Novel Arguments in Favor of the Substrate-Transport Model of Glucose-6-Phosphatase Isabelle Gerin, Gae ¨ tane Noe ¨ l, and Emile Van Schaftingen The purpose of this work was to discriminate between two models for glucose-6-phosphatase: one in which the enzyme has its catalytic site oriented toward the lumen of the endoplasmic reticulum, requiring transporters for glucose-6-phosphate, inorganic phosphate (Pi), and glucose (substrate-transport model), and a second one in which the hydrolysis of glucose-6-phosphate occurs inside the membrane (conformational model). We show that microsomes preloaded with yeast phosphoglucose isomerase catalyzed the detritiation of [2- 3 H]glucose-6- phosphate and that this reaction was inhibited by up to 90% by S3483, a compound known to inhibit glu- cose-6-phosphate hydrolysis in intact but not in deter- gent-treated microsomes. These results indicate that glucose-6-phosphate is transported to the lumen of the microsomes in an S3483-sensitive manner. Detritiation by intramicrosomal phosphoglucose isomerase was stim- ulated twofold by 1 mmol/l vanadate, a phosphatase inhibitor, indicating that glucose-6-phosphatase and the isomerase compete for the same intravesicular pool of glucose-6-phosphate. To investigate the site of release of Pi from glucose-6-phosphate, we incubated micro- somes with Pb 2 , which forms an insoluble complex with Pi, preventing its rapid exit from the microsomes. Under these conditions, 80% of the Pi that was formed after 5 min was intramicrosomal, compared with <10% in the absence of Pb 2 . We also show that, when incubated with glucose-6-phosphate and mannitol, glucose-6-phos- phatase formed mannitol-1-phosphate and that this non- physiological product was initially present within the microsomes before being released to the medium. These results indicate that the primary site of product release by glucose-6-phosphatase is the lumen of the endoplas- mic reticulum. Diabetes 50:1531–1538, 2001 S ince the finding that glucose-6-phosphatase, an enzyme playing a major role in glucose produc- tion during starvation, is associated with the endoplasmic reticulum (1), much work has been carried out to identify its constituents and to determine their respective role. According to the substrate-transport model (2,3), the glucose-6-phosphatase system comprises a relatively nonspecific hydrolase, the catalytic site of which is oriented toward the lumen of the endoplasmic reticu- lum, a specific transporter for glucose-6-phosphate and transporters for inorganic phosphate (Pi) and glucose (Fig. 1A). This model accounts for several kinetic observations (2–5), including the fact that glucose-6-phosphatase acts much better on glucose-6-phosphate than on mannose- 6-phosphate in intact liver microsomes, whereas it is about equally active on both substrates in detergent-treated (disrupted) microsomes. Furthermore, this model allows rationalization of the effects of chlorogenic acid and some of its derivatives (such as S3483, used in this work), which inhibit the hydrolysis of glucose-6-phosphate but not of inorganic pyrophosphate in intact microsomes, having no effect in detergent-treated microsomes. These compounds are thought to be inhibitors of the glucose-6-phosphate translocase (6 – 8). The substrate-transport model also accounts for the observation that two principal types of glucose-6-phos- phatase deficiency have been identified: one, in which the phosphohydrotase is deficient, is known as glycogen stor- age disease type Ia (rev. in 9); the other, in which no transport of glucose-6-phosphate can be demonstrated in liver microsomes (10), is GSD type Ib, or GSDIb. The cDNAs mutated in GSD Ia (11,12) and GSD Ib (13) have been identified. The second encodes a protein belonging to the same family as transporters (UhpT) and a putative receptor (UhpC) for hexose-6-phosphates. Although coex- pression of glucose-6-phosphatase with the protein mu- tated in GSD Ib in COS-1 cells reconstitutes some degree of glucose-6-phosphate transport into microsomes (14), a clear demonstration that the protein mutated in GSD Ib acts as an independent glucose-6-phosphate translocase is still missing, owing partly to difficulties in expressing the protein in heterologous systems (I.G., unpublished re- sults). Furthermore, measurements of the incorporation of radioactivity into liver microsomes incubated with radio- labeled glucose-6-phosphate show that only a fraction (at most 10%) of the liberated glucose or Pi can be found inside the microsomes (15,16). This led some authors to conclude that glucose-6-phosphate transport into the en- From the Laboratory of Physiological Chemistry, ICP and Universite ´ Catho- lique de Louvain, Brussels, Belgium. Address correspondence and reprint requests to E. Van Schaftingen, UCL 7539, Avenue Hippocrate 75, B-1200 Brussels, Belgium. E-mail: vanschaftingen@ bchm.ucl.ac.be. Received for publication 18 January 2001 and accepted in revised form 10 April 2001. GSD, glycogen storage disease; Pi, inorganic phosphate. DIABETES, VOL. 50, JULY 2001 1531