Fastacting Insulins

Biostructure-based drug design is not limited to design of low-molecular weight compounds based on knowledge of the structure of their biological targets. In the following text we are presenting an example on biostructure-based design of macromolecular drug molecules, i.e., insulin analogs.

(A) '"^f (B)

FIGURE 2.11 Structures of the ligand-binding core of the ionotropic glutamate receptor GluR2. (A) The open, unbound form of GluR2 (pdb-code 1FTO). (B) The NeuroSearch compound NS1209 stabilizes GluR2 in the open form (pdb-code 2CMO). (C) The endogenous ligand glutamate introduces domain closure of GluR2 by a "Venus flytrap" mechanism (pdb-code 1FTJ). (D) Various synthetic agonists (here Br-HIBO) also introduce domain closure in GluR2 (pdb-code 1M5C).

FIGURE 2.11 Structures of the ligand-binding core of the ionotropic glutamate receptor GluR2. (A) The open, unbound form of GluR2 (pdb-code 1FTO). (B) The NeuroSearch compound NS1209 stabilizes GluR2 in the open form (pdb-code 2CMO). (C) The endogenous ligand glutamate introduces domain closure of GluR2 by a "Venus flytrap" mechanism (pdb-code 1FTJ). (D) Various synthetic agonists (here Br-HIBO) also introduce domain closure in GluR2 (pdb-code 1M5C).

This design was made possible only by a detailed insight into the structure of insulin and the intermolecular interactions between the insulin molecules in the crystalline phase.

Insulin is a hormone produced in the pancreas and it is responsible for the regulation of glucose uptake and storage. Insulin is most often associated with diabetes mellitus, which is a disease causing hyperglycemia. Healthy people have a basal level of insulin in the bloodstream, but in response to intake of food or to cover glucose clearance from the blood, peaks of larger insulin concentrations appear throughout the 24 h of a day. Patients with diabetes may have difficulties in maintaining the proper insulin concentrations, basal as well as peak concentrations, and accordingly regulation of their insulin level is essential. Type I diabetes patients need insulin to supplement endogenously produced insulin, whereas Type II diabetes patients often are getting insulin in order to improve glycemic control.

FIGURE 2.12 Chemical structures of glutamate, the agonist Br-HIBO, and the antagonist NS1209.

FIGURE 2.12 Chemical structures of glutamate, the agonist Br-HIBO, and the antagonist NS1209.

Glutamate

Br-HIBO

NS1209

Glutamate

Br-HIBO

NS1209

Insulin has been available for treatment of diabetes since 1923, and the major form for administration is still by subcutaneous injection. Insulin therapy typically involves multiple doses of different forms of insulin to maintain near-physiological insulin (and thereby glucose) levels. Long-acting insulin maintains the basal insulin level over 24 h with a single administration, and fast- and short-acting insulin analogs, which are instantaneously absorbed, are used to meet the insulin requirements associated with food intake. Thus, the development of insulin formulations with tailored properties, e.g., a prolonged effect or a faster onset, has always had a high priority in insulin research.

With the biosynthesis of recombinant human insulin in the 1980s it became possible to optimize the insulin therapy by designing insulin analogs with optimal pharmacokinetic properties by changing the amino acid sequence of the insulin molecule. A rational design of insulin analogs was only possible because a large number of insulin structures were determined experimentally by NMR spectroscopy and x-ray crystallography. Insulin was one of the first proteins whose 3D structure was determined by x-ray crystallography, and today more than 200 insulin structures are available in the Protein Data Bank.

Insulin exists in the crystalline phase as hexamers, dimers, and monomers. Actually, several forms of hexamers, T6, T3R3, and R6, exist depending on the presence of zinc ions and phenol (Figure 2.13). The presence of several hexameric forms of insulin reflects that even in the crystalline form insulin exerts some kind of conformational flexibility. Both zinc ions and phenol are stabilizing the hexameric form of insulin and are accordingly added to insulin formulations in order to improve their stability. After subcutaneous injection the insulin hexamer dissociates to dimeric insulin and by further dissociation to monomeric insulin, which represents the bioactive form. Thus, from the beginning it was believed that shifting the equilibrium toward the monomeric form would lead to faster-acting insulins, whereas stabilization of the hexameric form would lead to longer-acting insulins.

The insulin molecule consists of two peptide chains, an A-chain of 21 residues and a B-chain of 31 residues. The A- and B-chains are connected by two disulfide bridges linking A7-B7 and A20-B19. With the introduction of recombinant DNA techniques the residues involved in interaction with the insulin receptor or involved in the hexamer vs. dimer stabilization were identified.

From the x-ray structure of hexamer insulin it was evident that the side chain of the HisB10 residue was involved in zinc binding and thereby in stabilizing the hexamer (Figure 2.13). Mutation of the B10 residue from His to Asp yielded an insulin analog being absorbed twice as rapidly as normal insulin. Unfortunately, this analog turned out to be mitogenic and thus not suitable for clinical use.

Based on various structural studies of insulin it could be concluded that the flexibility of the C-terminus is crucial for the binding of insulin to its receptor (Figure 2.14). It also became evident that the B24-B26 residues are stabilizing the dimer by making an intermolecular antiparallel

FIGURE 2.13 (A) Insulin R6 hexamer (pdb-code 1EV6) showing the threefold symmetry of the three dimers (red-green, cyan-orange, and magenta-blue). (B) Three HisB10 residues coordinate to a zinc ion.

FIGURE 2.14 (A) Human insulin (pdb-code 1ZNJ, green) and insulin aspart (pdb-code 1ZEG, yellow), showing the change in conformation of the C-terminus (marked by an asterisk) by the mutation ProB28Asp. (B) For human insulin ProB28 and for insulin aspart AspB28 and the neighboring LysB29 are shown as stick models. Heteroatoms are colored red and blue for oxygen and nitrogen, respectively.

FIGURE 2.14 (A) Human insulin (pdb-code 1ZNJ, green) and insulin aspart (pdb-code 1ZEG, yellow), showing the change in conformation of the C-terminus (marked by an asterisk) by the mutation ProB28Asp. (B) For human insulin ProB28 and for insulin aspart AspB28 and the neighboring LysB29 are shown as stick models. Heteroatoms are colored red and blue for oxygen and nitrogen, respectively.

beta-sheet and at the same time burying the nonpolar side chains. Removal of the B24-B30 residues yields monomeric insulin and eliminates the association of monomers into dimers and hexamers. Furthermore, molecular modeling studies showed that ProB28 was important for dimer formation, but apparently not involved in receptor binding. Mutation of B28 from Pro to Lys or Asp has a profound effect on dimerization. The ProB28Asp mutant is presently marketed as rapid-acting insulin named aspart (Novorapid).

The knowledge of the interactions, which stabilize the hexamer and dimer, has made it possible to engineer the properties of insulin to yield therapeutic insulin analogs with tailored properties. Homology modeling indicated that inversion of the residues B28 and B29 should give faster acting insulins (Figure 2.15). The double mutant, ProB28Lys + LysB29Pro, indeed had a faster onset than normal insulin and it was the first genetically engineered insulin analog to become available for clinical use. Generally, modifications of B29 have a less pronounced effect on dimerization than modification of ProB28. The double mutant is especially interesting because it has the same isoelectric point (pI) as human insulin as their amino acid compositions are identical, and thereby the same solubility. The double mutant called insulin lispro is marketed under the brand name Humalog.

Elements Earth Wollf

FIGURE 2.15 (A) Human insulin (pdb-code 1ZNJ, green) and insulin lispro (pdb-code 1LPH, yellow), showing the change in conformation of the C-terminus (marked by an asterisk). (B) The ProB28-LysB29 in human insulin and the LysB28-ProB29 in insulin lispro are shown as stick models. Heteroatoms are colored red and blue for oxygen and nitrogen, respectively. The side-chain nitrogen atom in LysB29 could not be observed in the x-ray structure.

FIGURE 2.15 (A) Human insulin (pdb-code 1ZNJ, green) and insulin lispro (pdb-code 1LPH, yellow), showing the change in conformation of the C-terminus (marked by an asterisk). (B) The ProB28-LysB29 in human insulin and the LysB28-ProB29 in insulin lispro are shown as stick models. Heteroatoms are colored red and blue for oxygen and nitrogen, respectively. The side-chain nitrogen atom in LysB29 could not be observed in the x-ray structure.

Insulin analogs with changes of ProB28, as the above-mentioned insulin analogs aspart and lispro, primarily exist in the monomeric form, which is known to be more vulnerable for unfolding and subsequent fibril formation. To avoid this problem, mutations were limited to the neighboring LysB29 and to residues in the N-terminus of the B-chain. It was already known that changes of the B1-B8 residues primarily affected the stabilization of the dimer form of insulin. The basic LysB29 was exchanged for the acidic, polar, and hydrophilic Glu, and the neutral, polar, and hydrophobic AsnB3 was exchanged for the basic, polar, and hydrophilic Lys. This double mutant has a slightly lower pi of 5.1 compared with human insulin with a pi of 5.5, and accordingly its solubility is also enhanced. The double mutant, AsnB3Lys + LysB29Glu, is especially interesting because higherorder forms dominate relative to the monomeric form, i.e., it is more stable, while still maintaining rapid dissociation to monomeric insulin. This double mutant is normally referred to as insulin glu-lisine and is marketed as Apidra.

Thus, the design of rapidly absorbed, fast-acting insulin analogs must be characterized as a clear success. The design of the above-mentioned insulin analogs were made possible because the large number of insulin structures provided the researchers with a detailed information about the molecular interactions responsible for receptor binding and the hexamer-dimer-monomer equilibrium.

Diabetes Sustenance

Diabetes Sustenance

Get All The Support And Guidance You Need To Be A Success At Dealing With Diabetes The Healthy Way. This Book Is One Of The Most Valuable Resources In The World When It Comes To Learning How Nutritional Supplements Can Control Sugar Levels.

Get My Free Ebook


Post a comment