Metformin Temporal and Localized Effects on Gut Glucose Metabolism Assessed Using 18F-FDG PET in Mice

In the course of metformin treatment, staging abdominal cancer lesions with F-18-FDG PET images is often hindered by the presence of a high bowel radioactivity. The present study aimed to verify the mechanism underlying this phenomenon. Methods: Fifty-three mice were submitted to dynamic acquisitions of F-18-FDG kinetics under fasting conditions. Three small-animal PET scans were obtained over a 4-mo study period. The animals were subdivided into 4 groups according to the following metformin administration protocol: group 1, untreated mice (n = 15); group 2, mice exposed to metformin treatment (750 mg/kg/d) for the 48 h before each PET study (pulsed, n = 10); group 3, mice treated for the whole study period (prolonged, n = 10); and group 4, mice in which prolonged treatment was interrupted 48 h before PET (interrupted, n = 8). The rate constant of F-18-FDG uptake was estimated by Patlak analysis. At the end of the study, the ileum and colon were harvested, washed, and counted ex vivo. Two further groups, of 5 animals each, were included to evaluate the effect of prolonged metformin treatment on phosphorylated adenosine monophosphate (AMP)-activated protein kinase (pAMPK) form and gene expression for thioredoxin-interacting protein (TXNIP). Results: Pulsed treatment did not modify gut tracer retention with respect to the untreated group. Conversely, prolonged treatment induced a progressive increase in F-18-FDG uptake that selectively involved the colonic wall, without any significant contamination of bowel content. This effect persisted after a complete drug washout in the interrupted group. These responses were paralleled by increased pAMPK availability and by reduced expression of TXNIP messenger RNA in colonic enterocytes exposed to prolonged metformin treatment. Conclusion: Metformin causes a selective increase in colonic F-18-FDG uptake. This effect appears after a relatively long period of treatment and persists soon after drug washout. Accordingly, the increased bowel glucose metabolism reflects a biologic response to chronic metformin treatment characterized by increased levels of pAMPK and reduced levels of TXNIP.