| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Department of Medicine, Austin Health and Northern Health, University of Melbourne, Heidelberg Heights; and 2Department of Physiology, Monash University, Clayton, Victoria, Australia
Submitted 30 June 2008 ; accepted in final form 2 September 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT EXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
endogenous glucose production; glucose intolerance
Glycerol enters the gluconeogenic pathway immediately before the step catalyzed by fructose-1,6-bisphosphatase (FBPase), a regulated enzyme in gluconeogenesis. As mentioned above, the rate of glycerol gluconeogenesis is increased approximately twofold in obese patients with type 2 diabetes (22, 27). Furthermore, when matched for plasma glycerol concentrations, patients with type 2 diabetes retained a higher rate of glycerol gluconeogenesis compared with nondiabetic subjects, which suggested an increase in the intrahepatic conversion of glycerol to glucose in diabetes. The study by Nurjhan et al. (22) suggested that this may be due to an increase in FBPase activity.
Our laboratory tested the role of FBPase in glucose production by generating a hemizygous transgenic mouse model (line 1) with a twofold overexpression of human FBPase in the liver. This resulted in an approximately threefold increase in glycerol gluconeogenesis both in the basal state and after a euglycemic hyperinsulinemic clamp compared with littermate controls (18). Interestingly, these transgenic mice had normal EGP and glucose tolerance (18). This was a surprising result, considering the human studies mentioned previously and results from our laboratory and others. These studies showed liver FBPase expression and/or activity to be double that of controls in several models of insulin resistance and obesity including New Zealand obese (NZO) mice (2), db/db mice (6), ob/ob mice (32, 33), fa/fa rats (39), and high-fat (HF) fed (HFF) mice and rats (2, 35), and in liver biopsies from patients with type 2 diabetes (9).
Other groups (38, 42, 43) have previously generated animal models overexpressing the other regulated gluconeogenic enzymes phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase (G-6-Pase) in the liver. When G-6-Pase was overexpressed in the liver of rats using adenoviral delivery, they became hyperinsulinemic and glucose intolerant (42). Mice that overexpressed PEPCK in the liver were glucose intolerant and had elevated levels of EGP (38, 43). Similarly, our group generated transgenic rats with renal PEPCK overexpression. These transgenic rats have impaired insulin suppression of EGP, causing peripheral (muscle and fat) insulin resistance (17, 31).
Considering these studies, the failure of a twofold overexpression of FBPase in line 1 transgenics to produce a phenotype was unexpected (18). However, as reported, this line of transgenic mice also overexpressed FBPase in the brain (18). Given the known effects of hypothalamic metabolism on EGP (23, 25), it was possible that the effects of liver overexpression may have been counteracted by the overexpression of FBPase in the brain. We suspected that site-of-integration effects may be responsible for this extrahepatic expression of our transgene. Previous studies (13, 26) have shown that it is possible to protect against site-of-integration effects with the use of "insulator sequences." These sequences are proposed to act by insulating the flanked transgenic sequence from nearby repressor/enhancer elements thereby resulting in tissue-specific expression (7). We took advantage of this technology and produced a new mouse line aiming for liver-specific transgene expression, using the same transgenic cassette as before (18) but flanking it with insulator sequences from chicken β-globin, named hypersensitive-4 (HS4; Ref. 26).
The current studies were performed to validate the results observed in the previous transgenic model (line 1) and to determine if higher levels of overexpression were necessary to observe a glucose intolerant phenotype. Additional aims were to further characterize the glucose tolerance phenotype in line 1 and the new insulated transgenic mice at two ages (24 wk and 1 yr old), after a 60% HF diet and when bred to homozygosity, to investigate whether FBPase has a specific role in EGP and glucose intolerance under these conditions.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
Genotyping of transgenic mice. To detect the presence of the TTR-FBPase transgene, a PCR using specific primers spanning the TTR promoter-FBPase junction was performed. This procedure and the primer sequences were described in detail previously (18).
Homozygous transgenic mice were identified by a two-step procedure. In the first step, the usual genotyping PCR was used to show the presence of the transgene. The second step utilized real-time PCR to differentiate between the hemizygous and homozygous transgenic mice. Primers were designed for SYBR green real-time PCR using Primer Express software (Applied Biosystems, Scoresby, VIC, Australia). The primer sequences for the 18S endogenous control were as follows: forward, 5'-CGG ACA CGG ACA GGA TTG ACA-3'; and reverse, 5'-ACA AAT CGC TCC ACC AAC TAA GA-3'. The primer sequences for detecting the genomic transgenic FBPase were as follows: forward, 5'-TGA CCC ATT TCA CTG ACA TTT CTC-3'; and reverse, 5'-CAG CCA TGC TTG AAC CGG-3'. The primers were positioned within introns to amplify genomic DNA and exclude potential contaminating RNA. Each SYBR green real-time PCR reaction used 0.19 nM of 18S, 56 nM of genomic transgenic FBPase primers, and 2.5 ng of DNA. A sample from each of the transgenic mice was analyzed by real-time PCR to differentiate homozygous from hemizygous transgenic mice. Samples were analyzed by the ABI Sequence Detection software and absolute quantification comparative Ct method (Applied Biosystems software). Transgenic samples were accepted as homozygous when approximately double the absolute quantitation of the hemizygous samples.
Housing and maintenance of mice.
The mice were housed at the University of Melbourne, Department of Medicine Animal Research Facility at the Heidelberg Repatriation Hospital, at room temperature (
20°C) under a 12-h light-dark cycle. All mice were supplied with tap water and fed ad libitum a standard laboratory chow (77% carbohydrate, 20% protein, and 3% fat), except for those used in the HFF studies, which were fed a 60% fat diet (22% carbohydrate, 18% protein, and 60% fat) for 12 wk from 12 wk of age. The mice were maintained in accordance with guidelines of the Austin Hospital Animal Ethics Committee (Approval No.: A2007.2752).
Physiological studies were performed on second generation hemizygous mice (TG+/–) and first generation homozygous mice (TG+/+) at 24 wk of age, except for the aging study mice (1 yr old). Control mice for all experiments were age-matched transgene-negative littermates from each line (NEG).
Determination of FBPase, PEPCK, and G-6-Pase mRNA levels. RNA was extracted from tissues using the Trizol method (Invitrogen, Mount Waverley, VIC, Australia). RNA was treated with DNaseI (Ambion, Scoresby, VIC, Australia) and cDNA synthesized using 1 µg of DNase-treated RNA and random primers with the Promega reverse transcription kit (Annandale, NSW, Australia).
Total FBPase mRNA levels (mouse and human) were quantitated using SYBR green real-time PCR. Primers for the β-actin endogenous control were described previously (18). Primers for total FBPase that matched the mouse sequence (GenBank: NM_019395) and human sequence of liver FBPase (GenBank: NM_000507) were forward 5'-AGC CTT CTG AGA AGG ATG CTC-3' and reverse 5'-GTC CAG CAT GAA GCA GTT GAC-3'. Transgenic FBPase mRNA levels were measured in a range of tissues by the real-time PCR method as described previously (18).
TaqMan gene expression assay kits (Applied Biosystems, Scoresby, VIC, Australia) were used to measure PEPCK (Mm00440636_m1) and G-6-Pase (Mm00839363_m1) mRNA levels using the ABI PRISM 7900 HT system (Applied Biosystems, Scoresby, VIC, Australia). Reactions containing 10 ng cDNA were analyzed by the ABI Sequence Detection software and relative quantification comparative Ct method (Applied Biosystems software).
Western blotting and enzyme assays for FBPase. Total FBPase protein levels were determined by Western blotting as described previously (18). The anti-rat FBPase primary antibody was a kind gift from Dr. Hideo Mizunuma (Akita University, Akita, Japan). Liver homogenates of 1 µg of total protein were loaded, and FBPase (37 kDa) was detected with the anti-rat liver FBPase antibody. GLUT2 protein was used as a loading control. Anti-GLUT2 (H-67) primary antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
FBPase enzyme activity was determined using a spectrophotometric assay as described previously (18).
Aging study. Mice were aged to 1 yr and underwent intraperitoneal glucose tolerance tests (IPGTTs).
HFF study.
Mice were fed a 60% HF diet (Specialty Feeds, Glen Forrest, WA, Australia) for 12 wk after 1 wk of acclimatization to the diet. The fat in the diet comprised of 80% saturated and 20% unsaturated fat. The mice were monitored weekly while on the diet to ensure they were gaining weight. After the mice were on the HF diet for 12 wk, IPGTTs were performed.
IPGTTs. IPGTTs were performed as previously described using a 2 g/kg glucose bolus, with blood samples taken for plasma glucose and insulin determination (17). These experiments were performed on hemizygous transgenic mice at 24 wk and 1 yr of age, the HFF mice, and the homozygous transgenic mice. After completion of the test, mice were killed by cervical dislocation, and the livers and other tissue were collected, quickly snap-frozen in liquid nitrogen, and stored at –70°C for further studies.
Intraperitoneal pyruvate tolerance tests.
Overnight fasted (16 h) mice were anesthetized with an intraperitoneal injection of pentobarbitone sodium (100 mg/kg). After a 5-µl basal blood sample was taken from the tail vein, the mice were injected intraperitoneally with 2 g/kg pyruvate (Sigma-Aldrich, Castle Hill, NSW, Australia). Subsequent blood samples were taken at 15, 30, 60, and 120 min, and blood glucose was determined using a Precision Q.I.D. glucometer (MediSense, Doncaster, VIC, Australia).
Plasma assays. Plasma glucose was determined using an Analox GM7 Micro-stat glucose analyzer and reagent (Helena Laboratories, Mount Waverley, VIC, Australia).
Plasma insulin was measured by a double antibody radioimmunoassay using a rat-specific insulin antibody (raised in guinea pigs) and highly purified rat insulin standard (Linco Research).
Glucose turnover.
Overnight (16 h) fasted mice were anesthetized with an IP injection of pentobarbitone sodium (100 mg/kg). Two catheters were inserted, one in the right jugular vein for tracer infusion and the other in the left carotid artery for blood sampling. A tracheostomy was also performed to prevent upper respiratory tract obstruction. A 2-min priming bolus (3.0 µCi) followed by a continuous (0.15 µCi/min) infusion of [6-3H]glucose was given for 120 min to measure basal whole body glucose turnover. Blood samples were collected at 90, 100, and 110 min. The blood was centrifuged, and the plasma was collected and stored at –20°C for measurement of glucose and radioactivity levels as described previously (3, 14, 18, 21).
Liver triglyceride levels.
Triacylglycerol content was analyzed as previously described (11). Lipid was extracted from 20 mg liver by a Folch extraction, the triacylglycerol was saponified in an ethanol-KOH solution at 60°C, and glycerol content was determined by enzymatic spectrophotometric analysis (free glycerol reagent; Sigma, Castle Hill, NSW, Australia).
Statistical analysis.
All results are expressed as means ± SE. Student's t-test was used to analyze differences between two independent groups. Area under the curve (AUC) was calculated using the trapezoidal rule. A general linear model (GLM) ANOVA with post hoc Tukey's test was used to determine significances in glucose and insulin levels between the groups (age, diet, and genotypes) in response to IPGTT repeated measures. P 
0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Initial characterization of new FBPase transgenic mice. Three founder mice were identified; one line did not overexpress the transgene (line 4) and two lines (lines 5 and 6) overexpressed the human liver FBPase transgene in the liver to similar levels as determined by Western blot analysis (data not shown). One of these lines, line 5, was used for the studies described in this article.
The endogenous TTR promoter predominately directs expression to the liver and low levels in the choroid plexus (47). However, when used as a transgenic promoter, it has also been documented to express at low levels in other tissues (1, 24, 40, 46). Therefore, the level of transgenic FBPase mRNA expression was assessed in a range of tissues from line 5 transgenic mice. Expression of the transgene was undetectable by real-time PCR analysis in the kidney, quadriceps, pancreas, intestine, stomach, heart, spleen, gonads, white adipose tissue, and brown adipose tissue (data not shown). In line 1 transgenic mice, there was an eightfold overexpression of total liver FBPase levels observed in the hypothalamus compared with negative littermates, while hypothalamic mRNA levels of FBPase in line 5 hemizygous transgenic mice were comparable with the negative littermates (Fig. 1C). Therefore, the current transgenic model had liver-specific overexpression of FBPase.
Line 5 hemizygous transgenic mice showed an approximately threefold overexpression of total liver FBPase mRNA (Fig. 1D) with concomitant increases in liver FBPase protein (Fig. 1E) and enzyme activity levels (Fig. 1F).
Effect of higher level of overexpression on transgenic phenotype. After an overnight fast, plasma glucose and insulin levels were not different in hemizygous transgenic mice compared with negative littermates (Table 1). Body weight and food intake were also reduced in both FBPase transgenic lines compared with negative littermates (Table 1). To determine if a higher level of overexpression is required to observe a glucose intolerant phenotype, 24-wk-old transgenic mice from line 1 (2-fold overexpression) and line 5 (3-fold overexpression) were challenged with a 2 g/kg bolus of glucose. There was no difference in glucose tolerance compared with the negative littermates even with the higher level of overexpression of FBPase in the liver (Fig. 2, B and D). Furthermore, the plasma insulin levels during the IPGTT were similar in the transgenic lines and their negative littermates (Fig. 2, C and E).
|
|
Effect of HF diet on transgenic phenotype.
Since elevated FBPase in spontaneous and diet-induced animal models has always been associated with excess weight gain or obesity, we assessed whether exposure to long-term HF feeding was necessary for a glucose intolerant phenotype to be observed in our transgenic mice. The HF diet elevated liver FBPase mRNA (Fig. 3A) and enzyme activity levels (Fig. 3B) in transgenic mice compared with the negative chow and negative HFF mice. Liver FBPase mRNA levels were even higher in the HFF line 5 transgenic mice (10-fold of negative) compared with the aged line 5 transgenic mice (4.5-fold of negative). In addition, the HFF diet resulted in 2.5-fold elevated levels in both liver PEPCK and G-6-Pase mRNA in the HFF negative and transgenic mice compared with those on a chow diet, with no differences in both these genes when the transgenics and negative littermates on the normal chow diet were compared (Fig. 3, G and H, respectively). Liver triglyceride levels were not different under chow conditions between the negative and transgenic mice from line 5 (data not shown). In response to a HF diet, liver triglyceride levels were approximately twofold higher in the HFF transgenics compared with the negative littermate mice on the same diet (61.9 ± 10.2 vs. 28 ± 7.6 µmol/g; P < 0.02). Interestingly, both line 1 and line 5 transgenic mice were clearly glucose intolerant after HFF compared with the negative littermates on the same diet (Fig. 3, C and E). Analysis with GLM ANOVA and post hoc testing also revealed that diet and genotype of the mice had a significant effect on plasma glucose levels during the IPGTT (P < 0.001, NEG HFF vs. line 1 and line 5 TG HFF). There were no significant differences when the two transgenic lines were compared against each other when both were placed on a HF diet. Although there was no effect of the HF diet on glucose tolerance in the negative mice (Fig. 3, C and E), it was sufficient to cause insulin resistance with significantly higher insulin levels during the IPGTT compared with the negative chow-fed mice (Fig. 3, D and F).
|
5.5-fold of negative) and the line 5 hemizygote transgenics (3-fold of negative; Fig. 4A). FBPase enzyme activity levels were measured in the liver and remained significantly higher in the homozygous mice than in the negative mice and tended to be higher than the hemizygous littermates (Fig. 4B).
|
A pyruvate tolerance test was performed to provide an indication of the rate of gluconeogenesis in the homozygous transgenic mice. The blood glucose levels were higher (Fig. 5A), suggestive of a higher basal rate of gluconeogenesis in the homozygous transgenic mice than the negative and hemizygous transgenic littermates. There was no difference in blood glucose levels observed between hemizygous transgenic mice and negative littermates.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
In light of this data, FBPase has recently received interest from pharmaceutical companies producing compounds to inhibit FBPase as a target for type 2 diabetes therapy (10, 16, 44, 45). These inhibitors decrease EGP in Zucker diabetic fatty rats, an obese diabetic animal model (10, 44), demonstrating a potential therapeutic benefit of FBPase inhibitor compounds for type 2 diabetes. Given the studies showing an increase in FBPase in models of obesity and insulin resistance, conditions that affect many other genes and biochemical pathways, it was important to determine whether liver-specific FBPase overexpression elevates EGP.
We previously generated transgenic mice (line 1) overexpressing human FBPase in the liver under the control of the TTR promoter (18). The mRNA, protein, and enzyme activity levels all showed a twofold overexpression of FBPase in the liver of these transgenic mice and an almost threefold increase in glycerol gluconeogenesis (18), consistent with the data from patients with type 2 diabetes (9) and our own work in the NZO mouse (2, 3). However, the transgenic mice had normal EGP and glucose tolerance, which was unexpected considering the impairments in glucose tolerance and insulin resistance in other transgenic models overexpressing gluconeogenic enzymes such as PEPCK (38, 43) and G-6-Pase (42) in the liver. Additionally, our laboratory (30, 31) has produced rats overexpressing PEPCK in the kidney and found that they displayed impaired insulin suppression of EGP and peripheral insulin resistance.
In addition to liver overexpression, our first transgenic model also expressed transgenic FBPase in the brain (18). We showed that these line 1 transgenic mice had lower body weights, which may have negated the effect of the transgene to cause glucose intolerance. Therefore, in this current study we wanted to produce mice overexpressing FBPase specifically in the liver. To accomplish this, the new transgenic mouse line (line 5) was generated with a construct that incorporated insulator sequences to confer specific expression to the liver by protecting the transgenic cassette from the position of integration effects when randomly inserted into the host genome (7). We achieved the desired liver transgene specificity without an increased expression in the hypothalamus and proposed that the use of insulator sequences may be a reasonable means by which a transgene can be protected from site-of-integration effects. Interestingly, line 5 transgenic mice still displayed reduced body weight and food intake compared with their negative littermates, which raises the possibility of a role for the liver in body weight and food intake regulation.
The effect of a chronic HF diet was tested, because, as mentioned earlier, increased levels of FBPase have been associated with glucose intolerance and increased rates of EGP when the animals were either obese or placed on a HF diet. With the introduction of high amounts of fat into their diet, the transgenic mice had much higher FBPase levels at 24 wk than the aged transgenic mice and became glucose intolerant compared with the negative littermate mice on the same diet. This suggested that lipid oversupply may be necessary for a glucose intolerant phenotype to be observed in these hemizygous FBPase transgenic mice. Despite the fact that a HF diet upregulated FBPase, many other genes are also upregulated with a HF diet including PEPCK and G-6-Pase, as well as increased lipid substrates such as triglycerides. Thus we believe that additional effects of the HF diet contributed to the impairment in glucose tolerance observed in the transgenic mice. The triglyceride levels in the livers of transgenic mice on a HF diet were higher compared with the negative mice on a HF diet, which may have contributed to the observed glucose intolerant phenotype.
Interestingly, we did observe a gene-dose effect on glucose metabolism, with homozygous line 5 transgenic mice developing mild, but significant, glucose intolerance and a substantially elevated EGP. This provides evidence that FBPase in the homozygous state can result in increased EGP, without the concomitant effects of HF feeding. While upregulated FBPase in the homozygous state appears to have a role, this is only evident when this enzyme is significantly elevated. When it is modestly elevated (2- to 3-fold as in these transgenic mice), it is clear that FBPase is not rate determining, and therefore, EGP is little affected. This implies that control of gluconeogenesis and EGP may reside elsewhere. It is possible that other gluconeogenic enzymes such as PEPCK may be more influential on the flux rate of gluconeogenesis and EGP. This is evident as mice overexpressing PEPCK in the liver were reported to have increased EGP in the hemizygous state and fasting glucose and insulin levels were double compared with controls with a significant impairment in glucose tolerance (38, 43). Furthermore, rats overexpressing G-6-Pase in the liver also showed glucose intolerance and hyperinsulinemia without the need for further stress on the system, such as a HF diet (42). We therefore suggest that under normal conditions FBPase may not be rate determining in EGP and that there may be a hierarchy of importance for these gluconeogenic enzymes, with FBPase being less important compared with PEPCK and G-6-Pase.
In summary, our study provides the first in vivo evidence in a transgenic mouse model that confirms a specific upregulation of FBPase in the liver causes glucose intolerance and a significant elevation in EGP. Therefore, if increased levels of FBPase in human type 2 diabetes are high enough to contribute to the observed increase in EGP, FBPase inhibitors would be a useful therapy.
|
GRANTS |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
* J. M. Favaloro and S. Andrikopoulos contributed equally to this work. 
|
REFERENCES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
), n = 21 for line 1 TG (gray circle), and n = 17 for line 5 TG (
)]. Dashed lines represent the HFF mice [n = 18 for NEG (
), n = 9 for line 1 TG (gray triangle), and n = 6 for line 5 TG (
)]. E: plasma glucose data during IPGTT expressed as AUC (*P < 0.05 vs. all other groups). F: plasma insulin levels expressed as AUC (*P < 0.03 vs. NEG chow-fed mice). G: liver PEPCK mRNA levels in chow and HFF mice (*P < 0.05 vs. NEG and line 5 TG). H: liver G-6-Pase mRNA levels in chow and HFF mice (*P < 0.05 vs. NEG and line 5 TG).
