Structural Mechanistic Studies

Overview

Structural studies of the insulin receptor have been limited largely to the cytoplasmic portion of the P-subunit. Within the cytoplasmic portion are 36 residues in the jux-tamembrane region (membrane-proximal), 295 residues in the tyrosine kinase domain, and 72 residues in the C-terminal tail (Fig. 1). Crystal structures of the insulin receptor kinase domain (IRK) have been determined in different states of phosphorylation [8,9] and for several mutants [10] (S. R. Hubbard, unpublished data). IRK shares a similar overall architecture with protein serine/threonine kinases [11], with an N-terminal lobe composed of a five-stranded antiparallel P-sheet and one a-helix, and a larger C-terminal lobe that is mainly a-helical (Fig. 2). ATP binds in the cleft between the two lobes, and the tyrosine-containing segment of a protein substrate interacts with residues in the C-terminal lobe. Several residues are highly conserved in all functional tyrosine kinases, including several glycines in the nucleotide binding loop; a lysine in P-strand 3 (P3); a glutamic acid

Flawe Antena

Figure 1 Overall architecture of the insulin receptor. The a-chains are extracellular and the P-chains pass through the plasma membrane. The major sites of tyrosine autophosphorylation are numbered according to Ebina et al. [2]. There are two fibronectin type III domains in the extracellular portion of the receptor. The N-terminal portion of the first domain is on the a-chain, and the C-terminal portion of this domain resides in the beginning of the P-chain, which is followed by a second, intact domain. Solid lines between the a-chains and between the a- and P-chains represent disulfide linkages.

Figure 1 Overall architecture of the insulin receptor. The a-chains are extracellular and the P-chains pass through the plasma membrane. The major sites of tyrosine autophosphorylation are numbered according to Ebina et al. [2]. There are two fibronectin type III domains in the extracellular portion of the receptor. The N-terminal portion of the first domain is on the a-chain, and the C-terminal portion of this domain resides in the beginning of the P-chain, which is followed by a second, intact domain. Solid lines between the a-chains and between the a- and P-chains represent disulfide linkages.

in a-helix C (aC); an aspartic acid, arginine, and asparagine in the catalytic loop; and, in the activation loop, a DFG motif in the beginning and a proline at the end (Figs. 2 and 4).

Although no crystal structure has been reported for the extracellular region of the insulin receptor, a crystal structure of the first three domains of the highly related IGF-1 receptor has been published [12]. This structure shows the probable site of interaction between the ligand (IGF-1 or insulin) and the L domains and cysteine-rich domain of the receptor. Recently, a low-resolution image reconstruction of the full-length insulin receptor with bound insulin has been determined using electron microscopy, providing a putative model for the spatial organization of the various domains in the intact receptor [13].

Receptor Activation Mechanism

The binding of insulin to its receptor induces a confor-mational change in the receptor, measured (among other techniques) as a reduction in the Stokes radius of the insulin-occupied versus -unoccupied receptor [14]. The physical mechanism by which binding of insulin to the extracellular portion of the receptor is transduced to the cytoplasmic portion is not well understood. From the image reconstruction

Figure 2 Ribbon diagram of the tyrosine kinase domain of the insulin receptor [9]. The a-helices are lettered, and the P-strands are numbered. The ATP analog (AMP-PNP) and the side chain of the substrate tyrosine are shown in ball-and-stick representation. The N terminus is denoted by N; the C terminus is not visible but follows shortly after aJ.

Figure 2 Ribbon diagram of the tyrosine kinase domain of the insulin receptor [9]. The a-helices are lettered, and the P-strands are numbered. The ATP analog (AMP-PNP) and the side chain of the substrate tyrosine are shown in ball-and-stick representation. The N terminus is denoted by N; the C terminus is not visible but follows shortly after aJ.

of the insulin-occupied receptor, a model for this mechanism has been proposed in which insulin binding produces a movement of the transmembrane helices toward one another, facilitating autophosphorylation between the kinase domains [15].

Once the insulin-triggered conformational change has been transmitted to the cytoplasmic domains, most biochemical data are consistent with tyrosine autophosphorylation proceeding via a trans mechanism, in which the tyrosine kinase domain of one P-chain phosphorylates tyrosines on the other P-chain within the same heterotetramer [16]. Simple modeling studies based on IRK crystal structures [8,9] indicate that the activation loop is too short to be autophosphorylated in cis, and solution studies support a trans mechanism for the activation loop [17]. Autophosphorylation of Tyr972 in the juxtamembrane region is also likely to occur in trans, for steric reasons, whereas autophospho-rylation of the C-terminal tyrosine sites could potentially occur in cis after trans-autophosphorylation of the activation loop.

Activation Loop Autoinhibition

A key mechanism by which the insulin receptor and other tyrosine and serine/threonine kinases regulate catalytic activity is through positioning of the activation loop [18]. The activation loop of the insulin receptor begins with a protein kinase-conserved 1150DFG motif and ends with tyrosine-kinase-conserved Pro1172. Between these conserved residues are three sites of tyrosine autophosphorylation: Tyr1158, Tyr1162, and Tyr1163. The activation loop in the crystal structure of the unphosphorylated, low-activity form of

Figure 3 Comparison of the activation loop conformations in unphosphorylated IRK (IRK-0P) [8] and tris-phosphorylated IRK (IRK-3P) [9]. The activation loop contains Asp1150, Tyr1158, Tyr1162, and Tyr1163. The catalytic loop contains Asp1132. The remainder of the protein in each case is represented by a molecular surface. For IRK-3P (night), the substrate peptide (containing Y(p)) and the ATP analog are shown, the latter of which is partially masked by the N-Terminal lobe of IRK-3P.

1RK-0P IRK-3P

Figure 3 Comparison of the activation loop conformations in unphosphorylated IRK (IRK-0P) [8] and tris-phosphorylated IRK (IRK-3P) [9]. The activation loop contains Asp1150, Tyr1158, Tyr1162, and Tyr1163. The catalytic loop contains Asp1132. The remainder of the protein in each case is represented by a molecular surface. For IRK-3P (night), the substrate peptide (containing Y(p)) and the ATP analog are shown, the latter of which is partially masked by the N-Terminal lobe of IRK-3P.

Figure 4 Properly configured active site in IRK for phosphoryl transfer derived from the ternary IRK-3P structure [9] with the following modifications: the Lys1030 side-chain rotamer and the ATP y-phosphate dihedral angle have been changed to coincide with those in the structure of ternary protein kinase A (PKA) [28]. Two active site Mg2+ ions are shown as spheres. Hydrogen bonds are represented as black dashed lines.

Figure 4 Properly configured active site in IRK for phosphoryl transfer derived from the ternary IRK-3P structure [9] with the following modifications: the Lys1030 side-chain rotamer and the ATP y-phosphate dihedral angle have been changed to coincide with those in the structure of ternary protein kinase A (PKA) [28]. Two active site Mg2+ ions are shown as spheres. Hydrogen bonds are represented as black dashed lines.

IRK (IRK-0P) traverses the cleft between the two kinase lobes, with Tyr1162 bound in the active site, hydrogen-bonded to conserved residues Asp1132 and Arg1136 of the catalytic loop (Figs. 3 and 4) [8]. This conformation of the activation loop obstructs the substrate and nucleotide binding sites via Tyr1162 (acting as a pseudosubstrate) and residues of the DFG motif, respectively. Solution studies of IRK indicate that in the presence of millimolar quantities of ATP (comparable to cellular levels), the activation loop is actually in equilibrium between inhibiting, gate-closed conformations, as represented by the IRK-0P crystal structure, and gate-open conformations in which the activation loop is displaced from the active site cleft [19].

The crystal structure of the tris-phosphorylated, activated form of IRK (IRK-3P), co-crystallized with a Mg-ATP analog and substrate peptide, reveals how autophosphorylation of the three activation loop tyrosines stabilizes a particular gate-open conformation that is optimal for catalysis (Figs. 3 and 4) [9,20]. More specifically, the autophosphorylation-induced rearrangement of the activation loop accomplishes four main mechanistic objectives: (1) release of steric constraints to Mg-ATP and substrate binding, (2) proper positioning of Asp1150 (DFG motif) for coordination of an active site Mg2+ ion, (3) proper positioning of Phe1151 (DFG motif) to facilitate lobe rotation for productive ATP binding, and (4) proper positioning of the end of the activation loop, which acts as a platform for substrate binding. The positions of residues in the catalytic loop remain virtually unchanged in the basal (unphosphorylated) and activated (tris-phosphorylated) states of IRK.

The activation loop conformation in the IRK-3P structure is stabilized by direct interactions between the phosphate groups of pTyr1162 and pTyr1163 and basic residues in the activation loop (Arg1164 and Arg1155, respectively) and by short, anti-parallel P-strand interactions between segments of the activation loop and the C-terminal lobe. These interactions are also observed in the structure of the related IGF-1 receptor kinase [21]. Interestingly, the phosphate group of pTyr1158 makes no contacts with other kinase residues and has been reported to bind a subset of Src homology 2(SH2) domain-containing proteins including SH2-By [22] and APS [23].

It has not been possible to crystallize IRK-0P with bound Mg-ATP analog and substrate peptide, presumably due to the high Km values for ATP and substrate peptide in the unphosphorylated state («1 and 2 mM, respectively) [24]. An approximate view of the initial, basal-state autophos-phorylation event was provided by the crystal structure of an IRK activation loop mutant, Asp1161 ^ Ala, with bound Mg-ATP analog [10]. This structure, in combination with biochemical studies [24,25], demonstrates that a single residue change can result in loss of pseudosubstrate autoinhibition by the activation loop. In the Asp1161 ^ Ala mutant, Km for ATP is decreased and kcat is increased to levels comparable to the tris-phosphorylated form [24], although Km for substrate peptide remains high.

The structure of this mutant with bound Mg-ATP analog shows that the activation loop is disengaged from the active site and mostly disordered. In the beginning of the activation loop, the residues of the DFG motif, particularly Phe1151, are not properly positioned for catalysis. Positioning of the Phe1151 side chain in a hydrophobic pocket below aC is critical for free rotation of the N-terminal lobe with respect to the C-terminal lobe, required for productive ATP binding, and for the independent (with respect to the N-terminal P-sheet) rotation of aC, which positions Glu1047 (aC) for salt-bridge formation with Lys1030 (P3) (Fig. 4) [10]. This structure indicates that despite release of activation loop autoinhibition (in this case through mutation), autophosphorylation of the activation loop tyrosines is still necessary to stabilize the loop in a configuration optimized for catalysis.

Diabetes 2

Diabetes 2

Diabetes is a disease that affects the way your body uses food. Normally, your body converts sugars, starches and other foods into a form of sugar called glucose. Your body uses glucose for fuel. The cells receive the glucose through the bloodstream. They then use insulin a hormone made by the pancreas to absorb the glucose, convert it into energy, and either use it or store it for later use. Learn more...

Get My Free Ebook


Post a comment