Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Mar 11, 2008; 105(10): 3921–3926.
Published online Mar 5, 2008. doi:  10.1073/pnas.0800293105
PMCID: PMC2268825
Medical Sciences

Txnip balances metabolic and growth signaling via PTEN disulfide reduction


Thioredoxin-interacting protein (Txnip) inhibits thioredoxin NADPH-dependent reduction of protein disulfides. Total Txnip knockout (TKO) mice adapted inappropriately to prolonged fasting by shifting fuel dependence of skeletal muscle and heart from fat and ketone bodies to glucose. TKO mice exhibited increased Akt signaling, insulin sensitivity, and glycolysis in oxidative tissues (skeletal muscle and hearts) but not in lipogenic tissues (liver and adipose tissue). The selective activation of Akt in skeletal muscle and hearts was associated with impaired mitochondrial fuel oxidation and the accumulation of oxidized (inactive) PTEN, whose activity depends on reduction of two critical cysteine residues. Whereas muscle- and heart-specific Txnip knockout mice recapitulated the metabolic phenotype exhibited by TKO mice, liver-specific Txnip knockout mice were similar to WT mice. Embryonic fibroblasts derived from knockout mice also accumulated oxidized (inactive) PTEN and had elevated Akt phosphorylation. In addition, they had faster growth rates and increased dependence on anaerobic glycolysis due to impaired mitochondrial fuel oxidation, and they were resistant to doxorubicin-facilitated respiration-dependent apoptosis. In the absence of Txnip, oxidative inactivation of PTEN and subsequent activation of Akt attenuated mitochondrial respiration, resulting in the accumulation of NADH, a competitive inhibitor of thioredoxin NADPH-reductive activation of PTEN. These findings indicate that, in nonlipogenic tissues, Txnip is required to maintain sufficient thioredoxin NADPH activity to reductively reactivate oxidized PTEN and oppose Akt downstream signaling.

Keywords: mitochondrial respiration, redox

Signals regulating metabolism and growth are coordinately integrated via the phosphoinositide 3-kinase (PI3K)/Akt signaling pathway (1). The pool size of phosphatidylinositol (3,4,5)-trisphosphate (PIP3) is regulated by the relative activities of PI3K and the phosphatidylinositol 3-phosphatase PTEN (2, 3). Oxidation of a small fraction of critical cysteine residues inactivates PTEN, thereby amplifying insulin and growth factor receptor signaling via expansion of the PIP3 pool, which activates the downstream signaling kinase Akt (4). After PTEN′s rapid reactivation via thioredoxin NADP(H)-mediated disulfide reduction, the transiently amplified PIP3 signal is attenuated by the phosphatase action of PTEN, allowing Akt activity to be restored to lower (basal) levels (5, 6).

The physiological synergy resulting from the confluence of metabolic and growth control signaling is epitomized by the anaerobic survival advantage provided by the cancer cell phenotype described by Warburg (7). As a result of diminished mitochondrial respiration, cancer cells rely almost exclusively on glucose as the fuel through anaerobic glycolysis, allowing them to proliferate in relatively hypoxic environments (7). Altered growth regulation has been linked to impaired mitochondrial respiration associated with the Warburg cancer cell phenotype via the activation of Akt caused by the accumulation of NADH, which indirectly inactivates PTEN by blocking thioredoxin NADP(H)-dependent activation of PTEN (8), whose redox-sensitive phosphatase activity is essential for opposing the activation of Akt (2, 9).

We show that the tumor suppressor thioredoxin-interacting protein (Txnip), a negative regulator of thioredoxin NADP(H)-dependent reduction of disulfide bonds (10), links mitochondrial respiration with glycolysis and cell growth (Warburg phenotype) by reductive reactivation of PTEN to levels essential to oppose Akt. Txnip was first identified as the gene at the murine Hyplip1 locus responsible for ketosis and hyperlipidemia (11). Because of a nonsense mutation in Txnip, Hyplip-1 mice do not express any functionally active Txnip protein (11) and become markedly hypertriglyceridemic and hypoglycemic when subjected to prolonged fasting (12). In nonlipogenic, oxidative tissue (i.e., skeletal muscle and hearts), Txnip deficiency impairs the ability of thioredoxin NADP(H)-dependent disulfide reduction to maintain PTEN in a sufficiently active state to oppose Akt downstream signaling. Unopposed, Akt activation in Txnip-deficient skeletal muscle and hearts prevents the switching of fuel away from glucose toward 3-hydroxybutyrate and fatty acids, resulting in an inappropriate metabolic adaptation to food deprivation (12, 13).


Metabolic Effects of Txnip Ablation Are Mediated Through Extrahepatic Tissues.

To gain an understanding of how Txnip facilitates appropriate adaptation to food deprivation, we used Cre-loxP-mediated gene recombination (14) to generate total and tissue-specific Txnip knockout mice. Details on derivation of Txnip floxed and knockout mice are described in supporting information (SI) Text. Genomic analysis and characterization of Txnip mRNA expression in the knockout mice are provided in SI Figs. 7 and 8. A genomic sequence containing the entire coding region of Txnip was subcloned into the targeting vector pNTK(A) (15) (a gift from Mark Magnuson, Vanderbilt University, Nashville, TN). Homozygous floxed (Txnipfl/fl) mice were used to produce total Txnip knockout (TKO), liver-specific Txnip knockout (LKO), and muscle-specific Txnip knockout (MKO) mice by breeding them, respectively, with zp3-cre mice (16), alb-cre mice (17), and mck-cre mice (18). Progeny were then backcrossed to C57BL/6 mice. TKO, LKO, and MKO mice showed no abnormalities in growth, viability, and reproduction.

Consistent with the phenotypes of other Txnip-deficient mice (1113), fasting TKO mice were hypertriglyceridemic (321.9 ± 71.4 mg/dl), hyperketonemic (5.6 ± 0.4 mM), and hypoglycemic (56.6 ± 12.0 mg/dl) (Fig. 1 A–C). Because the plasma insulin and glucagon levels of WT and TKO mice were similar (SI Fig. 9), the lower plasma glucose levels in TKO mice suggested that they are more insulin-sensitive. LKO mice were derived by breeding Txnipflox/flox mice with mice expressing cre recombinase driven by the hepatocyte-specific albumin promoter (15) and shown to exhibit liver-specific deletion of the Txnip gene (SI Fig. 8B). Fasted LKO mice did not show any of the metabolic alterations exhibited by TKO mice (Fig. 1 A–C), indicating that the loss of Txnip function in extrahepatic tissues is responsible for the fasting-induced metabolic alternations observed in TKO mice. This conclusion is further supported by the observation that fasted mice having selective deletion of Txnip in skeletal muscle and hearts (SI Fig. 8C) displayed a phenotype similar, albeit less severe, to that of TKO mice (Fig. 1 D–F): hypertriglyceridemia (223.3 ± 86.2 mg/dl), hyperketonemia (5.3 ± 1.0 mM), and hypoglycemia (82.8 ± 10.3 mg/dl). These data indicate that extrahepatic (skeletal muscle and heart) deficiency of Txnip is responsible for the inability of mice to appropriately modify fuel metabolism in response to fasting.

Fig. 1.
Plasma metabolic profile of fasting TKO, LKO, and MKO mice. (A–C) Plasma levels of triglycerides (A), 3-hydroxybutyrate (B), and glucose (C) in fasted WT, TKO, and LKO mice were determined (n = 9 per group). Results are presented as mean ± ...

Txnip Ablation Increases Extrahepatic Tissue Insulin Sensitivity and Protects Against High-Fat Diet-Induced Insulin Resistance.

Intraperitoneal glucose tolerance tests revealed that TKO and MKO mice were more glucose-tolerant than the WT mice (Fig. 2A). This interpretation was also supported by hyperinsulinemic–euglycemic clamp studies (Fig. 2 B–E) (19). Compared with WT littermates, TKO mice required higher rates of glucose infusion to maintain euglycemia during clamps (Fig. 2B), which corresponded to a 30% increase (P < 0.01) in insulin-stimulated whole-body glucose turnover (Fig. 2C). Both basal hepatic glucose production (+50%, P < 0.01) (Fig. 2D) and skeletal muscle insulin-stimulated glucose uptake (+68%, P < 0.01) (Fig. 2E) are increased in the TKO mice, suggesting that deletion of Txnip reduces hepatic insulin response while enhancing extrahepatic (skeletal muscle) insulin sensitivity.

Fig. 2.
Effects of Txnip ablation on glucose homeostasis and insulin sensitivity. (A) An i.p. glucose tolerance test (1 g/kg of body weight) was performed with fasted WT mice (filled diamonds), TKO mice (open squares), and MKO mice (filled triangles) (n = 4 per ...

Analysis of Akt phosphorylation, as an index of insulin sensitivity, showed that, in soleus muscle and hearts, deletion of Txnip was associated with increased Akt phosphorylation at Thr-308 in both basal and insulin-stimulated states (Fig. 3A). These findings are consistent with studies showing that Txnip regulates both the insulin-dependent and insulin-independent pathways of glucose uptake in human skeletal muscle (20).

Fig. 3.
Effects of Txnip ablation on Akt phosphorylation in chow and high-fat diet-fed TKO mice. (A) Insulin-stimulated Akt phosphorylation in soleus muscle and hearts of fasted mice (n = 4 per group) was measured. Samples were subject to Western blot analysis ...

The additional finding that Txnip deletion did not increase Akt phosphorylation in either liver (Fig. 3B) or white adipose tissue (Fig. 3C) indicates that the process responsible for the activation of Akt in response to Txnip deficiency occurs in oxidative tissue (skeletal muscle and hearts) but not in lipogenic tissue (liver and white adipose tissue). The concordant inability of Txnip deletion in liver to alter the metabolic phenotype (Fig. 1 A–C) or to activate Akt (Fig. 3B) supports the proposal that Txnip enhances glucose uptake and tolerance via activating Akt in oxidative tissue.

We examined whether increased insulin sensitivity in skeletal muscle and heart would protect TKO mice from high-fat diet-induced insulin resistance (21). After 12 weeks of high-fat diet consumption, there was no significant difference in weight gain between the groups (data not shown). Fasting plasma glucose levels in WT mice were increased by 80% (P < 0.01), from 100 to ≈180 mg/dl (Fig. 3D). In marked contrast, in TKO mice fasting plasma glucose levels were decreased by 40% (P < 0.01) (Fig. 3D). Although plasma insulin levels were not different between the groups (data not shown), phospho-Akt levels were two to three times greater in the heart and soleus muscle of TKO mice compared with those of control mice (Fig. 3E). These data suggest that extrahepatic insulin sensitivity is enhanced in TKO mice and that activation of Akt protects TKO mice against high-fat diet-induced insulin resistance.

Txnip Ablation Leads to Impaired Mitochondrial Fuel Oxidation and Increased Glycolysis.

Although both groups of mice exhibited similar rates of uptake of [114C]-d,l-3-hydroxybutyrate (Fig. 4A), soleus muscle from TKO mice displayed markedly impaired (60%, P < 0.01) oxidation of [114C]-d,l-3-hydroxybutyrate to 14CO2 (Fig. 4B). Moreover, soleus muscle from TKO mice also displayed impaired oxidation of both 14C-glucose (−43%, P < 0.05) (Fig. 4C) and [114C]oleic acid (−48%, P < 0.05) (Fig. 4D) to 14CO2, indicating that Txnip deletion impairs mitochondrial oxidation of all major fuels. Furthermore, elevated plasma lactate levels in fasted TKO mice (≈40%, P < 0.01) (Fig. 4E) suggest that impaired mitochondrial fuel oxidation caused fasted TKO mice to become more dependent on anaerobic glycolysis for energy.

Fig. 4.
Effects of Txnip ablation on glucose homeostasis, mitochondrial fuel metabolism, and PTEN oxidation. (A) Fasted mice were injected i.v. with 0.5 μCi of [1-14C]-3-hydroxybutyrate (n = 5 per group). After 30 min, skeletal muscle was isolated from ...

Txnip Ablation Increases PTEN Oxidation.

Because Txnip functions as a negative regulator of thioredoxin NADPH disulfide reduction (10), we sought to identify an established target of thioredoxin NADPH reduction that could account for the altered metabolic phenotype exhibited by oxidative tissue (i.e., skeletal muscle and hearts) of Txnip-deficient mice. The discovery that thioredoxin NADP(H)-dependent reduction is required to maintain the activity of PTEN (5, 6, 8) suggested a possible mechanistic link between Txnip ablation and the development of the respiration-deficient, glycolysis-dependent phenotype of skeletal muscle and hearts from fasted TKO mice. The active site of tumor suppressor PTEN contains two cysteines (Cys-71 and Cys-124) that must remain reduced to maintain phosphatase activity (9). Oxidation of these cysteines in response to activation of tyrosine kinases associated with insulin and growth factor receptors results in the formation of a disulfide bond, inhibiting PTEN activity (4). PTEN is then reactivated by thioredoxin NADP(H)-dependent reduction of PTEN active-site cysteines (5, 6, 8). This oscillatory oxidation–reduction of PTEN has been proposed to amplify PIP3 activation of Akt (4). Because NADH competitively inhibits thioredoxin NADP(H)-dependent reactivation of PTEN, increased NADH/NADP(H) in cancer cells caused by impaired mitochondrial respiration blocks the reductive reactivation of PTEN, thereby activating Akt (8). This proposal predicts that impaired mitochondrial oxidation would inactivate PTEN because of the inability to reduce its dual cysteine active site by thioredoxin NADP(H) reduction.

Indeed, comparison of the relative levels of reduced and nonreduced PTEN using Western blot analyses showed that soleus muscle from TKO mice, but not WT mice, accumulated oxidized (inactive) PTEN (Fig. 4F). After exhaustive reduction by 2-mercaptoethanol, the amount of immunodetectable PTEN was similar in soleus muscle homogenates from WT and TKO mice (Fig. 4F). In contrast, in the absence of disulfide reduction (no 2-mercaptoethanol added) the same soleus muscle samples from TKO mice exhibited an ≈50% reduction in immunodetectable PTEN compared with that of WT mice (Fig. 4F). These data clearly show that TKO soleus muscle extracts contain markedly more PTEN epitopes that were masked by conformational changes due to disulfide oxidation (i.e., disulfide reduction with 2-mercaptoethanol exposed these epitopes). Skeletal muscle-specific PTEN-ablated mice exhibit a phenotype similar to that of the TKO mice, further supporting the conclusion that Txnip affects Akt and insulin signaling by regulating the disulfide-reductive reactivation of oxidized PTEN (22).

Mouse Embryonic Fibroblasts (MEFs) from TKO Mice Exhibit Metabolic and Growth Characteristics of Warburg Cancer Cells.

To determine whether the activation of Akt in response to Txnip deficiency is sufficient to induce growth, MEFs were isolated from WT and TKO embryos and their growth and metabolic characteristics were compared. Similar to skeletal muscle and heart of TKO mice, MEFs lacking Txnip accumulated oxidized (inactive) PTEN (Fig. 5A) and activated phospho-Akt (Ser-473) (Fig. 5B). TKO fibroblasts replicated faster than WT fibroblasts (Fig. 6A); the exponential growth rate constants (k) for WT and TKO fibroblasts were 0.008 and 0.019 h−1, respectively. The additional findings showing that TKO MEFs exhibited increased glucose uptake (Fig. 6B), but reduced (80%) mitochondrial oxidation of 14C-glucose to 14CO2 (Fig. 6C), causing their medium to accumulate lactate (Fig. 6D), suggested that they depended on anaerobic glycolysis for energy in a manner similar to Txnip-deleted skeletal muscle and hearts.

Fig. 5.
Accumulation of oxidized PTEN and phospho-Akt in MEFs. (A) Protein extracts of MEFs from WT and TKO mice (n = 4 per group) were prepared and analyzed as described in Fig. 4F. Results are presented as mean ± SD. *, P < 0.05 (versus WT). ...
Fig. 6.
Expression of Warburg cancer cell phenotype by MEFs derived from TKO mice. (A) MEFs from WT (filled diamonds) and TKO (open squares) mice were seeded at the same density. The number of viable cells in each plate was counted at the time indicated. Each ...

Reduced mitochondrial respiration has been shown to enhance the resistance of cancer cells to the chemotherapeutic agent doxorubicin through activation of Akt (8). Similar to this finding, embryonic fibroblasts lacking Txnip displayed resistance to doxorubicin (Fig. 6E). These findings indicate that deletion of Txnip, previously identified as a tumor suppressor (23), is sufficient to induce a phenotype that is reminiscent of cancer cells, described by Warburg (7).


Tissue-specific deletion of functionally active Txnip showed that the inability of Txnip-null mice to appropriately adapt to prolonged food deprivation (12, 13) is due to impaired mitochondrial oxidation of fuels by extrahepatic tissue (soleus muscle and heart). Moreover, the observation that accumulation of oxidized (inactive) PTEN, an established substrate of the thioredoxin NADP(H) disulfide reduction (46), varied in parallel with the activation state of Akt provides a mechanistic explanation of how Txnip acts in concert with redox-facilitated regulation of insulin action, mitochondrial oxidation, and cell growth (9).

Previous studies have shown that oxidized inactivation of PTEN due to impaired mitochondrial respiration (8) or interruption of PTEN gene expression (22) increased Akt activation and glucose utilization. The active site of tumor suppressor PTEN contains two cysteines (Cys-71 and Cys-124) that must be reduced by thioredoxin NADP(H) disulfide reduction (46) to maintain phosphatase activity (2, 9). Recent studies showed that impairing mitochondrial respiration indirectly inactivates PTEN because of the accumulation of NADH, which competitively blocks NADPH-dependent thioredoxin-reductive PTEN activation (8), providing an explanation for the accumulation of oxidized (inactive) PTEN in skeletal muscle (Fig. 4F) and MEFs (Fig. 5A) of TKO mice, because both exhibit impaired mitochondrial fuel oxidation (Figs. 4 B–D and and66C, respectively). The data indicating that both mice (Fig. 2E) and MEFs (Fig. 6B) lacking Txnip exhibit increased glucose uptake and glycolytic formation of lactate further support the conclusion that NADH accumulated as a consequence of impaired mitochondrial fuel oxidation (Figs. 4 B–D and and66C). Both inactivation of PTEN [by reducing the activity of its downstream target PTEN-induced putative kinase 1 (PINK1) (2428)] and activation of Akt via Hif-1α-dependent (29, 30) and -independent (31) mechanisms impair mitochondrial respiration. Whether the Txnip-deficient phenotype is initiated by reduced mitochondrial respiration or by the inactivation of PTEN is unclear.

Our findings that Txnip ablation activates Akt in nonlipogenic tissues (soleus muscle and hearts; Fig. 3A) but not in lipogenic tissues (liver and adipose tissue; Fig. 3 B and C) suggest that the relative tissue capacity to generate NADPH through the pentose phosphate pathway may determine the susceptibility of thioredoxin NADP(H)-reductive activation of PTEN to NADH (anaerobic) inhibition. It seems likely that in nonlipogenic, oxidative tissue (e.g., skeletal muscle and heart), having reduced SREBP-mediated expression of genes important for NADPH production (32), Txnip acts to preserve NADPH by inhibiting thioredoxin NADP(H)-dependent reactions (10). Thus, we hypothesize that Txnip-mediated conservation of NADPH maintains essential thioredoxin NADP(H)-reductive functions [e.g., thioredoxin-reductive reactivation of oxidized PTEN (46, 33)] and prevents oxidation-induced injury (34, 35).

There is accumulating evidence that mitochondrial dysfunction may contribute to insulin resistance in diabetes-prone individuals (36). However, other studies suggest that mitochondrial respiratory impairment enhances insulin action and may be an effective therapeutic target for type 2 diabetes (37). This discrepancy may be reconciled by considering the difference in the underlying mechanisms leading to impaired mitochondrial function: decreased mitochondrial number versus decreased respiration rate (associated with accumulation of NADH). The accumulation of NADH, caused by impaired mitochondrial respiration, competes with NADPH, thereby inhibiting thioredoxin NADPH-dependent reductive activation of PTEN (8). As a result, impaired mitochondrial respiration increases Akt activation and insulin action. In contrast, diminished mitochondrial respiratory capacity caused by reduced mitochondria number would not necessarily cause NADH accumulation (36).

Our findings show that Txnip integrates energy metabolism and growth regulation by reciprocally modulating the activation states of PTEN and Akt by sensing the quantity and quality of electron transfer cofactors (e.g., NADH versus NADPH).

Materials and Methods

Animals and Diet.

Animals were kept in a 12-h dark–light cycle and fed standard rodent chow ad libitum unless otherwise stated. For the high-fat diet experiment, mice were fed a high-fat diet consisting of 60% fat, 20% carbohydrate, and 20% protein (D12492; Research Diets) for 12 weeks. All procedures described were approved by the Institutional Animal Care and Use Committee.

Plasma Metabolite Assays.

Blood was collected at 1,000 h from age-matched fasted animals in heparinized capillary tubes. Plasma samples were frozen in aliquots and stored at −20°C. Glucose and triglyceride levels were measured by using colorimetric kits (Wako). Plasma 3-hydroxybutyrate levels were determined by using an enzymatic assay kit from Stanbio Laboratory. Lactate was measured by using a colorimetric kit (Sigma).

Glucose Tolerance Test.

Mice were fasted for 18 h before i.p. injection with glucose (1 g/kg of body weight). Blood was collected from the tail, and glucose levels were determined by using a glucometer (Becton Dickinson).

Hyperinsulinemic–Euglycemic Clamps in Awake Mice.

After an overnight fast (≈18 h), a 2-h hyperinsulinemic–euglycemic clamp (150 milliunits/kg of body weight insulin prime followed by 2.5 milliunits/kg per minute insulin infusion) was performed in awake male TKO and WT C57BL/6 mice (38). Basal and insulin-stimulated whole-body glucose turnover was estimated with a continuous infusion of [33H]glucose before and during clamps. A bolus i.v. administration of 2-deoxy-d-[114C]glucose (NEN) was used during clamps to assess insulin-stimulated glucose uptake in skeletal muscle. Rates of basal hepatic glucose production and insulin-stimulated whole-body and skeletal muscle glucose uptake were determined as previously described (38). All procedures were performed at the Pennsylvania State University Mouse Metabolic Phenotyping Center and approved by the Pennsylvania State University Animal Care and Use Committee.

Oxidation of Glucose, 3-Hydroxybutyrate, and Oleate in Isolated Muscle.

Oxidation of glucose by isolated soleus muscle was performed as described (39). For 3-hydroxybutyrate oxidation, the incubation medium contained 16 mM d,l-3-hydroxybutyrate (0.5 μCi/ml), whereas, for oleate oxidation experiments, 4% BSA and 0.5 mM oleate (0.5 μCi/ml) were used. Vials were sealed with a rubber stopper, which was fitted with a center well. After 60 min, 0.35 ml of 1 M hyamine hydroxide (PerkinElmer Life Sciences) was injected through the rubber stopper into the center well, and the medium was acidified by injection of 1/10 volume of 1 M sulfuric acid. Liberated CO2 was collected overnight, and center wells were transferred to vials for liquid scintillation counting.

Measurement of Phospho-Akt Levels.

For the determination of Akt protein phosphorylation in soleus muscle, samples were collected 8 min after mice were injected (i.p.) with insulin (10 units/kg of body weight) (40). The tissue samples were homogenized in buffer A (100 mM Tris, pH 7.4/250 mM sucrose/1 mM sodium pyrophosphate/1 mM sodium orthovanadate/10 μg/ml leupeptin/10 μg/ml aprotinin/1 μM microcystin/1 mM PMSF/10 mM sodium fluoride). Tissue homogenates were centrifuged at 14,000 × g for 10 min at 4°C. Proteins in the supernatant were subjected to Western blot analysis with anti-total Akt and anti-phospho-Akt (Thr-308) antibody (Cell Signaling Technology), respectively.

Immunoblot of PTEN Under Reducing and Nonreducing Conditions.

MEFs or soleus muscle from male mice fasted overnight (14 weeks old, four per group) were homogenized in buffer containing 50 mM Na2HPO4 (pH 7.0), 1 mM EDTA, 10 mM N-ethylmaleimide, 10 mM iodoacetic acid, 5 mM NaF, 50 μg/ml aprotinin, 50 μg/ml leupeptin, and 1% Triton X-100. Twenty micrograms of protein per lane was resolved by SDS/PAGE under reducing and nonreducing conditions and then transferred onto a PVDF membrane. Blots were probed first with rabbit polyclonal antibodies against PTEN (Cell Signaling Technology) and then stripped and reprobed with mouse antibodies against tubulin (Sigma) as a loading control. The amount of protein was quantified by using ImageQuant software.

Cell Culture of Primary Embryonic Fibroblasts.

Primary embryonic fibroblasts were derived from 14-day-old embryos from three WT and three TKO mice. Differential expression of Txnip mRNA was confirmed by real-time quantitative PCR (SI Fig. 10). Primer sequences are listed in SI Text. Cells were maintained in DMEM (Mediatech) with 25 mM glucose, 4 mM l-glutamine, 1 mM sodium pyruvate, 100 units/ml penicillin and streptomycin, and 10% heat-inactivated FBS. For growth curve experiments, cells were seeded at a density of 7 × 104 cells per 60-mm plate. Every 24 h, the number of cells per plate was counted in quadruplicate for a total of 5 days. The medium was changed every 48 h. First-order rate constant was determined by using nonlinear least squares regression. Lactate concentration in the medium was determined by using a Lactate kit (Sigma) according to the manufacturer's instructions.

2-Deoxyglucose Uptake in MEFs.

MEFs were cultured to 80% confluence and rinsed twice with uptake buffer containing 140 mM NaCl, 2 mM KCl, 1 mM KH2PO4, 10 mM MgCl2, 1 mM CaCl, 5 mM l-alanine, and 10 mM Hepes/Tris (pH 7.4). The cells were then incubated with 0.5 μCi/ml [1-14C]-2-deoxyglucose at 37°C for 1 h.

Oxidation of Glucose by MEFs.

MEFs derived from WT and TKO mice were grown in T-25 flasks until 80% confluence. Cells were rinsed twice with Krebs–Henseleit buffer (5 mM glucose, 1.2 mM potassium phosphate, 1.2 mM magnesium sulfate, 4.7 mM sodium chloride, 2.5 mM calcium chloride, and 25 mM sodium bicarbonate) and incubated in 2 ml of Krebs–Henseleit buffer containing 0.5 μCi/ml [U-14C]glucose at 37°C for 2 h. At the end of incubation, medium was acidified with 300 μl of 1 M HCl, and carbon dioxide released was trapped with 350 μl of hyamine hydroxide. Radioactivity in CO2 was counted by using a scintillation counter. Protein content was determined from two mirror flasks and used for normalization of scintillation counts.

Survival of MEFs After Doxorubicin Treatment.

MEFs were grown until 80% confluence. Cells were incubated in medium containing 0.1 μM doxorubicin (Sigma) for 48 h. Percentage of survival was calculated by dividing viable cell counts from doxorubicin-exposed plates to control plate counts.

Statistical Analysis.

Statistical comparison of control versus Txnip knockout mice was done by using the unpaired two-tailed Student t test. Data are presented as mean ± SD for each group. Significance was accepted at P < 0.05.

Supplementary Material

Supporting Information:


We thank Ron Smith for help with genotyping of the mice and performing the initial phenotype characterization; Zhiyou Zhang and Hwi Jin Ko for help with the glucose clamp experiments; C. Ronald Kahn (Joslin Diabetes Center, Boston) for the mck-cre mice; Mark Magnuson for pNTK(A) vector; and Alan Attie, Jake Lusis, Peter Edwards, and Joe Witztum for helpful editorial comments. R.A.D. is a member of the Moores Cancer Center, University of California at San Diego, where the embryonic stem cells were transfected, screened, and cultured and gene-targeted mice were derived. This work was supported by the National Institutes of Health, the American Heart Association, the American Diabetes Association, a fellowship grant from the Rees-Steely Foundation, and the Pennsylvania State Department of Health (Tobacco Formula Fund).


The authors declare no conflict of interest.

This article contains supporting information online at www.pnas.org/cgi/content/full/0800293105/DC1.


1. Alessi DR, et al. Mechanism of activation of protein kinase B by insulin and IGF-1. EMBO J. 1996;15:6541–6551. [PMC free article] [PubMed]
2. Maehama T, Dixon JE. The tumor suppressor, PTEN/MMAC1, dephosphorylates the lipid second messenger, phosphatidylinositol 3,4,5-trisphosphate. J Biol Chem. 1998;273:13375–13378. [PubMed]
3. Cantley LC. The phosphoinositide 3-kinase pathway. Science. 2002;296:1655–1657. [PubMed]
4. Kwon J, et al. Reversible oxidation and inactivation of the tumor suppressor PTEN in cells stimulated with peptide growth factors. Proc Natl Acad Sci USA. 2004;101:16419–16424. [PMC free article] [PubMed]
5. Meuillet EJ, Mahadevan D, Berggren M, Coon A, Powis G. Thioredoxin-1 binds to the C2 domain of PTEN inhibiting PTEN′s lipid phosphatase activity and membrane binding: A mechanism for the functional loss of PTEN′s tumor suppressor activity. Arch Biochem Biophys. 2004;429:123–133. [PubMed]
6. Lee SR. Reversible inactivation of the tumor suppressor PTEN by H2O2. J Biol Chem. 2002;277:20336–20342. [PubMed]
7. Warburg O. On respiratory impairment in cancer cells. Science. 1956;124:269–270. [PubMed]
8. Pelicano H, et al. Mitochondrial respiration defects in cancer cells cause activation of Akt survival pathway through a redox-mediated mechanism. J Cell Biol. 2006;175:913–923. [PMC free article] [PubMed]
9. Maehama T, Taylor GS, Dixon JE. PTEN and myotubularin: Novel phosphoinositide phosphatases. Annu Rev Biochem. 2001;70:247–279. [PubMed]
10. Nishiyama A. Identification of thioredoxin-binding protein-2/vitamin D (3) up-regulated protein 1 as a negative regulator of thioredoxin function and expression. J Biol Chem. 1999;274:21645–21650. [PubMed]
11. Bodnar JS, et al. Positional cloning of the combined hyperlipidemia gene Hyplip1. Nat Genet. 2002;30:110–116. [PMC free article] [PubMed]
12. Hui TY, et al. Mice lacking thioredoxin interacting protein provide evidence linking cellular redox state to appropriate response to nutritional signals. J Biol Chem. 2004;279:24387–24393. [PubMed]
13. Oka S, et al. Impaired fatty acid utilization in thioredoxin binding protein-2 (TBP-2)-deficient mice: A unique animal model of Reye syndrome. FASEB J. 2006;20:121–123. [PubMed]
14. Gu H, Marth JD, Orban PC, Mossmann H, Rajewsky K. Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science. 1994;265:103–106. [PubMed]
15. Postic C, et al. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J Biol Chem. 1999;274:305–315. [PubMed]
16. Lewandoski M, Wassarman KM, Martin GR. Zp3-cre, a transgenic mouse line for the activation or inactivation of loxP-flanked target genes specifically in the female germ line. Curr Biol. 1997;7:148–151. [PubMed]
17. Ikegami T, et al. Model mice for tissue-specific deletion of the manganese superoxide dismutase (MnSOD) gene. Biochem Biophys Res Commun. 2002;296:729–736. [PubMed]
18. Bruning JC, et al. A muscle-specific insulin receptor knockout exhibits features of the metabolic syndrome of NIDDM without altering glucose tolerance. Mol Cell. 1998;2:559–569. [PubMed]
19. Cho H, et al. Insulin resistance and a diabetes mellitus-like syndrome in mice lacking the protein kinase Akt2 (PKB beta). Science. 2001;292:1728–1731. [PubMed]
20. Parikh H, et al. TXNIP regulates peripheral glucose metabolism in humans. PLoS Med. 2007;4:e158. [PMC free article] [PubMed]
21. An J, et al. Hepatic expression of malonyl-CoA decarboxylase reverses muscle, liver and whole-animal insulin resistance. Nat Med. 2004;10:268–274. [PubMed]
22. Wijesekara N, et al. Muscle-specific Pten deletion protects against insulin resistance and diabetes. Mol Cell Biol. 2005;25:1135–1145. [PMC free article] [PubMed]
23. Goldberg SF, et al. Melanoma metastasis suppression by chromosome 6: Evidence for a pathway regulated by CRSP3 and TXNIP. Cancer Res. 2003;63:432–440. [PubMed]
24. Scheele C, et al. The human PINK1 locus is regulated in vivo by a noncoding natural antisense RNA during modulation of mitochondrial function. BMC Genomics. 2007;8:74. [PMC free article] [PubMed]
25. Dodson MW, Guo M. Pink1, Parkin, DJ-1 and mitochondrial dysfunction in Parkinson's disease. Curr Opin Neurobiol. 2007;17:331–337. [PubMed]
26. Abeliovich A. Parkinson's disease: Prosurvival effects of PINK1. Nature. 2007;448:759–760. [PubMed]
27. Silvestri L, et al. Mitochondrial import and enzymatic activity of PINK1 mutants associated to recessive parkinsonism. Hum Mol Genet. 2005;14:3477–3492. [PubMed]
28. Healy DG, et al. The gene responsible for PARK6 Parkinson's disease, PINK1, does not influence common forms of parkinsonism. Ann Neurol. 2004;56:329–335. [PubMed]
29. Papandreou I, Cairns RA, Fontana L, Lim AL, Denko NC. HIF-1 mediates adaptation to hypoxia by actively downregulating mitochondrial oxygen consumption. Cell Metab. 2006;3:187–197. [PubMed]
30. Pore N, et al. Akt1 activation can augment hypoxia-inducible factor-1alpha expression by increasing protein translation through a mammalian target of rapamycin-independent pathway. Mol Cancer Res. 2006;4:471–479. [PubMed]
31. Gottlob K, et al. Inhibition of early apoptotic events by Akt/PKB is dependent on the first committed step of glycolysis and mitochondrial hexokinase. Genes Dev. 2001;15:1406–1418. [PMC free article] [PubMed]
32. Horton JD, Goldstein JL, Brown MS. SREBPs: Activators of the complete program of cholesterol and fatty acid synthesis in the liver. J Clin Invest. 2002;109:1125–1131. [PMC free article] [PubMed]
33. Song Z, Saghafi N, Gokhale V, Brabant M, Meuillet EJ. Regulation of the activity of the tumor suppressor PTEN by thioredoxin in Drosophila melanogaster. Exp Cell Res. 2007;313:1161–1171. [PMC free article] [PubMed]
34. Nishinaka Y, Masutani H, Nakamura H, Yodoi J. Regulatory roles of thioredoxin in oxidative stress-induced cellular responses. Redox Rep. 2001;6:289–295. [PubMed]
35. Schulze PC, De Keulenaer GW, Yoshioka J, Kassik KA, Lee RT. Vitamin D3-upregulated protein-1 (VDUP-1) regulates redox-dependent vascular smooth muscle cell proliferation through interaction with thioredoxin. Circ Res. 2002;91:689–695. [PubMed]
36. Lowell BB, Shulman GI. Mitochondrial dysfunction and type 2 diabetes. Science. 2005;307:384–387. [PubMed]
37. Pospisilik JA, et al. Targeted deletion of AIF decreases mitochondrial oxidative phosphorylation and protects from obesity and diabetes. Cell. 2007;131:476–491. [PubMed]
38. Kim H-J, et al. Differential effects of interleukin-6 and -10 on skeletal muscle and liver insulin action in vivo. Diabetes. 2004;53:1060–1067. [PubMed]
39. Hansen PA, Gulve EA, Holloszy JO. Suitability of 2-deoxyglucose for in vitro measurement of glucose transport activity in skeletal muscle. J Appl Physiol. 1994;76:979–985. [PubMed]
40. McClung JP, et al. Development of insulin resistance and obesity in mice overexpressing cellular glutathione peroxidase. Proc Natl Acad Sci USA. 2004;101:8852–8857. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...