Advertisement

Disruption of Bcr-Abl Coiled Coil Oligomerization by Design*

  • Andrew S. Dixon
    Affiliations
    Department of Pharmaceutics and Pharmaceutical Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108
    Search for articles by this author
  • Scott S. Pendley
    Affiliations
    Department of Pharmaceutics and Pharmaceutical Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108
    Search for articles by this author
  • Benjamin J. Bruno
    Affiliations
    Department of Pharmaceutics and Pharmaceutical Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108
    Search for articles by this author
  • David W. Woessner
    Affiliations
    Department of Pharmacology and Toxicology, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108
    Search for articles by this author
  • Adrian A. Shimpi
    Affiliations
    Juan Diego Catholic High School, Draper, Utah 84020
    Search for articles by this author
  • Thomas E. Cheatham III
    Correspondence
    To whom all other correspondence should be addressed: Dept. of Pharmaceutics and Pharmaceutical Chemistry, College of Pharmacy, University of Utah, 421 Wakara Way, Rm. 318, Salt Lake City, UT 84108. Fax: 801-585-3614.
    Affiliations
    Department of Pharmaceutics and Pharmaceutical Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108

    Department of Medicinal Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108
    Search for articles by this author
  • Carol S. Lim
    Correspondence
    To whom computational correspondence should be addressed.
    Affiliations
    Department of Pharmaceutics and Pharmaceutical Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84108
    Search for articles by this author
  • Author Footnotes
    * This work was supported, in whole or in part, by National Institutes of Health Grants CA129528 (to C. S. L.) and GM079383 (to T. E. C.). This work was also supported by an American Foundation for Pharmaceutical Education predoctoral fellowship (to A. S. D), Graduate Research Fellowship, University of Utah (to A. S. D.) the ALSAM Foundation (to A. A. S), and National Science Foundation Grant TG-MCA01S027 (to T. E. C.; for computer time).
    The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1 and S2.
Open AccessPublished:June 09, 2011DOI:https://doi.org/10.1074/jbc.M111.264903
      Oligomerization is an important regulatory mechanism for many proteins, including oncoproteins and other pathogenic proteins. The oncoprotein Bcr-Abl relies on oligomerization via its coiled coil domain for its kinase activity, suggesting that a designed coiled coil domain with enhanced binding to Bcr-Abl and reduced self-oligomerization would be therapeutically useful. Key mutations in the coiled coil domain of Bcr-Abl were identified that reduce homo-oligomerization through intermolecular charge-charge repulsion yet increase interaction with the Bcr-Abl coiled coil through additional salt bridges, resulting in an enhanced ability to disrupt the oligomeric state of Bcr-Abl. The mutations were modeled computationally to optimize the design. Assays performed in vitro confirmed the validity and functionality of the optimal mutations, which were found to exhibit reduced homo-oligomerization and increased binding to the Bcr-Abl coiled coil domain. Introduction of the mutant coiled coil into K562 cells resulted in decreased phosphorylation of Bcr-Abl, reduced cell proliferation, and increased caspase-3/7 activity and DNA segmentation. Importantly, the mutant coiled coil domain was more efficacious than the wild type in all experiments performed. The improved inhibition of Bcr-Abl through oligomeric disruption resulting from this modified coiled coil domain represents a viable alternative to small molecule inhibitors for therapeutic intervention.

      Introduction

      Coiled coil domains are ubiquitous protein structural motifs found in ∼10% of all eukaryotic proteins (
      • Taylor C.M.
      • Keating A.E.
      ). Characterized by a heptad repeat (sequence of seven amino acids) and association of two or more α-helices, coiled coils provide oligomerization capabilities (both homo- and hetero-oligomerization) useful for structural scaffolding and protein recognition. Coiled coil domains are critical for regulating many processes involved in the pathogenesis of various diseases (
      • Strauss H.M.
      • Keller S.
      ). Thus, rationally designed coiled coils can be used as therapeutics by interfering with the activity of a pathogenic protein through oligomeric disruption. One example is enfuvirtide (Fuzeon®), a fusion inhibitor that disrupts the helix bundle formation necessary for HIV-1 viral entry (
      • Matthews T.
      • Salgo M.
      • Greenberg M.
      • Chung J.
      • DeMasi R.
      • Bolognesi D.
      ), spawning the next generation of rationally designed fusion inhibitors (
      • Dwyer J.J.
      • Wilson K.L.
      • Davison D.K.
      • Freel S.A.
      • Seedorff J.E.
      • Wring S.A.
      • Tvermoes N.A.
      • Matthews T.J.
      • Greenberg M.L.
      • Delmedico M.K.
      ,
      • Otaka A.
      • Nakamura M.
      • Nameki D.
      • Kodama E.
      • Uchiyama S.
      • Nakamura S.
      • Nakano H.
      • Tamamura H.
      • Kobayashi Y.
      • Matsuoka M.
      • Fujii N.
      ,
      • Yan Z.
      • Tripet B.
      • Hodges R.S.
      ,
      • Eckert D.M.
      • Malashkevich V.N.
      • Hong L.H.
      • Carr P.A.
      • Kim P.S.
      ,
      • Sia S.K.
      • Carr P.A.
      • Cochran A.G.
      • Malashkevich V.N.
      • Kim P.S.
      ). The coiled coil reported here was designed to bind to the target (Bcr-Abl) better than the target protein binds to itself while exhibiting minimal homo-oligomerization.
      Coiled coils are very well characterized, and the high correlation between their sequence and structure (
      • Parry D.A.
      • Fraser R.D.
      • Squire J.M.
      ) is advantageous for rational design. Coiled coil oligomerization involves hydrophobic interactions, salt bridge formation, and helicity. In the heptad repeat, hydrophobic packing at the “a” and “d” positions (see Fig. 1, green residues) helps drive association with further stability provided by the residues at positions “g” and “e,” which are commonly charged and interact to form salt bridges (see Fig. 1, red and blue residues). Because charged residues are routinely found at the g and e positions, mutating these residues in a rational manner to add salt bridges to favor formation of hetero-oligomers (
      • Kammerer R.A.
      • Jaravine V.A.
      • Frank S.
      • Schulthess T.
      • Landwehr R.
      • Lustig A.
      • Garcia-Echeverria C.
      • Alexandrescu A.T.
      • Engel J.
      • Steinmetz M.O.
      ,
      • Spek E.J.
      • Bui A.H.
      • Lu M.
      • Kallenbach N.R.
      ,
      • Marqusee S.
      • Baldwin R.L.
      ) and charge-charge repulsions to reduce the formation of homo-oligomers (
      • Ryan S.J.
      • Kennan A.J.
      ,
      • Gurnon D.G.
      • Whitaker J.A.
      • Oakley M.G.
      ) can change the affinity and specificity of the coiled coil dimer.
      Figure thumbnail gr1
      FIGURE 1Ribbon diagrams (with corresponding helical wheel diagram below) of wild-type homodimer (A), wild-type-CCmut2 heterodimer (B), and CCmut2-CCmut2 homodimer (C). Gray ribbons (ribbon diagrams) or dots (helical wheel diagrams) represent the wild-type coiled coil domain, and cyan represents CCmut2. The side chains of key residues (Glu-34, Lys-39, Ser/Arg-41, Leu/Asp-45, Glu-46, Glu/Arg-48, Arg-53, Arg-55, and Gln/Glu-60) are shown as red (acidic), blue (basic), green (hydrophobic), yellow (serine), or black (glutamine) spheres (ribbon diagrams) or font (helical wheel diagrams). Dotted lines indicate possible ionic interactions, and solid lines indicate charge-charge repulsions. Ribbon diagrams were generated with UCSF Chimera starting with the Bcr coiled coil domain crystal structure (Protein Data Bank code 1K1F).
      Bcr-Abl exists primarily as a tetramer (more specifically as a dimer of dimers), facilitating the trans-autophosphorylation necessary to activate the tyrosine kinase domain (
      • Zhang X.
      • Subrahmanyam R.
      • Wong R.
      • Gross A.W.
      • Ren R.
      ,
      • Arlinghaus R.B.
      ). Oligomerization of Bcr-Abl is achieved through a coiled coil domain at the N terminus of the protein. Bcr-Abl constructs lacking the N-terminal coiled coil fail to induce chronic myelogenous leukemia (CML)
      The abbreviations used are: CML
      chronic myelogenous leukemia
      CC
      wild-type coiled coil domain from Bcr-Abl
      CCmut1
      mutant coiled coil domain with S41R, L45D, E48R, V49D, and Q60E mutations
      CCmut2
      mutant coiled coil domain with C38A, S41R, L45D, E48R, and Q60E mutations
      AMBER
      assisted model building with energy refinement
      MD
      molecular dynamics
      MM-PBSA
      molecular mechanics Poisson-Boltzmann/surface area
      NTA
      nuclear translocation assay
      CFA
      colony forming assay
      PS
      protein switch
      CrkL
      Crk-like
      EGFP
      enhanced GFP
      ANOVA
      analysis of variance.
      in a murine model (
      • Zhang X.
      • Subrahmanyam R.
      • Wong R.
      • Gross A.W.
      • Ren R.
      ,
      • He Y.
      • Wertheim J.A.
      • Xu L.
      • Miller J.P.
      • Karnell F.G.
      • Choi J.K.
      • Ren R.
      • Pear W.S.
      ), thus setting the stage for oligomeric disruption as a therapy. Proof of concept for using the coiled coil domain to inhibit Bcr-Abl activity has already been demonstrated through retroviral transfection of the wild-type coiled coil domain or addition of the purified protein attached to a cytoplasmic transduction peptide (
      • Mian A.A.
      • Oancea C.
      • Zhao Z.
      • Ottmann O.G.
      • Ruthardt M.
      ,
      • Beissert T.
      • Hundertmark A.
      • Kaburova V.
      • Travaglini L.
      • Mian A.A.
      • Nervi C.
      • Ruthardt M.
      ,
      • Beissert T.
      • Puccetti E.
      • Bianchini A.
      • Güller S.
      • Boehrer S.
      • Hoelzer D.
      • Ottmann O.G.
      • Nervi C.
      • Ruthardt M.
      ,
      • Huang S.F.
      • Liu D.B.
      • Zeng J.M.
      • Xiao Q.
      • Luo M.
      • Zhang W.P.
      • Tao K.
      • Wen J.P.
      • Huang Z.G.
      • Feng W.L.
      ,
      • Guo X.Y.
      • Cuillerot J.M.
      • Wang T.
      • Wu Y.
      • Arlinghaus R.
      • Claxton D.
      • Bachier C.
      • Greenberger J.
      • Colombowala I.
      • Deisseroth A.B.
      ). To further this approach, we designed and tested a mutant coiled coil (CCmut2) using both computational and in vitro experiments. This CCmut2 disfavors self-oligomerization (CCmut2:CCmut2) yet appears to bind more tightly to the target Bcr-Abl coiled coil domain (CCmut2:Bcr-Abl) than the native oligomerization partner of the target (Bcr-Abl:Bcr-Abl) and resulted in superior inhibition of Bcr-Abl.
      The coiled coil domain from Bcr-Abl consists of 72 amino acids composed of two α-helices (α1 (residues 5–15) and α2 (residues 28–67)) that form an N-shaped configuration with two parallel helices connected by a short linker (see Fig. 1A) (
      • Zhao X.
      • Ghaffari S.
      • Lodish H.
      • Malashkevich V.N.
      • Kim P.S.
      ). Upon dimerization, the resulting coiled coil has an antiparallel orientation with α2 at the core and α1 latching onto the backside of the opposing α2 helix (domain swapping) (
      • Taylor C.M.
      • Keating A.E.
      ,
      • Zhao X.
      • Ghaffari S.
      • Lodish H.
      • Malashkevich V.N.
      • Kim P.S.
      ). The majority of the dimer interface is composed of the classic knobs-in-holes type hydrophobic interactions from residues at the a and d positions (
      • Crick F.H.
      ). Further stabilization comes from four interchain salt bridges between residues in the α2 helices (Fig. 1A) as well as packing of aromatic residues from the α1 helix and opposing α2 helix. As seen in Fig. 1A, there are two uncharged residues (Ser-41 and Gln-60) that are in the appropriate position for the formation of salt bridges with two charged residues (Glu-32 and Glu-48). Thus, mutation of Ser-41 and Gln-60 to positively charged amino acids has the potential to provide two additional salt bridges with Bcr-Abl (Fig. 1B) and thus enhance binding. However, although this provides more salt bridges in the hetero-oligomer, these mutations alone are undesirable as they allow the formation of a greater number of salt bridges in the homo-oligomer. To reduce homo-oligomerization in the mutant coiled coil, residues proximal to charged residues on the opposing helix were considered as candidates for mutation to introduce charge-charge repulsion (
      • Ryan S.J.
      • Kennan A.J.
      ,
      • Gurnon D.G.
      • Whitaker J.A.
      • Oakley M.G.
      ). Leu-45 and Glu-48 were identified as two such residues (Fig. 1C). In addition, previous reports have incorporated a C38A mutation primarily for crystallization purposes (
      • Taylor C.M.
      • Keating A.E.
      ,
      • Zhao X.
      • Ghaffari S.
      • Lodish H.
      • Malashkevich V.N.
      • Kim P.S.
      ), and this mutation was also studied. Putative mutations were investigated first through molecular modeling and state-of-the art biomolecular simulation and free energy analysis to ascertain the impact on the coiled coil structure and stability. Such in silico methods provide a means to efficiently (and inexpensively) assess the influence of mutation. The resulting optimized mutant coiled coil (CCmut2) contains five mutations (C38A, S41R, L45D, E48R, and Q60E) and was further assessed in vitro. Taken together, these results demonstrate the effectiveness of the mutant coiled coil domain and, importantly, further illustrate the ability to rationally modify an existing coiled coil domain to improve therapeutic efficacy.

      EXPERIMENTAL PROCEDURES

      Computational Modeling and Simulation

      The initial model structure was obtained from the crystal structure (refined to 2.2-Å resolution) of the N-terminal oligomerization domain of Bcr-Abl (Protein Data Bank code 1K1F, chains A and B) (
      • Zhao X.
      • Ghaffari S.
      • Lodish H.
      • Malashkevich V.N.
      • Kim P.S.
      ). Selenomethionine residues were mutated to methionine, and position 39 was mutated back to cysteine to maintain consistency with the natural protein. Amino acid side chain mutations were introduced using DeepView (Swiss-PdbViewer) (
      • Guex N.
      • Peitsch M.C.
      ) and by hand with the LEAP module from AMBER 9 (
      • Case D.
      • Darden T.
      • Cheatham 3rd, T.
      • Simmerling C.
      • Wang J.
      • Duke R.
      • Luo R.
      • Merz K.
      • Pearlman D.
      • Crowley M.
      ). Molecular dynamics (MD) and free energy simulations were performed to assess the structure and stability of the model coiled coil structures. Such methods have proven useful for reproducing and predicting the structure of proteins, including the folding of small proteins and influence of amino acid side chain substitutions (
      • Kollman P.A.
      • Massova I.
      • Reyes C.
      • Kuhn B.
      • Huo S.
      • Chong L.
      • Lee M.
      • Lee T.
      • Duan Y.
      • Wang W.
      • Donini O.
      • Cieplak P.
      • Srinivasan J.
      • Case D.A.
      • Cheatham 3rd, T.E.
      ,
      • Huo S.
      • Massova I.
      • Kollman P.A.
      ,
      • Grant B.J.
      • Gorfe A.A.
      • McCammon J.A.
      ,
      • Lee E.H.
      • Hsin J.
      • Sotomayor M.
      • Comellas G.
      • Schulten K.
      ,
      • Meli M.
      • Colombo G.
      ,
      • Klepeis J.L.
      • Lindorff-Larsen K.
      • Dror R.O.
      • Shaw D.E.
      ,
      • Steinbrecher T.
      • Labahn A.
      ). All simulations were completed using the AMBER modeling suite (
      • Case D.
      • Darden T.
      • Cheatham 3rd, T.
      • Simmerling C.
      • Wang J.
      • Duke R.
      • Luo R.
      • Merz K.
      • Pearlman D.
      • Crowley M.
      ) and the AMBER ff03 protein force field (
      • Duan Y.
      • Wu C.
      • Chowdhury S.
      • Lee M.C.
      • Xiong G.
      • Zhang W.
      • Yang R.
      • Cieplak P.
      • Luo R.
      • Lee T.
      • Caldwell J.
      • Wang J.
      • Kollman P.
      ) with explicit solvent and counter-ions using standard simulation protocols, including minimization and ∼40–90 ns of MD sampling with Ewald treatments of the electrostatics (
      • Essmann U.
      • Perera L.
      • Berkowitz M.L.
      • Darden T.
      • Lee H.
      • Pedersen L.G.
      ). For further detail, refer to the supplemental data. Representative plots of the root-mean-squared deviations to the initial structure of the coiled coil during the MD simulations can be found in supplemental Fig. S1. The MD trajectory results (effectively the time history of the atomic motions of the model structures at different intervals over the simulation) were analyzed with various tools. Quantification and comparison of the relative helical content were measured by calculating mean residue ellipticities at 222 nm (representative of helix content in circular dichroism (CD) spectra) of five individual 500-ps average structures spanning the final 5 ns of simulation of each coiled coil dimer using the DichroCalc program (
      • Bulheller B.M.
      • Hirst J.D.
      ). Structural helical content (or percent helicity defined as the number of residues in an α-helix divided by the total number of residues) was also calculated based on secondary structure as determined by peptide backbone Ψ and Φ torsions from the final 10 ns of simulation using the DSSP method (
      • Rost B.
      • Sander C.
      ) as implemented in UCSF Chimera (
      • Pettersen E.F.
      • Goddard T.D.
      • Huang C.C.
      • Couch G.S.
      • Greenblatt D.M.
      • Meng E.C.
      • Ferrin T.E.
      ). Intrahelical hydrogen bonds (i.e. carbonyl oxygen of residue “i” to the amine nitrogen of residue “i + 4”) were calculated over the final 10 ns of MD with distances less than 3.5 Å indicative of a hydrogen bond. To estimate the relative binding free energies of the coils in the dimers, MM-PBSA as implemented in AMBER (
      • Srinivasan J.
      • Cheatham T.E.
      • Cieplak P.
      • Kollman P.A.
      • Case D.A.
      ) and described by Gohlke and Case (
      • Gohlke H.
      • Case D.A.
      ) was applied to independent MD trajectories for the dimer and individual monomers. In addition to the postprocessing of MD results, calculations of the relative free energy of binding with respect to the wild-type dimer (ΔΔGbinding) were completed using more detailed thermodynamic integration methods for the CCmut2 dimers (
      • Gouda H.
      • Kuntz I.D.
      • Case D.A.
      • Kollman P.A.
      ). On the basis of a thermodynamic cycle (see supplemental Fig. S2), the relative free energy of binding can be calculated by “mutating” the original protein (λ = 0) to incorporate designed amino acid side chain point mutations (λ = 1) in both the dimer and monomeric states over different λ states in silico. Incorporation of the five amino acid mutations considered using this approach was accomplished stepwise (see Fig. 1B). Two steps were required to incorporate all five mutations to form the heterodimer mutant (CC-CCmut2), and an additional two steps were required to perturb the transition dimer into the homodimer (CCmut2-CCmut2). Similarly, two steps were used to incorporate the five mutations in the unbound monomer. Relatively long (6-ns equilibration, >6-ns accumulation) MD simulations were performed at each λ for the thermodynamic integration. Further technical details are provided in the supplemental data.

      Construction of Plasmids and Mutagenesis

      The gene encoding the coiled coil domain from Bcr-Abl (
      • Dixon A.S.
      • Kakar M.
      • Schneider K.M.
      • Constance J.E.
      • Paullin B.C.
      • Lim C.S.
      ) was amplified via PCR with the primers 5′-tgtaactcgagttatggtggacccggtg-3′ and 5′-atgctctcgagaccggtcatagctcttc-3′ and subcloned into the XhoI site of the plasmid pEGFP-PS, an optimized protein switch (PS) (
      • Kakar M.
      • Davis J.R.
      • Kern S.E.
      • Lim C.S.
      ,
      • Kakar M.
      • Cadwallader A.B.
      • Davis J.R.
      • Lim C.S.
      ) for use in the nuclear translocation assay (
      • Dixon A.S.
      • Lim C.S.
      ), and pEGFP-C1 (Clontech) for other experiments. These plasmids, named pPS-CC and pEGFP-CC, respectively, were then used as the template DNA for site-directed mutagenesis using the QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA) to generate five amino acid mutations (S41R, L45D, E48R, V49D, and Q60E). The primers used for the mutagenesis were 5′-ggagcgctgcaaggcccgcattcggcgcgacgagcagcgggacaaccaggagcgcttccgcatgatctacctggagacgttgctggccaagg-3′ and the reverse complement. Another site-directed mutagenesis was carried out to make C38A and D49V mutations with the primers 5′-caggagctggagcgcgccaaggcccgcattcg-3′ (and reverse complement) and 5′-gcgacgagcagcgggtgaaccaggagcgcttcc-3′ (and reverse complement), respectively. The final constructs were termed pEGFP-CCmut2 and pPS-CCmut2. The primers 5′-cgcaagggagctcccatcatcatcatcatcatcttgaagttctttttcaaggtcctatggtggacccggtgggcttc-3′ and 5′-agcatggatccctaccggtcatagctcttcttttccttggc-3′ were used to PCR amplify wild-type and mutant coiled coil domains, which were subcloned into pMAL-c2x (New England Biolabs, Ipswich, MA) at the SacI and BamHI sites to generate pMAL-H6-PP-CC and pMAL-H6-PP-CCmut2 used for protein expression. The primers 5′-tgtaactcgagttatggtggacccggtg-3′ and 5′-atgctctcgagccggtcatagctcttc-3′ were used to PCR amplify both coiled coil genes (wild type and mutant), and each was subsequently subcloned into the XhoI site of pDsRed2-N1 (Clontech) to make pDsRed-CC and pDsRed-CCmut2. Similarly, the primers 5′-tgtaaggcccagccggccatggtggacccggtg-3′ and 5′-cggggcgcggccgcccggtcatagctcttcttttc-3′ were used to insert the genes into the plasmid pEFVP16 (mammalian two-hybrid prey vector containing the VP16 activation domain; obtained from Dr. T. H. Rabbitts, Leeds Institute of Molecular Medicine, Leeds, UK) at the SfiI and NotI sites to generate pEFVP16-CC and pEFVP16-CCmut2. The primers 5′-tgtaagaattcatggtggacccggtg-3′ and 5′-atgctgaattcaccggtcatagctcttc-3′ were used to insert the coiled coils into the vector pM1 (mammalian two-hybrid bait vector containing the Gal4 binding domain; obtained from Dr. T. H. Rabbitts) at the EcoRI site to generate pM1-CC and pM1-CCmut2.

      Cell Lines and Transient Transfection

      Cells were maintained in a 5% CO2 incubator at 37 °C. K562 (a kind gift from Kojo Elenitoba Johnson, University of Michigan) and COS-7 cells (ATCC) were grown in RPMI 1640 medium (Invitrogen) supplemented with 10% FBS (HyClone Laboratories, Logan, UT), 1% penicillin-streptomycin (Invitrogen), 0.1% gentamicin (HyClone), and 1% l-glutamine (HyClone). K562 cells were passaged every 2 days and maintained between 0.1 and 1 × 106 cells/ml. Amaxa Nucleofector II (Lonza Group Ltd., Basel, Switzerland) was used to transfect 2 × 106 cells with 5–8 μg of DNA in solution V following the manufacturer's recommended protocol and nucleofection program (T-013). COS-7 cells were passaged every 2–3 days and transfected 24 h after seeding the cells using Lipofectamine LTX (Invitrogen) as recommended by the supplier.

      Protein Purification and CD

      Fusion proteins consisting of maltose-binding protein (MBP), a His tag, a PreScission protease site, and CC or CCmut2 were expressed in BL21star DE3 Escherichia coli cells (Invitrogen) from pMAL-H6-PP-CC or pMAL-H6-PP-CCmut2 plasmids. The proteins were purified over amylose resin (New England Biolabs) with gravity flow and cleaved with PreScission protease (a kind gift from Dr. Chris Hill, University of Utah), and MBP-H6 was removed by purification over HisPur Cobalt resin (Thermo Scientific, Waltham, MA). Reverse phase HPLC was performed as a final purification before lyophilizing the protein. CC and CCmut2 purified proteins were confirmed through mass spectroscopy (Mass Spectroscopy and Proteomics Core Facilities, University of Utah). CD experiments were performed on an Aviv 410 CD spectrometer (Aviv Biomedical Inc., Lakewood, NJ). Measurements from 190 to 300 nm (1-nm steps) were taken on 10 μm protein solutions in PBS (50 mm sodium phosphate, 150 mm NaCl, 0.5 mm DTT, pH 7.2) in a 1-mm-path length cuvette. 3-s equilibration times were allowed prior to each measurement, and the signal was averaged over 3 s. The average of three scans was used for each solution. Thermal denaturation was monitored in a 1-cm cuvette at 222 nm every 2 °C from 10 to 95 °C and back down to 10 °C in 10 °C steps. The protein concentrations used were 10 μm CC, 10 μm CCmut2, and 5 μm CC + 5 μm CCmut2. In addition, a mixing cell cuvette was used with 2.5 μm CC in one chamber and 7.5 μm CCmut2 in the other chamber with spectra acquired prior to and after mixing. After mixing and before data acquisition, the sample was incubated at 80 °C for 10 min and then allowed to re-equilibrate at 10 °C.

      Nuclear Translocation Assay

      The nuclear translocation assay (NTA) was performed as described previously (
      • Dixon A.S.
      • Lim C.S.
      ). Briefly, this assay uses a nuclear localization-inducible protein switch fused to a protein of interest and measures its ability to translocate a second protein into the nucleus. Here, the protein fused to the protein switch was either the coiled coil domain or mutant coiled coil domain, and its ability to translocate a cotransfected coiled coil domain or mutant coiled coil domain into the nucleus was measured. 24 h after transient transfection of COS-7 cells, 200 nm dexamethasone or an equal volume of ethanol (carrier) was added to the cells and incubated for 1 h. 0.5 μl of H33342 (nuclear dye) was added and incubated for another 15 min before imaging the cells with a fluorescence microscope as described previously (
      • Dixon A.S.
      • Kakar M.
      • Schneider K.M.
      • Constance J.E.
      • Paullin B.C.
      • Lim C.S.
      ). The percentage of protein localized in the nucleus was quantified with and without ligand to determine the nuclear increase after ligand induction.

      Mammalian Two-hybrid Assay

      The pM1-CC or pM1-CCmut2 (bait), pEFVP16-CC or pEFVP16-CCmut2 (prey), pG5-Fluc (reporter; obtained from Dr. T. H. Rabbitts), and pRL-CMV (for normalization; obtained from Dr. T. H. Rabbitts) plasmids were cotransfected in RPMI 1640 medium without antibiotics into 7.5 × 104 COS-7 cells in a white 96-well plate (Cellstar, Greiner Bio-One, Monroe, NC) in a 10:10:10:1 ratio. 24 h after transfection, the medium was replaced with complete RPMI 1640 medium, and 48 h after transfection, both firefly and Renilla luminescence was measured using the Dual-Glo Luciferase Assay (Promega) reagents according to the manufacturer's recommendations. The mean from duplicate samples was taken from five separate experiments. pAD-SV40 and pBD-p53 (Stratagene) plasmids were used for the positive control, and pM1 lacking the coiled coil gene was used as the negative control. A relative response ratio was calculated by first normalizing the individual firefly luminescence to the Renilla luminescence. The negative control was subtracted from the mean of the duplicate experimental values and scaled by dividing by the difference between the positive and negative controls ((Experiment − Ctrl)/(Ctrl+ − Ctrl)).

      Cell Proliferation and Western Blotting

      48 h following transfection of K562 cells with pEGFP-C1, pEGFP-CC, or pEGFP-CCmut2, trypan blue exclusion was used to determine cell proliferation (cell viability). 3 × 106 cells were pelleted and resuspended in 600 μl of lysis buffer (62.5 mm Tris-HCl, 2% (w/v) SDS, 10% glycerol, 50 mm DTT, 0.01% (w/v) bromphenol blue). Standard Western blotting procedures were followed using antibodies to detect the phosphorylated (p-) forms of Bcr-Abl, STAT5, CrkL, as well as β-actin as a loading control. The primary antibodies (anti-p-Abl(Tyr-245), 73E5, Cell Signaling Technology; anti-p-STAT5(Tyr-694), E208, Abcam; anti-p-CrkL(Tyr-207), 3181, Cell Signaling Technology; and anti-actin, mAbcam 8226, Abcam) were detected with anti-mouse (Ab6814, Abcam) or anti-rabbit (7074, Cell Signaling Technology) HRP-conjugated antibodies before the addition of ChemiGlo (AlphaInnotech, Cell Biosciences, Santa Clara, CA) chemiluminescent substrate and detection with a FluorChem FC2 imager (AlphaInnotech). The resulting bands were quantified through densitometry and normalized to the β-actin bands.

      Colony Forming Assay

      24 h following transfection of pEGFP-C1, pEGFP-CC, or pEGFP-CCmut2, K562 cells were resuspended in Iscove's modified Dulbecco's medium (Stem Cell Technologies, Vancouver, British Columbia, Canada) with 2% FBS, and 1,000 cells were seeded in Methocult H4230 methylcellulose medium (Stem Cell Technologies) in the absence of cytokines. Imatinib mesylate (a kind gift from Novartis) was added to 1,000 untransfected K562 cells seeded in Methocult at the time of seeding. Each group of treated cells was seeded into two separate plates. Colony formation was assessed 7 days after seeding cells by counting colonies in a 200-μm2 area of the plate and calculating the mean number of colonies per treatment. Experiments were replicated at least three times and compared with control (cells transfected with pEGFP-C1).

      Caspase-3/7 Assay

      48 h following transfection of pEGFP-C1, pEGFP-CC, or pEGFP-CCmut2, 3 × 106 K562 cells were pelleted and resuspended in 50 μl of lysis buffer provided in the EnzChek Caspase-3/7 Assay Kit 2 (Invitrogen). Cells were frozen at −80 °C and then centrifuged at 5,000 × g for 5 min. Lysates were transferred to a black 96-well plate (Cellstar, Greiner Bio-One), and 50 μl of 2× 7-amino-4-methylcoumarin-DEVD substrate was added and incubated at room temperature for 30 min. Fluorescence was measured with excitation at 342 nm and emission at 441 nm on a SpectraMax M2 plate reader (Molecular Devices, Sunnyvale, CA).

      Fluorescence Microscopy and DNA Segmentation

      48 h following transfection of pEGFP-C1, pEGFP-CC, or pEGFP-CCmut2, K562 cells were transferred to 2-well live cell chambers, 1 μl of the nuclear stain H33342 was added, and the cells were incubated at 37 °C for 15 min. Cells were then analyzed with an inverted fluorescence microscope (Olympus IX701F, Scientific Instrument Co., Sunnyvale, CA) with a high quality narrow band GFP filter (excitation, HQ480/20 nm; emission, HQ510/20 nm; beam splitter Q4951p; Chroma Technology Corp., Brattleboro, VT). Cells were photographed with an F-view Monochrome charge-coupled device camera using a 60× objective and were selected based on EGFP fluorescence. The nuclei of at least 50 transfected cells (EGFP fluorescence) per group were classified as either healthy (round or kidney-shaped) or segmented (punctate) (
      • Barrett K.L.
      • Willingham J.M.
      • Garvin A.J.
      • Willingham M.C.
      ,
      • Willingham M.C.
      ), and the percentage of cells with segmented DNA was calculated.

      RESULTS

      Computational Modeling

      Rational design, molecular modeling, MD, and free energy simulations were used to predict favorable attributes, specificity, and energetics to facilitate the choice of coiled coil modifications that stabilize binding of the mutant coiled coil domain with Bcr-Abl (CCmut2-CC) while destabilizing self-oligomerization (CCmut2-CCmut2). Initial simulations monitored increases in α-helicity as an indirect correlate for improved free energy of binding of mutant pairs (
      • Jelesarov I.
      • Bosshard H.R.
      ,
      • Hodges R.S.
      • Saund A.K.
      • Chong P.C.
      • St-Pierre S.A.
      • Reid R.E.
      ) and focused on mutations to potentially improve salt bridge interactions, increase stability through formation of a disulfide bond, improve helicity in the backbone through alanine mutations to position “f,” and destabilize mutations to improve specificity.
      Alanine mutations in the peptide backbone of the C-terminal coiled coil region were designed at residues Gln-33, Gln-47, Phe-54, and Thr-61 to increase the α-helicity (
      • Jelesarov I.
      • Bosshard H.R.
      ,
      • Hodges R.S.
      • Saund A.K.
      • Chong P.C.
      • St-Pierre S.A.
      • Reid R.E.
      ,
      • Zitzewitz J.A.
      • Ibarra-Molero B.
      • Fishel D.R.
      • Terry K.L.
      • Matthews C.R.
      ). Comparisons of the helicity as measured by circular dichroism and secondary structure between the homodimer (CCmut-CCmut) and heterodimer (CCmut-CC) suggested that this design actually decreases the helicity and shows poor specificity for the heterodimer over the homodimer (see Table 1, row 3, CD and secondary structure). Visualization of the coiled coil monomer structure suggests that the designed alanine mutations have disrupted local intrahelical hydrogen bonds, which may affect secondary structure and protein folding.
      TABLE 1Overview of comparative helicity calculations for various mutations of Bcr coiled coil domain
      Circular dichroismSecondary structureHydrogen bonds
      Mutations[Θ]222S.D.HelicityS.D.i, i + 4 hbS.D.
      %%
      (1) None (wild type)
          Homodimer−20,94385866.61.756.62.7
      (2) S41R,E48R,Q60E
          Heterodimer−21,58067270.02.056.92.8
          Homodimer−20,96167866.52.959.42.7
      (3) Q33A,Q47A,F54A,T61A
          Heterodimer−20,88447766.32.560.62.5
          Homodimer−20,78776065.12.256.43.2
      (4) E52C
          Heterodimer−20,52179667.82.255.92.7
      (5) S41R,L45D,E48R,Q60E
          Heterodimer−20,3301,32966.81.952.93.0
          Homodimer−19,39867966.22.056.52.4
      (6) S41R,L45D,E48R,V49D,Q60E (CCmut1)
          Heterodimer−19,69676567.82.056.92.7
          Homodimer−20,35267169.12.559.22.8
      (7) C38A,S41R,E48R,Q60E
          Heterodimer−21,4939070.02.461.52.5
          Homodimer−21,46326570.02.661.42.5
      (8) C38A,S41R,L45D,E48R,Q60E (CCmut2)
          Heterodimer−21,18669567.72.459.12.8
          Homodimer−17,96450759.32.650.72.4
      (9) C38A,S41R,L45D,E48R,V49D,Q60E
          Heterodimer−19,65145664.22.554.02.5
          Homodimer−16,77031264.21.850.43.3
      The helicity of simulated coiled coil dimers using the calculated circular dichroism of the peptide, the percentage of the residue that is defined as α-helical according to Ψ and Φ backbone dihedral torsions (using the DSSP method), and the percentage of α-helical specific hydrogen bonds (hb) (<3.5 Å) formed between i and i + 4 residues at picosecond intervals over the final 10 ns of MD relative to the total number of potential interactions (the total number of residues minus 4) are shown. Row 6 contains the only set of mutations tested that indicated a more favorable homodimer than the heterodimer and was termed CCmut1 and used in subsequent MM-PBSA experiments as a negative control. Row 8 contains the set of mutations found to exhibit the optimal reduced homo-oligomerization paired with improved hetero-oligomerization and was termed CCmut2 for subsequent MM-PBSA and thermodynamic integration studies as well as in vitro experiments. The bold values distinguish the experimental values from the standard deviations. The italicized mutations are CCmut1 and CCmut2, which were modeled further and then experimentally tested.
      Cys-38 exists as an unbound free thiol in the native protein, and its close proximity to position 52 might allow the formation of a disulfide bond that could further stabilize the structure. An engineered disulfide was modeled by incorporating a cysteine residue at position 52. Visualization of the heterodimer and analysis of the structure helicity (see Table 1, row 4, CD) suggests that the geometry of the disulfide is not ideal and introduces structural disturbances.
      Three point mutations were designed in the wild-type monomer to improve binding to the oncoprotein: S41R, E48R, and Q60E. When these three designed point mutations were modeled, helicity of the homodimer and heterodimer mutants exceeded the wild-type dimer (see Table 1, compare row 1 with row 2), suggesting improved binding due to more favorable electrostatic interactions. The improved helicity of the mutant homodimer over the Bcr-Abl wild-type dimer, however, may result in decreased specificity for the heterodimer form. Destabilizing point mutations to improve the heterodimer specificity included designed aspartate mutations in the hydrophobic core (L45D and L49D) and were evaluated as both single and double mutations. Both mutations together decreased the secondary structure (and stability) of the heterodimer (see Table 1, rows 6 and 9), whereas L45D maintained heterodimer specificity (see Table 1, row 5). The final mutation converted an exposed thiol group (Cys-38) to alanine and, in conjunction with the four prior mutations, stabilized the heterodimer over both the mutant homodimer and the wild-type dimer (see Table 1, compare row 1 with row 8). Thus, the optimal mutant coiled coil domain (termed CCmut2) contains C38A, S41R, L45D, E48R, and Q60E mutations. Specificity as measured by helicity suggests that the mutant homodimer (CCmut2-CCmut2) is significantly less stable than the wild type (CC-CC) and the heterodimer (CC-CCmut2).
      MM-PBSA and thermodynamic integration energy calculation methodologies (the latter are more quantitative and accurate) were subsequently used to analyze the two sets of mutations. Both approaches validated the previous indications suggesting that CCmut2 exhibits improved heterodimer stability compared with the wild-type oligomerization (see Table 2). Furthermore, CCmut2 was shown to have reduced homo-oligomerization. Together, MM-PBSA and thermodynamic integration energy calculations suggest improved binding and specificity of CCmut2.
      TABLE 2MM-PBSA and thermodynamic relative free energy of binding results (in kcal/mol)
      MM-PBSAThermodynamic integration
      DimerΔΔGbindingS.E.ΔΔGbindingS.D.
      kcal/molkcal/mol
      CC-CC0.03.20.00.0
      CC-CCmut111.73.0NDND
      CCmut1-CCmut125.50.1NDND
      CC-CCmut2−1.33.1−1.10.5
      CCmut2-CCmut211.43.03.20.8
      MM-PBSA results were found by subtracting the absolute free energies of the unbound monomers from the calculated free energy of coiled coil dimer using separate MD trajectories. Thermodynamic integration calculations followed the scheme described in supplemental Fig. S2 using intermediate coiled coil dimers to build a consistent transition from the wild-type coiled coil dimer to the CCmut2 dimers. Results from both calculations are reported relative to the wild-type coiled coil dimer (CC-CC). ND, not determined.

      In Vitro Experiments Validate Design

      To substantiate the computational circular dichroism calculations, wavelength scans were performed on protein solutions of CCmut2, CC, and a mixture of CCmut2 and CC. As seen in Fig. 2A, all three protein solutions produced the typical pattern characteristic of α-helices with similar helicity. Given the relatively small number of mutations that distinguish CCmut2 from wild-type CC and because the mutations are designed to alter oligomerization while retaining helicity, it is reasonable to expect no major differences in their experimentally measured helicity. However, the modifications do clearly destabilize the mutant homodimer as made apparent in the thermal denaturation of the proteins (Fig. 2B). Although CC demonstrated a melting temperature (Tm) consistent with previous reports of 83 °C (
      • Zhao X.
      • Ghaffari S.
      • Lodish H.
      • Malashkevich V.N.
      • Kim P.S.
      ), a large decrease in Tm was found for the CCmut2 (Tm = 63 °C) (Fig. 2B), confirming the decreased stability of the mutant homo-oligomer. When CCmut2 was mixed with CC in a 1:1 ratio, two distinct melting transitions were evident (Fig. 2B, black triangles); however, a third Tm was not readily apparent. This suggests that the CC-CCmut2 heterodimer is either isoenergetic with one of the two homodimers or that it simply did not form. To further assess the formation of the heterodimer, a 3:1 (CCmut2-CC) ratio was studied in a mixing cell cuvette (Fig. 2C) both before and after mixing. The mixing shifted the curve most predominantly at lower temperatures and only slightly at higher temperatures, suggesting the formation of a new species (as seen in Fig. 2C). The shift in the curve at lower temperatures can be accounted for by the decreased CCmut2 concentration due to formation of heterodimers. The small difference in the curves at higher temperatures suggests the formation of hetero-oligomers that are nearly isoenergetic or slightly less stable than the CC homo-oligomers. Consistent with the computational modeling, the primary improvement made through these mutations is in the specificity granted by a less stable mutant homodimer while retaining the ability to oligomerize with wild type.
      Figure thumbnail gr2
      FIGURE 2Circular dichroism wavelength scans and thermal denaturation. A, α-helices exhibit a characteristic double absorption minimum at ≈208 and 222 nm. Gray circles, CC; black squares, CCmut2; black triangles, mixture (Mix). The lines represent the average from three scans. B, thermal denaturation curves in a 1-cm cuvette. The ratio of CC to CCmut2 used in the mixture was 1:1. Gray circles, CC; black squares, CCmut2; black triangles, mixture. C, thermal denaturation curves in a mixing cell cuvette using 2.5 μm CC and 7.5 μm CCmut2. Gray circles, separate (premixing); black squares, mixture. deg, degrees.
      The optimally designed mutant coiled coil was created through site-directed mutagenesis of a plasmid encoding the Bcr-Abl coiled coil for cell-based in vitro experiments and further validation. First, the NTA (
      • Dixon A.S.
      • Lim C.S.
      ) was used to study the interaction (Fig. 3A). This assay measures the ability of a nuclear localization-inducible PS fused to one form of the coiled coil domain to alter the nuclear localization of another form of the coiled coil domain. Essentially, the interaction between the coiled coil domains is indicated by an increase in fluorescence in the nucleus. The high level of nuclear translocation resulting from the mutant coiled coil domain (Fig. 3A, middle column) is likely to stem from both the reduced homo-oligomerization and the improved binding to the wild-type coiled coil domain. To specifically address whether the designed mutations limited the homo-oligomerization, the interaction between two mutant coiled coil domains (CCmut2-CCmut2) was assayed and found to be indistinguishable from the negative control (not included in the graph). These same interactions were also studied in a mammalian two-hybrid assay to further validate the results. Similar to the NTA results, the greatest binding was found between the mutant and wild-type coiled coil domains (CCmut2-CC; Fig. 3B, second column), and the homo-oligomerization between two mutants (CCmut2-CCmut2; Fig. 3B, third column) was not statistically distinguishable from the negative control (not included in the graph). Together, the NTA and two-hybrid results indicate that the proposed mutations reduced the homo-oligomerization of the mutant coiled coil domain and improved the binding to the wild-type coiled coil domain.
      Figure thumbnail gr3
      FIGURE 3Binding of homo- and heterodimers tested through NTA and mammalian two-hybrid assay. A, figure was modified from Dixon and Lim (
      • Dixon A.S.
      • Lim C.S.
      ). The NTA measures the ability of a nuclear localization-inducible PS fused to a protein of interest to cause a second interacting protein to be translocated into the nucleus and is monitored through fluorescence microscopy (fused to EGFP and DsRed, respectively). The assay was performed in COS-7 cells that do not contain Bcr-Abl. Each experiment was performed in triplicate with at least eight cells analyzed per experiment. Statistical significance was determined using one-way ANOVA with Tukey's post-test. *, p < 0.01; **, p < 0.001 compared with control (pDsRed2-N1/EGFP-PS-CC; not included in the graph). B, mammalian two-hybrid assay in COS-7 cells. Statistical significance was determined using one-way ANOVA with Tukey's post-test. *, p < 0.01 compared with CC-CC interaction (n = 5). Error bars represent ± S.D.
      Next, the oligomeric disruption of Bcr-Abl was indirectly measured through assaying the activity of Bcr-Abl. As the oligomeric state of Bcr-Abl is correlated to its activity, if the oligomerization is disrupted it will cause a reduction in Bcr-Abl activity (decrease in phosphorylation). After transfecting either the wild-type or mutant coiled coil domain into K562 cells, Western blotting with an antibody that specifically recognizes the phosphorylated form of Bcr-Abl was used to determine the activity. Under identical experimental conditions, the wild-type coiled coil domain had minimal effect on the level of phosphorylation of Bcr-Abl (Fig. 4A, third column), whereas the mutant coiled coil domain reduced the phosphorylation level to 35% (Fig. 4A, last column). Furthermore, the phosphorylation of two proteins known to be phosphorylated by Bcr-Abl, STAT5 and CrkL, was also tested. Again, the wild-type coiled coil domain had minimal effect, whereas the mutant coiled coil domain reduced the phosphorylation of both proteins (Fig. 4B, third and fourth columns). The decreased phosphorylation of Bcr-Abl suggests that the mutant coiled coil domain is capable of interacting with the endogenous Bcr-Abl and not just the isolated coiled coil domain as used in the previous NTA and two-hybrid experiments. Moreover, the decreased level of phosphorylation provides insight into the oligomeric disruption and inhibitory potential of the mutant coiled coil domain.
      Figure thumbnail gr4
      FIGURE 4Representative images of Western blots to detect phosphorylated form of Bcr-Abl (A) and two substrates of Bcr-Abl, STAT5 and CrkL (B). The phosphorylation of Bcr-Abl is indicative of the tyrosine kinase activity and is shown to be decreased by the addition of CCmut2 (percentage of p-Bcr-Abl from untreated K562 cells is indicated graphically). The proteins STAT5 and CrkL are also phosphorylated when Bcr-Abl is active and are secondary indicators of the Bcr-Abl activity. Western blotting followed by densitometry was replicated three times on lysates from three separate transfections. The level of p-Bcr-Abl, as a percentage of the untreated cells, is shown graphically in A, and the level of p-STAT5 and p-CrkL (±S.D.) is indicated above the representative images. Statistical significance was determined using one-way ANOVA with Tukey's post-test (n = 3). *, p < 0.05; **, p < 0.01 compared with cells transfected with EGFP. Error bars represent ± S.D.
      Inhibition of Bcr-Abl through oligomeric disruption (as with inhibition through tyrosine kinase inhibitors) should relieve the up-regulation of signaling pathways resulting in misregulated cell proliferation. The effect of the coiled coil domains on cell proliferation was measured through cell counts with trypan blue exclusion. Although the wild-type coiled coil domain demonstrated a slight effect on the number of proliferating cells, the mutant coiled coil domain was most effective at decreasing the number of proliferating cells (Fig. 5A). Furthermore, the effect on proliferation was measured via a colony forming assay, and again the mutant coiled coil domain was found to cause the greatest reduction in cell proliferation (Fig. 5B, column 2) similar to that seen with imatinib (Fig. 5B, column 3).
      Figure thumbnail gr5
      FIGURE 5Inhibition of Bcr-Abl through expression of CCmut2 results in decreased proliferation of K562 cells and activation of apoptosis. A, proliferation of K562 cells as determined by cell counts with trypan blue exclusion. B, proliferation of K562 cells as determined by colony forming assays. IM, imatinib mesylate. C, induction of apoptosis as measured through activation of caspase-3/7. For A–C, statistical significance was determined using one-way ANOVA with Tukey's post-test. *, p < 0.01; **, p < 0.001 compared with control (cells transfected with pEGFP-C1). Error bars represent ± S.D.
      As CML cells become dependent on the signaling pathways up-regulated by Bcr-Abl, the inhibition of Bcr-Abl and reprieve of that signaling should also induce apoptosis. As one indication of the ability of the coiled coil constructs to induce apoptosis, the activity of caspase-3/7 was measured in a fluorometric assay. In a trend similar to that found in all previous experiments, the mutant coiled coil domain again produced the greatest result and was the only construct able to induce the activation of caspase activity at a statistically significant level (Fig. 5C, third column). As expected, similar experiments with CCmut2 in cells that do not express Bcr-Abl (1471.1 and COS-7) did not show an increase in caspase activity (data not shown).
      Finally, as a measure of late stage apoptosis, DNA segmentation (of K562 nuclei) was measured (
      • Barrett K.L.
      • Willingham J.M.
      • Garvin A.J.
      • Willingham M.C.
      ,
      • Willingham M.C.
      ). Cells transfected with CCmut2 revealed segmented nuclei, a hallmark of apoptosis as shown in Fig. 6 (column 3, arrows). CC- and control (EGFP)-transfected cells had healthy (round) nuclei. The percentage of CCmut2-transfected cells demonstrating apoptosis was 29.4% compared with 4.29% for CC and 0.75% for EGFP control. Additionally, phase-contrast images (Fig. 6, column 1) demonstrated morphological changes indicative of apoptosis, including zeotic blebbing, cell shrinkage, and cell fragmentation (
      • Dixon A.S.
      • Kakar M.
      • Schneider K.M.
      • Constance J.E.
      • Paullin B.C.
      • Lim C.S.
      ,
      • Willingham M.C.
      ). Finally, in a control cell line (1471.1 cells), minimal DNA segmentation (less than 2%) was observed after EGFP, CC, or CCmut2 was transfected (data not shown). The inhibition of cell proliferation and the induction of apoptosis illustrate the therapeutic potential of oligomeric disruption through this modified coiled coil.
      Figure thumbnail gr6
      FIGURE 6Fluorescence microscopy for morphological evaluation of nuclei. Cells transfected with CCmut2 (bottom row) appear shrunken (column 1, arrows) or zeotic (column 1, arrowheads) and exhibit segmented (punctate) nuclei (column 3, arrows), hallmarks of late apoptotic cells. The far right column summarizes the percentage of transfected cells determined to have segmented DNA. Statistical significance was determined using one-way ANOVA with Tukey's post-test. *, p < 0.001 compared with control (cells transfected with pEGFP-C1).

      DISCUSSION

      Rational design, molecular modeling, MD simulation, and free energy analysis identified modifications to the Bcr-Abl coiled coil to improve interaction with Bcr-Abl while also reducing mutant homodimer (CCmut2-CCmut2) interactions. The optimal set of mutations served as the lead, reducing the need to test an overwhelming number of possible mutations or combination of mutations in vitro. The in vitro experiments performed with this construct confirmed the computational results and demonstrated that this designed mutant coiled coil has an enhanced capability to oligomerize with Bcr-Abl. The design incorporated charge-charge repulsions between two mutant coiled coil domains to reduce the homo-oligomerization, thereby making the mutant more available for interaction with the target, Bcr-Abl. Residues with the potential to form additional favorable electrostatic interactions with Bcr-Abl were also introduced to increase the binding affinity between the mutant and Bcr-Abl. Although further structural characterization could confirm that the hypothesized interactions are indeed occurring, both the computational modeling and in vitro experiments strongly indicate that the modifications to the coiled coil domain lead to a more specific, better binding coiled coil partner for Bcr-Abl.
      Although an isolated coiled coil domain from Bcr-Abl should in principal oligomerize with Bcr-Abl, the isolated coiled coil domain also has the ability to form homo-oligomers. Given that an isolated coiled coil domain is smaller, there is likely less entropic penalty for formation, and therefore it should be less effective as a therapeutic due to the decreased effective concentration of the monomer and need for dissociation of the dimer for activity. Our approach to a more potent therapeutic involved the design of a coiled coil domain with reduced ability to self-oligomerize while also exhibiting enhanced oligomerization with the target. This goal was confirmed through both computational modeling and in vitro experiments. Alternative treatments for CML are still needed because of the inability to eliminate CML stem cells, resistance to small molecule inhibitors, and ineffective treatment of advanced stages of the disease (
      • Martin M.G.
      • Dipersio J.F.
      • Uy G.L.
      ,
      • Santos F.P.
      • Ravandi F.
      ,
      • Kantarjian H.M.
      • Cortes J.
      • La Rosée P.
      • Hochhaus A.
      ,
      • Jamieson C.H.
      ,
      • Janssen J.J.
      • Schuurhuis G.J.
      • Terwijn M.
      • Ossenkoppele G.J.
      ). Given the importance the coiled coil domain has in the regulation of Bcr-Abl activity, it has long been hypothesized that this domain could be used therapeutically. This modified Bcr-Abl coiled coil domain has a heightened ability to inhibit the oncogenicity of Bcr-Abl and warrants further exploration as an alternative approach to treat CML.

      Acknowledgments

      We acknowledge the use of the DNA/Peptide Core (NCI Cancer Center Support Grant P30 CA042014, Huntsman Cancer Institute) and computer time from the Center for High Performance Computing at the University of Utah. We thank Dr. Debbie Eckert for technical assistance in protein purification and Jonathan Constance, Rian Davis, Mohanad Mossalam, Matthew Weinstock, and Dr. Michael Kay for scientific discussions.

      Supplementary Material

      REFERENCES

        • Taylor C.M.
        • Keating A.E.
        Biochemistry. 2005; 44: 16246-16256
        • Strauss H.M.
        • Keller S.
        Handb. Exp. Pharmacol. 2008; : 461-482
        • Matthews T.
        • Salgo M.
        • Greenberg M.
        • Chung J.
        • DeMasi R.
        • Bolognesi D.
        Nat. Rev. Drug Discov. 2004; 3: 215-225
        • Dwyer J.J.
        • Wilson K.L.
        • Davison D.K.
        • Freel S.A.
        • Seedorff J.E.
        • Wring S.A.
        • Tvermoes N.A.
        • Matthews T.J.
        • Greenberg M.L.
        • Delmedico M.K.
        Proc. Natl. Acad. Sci. U.S.A. 2007; 104: 12772-12777
        • Otaka A.
        • Nakamura M.
        • Nameki D.
        • Kodama E.
        • Uchiyama S.
        • Nakamura S.
        • Nakano H.
        • Tamamura H.
        • Kobayashi Y.
        • Matsuoka M.
        • Fujii N.
        Angew. Chem. Int. Ed. Engl. 2002; 41: 2937-2940
        • Yan Z.
        • Tripet B.
        • Hodges R.S.
        J. Struct. Biol. 2006; 155: 162-175
        • Eckert D.M.
        • Malashkevich V.N.
        • Hong L.H.
        • Carr P.A.
        • Kim P.S.
        Cell. 1999; 99: 103-115
        • Sia S.K.
        • Carr P.A.
        • Cochran A.G.
        • Malashkevich V.N.
        • Kim P.S.
        Proc. Natl. Acad. Sci. U.S.A. 2002; 99: 14664-14669
        • Parry D.A.
        • Fraser R.D.
        • Squire J.M.
        J. Struct. Biol. 2008; 163: 258-269
        • Kammerer R.A.
        • Jaravine V.A.
        • Frank S.
        • Schulthess T.
        • Landwehr R.
        • Lustig A.
        • Garcia-Echeverria C.
        • Alexandrescu A.T.
        • Engel J.
        • Steinmetz M.O.
        J. Biol. Chem. 2001; 276: 13685-13688
        • Spek E.J.
        • Bui A.H.
        • Lu M.
        • Kallenbach N.R.
        Protein Sci. 1998; 7: 2431-2437
        • Marqusee S.
        • Baldwin R.L.
        Proc. Natl. Acad. Sci. U.S.A. 1987; 84: 8898-8902
        • Ryan S.J.
        • Kennan A.J.
        J. Am. Chem. Soc. 2007; 129: 10255-10260
        • Gurnon D.G.
        • Whitaker J.A.
        • Oakley M.G.
        J. Am. Chem. Soc. 2003; 125: 7518-7519
        • Zhang X.
        • Subrahmanyam R.
        • Wong R.
        • Gross A.W.
        • Ren R.
        Mol. Cell. Biol. 2001; 21: 840-853
        • Arlinghaus R.B.
        Oncogene. 2002; 21: 8560-8567
        • He Y.
        • Wertheim J.A.
        • Xu L.
        • Miller J.P.
        • Karnell F.G.
        • Choi J.K.
        • Ren R.
        • Pear W.S.
        Blood. 2002; 99: 2957-2968
        • Mian A.A.
        • Oancea C.
        • Zhao Z.
        • Ottmann O.G.
        • Ruthardt M.
        Leukemia. 2009; 23: 2242-2247
        • Beissert T.
        • Hundertmark A.
        • Kaburova V.
        • Travaglini L.
        • Mian A.A.
        • Nervi C.
        • Ruthardt M.
        Int. J. Cancer. 2008; 122: 2744-2752
        • Beissert T.
        • Puccetti E.
        • Bianchini A.
        • Güller S.
        • Boehrer S.
        • Hoelzer D.
        • Ottmann O.G.
        • Nervi C.
        • Ruthardt M.
        Blood. 2003; 102: 2985-2993
        • Huang S.F.
        • Liu D.B.
        • Zeng J.M.
        • Xiao Q.
        • Luo M.
        • Zhang W.P.
        • Tao K.
        • Wen J.P.
        • Huang Z.G.
        • Feng W.L.
        Protein Expr. Purif. 2009; 64: 167-178
        • Guo X.Y.
        • Cuillerot J.M.
        • Wang T.
        • Wu Y.
        • Arlinghaus R.
        • Claxton D.
        • Bachier C.
        • Greenberger J.
        • Colombowala I.
        • Deisseroth A.B.
        Oncogene. 1998; 17: 825-833
        • Zhao X.
        • Ghaffari S.
        • Lodish H.
        • Malashkevich V.N.
        • Kim P.S.
        Nat. Struct. Biol. 2002; 9: 117-120
        • Crick F.H.
        Acta Crystallogr. 1953; 6: 689-697
        • Martin M.G.
        • Dipersio J.F.
        • Uy G.L.
        Leuk. Lymphoma. 2009; 50: 14-23
        • Guex N.
        • Peitsch M.C.
        Electrophoresis. 1997; 18: 2714-2723
        • Case D.
        • Darden T.
        • Cheatham 3rd, T.
        • Simmerling C.
        • Wang J.
        • Duke R.
        • Luo R.
        • Merz K.
        • Pearlman D.
        • Crowley M.
        AMBER. University of California, San Francisco2006
        • Kollman P.A.
        • Massova I.
        • Reyes C.
        • Kuhn B.
        • Huo S.
        • Chong L.
        • Lee M.
        • Lee T.
        • Duan Y.
        • Wang W.
        • Donini O.
        • Cieplak P.
        • Srinivasan J.
        • Case D.A.
        • Cheatham 3rd, T.E.
        Acc. Chem. Res. 2000; 33: 889-897
        • Huo S.
        • Massova I.
        • Kollman P.A.
        J. Comput. Chem. 2002; 23: 15-27
        • Grant B.J.
        • Gorfe A.A.
        • McCammon J.A.
        Curr. Opin. Struct. Biol. 2010; 20: 142-147
        • Lee E.H.
        • Hsin J.
        • Sotomayor M.
        • Comellas G.
        • Schulten K.
        Structure. 2009; 17: 1295-1306
        • Meli M.
        • Colombo G.
        Methods Mol. Biol. 2009; 570: 77-153
        • Klepeis J.L.
        • Lindorff-Larsen K.
        • Dror R.O.
        • Shaw D.E.
        Curr. Opin. Struct. Biol. 2009; 19: 120-127
        • Steinbrecher T.
        • Labahn A.
        Curr. Med. Chem. 2010; 17: 767-785
        • Duan Y.
        • Wu C.
        • Chowdhury S.
        • Lee M.C.
        • Xiong G.
        • Zhang W.
        • Yang R.
        • Cieplak P.
        • Luo R.
        • Lee T.
        • Caldwell J.
        • Wang J.
        • Kollman P.
        J. Comput. Chem. 2003; 24: 1999-2012
        • Essmann U.
        • Perera L.
        • Berkowitz M.L.
        • Darden T.
        • Lee H.
        • Pedersen L.G.
        J. Chem. Phys. 1995; 103: 8577-8593
        • Bulheller B.M.
        • Hirst J.D.
        Bioinformatics. 2009; 25: 539-540
        • Rost B.
        • Sander C.
        J. Mol. Biol. 1993; 232: 584-599
        • Pettersen E.F.
        • Goddard T.D.
        • Huang C.C.
        • Couch G.S.
        • Greenblatt D.M.
        • Meng E.C.
        • Ferrin T.E.
        J. Comput. Chem. 2004; 25: 1605-1612
        • Srinivasan J.
        • Cheatham T.E.
        • Cieplak P.
        • Kollman P.A.
        • Case D.A.
        J. Am. Chem. Soc. 1998; 120: 9401-9409
        • Gohlke H.
        • Case D.A.
        J. Comput Chem. 2004; 25: 238-250
        • Gouda H.
        • Kuntz I.D.
        • Case D.A.
        • Kollman P.A.
        Biopolymers. 2003; 68: 16-34
        • Dixon A.S.
        • Kakar M.
        • Schneider K.M.
        • Constance J.E.
        • Paullin B.C.
        • Lim C.S.
        J. Control Release. 2009; 140: 245-249
        • Kakar M.
        • Davis J.R.
        • Kern S.E.
        • Lim C.S.
        J. Control Release. 2007; 120: 220-232
        • Kakar M.
        • Cadwallader A.B.
        • Davis J.R.
        • Lim C.S.
        Pharm. Res. 2007; 24: 2146-2155
        • Dixon A.S.
        • Lim C.S.
        BioTechniques. 2010; 49: 519-524
        • Barrett K.L.
        • Willingham J.M.
        • Garvin A.J.
        • Willingham M.C.
        J. Histochem. Cytochem. 2001; 49: 821-832
        • Willingham M.C.
        J. Histochem. Cytochem. 1999; 47: 1101-1110
        • Jelesarov I.
        • Bosshard H.R.
        J. Mol. Biol. 1996; 263: 344-358
        • Hodges R.S.
        • Saund A.K.
        • Chong P.C.
        • St-Pierre S.A.
        • Reid R.E.
        J. Biol. Chem. 1981; 256: 1214-1224
        • Zitzewitz J.A.
        • Ibarra-Molero B.
        • Fishel D.R.
        • Terry K.L.
        • Matthews C.R.
        J. Mol. Biol. 2000; 296: 1105-1116
        • Santos F.P.
        • Ravandi F.
        Leuk. Lymphoma. 2009; 50: 16-26
        • Kantarjian H.M.
        • Cortes J.
        • La Rosée P.
        • Hochhaus A.
        Cancer. 2010; 116: 1419-1430
        • Jamieson C.H.
        Hematology Am. Soc. Hematol. Educ. Program. 2008; : 436-442
        • Janssen J.J.
        • Schuurhuis G.J.
        • Terwijn M.
        • Ossenkoppele G.J.
        Curr. Stem Cell Res. Ther. 2009; 4: 224-236