In silico and preclinical drug screening identifies dasatinib as a targeted therapy for T-ALL

OPEN LETTER TO THE EDITOR In silico and p eclinical d ug sc eening iden ifies dasa inib as a a ge ed he apy o T-ALL Blood Cance Jou nal (2017) 7, e604; doi:10.1038/bcj.2017.87; published online 8 Sep embe 2017 Sys ema ic de elopmen o combina ion chemo he apies has significan ly imp o ed he p ognosis o acu e lymphoblas ic leukemia (ALL). Pa ien s wi h T-cell ALL (T-ALL) s ill ha e less a o able ou comes, and he p ognosis o elapsed T-ALL is dismal. 1 In o de o find no el a ge ed he apies o T-ALL, we e ie ed he gene exp ession p ofiles o 4430 leukemia samples om he Gene Exp ession Omnibus (GEO) and pe o med an in silico d ug a ge sc eening, whe e he exp ession p ofiles we e compa ed wi h known d ug a ge s in he D ug signa u e da abase (DsigDB). 2,3 This compound lib a y con ains bo h FDA (Food and D ug Adminis a ion) –app o ed and no el in es iga- ional d ugs. S ong exp ession o S c amily y osine kinase LCK was de ec ed in T-ALL samples, exceeding ha o o he leukemias and no mal T cells (Figu e 1a), whe eas a LCK- a ge ing d ug, dasa inib, dec eased he kinase ac i i y o LCK o 1% in compa ison wi h he con ol a a 100 nMconcen a ion. 4 Dasa inib is known o ha e mul iple in acellula a ge s, and we no iced ha some o hem had a simila ly ele a ed exp ession in T-ALL pa ien s (Figu e 1b and Supplemen a y Figu e S1). The e o e, ou combina o ial d ug/ a ge sc eening sugges s dasa inib as a candida e a ge ed he apy o T-ALL pa ien s. To expe imen ally e alua e he po ency o dasa inib, we ea ed T-ALL cell lines wi h inc easing concen a ions o dasa inib (1– 1000 nM). A e 72 h, he mos significan esponse was seen in Figu e 1. The in silico and in i o esul s indica e dasa inib as a po en ial d ug o T-ALL wi h LCK as i s p ime a ge . (a)LCK exp ession in di e en hema ological sample g oups: acu e leukemia (n=4430), p e-B-ALL (n=1304), T-ALL (n=385) and T lymphoids (n=247). (b) The fil e ed lis o a ge s o dasa inib om in silico sc eening. The lis con ains a ge s wi h a lowe exp ession in no mal cells (myeloid, B lymphoid and T lymphoid) in compa ison wi h hei leukemic coun e pa s; a significan exp ession di e ence be ween T-ALL and T-lymphoid samples (adjus ed Po0.001 and a 41.25- old change) and dasa inib is capable o inhibi ing hem wi h high e ficiency. Pe cen age o con ol (POC) o Kd alues we e used o es ima e he e ficacy o dasa inib agains i s a ge s. The POC alue indica es he pe cen age o emaining ac i i y a e inhibi o ea men in compa ison wi h he un ea ed con ol sample. In addi ion o 0.5 μM, POC alues a 100 nMconcen a ions we e also a ailable o LCK and ABL1, and hey we e 1 and 0, espec i ely. (c) E ec o dasa inib o cell iabili y in se e al T-ALL cell lines measu ed by alama Blue assay a e 72 h o incuba ion in 10- old dasa inib dilu ion se ies (1–1000 nM). Values a e ela i e cell iabili es in compa ison wi h dime hyl sul oxide (DMSO) con ol and esul s a e median alues om h ee independen expe imen s pe o med in iplica e, excep o CCRF-CEM and HPB-ALL ha a e om wo independen expe imen s. (d) The exp ession o 21 po en ial dasa inib a ge s in Ju ka cells. Ba s indica e median alues. (e) The e ec o LCK knockdown o Ju ka cell p oli e a ion measu ed in ime se ies (0h, 24h, 48 and 72 h) wi h alama Blue assay. P oli e a ion end lines a e d awn h ough median alues. A ime poin 72 h, he p oli e a ion had dec eased by 14% in compa ison wi h he mock- ea ed con ol (P=0.0289, Mann–Whi ney U- es ). The da a consis o se en indi idual expe imen s pe o med in iplica e, and each ime poin is no malized o he 0 h ime poin . RT-qPCR and wes e n blo esul s show he e ficiency o LCK knockdown. ( ) The e ec o dasa inib on cell iabili y in he LCK-deficien Ju ka cell line in compa ison wi h he no mal Ju ka cell line measu ed by alama Blue assay a e 72 h o incuba ion in a 10- old dasa inib dilu ion se ies (1–1000 nM). The di e ence be ween he wo cell lines was s a is ically significan al eady a a 10 nMconcen a ion (P=0.014, Mann–Whi ney U- es ). The alues a e ela i e cell iabili ies in compa ison wi h he DMSO con ol, and he esul s a e he median alues om h ee independen expe imen s pe o med in iplica e. The e o ba s indica e 95% confidence in e als. Blood Cance Jou nal Ci a ion: Blood Cance Jou nal (2017) 7, e604; doi:10.1038/bcj.2017.87 www.na u e.com/bcj Ju ka cells, wi h a 31% iabili y dec ease a a 10 nMconcen a ion (n=3, P= 0.0039; Figu e 1c and Supplemen a y Figu e S2). As dasa inib inhibi s se e al kinases ha a e key egula o s o cellula p oli e a ion and iabili y, we chose a panel o likely candida es (9 a ge s based on he in silico sc eening and 12 o he well-known a ge s om he li e a u e), and analyzed hei exp ession in T-ALL cell lines by quan i a i e e e se ansc ip ase-PCR (RT-qPCR) and wes e n blo ing. LCK was he mos p ominen ly exp essed gene in T-ALL cell lines, whe eas FYN,ABL1,MAP2K5,MAP4K5 and LYN we e exp essed a lowe le els (Figu e 1d and Supplemen a y Figu e S3). Knockdown o LCK in a dasa inib-sensi i e cell line (Ju ka ) significan ly dec eased cell p oli e a ion (14% dec ease, P= 0.0289, n= 7, Figu e 1e), whe eas knockdown o FYN,ABL1, MAP2K5 and MAP4K5 had no significan e ec (Supplemen a y Figu es S4a–d). Impo an ly, Ju ka cells wi h educed LCK ac i i y due o a dele ion o exon 7 (cell line J.CaM1.6) los dasa inib sensi i i y (Figu e 1 ). Mo eo e , LCK knockdown did no cause s a is ically significan dec ease o p oli e a ion in ela i ely dasa inib-insensi i e P12-Ichikawa cell line (Supplemen a y Figu e S4e). These esul s sugges ha LCK is he p ime a ge o dasa inib in T-ALL. We nex pe o med ex i o d ug es ing o 22 p ima y T-ALL samples. In 6 cases (27%), he esponse o dasa inib was significan based on d ug sensi i i y sco es (DSS, using a cu o alue o 10, Figu e 2a). 5 Hal -maximal g ow h inhibi ion concen- a ions (IC 50 ) anged be ween 1.3 and 16 nM, whe eas he con ol samples had an IC 50 o 41000 nM(Supplemen a y Figu e S5). We also no ed a nega i e co ela ion be ween dasa inib and glucoco icoid DSS sco es (Supplemen a y Figu e S6). P e iously, dasa inib sensi i i y has been epo ed in T-ALL cases wi h NUP214-ABL1 usion. 6–8 In con as , none o he dasa inib esponde s in ou sample se ca ied he usion gene based on ei he genomic PCR o RNA-sequencing analysis (Supplemen a y Figu e S7). LCK was s ongly exp essed in ou ou o he fi e dasa inib- esponsi e pa ien samples, whe eas he exp ession o o he po en ial a ge s a ied om a low (LYN,ABL1) o medium le el (FYN,MAP2K5 and MAP4K5, Supplemen a y Figu e S8a). As LCK was also ela i ely s ongly exp essed in dasa inib-insensi i e pa ien samples, no co ela ion be ween dasa inib esponse and LCK exp ession was obse ed (Supplemen a y Figu e S8b). As T-ALL subg oups can be sepa a ed by exp ession o specific ansc ip ion ac o s, we explo ed whe he dasa inib sensi i i y was associa ed wi h any specific subg oups. 9 In addi ion o Ju ka cells ha belong o he TAL1 subg oup, fi e ou o six dasa inib- sensi i e pa ien samples showed ei he p ominen TAL1 exp es- sion (Figu e 2b) o ca ied SIL-TAL1 usion (da a no shown). The e was also inc eased exp ession o LMO2 and HOXA9/10 genes in some samples (Supplemen a y Figu e S9). 6q dele ions ha e also Figu e 2. Dasa inib-sensi i e subg oup o T-ALL samples. (a) D ug sensi i i y sco es (DSS) o dasa inib in a coho o 22 pa ien samples. These DSS alues a e calcula ed om g ow h inhibi ion measu emen s a e 72 h o ea men in a 10- old dasa inib dilu ion se ies (0.1–1000 nM), and a DSS alue o 10 was used as he h eshold o dasa inib sensi i i y. (b) The exp ession o he T-ALL sub ype defining ansc ip ion ac o TAL1 in T-ALL pa ien samples and cell lines measu ed by RT-qPCR. The h eshold o ec opic TAL1 exp ession (dashed line) is defined by he exp ession o TAL1 in SIL-TAL1 usion-posi i e cell lines, indica ed wi h black columns, and he s iped columns ep esen he dasa inib- sensi i e samples. 17 Pa ien sample 4 was p ocessed in a sepa a e RT-qPCR ba ch. (c) The hea map o LCK,FYN,MAP2K5,MAP4K5,ABL1 and LYN exp ession in a sepa a e GEO-based T-ALL sample da a se (n=385). Samples a e clus e ed based on he exp ession o T-ALL sub ype-defining ansc ip ion ac o s: TAL1 (n=61+103), NKX2-1 (n=18), TLX1 (n=33), TLX3 (n=51), HOXA (n=56) and LYL (n=63). (d)LCK exp ession in T-ALL sub ype clus e s. The di e ence be ween TAL1 clus e s and any o he clus e was s a is ically significan (Po0.001, Mann–Whi ney U- es ). Le e o he Edi o 2 been associa ed wi h he TAL1 subg oup, and h ee ou o six samples ca ied he 6q dele ion in cy ogene ic analyses (Supplemen a y Table S1). 9 When LCK exp ession was co ela ed wi h he subg oup in o ma ion in ou GEO-based gene exp ession da a se , a s a is ically significan co ela ion was ound be ween LCK exp ession and he TAL1 subg oup (Figu es 2c and d and Supplemen a y Figu e S10). Howe e , no di ec ansc ip ional egula ion o LCK by TAL1 was seen when he exp ession o TAL1 was knocked down (Supplemen a y Figu e S11, da a ep oduced om Sanda e al. 10 ). Taken oge he , he combina ion o in silico,in i o and ex i o da a indica e ha dasa inib exe s clinical u ili y beyond he o iginally sugges ed NUP214-ABL1 cases ha only ep esen 4– 10% o T-ALL pa ien s. 6,11 Ou esul s a e in line wi h he da a by F isman as e al. 12 who iden ified a sligh ly highe pe cen age o dasa inib-sensi i e pa ien s (30–40%) in hei T-ALL coho s. In addi ion o confi ming hei main findings, we expand on hem in wo di e en ways. Fi s , we iden ified LCK y osine kinase as he po en ial p ime a ge o dasa inib, whe eas F isman as e al. 12 did no ecognize any ecu en gene usions, mu a ions o ansc ip- ome p ofiles associa ed wi h dasa inib sensi i i y; a he , hey associa ed SRC exp ession and phospho yla ion s a us wi h dasa inib sensi i i y. They suppo ed hei hypo hesis by epo ing a posi i e co ela ion be ween dasa inib and SRC inhibi o KX2- 391 esponses in T-ALL pa ien samples. 12 Howe e , DsigDB iden ifies bo h SRC and LCK as a ge s o KX2-391. In ou GEO- based gene exp ession da a se and T-ALL cell lines, he exp ession o SRC was e y low when compa ed wi h LCK, and he e was no di e ence in he exp ession le els o T-ALL and no mal T-cell samples (Supplemen a y Figu e S1). A low le el o SRC exp ession was also obse ed in he da a se o F isman as e al. 12 Impo an ly, he LCK-mu an de i a i e o he Ju ka cell line ailed o espond o dasa inib, hus s ongly sugges ing he cen al ole played by LCK i sel . We ailed o de ec an associa ion be ween he phospho yla ion s a us o LCK kinase and dasa inib sensi i i y in T-ALL cell lines (da a no shown). Second, we show he e ha pa ien s belonging o he TAL1 subg oup a e he mos likely o espond o dasa inib, al hough no exclusi ely. A high exp ession o LCK associa es wi h he TAL1 subg oup, bu TAL1 does no di ec ly egula e LCK exp ession. As up o 60% o T-ALL pa ien s belong o he TAL1 subg oup and only 30–40% espond o dasa inib, he e is clea ly a need o mo e accu a e bioma ke s o be iden ified. Dasa inib is known o supp ess p oli e a ion o heal hy T cells by LCK inhibi ion. 13–15 We no iced a significan ly highe exp ession o LCK in T-ALL samples compa ed wi h heal hy T cells (a 1.46- old change, Po0.001), and a highe sensi i i y o T-ALL cells owa d dasa inib compa ed wi h heal hy bone ma ow cells. Al hough we iden ified LCK as he main a ge o dasa inib, ou esul s do no exclude he con ibu ion by o he kinases, including o he S c amily kinases. Many o he a ge candida es unc ion in ecep o signaling pa hways ela ed o cell p oli e a ion and su i al. LCK and FYN a e componen s o he T-cell ecep o (TCR) signaling cascade. 16 Du ing T-cell de elopmen , LCK is equi ed o he no mal de elopmen o hymocy es, whe eas in ma u e T cells, FYN is capable o ac i a ing se e al TCR signaling pa hways in he absence o LCK, including he Ras/ex acellula signal- egula ed kinase and phospha idylinosi ol 3-kinase pa hways. 16 In knock- down expe imen s, we obse ed ecip ocal eedback mechan- isms be ween LCK and FYN (da a no shown), in ag eemen wi h hei known pa ially o e lapping unc ions. Dasa inib is a y osine kinase inhibi o ha is cu en ly app o ed o ima inib- esis an Philadelphia ch omosome-posi i e (Ph+) ch onic myeloid leukemia and he second-line ea men o Ph+ ALL. We epo he e he po en ial u ili y o dasa inib in he ea men o a subse o T-ALL. As ou pa ien coho is ela i ely small, u he s udies a e needed o explo e he bioma ke findings and o deepen he mechanis ic basis be o e emba king on animal and human s udies. CONFLICT OF INTEREST The au ho s decla e no conflic o in e es . S Laukkanen 1 , T G ön oos 1 , P Pölönen 2 , H Kuusanmäki 3 , J Meh onen 2 , J Cloos 4 , G Ossenkoppele 5 , B Gje sen 6,7 , B Øys ein 6,7 , C Heckman 3 , M Heinäniemi 2 , M Kon o 8 and O Lohi 1 1 Tampe e Cen e o Child Heal h Resea ch, Uni e si y o Tampe e, and Tampe e Uni e si y Hospi al, Tampe e, Finland; 2 Ins i u e o Biomedicine, School o Medicine, Uni e si y o Eas e n Finland, Kuopio, Finland; 3 Ins i u e o Molecula Medicine Finland FIMM, Uni e si y o Helsinki, Helsinki, Finland; 4 Depa men o Pedia ic Oncology/Hema ology, VU Uni e si y Medical Cen e , Ams e dam, The Ne he lands; 5 Depa men o Hema ology, VU Uni e si y Medical Cen e , Ams e dam, The Ne he lands; 6 Depa men o Clinical Science, Hema ology Sec ion, Uni e si y o Be gen, Be gen, No way; 7 Depa men o In e nal Medicine, Hema ology Sec ion, Haukeland Uni e si y Hospi al, Be gen, No way and 8 Depa men o Hema ology, Hema ology Resea ch Uni Helsinki, Uni e si y o Helsinki, and Helsinki Uni e si y Cen al Hospi al Cance Cen e , Helsinki, Finland E-mail: [email p o ec ed].fi REFERENCES 1 Goldbe g JM, Sil e man LB, Le y DE, Dal on VK, Gelbe RD, Lehmann L e al. Childhood T-cell acu e lymphoblas ic leukemia: he Dana-Fa be Cance Ins i u e acu e lymphoblas ic leukemia conso ium expe ience. J Clin Oncol 2003; 21: 3616–3622. 2 Liuksiala T, Tei inen KJ, G anbe g K, Heinäniemi M, Annala M, Mäki M e al. O e exp ession o SNORD114-3 ma ks acu e p omyelocy ic leukemia. Leukemia 2014; 28:233–236. 3 Heinäniemi M, Vuo enmaa T, Teppo S, Kaikkonen MU, Bou y-Lii and M, Meh- onen J e al. T ansc ip ion-coupled gene ic ins abili y ma ks acu e lymphoblas ic leukemia s uc u al a ia ion ho spo s. eLi e 2016; 5: e13087. 4 Yoo M, Shin J, Kim J, Ryall KA, Lee K, Lee S e al. DSigDB: d ug signa u es da abase o gene se analysis. Bioin o ma ics 2015; 31: 3069–3071. 5 Yada B, Pemo ska T, Szwajda A, Kulesskiy E, Kon o M, Ka jalainen R e al. Quan i a i e sco ing o di e en ial d ug sensi i i y o indi idually op imized an icance he apies. Sci Rep 2014; 4: 5193. 6 Quin ás-Ca dama A, Tong W, Manshou i T, Vega F, Lennon PA, Cools J e al. Ac i i y o y osine kinase inhibi o s agains human NUP214-ABL1-posi i e T cell malignancies. Leukemia 2008; 22: 1117–1124. 7 Deenik W, Be e loo HB, an de Poel- an de Luy gaa de SC, Wa el MM, an Esse JW, Valk PJ e al. Rapid comple e cy ogene ic emission a e up on dasa inib mono he apy in a pa ien wi h a NUP214-ABL1-posi i e T-cell acu e lympho- blas ic leukemia. Leukemia 2009; 23: 627–629. 8 C ombe O, Las apes K, Zieske A, Mo ales-A ias J. Comple e mo phologic and molecula emission a e in oduc ion o dasa inib in he ea men o a pedia ic pa ien wi h -cell acu e lymphoblas ic leukemia and ABL1 amplifica ion. Pedia Blood Cance 2012; 59:333–334. 9 Gi a di T, Vicen e C, Cools J, De Kee smaecke K. The gene ics and molecula biology o T-ALL. Blood 2017; 129: 1113–1123. 10 Sanda T, Law on LN, Ba asa MI, Fan ZP, Kohlhamme H, Gu ie ez A e al. Co e ansc ip ional egula o y ci cui con olled by he TAL1 complex in human T cell acu e lymphoblas ic leukemia. Cance Cell 2012; 22:209–221. 11 Bu meis e T, Gökbuge N, Reinha d R, Riede H, Hoelze D, Schwa z S. NUP214- ABL1 in adul T-ALL: he GMALL s udy g oup expe ience. Blood 2006; 108: 3556–3559. 12 F isman as V, Dobay MP, Rinaldi A, Tchinda J, Dunn SH, Kunz J e al. Ex i o d ug esponse p ofiling de ec s ecu en sensi i i y pa e ns in d ug- esis an acu e lymphoblas ic leukemia. Blood 2017; 129: e26–e37. 13 Blake S, Hughes TP, May ho e G, Lyons AB. The S c/ABL kinase inhibi o dasa inib (BMS-354825) inhibi s unc ion o no mal human T-lymphocy es in i o. Clin Immunol 2008; 127: 330–339. Le e o he Edi o 3 Blood Cance Jou nal 14 Lee KC, Ouwehand I, Giannini AL, Thomas NS, Dibb NJ, Bijlmake s MJ. Lck is a key a ge o ima inib and dasa inib in T-cell ac i a ion. Leukemia 2010; 24:896–900. 15 Schade AE, Schie en GL, Townsend R, Jankowska AM, Susulic V, Zhang R e al. Dasa inib, a small-molecule p o ein y osine kinase inhibi o , inhibi s T-cell ac i- a ion and p oli e a ion. Blood 2008; 111: 1366–1377. 16 Salmond RJ, Filby A, Qu eshi I, Case a S, Zamoyska R. T-cell ecep o p oximal signaling ia he S c- amily kinases, Lck and Fyn, influences T-cell ac i a ion, di e en ia ion, and ole ance. Immunol Re 2009; 228:9–22. 17 Nagel S, Ven u ini L, Meye C, Kau mann M, Sche M, D exle HG e al. Mul iple mechanisms induce ec opic exp ession o LYL1 in subse s o T-ALL cell lines. Leuk Res 2010; 34: 521–528. This wo k is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons license, unless indica ed o he wise in he c edi line; i he ma e ial is no included unde he C ea i e Commons license, use s will need o ob ain pe mission om he license holde o ep oduce he ma e ial. To iew a copy o his license, isi h p://c ea i ecommons.o g/licenses/ by/4.0/ © The Au ho (s) 2017 Supplemen a y In o ma ion accompanies his pape on Blood Cance Jou nal websi e (h p://www.na u e.com/bcj). Le e o he Edi o 4 Blood Cance Jou nal