Binding Sites

From Chaperome

Please submit your thoughts and insite here (click on the edit button above).

Function

GRP94 suggested to have a dual functionality: first - as ER chaperone, and the second – binding and presentation of immunologically-important peptides. Peptide binding has been relatively well studied in vitro; however, its in vivo significance is disputed. The chaperone function has not been rigorously demonstrated as of yet; the protein has been often referred to as molecular chaperone probably due to its high homology to the cytosolic HSP90 and binding to folding intermediates. Induction of GRP94 expression by stresses and, on opposite, induction of the UPR upon GRP94 inhibition, also indicate a function in protein biosynthesis in ER.

Evidence of chaperone function: to be defined as chaperone (Hendrick and Hartl, 1993), a protein should 1. bind transiently to non-native form of substrate protein; 2. the binding and/or release of substrate should be regulated; and 3. the substrate’s stability, rate of folding or assembly should be influenced by the chaperone binding.

1. GRP94 has been shown to interact with a small set of substrates, including Ig chains(Melnick, 1992), MHC classII (Schaiff, 1992), c-erbB2 (Chavany, 1996), thyroglobulin (Muresan, 1997), a herpes-virus glycoprotein (Ramakrishnan, 1995), apolipoprotein B (Linnik, 1998), collagen (Ferreira, 1994), protein C (Katsumi, 1996), and bile-salt-dependent lipase Bruneau, 1998). For most of these, interaction was demonstrated by co-immunoprecipitation only; GRP94 was often found to interact with advanced folding intermediates or with incompletely assembled oligomers/complexes. Therefore, as for HSP90, GRP94’s recognition of protein substrates is probably through a structural determinant(s). However, there is no obvious similarity between folds of its putative substrates. By far the best studied example is Ig light chain – GRP94 has been shown to associate with late folding intermediates, before they became secretion-competent, but after BiP (Melnick, 1992 and 1994); BiP is known to bind incompletely oxidized early folding intermediates with exposed hydrophobic stretches.

2 and 3 – the in vivo regulation of GRP94 or the consequences of binding to the substrate proteins remain unknown. Inhibitors Rad and/or GA may induce similar conformational change in vivo as they do in vitro (Vogen, 2002), since little GRP94 was precipitated by 9G10 (antibody that distinguishes inhibitor-bound conformation of GRP94) from GA-treated cells (Chavany, 1996).

While the chaperone function of GRP94 awaits formal demonstration, it is clear that it is essential for biogenesis of certain cell-surface or secreted proteins. In GRP94-deficient B-cells, some Toll-like receptors (TLR1, 2 and 4) and a subset of integrins (CD11a/CD18, CD49d) were not expressed on cell surface, resulting in lack of response to LPS (Randow and Seed, 2001); re-expression of GRP94 restored LPS reactivity. GRP94 with deletions in negatively charged linker or KDEL sequence, as well as chimeric HSP90/GRP94 molecules were not able to rescue the deficiency. In Arabidopsis, the function of CLAVATA receptor complex was shown to depend on GRP94 function (Ishiguro, 2002). These observations suggest a function in either folding, subunit assembly, or cellular trafficking.

Peptide binding/antigen presentation: GRP94 is able to augment presentation of peptides to T cells (Srivastava, 1993). Direct evidence for peptide binding was obtained by identification of peptides in acid eluates of purified GRP94 (Li, 1993; Nieland, 1996) and direct binding of peptides to purified GRP94 (Blaschere, 1997; Wearsch, 1997). GRP94 dramatically increases peptide recognition by T cells by enhancement of peptide presentation to the adaptive arm of the immune system (Blaschere, 1997) and by stimulation of innate immunity (Singh-Jasuja, 2000). The GRP94-peptide complexes are taken up by professional antigen presenting cells via receptor mediated endocytosis (Binder, 2000) and the peptides are then re-presented on MHC class I molecules on the cell surface. Several VSV - derived peptides such as VSV8 and peptide A have bind GRP94 (Wearsh, 1997; Blaschere, 1997). VSV8 is an octamer (RGYVYQGL) from the N protein of VSV and is the dominant T cell viral epitope (Zhang, 1992). Peptide A is a non-immunogenic 15-mer (KRQIYTDLEMNRLGK) from the glycoprotein of the same virus. Other peptides include YVNTNMG from the core protein of hepatitis B virus (Meng, 2001)), the ovalbumin SIINFEKL (Breloer, 1998)) and a mouse leukemia tumor antigen IPGLPLSL from a mutated Akt (Ishii, 1999); except for relative hydrophilicity of most of good binders, no common sequence motif is obvious from comparing GRP94 binding peptides. No obvious size preference was found using radioactive peptides with photo-reactive side chains (Spee, 1997). The only sequence specificity found was that 9mers with basic or acidic amino acids in positions 2 and 9 bind relatively weakly to GRP94. Thus, neither sequences nor the structural features that are compatible with binding to GRP94 are known.

Peptide presentation by GRP94 has been contested; instead, GRP94 is suggested to function as a natural adjuvant in antitumor immune responses (Baker-LePain, 2002), mainly due to low stoichiometry of peptide binding and ability of N-terminal half of the protein to illicit immune response. A C-terminal peptide-binding site, hydrophobic in nature, was thought to be responsible for the peptide-binding by GRP94. The C-terminal peptide-binding site was identified by cross-linking of azido-modified peptide (Linderoth, 2000) and mapped using peptide-pyrene conjugates (Linderoth, 2001); aromatic and hydrophobic interactions appear to drive the peptide binding to this site. However, an N-terminal fragment was later found to efficiently bind relevant peptides in specific, Rad-dependent fashion (Vogen, 2002).

An N-terminal peptide binding site was identified via molecular docking simulation to be located on the concave face of the beta-sheet of the N-terminal domain (on the opposite face of the beta-sheet from the nucleotide/inhibitor binding pocket), in close proximity to a deep hydrophobic pocket (Gidalevitz, 2004) where bis-ANS binds(Wassenberg, 2000). An N-terminal fragment (aa 1 to 355 and aa 34 to 355) is sufficient for most of the known activities of the full length protein: peptide binding, inhibitor binding and conformational regulation. Residues 223–355 are dispensable for retention of bound peptide; peptide binding is hydrophilic in nature and sensitive to pH, consistent with the rather hydrophilic nature of binder peptides. His125 contacts the bound peptide and is important for peptide-binding activity (substitution to Asp abolishes peptide but not inhibitor binding) (Gidalevitz, 2004). Peptide binding to this site is specifically inhibited by Rad or GA; the binding is saturable, close to 1:1 stoichiometry, and has sequence specificity distinct from that of BiP (Vogen, 2002). The question of whether GRP94 can indeed present peptides via the N-terminal site awaits resolution.

See also:ER chaperones