The physiological significance of the presence of GLUT2 at the food‐facing pole of intestinal cells is addressed by a study of fructose absorption in GLUT2‐null and control mice submitted to different sugar diets. Confocal microscopy localization, protein and mRNA abundance, as well as tissue and membrane vesicle uptakes of fructose were assayed. GLUT2 was located in the basolateral membrane of mice fed a meal devoid of sugar or containing complex carbohydrates. In addition, the ingestion of a simple sugar meal promoted the massive recruitment of GLUT2 to the food‐facing membrane. Fructose uptake in brush‐border membrane vesicles from GLUT2‐null mice was half that of wild‐type mice and was similar to the cytochalasin B‐insensitive component, i.e. GLUT5‐mediated uptake. A 5 day consumption of sugar‐rich diets increased fructose uptake fivefold in wild‐type tissue rings when it only doubled in GLUT2‐null tissue. GLUT5 was estimated to contribute to 100 % of total uptake in wild‐type mice fed low‐sugar diets, falling to 60 and 40 % with glucose and fructose diets respectively; the complement was ensured by GLUT2 activity. The results indicate that basal sugar uptake is mediated by the resident food‐facing SGLT1 and GLUT5 transporters, whose mRNA abundances double in long‐term dietary adaptation. We also observe that a large improvement of intestinal absorption is promoted by the transient recruitment of food‐facing GLUT2, induced by the ingestion of a simple‐sugar meal. Thus, GLUT2 and GLUT5 could exert complementary roles in adapting the absorption capacity of the intestine to occasional or repeated loads of dietary sugars.

In intestinal cells, levels of the fructose transporter GLUT5 are increased by glucose and to a greater extent by fructose. We investigated the mechanism by which fructose increases GLUT5 expression. In Caco-2 cells, fructose and glucose increased activity of the -2500/+41 GLUT5 promoter to the same extent. cAMP also activated the GLUT5 promoter. However, if a protein kinase A inhibitor was used to block cAMP signalling, extensive GLUT5 mRNA degradation was observed, with no change in basal transcription levels demonstrating the involvement of cAMP in GLUT5 mRNA stability. Indeed, the half-life of GLUT5 mRNA was correlated ( R2=0.9913) with cellular cAMP levels. Fructose increased cAMP concentration more than glucose, accounting for the stronger effect of fructose when compared with that of glucose on GLUT5 production. We identified several complexes between GLUT5 3'-UTR RNA (where UTR stands for untranslated region) and cytosolic proteins that might participate in mRNA processing. Strong binding of a 140 kDa complex I was observed in sugar-deprived cells, with levels of binding lower in the presence of fructose and glucose by factors of 12 and 6 respectively. This may account for differences in the effects of fructose and glucose. In contrast, the amounts of two complexes of 96 and 48 kDa increased equally after stimulation with either glucose or fructose. Finally, PABP (polyadenylated-binding protein)-interacting protein 2, a destabilizing partner of PABP, was identified as a component of GLUT5 3'-UTR RNA-protein complexes. We conclude that the post-transcriptional regulation of GLUT5 by fructose involves increases in mRNA stability mediated by the cAMP pathway and Paip2 (PABP-interacting protein 2) binding.