Protein Kinase Inhibitors: Structural Insights Into ... - Ingenta Connect

Current Pharmaceutical Design, 2007, 13, 2751-2765

2751

Protein Kinase Inhibitors: Structural Insights Into Selectivity Ram Thaimattam1,* Rahul Banerjee1, Rajni Miglani1 and Javed Iqbal2 1

Department of Molecular Modeling and Drug Design and 2Discovery Chemistry, Dr. Reddy’s Laboratories Ltd., Discovery Research, Bollaram Road, Miyapur, Hyderabad 500 049, India Abstract: Protein kinases are involved in many diseases like cancer, inflammation, cardiovascular disease, and diabetes. They have become attractive target classes for drug development, making kinase inhibitors as important class of therapeutics. The success of smallmolecule ATP-competitive kinase inhibitors such as Gleevec, Iressa, and Tarceva has attracted much attention in the recent past. Kinases make use of ATP for phosphorylation of a specific residue(s) on their protein substrates. More than 400 X-ray structures of about 70 different kinases are publicly available. These structures provide insights into selectivity and mechanisms of inhibition. However, prediction of binding specificity of kinase inhibitors based on structural information alone appears to be insufficient. Here, we will review these observations to gain insights into the rules that govern protein kinase inhibitor selectivity.

Key Words: Protein kinases, kinase domain, protein kinase conformation, DFG motif, kinase inhibitors, ATP-competitive inhibitors, kinase inhibitor selectivity. INTRODUCTION Kinases play key roles in cancer, inflammation, diabetes and other diseases, making them one of the most important drug targets [1-2]. They are directly or indirectly involved in most signaling pathways. Aberrant kinase activity has been associated with these diseases. Sequencing of the human genome has revealed more than 500 genes encoding protein kinases [3]. All these enzymes make use of ATP for phosphorylation of a specific Ser/Thr/Tyr residue(s) on protein substrates. In many cases protein kinases themselves are substrates for phosphorylation. The success of small-molecule ATP-competitive inhibitors such as, imatinib (Gleevec) for the treatment of chronic myeloid leukemia (CML) and gastrointestinal stromal tumors (GIST), and gefitinib (Iressa) and erlotinib (Tarceva) for the treatment of non-small-cell lung cancer (NSCLC), confirm that this strategy is indeed effective [4-5]. X-ray crystal structures for a range of kinases are available in the Protein Data Bank (PDB) [6]. They provide structural basis for understanding kinase inhibition and facilitate structure-guided design of kinase-specific inhibitors, the details of which have been reviewed extensively elsewhere [7-11]. Although Gleevec is found to be a selective Abl kinase inhibitor, design of ATP-competitive inhibitors selective for a particular kinase appears to be quite challenging due to conserved protein fold and commonality of the catalytic mechanism across the kinase family. High selectivity of Gleevec (which is also known as imatinib mesylate or STI-571) is believed to be because of its ability to bind to a specific inactive conformation of Abl kinase. Gleevec is also a potent inhibitor of Platelet Derived Growth Factor Receptor (PDGFR) and c-Kit [12]. However, a recent study revealed that none of the ATP-competitive kinase inhibitors which are in the clinical trials or clinical use, including Gleevec, was really selective when tested against a panel of 113 kinases [13]. Here, we will review these observations, particularly focusing on ATP-competitive inhibitors, to understand the factors that govern protein kinase inhibitor selectivity.

much interest in various therapeutic areas such as inflammation and rheumatoid arthritis, cardiovascular diseases, diabetes, and neurological disorders. Inhibition of kinases such as Abl, EGFR, VEGFR, Src, B-raf, and Aurora has become a major area of therapeutic intervention for the treatment of various cancers [14-17]. GSK-3 and JNK are implicated in various diseases including diabetes, Alzheimer disease and CNS disorders [18-19]. PKCs, MAPK1/2, PI3K, p38 MAPK, and JNK have emerged as possible targets for cardioprotection [19]. Therefore, the ability to modulate kinase activity represents an attractive therapeutic strategy for the treatment of human illnesses. One of the major problems associated with the targeted cancer therapy has been the emergence of drug-resistant mutants. Gleevec, which is currently used for the treatment of CML, is found to be inactive against 33 clinically relevant Bcr-Abl mutants [20]. In contrast, most EGFR mutations predict favorable response to treatment with Iressa and Tarceva in non-small-cell lung cancer patients compared with the wild-type receptor; however, some EGFR mutations still confer drug resistance [21-22]. The molecular basis for such different sensitivity of EGFR mutants is not clear, as these inhibitors show similar in vitro inhibitory activities for wildtype and clinically relevant mutants [13].

TARGETING PROTEIN KINASES Protein kinases are rapidly becoming attractive targets for drug discovery, next to G protein–coupled receptors (GPCRs). Drugs such as Gleevec, Iressa and Tarceva have proven effective in the treatment of various cancers without the negative side effects of traditional chemotherapy. Recently, kinases have also attracted

3D STRUCTURE In the recent past, there has been an explosion in the number of kinase structures in the PDB. Kinase domain structures deposited in the PDB over last three years account for nearly 50% of the total number of entries for the same domain. Over 400 X-ray structures of the catalytic domain of about 70 different kinases are available in the PDB as of December 2005. The kinase domain is composed of two lobes joined by a peptide chain often referred to as “hinge” or “linker” (Fig. 1). The N-terminal lobe consists of mostly antiparallel -sheets and a conserved helix (helix C), while the C-terminal lobe is predominantly helical. ATP binds in a cleft formed between the two lobes of the kinase fold. The adenine group of ATP forms two hydrogen bonds with the hinge region, and the conserved Lys (from III) and Asp from DFG motif (the N-terminal part of activation loop) together with a Mg2+ ion are involved in stabilization of the transition-state, while P-loop (Glycine rich loop) acts as a flexible clamp stabilizing interactions with the triphosphate group (Fig. 1).

*Address correspondence to this author at the Department of Molecular Modeling and Drug Design, India; Tel: 91 40 2304 5439; Fax: 91 40 2304 5438; E-mail: [email protected]

KINASE REGULATION Activation loop plays a key role in catalytic regulation (Fig. 1). Phosphorylation of a Tyr/Ser/Thr residue(s) on this loop is required for full activity of many kinases [23-25]. The activation loop adopts

1381-6128/07 $50.00+.00

© 2007 Bentham Science Publishers Ltd.

2752 Current Pharmaceutical Design, 2007, Vol. 13, No. 27

Thaimattam et al.

Fig. (1). X-ray crystal structure of ATP bound to PKA (PDB code: 1ATP) showing distinct subdomains in the ATP binding site.

distinct conformations in active and inactive states in a phosphorylation dependent manner. The residue that undergoes phosphorylation is exposed in active form, whereas it is buried in case of inactive conformation. In addition, differences exist in the inter-lobe orientation, the disposition of helix C in the N-lobe, and the conformation of the N-terminal part of the activation loop, the DFG motif. In active form, the DFG motif adopts so-called 'DFG-in' conformation in such a way that the Phe residue from this motif is buried in a hydrophobic pocket (adjacent to the ATP binding site), while it adopts a ‘'DFG-out' conformation in inactive forms. An ion-pair interaction between the conserved amino acids Lys (from III) and Glu (located on helix C) is a characteristic feature of active conformation. This ion-pair interaction is absent in inactive conformations of many protein kinases but not in GleevecAbl (PDB code 1IEP) and few p38 complexes (1KV2, 1WBV etc). KINASE INHIBITION Structural information on kinases has grown enormously over the past few years relative to the first X-ray structure determined for protein kinase A (2CPK) in 1992 [26]. These structures provide insights into kinase inhibitor selectivity and mechanisms of inhibition. Strategies to inhibit protein kinases can be broadly classified into three: ATP-competitive inhibition, substrate-competitive inhibition, and allosteric inhibition. ATP-competitive inhibition of protein kinases is a widely studied topic [7-11]. ON012380, a smallmolecule Abl kinase inhibitor, competes with the substrate to block the activity. It was found to inhibit all of the clinically relevant BcrAbl mutants at concentrations less than 10 nM in vitro [27]. Allosteric inhibitors indirectly compete with ATP by stabilizing a kinase conformation that is incompatible with ATP. Well-known examples include p38 and MEK1/2 inhibitors that, instead of targeting the ATP binding site, bind to the adjacent hydrophobic pocket [28-29]. Allosteric binding at a site away from the ATP binding pocket has been identified in JNK. The crystal structure of JNK1 with pepJIP1, a peptide fragment of the scaffolding protein JIP1, reveals that pepJIP1 binding to JNK1 distorts the ATP binding cleft in such a way that the affinity of the kinase for ATP is reduced by three-fold [3031]. This review will focus on recent advances in understanding the selectivity of ATP-competitive inhibitors.

ATP-COMPETITIVE INHIBITORS Efforts towards the design of selective small-molecule kinase inhibitors have met with some success in spite of the fact that the overall protein fold is conserved in the kinase family. The kinase active site is also highly conserved among members of this family. It is comprised of several sub-sites that include hydrophobic pocket, adenine pocket, ribose pocket, phosphate groove, and specificity surface (Fig. 1). ATP does not occupy all these regions, and differences in these sub-sites can be utilized in the design of potent and selective kinase inhibitors. Targeting the Hydrophobic Pocket for Kinase Inhibition The hydrophobic region is formed as a lipophilic pocket located deep inside the ATP binding cleft towards helix C (Fig. 1). This space is not utilized by ATP itself, but ATP competitive inhibitors can take advantage of it for selective inhibition of a particular kinase. Quinazolines have been identified as potent inhibitors of various kinases such as Src, Aurora, p38 MAPK, EGFR, VEGFR, and ErbB2 [32-38]. Quinazoline based EGFR inhibitors (Iressa and Tarceva) are already in the market for the treatment of NSCLC, while several other analogues targeting different kinases are in discovery or in various developmental stages. It is noteworthy that by decorating the aniline ring alone selectivity against a particular kinase target can be significantly improved. There are also examples of distinct scaffolds inhibiting a particular kinase. Selective inhibitors of varying scaffolds have been developed for specific CDKs and Aurora kinase [39-40]. EGFR Kinase X-ray crystal structures of quinazoline based inhibitors bound to EGFR (Tarceva-1M17 and lapatinib-1XKK), p38 (1DI9), CDK2 (1DI8), and Aurora (2C6E) indicate that the binding pose of the quinazoline core is virtually similar in all of them. Tarceva (Fig. 2) is a potent and selective inhibitor of EGFR, whereas lapatinib is a potent EGFRErbB2 dual inhibitor [41]. In the crystal structure of Tarceva bound to EGFR (1M17, Fig. 3), the activation loop adopts a conformation similar to that of active kinases [42]. The aniline ring bearing acetylene moiety is inserted into the hydrophobic pocket. The N1 atom of the quinazoline is involved in hydrogen

Protein Kinase Inhibitors

Current Pharmaceutical Design, 2007, Vol. 13, No. 27

2753

F HN O O

O

O

HN N

N

O

N

O

N

O

Tarceva (1)

Cl

N Iressa (2)

Cl O O

F

S O

NH

HN O

Lapatinib (3)

N N

Fig. (2). Chemical structures of EGFR kinase inhibitors.

(a)

(b)

(c)

Fig. (3). X-ray structures of EGFR kinase with (a) Tarceva (inhibitor: orange C-atoms; enzyme: blue cartoon diagram), and (b) Lapatinib (cyan C-atoms and yellow cartoon). Water molecule connecting the inhibitor and the enzyme is shown in red sphere. The Phe residue from the DFG motif (green C-atoms) has different orientations in both complexes. (c) Stereoview of the overlay of both structures is shown using the same rendering and color scheme. Notice the differences in the conformations of the DFG motif and P-loop, and positioning of the helix C.


Thaimattam et al.

bonding interactions with the main chain NH in the hinge region, while N3 is connected to the side chain of the gatekeeper residue (Thr766 as in 1M17) through water-mediated interactions. For lack of these water mediated hydrogen bonding interactions, quinolines show poor binding in EGFR compared to the corresponding quinazolines [43]. In contrast, both quinolines and quinazolines are potent inhibitors of Aurora and VEGFR as the gatekeeper residue has a hydrophobic side chain in these kinases (Leu in Aurora, Val in VEGFR) [40, 44]. The water mediated interactions are also present in the crystal structure of a quinazoline bound to p38 (1DI9, gatekeeper residue: Thr) but not in CDK2 (1DI8) and Aurora (2C6E) in which the gatekeeper residue is hydrophobic (Phe in CDK and Leu in Aurora) [34, 35, 45]. The cyano group of SKI-606, a potent SrcAbl dual kinase inhibitor, might mimic the role of structural water observed in the crystal structures of quinazolines bound to EGFR and p38 [46]. In addition, there is a C-HO hydrogen bond between the acidic carbon (C2) and the main chain carbonyl O (Gln767 in 1M17), which is also present in the crystal structures of quinazolines bound to p38, CDK, and Aurora (1DI9, 1DI8, and 2C6E respectively) and lapatinibEGFR [47]. Unlike Tarceva, Lapatinib binds to an inactive-like conformation of EGFR (1XKK, [48]. The aniline ring bearing 3-fluorobenzyloxy group is deeply inserted into the hydrophobic pocket (Fig. 3). The orientation and hydrogen bonding interactions of Lapatinib are more or less similar to those found in TarcevaEGFR complex [42]. However, the overall conformation of the enzyme is significantly different in both complexes. The differences include relative orientation of the two lobes around the hinge region, the position of the helix C, and the DFG motif/activation loop conformation (Fig. 3c). In the crystal structure of Lapatinib bound to EGFR, the movement of helix C away from the active site results in the loss of a conserved Glu-Lys salt bridge (Glu738 and Lys721). Consequently, a bigger hydrophobic pocket is formed in which the bulkier fluorobenzyloxy group is accommodated comfortably. Aurora Kinase Members of the Aurora kinase family (Aurora A, B, and C) play a crucial role in mitosis by regulating chromosome segregation and cytokinesis [17]. Aberrant expression and activity of these en-

zymes have been indicated in many malignant tumors, including ovarian, colorectal, gastric, and breast cancer, and leukemia. Three Aurora kinase inhibitors have recently been described (ZM447439, Hesperadin, and VX-680, Fig. 4) [49]. None of these compounds show isoform specificity. The literature on selective inhibitors of Aurora kinases has been recently reviewed [40]. The crystal structure of compound (7) (Fig. 4) bound to Aurora A has been recently determined (2C6E) [34]. This class of inhibitors showed good selectivity for Aurora kinases. The orientation and hydrogen bonding interactions of compound (7) bound to Aurora A (Fig. 5) are similar to those of other quinazoline analogues bound to p38 (1DI9), CDK2 (1DI8), and EGFR (1M17 and 1XKK). In addition, the carbonyl O-atom and one of the pyrimidine N-atoms of the inhibitor form hydrogen bonding interactions with the conserved Lys161. Water-mediated hydrogen bonds involving the amide nitrogen of the inhibitor further stabilize the inhibitorenzyme complex. The DFG motif adopts a DFG-out conformation despite the introduction of the activating mutation, T287D. The ligand induced movement of helix C away from the ATP binding pocket is quite significant compared to the ADP complex (1MQ4). The additional space created by this movement is occupied by the extended substitution at the 4-position of the aniline ring, demonstrating the extent of induced fit (Fig. 5). p38 MAP Kinase p38 kinase plays a crucial role in regulating the production of proinflammatory cytokines, and elevated levels of these cytokines are associated with several autoimmune diseases, such as rheumatoid arthritis, diabetes and inflammatory bowel syndrome. The literature on p38 kinase inhibitors has been recently reviewed [50-51]. About 40 crystal structures of p38 kinase with various inhibitors are available in the PDB. The binding modes of SB203580 (1A9U) and compound (9) (1OUY) are partly similar (Fig. 6) [52-53]. In both structures, the fluorophenyl group is oriented towards the hydrophobic pocket in a similar fashion and the DFG motif has a typical `DFG in’ conformation, although the activation loop resembles that of an inactive kinase (Fig. 7). The pyridyl group of SB203580 forms a hydrogen bond with Met109 in the hinge region, and the imidazole N accepts

N

H N O

HN O

N

O

N

O

N

S

NH

O ZM-447439 (4)

O

N

Hesperadin (5)

O

HN

N

HN H N

N N

N

H N

N

S

N

O

O N

VX-680 (6) Fig. (4). Chemical structures of Aurora kinase inhibitors.

N

HN N

O

N

OH 7

O



2755

Fig. (5). Stereoview of the overlay of ADP (cyan C-atoms and green cartoon) and compound (7) (pink C-atoms and orange cartoon) bound to Aurora A as revealed by X-ray crystallography. Notice that helix C is moved away from the ATP binding pocket to accommodate the long and linear pyrimidinylbenzamide group.

F

F

F N

N

HN

N N

N H

N H

N

O

S

N

S

N

Cl SB203580 (8)

Cl

O

O

10

NH

Cl 9

O O

F

N

N N

N H

NH

HN

O

O

O 12

BIRB 796 (11)

N O

N

Cl

Fig. (6). Chemical structures of p38 MAP kinase inhibitors.

hydrogen bond from the conserved Lys53, while methylsulfinylphenyl ring makes favorable contacts with the P-loop. The quinazolinone and pyridol-pyrimidine classes of p38 MAP kinase inhibitors (eg compound (9)) show high specificity for p38 kinase over other MAP kinases. This is possible due to a peptide flip between Met109 and Gly110 to form an additional hydrogen bond with the

inhibitor. This conformational change is predicted to be energetically more favorable in p38 where the hinge region sequence is Met109-Gly110 than in other JNKs and ERKs where Gly110 is replaced with a bulkier residue [53]. However, conformations of the activation loop, the linker (hinge) and the P-loop in SB203590p38 are strikingly different compared to 9p38 (Fig. 7a).


Thaimattam et al.

(a)

(b)

(c)

Fig. (7). Stereoviews of the overlay of (a) SB203580 (green C-atoms and green cartoon) and 9 (cyan C-atoms and cyan cartoon), and (b) BIRB 796 (orange Catoms and orange cartoon) and compound (12) (pink C-atoms and pink cartoon) bound to p38 as revealed by their crystal structures; (c) Stereoview of the overlay of all four structures is shown using the same rendering and color. The Phe residue from the DFG motif is also shown. The activation loop in 12p38 complex is missing. Notice that the active site conformations are distinct in all these cases.



Bcr-Abl Kinase Structural studies reveal that Gleevec (Fig. 8) selectively binds to a specific inactive conformation of the Abl kinase domain through an induced-fit mechanism [55-56]. The pyridine ring of Gleevec lies in the adenine binding pocket and the rest of the molecule enters deep into the hydrophobic pocket, and wedges between the activation loop and helix C (1IEP, Fig. 9). The inhibitorenzyme complex is stabilized by as many as six hydrogen bonds and several van der Waals interactions. PD173955, a pyrido[2,3]pyrimidine based compound, binds to a conformation in which the activation loop resembles that of an active kinase (1M52) [57]. It is a more potent inhibitor of Abl than Gleevec, and structural data indicates that the greater potency may be due to its ability to bind to both active and inactive conformations of Abl [55-57]. This is due to the fact that PD173955 does not penetrate deeply into the hydrophobic pocket, and therefore does not interfere with the DFG motif residues. As a result, its binding is not influenced by the activation loop conformation. On the other hand, BMS-354825 and VX-680 which bind to the active form of Abl suffer from poor selectivity [58]. This led to a popular belief that to achieve higher selectivity it is advantageous to target an inactive kinase conformation. However, Gleevec was found to be inactive against several clinically relevant Bcr-Abl mutants [20]. Some of the mutated residues, although located far from the active site, still confer resistance towards Gleevec. Gleevec-resistant mutations are believed to destabilize the inactive conformation of Abl or directly obstruct the inhibitor binding [59-60]. To circumvent drug resistance back-up molecules were designed in such a way that they target both active and inactive forms of the Abl kinase, despite the fact that such inhibitors show poor selectivity [61]. BMS-354825 and AP23464 were found to inhibit clinically common Gleevec-resistant Bcr-Abl mutants [62-63]. Besides effectively blocking Abl kinase activity, Gleevec is found to be highly effective against other tyrosine kinases such as c-Kit and PDGFR [12]. Higher activity of c-Kit is implicated in the pathogenesis of human cancers, which is tightly regulated in normal cells. The structure of Gleevec bound to c-Kit (1T46) has been recently determined by X-ray crystallography [64]. The binding pose and hydrogen bonding interactions of Gleevec inside the active sites of both Abl and c-Kit are virtually similar. However, there

Regan and co-workers have found a highly potent and selective inhibitor (compound (10), Fig. 6) that, instead of targeting the ATP binding site, binds to the hydrophobic pocket (1KV1), and the inhibitor binding requires a large conformational change in the DFG motif which adopts a 'DFG-out' (inactive) conformation, in which the Phe169 (of the DFG motif) moves (by several Å) to a new position [28]. This movement exposes a large hydrophobic pocket which is occupied by the inhibitor. The urea NH groups form bifurcated hydrogen bonding interactions with Glu71 side chain. Authors have made few changes to compound (10) and identified BIRB 796, a picomolar inhibitor of p38 that showed 12,000-fold increase in binding affinity [28]. The orientation and hydrogen bonding interactions of BIRB 796 (1KV2, Fig. 7b) are more or less similar to those of 10 bound to p38. One of the urea NH groups of 10 forms hydrogen bonding interactions with Glu71 side chain. The tolyl substituent makes favorable interactions inside the hydrophobic pocket, while morpholino substituent forms hydrogen bond with the main chain NH in the hinge region. This hydrogen bond is equivalent to the one made by the N1 atom of the quinazolines discussed above. The conformation of the peptide in the hinge region is virtually similar to that observed in 10p38 despite the fact that peptide bond is flipped between Met109 and Gly110. The DFG motif conformation is also similar to that observed in 10p38. On the other hand, the active site conformation of p38 kinase with BIRB 796 is significantly different from that of SB203590 and 9 bound to p38 (Fig. 7c). A large conformational change is required for the DFG motif to accommodate 12 inside the active site of p38 (1W83, Fig. 7b) [54]. In this case too, the DFG motif adopts a 'DFG-out' conformation. The pyridine N forms hydrogen bonding interactions with the main chain NH of Met109 mimicking the role played by the morpholino group of BIRB 796. In addition, the phenyl ring bearing morpholine group makes favorable contacts inside the hydrophobic pocket, while amide NH group is hydrogen bonded to Glu71 side chain. The orientation and interactions of 12 inside p38 are more or less similar to those of BIRB 796 bound to p38; however, the conformations of the activation loop, linker, and P-loop are significantly different. Consequently, the active site conformation is markedly different from that of SB203590, 9, and BIRB 796 bound to p38 (Fig. 7c).

Cl

N N

N

N

N

N

O

N

S

N H

N

N

N

Gleevec (13) PD173955 (14) P O NH

N

H N S Cl

N N H

N

N

N

N

N

O BMS-354825 (15)

Fig. (8). Chemical structures of Bcr-Abl kinase inhibitors.

N

2757

N

OH AP23464 (16)

OH

O

Cl


Thaimattam et al.

Fig. (9). Stereoview of the overlay of Gleevec bound to Abl (cyan C-atoms and green cartoon) and c-Kit (pink C-atoms and yellow cartoon) as revealed by Xray diffraction studies. Notice the differences in the conformations of the DFG motif and P-loop, and positioning of the helix C in these structures.

H N

H N

N

N N

HN

N

N O

O

H N

H N

HN

O N

O NH

O

HN

O

N 17

18

19 N

Fig. (10). Chemical structures of CDK2 and Aurora kinase inhibitors.

are significant differences in the conformations of the P-loop and DFG motif, and positioning of helix C in these structures, although the same inhibitor is bound (Fig. 9). Interestingly, the DFG motif/activation loop in the inhibitor free crystal structure of c-Kit (1T45) is virtually similar to that of c-Kit with Gleevec [64]. Although Flt-3 is closely related to PDGF receptors and c-Kit, Gleevec does not inhibit Flt-3 [64-65]. On the other hand, Gleevec is a potent inhibitor of Flt-3(F691T) mutant [66]. This is due to the fact that the replacement of Phe691 at the gatekeeper position in the active site of Flt-3(F691T) mutant with a smaller residue (Thr) allows inhibitor access to the hydrophobic pocket. In summary, it may be concluded that the shape of ATP binding cleft depends on the relative orientation of the N- and C-terminal lobes with respect to each other, the conformational flexibility of the DFG motif and P-loop, and/or the movement of the helix C, which in turn depend on whether or not an inhibitor is present, and if so, whether it binds to ATP binding site or the hydrophobic pocket or both. Targeting the Phosphate Groove and the Specificity Surface for Kinase Inhibition The phosphate groove is surrounded by residues from the Ploop, catalytic loop and DFG motif. Inhibitors selective for CDK1 and p38 over CDK2 and ERK2 respectively were identified by utilizing the differences in the phosphate groove between the two enzymes [67-68]. Differences in the specificity surface are less utilized in the design of selective kinase inhibitors. It is a cleft that is exposed to solvent. The amino acid sequence and the conforma-

tion of the hinge lining this surface are not conserved in the kinase family. Aurora and Src kinases contain a single Gly insertion in the hinge region (residue 216 in Aurora A) compared to p38, GSK-3, and CDK2; whereas in Pim1 and PI3K, two residues are inserted [69-70]. These insertions change the shape of both adenine pocket and specificity surface. The differences in this region can also be exploited in the design selective inhibitors. Recent examples of the successful application of such an approach include identification of CDK2 and JNK inhibitors selective over Aurora and p38 respectively, which are discussed below. CDK2 and Aurora An additional residue insertion in the linker region of Aurora A changes conformation of the hinge in such a way that it favors substituents on the adenine mimicking core that protrude out to the solvent region in a planar fashion, while no such constraints are required for binding to CDK2 [71]. While pyrrolopyrazoles bearing planar substituents at the 3-position showed CDK2-Aurora dual activity (eg compound (17), Fig. (10), CDK2 IC50 = 1.08 μM and Aurora A IC50 = 2.14 μM), those with non-planar substituents were CDK2 selective (eg compound (18); CDK2 IC50 = 36 nM and Aurora A IC50 >10 μM). However, compound (19), a pyrrolopyrazole bearing a planar group at the 3-position, instead of showing dual kinase inhibitory activity, selectively inhibited Aurora (CDK2 IC50 >10 μM and Aurora A IC50 = 100 nM) [72]. Differences in the conformational flexibility in the phosphate binding region between the two enzymes may be the cause for this trend. The phenyl ring of 19 which is located in the phosphate binding groove is almost perpendicular to the pyrrolopyrazole (2BMC) [72]. The conserved



lysine (Lys162) that is located directly above the phenyl ring is possibly involved in cation/N-H interactions [47], and is within the hydrogen bonding distance from the carbonyl O-atom of the ligand. JNK Kinase SB203580 (Fig. 6), a known inhibitor of p38 (1A9U), binds to both p38 and JNK3 in a similar fashion [73]. The orientation and hydrogen bonding interactions of the inhibitor are more or less similar in both enzymes. Compound (20) (Fig. 11) too binds to JNK3 in a similar fashion [73]. It is a more potent inhibitor of JNK3 compared to JNK1 (JNK3 IC50 = 0.1 μM and p38 IC50 = 1.2 μM). The aniline ring fits into the hydrophobic pocket as the side chain of the gatekeeper amino acid Met146 makes room to accommodate the inhibitor. The indazole N atoms form hydrogen bonds with the backbone amino acids, Met149, and Glu147, while the anilino NH interacts with Lys93 through a water molecule. Overall kinase selectivity for the compounds in this series was good except for p38. All of them showed similar p38 inhibitory activity. Authors have elegantly exploited the differences in the specificity surface, and identified 21 which showed excellent selectivity for JNK3 over p38 (JNK3 IC50 = 3 nM and p38 IC50 = 900 nM). In the crystal (21JNK3: 2B1P), the 3-carboxylic acid of the inhibitor forms a hydrogen bond with Asn152 which is replaced with Asp112 in p38 kinase. KINASE SELECTIVITY As mentioned earlier, the success of Gleevec in the treatment of chronic myeloid leukemia (CML) and gastrointestinal stromal tumor (GIST) has attracted much attention in the recent past because of its excellent selectivity and ability to bind to a precise inactive conformation of Abl kinase, which led to a popular belief that inactive kinase conformation may be targeted for achieving higher selectivity. The only problem for concern so far has been the emergence of drug-resistant mutants [20]. Structural studies suggest that mutations in the kinase domain cause resistance to the Abl kinase inhibitor by impairing the flexibility of the kinase domain, restricting its ability to adopt a specific inactive conformation required for optimal Gleevec binding [55-57]. Like Gleevec, BAY-43-9006 and BIRB 796 (Fig. 12) too recognize a specific inactive conformation of B-Raf (1UWH) and p38 (1W83) respectively [16, 28]. The structural aspects of drug resistance to protein kinases and countermeasures to address this issue have been discussed by Daub and coworkers [61]. It is believed that drug resistance could be overcome with inhibitors that target the active conformation of Abl [74]. Dual AblSrc inhibitors such as BMS-354825, AZD-0530 and SKI-606 are reported to inhibit the active form of Abl kinase [75-77]. In addition, structural data suggest that kinase inhibitors targeting active form are also likely to bind to inactive forms of the same kinase [55-56]. However, mutations in Bcr-Abl kinase capable of conferring resistance to BMS-354825 have been identified, al-

though number of such occurrences is much less [58]. Griffin and co-workers have identified a novel selective inhibitor of Bcr-Abl, AMN107, which is significantly more potent than Gleevec, and active against a number of Gleevec-resistant Bcr-Abl mutants [78]. It shows high selectivity for Abl, c-Kit, and PDGFR when tested against a panel of 15 distinct kinases. Like Gleevec, it also recognizes a virtually similar inactive conformation of the Abl kinase. The differential activity of AMN107 is probably due to increased binding affinity through a number of weak and non-directional forces, particularly in the hydrophobic pocket, which makes it insensitive to conformational changes in the protein due to mutations, while the situation is quite opposite in case of Gleevec binding in which strong directional forces dominate the inhibitorenzyme interactions [78]. This indicates that AMN107 may be slightly more promiscuous than Gleevec, but it needs to be confirmed experimentally. Recently Lockhart and co-workers have shown that none of the kinase inhibitors which are in the clinical trials or clinical use, including Gleevec, was really selective when tested against a panel of 113 distinct kinases [13]. It is pertinent to mention that for most of the compounds specificity has been determined against small sets of kinases. Authors have described a novel experimental approach to assess the specificity of kinase inhibitors that directly and quantitatively measures binding to ATP site [13]. These investigations revealed many surprises, including tight binding of the p38 inhibitor BIRB 796 to a Gleevec-resistant form of the Abl (T315I), and binding of Gleevec to Lck. Compounds such as Gleevec, Lapatanib and Vatalanib bind very few off-target kinases, while others bind several off-target kinases (Table 1) in some cases with an affinity similar to that with which they bind their primary targets. Vatalanib, a potent VEGF receptor tyrosine kinase inhibitor, has shown encouraging results in colorectal cancer. X-ray crystal structure of AAL993, structurally related to Vatalanib, bound to VEGF-R2 revealed that it binds to an inactive form of the enzyme [79]. This binding mode, partly similar to that of Gleevec bound to Abl kinase, is probably responsible for high selectivity of AAL993 [55, 79]. BAY-43-9006 and BIRB 796 bind to several off-target kinases, and are not as selective as Gleevec, although they bind to a specific inactive conformation of their primary targets, B-Raf and p38 respectively (1UWH and 1KV2 respectively) [16, 28]. On the other hand, Tarceva (1M17) and Iressa show good selectivity in spite of the fact that they bind to the active form of EGFR [42]. SP600125, a small planar molecule, binds to several off-target kinases (Table 1). It is a potent inhibitor of JNK (1PMV), and does not inhibit p38 kinase, although several known p38 inhibitors were shown to possess JNK inhibitory activity as well [13, 80-81]. These observations indicate that the understanding of kinase selectivity and the utilization of such understanding in the design of selective ATPcompetitive inhibitors based on structural data alone is quite challenging.

Cl NH

H N

NH

N HN N

N

O N H

O

20 HO Fig. (11). Chemical structures of JNK inhibitors.

2759

21


Thaimattam et al.

Cl

Cl

HN

O

Cl

O O

HN

O

O

O N

CN

N N

O

N

N

O

N

N SKI-606 (23, Abl-Src)

AZD-0530 (22, Abl-Src) O Cl

O

F3C

N H

N

N

O

N H N

N H

N

N

H N

N H

N

O CF3

AMN107 (25, Bcr-Abl)

BAY-43-9006 (24, B-Raf) Cl HN

HN

CF3 N

O

N N

N

Vatalanib (26, VEGFR2)

F

H

N N H

O

N

N

Cl

N

N

O

Cl

PD180970 (29, Abl-Src)

SP600125 (28, JNK)

AAL993 (27, VEGFR2)

N

Fig. (12). Chemical structures of Bcr-Abl, B-Raf, VEGFR2, and JNK inhibitors. Table 1.

List of Some Known ATP-Competitive Kinase Inhibitors and Their Targets

Kinase Inhibitor

Number of kinase targetsa

BIRB 796 Gleevec Bay-43-9006 GW-572016 Vatalanib

Primary Target/s

PDB Code

Binding Conformation

17

p38

1KV2

inactive

4

Bcr-Abl

1IEP

inactive

15

B-Raf

1UWH

inactive

2

EGFR

1XKK

inactive

4

VEGFR2

-

inactive

Tarceva

9

EGFR

1M17

active

Iressa

5

EGFR

-

activeb

ZD-6474

20

VEGFR2,EGFR

-

activeb

SB203580

7

p38

1A9U

inactive

VX-680

34

Aurora

2F4J (Abl)

active

c

Abl-Src

1M52 (Abl)

active

JNK

1UKI (JNK1) 1PMV (JNK3)

active

BMS-354825

58

SP600125

17

a Binding constant 1μM [13, 58]; bBased on modeling studies (note: kinase inhibitors targeting active form are also likely to bind to inactive form/s of the same kinase); cTested against a panel of 148 kinases [58] and all others tested against a panel of 119 kinases [13].

We have examined the torsion angle distribution of the DFG motif residues in both active and inactive forms of a few selected kinases. Only those structures which were confirmed to be active or

inactive were considered in the analysis, including those analyzed recently by Ghosh and co-workers [23]. Active kinases contained the activation loop in an extended conformation, while inactive



forms adopted a folded conformation. The set of active kinases included 38 crystal structures from 26 distinct enzymes. The inactive set contained 47 crystal structures from 31 different kinases (Table 2). The torsion angels are clustered in three distinct regions of the - plot for the active kinases (Fig. 13a). Interestingly, the torsion angles of the Gly were also restricted. These observations indicate that in active kinases DFG motif adopts a specific conformation. The orientation of the Asp residue form the DFG motif is essential for stabilizing the ATPkinase complex through Mg2+ ion coordination. In contrast, there are no such restrictions on the DFG motif for inactive kinases, and therefore it can adopt several distinct conformations, thereby influencing the active site conformation (Fig. 13b). For example, there are at least four distinct active site conformations for p38 kinase (out of 40 structures deposited in the PDB) in which the activation loop resembles that of an inactive kinase (see above, Fig. 7). The shape of the active site is influenced by several factors which include access to allowed conformational states for the linker, P-loop, and DFG motif, which in turn depends on whether or not inhibitor targets the ATP binding pocket or the Table 2.

2761

hydrophobic pocket or both. Therefore, targeting an inactive conformation that is quite distinctive across the kinase family could be more advantageous for achieving higher selectivity. However, it is pertinent to mention here that understanding conformational flexibility of a protein from structural data and utilization of such an understanding in the design of selective inhibitors is quite challenging. Protein flexibility involves small conformational changes due to the movement of backbone and side-chain atoms, and/or largescale molecular motions, in which parts of the protein move as rigid bodies with respect to one another. Molecular dynamics simulations can be applied to explain protein structure flexibility [82]. A more practical approach to address this issue would be to determine the true selectivity of kinase inhibitors by screening them against a large number of kinase targets, and utilize the information to guide structural studies [13]. Such an approach has been successfully applied to identify inhibitors effective against Gleevec-resistant Abl(T315I) mutant. It is the one of the most frequently detected mutations in Gleevecresistant CML. T315I substitution at the gatekeeper position of Bcr-

List of Active and Inactive Kinase Crystal Structures Considered in the Torsion Angle Analysis of the DFG Motif Residues Active

Inactive

Kinase

PDB

Kinase

PDB 1UWH

Csk

1K9A

B-RAF

Lck

3LCK

IRK

1IRK

IRK

1GAG, 1IR3, 1RQQ

c-Src

2SRC, 2PTK, 1FMK, 1KSW

IGFR

1K3A

HCK

1QCF, 1AD5

EGFR

1M17, 1M14

c-KIT

1T45, 1T46

MAPK

1NXK

Titin

1TKI

Aurora A

1OL5, 1MQ4

CaMK-I

1A06

CDK2

1QMZ, 1GY3

Aurora A

2C6E, 1MUO

Syk

1XBA, 1XBB, 1XBC

PKB

1MRY, 1MRV

PKB

1O6K, 1O6L

GRK2

1OMW

ACK 1

1U46, 1U4D, 1U54

c-Abl

1OPJ, 1IEP, 1OPK, 1OPL

c-Kit

1PKG

CDK2

1HCK

PKA

1FMO, 1L3R, 1ATP

p38

1P38, 1KV1, 1KV2, 1W82, 1W83, 1WBS, 1WBV

PDK1

1H1W

JNK1

1UKH

p38

1CM8

JNK3

1JNK

VEGFR2

1VR2

Twitchin

1KOB 1NY3

Phk

2PHK

MAPKAPK2

Chk1

1IA8

TGF R1

1B6C

CK1

1CKJ

BTK

1K2P

CDK5

1H4L

CSK

1BYG

CDK6

1JOW

EphA2

1MQB

CK2

1DAW

EphB2

1JPA 1FGK

PknB

1MRU

FGFR1K

DAPK

1JKK

Flt-3

1RJB

SKY1P

1HOW

IGF1RK

1M7N

ERK2

2ERK

Tie2

1FVR

Musk

1LUF

FAK

1MP8

PAK1

1F3M

ERK2

1ERK

c-Met

1R1W


(a)

Thaimattam et al.

(b)

Fig. (13). - Angle distribution of the DFG motif residues in (a) active, and (b) inactive kinases. Notice that these angles are narrowly clustered in three distinct regions for the active kinases, while more scattered for inactive forms. Negative values of Asp residue (in active kinases) correspond to those structures in which the peptide bond of the preceding residue is flipped without disrupting the DFG motif conformation.

Fig. (14). Stereoview of the overlay of (a) Gleevec (green C-atoms and green cartoon) and (b) VX-680 (pink C-atoms and orange cartoon) bound to Abl as revealed by their crystal structures. Unlike Gleevec, VX-680 binds in such a way that it avoids steric clashes with the gatekeeper residue Thr315.

Abl directly interferes with Gleevec binding [83]. Although several compounds have been shown to inhibit some of the Gleevecresistant Abl forms, none of them was effective against Abl(T315I). The affinity of BMS-354825 and PD180970 is lowered by up to 80fold relative to the wild type Abl [58]. One way to override the drug-resistance to Abl(T315I) mutant is to identify inhibitors that do not depend on this structural element for binding by testing a set of known kinase inhibitors against a large number of kinase targets [13]. Kinase interaction maps for BIRB 796 and VX-680 reveal that they bind to several off-target kinases (Table 1), and more importantly, both of them inhibit the Gleevec- and BMS-354825-resistant Abl(T315I) mutant (Kd = 40 nM and 5 nM respectively) [13, 58]. VX-680 binds to Abl in such a way that it avoids close-contacts with the gatekeeper residue (2F4J, Fig. 14) [84]. CONCLUSIONS Modulation of kinase activity for therapeutic purposes still needs to be exploited. Inhibition of a single kinase alone may not be

sufficient to achieve a therapeutic benefit, while promiscuous small-molecule kinase inhibitors or cocktails of inhibitors that target several kinases may be more promising than selective agents. On the other hand, simultaneous inhibition of multiple targets might lead to adverse side effects. Crystal structures provide insights into selectivity and mechanisms of inhibition. Majority of kinases in active form have more or less similar active site conformation, while inactive states can adopt several distinct active site conformations. Protein rearrangements involving both main and side chain atoms play an important role in the ligand recognition. The gatekeeper residue and conformation of the DFG motif control inhibitor access to the hydrophobic pocket. Targeting a precise inactive kinase conformation that is quite unique across the family, and manipulating auxiliary inhibitorenzyme interactions could be more advantageous, both with respect to selectivity and for prevention of inhibitor resistance. However, it is pertinent to mention that binding specificity cannot be predicted on the basis of structural information alone. Experiments must be carried out to assess mo-



lecular specificity and identify off-target interactions, and this information should be utilized to guide lead optimization.

[23]

ACKNOWLEDGEMENTS We acknowledge Dr. N. Selvakumar and Dr. R. Rajagopalan for their support and encouragement.

[24]

REFERENCES References 85-87 are related articles recently published in Current Pharmaceutical Design.

[26]

[1]

[27]

[2] [3]

[4] [5] [6]

[7]

[8] [9]

[10]

[11]

[12]

[13]

[14] [15] [16]

[17]

[18]

[19] [20]

[21]

[22]

Dancey J, Sausville EA. Issues and progress with protein kinase inhibitors for cancer treatment. Nat Rev Drug Discov 2003; 2: 296313. Cohen P. Protein kinases-the major drugs targets of the twenty-first century? Nat Rev Drug Discov 2002; 1: 309-15. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science 2002; 298: 1912-34. Sawyers C. Targeted cancer therapy. Nature 2004; 294(432): 29497. Dowell J, Minna DJ, Kirkpatrick P. Erlotinib hydrochloride. Nat rev Drug Discov 2005; 4: 13-14. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The protein data bank. Nucl Acids Res 2000; 28(1): 23542. Stout JT, Foster GP, Matthews JD. High-throughput structural biology in drug discovery: Protein kinase. Curr Pharm Design 2004; 10: 1069-82. Mclnnes C, Fischer MP. Strategies for the design of potent and selective kinase inhibitors. Curr Pharm Design 2005; 11: 1845-63. Fischer MP. The design of drug candidate molecules as selective inhibitors of therapeutically relevant protein kinases. Curr Med Chem 2004; 11: 1563-83. Noble MEM, Endicott AJ, Johnson NL. Protein kinase inhibitors: insights into drug design from structure. Science 2004; 303: 180005. Cherry M, Williams HD. Recent kinase and kinase inhibitor X-ray structures: Mechanisms of inhibition and selectivity insights. Curr Med Chem 2004; 11: 663-73. Capdeville R, Buchdunger E, Zimmermann J, Matter A. Glivec (STI-571, imatinib), a rationally developed, targeted anticancer drug. Nat Rev Drug Discov 2002; 1: 493-502. Fabian AM, Biggs III HW, Treiber KD, Atteridge EC, Azimioara DM, Lockhart JD, et al. A small molecule-kinase interaction map for clinical kinase inhibitors. Nat Biotechnol 2005; 23(3): 329-36. Arora A, Scholar ME. Role of tyrosine kinase inhibitors in cancer therapy. J Pharmacol Exp Ther 2005; 315(3): 971-79. Biscardi JS, Tice DA, Parsons SJ. c-Src, receptor tyrosine kinases, and human cancer. Adv Cancer Res 1999; 76: 61-119. Wan CTP, Gamett JM, Roe MS, Lee S, Barford D, Marais R, et al. Mechanism of activation of the RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell 2004; 116: 855-67. Harrington AE, Bebbington D, Moore J, Rasmussen KR, Golec CMJ, Miler MK, et al. VX-680, a potent and selective smallmolecule inhibitor of the Aurora kinases, suppresses tumor growth in vivo. Nat Med 2004; 10(3): 262-67. Murphy E, Steenbergen C. Inhibition of GSK-3 as a target for cardioprotection: the importance of timing, location, duration and degree of inhibition. Expert Opin Ther Targets 2005; 9(3): 447-56. Vlahos CJ, McDowell SA, Clerk A. Kinases as therapeutic targets for heart failure. Nat Rev Drug Discov 2003; 2: 99-113. Hochhaus A, La Rosee P. Imatinib therapy in chronic myleogenous leukemia: Strategies to avoid and overcome resistance. Leukemia 2004; 18: 1321-31. Paez JG, Janne PA, Lee JC, Tracy S, Greulich H, Gabriel S, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 304 (5676): 1497-500. Dowell J, Minna DJ, Kirkpatrick. Erlotinib hydrochloride. Nature Rev Drug Discov 2004; 4: 13-14.

[25]

[28]

[29]

[30] [31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

2763

Nolen B, Taylor S, and Ghosh G. Regulation of protein kinases: controlling activity through activation segment conformation. Mol Cell 2004; 15: 661-75. Johnson LN, Noble ME, Owen DJ. Active and inactive protein kinases: structural basis for regulation. Cell 1996; 85: 149-58. Hubbard S. Protein tyrosine kinases: autoregulation and small molecule inhibition. Curr Opin Struc Biol 2002; 12: 735-741. Knighton DR, Zheng J, Ten Eyck LF, Ashford V, Xuong NH, Taylor SS, et al. crystal structure of the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science 1991; 253: 407-14. Gumireddy K, Baker JS, Cosenza CS, John P, Kang DA, Robell AK, et al. A non-ATP-competitive inhibitor of BCR-ABL overrides imatinib resistance. Proc Natl Acad Sci USA 2005; 102(6): 1992-97. Pargellis C, Tong l, Churchill L, Cirillo FP, Gilmore T, Graham GA, et al. Inhibition of p38 Map kinase by utilizing a novel allosteric binding site. Nat Struct Biol 2002; 9(4): 268-72. Ohren JF, Chen H, Pavlovsky A, Whitehead C, Zhang E, Kuffa P, et al. Structures of human MAP kinase kinase 1 (MEK1) and MEK2 describe novel noncompetitive kinase inhibition. Nat Struct Mol Biol 2004; 11(12): 1192-97. Bogoyevitch AM. Therapeutic promise of JNK ATPnoncompetitive inhibitors. Trends Mol Med 2005; 11(5): 232-39. Heo SY, Kim KS, Seo CII, Kim KY, Sung JB, Yang HC, et al. Structural basis for the selective inhibition of JNK1 by the scaffolding protein JIP1 and SP600125. EMBO J 2004; 23(11): 2185-95. Barlaam B, Fennell M, Germain H, Green T, Hennequin L, Costello G, et al. New heterocyclic analogues of 4-(2-chloro-5methoxyanilino)quinazolines as potent and selective c-Src kinase inhibitors. Bioorg Med Chem Lett 2005; 15(24): 5446-49. Heron MN, Anderson M, Blowers PD, Breed J, Eden MJ, Green S, et al. SAR and inhibitor complex structure determination of a novel class of potent and specific Aurora kinase inhibitors. Bioorg Med Chem Lett 2006; 16(5):1320-23. Jung HF, Pasquet G, Brempt CL, Lohmann MJ, Warin N, Renaud F, et al. Discovery of novel and potent thiazoloquinazoline as selective Aurora A and B kinase inhibitors. J Med Chem 2006; 49(3): 955-70. Cumming GJ, McKenzie LC, Bowden GS, Campbell D, Masters JD, Breed J, et al. Novel, potent and selective anilinoquinazoline and anilinopyrimidine inhibitors of p38 MAP kinase. Bioorg Med Chem Lett 2004; 14(21): 5389-94. Ballard P, Bradbury HR, Harris SC, Hennequin FAL, Hickinson M, Johnson DP, et al. Inhibitors of epidermal growth factor receptor tyrosine kinase: Novel C-5 substituted anilinoquinazolines designed to target the ribose pocket. Bioorg Med Chem Lett 2006; 16(6): 1633-37. Hennequin FL, Stokes SE, Thomas PA, Johnstone C, Wedge SR, Kendrew J, et al. Novel 4- anilinoquinazolines with C-7 basic side chains: Design and structure activity relationship of a series of potent, orally active, VEGF receptor tyrosine kinase inhibitors. J Med Chem 2002; 45(6): 1300-12. Ballard P, Bradbury HR, Hennequin FL, Hickinson MD, Klinowska T, Morgentin R, et al. 5-Substituted 4-anilinoquinazolines as potent, selective and orally active inhibitors of erbB2 receptor tyrosine kinase. Bioorg Med Chem Lett 2005; 15(19): 4226-29. Hirai H, Kawanishi N, Iwasawa Y. Recent advances in the development of selective small molecule inhibitors for cyclin-dependent kinases. Curr Top Med Chem 2005; 5(2): 169-79. Mortlock A, Keen JN, Jung HF, Heron MN, Foote MK, Wilkinson R, et al. Progress in the development of selective inhibitors of Aurora kinases. Curr Top Med Chem 2005; 5(2): 199-213. Zhang Y, Cockerill S, Guntrip S, Rusnak D, Wood E, Lackey K, et al. Synthesis and SAR of potent EGFR/erbB2 dual inhibitors. Bioorg Med Chem Lett 2004; 14(1): 111-14. Stamos J, Sliwkowski, MX, Eigenbrot C. Structure of the epidermal growth factor receptor kinase domain alone and in complex with a 4-anilinoquinazoline inhibitor. J Biol Chem 2002; 277: 46265-72. Kubo K, Shimizu T, Ohyama S, Murooka H, Nishitoba T, Kobayashi Y, et al. A novel series of 4-phenoxyquinolines: potent and


[44]

[45]

[46]

[47] [48]

[49] [50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61] [62]

[63]

highly selective inhibitors of PDGF receptor autophosphorylation. Bioorg Med Chem Lett 1997; 7(23): 2935-40. Hennequin LF, Thomas AP, Johnstone C, Lohmann JJ, Ogilvie DJ, Dukes M, et al. Design and structure-activity relationship of a new class of potent VEGF receptor tyrosine kinase inhibitors. J Med Chem 1999; 42(26): 5369-89. Shewchuk L, Hassell A, Wisely B, Rocque W, Holmes W, Veal J, Kuype FL. Binding mode of the 4-anilinoquinazoline class of protein kinase inhibitor: X-ray crystallographic studies of 4anilinoquinazolines bound to cyclin-dependent kinase 2 and p38 kinase. J Med Chem 2000; 43(1): 133-8. Thaimattam R, Daga PR, Banerjee R, Iqbal J. 3D-QSAR studies on c-Src kinase inhibitors and docking analyses of a potent dual kinase inhibitor of c-Src and c-Abl kinases. Bioorg Med Chem 2005; 13(15): 4704-12. Desiraju GR, Steiner T. The weak hydrogen bond in structural chemistry and biology. Oxford University Press: Oxford 1997. Wood RE, Truesdale TA, McDonald BO, Yuan D, Hassell A, Dickerson HS, et al. Unique structure for epidermal growth factor receptor bound to GW572016 (Lapatinib): Relationships among protein conformation, inhibitor off-rate, and receptor activity in tumor cells. Cancer Res 2004; 64(18): 6652-59. Keen N, Taylor S. Aurora-kinase inhibitors as anticancer agents. Nat Rev 2004, 4: 927-36. Lee MR, Dominguez C. MAP kinase p38 inhibitors: Cinical results and an intimate look at their interactions with p38 alpha protein. Curr Med Chem 2005; 12(25): 979-94. Peifer C, Wagner G, Laufer S. New approaches to the treatment of inflammatory disorders small molecule inhibitors of p38 MAP kinase. Curr Top Med Chem 2006; 6: 113-49. Wang Z, Canagarajah BJ, Boehm JC, Kassisa S, Cobb MH, Young PR, et al. Structural basis of inhibitor selectivity in MAP kinases. Structure 1998; 6: 1117-28. Fitzgerald CE, Patel SB, Becker JW, Cameron MP, Zaller D, Pikounis BV, et al. Structural basis for p38 MAP kinase quinazolinone and pyridol-pyrimidine inhibitor specificity. Nat Struct Biol 2003; 10(9): 764-69. Gill AL, Frederickson M, Cleasby A, Woodhead SJ, Carr M G, Woodhead AJ, et al. Identification of novel p38 MAP kinase inhibitors using fragment-based lead generation. J Med Chem 2005; 48: 414-26. Nagar B, Bornmann WG, Pellicena P, Schindler T, Veach DR, Miller WT, et al. Crystal structures of the kinase domain of c-Abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res 2002; 62: 4236-43. Schindler T, Bornmann W, Pellicena P, Miller TW, Clarkson B, Kuriyan J. Structural mechanism of STI-571 inhibition of abelson tyrosine kinase. Science 2000; 289: 1938-42. Nagar B, Bornmann GW, Pellicena P. Crystal structures of the kinase domain of c-Abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res 2002; 62: 4236-43. Carter AT, Wodicka ML, Shah PN, Velasco MA, Fabian AM, Lockhart JD, et al. Inhibition of drug-resistant mutants of ABL, KIT, and EGF receptor kinases. Proc Natl Acad Sci USA 2005; 102(31): 11011-16. Corbin AS, La Rosee P, Stoffregen EP, Druker BJ, Deininger MW. Several Bcr-Abl kinases mutants associated with imatinib mesylate resistance remain sensitive to imatinib. Blood 2003; 101: 4611-14. Shah NP, Nicoll JM, Nagar B, Gorre ME, Paquette RL, Kuriyan J, Sawyers CL. Multiple BCR-ABL kinase domain mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia. Cancer Cell 2002; 2: 117-25. Daub H, Specht K, Ullrich A. Strategies to overcome resistance to targeted protein kinase inhibitors. Nat Rev 2004; 3: 1001-10. Shah NP, Tran C, Lee FY, Chen P, Norris D, Saywers CL. Overriding imatinib resistance with a novel ABL kinase inhibitor. Science 2004; 305(5682): 399-401. Hare T, Pollock R, Stoffregen PE, Keats JA, Abdullah OM, Moseson EM, et al. Inhibition of wild-type and mutant Bcr-Abl by AP23464, a potent ATP-based oncogenic protein kinase inhibitor: implications for CML. Blood 2004; 104: 2532-39.

Thaimattam et al. [64]

[65]

[66]

[67] [68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78]

[79]

[80]

[81]

Mol DC, Dougan RD, Schneider RT, Skene JR, Sang B, Wilson PK et al. Structural basis for the autoinhibition and STI-571 inhibition of c-Kit tyrosine kinase. J Biol Chem 2004; 279(30): 3165563. Griffith J, Black J, Faerman C, Swenson L, Wynn M, Lu F, et al. The structural basis for autoinhibition of FLT3 by the juxtamembrane domain. Mol Cell 2004; 13(2): 169-78. Bohmer DF, Karagyozov L, Uecker A, Serve H, Botzki A, Mahboobi S, et al. A single amino acid exchange inverts susceptibility of related receptor tyrosine kinases for the ATP site inhibitor STI571. J Biol Chem 2003; 278(7): 5148-55. Mclnnes C, Fischer MP. Strategies for the design of potent and selective kinase inhibitors. Curr Pharm Design 2005; 11: 1845-63. Wang Z, Canagarajah BJ, Boehm JC, Kassisa S, Cobb MH, Young PR, et al. Structural basis of inhibitor selectivity in MAP kinases. Structure 1998; 6: 1117-28. Kumar A, Mandiyan V, Suzuki Y, Artis RD, Ibrahim P, Bremer R, et al. Crystal structures of proto-oncogene kinase Pim1: A target of aberrant somatic hypermutations in diffuse large cell lymphoma. J Mol Biol 2005; 348: 183-93. Jacobs DM, Black J, Futer O, Swenson L, Hare B, Fleming M, et al. Pim-1 ligand-bound structures reveal the mechanism of serine/threonine kinase inhibition by LY294002. J Biol Chem 2005; 280(14): 13728-34. Pevarello P, Fancelli D, Vulpetti A, Amici R, Villa M, Pittala V, et al. 3-Amino-1,4,5,6-tetrahydropyrrolo[3,4-C]pyrazoles: A new class of CDK2 inhibitors. Bioorg Med Chem Lett 2006; 16(4): 1084-90. Fancelli D, Berta D, Bindi S, Cameron A, Cappella P, Carpinelli P, et al. Potent and selective Aurora inhibitors identified by the expansion of a novel scaffold for protein kinase inhibition. J Med Chem 2005; 48: 3080-84. Swahn MB, Huerta F, Kallin E, Malmstro J, Malmstrom J, Weigelt T, et al. Design and synthesis of 6- anilinoindazoles as selective inhibitors of c-Jun N-terminal kinase-3. Bioorg Med Chem Lett 2005; 15(22): 5095-9. Cowan-Jacob SW, Guez V, Griffin DJ, Fabbro WP, Manley WP, Fabbro D, et al. Imatinib (STI-571) resistance in chronic myelogenous leukemia: Molecular basis of the underlying mechanisms and potential strategies for treatment Bcr-Abl kinase mutations and drug resistance to imatinib (STI-571) in chronic myelogenous leukemia. Mini Rev Med Chem 2004; 4: 285-99. Manley WP, Cowan-Jacob WS, Mestan J. Advances in the structural biology, design and clinical development of Bcr-Abl kinase inhibitors for the treatment of chronic myeloid leukaemia. Biochim Biophys Acta 2005; 1754: 3-13. Golas MJ, Arndt K, Etienne C, Lucas J, Nardin D, Boschelli1 F, et al. SKI-606, a 4-anilino-3-quinolinecarbonitrile dual inhibitor of Src and Abl kinases, is a potent antiproliferative agent against chronic myelogenous leukemia cells in culture and causes regression of K562 xenografts in nude mice. Cancer Res 2003; 63: 37581. Lombardo JL, Lee YF, Chen P, Norris D, Barrish CJ, Behnia K, et al. Discovery of N-(2-Chloro-6-methylphenyl)-2-(6-(4-(2-hydroxyethyl)-piperazin-1-yl)-2-methylpyrimidin-4-ylamino)thiazole-5-carboxamide (BMS-354825), a dual Src/Abl kinase inhibitor with potent antitumor activity in preclinical assays. J Med Chem 2004; 47: 6658-61. Weisberg E, Manley WP, Breitenstein W, Bruggen J, Cowan-Jacob WS, Griffin DJ, et al. Characterization of AMN107, A selective inhibitor of native and mutant Bcr-Abl. Cancer Cell 2004; 7: 129-41. Manley WP, Bold G, Bruggen J, Fendrich G, Furet P, Mestan J, et al. Advances in the structural biology, design and clinical development of VEGF-R kinase inhibitors for the treatment of angiogenesis. Biochim Biophys Acta 2004; 1697: 17-27. Bennett LB, Sasaki T, Murray WB, O'Leary CE, Sakata TS. SP600125, an anthrapyrazolone inhibitor of Jun N-terminal kinase. Proc Natl Acad Sci USA 2001; 98(24): 13681-86. Scapin G, Patel SB, Lisnock J, Becker JW, LoGrasso P. The structure of JNK3 in complex with small molecule inhibitors: Structural basis for potency and selectivity. Chem Biol 2003; 10(8): 705-12.

Protein Kinase Inhibitors [82]

[83]

[84]

Allen TW, Andersen OS, Roux B. On the importance of atomic fluctuations, protein flexibility, and solvent in ion permeation. J Gen Physiol 2004; 124: 679-90. Burgess RM, Skaggs JB, Shah PN, Lee FY, Sawyers LC. Comparative analysis of two clinically active BCR-ABL kinase inhibitors reveal the role of conformation-specific binding in resistance. Proc Natl Acad Sci USA 2005; 102(9): 3395-00. Young AM, Shah PN, Chao HL, Seeliger M, Milanov VZ, Biggs HW, et al. Structure of the kinase domain of an imatinib-resistant


[85] [86]

[87]

2765

Abl mutant in complex with the Aurora kinase inhibitor VX-680. Cancer Res 2006; 66(2): 1007-14. Medinger M, Drevs J. Receptor tyrosine kinases and anticancer therapy. Curr Pharm Des 2005; 11(9): 1139-49. Pumfery A, de la Fuente C, Berro R, Nekhai S, Kashanchi F, Chao SH. Potential use of pharmacological cyclin-dependent kinase inhibitors as anti-HIV therapeutics. Curr Pharm Des 2006; 12(16): 1949-61. Schang LM. Herpes simplex viruses in antiviral drug discovery. Curr Pharm Des 2006; 12(11): 1357-70.