Reassessment of an Innovative Insulin Analogue Excludes Protracted Action yet Highlights the Distinction between External and Internal Diselenide Bridges

Abstract

Long-acting insulin analogues represent the most prescribed class of therapeutic proteins. An innovative design strategy was recently proposed: diselenide substitution of an external disulfide bridge. This approach exploited the distinctive physicochemical properties of selenocysteine (U). Relative to wild type (WT), Se-insulin[C7U^A , C7U^B ] was reported to be protected from proteolysis by insulin-degrading enzyme (IDE), predicting prolonged activity. Because of this strategy's novelty and potential clinical importance, we sought to validate these findings and test their therapeutic utility in an animal model of diabetes mellitus. Surprisingly, the analogue did not exhibit enhanced stability, and its susceptibility to cleavage by either IDE or a canonical serine protease (glutamyl endopeptidase Glu-C) was similar to WT. Moreover, the analogue's pharmacodynamic profile in rats was not prolonged relative to a rapid-acting clinical analogue (insulin lispro). Although [C7U^A , C7U^B ] does not confer protracted action, nonetheless its comparison to internal diselenide bridges promises to provide broad biophysical insight.

Keywords: chemical protein synthesis; insulin; oxidative protein folding; selenocysteine; selenoprotein.

Figure 1.

Structural features of insulin and IDE. a. 3-D structure of insulin contains α-helical subdomain stabilized by three disulfide bonds: A6–A11, A7–B7 and A20–B19 as indicated as sticks (PDB: 3w7y). b. Ribbon diagram of substrate free-IDE (PDB: 2jg4). c. Stereo view of IDE-insulin complex (PDB: 2wby). IDE-bound insulin molecule shown in pink.

Figure 2.

Structural characterization and thermodynamic stability of insulin analogues. a. CD spectra of Se-bovine insulin[C7U^A, C7U^B] (Se-Ins, blue) and bovine insulin (green). b. Guanidine denaturation assays of insulin analogues monitored by ellipticity at 222 nm; color code as in panel a. The resulting stability values are given in Table 1.

Figure 3.

Stability assays using insulin-degrading enzyme (IDE). a. Studies of human IDE. b. Studies of rat IDE; color code as in panel a Quantitative and statistical analyses are provided in SI Fig. S3.

Figure 4.

Receptor-binding assay for insulin receptor (IR)^[22b] for the Se-insulin analogue (●) compared to bovine ins (■). The two proteins present a similar binding affinity to IR with K_d = 130 ± 20 pM and 110 ± 10 pM respectively.

Figure 5.

Rat studies of insulin analogues. a. time course of [blood glucose] following SQ injection. Insulin lispro (■, n=15); bovine ins (▲, n=6); bovine Se-ins (●, n=6); diluent (◆, 100 μL, n=6). The dose of WT or modified insulin was 15 μg per 300-gram rat. b. time course of [blood glucose] following IV injection. Insulin lispro (n=20); bovine ins (n=12); bovine Se-ins (n=12); diluent (100 μL, n=5). This study used 10 μg insulin per 300-gram rat. (c) and (d), normalized blood glucose profiles for the data in panels a and b, respectively. Quantitative and statistical analyses are provided in SI Fig. S10–S12.

Scheme 1.

Synthetic strategy for making bovine Se-insulin[C7U^A, C7U^B] analogue. The control bovine WT insulin was synthesized adapting a similar strategy. DesDi represents Lys^B28-des-[B29,B30]-insulin; DOI – des-octapeptide[B23-B30]-bovine insulin, TFA – Trifluoroacetic acid.