Molecular Cancer Therapeutics Molecular Diagnostics in Cancer Therapeutic Development: Fulfilling the Promise of Personalized Medicine Targeting the PI3-Kinase Pathway in Cancer
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McLean, S. R.
Right arrow Articles by Boshoff, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McLean, S. R.
Right arrow Articles by Boshoff, C.
Mol Cancer Ther. 2005;4:2008-2015
© 2005 American Association for Cancer Research

Imatinib binding and cKIT inhibition is abrogated by the cKIT kinase domain I missense mutation Val654Ala

Sean R. McLean1, Mali Gana-Weisz1, Basil Hartzoulakis5, Richard Frow1,3, Jeremy Whelan4, David Selwood2 and Chris Boshoff1

1 Cancer Research UK Viral Oncology Group and 2 Biological and Medicinal Chemistry Group, Wolfson Institute for Biomedical Research, University College London; 3 Department of Histopathology, Royal Free and University College London Hospitals; 4 University College London Hospitals, London, United Kingdom and 5 Discovery Ltd., Cambridge, United Kingdom

Requests for reprints: David Selwood, Biological and Medicinal Chemistry Group, Wolfson Institute for Biomedical Research, University College London, London WC1E 6BT, United Kingdom. E-mail: d.selwood{at}ucl.ac.uk


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Several activating mutations in the cKIT receptor tyrosine kinase are associated with the development and progression of gastrointestinal stromal tumors (GIST). Treatment of GIST with the tyrosine kinase inhibitor imatinib (Gleevec, STI571; Novartis, Basel, Switzerland) increases patient survival. However, many patients develop resistance to imatinib following initial responses. We sequenced cKIT exons from two patients with GIST after the development of imatinib resistance, revealing a point mutation in kinase domain I (exon 13), Val654Ala, which has been associated previously with relapse and resistance. Molecular modeling of cKIT-imatinib complexes shows that this residue is located in the drug-binding site and that the Val654Ala mutation disrupts drug binding by removing hydrophobic contacts with the central diaminophenyl ring of imatinib. Loss of these contacts results in a destabilizing effect on two key hydrogen bonds between imatinib and Asp310 and Thr670 of cKIT. Calculations based on published crystallography data show an estimated destabilization energy of 2.25 kcal/mol in the Val654Ala cKIT compared with wild type. When present on the same cKIT allele as an oncogenic mutation, the Val654Ala mutation abolishes imatinib-mediated inhibition of cKIT phosphoactivation in vitro. These results highlight some of the structural and functional consequences of the Val654Ala mutation in relapsing imatinib-resistant GIST and emphasize the importance of tumor genetics in drug development and patient-specific cancer treatment regimens. [Mol Cancer Ther 2005;4(12):2005–15]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Gastrointestinal stromal tumor (GIST) is the most common mesenchymal tumor of the gastrointestinal tract and is thought to arise from the interstitial cells of Cajal, the pacemaker cells of the gut (1). Oncogenic mutations in the transmembrane receptor tyrosine kinase (RTK) cKIT are implicated as an important event in the pathogenesis of up to 90% of all GIST (2, 3) and are also associated with the development of mast cell leukemia (4) and testicular seminoma (5).

cKIT is a type III RTK containing five extracellular NH2-terminal immunoglobulin-like domains and an intracellular split kinase motif that includes ATP-binding and phosphotransferase motifs in kinase domains I and II, respectively (6). The first two NH2-terminal immunoglobulin-like domains confer specificity to the cKIT ligand steel factor/stem cell factor, and the fourth is associated with receptor dimerization (7). Following stem cell factor binding and receptor dimerization, tyrosine residues in the juxtamembrane region (JMR) of one cKIT monomer become the substrate for transphosphorylation by the kinase domain of the other monomer (8). Tyrosine phosphorylation, primarily at Tyr568, moves the autoinhibitory JMR out of the ATP-binding/phosphotransferase kinase motif and activates the cKIT kinase domain. In the autoinhibited conformation of cKIT, Trp557 of the JMR prevents Asp810 (D), Phe811 (F), and Gly812 (G) from forming the DFG motif that is necessary for formation of the kinase-active conformation (9). Kinase-active cKIT is responsible for relaying the extracellular signal (stem cell factor binding) to the intracellular compartment (10). Downstream cKIT signaling involves the phosphatidylinositol 3-kinase (PI3K), mitogen-activated protein kinase (MAPK), and Janus-activated kinase/signal transducers and activators of transcription (JAK/STAT) pathways (11), which are associated with increased cell proliferation and resistance to apoptosis (12). Constitutive activation of these pathways via oncogenic mutation is thought to play a critical role in interstitial cells of Cajal transformation and development of GIST (1).

Numerous reports have described unique mutations in cKIT, which cause constitutive activation of the cKIT kinase domain (6). The most common (66% of cases) involves abrogation of autokinase inhibition through a deletion in the JMR (exon 11; ref. 3). Alternative gain-of-function cKIT mutations include point mutations in exon 9 of the dimerization domain (13% of cases), exon 13 of the ATP-binding domain, or exon 17 of the phosphotransferase domain (both <3% of cases; ref. 13). Moreover, gene expression studies have shown that cKIT is up-regulated in GIST tumor samples leading to exacerbated downstream signaling (14). Due to its conspicuous location on the plasma membrane, high frequency of mutation, and key role in oncogenesis, cKIT is an attractive target for the development of small-molecule inhibitors.

Imatinib (Gleevec, STI571, Novartis, Basel, Switzerland) is an inhibitor of RTKs, including platelet-derived growth factor receptor {alpha} (PDGFRA), cABL, and cKIT (15), which was originally developed to treat chronic myeloid leukemia (CML) by blocking the ATP-binding pocket of the mutant kinase produced by the oncogenic BCR-ABL fusion gene (16). However, it was shortly afterward shown that imatinib can also be used to treat GIST (17) and leads to improved survival (18). Imatinib-induced tumor shrinkage is thought to result mainly from cKIT inhibition; however, inhibition of related RTKs, such as PDGFRA, could also play a role in the antitumor activity of this drug. Structural studies have shown that imatinib binds to the cKIT kinase domain I and abolishes the kinase-active conformation (9). Furthermore, imatinib inhibits cKIT phosphorylation and activation in vitro (19) leading to decreased downstream proliferative signals and increased apoptotic activity in GIST cell lines (11, 19).

Although imatinib treatment represents a major step forward in the management of GIST, most patients develop imatinib resistance, resulting in disease progression (18). Several point mutations have been identified to associate with imatinib resistance in vitro and in vivo (20). These point mutations include the Thr670Ile mutation at the "gatekeeper" residue Thr670 of the kinase domain (21) and two distinct mutations (Tyr823Asp and Asp816Val) that occur in the kinase domain activation loop (22, 23). Recently, a Val654Ala mutation in the kinase domain I of the same oncogenic cKIT allele was shown to be associated with imatinib resistance and relapse in GIST patients (24). These findings highlight the importance of constitutive cKIT signaling in GIST persistence.

In this study, we corroborate the clinical significance of the Val654Ala mutation reported previously by Chen et al. and next investigate the structural and functional significance of this mutation in conferring resistance to imatinib. We show through molecular modeling of the imatinib-cKIT complex that Val654Ala removes key contacts between the cKIT kinase domain I and imatinib, resulting in decreased binding affinity. Moreover, we show that, although an oncogenic cKIT mutant is inactive in the presence of imatinib, a cKIT allele harboring both the oncogenic mutation and the Val654Ala mutation remains constitutively active. These results provide insight into the structural and functional consequences of genetic changes that occur in the development of imatinib-resistant GIST.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Clinical Information
Patient A (female; date of birth, March 18, 1942) and patient B (female; date of birth, April 22, 1956) were diagnosed with GIST localizing to the stomach, liver, and omental regions and the stomach and omental regions, respectively. Both patients obtained good partial responses on 400 mg/d imatinib (treatment start dates, August 20, 2001 for patient A and August 10, 2001 for patient B). After a period of 12 and 14 months for patients A and B, respectively, both developed progressive disease. Despite increased dosages of imatinib (to 600 and 800 mg/d for patients A and B, respectively), disease progression was unimpeded and both patients eventually succumbed. Both patients provided written informed consent to perform ultrasound-guided biopsies at the time of disease progression to compare cKIT mutational status in the recurrent tumor masses with that of the original surgical diagnostic biopsies.

Genomic Sequence Analysis
Genomic DNA was extracted from four paraffin-embedded GIST biopsies using QIAamp DNA Mini kits (Qiagen, Inc., West Sussex, United Kingdom) according to the manufacturer's instructions. Individual cKIT exons were amplified to assay mutation status with primers designed to anneal on either side of the exon (Table 1 ). PCR product size was determined using DNA 7500 Lab-on-a-Chip assay on the Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA) according to the manufacturer's instructions. To sequence PCR products, amplicons were run on agarose electrophoresis gels, extracted with QIAquick columns (Qiagen), and cloned into pcDNA3 TOPO (Invitrogen, Paisley, United Kingdom). In all cases, sequencing was done with M13-20 forward sequencing primer and the reverse primer used in the PCR amplification.


View this table:
[in this window]
[in a new window]
  		Table 1. cKIT exons sequenced and primers used for PCR amplification

 
cKIT Cloning
cDNA for wild-type (WT) cKIT was generously donated by Dr. A. Moses (Oregon Health Sciences University, Portland, OR). Mutant cKIT cDNAs were constructed with the QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) and the following respective primers: constitutively active cKIT mutant (Lys642Glu; 5'-GGAAGCCCTCATGTCTGAACTCGAAGTCCTGAGTTACCTTGGTAATC and 5'-GATTACCAAGGTAACTCAGGACTTCGAGTTCAGACATGAGGGCTTCC), imatinib-resistant cKIT mutant (Val654Ala; 5'-CCTTGGTAATCACATGAATATTGCGAATCTACTTGGAGCCTGCACCATTGG and 5'-CCAATGGTGCAGGCTCCAAGTAGATTCGCAATATTCATGTGATTACCAAGG), and imatinib-resistant/constitutively active cKIT double mutant (Lys642Glu/Val654Ala) cloned from Val654Ala cDNA using the primers described above.

Cell Culture, Immunoprecipitation, and Western Blotting
Human embryonic kidney 293T cells were plated at a density of 2.5 x 106 per 10-cm dish and grown in DMEM (Invitrogen) plus 10% fetal bovine serum (Sigma, Dorset, United Kingdom). Twenty-four hours later, the cells were transfected with 5 µg of either pcDNA cKIT WT, pcDNA cKIT Lys642Glu, pcDNA cKIT Val654Ala, or pcDNA cKIT Lys6452Glu/Val654Ala using FuGENE 6 transfection reagent (Roche, East Sussex, United Kingdom) according to the manufacturer's instructions. Twenty-four hours after transfection, the medium was changed to 0.5% fetal bovine serum-DMEM. One day after the medium change, the cells were treated for 4 hours with imatinib (0–1 µmol/L) and then harvested and washed in PBS. All subsequent procedures were carried out on ice or in a 4°C cold room. The cells were lysed in lysis buffer [20 mmol/L Tris-HCl (pH 8.0), 137 mmol/L NaCl, 10% glycerol, 1% NP40, phosphatase inhibitors cocktails 1 and 2 (1:100 dilution; Sigma), and protease inhibitor cocktail (1:100 dilution; Sigma)] and protein samples (500 µg) were precleared for 30 minutes with protein G-Sepharose beads (Amersham Pharmacia Biotech Ltd., Uppsala, Sweden) and then centrifuged. Supernatants were incubated with anti-cKIT antibody (sc-17806, Santa Cruz Biotechnology, Santa Cruz, CA) for 2 hours on a rotating platform. Protein G-Sepharose (40 µL) was then added to the mixture and incubated/rotated for a further 1.5 hours. Beads were washed four times with lysis buffer, denatured in SDS loading buffer with ß-mercaptoethanol, and separated by 8% PAGE. Separated protein was transferred onto polyvinylidene difluoride membrane and immunoblotting was done at 4°C overnight with a horseradish peroxidase–conjugated anti-phosphotyrosine antibody (BD Biosciences, Palo Alto, CA) for phospho-cKIT and an anti-cKIT antibody (sc-17806, Santa Cruz Biotechnology) for input cKIT levels.

Molecular Modeling
General. Molecules were analyzed/visualized using SYBYL 7.0 (Tripos, Inc., St. Louis, MO). Electrostatic charges for the protein and the inhibitor were calculated with the Gasteiger-Hückel method (25). Structures were minimized using the Amber force field (version 7.0 or FF02) with a steepest descent gradient of 200 iterations followed by a conjugate gradient of 0.01 kcal/mol or a maximum of 10K iterations as termination criteria. During minimization, an 8-Å nonbonded cutoff was applied. The same force field and variables were used to calculate binding energies through the DOCK algorithm available in the same software suite.

Homology and Imatinib-Binding Site Construction. A computer model of the cKIT protein tyrosine kinase domain (Gly601-Glu925) excluding the kinase insert regions (Cys691-Leu764) was generated using the program COMPOSER (26), part of the molecular modeling package SYBYL version 6.92. Two crystal structures from the SYBYL Binary Protein Data Bank (PRODAT) library were selected as templates [i.e., the insulin receptor kinase (1irk; identity, 31.5%) and murine cABL kinase (1iep; identity, 29.5%)]. Both structures contain the activation loop in the inactive conformation and the latter has the inhibitor bound. Based on optimal sequence alignments, the structurally conserved regions were determined by an iterative approach, improving both the multiple alignment and the subsequent structurally conserved region framework by pairwise Needleman and Wunsch dynamic programming procedures with a similarity matrix constructed from inter-C distances. The backbone of each structurally conserved region of cKIT kinase was then built by fitting the corresponding structurally conserved region from one of the known homologues to the appropriate region of the framework. Some of the structurally conserved region had to be extended manually to secure that important features were incorporated in the model. Side chains were added using a knowledge-based approach, taking into account the backbone secondary structure and the side chains at the corresponding residues of the templates. Structurally variable regions of the model were constructed by the loop search algorithm within SYBYL. A few manual interventions were necessary to correct backbone geometry, unacceptable side chain conformations, or local side chain clashes. The software was unable to create an acceptable model of the c-helical region (Ser628-His650). The corresponding sequence from insulin receptor kinase (Ser1035-His1057) was used as a template after the necessary mutations. The final structure was minimized and submitted for quality control to the WHATIF server.6

Imatinib Docking. Our cKIT model was aligned with the murine cABL crystal structure and the imatinib molecule was transferred to the model. The inhibitor was assigned AMBER force field atoms and the complex was minimized up to a root mean square gradient of 0.01 kcal/mol (conjugate gradient).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The cKIT Missense Mutation Val654Ala Is Present in Relapsing GIST Tumor Biopsies following Initial Response to Imatinib Treatment
Two GIST patients were identified presenting with progressive GISTs following initial tumor regression with imatinib (Table 1; Fig. 1A ). To examine the mutational status of cKIT before and after imatinib treatment in these patients, genomic DNA was extracted from tumor biopsy material and selected cKIT exons were PCR amplified and sequenced. Exons 9, 11 to 14, 16, and 17 were chosen because they encode either intracellular functional domains or peptide sequences homologous with regions of BCR-ABL mutated in imatinib-resistant chronic myeloid leukemia (27). Paraffin-embedded biopsy material taken before treatment was used as control material. Sequencing for both pretreatment and posttreatment samples showed a 13-residue amino acid deletion in patient A (Val561-Asp573) and a 6-residue deletion in patient B (Glu556-Val561) located within the juxtamembrane region (data not shown). Sequencing of biopsies taken following imatinib resistance revealed substitution of nucleotide 1,982 of the genomic sequence encoding cKIT from thymine to cytosine in both patients (Fig. 1B). This substitution corresponds to a missense mutation in the protein sequence, with Val654 being replaced by alanine. Peptide sequences for members of the type III RTK family (KIT, ABL, FLT-3, and PDGFRA) and representatives from other RTK families [SRC, vascular endothelial growth factor receptor-2 (VEGFR2), insulin receptor (INSR), fibroblast growth factor receptor 1 (FGFR1), and epidermal growth factor receptor (EGFR)] were aligned and Val654 was found to be highly conserved in these RTKs (Fig. 1C).



View larger version (74K):
[in this window]
[in a new window]
  		Figure 1. cKIT sequencing in GIST before and after imatinib treatment reveals point mutation in conserved residue Val654. A, computed tomography scans from a representative patient showing initial tumor mass, tumor regression following imatinib treatment at 400 mg/d (later increased to 800 mg/d), and relapsed tumor. B, sequencing chromatogram of cKIT exon 13 before (i) and after (ii) administration of imatinib. Arrow, point mutated residue T1982C. C, peptide sequence alignments show the conservation of Val654 (asterisk) in type III RTK family (KIT, ABL, FLT-3, and PDGFRa) and representatives from other RTK families (SRC, VEGFR2, INSR, FGFR1, and EGFR). Conserved amino acids are color coded according to residue identity. NCBI protein accession codes are shown in parentheses. Conserved residue number is shown in column 2. Alignments were done with BioEdit version 7.0.5 (http://www.mbio.ncsu.edu/BioEdit/).

 
Imatinib/cKIT-Binding Models Show Loss of Stabilizing Contacts in the Val654Ala cKIT Mutant Compared with WT cKIT
To investigate the effect of the Val654Ala on cKIT protein structure, we created a model of the protein based on the crystal structures of related RTKs. The model was based on insulin receptor kinase and cABL both in the inactive conformation. The choice of these two structures was founded on the assumption that inactive structures contain the most relevant information for building the cKIT-imatinib interaction model. Similar approaches have been used to model the imatinib resistance for other cKIT mutants (28).

The model required manual construction of the c-helical region of cKIT (Fig. 2A ). This was necessary because the c-helical region built by COMPOSER was too short and did not present an acceptable conformation for the Glu640 side chain. Glu640 forms a critical hydrogen bond with Lys823 conserved in similar kinases (29) that bridges the c-helix to the ß-sheet hinge region.



View larger version (78K):
[in this window]
[in a new window]
  		Figure 2. Traces of cKIT model and imatinib binding. A, model of the inactive cKIT kinase. The C{alpha} tube (yellow) illustrates the well-conserved kinase structural features, including the A-loop and the c-helical region. The predicted imatinib (CPK spacefill representation) binding conformation is characterized by an extensive range of contacts with the kinase but does not interact with the activation loop (red). B, schematic drawing of the interactions between cKIT and imatinib generated by LIGPLOT. Residues forming van der Walls interactions are illustrated (red hemispheres). Residues participating in hydrogen bonds (dotted green) are shown (ball-and-stick representation) along with the relevant donor-acceptor distance given in Å. Box, the only false interaction assigned in our model. C, comparison between the model (yellow tube) and the crystal structure (Protein Data Bank code 1T46; green trace). Key residues around the drug are depicted according to the same color scheme. The relative position of the drug is also depicted (red for the model and magenta for the crystal structure). The weaknesses of the model are highlighted; shorter N-lobe and the position/orientation of Tyr609 and Phe811 (spacefill atoms). D, detail of the imatinib (red) binding region with cKIT. Val654 (blue spheres) is in a key position supporting the correct alignment of the key hydrogen bonds to Asp810 and Thr670. Important hydrogen bonds (green) and salt bridges (orange) are indicated; for distances, see B.

 
The quality of the final model was assessed and deemed acceptable using the WHATIF software server, with only 9% of the amino acids adopting forbidden Psi/Phi combination in the Ramachadran plot (data not shown). All the main features of an archetypical kinase domain were successfully reproduced (Fig. 2A) and the conformation of the activation loop was in accordance with related structures of inactivated kinases (30).

During completion of this work, Mol et al. (31) resolved the coordinates of the cKIT-imatinib complex (Protein Data Bank code 1t46). Our model was in excellent agreement with the crystal structure; however, three weaknesses were identified. The outer P-loop in the N-lobe region was shorter than in the X-ray structure (Fig. 2C). This problem was due to the inability of the software to correctly align the cKIT kinase insert region sequence with our template structures. As a result of this shortcoming, Tyr609 was considered to interact with the imatinib pyridine ring (Fig. 2B and C). The crystal structure showed this interaction to be false with Leu595, providing the hydrophobic shielding necessary for the pyridine ring in the same coordinates. Finally, the geometry of the Phe811 interaction with the drug is different (Fig. 2C and D). In the model, the phenyl ring is accommodated in a {pi}-stacking interaction with the pyrimidine ring, whereas in the crystal structure the pyrimidine ring is slightly further away and its plane is vertical to the phenylalanine ring. All the other interactions, including the hydrophobic contacts of Val654, were modeled very well with all the key amino acids located within 0.5 to 2.5 Å of the crystal structure coordinates. The root mean square value for the {alpha}-carbon atoms of the drug-binding region was 1.9 Å, indicating a good crystal structure mimicry by the model.

The model indicated that Val654 interacted with the central diaminophenyl ring of imatinib and contributed significantly to the overall binding energy. Additionally, the hydrophobic bulk of the side chain seems to be necessary to force imatinib in a position where it can form the two key hydrogen bonds to the backbone NH of Asp810 and the carboxylic group of Glu640. The geometry of this bond can be easily disrupted in the presence of the Val654Ala mutation as the drug moves closer to the alanine residue (Fig. 2D). This disruption was observed during minimization where the amide oxygen of imatinib moved out of range from the Asp810 backbone NH and changed orientation significantly. In a similar fashion, the length of the second crucial hydrogen bond to Thr670 increased significantly from 1.8 to 2.9 Å. A comprehensive schematic of the contacts between imatinib and the protein is presented in Fig. 2B. The publication of the crystal structure offered the opportunity to quantify our original observations by calculating the relative binding energies of the drug in the WT and the Val654Ala mutant form of the receptor. Using established methodology (32, 33) and the cKIT crystal structure data, we obtained a difference of 2.25 kcal/mol between the binding energies of the WT and Val654Ala mutant for imatinib (Table 2 ).


View this table:
[in this window]
[in a new window]
  		Table 2. Relative binding energies of imatinib to cKIT as calculated with the Amber force field

 
An Intra-allelic Val654Ala Mutation Abolishes Imatinib-Mediated cKIT Deactivation of the Oncogenic cKIT Mutant Lys642Glu
To investigate the functional consequences of the Val654Ala mutation on imatinib-mediated cKIT inhibition, a series of cKIT mutants were cloned and expressed in HEK 239T cells. We used a 293T cell line to assess the phosphorylation status of cKIT rather than a GIST cell line to avoid masking of results by endogenous cKIT expression. Although previous studies indicate that selected downstream signaling pathways activated by different gain-of-function cKIT mutants may differ, tyrosine phosphorylation of cKIT is a consistent feature for all gain-of-function cKIT mutants (11), including the Lys642Glu mutant (34). The Lys642Glu mutation likely interferes with the kinase autoinhibitory mechanism by removing the key hydrogen bond between Lys642 and the backbone carbonyl of Thr574 in the JMR. This interaction is involved in supporting the autoinhibitory peptide under the c-helical region. Thus, the Lys642Glu mutant would not require the activating phosphorylation of Tyr568 because the JMR will be permanently destabilized away from its inhibitory conformation out of the kinase active center. Imatinib does not interact with the JMR when binding to oncogenic cKIT and so should bind cKIt oncogenes with JMR deletions in a similar manner to cKIT Lys642Glu. Hence, despite occupying a low percentage of naturally occurring cKIT mutations, the Lys642Glu mutant is a useful tool for studying cKIT-imatinib interactions.

After validating transfection and expression of the cKIT constructs (Fig. 3, input lanes ), the activation of cKIT in the presence or absence of imatinib was assessed. Immunoprecipitation of total cKIT followed by Western blotting with a pan-phosphotyrosine horseradish peroxidase–conjugated antibody was used to detect phosphorylated (activated) cKIT levels in each treatment group (Fig. 3). Whereas phospho-activation of WT Lys642Glu mutant cKIT was diminished in the presence of imatinib, cKIT constructs with intra-allelic Val654Ala mutations (Val654Ala and Lys642Glu/Val654Ala) remained active despite 1 µmol/L imatinib dosages (Fig. 3).



View larger version (16K):
[in this window]
[in a new window]
  		Figure 3. Intra-allelic Val654Ala abolishes imatinib-mediated cKIT inactivation. Top, immunoprecipitations (IP) show imatinib-mediated inactivation of cKIT by decreased levels of phosphorylated cKIT in both WT and oncogenic mutant (Lys642Glu) in treated versus untreated lanes. No change in the level of mature cKIT occurs in either Val654Ala or Lys642Glu/Val654Ala, indicating that intra-allelic Val654Ala abolishes imatinib-mediated cKIT inactivation. Bottom, equal levels of cKIT expression by Western blot. Migration end points for 150- and 100-kDa standards are shown (right). The two bands for cKIT are the 145-kDa mature form and the 125-kDa immature form.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Recent studies in targeted cancer therapy have illustrated how the mutation status of the target receptor is valuable in predicting tumor response to specific tyrosine kinase inhibitors (35). For patients with non–small cell lung cancer treated with the anti-EGFR tyrosine kinase inhibitor gefitinib (Iressa, AstraZeneca, Macclesfield, United Kingdom), selected EGFR1 mutations correlate with better therapeutic response (36, 37). In the context of GIST, specific mutations in cKIT are associated with better initial responses to imatinib (3, 38). Furthermore, SU11248 (Pfizer, Cambridge, MA; ref. 20), a tyrosine kinase inhibitor that inhibits VEGFR2, cKIT, PDGFR, and FLT3, has been shown to have clinical benefit in patients that have developed imatinib resistance (39). Moreover, imatinib resistance is not limited to patients with GIST and was initially reported in CML patients as being linked with specific BCR-ABL mutations (27, 40). These findings indicate that RTK genotyping in patients would be critical to establishing personalized treatment regimens and may be a successful tool in preventing the development of drug resistance and improving outcomes.

Here, we show that individuals with GIST who initially responded to imatinib developed identical mutations at time of disease progression to that recently described from another single center (24). This corroborates the importance of the Val654Ala mutation in conferring resistance to imatinib. Using the structural information available, we determined how the Val654Ala mutation could abolish cKIT inhibition by imatinib. On initial observation, a conservative mutation from valine to alanine should be unlikely to induce any large conformational changes that would render the pocket unsuitable for imatinib binding. However, a detailed inspection of the minimization progression showed that Val654 stabilizes the position of the central diaminophenyl ring of the drug and rendered imatinib in an optimal position for critical H-bonds with Asp810 and Thr670 to form. A further contribution to the effects of the Val654Ala mutation could originate from the fact that alanine is unable to shield the central hydrophobic region of the binding pocket from solvent molecules (41). Water molecules intruding the pocket from this direction would destabilize the protein-drug complex. Our binding energy calculations show that the mutation destabilizes the cKIT-imatinib complex by 2.25 kcal/mol. In terms of differences in the Gibbs free energy of binding ({Delta}({Delta}G)binding = –RTln[(Kd)V654A / (Kd)wt]), this represents an unfavorable change in the binding constant between 1 and 2 orders of magnitude (42). Such a reduction could be sufficient to render imatinib less effective in physiologic concentrations associated with current standard dosages. Interestingly, Val654 is buried under Asn655 and both residues are located in the 10–amino acid region between the c-helix and the hinge loop (Fig. 2D), regions that are critical in regulation of cKIT activation. The conspicuous location of this residue and observed conservation of Val654 in type III and other family RTKs indicates that it could be involved in RTK regulatory mechanisms or mutated in other malignancies treated with imatinib.

RTKs occupy an apical position in proliferative and antiapoptotic signaling cascade hierarchies, and oncogenic signals from mutant cKIT have been shown to include several pathways critical to oncogenesis, including MAPK p42/44, AKT, and various STAT pathways (11). Some of the same pathways can be activated by PDGFRA, which is also a target for imatinib (15) and has been shown to harbor gain-of-function mutations in GIST lacking cKIT mutations (3). Coupled with the observation that cKIT mutations occur early in GIST development (43) and that imatinib-resistant tumors can harbor the Val654Ala or other cKIT resistance mutations (21, 22), the emergence of Val654Ala during imatinib treatment may be sufficient to invoke the relapse observed in imatinib-resistant GIST tumors. Constitutive activation of Lys642Glu/Val654Ala cKIT in the presence of imatinib in vitro could account for the imatinib-resistant phenotype in patients with relapsing GIST and supports the hypothesis that Val654Ala plays a critical role in the relapse and maintenance of imatinib-resistant tumors.

Overcoming the development of therapeutic resistance in tumors requires a diverse arsenal of drugs and administration protocols (44). Recent studies have indicated that downstream signal transduction may vary based on the cKIT mutant type (11) and that unique gene expression profiles are associated with different cKIT-mutant and PDGFRA-mutant types in GIST (45). As cKIT and PDGFRA mutations are implicated in the majority of GIST cases (45), perhaps the most broadly applicable complementary therapeutic to imatinib should also target cKIT and PDGFR to halt vital oncogenic signaling at the source.

The development of new therapeutic cKIT inhibitors requires consideration of how both the active and the inactive conformations of cKIT are bound by the drug. Imatinib does not bind the inactive (autoinhibited) conformation of cKIT, as the piperazine group of imatinib binds in the same region as the juxtamembrane region inhibitory domain and therefore interferes with one of the natural regulatory mechanisms of cKIT (9). Currently, most drugs under development target the active form of the enzyme (e.g., PP1, OSI930, and SU11248; see Fig. 4 ); however, the development of drugs like PD173955 that bind both inactive and active forms of the enzyme (46, 47) could be more effective as a frontline treatment by delaying the onset of conformation-specific drug resistance mutations. Alternatively, some features of the imatinib molecule could be removed or modified in an effort to address the Val654Ala mutation itself and tailor therapeutics to patient cKIT genotype. The introduction of additional interactions in regions presumably unaffected by the Val654Ala mutation is one possibility (e.g., replacing the piperazine moiety with a small peptide or peptidomimetic that resembles the autoinhibitory JMR). Such an approach would target two different binding regions of a kinase and should be successful in overcoming mutations that do not alter the equilibrium between active and inactive forms of the kinase. This technique has been used previously to inhibit the active form of insulin receptor kinase and could represent an opportunity to develop effective combination therapies for GIST (48). Finally, various analogues of imatinib and other inhibitors known to further stabilize the inactive state of similar RTKs (e.g., PD173955; Fig. 4; ref. 49) should be tested for efficacy in inactivating both the WT and the Val654Ala mutant. The results could provide more information about the effect of a mutation in the binding pocket of cKIT and perhaps offer some insight as to whether there is any functional or therapeutic significance to the highly conserved Val654 in type III RTKs.



View larger version (41K):
[in this window]
[in a new window]
  		Figure 4. Structures of known cKIT inhibitors. Background color indicates different conformation of the kinase during binding: kinase with the activation loop (Fig. 2A, red) in the inactive conformation (aqua), active or inactive conformation (green), and ATP-competitive RTK antagonists that bind only the active conformation of the enzyme (yellow).

 

   Acknowledgments
 
We thank the patients involved in this study and Drs. D. Lagos and S. Henderson for helpful revision of the article.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

6 http://swift.cmbi.kun.nl/WIWWWI/. Back

Received 3/11/05; revised 9/28/05; accepted 10/19/05.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Sircar K, Hewlett BR, Huizinga JD, et al. Interstitial cells of Cajal as precursors of gastrointestinal stromal tumors. Am J Surg Pathol 1999;23:377–89.[CrossRef][Medline]
  2. Hirota S, Isozaki K, Moriyama Y, et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 1998;279:577–80.[Abstract/Free Full Text]
  3. Heinrich MC, Corless CL, Demetri GD, et al. Kinase mutations and imatinib response in patients with metastatic gastrointestinal stromal tumor. J Clin Oncol 2003;21:4342–9.[Abstract/Free Full Text]
  4. Nagata H, Worobec AS, Oh CK, et al. Identification of a point mutation in the catalytic domain of the protooncogene c-kit in peripheral blood mononuclear cells of patients who have mastocytosis with an associated hematologic disorder. Proc Natl Acad Sci U S A 1995;92:10560–4.[Abstract/Free Full Text]
  5. Przygodzki RM, Hubbs AE, Zhao FQ, O'Leary TJ. Primary mediastinal seminomas: evidence of single and multiple KIT mutations. Lab Invest 2002;82:1369–75.[Medline]
  6. Dibb NJ, Dilworth SM, Mol CD. Switching on kinases: oncogenic activation of BRAF and the PDGFR family. Nat Rev Cancer 2004;4:718–27.[CrossRef][Medline]
  7. Blechman JM, Lev S, Barg J, et al. The fourth immunoglobulin domain of the stem cell factor receptor couples ligand binding to signal transduction. Cell 1995;80:103–13.[CrossRef][Medline]
  8. Lev S, Yarden Y, Givol D. Dimerization and activation of the kit receptor by monovalent and bivalent binding of the stem cell factor. J Biol Chem 1992;267:15970–7.[Abstract/Free Full Text]
  9. Mol CD, Dougan DR, Schneider TR, et al. Structural basis for the autoinhibition and STI-571 inhibition of c-Kit tyrosine kinase. J Biol Chem 2004;279:31655–63.[Abstract/Free Full Text]
  10. Sattler M, Salgia R. Targeting c-Kit mutations: basic science to novel therapies. Leuk Res 2004;28 Suppl 1:S11–20.[Medline]
  11. Duensing A, Medeiros F, McConarty B, et al. Mechanisms of oncogenic KIT signal transduction in primary gastrointestinal stromal tumors (GISTs). Oncogene 2004;23:3999–4006.[CrossRef][Medline]
  12. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004;10:789–99.[CrossRef][Medline]
  13. Corless CL, Fletcher JA, Heinrich MC. Biology of gastrointestinal stromal tumors. J Clin Oncol 2004;22:3813–25.[Abstract/Free Full Text]
  14. Nielsen TO, West RB, Linn SC, et al. Molecular characterisation of soft tissue tumours: a gene expression study. Lancet 2002;359:1301–7.[CrossRef][Medline]
  15. Heinrich MC, Griffith DJ, Druker BJ, et al. Inhibition of c-kit receptor tyrosine kinase activity by STI 571, a selective tyrosine kinase inhibitor. Blood 2000;96:925–32.[Abstract/Free Full Text]
  16. Druker BJ, Sawyers CL, Kantarjian H, et al. Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 2001;344:1038–42.[Abstract/Free Full Text]
  17. Joensuu H, Roberts PJ, Sarlomo-Rikala M, et al. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001;344:1052–6.[Free Full Text]
  18. Verweij J, Casali PG, Zalcberg J, et al. Progression-free survival in gastrointestinal stromal tumours with high-dose imatinib: randomised trial. Lancet 2004;364:1127–34.[CrossRef][Medline]
  19. Tuveson DA, Willis NA, Jacks T, et al. STI571 inactivation of the gastrointestinal stromal tumor c-KIT oncoprotein: biological and clinical implications. Oncogene 2001;20:5054–8.[CrossRef][Medline]
  20. Daub H, Specht K, Ullrich A. Strategies to overcome resistance to targeted protein kinase inhibitors. Nat Rev Drug Discov 2004;3:1001–10.[CrossRef][Medline]
  21. Tamborini E, Bonadiman L, Greco A, et al. A new mutation in the KIT ATP pocket causes acquired resistance to imatinib in a gastrointestinal stromal tumor patient. Gastroenterology 2004;127:294–9.[CrossRef][Medline]
  22. Wakai T, Kanda T, Hirota S, et al. Late resistance to imatinib therapy in a metastatic gastrointestinal stromal tumour is associated with a second KIT mutation. Br J Cancer 2004;90:2059–61.[Medline]
  23. Ma Y, Zeng S, Metcalfe DD, et al. The c-KIT mutation causing human mastocytosis is resistant to STI571 and other KIT kinase inhibitors; kinases with enzymatic site mutations show different inhibitor sensitivity profiles than wild-type kinases and those with regulatory-type mutations. Blood 2002;99:1741–4.[Abstract/Free Full Text]
  24. Chen LL, Trent JC, Wu EF, et al. A missense mutation in KIT kinase domain 1 correlates with imatinib resistance in gastrointestinal stromal tumors. Cancer Res 2004;64:5913–9.[Abstract/Free Full Text]
  25. Gasteiger J, Marsilli M. Iterative partial equalization of orbital electronegativity—a rapid access to atomic charges. Tetrahedro 1980;36:3219–28.[CrossRef]
  26. Sutcliffe MJ, Haneef I, Carney D, Blundell TL. Knowledge based modelling of homologous proteins. Part I. Three-dimensional frameworks derived from the simultaneous superposition of multiple structures. Protein Eng 1987;1:377–84.[Abstract/Free Full Text]
  27. Gorre ME, Mohammed M, Ellwood K, et al. Clinical resistance to STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification. Science 2001;293:876–80.[Abstract/Free Full Text]
  28. Foster R, Griffith R, Ferrao P, Ashman L. Molecular basis of the constitutive activity and STI571 resistance of Asp816Val mutant KIT receptor tyrosine kinase. J Mol Graph Model 2004;23:139–52.[CrossRef][Medline]
  29. Hubbard SR. Juxtamembrane autoinhibition in receptor tyrosine kinases. Nat Rev Mol Cell Biol 2004;5:464–71.[CrossRef][Medline]
  30. Nagar B, Bornmann WG, Pellicena P, et al. Crystal structures of the kinase domain of c-Abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res 2002;62:4236–43.[Abstract/Free Full Text]
  31. Mol CD, Fabbro D, Hosfield DJ. Structural insights into the conformational selectivity of STI-571 and related kinase inhibitors. Curr Opin Drug Discov Dev 2004;7:639–48.[Medline]
  32. Kalani MY, Vaidehi N, Hall SE, et al. The predicted 3D structure of the human D2 dopamine receptor and the binding site and binding affinities for agonists and antagonists. Proc Natl Acad Sci U S A 2004;101:3815–20.[Abstract/Free Full Text]
  33. Ai N, DeLisle RK, Yu SJ, Welsh WJ. Computational models for predicting the binding affinities of ligands for the wild-type androgen receptor and a mutated variant associated with human prostate cancer. Chem Res Toxicol 2003;16:1652–60.[Medline]
  34. Lux ML, Rubin BP, Biase TL, et al. KIT extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am J Pathol 2000;156:791–5.[Abstract/Free Full Text]
  35. Arteaga CL, Baselga J. Tyrosine kinase inhibitors: why does the current process of clinical development not apply to them? Cancer Cell 2004;5:525–31.[CrossRef][Medline]
  36. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129–39.[Abstract/Free Full Text]
  37. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497–500.[Abstract/Free Full Text]
  38. Debiec-Rychter M, Dumez H, Judson I, et al. Use of c-KIT/PDGFRA mutational analysis to predict the clinical response to imatinib in patients with advanced gastrointestinal stromal tumours entered on phase I and II studies of the EORTC Soft Tissue and Bone Sarcoma Group. Eur J Cancer 2004;40:689–95.[CrossRef][Medline]
  39. Demetri GD, Desai J, Fletcher JA, et al. SU11248, a multi-targeted tyrosine kinase inhibitor, can overcome imatinib (IM) resistance caused by diverse genomic mechanisms in patients (pts) with metastatic gastrointestinal stromal tumor (GIST). J Clin Oncol 2004;22:3001–1.
  40. Shah NP, Nicoll JM, Nagar B, et al. Multiple BCR-ABL kinase domain mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia. Cancer Cell 2002;2:117–25.[CrossRef][Medline]
  41. Gambacorti-Passerini CB, Gunby RH, Piazza R, et al. Molecular mechanisms of resistance to imatinib in Philadelphia-chromosome-positive leukaemias. Lancet Oncol 2003;4:75–85.[CrossRef][Medline]
  42. Ai N, DeLisle RK, Yu SJ, Welsh WJ. Computational models for predicting the binding affinities of ligands for the wild-type androgen receptor and a mutated variant associated with human prostate cancer. Chem Res Toxicol 2003;16:1652–60.[Medline]
  43. Logrono R, Jones DV, Faruqi S, Bhutani MS. Recent advances in cell biology, diagnosis, and therapy of gastrointestinal stromal tumor (GIST). Cancer Biol Ther 2004;3:251–8.[Medline]
  44. Courtneidge SA. Cancer: escape from inhibition. Nature 2003;422:827–8.[Medline]
  45. Subramanian S, West RB, Corless CL, et al. Gastrointestinal stromal tumors (GISTs) with KIT and PDGFRA mutations have distinct gene expression profiles. Oncogene 2004;23:7780–90.[CrossRef][Medline]
  46. Krystal GW, Honsawek S, Kiewlich D, et al. Indolinone tyrosine kinase inhibitors block Kit activation and growth of small cell lung cancer cells. Cancer Res 2001;61:3660–8.[Abstract/Free Full Text]
  47. Tatton L, Morley GM, Chopra R, Khwaja A. The Src-selective kinase inhibitor PP1 also inhibits Kit and Bcr-Abl tyrosine kinases. J Biol Chem 2003;278:4847–53.[Abstract/Free Full Text]
  48. Parang K, Till JH, Ablooglu AJ, et al. Mechanism-based design of a protein kinase inhibitor. Nat Struct Biol 2001;8:37–41.[CrossRef][Medline]
  49. Nagar B, Bornmann WG, Pellicena P, et al. Crystal structures of the kinase domain of c-Abl in complex with the small molecule inhibitors PD173955 and imatinib (STI-571). Cancer Res 2002;62:4236–43.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Molecular Cancer TherapeuticsHome page
K. G. Roberts, A. F. Odell, E. M. Byrnes, R. M. Baleato, R. Griffith, A. B. Lyons, and L. K. Ashman
Resistance to c-KIT kinase inhibitors conferred by V654A mutation
Mol. Cancer Ther., March 1, 2007; 6(3): 1159 - 1166.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
M. C. Heinrich, C. L. Corless, C. D. Blanke, G. D. Demetri, H. Joensuu, P. J. Roberts, B. L. Eisenberg, M. von Mehren, C. D.M. Fletcher, K. Sandau, et al.
Molecular Correlates of Imatinib Resistance in Gastrointestinal Stromal Tumors
J. Clin. Oncol., October 10, 2006; 24(29): 4764 - 4774.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by McLean, S. R.
Right arrow Articles by Boshoff, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McLean, S. R.
Right arrow Articles by Boshoff, C.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online