Structure-Based Design of Specific Inhibitors of Janus Kinase 3 As Apoptosis-Inducing Antileukemic Agents1

Home , HCK, Janus kinase, Janus kinase 3, TEC (gene)

Vol. 5, 1569–1582, June 1999 Clinical Cancer Research 1569

Structure-based Design of Specific Inhibitors of Janus Kinase 3 as Apoptosis-inducing Antileukemic Agents1

Elise A. Sudbeck, Xing-Ping Liu, predicted to bind less strongly, with estimated Kis ranging from Rama Krishna Narla, Sandeep Mahajan, 28 to 72 ␮M. These compounds did not show any significant Sutapa Ghosh, Chen Mao, and Fatih M. Uckun2 JAK3 inhibition in kinase assays. Furthermore, the lead dimethoxyquinazoline compound, WHI-P131, which showed potent Parker Hughes Cancer Center [R. K. N., F. M. U.], Departments of JAK3-inhibitory activity (IC of 78 ␮M), did not inhibit JAK1 Structural Biology [E. A. S., S. G., C. M.], Experimental Oncology 50 [R. K. N., F. M. U.], Biochemistry [S. M.], and Chemistry [X-P. L.], and JAK2, the ZAP/SYK family tyrosine kinase SYK, the TEC and Drug Discovery Program [E. A. S., X-P. L., S. G., C. M., family tyrosine kinase BTK, the SRC family tyrosine kinase F. M. U.], Hughes Institute, St. Paul, Minnesota 55113 LYN, or the receptor family tyrosine kinase insulin receptor kinase, even at concentrations as high as 350 ␮M. WHI-P131 induced apoptosis in JAK3-expressing human leukemia cell ABSTRACT lines NALM-6 and LC1;19 but not in melanoma (M24-MET) A novel homology model of the kinase domain of Janus or squamous carcinoma (SQ20B) cells. Leukemia cells were kinase (JAK) 3 was used for the structure-based design of not killed by dimethoxyquinazoline compounds that were in- dimethoxyquinazoline compounds with potent and specific active against JAK3. WHI-P131 inhibited the clonogenic inhibitory activity against JAK3. The active site of JAK3 in growth of JAK3-positive leukemia cell lines DAUDI, RAMOS, ؋ ؋ this homology model measures roughly 8 Å 11 Å 20 Å, LC1;19, NALM-6, MOLT-3, and HL-60 (but not JAK3-nega- ϳ 3 with a volume of 530 Å available for inhibitor binding. tive BT-20 breast cancer, M24-MET melanoma, or SQ20B Modeling studies indicated that 4-(phenyl)-amino-6,7-dime- squamous carcinoma cell lines) in a concentration-dependent thoxyquinazoline (parent compound WHI-258) would likely fashion. Potent and specific inhibitors of JAK3 such as WHI- fit into the catalytic site of JAK3 and that derivatives of this P131 may provide the basis for the design of new treatment ؅ compound that contain an OH group at the 4 position of the strategies against acute lymphoblastic leukemia, the most com- phenyl ring would more strongly bind to JAK3 because of mon form of childhood cancer. added interactions with Asp-967, a key residue in the catalytic site of JAK3. These predictions were consistent with INTRODUCTION docking studies indicating that compounds containing a 3 4؅-OH group, WHI-P131 [4-(4؅-hydroxyphenyl)-amino- STATs constitute a family of DNA-binding proteins that 6,7-dimethoxyquinazoline], WHI-P154 [4-(3؅-bromo-4؅-hy- reside in the cytoplasm until they are activated by tyrosine droxylphenyl)-amino-6,7-dimethoxyquinazoline], and WHI- phosphorylation. This phosphorylation event is catalyzed by P97 [4-(3؅,5؅-dibromo-4؅-hydroxylphenyl)-amino-6,7-dime- members of the Janus family of tyrosine kinases, including thoxyquinazoline], were likely to bind favorably to JAK3, JAK3 (1, 2). The dual role of STATs as signaling molecules and transcription factors is reflected in their structure. All STAT with estimated K s ranging from 0.6 to 2.3 ␮M. These com- i proteins contain a DNA-binding domain, an SH2 domain, and a pounds inhibited JAK3 in immune complex kinase assays in transactivation domain necessary for transcriptional induction. a dose-dependent fashion. In contrast, compounds lacking In unstimulated cells, latent forms of STATs are predominantly the 4؅-OH group, WHI-P79 [4-(3؅-bromophenyl)-amino-6,7- localized in the cytoplasm. Ligand binding induces STAT pro- -dimethoxyquinazoline], WHI-P111 [4-(3؅-bromo-4؅-methyl teins to bind with their SH2 domains to the tyrosine-phosphor- -phenyl)-amino-6,7-dimethoxyquinazoline], WHI-P112 [4-(2؅,5؅ ylated motifs in the intracellular domains of various transmem- dibromophenyl)-amino-6,7-dimethoxyquinazoline], WHI-P132 brane cell surface receptors (3, 4). Once STATs are bound to 2؅-hydroxylphenyl)-amino-6,7-dimethoxyquinazoline], and)4-] receptors, the receptor-associated JAKs phosphorylate STATs WHI-P258 [4-(phenyl)-amino-6,7-dimethoxyquinazoline], were on a single tyrosine residue located near the SH2 domain. Two STATs then dimerize through specific reciprocal SH2-phospho- tyrosine interactions. The dimerized STAT proteins can also

Received 1/5/99; revised 3/2/99; accepted 3/2/99. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to 3 The abbreviations used are: STAT, signal transducers and activators of indicate this fact. transcription; JAK, Janus kinase; ALL, acute lymphoblastic leukemia; 1 Supported in part by a special grant from the Parker Hughes Trust. IRK, insulin receptor kinase; FGFR, fibroblast growth factor receptor; F. M. U. is a Stohlman Scholar of the Leukemia Society of America. This ZAP, ␨ associated protein; SYK, spleen tyrosine kinase; TEC, tyrosine work was presented in part at the 1998 Annual Meeting of the American kinase expressed in hepatocellular carcinoma; BTK, Bruton’s tyrosine Society of Hematology held December 4–8, 1998, in Miami Beach, FL, kinase; SRC, cellular homologue of oncogene product from Rous avian and published in abstract form in: Blood, 92 (Suppl. I): 599a, 1998. sarcoma virus; LYN, PTK related to LCK and YES; PTK, protein 2 To whom requests for reprints should be addressed, at Hughes Insti- tyrosine kinase; EMSA, electrophoretic mobility-shift assay; NAO, 10- tute, 2665 Long Lake Road, Suite 330, St. Paul, MN 55113. Phone: n-nonyl-acridine orange; B-ALL, B-cell ALL; IL, interleukin; TdT, (651) 697-9228; Fax: (651) 697-1042. terminal dideoxynucleotidyl transferase.

Fig. 1 A, model of JAK3 showing molecular surface of protein (blue) and catalytic (ATP-binding) site (yellow). B, ribbon representation (C␣ backbone) of the homology model of the JAK3 kinase domain. The WHI-P131 molecule is shown as a space-filling model in the catalytic site of JAK3. C, close-up view of catalytic site of JAK3 model with docked quinazoline inhibitor WHI-P131 (green). Residues and inhibitor are shown as space-filling atoms. The solvent-exposed opening of the catalytic site has dimensions to allow a relatively planar inhibitor to enter and bind to JAK3. The opening of the pocketis defined by residues Pro-906, Ser-907, Gly-908, Asp-912, Arg-953, Gly-829, Leu-828, and Tyr-904 (blue residues). The far wall deep inside the pocket is lined with Leu-905 (C␣ backbone), Glu-903, Met-902, Lys-905, and Asp-967 (pink residues), and the floor of the pocket is lined by Leu-905 (side chain), Val-884, Leu-956, and Ala-966 (yellow residues). Residues defining the roof of the pocket include Leu-828, Gly-829, Lys-830, and Gly-831 (topmost blue residues). This figure was prepared using the InsightII program (InsightII, Molecular Simulations Inc. San Diego, CA, 1996).

form complexes with other DNA-binding proteins. The STAT dimethoxyquinazoline] inhibited JAK3 but not JAK1 or JAK2. dimers/complexes next translocate to the nucleus and use their Similarly, the ZAP/SYK family tyrosine kinase SYK, the TEC DNA-binding domain to interact with DNA response elements family tyrosine kinase BTK, the SRC family tyrosine kinase in promoters of target genes (5). STATs then interact directly or LYN, and the receptor family tyrosine kinase IRK were not indirectly, via their transactivation domain, with components of inhibited by WHI-P131. WHI-P131 induced apoptosis in JAK3- the RNA polymerase II complex to activate transcription of expressing human leukemia cell lines but not in melanoma or target genes. Different ligands use specific JAK and STAT squamous carcinoma cells. Leukemia cells were not killed by family members; thus, utilization of this pathway mandates dimethoxyquinazoline compounds that were inactive against specificity in signaling cascades and contributes to a diverse JAK3. Potent and specific inhibitors of JAK3, such as the array of cellular responses (3, 4). JAKs, including JAK3, are dimethoxyquinazoline compound WHI-P131, may provide the abundantly expressed in primary leukemic cells from children basis for the design of new treatment strategies against ALL. with ALL, the most common form of childhood cancer, and recent studies have correlated STAT activation in ALL cells with signals regulating apoptosis (5–8). MATERIALS AND METHODS Here, we used a novel homology model of the kinase Constructing a Homology Model for the JAK3 Kinase domain of JAK3 to design compounds with potent and specific Domain. Because the three-dimensional coordinates of the JAK3-inhibitory activity as potential antileukemic agents. The JAK3 kinase domain are currently unknown, a structural model lead compound WHI-P131 [4-(4Ј-hydroxyphenyl)-amino-6,7- of JAK3 was required for a docking analysis of JAK3 inhibitors.

A homology model of JAK3 was constructed (Fig. 1) by using receptor (15) and BTK (16) was described previously ]. The known coordinates of homologous kinase domains as refer- various docked positions of each compound were evaluated ences. The JAK3 homology model was built by first obtaining using a Ludi (17) scoring procedure in InsightII (InsightII, the protein sequence of JAK3 (Swiss-Prot database access no. Molecular Simulations Inc. San Diego, CA, 1996), which esti-

P52333; University of Geneva, Geneva, Switzerland) from mated a binding constant, Ki, taking into account the predicted GenBank (National Center for Biotechnology Information, lipophilic, hydrogen bonding, and van der Waals interactions Bethesda, MD) and determining the most reasonable sequence between the inhibitor and the protein. A comparison of the alignment for the JAK3 kinase domain relative to some template catalytic site residues of several different PTKs was made by coordinates [known kinase structures such as HCK (9), FGFR manually superimposing crystal structure coordinates of the (10, 11), and IRK (12)]. This was accomplished by first super- kinase domains of IRK (12) and HCK (9) and models of JAK1, imposing the C␣ coordinates of the kinase domains of HCK, JAK2, JAK3, BTK (16), and SYK4 and then identifying features FGFR, and IRK using the InsightII program (InsightII, Molec- in the active site that were unique to JAK3 (Figs. 2C and 3). ular Simulations Inc. San Diego, CA, 1996) to provide the best Chemical Synthesis of Quinazoline Derivatives. The overall structural comparison. The sequences were then aligned compounds listed in Table 1 were synthesized and characterized based on the superimposition of their structures (amino acid using procedures published previously (18). sequences were aligned together if their C␣ positions were Immune Complex Kinase Assays. Sf21 (IPLB-SF21- spatially related to each other). The alignment accommodated AE) cells (19), derived from the ovarian tissue of the fall features, such as loops in a protein, that differed from the other armyworm Spodotera frugiperda, were obtained from In- protein sequences. The structural superimposition was per- vitrogen (Carlsbad, CA) and maintained at 26–28°C in formed using the Homology module of the InsightII program Grace’s insect cell medium supplemented with 10% fetal and a Silicon Graphics INDIGO2 computer (Silicon Graphics, bovine serum and 1.0% antibiotic/antimycotic (Life Technol- Mountain View, CA). The sequence alignment was performed ogies, Inc.). Stock cells were maintained in suspension at manually and produced a sequence variation profile for each 0.2 ϫ 106-1.6 ϫ 106 cells/ml in a total culture volume of 600 superimposed C␣ position. The sequence variation profile ml in 1-liter Bellco spinner flasks at 60–90 rpm. Cell viabil- served as a basis for the subsequent sequence alignment of the ity was maintained at 95–100%, as determined by trypan blue JAK3 kinase with the other three proteins. In this procedure, the dye exclusion. Sf21 cells were infected with a baculovirus sequence of JAK3 was entered and aligned with the three known expression vector for BTK, SYK, JAK1, JAK2, or JAK3, as kinase proteins based on the sequence variation profiles de- reported previously (16). Cells were harvested and lysed [10 scribed previously. Next, a set of three-dimensional coordinates mM Tris (pH 7.6), 100 mM NaCl, 1% NP40, 10% glycerol, 50 ␮ ␮ was assigned to the JAK3 kinase sequence using the three- mM NaF, 100 M Na3VO4,50 g/ml phenylmethylsulfonyl dimensional coordinates of HCK as a template and the Homol- fluoride, 10 ␮g/ml aprotonin, and 10 ␮g/ml leupeptin], the ogy module within the InsightII program (InsightII, Molecular kinases were immunoprecipitated from the lysates, and their Simulations Inc. San Diego, CA, 1996). The coordinates for a enzymatic activity was assayed, as reported previously (19– loop region where a sequence insertion occurs (relative to HCK 23). The immunoprecipitates were subjected to Western blot without the loop) were chosen from a limited number of possi- analysis, as described previously (19, 20). bilities automatically generated by the computer program and For IRK assays, HepG2 human hepatoma cells grown to manually adjusted to a more ideal geometry using the program ϳ80% confluency were washed once with serum-free DMEM and

CHAIN (13). Finally, the constructed model of the JAK3 kinase starved for3hat37°C in a CO2 incubator. Subsequently, cells domain was subjected to energy minimization using the X- were stimulated with insulin (Eli Lilly and Co., Indianapolis, IN; 10 PLOR program (14) so that any steric strain introduced during units/ml, 10 ϫ 106 cells) for 10 min at room temperature. Follow- the model-building process could be relieved. The model was ing this IRK activation step, cells were washed once with serum- screened for unfavorable steric contacts and, if necessary, such free medium and lysed in NP40 buffer, and IRK was immunopre- side chains were remodeled either by using a rotamer library cipitated from the lysates with an anti-IR␤ antibody (Santa Cruz database or by manually rotating the respective side chains. The Biotechnology, Santa Cruz, CA; polyclonal IgG). Prior to perform- procedure for homology model construction was repeated for ing the immune complex kinase assays, we equilibrated the beads JAK1 (Swiss-Prot database access no. P23458) and JAK2 (Gen- with the kinase buffer [30 mM HEPES (pH 7.4), 30 mM NaCl, 8 mM

Bank database access no. AF005216) using the JAK3 model as MgCl2,and4mM MnCl2]. LYN was immunoprecipitated from a structural template. The energy-minimized homology models whole cell lysates of NALM-6 human leukemia cells, as reported of JAK1, JAK2, and JAK3 were then used, in conjunction with previously (23, 24). energy-minimized structural models of dimethoxyquinazoline In JAK3 immune complex kinase assays (16, 21), KL-2 compounds, for modeling studies of JAK-dimethoxyquinazoline EBV-transformed human lymphoblastoid B cells (native JAK3 complexes. kinase assays) or insect ovary cells (recombinant JAK3 kinase Docking Procedure Using Homology Model of JAK3 assays) were lysed with NP40 lysis buffer [50 mM Tris (pH 8), Kinase Domain. Modeling of the JAK3-dimethoxyquinazo- 150 mM NaCl, 5 mM EDTA, 1% NP40, 100 ␮M sodium or- line complexes was accomplished using the Docking module thovanadate, 100 ␮M sodium molybdate, 8 ␮g/ml aprotinin, 5 within the program InsightII (InsightII, Molecular Simulations Inc. San Diego, CA, 1996) and using the Affinity suite of programs for automatically docking an inhibitor into a protein binding site [a similar procedure for epidermal growth factor 4 C. Mao, unpublished data.

␮g/ml leupeptin, and 500 ␮M phenylmethylsulfonyl fluoride] and centrifuged 10 min at 13,000 ϫ g to remove insoluble material. Samples were immunoprecipitated with antisera prepared against JAK3. The antisera were diluted and immune complexes collected by incubation with 15 ␮l of protein A- Sepharose. After four washes with NP40 lysis buffer, the protein A-Sepharose beads were washed once in kinase buffer [20 mM

MOPS (pH 7)-10 mM MgCl2] and resuspended in the same buffer. Reactions were initiated by the addition of 25 ␮Ci of [␥-32P]ATP (5000 Ci/mmol) and unlabeled ATP to a final concentration of 5 ␮M. Reactions were terminated by boiling for 4 min in SDS sample buffer. Samples were run on 9.5% SDS polyacrylamide gels, and labeled proteins were detected by autoradiography. Following electrophoresis, kinase gels were dried onto Whatman 3M filter paper and subjected to phospho- rimaging on a Molecular Imager (Bio-Rad, Hercules, CA) as well as autoradiography on film. For each drug concentration, a kinase activity index was determined by comparing the kinase activity in phosphorimager units to that of the baseline sample. In some experiments, cold kinase assays were performed, as described previously (25). EMSAs. EMSAs were performed to examine the effects of dimethoxyquinazoline compounds on cytokine-induced STAT activation in 32Dc11/IL2R␤ cells (a gift from Dr. James Ihle, St. Jude Children’s Research Hospital), as described previously (21). Mitochondrial Membrane Potential Assessment. To measure the changes in mitochondria, we incubated cells with WHI-P131 at concentrations ranging from 7.4 ␮g/ml (25 ␮M)to30 ␮g/ml (200 ␮M) for 24 or 48 h; the cells were then stained with specific fluorescent dyes and analyzed with flow cytometer. Mito- chondrial membrane potential (⌬⌿m) was measured using two dyes including a lipophilic cation 5,5Ј,6,6Ј-tetrachloro-1,1Ј,3,3Ј- tetraethlybenzimidazolylcarbocyanine iodide (JC-1) and a cyanine Ј Ј Ј dye, 1,1 ,3,3,3 ,3 -hexamethylindodicarbocyanine iodide (DiIC1; Refs. 26–28) obtained from Molecular Probes (Eugene, OR). JC-1 is a monomer at 527 nm after being excited at 490 nm; with polarization of ⌬⌿m, J-aggregates are formed that shift emission to 590 nm (29). This can be detected on a flow cytometer by assessing the green signal (at 527 nm) and green-orange signal (at 590 nm) simultaneously, creating an index of the number of cells polarized

and depolarized mitochondria. DiIC1, a cyanine dye that is am- phipathic and cationic, concentrates in energized mitochondria and has been used in a variety of studies to measure the mitochondrial

membrane potential (26–28). Cells were also stained with DiICI1 at 40 nM concentration for 30 min in the dark as described for JC-1.

molecules (royal blue) or represent regions just large enough to occupy solvent molecules (pink, yellow, and light blue). A model of WHI-P131 Fig. 2 A, model of unoccupied space in the catalytic (ATP binding) docked into the catalytic site is shown in white, superimposed on the site of a JAK3 homology model. Shown in green is the binding site for green region. B, model of the catalytic site of JAK3 with quinazolines ATP and the most likely binding site for dimethoxyquinazoline inhibi- WHI-P131 (multicolor), WHI-P132 (pink), and WHI-P154 (yellow). tors. The green kinase active site region represents a total volume of Each compound fits into the binding site but WHI-P132 (shown to be ϳ530 Å. Modeling studies showed that an inhibitor or a portion of an inactive against JAK3 in biological assays) lacks a OH group that is in inhibitor with significant binding to this region would occupy a volume a location to bind with Asp-967. WHI-P131 and WHI-P154, with OH Ͻ530 Å3 and have molecular dimensions compatible with the shape of groups at the C4Ј position of the phenyl ring, are able to form a the binding site region. Other regions near the binding site that show favorable interaction with Asp-967 of JAK3, which may contribute to measurable unoccupied volume are shown in royal blue, pink, yellow, their enhanced inhibition activity. C, features of dimethoxyquinazoline and light blue. These binding regions are either unavailable to inhibitor derivatives that are predicted to aid binding to JAK3 catalytic site.

Fig. 3 A, structural comparison of nonconserved residues in the catalytic sites of five different PTKs: JAK3 (pink), BTK (red), SYK (light blue), IRK (dark blue), and HCK (yellow). Residues within 5 Å of the docked JAK3 inhibitor, WHI-P131 (white), are shown as rod-shaped side chains. The C␣ backbone of JAK3 is shown as a thin pink line, for perspective. Regions A–F cor- respond to areas containing nonconserved residues in the catalytic site (see B and “Re- sults and Discussion”). Crys- tal structure coordinates of HCK (9) and IRK (36) and homology models of JAK3, BTK (16), and SYK were used for the structural analysis. B, nonconserved residues in the catalytic sites of eight different PTKs. Regions A–F refer to locations in the catalytic site, which are illustrated in A.

Table 1 Predicted interaction of protonated quinazolines with JAK3 kinase active site and measured inhibition values (IC50) from JAK3 kinase assays

No. of H bond Lipophilic Contact Total binding Estimated Molecular surface Molecular IC50 Ј Ј Ј Ј ␮ 2 3 ␮ Compound R5 R4 R3 R2 H bonds score score score score Ki ( M) area (Å ) volume (Å ) ( M) WHI-P131 H OH H H 3 188 476 64 568 2.3 276 261 9.1 WHI-P97 Br OH Br H 3 156 559 65 622 0.6 314 307 11.0 WHI-P154 H OH Br H 3 171 512 64 587 1.4 296 284 27.9 WHI-P79 H H Br H 1 9 531 63 444 36 278 272 Ͼ300 WHI-P132 H H H OH 1 82 476 66 462 25 269 265 Ͼ300 Ͼ WHI-P111 H CH3 Br H 1 52 502 58 458 46 309 291 300 WHI-P112 Br H H Br 0 0 541 62 445 52 306 297 Ͼ300 WHI-P258 H H H H 0 0 510 64 414 72 266 252 Ͼ300

The cells were analyzed using a Vantage Becton Dickinson (San the mitochondrial phospholipid cardiolipin, which, in turn, has Jose, CA) cell sorter equipped with HeNe laser with excitation at been extensively used to provide an index of mitochondrial 635 nm, and the fluorescence was measured at 666 nm. mass (30). Mitochondrial Mass Determination. Relative mito- Human Leukemia and Cancer Cell Lines. The follow- chondrial mass was measured by using Becton Dickinson Cali- ing cell lines were used in various biological assays: NALM-6 bur flow cytometry and the fluorescent stain NAO, which binds (pre-B-ALL), LC1;19 (pre-B-ALL), DAUDI (B-ALL),

RAMOS (B-ALL), MOLT-3 (T-cell ALL), HL60 (acute my- bind to the catalytic site, which results in an attenuation of PTK elogenous leukemia), BT-20 (breast cancer), M24-MET (mela- activity by inhibiting ATP binding. An analysis of the JAK3 noma), SQ20B (squamous cell carcinoma), and PC3 (prostate model revealed specific features of the catalytic site, which can cancer). These cell lines were maintained in culture, as reported be described as a quadrilateral-shaped pocket (Fig. 1C). The previously (15, 16, 19, 23, 31, 32). Cells were seeded in six-well opening of the pocket is defined by residues Pro-906, Ser-907, tissue culture plates at a density of 50 ϫ 104 cells/well in a Gly-908, Asp-912, Arg-953, Gly-829, Leu-828, and Tyr-904 treatment medium containing various concentrations of WHI- (Fig. 1C, blue residues). The far wall deep inside the pocket is P131 and incubated for 24–48 h at 37°C in a humidified 5% lined with Leu-905 (C␣ backbone), Glu-903, Met-902, Lys-905,

CO2 atmosphere. and Asp-967 (Fig. 1C, pink residues), and the floor of the pocket Apoptosis Assays. Cells were examined for apoptotic is lined by Leu-905 (side chain), Val-884, Leu-956, and Ala-966 changes after treatment with WHI-P131 by the in situ TdT- (Fig. 1C, yellow residues). Residues defining the roof of the mediated dUTP end-labeling assay using the ApopTag apopto- pocket include Leu-828, Gly-829, Lys-830, and Gly-831 (Fig. sis detection kit (Oncor, Gaithersburg, MD) according to the 1C, topmost blue residues). Figs. 1C and 2A illustrate that the manufacturer’s recommendations, as detailed in our earlier re- catalytic site of the JAK3 model has approximate dimensions of ports (32, 33). 8Åϫ 11 Å ϫ 20 Å and an available volume for binding of To detect apoptotic fragmentation of DNA, we harvested ϳ530 Å3. According to the model, the solvent exposed opening cells after a 24-h exposure at 37°C to WHI-P131 or other to the binding region would allow inhibitors to enter and bind if dimethoxyquinazoline compounds at 1, 3, and/or 10 ␮M con- the molecule contained some planarity. centrations. DNA was prepared from Triton X-100 lysates for Although most of the catalytic site residues of the JAK3 analysis of fragmentation (20). In brief, cells were lysed in kinase domain were conserved relative to other PTKs, a few hypotonic 10 mmol/liter Tris-HCl (pH 7.4), 1 mmol/liter EDTA, specific variations were observed (Fig. 3). These differences and 0.2% Triton X-100 detergent and subsequently centrifuged include an alanine residue in BTK, IRK, and HCK/LYN (Fig. at 11,000 ϫ g. To detect apoptosis-associated DNA fragmenta- 3A, region A) that changes to Glu in SYK and Pro-906 in JAK3. tion, we electrophoresed supernatants on a 1.2% agarose gel, At region B, a tyrosine residue is conserved in JAK3 (Tyr-904), and the DNA fragments were visualized by UV light after BTK, and LYN but changes to Phe in HCK (which is the only staining with ethidium bromide. apparent residue difference between HCK and LYN relevant to Clonogenic Assays. The antileukemic activity of WHI- inhibitor binding), Met in SYK, and Leu in IRK. Region C P131 against clonogenic tumor cells was examined using a shows a methionine residue that is conserved in BTK, IRK, and methylcellulose colony assay system (34, 35). In brief, cells (105 HCK/LYN but changes to Leu-905 in JAK3 and Ala in SYK. cells/ml in RPMI-10% fetal bovine serum) were treated over- Region D shows Met-902 in JAK3, which is conserved in SYK night at 37°C with WHI-P131 at varying concentrations. After and IRK but changes to Thr in BTK and to a much smaller treatment, cells were washed twice, plated at 104 or 105 cells/ml residue, Ala, in LYN and HCK. This Met-902 residue in JAK3, in RPMI, 10% fetal bovine serum, and 0.9% methylcellulose in which is located on the back wall of the pocket and protrudes in Petri dishes and cultured for 7 days at 37°C in a humidified 5% toward the center of the pocket volume, can significantly affect

CO2 incubator. Subsequently, leukemic cell (or tumor cell) the shape of the binding pocket. At this location, the extended colonies were enumerated using an inverted phase-contrast mi- conformation of the Met-902 side chain can hinder the close croscope. The percentage inhibition of colony formation was contact of inhibitors with residues lining the back wall of the calculated using the following formula: pocket and with the hinge region, relative to other kinases with smaller residues here such as BTK (Thr) and HCK/LYN (Ala). % inhibition Ala-966 in region E is conserved in HCK/LYN but changes to Gly in IRK and to the more hydrophilic residue Ser in BTK and Mean no. of colonies in test culture ϭ 1 Ϫ ϫ 100 SYK. Region F, which is farther away from the inhibitor loca- Mean no. of colonies in control culture tion, is the least conserved region of the catalytic site and contains Asp-912 in JAK3, Asn in BTK, Lys in SYK, Ser in RESULTS AND DISCUSSION IRK, and Asp in HCK/LYN (Fig. 3). These residue identity Homology Model of JAK3 Kinase Domain. The three- differences between tyrosine kinases provide the basis for de- dimensional coordinates of JAK3 used in the protein/inhibitor signing selective inhibitors of the JAK3 kinase domain. modeling studies were constructed based on a structural align- Structure-based Design and Synthesis of JAK3 Inhibi- ment with the sequences of known crystal structures of the tors. A computer docking procedure was used to predict how kinase domains of three PTKs: HCK (9), FGFR (11, 16), and well potential inhibitors could fit into and bind to the catalytic IRK (12, 36), as detailed in “Materials and Methods.” Fig. 1, A site of JAK3 and result in kinase inhibition (Fig. 2B). The and B, shows the homology model of the JAK3 kinase domain, dimethoxyquinazoline compound WHI-P258 [4-(phenyl)-ami-

which is composed of an NH2-terminal lobe and a COOH- no-6,7-dimethoxyquinazoline] contains two methoxy groups on terminal lobe, which are linked by a hinge region near the the quinazoline moiety but no other ring substituents. Molecular catalytic (ATP-binding) site. The catalytic site is a pocket lo- modeling studies using the homology model of JAK3 kinase cated in the central region of the kinase domain, which is domain suggested that WHI-P258 would fit into the catalytic defined by two ␤-sheets at the interface between the N and C site of JAK3 but probably would not bind very tightly due to lobes. The opening to the catalytic site is solvent accessible and limited hydrogen-bonding interactions. Asp-967, a key residue facilitates binding of ATP. Small molecule inhibitors can also in the catalytic site of JAK3, can form a hydrogen bond with

Fig. 5 Specificity of WHI-P131. JAK3, SYK, and BTK immunoprecipitated from Sf21 insect ovary cells transfected with the appropriate Fig. 4 Effects of WHI-P131 on the tyrosine kinase activity of JAK3. A–D, baculovirus expression vectors, LYN immunoprecipitated from JAK3, JAK1, and JAK2 immunoprecipitated from Sf21 insect ovary cells NALM-6 human B-lineage ALL cells, and IRK immunoprecipitated transfected with the appropriate baculovirus expression vectors were treated from HepG2 hepatoma cells were treated with WHI-P131 and then with WHI-P131, then subjected to in vitro kinase assays, as described in subjected to in vitro kinase assays, as described in “Materials and “Materials and Methods.” The enzymatic activity of JAKs was determined Methods.” by measuring autophosphorylation in a 10-min kinase assay, as described in “Materials and Methods.” The kinase activity (KA) levels were expressed as percentage of baseline activity (%CON). E, EMSAs of 32Dc22-IL-2R␤ cells. WHI-P131 (10 ␮g/ml ϭ 33.6 ␮M) and WHI-P154 (10 ␮g/ml ϭ 26.6 pounds is provided in Table 1. A summary of structural ␮M) but not WHI-P132 (10 ␮g/ml ϭ 33.6 ␮M) inhibited IL-2-triggered JAK-3-dependent STAT activation but not IL-3-triggered JAK-1-/JAK-2- features of the designed dimethoxyquinazoline compounds dependent STAT activation in 32Dc11-IL-2R␤ cells. that were observed to be relevant for binding to the catalytic site of JAK3 is shown in Fig. 2C. The approximate molecular volumes of the compounds in Table 1 range from 252 to 307 Å, which are small enough to fit into the 530-Å3 binding site molecules binding to the catalytic site, if such molecules contain of JAK3 kinase. Table 1 also lists the results of molecular a hydrogen bond donor group such as a OH group. WHI-P258, modeling studies, including estimated binding constants (i.e., however, does not contain a OH group and, therefore, would not Kis) for the compounds that were docked into the JAK3 interact as favorably with Asp-967. We postulated that the catalytic site. The compounds that were evaluated in docking presence of a OH group at the 4Ј position of the phenyl ring of studies contain substitutions of similar functional groups at WHI-P258 would result in stronger binding to JAK3 because of different positions on the phenyl ring. added interactions with Asp-967. A series of dimethoxyquina- The conformations of the energy-minimized docked mod- zoline compounds were designed and synthesized to test this els of the compounds listed in Table 1 were relatively planar, hypothesis. with dihedral angles of ϳ4–18° between the phenyl ring and An estimation of the molecular volume for the com- quinazoline ring system. This conformation allows the molecule

Fig. 6 WHI-P131 depolarizes mitochondrial membranes in a concentration-dependent fashion without affecting the mitochondrial mass. NALM-6

human leukemic cells were incubated with indicated concentrations of WHI-P131 for 48 h, stained with DiIC1 to assess the mitochondrial membrane potential (⌬␺m) or NAO to detect the mitochondrial mass, and then analyzed with cell sorter equipped with HeNe laser. WHI-P131 caused a progressive increase in depolarized mitochondria (as indicated by M1 in A) with increasing concentrations. At similar concentrations, no significant change in mitochondrial mass (B) was detected. C, cells were stained with JC-1 for simultaneous analysis of mitochondrial mass (green fluorescence)

Fig. 7 WHI-P131 induces apoptosis in NALM-6 leukemia cells. A, cells were incubated with 10 ␮M-500 ␮M WHI-P131 for 24 h and then processed for in situ apoptosis assays. The percentage of apoptotic cells was determined by examining an average of 1380 cells per 10 fields per sample. Data points, means from counts obtained in two independent experiments, each performed in duplicate; vertical bars, SD. B.1 and B.2, NALM-6 cells were incubated with 100 ␮M of WHI-P131 for 48 h, processed for the in situ apoptosis assay, and analyzed with a laser scanning confocal microscope. When compared with controls treated with DMSO (0.1%; B.1), several of the cells incubated with WHI- P131 (B.2) showed apoptotic nuclei (yellow fluorescence). Red fluorescence represents nuclei stained with propidium iodide.

to fit more easily into the catalytic site of JAK3. All of the listed hydrogen bond with the quinazoline ring nitrogen, which may compounds contain a ring nitrogen (N1), which can form a contribute to a significantly lower affinity of WHI-P132 for the hydrogen bond with NH of Leu-905 in the hinge region of catalytic site of JAK3. The relatively large bromine substituents JAK3. When N1 is protonated, the NH can instead interact with (WHI-P97 and WHI-P154) can increase the molecular surface the carbonyl group in Leu-905 of JAK3. The presence of a OH area in contact with binding site residues if the molecule can fit group at the 4Ј position on the phenyl ring was anticipated to be into the binding site. Modeling of WHI-P154 and WHI-P97 particularly important for binding to the catalytic site of JAK3. showed that there is enough room to accommodate the bromine ϭ ␮ ϭ WHI-P131 (estimated Ki 2.3 M), WHI-P154 (estimated Ki groups if the phenyl ring is tilted slightly relative to the fused ␮ ϭ ␮ 1.4 M), and WHI-P97 (estimated Ki 0.6 M) shown in Table ring group of the molecule. The results from the modeling 1 were predicted to have favorable binding to JAK3 and potent studies prompted the hypothesis that WHI-P131, WHI-P154, JAK3 inhibitory activity because they contain a 4Ј-OH group on and WHI-P97 would exhibit potent JAK3-inhibitory activity. To the phenyl ring that can form a hydrogen bond with Asp-967 of test this hypothesis and validate the predictive value of the JAK3, contributing to enhanced binding. However, the 2Ј-OH described JAK3 homology model, we synthesized WHI-P131, group of WHI-P132 is not in the right orientation to interact WHI-P154, WHI-P97, and five other dimethoxyquinazoline with Asp-967, and it would probably form an intramolecular compounds, listed in Table 1.

and mitochondrial transmembrane potential (red/orange fluorescence). Untreated NALM-6 cells (D.1) as well as NALM-6 cells treated with 50 ␮M of WHI-P131 for 24 h (D.2) were incubated with JC-1 and analyzed by confocal laser scanning microscopy. Mitochondria of control cells showed a higher membrane potential (⌬␺m), as indicated by brighter JC-1 red fluorescence. Treatment of cells with WHI-P131 reduced mitochondrial membrane ⌬␺m as indicated by a substantial decrease in JC-1 red fluorescence.

Fig. 8 WHI-P131 induces apoptotic DNA fragmentation in human leukemia cells. DNA from Triton X-100 lysates of control and drug-treated cells was analyzed for fragmentation, as described previously (20). A, NALM-6 human B-lineage ALL cells were treated for 24 h at 37°C with the listed dimethoxyquinazoline compounds at 1, 3, or 10 ␮M final concentrations. WHI-P131 was used as the lead JAK3 inhibitory compound, and all other compounds were included as controls that lacked JAK3-inhibitory activity. B, LC1;19 human B-lineage ALL cells and the control cell lines SQ20B and M24-MET were treated for 24 h at 37°C with WHI-P131 at 1 or 3 ␮M final concentrations.

Inhibition of JAK3 by Rationally Designed Dime- icant JAK3-inhibitory activity at micromolar concentrations thoxyquinazoline Compounds. We first used immune com- (which was not the case for the other compounds, which had ␮ plex kinase assays to compare the effects of the synthesized estimated Kis ranging from 25 to 72 M), inhibited JAK3 in dimethoxyquinazoline compounds on the enzymatic activity of concentration-dependent fashion. The measured IC50s were 9.1 human JAK3 immunoprecipitated from the KL-2 EBV-trans- ␮M for WHI-P131, 11.0 ␮M for WHI-P97, and 27.9 ␮M for formed human lymphoblastoid B-cell line. WHI-P131, WHI- WHI-P154, but Ͼ300 ␮M for all of the other dimethoxyquina-

P154, and WHI-P97, which had very similar estimated Kis, zoline compounds (Table 1). WHI-P131 and WHI-P154 were ranging from 0.6 to 2.3 ␮M and were predicted to show signif- also tested against recombinant murine JAK3 expressed in a

Table 2 Effects of WHI-P131 against clonogenic leukemic cells WHI-P131 Cell linesa Experiment concentration no. (␮M) Mean no. of colonies/105 cells % inhibition Mean no. of colonies/105 cells % inhibition NALM-6 (pre B-ALL) BT20 (breast cancer)

1 0 2890 (2660, 3120) NA 2676 (2712, 2640) NA 0.1 2970 (2756, 3184) 0 ND ND 1 3180 (3080, 3136) 0 ND ND 10 1932 (1864, 2000) 33.2 3298 (2940, 3656) 0 100 2 (2, 2) Ͼ99.9 2190 (1632, 2748) 18.2

DAUDI (B-ALL) M24-MET (melanoma)

2 0 3950 (3100, 4800) NA 177 (120, 235) NA 0.3 2030 (1700, 2360) 48.6 312 (238, 386) 0 1 2136 (1568, 2704) 45.9 287 (157, 418) 0 3 1406 (988, 1824) 64.4 390 (280, 500) 0 10 1149 (1054, 1244) 70.9 301 (249, 353) 0 30 29 (12, 46) 98.0 599 (534, 664) 0

RAMOS (B-ALL) SQ20B (squamous carcinoma)

3 0 1286 (1164, 1408) NA 754 (452, 1056 NA 30 2 (0, 4) 99.8 838 (600, 1076) 0

MOLT-3 (T-ALL) HL-60 (AML)

0 1322 (1252, 1392) NA 1854 (1648, 2060) NA 100 1 (1, 1) Ͼ99.9 1 (1, 1) Ͼ99.9 a Cells were treated with WHI-P131 and then assayed for colony formation, as described in “Materials and Methods.” NA, not applicable; ND, not determined. T-ALL, T-cell ALL; AML, acute myelageneous leukemia.

baculovirus vector expression system and inhibited JAK3 in a interpretation of these remarkable differences in sensitivity of ␮ concentration-dependent fashion with an IC50 of 23.2 g/ml JAK3 versus JAK1 and JAK2 to WHI-131 and WHI-P154 must (ϳ78 ␮M; Fig. 4A) and 48.1 ␮g/ml (ϳ128 ␮M; Fig. 4B), respec- await determination of the X-ray crystal structures of these tively. The ability of WHI-P131 and WHI-P154 to inhibit re- kinases because simple amino acid discrepancies in their cata- combinant JAK3 was confirmed in four independent experi- lytic sites could result in pronounced structural differences. ments. These kinase assay results are consistent with our Specificity of WHI-P131 as a Tyrosine Kinase Inhibi- modeling studies described above. tor. Compound WHI-P131 was selected for additional exper- Importantly, WHI-P131 and WHI-P154 did not exhibit any iments designed to examine the sensitivity of non-Janus family detectable inhibitory activity against recombinant JAK1 or PTKs to this novel dimethoxyquinazoline class of JAK3 inhib- JAK2 in immune complex kinase assays (Fig. 4, C and D). itors. The inhibitory activity of WHI-P131 against JAK3 was EMSAs were also performed to confirm the JAK3 specificity of specific because it did not affect the enzymatic activity of other these dimethoxyquinazoline compounds by examining their ef- PTKs (Table 1; Fig. 5), including the ZAP/SYK family tyrosine fects on cytokine-induced STAT activation in 32Dc11/IL2R␤ kinase SYK (Fig. 5C), the TEC family tyrosine kinase BTK cells. As shown in Fig. 4E, both WHI-P131 (10 ␮g/ml ϭ 33.6 (Fig. 5D), the SRC family tyrosine kinase LYN (Fig. 5E), and ␮M) and WHI-P154 (10 ␮g/ml ϭ 26.6 ␮M) but not the control the receptor family tyrosine kinase IRK (Fig. 5F), even at compound WHI-P132 (10 ␮g/ml ϭ 33.6 ␮M) inhibited JAK3- concentrations as high as 350 ␮M. dependent STAT activation after stimulation with IL-2, but they A structural analysis of these PTKs was performed using did not affect JAK1/JAK2-dependent STAT activation after the crystal structures of HCK (which served as a homology stimulation with IL-3. Modeling studies suggest that this ex- model for LYN; Ref. 9) and IRK (36) and constructed homology quisite JAK3 specificity could in part be due to an alanine models of JAK3, BTK, and SYK. This analysis revealed some residue (Ala-966) that is present in the catalytic site of JAK3 but nonconserved residues located in the catalytic binding site of the changes to glycine in JAK1 and JAK2. This alanine group, different tyrosine kinases, which may contribute to the specific- which is positioned near the phenyl ring of the bound dime- ity of WHI-P131 (Fig. 3). One such residue, which is located thoxyquinazoline compounds, can provide greater hydrophobic closest to the docked inhibitors, is Ala-966 in JAK3 (Fig. 3, contact with the phenyl group and, thus, can contribute to higher region E), which may provide the most favorable molecular affinity relative to the smaller glycine residue in this region of surface contact with the hydrophobic phenyl ring of WHI-P131. the binding site in JAK1 and JAK2. However, an accurate The fact that WHI-P131 did not inhibit LYN, although LYN

contains the Ala residue conserved in JAK3 (Ala-966), suggests significant decrease in mitochondrial transmembrane potential that other factors (residue differences) contribute to this selec- in NALM-6 human leukemia cells. tivity. Other nonconserved residues in the catalytic site of tyro- We next used the in situ TdT-mediated labeling of 3Ј-OH sine kinases are shown in Fig. 3 (regions A–F). All of these termini with digoxigenin-conjugated UTP assay method com- differences in residues, especially residues that directly contact bined with confocal laser scanning microscopy to confirm that the bound inhibitor, may play an important role in the observed WHI-P131 can induce apoptosis in leukemia cells. At 48 h after specificity of WHI-P131 for JAK3. treatment with WHI-P131 at concentrations ranging from 10 to Antileukemic Activity of WHI-P131. We hypothesized 500 ␮M, NALM-6 cells were examined for digoxigenin-dUTP that compound WHI-P131 would exhibit significant cytotoxic incorporation using FITC-conjugated antidigoxigenin (green activity against JAK3-expressing human leukemia cells. To test fluorescence) and propidium iodide counterstaining (red fluo- this hypothesis, we first examined leukemic cells exposed to this rescence). The percentage of apoptotic cells increased in a

novel JAK3 inhibitor for apoptosis-associated changes in mito- concentration-dependent fashion with an average EC50 of 84.6 chondrial membrane potential (⌬⌿m) and mitochondrial mass ␮M (Fig. 7A). Fig. 7, B.1 and B.2, depicts the two-color confocal using specific fluorescent mitochondrial probes and multipa- microscopy images of vehicle-treated control cells and cells rameter flow cytometry. To measure changes in ⌬⌿m, we used treated with 100 ␮M WHI-P131. WHI-P131-treated cells

DiIC1 (which accumulates in energized mitochondria), whereas showed apoptotic yellow nuclei (yellow indicates superimposed the mitochondrial mass was determined by staining the cells green and red fluorescence; Fig. 7B.2). Further evidence for with NAO, a fluorescent dye that binds to the mitochondrial apoptosis was observed in DNA fragmentation assays. Because inner membrane independent of energetic state. Treatment of of their exquisite sensitivity in detecting DNA fragments re- NALM-6 leukemia cells with WHI-P131 at 7.4 ␮g/ml (25 ␮M) leased from a small percentage of apoptotic cells, the DNA gel to 60 ␮g/ml (200 ␮M) for 24 or 48 h increased the number of assays of apoptosis are uniquely suited to examine the nonspe- depolarized mitochondria in a concentration- and time-depen- cific toxicity of new antileukemic agents. Fig. 8A demonstrates

dent manner as determined by flow cytometry using DiIC1 that supernatants from NALM-6 leukemia cells, treated with 1 (Refs. 26–28; Fig. 6A). As shown in Fig. 6A, the fraction of or 3 ␮M WHI-P131, contained oligonucleosome-length DNA

DiIC1-negative cells with depolarized mitochondria increased fragments with a “ladder-like” fragmentation pattern consistent from 1.3% in vehicle treated control cells to 81.6% in cells with apoptosis, whereas no DNA fragments were detected in ␮ treated with 200 M WHI-P131 for 48 h. The average EC50s for supernatants of NALM-6 cells treated with structurally similar WHI-P131-induced depolarization of mitochondria, as meas- dimethoxyquinazoline compounds that lacked JAK3-inhibitory ␮ ured by decreased DiIC1 staining, were 79.3 M fora24h activity. Unlike JAK3-positive leukemia cells (NALM-6 cells in treatment and 58.4 ␮M for a 48 h treatment. The observed Fig. 8A and LC1;19 cells in Fig. 8B), JAK3-negative SQ20B changes in ⌬⌿m were not due to loss in mitochondrial mass, as squamous carcinoma cells and M24-MET melanoma cells did confirmed by a virtually identical staining intensity of NAO in not show any evidence of apoptotic DNA fragmentation after the treated and untreated NALM-6 cells (Fig. 6B). To further treatment with WHI-P131 (Fig. 8B). Taken together, these re- confirm this relative change in ⌬⌿m, we used JC-1, a mito- sults provided experimental evidence that the JAK3-specific chondrial dye, which normally exists in solution as a monomer tyrosine kinase inhibitor WHI-P131 results in depolarization of emitting green fluorescence and assumes a dimeric configura- the mitochondrial membrane and triggers apoptotic death in tion emitting red fluorescence in a reaction driven by mitochon- human B-lineage ALL cells, as evidenced by the ladder-like drial transmembrane potential (29). Thus, the use of JC-1 allows fragmentation pattern of nuclear DNA and digoxigenin-11-UTP simultaneous analysis of mitochondrial mass (green fluores- labeling of the exposed 3Ј-hydroxyl end of the fragmented cence) and mitochondrial transmembrane potential (red/orange nuclear DNA in the presence of TdT. fluorescence). After treatment of NALM-6 cells with WHI-P131 We also examined the antileukemic activity of WHI-P131 at increasing concentrations ranging from 25 ␮M to 200 ␮M and by determining its ability to inhibit the in vitro clonogenic with increasing duration of exposure of 24 or 48 h, we observed growth of the ALL cell lines NALM-6, DAUDI, LC1;19, a progressive dissociation between ⌬⌿m and mitochondrial RAMOS, MOLT-3, and the acute myelogenous leukemia cell mass, with decrement in JC-1 red/orange fluorescence without a line HL-60. As detailed in Table 2, WHI-P131 inhibited clono-

significant corresponding drop in JC-1 green fluorescence (Fig. genic growth in a concentration-dependent fashion with EC50s 6, C and D). As shown in Fig. 6C, the fraction of JC-1 red/ of 24.4 ␮M for NALM-6 cells and 18.8 ␮M for DAUDI cells. At orange fluorescence-positive cells decreased from 79.2% in 100 ␮M, WHI-P131 inhibited the in vitro colony formation by vehicle-treated control cells to 16.9% in cells treated with 200 these leukemia cell lines by Ͼ99%. In contrast, WHI-P131 did ␮M WHI-P131 for 48 h. The corresponding values for JC-1 not inhibit the clonogenic growth of JAK3-negative M24-MET green fluorescence were 99.3% for vehicle-treated cells and melanoma or SQ20B squamous carcinoma cell lines (Table 2). 99.8% for WHI-P131-treated (200 ␮M for 48 h) cells. The In summary, our study demonstrates that a novel homology

average EC50s for WHI-P131 induced depolarization of mito- model of the JAK3 kinase domain can be used for structure- chondria, as measured by decreased JC-1 red/orange fluores- based design and synthesis of potent and specific inhibitors of cence were 94.2 ␮M for a 24 h treatment and 50.4 ␮M fora48h JAK3. The reported lead compound WHI-P131 inhibited JAK3 treatment. Fig. 6D compares the single color (red/orange) fluo- but not other PTKs including JAK2, SYK, BTK, LYN, and IRK. rescent confocal images of vehicle-treated and WHI-P131- ALL cells express JAK2; a JAK2 inhibitor has been reported to treated (100 ␮M ϫ 48 h) NALM-6 cells stained with JC-1. exhibit antileukemic activity (37). Similarly, the SRC family These results collectively demonstrate that WHI-P131 causes a PTK LYN, the ZAP/SYK family PTK SYK, and the TEC

family PTK BTK are expressed in ALL cells and affect their 6. Jurlander, J., Lai, C. F., Tan, J., Chou, C. C., Geisler, C. H., Schriber, adhesion, proliferation, and survival (19, 23, 38, 39). IRK is the J., Blumenson, L. E., Narula, S. K., Baumann, H., and Caligiuri, M. A. only member of the receptor PTK family that has been detected Characterization of interleukin-10 receptor expression on B-cell chronic lymphocytic leukemia cells. Blood, 89: 4146–4152, 1997. in leukemic cells, especially pre-B ALL cells with a t(1;19) 7. Kaneko, S., Suzuki, N., Koizumi, H., S. Yamamoto, S., and Sakane, translocation (40–42). Because our JAK3-specific inhibitor T. Rescue by cytokines of apoptotic cell death induced by IL-2 depri- WHI-P131 does not inhibit these tyrosine kinases, its ability to vation of human antigen-specific T cell clones. Clin. Exp. Immunol., kill ALL cells cannot be attributed to a nonspecific inhibition of 109: 185–193, 1997. JAK2, LYN, SYK, BTK, or IRK in these cells. The use of this 8. Nakamura, N., Chin, H., Miyasaka, N., and Miura, O. An epidermal compound in biological assays confirmed that JAK3 is a vital growth factor receptor/Jak2 tyrosine kinase domain chimera induces tyrosine phosphorylation of Stat5 and transduces a growth signal in therapeutic target in ALL and demonstrated that WHI-P131 hematopoietic cells. J. Biol. Chem., 271: 19483–19488, 1996. triggers apoptosis in leukemia cells. Potent and specific inhib- 9. Sicheri, F., Moarefi, I., and Kuriyan, J. Crystal structure of the Src itors of JAK3, such as the dimethoxyquinazoline compound family tyrosine kinase Hck. Nature (Lond.), 385: 602–609, 1997. WHI-P131, may, therefore, provide the basis for the design of 10. Mohammadi, M., Schlessinger, J., and Hubbard, S. R. Structure of new treatment strategies against ALL, which is the most com- the FGF receptor tyrosine kinase domain reveals a novel autoinhibitory mon form of childhood cancer. mechanism. Cell, 86: 577–587, 1996. Finally, our homology model uniquely indicates that the 11. Mohammadi, M., McMahon, G., Sun, L., Tang, C., Hirth, P., Yeh, ϫ ϫ B. K., Hubbard, S. R., and Schlessinger, J. Structures of the tyrosine active site of JAK3 measures approximately 8 Å 11 Å 20 kinase domain of fibroblast growth factor receptor in complex with 3 Å with an ϳ530-Å volume available for inhibitor binding. Our inhibitors. Science (Washington DC), 276: 955–960, 1997. modeling studies using the constructed homology model of 12. Hubbard, S. R., Wei, L., Ellis, L., and Hendrickson, W. A. Crystal JAK3 kinase also showed that there is significant opportunity structure of the tyrosine kinase domain of the human insulin receptor. for improvement of the quinazoline inhibitors. The JAK3 model Nature (Lond.), 372: 746–754, 1994. shows that there is additional volume in the ATP-binding site 13. Sack, J. S. CHAIN: a crystallographic modeling program. J. Mol. Graphics, 6: 244–245, 1988. which can be better used by quinazoline derivatives. The aver- 14. Bru¨nger, A. T. X-PLOR, A System for X-Ray Crystallography and age molecular volume of our dimethoxyquinazoline compounds NMR, Ed. 3.1. New Haven, CT: Yale University Press, 1992. is 277 Å3, which is well below the estimated total volume of the 3 15. Ghosh, S., Zheng, Y., Jun, X., Narla, R., Mahajan, S., Navara, C., binding site, 530 Å . This leaves opportunities for the design of Mao, C., Sudbeck, E. A., and Uckun, F. M. ␣-Cyano-␤-hydroxy-␤- new inhibitors which have slightly larger functional groups at methyl-N-[4-(trifluoromethoxy)phenyl]propenamide: an inhibitor of the the 2Ј and 3Ј positions of the phenyl ring. Structural and chem- EGF receptor tyrosine kinase with potent cytotoxic activity against ical features of dimethoxyquinazoline compounds, which are breast cancer cells. Clin. Cancer Res., 4: 2657–2668, 1998. proposed to facilitate their binding to the Jak3 catalytic site, 16. Mahajan, S., Ghosh, S., Sudbeck, E. A., Zheng, Y., Downs, S., Hupke, M., and Uckun, F. M. Rational design and synthesis of a novel include the following features, which are illustrated in Fig. 2C: antileukemic agent targeting Bruton’s tyrosine kinase (BTK): LFM-A13 (a) the presence of a 4Ј-OH group on the phenyl ring; (b) the [␣-cyano-␤-hydroxy-␤-methyl-N-(2,5-dibromophenyl)propenamide]. presence of a hydrogen bond acceptor (N, carbonyl, OH) near J. Biol. Chem., 274: 9587–9599, 1999. the Leu-905 NH or a hydrogen bond donor (NH, OH) near the 17. Bohm, H. J. The development of a simple empirical scoring func- Leu-905 carbonyl; (c) a relatively planar molecular shape to tion to estimate the binding constant for a protein-ligand complex of known three-dimensional structure. J. Comput. Aided Mol. Des., 8: allow access to the binding site; (d) the ability to fit into a 3 243–256, 1994. 530-Å space defined by the residues lining the JAK3 catalytic 18. Rewcastle, G. W., Denny, W. A., Bridges, A. J., Zhou, H., Cody, site. These predicted binding preferences to JAK3 residues in D. R., McMichael, A., and Fry, D. W. Tyrosine kinase inhibitors. 5. the catalytic site can be used for the design of new and more Synthesis and structure-activity relationships for 4-[(phenylmethyl) potent inhibitors of JAK3 as antileukemic agents. amino]-and 4-(phenylamino)quinazolines as potent adenosine 5Ј- triphosphate binding site inhibitors of the tyrosinekinase domain of the epidermal growth factor receptor. J. Med. Chem., 38: 3482–3487, 1995. 19. Vassilev, A., Ozer, Z., Navara, C., Mahajan, S., and Uckun, F. M. REFERENCES Bruton’s tyrosine kianse (BTK) as an inhibitor of the Fas/CD95 death 1. Ihle, J. N. The Janus protein tyrosine kinase family and its role in inducing signaling complex (DISC). J. Biol. Chem., 274: 1646–1656, cytokine signaling. Adv. Immunol., 60: 1–35, 1995. 1999. 2. Witthuhn, B. A., Williams, M. D., Kerawalla, H., and Uckun, F. M. 20. Uckun, F. M., Waddick, K. G., Mahajan, S., Jun, X., Takata, M., Differential substrate recognition capabilities of Janus family protein Bolen, J., and Kurosaki, T. Bruton’s tyrosine kinase as a mediator of tyrosine kinases within the interleukin 2 receptor (Il2R) system: Jak3 as radiation-induced apoptosis in B-lineage lymphoid cells. Science a potential molecular target for treatment of leukemias with a hyperac- (Washington DC), 22: 1096–1100, 1996. tive Jak-Stat signaling machinery. Leuk. Lymphoma, 32: 289–297, 21. Goodman, P. A., Niehoff, L. B., and Uckun, F. M. Role of tyrosine 1999. kinases in induction of the c-jun proto-oncogene in irradiated B-lineage 3. Horvath, C. M., and Darnell, J. E. The state of the STATs: recent lymphoid cells. J. Biol. Chem., 273: 17742–17748, 1998. developments in the study of signal transduction to the nucleus. Curr. 22. Mahajan, S., Fargnoli, J., Burkhardt, A. L., Kut, S. A., Saouaf, Opin. Cell. Biol., 9: 233–239, 1997. S. J., and Bolen, J. B. Src family protein tyrosine kinases induce 4. Levy, D. E. The house that Jak/STAT built. Cytokine Growth Factor autoactivation of Bruton’s tyrosine kinase. Mol. Cell. Biol., 15: Rev., 8: 81–90, 1997. 5304–5311, 1995. 5. Demoulin, J. B., Uyttenhove, C., Van Roost, E., DeLestre, B., 23. Uckun, F. M., Evans, W. E., Forsyth, C. J., Waddick, K. G., Donckers, D., Van Snick, J., and Renauld, J. C. A single tyrosine of the Ahlgren, L. T., Chelstrom, L. M., Burkhardt, A., Bolen, J., and Myers, interleukin-9 (IL-9) receptor is required for STAT activation, antiapo- D. E. Biotherapy of B-cell precursor leukemia by targeting genistein to ptotic activity, and growth regulation by IL-9. Mol. Cell. Biol., 16: CD19-associated tyrosine kinases. Science (Washington DC), 267: 4710–4716, 1996. 886–891, 1995.

24. Uckun, F. M., Tuel-Ahlgren, L., Waddick, K. G., Jun, X., Jin, J., 34. Uckun, F. M., Kazimiera, J. G-P., Kersey, J. H., Houston, L., and Myers, D. E., Rowley, R. B., Burkhardt, A. L., and Bolen, J. B. Physical Vallera, D. A. Use of a novel colony assay to evaluate the cytotox- and functional interactions between Lyn and p34CDC2 kinases in icity of an immunotoxin containing pokeweed antiviral protein irradiated human B-cell precursors. J. Biol. Chem., 271: 6396–6397, against blast progenitor cells freshly obtained from patients with 1996. common B-lineage acute lymphoblastic leukemia. J. Exp. Med., 163: 25. Uckun, F. M., Jun, X., Narla, R. K., Zeren, T., Venkatachalam, T., 347–368, 1986. Waddick, K., Rostostev, A., and Myers, D. E. Cytotoxic activity of 35. Messinger, Y., Yanishevski, Y., Ek, O., Zeren, T., Waurzyniak, B., EGF-genistein against human breast cancer cells. Clin. Cancer Res., 4: Gunther, R., Chelstrom, L., Chandan-Langlie, M., Schneider, E., Myers, 901–912, 1998. D. E., Evans, W., and Uckun, F. M. In vivo toxicity and pharmacokinetic 26. Fujii, H., Cody, S. H., Seydel, U., Papadimitriou, J. M., Wood, features of B43 (anti-CD19)-genistein immunoconjugate in nonhuman D. J., and Zheng, M. H. Recording of mitochondrial transmembrane primates. Clin. Cancer Res., 4: 165–170, 1998. potential and volume in cultured rat osteoclasts by confocal laser scan- 36. Hubbard, S. R. Crystal structure of the activated insulin receptor ning microscopy. Histochem J., 29: 571–581, 1997. tyrosine kinase in complex with peptide substrate and ATP analog. 27. Mancini, M., Anderson, B. O., Caldwell, E., Sedghinasab, M., Paty, EMBO J., 16: 5572–5581, 1997. P. B., and Hockenbery, D. M. Mitochondrial proliferation and paradoxical membrane depolarization during terminal differentiation and apoptosis in a 37. Meydan, N., Grunberger, T., Dadi, H., Shahar, M., Arpaia, E., human colon carcinoma cell line. J. Cell Biol., 138: 449–469, 1997. Lapidot, Z., Leeder, J. S., Freedman, M., Cohen, A., Gazit, A., Levitzki, A., and Roifman, C. M. Inhibition of acute lymphoblastic leukaemia by 28. Petit, P. X., Lecoeur, H., Zorn, E., Dauguet, C., Mignotte, B., and a Jak-2 inhibitor. Nature (Lond.), 379: 645–648, 1996. Gougeon, M. L. Alterations in mitochondrial structure and function are early events of dexamethasone-induced thymocyte apoptosis. J. Cell 38. Kristupaitis, D., Dibirdik, I., Vassilev, A., Mahajan, S., Kurosaki, Biol., 130: 157–167, 1995. T., Chu, A., Tuel-Ahlgren, L., Tuong, D., Pond, D., Luben, R., and Uckun, F. M. Electromagnetic field-induced stimulation of Bruton’s 29. Smiley, S. T., Reers, M., Mottola-Hartson, C., Lin, M., Chen, A., tyrosine kinase (BTK). J. Biol. Chem., 273: 9119–9123, 1998. Smith, T. W., Steele, G. D., and Chen, L. Intracellular heterogeneity in mitochondrial membrane potentials revealed by J-aggregate-forming li- 39. Xiao, J., Messinger, Y., Jin, J., Myers, D. E., Bolen, J. B., and pophilic cation JC-1. Proc. Natl. Acad. Sci. USA, 88: 3671–3675, 1991. Uckun, F. M. Signal transduction through the ␤1 integrin family surface 30. Maftah, A., Petit, J. M., Ratinaud, M. H., and Julien, R. 10-N nonyl adhesion molecules VLA-4 and VLA-5 of human B-cell precursors acridine orange: a fluorescent probe which stains mitochondria inde- activates CD19 receptor-associated protein-tyrosine kinases. J. Biol. pendently of their energetic state. Biochem. Biophys. Res. Commun., Chem., 271: 7659–7664, 1996. 164: 185–190, 1989. 40. Kaplan, G. C., Pillion, D. J., Rutter, W. J., Kim, H., and Barker, P. E. 31. Narla, R. K., Liu, X-P., Myers, D. E., and Uckun, F. M. 4-(3Ј- Insulin receptor overexpression in a human pre-B acute lymphocytic leu- Bromo-4Ј hydroxylphenyl)-amino-6,7-dimethoxyquinazoline: a novel kemia cell line with a t(1;19) chromosome translocation near the INSR quinazoline derivative with potent cytotoxic activity against human locus. Biochem. Biophys. Res. Commun., 159: 1275–1282, 1989. glioblastoma cells. Clin. Cancer Res., 4: 1405–1414, 1998. 41. Newman, J. D., Harrison, L. C., Eckardt, G. S., and Jack, I. 32. Zhu, D-M., Narla, R. K., Fang, W-H., Chia, N-C., and Uckun, F. M. Enhanced insulin-receptor tyrosine kinase activity associated with chro- Calphostin C triggers calcium-dependent apoptosis in human acute mosomal translocation (1;19) in a pre-B-cell leukemia line. Int. J. lymphoblastic leukemia cells. Clin. Cancer Res., 4: 2967–2976, 1998. Cancer, 50: 500–504, 1992. 33. D’Cruz, O., P., G., and Uckun, F. M. Spermicidal activity of 42. Bushkin, I., and Zick, Y. Alterations in insulin receptor kinase metallocene complexes containing vanadium (IV)1 in humans. Biol. activity during differentiation of HL-60 cells. Biochem. Biophys. Res. Reprod., 58: 1515–1526, 1998. Commun., 172: 676–682, 1990.

Downloaded from clincancerres.aacrjournals.org on October 1, 2021. © 1999 American Association for Cancer Research. Structure-based Design of Specific Inhibitors of Janus Kinase 3 as Apoptosis-inducing Antileukemic Agents

Elise A. Sudbeck, Xing-Ping Liu, Rama Krishna Narla, et al.

Clin Cancer Res 1999;5:1569-1582.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/5/6/1569

Cited articles This article cites 41 articles, 21 of which you can access for free at: http://clincancerres.aacrjournals.org/content/5/6/1569.full#ref-list-1

Citing articles This article has been cited by 27 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/5/6/1569.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 Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/5/6/1569. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.