by Daniel F. Berenson, Allison R. Weiss and Michael A. Weiss

The engineering of insulin analogues to better treat diabetes mellitus represents a triumph of structure-based protein design. Based on classical structures of insulin hexamers, such engineering originally focused on destabilisation of the dimer interface. Pharmacokinetic effects of such destabilisation enabled development of rapid-acting formulations for use at mealtime and in pumps. Here, we review complementary efforts to stabilise the insulin hexamer and its higher-order self-assembly as a means to obtain protracted action and enhance basal glycaemic control. In particular, we describe a recently proposed novel approach based on pH-dependent zinc-mediated supramolecular assembly of insulin hexamers. Because this approach exploits a metal-ion-binding motif that is ubiquitous among diverse mammalian proteins, such assembly may find broad application as a nanobiotechnology.

Insulin and analogues

Elucidation of the crystal structure of insulin in 1969 represented a landmark in biology [1], [Figure 1]. This hexameric structure provided a framework for analysis of multiple biological issues, including the biosynthesis and storage of insulin in pancreatic β-cells. Despite insulin’s small size (51 amino acids), the hexamer exhibits key features of globular proteins in general: well-defined secondary structure, tertiary organisation about a hydrophobic core, formation of specific interfaces for self-assembly, and capacity for conformational change [2]. The crystal structure of the zinc insulin hexamer has also been central to the optimisation of its pharmaceutical formulation in the treatment of diabetes mellitus [3].

Rational principles of protein folding and assembly were first applied to the goal of accelerating hexamer disassembly to optimise pharmacokinetics [4]. Such efforts were based on the premise that more rapid disassembly in the subcutaneous depot would facilitate capillary absorption [5]. Multiple amino-acid substitutions were found at subunit interfaces that weakened or prevented the formation of dimers and hexamers. A subset of such substitutions was further found to be compatible with native biological activity and amenable to stable pharmaceutical formulation. There are three such rapid-acting analogues in current use [Table 1]. These products have been proven safe and effective in multi-injection regimens [6] and for use in insulin pumps [7]. The diversity of molecular changes conferring more rapid insulin absorption reflects the multiple ways that an elegantly crafted assembly may be rendered less stable with maintenance of biological activity.

How can protracted action be achieved [8]? The converse goal of targeted stabilisation of the insulin hexamer poses a more subtle challenge [3]. Insulin glargine (the active component of Lantus; Sanofi-Aventis) exploits a phenomenon that is robust to the details of molecular structure: isoelectric precipitation, a reversible transition to insolubility that classically occurs between pH 5 and 6 where wild-type insulin exhibits little or no net charge [9]. Raising insulin’s isoelectric point by addition of basic residues causes insulin to precipitate at physiologic pH, forming a subcutaneous depot with protracted absorption [10]. Insulin detemir (the active component of Levemir; Novo-Nordisk) contains a prosthetic fatty acyl group on LysB29, intended to mediate binding to serum albumin and hence provide a circulating depot [11]. The tethered moiety serendipitously also enhances the stability of hexamers of the modified insulin [12]. Thus, non-standard modification of the insulin molecule provides an opportunity to complement the structural strategies implicitly optimised by evolution.

A new approach to the engineering of a prolonged subcutaneous protein depot is based on principles of supramolecular chemistry. This approach exploits design features of proteins that are ubiquitous among intracellular proteins. Designated "zinc staples" [13], such supramolecular engineering offers three potential advantages: ease of manufacturability with standard microbiological processes, compatibility with current formulation chemistries and their exipients, and exploitation of the established mechanisms operative in Lantus and Levemir. Because the amino-acid substitutions in the stapled analogue lie within an α-helix that modulates the binding of insulin to the mitogenic IGF receptor[14], the analogue also exhibits enhanced selectivity for the insulin receptor and so may be associated with a reduced risk of carcinogenesis [15]. Zinc stapling promises to provide a general approach to the rational design of protein deposits as a therapeutic nanobiotechnology.

Bottom-up supramolecular assembly

Zinc stapling is a form of bottom-up supramolecular assembly that takes advantage of the intrinsic tendency of some materials to amalgamate via molecular self-recognition. Physico-chemical interactions that mediate molecular recognition at complementary surfaces include hydrogen bonding, donor-acceptor interactions, van der Waals forces, and the hydrophobic effect. Nature is rich with examples of systems that employ bottom-up self-assembly with exquisite precision. Archetypal supramolecular polymers include cytoskeletal microfilaments (actin) and microtubules (tubulin). The cell’s ability to regulate polymerisation and depolymerisation of these fibres provides a key mechanism underlying cell motility and mitosis. The growing database of biological structures and their nanoscale organisation provides a catalogue of what may in principle be achieved through bottom-up protein engineering.
Fine tuning of such assembly may be effected by ligand-directed switches. Complex chemical surfaces with switchable elements mediate hierarchical assembly in predictable ways. Indeed, diverse supramolecular structures are mediated by coordination with metal ions, especially divalent cations such as Zn²⁺ and Ca²⁺ [13] as exemplified by the assembly of virus capsids [16]. "Zinc zipper" adhesion in staphylococcal biofilms is a related form of supramolecular assembly that contributes to antibiotic resistance [13]. Metal-ion-coordinated supramolecular assembly has an important advantage: the ligand can act as a switch to alter the oligomeric state of a peptide or protein, thus permitting biological control through regulation of ion concentrations. This mechanism provides a convenient tool for nano-scale engineering.

Zinc finger motifs

Novel zinc-binding sites may be introduced into proteins [17] via "zinc finger" motifs [18]. Zinc fingers are ancient zinc-binding mini-domains that coordinate the Zn2+ through cysteine and histidine residues. Their widespread occurrence in eukaryotic genomes is likely a consequence of the high structural stability of the mini-domains as a combinatorial platform for sequence-specific DNA and RNA recognition. The physico-chemical principles of designed metal-ion-binding sites [17] has found recent application in the engineering of an insulin supramolecular assembly whose pharmacokinetic properties may be of potential therapeutic benefit. Such exploits the pH sensitivity of metal-ion binding (which required side-chain deprotonation).

"Zinc-stapled" insulin analogues

A portion of a zinc finger may be introduced into insulin by substituting histidine residues at positions (i, i+4) on the surface of an α-helix. Application of this scheme employed paired substitutions Glu^A4 His and Thr^A8 His. In such variant hexamers (a dimer of trimers) three insulin protomers were predicted to expose partial novel zinc-binding sites along one symmetry-related surface ("facing up") whereas the other three were predicted to expose such sites at the opposite surface ("facing down"). Each HX₃H element provides a half-site, requiring residues from an adjoining hexamer for completion of a tetrahedral metal-ion binding site. Thus, the half-binding sites on the top face of one hexamer could align with the half sites on the bottom face of the next hexamer in a stacked array of hexamers to form coordination sites. The predicted model is illustrated in schematic fashion in Figure 2A. In this scheme there are five zinc ions per hexamer: the two interior zinc ions coordinated by His^B10 at the trimer axis ("axial zinc ions) as in the classical Hodgkin structures [1] and three novel interfacial zinc ions between successive hexamers. In principle such a "zinc stapled" arrangement could lead to successive stacking of myriad insulin hexamers. The attractiveness of this potential model of supramolecular assembly in a subcutaneous depot was enhanced by the favourable receptor-binding properties of [HisA4, HisA8]-human insulin: native affinity for the lectin-purified insulin receptor with decreased cross-binding to the IGF receptor [13].

The design of a zinc-stapled insulin analogue as a supramolecular assembly was validated by X-ray crystallography [Figure 2B and 2C]. The crystal lattice indeed exhibited stacks of successive insulin hexamers bridged by interfacial zinc ions as originally envisaged. Details of the interfacial coordination differed from the predicted model. Whereas a tetrahedral binding site consisting of four imidazole rings was envisaged, the crystal structure revealed that this non-classical site actually consists of three His residues and one chloride anion, itself part of a water-linked network of hydrogen bonds at the protein surface. The tertiary structure, mode of hexamer assembly, and nature of the interior axial zinc coordination were nonetheless similar to the corresponding features of wild-type insulin hexamers. We speculate that the crystal lattice provides a model for the higher-order self-assembly of the variant hexamers on subcutaneous injection as a potential long-term depot of bioactive insulin.

The pharmacodynamic properties of [His^A4, His^A8]-human insulin were tested in male Lewis rats rendered diabetic by the β-cell poison steptozotocin. Such studies demonstrated native potency with prolonged action in accordance with the original design goals [13]. The analogue was formulated as a clear unbuffered solution at pH 4 as employed in the clinical formulation of insulin glargine [Lantus; Table 1A]. The expected protonation of HisA4, HisA8, and HisB10 under such conditions was predicted to preclude binding of zinc ions. Further, because the A4 and A8 substitutions also raise the isoelectric point (by removal of the negative charge of GluA4 and addition of the titratable imidazole groups), a change in pH to 7.4 in the subcutaneous depot was predicted to promote three concurrent processes: classical zinc-mediated assembly of hexamers, novel supramolecular zinc-stapled stacking of hexamers, and their isoelectric precipitation. Such extensive self-association and precipitation presumably underlies the protracted pharmacodynamics observed in the Lewis rats. The extent of prolongation is similar to that achieved by Lantus in this model. It is not known whether [HisA4, HisA8]-human insulin forms micro-crystals or amorphous precipitates in the subcutaneous depot.

A potential advantage of this novel insulin analogue relative to current basal insulin products is its superior receptor specificity. Patient anxiety has been raised by the increased mitogenicity of current basal insulin products [19] and potential association with cancer [20]. Since [HisA4, HisA8]-human insulin exhibits negligible affinity for the mitogenic IGF receptor [13], it is possible that this analogue would be safer than even human insulin on long-term use in patients with insulin resistance.

Concluding remarks

Engineered zinc binding sites provide a general strategy to insert a ligand-regulated switch to promote supramolecular assembly. Such design may have a wide-ranging impact in protein therapeutics. Modulation of the intrinsic self-assembly properties of proteins can circumvent the need for slow-release polymers in the optimisation of pharmacokinetic properties. Therapeutic applications of designed proteins are poised to grow exponentially.

Characterisation of the structure and function of proteins provides only a starting point for clinical translation. Formulations must be stable and provide appropriate pharmacokinetic and pharmacodynamic properties [21]. We envisage the subcutaneous protein depot as a novel biomaterial that is tractable by fundamental physico-chemical principles. Rational design of a zinc-stapled insulin analogue exemplifies a general approach to a ligand-regulated switch between a soluble injectable state and an insoluble depot state. Such a strategy may in principle be combined with other engineering goals, such as enhanced resistance to degradation. Such protein-based nanobiotechnology promises to have broad clinical impact in the coming decade.

References

1. Adams MJ et al. Nature 1969; 224: 491-495

2. Baker EN et al. Philos Trans R Soc Lond B Biol.Sci 1988; 319: 369-456

3. Dodson GG et al. Biochem Soc Trans 1993; 21: 609-614

4. Brems DN et al. Protein Eng 1992; 5: 527-533

5. DeFelippis MR et al. Crit Rev Ther Drug Carrier Syst 2001; 18: 201-264

6. Garg S et al. Endocr Pract 2010; 16: 486-505

7. Colquitt J et al. Diab Med 2003; 20: 863-866

8. Pollock RF. et al. Curr Diab Rev 2011; 7: 61-64

9. Goykhman S et al. Expert Opin Pharmacother 2009; 10: 705-718

10. Heinemann L et al. Diab Care 2000; 23: 644-649

11. Havelund S et al. Pharm Res 2004; 21: 1498-1504

12. Whittingham JL et al. Biochemistry 2004; 43: 5987-5995

13. Phillips NB et al. J Biol Chem 2010; 285: 11755-11759

14. Zhao M et al. J Biol Chem 2009; 284: 32178-32187

15. Werner H et al. Expert Opin Biol Ther 2011; 11: 199-209

16. Caspar DL and Namba K. Adv Biophys 1990; 26: 157-185

17. Regan L. Annu Rev Biophys Biomol Struct 1993; 22: 257-281

18. Berg JM. Proc Natl Acad Sci USA 1988; 85: 99-102

19. Zib I. and Raskin P. Diab Obes Metab 2006; 8: 611-620

20. Dawson LK and Hamilton LA. Am J Health Syst Pharm 2010; 67: 2025-2031

21. Leader B et al. Nat Rev Drug Discov 2008; 7: 21-39

22. Derewenda U et al. Br Med Bull 1989; 45: 4-18

The authors

Daniel F. Berenson ¹, Allison R. Weiss ², and Michael A. Weiss ³*

¹ Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, CT 06511, USA

² National Institute of Diabetes, Digestive, and Kidney Diseases, National Institutes of Health, 10 Center Drive (RM 7C432B) Bethesda, MD 20814, USA

³ Departments of Biochemistry, Biomedical Engineering and Medicine, Case Western Reserve University School of Medicine
10900 Euclid Avenue, Cleveland, OH 44106, USA

*For correspondence:

e-mail: michael.weiss@case.edu