Role of Hsp90 machinery in native folding and activity of steroid/nuclear receptors
	Components of Hsp90 machinery
	Hormone binding and activity of steroid receptors is mediated by the Hsp90 chaperone machinery (Pratt and Toft, Endocrine Rev., 18(3), 306-360, 1997; Pratt and Toft, Exp Biol Med, Maywood., 228:111-133, 2003). Hsp90 is necessary for native folding of the receptors and for keeping the hydrophobic ligand-binding pocket (cleft) in an open, ligand-binding competent conformation (Pratt et al., J Biol Chem. 283:22885-9, 2008). A minimal set of proteins that are required for reconstitution of the receptor-Hsp90 complex and for uncompromised ligand-binding activity are Hsp90, Hsp40, Hsp70, Hop and p23 (ibid). AB Vector has reconstructed this machinery in a baculovirus expression system (details). Initially Hsp40 binds to the ligand-binding domain of the receptor, followed by Hsp70. This primes the receptor for interaction with Hsp90 in complex with Hop. Hop, while not essential for the ligand-binding cleft to open, binds independently via the N-terminal TRP domain to Hsp70 and via a central TRP domain to Hsp90. After Hop-mediated Hsp90 binding to Hsp70 and the receptor, an ATP and K+ dependent event leads to the opening of the receptor cleft enabling access to hydrophobic ligands. Following the cleft-opening step, Hop and some Hsp70 dissociates from the complex. The receptor-bound Hsp90 is now in its ATP-dependent conformation and can be bound by p23, which stabilizes the chaperone in that conformation, preventing disassembly of the receptor-Hsp90 complex. Hsp90 appears to associate with steroid receptors at about a ratio 2:1 (Denis et al., J. Biol., Chem., 262, 11803-6, 1987; Iannotti et al., Arch. Biochem. Biophys., 264, 54-60, 1988; Renoir et al., Biochemistry, 23, 6016-23, 1984).
	Hsp90 implications for receptor activity and specificity
	Several Hsp90-dependent receptors exhibit ligand-binding activity in the absence of Hsp90. However, their specific activity is compromised because in the absence of Hsp90 the ligand-binding pocket is predominantly closed, opening only very transiently in the course of molecular breathing. On the contrary, when stable complexes are assembled with Hsp90, nearly all ligand-binding pockets remain open at all times and low concentrations of steroid are sufficient for binding (Pratt et al., J Biol Chem. 283:22885-9, 2008). Since the process of ligand entering a stably open pocket as compared to entering breathing, only partially open pocket is different, ligand selectivity may differ contingent upon size, shape and charge of the ligands. For instance, larger ligands may be at disadvantage entering a half-open pocket as compared to smaller ligands. It is expected that the kinetics of binding also will be different.
	Hsp90 implications for receptor stability in HTS assays
	Target protein aggregation is a common problem that leads to variability in drug screening assays (Taylor et al., J Biomol Screen., 5(4):213-26, 2000). Unliganded and apo-Hsp90 receptors are structurally labile and inherently unstable and prone to aggregation. In order to prevent protein aggregation during the storage, commercially available steroid receptors contain various excipients such as high concentrations of salt and detergents (Estrogen receptors alpha and beta), urea (Progesterone receptor) and ammonium sulfate (Androgen receptor). However, adding such receptors into assay mixtures that do not contain urea, high salt or detergents could initiate protein aggregation. Furthermore, detergents may interfere with the properties of hydrophobic ligand binding pockets of the receptors or with the hydrophobic ligands. In contrast, receptor/Hsp90 complexes are less structurally labile as they are stabilized by Hsp90 and are less prone to aggregation. Furthermore, Hsp90 is a highly soluble protein and can help keep bound proteins in solution. AB Vector has formulated a panel of receptor/Hsp90 complexes without the use of chaotropic agents, detergents and high salt concentrations, representing a more physiological environment. All receptor/Hsp90 complexes are available in similar formulations. This facilitates easier comparison between different receptors when compared to the receptors from other suppliers that are available in a wide range of formulations. It should be noted that the yield of recombinantly expressed active steroid receptors is low in mammalian, insect, and yeast systems, and it is nearly impossible to produce active steroid receptors in E.coli (Srinivasan G., Mol Endocrinol. 6(6):857-60, 1992). Better yields were obtained in insect cell expression system infected with recombinant baculoviruses and it remains the system of choice for commercial production of receptors for drug screening. When steroid and nuclear receptors are overexpressed in insect cells without additional Hsp90, the host cell protein folding capacity is overwhelmed and the receptors are produced largely in a misfolded conformation. This results in formation of soluble and insoluble protein aggregates (inclusion bodies). Typically, full or partial disruption of aggregates is achieved using chaotropic agents and high salt concentrations for protein extraction, purification and storage. Though this often yields soluble protein, it usually does not result in a native conformation of all protein molecules. This is especially true for labile proteins, such as steroid receptors. Therefore, receptor aggregation during storage or at mixing with the components of drug screening assays often is problematic. Another reason for the misfolding of recombinant steroid receptors is that insect molecular chaperones are not fully competent in folding steroid receptors. Evolutionary, Hsp90 function is domain and kingdom specific and, to a degree, phylum-specific. For instance, E.coli Hsp90 cannot facilitate native folding of steroid receptors. Therefore, human steroid receptors are inactive if expressed in E.coli (Jaglaguier et al., J Steroid Biochem Mol Biol., 57: 43-50, 1996). Plant molecular chaperones are also incapable of supporting mammalian steroid receptor activity. For instance, glucocorticoid receptor (GR) translated in wheat germ extract was not associated with wheat Hsp90 and did not bind steroid with high affinity, whereas GR translated in rabbit reticulocyte lysate was in complex with rabbit Hsp90 and attained high affinity steroid binding conformation (Pratt et al., Nihon Naibunpi Gakkai Zasshi, 66:1185-97, 1990). Most human receptors exhibit ligand-binding activity if expressed in insect cells. However, even insect Hsp90 does not interact with human steroid receptors as well as human or even avian Hsp90. Therefore ligand binding is not as competent. For example, when human GR, chicken Hsp90 and insect Hsp90 were co-expressed in insect cells, GR ligand will preferentially bind to GR/chicken Hsp90 complexes versus the GR/insect Hsp90 complexes (Cadepond et al., Proc. Natl. Acad. Sci. USA, 90:10434-8, 1993). Likewise, we observed weak interaction of human steroid receptors with indigenous insect Hsp90 as compared to the receptors interaction with recombinant human Hsp90 expressed in insect cells (details).