Separation and characterization of trypsin and carboxypeptidase B-digested products of Met-Lys-human proinsulin

Insulin related compounds and identification

Insulin-related compounds (IRCs), which originate during the expression and purification of human insulin using recombinant Escherichia coli, were purified and identified. We investigated the identity of IRCs and their origin. We also presented methods for inhibiting IRC formation. The strains used in this report were E. coli B5K and E. coli H27R. E. coli B5K had a 6-amino acid-fused peptide at the N-terminus of proinsulin, and E. coli H27R had a 28-amino acid-fused peptide at the N-terminus of proinsulin. We investigated the identity of IRCs and their origin by mainly using High Performance Liquid Chromatography (HPLC). The well-known IRCs, desamido human insulin and desthreonine human insulin, formed in both strains. In addition to these two IRCs, the B5K strain produced three different IRCs, Arg(A(0))-insulin (IRC 1), prepeptide-insulin (IRC 2), and Glu(A(22))-insulin (IRC 3). The amounts of IRC 1, IRC 2, IRC 3 were approximately 0.1-0.3% after final purification step. Among these IRCs, Arg(A(0))-insulin, prepeptide-insulin, and desthreonine insulin originated from incomplete enzyme reaction. Glu(A(22))-insulin was formed when we used a double stop codon during the expression of preproinsulin; that is, it was formed by the misreading of the first stop codon through the amber mutation. The major IRCs of the H27R strain were human insulin fragment (B1-B21) (IRC 4), and A9(Ser→Asn) amino acid single mutation human insulin (IRC 5), Arg(B(31))-insulin (IRC 6). Human insulin fragment (B1-B21) was formed by β-mercaptoethanol, which was added during refolding. It formed when the disulfide bonds between A-chain and B-chain of human insulin were cut by β-mercaptoethanol, followed by cleavage of the B-chain by trypsin and carboxypeptidase B. A9(Ser→Asn) amino acid single mutation human insulin originated from the mistranslation of A9 serine, such that asparagine was translated instead of serine. Arg(B(31))-insulin originated from incomplete enzyme reaction. The amount of IRC 4 was 10-15% after enzyme reaction. The amounts of IRC 5, IRC 6 were around 0.2% after final purification step. We present methods for inhibiting the formation of IRCs by controlling the amount of enzyme, controlling the rate of enzyme reaction, using a single stop codon, using hydrogen peroxide (H(2)O(2)) to inhibit β-mercaptoethanol, and modifying the A9 codon.

Application of size-exclusion chromatography in the investigation of the in vitro stability of proinsulin and its cleaved metabolites in human serum and plasma

To help ensure reliability of proinsulin measurements and define the optimal matrix for conducting routine bioanalysis of this prognostic biomarker, we undertook a systematic evaluation of its in vitro stability. For this study, we subjected mono-radioiodinated forms of hPI and its cleaved metabolites to size-exclusion chromatography (FPLC-SEC employing a Superdex-75 10/30 HR column) to characterize their elution profiles following incubation in human serum and plasma. We determined that intact hPI is a substrate for serine-like protease(s) that are present in human serum. Furthermore, RIA analysis of the elution profile of unlabeled peptide demonstrated that the B-C junction is cleaved preferentially. Thus, in vitro degradation of hPI represents a potential pathway for the formation of cleaved metabolites. Our findings confirmed that EDTA plasma is the preferred matrix for quantitative determination of intact hPI and its cleaved metabolites. We concluded the SEC strategy employed in this study is broadly applicable to evaluating the in vitro stability of other peptides/proteins of diagnostic or therapeutic interest.