BiP (PDB ID: 3IUC)
Created by Christine Kelley
BiP, also known as HSP70 and GRP78, is an essential ER chaperone which functions in anterograde transport of nascent chains during cotranslational translocation and retrograde transport of misfolded proteins across the ER membrane (Hass, IG, 1994). It performs its function by attaching to the hydrophobic regions of the nascent chain to prevent it from folding prematurely. This enhances its ability to fold correctly with the hydrophobic regions on the inside of the structure (3, 7). This also helps to avoid the activation of the ERAD pathway which is instigated by the exposure of hydrophobic residues of misfolded proteins as BiP shelters them from detection (10). It has a molecular weight of 72.46kDa and an isoelectric point of 5.09 (8).
BiP consists of two almost identical subunits, the A and C chains which are 379 and 388 amino acids long, respectively. The N-terminus is the nucleotide binding domain, which binds ATP or ADP, and the C-terminus represents the substrate binding domain, which attaches to the nascent chains (1, 18). The nucleotide binding domain is linked to the substrate binding domain by a flexible linker, connecting the N terminus and the C terminus (18).
The two subunits of BiP are constituted of approximately 27% alpha helices and 36% beta sheet secondary structure, with the rest being random coils and turns. Most are aliphatic, with the polar residues on the surface and the nonpolar amino acids facing the interior, mediating the hydrophobic collapse and favorable interactions with the polar lumen (18). The beta strands' amino acids alternate in polarity to create their aliphatic character with a polar and nonpolar face while the alpha helices also have this face character (8). Here, the beta sheets are represented in yellow, the alpha helices in red, random coils in green and turns in purple. The calcium ions are in grey and the ADP molecules in blue.
Hydrogen bonds exist between the extensive secondary structures in this protein. There are intrastrand hydrogen bonds present in all of the alpha helices, connecting every 10 residues and interstrand hydrogen bonds between the beta strands to form beta sheets. (18). These hydrogen bonds decrease the polarity of the residues' carboxyl and amino groups, allowing them to stack closer together and to allow for hydrophobic collapse.
Some studies show that the Threonine-37, Threonine-229 and Glutamic Acid-201 residues are involved in the intrinsic ATPase activity of BiP. Mutations of these amino acids do not inhibit ATP binding, but do inhibit hydrolysis and thus the protein remains bound to the polypeptide fragment. These studies argue that point mutations of E201G reduced the ATPase activity dramatically due to the lose of negative charge. The Glu201 amino acid interacts with the divalent cation which is required for nucleotide binding. Mutating it to a glycine does not inhibit binding because there is less steric hinderance despite the fact that it loses the negative charge of the acidic side chain. It decreases the ATPase activity as it cannot successfully interact with the ATP and cause the conformational changes required for ATPase activity (6, 16). Mutations in Thr-37, such as T37G allow binding but the conformational change is inhibited and thus the ATPase mechanism is defective. Failing to undergo the conformational change prevents the protein from transducing the information from the nucleotide binding domain to the peptide binding domain, which would normally tighten its fit on the peptide had the ATP been hydrolyzed. Finally, another class of mutant is the hydrolysis mutation exhibited by a change in T229. These mutants, such as T229G, can bind ATP, exhibit the ATP induced conformational change, but cannot hydrolyze it. This is probably because the threonine's side chain interacts with the gamma phosphate of the ATP and removal of this side chain reduces its ability to remove the phosphate (16).
Also, Thr-13, which is not present on this graphical representation of the protein, is thought to hydrogen bond to the alpha-phosphate of ATP to stabilize it within the ATP binding pocket (6, 16). Other residues involved in ATP binding are G-226 and G-227. When mutated to aspartic acids the protein is unable to bind the ATP as the introduction of a large and charged side chain repels the negative charges on the phosphates. Mutations to other amino acids with side chains, as compared to glycine's hydrogen, also increase ATP's KD (16). Furthermore, T-37, Y-39 and G-227 hydrogen bond the beta phosphate of ATP to stabilize it in the ATP cleft while S-300 and R-297 hydrogen bond the adenosine ring. Further stabilizing the substrate, S-365, E-293 and K-296 hydrogen bond to the ribose ring (18).
Other amino acids of BiP are involved in complexing with the four calcium ions present. His-252 and Asp-257 of chain A and Gly-315 of chain C are responsible for one such interaction, with the other calcium in chain C being complexed by the same residues on the opposite chains (Wisniewska, M., et. al., 2010). The latter two are at the base of their respective alpha helix 6, and held by the influence of E-256 and D-257's carboxyl groups near the ATP binding site, as well as the negative charges on the phosphates(18).
Also present at the surface of the nucleotide binding domain is Arginine-197, which interacts both with cochaperones, which accelerate ATP hydrolysis, and the substrate binding domain. Substituting a neutral or acidic amino acid in place of the positively charged basic argenine does enhance the ATPase activity from its wild type efficiency, but negatively effects the interaction of the substrate binding and nucleotide binding domains. This change reduced the ability of the protein to release the polypeptide substrate (1).
BiP binds to the nascent polypeptide chain being translocated into the ER when it has ATP bound to its nucleotide binding domain. This ATP is then hydrolyzed to ADP due to the intrinsic ATPase within BiP, and this triggers a conformational change in the substrate binding domain which tightens its hold on the peptide chain. This affinity holds until the ADP is exchanged with ATP, at which point the peptide is released due to the increase in KD (17). The conformations of BiP with ATP bound have not yet been resolved, but it is known that a conformational change must be present because ATPase activity is not sufficient for peptide release alone, as seen when studying conformational mutants (16, 18). The T37G mutant studied exhibits the same KD as the wild type so the researchers found that the conformation change does not change the stability of the nucleotide in the binding pocket but instead transduces a signal from the nucleotide binding domain to the substrate binding domain for either a tighter bond (when ATP is hydrolyzed) or to release the peptide (when ATP replaces ADP) (16). As the substrate binding domain is at the interface of the two subunits, the conformational change is probably a hinge-like mechanism which brings the two closer together, tightening the attachment to the peptide.
BiP tends to exist in the ER lumen with ATP bound and its ATPase activity is only activated once it binds to its polypeptide substrate (17, 18). Therefore, it is the nucleotide binding domain which mediates the structure of the substrate binding domain through ligand induced conformational changes. Furthermore, the peptide binding domain also exhibits allosteric feedback inhibition of the ATPase activity when no peptide is bound. This affects the rate, but not the mechanism that the nucleotide binding domain exhibits (6, 18).
The conformational changes exhibited by the nucleotide binding domain occur mostly due to the Arg-197 residue. This amino acid interacts with cochaperones which accelerates the rate of ATP hydrolysis and also interacts with the substrate binding domain, transducing the signal to either tighten or release its grip on the peptide. Substituting a neutral or acidic amino acid for the positive basic argenine increases the ATPase activity but negatively affects the interaction of the nucleotide binding domain with the substrate binding domain. Therefore, while it may still hydrolyze the ATP, it cannot produce the needed conformational change to release the protein (Awad, W., Estrada, I., Shen, Y., Hendershoot, L. M., 2008).
BiP also has associated peptides which complex with it to mediate the ATPase activity as well as nucleotide exchange. Examples of these peptides is the BiP associated protein (BAP), GrpE and ERdj (PDB ID 2YUA. These proteins function to regulate ATPase activity, strengthening the grip on the peptide, and act as nucleotide exchange factors, replacing ADP with ATP to release the nascent chain (1, 4, 9). The intrinsic ATPase of BiP is not efficient enough to bind the peptides as fast as is necessary to prevent misfolding and aggregation, therefore, these other peptides are needed to increase its velocity (16).
In addition to ADP and ATP, BiP also binds metal ions as prosthetic groups. Magnesium associates in the nucleotide binding site and calcium more in two pockets near the ATP binding domain. Mutations at the magnesium binding sites decrease the binding affinity of BiP for nucleotides dramatically. One study cites D10 and D199 for these interactions, however, these residue locations do not correspond with the sequence from the PDB. They studied an HSP70 isoform and since the HSP70 proteins all share the same ATPase sites it is logical to assume that these residues have corresponding amino acids in BiP's sequence. Mutating these aspartic acids decreases the Kcat to less than 1% of the wild type efficiency, showing that the association of magnesium is critical to ATPase function (17). The magnesium is also crucial for hydrolysis because it orients the phosphates in such a way that the Thr-203's hydroxide is capable of nucleophilic attack on the phosphoric anhydride bond, releasing inorganic phosphate as it hydrolyzes the ATP to ADP. In this crystallized structure, the conformational change due to the hydrolysis has already occurred and the Thr-203 is oriented so that it's hydroxide is facing opposite to the ADP; however, through mutation studies it is strongly held that this residue is responsible for the nucleophilic attack (6). BiP can also be considered to be a calcium storage protein and the binding of this cation has strong effects on the binding of the nucleotides. In one study, the association of calcium was shown to lower the KD for ATP by more than a factor of 10 and ADP by almost 1000. Furthermore, the binding of ATP also decreases the KD for calcium binding by 40-fold and almost ADP binding decreases it by a factor of 900 (11). The association of magnesium in increasing the ATPase activity and calcium in decreasing the Kd of the nucleotides acts to enhance the activity of BiP by increasing its efficiency and speed.
The E. Coli chaperone DnaK (PDB ID 1DKG) is the prokaryotic homolog of BiP. DnaK has 638aa and BiP 654aa. According to PDB, it has a 50% sequence similarity to BiP and has a similar function in preventing protein aggregation through the prokaryotic UPR. Additionally, it can stimulate the refolding of denatured proteins in the cytosol, which BiP is unable to do (2). While it binds ATP as BiP does, it has a very different structure which is exacerbated by its two extended alpha helices (15).
The protein alpha actin (PDB ID 2FXU) has an 83% structural similarity to BiP's individual subunits while a surprisingly low 10% sequence similarity (Gaut, J. R., Hendershot, L. M., 1993). Here, the A subunit of BiP is superimposed with alpha actin showing the striking conformational similarity. Similarity. It has a very similar ATP binding pocket and is constituted mostly of alpha helices, as BiP is (12). Also, they have practically identical ATP binding pockets and they contain many of the same residues (Gaut, J. R., Hendershot, L. M., 1993). However, despite the structural similarities, its function is much different than the chaperone. It exists in the cytosol of skeletal muscle and Is primarily responsible for the contractile fibers and can also induce muscle formation (19).
Another example is HspBP-1 (PDB ID 1XQR), which has a 70% sequence identity with BiP, however, it consists mostly of alpha helices compared to the approximate equity of helices and beta sheets in BiP. This protein functions as a nucleotide exchange factor and as a result, complexes with the ATPase domain of the HSP70 proteins (14). The structure of the HspBP-1 - BiP ATPase fragment complex has been resolved (PDB ID 1XQS) (13).
Finally, no drugs have currently been developed which target BiP for therapeutic use; however, there are groups investigating this protein as a target to regulate an abnormal unfolded protein response (UPR) which has been indicated in cancer and genetic disorders (5).