Translating glucose tolerance data from mice to humans: Insights from stable isotope labelled glucose tolerance tests

Abstract

Objective: The glucose tolerance test (GTT) is widely used in human and animal biomedical and pharmaceutical research. Despite its prevalent use, particularly in mouse metabolic phenotyping, to the best of our knowledge we are not aware of any studies that have attempted to qualitatively compare the metabolic events during a GTT in mice with those performed in humans.

Methods: Stable isotope labelled oral glucose tolerance tests (siOGTTs; [6,6-²H₂]glucose) were performed in both human and mouse cohorts to provide greater resolution into postprandial glucose kinetics. The siOGTT allows for the partitioning of circulating glucose into that derived from exogenous and endogenous sources. Young adults spanning the spectrum of normal glucose tolerance (n = 221), impaired fasting (n = 14), and impaired glucose tolerance (n = 19) underwent a 75g siOGTT, whereas a 50 mg siOGTT was performed on chow (n = 43) and high-fat high-sucrose fed C57Bl6 male mice (n = 46).

Results: During the siOGTT in humans, there is a long period (>3hr) of glucose absorption and, accordingly, a large, sustained insulin response and robust suppression of lipolysis and endogenous glucose production (EGP), even in the presence of glucose intolerance. In contrast, mice appear to be highly reliant on glucose effectiveness to clear exogenous glucose and experience only modest, transient insulin responses with little, if any, suppression of EGP. In addition to the impaired stimulation of glucose uptake, mice with the worst glucose tolerance appear to have a paradoxical and persistent rise in EGP during the OGTT, likely related to handling stress.

Conclusions: The metabolic response to the OGTT in mice and humans is highly divergent. The potential reasons for these differences and their impact on the interpretation of mouse glucose tolerance data and their translation to humans are discussed.

Keywords: Endogenous glucose production; Human; Mouse; OGTT; Stable isotope.

Figure 1

In humans, glucose (A), insulin (B), FFA (C), exogenous glucose enrichment (D), exogenous glucose concentrations (E) and endogenous glucose concentrations (F) during the OGTT in individuals with NGT (n = 221), IFG (n = 14) and IGT (n = 19). Data are mean ± SEM. Data were analysed by two-way repeated measures ANOVA. Significant (P < 0.05) main effects for time and group as well as time × group interactions were found for all data. ∗P < 0.05 vs. NGT; †P < 0.05 vs. IGT; ‡P < 0.05 vs. IFG.

Figure 2

In mice, glucose (A), glucose iAUC (B), insulin (C), FFA (D), exogenous glucose enrichment (E), exogenous glucose concentrations (F), endogenous glucose concentrations (G) and the change in endogenous glucose concentration from basal (H) during the OGTT in chow (n = 43) and HFHS (n = 46) fed conditions. Individual data (thin blue and red lines) and the mean (thick blue and red lines) are presented in A and C–I. Individual data and mean ± SEM are presented in B. Data in A and C–I were analysed by two-way repeated measures ANOVA. Significant (P < 0.05) main effects for time and group as well as time × group interactions were found for all data. Data in B were analysed by an independent t-test. ∗P < 0.001 vs HFHS.

Figure 3

In mice, glucose (A & B), exogenous glucose enrichment (C & D), exogenous glucose concentrations (E & F), endogenous glucose concentrations (G & H), change in endogenous glucose concentration from basal (I & J), insulin (K & L) and FFA (M & N) during the OGTT in chow (A, C, E, G, I, K, M) and HFHS (B, D, F, H, J, L, N) fed conditions with the lowest and highest glucose iAUC. N = 4 for each group. Data are mean ± SEM. Data were analysed by two-way repeated measures ANOVA. Significant (P < 0.05) main effects for time and group as well as time × group interactions were found for data in A, B, E, F, and H. Significant (P < 0.05) main effects for time and time × group interactions were found for data in G and L. Significant (P < 0.05) main effects for time and group were found for data in C. Significant (P < 0.05) main effects for time were found for data in D, K, and M. ∗P < 0.05 vs low; ∗∗P < 0.01 vs low.

Figure 4

In mice, glucose (A), insulin (B) and FFA (C) concentrations during a sham water gavage (n = 8). Data are mean ± SEM. Data were analysed by one-way repeated measures ANOVA. The ANOVA for the glucose data in (A) was statistically significant (P = 0.02). ∗P < 0.05 vs. 0 min; ∗∗P < 0.0001 vs. 0 min.