Drug discovery and the human kinome: Recent trends

Pharmacology & Therapeutics 130 (2011) 144–156

Contents lists available at ScienceDirect

Pharmacology & Therapeutics j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p h a r m t h e r a

Drug discovery and the human kinome: Recent trends Richard Eglen ⁎, Terry Reisine 1 Bio-discovery, 940 Winter St., Waltham, MA 02451-1457, United States

a r t i c l e

i n f o

Keywords: Allosterism Protein–protein interactions Kinase translocation Growth factor agonists/antagonists

a b s t r a c t A major new trend in drugs targeted at protein kinases is the discovery of allosteric modulators. These compounds differ from ATP-centric drugs in that they do not compete with ATP for binding to the catalytic domain, generally acting by inducing conformational changes to modulate activity. They could provide a number of advantages over more classical protein kinase drugs. For example, they are likely to be more selective, since they bind to unique regions of the kinase and may be useful in overcoming resistance that has developed to drugs that compete with ATP. They offer the ability of activating the kinases either by removing factors that inhibit kinase activity or by simply producing changes to the enzyme to foster catalytic activity. Furthermore, they provide more subtle modulation of kinase activity than simply blocking ATP access to inhibit activity. One hurdle to overcome in discovering these compounds is that allosteric modulators may need to inhibit protein–protein interactions; generally difficult to accomplish with small molecules. Despite the technical problems of identifying allosteric modulators, major gains have been made in identifying allosteric inhibitors and activators of the growth factor receptors as well as soluble tyrosine and serine/threonine kinases and some of these drugs are now in various stages of clinical trials. This review will focus on the discovery of novel allosteric modulators of protein kinases and drug discovery approaches that have been employed to identify such compounds. © 2011 Elsevier Inc. All rights reserved.

Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Discovering small molecules that act like large proteins . . . . . 3. Allosteric modulators of soluble protein kinases . . . . . . . . . 4. Assays to discover novel allosteric modulators of protein kinases . 5. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

144 145 149 146 153 148 154 149 154 149 154 150

1. Introduction

Abbreviations: ATP, adenosine triphosphate; AKT, v-akt murine thymoma viral oncogene homolog; BRAF, v-raf murine sarcoma viral oncogene homolog B1; BDNF, brain-derived neurotrophic factor; CNS, central nervous system; EGF, epidermal growth factor; EPO, erythropoietin; ERK, extracellular-signal-regulated kinase; FRET, Förster resonance energy transfer; IL, interleukin; JAK-STAT, Janus kinase-signal transducer and activator of transcription; MAP kinase, mitogen-activated protein kinase; NGF, nerve growth factor; NMR, nuclear magnetic resonance; PDK1, PDK1, 3phosphoinositide-dependent kinase-1. ⁎ Corresponding author at: Bio-discovery, PerkinElmer, 940 Winter St., Waltham, MA 02451-1457, United States. Tel.: 781 663 5599; fax: 781 663 5984. E-mail address: [email protected] (R. Eglen). 1 Terry Reisine, PhD is an independent consultant. 0163-7258/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.pharmthera.2011.01.007

Protein kinases are a family of enzymes involved in signal transduction in every human cell. The enzymes detect both external and internal stimuli to cells and produce their functions by phosphorylating proteins. This process initiates and propagates information flow to allow cells to respond to their changing environment. This family of proteins is essential for normal physiology and when dysfunctional leads to abnormal cellular activity and disease. The protein kinase family is a major target for drug discovery by the pharmaceutical industry (Simpson et al., 2009; Eglen & Reisine, 2009, 2010). A large number of protein kinase inhibitors are either in clinical development or have been approved for marketing by the FDA to treat a wide variety of diseases including cancer, inflammation, diabetes, immunodeficiency and CNS disorders (see Eglen & Reisine,

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

2009, 2010). Many of these drugs have improved survival and quality of life of cancer patients as well as in individuals suffering from other complications. In general, kinase inhibitors have been classified into four different types based on their mechanisms of action (see Zhang et al., 2009; Eglen & Reisine, 2010). Type 1 inhibitors work via a classical mechanism of action to block the access of ATP to the catalytic domain of the kinase in a competitive manner. Type 2 inhibitors interact with kinases in a somewhat different manner, specifically binding to the inactive form of the kinase to prevent the activation process, much in the manner of Gleevec. Type 3 inhibitors, which will be the major focus of this review act via allosteric sites to block the activity. As defined by Zhang et al. (2009) and others, allosteric modulators interact with “A site distinct from the enzyme active site... [to] regulate[s] enzyme activity”. This can mean interacting with sites near the active site but not within the catalytic domain or more distal sites such as regions involved in the regulatory subunit interaction with the catalytic domain of the cAMP dependent protein kinase, or even the growth factor binding sites on the N-terminal transmembrane domain receptors that affect conformational changes in the Cterminal catalytic domain to affect enzyme activity. Thus, allosterism encompasses a large gamut of mechanisms of kinase regulation. Finally, type 4 inhibitors are primarily covalent inhibitors of kinases that target active sites. While much of the classical discovery of the protein kinase drugs has targeted regions around the ATP binding sites to identify inhibitors, emerging trends have focused on identifying allosteric modulators. The interest in developing allosteric modulators is that such drugs may provide unique advantages over more classically developed compounds (Li et al., 2004; Noble et al., 2004; Zhang et al., 2009; Eglen & Reisine, 2009, 2010). First, they offer the possibility of greater selectivity because they target sites in kinases more unique in sequence and structure than those compounds that bind to the regions near the ATP binding domain, which are generally more conserved amongst protein kinases. Greater selectivity might be expected to reduce the sideeffects compared to the more pervasive kinase inhibitors. The selectivity of the allosteric modulators can also provide approaches to differentially regulate the subtypes of a kinase within a subfamily, which may have similar or the same substrates and high overall sequence similarity. Secondly, allosteric modulators hold a promise in selectively targeting mutant forms of disease causing protein kinases and in overcoming resistance of the kinases to the ATP binding competitive drugs. As described elsewhere (Cohen, 2002; Bardelli et al., 2003; Dancy & Sausville, 2003; Noble et al., 2004; Pearson et al., 2006; Zhang et al., 2009), many diseases, notably proliferative diseases such as cancer, are caused in part or completely by mutations that generate constitutive kinase activity. The mutations can change the conformation of the kinase and drugs that selectively interact with the mutant form of the kinase may block the activity of the disease causing enzyme while having less or no effect on the natural form of enzyme preserving normal function. For example, this has been shown to be the case for the serine/ threonine kinase BRAFV600E which causes almost half of the malignant melanomas and is the most common disease causing mutant kinase (Tsai et al., 2008). The mutation causes constitutive activity of the kinase and continuous stimulation of the downstream MAPkinase/ ERK signaling pathway. The small molecule drug PLX4032 (see Fig. 1 for structure) selectively inhibits BRAFV600E and is much less potent against the wild type kinase or any other kinase and blocks the MAPkinase/ERK signaling pathway only in cells expressing the mutant kinase both in vitro and in vivo. This drug selectively targets the disease causing kinase providing an incredible level of specificity over any other protein in the body and this drug. PLX4032 is currently being tested in the clinic and has shown great promise in treating

145

melanomas which have previously not been effectively treated by other drugs (Flaherty et al., 2009). While this drug is not allosteric in action, the allosteric inhibitors are more likely to show this profile because in general they are much more selective in targeting kinase specific conformations than the more classical ATP competitive inhibitors. Thirdly, while most of the drug discovery activities against protein kinases have focused on identifying the inhibitors of kinase activity and ATP binding, allosterism allows for the identification of compounds that could result in activators of kinases. This may be particularly relevant for the family of receptor kinases such as the growth factor receptors, where binding of large proteins to the extracellular domains induces conformational changes that activate the intracellular catalytic activity of the kinase. Some of these growth factors, such as BDNF and NGF, have an important therapeutic value in abrogating neurodegeneration due to a supporting role in neuronal survival and blocking disease progression in Alzheimer's and Parkinson's diseases. Large growth factors are not generally good drug candidates and are difficult to optimize for CNS penetration. Novel technologies have now been developed to allow for the identification of small molecules that bind to similar regions of the growth factors on their receptors and cause kinase activation, providing approaches to identify new growth factor receptor modulators with optimal pharmacokinetic properties. Finally, small molecule allosteric modulators can provide subtle regulation of kinases controlled by multiple endogenous factors. For example, the cyclin dependent kinases (CDK) are regulated by both endogenous protein activators (cyclins) and inhibitors (CDKI) (Roy & Sausville, 2001). Small molecules could affect the balance of CDK control by these endogenous factors to cause cell apoptosis, something not easily done with the ATP-centric drugs. Developing small molecule regulators of the endogenous factors controlling protein kinases can require the use of approaches to identify compounds that inhibit protein–protein interaction. Once considered a difficult, if not impossible approach, numerous examples have in fact become available (White et al., 2008; Arkin & Whitty, 2009). New technologies have been adapted to discover protein– protein inhibitors (PPI) in a high throughput screening (HTS) format, as discussed below. Furthermore, there is good evidence that allosteric sites are ‘druggable’ (Hajduk et al., 2005; Fuller et al., 2009) and that some of the same structure–function analysis the industry has employed to discover ATP binding site inhibitors can actually be used to develop the allosteric regulators. The focus of this review, rather than discussing protein kinase drug discovery as a whole, will attempt to describe the innovations that provide the basis of drug development that targets allosteric regulators. We will be liberal in the use of the term allosteric modulator to encompass factors that affect kinase activity through mechanisms independent of a direct ATP binding site competition to include molecules binding to sites outside of the catalytic domain, such as the growth factors that affect kinase conformation or dimer formation to regulate activity. Importantly, using the knowledge of kinase function and its regulation, new technologies have been developed to exploit the utility of these advances to foster a new generation of drugs for the future that may not only provide advantages over the drugs developed to date, but may lead to new compounds as tools for defining their biological function. 2. Discovering small molecules that act like large proteins Receptor tyrosine kinases (RTKs) are a major subfamily of kinases that mediate the biological effects of many growth factors. Unlike the soluble kinases, this family consists of the integral membrane proteins containing an extracellular domain that binds the growth factors and intracellular domains which contain the tyrosine kinase catalytic activity. The binding of the growth factor to the allosteric regions in

146

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

brain permeability. Clearly, the development of small molecule agonists of BDNF could overcome hurdles in using this protein as a CNS therapeutic to treat brain diseases. BDNF induces its action via the receptor kinase tropomyosinrelated kinase (Trk) receptor B (TrkB)(Chao, 2003). While little information existed on the regions of TrkB involved in BDNF binding and activation, structural information is available on the regions of BDNF involved in TrkB activation. In fact, the peptides modeled to some of the loop regions of BDNF have been made that mimic BDNF activity (O'Leary & Hughes, 2003; Williams et al., 2005). Furthermore chimeric BDNF–NGF proteins revealed discrete regions of BDNF involved in the activation of TrkB, in contrast to TrkA, the receptor that mediates the NGF effects (Ibanez et al., 1991, 1993). Using this information, Massa et al. (2010) developed a pharmacophore model of the regions of BDNF needed for activity, and employed virtual high throughput screening to identify small molecules with structural similarities of the BDNF pharmacophore. They then screened small sets of compounds in low throughput screening for the BDNF activity, first by their ability to promote the survival of primary brain neurons in culture, then by their ability to stimulate the TrkB receptor to induce the phosphorylation of its activation loop. They also tested small molecule allosteric agonists for their selectivity in stimulating TrkB signaling pathways to promote cell survival in vitro. From these studies, a highly effective small molecule allosteric agonist at TrkB was identified, LM22A-4 (see Fig. 1 for structure). It was tested in multiple disease models of neuronal protection and shown to be as effective as BDNF. In particular, LM22A-4 protected animals in a model of traumatic brain injury (TBI), and significantly improved motor performance. This is important because presently there is no treatment of TBI nor is there any effective therapeutic to block neurodegeneration in a host of CNS diseases. Interestingly, TrkB, like other RTKs requires dimerization for full activation, both for the autophosphorylation and optimal catalytic activity. Dimers of TrkB form when two BDNF molecules are bound and crosslinked to the receptor complex. In fact, the requirement for dimer formation is one characteristic of the growth factor receptor signaling that would be expected to present a hurdle in use of the small molecule ligands as growth factor agonists, since how would the

the extracellular domain induces oligomerization of the receptor and conformational changes to increase catalytic activity to induce autophosphorylation of the receptor itself to heighten the catalytic activity as well as the phosphorylation of downstream substrates including transcription factors to affect long term cell activity. The allosteric regions can be employed as targets for the discovery of therapeutics to block activity. In fact, most drug discoveries targeting the RTKs have focused on identifying inhibitors since the over-expression or over-activity of these receptors is associated with a number of proliferative diseases. This is most clearly seen with the EGF receptor (EGFR) which is linked to breast cancer (DiGiovanna et al., 2005). Antibodies have been developed to target the allosteric regions of the extracellular domains of EGFR as therapeutics to block the growth factor activation and in the case of EGFR, antibodies have been developed to block other allosteric sites to prevent the interaction of monomeric forms of the EGFR with other subunits (Dancy & Sausville, 2003; Piccart-Gebhart et al., 2005). Small molecule inhibitors have also been developed such as Iressa (Barker et al., 2001) and Tarceva, (Perez-Soler et al., 2004) but these focus on the catalytic domain of the kinase, not binding to the allosteric sites. Generally, small molecule allosteric inhibitors or activators of these receptor-kinases have not been identified. 2.1. Allosteric activators of growth factor receptors While most therapeutic approaches have attempted to identify drugs that inhibit RTKs, there is a large family of the growth factor receptor kinases for which allosteric activators could have important therapeutic uses. Specifically, BDNF is known to support and facilitate neuronal growth and survival (Kaplan & Miller, 2000; Chao, 2003; Huang & Reichardt, 2003). Loss of BDNF in the brain has been linked to neurodegeneration in a number of diseases including Alzheimer's, Huntington's and Parkinson's diseases (Fumagalli et al., 2006; Zuccato & Cattaneo, 2007; Schindowski et al., 2008). The reversal or attenuation of neuronal loss has been found in the animal models of these diseases after the BDNF treatment. BDNF is not easy to employ as a therapeutic to treat CNS diseases because of the proteins' limited

F O

Cl

F N

N SO2 H

N H

PLX 4032 - CID42611257 NH O

HO

OH

O

O

HO

N H

N H

OH

O N H

O

NH

L-783,281 CID:3013166 OH

LM22A-4 Fig. 1. Structures of allosteric modulators.Structures of a number of allosteric modulators are described. Some structures were taken from the Pubmed database and the CID numbers are included. Others are taken from the articles from which they are published and the references included.

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

147

NH O OH HO O N H

L-783,281 CID:3013166

CO2CH3 HN

O

CF3CO2H

N HN

NH2

Ro26-4550 CID16760522

O

CH3

N N

N H

N HO2C

O

H N

NH NH2

O O

Cl Cl

SP4206 Fig. 1 (continued).

small molecules physically bridge the TrkB subunits. The finding that a small molecule can activate TrkB and induce phosphorylation and downstream signaling either implies that the small molecule can induce conformational changes in the receptor to activate the kinase catalytic domain independent of the dimer formation or somehow the small molecule agonist induces dimerization to activate signaling. In fact, studies by Tian et al. (1998) have indicated that small molecules can induce the dimerization of the growth factor receptors such as the granulocyte colony stimulating factor (G-CSF) receptor to increase downstream signaling pathways.

Importantly, the studies of Massa et al. (2010) suggest that it is possible to discover small molecule allosteric activators of RTKs using a combination of pharmacophore modeling, 3D database and virtual high throughput screening and low throughput functional screening. The work also may provide the foundation for the further development of small molecule BDNF ligands that could have widespread utility in treating neurological disorders. The study implies that there is no reason that the approach could not be employed for other growth factors to identify agonists to provide the development of a new generation of pharmacological agents both to study the function

148

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

O N H

NH

N

N

O

O

Ro 08-2750 CONH2

N HN

N

OCF3

GNF-2 CID5311510

NH2

N O

N HN N

MK-2206 N N N N HN H3C H3C H3C

N

CH3

Akt-I-1 H2N

CH3 CH3

N N

Akt-I-1,2 Fig. 1 (continued).

of these receptors in vivo and potentially to the transition into new potential therapeutics. While studies with BDNF focused on developing ligands based on modeling of the growth factor, Zhang and associates (Zhang et al., 1999; Qureshi et al., 2000; Salituro et al., 2001) identified allosteric activators of the insulin receptor kinase using more conventional receptor screening approaches testing natural product libraries for agonists. The insulin receptor, like other RTKs consists of an extracellular domain

that binds insulin and intracellular domains with the kinase activity. To identify the small molecule demethylasterriquinone B-1 (L-783,281) (see Fig. 1 for structure) and its analogs as allosteric activators of the receptor, Zhang et al. (1999) expressed the recombinant receptor in CHO cells, screened natural products and measured insulin receptor kinase activity as a response. L-783,281 stimulated the activity of the cloned insulin receptor with EC50 values of 3–6 uM. Furthermore, L-783,281 effectively stimulated insulin receptor activity in the liver and activated downstream signaling pathways of the insulin receptor including PI-3-kinase and phosphorylation of the Akt kinase. The allosteric activator had limited or no effect on the IGF-I receptor, EGFR or PDGF receptor kinase activity, suggesting it was selective for insulin receptor activation. Like insulin, L-783,281 stimulated insulin uptake into adipocytes and muscle cells in vitro. More importantly, it lowered glucose in the diabetic db/db mice and corrected the hyperinsulinemia. Also, in the ob/ob mice, a model of obesity in which the mice exhibit extreme hyperinsulinemia and hyperglycemia, L-783,281 suppressed the elevated insulin levels, suggesting that the compound may be effective in treating type 2 diabetes. Mechanism studies suggest that L-783,281 binds to the beta subunit of the insulin receptor containing the tyrosine kinase domain, rather than acting via the extracellular insulin binding domain. This was suggested by studies showing that the small molecule did not displace insulin binding to the receptor, could activate the kinase domain in cells, and activated a receptor chimera consisting of the tyrosine kinase domain linked to an insulin receptor-related receptor that does not respond to insulin. Studies by Qureshi et al. (2000) indicated that these insulin allosteric activators are likely to bind to the inactivate insulin receptor kinase domain to induce conformational changes to remove auto-inhibition of the kinase activity to allow ATP access to the active site. These studies were important because they were one of the first evidences of the small molecule allosteric activators of an RTK. Furthermore, L-783,281 and some of its analogs are orally available providing a potential insulin substitute for the treatment of diabetes. In addition, because of their mode of action in bypassing the extracellular insulin binding domain, these small molecule allosteric activators may be useful in treating some forms of insulin resistance which are known to occur due to either mutations in the extracellular insulin binding domain that prevent insulin binding or because of the reduced expression of the insulin binding domain. In fact, studies by Li et al. (2001) showed that these compounds effectively stimulated mutated forms of the insulin receptor kinase which were no longer responsive to insulin. Lastly, while the method employed was laborious and costly, these studies provide an approach to possibly identify small molecular allosteric activators of other receptor kinases and indicate that direct interaction with the same sites as the growth factor are not necessary for activating these receptor kinases. Thus, in summary, despite the large size of the protein growth factors and potentially large number of sites on the receptor extracellular domains to which they may bind to activate kinases, it is possible to design and discover small molecule agonists at these receptors to mimic the actions of the growth factors. This has shown to be the case for BDNF (Massa et al., 2010), EPO (Qureshi et al., 1999), growth hormone receptor (Guo et al., 2000) and G-CSF (Tian et al., 1998) receptors. This suggests that it could be possible to use the allosteric sites to discover small molecule drugs that stimulate these pathways which could provide important therapeutic advantages over the use of protein therapeutics. 2.2. Allosteric inhibitors of growth factor receptors One approach to inhibit the growth factor receptors via allosteric mechanisms is to prevent the activation either by blocking the dimer formation and/or preventing the conformational changes needed for

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

catalytic activation. Antibodies can block the dimer formation by preventing growth factor binding to the receptors. Generally, this is more difficult to accomplish with the small molecules because the growth factors are generally large polypeptides and have multiple points of contact with the extracellular domains of RTKs, making it difficult to design small molecules that could block enough sites to diminish protein–protein interaction. One approach that can be used to design the inhibitors of growth factor binding to RTKs employs pharmacophore modeling, much like that described above in studies by Massa et al. (2010). A technology that was first employed on IL-2 receptors, but which can be applied to RTKs to design allosteric inhibitors was described by Arkin and Wells (2004). IL-2 receptors are not RTKs. However, cytokines, like IL-2 are proteins and they bind to the N-terminal regions of the transmembrane spanning IL-2 receptors. Furthermore, the binding of IL-2 induces oligomerization to activate the cytokine receptor, much like that found in RTKs. Thus, the basics of IL-2 binding and activation of its receptor is similar to that found with many RTKs and therefore approaches used to develop small molecule inhibitors of IL-2 binding could have applications in developing small molecule inhibitors of RTKs. Arkin and Wells (2004) reasoned that because structural information identified “hot spots”, of limited regions of contact between IL-2 and its receptor, one could employ the structural information of this interaction to design pharmacophores of IL-2 that could lead to small molecule inhibitors of the protein–protein interaction to block the binding of IL-2 to its receptor to prevent dimer formation. In fact, Emerson et al. (2003) employed NMR to identify residues in IL-2 needed for binding to the IL-2 receptor (IL-2R). Using this information, this group designed a rigid peptidomimetic that simulated the IL-õ2 pharmacophore and showed that it blocked the binding of IL-2 to its receptor. They (Tilley et al., 1997; Emerson et al., 2003) also designed a small molecule acylphenylalanine derivative corresponding to pharmacophore of IL-2 and identified a small molecule compound (Ro26-4550) (see Fig. 1 for structure) that blocked IL-2 binding to the IL-2R with an IC50 of 3 uM. Interestingly, the NMR studies showed that this relatively potent IL-2 inhibitor is bound to IL-2 itself, rather than directly interacting with the receptor, suggesting the molecule masked sites on the IL-2 needed for binding to its receptor. Based on these findings, Wells and associates developed a tethering technology to identify potent small molecule inhibitors of IL-2. They (Arkin et al., 2003) mutated IL-2 by substituting cysteine residues around the Ro26-4550 binding site in IL-2. They then screened the mutant IL-2 with a library of small molecules tethered to disulfide linkers. Compounds in the library that bound to the Ro26-4550 site with reasonable affinity were able to form disulfide bridges with the cysteine residues incorporated into the receptor and bound ligands were determined by mass spectroscopy. In follow up studies using this technology, Raimundo et al. (2004) and Thanos et al. (2006) identified a small molecule compound (SP4206) (see Fig. 1 for structure) that bound to IL-2 with a 60 nM affinity. Molecular dynamic studies suggest that the compound binds to the low-energy conformations of IL-2 (Thanos et al., 2006). SP4206 blocked IL-2 binding to IL-2R with 70 nM affinity. Importantly, it blocked IL-2 induced STAT phosphorylation in cells expressing IL-2R with an affinity of 3 uM. Interestingly, this concept of designing small molecules that bind to cytokines to block the actions of these proteins, has in fact been employed to identify small molecule allosteric inhibitors of the NGF regulated RTKs. Niederhauser et al. (2000) identified a small molecule, Ro 08-2750 (see Fig. 1 for structure) that binds to the NGF dimers. Importantly, the NGF dimers bind to TrkA and p75NTR. Both receptors form dimers in response to the NGF binding with TrkA receptor activation promoting survival of neurons while the activation of p75NTR causes apoptosis and cell death. Niederhauser et al. (2000) showed that at low concentrations, Ro 08-2750 interacted with the NGF to block the binding of NGF to p75NTR but not to TrkA. As a consequence, the NGF induced apoptosis through its activation of p75NTR was abolished while

149

the ability of NGF to promote neuronal survival and neurite formation through the stimulation of TrkA was not affected. The authors proposed that Ro 08-2750 produced conformational changes in NGF presumably to reduce the epitopes needed for binding to the p75NTR while maintaining the pharmacophore needed for binding to TrkA. As a consequence, Ro 08-2750 can be viewed as an allosteric modulator that modifies the NGF to selectively activate an RTK without affecting other signaling pathways, much like a designer drug. Thus, one approach to allosterically modulate RTKs is to devise small molecules that bind to the pharmacophore of the growth factor that normally regulates the RTK rather than devising small molecules targeting the RTK itself. Furthermore, as shown in studies by Maliartchouk et al. (2007), small peptidomimetics based on the pharmacophore of the growth factors, NGF in this case, can be designed that directly stimulate TrkA to promote neuronal survival indicating that it should be possible to discover small molecule agonists that directly bind to the extracellular domain of the TrkA to allosterically increase RTK activity. Similarly, one might expect the same approach to be used to design small molecule allosteric inhibitors that prevent the binding of the growth factor. A second approach to develop allosteric inhibitors of RTKs focused at interacellular sites is to block the interaction of the catalytic domains of the RTK dimers to prevent autophosphorylation and activation. This in fact was the approach used by Zhang et al. (2007) to develop novel allosteric inhibitors of the EGF receptor. For this receptor, dimers form when EGF binds to the extracellular domain of the receptor. When this happens, the catalytic domains interact and the C-terminal region of one monomer binds to the N-terminal region of the other catalytic domain of the other to induce activity. An endogenous protein, mitogen-induced gene 6 (MIG6), binds to the C-terminal region of the kinase domains preventing the interaction of the dimer catalytic domains and prevents activation. Importantly, the mutant forms of the EGF receptor that are constitutively active still require this asymmetric dimer formation for activity and the process is regulated by MIG6. Zhang et al. (2007) suggested that small molecules simulating the actions of MIG6 could provide a new generation of anti-cancer drugs highly selective for the EGF receptors. In fact, follow up studies have further defined the interaction of the catalytic domains of the EGF receptors using both crystallography and molecular simulation studies providing structural basis for developing selective small molecule inhibitors that could be effective in treating EGF receptor related cancers (Jura et al., 2009a, 2009b). In summary, if structural and mutagenesis data exists on an individual RTK and/or its growth factor, it should be possible to employ the Tethering technology (Arkin & Wells, 2004) as well as other structural approaches described above to design and develop small molecule allosteric inhibitors against a number of RTKs. The tethering approach is a unique technology that can be employed to develop small molecule modulators of the protein–protein interaction. It has been employed to discover novel allosteric regulators of the cytokine receptors and enzymes such as the caspases (Hardy et al., 2004) and should be amenable for developing inhibitors of RTKs providing a rational approach for drug discovery against this family of kinases and also providing a new generation of drugs to treat disorders involving over active growth factor and their receptors such as in the case of proliferative disorders like cancer and inflammation as well as CNS diseases where certain growth factors and cytokines cause neurodegeneration and cell death. 3. Allosteric modulators of soluble protein kinases 3.1. Allosteric modulators of the tyrosine kinase Bcr–Abl overcome Gleevec resistance While the development of allosteric modulators against RTKs holds great promise in the future for the development of novel drugs,

150

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

extensive studies have also focused on developing allosteric inhibitors and activators of soluble tyrosine and serine/threonine kinases. One of the more successful targets that have led to clinically useful drugs is the –Bcr-Abl fusion protein, which has been a focus for novel tyrosine kinase drug discovery over the last few decades. The Bcr fusion to Abl kinase creates constitutive activity and causes chronic myelogenous leukemia (CML). Gleevec, a novel type 2 kinase inhibitor was originally targeted to this mutant kinase (Zhang et al., 2009). It was one of the first molecules found to bind near the ATP binding region of the inactive form of this or any kinase. Its binding prevented the opening of the activation loop needed for catalytic activity thus forcing the kinase to remain in an inactive state. In fact, Gleevec selectively binds to the inactivate state of this mutant enzyme and was one of the first drugs approved using this unique mechanism of action. While in treating CML and approved by FDA, some advanced–stage patients relapse as a result of the emergence of clones expressing Gleevec-resistant forms of Bcr–Abl (Gorre et al., 2001; Shah et al., 2004). Specifically, some mutations occur in the kinase to reduce the binding of Gleevec. This can be a particular problem in anti-cancer drug development since cells have an incredible ability to exert mechanisms to resist drug effects. It is as if cells can generate multiple forms of a particular kinase with subtle mutations that do not affect the activity but seem predestined to overcome any chemical forms of inhibition. When a drug like Gleevec inhibits Bcr–Abl, other clones to which Gleevec can't affect take over the activity. In a continuing battle to find inhibitors of Bcr–Abl, the crystal structure analysis of Gleevec–Bcr–Abl binding was used as a basis, to identify the newer Bcr–Abl inhibitors nilotinib (AMN107) and dasatinib (BMS-354825) which have been approved by FDA (Shah et al., 2004; Weisberg et al., 2005). These compounds interact with the ATP-binding region of the Bcr–Abl with much higher affinity than Gleevec and were effective in inhibiting Bcr–Abl with mutations that confer Gleevec resistance. However, even with these newer drugs, some patients become resistant to treatment because of the ‘gatekeeper’ T315I mutation, located in the center of the ATP-binding cleft (Bradeen et al., 2006; Zhang et al., 2010). This mutation, within the ATP binding site substitutes a large isoleucine for the threonine needed for Gleevec binding but does not affect ATP binding, so that Gleevec and the newer drugs developed don't work at inhibiting this mutant kinase. Using an entirely different approach to those employed in developing previous Bcr–Abl inhibitors, Adrian et al. (2006) and Zhang et al. (2010) developed GNF-2 (see Fig. 1 for structure) and GNF-5, which are allosteric, non-ATP competitive inhibitors of Bcr– Abl. GNF-2 was identified in an unbiased cytotoxicity assays using the Bcr–Abl expressing cells. GNF-2 inhibited Bcr–Abl activity and reduced Stat5 tyrosine phosphorylation in cells but did not affect the Bcr–Abl kinase activity in vitro nor did it affect the activity of a number of other kinases in vitro. It did bind to the Bcr–Abl in vitro but did not affect the binding of ATP or Gleevec to the kinase. GNF-2 binds to the myristoyl pocket in Bcr–Abl and its binding to the kinase was selectively blocked in the competition assays by myristoylated peptides. Furthermore, the mutations in the myristoylated pocket prevented GNF-2 from inhibiting the proliferation of cells expressing the mutant kinase but did not affect the inhibitory ability of Gleevec. Subsequent crystal structure analysis and NMR spectroscopy have shown that GNF-2 binds to the myristoylated pocket while Gleevec bound near the ATP binding domain. Importantly, GNF-2 allosterism did not affect Gleevec binding but facilitated Gleevec's ability to inhibit Bcr–Abl. GNF-2 was found to produce conformational changes in the kinase to preserve the inactive form of the enzyme to reduce ATP binding as assessed through the use of hydrogen-exchange mass spectroscopy. Furthermore, GNF-5, an analog of GNF-2, with Gleevec produced additive inhibition of the kinase and overcame resistance conferred by the T315I mutation as well as other mutations. GNF-5

was also effective in a murine model of the Bcr–Abl induced leukemia. The use of these novel allosteric inhibitors not only provides an approach to treat CML more effectively than presently available, but also supports the utility of the allosteric modulators in overcoming resistance to more classical protein kinase drug inhibitors. 3.2. Allosteric inhibitors of the serine/threonine kinase, Akt Akt consists of a family of three serine/threonine protein kinases which have high sequence similarity, especially in their catalytic domains. The enzymes have a critical role in proliferative diseases such as cancer and have been a major target of the pharmaceutical industry in its efforts to develop newer and more effective chemotherapeutic agents (Carnero et al., 2008; Tokunaga et al., 2008). The Akt kinases have an unusual N-terminal pleckstrin homology (PH) domain which is important for anchoring the enzymes to the cell membrane where they come in contact with phosphoinositide dependent kinase 1 (PDK-1) which phosphorylates the Akt kinases on the activation loop to produce continued catalytic activity. When not bound to the plasma membrane, the PH domain occludes the activation loops of the Akt kinase preventing PDK1 activation and maintaining the kinases in a closed conformation. The high sequence similarity of the Akt isoforms in the kinase domain suggested that targeting the PH domain may be a viable approach to develop subtype selective inhibitors of the Akt isoforms. Studies by Barnett et al. (2005) and his colleagues focused on identifying selective Akt kinase inhibitors. They employed an Akt activity assay in screening a small molecule library and by luck identified several compounds (Akt-I-1 and Akt-I-1,2) (see Fig. 1 for structure) that inhibited Akt1 in an ATP and substrate noncompetitive manner. Neither compound affected the activity of Akt3 nor a large number of other protein kinases. Both inhibitors effectively kill the tumor cell lines overexpressing Akt1 in vitro. They were effective alone and also synergized with other chemotherapeutic agents (Hirai et al., 2010). They were also effective in vivo in a tumor xenograft model (LNCaP prostate cancer xenografts) (Cherrin et al., 2010). Furthermore, one of the inhibitors (MK-2206) was recently shown to be safe in humans in Phase 1 studies after oral administration (Lindsley, 2010). The selectivity of these compounds was remarkable given the high similarity of the Akt isoforms. Subsequent studies showed the compounds bind to the PH domain maintaining the kinase in a closed formation preventing PDK1 phosphorylation of Akt1 and reducing Akt1 phosphorylation of the downstream substrates. The evidence that these compounds acted as allosteric inhibitors of Akt1 and bound to the PH domain was supported by the FRET based studies. Interestingly, these inhibitors may also be working by affecting the translocation of Akt. The studies by Calleja et al. (2010) showed that these inhibitors blocked the translocation of a GFP-Akt1-mRFP to the plasma membrane induced by PDGF, presumably by hindering the association of PH with anchoring sites in the membrane. This translocation is necessary for localizing Akt1 with PDK1 to cause the activation of Akt1. Interestingly, Kim et al. (2010) have also identified the allosteric inhibitors of Akt1 that bind to the PH domain. The small molecule blocks the translocation of Akt to the cell membrane, to inhibit the activity and prevent the proliferation of tumor cells that exhibit high activity. It was also effective in vivo in reducing the sizes of the ovarian and pancreatic tumors in mice that have high Akt activity. They also found that their compound inhibited the activity of mutant forms of Akt, in particular Akt E17K mutation. This mutation is found in the human breast, colorectal and ovarian cancers and is in the PH domain of the kinase. By enhancing the electrostatic interaction of Akt1 with phosphoinositide in the plasma membrane, the mutation causes “pathological” localization of the kinase to the plasma membrane, to cause the cell transformation and leukemia in mice. Interestingly, Carpten et al. (2007) have shown this mutation to

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

reduce the affinity of Akt1 for MK-2206 by over 4-fold suggesting that the amino acid substitution affects the conformation of the PH binding pocket for this small molecule. In contrast, the Akt allosteric inhibitor identified by Kim et al. (2010) was equally effective in inhibiting the mutant and wild type Akt suggesting it may overcome potential resistance that may form to MK-2206. The results with the PH domain allosteric inhibitors and their ability to block the translocation of Akt to inhibit the activity suggest that a potently important direction for the discovery of novel, highly selective allosteric kinase inhibitors could focus on modulating the intracellular movement of the enzymes, rather than purely focusing on enzyme conformation, since the importance of the activity of many kinases is dependent on their subcellular distribution. 3.3. Allosteric inhibitors of other serine/threonine kinases In addition to Akt, unique allosteric modulators have been discovered against a number of other serine/threonine protein kinases as described previously (Eglen & Reisine, 2009, 2010). For example, much progress has been made in developing novel regulators of the Checkpoint kinase 1 (CHK1). This enzyme controls the movement from S to G2 phase in cell cycle and is important for the DNA damage control. As such, it is important in cancer treatment and sensitivity of malignant cells to chemotherapeutics. Thus, the allosteric inhibitors of this kinase have been identified by first screening chemical libraries in classical kinase phosphorylation assays and then picking out “hits” based on their lack of dependence on ATP, that is they were not competitive with ATP (Converso et al., 2009). Recent studies by Vanderpool et al. (2009) have now described two new allosteric CHK1 inhibitors which bind near the substrate binding site of the kinase but which do not compete directly with the substrate binding. They have been proposed to produce conformational changes that subsequently hinder the substrate interaction in the catalytic domain. These findings are particularly interesting since relatively few small molecule drugs have been identified that can affect substrate binding either directly or indirectly and such inhibitors might be expected to confer a high degree of selectivity, which is certainly needed when regulating a kinase controlling cell cycle. p38 MAPK has been another serine/threonine kinase for which extensive efforts have been made to discover novel and highly selective allosteric inhibitors because of its important role in the disease (see Eglen & Reisine, 2010). The allosteric inhibitors have been identified that block the translocation of this kinase to the nucleus in cells preventing its access to target sites (Almholt et al., 2004; Trask et al., 2009) and by nature these inhibitors are not competitive with ATP. Furthermore, Diskin et al. (2008) have identified novel lipid allosteric binding sites on p38 kinase that can lead to the inhibition of activity. Furthermore, Simard et al. (2009) developed a novel assay to measure movement of the activation loop of p38 which they used to identify novel type 2 inhibitors of the kinase there were highly selective. Similarly, major progress has been made in developing novel allosteric modulators of the phosphoinsositide-dependent protein kinase-1 (PDK1), which can control the activation of a large number of protein kinase including the protein kinase A and C families by catalyzing the phosphorylation of residues in the activation loop (Gao & Harris, 2006). PDK1 has been found to have a docking site that binds the substrates, the so called PDK1 interacting fragment (PIF) site. PIF sequences are found in many PDK1 substrates. In fact, the PIF binding site in PDK1 is a hydrophobic groove, uniquely located N-terminal to the catalytic site and allows this kinase to sense other target kinases as substrates. Importantly, the binding of substrate to this docking site activates PDK1 by inducing conformational changes in the catalytic domain. Thus, the binding of a substrate kinase to PDK1 via the substrates PIF motif stimulates the PDK1 activity which then phosphorylates the activation loops of the substrate kinase (Engel et al., 2006; Gao & Harris, 2006).

151

A polypeptide corresponding to the PIF motif was shown to bind to PDK1, by surface plasmon resonance and activate the kinase and X-ray crystallography has been employed to study the mode of interaction of the activating PIF-tide and PDK1 as a basis for small molecule drug discovery (Gao & Harris, 2006). Using this structural information, Engel et al. (2006) was able to devise a strategy to identify the allosteric activators of PDK1 that bound to the PIF docking site. This involved in silico compound screening and testing the ability of PDK1 to phosphorylate an artificial substrate consisting of segments of the activation loop of PKB. Several small molecule compounds where identified that activated PDK1 and this activation was blocked by the PIF-tide suggesting that the small molecules are bound to the PIF docking site. Similarly, Stockman et al. (2009) employed library screening using NMR analysis to identify the compounds that bound to the PDK1–PIF binding domain. The compounds identified as binding were analyzed for whether they were ATP competitive or selectively blocked by the PIF peptides and then subjected to functional assays involving the ADP depletion assays. Using this approach, these authors also identified the potential allosteric activators of PDK1. To further identify the allosteric inhibitors of PDK1, Bobkova et al. (2010) a TR-FRET assay was employed using a fusion of the fragment of the activation loop substrate to screen small molecule libraries. Using the assay, small molecule alkaloids were identified that bound to the PIF domain in PDK1 and inhibited the kinase activity. In fact, these allosteric inhibitors were highly selective, since they were unable to affect the activity of over 96 other protein kinases. Since PDK1 is a master kinase involved in regulating a larger number of other serine–threonine kinases and is involved in a number of proliferative diseases, such highly selective allosteric inhibitors may be developed into unique and highly value anti-cancer therapies. 3.4. Allosteric inhibitors that affect protein kinase translocation The studies of other protein kinases have shown the importance of translocation in both normal activity and dysfunction in a disease. The importance of the translocation of the protein kinases with regard to function was most clearly shown for protein kinase C (PKC) (Budas et al., 2007). The PKC family consists of a number of different isoforms, many of which have distinct functions in cells. The kinases interact with the receptors for the activated C-kinase (RACK) proteins. The RACKs interact with the allosteric sites in the PKCs to stabilize the kinases in an open activated form and transport the enzymes to the subcellular compartments where they have access to their target substrates (Budas et al., 2007). The functional role of the RACKs and the translocation in PKC function was shown by the discovery of the allosteric peptide inhibitors of PKC–RACK association that revert the PKCs to an inactive form and reduce the access of the kinases to their targets. The peptide inhibitors corresponding to the unique contact sites of each PKC isoform with their RACKs were developed that could be used to investigate the selective functions of each isoform and in some cases have been developed into therapeutics (Liron et al., 2007). For example, a peptide inhibitor of δPKC–RACK was discovered that was effective in reducing the infarct size in the animal models of myocardial infarction and stroke. The peptide was found safe in human trials and is now undergoing later stage efficacy studies for the treatment of cardiovascular disorders (Budas et al., 2007). Peptidomimetics targeting other PKC isoform– RACK complexes have been shown in preclinical studies to be useful in treating cardiac hypertrophy, ischemia and some pain responses. Thus, the compounds, in particular small molecules targeting the allosteric sites of the RACK–PKC interaction could have a number of potentially useful therapeutic applications. Similarly, translocation is a critical factor in the functions of cAMP dependent protein kinase (PKA) (Patel et al., 2010). The translocation is mediated by a family of 13 different A kinase anchoring proteins (AKAP) which both transport and anchor PKA to target substrates and the site of

152

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

cAMP gradients in cells (Ruehr et al., 2004; Patel et al., 2010). AKAPs interact with allosteric sites in PKA, much like RACKs bind to the allosteric regions of PKC. They interact with the regulatory subunits of PKA and the peptides corresponding to the PKA binding region of AKAP exhibit low nM affinities for the regulatory regions. The binding has been evaluated by NMR showing the dynamics unique for the protein– protein interaction (Kinderman et al., 2006; Sarma et al., 2010). Crystal structure analysis has also been made on the AKAP binding to the regulatory subunits revealing that AKAP and peptide analogs interact with a groove on the surface of the regulatory subunit of PKA involving a strong hydrophobic protein–protein interaction (Kinderman et al., 2006; Taylor et al., 2008). A recent study by Patel et al. (2010) further established the importance of AKAP in PKA function in studies in which a peptide inhibitor of the interaction blocked the PKA translocation in cardiac myocytes, reduced the kinase phosphorylation of selected targets and affected myocyte contractility, heart rate and ventricular pressure. Furthermore, recent studies have shown that SNPs are present in some of the PKA binding domains of AKAP, one of which increases AKAP–PKA binding. This SNP has been shown to increase the risk for sudden cardiac death and ventricular arrhythmias. This suggests that allosteric inhibitors targeting this particular AKAP–PKA might have a therapeutic benefit, supporting the role of the allosteric modulatory sites of transporters of kinases as potential targets for the novel drug discovery. In fact, the translocation between cellular compartments appears to be a common and important factor in the function of many protein kinases, including the G protein linked receptor kinases (GRKs), p38 MAPK and others. In all cases, the translocation process involves the allosteric sites on the protein kinases to which the cellular proteins involved in the transport of the kinases bind to both affect the conformation of the kinase and anchor the enzymes to sites near their target substrates (see Trask et al., 2009; Eglen & Reisine, 2010). The assays that measure the protein kinase translocation could provide directed approaches for the discovery of the allosteric inhibitors of a number of different protein kinases. For example, Almholt et al. (2004) developed cellular distribution assay for p38 MAPK using fluorescent tags and confocal microscopy to measure the movement of the kinase in an HTS format. Similar assays have been developed using a complementation assay format and secondary assays can be employed to distinguish the direct inhibitors of the kinase catalytic activity from the allosteric inhibitors (see Eglen & Reisine, 2009, 2010). 3.5. Allosteric modulators of G protein linked receptor kinases (GRK)s G protein linked receptor kinases (GRK) are a family of 7 serine/ threonine kinases that are critical regulators of the G protein linked receptor (GPCR) function and as such are important in neurotransmission, and the action of hormones and growth factors (Krupnick & Benovic, 1998; Pitcher et al., 1998). These kinases recognize activated, agonist occupied GPCRs (Boguth et al., 2010) and catalyze the phosphorylation of the receptors to attract β-arrestins which uncouple the receptors from G proteins, terminating the signaling through the G protein pathways. GPCRs then couple to other signaling pathways, in some cases to growth factor signaling, to alter their functional profile and change the biological actions of the activating transmitter or hormone on the target cells (Lefkowitz & Shenoy, 2005). Thus, GRKs restrict the activities of the GPCRs and are responsible for generating alternative cellular regulation by these cell surface receptors. GRKs are soluble and contain catalytic domains similar to most other kinases. However, they have an abundance of allosteric regulatory sites that are involved in binding and recognizing target GPCRs as well as the regions needed to anchor the kinase in the plasma membrane to allow for association with the receptors (Huang et al., 2009; Boguth et al., 2010). Furthermore, GRKs are recruited to

membrane bound GPCRs by βγ subunits. These subunits are released from the G protein following the agonist activation of the GPCR to interact with multiple elements in signaling pathways including enzymes, ionic channels and GRKs to turn off continued stimulation. The βγ subunit bind to additional allosteric sites in the GRKs to allow the kinases to associate with the receptors and the binding also induces allosterism to catalyze phosphorylation of the intracellular sites in the receptor (Pitcher et al., 1992; Tesmer et al., 2010). Studies by Boguth et al. (2010) have identified some of the structural elements of GRKs, in this case GRK6, by crystallography, that are responsible to the association with GPCRs and have identified the physical nature of some of the allosteric sites in the kinase. In particular, GRKs have unique N-terminal regions that make direct contact with GPCRs and thus are responsible for receptor recognition and may also be involved in detecting agonist bound conformations of the receptor. GPCRs bind to allosteric sites on GRKs to induce conformational changes in the GRKs to promote activity. Adjacent to the N-terminus is a highly basic, relatively flat region essential for the association of GRKs with phospholipids in the plasma membrane and as such are important for anchoring the kinase near the target GPCR. In addition to these regions, GRKs have unique C-terminal regions also involved in membrane targeting and possibly interaction with other proteins, such as the βγ subunits and therefore contains allosteric sites. Both peptides and small molecules have been identified that affect the GRK activity via the allosteric mechanisms. The regions of GRK2 which bind βγ subunits have been identified and the peptides corresponding to these regions block the activity of the G proteins and also prevent the recruitment of GRK to the plasma membrane following receptor activation (Rockman et al., 2002; Hata & Koch, 2003). Furthermore, small molecules that bind to βγ, including M119 and gallein (Davis et al., 2005; Bonacci et al., 2006; Lehmann et al., 2008) block the association of the G proteins to GRK2 in the cell free systems and in HL60 cells (Casey et al., 2010). Thus, M119 and gallein are the allosteric modulators of GRKs and act by binding to the regions of G proteins that are needed to recruit the GRKs to the cell membrane to phosphorylate GPCRs. These small molecules have been used to show the potential therapeutic importance of developing drugs targeting GRK (Casey et al., 2010). β-adrenergic receptors and excessive βγ subunit activity have been associated with pathophysiological mechanisms involved in heart failure (Bristow et al., 1982; Ungerer et al., 1993; Koch et al., 1995; Iaccarino et al., 1999; Rockman et al., 2002; Hata et al., 2006; Matkovich et al., 2006; Raake et al., 2008; Dorn, 2009). Excessive stimulation of β-adrenergic receptors can result in the excessive release of βγ subunit and extensive recruitment of GRK2 to the plasma membrane to phosphorylate and desensitize and eventually downregulate the β-receptor. Over-expressing GRK2 in the cardiac tissue can also lead to heart failure while the knockdown of the kinase can be cardioprotective. Furthermore, the peptides that block βγ subunit interaction with GRKs have been found effective in improving the cardiac function in the models of heart failure. Similarly, Casey et al. (2010) found that the small molecules M119 and gallein, which block βγ subunit interaction with GRKs, blocked the β-agonist induced recruitment of GRK2 to the cell membranes of the cardiomyocytes. They were also cardioprotective in several models of the heart failure. Specifically, continuous administration of isoproterenol will cause receptor desensitization and cardiac dysfunction whereas cotreatment with M119 blocked the cardiac dysfunction and maintained normal contractile activity. Furthermore, in a genetic model of heart failure in mice with cardiac overexpression of calsequestrin, gallein administration prevented cardiac degeneration. These studies suggest that small molecule drugs that block GRK2 translocation to the cell membrane could be effective in treating cardiac disorders. Importantly, GRK over-activity has been linked to a number of disorders and diseases including opiate tolerance development and

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

addiction. Since GRKs are generally inactive and are only stimulated when associating with GPCRs, then compounds targeting allosteric sites on GRKs involved in coupling to GPCRs and G proteins could be effective in a host of disorders by blocking the accessibility of the kinase to the target substrate without affecting the activity per se of the catalytic domain. Furthermore, if GRKs show selectivity in which GPCRs they associate with, then the allosteric modulators could be designed to selectively affect the GRK regulation of different GPCRs allowing for focused regulation of the actions of the specific neurotransmitters, hormones or growth factors. There are assays that can be employed to discover novel inhibitors of GRK translocation. Tagging the GRKs with fluorescent molecules and tracking movement using confocal microscopy, much like what has been done previously with β-arrestin (Ferguson & Caron, 2004) can be employed to study GRK recruitment to the cell membrane in response to the GPCR activation and a number of drug discovery programs in the pharmaceutical industry have develop HTS confocal screening technologies for the drug discovery. Similarly, complementation assays have been developed to measure protein translocation in cells (Eglen, 2005; Fung et al., 2006; Zhao et al., 2008). 4. Assays to discover novel allosteric modulators of protein kinases Numerous assays have been employed for the discovery of drugs targeting protein kinases, such as standard phosphorylation assays as well as more recently developed ADP depletion or formation assays which have been described elsewhere (Eglen & Reisine, 2009, 2010). The readout of these assays, when combined with the structural analysis, has resulted in the discovery and development of a number of useful and effective drugs. The mode of action of the compounds identified can generally be distinguished in secondary assays in which the effectiveness of the inhibitor is assessed with respect to ATP concentration. Compound binding can then be further evaluated using surface plasmon resonance spectroscopy, mass spectroscopy and even crystallography to facilitate the rational design of drugs. While such assays have been employed to discover some allosteric modulators, newer technologies have now been developed to identify novel allosteric modulators. 4.1. Receptor tyrosine kinase assays for allosteric modulators The majority of drugs targeting RTKs are either protein therapeutics, such as antibodies targeting the extracellular domains to block growth factor interaction, or small molecules targeting the ATP binding domains of the catalytic subunit of the kinase. As indicated above, there are some examples of small molecule allosteric activators of the receptor tyrosine kinases which mimic the actions of the protein growth factors as well as the small molecule inhibitors. The small molecule allosteric inhibitors could be particularly useful clinically because of their increased selectivity compared to the ATP competitive drugs. The assays to measure the RTK activity are available that either detect receptor phosphorylation as a response to receptor activation or employ downstream readouts of an activity, such as the phosphorylation of the transcription factors. More recently, an RTK assay was developed that employed a β-galactosidase complementation assay and the assay is referred to as the PathHunter (Olson & Eglen, 2007; Eglen, 2007). The assay employs two fragments of the enzyme β-galactosidase that when separated have no enzymatic activity but when in close proximity recombine to have full enzyme activity which can result in a highly amplified luminescent response. For the RTK assay, one of the fragments of the β-galactosidase, enzyme acceptor (EA), is associated with an SH domain while the smaller fragment, Prolabel, is embedded in the C-terminus of the receptor tyrosine kinase. When the kinase is stimulated by the growth factor, the catalytic domain is phosphorylated attracting the SH-EA fragment which complements with the Prolabel in

153

the receptor to form an active β-galactosidase. The assay technology can also be employed to measure the dimerization of the RTKs using an approach developed by Wehrman et al. (2002, 2005, 2006, 2007). In this format, EA and Prolabel are incorporated into the C-terminal regions of the different subunits of the receptor tyrosine kinase. When dimerization occurs, the fragments can recombine to form active β-galactosidase and luminescent response. This assay can then identify the allosteric antagonists that block the dimer formation to inhibit receptor kinase activity. Similarly, Leuchowius et al. (2010) and Weibrecht et al. (2010) have developed a proximity ligation assay to measure the dimer formation and RTK activity. Unlike the previously described assay, this approach involves the use of antibodies, one selective for the RTK intracellular domain and the other targeting the tyrosine phosphorylation sites. The antibodies are conjugated to oligonucleotides and when the antibodies are in close proximity, the oligonucleotides serve as templates for ligating additional oligonucleotides into a cyclic DNA construct and the antibody attached oligonucleotides are extended by rolling circle amplification. Hybridizing fluorophore-labeled oligonucleotides provided and amplified the reaction product revealing an intense fluorescent signal. Due to the intensity of the signal, the assay can be employed in cells expressing low levels of receptor such as in the primary cells to study the native receptor tyrosine kinases as well as the recombinant receptors and the assay has to be adapted to an HTS format. This provides an important advantage if there are cellular factors affecting the receptor dimer formation or activity that are not easily included in the recombinant receptor assay formats and if the receptor density and expression levels are a critical factor in the drug discovery, as they are in the GPCR drug discovery. Disadvantages are that the proximity ligation assay requires the use of antibodies against the receptor, is not homogenous, and is more laborious to perform as a result. 4.2. Assays for allosteric modulators of soluble kinases — subunit interactions Many soluble protein kinases either consist of subunits or protein modulators that bind to the kinase to regulate catalytic activity. These modulators can interact with the allosteric sites to affect the conformation of the kinase to either increase or diminish the activity. Consequently, the small molecules targeting those allosteric sites can affect the activity either by blocking the interaction of the protein modulator or by inducing the conformational changes in the kinase to alter the catalytic activity. The assays to measure the small molecule interaction with those sites can be employed to discover novel allosteric modulators of the kinases. Cyclic AMP dependent protein kinase has been employed extensively to develop novel assays for the allosteric modulators. This kinase consists of regulator (R) and catalytic (C) subunits with R suppressing activity of C. The cAMP binding to R causes dissociation of the complex and increasing activity. Thus, the small molecules targeting the cAMP binding site on R as well as multiple other sites on R and even the R binding site in C not only can affect the catalytic activity but also modulate the R–C association. A commonly employed protein–protein interaction assay to measure the R–C association is the AlphaScreen/AlphaLISA assay. The assay consists of acceptor and donor beads coated with either streptavidin or anti-GST antibodies. C is biotinylated and R is constructed with a GST tag as described in Gesellchen et al. (2006). In this bead proximity assay, an intense luminescent response occurs when the R–C complex forms and the response is lost when the cAMP or its analogs are added to cause the complex to dissociate. The assay is homogenous and is easily formatted for the cell free based HTS. Importantly, the assay format could be employed for most protein–protein interactions and therefore could be used to discover the allosteric modulators of a large number of different protein kinases. In fact, recent studies by Bobkova et al. (2010) used the

154

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

AlphaScreen to identify the novel allosteric inhibitors of PDK1 using ultra high throughput screening. Similarly, bioluminescent resonance energy transfer (BRET) assay formats can be used to measure R and C interaction and protein–protein interactions in general. As described by Gesellchen et al. (2006), the assay can be set up by tagging R or C with either Renilla luciferase as a bioluminescent donor and GFP as a the acceptor and the constructs can then be expressed in cells by transfection. This provides a homogenous cell based assay to measure protein–protein interaction allowing the determination of both of the molecules that can affect haloenzyme formation and those that are cell permeable. Both the AlphaScreen/LISA and BRET assays are robust and highly sensitive and have been used extensively for compound screening. However, they require the modification of the target proteins. These can involve relatively large additions such as either GST or GFP. While these modifications do not seem to affect the measurement of the PKA subunit interaction, such modifications could affect other target proteins especially if the additions are larger than the target proteins themselves. The technologies to measure the physical interaction of the kinase subunits or regulators can be employed and that do not require the modification of the interacting proteins. These include simple spectrophotometric analysis and surface plasmon resonance. However, these assays, which are cell free, are not easily employed for HTS like the AlphaScreen and BRET assay formats. Interestingly, Saldanha et al. (2006) developed an assay for PKA to identify the allosteric modulators based on a fluorescence polarization (FP) format. The assay uses the endogenous inhibitor of protein kinase (PKI) as a probe for measuring R and C interaction. Both PKI and R compete for similar sites on C. R exhibits a much higher affinity for C than PKI, thus when R–C are complexed, PKI does not bind whereas when C is free, PKI can bind. Thus, these investigators coupled a fragment of PKI with carboxyfluorescien and developed a PPI assay that they adapted for HTS for the allosteric activators of PKA. The assay is homogenous, highly sensitive and, unlike the AlphaScreen and BRET assays, does not require the modification of the target kinase to be implemented. Conceivably the FP assay format can be employed to discover the novel small molecule allosteric inhibitors and activators of other protein kinases. In fact, an FP assay was used to identify the small molecule inhibitors of JIP1 binding to the kinase JNK1 (Chen et al., 2009). JIP1 is a protein activator of JNK1 and binds to the allosteric sites on the kinase. These authors labeled JIP1 with the fluorescent molecule TAMRA and developed a binding assay which they screened with the small molecule libraries for the allosteric inhibitors of this kinase. Similar types of assays could be employed for the IKKB regulation by NEMO and the CDK family as well as likely a number of other kinases.

5. Summary Studies on the kinase structure and function have led to the advances in knowledge on the molecular properties of this protein family and facilitated the discovery of many effective drugs to treat cancer and other diseases. Newer technologies allow a broader scope of drug development against kinases, especially the development of the allosteric inhibitors. This class of drugs provides opportunities to identify more selective drugs as well as those able to modulate the kinase activity in ways additional to those previously developed. First, it is possible to activate the kinases with the allosteric modulators which have importance in treating disorders associated with a lack of activity, such as in a neurodegenerative disease. Secondly, it provides ways to block the activation of some kinases without affecting basal activity, which is in itself, physiologically important. Thirdly, it is now possible to modulate intracellular movement of kinases to specific locations to inhibit the activity of some functions but not others, in effect, making it feasible to develop highly targeted kinase drugs. Finally, and potentially most important, is the potential development of the small molecule

allosteric agonists and antagonists of the growth factor receptors opening up a new field of receptor targeted therapies. Acknowledgments The authors wish to thank Crist Filer for preparing the structures shown in Fig. 1. References Adrian, F. J., et al. (2006). Allosteric inhibitors of Bcr–abl-dependent cell proliferation. Nat Chem Biol 2, 95−102. Almholt, D. L., Loechel, F., Nielsen, S. J., et al. (2004). Nuclear export inhibitors and kinase inhibitors identified using a MAPK-activated protein kinase 2 redistribution screen. Assay Drug Dev Technol 2, 7−20. Arkin, M., & Wells, J. (2004). Small molecule inhibitors of protein–protein interactions: progressing towards a dream. Nat Rev Drug Discov 3, 301−317. Arkin, M., & Whitty, A. (2009). The road less traveled: modulating signal transduction enzymes by inhibiting their protein–protein interactions. Curr Opin Chem Biol 13, 284−290. Arkin, M. R., et al. (2003). Binding of small molecules to an adaptive protein: protein interface. Proc Natl Acad Sci USA 100, 1603−1608. Bardelli, A., Parsons, D., Silliman, N., Ptak, H., Szabo, S., Saha, S., et al. (2003). Mutational analysis of the tyrosine kinome in colorectal cancers. Science 300, 949−950. Barker, A. J., Gibson, K. H., Grundy, W., et al. (2001). Studies leading to the identification of ZD1839 (IRESSA): an orally active, selective epidermal growth factor receptor tyrosine kinase inhibitor targeted to the treatment of cancer. Bioorg Med Chem Lett 11, 1911−1914. Barnett, S., et al. (2005). Identification and characterization of pleckstrin-homologydomain dependent and isoenzyme-specific Akt inhibitors. Biochem J 385, 399−408. Bobkova, E. V., Weber, M. J., Xu, Z., Zhang, Y. L., Jung, J., Blume-Jensen, P., et al. (2010). Discovery of PDK1 kinase inhibitors with a novel mechanism of action by ultrahigh throughput screening. J Biol Chem 285(24), 18838−18846. Boguth, C., Singh, P., Huang, C., & Tesmer, J. (2010). Molecular basis for activation of G protein-coupled receptor kinases. EMBO J 29, 3249−3259. Bonacci, T. M., Mathews, J. L., Yuan, C. J., Lehmann, D. M., Malik, S., Wu, D. Q., et al. (2006). Differential targeting of G beta gamma-subunit signaling with small molecules. Science 312(5772), 443−446. Bradeen, H. A., et al. (2006). Comparison of imatinib mesylate, dasatinib (BMS-354825), and nilotinib (AMN107) in an N-ethyl-N-nitrosourea (ENU)-based mutagenesis screen: high efficacy of drug combinations. Blood 108, 2332−2338. Bristow, M. R., Ginsburg, R., Minobe, W., Cubicciotti, R. S., Sageman, W. S., Lurie, K., et al. (1982). Decreased catecholamine sensitivity and beta-adrenergic-receptor density in failing human hearts. N Engl J Med 307, 205−211. Budas, G., Koyanagi, T., Churchill, E., & Mochly-Rosen, D. (2007). Competitive inhibitors and allosteric activators of protein kinase C isoenzymes: a personal account and progress report on transferring academic discoveries to the clinic. Biochem Soc Trans 35, 1021−1025. Calleja, V., Laguerre, M., Parker, P., & Larijani, B. (2010). Role of a novel ph-kinase domain interface in PKB/Akt regulation: structural mechanism for allosteric inhibition. PLoS Biol 7(1), e1000017. Carnero, A., Blanco-Aparicio, C., Renner, O., Link, W., & Leal, M. (2008). The PTEN/PI3K/AKT signaling pathway in cancer, therapeutic implications. Curr Cancer Drug Targets 8, 187−198. Carpten, J. D., Faber, A. L., Horn, C., Donoho, G. P., Briggs, S. L., Robbins, C. M., et al. (2007). A transforming mutation in the pleckstrin homology domain of AKT1 in cancer. Nature 448, 439−444. Casey, L., et al. (2010). Small molecule disruption of gbeta/gamma signaling inhibits the progression of heart failure. Circ Res 107, 532−539. Chao, M. V. (2003). Neurotrophins and their receptors: a convergence point for many signaling pathways. Nat Rev Neurosci 4(4), 299−309. Chen, T., Kablaoui, N., Little, J., et al. (2009). Identification of small-molecule inhibitors of the JIP–JNK interaction. Biochem J 420, 283−294. Cherrin, C., Haskell, K., Howell, B., Jones, R., Leander, K., Robinson, R., et al. (2010). An allosteric Akt inhibitor effectively blocks Akt signaling and tumor growth with only transient effects on glucose and insulin levels in vivo. Cancer Biol Ther 4(9(7)). Cohen, P. (2002). Protein kinases — the major drug targets of the twenty-first century? Nat Rev Drug Discov 1, 309−315. Converso, A., Hartingh, T., Garbaccio, R. M., et al. (2009). Development of thioquinazolinones, allosteric Chk1 kinase inhibitors. Bioorg Med Chem Lett 19, 1240−1244. Dancy, J., & Sausville, E. A. (2003). Issues and progress with protein kinase inhibitors for cancer treatment. Nat Rev Drug Discovery 2, 296−313. Davis, T. L., Bonacci, T. M., Sprang, S. R., & Smrcka, A. V. (2005). Structural and molecular characterization of a preferred protein interaction surface on G protein beta gamma subunits. Biochemistry 44(31), 10593−10604. DiGiovanna, M. P., Stern, D. F., Edgerton, S. M., et al. (2005). Relationship of epidermal growth factor receptor expression to ErbB-2 signaling activity and prognosis in breast cancer patients. J Clin Oncol 23, 1152−1160. Diskin, R., Engelberg, D., & Livnah, O. (2008). A novel lipid binding site formed by the MAP kinase insert in p38α. J Mol Biol 375, 70−79. Dorn, G. W. I. I. (2009). GRK mythology: G-protein receptor kinases in cardiovascular disease. J Mol Med 87, 455−463.

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156 Eglen, R. M. (2005). Functional G protein-coupled receptor assays for primary and secondary screening. Comb Chem High Throughput Screen 8(4), 311−318. Eglen, R. M. (2007). Assessing GPCR activation using protein complementation: a novel technique for HTS. Biochem Soc Trans 35(Pt 4), 746−748. Eglen, R. M., & Reisine, T. (2009). The current status of drug discovery against the human kinome. Assay Drug Dev Technol 7, 22−43. Eglen, R. M., & Reisine, T. (2010). Human kinome drug discovery and the emerging importance of atypical allosteric inhibitors. Expert Opin Drug Discov 5(3), 277−290. Emerson, S. D., Palermo, R., Liu, C. M., Tilley, J. W., Chen, L., Danho, W., et al. (2003). NMR characterization of interleukin-2 in complexes with the IL-2Ralpha receptor component, and with low molecular weight compounds that inhibit the IL-2/IL– Ralpha interaction. Protein Sci 12(4), 811−822. Engel, M., et al. (2006). Allosteric activation of the protein kinase PDK1 with low molecular weight compounds. EMBO J 25, 5469−5480. Ferguson, S., & Caron, M. (2004). Green fluorescent protein-tagged beta-arrestin translocation as a measure of G protein-coupled receptor activation. Methods Mol Biol 237, 121−126. Flaherty, K., Puzanov, I., Sosman, J., Kim, K., Ribas, A., McArthur, G., et al. (2009). Phase I study of PLX4032: proof of concept for V600E BRAF mutation as a therapeutic target in human cancer. J Clin Oncol (Meeting Abstracts) 27, 9000. Fuller, J. C., Burgoyne, N. J., & Jackson, R. M. (2009). Predicting druggable binding sites at the protein–protein interface. Drug Discov Today 14, 155−161. Fumagalli, F., Racagni, G., & Riva, M. A. (2006). Shedding light into the role of BDNF in the pharmacotherapy of Parkinson's disease. Pharmacogenomics J 6(2), 95−104. Fung, P., Peng, K., Kobel, P., Dotimas, H., Kauffman, L., Olson, K., et al. (2006). A homogeneous cell-based assay to measure nuclear translocation using beta-galactosidase enzyme fragment complementation. Assay Drug Dev Technol 4(3), 263−272. Gao, X., & Harris, T. (2006). Steady state kinetic mechanism of PDK1. J Biol Chem 281 (31), 21670−21681. Gesellchen, F., Prinz, A., Zimmermann, B., & Herberg, F. W. (2006). Quantification of cAMP antagonist action in vitro and in living cells. Eur J Cell Biol 85(7), 663−672. Gorre, M., Mohammed, M., Ellwood, K., Hsu, N., Paquette, R., Rao, P., et al. (2001). Clinical resistance to STI-571 cancer therapy caused by BCR–ABL gene mutation or amplification. Science 293, 876−880. Guo, Z., Zhou, D., & Schultz, P. (2000). Designing small-molecule switches for protein– protein interactions. Science 288, 2042−2045. Hajduk, P., Huth, J., & Fesik, S. (2005). Druggability indices for protein targets derived from NMR-based screening data. J Med Chem 48(7), 2518−2525. Hardy, J., Lam, J., Nguyen, J., O'Brien, T., & Wells, J. (2004). Discovery of an allosteric site in the caspases. PNAS 101, 12461−12466. Hata, J. A., & Koch, W. J. (2003). Phosphorylation of G protein-coupled receptors: GPCR kinases in heart disease. Mol Interv 3, 264−272. Hata, J. A., Williams, M. L., Schroder, J. N., Lima, B., Keys, J. R., Blaxall, B. C., et al. (2006). Lymphocyte levels of GRK2 (betaARK1) mirror changes in the LVAD-supported failing human heart: lower GRK2 associated with improved beta-adrenergic signaling after mechanical unloading. J Card Fail 12, 360−368. Hirai, H., et al. (2010). MK-2206, an allosteric Akt inhibitor, enhances antitumor efficacy by standard chemotherapeutic agents or molecular targeted drugs in vitro and in vivo. Mol Cancer Ther 9(7), 1956−1967. Huang, E. J., & Reichardt, L. F. (2003). Trk receptors: roles in neuronal signal transduction. Annu Rev Biochem 72, 609−642. Huang, C., Yoshino-Koh, K., & Tesmer, J. A. (2009). Surface of the kinase domain critical for the allosteric activation of G protein-coupled receptor kinases. JBC 284, 17206−17215. Iaccarino, G., Dolber, P. C., Lefkowitz, R. J., & Koch, W. J. (1999). Beta-adrenergic receptor kinase-1 levels in catecholamine-induced myocardial hypertrophy: regulation by beta- but not alpha1-adrenergic stimulation. Hypertension 33, 396−401. Ibanez, C. F., Ebendal, T., & Persson, H. (1991). Chimeric molecules with multiple neurotrophic activities reveal structural elements determining the specificities of NGF and BDNF. EMBO J 10(8), 2105−2110. Ibanez, C. F., Ilag, L. L., Murray-Rust, J., & Persson, H. (1993). An extended surface of binding to Trk tyrosine kinase receptors in NGF and BDNF allows the engineering of a multifunctional pan neurotrophin. EMBO J 12(6), 2281−2293. Jura, N., Endres, N. F., Engel, K., Deindl, S., Das, R., Lamers, M. H., et al. (2009). Mechanism for activation of the EGF receptor catalytic domain by the juxtamembrane segment. Cell 137(7), 1293−1307. Jura, N., Shan, Y., Cao, X., Shaw, D. E., & Kuriyan, J. (2009). Structural analysis of the catalytically inactive kinase domain of the human EGF receptor 3. Proc Natl Acad Sci USA 106(51), 21608−21613. Kaplan, D. R., & Miller, F. D. (2000). Neurotrophin signal transduction in the nervous system. Curr Opin Neurobiol 10(3), 381−391. Kim, D., et al. (2010). A small molecule inhibits Akt through direct binding to Akt and prevents Akt membrane translocation. J Biol Chem 285, 8383−8394. Kinderman, F. S., Kim, C., von Daake, S., Ma, Y., Pham, B. Q., Spraggon, G., et al. (2006). A dynamic mechanism for AKAP binding to RII isoforms of cAMP-dependent protein kinase. Mol Cell 24, 397−408. Koch, W., Rockman, H., Samama, P., Hamilton, R., Bond, R., Milano, C., et al. (1995). Cardiac function in mice overexpressing the beta-adrenergic receptor kinase or a beta-ARK inhibitor. Science 268, 1350−1353. Krupnick, J. G., & Benovic, J. L. (1998). The role of receptor kinases and arrestins in G protein-coupled receptor regulation. Annu Rev Pharmacol Toxicol 38, 289−319. Lefkowitz, R. J., & Shenoy, S. K. (2005). Transduction of receptor signals by betaarrestins. Science 308, 512−517. Lehmann, D. M., Seneviratne, A., & Smrcka, A. V. (2008). Small molecule disruption of G protein beta gamma subunit signaling inhibits neutrophil chemotaxis and inflammation. Mol Pharmacol 73(2), 410−418.

155

Leuchowius, K. J., Jarvius, M., Wickström, M., Rickardson, L., Landegren, U., Larsson, R., et al. (2010). High content screening for inhibitors of protein interactions and posttranslational modifications in primary cells by proximity ligation. Mol Cell Proteomics 9(1), 178−183. Li, B., Liu, Y., Uno, T., & Gray, N. (2004). Creating chemical diversity to target protein kinases. Comb Chem High Throughput Screen 7(5), 453−472. Li, M., Youngren, J., Manchem, V., Kozlowski, M., Zhang, B., Maddux, B., et al. (2001). Small molecule insulin receptor activators potentiate insulin action in insulinresistant cells. Diabetes 50, 2323−2328. Lindsley, C. W. (2010). The Akt/PKB family of protein kinases: a review of small molecule inhibitors and progress towards target validation: a 2009 update. Curr Top Med Chem 10(4), 458−477. Liron, T., Chen, L., Khaner, H., Vallentin, A., & Mochly-Rosen, D. (2007). Rational design of a selective antagonist of ε protein kinase C. J Mol Cell Cardiol 42(4), 835−841. Maliartchouk, S., Feng, Y., Ivanisevic, L., Debeir, T., Cuello, C., Burgess, K., et al. (2007). A designed peptidomimetic agonistic ligand of TrkA nerve growth factor receptors. Mol Pharm 57, 385−391. Massa, S., et al. (2010). Small molecule BDNF mimetics activate TrkB signaling and prevent neuronal degeneration in rodents. J Clin Invest 120(5), 1774−1785. Matkovich, S. J., Diwan, A., Klanke, J. L., Hammer, D. J., Marreez, Y., Odley, A. M., et al. (2006). Cardiac-specificablation of G-protein receptor kinase 2 redefines its roles in heart development and beta-adrenergic signaling. Circ Res 99, 996−1003. Niederhauser, O., Mangold, M., Schubenel, R., Kusznir, E., Schmidt, D., & Hertel, C. (2000). NGF ligand alters NGF signaling via p75NTR and TrkA. J Neurosci Res 61, 263−272. Noble, M., Endicott, J., & Johnson, L. (2004). Protein kinase inhibitors: insights into drug design from structure. Science 303, 1800−1805. O'Leary, P. D., & Hughes, R. A. (2003). Design of potent peptide mimetics of brainderived neurotrophic factor. J Biol Chem 278(28), 25738−25744. Olson, K. R., & Eglen, R. M. (2007). Beta galactosidase complementation: a cell-based luminescent assay platform for drug discovery. Assay Drug Dev Technol 5(1), 137−144. Patel, H. H., Hamuro, L. L., Chun, B. J., Kawaraguchi, Y., Quick, A., Rebolledo, B., et al. (2010). Disruption of protein kinase A localization using a trans-activator of transcription (TAT)-conjugated A-kinase-anchoring peptide reduces cardiac function. J Biol Chem 285(36), 27632−27640. Pearson, M., Garcia-Echeverria, C., & Fabbro, D. (2006). Protein tyrosine kinases as target for cancer and other indications. In D. Fabbro, & F. McCormick (Eds.), Protein Tyrosine Kinases: From Inhibitors to Useful Drugs (pp. 1−31). Totowa NJ: Humana Press. Perez-Soler, R., Chachoua, A., Hammond, L. A., et al. (2004). Determinants of tumor response and survival with erlotinib in patients with non-small-cell lung cancer. J Clin Oncol 22, 3238−3247. Piccart-Gebhart, M., Procter, M., Leyland-Jones, B., et al. (2005). Herceptin Adjuvant (HERA) Trial Study Team. N Engl J Med 353, 1659−1672. Pitcher, J. A., Freedman, N. J., & Lefkowitz, R. J. (1998). G protein-coupled receptor kinases. Annu Rev Biochem 67, 653−692. Pitcher, J. A., Inglese, J., Higgins, J. B., Arriza, J. L., Casey, P. J., Kim, C., et al. (1992). Role of beta gamma subunits of G proteins in targeting the beta-adrenergic receptor kinase to membrane-bound receptors. Science 257, 1264−1267. Qureshi, S., et al. (1999). Mimicry of erythropoietin by a nonpeptide molecule. PNAS 96, 12156−12161. Qureshi, S., et al. (2000). Activation of insulin signal transduction pathway and antidiabetic activity of small molecule insulin receptor activators. J Biol Chem 275, 36590−36595. Raake, P. W., Vinge, L. E., Gao, E., Boucher, M., Rengo, G., Chen, X., et al. (2008). G protein-coupled receptor kinase 2 ablation in cardiac myocytes before or after myocardial infarction prevents heart failure. Circ Res 103, 413−422. Raimundo, B., et al. (2004). Integrating fragment assembly and biophysical methods in the chemical advancement of small-molecule antagonists of IL-2: an approach for inhibiting protein–protein interactions. J Med Chem 47, 3111−3130. Rockman, H. A., Koch, W. J., & Lefkowitz, R. J. (2002). Seven-transmembrane-spanning receptors and heart function. Nature 415, 206−212. Roy, K., & Sausville, E. (2001). Early development of cyclin dependent kinase modulators. Curr Pharm Des 7, 1669−1687. Ruehr, M. L., Russell, M. A., & Bond, M. (2004). A-kinase anchoring protein targeting of protein kinase A in the heart. J Mol Cell Cardiol 37, 653−665. Saldanha, S., Kaler, G., Cottam, H., Abagyan, R., & Taylor, S. S. (2006). Assay principle for modulators of protein–protein interactions and its application to non-ATPcompetitive ligands targeting protein kinase A. Anal Chem 78, 8265−8272. Salituro, G., Pelaez, F., & Zhang, B. (2001). Discovery of a small molecule insulin receptor activator. Recent Prog Horm Res 56, 107−126. Sarma, G. N., Kinderman, F. S., Kim, C., von Daake, S., Chen, L., Wang, B. C., et al. (2010). Structure of D-AKAP2:PKA RI complex: insights into AKAP specificity and selectivity. Structure 18, 155−166. Schindowski, K., Belarbi, K., & Buee, L. (2008). Neurotrophic factors in Alzheimer's disease: role of axonal transport. Genes Brain Behav 7(suppl 1), 43−56. Shah, N., Tran, C., Lee, F., Chen, P., Norris, D., & Sawyers, C. L. (2004). Overriding imatinib resistance with a novel ABL kinase inhibitor. Science 305, 399−401. Simard, J. R., et al. (2009). A new screening assay for allosteric inhibitors of cSrc. Nat Chem Biol 5, 394−396. Simpson, G., Hughes, J., Washio, Y., & Bertrand, S. (2009). Direct small-molecule kinase activation: novel approaches for a new era of drug discovery. Curr Opin Drug Discov Dev 12(5), 585−596. Stockman, B. J., et al. (2009). Identification of allosteric PIF-pocket ligands for PDK1 using NMR-based fragment screening and 1H–15N TROSY experiments. Chem Biol Drug Des 73, 179−188.

156

R. Eglen, T. Reisine / Pharmacology & Therapeutics 130 (2011) 144–156

Taylor, S. S., Kim, C., Cheng, C., Brown, S., Wu, J., & Natarajan, K. (2008). Signaling through cAMP and cAMP-dependent Protein Kinase: Diverse Strategies for Drug Design. Biochim Biophys Acta 1784, 16−26. Tesmer, J., Tesmer, V., Lodowski, D., Steinhagen, H., & Huber, J. (2010). Structure of human G protein-coupled receptor kinase 2 in complex with the kinase inhibitor balanol. J Med Chem 53, 1867−1870. Thanos, C., DeLano, W., & Wells, J. (2006). Hot-spot mimicry of a cytokine receptor by a small molecule. PNAS 103(42), 15422−15427. Tian, S., Lamb, P., King, A., Miller, S., Kessler, L., Luengo, J., et al. (1998). A small, nonpeptidyl mimic of granulocyte colony-stimulating factor. Science 281, 257−259. Tilley, J. W., et al. (1997). Identification of a small molecule inhibitor of the IL-2/IL-2Rα receptor interaction which binds to IL-2. J Am Chem Soc 119, 7589−7590. Tokunaga, E., Oki, E., Egashira, A., et al. (2008). Deregulation of the Akt pathway in human cancer. Curr Cancer Drug Targets 8, 27−36. Trask, O. J., Nickischer, D., Burton, A., et al. (2009). High-throughput automated confocal microscopy imaging screen of a kinase-focused library to identify p38 mitogenactivated protein kinase inhibitors using the GE InCell 3000 analyzer. Methods Mol Biol 565, 159−186. Tsai, J., et al. (2008). Discovery of a selective inhibitor of oncogenic B-Raf kinase with potent antimelanoma activity. Proc Natl Acad Sci 105, 3041−3046. Ungerer, M., Bohm, M., Elce, J. S., Erdmann, E., & Lohse, M. J. (1993). Altered expression of beta-adrenergic receptor kinase and beta 1-adrenergic receptors in the failing human heart. Circulation 87, 454−463. Vanderpool, D., et al. (2009). Characterization of the CHK1 allosteric inhibitor binding site. Biochemistry 48, 9823−9830. Wehrman, T. S., Casipit, C. L., Gewertz, N. M., & Blau, H. M. (2005). Enzymatic detection of protein translocation. Nat Methods 2(7), 521−527. Wehrman, T., He, X., Raab, B., Dukipatti, A., Blau, H., & Garcia, K. C. (2007). Structural and mechanistic insights into nerve growth factor interactions with the TrkA and p75 receptors. Neuron 53(1), 25−38.

Wehrman, T., Kleaveland, B., Her, J. H., Balint, R. F., & Blau, H. M. (2002). Protein–protein interactions monitored in mammalian cells via complementation of beta-lactamase enzyme fragments. Proc Natl Acad Sci USA 99(6), 3469−3474. Wehrman, T. S., Raab, W. J., Casipit, C. L., Doyonnas, R., Pomerantz, J. H., & Blau, H. M. (2006). A system for quantifying dynamic protein interactions defines a role for Herceptin in modulating ErbB2 interactions. Proc Natl Acad Sci USA 103(50), 19063−19068. Weibrecht, I., et al. (2010). Review: proximity ligation assays: a recent addition to the proteomics toolbox. Expert Rev Proteomics 7, 401−409. Weisberg, E., et al. (2005). Characterization of AMN107, a selective inhibitor of native and mutant Bcr–Abl. Cancer Cell 7, 129−141. White, A., Westwell, A., & Brahemi, G. (2008). Protein–protein interactions as targets for small-molecule therapeutics in cancer. Expert Rev Mol Med 10, 1−14. Williams, G., Williams, E. J., Maison, P., Pangalos, M. N., Walsh, F. S., & Doherty, P. (2005). Overcoming the inhibitors of myelin with a novel neurotrophin strategy. J Biol Chem 280(7), 5862−5869. Zhang, X., Pickin, K., Bose, R., Jura, N., Cole, P., & Kuriyan, J. (2007). Inhibition of the EGF receptor by binding of MIG6 to an activating kinase domain interface. Nature 450, 741−745. Zhang, B., Salituro, G., Szalkowski, D., Li, Z., Zhang, Y., Royo, I., et al. (1999). Discovery of a small molecule insulin mimetic with antidiabetic activity in mice. Science 284, 974−977. Zhang, J., Yang, P., & Gray, N. (2009). Targeting cancer with small molecule kinase inhibitors. Nat Rev Cancer 9, 28−39. Zhang, J., et al. (2010). Targeting Bcr–Abl by combining allosteric with ATP-binding-site inhibitors. Nature 463, 501−507. Zhao, X., Jones, A., Olson, K. R., Peng, K., Wehrman, T., Park, A., et al. (2008). A homogeneous enzyme fragment complementation-based beta-arrestin translocation assay for high-throughput screening of G-protein-coupled receptors. J Biomol Screen 13(8), 737−747. Zuccato, C., & Cattaneo, E. (2007). Role of brain-derived neurotrophic factor in Huntington's disease. Prog Neurobiol 81(5–6), 294−330.