G_Protein_Coupled_Glucagon

The binding of the hormone glucagon to the human class B G-protein coupled glucagon receptor (GCGR) leads to a signal cascade that raises blood glucose levels through gluconeogenesis (1). The use of allosteric antagonists such as N-[(4-{(1S)-1-[3(3, 5-Dichlorophenyl)-5-(6-methoxynaphthalen-2-yl)-1H-pyrazol-1-yl]ethyl}phenyl)carbonyl]-β-alanine (MK-0983) to limit the function of the glucagon receptor has been shown to greatly reduce blood glucose levels in mouse models (2). Residues 1-135 of the N-terminus and residues 418-477 of the C-terminus were removed from a thermostabilized GCGR and residues 256-258 were replaced with T4-lysozyme (T4L) to produce a chimeric glucagon receptor called GCGR-StaR(136–417)–T4L (PDB ID: 5EE7) (3, 4). The chimeric receptor, at 452 residues long, has an isoelectric point (pI) of 9.53 and a molecular weight of 51212.62 Da (4, 5). GCGR-StaR(136–417)–T4L was crystallized in the presence of MK-0893 to determine the interactions that take place between GCGR-StaR(136–417)–T4L and MK-0893 (3).

When glucagon, which originates from alpha cells of Islets of Langerhans in the pancreas, binds a GCGR, the GCGR undergoes a conformational change to activate its associated G protein (1). The G protein then activates adenylyl cyclase, which produces the second messenger cAMP. The second messenger then activates various responses such as promoting gluconeogenesis in the liver. Hyperglucagonemia produces extra activity in the pathway, contributing along with insulin resistance to the high blood glucose levels seen in type 2 diabetes mellitus patients (6).

Wild-type GCGRs are polypeptide monomers composed of an N-terminal extracellular domain (ECD) and a seven transmembrane (7TM) domain (7). The 7TM domain contains seven membrane-spanning alpha helices with three extracellular loop (ECL) regions and three intracellular loop (ICL) regions. The high content of hydrophobic amino acids including leucine, valine, and alanine improves the ability of the helices to cross the hydrophobic bilayer of the cell membrane (7). The helical secondary structure neutralizes the polar amide moieties of the peptide backbone, further improving the peptide’s ability to cross the hydrophobic bilayer of the cell membrane. Here, primary and secondary structure demonstrate the importance of structure for function, as it is crucial that the GCGRs are able to embed into the membrane, otherwise they would not be able to bind to the extracellular glucagon (7).

The membrane-spanning helices of the 7TM domain and the N-terminal ECD form a binding pocket for glucagon (7). Various residues within the 7TM helices, including Tyr-149, Val-191, Gln-232, Glu-362 and Leu-386, are important in creating the correct binding pocket shape for glucagon (7). Mutations at any of these residues significantly decreases binding affinity. Outside of the helices, Asp-208 and Trp-215 of ECL1 and Trp-295 and Asn-298 of ECL2 stabilize the resulting conformation when glucagon binds to the GCGR (7). Trp-36 of the N-terminal ECD, which also aids in forming the binding pocket, forms a particularly important hydrophobic interaction with the C-terminus of glucagon (7).

The interactions between the antagonist MK-0893 and a class B GCGR prevent the GCGR from transmitting a signal, even in the presence of glucagon (3). The crystallization of GCGR-StaR(136–417)–T4L in the presence of MK-0893 allowed for the elucidation of the specific interactions at play. It should be noted that the structural modifications of GCGR-StaR(136–417)–T4L did not affect binding of MK-0893 when compared to the wild type GCGR (3).

MK-0893 binds to GCGR-StaR(136–417)–T4L between transmembrane helix six (TM6) and TM7 (3). The amide group of MK-0893 engages in polar interactions with Lys-349 and Ser-350. The carboxyl group of MK-0893 forms an ionic interaction with Arg-346 and a polar interaction with Asn-404 (3). The combination of these interactions prevents TM6 from moving outward, such that a GCGR would not be able to undergo the glucagon-induced conformational change required to activate its associated G-protein (3). As a result, signal transduction is interrupted.

The interactions between MK-0893 and GCGR-StaR(136–417)–T4L illustrate that structure determines not only function, but also how function can be inhibited. The different moieties of MK-0893 compliment the different residues present in the allosteric binding pocket between TM6 and TM7. Mutations of any of these residues were shown to significantly decrease binding and weaken inhibition (3).

In GCGR-StaR(136–417)–T4L, residues 256-258 of ICL3 were replaced with T4L (PDB ID: 164L) (3, 4). The addition of a highly soluble enzyme such as T4L into a GCGR acts to increase the overall solubility of the GCGR, making crystallization of the receptor easier (8). The insertion likely directly aided in crystallization by also providing better protein-protein contact surfaces for crystallization to occur (8). Crystallization was also improved by the addition of oleic acid, an 18-carbon long fatty acid, to the crystallization buffer (4). Several molecules of oleic acid were found to be interacting with GCGR-StaR(136–417)–T4L in its crystallized form(4).

Wild-type T4L is a polypeptide monomer consisting of 164 residues, with a molecular weight of 18700 Da (9, 10). In comparison to GCGR-StaR(136–417)–T4L, the PSI-BLAST database assigns an E value of 7e-114 to free T4L (11). E values are determined by an algorithm that compares a query sequence, GCGR-StaR(136–417)–T4L in this case, to other primary structures and assigning “gaps,” which are amino acid sequences found in the other primary structures but not found in the query sequence. The fewer and smaller the gaps, the smaller the E value, with a value less than 0.05 considered significant. An E value of 7e-114 for T4L suggests that its amino acid is highly similar to the sequence of GCGR-StaR(136–417)–T4L.

Residues 18-34 of wild-type T4L form an anti-parallel sheet of three strands (10). Over half of T4L is made up of alpha helices, the longest of which is a 20-residue helix connecting the two domains of the lysozyme (10). The other helices vary in length from six residues to 13 residues (10).

In addition to having near identical primary structure, wild-type T4L and the T4L domains of GCGR-StaR(136–417)–T4L have almost identical secondary and tertiary structure (10). In comparison to GCGR-StaR(136–417)–T4L, the Dali Server Protein Database assigns wild-type T4L a Z-score of 28.0 (12). The Dali Server compares tertiary structures of proteins by calculating the differences in intramolecular distances between proteins through a sums-of-pairs method. The calculations result in a Z-score that represents the similarity of tertiary structure. The greater the similarity in folding, the larger the Z-score, with a score larger than 2 considered significant (12). A Z-score of 28.0 for T4L suggests that its tertiary structure is highly similar to that of GCGR-StaR(136–417)–T4L. The high Z-score supports that the T4L did not undergo any significant folding changes as a result of being inserted into ICL3 of GCGR-StaR(136–417)–T4L.

The difference between T4L and GCGR-StaR(136–417)–T4L lies in that GCGR-StaR(136–417)–T4L contains an additional domain not present in T4L (3). The 7TM domain of GCGR-StaR(136–417)–T4L is composed mainly of hydrophobic residues (7). This contrasts T4L, in which polar residues make up the majority of the primary structure (4).

The difference in primary structure leads to a difference in tertiary structure as well. The relatively less numerous hydrophobic residues of T4L form a small hydrophobic core, and the protein takes a globular shape to allow its many polar residues to have contact with the aqueous solvent (10). GCGR-StaR(136–417)–T4L possesses lipophilic alpha helices that are most stable when hidden from the aqueous solvent by other hydrophobic substances (7). It was for this reason that oleic acid was added to the GCGR-StaR(136–417)–T4L crystallization buffer.

GCGR-StaR(136–417)–T4L and T4L perform different functions, a result of the differences in their structure. The 7TM domain of GCGR-StaR(136–417)–T4L allows it to successfully incorporate into plasma membranes, which is crucial for its ability to function as a glucagon receptor (1). The globular wild type T4L, with its hydrophobic core and hydrophilic outer layer, is a free cytosolic protein (10). It cannot embed in the plasma membrane and therefore cannot take part in the signal transduction activated by glucagon binding.

Even if wild type T4L could embed in the plasma membrane to gain access to glucagon, it cannot bind to it. Wild type GCGRs bind to glucagon via the 7TM domain and ECD, both of which are absent in T4L (7). The N-terminus of GCGR-StaR(136–417)–T4L was truncated for easier crystallization, removing the ECD from the receptor (3). The loss of the ECD significantly reduces glucagon binding, but due to the 7TM binding pocket, binding is still possible for GCGR-StaR(136–417)–T4L, unlike for T4L (3).

The use of allosteric antagonists to inhibit GCGR activity has been studied and is seen as a potential therapy to treat hyperglucagonemia (2). GCGR-StaR(136–417)–T4L, a GCGR whose N-terminus and C-terminus were truncated and with residues 256-258 being replaced with wild type T4L, was crystallized while interacting with the allosteric antagonist MK-0893 (3).

The 7TM domain of GCGR-StaR(136–417)–T4L is made of seven connected alpha helices with high hydrophobic residue content that allow them to embed into the plasma membrane (7). The 7TM domain along with the ECD that was cleaved from the N-terminus form the binding pocket necessary to bind glucagon (7).

Data gathered from crystallization of GCGR-StaR(136–417)–T4L with MK-0893 suggests that polar and ionic interactions between MK-0893 and several residues of GCGR-StaR(136–417)–T4L prevent TM6 of GCGR-StaR(136–417)–T4L from moving outward (3). The result is the inability of GCGR-StaR(136–417)–T4L to undergo a conformational change to activate an associated G-protein, interrupting signal transduction (3). This insight into how MK-0893 inhibits GCGR function may be used in the future to improve treatment of type 2 diabetes mellitus patients and others suffering from hyperglucagonemia.