adam J. Dupuy, University of Iowa, 375 Newton Road, tion of these phenotype-driving mutations. This method has 3202 MERF, Iowa City, IA 52242. Phone: 319-335-8090; Fax: 319-335-7198; been used to select for specific phenotypes using ex vivo cell- E-mail: [email protected]; and Christopher S. Stipp, – [email protected] based assays (15 17), as well as to drive the development of drug-resistant tumors in engineered mouse models (18, 19). Cancer Res 2019;79:5074–87 However, previous ex vivo approaches using human cells have doi: 10.1158/0008-5472.CAN-19-0244 been limited by the relative inefficiency of delivering both Ó2019 American Association for Cancer Research. transposon and transposase vectors to cells. Moreover, prior 5074 Cancer Res; 79(19) October 1, 2019 Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2019 American Association for Cancer Research. Published OnlineFirst August 15, 2019; DOI: 10.1158/0008-5472.CAN-19-0244 DBL GEFs Drive Resistance to MAPK Inhibitors in Melanoma studies have required the isolation of clonal cell populations to RAF1, RAC1) or from Transomics Technology human cDNA identify insertional mutations associated with the desired phe- clones (MCF2, VAV1). Cloned cDNAs were inserted into pig- notype, a process that greatly reduces screen throughput. Col- gyBac expression vectors containing a human Ef1a promoter lectively, these challenges have limited the broader application along with an IRES-puromycin-polyA cassette. Stable cell of SB mutagenesis in ex vivo screening approaches. lines were obtained by cotransfection of each vector with a We present here the results of three SB mutagenesis drug piggyBac transposase expression vector via Effectene (Qiagen) resistance screens conducted in human BRAFV600E-mutant mel- transfection reagent. Overexpression was assessed via RT-qPCR anoma cells to identify novel drivers of resistance to either and immunoblot. See list of primers and antibodies for BRAFi treatment alone or BRAFi/MEKi combination treatment. specifics. Importantly, we detected recurrent N-terminal truncations of BRAF as a driver of BRAFi resistance, a mechanism previously RNA interference associated with BRAFi resistance in human melanoma (6). We RHOA and RHOC constructs had a pLKO backbone (Sigma- also identified MCF2, VAV1, PDGFRB, and N-terminally trun- Aldrich). RAC1 constructs had a pZIP-mCMV vector backbone cated RAF1 as drivers of BRAFi resistance. We experimentally (Transomics Technologies). A nontargeting shRNA in the appro- verified the ability of candidates to drive drug resistance in priate vector backbone was included to produce vector control independent melanoma cell lines, and analysis of transcrip- cell lines. Cells were maintained as stably transduced, polyclonal tome data from clinical progression samples revealed that populations. See Supplementary Information for RNAi-targeting overexpression of our candidates is associated with BRAFi sequences. resistance in human patients. Finally, we elucidate a mecha- nism through which the DBL family members MCF2 and VAV1 Cell culture act to drive drug resistance and show that this mechanism can The following cell lines were obtained from the ATCC: A375 be blocked with saracatinib, an inhibitor of the Src family. (CRL-1619), SKMEL28 (HTB-72), A101D (CRL-7898), and These findings demonstrate the utility of our genetic screening 451Lu (CRL-2813). Clones of vemurafenib-resistant A375 were fi approach to identify clinically relevant drivers of drug resis- veri ed as A375 using short tandem repeat analysis performed tance, as well as the potential for discovering new therapeutic by IDEXX Bioanalytics. Mycoplasma testing is conducted period- approaches to reverse or prevent its occurrence. ically on all cell lines using a PCR-based method to detect the mycoplasma genome. All cell lines were grown in DMEM (Gibco) supplemented with Materials and Methods penicillin/streptomycin (Gibco) and 10% FBS (Gibco). Sponta- SB mutagenesis screen neous vemurafenib-resistant clones and populations were created – – m A375 cells were stably transfected via Effectene (Qiagen) cou- after 2 3 and 4 6 weeks cultured in 3 m vemurafenib, respec- pled with a piggyBac transposase integration system (20) with tively. Clones were isolated via cloning rings. All spontaneously Ef1a-SB100 transgene. After puro selection, SB100-expressing resistant clones and populations were maintained in media with m cells were transfected with the pT2-Onc3 transposon plas- 3 m vemurafenib. mid (21). Forty-eight hours later, 1 Â 106 cells of SB100 þ T2/ CellTiter blue viability assay Onc3 were plated on 10 cm plates. Cells were subsequently Cells were plated in triplicate in 96-well plates at a density of treated with vemurafenib (5 mm), vemurafenib (5 mm) and 5 Â 102 to 5 Â 103 cells per well depending on the cell line. The cobimetinib (5 nm), or vehicle (DMSO, 0.2%) 24 hours after CellTiter-Blue