Total Chemical Synthesis of Palmitoyl-Conjugated Insulin

Abstract

Commercially available insulins are manufactured by recombinant methods for the treatment of diabetes. Long-acting insulin drugs (e.g., detemir and degludec) are obtained by fatty acid conjugation at LysB29 ε-amine of insulin via acid-amide coupling. There are three amine groups in insulin, and they all react with fatty acids in alkaline conditions. Due to the lack of selectivity, such conjugation reactions produce non-desired byproducts. We designed and chemically synthesized a novel thiol-insulin scaffold (CysB²⁹-insulin II), by replacing the Lys^B29 residue in insulin with the Cys^B29 residue. Then, we conjugated a fatty acid moiety (palmitic acid, C16) to CysB²⁹-insulin II by a highly efficient and selective thiol-maleimide conjugation reaction. We obtained the target peptide (palmitoyl-insulin) rapidly within 5 min without significant byproducts. The palmitoyl-insulin is shown to be structurally similar to insulin and biologically active both in vitro and in vivo. Importantly, unlike native insulin, palmitoyl-insulin is slow and long-acting.

Figure 1

(A) Chemical synthesis of palmitoyl-insulin. (a) Cys(tBu)^B29-insulin I was treated with TFMSA (10%) and anisole (5%) in TFA to obtain Cys^B29-insulin II. (b) 1-(2-Aminoethyl)maleimide was reacted with palmitic acid N-hydroxysuccinimide in DCM in the presence of DIEA to afford N-(2-(N-palmitoylaminoethyl)maleimide) III. (c) Lipidation of Cys^B29-insulin II was performed in DMF with collidine to generate palmitoyl-insulin IV. (B) HPLC profile of purified palmitoyl-insulin IV. (C) MALDI-TOF mass spectroscopy of purified palmitoyl-insulin IV.

Figure 2

Competition binding of (A) palmitoyl-insulin and insulin detemir against Eu-labeled insulin to insulin receptor isoform-B (IR-B) and (B) insulin, palmitoyl-insulin, and insulin detemir against the Eu-labeled IGF-1 to IGF-1 receptor. (C) Binding affinity in IC₅₀ values is shown in the table. Results are expressed as a ratio of binding percentage in the presence/absence of the competing ligand (%B/B₀). Curves are plotted from at least three separate experiments, where each data point was performed in triplicate. Error bars are shown when greater than the size of the symbols.

Figure 3

Blood glucose concentration in response to equimolar concentration of insulin (Actrapid), detemir, or palmitoyl-insulin administered IP (n = 10 per treatment). Data are presented as mean ± SEM. ^ap < 0.05 palmitoyl-insulin vs insulin (Actrapid); ^bp < 0.05 palmitoyl-insulin vs detemir; and ^cp < 0.05 insulin (Actrapid) vs detemir. (A) Absolute glucose levels over 180 min. Two-way ANOVA demonstrated significant effects of treatment, time, and an interaction between the two parameters on glucose response (p < 0.05). Tukey’s post-hoc test showed that palmitoyl-insulin maintained reduced glucose levels compared to insulin (Actrapid) from 120 to 180 min. (B) Change from baseline glucose levels with the same statistically significant effects established.

Figure 4

Blood glucose concentration in response to equimolar concentration of insulin (Actrapid), detemir, or palmitoyl-insulin administered SC (n = 5 per treatment). Data are presented as mean ± SEM. ^ap < 0.05 palmitoyl-insulin vs insulin (Actrapid); ^bp < 0.05 palmitoyl-insulin vs detemir; and ^cp < 0.05 insulin (Actrapid) vs detemir. (A) Absolute glucose levels over 180 min. Two-way ANOVA demonstrated significant effects of treatment, time, and an interaction between the two parameters on glucose response (p < 0.05). Tukey’s post-hoc test showed that palmitoyl-insulin did not achieve reduced glucose levels compared to insulin (Actrapid) or detemir. (B) Change from baseline glucose levels.

Figure 5

CD spectra of insulin and palmitoyl-insulin in 20% trifluoroethanol/water.