Patent Application
20060089401
Protein kinases are a large class of enzymes which catalyze the transfer of the γ-phosphate from ATP to the hydroxyl group on the side chain of Ser/Thr or Tyr in proteins and peptides and are intimately involved in the control of various important cell functions, perhaps most notably: signal transduction, differentiation and proliferation. There are estimated to be about 2,000 distinct protein kinases in the human body (Hunter, 1987, 1994, Hanks & Hunter, 1995), and although each of these phosphorylate particular protein/peptide substrates, they all bind the same second substrate ATP in a highly conserved pocket.
Inhibitors of various known protein kinases could have a variety of therapeutic applications provided sufficient selectivity, and acceptable in vivo pharmacological properties, can be incorporated into such inhibitors (Levitzki, 1996a). Perhaps the most promising potential therapeutic use for protein kinase inhibitors is as anti-cancer agents. This potential application for protein tyrosine kinase (“PTK”) inhibitors has been highlighted in many recent reviews (e.g. Lawrence & Hiu, 1998, Kolibaba & Druker, 1997, Showalter & Kraker, 1997, Patrick & Heimbrook, 1996, Groundwater et al., 1996, Levitzki, 1995). The foundation for this application is based partly upon the fact that about 50% of the known oncogene products are PTKs and their kinase activity has been shown to lead to cell transformation (Yamamoto, 1993).
The PTKs can be classified into two categories (Courtneidge, 1994), the membrane receptor PTKs (e.g. growth factor receptor PTKs) and the non-receptor PTKs (e.g. the src family of proto-oncogene products). There are at least 9 members of the src family of non-receptor PTKs with pp60c-src (hereafter referred to simply as “src”) being the prototype PTK of the family wherein the approximately 300 amino acid catalytic domains are highly conserved (Rudd et al., 1993, Courtneidge, 1994). The hyperactivation of src has been reported in a number of human cancers, including those of the colon (Mao et al., 1997, Talamonti et al., 1993), breast (Luttrell et al., 1994), lung (Mazurenko et a, 1992), bladder (Fanning et al., 1992) and skin (Barnekow et al., 1987) as well as in gastric cancer (Takeshima et al., 1991), hairy cell leukemia (Lynch et al., 1993) and neuroblastoma (Bjelfman et al., 1990). Overstimulated cell proliferation signals from transmembrane receptors (e.g. EGFR and p185HER2/Neu) to the cell interior also appears to pass through src (Mao et al., 1997, Parsons & Parsons, 1997, Bjorge et al., 1996, Taylor & Shalloway, 1996). Consequently, it has recently been proposed that src is a universal target for cancer therapy (Levitzki, 1996) because its hyperactivation (without mutation) is involved in tumor initiation, progression, and metastasis for many important human tumor types.
In view of the large and growing potential for inhibitors of various protein kinases, a variety of approaches to obtaining useful inhibitors is needed. The status of the discovery of PTK inhibitors (Lawrence & Niu, 1988, Showalter & Kraker, 1997, Patrick & Heimbrook, 1996, Groundwater et al., 1996, Budde et al., 1995, Levitzki & Gazit, 1995) has been extensively reviewed. Random screening efforts have been successful in identifying non-peptide protein kinase inhibitors but the vast majority of these inhibitors bind in the highly conserved ATP binding site. A notable recent example of such non-peptide, ATP-competitive, inhibitors are the 4-anilinoquinazolines and analogs, which were shown to be effective against the epidermal growth factor receptor PTK (EGFRTK) (e.g. Rewcastle et al., 1996). Although this class of inhibitors was reported to be selective for the EGFR PTK vs. six other PTKs (including src, Fry et al., 1994) it is unknown what their effect is on most of the remaining 2,000 protein kinases that all bind ATP as well as a large number of other ATP, ADP, GTP, GDP, etc. utilizing proteins in the body. Therefore, potential side effects from PTK inhibitor drugs that mimic ATP, which might only be discovered after expensive animal toxicity studies or human clinical trials, are still a serious concern. Also, although this class of compounds was a nice discovery and is undergoing further exploration, these compounds do not provide a rational and general solution to obtaining non-peptide inhibitors for any desired PTK, e.g. in this case src. The risk of insufficient specificity in vivo with ATP-competitive PTK inhibitors has also been noted by others, along with the inherent three order of magnitude reduction in potency these inhibitors display when competing with the mM levels of intracellular ATP rather than the μM levels used in the isolated enzyme assays (e.g. see Lawrence & Niu, 1998, Hanke et al., 1996, Kelloff et al., 1996).
An older, and more extensively studied, class of non-peptide PTK inhibitors is erbstatin and the related tyrphostins (see reviews). This class of inhibitors is active against the receptor PTKs and their mode of inhibition is complex but does not appear to involve binding in the peptide substrate specificity site regions of the active site (Hsu et al., 1992, Posner et al., 1994). Furthermore, this class of inhibitorsis inactive against the isolated PTK when the unnatural assay metal Mn2+ is replaced with the natural Mg2+ (Hsu et al., 1992), is chemically unstable (Budde et al., 1995, Ramdas et al., 1995 & 1994), and is known to be cytotoxic to normal and neoplastic cells by crosslinking proteins (Stanwell et al., 1995 & 1996) as well as to inhibit cell growth by disrupting mitochondria rather than PTK inhibition (Burger et al., 1995).
An important contribution to the protein kinase field has been the x-ray structural work with the serine kinase cAMP-dependent protein kinase (“PKA”) bound to the 20-residue peptide derived from the heat stable inhibitor protein, PKI(5-24), and Mg2ATP (Taylor et al., 1993). This structural work is particularly valuable because PKA is considered to be a prototype for the entire family of protein kinases since protein kinases have evolved from a single ancestral protein kinase. Sequence alignments of PKA with other serine and tyrosine kinases have identified a conserved catalytic core of about 260 residues and 11 highly conserved residues within this core (Taylor et al., 1993). Two highly conserved residues of particular note for the work proposed herein are the general base Asp-166, which is proposed to interact with the substrate OH and the positively charged residue, Lys-168 for serine kinases and an Arg for tyrosine kinases (Knighton et al., 1993), which is proposed to interact with the γ-phosphate of ATP to help catalyze transfer of this phosphate. Two additional important PKA crystal structures have been reported (Madhusudan et al., 1994), one for the ternary PKA:ADP:PKI(5-24) complex wherein the PKI Ala 21 has been replaced with Ser (thereby becoming a substrate), and one for the binary PKA:PKI(5-24) complex wherein the PKI Ala 21 has been replaced with phosphoserine (an end product inhibitor). The ternary complex shows the serine OH donating a H-bond to Asp-166 and accepting a H-bond from the side chain of Lys 168. The binary complex shows the phosphate group of phosphoserine forming a salt bridge with the Lys-168 side chain and within H-bonding distance of the Asp-166 carboxyl group. These structures support the earlier proposed roles for Asp-166 and Lys-168 in the catalytic mechanism.
The x-ray structures of PKA show that the enzyme consists of two lobes where the smaller lobe binds ATP and the larger lobe binds the peptide substrate. Catalysis occurs at the cleft between the lobes. The crystallographic and solution structural studies with PKA have indicated that the enzyme undergoes major conformational changes from an “open” form to the “closed” catalytically active form as it binds the substrates (Cox et al., 1994). These conformational changes are presumed to involve the closing of the cleft between the two lobes as the substrates bind and bring the γ-phosphate of ATP and the Ser OH in closer proximity for direct transfer of the phosphate.
However, the inhibitors of protein kinases still lack the specificity and potency desired for therapeutic use. Due to the key roles played by protein kinases in a number of different diseases, including cancer, psoriasis, atherosclerosis, and their role in regulating immune system activity, inhibitors of specific protein kinases are needed. The present invention provides a novel approach for designing protein kinase inhibitors, which are more potent as well as being more specific for the targeted pathways.
The present invention provides a method for identifying inhibitors of protein kinases. A first module having a one or more functional groups for binding to catalytic residues of the protein kinase is combined with a second module which provides a non-peptide scaffold. Combinations of the first and second modules which inhibit protein kinase activity are selected.
The present invention also provides a method of inhibiting a protein kinase. The protein kinase is contacted by a compound comprising a first module having a functionality for binding to catalytic residues of the protein kinase and a second module which provides a non-peptide scaffold. The combination of the first and second modules inhibits the protein kinase activity.
In a further embodiment, the invention provides a non-peptide protein tyrosine kinase inhibitor having the formula I:
wherein R1 is H or OH, R2 is H or OH, R3 is OH or H, and R4 is CH2, CH(CH3)(R-configuration), or CH(CH3)(S-configuration), R5 is OCH3, H, or OH, R6 is OCH3, F, H, or OH, and R7 is OCH3, H, OH, CO2H, CO2CH3, CH2CO2H, or CH2CO2CH3.
The present invention also provides a non-peptide protein tyrosine kinase inhibitor having the formula II:
wherein R1 is OH or H, R2 is OH or H, R3 is OH or H, R4 is OH or H, R5 is OH, OMe, or H, R6 is OH, OCH3, or H, R7 is OH, OCH3, or H, and X is 0 or 1.
In yet another embodiment, the present invention provides a method of treating a condition, responsive to a protein kinase inhibitor, in a patient. A protein kinase inhibitor is administered to a patient. The protein kinase inhibitor has a first module having a functionality for binding to catalytic residues of the protein kinase and a second module which provides a non-peptide scaffold. The combination of the first and second modules inhibits protein kinase activity in the patient.
The present invention provides a method for identifying inhibitors of protein kinases. The general modular strategy for the development of non-peptide PTK inhibitors is outlined in
Validation of the scaffolds at the level of Step 2 before undertaking the resource intensive combinatorial library synthesis and testing of Step 3 is important for three reasons: 1) to develop the chemistry for appending the specificity element (Sn) side chains; 2) to determine that these inhibitors are not ATP-competitive; and 3) to determine that the potency is a result of the side chain Sn properties and attachment points as would be expected based upon the working model for the src:inhibitor complex (this provides some confidence that rationally guided choices can be made for the ranges of individual selectivity elements Sn to include in the focused libraries of Step 3).
It is in Step 3 that high potency and specificity for a particular PTK is anticipated because numerous combinations of M1 functional groups (and close analogs M1 modules) with selectivity elements (Sn) will be evaluated experimentally via combinatorial chemistry and high-throughput screening. Potency and selectivity may be further increased if necessary by appending additional specificity elements (see optional selectivity elements (Sn) in
In each of the Steps 1-3, molecular modeling studies with the IRTK:peptide:AMP-PNP crystal structure, the model of the src:peptide complex and the models for the src complex with the individual families of inhibitors based upon a particular scaffold will be used as qualitative guides. These modeling studies have been remarkably helpful thus far in guiding the inhibitor design efforts as detailed later. Combining structure-based design and combinatorial chemistry technologies in this fashion provides a synergy where the major individual deficiencies of these technologies used in isolation are addressed by the strengths of the other. The major deficiency of structure-based design is the difficulty in quantitatively predicting ligand binding affinities, which is particularly challenging due to the complex effects of solvation and entropy (Ajay & Murcko, 1995). The major strength of structure-based design is its capability to predict what types of molecules are likely to be good ligands. Structure-based design can determine the rough boundaries (proteins have some flexibility which need to be taken into account) for molecular size and shape as well as indicate where hydrophobic, hydrogen-bonding and ionic interactions are likely to occur. On the other hand, the major deficiency of combinatorial chemistry is that “molecular space” for drug-sized molecules (i.e. MW approximately 500 or less) is so large that one could not hope to sample all of this molecular space with a high density of coverage in a reasonable sized combinatorial library. A recent estimate (Bohacek et al., 1996) of the number of possible compounds containing up to 30 atoms chosen only from carbon, nitrogen, oxygen and sulfur (in addition to hydrogen) is 1060 compounds. This includes compounds in the molecular weight range of typical drug molecules and still does not include additional diversity provided by other atoms, e.g. halogens. Consequently, additional constraints need to be used to identify regions of molecular space where particular drug candidates are likely to be located. Structure-based design can drastically reduce the volume of molecular space that needs to be explored by identifying the types of molecules which have a higher probability of being good ligands. The inability to quantitatively predict which of these “focused” combinatorial library members will in fact be the tightest binding ligands (i.e. the quantitation problem) is then resolved by employing an efficient combinatorial synthesis and high-throughput testing of the library.
In the earlier efforts to design peptide based serine and tyrosine kinase inhibitors, PKA was used as a convenient qualitative model for designing the protein kinase inhibitor module M1 for interaction with the conserved catalytic residues. There is much more structural and kinetic information available for PKA than any other protein kinase.
The crystal structure of PKA complexed with Mg2ATP and a pseudosubstrate (i.e. OH replaced with H) peptide inhibitor (PKI 5-24 amide) has been solved (Zheng et al., 1993) and the active site interactions near the P 0 Ala of this inhibitor are shown in
This crystal structure shows Mg2ATP bound to the small lobe of PKA and a 20-residue pseudosubstrate peptide inhibitor bound to the large lobe with the overall conformation of the enzyme in the closed (i.e. the two lobes are touching) and activated state. The distances between the P 0 Ala side chain carbon and the nearby heavy atoms in the complex are shown in AO in
Since the PKA catalytic residues Asp-166 and Lys-168 are completely conserved in all serine kinases, and the tyrosine kinases only differ by the substitution of Arg for Lys-168 (Taylor et al., 1993), this region of the active site was chosen, along with the adjoining MgATP or MgADP, to target a selection of inhibitor functional groups which could serve as M1 and be broadly useful for developing inhibitors for the entire protein kinase family. By targeting M1 to the region of the active site adjacent to the nucleotide, an orientation point is provided for the non-peptide inhibitors which can extend into the peptide binding specificity sites without always competing with ATP/ADP binding.
A selection of functional groups which could be utilized as M1 was identified first because, although this region of the active site is very highly conserved, it was expected that each particular protein kinase will still display some differing preferences across this selection due to small variations in the active site conformations and adjoining residues. Furthermore, the rank order preference among this selection of M1 modules may change somewhat as the M1 module is appended to different non-peptide scaffolds. This expectation is based upon the potential for each non-peptide scaffold to bind in somewhat different orientations with each individual protein kinase and with each particular set of selectivity element (Sn) side chains. Pentapeptide scaffolds were chosen for the initial screening of functional groups for M1 because the binding orientation of these larger peptide scaffolds is likely to be very consistent and predictable (i.e. closely resembling that observed by x-ray) throughout the series and could be more confidently assumed to position each tested M1 functionality adjacent to the conserved catalytic residues as intended. Consequently, the goal of this earlier peptide-based work was to identify a collection of M1 functional groups which can be used, not only for the initial screening of non-scaffolds (Step 1), but also as an initial set of M1 side chains which can be further expanded via close analogs and thereby optimized simultaneously with the other side chains in the final non-peptide combinatorial libraries (Step 3).
In order to model the candidate M1 functional groups in this conserved catalytic region of the PKA active site, they were built onto the P 0 Ala position in the PKA ternary structure using the SYBYL molecular modeling package (Tripos) on a Silicone Graphics workstation as indicated in
A crystal structure of PKA with MgATP and an inhibitor bound in a more “open” conformation was not available, so initial modeling studies were carried out on the MgADP bound form of PKA derived from the ternary complex illustrated in
In order to allow molecular mechanics calculations to be done with the Silicone Graphics workstation in a reasonable amount of time, two layers of residues were carved out from the PKA ternary structure which are surrounding the PKA active site, along with the peptide inhibitor and Mg2ADP. The M1 functional groups were then appended to the P 0 Ala side chain and the entire PKA active site:Mg2ADP:modified peptide inhibitor complex was then subjected to 300 iterations of molecular mechanics minimization using the Tripos force field with a distance dependent dielectric constant after assigning appropriate formal charges and calculating Gasteiger Marsili point charges using SYBYL. Setting the maximum number of iterations at 300 was sufficient to remove any serious strain in the complexes and yet not allow the overall structure to “drift” significantly from the starting x-ray structure if convergence is not reached. These minimized complexes were then visually evaluated to determine if the appended individual M1 functional groups were able to engage in favorable interactions with the conserved catalytic residues and/or Mg2ADP. This visual evaluation involved, among other standard interaction evaluations, measuring atom-atom distances to determine if hydrogen bonds and ionic interactions were being favorably formed.
Favorable intermolecular interactions between an individual M1 functionality and the conserved catalytic residues or Mg2ADP does not necessarily mean an enhanced binding affinity will be observed for the new inhibitor. Unfavorable desolvation of both the polar M1 functionality and the polar PKA active site residues (as well as complex entropy effects) are not included in this analysis and may reduce the net binding affinity to the point that the modified inhibitor may even be less potent that the corresponding P 0 Ala inhibitor, even though the appended M1 functionality is interacting with the conserved catalytic residues and/or MgADP (or MgATP) as intended. Even in cases where this desolvation penalty results in no net increase in binding affinity, these M1 functional groups are still useful as an orienting groups for correctly positioning the non-peptide inhibitor analogs in the protein kinase active site. Positioning these polar functional groups elsewhere within the active site (assuming the groups are tethered so as not to be able to extend into bulk solvent while the scaffold is favorably bound in the active site) is likely to result in a reduced binding affinity because the groups were specifically designed and selected based upon their demonstrated ability (while appropriately tethered to pentapeptide scaffolds) to be accepted adjacent to the conserved catalytic residues and MgADP/MgATP. If a particular M1 functionality does not correctly position a non-peptide scaffold in Step 1 then attempts to improve the potency by rationally appending initial specificity elements in Step 2 would likely fail.
None of the literature protein kinase assay procedures contain added ADP. A typical PKA literature assay procedure (Glass et al., 1989) was modified by adding 10% as much ADP as the ATP concentration used to reflect the natural 1/10 ratio in the cell. This protein kinase assay is hereinafter referred to as the “Literature Mimetic” assay. It has been used for PKA as well as for the cyclic AMP dependent kinase src. An examination of the literature, and commercially available protein kinase assays, showed that there is poor consistency from lab to lab and company to company and that all of these assays use physical chemical conditions which differ considerably from those known to exist inside cells. Since inhibition of intracellular protein kinases is the ultimate goal for drug discovery, new protein kinase assays have been developed which come much closer to mimicking the overall cytosolic physical chemical conditions known to exist inside cells. The development of these “Cellular Mimetic” protein kinase assays, is described herein, along with a novel method for determining which form of a protein kinase a given inhibitor binds to best (the STAIRe method). Data was collected correlating the activity of the new non-peptide src inhibitors in the Cellular Mimetic assay with that obtained in the LA25 src transformed cell line (see below).
When these two assay conditions were applied to some of the pentapeptide-based PKA inhibitors, which were designed as illustrated in
The standard pentapeptide sequence chosen for the majority of PKA inhibitors in Table I was derived from the pseudosubstrate sequence of the peptide inhibitor which was bound to PKA, when the crystal structure illustrated in
The collective results in Tables I and II show that both the serine kinase PKA and the PTK src can accommodate a variety of large polar M1 functional groups at the P 0 phosphorylation position. Furthermore, using the STAIRe methodology (see Choi et al. 1996), the sulfamic acid inhibitor 8, and related inhibitors, were shown to actually bind best when MgATP (not MgADP or no nucleotide) is also bound. This was a somewhat surprising result since these inhibitors are analogs of the “end product inhibitors” 1 and 12 which must bind simultaneously with MgADP, following phosphate transfer in the generally accepted reaction mechanism for protein kinases.
These results also demonstrate that both PKA and src can show a large difference in binding affinity for structurally very similar inhibitors. For example, the sulfamic acid PKA inhibitor 8 (Table I) has a Ki of 0.16 μM under Literature Mimetic assay conditions (L) whereas the isosteric sulfonamide 7 is 1,875 times less potent (Ki=300 μM). The sulfamic acid inhibitor 8 is also isosteric with the end product phosphate inhibitor 1, yet it binds much more tightly under both Literature Mimetic assay conditions (31 times) and Cellular Mimetic (C) assay conditions (108 times). The beneficial effect of an oxygen atom positioned analogously to the oxygen in the substrate Ser is illustrated by comparison of phosphonate 2 to phosphate 1 and also ether 6 to phosphate 1. The oxygen atom can also be positioned as a serine-like OH side chain and enhance binding (compare 2 to 3A and 4A) wherein the closer serine mimic 4A is the more active. The difference in activity of the diasteromeric inhibitors 3A or B and 4A or B suggests that a specific interaction with the active site catalytic residue Asp-166 may in fact be occurring as intended in the M1 design (
*L = Literature Mimetic
C = Cellular Mimetic
The structure identified in Table I as Ac-Arg-Arg-Gly-Ala bonded to M1-Ile-NH2 is SEQ.ID. No.3.
The src inhibition results (Table II) show that the end product inhibitor 12 drops in activity upon going from Literature Mimetic assay conditions to the higher ionic strength Cellular Mimetic assay conditions analogous to the PKA end product inhibitor 1. However, whereas all of the PKA inhibitors with polar M1 functional groups were less active under Cellular Mimetic assay conditions than three of the src inhibitors 14, 15, and 17 which held their activity under these higher ionic strength assay conditions. Also, the hydroxyphosphonate src inhibitor 13 (a mixture of the R and S diastereomers) is analogous to the PKA inhibitor 3A and both are roughly in the same activity range as their corresponding end product inhibitors, 12 and 1 respectively, under Literature Mimetic assay conditions. Shortening the side chain length in the phosphonate src inhibitor 13 by one carbon atom (and necessarily removing the attached OH at the same time) to give 14 improved the activity (analogous to the PKA inhibitor comparison 3 to 4) and, more importantly, resulted in equivalent activity under Cellular Mimetic assay conditions. The src results with 16-19 (particularly 17, see later for an analogous α-tricarbonyl acid M1 analog appended to non-peptide src inhibitors) also suggests that similar amides may be useful M1 functional groups to explore with non-peptide src inhibitors.
Non-peptide src inhibitors are preferred to peptide scaffold based compounds, partly because some of these inhibitors have a dual effect on src. For example, phosphonate inhibitor 14 not only inhibits src by competitively binding in the active site but it also activates src by binding to the SH2 site thereby releasing the intramolecular autoinhibition mechanism (Xu et al., 1997). This opposing effect gives an unusual IC50 curve for 14 wherein at low inhibitor concentrations src is stimulated (to a maximum of 70%) in a smooth dose-response fashion (due to initial tighter SH2 binding) followed by a typical IC50 inhibition curve at higher inhibition or concentration (due to lower affinity blockade of the active site). This opposing activation effect of the pentapeptide inhibitors makes the inhibitors appear to be less potent active site inhibitors than they in fact are, and makes it difficult to accurately rank M1 groups while appended to this pentapeptide scaffold. However, the better M1 groups identified with the src pentapeptide scaffold must still be accommodated in the catalytic region of the active site and hence are useful orienting groups for the ongoing non-pepitde src inhibitor studies as intended. Since PKA does not have an SH2 domain, this complication is not a factor in interpreting the PKA pentapeptide inhibitor M1 testing data.
The structure identified in Table II as Ac-Ile-Tyr bonded to M1-Gly-Glu-Phe-NH2 is SEQ. ID. No. 2.
The results in Tables I and II also show how much effect the assay conditions can have on both inhibitor potencies and the rank order of activity. For example, as shown in Table I, switching from the Literature Mimetic (L) assay conditions to the Cellular Mimetic (C) assay conditions can change the potency from as little as 3-fold (inhibitor 10) to as much as 108-fold (inhibitor 1). Also, whereas inhibitor 10 is less potent than 1 under Literature Mimetic conditions, it is more potent under Cellular Mimetic conditions. The src inhibitor data presented in Table II shows that many of the inhibitors lose their potency upon going from Literature Mimetic assay conditions to Cellular Mimetic assay conditions. The rank order of potency against src is also sensitive to the assay conditions. Whereas inhibitor 18 is more potent than inhibitor 17 under Literature Mimetic conditions, the opposite is true under Cellular Mimetic conditions. Since activity within cells is the goal, the Cellular Mimetic src assay was selected as the standard assay for testing potential non-peptide src inhibitors. Activity within the Cellular Mimetic assay is a necessary, but not sufficient, condition for activity within cells. As will be described later, the Cellular Mimetic src assay will be followed up with cell culture assays wherein cell penetration, metabolism, and binding to other cellular components are also factors in the measured potency.
The next class of M1 functionality which was explored was the boronic acid group. This functional group is an intriguing candidate for M1 for a number of reasons: 1) It can exist in a non-ionic state so that it should not prevent passive absorption of non-peptide inhibitors across cell membranes. 2) The planar, trigonal, boron acids might form reversible tetrahedral covalent borate complexes (a well known property of boronic acids, see Loomis & Durst, 1992) through their vacant 2p orbitals with anions present in the protein kinase active site such as the catalytic Asp carboxyl group or the ATP/ADP terminal phosphate oxygen atoms. This ability to form borate complexes with active site nucleophiles has been extensively utilized to develop slow binding inhibitors of serine proteases (e.g. see Kettner & Shenvi, 1984), where the nucleophilic serine OH forms a covalent bond with the vacant 2p orbital in the boronic acid resulting in a tetrahedral borate complex (e.g. see Skordalakes et al., 1997). Also, an intramolecular complex of a boronic acid with a urea NH2 was used to prepare transition state analogs inhibitors of dihydroorotase (Kinder et al., 1990). 3) Boronic acids act as Lewis acids and are converted to tetrahedral hydrates in water by forming borate complexes with water or hydroxide ions. Therefore, it is also possible that these boronic acid hydrates may function as phosphate mimics and M1 modules as proposed in
In the area of pentapeptide-based PKA inhibitors, the boronic acid functionality has been prepared and tested as a potential M1 module utilizing the four inhibitors 21-24 shown in Table III (see Hsiao & Hangauer, 1998, for some of the chemistry used to prepare these compounds).
*Very distorted IC50 curve: Suggests Inhibitor is also a substrate.
L = Literature Mimetic Assay Conditions.
Inh = Inhibition.
C = Cellular Mimetic Assay Conditions.
Sti = Stimulation.
The structure identified in Table III as Ac-RRGXI-NH2 is SEQ. ID. No. 4. While testing these boronic acid-containing PKA inhibitors, the corresponding pentapeptide pseudosubstrate inhibitor 20 was included as an internal control while investigating time-dependent inhibition as shown in Table III. Under Literature Mimetic assay conditions, and no preincubation, the initial results suggested that the shortest chain L-amino acid 21 was binding with the same affinity as the pseudosubstrate inhibitor 20 (i.e. Ki approximately 9 μM). As this side chain was increased in length (to 23 and then 24) binding affinity appeared to decrease. When the stereochemistry of the unnatural amino acid was inverted from L in 21 to D in 22, binding affinity appeared to increase 3-fold. This improvement in binding may occur as a result that the boronic acid OH in 21 which is positioned at the same chain length as L-homoserine, whereas the natural substrate, L-serine, has a one carbon shorter side chain. Modeling results with the PKA ternary structure indicated that the boronic acid OH can be retracted back somewhat by inverting the α-carbon stereochemistry from L in 21 to D in 22 and then repositioning the side chain to more closely mimic the positioning of the natural substrate L-serine OH adjacent to the catalytic residues (Asp-166 and Arg-168). The modeling results were subsequently supported by the finding that, upon incubation of PKA with these inhibitors for up to 4 hours without adding the competing peptide substrate (Kemptamide: LRRASLG-NH2) (SEQ. ID. No. 5), both 21 and 22 function as substrates with the D-diastereomer 22 being phosphorylated faster.
The fact that these boronic acid inhibitors are also substrates, became much more obvious by the greatly distorted IC50 curves obtained under the Cellular Mimetic conditions, both with and without preincubation (both PKA and src are more active enzymes under the Cellular Mimetic conditions than under Literature Mimetic conditions). In the assay used to obtain these results, the p32 phosphorylated Kemptamide product (25 generated from γ-p32 ATP) was isolated at the end of the substrate incubation period by binding to phosphocellulose filter paper via the three cationic groups (two Arg residues and the N-terminus) and the level of phosphorylated product isolated on the paper is then measured by liquid scintillation counting (cpm units). The boronic acid inhibitors 21-24 have two Arg residues in their sequence and therefore will bind to the phosphocellulose paper in addition to Kemptamide (although not as consistently or completely due to one less positive charge). Consequently, when analyzed as inhibitors, the amount of phosphorylated Kemptamide produced was not only counted, but also the amount of phosphorylated inhibitor simultaneously produced (e.g. see 26 below). The net result is that distorted IC50 curves are obtained which show net “stimulation” at higher inhibitor concentrations in some cases. The D diastereomer 22 gives the greatest apparent “stimulation” (71%) when preincubated with PKA for 4 hours under Cellular Mimetic conditions followed by the L diastereomer 21 (19%) and then the one carbon homolog 23 (5%) indicating that all three compounds are substrates for PKA (Table III). The underlying substrate behavior of these “inhibitors” makes an accurate measurement of their inhibition potency impossible with the current assay. However, it does appear from the data that homologating/extending the boronic acid functionality with CH2 groups (homologations/extensions with boronic acid non-peptide src inhibitors may also be carried out) decreases the binding affinity and ability of the compound to function as a substrate.
Phosphorylated Kemptamide is SEQ. ID. No. 6. The boronic acid “inhibitors” 21 and 22 were shown to be substrates for PKA by running the same assay, but without adding Kemptamide, and stopping the reaction at various time points as shown in
The src and PKA pentapeptide scaffold tethered M1 evaluations described above have resulted in the identification of a variety of orienting M1 groups which could be used for screening potential non-peptide scaffolds as indicated in Step 1 (
The most useful crystal structures available for the design of non-peptide src inhibitors, which do not compete with ATP, are the native src structure and the IRTK:peptide:AMP-PNP ternary structure. For all of the modeling studies discussed below, the SYBYL molecular modeling software package is used on a Silicone Graphics Workstation.
Since the src and IRTK structures are only used as qualitative guides in designing the non-peptide scaffolds and combinatorial libraries, the active sites along with two layers of surrounding residues were carved out from the native src and IRTK ternary structures, analogous to the previous PKA modeling studies. The IRTK:peptide:AMP-PNP ternary structure active site region was used as the template structure to guide the building of the src residue sequence 424-418 back onto the src structure using the comparative homology modeling technique (see Hutchins & Greer, 1991). These residues were disordered in the native src crystal structure and therefore not visible by x-ray. The residues were reintroduced because they help form the P+1 to P+3 binding sites for peptide substrates which are important for some of the modeling studies. The analogous residues in the IRTK ternary structure are seen by x-ray and directly interact with the bound peptide substrate. In fact, it is probably the presence of the bound peptide substrate which induces order in the positioning of this sequence so that it is visible by x-ray. The src pentapeptide substrate Ac-Ile-Tyr-Gly-Glu-Phe-NH2 (SEQ. ID. No. 1) (Nair et al., 1995) was then docked into the src active site again using the IRTK ternary structure as a template. Small adjustments were then manually made to partially clean up this complex, all of the hydrogen atoms were added, appropriate formal and partial charges (calculated via the Gasteiger Marsili method) were added, and then the entire complex was subjected to 300 iterations of molecular mechanics minimization using the Tripos force field, analogous to the previous PKA modeling procedure. A schematic representation of this modeled complex is given in
As shown in
The src:peptide substrate complex can now be used to model potential non-peptide scaffolds and determine preferred substitution positions for the specificity elements, all with an appropriately attached M1 functionality, before choosing new scaffolds to experimentally evaluate. The IRTK:peptide:AMP-PNP ternary structure can also be used to model these potential scaffolds and preferred substitution positions. These scaffolds have broad utility for the development of selective PTK inhibitors by further developing them with appropriate specificity elements following the strategy outlined in
The first non-peptide scaffold evaluated with this src:peptide substrate model was the naphthalene scaffold. This is the first use of bicyclic aromatic scaffolds for non-peptide PTK inhibitors, which do not compete with ATP. The utility of a naphthalene scaffold for this purpose was demonstrated by developing a non-peptide inhibitor of the IRTK and EGF receptor PTK (Saperstein et al., 1989). The IRTK ternary complexes were subsequently used to adapt this scaffold for src inhibition (see Marsilje et al., 2000). The naphthalene scaffold was docked into the src active site by first carrying out a least squares fitting of atoms a-d onto the peptide substrate as indicated in
In all of these modeled complexes selectivity element S2 consists of various hydrophobic side chains which can bind in the same pocket as the substrate P−1 Ile side chain and selectivity element S3 consists of various molecular fragments which can bind in the P+1 to P+3 region of the peptide substrate binding sites (
The src results using M1 functional groups to experimentally identify promising non-peptide scaffolds are listed in Table IV. The data in Table IV allows a number of conclusions to be drawn: 1) Low, but measurable, inhibition potency can be obtained with an appropriate M1 group attached to a scaffold (e.g. 27 and 38). 2). A 1 mM inhibitor concentration for this type of screening is higher than desirable but 100 μM is too low. Screening of compounds with scaffolds bearing an M1 group would optimally be conducted at 500 μM. 3) The boronic acid, sulfamic acid and phosphonic acid M1 functional groups, which had been identified using the PKA pentapeptide scaffold (22, Table III and 8, Table I) or the src pentapeptide scaffold (14, Table II), respectively, give measurable activity when placed at the 2-position of the naphthalene ring (27, 28 and 30, respectively), the preferred position for M1 identified in the model naphthalene inhibitor:src complex (
Since the bare scaffolds, with only an M1 group appended, often have low binding affinity, the IC50 and Ki values for the 2-naphthalene boronic acid and sulfamic acid inhibitors were determined to ensure that a typical dose/response IC50 curve is obtained. This analysis provided the typical shape dose/response curves seen with more potent inhibitors. The IC50 and Ki values of these simple inhibitors also confirmed that the boronic acid inhibitor 27 is the more potent than the sulfamic acid analog 28 and has a Ki of about 554 μM.
The next issue addressed with these simple inhibitors before proceeding to elaborate them further was their mode of inhibition, specifically whether they are ATP-competitive inhibitors. In the case of the naphthalene inhibitors 27 and 28, their IC50 values were monitored as the ATP concentration was increased in three steps up to 1 mM. As a comparison, the IC50 of the pentapeptide phosphonic acid src inhibitor 14 (Table II) was also monitored. If any of these inhibitors were competing with ATP, then their IC50 values should have increased proportionally with the ATP concentration (i.e. the dashed line). As shown, the IC50 values for all three inhibitors remained essentially constant as the ATP concentration was increased demonstrating that they are not ATP-competitive inhibitors. A very similar, but much less costly (commercial src is expensive), analysis was conducted with the indole boronic acid inhibitor 38. In this case, the % inhibition was monitored with 38 at a constant 500 μM inhibitor concentration but with increasing ATP concentrations of 200, 500 and 1,000 μM. Once again the inhibitor potency was not reduced by the increasing ATP concentration demonstrating that 38 is also non-ATP competitive.
The initial results obtained in Step 1 suggests that it is possible to identify promising scaffolds for further elaboration with this procedure. The biggest uncertainty with Step 1 is that some of the scaffolds identified in this way might not be binding in the fashion suggested by the prior modeling evaluations. This is essentially a “false positive” problem. These “false positives” will likely fail in Step 2, when they are evaluated for improved binding using the modeled complexes as a guide. Some false positive results can be accepted in Step 1 because the bare scaffolds with only the M1 group attached are easily obtained. For further inhibitor development, one may return to Step 1 each time new scaffolds are needed to carry through Steps 2 and 3. The best M1 generated can be used each time Step 1 is repeated. Currently, the boronic acid M1 group is preferred since it has a proven ability to give measurable activity with bare scaffolds. Also the boronic acid M1 group offers multiple interesting possibilities for covalent and non-covalent interactions with the conserved catalytic residues since it can: 1) hydrate, 2) form borate complexes with electron rich active site atoms, and/or 3) be phosphorylated and then react with active site nucleophiles or engage in additional non-covalent interactions. From the data in Table IV, the naphthalene and indole scaffolds were chosen as M2 for the first efforts in Step 2 (the biphenyl scaffold is also a preferred scaffold). It is also worth mentioning that naphthylalanine and analogs can be successfully substituted for the P 0 tyrosine in src peptide substrates (e.g. see Alfaro-Lopez et al., 1998) further supporting the notion that naphthalene and related scaffolds can bind at the P 0 site.
In comparing the naphthalene vs. indole scaffold results with a boronic acid M1 group (i.e. 27 vs. 38, Table IV) the indole hydrogen bond donating NH and/or the adjacent ester group appeared to be the reason for the enhanced potency. Consequently, for Step 2 one of the first attempts was to add a hydroxyl group and an amide (with S2) to the naphthalene scaffold at the adjacent positions as suggested by the modeling results (
Inhibitor 50 and analogs (Huang et al., 1995), were of particular interest because the iminochromene scaffold is closely related to the naphthalene scaffold and its binding mode would be expected to be very similar based upon the model (
The data in Table V allow the following conclusions to be drawn: 1) Adding an amide extension onto both the naphthalene and indole scaffolds can increase potency as predicted by the models for these scaffolds bound in the src active site (approximately 5-fold in the cases of 42 vs. 43-meta and 47 vs. 48). 2) Adding a hydroxyl group to the naphthalene scaffold adjacent to the amide increases potency (about 5-fold, 43-meta vs. 44) as predicted by the src model, and also suggests Asn 468 does hydrogen bond with this OH. 3) Moving the M1 OH group from the position predicted to be best in the src model to the adjacent position reduces potency by one order of magnitude (43-meta to 45). 4) The indole scaffold is less responsive than the naphthalene scaffold to the regiochemistry of the hydroxyaniline extension (48 vs. 43). 5) The naphthalene and the indole scaffolds accept the one carbon homologation provide by using hydroxybenzylamides (46 vs. 43 and 49 vs. 48). 6) The two M1 hydroxy regioisomers of the naphthalene scaffold are both non-ATP competitive (see Marsilje 2000). 7) All of the methyl hydroxyaniline and hydroxybenzylamide inhibitors were found to be less active suggesting that the hydroxyl group in the amide extension is functioning as a hydrogen bond donor. In this regard, it is worth mentioning that in another src peptide substrate combinatorial library study Ser and Thr were identified as two of the most preferred residues at the P+2 position (Alfaro-Lopez et al., 1998) suggesting that there are other binding opportunities for an amide extension OH other than to the Phe424-Ala 422 peptide bonds suggested by the modeling studies. 8) The most potent non-ATP competitive, non-peptide, src inhibitor previously disclosed in the literature (50) is not nearly as potent as reported when tested under the Cellular Mimetic assay conditions (IC50=118 nM reported by Huang et al., 1995 vs only 30% inhibition at 100 μM) and is less potent than a number of the current inhibitors (e.g., 43-meta) including the most analogous inhibitor (50 vs. 45). The structure-activity-relationship (SAR) reported for hydroxy regioisomers of 50 in their assay (Huang et al., 1995) also does not correspond with the SAR which was obtained for the related naphthalene inhibitors. For example, the iminochromene analog of the most potent naphthalene inhibitor 43-meta is 230-fold less potent than 50 in their src assay. An important advantage of the naphthalene scaffold over the iminochromene scaffold is that it allows a highly desirable S2 specificity element to be added for accessing the P−1 hydrophobic site (see
The inhibitor potencies in the src Cellular Mimetic assay can be further calibrated against other literature non-ATP, non-peptide src inhibitors. Two additional examples are 51 (ST 638, Shiraishi et al., 1989) which is a member of the “tyrphostin” family of erbstatin analogs (see Lawrence & Niu, 1998) and the natural product PTK inhibitor piceatannol 52 (Thakkar et al., 1993). In the Cellular Mimetic assay, all of these known inhibitors are less potent than had been reported suggesting that the assay is particularly demanding in terms of achieving high potency. The initial testing of src inhibitors is carried out using a single concentration (in triplicate) because commercial src is too expensive to do full IC50 curves on every inhibitor. It should be mentioned, however, that an IC50 dose response curve is not linear and the difference between approximately 50% inhibition at 100 μM and a approximately 90% inhibition at 100 μM is actually a factor of 10 and not a factor of 2 (e.g. 45 vs. 43-meta). Consequently, the literature src inhibitors 50-52 are greater than an order-of-magnitude less active than the currently most potent inhibitor 43-meta.
The discrepancies found within the literature reporting the potency of these inhibitors, the sensitivity to assay conditions described earlier with the PKA inhibitors, and the lack of consistency among numerous labs and commercial protein kinase assay kits highlights this overlooked, but crucial, problem in the field. Although the inhibitors produced by the present invention may be more potent under other assay conditions, the Cellular Mimetic assay should be used, which mimics the intracellular physical chemical conditions as closely as possible, as the primary potency and rank order guide for evaluating the inhibitors before choosing compounds to proceed to whole cell or tissue assays. As will be discussed in more detail later, the most potent naphthalene-based inhibitor thus far from the Cellular Mimetic assay (i.e. 43-meta, IC50=18 μM and Ki=10 μM) is also effective in specifically blocking pp60v-src stimulated cell proliferation with a similar IC50 of approximately 25 μM. This result suggests that not only is the Cellular Mimetic src assay predictive, but also that this class of naphthalene-based inhibitors can readily pass through cell membranes and inhibit intracellular src.
Analogs of a number of the naphthalene and indole inhibitors above can be prepared with the boronic acid M1 group in place of the M1 OH and/or with a S2 hydrophobic specificity element attached for binding in the src P−1 site as illustrated in
Some of the chemistry used to prepare the naphthalene inhibitors is described in Marsilje 2000. For attaching a boronic acid functionality in place of a M1 hydroxyl groups in the src inhibitors from Table V, the Pd (0)-catalyzed cross-coupling methodology was used wherein either an aryl triflate (Ishiyama et al., 1997) or an aryl halide (Ishiyama, 1995) can be coupled with the commercially available pinacol ester of diboron. An illustrative example was recently completed as shown in
The example shown in
The naphthalene chemistry can be converted to the solid phase in preparation for synthesizing combinatorial libraries of this scaffold in a 96-well plate format. Thus far, model chemistry has been carried out on the less active naphthalene regioisomer represented by 44 because this compound is readily obtained from commercially available 3,5-dihydroxy-2-naphthoic acid as describe in Marsilje 2000. The successful model reactions to date are shown in
These model reactions demonstrate that it is possible to couple the naphthalene scaffold to the Wang resin (63) and then carry out chemistry on the triflate [e.g., in this case the Pd (0)-catalyzed cross-coupling to the boronic ester 64 ] followed by cleavage under mild conditions (65). The ester in 63 can also be saponified for subsequent coupling reactions to form amides containing the S3 selectivity elements.
The naphthalene scaffold currently provides three diversity sites to be explored in the combinatorial libraries, M1, S2 and S3. Solid phase combinatorial chemistry with 96-well plate reactors is similar to that used in previous studies (Pavia et al., 1996). The greatest number and diversity of side chains will be used for S3 followed by S2 and then M1 for the reasons discussed earlier. One possible overall synthetic strategy, based upon the synthetic model studies above, for preparing these libraries is shown in
Of course if problems arise with this route, there are many other possibilities. For example, if the Mitsunobu coupling to give 67 proceeds in too low a yield (due to the increased steric congestion of the added adjacent allyl group-but perhaps not a problem given the 92% loading obtained in
At least 14 S2 hydrophobic side chains (includes linear, branched and cyclic) are identified for further study (28 side chains, if the corresponding alkenes are also explored) based upon the modeling of candidate side chains into the P−1 site of the src model (
The indole scaffold can be developed into a combinatorial library in much the same fashion. In this case, the indole NH would be used as the tether point for attachment to the Wang (or other) resin since the analogous Minsunobu reaction is known (Bhagwat & Gude, 1994). A large amount of synthetic methodology has been developed for the synthesis of substituted indoles and have designed a route to include the S2 hydrophobic side chain (see
The triflate functionality formed in reaction 2 from intermediate 69 (
When the amine 72 is available, the known M1 modules (e.g. the sulfamic acid from src inhibitor 28 Table V and amide-acid 17 Table III) can be evaluated with this more developed scaffold and evaluate some new amine derivatives as potential M1 modules. For example the hydrated tricarbonyl amide M1 group shown in structure 73 (and its non-hydrated precursor) is accessible via the synthetic methodology (see Lai et al., 1996) and could form a variety of interesting interactions with the conserved catalytic residues.
Following the modeling procedure described above, a the series of hydroxy-containing analogs of the boronic acid M1 group shown in
In a preferred embodiment of the invention, the first module is produced by attaching the first module to a peptide scaffold. One or more functional groups are identified which preferentially bind to catalytic residues of the protein kinase. Further, the first module is combined with the second module so that the second module substitutes for the peptide scaffold.
Preferred first modules have a functional group such as boronic acid, a hydroxyl group, phosphonic acid, sulfamic acid, a guanidino group, carboxylic acid, an aldehyde, an amide, and hydroxymethylphosphonic acid. The compounds of the present invention may have two or more functional groups within the first module. More preferred modules are boronic acid groups, a hydroxyl group, or an amide group. An even more preferred amide group is a vicinal tricarbonyl amide.
Preferred second modules include indole, naphthalene, biphenyl, isoquinoline, benzofuran, and benzothiophene. More preferred second modules are an indole or naphthalene. In some embodiments of the invention more than one first module may be bound to the second module. In addition, the first module may have a linear chain comprising between one and three carbon atoms which links the first module to the second module. In alternative embodiments, one of the carbon atoms in the linear chain is substituted with a nitrogen, oxygen, or sulfur atom.
The methods and compounds of the invention are broadly applicable to any protein kinase. Preferred protein kinases are protein tyrosine kinases and protein serine kinases. Preferred protein tyrosine kinases are pp60c-src, p56lck, ZAP kinase, platelet derived growth factor receptor tyrosine kinase, Bcr-Abl, VEGF (vascular endothelial growth factor) receptor tyrosine kinase, and epidermal growth factor receptor tyrosine kinase, and epidermal growth factor receptor-like tyrosine kinases. A more preferred protein tyrosine kinase is pp60c-src. Preferred serine protein kinases include MAP (mitogen activated protein) kinase, protein kinase C, and CDK (cyclin dependent protein kinase).
The method of the present invention may further consist of adding one or more specificity side chain elements to the combination of the first and second modules. Specificity side chains can increase potency and specificity of the inhibitor.
Once a promising second module is identified it is not necessary to repeat all the steps of the method. Rather, the first module, specificity side chains, or a combination of the two may be modified to improve the original inhibitor, i.e an inhibitor which has an increased ability to inhibit protein kinase activity when compared to the unmodified first inhibitor.
The present method is designed to preferentially provide protein kinase inhibitors which do not act by inhibiting ATP binding to the protein kinase. Inhibitors of protein kinases may be potent but often lack specificity and are therefore often not good drug candidates. Therefore, protein kinase inhibitors which inhibit protein kinase activity but do not inhibit or only weakly inhibit ATP binding to the protein kinase are preferred.
The present invention also provides a method for testing compounds for an ability to inhibit protein kinase activity. Compounds are produced as described above. The activity of the protein kinase is measured in the presence of the inhibitor at the same temperature, pH, ionic strength, osmolarity, and free magnesium concentration as found in a cell which expresses the protein kinase. The level of protein kinase activity is compared to the level of activity from the protein kinase without the presence of the inhibitor. Such an assay system which mimics physiological conditions provides the most relevant inhibition data. The assay may be conducted in an automated assay system. Furthermore, the assay may be combined with a combinatorial chemistry method to rapidly screen numerous candidates.
The Pierce 96-well plate non-radioactive ELISA PTK assay method may be adapted to the Cellular Mimetic assay conditions for initial src screening of the 96-well plate combinatorial libraries. This high throughput assay utilizes the same RR-SRC peptide substrate, except that it is biotinylated so that it can be attached to the NeutrAvidin-coated wells in their commercial 96-well plates. This high throughput inhibition assay can be run by incubating src with the RR-SRC substrate pre-bound to the wells followed by adding their anti-phosphotyrosine antibody (PY20)-horseradish peroxidase (HRP) conjugate and their HRP substrate to quantitate the level of phospho-RR-SRC produced via measuring the level of HRP product with a 96-well plate UV reader. Standard low throughput P32-ATP radioactive assays have been used, but a 96-well plate format is preferred, especially with a non-radioactive assay if possible. As very potent src inhibitors are developed, a panel of protein kinase assays could be set up with up to approximately 6 commercially available protein kinases (mostly PTKs), using the Cellular Mimetic protein kinase assay conditions, and test these inhibitors across the panel to obtain an initial assessment of specificity. A more complete specificity assessment, involving the full approximately 2,000 protein kinases, will need to be conducted in cell culture and in vivo through additional collaborations at the appropriate time.
Active src inhibitors can be studied in a set of side-by-side cell-based assays using normal rat kidney (NRK) cells and a temperature-sensitive pp60v-src transformant of this cell line (LA25). The LA25 transformant engages in anchorage- and serum-independent growth at the “permissive” temperature of 33° C. due to activation of pp60v-src but not at the “non-permissive” temperature of 40° C. at which pp60v-src is not activated (Li et al., 1996). The use of this pair of closely related cell lines for testing the src inhibitors at both the permissive and non-permissive temperatures allows one to determine if a given src inhibitor is blocking cell growth due to specific blockade of the src signaling pathway, by a different mechanism or by a general cytotoxic effect. Results from initial testing of the non-peptide src inhibitor 43-meta (Table V) in this pair of cell lines are shown in
As shown in this graph the growth of the LA25 cells at the permissive temperature of 33 ° C. is inhibited by approximately 50% at a 25 μM concentration of 43-meta relative to the LA25 cell growth at the non-permissive 40 ° C. as a control. The lack of cell toxicity of 43-meta is evidenced by the fact that as its concentration is increased up to 400 μM, the basal growth of the NRK non-transformed cells, the LA25 cells at the non-permissive 40 ° C. and the LA20 cells at the permissive temperature of 33 ° C. (but with pp60v-src fully inhibited by 43-meta) not only does not decrease but actually increases somewhat (presumably due to a non-src related activity of this compound). Since the 43-meta solutions were prepared with a low concentration of DMSO for solubilization, a DMSO control was also run at the same concentration. A more complete dose/response curve centered around 25 μM will be run.
In another embodiment, the present invention provides a method of inhibiting a protein kinase. The protein kinase is contacted with a compound having a first module which has a functionality for binding to catalytic residues of the protein kinase and a second module which provides a non-peptide scaffold. The combination of the first and second modules inhibits the protein kinase activity.
A preferred non-peptide protein tyrosine kinase inhibitor provided by the present invention has the following formula I:
wherein R1 is H or OH, R2 is H or OH, R3 is OH or H, and R4 is CH2, CH(CH3)(R-configuration), or CH(CH3)(S-configuration), R5 is OCH3, H, or OH, R6 is OCH3, F, H, or OH, and R7 is OCH3, H, OH, CO2H, CO2CH3, CH2CO2H, or CH2CO2CH3. In a more preferred embodiment, the non-peptide protein tyrosine kinase inhibitor inhibits the activity of pp60c-src tyrosine kinase.
Another preferred non-peptide protein tyrosine kinase inhibitor has the following formula II:
where R1 is OH or H, R2 is OH or H, R3 is OH or H, R4 is OH or H, R5 is OH, OMe, or H, R6 is OH, OMe, or H, R7 is OH, OMe, or H, and X is 0 or 1. In a more preferred embodiment, the non-peptide protein tyrosine kinase inhibitor has the above structure and R1 is OH, R2 is OH, R3 is H, R4 is H, R5 is OMe, R6 is H, R7 is H, and X is 1.
The present invention further provides a method of treating a condition, responsive to a protein kinase inhibitor, in a patient. An effective dose of a protein kinase inhibitor is administered to a patient. The protein kinase inhibitor has a first module having a functionality for binding to catalytic residues of the protein kinase and a second module which provides a non-peptide scaffold, where the combination of the first and second modules inhibits protein kinase activity.
Finally, promising src inhibitors can be screened in primary human tumor tissue assays, particularly to look for synergy with other known anti-cancer drugs.
The crystal structure of the autoinhibited human IRTK catalytic domain (Hubbard et al., 1994) was used to carry out qualitative molecular modeling studies (SYBYL™, 6.4, Tripos Inc., St. Louis) wherein a naphthalene ring was superimposed upon the IRTK Tyr 1,162. The IRTK region containing Tyr 1,162 folds back into the active site, with Tyr 1,162 positioned analogous to a phosphorylated Tyr in a peptide substrate, thereby autoinhibiting the tyrosine kinase. This superimposition indicated that an amide carbonyl should be placed at the 2-position (Scheme 1) of the
naphthalene ring to mimic the Tyr 1,162 carbonyl and a hydroxyl group should be positioned at the 6-position to mimic the Tyr 1,162 hydroxyl group. These modeling studies also indicated that a hydroxyl group at the 3-position could mimic the Tyr 1,162 NH.
In order to test these design concepts experimentally, the 2-position carbonyl group was appended as either a methyl ester or as a series of amides (Table VI). The hydroxy N-phenyl (X=0) and N-benzyl (X=1) amides were chosen based upon the increase in pp60c-src inhibitor potency observed with iminochromene analogs containing appended hydroxy N-phenyl amide side-chains (Huang et al., 1995). Analogs where the 6-hydroxyl group was either deleted or moved were also prepared to determine if a drop in potency occurs as predicted from the modeling studies.
The series of 2-carbonyl-3,5-dihydroxy naphthalene inhibitors (1a, 2a-2d, 2i-2l, 2o-2p) and 2-carbonyl-3,7-dihydroxy naphthalene inhibitors (1c, 2t-2u) were synthesized from commercially available (Aldrich) 3,5-dihydroxy-2-naphthoic acid and 3,7-dihydroxy-2-naphthoic acid, respectively. The methyl esters 1a and 1c were obtained by refluxing the respective acid starting materials for 48 h in methanol pre-saturated with HCl gas. The amides (2a-2d, 2i-2l, 2o-2p, 2t-2u) were synthesized by coupling the respective carboxylic acid with commercially available (Aldrich or Lancaster) amines using one of two methods. The first method utilized the NBS/PPh3 methodology as described by Froyen (Froyen, 1997). The second method utilized IIDQ (Aldrich) as the coupling reagent. The carboxylic acid was first reacted with 1.0 eq. IIDQ in anhydrous DMF at room temperature for 24 h. The respective amine (2.0 eq.) was then added neat and the reaction was heated to 80 ° C. for 2-6 hours. After aqueous workup, purification was achieved by silica gel chromatography and precipitation from CH2Cl2/hexane, followed by preparative C-18 RP-HPLC (CH3CN/H2O), if necessary. The benzyl amines were commercially available only as their corresponding hydroxyl protected methyl ethers. Consequently, after amide formation, the hydroxyl groups were deprotected by treatment with 6 eq. BBr3 in DCM for 1 minute at −78° C. followed by 1 hour at room temperature.
Table VI Footnotes:
aThe previously described assay procedure (Lai et al., 1998) was used with the following assay components, final concentrations and conditions: 50.0 mM MOPS, 4.02 mM MgCl2, 6.00 mM K3citrate (used as a Mg2+ buffer to stabilize the free Mg2+ at 0.5 mM), 99.0 mM KCl, 10.0 mM 2-mercaptoethanol, 198 μM ATP, 19.8 μM ADP, 10 U full length human purified recombinant pp60c-src (Upstate Biotechnology
bAll new compounds were characterized by proton NMR, EI or FAB(+) MS and are pure by TLC.
cN/A = Not applicable, n.t. = Not tested.
dATP-competitive.
The series of 2-carbonyl, 3,6-dihydroxy naphthalene inhibitors (1b, 2e-2h,
2m-2n, 2q-2s) were synthesized from 3,6-dihydroxy-2-naphthoic acid 6 using the methods described above. The synthesis of intermediate 6 that was developed is shown in Scheme 2 beginning with commercially available 2,7-dihydroxynaphthalene 3 (Aldrich).
Compound 1d was synthesized from 3-amino-2-naphthoic acid (Aldrich) by reaction with TMS-diazomethane in DCM at room temperature. Compound 2v was synthesized from 6-hydroxy-2-naphthoic acid (Aldrich) using the amidation method described by Froyen (Froyen, 1997).
Kinase assay conditions have been shown to influence the measured inhibitory activity (Lawrence et al., 1998). Consequently, in order to accurately determine the relative potency of the newly designed class of pp60c-src inhibitors, the inhibitory activity of four previously published, non-ATP competitive PTK inhibitors, was also tested. Piceatannol, ST-638, and Tyrphostin A47 were chosen because they are commercially available (Sigma or Calbiochem) and are representative of the spectrum of known non-ATP competitive PTK inhibitors. Emodin (Calbiochem) is ATP-competitive when analyzed with the tyrosine kinase p56lck. Previously, iminochromene 9TA was the most potent non-ATP competitive pp60c-src inhibitor reported (Huang et al., 1995). Since iminochromene 9TA was not commercially available, it was synthesized using a novel route by converting 3-aminophenol to the corresponding TBDMS ether (1.1 eq. TBDMS-Cl, 1.1 eq. DIEA, 5 mol % DMAP, DMF, 24 h, rt, 71%). The resulting aniline was coupled using 2.0 eq. of cyanoacetic acid (1.1 eq. EDCI, 1.1 eq. TEA, DMF, 18 h, 75° C., 70%). Condensation of the resulting amide with 1.2 eq. of 2,3-dihydroxybenzaldehyde (cat. piperidine, abs. EtOH, 2 h, 60° C.) followed by deprotection (1.1 eq. TBAF, THF, 15 m, 43% overall) gave iminochromene 9TA with satisfactory elemental, FAB(+)MS and 1H NMR analysis after purification by flash chromatography (10: 1, DCM:MeOH).
The inhibitory activities shown in Table VI for compounds 1a-d and 2a-2v were determined using purified, full length, human recombinant pp60c-src. Due to the number of compounds tested, and the associated cost, their rank order potencies were first determined at a constant inhibitor concentration (100 μM). As predicted by the modeling studies, based upon analogy to the IRTK Tyr 1,162 hydroxyl group, a preference for positioning the naphthalene hydroxyl group on carbon 6 vs. 5 or 7 was observed in both the ester (1b, 47% vs. 1a, 5% & 1c, 19%) and amide (e.g. 2f, 89% vs. 2b, 51% & 2t, 68%) series. The prediction that attaching a hydroxyl group at naphthalene carbon 3 (mimicking the Tyr 1,162 NH) would improve potency was also confirmed (2f, 89% vs. 2v, 45%). Finally, the prediction that extending the inhibitor as an amide at the 2-position (mimicking the peptide bond) could further improve potency was confirmed as well (e.g. 2f, 89% vs. 1b, 47%).
The data provided in Table VI demonstrate that moving the hydroxyl group from the optimal 6-position to the adjacent naphthalene carbon 5 results in a different structure activity profile with regard to the optimal concurrent positioning of the hydroxyl group(s) in the amide side chain (e.g. 2f/2g vs. 2b/2c). Also of note is the replacement of the amide side chain hydroxyl group with a corresponding methoxy group in compounds 2i-2n. In the case of the N-phenyl amides (2i-2j), their activity, relative to the corresponding hydroxy amides (2b-2c), was not reduced as significantly as in the case of the N-benzyl amides (2k-2n vs. 2o-2q, 2s). This suggests that in the benzyl derivatives, the amide side chain hydroxyl groups either interact with the enzyme as hydrogen bond donors, or the methoxy groups are too large to fit in the binding site.
A more quantitative analysis of the selectivity for positioning a hydroxyl group on carbon 6 vs. 5 is provided by comparing the IC50 values of 2f (16 μM) vs. 2b (150 μM), respectively. These results also confirm that a drop in % inhibition from approximately 90% to approximately 50% represents an order of magnitude difference in potency, as expected. Similarly, a drop in % inhibition from approximately 50% to 10% would represent another order of magnitude difference in potency.
A direct comparison of the most potent inhibitor from this series, compound 2f, with the five previously reported PTK inhibitors shown in Table VI demonstrates that, under these assay conditions, 2f is more potent by one to two orders of magnitude. Interestingly, iminochromene 9TA was previously reported (Huang et al., 1995) to have an IC50 of 118 nM against pp60c-src and was the most potent known non-ATP competitive pp60c-src inhibitor, but under the current assay conditions only a 30% inhibition at 100 μM was observed. These results re-emphasize (Lawrence et al., 1998) the importance of comparing protein kinase inhibitors under identical assay conditions.
A goal of these studies was to obtain non-peptide pp60c-src inhibitors which do not compete with ATP. Consequently the % inhibition of pp60c-src by 2f and 2b at constant inhibitor concentrations was monitored as a function of increasing [ATP] up to a cellular mimetic 1 mM level. Since the [ATP] had little effect on the % inhibition, both 2f and 2b are non-competitive inhibitors with respect to ATP. The % inhibition was measured using ATP concentrations of 200, 500 and 1,000 μM while holding the inhibitor concentration constant. If the inhibitor is directly competing with ATP, then this 5-fold overall increase in [ATP] is equivalent to decreasing the inhibitor concentration 5-fold in terms of the effect on % inhibition. Consequently, the % inhibition should decrease to the value observed in the IC50 dose-response curve (obtained with 200 μM ATP) for ⅕ of the set inhibitor concentration used in this experiment if direct competition with ATP is occurring. For inhibitor 2f (set at 25 μM) a 62% (±5), 54% (±3) and 50% (±1) inhibition at 200 μM, 500 μM and 1,000 μM ATP, respectively, was obtained whereas the level of inhibition should have dropped to approximately 20% at 1,000 μM ATP if direct competition with ATP were occurring. Similarly, for inhibitor 2b (set at 300 μM) an 84% (±1), 81% (±1) and 77% (≅2) inhibition at 200 μM, 500 μM and 1,000 μM ATP, respectively, was obtained. The high cost of many kinases has stimulated other researchers to monitor inhibitor potency as a function of increasing [ATP] for the same purpose (Saperstein et al., 1989; Burke et al., 1993; Davis et al., 1989; Davis et al., 1992; Faltynek et al., 1995; and Sawutz et al., 1996).
In summary, structure-based design has produced a series of hydroxynaphthalene pp60c-src non-peptide inhibitors that do not compete with ATP. An extension of these design concepts from the naphthalene scaffold to an indole scaffold is reported in the following Example.
In the preceding example, the structure-based design of a series of pp60c-src inhibitors utilizing a naphthalene scaffold is described. These compounds were designed to bind in the peptide substrate site because of the potential for greater selectivity and efficacy in a cellular environment relative to the alternative ATP
substrate site. This example presents an extension of these design concepts to a series of pp60c-src inhibitors based upon an indole scaffold. Once again the crystal structure of the autoinhibited insulin receptor PTK (IRTK) was used to carry out qualitative molecular modeling studies, except in this case an indole ring was superimposed upon the IRTK Tyr 1,162. This superimposition indicated that the indole NH can mimic the Tyr 1,162 NH, a carbonyl should be placed at the 2-position, and a hydroxyl group should be placed at the 5 position to mimic the Tyr 1,162 carbonyl and OH, respectively (Scheme 1). The conceptual cyclization of Tyr 1,162 to the smaller 5-membered ring of an indole illustrated in Scheme 1, relative to a 6-membered ring in the case of the naphthalene scaffold (Karni et al., 1999), results in a movement of the optimal positioning of the OH from carbon 6 in the naphthalene scaffold to carbon 5 in the indole scaffold.
The indole amide derivatives containing hydroxy phenyl/benzyl side chains 2d-f, 2j-l (Table VII), respectively, were selected based upon the increase in pp60c-src inhibitor potency observed for the analogous naphthalene-based hydroxy phenyl amides reported in the previous example. The corresponding methyl ethers 2a-c,g-i,v are precursors in the synthesis. The additional analogs shown in Table VII were prepared to begin expanding the range of side chains beyond the hydroxy/methoxy groups that have now been extensively probed with both the indole and naphthalene scaffolds.
The indole amides containing only hydroxy or methoxy side chains were synthesized as illustrated:
The 2-indolecarboxylic acid derivative, the methoxyphenyl amine (1.1 eq, Aldrich, Lancaster or Fluka), and the coupling reagent PyBOP (benzotriazol-1-yloxy)tripyrrolidino-phosphonium-hexafluorophosphate) (1 eq, Fluka) were dissolved in anhydrous DMF. The solution was cooled to 0° C. under argon and then diisopropylethylamine (DIEA, 3 eq) was added. The reaction was stirred at 0° C. for 1 min followed by 1 hour at room temperature. After workup the residue was purified by silica gel chromatography.
The methyl ethers were cleaved with boron tribromide (1 M in DCM, Aldrich) when desired. The indole amide methyl ether was suspended in dry DCM and cooled to −78° C. under argon. One equivalent of BBr3 was added for each heteroatom in the starting material plus one excess equivalent. The resulting dark red solution was stirred at −78° C. for 30 min and then at room temperature for 1-2 hours. The reaction was quenched with water (10 min) before workup.
aAll compounds were tested as described in the preceding Example.
bAll compounds were characterized by proton NMR, FAB(+) MS and are pure by TLC.
cN/A = Not applicable.
Using this synthetic route, the series of 5-hydroxyindole amide inhibitors 2a-m,y,z were prepared from 5-hydroxy-2-indolecarboxylic acid. The 4- and 6-hydroxyindole amides (2x,u, respectively) were synthesized from methyl 4-methoxy-2-indolecarboxylate and methyl 6-methoxy-2-indolecarboxylate, respectively. The 5,6-dihydroxyindole amide 2t was prepared from ethyl 5,6-dimethoxyindole-2-carboxylate. Sonication of the esters in 1 N NaOH for 1 h provided the corresponding carboxylic acids for coupling. The des-hydroxy indole amides 2v,w were synthesized from indole-2-carboxylic acid. All of the indole starting materials were commercially available (Aldrich or Lancaster).
The fluoro inhibitors 2r,s were likewise prepared from the corresponding fluorophenyl amines (Aldrich). The inhibitors containing esters or carboxylic acids on the amide side chain, 2n-q, were prepared from the corresponding amino carboxylic acids (Aldrich). The side chain carboxylic acid was first protected as a methyl ester (anh. MeOH pre-saturated with HCl, reflux, 1d), followed by PyBOP coupling (as above), then saponification back to the carboxylic acid when desired.
The methyl ester 1a was prepared by refluxing a solution of the carboxylic acid overnight in anhydrous methanol pre-saturated with HCl gas. The ethyl ester 1b was prepared by BBr3 deprotection of ethyl 5,6-dimethoxyindole-2-carboxylate as above. All of the inhibitors listed in Table VII were purified by silica gel chromatography.
As in Marsilje 2000, the rank order activity of this series of pp60c-src inhibitors was first determined at a constant inhibitor concentration (Table VII). The same inhibitor concentration (100 μM) was used for the current indole series of inhibitors, the previous naphthalene series of inhibitors, and five non-ATP competitive literature PTK inhibitors. This allowed an efficient rank order comparison of 59 compounds in total under identical assay conditions.
The modeling studies predicted that a hydroxy group at carbon 5 of the indole scaffold would be optimal. Comparison of the 5-hydroxy indole inhibitor 2k (74%) with the analogous 6-hydroxy indole inhibitor 2u (56%) and 4-hydroxy indole inhibitor 2x (60%) confirms this prediction, although the preference is not strong. The prediction that a hydroxy group at carbon 5 will improve the activity (relative to no hydroxy group) is confirmed by comparing the 5-hydroxy indole inhibitor 2l (54%) with the corresponding des-hydroxy inhibitor 2w (36%).
Extending the indole inhibitors as aryl amides at carbon 2 improved potency, as expected based upon the previous naphthalene inhibitors. For example, the meta-hydroxybenzyl amide indole 2k gives 74% inhibition at 100 μM whereas the analogous methyl ester 1a gives only 40% inhibition at 500 μM. Interestingly, comparing the 5,6-dihydroxy ethyl ester 1b (28%) to the corresponding aryl amide 2t (26%) shows that the simultaneous presence of the second hydroxy at carbon 6 prevents the potency enhancement normally provided by the otherwise preferred meta-hydroxybenzyl amide side chain. This amide side chain was the best of the current series when the 5-hydroxyl group is present alone (2k, 74%) and still gave good inhibition when a 6-hydroxy group was present alone (2u, 56%). Also, the simultaneous presence of two hydroxy groups at carbons 5 and 6 seems well tolerated in the absence of an amide side chain (1b vs. 1a and 2e). This data suggests that a change in the binding orientation of the indole scaffold may have occurred due to the presence of the second hydroxy group and that a different amide side chain may now be preferred. The optimal combination of side chains at carbons 4-7 (including functional group replacements for hydroxy groups (Lai et al., 1999)) and amide side chains is currently under investigation.
In general, the indole scaffold structure-activity-relationships (“SARs”) revealed by the data in Table VII parallels that reported in the preceding paper for the naphthalene scaffold. In both cases positioning a hydroxy group on the scaffold analogous to the Tyr 1,162 OH, as identified by modeling studies, provided the highest potency. Moving this hydroxy group to one of the adjacent carbons reduced the potency, but not dramatically, in both cases. Extending both scaffolds with aryl amides at the position identified by the modeling studies to mimic the Tyr 1,162 peptide bond improved the potency. With both scaffolds, substitution of a methoxy group for the hydroxy groups on the amide side chain usually reduced potency, and did so to a greater extent with the longer benzylamide side chain (e.g. 2k, 74% vs. 2h, 7% compared to 2e, 30% vs. 2b, 21%). The major difference in the SARs for these two scaffolds is that the 5-hydroxyindole scaffold prefers the longer m-hydroxybenzyl amide side chain (2k, 74% vs. 2e, 30%) whereas the analogous 3,6-dihydroxynaphthalene scaffold prefers the shorter amide side chain derived from m-hydroxyaniline. The 5-hydroxyindole scaffold showed essentially no preference for the position of the hydroxyl group on the shorter amide side chain (2d-f) whereas with the longer hydroxybenzyl amide side chain a significant preference for the meta position was observed (2j-l). In the case of the 3,6-dihydroxynaphthalene scaffold the opposite was observed.
Additional molecular modeling studies were carried out to further probe the preference for a longer amide side chain with the indole scaffold. The most active naphthalene inhibitor 3 from the previous report was used as a template upon which the analogous indole inhibitor 2e and the homologated indole inhibitor 2k were superimposed. The three most important side chain functional groups in naphthalene inhibitor 3 are considered to be the 6-hydroxy group (hydrogen-bond donor and
acceptor), the hydrogen from the 3-hydroxy group (hydrogen-bond donor), and the side chain hydroxy group (hydrogen-bond acceptor) based upon the rational design and SAR for both series of inhibitors. This three point pharmacophore model is identified in both series by asterisks in Scheme 3.
The “multifit”, energy minimization and “fit atoms” facilities within SYBYL™ (6.4, Tripos, St. Louis) were used in sequence to superimpose 2e and 2k onto 3. This overall fitting process was carried out with spring constants (multifit) and weights (fit atoms) chosen such that the highest emphasis was on optimally superimposing the scaffold pharmacophore oxygen and hydrogen atoms (100), followed by the side chain oxygen atoms (10) and then the intervening amide bond (1). The “multifit” process adjusted the conformations for maximum pharmacophore fit, the subsequent minimization produced the nearest local minimum energy conformations and finally the “fit atoms” process produced the best pharmacophore superimposition of these minimized conformations. As expected, the scaffold pharmacophore oxygen and hydrogen atoms of both 2e and 2k superimposed closely and similarly upon the corresponding atoms in 3 (all within approximately 0.50 Ao). However, the side chain pharmacophore oxygen atoms of 2e and 2k differed significantly in their superimposition on the corresponding oxygen atom of 3, with displacements of 1.8 Ao vs. only 0.08 Ao respectively. This close fit of the three key pharmacophore sites between 2k and 3 provides a rationalization for their potency differing by only a factor of 2.4 (IC50 values of 38 μM vs. 16 μM, respectively).
Extending the amide side chain by another carbon atom reduced the activity (2m, 21% vs. 2l, 54%). Adding a methyl group to the benzylic carbon of 2k, in either stereochemistry, greatly reduced the activity (2y, 15% & 2z, 13% vs. 2k, 74%). Replacing the side chain hydroxy group (in the para position) with a carboxylate anion (2n, 0% vs. 2l, 54% and 2p, 7% vs. 2f, 45%) reduced the activity, whereas the corresponding methyl esters (2o, 11% & 2q, 32%, respectively) showed a smaller loss of potency. Importantly, replacing the side chain hydroxy group with a fluorine maintained much of the potency (2s, 57% vs. 2k, 74% and 2r, 21% vs. 2e, 30%). Consequently, the fluoro analog 2s has only one hydroxy group left for potential Phase II metabolism (e.g. glucuronide formation), and this remaining hydroxy group is a current target for replacement (Lai et al., 1998).
Using the same method as in the preceding example (Marsilje2000), the most potent inhibitor from the current indole series (2k) was analyzed for ATP competition by monitoring the % inhibition at increasing [ATP] while holding the inhibitor concentration constant. Since the [ATP] had little effect on the % inhibition (The % inhibition was 46% and 41% with 2k at 45 μM and [ATP] at 200 μM or 1,000 μM, respectively.), 2k is non-competitive with respect to ATP under these assay conditions.
In summary, an indole scaffold has been designed, and an initial SAR carried out, for the development of non-ATP competitive pp60c-src inhibitors. The potency of the best indole-based inhibitor from the current series was found to be close to that of the best naphthalene-based inhibitor. The % inhibition was 46% and 41% with 2k at 45 μM and [ATP] at 200 μM or 1,000 μM, respectively.
The following results show the synthesis and testing of indole derived protein kinase inhibitors. Four reaction schemes are provided and separately followed by experimental details for the preparation of the final product of each of these reaction schemes. These final products are examples of indole-base tyrosine kinase inhibitors wherein the synthesis with certain groups on the aryl ring of the indole moiety is illustrated (boronic acid, Scheme 1; OH, Scheme 2; an aliphatic amide extension, Scheme 3; and a phosphonic acid Scheme 4).
Dissolved 3.50 g 5-hydroxy-2-indolecarboxylic acid in anh. MeOH pre-saturated with HCl gas. Refluxed for 48 hours. Concentrated in vacuo and triturated with AcCN×3 to remove residual acid. Filtered through silica plug with EtOAc to remove baseline contamination. Recovered 4.32 g (quant. yield) TLC Rf=0.78 (EtOAc) 1H NMR (DMSO-d6): 3.82 (s, 3H), 6.78 (d, J=8.8 Hz, 1H), 6.88 (s, 1H), 6.93 (s, 1H), 7.23 (d, J=8.8 Hz, 1H), 8.90 (s, 1H) 11.62 (s, 1H) FAB(+) MS m/e 191.9 (M+1)
Added 150 ml anh. DCM to 3.24 g (17 mmol) methyl 5-hydroxy-2-indolecarboxylate (1) and 6.67 g (18.7 mm) n-phenyl trifluoromethane sulfonamide at 0° C. Added 2.6 ml triethylamine dropwise at which point clear yellow solution formed. Stirred at 0° C. for 1 hour. Warmed to room temperature and stirred for 2 hours. Concentrated in vacuo and purified through silica gel column (1/1 EtOAc/hexanes). Recovered 4.69 g (86%). TLC Rf=0.63 (1/1 EtOAc/hexanes). HPLC Rf=20.879 1H NMR (DMSO-d6): 3.87 (s, 3H), 7.25 (s, 1H), 7.31 (d, J=9.2 Hz, 1H), 7.55 (d, J=9.2 Hz, 1H), 7.80 (s, 1H), 12.34 s, 1H) FAB(+) MS m/e 323.1 (M+1).
500 mg 1.55 mmol methyl 5-[(trifluoromethyl)sulfonyloxy]indole-2-carboxylate (2), 37.9 mg (0.05 mmol) PdCl2 (dppf), 432 mg (1.7 mmol) bispinacolatodiboron, 454.8 mg (4.65 mmol) potassium acetate, and 25.7 mg (0.05 mmol) dppf were added to a flask and vacuum dried at 40° C. for 2 hours. Added 20 ml anh dioxane and heated to 80° C. overnight. Reaction turned black as Pd black precipitated out. Filtered off catalyst and ran silica plug to remove baseline impurities. TLC Rf=0.51 (1/4 EtOAc/Hexane) Crude product was taken through to next reaction.
391.2 mg (1.3 mmol) methyl 5-methylindole-2-carboxylate, 4,4,5,5-tetramethyl-1,3,2-dioxaborolanemethyl (3) was dissolved in EtOAc. 0.25 ml (2.6 mmol) diethanolamine was added, and the reaction was stirred at room temperature overnight. The white precipitate which formed was filtered and sonicated in 1 N HCl. The resulting white precipitate was filtered, dissolved in MeOH, and concentrated in vacuo. Recovered 36.6 mg (13%). HPLC Rf=13.912, 1H NMR (DMSO-d6): 3.85 (s, 3H), 7.15 s, (1H), 7.36 (d, J=8.4 Hz, 1H), 7.67 (d, J=8.4 Hz, 1H), 7.87 (s, 1H), 8.14 (s, 1H), 11.91 (s, 1H).
Dissolved 2.00 g (11.3 mmol) 5-hydroxy-2-indolecarboxylic acid, 1.6 ml (12.4 mmol) 3 methoxybenzylamine, and 5.87 g (11.3 mmol) PyBOP in 10 ml anh. DMF. Cooled to 0° C. and added 5.9 ml (33.9 mmol) DIEA. Stirred for 5 min at 0° C. and allowed to warm to room temperature for 1 hour. Recovered 2.83 g (85% yield) TLC Rf=0.34 (1/1 EtOAc/hexanes) 1H NMR (DMSO-d6): 3.70 (s, 3H), 4.43 (d, J=4.4 Hz, 2H) 6.69 (d, J=8.8 Hz, 1H), 6.78 (d, J=7.7 Hz, 1H), 6.83 (s, 1H), 6.86 (s, 1H), 6.94 (s, 1H), 7.20 (m, 3H), 8.92 (t, J=4.4 Hz, 1H), 11.36 (s, 1H) FAB(+) MS m/e 297.3 (M+1)
Added 20 ml anh. DCM to 200 mg (0.67 mmol) (5-hydroxyindol-2-yl)-N-[(3-methoxyphenyl)methyl]carboxyamide(5) and cooled to −78° C. under argon. Added 4.0 ml (4.0 mmol, 6 eq) BBr3. Held at −78° C. for 5 min and warmed to rt. After 90 min at rt, quenched with H2O and stirred for 10 min. Diluted rxn mix with EtOAc and washed with NaHCO3 and brine. Dried organic layer over MgSO4 and concentrated in vacuo. Ran through silica plug to remove baseline contamination.
Recovered X mg. (80% yield.) TLC Rf=0.21 (1/1 EtOAc/hexanes). 1H NMR (DMSO-d6): 4.38 (d, J=4.8 Hz, 2H), 6.59 (d, J=8.8 Hz, 1H), 6.71 (m, 3H) 6.83 (d, J=1.8 Hz, 1H), 6.94 (s, 1H), 7.08 (dd, J=7.7 Hz, 1H), 7.19 (d, J=8.8 Hz, 1H), 8.84 (t, J=5.9 Hz), 11.28, (s, 1H). FAB(+) MS m/e 283.2 (M+1)
N-(1-carbamoyl-2-methylbutyl)(5-hydroxyindol-2-yl)carboxyamide (7)
100 mg (0.56 mmol) 5-hydroxy-2-indolecarboxylic acid, 103.4 mg (0.62 mmol, 1.1 eq) L-isoleucinamide, and 291 mg (0.56 mmol, 1 eq) PyBOP were all dissolved in 1 ml anh DMF. The solution was cooled to 0° C. and 0.3 ml (1.68 mmol, 3 eq) DIEA was added. The reaction mixture was stirred for 1 min at 0° C. and at room temperature for 1 hour. The reaction was then diluted with EtOAc and washed with 1 N HCl×3 and saturated NaHCO3×3. The organic layer was dried over MgSO4, and concentrated in vacuo to give 166.7 mg (91% yield.) TLC Rf=0.08 (1/1 EtOAc/hexanes). 1H NMR (DMSO-d6): 0.83 (m, 6H), 1.15 (m, 2H), 1.68 (m, 1H), 1.83 (m, 1H), 4.29 (t, J=8.8 Hz, 1H), 6.69 (d, J=8.5 Hz, 1H), 6.83 (s, 1H), 7.01 (s, 1H), 7.06 (s, 1H), 7.19 (d, J=8.4 Hz, 1H), 7.48, (s, 1H), 8.00 (d, 9.2 Hz, 1H), 8.76 (s, 1H), 11.3, (s, 1H). FAB(+) MS m/e 290.1 (M+1)
200 mg (0.62 mmol) methyl 5-[(trifluoromethyl)sulfonyloxy]indole-2-carboxylate (2), 195.8 mg (0.74 mmol,1.2 eq) dibenzylphosphite, 0.14 ml (0.81 mmol, 1.3 eq) DIEA, and 35.7 mg (0.03 mmol, 5 mol %) Pd(PPh3)4 were all dissolved in anh AcCN under argon. The reaction mix was heated to 80° C. overnight. The solvent was removed under reduced pressure, and the title compound was isolated by silica gel chromatography. 130 mg (50% yield). TLC Rf=0.28 (1/1 EtOAc/hexanes) 1H NMR (DMSO-d6): 3.85 (s, 3H), 4.98-5.01 (m, 4H), 7.28-7.32 (m, 11H), 7.53-7.55 (m, 2H), 8.17 (d, J=14.6 Hz, 1H) 31P NMR (DMSO-d6): 23.89.
Methyl 5-dibenzylphosphorylindole-2-carboxylate (8) (125 mg) was dissolved in 10 ml MeOH. 20 mg Pd-C was added and the mixture was hydrogenated in a Parr apparatus overnight. Filtered off catalyst and removed solvent under reduced pressure. Obtained 72.5 mg (73% yield). TLC Rf=baseline in EtOAc. 1H NMR (DMSO-d6): 3.84 (s, 3H), 7.24 (s, 1H), 7.44-7.49 (m, 2H), 8.01 (d, J=14.3 Hz, 1H) 12.11 (s, 1H) 31P NMR (DMSO-d6): 17.22.
The ester compounds in this example could be increased in potency by converting the ester to an amide and/or adding additional specificity elements.
The synthesis of some further elaborated indole inhibitors is illustrated below. These syntheses should result in compounds with greater potency against pp60c-src and other tyrosine kinases. The methyl ester group can be subsequently converted into various amide derivatives to increase potency.
There is considerable recent literature support for targeting pp60c-src (Src) as a broadly useful approach to cancer therapy without resulting in serious toxicity. For example, tumors that display enhanced EGF receptor PTK signaling, or overexpress the related Her-2/neu receptor, have constitutively activated Src and enhanced tumor invasiveness. Inhibition of Src in these cells induces growth arrest, triggers apoptosis, and reverses the transformed phenotype (Karni et al., 1999). It is known that abnormally elevated Src activity allows transformed cells to grow in an anchorage-independent fashion. This is apparently caused by the fact that extracellular matrix signaling elevates Src activity in the FAK/Src pathway, in a coordinated fashion with mitogenic signaling, and thereby blocks an apoptotic mechanism which would normally have been activated. Consequently FAK/Src inhibition in tumor cells may induce apoptosis because the apoptotic mechanism which would have normally become activated upon breaking free from the extracellular matrix would be induced (Hisano et al., 1997). Additionally, reduced VEGF mRNA expression was noted upon Src inhibition and tumors derived from these Src-inhibited cell lines showed reduced angiogenic development (Ellis et al., 1998).
The issue of potential toxicity of Src inhibition has been addressed with very promising results. For example, a knock-out of the Src gene in mice led to only one defect, namely osteoclasts that fail to form ruffled borders and consequently do not resorb bone. However, the osteoclast bone resorb function was rescued in these mice by inserting a kinase defective Src gene (Schwartzberg et al., 1997). This suggested that Src kinase activity can be inhibited in vivo without triggering the only known toxicity because the presence of the Src protein is apparently sufficient to recruit and activate other PTKs (which are essential for maintaining osteoclast function) in an osteoclast essential signaling complex.
Src has been proposed to be a “universal” target for cancer therapy since it has been found to be over activated in a growing number of human tumors, in addition those noted above (Levitzki, 1996). The potential benefits of Src inhibition for cancer therapy appear to be four-fold based upon the cited, and additional, literature. They are: 1) inhibition of uncontrolled cell growth caused by autocrine growth factor loop effects, etc.; 2) inhibition of metastasis due to triggering apoptosis upon breaking free from the cell matrix; 3) inhibition of tumor angiogenesis via reduced VEGF levels; and 4) low toxicity.
The initial non-peptide Src inhibitors have also shown very encouraging results in four different series of cell culture assays. 1) In the NIH 60-tumor cell panel assay, broad activity (as one would expect for a Src inhibitor) was seen against the tumor cell lines, including the prostate lines. For example, three of the inhibitors gave the following growth inhibition IC50 values against the NIH prostate cancer cell lines: TOM 2-32 (PC-3, 15 μM; DU-145, 38 μM), TOM 2-47 (PC-3, 19 μM), KLM 2-31 (PC-3, 39 μM; DU-145, >100 μM). 2) In the v-Src transformed normal rat kidney cell line (LA25) TOM 2-47 & TOM 2-32 specifically blocked the v-Src induced cell growth without inhibiting the normal growth of the parent non-transformed cells. This result showed that the inhibitors do not affect normal cells but are effective in blocking Src induced cell transformation. 3) The Src inhibitors to the cancer drugs etoposide, taxol, doxorubicin and cisplatin in ovarian tumors from three different patients and an abdominal carcinoma from another patient. In all cases, the Src inhibitors were at least as effective, and typically more effective, than the known cancer drugs, with full efficacy seen at the lowest dose tested (3 μM). As a representative example, a comparison of taxol and doxorubicin (they were more effective than etoposide & cisplatin in this particular tumor cell culture) with the three Src inhibitors mentioned above utilizing ovarian tumor cells from tumor N015 is shown in
Overall, the cell data obtained thus far shows what one might expect for Src inhibitors, i.e. broad activity against many cancer cell lines with little or no normal cell toxicity.
The preliminary Src inhibitors are lead structures from which it is possible to design more potent and selective inhibitors. In addition to utilizing the tyrosine kinase crystal structures, molecular modeling studies can be carried out with the natural product tyrosine kinase inhibitor damnacanthal (Faltynek et al., 1995) to investigate its peptide-competitive binding mode. These additional modeling studies are enable one to design further analogs of Src inhibitors wherein the key pharmacophore elements of damnacanthal are incorporated into the new inhibitors. Their syntheses will be undertaken and the isolated Src testing done as reported (Marsilje 2000).
Although preferred embodiments have been depicted and described in detail herein, it will be apparent to those skilled in the relevant art that various modifications, additions, substitutions, and the like can be made without departing from the spirit of the invention and these are therefore considered to be within the scope of the invention as defined in the claims which follow.
The following references which were cited herein, are hereby incorporated by reference into this application:
Ajay, Murcko, M. A. (1995) Computational Methods to Predict Binding Free Energy in Ligand-Receptor Complexes. J. Med. Chem., 38, 4953-4967.
Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K. & Watson, J. D. (1994) Molecular Biology Of The Cell, 3rd ed., Garland Publishing, Inc., New York, pp 97, 508 & 667.
Alfaro-Lopez, J., Yuan, W., Phan, B. C., Kamath, J., Lou, Q., Lam, K. S., Hruby, V. J. (1998) Discovery of a Novel Series of Potent and Selective Substrate-Based Inhibitors of p60c-src Protein Tyrosine Kinase: Conformational and Topographical Constraints in Peptide Design. J. Med. Chem., 41, 2252-2260.
Backes, B. J., Virgilo, A. A., Ellman, J. A. (1996) Activation Method to Prepare a Highly Reactive Acylsulfonamide “Safety-Catch” Linker for Solid-Phase Synthesis. J. Am. Chem. Soc., 118, 3055-3056.
Baggio, R., Elbaum, D., Kanyo, Z. F., Carroll, P. J., Cavalli, C., Ash, D. E., Christianson, D. W. (1997) Inhibition of Mn2+-Arginase by Borate Leads to the Design of a Transition State Analog Inhibitor, 2(S)-Amino-6-boronohexanoic Acid. J. Am. Chem. Soc., 119, 8107-8108.
Barnekow, A.; Paul, E.; Schartl, M. (1987) Expression of the c-src protooncogene in human skin tumors. Cancer Res., 47, 235-240.
Benson, W. H., Birge, W. J., Dorough, H. W. (1984) Environ. Toxicol. Chem., 3, 209. Chem. Abstr. 101:124626g.
Bhagwat, S. S., Gude, C. (1994) N-Alkylation of indole ring using Mitsunobu reaction. Tet. Lett., 35, 1847-1850.
Bjorge, J. D., O'Connor, T. J., Fujita, D. J. (1996) Activation of human pp60c-src. B iochemistry & Cell Biology, 74, 477-484.
Bjelfman, C.; Hedborg, F.; Johansson, I.; Nordenskjold, M.; Pahlman, S. (1990) Expression of the neuronal for of pp60c-src in neuroblastoma in relation to clinical stage and prognosis. Cancer Res, 50, 6908-6914.
Bohacek, R. S., McMartin, C., Guida, W. C. (1996) The Art and Practice of Structure-Based Drug Design: A Molecular Modeling Perspective. Medicinal Research Reviews, 16, 3-50 (see p. 43).
Boyd, M. R., Paull, K. D. (1995) Some practical considerations and applications fo the National Cancer Institute in vitro anticancer drug discovery screen. Drug Development Research, 34, 91-109.
Brooks, S. P. J. & Storey, K. B. (1992) Bound and Determined: A Computer Program for Making Buffers of Defined Ion Concentrations. Analytical Biochemistry, 201, 119-126.
Brown, D. (1997) Future Pathways for Combinatorial Chemistry. Molecular Diversity, 2(4), 217-222.
Budde, R. J. A., McMurray, J. S., Saya, H., Gallick, G. E. & Levin, V. A. (1995) Discovery, Development, and Testing of Substrates and Inhibitors of pp60c-src. International Journal of Pharmacognosy, 33, 27-34.
Budde, R. J. A., Ke, S., Levin, V. A. (1994) Activity of pp60c-src in 60 different cell lines derived from human tumors. Cancer Biochem. Biophys., 14, 171-175.
Burger, A. M., Kaur, G., Alley, M. C., Supko, J. G., Malspeis, L., Grever, M. R. & Sausville, E. A. (1995) Tyrphostin AG17, [(3,5-Di-tert-butyl-4-hydroxybenzylidene)-malonitrile], inhibits cell growth by disrupting mitochondria. Cancer Research, 55, 2794-2799.
Burke, T. R.; Lim, B.; Marquez, V. E.; Li, Z-H.; Bolen, J. B.; Stefanova, I.; Horak, I. D. (1993) J. Med. Chem. 36, 425.
Choi, S. (1999), Ph.D. Thesis SUNY at Buffalo, Buffalo, N.Y.
Cooper, C. M. (1990) Oncogenes. Jones and Bartlett Publishers, Boston, Mass.
Coughlin, J. R. (1996) Inorganic borates-chemistry, human exposure, and health and regulatory guidelines. J. Trace Elements in Experimental Medicine, 9, 137-151.
Courtneidge, S. A. (1994) Protein tyrosine kinases, with emphasis on the Src family. Seminars in Cancer Biology, 5, 239-246.
Cox, S., Radzio-Andzelm, E. & Taylor, S. S. (1994) Domain movements in protein kinases. Current Opinion in Structural Biology, 4(6), 893-901.
Culver, B. D., Hubbard, S. A. (1996) Inorganic boron health effects in humans—and aid to risk assessment and clinical judgment. J. Trace Elements in Experimental Medicine, 9, 175-184.
Davis, P. D.; Hill, C. H.; Keech, E.; Lawton, G.; Nixon, J. S.; Sedgwick, A. D.; Wadsworth, J.; Westmacott, D.; Wilkinson, S. E. (1989) FEBS Lett. 259(1), 61.
Davis, P. D.; Elliott, L. H.; Harris, W.; Hill, C. H.; Hurst, S. A.; Keech, E.; Kumar, M. K. H.; Lawton, G.; Nixon, J. S.; Wilkinson, S. E. (1992) J. Med. Chem. 35, 994.
Ellis, L. M., Staley, C. A., Liu, W., Fleming, R. Y., Parikh, N. U., Bucana, C. D., & Gallick, G. E. (1998) Down-regulation of vascular endothelial growth factor in a human colon carcinoma cell line transfected with an antisense expression vector specific for c-src. Journal of Biological Chemistry 273 (2): 1052-1057.
Ezquerra, J., Pedregal, C., Lamas, C., Barluenga, J., Perez, M., Garcia-Martin, M. A., Gonzalez, J. M. (1996) Efficient reagents for the synthesis of 5-, 7-, and 5,7-substituted indoles starting from aromatic amines: scope and limitations. J. Org. Chem., 61, 5804-5812.
Faltynek, C., et al. (1995) Damnacanthal is a highly potent, selective inhibitor of p56lck tyrosine kinase activity. Biochemistry 34, 12404-12410.
Faltynek, C. R.; Wang, S.; Miller, D.; Mauvais, P.; Gauvin, B.; Reid, J.; Xie, W.; Hoekstra, S.; Juniewicz, P.; Sarup, J.; Lehr, R.; Sawutz, D. G.; Murphy, D. J. (1995) Enzyme Inhibition 9, 111.
Fanning, P.; Bulovas, K.; Saini, K. S.; Libertino, J. A.; Joyce, A. D.; Summerhayes, I. C. (1992) Elevated expression of pp60c-src in low grade human bladder carcinoma. Cancer Research, 52, 1457-1462.
Fredenhagen, A.; Mett, H.; Meyer, T.; Buchdunger, E.; Regenass, U.; Roggo, B. E.; Petersen, F. J. (1995) Antibiotics 48, 1355.
Froyen, P. (1997) Tetrahedron Lett. 38(30), 5359.
Fry, D. W., Kraker, A. J., McMichael, A., Ambroso, L. A., Nelson, J. M. Leopold, W. R., Connors, R. W. & Bridges, A. J. (1994) A Specific Inhibitor of the Epidermal Growth Factor Receptor Tyrosine Kinase. Science, 265, 1093-1095.
Glass, D. B., Cheng, H.-C., Mende-Mueller, L., Reed. J. & Walsh, D. A. (1989) Primary structure determinants essential for potent inhibition of cAMP-dependent protein kinase by inhibitory peptides corresponding to the active portion of the heat-stable inhibitor protein. J. Biol. Chem., 264, 8802-8810.
Groundwater, P. W., Solomons, K. R. H., Drewe, J. A. & Munawar, M. A. (1996) Protein Tyrosine Kinase Inhibitors. Progress in Medicinal Chemistry, 33, 233-329.
Hanks, S. K. & Hunter, T. (1995) Protein kinases. 6. The eukaryotic protein kinase superfamily: Kinase (catalytic) domain structure and classification. FASEB J., 9, 576-596.
Hanke, J. H., Gardner, J. P., Dow, R. L., Changelian, P. S., Brissette, W. H., Weringer, E. J., Pollok, B. A. & Connelly, P. A. (1996) Discovery of a novel, potent, and Src family-selective tyrosine kinase inhibitor. J. Biol. Chem., 271, 695-701.
Hisano, C., Nakano, S., Fujishima, H., Masumoto, N., Tatsumoto, T., & Niho. Y. (1997) src oncogene inhibits detachment-induced apoptosis through constitutive activation of p125FAK in HAG-1 human epithelial cells. Proc. Annu. Meet. Am. Assoc. Cancer Res. 38:A1925.
Hsiao, G. K., Hangauer, D. G. (1998) A Facile Synthesis of tert-Butyl 2-[(Benzyloxycarbonyl)amino]-3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)propionate: An Orthogonally Protected Boronic Acid Analog of Aspartic Acid. Synthesis, 1043-1046.
Hsu, C-Y., J., Jacoski, M. V., Maguire, M. P., Spada, A. P. & Zilberstein, A. (1992) Inhibition Kinetics and Selectivity of the Tyrosine Kinase Inhibitor Erbstatin and a Pyridone-based Analog. Biochemical Pharmacology, 43, 241-2477.
Huang, C-K., Wu, F-Y., Ai, Y-X. (1995) Polyhydroxylated 3-(N-phenyl) carbamoyl-2-iminochromene derivatives as potent inhibitors of tyrosine kinase p60c-src. Bioorg. & Med. Chem. Lett., 5, 2423-2428.
Hubbard, S. R., Wei, L, Ellis, L, & Hendrickson, W. A. (1994) Crystal structure of the tyrosine kinase domain of the human insulin receptor, Nature, 372, 746-754.
Hubbard, S. R. (1997) Crystal structure of the activated insulin receptor tyrosine kinase in complex with peptide substrate and ATP analog. The EMBO Journal, 16, 5572-5581.
Hughes, R. L., Smith, I. C., Lawless, E. W. (1967) Production of the Boranes and Related Rearch, Holtzman R. T., Ed., Academic Press, New York, pp 291-294.
Hunter, T. (1987) A thousand and one protein kinases. Cell, 50, 823-829.
Hunter, T. (1994) 1001 protein kinases redux-towards 2000. Seminars in Cell Biology, 5, 367-376.
Hunter, T. (1998) The Croonian Lecture 1997. The phosphorylation of proteins on tyrosine: its role in cell growth and disease. Philosophical Transactions of the Royal Society of London—Series B: Biological Sciences 353 (1368):583-605.
Hutchins, C., Greer, J. (1991) Comparative modeling of proteins in the design of novel renin inhibitors. Critical Reviews in Biochemistry & Molecular Biology, 26, 77-127.
Ishiyama, T., Murata, M., Miyaura, N. (1995) Palladium(0)-catalyzed cross-coupling reaction of alkoxydboron with haloarenes: A direct procedure for arylboronic esters. J. Org. Chem., 60, 7508-7510.
Ishiyama, T., Itoh, Y., Kitano, T., Miyaura, N. (1997) Synthesis of arylboronates via the palladium(0)-catalyzed cross-coupling reaction of tetra(alkoxo)diborons with aryl triflates. Tet. Lett., 38, 3447-3450.
Karni, R., Jove R., & Levitzki A. (1999) Inhibition of pp60c-src reduces Bcl-X expression and reverses the transformed phenotype of cells overexpressing EGF and HER-2 receptors. Oncogene 18(33): 4654-4662.
Kelloff, G. J., Fay, J. R., Steele, V. E., Lubet, R. A., Boone, C. W., Crowell, J. A. (1996) Epidermal growth factor receptor tyrosine kinase inhibitors as potential cancer chemopreventatives. Cancer Epidemiology, Biomarkers & Prevention, 5, 657-666.
Kettner, C. A., Shenvi, A. B. (1984) Inhibition of the Serine Proteases Leukocyte Elastase, Pancreatic Elastase, Cathepsin G, and Chymotrypsin by Peptide Boronic Acids. J. Biol. Chem., 259, 15106-15114.
Kim. M. H., Lai, J. H. & Hangauer, D. G. (1994) Tetrapeptide tyrosine kinase inhibitors: Enantioselective synthesis of p-hydroxymethyl-L-phenylalanine, incorporation into a tetrapeptide, and subsequent elaboration into p-(R,S-hydroxyphosphonomethyl)-L-phenylalanine. Int. J. Peptide Protein Res., 44, 457-465.
Kinder, D. H., Frank, S. K., Ames, M. M. (1990) Analogues of Carbamyl Aspartate as Inhibitors of Dihydrooratase: Preparation of Boronic Acid Transition-State Analogues and a Zinc Chelator Carbamylhomocysteine. J. Med. Chem., 33, 819-823.
Klein, G. (1990) Multistep emancipation of tumors from growth control: can it be curbed in a single step? BioEssays, 12, 347-350.
Knighton, D. R., Cadena, D. L., Zheng, J., Ten Eyck, L. F., Taylor, S. S. & Sowadski, J. M. (1993) Structural features that specify tyrosine activity deduced from homology modeling of the epidermal growth factor receptor. Proc. Natl. Acad. Sci. U.S.A., 90(11), 5001-5.
Kolibaba, K. S. & Druker, B. J. (1997) Protein tyrosine kinases and cancer. Biochimica et Biophysica Acta, 1333: F217-F248.
Lai, J. H., Marsilje, T. M., Choi, S., Nair, S. A., Hangauer, D. G. (1998) The design, synthesis and activity of pentapeptide pp60c-src inhibitors containing L-phosphotyrosine mimics. J. Peptide Res., 51, 271-281.
Lai, J. H., Pham, H. & Hangauer, D. G. (1996) Synthesis of a Vicinal Tricarbonyl Amide Derivative of L-Phenylalanine. J. Org. Chem., 61, 1872-1874.
Lam, K. S. (1997) Application of Combinatorial Library Methods in Cancer Research an Drug Discovery. Anti-Cancer Drug Design, 12(3), 145-167.
Lawrence, D. S. & Niu, J. (1998) Protein Kinase Inhibitors: The Tyrosine-Specific Protein Kinases. Pharmacol. Ther., 77(2), 81-114.
Levitzki, A. (1996a) Targeting signal transduction for disease therapy. Current Opinion in Cell Biology, 8, 239-244.
Levitzki, A. (1996b) SRC as a target for anti-cancer drugs. Anti-Cancer Drug Design, 11, 175-182.
Levitzki, A.; Gazit, A. (1995) Tyrosine Kinase Inhibition: An Approach to Drug Development. Science, 267, 1782-1788.
Li, H., Liu, T. F., Lazrak, A., Peracchia, C., Goldberg, G. S., Lampe, P. D., Johnson, R. G. (1996) Properties and regulation of gap junctional hemichannels in the plasma membranes of cultured cells. J. Cell. Biol., 134, 1019-1030.
Loomis, W. D. & Durst, R. W. (1992) Chemistry and biology of boron. BioFactors, 3, 229-239.
Lou, Q., Leftwich, M. E., McKay, T., Salmon, S. E., Rychetsky, L. & Lam, K. S. (1997) Potent Pseudosubstrate-based Peptide Inhibitors for p60c-src Protein Tyrosine Kinase. Cancer Research, 57(10), 1877-1881.
Lou, Q., Leftwich, M. E. & Lam, K. S. (1996) Identification of GIYWHHY as a Novel Peptide Substrate for Human p60c-src Protein Tyrosine Kinase. Biorganic & Medicinal Chemistry, 4, 677-682.
Luttrell, D. K.; Lee, A.; Lansing, T. J.; Crosby, R. M.; Jung, K. D.; Willard, D.; Luther, M.; Rodriguez, M.; Berman, J.; Gilmer, T. M. (1994) Involvement of pp60c-src with two major signaling pathways in human breast cancer. Proc. Natl. Acad. Sci. USA, 91, 83-87.
Lynch, S. A.; Brugge, J. S.; Fromowitz, F.; Glantz, L.; Wang, P.; Caruso, R.; Viola, M. V. (1993) Increased expression of the src proto-oncogene in hairy cell leukemia and a subgroup of B-cell lymphomas. Leukemia, 7, 1416-1422.
Madhusudan, Trafny, E. A., Xuong, N-H, Adams, J. A., Ten Eyck, L. F., Taylor, S. S. & Sowadski, J. M. (1994) cAMP-dependent protein kinase: Crystallographic insights into substrate recognition and phosphotransfer. Protein Science, 3, 176-187.
Mao, W. G., Irby, R., Coppola, D., Fu, L., Turner, J. (1997) Activation of c-src by receptor tyrosine kinases in human colon cancer cells with high metastatic potential. Oncogene, 15, 3083-3090.
Marsilje, T. H., Milkiewicz, K. L., & Hangauer, D. L. (2000) The design, synthesis and activity of non-ATP competitive inhibitors of pp60c-src tyrosine kinase 1. Hydroxynaphthalene Derivatives. Bioorganic and Medicinal Chemistry Letters, in press.
Marx, J. (1990) Oncogenes evoke new cancer therapies. Science, 249, 1376-1378.
National Cancer Institute (1989) Survey of Compounds which have been tested for carcinogenic activity. NIH Publication No. 49-468, p. 16.
Matteson, D. S., Kandil, S. A., Soundararajan, R. (1990) Synthesis of Asymmetrically Deuterated Glycerol and Dibenzylglyceraldehyde via Boronic Esters. J. Am. Chem. Soc., 112, 3964-3969.
Matteson, D. S. (1988) Acc. Chem. Res., 21, 294-300.
Matteson, D. S., Kandil, A. A. (1987) Conversion of α-halo boronic esters to inverted α-(methylsulfonyl)oxy boronic esters. J. Org. Chem., 52, 5121-5124.
Matteson, D. S., Soloway, A. H., Tomlinson, D. W., Campbell, J. D., Nixon, G. A. (1964) J. Med. Chem., 7, 640.
Mazurenko, N. N.; Kogen, E. A.; Zborovskaya, I. B.; Kisseljov, F. L. (1992) Expression of pp60c-src in human small cell and non-small cell long carcinomas. European J. of Cancer, 28, 372-377.
Milkiewicz, K.; Marsilje, T.; Woodward Jr, R.; Bifulco Jr, N.; Hangauer, M.; Hangauer, D.G. (2000) The design, synthesis and activity of non-ATP competitive inhibitors of pp60c-src tyrosine kinase 2. Hydroxyindole Derivatives. Bioorganic and Medicinal Chemistry Letters, in press.
Mohammadi, M., Schlessinger, J., Hubbard, S. R. (1996) Structure of the FGF Receptor Tyrosine Kinase Domain Reveals a Novel Autoinhibitory Mechanism. Cell, 86, 577-587.
Mohammadi, M., McMahon, G., Li, S., Tang, C., Hirth, P., Yeh, B. K., Hubbard, S. R., Schlessinger, J. (1997) Structures of the Tyrosine Kinase Domain of Fibroblast Growth Factor Receptor in Complex with Inhibitors. Science, 276, 955-960.
Morin, C. (1994) The Chemistry of Boron Analogues of Biomolecules. Tetrahedron, 50, 12521-12569.
Murakami, Y., Otsuka, K. Wada, Y., Morikawa, A. (1990) The partial oxidation of ethane over a B2O3-Al2O3 catalyst. Bull. Chem. Soc. Jpn., 63, 340-346.
Nair, S. A., Kim, M. K., Warren, S. D., Choi, S., Songyang, Z., Cantley, L. C. & Hangauer, D. G. (1995). Identification of Efficient Pentapeptide Substrates for the Tyrosine Kinase pp60c-src. J. Med. Chem., 38, 4276-4283.
Nair, S. A., Lee, B. & Hangauer, D. G. (1995b). Synthesis of Orthogonally Protected L-Homocysteine and L-2-Amino-4-phosphonobutanoic Acid From L-Homoserine. Synthesis, 7, 810-814.
Nielsen, F. H. (1997) Boron in human and animal nutrition. Plant & Soil, 193, 199-208.
Otsuka, K., Uragami, Y., Hatano, M. (1992) Tile partial oxidation of ethane to acetaldehyde. Catalysis Today, 13, 667-672.
Parsons, J. T. & Parsons, S. J. (1997) Src family protein tyrosine kinases: cooperating with growth factor and adhesion signaling pathways. Current Opinion in Cell Biology, 9, 187-192.
Patrick, D. R. & Heimbrook, D. C. (1996) Protein Kinase Inhibitors For The Treatment of Cancer. Drug Discovery Today, 1, 325-330.
Pavia, M. R., Cohen, M. P., Dilley, G. J., Dubuc, G. R., Durgin, T. L., Forman, F. W., Hediger, M. E., Milot, G., Powers, T. S., Sucholeiki, I., Zhou, S. & Hangauer, D. G. (1996) The Design and Synthesis of Substituted Biphenyl Libraries. Biorganic & Medicinal Chemistry, 4, 659-666. Posner, I., Engel, M., Gazit, A. & Levitzki, A. (1994) Kinetics of Inhibition by Tyrphostins of the Tyrosine Kinase Activity of the Epidernzal Growth Factor Receptor and Analysis by a New Computer Program. Molecular Pharmacology, 45, 673-683.
Powis, G. (1991) Signal targets for anticancer drug development. TIPS, 188-194.
Ramdas, L., Obeyesekere, N. U., McMurray, J. S., Gallick, G. E., Seifert, W. E. Jr. & Budde, R. J. (1995) A tyrphostin-derived inhibitor of protein tyrosine kinases: isolation and characterization. Archives of Biochemistry & Biophysics, 323, 237-242.
Ramdas, L., McMurray, J. S. & Budde, R. J. (1994) The degree of inhibition of protein tyrosine kinase activity by tyrphostin 23 and 25 is related to their instability. Cancer Research, 54, 867-869.
Rewcastle, G. W., Palmer, B. D., Thompson, A. M., Bridges, A. J., Cody, D. R., Zhou, H. Fry, D. W., McMichael, A. & Denny, W. A. (1996) Tyrosine Kinase Inhibitors. 10. Isomeric 4-[(3-Bromophenyl)amino]pyrido[d]-pyrimidines Are Potent ATP Binding Site Inhibitors of the Tyrosine Kinase Function of the Epidermal Growth Factor Receptor. J. Med. Chem., 39, 1823-1835.
Rudd, C. E.; Janssen, O.; Prasad, K. V. S.; Raab, M.; da Silva, A.; Telfer, J. C.; Yamamoto, M. (1993) src-related protein tyrosine kinases and their surface receptors. Biochimica et Biophysica Acta, 1155, 239-266.
Saperstein, R., Vicario, P. P., Strout, H. V., Brady, E., Slater, E. E., Greenlee, W. J., Ondeyka, D. L., Patchett, A. A. & Hangauer, D. G. (1989) Design of a selective insulin receptor tyrosine kinase inhibitor and its effect on glucose uptake and metabolism in intact cells. Biochemistry, 28, 5694-5701.
Sawutz, D. G.; Bode, D. C.; Briggs, G. M.; Reid, J. R.; Canniff, P.; Caldwell, L.; Faltynek, C. R.; Miller, D.; Dunn, J. A.; Garavilla, L.; Guiles, J. W.; Weigelt, C.; Michne, W.; Treasurywala, A. M.; Silver, P. J. (1996) Biochem. Pharmacol. 51, 1631.
Schwartzberg, P. L., et al. (1997) Rescue of osteoclast function by transgenic expression of kinase-deficient Src in src −/− mutant mice. Genes & Development 11: 2835-2844.
Shiraishi, T., Owada, M. K., Tatsuka, M., Yamashita, T., Watanabe, K., Kakunaga, T. (1989) Specific Inhibitors of Tyrosine-specific Protein Kinases: Properties of 4-hydroxycinnamamide derivatives in vitro. Cancer Research, 49, 2374-2378.
Showalter, H. H. & Kraker, A. J. (1997) Small molecule inhibitors of the platelet-derived growth factor receptor, the fibroblast growth factor receptor, and src family tyrosine kinases. Pharmacology & Therapeutics, 76, 55-71.
Sicheri, F., Moarefi, I. & Kuriyan, J. (1997) Crystal structure of the Src family tyrosine kinase Hck. Nature, 385, 602-609.
Skordalakes, E., Tyrell, R., Elgendy, S., Goodwin, C. A., Green, D., Dodson, G., Scully, M. F., Freyssinet, J-M. H., Kakkar, V. V., Deadman, J. J. (1997) Crystallographic Structures of Human α-Thrombin Complexed to Peptide Boronic Acids Lacking a Positive Charge at P1. Evidence of Novel Interactions. J. Am. Chem. Soc., 119, 9935-9936.
Snyder, H. R., Kuck, J. A., Johnson, J. R. (1938) J. Am. Chem. Soc., 60, 105.
Soloway, A. H., Whitman, B., Messer, J. R. (1962) J. Med. Pharm. Chem., 7, 640.
Soloway, A. H., Whitman, B., Messer, J. R. (1960) J. Pharmacology and Experimental Therapeutics, 129, 310-314.
Soloway, A. H. (1958) Science, 128, 1572.
Songyang, Z, Blechner, S., Hoagland, N., Hoekstra, M. F., Piwnica-Worms, H. & Cantley, L. C. (1994) Use of an oriented peptide library to determine the optimal substrates of protein kinases. Current Biology, 4, 973-982.
Songyang, Z., Carraway III, K. L., Eck, M. J., Harrison, S. C., Feldman, R. A., Mohammadl, M., Schlessinger, J., Hubbard, S. R., Smith, D. P., Eng. C., Lorenzo, J. J., Ponder, B. A. J., Mayer, B. J. & Cantley, L. C. (1995) Protein tyrosine kinases and SH2 domains have overlapping specificities. Nature, 373, 536-539.
Staley, C. A.; Parikh, N. U.; Gallick, G. E. (1997) Cell Growth & Differentiation 8(3), 269.
Stanwell, C., Burke, T. R. & Yuspa, S. H. (1995) Erbstatin Analogue Methyl 2,5-dihydrocinnamate Cross-links Proteins and is Cytotoxic to Normal and Neoplastic Epithelial Cells by a Mechanism Independent of Tyrosine Kinase Inhibition. Cancer Research, 55, 4950-4956.
Stanwell, C., Ye, B. & Burke, T. R. (1996) Cell Protein Cross-linking by Erbstatin and Related Compounds. Biochemical Pharmacology, 52, 475-480.
Takeshima, E.; Hamaguchi, M.; Watanbe, T.; Akiyama, S.; Kataoka, M.; Ohnishi, Y.; Xiao, H.; Hagai, Y., Taka, H. (1991) Aberrant elevation of tyrosine-specific phosphorylation in human gastric cancer cells. Japan J. Cancer Res., 82, 1428-1435.
Talamonti, M. S.; Roh, M. S.; Curley, S. A.; Gallick, G. E. (1993) Increase in activity and level of pp60c-src in progressive stages of human colorectal cancer. J. of Clinical Investigation, 91, 53-60.
Taniyama, K., Fujiwara, H., Kuno, T., Saito, N., Shuntoh, H., Sakaue, M. (1989) Acute and subacute toxicity of 10B-paraboronophenylalanine. Pigment Cell Research, 2, 291-296.
Taylor, S. J., Shalloway, D. (1996) Src and the control of cell division. Bioessays, 18, 9-11.
Taylor, S. S., Knighton, D. R., Zheng, J., Sowadski, J. M., Gibbs, C. S. & Zoller, M. J. (1993) A template for the protein kinase family. Trends in Biochemical Sciences, 18(3), 84-9.
Taylor, S. S., Radzio-Andzelm, E. (1994) Three protein kinase structures define a common motif. Stucture, 2, 345-355.
Thakkar, K., Geahlen, R. L., Cushman, M. (1993) Synthesis and protein-tyrosine kinase inhibitory activity of polyhydroxylated stilbene analogs of piceatannol. J. Med. Chem., 36, 2950-2955.
Weinberg, R. A. (1989) Oncogenes, antioncogenes, and the molecular basis of multistep carcinogenesis. Cancer Research, 49, 3713-3721.
Wolfe, J. P., Ahman, J., Sadighi, J. P., Singer, R. A., Buchwald, S. L. (1997) An ammonia equivalent for the palladium-catalyzed amination of aryl halides and triflates. Tet. Lett., 38, 6367-6370.
Wong, T. W. & Goldberg, A. R. (1984) Kinetics and mechanism of angiotensin phosphorylation by the transforming gene product of Rous Sarcoma Virus. J. Biol. Chem., 259, 3127-3131.
Xu, W., Harrison, S. C. & Eck, M. J. (1997) Three-dimensional structure of the tyrosine kinase c-Src. Nature, 385, 595-602.
Yamaguchi, H. & Hendrickson, W. A. (1996) Structural basis for activation of human lymphocyte kinase Lck upon tyrosine phosphorylation. Nature, 384, 484-489.
Yamamoto, T. (1993) Molecular Basis of Cancer: Oncogenes and Tumor Suppresor Genes. Microbiol. Immunol. 37, 11-22.
Zheng, J., Knighton, D. R., Ten Eyck, L. R., Karlsson, R., Xuong, N-H., Taylor, S. S. & Sowadski, J. M. (1993) Crystal structure of the catalytic subunit of cAMP-dependent protein kinase complexed with MgATP and peptide inhibitor. Biochemistry, 32, 2154-61.
Zhanpeisov, N. U., Otsuka, K. (1992) Cluster quantum chemical study of the mechanism of selective oxidation of ethane to acetaldellyde on boron-phosphorous mixed oxide catalysts. React. Kinet. Catal. Lett., 48, 301-308.
This patent application is a continuation of U.S. Ser. No. 09/482,585 filed Jan. 13, 2000, which claims the benefit of U.S. Provisional Patent Application Ser. No. 60/115,643, filed Jan. 13, 1999. The entire contents of each of the above identified applications is hereby incorporated by reference.
| Number | Date | Country | |
|---|---|---|---|
| 60115643 | Jan 1999 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 09482585 | Jan 2000 | US |
| Child | 11261858 | Oct 2005 | US |
