Stability of glucagon-like peptide 1 and glucagon in human plasma

Abstract

To investigate the stability of glucagon-like peptide 1 (GLP-1) and glucagon in plasma under short- and long-term storage conditions. Pooled human plasma (n=20), to which a dipeptidyl peptidase 4 (DPP4) inhibitor and aprotinin were added, was spiked with synthetic GLP-1 (intact, 7-36NH2 as well as the primary metabolite, GLP-1 9-36NH2) or glucagon. Peptide recoveries were measured in samples kept for 1 and 3 h at room temperature or on ice, treated with various enzyme inhibitors, after up to three thawing-refreezing cycles, and after storage at -20 and -80 °C for up to 1 year. Recoveries were unaffected by freezing cycles or if plasma was stored on ice for up to 3 h, but were impaired when samples stood at RT for more than 1 h. Recovery of intact GLP-1 increased by addition of a DPP4 inhibitor (no ice), but was not further improved by neutral endopeptidase 24.11 inhibitor or an inhibitor cocktail. GLP-1, but not glucagon, was stable for at least 1 year. Surprisingly, the recovery of glucagon was reduced by almost 50% by freezing compared with immediate analysis, regardless of storage time. Plasma handling procedures can significantly influence results of subsequent hormone analysis. Our data support addition of DPP4 inhibitor for GLP-1 measurement as well as cooling on ice of both GLP-1 and glucagon. Freeze-thaw cycles did not significantly affect stability of GLP-1 or glucagon. Long-term storage may affect glucagon levels regardless of storage temperature and results should be interpreted with caution.

Keywords: degradation; glucagon; glucagon-like peptide-1; samples handling; storage.

© 2015 The authors.

Figures

Figure 1

The impact of storage conditions with or without ice for 1 and 3 h, respectively, on stability of added (10 and 40 pmol/l) GLP-1 and glucagon in human plasma to which DPP4 inhibitor and aprotinin were added. (A) 7–36NH2 GLP-1 isoform spiked in human plasma and measured with a C-terminal-specific RIA. (B) 7–36NH2 GLP-1 isoform spiked in human plasma and measured with a sandwich ELISA specific for intact GLP-1. (C) 9–36NH2 GLP-1 isoform spiked in human plasma and measured with a C-terminal-specific RIA. (D) Glucagon spiked in human plasma and measured with a C-terminal RIA. (E) Glucagon spiked in human plasma and measured with an N-terminal RIA. Each result represents mean±s.d. of six replicated determinations of GLP-1 and glucagon measured with different assays and normalized to results measured after immediate extraction. *P<0.05 tested by one-way ANOVA for repeated measurement with post hoc Bonferroni correction. Standard curve CV ranging from 4 to 10%.

Figure 2

(A) DPP4 activity in human plasma on ice (left) or at RT (right) with (grey bars) or without (white bars) addition of DPP4 inhibitor (valine pyrrolidide 0.01 mM). *P<0.05 between treatment and #P<0.05 between samples on ice and RT tested with a Student's paired t-test. Each result represents mean±s.d. of eight replicated determinations. (B) Enzyme inhibitors illustrates the recovery (%) of intact GLP-1 in human plasma (left part) or in buffer (right part) treated with respectively DPP4 inhibitor (black), DPP4 inhibitor+NEP 24.11 inhibitor (grey), and an inhibitor cocktail (grey with dots). No significant changes in recovery between treatments but as expected between human plasma and buffer (matrix effect and solvent extraction of human plasma). Samples were kept at RT for 1 h. Similar data were obtained when samples were kept on ice (data not shown). Each result represents mean±s.d. of eight replicated determinations. Standard curve CV ranging from 5 to 9%.

Figure 3

The impact of refreezing cycles on recovery of added (10 and 40 pmol/l) GLP-1 and glucagon in human plasma. (A) 7–36NH2 GLP-1 isoform spiked in human plasma and measured with a C-terminal-specific RIA. (B) 7–36NH2 GLP-1 isoform spiked in human plasma and measured with a sandwich ELISA specific for intact GLP-1. (C) 9–36NH2 GLP-1 isoform spiked in human plasma and measured with a with a C-terminal-specific RIA. (D) Glucagon spiked in human plasma and measured with a C-terminal-specific RIA. (E) Glucagon spiked in human plasma and measured with a N-terminal-specific RIA. Each result represents mean±s.d. of six replicated determinations of GLP-1 and glucagon measured with different assays and normalized to samples extracted immediately after peptide addition. Standard curve CV ranging from 4 to 10%.

Figure 4

The impact of long-term storage at −20 °C (black circle) and at −80 °C (black square), respectively, on the recovery of added (40 pmol/l) GLP-1 and glucagon in human plasma. (A) 7–36NH2 GLP-1 isoform spiked in human plasma and measured with a sandwich ELISA specific for intact GLP-1. (B) Glucagon spiked in human plasma and measured with a C-terminal-specific RIA. (C) Glucagon spiked in human plasma and measured with a N-terminal-specific RIA. Each result represents mean±s.d. of eight replicated determinations of GLP-1 and glucagon measured with different assays and normalized to result from samples extracted immediately after addition. *P<0.05 tested by one-way ANOVA for repeated measurement with post hoc Bonferroni correction. 10 pmol/l data not shown but was similar. Standard curve CV ranging from 6 to 13%.