Yeast secretory expression of insulin precursors

Abstract

Since the 1980s, recombinant human insulin for the treatment of diabetes mellitus has been produced using either the yeast Saccharomyces cerevisiae or the prokaryote Escherichia coli. Here, development of the insulin secretory expression system in S. cerevisiae and its subsequent optimisation is described. Expression of proinsulin in S. cerevisiae does not result in efficient secretion of proinsulin or insulin. However, expression of a cDNA encoding a proinsulin-like molecule with deletion of threonine^B30 as a fusion protein with the S. cerevisiaeα-factor prepro-peptide (leader), followed either by replacement of the human proinsulin C-peptide with a small C-peptide (e.g. AAK), or by direct fusion of lysine^B29 to glycine^A1, results in the efficient secretion of folded single-chain proinsulin-like molecules to the culture supernatant. The secreted single-chain insulin precursor can then be purified and subsequently converted to human insulin by tryptic transpeptidation in organic–aqueous medium in the presence of a threonine ester. The leader confers secretory competence to the insulin precursor, and constructed (synthetic) leaders have been developed for efficient secretory expression of the insulin precursor in the yeasts S. cerevisiae and Pichia pastories. The Kex2 endoprotease, specific for dibasic sites, cleaves the leader-insulin precursor fusion protein in the late secretory pathway and the folded insulin precursor is secreted to the culture supernatant. However, the Kex2 endoprotease processing of the pro-peptide-insulin precursor fusion protein is incomplete and a significant part of the pro-peptide-insulin precursor fusion protein is secreted to the culture supernatant in a hyperglycosylated form. A spacer peptide localised between the leader and the insulin precursor has been developed to optimise Kex2 endoprotease processing and insulin precursor fermentation yield.