A convenient assay method for the quality control of peptides and proteins

Abstract

The development of a convenient and very accurate procedure with which to discriminate among subsets of structurally similar peptides and proteins, and measure enantiomeric purities with very good accuracy, has been described in a series of recent articles. A factor preventing its general application to all peptide forms is that comparisons were originally limited to closed subsets of structurally similar types, e.g., dipeptides, tripeptides, and insulin drug forms. In the most recent of these articles, a modification to the method was described which did enable the comparisons to be extended between sets, in particular the di-and tripeptides. That same modification is extended even further in this article to include additional di- and tripeptides, glycylglycine oligomers, insulin drug forms, and neuropeptides. The same principal component analysis treatment used for data reduction and statistical comparisons in prior work enables the discrimination among 49 of the total of 51 analytes investigated.

Clustering of CD spectral data as a prototype QSAR model for neuropeptides

“…13 Reagent grade D-histidine, reported to have an EP better than 99.8%, was obtained from Sigma Chemical Co. Reagent grade CuSO4‚5H2O was obtained from Fisher Scientific.…”

“…The motivation for this study and previous related studies on peptides and proteins [10][11][12][13] was to address the lack of emphasis that is given to causal relationships between chirality and function, especially since molecular chirality is so commonplace among natural and synthetic drug substances. Protein receptor sites are chiral.…”

“…13 Mathematical algorithms were derived with which to determine their chemical and enantiomeric purities. 12,13 The absorbance property required for CD to occur in the visible range, was provided by first binding the peptides to Cu(II) ion in pH 13 aqueous solution. Visible absorbance spectra of Cu(II) complexes are comprised of three electronic transitions 16 all of which are CD-active.…”

“…Major accomplishments from these articles that are of relevance to the goals of this one are the ability to discriminate between human and porcine insulins 12 (two 51 amino acid residue proteins whose sequences differ in the identity of one acid at the B30 terminal amino acid only) and between human insulin and the human LysPro variant form in which the order of the B29 L-lysine and B30 L-proline are reversed, and to achieve a level of quantitative analytical selectivity that approaches specificity among 51 peptides. 13 The specific objective of this article was to derive a cluster correlation diagram between visible range CD data on one hand and molecular stuctures, and their corresponding clinical therapeutic functions, on the other. Success would mean we will have a prototypical in vivo predictive QSAR model of some general applicability.…”

“…CD spectral data are taken from the previous companion article. 13 Only tripeptides and longer oligopeptides are included to test the general practicality of a clustering algorithm. Special mention is given to the subgroup of 19 neuropeptides, classified by their particular clinical functional actions.…”