Blood sugar levels for some diabetics can appear to imitate a roller-coaster ride. Tight regulation of blood glucose levels is vital to avoid the fatal long-term complications of diabetes. At present, injecting insulin frequently fails to provide such a degree of control. In the August issue of Nature Biotechnology, researchers at Eli Lilly report how they generated novel crystal forms of insulin, which can provide longer and smoother control of blood glucose.

In the 1930s, the concept of combining insulin with zinc and protamine (an arginine-rich basic protein) led to the development of the Neutral Protamine Hagedorn (NPH) insulin preparation, which gave improved sustained action for diabetics. NPH insulin, a microcrystalline suspension, is used today to provide medium-to long-term activity. However, the duration of action is too short to provide basal therapy with a single injection per day, and its pharmacodynamic profile exhibits a large peak. Brader et al. sought to retain the favourable properties of NPH insulin, yet correct its pharmacokinetic shortcomings, by manipulating the release rate of an NPH-like formulation.

The researchers co-crystallized human insulin (HI) with the less-soluble insulin derivative octanoyl-Nɛ-LysB29-HI (C8-HI) in the presence of zinc and protamine. By altering the ratio of the soluble HI and less-soluble C8-HI versions of insulin, they were able to build stable co-crystals that released insulin at a slower rate than standard formulations of NPH. Low crystal solubility in the interstitial fluid is essential for controlled delivery. In dogs with transient experimental diabetes, a single injection of the co-crystals provided a sustained control of blood glucose levels for 24 hours, with nearly ideal pharmacodynamics.

The lipophilic moiety does not interfere with the order of the hexamer building blocks of insulin crystals, and confers a proportionate decrease in aqueous solubility. Both HI and C8-HI are biologically active. The key to the superior release profile is the proportion of the octanoyl derivative relative to unmodified protein.

Could this strategy be applied to the delivery of other proteins? Because conjugation of a lipophilic moiety to the protein might interfere with its conformation and render co-crystallization and solubility difficult to control, it is likely that the success of this approach will be determined on a case-by-case basis. Other sophisticated drug delivery technologies, including hydrogels, liposomes and microspheres, provide a matrix from which a drug can be released in a controlled fashion. However, these complicated formulations often reduce the efficacy of the therapeutic protein or render mass production of the drug difficult. Tweaking the drug itself, by altering the composition of the protein crystal lattice without losing efficacy, might be a simpler and more effective alternative.