Molecular Cell Biology

BMX-Mediated Regulation of Multiple TyrosineKinases Contributes to Castration Resistancein Prostate CancerSen Chen1, Changmeng Cai2, Adam G. Sowalsky3, Huihui Ye4, Fen Ma1, Xin Yuan1,Nicholas I. Simon1, Nathanael S. Gray5,6, and Steven P. Balk1

Abstract

Prostate cancer responds to therapies that suppressandrogen receptor (AR) activity (androgen deprivationtherapy, ADT) but invariably progresses to castration-resis-tant prostate cancer (CRPC). The Tec family nonreceptortyrosine kinase BMX is activated downstream of PI3K andhas been implicated in regulation of multiple pathways andin the development of cancers including prostate cancer.However, its precise mechanisms of action, and particularlyits endogenous substrates, remain to be established. Here,we demonstrate that BMX expression in prostate cancer issuppressed directly by AR via binding to the BMX gene andthat BMX expression is subsequently rapidly increased inresponse to ADT. BMX contributed to CRPC developmentin cell line and xenograft models by positively regulatingthe activities of multiple receptor tyrosine kinases throughphosphorylation of a phosphotyrosine-tyrosine (pYY)motif in their activation loop, generating pYpY that isrequired for full kinase activity. To assess BMX activity in

vivo, we generated a BMX substrate–specific antibody (anti-pYpY) and found that its reactivity correlated with BMXexpression in clinical samples, supporting pYY as an in vivosubstrate. Inhibition of BMX with ibrutinib (developedas an inhibitor of the related Tec kinase BTK) or anotherBMX inhibitor BMX-IN-1 markedly enhanced the responseto castration in a prostate cancer xenograft model. Thesedata indicate that increased BMX in response to ADT con-tributes to enhanced tyrosine kinase signaling and thesubsequent emergence of CRPC, and that combinationtherapies targeting AR and BMX may be effective in a subsetof patients.

Significance: The tyrosine kinase BMX is negatively reg-ulated by androgen and contributes to castration-resistantprostate cancer by enhancing the phosphorylation andactivation of multiple receptor tyrosine kinases followingADT. Cancer Res; 78(18); 5203–15. �2018 AACR.

IntroductionTec kinases are a family of nonreceptor tyrosine kinases

expressed primarily in hematopoietic cells and are related instructure to SRC kinases in having SH3 followed by SH2 andtyrosine kinase domains, but they lack the C-terminal tyrosinethat negatively regulates SRC kinases (1). The Tec kinases arealso unique in having a pleckstrin homology (PH) domain,which is located at the N-terminus and mediates membrane

targeting in response to PI3K activation and subsequent SRC-mediated phosphorylation of a tyrosine in the kinase domainthat activates the enzyme. BMX (bone marrow tyrosine kinasegene on chromosome X), also termed epithelial tyrosine kinase(ETK), in contrast to other Tec kinases, is broadly expressed bycell types outside the hematopoietic lineage, including in arte-rial endothelium and epithelial cells (2–5). Similarly to otherTec kinases, BMX can be activated downstream of PI3K by PHdomain–mediated membrane targeting and SRC-mediatedphosphorylation of a kinase domain tyrosine (Y566; refs. 1, 6),and may alternatively be recruited to the membrane byfocal adhesion kinase (FAK; ref. 7). BMX-deficient mice haveonly modest defects related to angiogenesis and inflammation(3, 8–11). However, increasing evidence indicates that BMXhas roles in modulating multiple cellular processes includingproliferation, differentiation, motility, and apoptosis (12–20),and BMX has been implicated in several cancers (18, 20–24).BMX has been reported to directly or indirectly regulate theactivity of proteins including TNFR2, PAK1, TP53, PIM1, andSTAT3 (10, 13, 15, 25–27), and was recently reported to directlyphosphorylate BAK (19).

Prostate cancer is themost commonnoncutaneousmalignancyinmen. The androgen receptor (AR) plays a central role in prostatecancer, andmost patients withmetastatic prostate cancer respondinitially to androgen deprivation therapy (ADT). However, theyinvariably relapse withmetastatic disease despite castrate levels ofandrogen (castration-resistant prostate cancer, CRPC), and

1Hematology-Oncology Division, Department of Medicine, and Cancer Center,Beth Israel Deaconess Medical Center, Boston, Massachusetts. 2Center forPersonalized Cancer Therapy, University of Massachusetts Boston, Boston,Massachusetts. 3Laboratory of Genitourinary Cancer Pathogenesis, NationalCancer Institute, NIH, Bethesda, Maryland. 4Department of Pathology, BethIsrael Deaconess Medical Center, Boston, Massachusetts. 5Department of Can-cer Biology, Dana-Farber Cancer Institute, Boston, Massachusetts. 6Departmentof Biological Chemistry and Molecular Pharmacology, Harvard Medical School,Boston, Massachusetts.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Authors: Steven P. Balk, Beth Israel Deaconess MedicalCenter, 330 Brookline Avenue, Boston, MA 02215. Phone: 617-735-2065;Fax: 617-735-2050; E-mail: sbalk@bidmc.harvard.edu; and Sen Chen, BethIsrael Deaconess Medical Center, 330 Brookline Avenue, Boston, MA 02215.Phone: 617-735-2071; E-mail: schen2@bidmc.harvard.edu

doi: 10.1158/0008-5472.CAN-17-3615

�2018 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 5203

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

treatment of this advanced stage of the disease is a therapeuticchallenge. Responses can be obtained by further suppression ofandrogen synthesis using agents such as abiraterone or by ARantagonists such as enzalutamide, but most of these patientsstill relapse within 1 to 2 years, likely via multiple mechanisms(28–31). Further responsesmaybeobtainedwith taxanes, radium223, or immunotherapy in subsets of patients, but these also arenot generally durable. BMX is expressed inprimary prostate cancerand metastatic CRPC (2, 4, 22), and transgenic overexpression ofBMX in mouse prostate epithelium results in lesions resemblingprostate intraepithelial neoplasia (PIN; ref. 22). BMX expressionmay be increased in response to ADT, and its ectopic overexpres-sion can confer resistance to castration (23). In addition to itsincreased expression, BMX in prostate cancer cells may be acti-vated downstream of PTEN loss, EGF family receptors, IL6, andseveral neuropeptides (6, 32, 33).

The progression of prostate cancer to higher grades and tometastatic CRPC also is associated with increased MAPK pathwayactivation (34), which in many cancers is mediated by receptortyrosine kinases (RTK), but activating mutations or copy-numberchanges in RTKs are uncommon in primary or advanced prostatecancer (35). We previously identified phosphotyrosine-tyrosine(pYY) as a novel substrate motif for BMX, with BMX generatingpYpY (36). Significantly, this motif is found in the activation loopofmultiple tyrosinekinases, andwe found thatBMXcouldenhancethe activation of these kinases by phosphorylating this secondtyrosine, which is required for full activity. The physiologic impor-tance of this activity is supported by increased expression of insulinreceptor in BMX-deficient mice (36), but whether BMX regulatesRTK activity in cancer is unknown. We hypothesized that increasedBMX in response to ADT may function to augment basal andgrowth factor-stimulated RTK activity and thereby contribute toCRPC. Our objectives in this study were to test this hypothesis andto assess the efficacy of BMX inhibition in combination with ADT.

Materials and MethodsAntibodies and plasmids

Anti-pFAK (pY576/577), anti-pIGF1R (pY1165/1166)/InsR(pY1189/1190), anti-pMET (pY1234/1235), anti-pS6s235/236,anti-pPYK2Y579/580, and anti-pErbB2Y1221/1222 were fromCell Signaling Technology. Anti-Flag M2, anti-FAK, anti-pFAK(pY576), anti-pFAK (pY577), anti-BMX (H-220, sc20711, rabbitpolyclonal), and anti–b-actin were from Santa Cruz Biotechnol-ogy. Anti-pTyr (4G10) and anti-tubulin were from Millipore.3xFlag-FAKwas produced by inserting FAK (pCMV-SPORT6-FAK,Open Biosystems) into p3XFlag-CMV (Sigma) between HindIIIand BamH1 sites, and mutants were then generated using Quik-Change Site–Directed Mutagenesis Kit (Stratagene). The BMXsubstrate antibody was generated in collaboration with CellSignaling Technology. Rabbits were immunized with a BMXmotif peptide, XX(D/E/S/T)pYpYXX (where X is any aminoacid). The antibodies then were depleted of reactivity to theunphosphorylated and single phosphorylated peptides, andabsorbed and eluted from resin conjugated with the duallyphosphorylated peptide.

Cell culture and transfectionHEK293 and VCaP cells were cultured in DMEM/10% FBS

(Hyclone). LNCaP, DU145, and CWR22 RV1 cells were culturedin RPMI-1640/10% FBS. Cells were obtained from the ATCC

and used within 6 months of thawing for studies. They wereauthenticated by short tandem repeat profiling and were free ofMycoplasma. For three-dimensional (3D) cultures, approximately4,000 cells were seeded onto chamber slides (LAB-TEK) coatedwith growth factor–reduced Matrigel (BD Biosciences) and thencultured in basal medium (RPMI1640 or DMEM) replenishedwith 2% FBS and 2% Matrigel. The medium was replaced every4 days. Transfections were performed using Lipofectamine 2000(Invitrogen) according to manufacturer instructions. For cellproliferation assays, cells were grown with or without drugtreatment for 72 hours. Viable cells were counted or analyzedusing Cell Counting Kit-8 (Dojindo).

Immunoblotting and immunoprecipitationCells were lysed with RIPA buffer (Thermo Scientific) contain-

ing Pierce protease inhibitor and phosphatase inhibitor cocktails(Thermo Scientific), with 1 mmol/L of sodium orthovanadateadded when necessary. Lysates were sonicated for 10 seconds andcentrifuged at 13,000 rpm at 4�C for 15 minutes. Protein extractswere mixed with 2X Laemmli sample buffer and boiled for 5minutes and then resolvedon4%–15%Mini-ProteanTGXprecastgels (Bio-Rad) followed by membrane transfer. Membranes wereblocked with 5% milk (or 5% BSA for phosphospecific Abs) inTBS/0.1% Tween 20 (TBS/T) at room temperature for 1 hour andincubated with primary antibodies overnight at 4�C. Membraneswere then incubated with secondary antibodies at room temper-ature for 1hour anddeveloped by ECL. For immunoprecipitation,equal amounts of protein extracts (1–5mg) weremixed with 5 mgof pYpY antibody and 20 mL of protein A agarose beads andincubated at 4�C overnight with continuous agitation. The beadswerewashed extensivelywith RIPAbuffer, followed by TBSbuffer,and beads were eluted with 2� Laemmli sample buffer.

Immunohistochemistry in xenograft, patient samples, andprostate cancer tissue microarrays

Tissues were fixed with 10% formalin and then paraffin-embedded. The sections were analyzed by hematoxylin andeosin staining and by standard IHC staining using the pYpYantibody (clone BL10920, BL10921; Cell Signaling Technolo-gy). IHC for BMX was carried out with a mouse monoclonalIgG1 antibody (BD Biosciences #610792) versus nonspecificIgG1 (DakoCytomation). CRPC patient samples were from aclinical trial of abiraterone combined with dutasteride (28).Written informed consent was obtained from patients forresearch use of excess tissue, and the studies were conductedin accordance with recognized ethical guidelines as outlined inthe U.S. Common Rule. Tissue microarrays (TMA) were from USBiomax. The immunostained TMAs were then evaluated andscored by a study pathologist (X. Yuan).

RT-PCR and chromatin immunoprecipitation assayQuantitative real-time RT-PCR amplification was performed

on RNA extracted from tissue samples or cell lines using anRNeasy Mini Kit (Qiagen). RNA (50 ng) was used for eachreaction, and the result was normalized by coamplification of18S RNA. Reactions were performed on an ABI Prism 7700Sequence Detection System using Taqman one-step RT-PCRreagents. For chromatin immunoprecipitation (ChIP) assay,cells were formalin fixed, lysed, and sonicated. Anti-AR (SantaCruz Biotechnology), anti-p300 (Santa Cruz Biotechnology),anti-FOXA1 (Abcam), anti-OCT1 (Santa Cruz Biotechnology),

Chen et al.

Cancer Res; 78(18) September 15, 2018 Cancer Research5204

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

anti-LSD1 (Abcam), anti-H3K4me1 (Abcam), anti-H3K4me2(Upstate), anti-H3K4me3 (Abcam), or rabbit IgG (Santa CruzBiotechnology) were used to precipitate chromatin fragmentsfrom cell extracts. Quantitative real-time PCR was used toanalyze binding to the ABS1 and ABS2. We used real-timequantitative PCR (SYBR Green) to amplify the DNA fragmentin the antibody-precipitated DNA and the unprecipitated inputDNA to calculate DCT values. The RQ values (RQ ¼ 2�DCT) arepresented and reflect the precipitated DNA as a percentage ofthe input DNA. Results are represented as mean � STD forreplicate samples. Data are representative of at least threeexperiments. Significant differences are indicated (�) in theexperiments.

Mouse xenograft studiesAll animal experimentswere approvedbyBeth IsraelDeaconess

Medical Center's (BIDMC) Institutional Animal Care and UseCommittee and were performed in accordance with institutionaland national guidelines. About 2 to 5 million prostate cancercells (CWR-22RV1, DU145, and VCaP) were suspended in serum-free medium supplemented with 50% Matrigel (BD Biosciences)and were implanted subcutaneously into the dorsal flank of4- to 6-week-oldmale ICR SCIDmice (Taconic). Once the tumorsreach approximately 250 mm3 in size, mice were injected intra-peritoneally with BMX-IN-1, ibrutinib, or vehicle at indicateddoses once per day. Tumor volumes were measured twice weeklyusing calipers. At the end of the study, all mice were humanelyeuthanized, and tumors were collected and analyzed. For VCaPxenograft model, mice were castrated once the tumor reachedapproximately 250 mm3, and the BMX inhibitor therapy wasimmediately started.

Analysis of public domain RNA-seq and microarray dataThe 75th-percentile normalized gene-level RSEM values

for the prostate TCGA tumor and normal cases (mRNA-seqV2) were downloaded directly from the NCI Genomic DataCommons. CRPC data were downloaded from dbGaP forthe Beltran 2016 (37) and Robinson 2015 (38) datasets(phs000909 and phs000915). Unpublished RNA-seq data cor-responding to the LuCaP cases described previously (39) weregraciously provided by Dr. Eva Corey (University of Washing-ton). For the Beltran, Robinson, and LuCaP datasets, CRPCcases were identified by excluding cases previously marked asneuroendocrine or otherwise having high levels of SYP orCHGA mRNA expression. LuCaP data were preprocessed bythe removal of ambiguous reads originating from mouse tissueas described previously (40). All datasets were then processedthrough the TCGA mRNA-seq V2 alignment protocol, in whichraw data were adaptor trimmed, aligned with MapSplice, sortedby read group and reference name with SamTools, translatedand filtered to the transcriptome, quantified with RSEM, andnormalized to the 75th percentile of each sample. Outlieranalyses were performed using GraphPad Prism version 7. Formicroarrays, log2 microarray data of 4 LuCaP PDX cases (41)were downloaded from GEO, accession ID GSE85672. Valuescorresponding to BMX were grouped by condition and com-pared within each group by one-way ANOVA. Samples withineach group were compared against each other using Tukeymultiple comparison test. Statistical analyses were performedusing GraphPad Prism version 7.

Statistical analysisResults are expressed as mean � SE. Statistical significance

was determined by a two-sided Student t test, with P < 0.05considered statistically significant.

ResultsBMX inhibition suppresses the growth of CRPC cells in vitroand in vivo

Previous studies using BMX inhibitors or RNAi have indi-cated that blocking BMX can suppress the growth or induceapoptosis of prostate cancer cell lines and xenografts (42–44).We also previously reported on development of a Tec kinase(BMX and BTK) selective irreversible covalent inhibitor(BMX-IN-1, Bi1) and showed that it could suppress the growthand/or induce apoptosis of several prostate cancer cell lines(45). Consistent with this previous study, BMX-IN-1 sup-pressed the proliferation of CWR22Rv1 cells (AR-dependentbut androgen-independent prostate cancer cell line) with anIC50 of approximately 5 mmol/L, whereas the reversible non-covalent analogue of BMX-IN-1 (BMX-IN-1R, Bi1R) was muchless active (Fig. 1A). Ibrutinib is a covalent inhibitor of BMXand BTK, which can also target EGFR family kinases at approx-imately 10-fold higher concentrations, and is an approvedBTK inhibitor used for the treatment of B-cell malignancies(BTK not being expressed at detectable levels in prostate cancercell lines, see below). It had similar effects on proliferation(Fig. 1B), although the concentrations required were sub-stantially higher than for suppression of BTK in CLL cells(10–100 nmol/L). More dramatically, BMX-IN-1 and ibrutinibblocked CWR22Rv1 cell colony formation in Matrigel-based3D cultures (Fig. 1C), with the compounds reducing bothcolony number and size at 2 weeks (Fig. 1D and E).

To determine whether BMX may similarly contribute toprostate cancer growth in vivo, we generated subcutaneousCWR22Rv1 xenografts in immunodeficient male mice. Whenxenografts reached approximately 250 mm3, mice were ran-domized to daily treatment with BMX-IN-1 (50 or 100 mg/kg),ibrutinib (150 mg/kg), or vehicle by i.p. injection. Both BMX-IN-1 (Fig. 1F) and ibrutinib (Fig. 1G) significantly suppressedthe growth of these CWR22Rv1 xenografts. This response wasnot clearly related to AR, as BMX-IN-1 also suppressed thegrowth of xenografts derived from the AR-negative DU145prostate cancer cell line (Fig. 1H).

To determine whether these drug-induced effects are con-sistent with what is observed upon inhibiting BMX expressionin vivo, we also generated CWR22Rv1 sublines expressingdoxycycline-inducible BMX shRNA. In two independent lines,doxycycline induction of the BMX shRNA resulted in reducedgrowth in both two-dimensional cultures and in Matrigel-based3D models when compared with uninduced counterparts (Sup-plementary Fig. S1A and S1B). Of note, basal BMX levels inthese lines differed, and the growth rates prior to (as well asafter) doxycycline induction of the shRNA in the two lines werecorrelated with BMX expression, consistent with a growth-stimulatory function. Finally, we injected cells from one ofthese lines (line 12) subcutaneously into immunodeficientmale mice and randomized the mice to receive doxycycline orvehicle in the drinking water. Consistent with the in vitro results,doxycycline suppressed growth of these CWR22Rv1 xenografts(Supplementary Fig. S1C).

BMX Regulation of RTKs in Prostate Cancer

www.aacrjournals.org Cancer Res; 78(18) September 15, 2018 5205

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

BMX inhibition in prostate cancer cells blocks thedownstream dual tyrosine phosphorylation of multipletyrosine kinases in vitro

We previously reported that BMX showed a dramatic pref-erence for substrates with a priming phosphotyrosine at posi-tion –1 (pYY), to yield pYpY (36). The activation loop ofmultiple tyrosine kinases contains a tandem YY that must bedually phosphorylated for full activity (46, 47), and we further

found that BMX could thereby enhance the activity of multipletyrosine kinases through phosphorylation of the second tyro-sine in the activation loop after a priming tyrosine autopho-sphorylation at the –1 position (36). Based on these findings,we proposed that BMX, which is activated downstream of PI3Kand SRC, functions to modulate tyrosine kinase activity andsubsequently downstream signaling (see schematic, Supple-mentary Fig. S2). Consistent with this function, BMX depletion

Figure 1.

BMX inhibition suppresses the growth of CRPC cells in vitro and in vivo. A, CWR22RV1 cells were treated with BMX-IN-1 (Bi1) or BMX-IN-1R (BilR) as anegative control for 72 hours and then measured for viability (mean � SE, n ¼ 4). B, CWR22RV1 cells were treated with ibrutinib for 72 hours andthen measured for viability (mean � SE, n ¼ 4). CWR22RV1 cells were cultured in Matrigel-based 3D system with indicated BMX-IN-1 or controltreatment for 2 to 3 weeks (C) and analyzed for colony number (left) and average diameters (right; mean � SE; n ¼ 5 for left plot and n ¼ 10 for right plot;D). E, CWR22RV1 cells were cultured in Matrigel-based 3D system with ibrutinib or control treatment for 2 weeks, and colonies were analyzed fornumbers and average diameters (mean � SE; n ¼ 8 for left plot and n ¼ 20 for right plot). F, CWR22RV1 xenograft–bearing mice were administered 0,50, or 100 mg/kg/d of BMX-IN-1 (Bi1), and growth was monitored (mean � SE; n ¼ 9 for vehicle, n ¼ 5 for 100 mg/kg/d, and n ¼ 6 for 50 mg/kg/d).G, CWR22RV1 xenograft–bearing mice were administered 0 or 150 mg/kg/d of ibrutinib, and tumor growth was monitored (mean � SE; n ¼ 9 forvehicle, n ¼ 5 for 150 mg/kg/d of ibrutinib). H, DU145 xenograft–bearing mice were administered with 0 or 100 mg/kg/d of BMX-IN-1 (Bi1), andgrowth was monitored (mean � SE; n ¼ 5). � , P < 0.05; �� , P < 0.01; ��� , P < 0.001.

Chen et al.

Cancer Res; 78(18) September 15, 2018 Cancer Research5206

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

by siRNA in CWR22Rv1 cells impaired activation of the MAPKand PI3K pathways in response to serum stimulation (Fig. 2A).This impairment was observed with a pool of BMX siRNA andwas more dramatic using a single BMX siRNA that moreeffectively depleted BMX (Supplementary Fig. S3 and Fig. 2A).

We next assessed the effects of the BMX inhibition on prostatecancer in vivo. Similarly to the in vitro results, by immunoblotting,we found that levels of phospho-S6 (serine 235, 236) and phos-pho-ERK (threonine 202, tyrosine 204) were reduced after 3 daysof therapy with BMX-IN-1 in the CWR22Rv1 xenografts (Fig. 2B),supporting a block in both the PI3K and MAPK pathways. There-fore, we next determined the effect of BMX inhibitors on thephosphorylation of individual tyrosine kinases in CWR22Rv1and DU145 prostate cancer cells. BMX inhibitors diminishedpYpY levels on FAK and on insulin-stimulated Insulin Recep-tor/InsulinGrowth Factor (IR/IGFR1), with a resultant decrease inS6 phosphorylation (Fig. 2C–E). BMX-IN-1 and ibrutinib alsosuppressed pYpY levels inMET after HGF stimulation (Fig. 2F), aswell as the increase in SOX9 that occurs downstream of METactivation (Fig. 2G; ref. 48). Taken together, these findings sug-gested that BMXmay contribute to driving prostate cancer growthin vivo through this enhancement of tyrosine kinase activation.

Generation of a BMX substrate antibodyIn addition to a pY at the –1 position, we previously found

that BMX showed a preference for acidic residues at position�2(also consistent with the activation loop of multiple tyrosinekinases; ref. 36). Therefore, in collaboration with Cell Signaling

Technology, we generated an affinity-purified rabbit antibodythat recognizes (D/E/S/T)pYpY (herein referred to as pYpYantibody) as a potential reagent to assess for BMX activity(Fig. 3A). In order to determine the specificity of this antibody(which was extensively absorbed against the unphosphoryla-ted and monophosphorylated peptides), we first cotransfectedFAK, a BMX substrate, with empty vector (EV), BMX wild type(WT), or BMX kinase dead (KD), and then treated with orwithout ibrutinib. BMX WT increased levels of dual-phosphor-ylated pY576,p577 FAK, as assessed by a specific phospho-FAKpY576, p577 antibody, and markedly increased reactivitywith the pYpY antibody (Fig. 3B). The reactivity was notincreased by KD BMX and was markedly suppressed byibrutinib. We further mutated FAK at tyrosines 576 and/or577, and found that both single and double tyrosine mutationscompletely abrogated the pYpY signal (as well as reactivity withthe phospho-FAK pY576, p577 antibody; Fig. 3C). Moreover,the 4G10 antibody (recognizing pY) showed modest basalreactivity that was increased by BMX on the control FAKconstructs, but not on the Y576A or Y577A constructs.

We next used the pYpY antibody to assess the effects of BMXon a series of tyrosine kinases. Serum-starved CWR22Rv1 cellswere treated overnight with BMX-IN-1 or vehicle, and thentreated with serum for 10 minutes. Proteins were then immu-noprecipitated with the pYpY antibody followed by immuno-blotting using individual antibodies that recognize the pYpYin specific tyrosine kinases including MET, FAK, PYK, ERBB2,and BMX itself. For each kinase, the BMX inhibitor decreased

Figure 2.

BMX inhibition in prostate cancercells blocks the downstream dualtyrosine phosphorylation signal ofmultiple tyrosine kinases in vitro.A, CWR22Rv1 cells were treatedwith BMX or control siRNA for48 hours, serum starved overnight,and then stimulated by additionof FBS to 10% for 10 minutes.B, CWR22 RV1 tumor–bearing micewere administered with BMX-IN-1for 3 days; biopsies prior to andafter treatment wereimmunoblotted for indicatedproteins. C and D, CWR22RV1 cellsin complete medium were treatedwith BMX-IN-1 (C) or ibrutinib (D)for 4 hours.E,CWR22RV1 cellswereserum starved for 24 hours, thenpreincubated with BMX-IN-1 inserum-freemedium for 2 hours, andthen stimulated with insulin for10 minutes. F and G, DU145 cellsserum starved for 24 hours werepreincubatedwith BMX inhibitors inserum-free medium for 2 hours,then stimulated with HGF, andimmunoblotted for pMET after10 minutes (F) or SOX9 after3 hours (G).

BMX Regulation of RTKs in Prostate Cancer

www.aacrjournals.org Cancer Res; 78(18) September 15, 2018 5207

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

the pYpY signal in the serum-starved and serum-stimulatedcells (Fig. 3D). We also immunoblotted the pYpY-immuno-precipitated proteins with the pYpY antibody or with 4G10(recognizing a single phosphotyrosine). BMX inhibition sup-pressed dual tyrosine phosphorylation but failed to inhibitsingle tyrosine phosphorylation, further supporting the sub-strate specificity of BMX (Fig. 3E).

Reactivity with BMX substrate antibody correlates with BMXlevel and activity in vivo

We next performed IHC for BMX and with the pYpY antibodyon castration-resistant VCaP xenografts (which express increas-ed levels of BMX, see below) and patient samples. For the BMXIHC, we used a mouse monoclonal that had been validatedpreviously for IHC (24) and further confirmed its specificity byimmunoblotting and by IHC on CWR22Rv1 cell blocks fromcells treated with BMX siRNA (Supplementary Fig. S4A andS2B). The castration-resistant VCaP xenografts were positive forBMX and with the pYpY antibody, and the latter staining couldbe competed with excess pYpY peptide (Fig. 4A). To furtherassess BMX function in vivo, we biopsied castration-resistantVCaP xenografts and then treated for 4 days with BMX-IN-1(100 mg/kg/day). Significantly, the pYpY signal by IHC wassubstantially reduced in the BMX-IN-1–treated xenograft tumorscompared with the pretreated counterparts from the same mice

(Fig. 4B). We next stained several CRPC bone marrow biopsiesand a primary prostate cancer TMA (Fig. 4C and D). Impor-tantly, the pYpY signal correlated with BMX expression in thebiopsies and the TMA (Pearson coefficient 0.7743 in theTMA). Figure 4E shows the distribution of staining in the TMA.Taken together, these data strongly support the conclusion thatBMX functions in vivo to phosphorylate pYY substrates includingmultiple tyrosine kinases, and may thereby contribute to theincreased signaling through downstream pathways.

BMX is upregulated in castration-resistant and abiraterone-resistant prostate cancer

A previous study found that BMX protein was increased in aseries of castration-resistant transurethral resection of prostatesamples versus untreated samples, and in prostate cancer xeno-grafts and mouse prostate after castration (23). We thereforeexamined BMX mRNA levels in primary prostate cancer andmetastatic CRPC RNA-seq datasets. Significantly, median BMXmRNA levels in the TCGA dataset of primary prostate cancerwere low and not increased relative to normal prostate, althoughthere was a substantial outlier population (Fig. 5A; ref. 49).BMX mRNA levels were higher in two metastatic CRPC datasets,although they were still lower than for a series of other kinases(Fig. 5B). Consistent with these results, although overall levelswere low, BMX mRNA was increased in CRPC bone marrow

Figure 3.

Generation of a dual phospho-tyrosine (pYpY) antibody that indicates BMX activity. A, Schematic for (D/E/S/T)pYpY BMX substrate antibody generation.B, HEK 293 cells double transfected with WT FAK and WT or KD BMX (or EV) were treated with or without 1 mmol/L ibrutinib for 4 hours. C, HEK293 cells were double transfected with WT or mutant FAK (or EV), and with WT or KD BMX. D, CWR22RV1 cells serum starved for 24 hours were pretreatedwith BMX-IN-1 (Bi1) for 4 hours and then stimulated with or without FBS (to 10%) for 10 minutes. Proteins were immunoprecipitated with the pYpYantibody and then blotted for dual tyrosine phosphorylation of the indicated tyrosine kinases or for total BMX. Arrowheads, predicted molecule weights.E, CWR22RV1 cells treated with or without 5 mmol/L of BMX-IN-1 were immunoprecipitated using the pYpY BMX substrate antibody. Whole cell lysate(WCL) or immunoprecipitated proteins were then blotted with the pYpY (left) or phospho-tyrosine (4G10; right) antibodies.

Chen et al.

Cancer Res; 78(18) September 15, 2018 Cancer Research5208

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

metastases versus primary untreated prostate cancer samples onAffymetrix microarrays we reported previously (SupplementaryFig. S5; ref. 50).

To further trace the effects of androgen deprivation on BMX,we established androgen-dependent VCaP xenografts and exam-ined BMX expression prior to castration (androgen dependent,AD), 4 days after castration (post-cast), and in the relapsedCRPC stage. Both Affymetrix microarray data (Fig. 5C, top plotheatmap) and quantitative reverse transcriptase PCR (qRT-PCR)validation data (Fig. 5C, bottom plot) showed that BMX mRNAwas upregulated immediately after castration and persisted atthe CRPC stage. Strikingly, corresponding to the increase ofBMX expression with tumor progression to CRPC, staining withthe pYpY antibody was also markedly intensified (Fig. 5D). It isalso noteworthy that the pYpY signal was diffusely expressed inthe cytoplasm prior to castration, started to be enriched in theplasma membrane area immediately after castration, and washighly abundant in plasma membrane at the CRPC stage,consistent with membrane-localized RTKs downstream of BMXbeing activated during the progression to CRPC.

To determine whether BMX expression was altered inresponse to further androgen deprivation with abiraterone, wenext administered abiraterone to a cohort of castrated micebearing castration-resistant VCaP xenografts. This treatmentresulted in tumor regression or stabilization for approximately4 to 6 weeks, followed by progression, at which time theseabiraterone-resistant VCaP xenografts were similarly analyzed

for BMX expression in comparison with tumor biopsies takenprior to initiating abiraterone treatment. Results from qRT-PCRshowed that the abiraterone treatment further increased BMXexpression above the levels in the CRPC xenografts (Fig. 5E).Finally, we examined BMX expression in a series of LuCaPpatient-derived xenografts (PDX) that were given AR-targetedtherapy (41). In a castration-sensitive PDX (LuCaP-77), BMXmRNA was significantly increased in tumors that progressedafter castration or abiraterone treatment (Fig. 5F). In contrast,BMX mRNA was not increased by castration or abiraterone inanother castration-sensitive PDX (LuCaP-136). In the castra-tion-resistant LuCaP-96CR (established in castrated mice), treat-ment with abiraterone increased BMX during an initial responseafter 7 days, whereas there was no effect in another castration-resistant PDX (LuCaP-35CR). Overall, these findings indicatethat increased BMX may contribute to prostate cancer progres-sion after ADT in at least a subset of tumors.

AR negatively regulates BMX gene transcriptionA previous study found that BMX was repressed by androgen

in LNCaP cells and found by ChIP that AR was recruited to sitesin the BMX gene, suggesting AR may directly repress the BMXgene (23). We examined our previous Affymetrix microarraydata generated to identify genes that are stimulated versusrepressed by androgen in vitro in VCaP cells and in a sublineof VCaP derived from a castration-resistant xenograft (VCS2cells; ref. 51), and found that BMX also was repressed by DHT

Figure 4.

Reactivity with pYpY antibody correlates withBMX activity in vivo. A, Castration-resistant VCaPxenografts were immunostained for total BMX orpYpY, the latter with or without (D/E/S/T)pYpYpeptide blocking. B, Castration-resistant VCaPxenograft–bearing mice were administered100 mg/kg/d of BMX-IN-1 for 4 days. Pretreatmentbiopsies and posttreatment tissues wereimmunostained for pYpY. C, Bone metastasisbiopsies from patients with CRPC wereimmunostained for pYpY and BMX. D and E, TMAsof human prostate cancer were immunostainedfor pYpY and BMX (D) and scored for correlationbetween pYpY and BMX (E).

BMX Regulation of RTKs in Prostate Cancer

www.aacrjournals.org Cancer Res; 78(18) September 15, 2018 5209

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

in both cell lines (Supplementary Fig. S6). To confirm thisresult, VCaP cells in steroid-depleted medium were treated withDHT for 24 hours. This treatment markedly decreased BMX

mRNA, and this decrease was blunted by cotreatment with anAR antagonist (bicalutamide; Fig. 6A). Analysis by immuno-blotting similarly showed BMX protein was markedly reduced

Figure 5.

BMX is upregulated in castration-resistant and abiraterone-resistantprostate cancer. A, Tukey method fora box-and-whiskers plot of normal(N ¼ 53) and tumor (N ¼ 493) RNAexpression data from the prostatecancer TCGA, where the boxes depictthe 25th percentile, median, and 75thpercentile, and whiskers depict thelowest value and the 75th percentileþ1.5 times the interquartile range. Dotsdepict outlier values greater than the75th percentile þ 1.5 times theinterquartile range. B, Box-and-whiskers plots for three metastaticCRPC RNA-seq data sets. For BMXexpression, whiskers depict the 10th to90th percentiles, whereas open circlesdepict outlier values outside the 10thand 90th percentiles. For all othergenes, whiskers depict the minimumand maximum values. C, VCaPxenografts were biopsied at threestages: androgen-dependent tumor(AD), 4 days after castration (post-cast), and castration-resistant relapsedtumor (CRPC). mRNAs were extractedfrom the biopsies of tumors andanalyzed on Agilent microarrays.BMX expression is shown by heatmap(top plot) and validated by qRT-PCRanalysis (bottom plot). D,Representative VCaP xenograftbiopsied at different stages wasimmunostained for pYpY.E, Castration-resistant VCaPxenograft–bearing mice were treatedwith abiraterone (i.p. 1 mg/day) forthe first week and then 2 mg/day untilprogression. BMX in preabirateronebiopsies versus at progression wasanalyzed for BMX mRNA expressionby qRT-PCR. F, BMX Affymetrixexpression prior to therapy (control),after 7 days of abiraterone (d7), orat relapse on abiraterone or aftercastration (end of study, EOS) inLuCaP PDX models (41).

Chen et al.

Cancer Res; 78(18) September 15, 2018 Cancer Research5210

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

by DHT and could be further increased by treatment with theAR antagonist enzalutamide (presumably reflecting blockade ofresidual androgen in the steroid-depleted medium; Fig. 6B).Consistent with AR-mediated repression of BMX, the expres-sion of BMX protein under basal conditions (complete mediumcontaining androgen) was markedly lower in the AR-positiveLNCaP and VCaP cell lines versus the AR-negative PC3 andDU145 cell lines, although it was also higher in the AR-positiveCWR22Rv1 cell line (which expresses high levels of an AR splicevariant; Fig. 6C). We further confirmed that this band was BMX

by immunoblotting after treatment with a series of BMX siRNA(Supplementary Fig. S7). BTK was undetectable in any of theprostate cancer lines (Fig. 6C).

We next examined our previous AR ChIP-seq data in VCaPcells (52) and identified two AR-binding sites linked to the BMXgene, designated as ABS1 and ABS2, which are located in intron1 and overlapping exon 5, respectively (Fig. 6D; Supplement-ary Fig. S8). To confirm these binding sites in VCaP cells, wedesigned primers spanning each region and performed ChIPcoupled with quantitative PCR (ChIP-qPCR) to measure AR

Figure 6.

ARnegatively regulatesBMXgene transcription.A,VCaP cells in steroid-depletedmediumwere treatedwithDHT,without orwith 10mmol/L of bicalutamide (Bic) for24hours, andBMXmRNAwasmeasured usingqRT-PCR.B,VCaP cellswere cultured in charcoal-stripped serum (CSS)–containingmedium (androgendepleted)withor without 10 nmol/L DHT for 72 hours (left plot). VCaP cells in charcoal-stripped serum-containing medium for 48 hours were treated with vehicle, DHT, orenzalutamide (ENZ; 10 mmol/L) for 24 hours. C, BMX and BTK protein expression in a series of prostate cancer cell lines (LNCaP-BMX cells are stably transfectedwith BMX). D, AR-binding sites in BMX loci after 4-hour DHT stimulation identified by AR ChIP-seq. E, VCaP cells in charcoal-stripped serum-containingmedium were treated with or without 10 nmol/L DHT for 4 hours, followed by CHIP with the indicated antibodies and qPCR for the ABS1 (left) and ABS2 (right)sites (mean � SD of at least three biological replicates). F, VCaP cells were treated with LSD1 inhibitor S2101 (100 mmol/L) for 4 hours with/without10 nmol/L DHT, and effects on BMX mRNA were analyzed by qRT-PCR (mean � SD; n ¼ 3). G, VCaP cells were treated with or without 10 nmol/L DHT for4 hours, followed by CHIP-qPCR for FOXA1, p300, or OCT1 binding to ABS1 (left) and ABS2 (right) sites (mean � SD of at least three biological repeats).

BMX Regulation of RTKs in Prostate Cancer

www.aacrjournals.org Cancer Res; 78(18) September 15, 2018 5211

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

binding. DHT induced significant AR binding at both sites,which was more dramatic on ABS2 (Fig. 6E). To determinewhether AR binding had local effects on chromatin, wenext assessed for changes in histone marks associated withactive transcription (H3K4 mono-, di-, and trimethylation).Substantial H3K4 monomethylation (H3K4me1) was de-tected at both sites, H3K4 dimethylation (H3K4me2) wasabundant at the ABS2 site but minimal at ABS1 site, andtrimethylation (H3K4me3, generally associated with genepromoters) was only observed at the ABS2 site (Fig. 6E).Significantly, DHT treatment (10 nmol/L for 4 hours) causeda decrease in H3K4 methylation at both sites, consistent withtranscriptional repression.

As demethylation of H3K4me1 and H3K4me2 is mediatedby lysine-specific demethylase 1 (LSD1, KDM1A), we alsoexamined LSD1 binding and found that it was higher at theABS1 site prior to DHT, but markedly increased at the ABS2 sitein response to DHT (Fig. 6E). We have similarly reported anassociation between AR-mediated LSD1 recruitment and H3K4demethylation at several other AR-repressed genes includingthe AR gene (51). Consistent with a role for LSD1, treatmentwith an LSD1 inhibitor (S2102) prevented the DHT-inducedsuppression of BMX (Fig. 6F). Finally, to further characterizethese two sites, we examined binding of FOXA1 (AR pioneerfactor), p300 (major AR coactivator), and OCT1 (transcriptionfactor frequently associated with AR at enhancer sites). Sub-stantial basal levels of FOXA1 were associated with both sites,and these levels were markedly decreased at both sites inresponse to DHT (Fig. 6G). In contrast, basal OCT1 andp300 were higher at ABS2 and were decreased in response toDHT at ABS2, but not ABS1. By immunoblotting, we found thatthe DHT treatment did not change total cellular levels of LSD1(or CoREST), FOXA1, or p300, whereas AR protein was initiallyincreased at 6 hours and then markedly decreased at 24 hours(consistent with previous data showing that DHT rapidly sta-bilizes AR protein, but subsequently causes repression of the ARgene; Supplementary Fig. S9A and S9B, respectively; ref. 51).Overall, these findings support the conclusion that ABS2 con-tains a BMX enhancer that is negatively regulated by the agonist-liganded AR and contributes to the increase of BMX after ADT.

BMX is a driver of castration resistance in vitro and in vivoOur data so far have shown that BMX is upregulated immedi-

ately after castration and that this is associated with increasedlevels pYpY, suggesting a role for BMX at early stages in thedevelopment of castration resistance. In order to determine wheth-er BMX contributes to tumor progression after castration, we firsttreated androgen-starved VCaP cells with BMX inhibitors to seewhether BMX inhibition could further suppress cell recovery. BothBMX-IN-1 and ibrutinib dose dependently decreased cell recovery(Fig. 7A and B), and there was at least an additive effect whenibrutinib was combined with enzalutamide to further suppress ARactivity (Fig. 7C). The results from Matrigel-based 3D culturesfurther supported the combinatorial effect between BMX inhibi-tors and AR antagonist, although VCaP cells did not generate well-formed colonies in this 3D model (Fig. 7D). Most strikingly,simultaneous administration of BMX inhibitors (BMX-IN-1 oribrutinib) along with castration in VCaP xenografts led to markedtumor regression compared with castration alone (Fig. 7E).

We showed previously that progression to castration resis-tance in the VCaP xenograft model is associated with restora-

tion of AR activity (53, 54). A previous study indicated thatBMX may directly or indirectly interact with AR and enhance itsstability or activity, possibly through tyrosine phosphorylation,suggesting that BMX may be contributing to AR activity inCRPC (23). To test this hypothesis, we treated a cohort ofcastration-resistant VCaP xenografts with BMX-IN-1 for 4 daysand assessed for gene expression in pretreatment biopsiesversus posttreatment tumor. Significantly, IHC for cleavedcaspase-3 showed that this 4-day treatment with BMX-IN-1,or with ibrutinib, resulted in a marked increase in apoptosis(Fig. 7F). By qRT-PCR, we found decreased expression ofAR-regulated genes (PSA, TMPRSS2, NKX3.1), but also founddecreases in AR mRNA, indicating that BMX inhibition maydecrease AR activity by transcriptional as well as posttransla-tional mechanisms (Fig. 7G). In any case, taken together thesedata support the conclusion that BMX plays a role at earlystages in the development of CRPC, with increased phosphor-ylation of regulatory sites in multiple tyrosine kinases being amechanism that contributes to its survival function.

DiscussionBMX is activated downstream of PTEN loss and PI3K path-

way stimulation, and has been implicated in several cancersincluding prostate cancer, but mechanisms through which itmay drive tumor growth remain to be established. We previ-ously used a positional scanning peptide library screeningapproach to determine that BMX had a marked preference fora priming phosphotyrosine in the –1 position, which indicatedthat BMX substrates may include multiple tyrosine kinases thatare fully activated by pYpY sites in the kinase domain (36).However, the extent to which BMX functions in vivo throughthis mechanism has not been clear. In this study, we initiallyfound that BMX inhibitors decrease pYpY levels in FAK, IR, andMET in prostate cancer cell lines. We then generated andcharacterized a BMX motif antibody against a (D/E/S/T)pYpYpeptide and showed that treatment of VCaP xenografts with aBMX inhibitor caused a rapid decrease in the level of pYpY.Moreover, immunostaining of clinical samples showed a strongcorrelation between BMX and pYpY levels, further supportingthe conclusion that pYY is a BMX substrate in vivo.

Together, these findings indicate that BMX contributes toactivation of tyrosine kinase signaling in vivo in prostate can-cer, with subsequent increases in MAPK and PI3K pathwayactivation. Specifically, we propose a positive feedback loopwherein BMX enhances the activation of multiple RTKs byphosphorylation of pYY sites in their activation loop (withphosphorylation of the –1 tyrosine being mediated by auto-phosphorylation), and that subsequent activation of RAS, PI3K,and SRC further enhances BMX activity. It should be noted thatBMX may be activated by alternative mechanisms (7, 32, 33)and may have additional direct substrates. In particular, BMXhas been linked to increased STAT3 phosphorylation in bothprostate cancer and glioblastoma (22, 24, 27, 44), which couldbe direct (although STAT3 does not have the YY motif) orindirect through RTKs. Other potential BMX substrates thatmay contribute to our observed responses to BMX inhibitioninclude AR (23) and BAK (19).

Consistent with a previous report (23), we also found thatBMX expression was repressed by androgen and increased byandrogen deprivation. By analysis of ChIP-seq data, we then

Chen et al.

Cancer Res; 78(18) September 15, 2018 Cancer Research5212

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

identified and validated two AR-binding sites in the BMXgene, which were distinct from those examined in the previousstudy. Both sites had features of enhancers that were repressedin response to androgen, with the ABS2 (overlapping exon 5)appearing to be most repressed in response to androgen. Thesefindings indicate that AR is a direct negative regulator ofBMX and that increased expression of BMX may be a very

rapid adaptation in response to castration that contributes totumor cell survival and eventual emergence of CRPC. Indeed,we found that BMX expression was markedly increased at4 days after castration in a series of VCaP xenografts. Moreover,this expression persisted when xenografts became castration-resistant, and was then further increased in xenografts thatdeveloped resistance to abiraterone. Significantly, this increase

Figure 7.

BMX is a driver of castration resistance in vitroand in vivo. A and B, VCaP cells in charcoal-stripped serum (CSS)-containing medium weretreated with BMX-IN-1 (A) or ibrutinib (B) for 5days and counted (mean � SE, n ¼ 4). C, VCaPcells in charcoal-stripped serum-containingmedium were treated with 5 mmol/Lenzalutamide (ENZ), 5 mmol/L ibrutinib (Ibr), orthe combination (Combo) for 5 days (mean �SE; n ¼ 4). D, Representative images of VCaPcells cultured for 2 to 3 weeks in Matrigel-based 3D system with 5 mmol/L enzalutamide,BMX-IN-1, and/or ibrutinib. E, VCaPsubcutaneous xenografts were established inICR-SCID mice. The mice were castrated oncethe tumors reached 500 mm3 and thenadministered vehicle (n ¼ 13), ibrutinib (n ¼ 8),or BMX-IN-1 (n¼ 8; each at 100 mg/kg/d) by i.p.injection. The results are mean � SE. F and G,Castration-resistant VCaP xenografts wereestablished and then treated for 4 dayswith vehicle or 100 mg/kg/d BMX-IN-1 oribrutinib via i.p. injection. F, Representativesections were stained for cleaved caspase-3.G, Biopsies prior to or post 4-day treatmentwere analyzed for PSA, TMPRSS2, NKX3.1, or ARmRNA expression by qRT-PCR.

BMX Regulation of RTKs in Prostate Cancer

www.aacrjournals.org Cancer Res; 78(18) September 15, 2018 5213

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

in BMX was associated with an increase in staining with theBMX substrate antibody.

Based on these findings, we hypothesized that treatmentwith a BMX inhibitor may enhance responses to ADT. Indeed,BMX inhibition markedly decreased the recovery of androgen-deprived VCaP cells in vitro. Moreover, treatment with BMXinhibitors (ibrutinib or BMX-IN-1) dramatically enhanced theresponse to castration in VCaP xenografts. It should again benoted that these inhibitors also target BTK, but this kinase isB-cell specific and not expressed at detectable levels in theprostate cancer cells we have examined. However, we cannotrule out effects on another Tec kinase, or the possibility thatother effects of these inhibitors contribute to their efficacy.Nonetheless, these findings indicate that BMX contributes totyrosine kinase pathway activation in at least a subset ofprimary prostate cancer, and that its contribution may begreater in CRPC. Moreover, the rapid induction of BMX afterADT suggests that BMX may be particularly critical for tumorcell survival immediately after ADT. Therefore, these studiesprovide support for clinical trials of BMX inhibitors (such asibrutinib) in CRPC, possibly with selection for patients basedon BMX expression or activity. In addition, these studies suggestthat combining BMX inhibition with initial ADT or withsecondary ADT (abiraterone or enzalutamide) may be partic-ularly effective.

Disclosure of Potential Conflicts of InterestS. Chen has an ownership interest in a patent. H. Ye is a consultant/

advisory board member for Janssen. N.S. Gray has an ownership interest(including stock, patents, etc.) in DFCI. S.P. Balk has an ownership interest(including stock, patents, etc.) in a patent and is a consultant/advisory board

member for Sanofi and Pfizer. No potential conflicts of interest were dis-closed by the other authors.

Authors' ContributionsConception and design: S. Chen, C. Cai, S.P. BalkDevelopment of methodology: S. Chen, C. CaiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): S. Chen, H. Ye, X. Yuan, N.I. SimonAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S. Chen, C. Cai, A.G. Sowalsky, H. Ye, F. Ma, X. Yuan,S.P. BalkWriting, review, and/or revision of the manuscript: S. Chen, C. Cai,A.G. Sowalsky, H. Ye, F. Ma, N.S. Gray, S.P. BalkAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): S. Chen, H. Ye, N.I. Simon, N.S. GrayStudy supervision: N.S. Gray, S.P. Balk

AcknowledgmentsPortions of this research utilized the computational resources of the NIH

HPC Biowulf cluster.This work was supported by grants to S.P. Balk (NIH P50 CA090381,

NIH P01 CA163227, R01 CA168393, DoD W81XWH-09-1-0435, andW81XWH-16-1-0431), S. Chen (DoDW81XWH-14-1-0126), H. Ye (ProstateCancer Foundation Young Investigator Award), X. Yuan (R01 CA168393),and A. G. Sowalsky (supported by the Intramural Research Program of theNational Institutes of Health, National Cancer Institute).

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received November 20, 2017; revised May 21, 2018; accepted July 11, 2018;published first July 16, 2018.

References1. Qiu Y, Kung HJ. Signaling network of the Btk family kinases. Oncogene

2000;19:5651–61.2. Chott A, Sun Z,Morganstern D, Pan J, Li T, Susani M, et al. Tyrosine kinases

expressed in vivo by human prostate cancer bone marrow metastases andloss of the type 1 insulin-like growth factor receptor. Am J Pathol 1999;155:1271–9.

3. Rajantie I, Ekman N, Iljin K, Arighi E, Gunji Y, Kaukonen J, et al. Bmxtyrosine kinase has a redundant function downstream of angiopoietinand vascular endothelial growth factor receptors in arterial endothelium.Mol Cell Biol 2001;21:4647–55.

4. Robinson D, He F, Pretlow T, Kung HJ. A tyrosine kinase profile ofprostate carcinoma. Proc Natl Acad Sci U S A 1996;93:5958–62.

5. Tamagnone L, Lahtinen I, Mustonen T, Virtaneva K, Francis F, MuscatelliF, et al. BMX, a novel nonreceptor tyrosine kinase gene of the BTK/ITK/TEC/TXK family located in chromosome Xp22.2. Oncogene 1994;9:3683–8.

6. Qiu Y, Robinson D, Pretlow TG, Kung HJ. Etk/Bmx, a tyrosine kinasewith a pleckstrin-homology domain, is an effector of phosphatidylino-sitol 3'-kinase and is involved in interleukin 6-induced neuroendocrinedifferentiation of prostate cancer cells. Proc Natl Acad Sci U S A 1998;95:3644–9.

7. Chen R, Kim O, Li M, Xiong X, Guan JL, Kung HJ, et al. Regulation of thePH-domain-containing tyrosine kinase Etk by focal adhesion kinasethrough the FERM domain. Nat Cell Biol 2001;3:439–44.

8. He Y, Luo Y, Tang S, Rajantie I, Salven P, Heil M, et al. Critical functionof Bmx/Etk in ischemia-mediated arteriogenesis and angiogenesis. J ClinInvest 2006;116:2344–55.

9. Zhang R, Xu Y, EkmanN,WuZ,Wu J, AlitaloK, et al. Etk/Bmx transactivatesvascular endothelial growth factor 2 and recruits phosphatidylinositol3-kinase tomediate the tumor necrosis factor-inducedangiogenic pathway.J Biol Chem 2003;278:51267–76.

10. Pan S, An P, Zhang R, He X, Yin G, Min W. Etk/Bmx as a tumor necrosisfactor receptor type 2-specific kinase: role in endothelial cell migration andangiogenesis. Mol Cell Biol 2002;22:7512–23.

11. Gottar-Guillier M, Dodeller F, Huesken D, Iourgenko V, Mickanin C,Labow M, et al. The tyrosine kinase BMX is an essential mediator ofinflammatory arthritis in a kinase-independent manner. J Immunol2011;186:6014–23.

12. Tu T, Thotala D, Geng L, Hallahan DE, Willey CD. Bone marrow X kinase-mediated signal transduction in irradiated vascular endothelium. CancerRes 2008;68:2861–9.

13. Jiang T, Guo Z, Dai B, Kang M, Ann DK, Kung HJ, et al. Bi-directionalregulation between tyrosine kinase Etk/BMX and tumor suppressor p53 inresponse to DNA damage. J Biol Chem 2004;279:50181–9.

14. KimO,Yang J,Qiu Y. Selective activationof small GTPase RhoAby tyrosinekinase Etk through its pleckstrin homology domain. J Biol Chem2002;277:30066–71.

15. Bagheri-Yarmand R, Mandal M, Taludker AH, Wang RA, Vadlamudi RK,Kung HJ, et al. Etk/Bmx tyrosine kinase activates Pak1 and regulatestumorigenicity of breast cancer cells. J Biol Chem 2001;276:29403–9.

16. Chau CH, Chen KY, Deng HT, Kim KJ, Hosoya K, Terasaki T, et al.Coordinating Etk/Bmx activation and VEGF upregulation to promote cellsurvival and proliferation. Oncogene 2002;21:8817–29.

17. Semaan N, Alsaleh G, Gottenberg JE, Wachsmann D, Sibilia J. Etk/BMX, aBtk family tyrosine kinase, and Mal contribute to the cross-talk betweenMyD88 and FAK pathways. J Immunol 2008;180:3485–91.

18. Potter DS, Galvin M, Brown S, Lallo A, Hodgkinson CL, Blackhall F, et al.Inhibition of PI3K/BMX cell survival pathway sensitizes to BH3 mimeticsin SCLC. Mol Cancer Ther 2016;15:1248–60.

19. Fox JL, Storey A. BMXnegatively regulates BAK function, thereby increasingapoptotic resistance to chemotherapeutic drugs. Cancer Res 2015;75:1345–55.

Chen et al.

Cancer Res; 78(18) September 15, 2018 Cancer Research5214

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

20. PotterDS, Kelly P,DennenyO, Juvin V, Stephens LR,Dive C, et al. BMX actsdownstream of PI3K to promote colorectal cancer cell survival and path-way inhibition sensitizes to the BH3 mimetic ABT-737. Neoplasia 2014;16:147–57.

21. Paavonen K, Ekman N, Wirzenius M, Rajantie I, Poutanen M, Alitalo K.Bmx tyrosine kinase transgene induces skin hyperplasia, inflammatoryangiogenesis, and accelerated wound healing. Mol Biol Cell 2004;15:4226–33.

22. Dai B, KimO, Xie Y, Guo Z, Xu K,Wang B, et al. Tyrosine kinase Etk/BMX isup-regulated in human prostate cancer and its overexpression inducesprostate intraepithelial neoplasia inmouse. Cancer Res 2006;66:8058–64.

23. Dai B, Chen H, Guo S, Yang X, Linn DE, Sun F, et al. Compensatoryupregulation of tyrosine kinase Etk/BMX in response to androgen depri-vation promotes castration-resistant growth of prostate cancer cells. CancerRes 2010;70:5587–96.

24. Guryanova OA, Wu Q, Cheng L, Lathia JD, Huang Z, Yang J, et al.Nonreceptor tyrosine kinase BMXmaintains self-renewal and tumorigenicpotential of glioblastoma stem cells by activating STAT3. Cancer Cell2011;19:498–511.

25. Xie Y, Xu K, Dai B, Guo Z, Jiang T, Chen H, et al. The 44 kDa Pim-1 kinasedirectly interacts with tyrosine kinase Etk/BMX and protects human pros-tate cancer cells from apoptosis induced by chemotherapeutic drugs.Oncogene 2006;25:70–8.

26. Saharinen P, Ekman N, Sarvas K, Parker P, Alitalo K, Silvennoinen O. TheBmx tyrosine kinase induces activationof the Stat signalingpathway,whichis specifically inhibited by protein kinase Cdelta. Blood 1997;90:4341–53.

27. Tsai YT, Su YH, Fang SS, Huang TN, Qiu Y, Jou YS, et al. Etk, a Btk familytyrosine kinase, mediates cellular transformation by linking Src to STAT3activation. Mol Cell Biol 2000;20:2043–54.

28. Chen EJ, Sowalsky AG, Gao S, Cai C, Voznesensky O, Schaefer R, et al.Abiraterone treatment in castration-resistant prostate cancer selects forprogesterone responsive mutant androgen receptors. Clin Cancer Res2015;21:1273–80.

29. Ware KE,Garcia-BlancoMA, ArmstrongAJ, DehmSM. Biologic and clinicalsignificance of androgen receptor variants in castration resistant prostatecancer. Endocr Relat Cancer 2014;21:T87–T103.

30. Yuan X, Cai C, Chen S, Chen S, Yu Z, Balk SP. Androgen receptor functionsin castration-resistant prostate cancer andmechanisms of resistance to newagents targeting the androgen axis. Oncogene 2014;33:2815–25.

31. Watson PA, Arora VK, Sawyers CL. Emerging mechanisms of resistance toandrogen receptor inhibitors in prostate cancer. Nat Rev Cancer 2015;15:701–11.

32. Lee LF, Guan J, Qiu Y, Kung HJ. Neuropeptide-induced androgen inde-pendence in prostate cancer cells: roles of nonreceptor tyrosine kinasesEtk/Bmx, Src, and focal adhesion kinase. Mol Cell Biol 2001;21:8385–97.

33. Jiang X, Borgesi RA, McKnight NC, Kaur R, Carpenter CL, Balk SP. Activa-tion of nonreceptor tyrosine kinase Bmx/Etk mediated by phosphoinosi-tide 3-kinase, epidermal growth factor receptor, and ErbB3 in prostatecancer cells. J Biol Chem 2007;282:32689–98.

34. Weber MJ, Gioeli D. Ras signaling in prostate cancer progression. J CellBiochem 2004;91:13–25.

35. Robinson D, Van Allen EM, Wu YM, Schultz N, Lonigro RJ, Mosquera JM,et al. Integrative clinical genomics of advanced prostate cancer. Cell 2015;161:1215–28.

36. Chen S, Jiang X, Gewinner CA, Asara JM, Simon NI, Cai C, et al. Tyrosinekinase BMX phosphorylates phosphotyrosine-primed motif mediatingthe activation ofmultiple receptor tyrosine kinases. Sci Signal 2013;6:ra40.

37. Beltran H, Prandi D, Mosquera JM, Benelli M, Puca L, Cyrta J, et al.Divergent clonal evolution of castration-resistant neuroendocrine prostatecancer. Nat Med 2016;22:298–305.

38. Robinson D, Van Allen Eliezer M, Wu Y-M, Schultz N, Lonigro Robert J,Mosquera J-M, et al. Integrative clinical genomics of advanced prostatecancer. Cell 2015;161:1215–28.

39. Zhang X, Coleman IM, Brown LG, True LD, Kollath L, Lucas JM, et al.SRRM4 expression and the loss of REST activity may promote the emer-gence of the neuroendocrine phenotype in castration-resistant prostatecancer. Clin Cancer Res 2015;21:4698–708.

40. Ahdesmaki MJ, Gray SR, Johnson JH, Lai Z. Disambiguate: an open-sourceapplication for disambiguating two species in next generation sequencingdata from grafted samples. F1000Res 2016;5:2741.

41. LamHM,McMullin R, NguyenHM, Coleman I, GormleyM, Gulati R, et al.Characterization of an abiraterone ultraresponsive phenotype in castra-tion-resistant prostate cancer patient-derived xenografts. Clin Cancer Res2017;23:2301–12.

42. Dai B, Chen H, Guo S, Yang X, Linn DE, Sun F, et al. Compensatoryupregulation of tyrosine kinase Etk/BMX in response to androgen depri-vation promotes castration-resistant growth of prostate cancer cells. CancerRes 2010;70:5587–96.

43. Guo W, Liu R, Bhardwaj G, Ma AH, Changou C, Yang JC, et al. CTA095, anovel Etk and Src dual inhibitor, induces apoptosis in prostate cancer cellsand overcomes resistance to Src inhibitors. PLoS One 2013;8:e70910.

44. Guo W, Liu R, Bhardwaj G, Yang JC, Changou C, Ma AH, et al. TargetingBtk/Etk of prostate cancer cells by a novel dual inhibitor. Cell Death Dis2014;5:e1409.

45. Liu F, Zhang X, Weisberg E, Chen S, Hur W, Wu H, et al. Discovery of aselective irreversible BMX inhibitor for prostate cancer. ACS Chem Biol2013;8:1423–8.

46. Calalb MB, Polte TR, Hanks SK. Tyrosine phosphorylation of focal adhe-sion kinase at sites in the catalytic domain regulates kinase activity: a rolefor Src family kinases. Mol Cell Biol 1995;15:954–63.

47. Favelyukis S, Till JH, Hubbard SR, Miller WT. Structure and autoregulationof the insulin-like growth factor 1 receptor kinase. Nat Struct Biol 2001;8:1058–63.

48. van Leenders GJ, Sookhlall R, Teubel WJ, de Ridder CM, Reneman S,Sacchetti A, et al. Activation of c-MET induces a stem-like phenotype inhuman prostate cancer. PLoS One 2011;6:e26753.

49. Cancer Genome Atlas Research N. The molecular taxonomy of primaryprostate cancer. Cell 2015;163:1011–25.

50. StanbroughM, Bubley GJ, Ross K, Golub TR, RubinMA, Penning TM, et al.Increased expression of genes converting adrenal androgens to testosteronein androgen-independent prostate cancer. Cancer Res 2006;66:2815–25.

51. Cai C,HeHH,Chen S, Coleman I,WangH, FangZ, et al. Androgen receptorgene expression in prostate cancer is directly suppressed by the androgenreceptor through recruitment of lysine-specific demethylase 1. Cancer Cell2011;20:457–71.

52. Cai C,HeHH,Gao S, Chen S, YuZ,GaoY, et al. Lysine-specific demethylase1 has dual functions as a major regulator of androgen receptor transcrip-tional activity. Cell reports 2014;9:1618–27.

53. Cai C, Wang H, Xu Y, Chen S, Balk SP. Reactivation of androgen receptor-regulated TMPRSS2:ERG gene expression in castration-resistant prostatecancer. Cancer Res 2009;69:6027–32.

54. Gao S, Ye H, Gerrin S, Wang H, Sharma A, Chen S, et al. ErbB2 signalingincreases androgen receptor expression in abiraterone-resistant prostatecancer. Clin Cancer Res 2016;22:3672–82.

www.aacrjournals.org Cancer Res; 78(18) September 15, 2018 5215

BMX Regulation of RTKs in Prostate Cancer

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615

2018;78:5203-5215. Published OnlineFirst July 16, 2018.Cancer Res   Sen Chen, Changmeng Cai, Adam G. Sowalsky, et al.   to Castration Resistance in Prostate CancerBMX-Mediated Regulation of Multiple Tyrosine Kinases Contributes

  Updated version

  10.1158/0008-5472.CAN-17-3615doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://cancerres.aacrjournals.org/content/suppl/2018/09/27/0008-5472.CAN-17-3615.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/78/18/5203.full#ref-list-1

This article cites 54 articles, 30 of which you can access for free at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.org

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/78/18/5203To request permission to re-use all or part of this article, use this link

on November 8, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2018; DOI: 10.1158/0008-5472.CAN-17-3615