Metabolic Impact of Adenovirus-mediated Overexpression of the Glucose-6-phosphatase Catalytic Subunit in Hepatocytes*

Glucose-6-phosphatase (G6Pase) catalyzes the hydrolysis of glucose 6-phosphate (Glu-6-P) to free glucose and, as the last step in gluconeogenesis and glycogenolysis in liver, is thought to play an important role in glucose homeostasis. G6Pase activity appears to be conferred by a set of proteins localized to the endoplasmic reticulum, including a glucose-6-phosphate translocase, a G6Pase phosphohydrolase or catalytic subunit, and glucose and inorganic phosphate transporters in the endoplasmic reticulum membrane. In the current study, we used a recombinant adenovirus containing the cDNA encoding the G6Pase catalytic subunit (AdCMV-G6Pase) to evaluate the metabolic impact of overexpression of the enzyme in primary hepatocytes. We found that AdCMV-G6Pase-treated liver cells contain significantly less glycogen and Glu-6-P, but unchanged UDP-glucose levels, relative to control cells. Further, the glycogen synthase activity state was closely correlated with Glu-6-P levels over a wide range of glucose concentrations in both G6Pase-overexpressing and control cells. The reduction in glycogen synthesis in AdCMV-G6Pase-treated hepatocytes is therefore not a function of decreased substrate availability but rather occurs because of the regulatory effects of Glu-6-P on glycogen synthase activity. We also found that AdCMV-G6Pase-treated-cells had significantly lower rates of lactate production and [3-3H]glucose usage, coupled with enhanced rates of gluconeogenesis and Glu-6-P hydrolysis. We conclude that overexpression of the G6Pase catalytic subunit alone is sufficient to activate flux through the G6Pase system in liver cells. Further, hepatocytes treated with AdCMV-G6Pase exhibit a metabolic profile resembling that of liver cells from patients or animals with non-insulin-dependent diabetes mellitus, suggesting that dysregulation of the catalytic subunit of G6Pase could contribute to the etiology of the disease.

In liver cells, hydrolysis of Glu-6-P is carried out by glucose-6-phosphatase (G6Pase), 1 an activity comprising a set of discrete proteins localized to the endoplasmic reticulum (ER) (1). These include a glucose-6-phosphate translocase that transports Glu-6-P from the cytoplasm to the lumen of the ER, the G6Pase phosphohydrolase or catalytic subunit that appears to reside in the lumen, and putative ER membrane-localized glucose and inorganic phosphate transporters that allow removal of the products of the reaction. To date, cDNA clones for the G6Pase catalytic subunit from several species (2)(3)(4) and a putative ER membrane glucose transporter termed GLUT-7 (5) have been isolated and analyzed, but the glucose-6-phosphate translocase remains uncharacterized. The existence of the translocase as a discrete entity is strongly supported by genetic studies (6), and the translocase and phosphohydrolase activities appear to be tightly linked (7).
Glucose causes activation of glycogen synthesis in the liver. This effect appears to be mediated by accumulation of Glu-6-P, the levels of which are directly correlated with the activation state of the rate-limiting enzyme for glycogenesis, glycogen synthase (8 -10). The Glu-6-P concentration in hepatocytes is determined in part by the opposing actions of G6Pase and the glucose-phosphorylating enzyme glucokinase. Glucokinase expression has been shown to decrease, and expression of the G6Pase catalytic subunit to increase, in several animal models of diabetes (11)(12)(13)(14), and these changes have been correlated with an increase in hepatic glucose production that appears to contribute to the diabetic phenotype, leading to the suggestion that the ratio of glucokinase:G6Pase is an important determinant of the metabolic fate of glucose in liver cells (12). While such a model is logical, the fact that G6Pase activity is imparted by a set of proteins, any of which could be rate-determining for the overall reaction, suggests that it should not be assumed that changes in expression of the G6Pase catalytic subunit necessarily correlate with altered rates of Glu-6-P hydrolysis in intact cells. Furthermore, the relative importance of changes in glucokinase versus G6Pase expression in liver has not been adequately defined. In fact, overexpression of glucokinase alone in isolated hepatocytes results in a large stimulation in glycogen synthesis, occurring in concert with increased Glu-6-P levels and an enhanced activation state of glycogen synthase (15,16). A further complication is introduced by the recent finding that adenovirus-mediated expression of glucokinase in hepatoma cells that are otherwise deficient in this enzyme activity leads to a glucose-stimulated increase in G6Pase expression (17). Thus, G6Pase gene expression appears to be positively regulated by glucose but negatively regulated by insulin (13,14,17,18). The fact that a large number of other enzymes of glycolysis and gluconeogenesis are regulated by these two effectors, including glucokinase, which is highly sensitive to induction by insulin (19), makes it difficult to isolate the contribution of changes in G6Pase expression to regulation of glucose metabolism in whole animal models.
In this study, we have utilized a recombinant adenovirus containing the cDNA encoding the catalytic subunit of G6Pase (20) to provide specific insight into the metabolic impact of its overexpression in primary hepatocytes. We demonstrate that overexpression of this single protein has a pronounced effect on glycolytic and gluconeogenic flux and on glycogen synthesis. The profound reduction in glycogen accumulation observed in the face of reduced Glu-6-P levels in G6Pase-overexpressing hepatocytes provides further support for an essential role for this intermediate in the activation of glycogen synthesis.
After treatment with recombinant viruses, the medium was replaced with Dulbecco's modified Eagle's medium supplemented with 1 mM glucose, 1 nM dexamethasone, and 1 nM insulin. Cells were incubated for 16 h prior to the experimental manipulations detailed in the legend to Table I. At the end of each experimental manipulation, cell monolayers were frozen in liquid N 2 until analysis.
Enzyme Activity Assay-To measure enzyme activities, 100 l of homogenization buffer consisting of 10 mM Tris-HCl (pH 7.0), 150 mM KF, 15 mM EDTA, 600 mM sucrose, 15 mM 2-mercaptoethanol, 10 g/ml leupeptin, 1 mM benzamidine, and 1 mM phenylmethylsulfonyl fluoride was added to frozen plates containing the cell monolayers, and cells were collected with a plastic scraper. Cell bursting was caused by thawing, except in the G6Pase activity assay, where sonication was used, and checked under light microscopy for completeness. Homogenates were collected in Eppendorf tubes and centrifuged at 10,000 ϫ g for 15 min at 4°C, and supernatants and/or pellets were used for determinations. Protein concentration was measured as described by Bradford (24) using the Bio-Rad assay reagent. Glucose phosphorylating activity was measured spectrophotometrically in 10,000 ϫ g supernatant fractions of hepatocyte extracts at 37°C in the presence of 100 mM and 0.5 mM glucose as described (25). Glycogen synthase activity was measured in the presence or absence of 6.6 mM Glu-6-P as described (26) in both supernatant and pellet fractions. The activity measured in the absence of Glu-6-P represents the active form of the enzyme, whereas the activity tested in the presence of 6.6 mM Glu-6-P is a measure of total activity. The ratio of these two activities is known as the glycogen synthase activity ratio and is an estimate of the degree of activation of the enzyme. G6Pase activity was measured in both supernatant and pellet fractions of sonicated homogenates as described (27).
Metabolite Determinations-Glycogen content was measured by scraping cells into 30% KOH and boiling the extract for 15 min. Glycogen was measured as described (28). The intracellular concentrations of Glu-6-P and UDP-glucose were measured by previously described spectrophotometric assays (29,30). L-Lactate levels in the incubation medium were measured as described (31).
Measurements of Glycolytic and Gluconeogenic Flux-Hepatocytes were left untreated or treated with AdCMV-G6Pase and, after removal of the virus, incubated for 16 h in the supplemented Dulbecco's modified Eagle's medium described earlier. Glucose usage in AdCMV-G6Pase or untreated hepatocytes was assayed by incubation of cells with 5 or 20 mM [3-3 H]glucose and measurement of 3 H 2 O production as described previously (32). In separate assays, 3 H 2 O incorporation into glucose was measured as recently described (20). Briefly, cells were washed with Krebs-bicarbonate buffer containing 0.5% bovine serum albumin and incubated for 3 h in the same buffer containing either 5 or 20 mM glucose plus 3 H 2 O (1.25 mCi/100 ml of medium; NEN Life Science Products). Media samples were collected, and 3 H-labeled glucose was separated from other labeled media components by silica gel thin layer chromatography. Glucose production was assayed in hepatocytes incubated for 16 h in the supplemented Dulbecco's modified Eagle's medium described earlier. Cells were then incubated either in a medium containing 5 mM alanine, glycine, and glutamine or in another medium containing 10 mM lactate and 1 mM pyruvate for 2 h. Media were collected for glucose determination.
Effect of G6Pase Overexpression on Glucose Phosphorylation and Glucose 6-Phosphate Levels-We next evaluated the effect of AdCMV-G6Pase treatment on total glucose phosphorylating capacity and Glu-6-P levels in hepatocytes. As shown in Table  I, overexpression of G6Pase had no effect on total glucose phosphorylating capacity in hepatocyte extracts, measured at either 0.5 or 100 mM glucose, indicating that the AdCMV-G6Pase virus allows us to specifically manipulate G6Pase activity in liver cells without affecting glucose phosphorylation. Untreated, AdCMV-G6Pase-treated, or AdCMV-␤Gal-treated cells preincubated for 16 h in 1 mM glucose and then transferred to media containing 1, 5, 10, or 25 mM glucose for 2 h exhibited a glucose concentration-dependent increase in Glu-6-P levels. However, AdCMV-G6Pase-treated cells had lower Glu-6-P levels for all conditions studied than either control group, the effect becoming more pronounced with increasing media glucose concentration (Fig. 1). The maximal difference was observed at 25 mM glucose, where Glu-6-P levels were reduced by 25% in AdCMV-G6Pase-treated cells relative to untreated or AdCMV-␤Gal-treated control cells.
Effect of G6Pase Overexpression on Gluconeogenic Flux-Flux through G6Pase in untreated and AdCMV-G6Pasetreated hepatocytes was measured either by incorporation of  3 H incorporation into glucose at 5 mM glucose and a 6-fold increase at 20 mM glucose relative to untreated cells incubated at the same glucose concentrations (Fig. 2). We have previously shown that treatment of INS-1 cells with AdCMV-␤Gal has no effect on 3 H 2 O incorporation into glucose relative to untreated control cells (20), so this control was not included in the present study. Thus, overexpression of the catalytic subunit of G6Pase alone is sufficient to cause a large increase in Glu-6-P hydrolysis and flux through the G6Pase system in hepatocytes. Total gluconeogenic flux was measured in the presence of gluconeogenic amino acids (a mixture of 5 mM alanine, 5 mM glycine, and 5 mM glutamine) or in medium supplemented with 10 mM lactate, 1 mM pyruvate by measurement of the media glucose concentration. Experiments were conducted by preincubation of AdCMV-G6Pase-treated or untreated cells in 1 mM glucose for 16 h, followed by a 2-h incubation in either of the two gluconeogenic media. As shown in Fig. 3, both untreated and AdCMV-G6Pase-treated hepatocytes produced more glucose when incubated with the lactate/ pyruvate mixture than with the mixture of gluconeogenic amino acids. Further, AdCMV-G6Pase-treated cells produced 40% more glucose from the amino acids and 47% more from lactate/pyruvate than did untreated cells. Overexpression of the catalytic subunit of G6Pase alone therefore results in enhanced glucose production from gluconeogenic precursors, suggesting that the G6Pase step participates in control of the overall rate of gluconeogenesis in isolated hepatocytes.

Effect of G6Pase
Overexpression on Glycolytic Flux-The impact of overexpression of the G6Pase catalytic subunit on glycolytic flux was determined by measurement of lactate production and by incubation of cells with [3-3 H]glucose and measurement of glucose usage as a function of 3 H 2 O production. Lactate accumulation measured in the media of cells incubated for 8 h in the presence of 1, 5, 10, or 25 mM glucose was found to increase in a glucose concentration-dependent manner in untreated and AdCMV-G6Pase-treated hepatocytes, but this increase was significantly attenuated in AdCMV-G6Pase-treated cells (Fig. 4). The effect of AdCMV-G6Pase on lactate production resembled its effect on Glu-6-P levels (Fig.  1), with a maximal difference in lactate accumulation of 32% observed at 25 mM glucose, similar to the 25% difference in Hepatocytes were treated with AdCMV-G6Pase (Ⅺ) or AdCMV-␤Gal (s) or left untreated (f) and were incubated in different glucose concentrations as described in the legend to Table I. Cells were then collected for assay of intracellular Glu-6-P levels. Data represent the mean Ϯ S.E. for four independent experiments. The asterisks indicate statistical significance at the following levels: *, p Ͻ 0.05; **, p Ͻ 0.01 for comparisons between the AdCMV-G6Pase and the untreated or AdCMV-␤Gal control groups.
Glu-6-P levels at this glucose concentration. [3-3 H]glucose usage was reduced by 51% at 5 mM glucose and 49% at 20 mM glucose in AdCMV-G6Pase-treated hepatocytes relative to untreated control cells (Fig. 5). We have previously demonstrated that the AdCMV-␤Gal virus has no effect on glucose usage or lactate production in hepatocytes relative to untreated cells (16), and this control was therefore not included in the current study. Thus, while the effect of G6Pase overexpression on glucose usage appeared slightly larger than its effect on lactate accumulation, particularly at low glucose, a significant decrease in glycolytic flux is indicated by both methods.
Effect of G6Pase Overexpression on UDP-Glucose Levels and Glycogen Synthesis-To evaluate the effect of G6Pase overexpression on glycogen synthesis, a number of assays were conducted. First, we determined the levels of the immediate precursor for glycogen synthesis, UDP-glucose. Using an experimental approach identical to that described earlier for the determination of Glu-6-P levels, we found no significant differences in UDP-glucose levels in AdCMV-G6Pase-treated compared with untreated hepatocytes at any concentration of glucose tested (Fig. 6). Next, we measured glycogen content in groups of hepatocytes treated identically as for the Glu-6-P and UDP-glucose assays. Glycogen content increased in a glucose concentration-dependent manner in untreated, AdCMV-␤Galtreated, and AdCMV-G6Pase-treated cells, but glycogenesis was markedly reduced at all glucose concentrations in the AdCMV-G6Pase-treated group (Fig. 7). The most dramatic difference was observed at 25 mM glucose, where AdCMV-G6Pasetreated cells contained only 45% of the glycogen found in either group of control cells.
Effect of Overexpression of G6Pase on the Activation State of Glycogen Synthase-In light of the large effect of G6Pase expression on glycogen synthesis, we next investigated the activation state of the rate-limiting enzyme for glycogenesis, glycogen synthase. Total glycogen synthase activity, measured in the presence of 6.6 mM Glu-6-P, was not different in untreated and AdCMV-G6Pase-treated hepatocytes (1.35 Ϯ 0.04 versus 1.30 Ϯ 0.04 milliunits/10 6 cells, respectively), and the amount of glycogen synthase protein measured by immunoblotting was also the same in the two groups (data not shown). The glycogen synthase activity ratio is calculated as the ratio of enzyme activity measured in the absence of Glu-6-P to that in the presence of Glu-6-P, and it is a measure of the activation state of the enzyme. This activity ratio was determined for glycogen synthase recovered in the 10,000 ϫ g supernatant and pellet fractions of extracts prepared from hepatocytes incubated with increasing concentrations of glucose, as described earlier for the Glu-6-P, UDP-glucose, and glycogen measurements. The glycogen synthase activity ratio increased in a glucose concentrationdependent fashion in both untreated and AdCMV-G6Pase-treated hepatocytes and in both the supernatant and pellet fractions, but cells treated with AdCMV-G6Pase experienced a much smaller increase in activity ratio in both subcellular fractions (Figs. 8, A and B). We have previously shown that a control virus had no effect on the glycogen synthase activity ratio in hepatocytes relative to untreated cells (15), so the AdCMV-␤Gal virus control was not included in the current study. The changes in the activity ratio in the supernatant fraction were compared with the changes in Glu-6-P levels described in Fig. 1, and these two parameters are shown to be tightly correlated (Fig. 9). The same correlation also holds true for the enzyme present in the 10,000 ϫ g pellet (data not shown) Thus, G6Pase overexpression reduces the amount of glycogen synthase that is in an active state, a change that is well correlated with the intracellular level of Glu-6-P. DISCUSSION In this study, we have investigated the metabolic impact of overexpression of the catalytic subunit of G6Pase in cultured rat hepatocytes. The motivation for this work was to address two fundamental issues. First, a significant body of evidence, beginning with the study of Ciudad et al. (8), has accumulated in support of the notion that a glucose-induced increase in  Table I and collected for an assay of intracellular UDP-glucose levels. Data represent the mean Ϯ S.E. for four independent experiments.

FIG. 7. Effects of G6Pase overexpression on glycogen content.
Hepatocytes were treated with AdCMV-G6Pase (Ⅺ) or AdCMV-␤Gal (s) or left untreated (f) and were incubated in different glucose concentrations as described in the legend to Table I. Cells were then collected and assayed for glycogen content. Data represent the mean Ϯ S.E. for four independent experiments. The asterisks indicate statistical significance at the following levels: *, p Ͻ 0.05; **, p Ͻ 0.01; and ***, p Ͻ 0.001 for comparisons between AdCMV-G6Pase and untreated or Ad-CMV-␤Gal-treated hepatocytes.
Glu-6-P levels is an essential event for the activation of glycogen synthase and thus glycogen deposition. This implies that maneuvers designed to specifically lower Glu-6-P levels will block the activation of glycogen synthase during glucose stimulation of hepatocytes leading to a decrease in glycogen synthesis, but this idea had never been tested prior to this study. Second, studies in a variety of animal models of diabetes have led to the suggestion that lowering of the glucokinase:G6Pase enzyme activity ratio, whether by a decrease in glucokinase, an increase in G6Pase, or some combination of the two, will cause an increase in hepatic glucose output and contribute to the etiology of diabetes (11)(12)(13). However, to date it had not been possible to isolate the effect of overexpression of the catalytic subunit of G6Pase on glucose utilization and storage and hepatic glucose production. This has been accomplished in the current study by use of the recombinant adenovirus system for expression of G6Pase in isolated hepatocytes, a maneuver that did not alter the glucose phosphorylating capacity.
With regard to the first issue raised, our results provide fresh insight into the role of Glu-6-P in the control of glycogen synthase activity and glycogen synthesis. In previous work, we demonstrated that adenovirus-mediated overexpression of glucokinase in hepatocytes caused a large glucose concentrationdependent increase in Glu-6-P levels relative to untreated control hepatocytes. The increase in Glu-6-P was tightly correlated with large increases in the activation state of glycogen synthase and the amount of glycogen synthesized (15,16). Here, we show that the opposite is also true. When Glu-6-P levels are decreased by overexpression of G6Pase, glycogen synthase is less active and the synthesis of glycogen is greatly diminished. The effect of G6Pase on glycogen synthase activity state and glycogen synthesis is tightly correlated with the concentration of Glu-6-P. Further, these studies define a threshold level of Glu-6-P for activation of glycogen synthase, since the enzyme is not active at Glu-6-P concentrations of less than 0.2 mM. As has been shown in vivo, the glycogen synthase activation system seems to be exquisitely sensitive to changes in Glu-6-P intracellular concentration in the range 0.2-0.3 mM (9). We conclude that the reduction in glycogen synthesis that occurs in response to overexpression of the G6Pase catalytic subunit does not appear to be a result of decreased substrate availability but rather is a function of the regulatory effects of Glu-6-P on glycogen synthase activation state. This statement is partly based on our finding that the levels of the immediate precursor for glycogen synthesis, UDP-glucose, were the same in control and AdCMV-G6Pase-treated hepatocytes at all glucose concentrations used in this study.
These experiments also suggest that the main role of Glu-6-P in the liver is to trigger the permanent activation of glycogen synthase by favoring its dephosphorylation. In vitro studies have shown that Glu-6-P binding to glycogen synthase, in addition to causing allosteric activation of the enzyme, renders it more susceptible to dephosphorylation and covalent activation by purified phosphatases 1 and 2A (33). Therefore, we may postulate the following mechanism: the binding of Glu-6-P to glycogen synthase alters its structure in a way that facilitates the covalent dephosphorylation and activation of the enzyme. This produces changes in the glycogen synthase activation state, which are is responsible for the enhanced glycogen synthesis. The capacity of Glu-6-P to activate the enzyme through the allosteric mechanism probably plays a minor, transient role, since at the same time that the enzyme is allosterically activated it is rapidly fixed into a more active, less phosphorylated state, by the action of phosphatases.
Classically, there is debate as to whether the stimulation of the synthesis of glycogen in the liver is brought about by a "pull" (34) or by a "push" (35) mechanism. According to the pull mechanism, glucose administration provokes a sequential inactivation of phosphorylase and activation of synthase, with the latter activation decreasing Glu-6-P and UDP-glucose levels as a result of a pull into glycogen. The push mechanism is based on a simulated metabolic model. It predicts that glucose induces glycogen deposition by a mechanism independent of synthase activation but requiring an increase in Glu-6-P and FIG. 9. Correlation between glycogen synthase activity ratio and Glu-6-P levels. Glu-6-P levels measured in Fig. 1 were plotted against the glycogen synthase activity ratio values achieved at the same glucose concentration measured in the 10,000 ϫ g supernatant fraction. Ⅺ, AdCMV-G6Pase-treated hepatocytes; f, untreated hepatocytes. The regression coefficient was 0.98 for the plot.
FIG. 8. Effects of G6Pase overexpression on glycogen synthase activity ratio. Hepatocytes were treated with AdCMV-G6Pase (Ⅺ) or left untreated (f) and were incubated in different glucose concentrations as described in the legend to Table I. Cells were then collected for an assay of glycogen synthase activity. A, glycogen synthase activity ratio (glycogen synthase activity measured in the absence of Glu-6-P divided by the activity measured in the presence of 6.6 mM Glu-6-P) in the 10,000 ϫ g supernatant fraction of hepatocyte homogenates. B, glycogen synthase activity ratio in the 10,000 ϫ g pellet fraction of hepatocyte homogenates. Data represent the mean Ϯ S.E. for four independent experiments. The asterisks indicate statistical significance at the following levels: *, p Ͻ 0.05; **, p Ͻ 0.01 for comparisons with untreated hepatocytes. ultimately in UDP-glucose concentrations. Our results indicate that hepatic glycogenesis occurs by a combination of these two mechanisms. It can be considered that glycogen synthesis is a pull process in the sense that the activation of glycogen synthase, by covalent dephosphorylation, stimulates the incorporation of glucosyl units from UDP-glucose into glycogen. The push comes from the increase in intracellular concentration of Glu-6-P in response to glucose, which triggers the covalent activation of glycogen synthase. When, as in the G6Pase-overexpressing cell, Glu-6-P levels decrease, protein phosphatases are less able to dephosphorylate glycogen synthase, and the enzyme remains in a lower activation state. The net result is that less glycogen can be synthesized.
With regard to the second issue raised, overproduction of glucose by the liver is a major cause of fasting hyperglycemia in non-insulin-dependent diabetes mellitus (13). It has been hypothesized that augmented G6Pase activity may contribute to increased hepatic glucose production and onset of the disease (11-14, 17, 18). Herein we demonstrate that overexpression of the G6Pase catalytic subunit alone is sufficient to markedly reduce glucose utilization and storage and to augment glucose production. Thus it appears that other components of the G6Pase complex such as the glucose-6-phosphate translocase and ER glucose transporter do not prevent an increase in Glu-6-P hydrolysis when the catalytic subunit alone is overexpressed. In addition, we show that by overexpressing G6Pase in cultured hepatocytes, a metabolic pattern is obtained opposite to that encountered in cases of deficient activity of the enzyme. The lack of the catalytic subunit in human patients of glycogen storage disease type 1a or in mice in which the G6Pase gene is knocked out by homologous recombination results in glycogen accumulation in the liver and severe hypoglycemia and lactic acidemia (7,36). Overexpression of the G6Pase catalytic subunit promotes the reverse situation: a reduction in glycogen synthesis and an increased glucose production from lactate. We conclude that overexpression of G6Pase in hepatocytes has a potent metabolic impact. Cells overexpressing G6Pase accumulate much less glycogen, produce less lactic acid, and are more effective in producing glucose from gluconeogenic substrates than normal cells. These findings are consistent with our recent report showing that 10-fold overexpression of the G6Pase catalytic subunit in the insulinoma cell line INS-1 caused a 4.2-fold increase in 3 H 2 O incorporation into glucose, a 32% reduction in glucose usage, and a proportional 30% decrease in insulin secretion (20). Based on these results, it would appear that changes in expression of the G6Pase catalytic subunit reported in liver and islets of various models of diabetes (11)(12)(13)(14)37) are likely to correlate with altered glucose metabolism and that increased expression of this gene could contribute to the onset of diabetes. We have recently adapted the recombinant adenovirus technology for studies in intact animals (38,39). This approach should provide a precise mechanism for altering the expression of G6Pase in liver of whole animals, thereby providing further insight into the role of this enzyme in glucose homeostasis and pathogenesis of diabetes.