A dual role of the N-terminal FQQI motif in GLUT4 trafficking

Background and aims: Insulin stimulation of muscle and adipose cells results in a rapid and reversible translocation of GLUT4 from storage vesicles to the plasma membrane. A N-terminal phenylalanine-containing amino acid motif (F5QQI) is required for the endocytosis and subsequent sorting of GLUT4 into the vesicles. The F5A-mutant of GLUT4 accumulates at the cell-surface in an insulin-independent manner. In previous experiments, we identified the FQQI-motif as binding partner for the heterotetrameric adapter protein complexes AP-1 and AP-2. Our data suggested that the F5-motif is involved in endocytosis of the transporter via an AP-2-mediated-pathway. Analysis of the F5A-GLUT4 recycling kinetics indicated a role of the phenylalanine motif and AP-1 in a post-endosomal sorting step. In order to prove an involvement of AP-1 in GLUT4 recycling, we analyzed GLUT4 targeting in adipose cells lacking either AP-1 or AP-2 respectively.

Material and methods: We used N-terminal GST-GLUT4 fusion proteins (F5, F5A, F5Y) in GST-pulldown assays to map the GLUT4/µ1-interaction with in vitro translated µ1-adaptin. We established a lentivirus-expression system to introduce an epitope-tagged HA-GLUT4-GFP reporter into 3T3-L1 adipocytes. To identify the precise role of the AP's in GLUT4 recycling, we generated AP-1 and AP-2 knockdown cells by RNA interference (RNAi). The subcellular localization of HA-GLUT4-GFP and endogenous GLUT4 was analyzed by confocal microscopy and biochemical methods.

Results: We could confirm the interaction of the N-terminal phenylalanine motif of GLUT4 and µ1-adaptin on the protein level. Efficiency of the lentiviral transfection of 3T3-L1 cells with the GLUT4 construct was 80%. The HA-tagged GLUT4-GFP is translocated to the plasma membrane after insulin stimulation. Furthermore, we were successful in designing siRNA oligonucleotides for specific knockdown of mouse AP-1 and AP-2 complexes, respectively. The analysis of the HA-GLUT4-GFP reporter in AP-1 knockdown adipocytes revealed that AP-1 depletion leads to a substantial, insulin-independent accumulation of the HA-GLUT4-GFP reporter at the cell surface. This effect was reproducible for endogenous GLUT4 in AP-1 depleted cells. In addition, we measured an increased basal glucose uptake in AP-1 knockdown 3T3-L1 adipocytes, indicating that endogenous GLUT4 also accumulates at the cell surface in response to AP-1 depletion.

Conclusion: Our data clearly indicate a role of AP-1 and AP-2 in GLUT4 sorting, and a dual role of the F5QQI motif in GLUT4 trafficking and endocytosis.