Improved insulins for diabetics. Successful treatment of diabetes mellitus relies on the
continuous control of blood glucose levels. This is a major challenge, because the pharmacokinetics of traditional insulin preparations that are injected subcutaneously do not
match the physiological profiles of insulin secreted by the pancreas. In particular, there is a need to accelerate the rate of arrival of insulin into the blood from the site of injection, and to prolong its lifetime once it is there. The observation of a lag phase in the absorption of subcutaneously injected insulin and the fact that, in concentrated preparations, insulin self-associates to form dimers and hexamers, led to the idea that insulins with a reduced tendency to aggregate might be absorbed more rapidly. Selfassembly is not necessary for biological activity
because insulin binds to its receptor as a monomer. The surfaces that are used in this self-assembly were readily identified from the crystal structure of the insulin hexamer. Protein engineering was used successfully to disrupt these interfaces through mutations that either juxtapose negative charges, such as when proline 28 in the B chain (Pro B28) is replaced by an Asp residue, or by destabilizing the intermolecular β-sheet interactions that are formed by
the B-chain carboxyl termini, such as when ProB28 and LysB29 are exchanged. These fast-acting monomeric insulins retain biological activity and are now being used for therapy.