Insulin is an example of a protein that has a quaternary structure (i.e., it normally exists in a self-associated form rather than as a monomer). Insulin exists as a monomer only at a very low concentration (<0.1 |M, ~0.6 mg/ ml). At higher concentrations, insulin exists as a dimer. The dimers are believed to result from the hydrophobic association of the B23 to B28 regions on insulin monomers. In the presence of zinc ions and in the pH range 4.0 to 8.0, three dimers come together to form a hexamer. At concentrations of 2 mM and above, the hexamer is formed at neutral pH without the assistance of zinc ions. The insulin formulations on the market predominantly are either neutral solutions or suspensions of zinc insulin and exist primarily in hex-americ form104; insulin concentrations in blood are sufficiently low so that insulin circulates and brings about its biological effects as a monomer.105 However, there is no anomaly here because at low concentrations, insulin dissociates into monomers. Therefore, in solution insulin may exist as an equilibrium mixture of monomers, dimers, tetramers, hexamers, or higher associated states, the relative amounts dependent on factors such as concen tration of insulin, pH, nature and concentration of metal ions, purity, processing methods, storage temperature, or ionic strength.17106 Many of the recently introduced monomeric insulin analogues also exist in hexameric form in the formulation but dissociate into monomers more easily on administration. The diameters of the insulin monomer, dimer, tetramer, and hex-amer are approximately 30, 39, 50, and 59 A, respectively.
Chelation of zinc ions by EDTA has been reported to cause hexamers to deaggregate to dimers. Because three dimers result from dissociation of hexamer, the enzymatic degradation of insulin by a-chymotrypsin was enhanced threefold in the presence of EDTA.107 Similarly, sodium glycocho-late, a bile salt, may be capable of dissociating insulin oligomers to monomers, and this may partly explain the role of bile salts as enhancers of insulin bioavailability across mucosal barriers. In studies with zinc insulin (hexamers) and sodium insulin (dimers), the rate of degradation by a-chymot-rypsin in the presence of bile salts was increased by a factor of 5.4- and 2.1-fold, respectively. These values are close to the 6- and 2-fold increases that would be expected by the complete dissociation of hexamers and dimers to monomers.108
It may, however, be noted that covalent, higher molecular weight transformation products can result on long-term storage of commercial insulin preparations. Accelerated stability studies indicated that the main product is covalent insulin dimer, but in protamine-containing products, the formation of covalent insulin-protamine also takes place. At temperatures greater than 25°C, covalent oligomers and polymers can also form.109 Although the formation of such covalent aggregates is slower than the chemical decomposition of insulin, their presence may lead to immunological side effects.
Aggregation of zinc-free insulin as a function of concentration of sodium chloride (10 to 100 mM), pH value (7.5 to 10.5), and insulin concentration (1.8 to 13.4 mg/ml) has been investigated. At the lowest pH and the highest salt concentration (pH 7.5, 100 mM NaCl, 12 mg/ml insulin), the weight-average molar mass was close to that of a hexamer. In contrast, the weight-average molar mass was close to that of monomer at the highest pH and lowest salt concentration (pH 10.5, 10 mM NaCl, 1.9 mg/ml). Because the monomer carries two negative charges at pH 7.5 and six negative charges at pH 10.5, the results can be explained in terms of electrostatic repulsion between insulin monomers. As the pH is reduced, protein charge-charge repulsions are reduced, shifting the equilibrium toward oligomers. Similarly, increased ionic strength screens the charge repulsions, favoring the formation of oligomers. However, it may be noted that, in all cases, a distribution of oligomers is present with a relative Gaussian width of about 30%.110
Problems relating to aggregation of insulin result not from the dimers or hexamers (provided they are noncovalent), but from the formation of insoluble precipitates, often referred to as fibrillation of insulin. Formation of such insoluble aggregates is a major obstacle to the development of insulin infusion systems because the aggregates result in blockage of tubing, membranes, and pumps. Insulin aggregates can develop in both implantable and portable systems, and the reservoirs may have to be flushed or changed every 3 to 4 days.17 Obviously, this makes it difficult to develop implantable systems. Any irregularities in the tubings, such as the microscopic barbs and distortions produced while sectioning capillary bore polytetrafluoroethylene tubing, have been reported to induce aggregation of insulin. Surface irregularities such as these may also attract platelets from the bloodstream to the roughened regions, which may contribute to insulin aggregation.111
The mechanism of formation of insoluble aggregates in lyophilized insulin at an elevated temperature and a high humidity has been investigated. The aggregates were formed by noncovalent interactions and p-elimination of cystines, followed by thiol-catalyzed disulfide interchange in the solid state. Insulin can thus be stabilized by lowering the formation of thiols or transforming them chemically to nonreactive species. Controlling the humidity will also help to stabilize insulin in the solid state.112 Bovine insulin is more susceptible to fibrillation compared to porcine or human insulin.
In the electron microscope, insulin fibrils are seen as long fibers with a diameter of 10 to 50 nm. Although the exact mechanism of fibril formation is not known, the main driving force appears to be shielding of hydrophobic surfaces and possible formation of a p-sheet to further stabilize the fibrillar structure. It is believed that formation of fibril nuclei first requires the mono-merization of oligomeric insulin. This is because the hydrophobic interfaces of insulin monomers are normally buried in the dimer or the hexamer. By monomerization, these hydrophobic surfaces are exposed, allowing monomers to associate. Therefore, it seems logical that prevention of insulin dissociation will stabilize the molecule against fibrillation. As an example, this could be achieved by addition of surplus zinc ions, which will further stabilize the hexameric structure. Also, blockage of hydrophobic surfaces by addition of surfactant can also counter insulin fibrillation.104,113 A study has proposed that a partially folded intermediate is the precursor for association and eventually fibrillation.114
Marketed insulin preparations often show frosting, which is the formation of a finely divided precipitate on the walls of the containers. The process can be accelerated by the presence of a large headspace within the vial, suggesting that denaturation at the air-water interface is involved. Other factors that may contribute to frosting include zinc concentration, pH, and the presence or absence of additives.115 The frosting is most likely a macroscopic manifestation of insulin fibrillation and is observed more with human NPH insulin. Because human insulin contains more monomeric insulin than animal insulins, fibrillation is more likely in accordance with the mechanism of fibrillation discussed above.104
Chemical modification of insulin to produce sulfated insulin has also been reported to produce a nonaggregating insulin. The introduction of charged sulfate groups in the molecule disrupt the interaction between hydrophobic regions of insulin molecules, thereby preventing aggregation.116 Another approach to reducing the self-association of insulin is the development of a Lys, Pro analogue of human insulin,117 which is resistant to self-association. The development of this analogue and its pharmacokinetics, as well as other monomeric insulin analogues that have recently become available, are discussed in Chapters 1 and 6.
Was this article helpful?