AT13148 Is a Novel, Oral Multi-AGC Kinase Inhibitor with Potent Pharmacodynamic and Antitumor Activity

Published OnlineFirst July 15, 2012; DOI: 10.1158/1078-0432.CCR-11-3313

Clinical Cancer Cancer Therapy: Preclinical Research

Timothy A. Yap1, Mike I. Walton1, Kyla M. Grimshaw2, Robert H. te Poele1, Paul D. Eve1, Melanie R. Valenti1, Alexis K. de Haven Brandon1, Vanessa Martins1, Anna Zetterlund1, Simon P. Heaton1, Kathrin Heinzmann1, Paul S. Jones3, Ruth E. Feltell2, Matthias Reule2, Steven J. Woodhead2, Thomas G. Davies2, John F. Lyons2, Florence I. Raynaud1, Suzanne A. Eccles1, Paul Workman1, Neil T. Thompson2, and Michelle D. Garrett1

Abstract Purpose: Deregulated phosphatidylinositol 3-kinase pathway signaling through AGC kinases including AKT, p70S6 kinase, PKA, SGK and Rho kinase is a key driver of multiple cancers. The simultaneous inhibition of multiple AGC kinases may increase antitumor activity and minimize clinical resistance compared with a single pathway component. Experimental Design: We investigated the detailed pharmacology and antitumor activity of the novel clinical drug candidate AT13148, an oral ATP-competitive multi-AGC kinase inhibitor. Gene expression microarray studies were undertaken to characterize the molecular mechanisms of action of AT13148. Results: AT13148 caused substantial blockade of AKT, p70S6K, PKA, ROCK, and SGK substrate phosphorylation and induced apoptosis in a concentration and time-dependent manner in cancer cells with clinically relevant genetic defects in vitro and in vivo. Antitumor efficacy in HER2-positive, PIK3CA- mutant BT474 breast, PTEN-deficient PC3 human prostate cancer, and PTEN-deficient MES-SA uterine tumor xenografts was shown. We show for the first time that induction of AKT phosphorylation at serine 473 by AT13148, as reported for other ATP-competitive inhibitors of AKT, is not a therapeutically relevant reactivation step. Gene expression studies showed that AT13148 has a predominant effect on apoptosis genes, whereas the selective AKT inhibitor CCT128930 modulates cell-cycle genes. Induction of upstream regulators including IRS2 and PIK3IP1 as a result of compensatory feedback loops was observed. Conclusions: The clinical candidate AT13148 is a novel oral multi-AGC kinase inhibitor with potent pharmacodynamic and antitumor activity, which shows a distinct mechanism of action from other AKT inhibitors. AT13148 will now be assessed in a first-in-human phase I trial. Clin Cancer Res; 18(14); 3912–23. 2012 AACR.

Introduction threonine kinases, including several members of the AGC The class I phosphoinositide 3-kinases (PI3K) are key kinase family, such as AKT, also known as protein kinase B mediators of intracellular signaling between the membrane- (PKB), phosphoinositide-dependent kinase 1 (PDK1), bound receptor tyrosine kinases (RTKs) and downstream p70S6 kinase (p70S6K), p90 ribosomal S6 kinase (RSK), effector molecules, which control many vital cellular func- serum- and glucocorticoid-induced kinase (SGK), and Rho tions, including survival, growth, proliferation, and motility kinase (ROCK; refs. 3, 4). The PI3K-AKT axis of this signaling (1, 2). Downstream of these PI3Ks lies a network of serine/ network is hyperactivated in multiple cancers through different mechanisms, including the deregulation of upstream RTKs, for example, insulin-like growth factor-1 receptor 1 Authors' Affiliations: Cancer Research UK Cancer Therapeutics Unit, (IGF-1R), and genetic alterations of PIK3CA, PTEN, or AKT Division of Cancer Therapeutics, The Institute of Cancer Research, Sutton; 2Astex Pharmaceuticals, Cambridge; and 3Cancer Research UK Drug genes, AKT1, 2, and 3 (1, 2). Thus, pharmacologic inhibition Development Office, Research Operations and Funding, London, United of this pathway is an area of great therapeutic interest (5). Kingdom Several drugs targeting the PI3K-AKT pathway are cur- Note: Supplementary data for this article are available at Clinical Cancer rently in clinical development, including inhibitors of PI3K, Research Online (http://clincancerres.aacrjournals.org/). AKT, and mTORC1/2 (5, 6). However, inhibiting PI3K-AKT Corresponding Author: Michelle D. Garrett, Cancer Research UK Cancer signaling at a single node has shown relatively limited Therapeutics Unit, Division of Cancer Therapeutics, The Institute of Cancer Research, Haddow Laboratories, 15 Cotswold Road, Sutton, Surrey SM2 clinical efficacy to date. There are several possible explana- 5NG, United Kingdom. Phone: 44-20-8722-4352; Fax: 44-20-8722-4126; tions for this. First, AKT inhibition has been shown to E-mail: [email protected] relieve feedback suppression of RTK expression and activity, doi: 10.1158/1078-0432.CCR-11-3313 which may attenuate antitumor activity (7). Second, PI3K 2012 American Association for Cancer Research. deregulation may promote cancer through both AKT-

3912 Clin Cancer Res; 18(14) July 15, 2012

AT13148, a Potent, Oral AGC Kinase Inhibitor

ies that characterize the underlying molecular mechanisms Translational Relevance of action of AT13148, including compensatory feedback Deregulated phosphoinositide 3-kinase (PI3K)-AKT loops that reveal differences compared with a more selective pathway signaling through AGC kinases is implicated AKT inhibitor. On the basis of these studies, AT13148 is in many cancers. The simultaneous inhibition of mul- now undergoing preclinical development before entry into tiple AGC kinases may increase antitumor activity and phase I clinical trials. minimize clinical resistance compared with a single kinase target. The clinical candidate AT13148 is a novel, Materials and Methods oral, multi-AGC kinase inhibitor, which has potent Cell culture and reagents pharmacodynamic and antitumor activity in human All human cancer cell lines were purchased from the tumor xenografts with clinically relevant genetic defects American Type Culture Collection and grown in their in vitro and in vivo, and shows a distinct mechanism of recommended culture medium, containing 10% FBS at action from selective AKT inhibitors. AT13148 showed 37C in an atmosphere of 5% CO and passaged for less linear pharmacokinetics, achieved therapeutically active 2 than 6 months. AT13148 (16), CCT128930 (17), and drug concentrations, and induced biomarker changes LY294002 (Calbiochem, Merck Biosciences) were made up consistent with AGC inhibition in human tumor xeno- as 10 mmol/L stocks in dimethyl sulfoxide (DMSO). grafts. Moreover, we show for the first time that induction of AKT phosphorylation at serine 473 by AT13148, In vitro kinase assays as reported for other ATP-competitive inhibitors of AKT, AT13148 was assayed against 40 kinases (National Centre is not a therapeutically relevant reactivation step for this for Kinase Profiling, Dundee, UK) and the percentage inhi- compound. These detailed preclinical and mechanistic bition at 10 mmol/L of AT13148 was determined. Individual data will facilitate the forthcoming first-in-human phase IC values were measured for selected kinases using ATP I trial of AT13148. 50 concentrations equivalent to the Km for each enzyme (Invitrogen).

Protein immunoblotting and immunoassay dependent and AKT-independent mechanisms, the latter Cells were harvested, lysates prepared, protein estima- involving the AGC kinases PDK1 and SGK (8). Third, tions conducted, and Western blots undertaken as described inhibition of a single node such as PI3K or AKT may allow (18), using the following antibodies: pSer473 AKT, AKT, clinical resistance as reported for the selective BRAF inhib- pSer9 GSK3b, GSK3b, pSer235/236 S6 ribosomal protein itor vemurafenib (PLX4032, Roche; ref. 9). Furthermore, (S6RP), S6RP, pSer330 NDRG, NDRG, pSer157 VASP, concurrent blockade of multiple components of the PI3K VASP, pSer19 MLC2, MLC2, pThr24 FOX01/pThr32 network may have greater therapeutic value than inhibition FOX03a, FOX01, PRAS40, cleaved PARP, IRS2, pThr1135 of any single target (10). Thus, the simultaneous inhibition Rictor, Rictor, cyclin E2, c-MYC (Cell Signaling Technolo- of several essential nodes of the PI3K signaling network may gy), PIK3IP1 (Abcam), pThr246 PRAS40 (Upstate), cleaved provide greater overall suppression of key pathways, with caspase-3 (Epitomics), cyclin D1, p27, and glyceraldehyde- the potential for improved therapeutic efficacy across a 3-phosphate dehydrogenase (GAPDH; Neomarkers). broader range of cancer types with less opportunity for Western blot analysis of MES-SA cells was conducted as resistance to develop. above, except that samples were lysed in 40 mmol/L Tris/ A possible strategy for such a targeted combinatorial HCl pH 7.5, 274 mmol/L NaCl, 2% Triton-X-100, and 20% blockade is the development of a drug that inhibits AKT glycerol. Samples were resolved by SDS-PAGE, blotted onto together with other key AGC kinases that form part of the nitrocellulose filters, blocked with Odyssey blocking buffer PI3K signaling network. In this paper, we present the first (LI-COR Biosciences) and incubated overnight with the detailed pharmacologic characterization of the novel, oral respective antibodies (pSer9 GSK3b, GSK3b, pSer240/244 clinical drug candidate AT13148, a potent multi-AGC S6RP, S6RP, pThr1462 tuberin, tuberin, and cleaved PARP kinase inhibitor. AT13148 was identified using high- (Cell Signaling Technology). Primary antibodies were throughput X-ray crystallography and fragment-based lead labeled with infrared IRDye-labeled secondary antibodies discovery techniques (11–14). We show that this orally (LI-COR Biosciences) and proteins visualized using the bioavailable drug candidate achieves active tumor expo- Odyssey Infrared Imager (LI-COR Biosciences). sures, induces robust pharmacodynamic biomarker mod- For studies using the Meso Scale Discovery (MSD) elec- ulation and apoptosis in cancer cells with clinically relevant trochemoluminescence platform, U87MG cell lysates were genetic defects both in vitro and in vivo, and exhibits anti- prepared as described earlier and probed for pSer473 AKT, tumor efficacy in multiple human tumor xenograft models. AKT, pSer9 GSK3b, GSK3b, pThr421/424 p70S6K, and Moreover, we show that AT13148-induced hyperphosphor- p70S6K according to the manufacturer’s instructions (19). ylation of Ser473 AKT is not a therapeutically relevant reactivation step for this compound, in contrast to previous Cell-cycle effects and Annexin V staining suggestions for other ATP-competitive inhibitors of AKT Following drug or vehicle treatment, U87MG cells were (15). Finally, we describe gene expression microarray stud- labeled with either bromodeoxyuridine or propidium

www.aacrjournals.org Clin Cancer Res; 18(14) July 15, 2012 3913

Yap et al.

iodide as described and analyzed by ﬂow cytometry (18). The proportion of cells in each phase of the cell cycle was A determined using either the Cell Quest Pro Software pack- CI age (BD Biosciences) or WinMDI 2.8. Annexin V staining OH was carried out according to the manufacturer’s instructions NH2 (BD PharMingen)

In vivo studies Pharmacokinetic and pharmacodynamic analyses. All procedures were in accordance with UK Home Office reg- AT13148 ulations under the Animals (Scientific Procedures) Act 1986, approved by The Institute of Cancer Research’s Ethics N N Committee and in accordance with published guidelines H (20). Mice were allowed access to food and water ad libitum. For pharmacokinetic analysis, male athymic BALB/c mice B were obtained from Harlan. AT13148 was formulated in 10% DMSO, 1% Tween-20, and 89% saline and administered at 5 mg/kg i.v. or p.o. Duplicate samples of heparin- ized whole blood were collected by cardiac puncture at 1, 2, 4, 6, 8, 16, 24, and 72 hours after dosing. Plasma and tissues Glycine loop (liver, kidney, spleen, and muscle were also taken) were prepared and frozen at 20C until analysis. AT13148 was Tyr 122 extracted from plasma and tissues using acetonitrile containing an internal standard and quantified using a liquid Asn 171 chromatography tandem mass spectrometry (LC-MS/MS) Met 173 method and appropriate standard curves. Pharmacokinetic parameters were determined using WinNonLin software version 5.2. To assess pharmacokinetic and pharmacodynamic rela- tionships, a single dose of AT13148 (30, 40, or 50 mg/kg Figure 1. The structure and biologic activity of AT13148. A, chemical p.o.) was given to groups of 3 female athymic (CrTac:Ncr- structure of AT13148. B, crystal structure of AT13148 bound to Fox1nu) mice bearing established subcutaneous xeno- recombinant human PKA-AKT chimera. graftsofMES-SA,BT474,ormale athymic mice bearing PC3 xenografts. The drug was formulated in vehicle containing 10% DMSO 1% Tween-20 in 89% saline. described (17) 3 times weekly. For the MES-SA study, Plasma and tumor samples (n ¼ 3) were obtained at %T/C represents mean tumor volume of treated animals 2, 6, and 24 hours after compound or vehicle adminis- divided by mean control tumor volume, expressed as a tration. Tumors were divided into 2 halves, snap frozen, percentage on any particular day. For the BT474 and PC3 stored at 20C and each half used for pharmacokinetic studies, %T/C represents mean tumor weight of treated and pharmacodynamic analyses, respectively. For phar- animals divided by mean control tumor weight at the end macokinetic analysis, tissue samples were first homoge- of the experiment, expressed as a percentage. nized in 5 volumes (w/v) of acetonitrile/water (50/50). AT13148 was extracted from plasma and tissue homo- Statistical analyses genates and quantified as described earlier. Statistical significance was determined using 1- or 2-tailed For pharmacodynamic studies, tumors were ground to a t tests as appropriate with GraphPad Prism 5.0. powder under liquid nitrogen, lysed and centrifuged to Other methods including those for crystallography and remove debris. Protein content was measured using BCA microarrays studies are detailed in the Supplementary or Bradford reagent and samples evaluated by Western Material. blots or by MSD analysis as described earlier. Results Efficacy studies Identification of a potent ATP-competitive inhibitor of Human MES-SA, BT474, and PC3 tumor cells were key AGC kinases injected s.c. into the right flank of female or male athymic AT13148 was discovered using fragment-based screening mice. When tumors reached 100 mm3 mean volume, combined with structure-based design as previously describ- animals were randomized and treated with vehicle (10% ed (refs. 13 and 16; Fig. 1A). The structure of AT13148 bound DMSO, 1% Tween-20, and 89% saline) or AT13148 p.o. in the ATP pocket of the PKA-AKT chimera was solved by using the dosage schedules detailed in the figures. Body X-ray crystallography (Fig. 1B). The use of this PKA-based weight and tumor size were determined as previously surrogate provides a robust and validated structural system

3914 Clin Cancer Res; 18(14) July 15, 2012 Clinical Cancer Research

AT13148, a Potent, Oral AGC Kinase Inhibitor

for understanding interactions between inhibitors and AKT potent inhibitor of the AGC kinases p70S6K, PKA, ROCKI, (14). As expected, the structure shows that AT13148 fulﬁlls and ROCKII that also potently inhibits the related family the requirements of the canonical 3-point pharmacophore members AKT1, AKT2, AKT3, RSK1, and SGK3 (Supple- needed for potent binding to the ATP site of AKT, forming mentary Table S2). hydrogen-bonding interactions with the kinase hinge, elec- trostatic interactions with the ribose site, and hydrophobic AT13148 inhibits the proliferation and AGC kinase contacts with a lipophilic pocket in the glycine-rich loop. activity of cancer cells Screening of AT13148 against a panel of kinases at 10 AT13148 potently inhibited proliferation with GI50 mmol/L revealed >80% inhibition of the structurally related values of 1.5 to 3.8 mmol/L across a selected panel of cancer AGC kinases AKT, PKA, ROCK2, p70S6K, MSK, RSK1/2, and cell lines (Supplementary Table S3) representing common SGK (Supplementary Table S1). Further studies showed that human malignancies with deregulation of PI3K-AKT- IC50 values for p70S6K, PKA, ROCKI, and ROCKII were all mTOR or RAS-RAF pathways. The effect of 1-hour exposure less than 10 nmol/L and those for AKT1, 2, and 3 were 38, to AT13148 on AKT and p70S6K signaling was initially 402, and 50 nmol/L, respectively (Supplementary Table explored in PTEN-deﬁcient U87MG glioblastoma cells (Fig. S2). For the related AGC kinases RSK1 and SGK3, the 2A). Marked induction of pSer473 AKT occurred at all IC50 values were 85 and 63 nmol/L, respectively. In contrast, concentrations. Nevertheless, phosphorylation of the 2 AKT IC50 values for the non-AGC kinases CHK2 and Aurora B substrates GSK3b and PRAS40 was inhibited at AT13148 were both greater than 800 nmol/L. Therefore, AT13148 is a concentrations >1 and 5 mmol/L AT13148, respectively.

A C Concentration (µmol/L) Time (h) AT13148 (μmol/L) D 0.05 0.1 0.5 1 5 10 20 LY Time (h) D 0.5 1 2 4 8 16 24 LY D

pSer473 AKT pSer473 AKT

AKT AKT

pSer9 GSK3β pSer9 GSK3β

GSK3β GSK3β

pThr246 PRAS40 pThr246 PRAS40

PRAS40 PRAS40

pSer235/236 S6RP pSer235/236 S6RP

S6RP S6RP

GAPDH GAPDH BD Concentration (µmol/L) Concentration (µmol/L) AT13148 (μmol/L) D 0.05 0.1 0.5 1.0 5.0 10 20 LY D AT13148 (μmol/L) C 0.03 0.1 0.3 1 3 10 30 C pSer9 GSK3β pSer157 VASP GSK3β VASP pSer240/244 S6RP pSer19 MLC2 S6RP MLC2 pThr1462 Tuberin

pSer330 NDRG1 Tuberin

NDRG1

GAPDH

Figure 2. The effect of AT13148 exposure on AGC kinase biomarker expression. PTEN-deﬁcient U87MG human glioblastoma cells were incubated with AT13148 for 1 hour (A and B) at the concentrations indicated or (C) with 10 mmol/L AT13148 for the times indicated. D, PTEN-deﬁcient MES-SA human uterine sarcoma cells were incubated for 1 hour with AT13148 at the concentrations indicated. Immunoblotting was carried out for the proteins indicated. GAPDH was used as a loading control. C, no treatment control; D, DMSO vehicle control; LY, positive control (PI3K inhibitor LY294002, 30 mmol/L).

www.aacrjournals.org Clin Cancer Res; 18(14) July 15, 2012 3915

Yap et al.

Phosphorylation of the p70S6K substrate, Ser235/236 From these studies it can be concluded that although S6RP, was also inhibited at concentrations >5 mmol/L. Total AT13148 induces phosphorylation on Ser473 AKT, com- protein levels remained generally constant throughout this pound removal does not cause downstream phosphory- study, apart from PRAS40, which increased with AT13148 lation signals to recover rapidly or to rebound to greater treatment (Fig. 2A). than pretreatment levels. Immunoblotting after 1-hour treatment with AT13148 in U87MG cells also showed clear inhibition of phos- Pharmacokinetics and pharmacodynamics of AT13148 phorylation on direct substrates of the AGC kinases To determine if therapeutically active concentrations of PKA, ROCK, and SGK, namely Ser157 VASP, Ser19 MLC2, AT13148 can be achieved in vivo, the pharmacokinetic and Ser330 NDRG1, respectively (Fig. 2B). Inhibition of profile of this drug candidate was investigated in BALB/c pSer157 VASP and pSer19 MLC2 was observed from mice. After administration of 5 mg/kg i.v., AT13148 showed 0.5 mmol/L AT13148, whereas pSer330 NDRG1 was a low plasma clearance of 1.68 L/h/kg, which is less than inhibited from 5 mmol/L AT13148. Expression of total half liver blood flow, and a large volume of distribution of protein for all 3 AGC kinase substrates remained constant 9.05 L/kg with a terminal half-life of 2.83 hours (Fig. 3A; up to 20 mmol/L AT13148. Supplementary Table S4). Oral drug administration of We next determined the effect of 10 mmol/L AT13148 on 5 mg/kg of AT13148 resulted in complete bioavailability. AKT and p70S6K signaling output over time in U87MG cells Increasing oral doses from 5 to 50 mg/kg showed linear (Fig. 2C). Induction of pSer473 AKT was detected at the pharmacokinetics, with plasma AUC0–¥ increasing in pro- earliest time point assayed (0.5 hours) and sustained portion with dose (Fig. 3B). throughout the compound exposure time. Decreases in Figure 3C shows the concentrations of AT13148 achieved both pSer9 GSK3b and pThr246 PRAS40 were also observed in athymic mouse plasma and HER2-positive PIK3CA- from 0.5 to 24 hours, while a decrease in pSer235/236 S6RP mutant BT474 human breast cancer xenografts after the was first detected at 1 hour. Total protein levels remained administration of 2 daily doses of 40 mg/kg p.o. Consistent generally constant throughout the time course. with the large volume of distribution described, tumor Inhibition of AKT and p70S6K signaling was also AT13148 concentrations greatly exceeded plasma concen- observed in PTEN-deficient MES-SA cells after treatment trations at 2, 6, and 24 hours with tumor:plasma ratios of with AT13148 (Fig. 2D). A 1-hour drug exposure markedly 8.5, 7.0, and 13.6, respectively. Moreover, tumor AT13148 inhibited phosphorylation of the AKT substrates GSK3b concentrations were at least 9 times greater than the in vitro and tuberin, and the downstream p70S6K substrate S6RP, at GI50 value of 1.8 mmol/L for this cell line, maintained over AT13148 concentrations of 3 mmol/L or above, with min- 24 hours. imal effects on total protein levels. Taken together, these Pharmacodynamic biomarker changes measured in the biomarker data clearly show that AT13148 can inhibit a same BT474 xenografts by MSD immunoassay are shown number of AGC kinases, including AKT, in human tumor in Fig. 3D. There was an increase in pS473 AKT at 2, 6, and cell lines in vitro. 24 hours after AT13148 treatment, consistent with in vitro observations (Fig. 2C). Importantly, phosphorylation of Phosphorylation of AKT substrates remains the AKT substrate GSK3b and the downstream target suppressed, despite pSer473 AKT induction by p70S6K was significantly decreased, consistent with sus- AT13148 tained inhibition of AKT activity by AT13148 in vivo at this ATP competitive inhibitors of AKT induce phosphoryla- dose. Comparable studies in PTEN-deficient PC3 human tion on Ser473 of this kinase and this may have therapeutic prostate tumor xenografts also showed high AT13148 implications (15). Specifically, this phosphorylated form tumor:plasma ratios and significant decreases in the phos- of AKT has been shown to be hyperactive in vitro when phorylation of AKT biomarkers (Supplementary Fig. S2A dissociated from the inhibitor, thus potentially leading to and S2B). Clear inhibition of phosphorylation of the AKT activation of AKT targets in cells, and in turn promoting substrates GSK3b, tuberin, and the p70S6K target S6RP were oncogenesis. To address this, we monitored both pSer473 also observed in PTEN-deficient MES-SA human uterine AKT and phosphorylation of targets downstream of AKT tumor xenografts after treatment with 40 and 50 mg/kg in U87MG cells in vitro after exposure to AT13148 at 1 p.o. of AT13148 over 24 hours (Fig. 4A; Supplementary and 10 mmol/L for 1 and 24 hours, followed by the Fig. S3). Importantly, induction of cleaved PARP was removal of compound for 0, 4, 8, or 24 hours (Supple- observed at these doses of AT13148 over 24 hours, indi- mentary Fig. S1A and S1B). Analysis of total and phos- cating that AT13148 induces apoptosis in solid tumors. pho-protein signals using the quantitative MSD electro- Taken together, these data suggest that AT13148 marked- luminescence immunoassay revealed that the pSer473 ly inhibits the activity of both AKT and p70S6K AGC kinases AKT signal was strongly induced in U87MG cells at both in human tumor xenografts with differentially activated concentrations and all time points, and was sustained PI3K pathways after oral administration. even after 24 hours of compound removal. In contrast, both the pSer9 GSK3b and pThr421/424 S6K signals were In vivo antitumor activity of AT13148 inhibited under the same conditions and only showed Following the demonstration of clear in vitro activity and partial recovery up to 24 hours after compound removal. promising in vivo pharmacokinetic and pharmacodynamic

3916 Clin Cancer Res; 18(14) July 15, 2012 Clinical Cancer Research

AT13148, a Potent, Oral AGC Kinase Inhibitor

A B 1,000 150 ) 1 100 − 100 mol/L) 10 μ mol/Lh μ ( 1

40 mg/kg p.o. 0– ∞ 50

0.1 5 mg/kg i.v. Spleen 40 mg/kg p.o. AT13148 ( AT13148 5 mg/kg p.o. Liver 40 mg/kg p.o. AUC 0.01 0 0 2 4 6 820 24 0 10 20 30 40 50 Time (h) Dose (mg/kg p.o.) C D pSer473 AKT:total AKT 10 * pSer9 GSK3β:total GSK3β pThr421/424 p70S6K:total p70S6K 8 Plasma (40 mg/kg p.o.) 6 BT474 tumor (40 mg/kg p.o.) 4 ** 50 2 1.50

40 1.25

30 1.00 mol/L) μ 0.75 20 **** *** 0.50

AT13148 ( AT13148 10 0.25

0 Ratio of phospho/total signal (normalized to control) 0.00 2 6 24 Cont 2 h 6 h 24 h Time (h) AT13148 (40 mg/kg p.o.)

Figure 3. Characterization of the pharmacokinetic properties and pharmacodynamic effects of AT13148 in vivo. A, plasma and tissue pharmacokinetics of AT13148 in mice following i.v. and oral (p.o.) administration. Values are mean SE for 3 to 5 mice per time point. B, relationship between dose and exposure for AT13148 after single-dose oral administration in mice. Data from 2 independent experiments (open and closed circles). C, concentrations of AT13148 measured in plasma (open bar) and HER2-positive, PIK3CA-mutant BT474 breast cancer xenografts (closed bar) taken 2, 6, and 24 hours after the second dose of AT13148 given p.o. at 40 mg/kg on 2 consecutive days of treatment (i.e., 1 cycle). Dotted line represents 96-hour GI50 value (1.6 mmol/L) in BT474 cells in vitro. Bars show mean SE for 3 determinations. D, quantification of pharmacodynamic biomarker changes (ratio of phospho vs. total protein) measured using the MSD electrochemoluminescent platform for the tumors shown in C. Dashed line represents this ratio in vehicle-treated controls as 1.0. Values are mean SE for 3 determinations. Statistics: , P < 0.05; , P < 0.01; , P < 0.001 significantly different from control. Cont, control. properties indicating target modulation, the antitumor 3A–D and Fig. 4A; Supplementary Figs. S2A, S2B, and S3), activity of AT13148 was assessed in multiple human tumor which indicate that for this oral route of administration, xenograft models. Figure 4B shows that AT13148 markedly plasma drug concentrations were at or just above the in inhibited the growth of established MES-SA human uterine vitro GI50 values, whereas tumor drug concentrations sarcoma tumor xenografts at the same doses of 40 and 50 greatly exceeded GI50 values for at least 24 hours for mg/kg that also resulted in pharmacodynamic changes these tumor cells. AT13148 also showed growth inhibi- indicative of target engagement (Fig. 4A; Supplementary tion in the mutant KRAS A549 lung adenocarcinoma Fig. S3); the percentage treated/control (%T/C) was 41% xenograft model (Supplementary Fig. S4). Tumor growth and 54% at 40 and 50 mg/kg p.o. of AT13148, respectively, was significantly inhibited in the MES-SA, BT474, and when measured on day 11 of treatment. PC3 tumor xenografts (P < 0.01), with marked growth Further antitumor studies were undertaken in estab- inhibition in A549 (P ¼ 0.0788) and minor animal lished BT474 (Fig. 4C) and PC3 human tumor xenograft weight loss in all studies (Supplementary Figs. S2D, models (Supplementary Fig. S2C). AT13148 inhibited the S4B, S5A, and S5B). Our results therefore clearly show growth of both these models giving T/C values of 35.8% that orally administered AT13148 induces sustained inhi- on day 26 and 65.4% on day 27, for BT474 and PC3 bition of the AGC kinases AKT and p70S6K and exhibits respectively. These results are consistent with both marked antitumor effects in 4 genetically relevant human the pharmacokinetic and pharmacodynamic data (Figs. tumor xenograft models.

www.aacrjournals.org Clin Cancer Res; 18(14) July 15, 2012 3917

Yap et al.

A 50 mg/kg AT13148

C1 6 h 8 h 16 h 24 h C2 6 h 8 h 16 h 24 h C3 Figure 4. Pharmacodynamic biomarker and antitumor activity of β pSer9 GSK3 AT13148 in human tumor GSK3β xenografts. A, effect of a single dose of 50 mg/kg AT13148 given p. pSer240/244 S6RP o. on pharmacodynamic biomarkers in PTEN-deﬁcient S6RP MES-SA human uterine sarcoma pThr1462 Tuberin xenografts. Protein expression was assessed by immunoblotting Tuberin using GAPDH as a loading control. B, antitumor activity of AT13148 at Cleaved PARP 40 or 50 mg/kg p.o. on an intermittent schedule [every 3 days GAPDH or (daily on 2 consecutive days) every 5 days] in established MES- BC SA human uterine sarcoma MES-SA xenograft BT474 xenograft xenografts. C, antitumor activity of AT13148 at 40 mg/kg p.o. on 2 800 800 Vehicle consecutive days with 3 days rest 50 mg/kg q3d p.o. 700 Vehicle control between treatments in HER2- 600 40 mg/kg qdx2 p.o. 600 40 mg/kg p.o. positive, PIK3CA-mutant BT474 500 human breast cancer xenografts. 400 400 Values are mean SE for 5 to 8

(% d 0) (% 300 ** d 0) (% ** mice per time point. Statistics: , P 200 ** 200 < 0.01 signiﬁcantly different from 100 Relative tumor volume Relative tumor volume Relative tumor control. 0 0 0 2 4 6 8 10 12 0 2 4 6 8 10 12 14 16 18 20 22 24 26 Time (d) Time (d)

Gene expression microarray analysis sion was also confirmed for p27KIP1 at the protein level (Fig. The molecular and cellular responses of the multi-AGC 5C). Expression of the FOXO1 gene is itself transcriptionally kinase inhibitor AT13148 drug candidate were explored regulated by FOXO3a (23, 24), and FOXO1 expression was by gene expression microarray and compared with increased upon treatment with both compounds by gene those of the previously described AKT-selective inhibitor expression microarray and confirmed at the protein level by CCT128930 (17) after 6-hour treatment of U87MG cells at Western blot analysis (Figs. 5C and 6B; Supplementary 0.1 mmol/L, 1 GI50, and 3 GI50 concentrations (Figs. 5A– Table S5A and S5B). D and 6A–C). The expression levels of 563 genes were Interestingly, the gene expression network analysis also significantly altered after AT13148 treatment, whereas the revealed evidence of the induction of both negative and expression levels of 898 genes were significantly changed positive feedback mechanisms (Fig. 6C). Both PIK3CA with CCT128930 (Fig. 5A and B; Supplementary Table S5A and IRS2 can be transcriptionally regulated by FOXO3a and S5B). Importantly, there was an overlap of 147 genes and induction of both has been reported after treatment that exhibited significantly altered expression in response of cells with PI3K inhibitors (25–27). This could poten- to both AT13148 and CCT128930 (Fig. 5B). tially lead to reactivation of the PI3K/AKT pathway. We The effect of each compound on AKT pathway signaling observed increased expression of both PIK3CA and IRS2 was also simultaneously confirmed by Western blot analysis genes after treatment with AT13148 and CCT128930 (Fig. (Fig. 5C). Effects on AKT signaling were reflected by per- 6C and Supplementary Table S5A and S5B). We also turbations in the gene expression network (Fig. 6A–C). noted a concentration-dependent band shift downward Specifically, inhibition of AKT has been reported to block of IRS2 on the Western blots for CCT128930, which was phosphorylation of the forkhead transcription factors less marked for AT13148 (Fig. 5C). We confirmed using FOXO3a and FOXO1, which then translocate to the nucleus lambda phosphatase that thisbandshiftisbecauseof and activate transcription of downstream targets. The neg- hypophosphorylation of IRS2 (Fig. 5D). AKT RNA inter- ative cell-cycle regulators p27KIP1 and cyclin G2, encoded by ference (RNAi) studies confirmed that this band shift was CDKN1B and CCNG2, respectively (21, 22) are both FOXO AKT-dependent (Fig. 5D). The more specific AKT inhib- targets and their gene expression levels were upregulated in itor CCT128930 induced a greater mobility shift and response to both compounds, although the level of induc- marked accumulation of IRS2 protein, compared with tion appeared stronger with CCT128930 (Fig. 6A; Supple- that observed for AT13148 or AKT RNAi, suggesting a mentary Table S5A and S5B). The upregulation in expres- greater blockade on targets that impact on IRS2. However,

3918 Clin Cancer Res; 18(14) July 15, 2012 Clinical Cancer Research

AT13148, a Potent, Oral AGC Kinase Inhibitor

A B

3 CCT128930 AT13148 CCT128930 0.1 µmol/L AT13148 0.1 µmol/L

CCT128930 1 × GI50 (6.3 µmol/L) 751 147 416

AT13148 1 × GI50 (3.3 µmol/L) Fold change CCT128930 3 × GI50 (18.9 µmol/L) AT13148 3 × IC (9.9 µmol/L) 50 0.3 CD CCT128930 AT13148 CCT128930 AT13148 AKT RNAi D + 0.1 1 × IC 3 × IC 0.1 1 × IC 3 × IC 0.1 1 × IC 3 × IC 0.1 1 × IC 3 × IC D 10 30 D µmol/L 50 50 µmol/L 50 50 D 50 50 50 50 λ µmol/L µmol/L PP NA NT nmol/L nmol/L

pSer473 AKT IRS2 pSer473 AKT

AKT PIK3IP1 AKT

β pSer9 GSK3 pThr1135 pThr246 RICTOR PRAS40 β Total GSK3 RICTOR

pThr246 PRAS40 PRAS40 Cyclin E2

IRS2 PRAS40 Cyclin D1

pThr32 FOXO3a p27Kip1 pThr24 FOXO1 PIK3IP1 GAPDH

FOXO1

GAPDH

Figure 5. Comparison of the effects of 6-hour treatment of AT13148 versus CCT128930 on gene and protein expression in U87MG human glioblastoma cells. A, heat map showing the significant gene expression changes (P < 0.05) induced in PTEN-deficient U87MG glioblastoma cells by AT13148 and CCT128930. B, Venn diagram quantifying the gene expression changes shown in A. RNA samples from 3 independent experiments were analyzed using the Agilent Two- Color Microarray-Based Gene Expression Analysis protocol (version 5.7). C, U87MG cells were treated for 6 hours with AT13148 or CCT128930 at the concentrations shown, lysates prepared, and immunoblotted for the proteins indicated. D, DMSO control. D, effect of lambda phosphatase treatment and RNAi knockdown of AKT1, 2, and 3 on the expression of selected AKT biomarkers and IRS2 bandshift mobility in U87MG human glioblastoma cells. Protein expression was assessed by immunoblotting as indicated. D, DMSO control; l, lambda phosphatase treatment; NA, no addition; NT, nontargeting control siRNA oligo. this loss of phosphorylation on IRS2 does not lead to P value) GO categories, which is consistent with gene expres- downstream reactivation of the PI3K-AKT pathway, as can sion changes previously observed with PI3K inhibitors be seen by the continued suppression of phosphorylation (refs. 33, 34; Supplementary Table S6B). The effects of of AKT substrates (Fig. 5C). Expression of the tumor CCT128930 on the expression of cell-cycle genes were gen- suppressor PIK3IP1, a potential transcriptional target of erally greater than with AT13148 (Fig. 6A). For example, the FOXO1 and CREB1 (28), was also induced at the gene downregulation of positive cell-cycle regulators such as transcription level following treatment with both com- CDC25A, CDC6,andCCNE1 was more pronounced with pounds (Fig. 6C; Supplementary Table S5A and S5B). CCT128930 than AT13148. However, in contrast to the This induction was confirmed at the protein level (Fig. downregulation of CCND1 in response to CCT128930, a 5C). PIK3IP1 has been shown to decrease PI3K p110a modest but significant and reproducible increase was activity both in vitro and in vivo (29, 30). observed with AT13148, which was confirmed by TaqMan Gene ontology (GO) analysis (31, 32) for the biologic (data not shown). The downregulation of cyclin D1 by processes of the 147 common genes that showed signifi- CCT128930 was also confirmed at the protein level, whereas cantly altered expression in response to both AT13148 and AT13148 had no effect on the level of cyclin D1 protein (Fig. CCT128930 revealed that most genes affected were involved 5C; Supplementary Fig. S6A and S6B). CCT128930 also in regulation of the cell cycle, and apoptosis (Figs. 5B and 6A caused a decrease in the expression of the proapoptotic and B; Supplementary Table S6A–S6C). GO analysis for the marker c-MYC, both at the gene and protein level, whereas genes showing altered expression with the specific AKT AT13148 did not (Fig. 5B; Supplementary Fig. S6A and S6B). inhibitor CCT128930 revealed that "cell cycle" was the most In contrast to CCT128930, the GO analysis for AT13148 common term in the top 20 most statistically significant (by identified "cell death," "programmed cell death," and

www.aacrjournals.org Clin Cancer Res; 18(14) July 15, 2012 3919

Yap et al.

A Cell-cycle genes

G1 G2–M

1 2 3 4 5 6 ID4 ID4 BLM PML PML PML PML OIP5 PIM2 PIM1 BCL2 SKP2 TIPIN TIPIN BCL2 BUB3 SKP2 EXO1 EXO1 RRS1 CDK6 KIF15 CDC6 CDC7 CDK6 EGFR CDC7 CDC6 EGFR PINX1 NIPBL LATS2 LATS2 LATS2 LATS2 SPC25 TUBB3 FBXO5 FBXO5 STAG2 BRCA1 FOXN3 FOXN3 FOXN3 SGOL2 XRCC2 CCNE1 CCNB2 CCND1 CCNE1 CCNG2 CCNG2 CCNG1 CCND1 HMGA2 FBXO31 CLASP2 TUBB2A FBXO31 CDC25A PPP3CB ADAM17 CDC25A PPP3CB CDKN1B ADAM17 CDKN1C MAPRE2 CDKN1B CDKN1C PSMC3IP SUV39H2 MAP3K11 ANAPC10 MAP3K11 GADD45A

B Apoptosis genes

1 2 3 4 5 6 IL7 JUN JMY XPA PML PML BMF MNT MYC PPIF IER3 MITF MITF MITF PIM2 PIM3 PIM1 TRIO BCL3 STK4 BCL6 ETS1 BCL2 ETS1 DLX1 FGF2 DLC1 DLC1 SKP2 VAV3 TERT BTG1 BBC3 BBC3 FGD3 POLB SGK3 GAS1 BDNF EGFR GDF5 IGF1R BIRC3 TIMP3 TRAF1 P2RX7 P2RX7 NR4A2 NR4A1 THBS1 CASP6 MCF2L MCF2L HSPA5 DAPK1 PTGS2 PTGS2 BARD1 BRCA1 DUSP1 NUAK2 DYRK2 FOXC1 FOXC1 XRCC2 CARD8 DEDD2 NUPR1 FOXO1 FOXO1 BNIP3L GNRH1 CITED2 PIK3CA STK17B BCL11B AMIGO2 TUBB2C HSPA1A RB1CC1 MAP2K6 ADAM17 CDKN1B KCNMA1 BCL2L11 PLEKHF1 GPR109B MAP3K11 TP53INP1 HERPUD1 SERPINB2 SERPINB2

Key 1. CCT128930 0.1 µmol/L

2. CCT128930 ×1 IC50 3. CCT128930 ×3 IC50 4. AT13148 0.1 µmol/L

5. AT13148 ×1 IC50 6. AT13148 ×3 IC50 C PI3K pathway genes

IKBKB RPS6KA3 NFKB1 GSK3A PTEN RPS6KB1 YWHAE 3 CCND1 CCND1 MYC FOXO1 CDKN1B

Fold change FOXO1 PIK3CA 0.3 IGFBP1 Key IGF1R 1. CCT128930 0.1 µmol/L CCNG2 2. CCT128930 ×1 IC50 3. CCT128930 ×3 IC CCNG2 50 4. AT13148 0.1 µmol/L IRS2 5. AT13148 ×1 IC50 PIK3IP1 6. AT13148 ×3 IC50 1 2 3 4 5 6

Figure 6. Gene ontology analysis for the biologic processes of the 147 common genes that showed signiﬁcantly altered expression in response to 6-hour treatment of both AT13148 and CCT128930 in U87MG human glioblastoma cells. A, cell-cycle genes; B, genes involved in apoptosis; C, PI3K pathway genes.

3920 Clin Cancer Res; 18(14) July 15, 2012 Clinical Cancer Research

AT13148, a Potent, Oral AGC Kinase Inhibitor

"apoptosis" as the most common terms in the 20 GO and pathway blockade. We have showed that AT13148 has categories with the most statistically significant P values. antiproliferative activity in a range of in vitro models har- (Supplementary Table S6C). These data are consistent with boring different genetic abnormalities, including pathogen- the potent induction of apoptosis seen after treatment with ic PTEN, KRAS, PIK3CA, and HER2 aberrations (Supple- AT13148 both in vitro and in vivo. Furthermore, a direct mentary Table S3). Interestingly, the GI50 values obtained comparison of the cellular effects of AT13148 (GI50 ¼ 3.3 from these tumor cell lines were broadly similar despite the mmol/L, Supplementary Table S3) versus CCT128930 (GI50 different oncogenic alterations and therefore we intend to ¼ 6.3 mmol/L; ref. 17) identified a marked increased in expand these observations to a much larger panel of human PARP and caspase-3 cleavage with AT13148 versus cancer cell lines. These results may be in part because of the CCT128930 at approximately equipotent doses of 10 and simultaneous blockade of different AGC kinases, reducing 20 mmol/L, respectively (Supplementary Fig. S6A and S6B). opportunities for the disruption of negative feedback loops This was further emphasized by the fact that AT13148 and cross-talk with other key signaling pathways, and thus showed a significantly greater degree of Annexin V staining attenuating the potential for intrinsic resistance. Further- than CCT128930 at equipotent concentrations, again indic- more, in our in vitro studies with PTEN-deficient U87MG ative of a higher level of apoptosis with AT13148 (Supple- human glioblastoma and similarly PTEN-deficient MES-SA mentary Fig. S6C). Moreover, AT13148 inhibited the phos- uterine sarcoma tumor cell lines, we have shown that phorylation of the AKT substrates GSK3b, PRAS40, FOXO1, AT13148 causes substantial blockade of AKT, p70S6K, PKA, FOXO3a at similar concentrations to CCT128930 (Fig. 5C; ROCK, and SGK substrate phosphorylation in both a con- Supplementary Fig. S6A and S6B), but was 10-fold more centration- and time-dependent manner, confirming that potent against the ROCK substrate MCL2 (Supplementary AT13148 can simultaneously inhibit multiple AGC kinases Fig. S6A and S6B), suggesting that inhibition of this AGC in these cancer cells (Fig. 2A–D). kinase may contribute to the increased apoptotic cell death. Having showed promising pathway modulation and Furthermore, there was a clear difference in the cell-cycle antiproliferative effects on cancer cells in vitro, our subse- effects of the 2 compounds, with CCT128930 causing a quent pharmacokinetic studies showed that drug exposure predominant G1/S arrest with loss of S-phase at increasing is related linearly to the administered oral dose of AT13148 concentrations, in contrast to a predominant G2/M arrest (Fig. 3A and B). Furthermore, oral administration of seen with AT13148 (Supplementary Figs. S6D, S6E, and AT13148 gave high tumor:plasma concentrations for at S7). least 24 hours in mice bearing HER2-positive, PIK3CA- Summarizing this gene network analysis, both the AGC mutant BT474 breast or PTEN-deficient PC3 human pros- kinase inhibitor AT13148 and the more AKT-selective tate cancer xenografts (Fig. 3C; Supplementary Fig. S2A). CCT128930 show molecular effects in cancer cells consis- These exposures greatly exceeded the in vitro antiprolifera- tent with blockade of AKT signaling, leading to changes in tive GI50 values that would be predicted to produce phar- gene expression that include induction of upstream regu- macodynamic biomarker modulation, pathway blockade, lators. However, the 2 agents clearly also have distinct effects and antitumor efficacy. Indeed, inhibition of signaling in cancer cells. Although CCT128930 primarily modulates output was confirmed by the marked inhibition of phos- genes in the network regulating cell cycle and causes a G1 phorylation on both AKT and p70S6K substrates for up to phase arrest, AT13148 has a predominant effect on apo- 24 hours in both of these cancer models, as well as in PTEN- ptosis genes and causes a greater apoptotic phenotype, with deficient MES-SA uterine tumor xenografts (Figs. 3D a secondary effect on cell cycle at the G2–M phase. and 4A; Supplementary Figs. S2B and S3). Subsequently, oral efficacy studies with AT13148 showed antitumor effects in all 3 clinically relevant human tumor xenografts (Fig. 4B Discussion and C; Supplementary Fig. S2C). These data indicate that PI3K signaling is commonly deregulated in cancer and AT13148 exhibits promising single-agent antitumor activity the oncogenic signal is transmitted predominantly through after oral administration and support its clinical evaluation. AGC kinases, such as AKT, p70S6K, PDK1, SGK, and ROCK Apart from a monotherapy drug development strategy, (3). We report for the first time the detailed biologic activity AT13148 may also be considered for rational combination of a novel, potent, oral clinical drug candidate AT13148, regimens, especially with other targeted therapies that which is a multi-AGC kinase inhibitor discovered using induce the activation of compensatory pathways, for exam- fragment-based screening combined with structure-based ple AKT phosphorylation after mTORC1 inhibition design. We have showed that AT13148 is a potent inhibitor observed with everolimus (37). In keeping with other AKT of selected AGC kinases including AKT, p70S6K, PKA, SGK, inhibitors currently in the clinic, toxicologic studies with and ROCK (Supplementary Table S2). AT13148 revealed some early hyperglycemia, but the effects In this study, we have used the pharmacologic audit trail were largely equivocal (data not shown). As might be that we originally conceptualized and subsequently advo- expected from its target kinase profile, AT13148 also cated (35, 36) to guide the biomarker-driven drug discovery showed vascular smooth muscle contraction, hypotension, and development of AT13148. Thus, we incorporated and tachycardia (data not shown), but these perturbations detailed pharmacokinetic and pharmacodynamic studies returned to normal after repeat dosing, suggesting an adap- to confirm adequate drug exposure with concomitant target tive response. Gene expression microarray analysis of

www.aacrjournals.org Clin Cancer Res; 18(14) July 15, 2012 3921

Yap et al.

normal tissues might help to identify those genes associated cyclin E and CDC6, and upregulation of negative cell-cycle with drug toxicity. regulators including p27KIP1 in response to treatment with We have shown that robust inhibition of AGC kinase both AT13148 and CCT128930, correlated with a substan- activity occurs in cancer cells both in vitro and in vivo, despite tial decrease in S-phase cells. However, although the observed induction of phosphorylation on Ser473 of CCT128930, like other inhibitors of PI3K/AKT signaling, AKT by AT13148, which binds into the ATP pocket of this gave a predominant G1 arrest, AT13148 did not. This may be kinase. This type of induction has been seen with other ATP- explained by the fact that in contrast to CCT128930, competitive inhibitors of AKT and shown to be because of AT13148 did not cause decreased expression of the major direct inhibitor binding, rather than a regulatory pathway G1 regulator, cyclin D1 at either the gene expression or feedback mechanism (15). Furthermore, it has been protein level (Fig. 5C; Supplementary Fig. S6A and S6B). showed that this phosphorylated form of AKT is hyperactive Consequently, there was an equal distribution of cells either suggesting that in vivo treatment with an ATP competitive side of S phase, expressed as an increase in the G2/M phase inhibitor of AKT may promote tumor cell growth (15). (Supplementary Figs. S6D, S6E, and S7). Conversely, the However, our studies have shown that despite the induction most enriched population of genes showing altered expres- of pSer473 AKT, the removal of AT13148 from cancer cells sion with AT13148 but not CCT128930 are those involved in vitro does not lead to increased phosphorylation of AKT in the control of apoptosis. This molecular phenotype is substrates (Supplementary Fig. S1A and S1B). Furthermore, recapitulated at the cellular level where a much greater our in vivo pharmacodynamic and efficacy studies with induction of apoptosis is observed with AT13148 compared AT13148 (Figs. 3D and 4A–C; Supplementary Figs. S2– with CCT128930 at the same concentrations (Supplemen- S4), using the doses and schedules presented in this paper, tary Fig. S6A–S6C). We hypothesize that these differences indicate that AKT is not hyperactivated and does not pro- are a result of the targeting of several key AGC kinases by mote tumor cell growth but rather, signaling output and AT13148, in contrast to the more AKT-specific effects of tumor growth are inhibited. It is not possible to conclude at CCT128930. this point whether the observed inhibition of AKT signaling In conclusion, we have disclosed here for the first time the output is because of the fact that AT13148 inhibits multiple detailed mechanism of action and therapeutic potential of AGC kinases, or is associated with the pharmacologic prop- the novel, potent, multi-AGC kinase inhibitor, and oral erties of this inhibitor. However, our data provide evidence drug candidate AT13148. We report the preclinical phar- both in vitro and in vivo that the AKT pathway is inhibited macologic audit trail for AT13148 that supports its clinical rather than activated with the AGC kinase inhibitor development, including the pharmacokinetic—pharmaco- AT13148 and that such an inhibitor approach is a viable dynamic–antitumor activity relationship in clinically rele- therapeutic anticancer strategy. vant human tumor xenografts. In addition, our detailed Our gene expression microarray studies in PTEN-deficient gene expression microarray analysis has revealed that U87MG human glioblastoma cells identified an overlap of AT13148 shows a distinct gene expression profile that 147 genes that exhibited significantly altered expression in correlates with a marked apoptotic rather than cytostatic response to both AT13148 and CCT128930 (Fig. 5B). This phenotype, emphasizing the functional differences suggests a component of shared mechanism of action on the between its properties as a multi-AGC kinase inhibitor in gene network that correlated with inhibition of the IGF- contrast to a more AKT-selective inhibitor. In view of the PI3K-AKT-mTOR pathway (Figs. 5C and 6A–C). Of interest, potential mechanistic advantages detailed above, and the increased expression of upstream positive regulators was potent antitumor activity observed at well-tolerated doses observed especially of IRS2, which also seems to be at least against established human tumor xenografts with clinically partly regulated by FOXO1 and FOXO3a. The increased relevant genetic drivers, the clinical use of such an AGC expression of genes encoding IRS2 and PI3K p110a could kinase inhibitor strategy will now be assessed in a first-in- potentially lead to reactivation of the pathway, as has been human phase I trial of AT13148. shown previously for IRS2 with the pan-class I PI3K inhibitor, GDC-0941 (34). However, although we observed Disclosure of Potential Conflicts of Interest increased Ser473 phosphorylation on AKT with both T.A. Yap, M.I. Walton, R.H. te Poele, P.D. Eve, M.R. Valenti, A.K. de Haven Brandon, V. Martins, A. Zetterlund, S.P. Heaton, K. Heinzmann, F.I. Ray- CCT128930 and AT13148 (Fig. 5C), downstream targets naud, S.A. Eccles, P. Workman, and M.D. Garrett are current or former were still dephosphorylated and the pathway remained employees of The Institute of Cancer Research, which has a commercial interest in the development of AKT inhibitors, including AT13148, and inactive. operates a rewards for inventors scheme. K.M. Grimshaw, R. Feltell, M. Reule, Both compounds altered the expression of genes involved S.J. Woodhead, T.G. Davies, J.F. Lyons, and N.T. Thompson are current or in cell-cycle regulation and apoptosis (Figs. 5C and 6A and former employees of Astex Therapeutics, which also has a commercial interest in the development of AKT inhibitors including AT13148. Both B; Supplementary Fig. S6). The enrichment for cell-cycle Astex Therapeutics and The Institute of Cancer Research have been involved genes is very similar to that seen with the dual pan-class I in a commercial collaboration with Cancer Research Technology Limited PI3K/mTOR inhibitor PI-103 (33), suggesting that these (CRT) to discover and develop inhibitors of AKT and intellectual property arising from this program has been licensed to AstraZeneca. P. Workman has gene expression changes are pathway related. However, a commercial research grant from Yamanouchi (now Astellas), Piramed effects were greater for CCT128930 than AT13148, consis- Pharma, and Astex Pharmaceuticals; ownership interest (including patents) from Piramed Pharma (acquired by Roche) and Chroma Therapeutics; and tent with the former being a more AKT-selective compound. is a consultant/advisory board member for Piramed Pharma, Chroma Downregulation of positive cell-cycle regulators, such as Therapeutics, Novartis, Wilex, and Nextech Ventures.

3922 Clin Cancer Res; 18(14) July 15, 2012 Clinical Cancer Research

AT13148, a Potent, Oral AGC Kinase Inhibitor

Grant Support Therapeutics. The authors acknowledge NHS funding to the NIHR Biomed- Grant support was provided to T.A. Yap, M.I. Walton, P.D. Eve, M.R. ical Research Centre. Valenti, A.K. de Haven Brandon, V. Martins, A. Zetterlund, F.I. Raynaud, S.A. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked Eccles, P. Workman, and M.D. Garrett by Cancer Research UK (CR-UK) grant advertisement C309/A8274; to S.P. Heaton by CR-UK grant C51/A6883; to R.H. te Poele by in accordance with 18 U.S.C. Section 1734 solely to indicate CR-UK grant C51/A7401; to K. Heinzmann by Marie Curie Early Stage this fact. Funding; and to P. Workman by CR-UK grant number C309/A8992. P. Workman is a Cancer Research Life Fellow. Additional support was provided to S.A. Eccles and M.D. Garrett by The Institute of Cancer Research. This work Received December 23, 2011; revised May 1, 2012; accepted May 15, 2012; was carried out as part of a funded research collaboration with Astex published OnlineFirst July 10, 2012.

References 1. Engelman JA. Targeting PI3K signalling in cancer: opportunities, 20. Workman P, Aboagye EO, Balkwill F, Balmain A, Bruder G, Chaplin DJ, challenges and limitations. Nat Rev Cancer 2009;9:550–62. et al. Guidelines for the welfare and use of animals in cancer research. 2. Yap TA, Garrett MD, Walton MI, Raynaud F, de Bono JS, Workman P. Br J Cancer 2010;102:1555–77. Targeting the PI3K-AKT-mTOR pathway: progress, pitfalls, and pro- 21. Chen J, Yusuf I, Andersen HM, Fruman DA. FOXO transcription factors mises. Curr Opin Pharmacol 2008;8:393–412. cooperate with delta EF1 to activate growth suppressive genes in B 3. Pearce LR, Komander D, Alessi DR. The nuts and bolts of AGC protein lymphocytes. J Immunol 2006;176:2711–21. kinases. Nat Rev Mol Cell Biol 2010;11:9–22. 22. Martinez-Gac L, Marques M, Garcia Z, Campanero MR, Carrera AC. 4. Manning BD, Cantley LC. AKT/PKB signaling: navigating downstream. Control of cyclin G2 mRNA expression by forkhead transcription Cell 2007;129:1261–74. factors: novel mechanism for cell cycle control by phosphoinositide 5. Courtney KD, Corcoran RB, Engelman JA. The PI3K pathway as drug 3-kinase and forkhead. Mol Cell Biol 2004;24:2181–9. target in human cancer. J Clin Oncol 2010;28:1075–83. 23. Al-Mubarak B, Soriano FX, Hardingham GE. Synaptic NMDAR activity 6. Liu P, Cheng H, Roberts TM, Zhao JJ. Targeting the phosphoinositide suppresses FOXO1 expression via a cis-acting FOXO binding site: 3-kinase pathway in cancer. Nat Rev Drug Discov 2009;8:627–44. FOXO1 is a FOXO target gene. Channels (Austin, Tex) 2009;3:233–8. 7. Chandarlapaty S, Sawai A, Scaltriti M, Rodrik-Outmezguine V, Grbo- 24. Essaghir A, Dif N, Marbehant CY, Coffer PJ, Demoulin JB. The vic-Huezo O, Serra V, et al. AKT inhibition relieves feedback suppres- transcription of FOXO genes is stimulated by FOXO3 and repressed sion of receptor tyrosine kinase expression and activity. Cancer Cell by growth factors. J Biol Chem 2009;284:10334–42. 2011;19:58–71. 25. Kel A, Voss N, Jauregui R, Kel-Margoulis O, Wingender E. Beyond 8. Vasudevan KM, Barbie DA, Davies MA, Rabinovsky R, McNear CJ, Kim microarrays: finding key transcription factors controlling signalF trans- JJ, et al. AKT-independent signaling downstream of oncogenic duction pathways. BMC Bioinformatics 2006;7 Suppl 2:S13. PIK3CA mutations in human cancer. Cancer Cell 2009;16:21–32. 26. Nakagawa Y, Shimano H, Yoshikawa T, Ide T, Tamura M, Furusawa M, 9. Nazarian R, Shi H, Wang Q, Kong X, Koya RC, Lee H, et al. Melanomas et al. TFE3 transcriptionally activates hepatic IRS-2, participates in acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS insulin signaling and ameliorates diabetes. Nat Med 2006;12:107–1. upregulation. Nature 2010;468:973–7. 27. Hui RC, Gomes AR, Constantinidou D, Costa JR, Karadedou CT, 10. Fan QW, Knight ZA, Goldenberg DD, Yu W, Mostov KE, Stokoe D, et al. Fernandez de Mattos S, et al. The forkhead transcription factor A dual PI3 kinase/mTOR inhibitor reveals emergent efficacy in glioma. FOXO3a increases PI3K/Akt activity in drug-resistant leukaemic cells Cancer Cell 2006;9:341–9. through induction of PIK3CA expression. Mol Cell Biol 2008;28: 11. Yang J, Cron P, Good VM, Thompson V, Hemmings BA, Barford D. 5886–98. Crystal structure of an activated Akt/protein kinase B ternary complex 28. Mani KM, Lefebvre C, Wang K, Lim WK, Basso K, Dalla-Favera R, et al. with GSK3-peptide and AMP-PNP. Nat Struct Biol 2002;9:940–4. A systems biology approach to prediction of oncogenes and molecular 12. Donald A, McHardy T, Rowlands MG, Hunter LJ, Davies TG, Berdini V, perturbation targets in B-cell lymphomas. Mol Syst Biol 2008;4:169. et al. Rapid evolution of 6-phenylpurine inhibitors of protein kinase B 29. He X, Zhu Z, Johnson C, Stoops J, Eaker AE, Bowen W, et al. PIK3IP1, a through structure-based design. J Med Chem 2007;50:2289–92. negative regulator of PI3K, suppresses the development of hepato- 13. Saxty G, Woodhead SJ, Berdini V, Davies TG, Verdonk ML, Wyatt PG, cellular carcinoma. Cancer Res 2008;68:5591–8. et al. Identification of inhibitors of protein kinase B using fragment- 30. Zhu Z, He X, Johnson C, Stoops J, Eaker AE, Stoffer DS, et al. PI3K is based lead discovery. J Med Chem 2007;50:2293–6. negatively regulated by PIK3IP1, a novel p110 interacting protein. 14. Davies TG, Verdonk ML, Graham B, Saalau-Bethell S, Hamlett CC, Biochem Biophys Res Commun 2007;358:66–72. McHardy T, et al. A structural comparison of inhibitor binding to PKB, 31. DAVID Bioinformatics Resource 6.7. Available from: http://david.abcc. PKA and PKA-PKB chimera. J Mol Biol 2007;367:882–94. ncifcrf.gov/home.jsp. 15. Okuzumi T, Fiedler D, Zhang C, Gray DC, Aizenstein B, Hoffman R, et al. 32. GeneGo. Available from: http://www.genego.com/. Inhibitor hijacking of Akt activation. Nat Chem Biol 2009;5:484–93. 33. Guillard S, Clarke PA, Te Poele R, Mohri Z, Bjerke L, Valenti M, et al. 16. Woodhead SJ, Rees DC, Frederickson M, Grimshaw KM, inventors; Molecular pharmacology of phosphatidylinositol 3-kinase inhibition in Astex Therapeutics Limited, The Institute of Cancer Research: Royal human glioma. Cell Cycle (Georgetown, TX) 2009;8:443–53. Cancer Hospital, Cancer Research Technology Limited, assignees. 34. O'Brien C, Wallin JJ, Sampath D, GuhaThakurta D, Savage H, Pun- Patent WO20081/110846. noose EA, et al. Predictive biomarkers of sensitivity to the phospha- 17. Yap TA, Walton MI, Hunter LJ, Valenti M, de Haven Brandon A, Eve PD, tidylinositol 30 kinase inhibitor GDC-0941 in breast cancer preclinical et al. Preclinical pharmacology, antitumor activity, and development models. Clin Cancer Res 2010;16:3670–83. of pharmacodynamic markers for the novel, potent AKT inhibitor 35. Workman P. How much gets there and what does it do?: The need CCT128930. Mol Cancer Ther 2011;10:360–71. for better pharmacokinetic and pharmacodynamic endpoints in con- 18. Walton MI, Eve PD, Hayes A, Valenti M, De Haven Brandon A, Box G, temporary drug discovery and development. Curr Pharm Des 2003;9: et al. The preclinical pharmacology and therapeutic activity of the novel 891–902. CHK1 inhibitor SAR-020106. Mol Cancer Ther 2010;9:89–100. 36. Yap TA, Sandhu SK, Workman P, de Bono JS. Envisioning the future of 19. Gowan SM, Hardcastle A, Hallsworth AE, Valenti MR, Hunter LJ, de early anticancer drug development. Nat Rev Cancer 2010;10:514–23. Haven Brandon AK, et al. Application of meso scale technology for the 37. O'Reilly KE, Rojo F, She QB, Solit D, Mills GB, Smith D, et al. mTOR measurement of phosphoproteins in human tumor xenografts. Assay inhibition induces upstream receptor tyrosine kinase signaling and Drug Dev Technol 2007;5:391–401. activates Akt. Cancer Res 2006;66:1500–8.

www.aacrjournals.org Clin Cancer Res; 18(14) July 15, 2012 3923

AT13148 Is a Novel, Oral Multi-AGC Kinase Inhibitor with Potent Pharmacodynamic and Antitumor Activity

Timothy A. Yap, Mike I. Walton, Kyla M. Grimshaw, et al.

Clin Cancer Res 2012;18:3912-3923. Published OnlineFirst July 15, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-11-3313

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2012/07/10/1078-0432.CCR-11-3313.DC1

Cited articles This article cites 34 articles, 11 of which you can access for free at: http://clincancerres.aacrjournals.org/content/18/14/3912.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/18/14/3912.full#related-urls

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/18/14/3912. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.