2849 Analysis of the cancer gene targeting of clinical kinase inhibitor drugs by combining cellular and biochemical profiling

Joost C.M. Uitdehaag1, Jeroen A.D.M. de Roos1, Antoon M. van Doornmalen1, Martine B.W. Prinsen1, Jos de Man1, Yoshinori Tanizawa2, Yusuke Kawase2, Kohichiro Yoshino2, Rogier C. Buijsman1, Guido J.R. Zaman1 1Netherlands Translational Research Center B.V., Molenstraat 110, 5342 CC Oss, The Netherlands; 2Carna Biosciences Inc., 3rd Floor, BMA, 1-5-5 Minatojima-Minamimachi, Kobe, Japan. T: +31412700500 W: www.ntrc.nl and www.carnabio.com E: [email protected]

Introduction 100%100% inh. inh. Bosutinib Cisplatin 1 nM PPonatinibonatinib Sunitinib Tofacitinib • We have systematically compared the Crizotinib Fasudil Axitinib biochemical and cellular potency and selectivity Dasatinib Nilotinib Nilotinib of all kinase inhibitor drugs on the market Pazopanib Pazopanib Etoposide Fluorouracil RegorafSorafenibenib Axitinib • All kinase inhibitor drugs were profi led in a panel SorafVandetanibenib Bosutinib VCabozantinibandetanib Sunitinib 10,000 nM 0% inh. Crizotinib of more than 300 kinase activity assays at Carna CabozantinibDabrafenib DabrafRuxolitinibenib 0% inh. 1 Regorafenib Biosciences Ruxolitinib Erlotinib Erlotinib Lapatinib Gefitinib Gefitinib Dasatinib Afatinib Lapatinib • In addition, all kinase inhibitor drugs were Tofacitinib Bortezomib AfFatinibasudil Vincristine Docetaxel profi led in a panel of highly reproducible ToTrfacitinibametinib FasudilEverolimus Doxorubicin_v3 proliferation assays on 44 well-characterized cell TrTametinibemsirolimus Doxorubicin_v1 Doxorubicin_v2 EvVeerolimumurafenibs SHP−77 BT−549 BT−20 J82 O 786−0 A−498 DLD−1 HCT−15 AN3 CA NCI−H82 SJCRH30 A−204 K−562 RPMI−7951 SW48 SW620 LS 174T HCT−116 NCI−H460 A−549 MG−63 A−427 SK−N−AS A375 A BxPC−3 LoV 769−P A−172 Hs 578T Me Wo SW480 U−2 OS U−87 MG C−33 A CAL 27 F A P SR CCRF−CEM Ju MOL A−1 aDu U−565

2 CHN VCAR−3 r kat E6.1

lines (Oncolines™) at NTRC Imatinib o

Temsirolimus T−4 AL K MET JAK2 ROCK 1 HER2 EGFR KI T BRAF PDGFR α VEGFR 2 ABL Vemurafenib Imatinib • The results allow comparison of existing AL K MET JAK2 ROCK 1 HER2 EGFR KI T BRAF PDGFR α VEGFR 2 ABL

therapies, and defi ne hypotheses for specifi c Figure 2 Clustering of the response in kinase activity assays. %-inhibition at 10 µM inhibitor 10 Figure 3 Clustering of cell proliferation responses ( logIC50) In red non kinase drugs. target populations for new therapies concentration.

ABL inhibitors JAK inhibitors 100000 0.001 0.001 RAF Trametinib Afatinib 100 100 RAF inhibitors inhibitors Imatinib 10000 EGFR 80 80 Ruxolitinib inhibitors Dasatinib 20000 EGFR imatinib 60 60 Tofacitinib 1000 SMAD4 0.01 0.01 ● Nilotinib inhibitors 40 40 ● MAP2K4 dabrafenib imatinib gefitinib ●CTNNB1 CDKN2A alue 20 Ponatinib 20 100 gefitinib ● VHL alue

in the cell panel (nM) dabrafenib 50 Bosutinib in the cell panel (nM)) p−v

0 0 p−v NRAS NOTCH 50 2000 MYC IC 10 ABL inhibitors TP53 MAP2K4 CDKN2a.P14.

IC 0.1 ABL 0.1 mTOR inh. MSH2 ABL BRAF RB1 CDKN2A

1 inhibitors STK11 FBXW7 MEK inh. APCSMARCA4 TP53 KRAS SMARCA4 decreasing general cytooxicity FBXW7 CDKN2a.P14. PIK3CANOMLH1TCH APC PIK3KR1 mTOR inh. SMAD4 PTENRB1 BRAFABL (average PIK3KR1STK11 PIK3CA MLH1 NRASCTNNB1 most potent 0,1 KRAS 200 MYC PTEN VHL 1 1 0,1 1 10 100 1000 4,00 3,50 3,00 2,50 2,00 1,50 1,00 0,50 0,00 0.01 0.1 110 100 ABbiochemical potency (target IC50, nM) increased biochemical selectivity (selectivity entropy) 0.1 110 drug IC effect drug IC50 effect RAF / MEK inhibitors PDGFR / VEGFR inhibitors 50

100 100 Figure 4 Correlation between the biochemical and cellular activities of inhibitors: Figure 5 Anova analysis for drug response markers reveals novel associations for MEK and EGFR Dabrafenib 80 80 Axitinib (A) potency (B) selectivity. inhibitors. 60 Vemurafenib 60 Cabozantinib 40 40 Trametinib Pazopanib 20 (MEK) 20 Regorafenib 0 0 Sorafenib BRAF ABL EGFR Conclusions Sunitinib 1,E-12 Vandetanib 0,000001 1,E-04 imatinib gefitinib • Cell line panel profi ling reveals that MEK 0,00001 dabrafenib 1,E-10 1,E-03 inhibitors can be benefi cial in β-catenin mutant 0,0001 1,E-08 cabozantinib erlotinib vemurafenib afatinib 0,001 1,E-06 1,E-02 cancers and EGFR inhibitors in SMAD4 mutant EGFR inhibitors other kinase inhibitors ponatinib

p-value p-value p-value lapatinib axitinib 100 100 1,E-04 nilotinib vandetanib Afatinib 0,01 cancers 80 80 sorafenib Crizotinib 1,E-01 trametinib (ALK) pazopanib 60 Erlotinib 60 1,E-02 dasatinib 0,1 dasatinib trametinib 40 40 Everolimus • Biochemical selectivity and potency are prime Gefitinib bosutinib dasatinib (mTOR) everolimus 1,E+00 20 20 1 1,E+00 Lapatinib Temsirolimus 0,001 1 1000 0,001 1 1000 0,1 1 10 0 0 (mTOR) determinants of the target effi cacy of kinase drug IC50 effect drug IC50 effect drug IC50 effect Fasudil (ROCK) inhibitor drugs Figure 6 Comparison of targeting efficacy of marketed BRAF, ABL and EGFR inhibitors. • Integrated biochemical and cellular profi ling Figure 1 Inhibitors used in this study and their biochemical References can help to prioritize compounds and provide selectivity profiles. 1. Kitagawa et al. Activity-based kinase profi ling of approved inhibitors. Genes to cells 18: 110-122 (2013). 2. Uitdehaag et al. Comparison of the cancer gene targeting and biochemical selectivities of all targeted kinase inhibitors approved for clinical use. PLOS ONE 9: e92146 (2014). directions for optimization

NETHERLANDS TRANSLATIONAL RESEARCH CENTER B.V.