|Tumor differentiation factor; Protein-protein interactions; Molecular modeling
|Tumor differentiation factor (TDF) is a recently identified protein, produced in the brain. Work in our lab identified TDF in the brain, specifically in the pituitary, but also in other regions. TDF immunostaining specifically co-localized with markers specific for neurons, but not with markers specific to astrocytes .
|The TDF protein produced by the pituitary is likely secreted in the blood stream and targets breast and prostate tissue . In vitro work using proteomics suggested that the TDF receptor candidates are members of the heat shock family, specifically heat shock protein 70 (HSP70) and heat shock protein 90 (HSP90) [3-5].
|Proteomics allows identification and characterization of proteins in large scale [6-10] and TDF receptor candidates (and TDF ligand) were identified by affinity chromatography followed by mass spectrometry [3-5,11]. Our earlier studies also indicated that TDF may have an inducible receptor, composed of HSP70 and HSP90 [3-5]. Furthermore, we also suggested that TDF signaling depends on the interaction of HSP70 and HSP90 proteins with TDF . However, the function of TDF is still not understood and many of these studies still need additional theoretical and experimental confirmation. In addition, the TDF crystal structure has not been established, thus making us rely mostly on structural biology-based work . Therefore, a better understanding of TDF requires additional studies. To further understand the function of TDF, we employed structural biology to investigate the interaction of HSP70, HSP90 and TDF. Here we provide additional insights about the possible interaction between TDF and its potential HSP70 and HSP90 receptors.
|For protein interaction and docking experiments, we have taken the homology model of 3C7NB, a member of the HSP70 as a ligand protein  and HSP90-beta as receptor protein. The 3C7NB is the open, weakly ADP bound form . For HSP90, we have selected the homology model based on gi20149594 and template crystal structure 2IOQ Chain B . This model receptor protein is developed using Swiss model [16,17]. 2IOQ is an open form of HSP90.
|Protein-protein docking was carried out using GRAMM-X Docking Web Server v.1.2.0 [18,19] and verified by Patch dock and Fire dock servers [20-23]. Descriptions of these docking experiments are described elsewhere [3-5]. For the TDF-HSP90 interaction, we used the same homology model of HSP90 (2IOQ.pdb) as receptor protein , but used the model TDF structure as a ligand protein [11,12]. This model TDF protein was developed using I-Tasser server [24,25]. In our previous experiments, we have used TDF-P1 as a ligand .
|To further substantiate we have performed another set of experiment using HSP90 model receptor based on template structure 2CG9B.PDB (Chain B) . That part could be found in the supplemental materials (Figures S1-S6).
|Results and Discussion
|HSP70 proteins may interact with the proteins from the HSP90 family. Both proteins were also experimentally determined that interact with TDF. To investigate the details on TDF-HSP70-HSP90 interaction, we conducted docking experiments involving HSP70 and HSP90 proteins, and TDF-HSP90 proteins. HSP90 is composed of N Terminal Domain (NTD), Middle Domain (MD) and C Terminal Domain (CTD). HSP90 forms a dimer at the C-terminal domain . Figure 1A describes the structure of the model receptor protein (HSP90, based on 2IOQ.pdb) colored from N-terminal to C-terminal. Three receptor cavities for potential ligand-binding are also labeled by numbers in this figure. Figure 1B displayed the hydrophobic surface of the model receptor protein. From Figure 1B it is clear that the ligand binding pockets are somewhat hydrophobic in nature. Figure 1C described the structural aspects of model ligand protein. The structure of the model ligand protein is based on HSP70 (HSPA8, 3C7NB.pdb) (Figure 1C). The structure of the model HSP70 is composed of three parts N-terminal nucleotide binding domain (NBD), substrate binding domain (SBD), and a linker that connects the SBD and NBD.
|Figures 2 and 3 display the two tentatively identified docking sites (cavity 3 and cavity 1 of Figure 1A) of HSP70 ligand on the HSP90 model receptor model. Interfacial residues are also displayed in Figures 2C-2D and 3C-3D. These docking sites are identified using GRAMM-X docking server. One additional ligand protein binding pocket (cavity 2 of Figure 1A) as displayed in Figure 4 is identified using Patch dock and Fire dock server. Figure 4C and 4D are the residues of the receptor and ligand protein at the interfacial region. In Figure 2, the NBD of the ligand protein is docked onto the receptor cavity 3. In Figure 3, the SBD of the ligand protein is docked onto the receptor cavity 1. In Figure 4, the SBD is docked onto the receptor protein cavity 2. From the above computational study we can see that the SBD of HSP70 is most frequently bound to HSP90. An additional feature is observed while identifying the tentative docking site using Patch dock and Fire dock. In this particular instance, the tentative docking site of the ligand protein is traversed between the receptor protein cavity 2 and 3 of Figure 1A. When docking by GRAMM-X, the ligand protein has a higher tendency to dock in cavity 3 than cavity 1 of model receptor based on 2IOQ. In case of patch dock and fire dock web-server however, the ligand protein is mostly docked at cavity 2 and 3 than cavity 1. All these figures are made using Accelrys Discovery Studio 3.5 . Overall, we have found three tentative docking sites for interaction between HSP70 and HSP90 and that From Figure 1B, we can assume that these pockets are somewhat hydrophobic in nature.
|Figure 5 displays the predicted protein-protein interaction network for HSP70 kDa protein 8 (HSPA8). These analyses are performed using the The Biological General Repository for Interaction Datasets (BioGRID) [29,30] and the graphic views are generated using the web based BioGRID graphical network viewer [http://thebiogrid.org/]. The numbers of total interactions are 414 (Figure 5A). If “interactor interactions” are included then the total numbers of interactions for HSPA8 are 7160 (Figure 5B).
|We also investigated the interaction between TDF ligand protein [11,12] and HSP90 receptor based on template structure 2IOQ. pdb . In this case, when docking by GRAMM-X, the ligand TD protein has a higher tendency to dock in cavity 3 than cavity 1 of model receptor. And in this case the TDF protein does not seem to dock strongly at cavity 2 of the model receptor. These docking site using GRAMM-X webserver are displayed in Figure 6A. Interfacial residues of TDF structure are displayed in Figures 6B. In case of patch dock and fire dock web-server the ligand TDF protein however mostly docked at cavity 3 than cavity 1 and 2. Figures 6A and 6B were made using Accelrys Discovery Studio 3.5 .
|The protein-protein interaction network (physical and genetic) for HSP90 protein is shown in Figure 7. The numbers of total predicted interactions are 906 (Figure 7A). These calculations are based on interspecies interactions and self-interaction. If “interactor interactions” are included (Figure 7B) then the total numbers of interactions are 8456 [29,30].
|Overall, the current data suggest that TDF may indeed interact, in addition to HSP70, with HSP90. As we demonstrated previously, TDF may activate a pathway that is specific to breast and prostate cancer cells but not sensitive to other cancer or normal fibroblast or fibroblast-like cells. Our previous report also suggested that TDF-R may be a multisubunit inducible receptor, composed of HSP70 and HSP90, and that TDF signaling depends on the interaction of these proteins. Thus TDF interacts with its receptors and induces cell differentiation through a unique, non-steroid mechanism [3-5,12].
|From the above structural study we can speculate that TDF protein has a higher tendency to bind to the MD of HSP90 receptor protein. The client proteins and co-chaperones are supposed to bind in the MD of HSP90 [31-35]. The NTD of HSP90 contains a lid structure that is closed in ATP bound conformation but remains open in the ADP bound and apo forms . In the case of HSP70, we find that the ligand typically binds to the SBD of HSP70, which is consistent with earlier reported findings [37,38]. Nevertheless, some binding of the ligand protein is found in the linker region of HSP70, and this type of binding has also been reported in the literature [39,40]. For HSP70, the α helical lid in the SBD plays a role in substrate binding. Here, the substrate is encompassed by the lid in a cavity comprised of β sheets. Structural reordering of helical lid is possible before and after substrate binding . Ongoing experimental investigations in our laboratory will hopefully shed some light in the current interactions identified by structural biology.
|Although HSP90 system is largely studied, there are many unanswered questions [42-45]. To our knowledge, complex formation between Hsp90 and Hsp70 are mediated through co-chaperones such as Hop/p60 or Sti1 in yeast [46-48]; no strong experimental evidence or crystal structure is available at this time to verify direct interactions of Hsp70’s with hsp90’s. The present work proposes the idea of a complex formation between HSP70 and HSP90 (as a TDF receptor), and a recent study on multiple myeloma also suggests the possibility of such direct interaction between Hsp70 and Hsp90 . Nevertheless, since no experimental evidence is available to verify this interaction, this proposal remains an open topic for further investigation.
|This work was supported in part by the Keep A Breast Foundation (KEABF-375-35054) and by support from the U.S. Army research office through the Defense University Research Instrumentation Program (DURIP grant #W911NF-11-1-0304). This work was also supported in part by Volverine Packing Co (Jim & Marsha Bonahoom) and Bonhomie Wine imports (Charlie & Claire Woods). CCD was supported during the summer of 2013 by the David A. Walsh ’67 fellowship.
- Woods AG, Sokolowska I, Deinhardt K, Sandu C, Darie CC (2013) Identification of tumor differentiation factor (TDF) in select CNS neurons. Brain Struct Funct.
- Platica M, Ivan E, Holland JF, Ionescu A, Chen S, et al. (2004) A pituitary gene encodes a protein that produces differentiation of breast and prostate cancer cells. Proc Natl Acad Sci U S A 101: 1560-1565.
- Sokolowska I, Woods AG, Gawinowicz MA, Roy U, Darie CC (2013) Characterization of tumor differentiation factor (TDF) and its receptor (TDF-R). Cell Mol Life Sci 70: 2835-2848.
- Sokolowska I, Woods AG, Gawinowicz MA, Roy U, Darie CC (2012) Identification of a potential tumor differentiation factor receptor candidate in prostate cancer cells. FEBS J 279: 2579-2594.
- Sokolowska I, Woods AG, Gawinowicz MA, Roy U, Darie CC (2012) Identification of potential tumor differentiation factor (TDF) receptor from steroid-responsive and steroid-resistant breast cancer cells. J Biol Chem 287: 1719-1733.
- NgounouWetie AG, Sokolowska I, Woods AG, Wormwood KL, Dao S, et al. (2013) Automated mass spectrometry-based functional assay for the routine analysis of the secretome. J Lab Autom 18: 19-29.
- Sokolowska I, NgounouWetie AG, Woods AG, Darie CC (2012) Automatic determination of disulfide bridges in proteins. J Lab Autom 17: 408-416.
- Sokolowska I, Woods AG, Wagner J, Dorler J, Wormwood K, et al. (2011) Mass spectrometry for proteomics-based investigation of oxidative stress and heat shock proteins. Oxidative Stress: Diagnostics, Prevention, and Therapy 1083: 369-411.
- Woods AG, Sokolowska I, Yakubu R, Butkiewicz M, LaFleur M, et al. (2011) Blue native page and mass spectrometry as an approach for the investigation of stable and transient protein-protein interactions. Oxidative Stress: Diagnostics, Prevention, and Therapy 1083: 341-367.
- Darie CC,Shetty V, Spellman DS, Zhang G, Xu C, et al. (2008) Blue Native Page and Mass Spectrometry Analysis of Ephrin Stimulation-Dependent Protein-Protein Interactions in Ng108-Ephb2 Cells. Applications of Mass Spectrometry in Life Safety 3-22.
- Roy U, Sokolowska I, Woods AG, Darie CC (2012) Structural investigation of tumor differentiation factor. Biotechnol Appl Biochem 59: 445-450.
- Roy U, Sokolowska I, Woods AG, Darie CC (2013) Tumor Differentiation Factor (TDF) and its receptor (TDF-R): Is TDF-R an inducible complex with multiple docking sites? Mod Chem appl1: 108.
- Roy U, Woods AG, Sokolowska I, Darie CC (2013) Structural evaluation and analyses of tumor differentiation factor. Protein J 32: 512-518.
- Schuermann JP, Jiang J, Cuellar J, Llorca O, Wang L, et al. (2008) Structure of the Hsp110:Hsc70 nucleotide exchange machine. Mol Cell 31: 232-243.
- Shiau AK, Harris SF, Southworth DR, Agard DA (2006) Structural Analysis of E. coli hsp90 reveals dramatic nucleotide-dependent conformational rearrangements. Cell 127: 329-340.
- Arnold K, Bordoli L, Kopp J, Schwede T (2006) The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22: 195-201.
- Kiefer F, Arnold K, Künzli M, Bordoli L, Schwede T (2009) The SWISS-MODEL Repository and associated resources. Nucleic Acids Res 37: D387-392.
- Tovchigrechko A, Vakser IA (2005) Development and testing of an automated approach to protein docking. Proteins 60: 296-301.
- Tovchigrechko A, Vakser IA (2006) GRAMM-X public web server for protein-protein docking. Nucleic Acids Res 34: W310-314.
- Duhovny D, Nussinov R, Wolfson HJ (2002) Efficient Unbound Docking of Rigid Molecules. Algorithms in Bioinformatics,Lecture Notes in Computer Science 2452: 185-200.
- Schneidman-Duhovny D, Inbar Y, Nussinov R, Wolfson HJ (2005) PatchDock and SymmDock: servers for rigid and symmetric docking. Nucleic Acids Res 33: W363-367.
- Andrusier N, Nussinov R, Wolfson HJ (2007) FireDock: fast interaction refinement in molecular docking. Proteins 69: 139-159.
- Mashiach E, Schneidman-Duhovny D, Andrusier N, Nussinov R, Wolfson HJ (2008) FireDock: a web server for fast interaction refinement in molecular docking. Nucleic Acids Res 36: W229-232.
- Roy A, Kucukural A, Zhang Y (2010) I-TASSER: a unified platform for automated protein structure and function prediction. Nat Protoc 5: 725-738.
- Zhang Y (2008) I-TASSER server for protein 3D structure prediction. BMC Bioinformatics 9: 40.
- Ali MM, Roe SM, Vaughan CK, Meyer P, Panaretou B, et al. (2006) Crystal structure of an Hsp90-nucleotide-p23/Sba1 closed chaperone complex. Nature 440: 1013-1017.
- Pearl LH, Prodromou C (2000) Structure and in vivo function of Hsp90. CurrOpinStruct Biol 10: 46-51.
- Stark C, Breitkreutz BJ, Reguly T, Boucher L, Breitkreutz A, et al. (2006) BioGRID: a general repository for interaction datasets. Nucleic Acids Res 34: D535-539.
- Li J, Soroka J, Buchner J (2012) The Hsp90 chaperone machinery: conformational dynamics and regulation by co-chaperones. BiochimBiophys Acta 1823: 624-635.
- Meyer P, Prodromou C, Liao C, Hu B, Roe SM, et al. (2004) Structural basis for recruitment of the ATPase activator Aha1 to the Hsp90 chaperone machinery. EMBO J 23: 1402-1410.
- Sato S, Fujita N, Tsuruo T (2000) Modulation of Akt kinase activity by binding to Hsp90. ProcNatl Acad Sci U S A 97: 10832-10837.
- Street TO, Lavery LA, Agard DA (2011) Substrate binding drives large-scale conformational changes in the Hsp90 molecular chaperone. Mol Cell 42: 96-105.
- Fontana J, Fulton D, Chen Y, Fairchild TA, McCabe TJ, et al. (2002) Domain mapping studies reveal that the M domain of hsp90 serves as a molecular scaffold to regulate Akt-dependent phosphorylation of endothelial nitric oxide synthase and NO release. Circ Res 90: 866-873.
- Blacklock K, Verkhivker GM (2014) Allosteric regulation of the Hsp90 dynamics and stability by client recruiter cochaperones: protein structure network modeling. PLoS One 9: e86547.
- Zhu X, Zhao X, Burkholder WF, Gragerov A, Ogata CM, et al. (1996) Structural analysis of substrate binding by the molecular chaperone DnaK. Science 272: 1606-1614.
- Mayer MP, Bukau B (2005) Hsp70 chaperones: cellular functions and molecular mechanism. Cell Mol Life Sci 62: 670-684.
- NgounouWetie AG, Sokolowska I, Woods AG, Roy U, Deinhardt K, et al. (2014) Protein-protein interactions: switch from classical methods to proteomics and bioinformatics-based approaches. Cell Mol Life Sci 71: 205-228.
- Yu H, Li S, Yang C, Wei M, Song C, et al. (2011) Homology model and potential virus-capsid binding site of a putative HEV receptor Grp78. J Mol Model 17: 987-995.
- Schlecht R, Erbse AH, Bukau B, Mayer MP (2011) Mechanics of Hsp70 chaperones enables differential interaction with client proteins. Nat StructMol Biol 18: 345-351.
- Wang AM, Miyata Y, Klinedinst S, Peng HM, Chua JP, et al. (2013) Activation of Hsp70 reduces neurotoxicity by promoting polyglutamine protein degradation. Nat Chem Biol 9: 112-118.
- Mahalingam D, Swords R, Carew JS, Nawrocki ST, Bhalla K, et al. (2009) Targeting HSP90 for cancer therapy. Br J Cancer 100: 1523-1529.
- Leach MD, Klipp E, Cowen LE, Brown AJ (2012) Fungal Hsp90: a biological transistor that tunes cellular outputs to thermal inputs. Nat Rev Microbiol 10: 693-704.
- Li J, Buchner J (2013) Structure, function and regulation of the hsp90 machinery. Biomed J 36: 106-117.
- Zhao R, Houry WA (2007) Molecular interaction network of the Hsp90 chaperone system. Adv Exp Med Biol 594: 27-36.
- Pratt WB, Morishima Y, Osawa Y (2008) The Hsp90 chaperone machinery regulates signaling by modulating ligand binding clefts. J Biol Chem 283: 22885-22889.
- Johnson BD, Schumacher RJ, Ross ED, Toft DO (1998) Hop modulates Hsp70/Hsp90 interactions in protein folding. J Biol Chem 273: 3679-3686.
- Chatterjee M, Andrulis M, Stühmer T, Müller E, Hofmann C, et al. (2013) The PI3K/Akt signaling pathway regulates the expression of Hsp70, which critically contributes to Hsp90-chaperone function and tumor cell survival in multiple myeloma. Haematologica 98: 1132-1141.