Pten (phosphatase and tensin homologue gene) haploinsufficiency promotes insulin hypersensitivity
- 1.3k Downloads
Insulin controls glucose metabolism via multiple signalling pathways, including the phosphatidylinositol 3-kinase (PI3K) pathway in muscle and adipose tissue. The protein/lipid phosphatase Pten (phosphatase and tensin homologue deleted on chromosome 10) attenuates PI3K signalling by dephosphorylating the phosphatidylinositol 3,4,5-trisphosphate generated by PI3K. The current study was aimed at investigating the effect of haploinsufficiency for Pten on insulin-stimulated glucose uptake.
Materials and methods
Insulin sensitivity in Pten heterozygous (Pten +/−) mice was investigated in i.p. insulin challenge and glucose tolerance tests. Glucose uptake was monitored in vitro in primary cultures of myocytes from Pten +/− mice, and in vivo by positron emission tomography. The phosphorylation status of protein kinase B (PKB/Akt), a downstream signalling protein in the PI3K pathway, and glycogen synthase kinase 3β (GSK3β), a substrate of PKB/Akt, was determined by western immunoblotting.
Following i.p. insulin challenge, blood glucose levels in Pten +/− mice remained depressed for up to 120 min, whereas glucose levels in wild-type mice began to recover after approximately 30 min. After glucose challenge, blood glucose returned to normal about twice as rapidly in Pten +/− mice. Enhanced glucose uptake was observed both in Pten +/− myocytes and in skeletal muscle of Pten +/− mice by PET. PKB and GSK3β phosphorylation was enhanced and prolonged in Pten +/− myocytes.
Pten is a key negative regulator of insulin-stimulated glucose uptake in vitro and in vivo. The partial reduction of Pten due to Pten haploinsufficiency is enough to elicit enhanced insulin sensitivity and glucose tolerance in Pten +/− mice.
KeywordsGlucose uptake Insulin hypersensitivity Insulin sensitivity Pten haploinsufficiency
glycogen synthase kinase 3
positron emission tomography
protein kinase B
phosphatase and tensin homologue
SH2-containing inositol 5′-phosphatase 2 (also known as INPP5D)
Type 2 diabetes mellitus is a multifactorial disease with a complex pathophysiology that includes defects in insulin production and a failure in peripheral tissues to take up glucose in response to insulin. Signalling via phosphatidylinositol 3-kinase (PI3K) is a key pathway in the regulation of glucose uptake by insulin . Activation of PI3K via insulin receptor substrate-1 causes PI3K to phosphorylate inositol-containing phospholipids at the 3′-position of the inositol ring, increasing the levels of the lipid messenger phosphatidylinositol 3,4,5-trisphosphate (PIP3), among others. PIP3 activate a phosphoinositide kinases, which in turn phosphorylate and activate protein kinase B (PKB), also known as Akt, a key effector kinase for many of the downstream metabolic effects initiated by insulin via PI3K. The phosphatase and tensin homologue (PTEN) is a phosphatase which recognises both protein and lipid substrates. As a lipid phosphatase, PTEN dephosphorylates PIP3 at the 3′-position of the inositol ring, thereby acting as an antagonist to PI3K signalling [2, 3]. PTEN was originally identified as a tumour suppressor gene, and is one of the most commonly mutated genes in human cancer [4, 5, 6]. Mutation or deletion of PTEN has been identified in many cancers including prostate, brain, breast, endometrium and skin [4, 5, 6, 7, 8].
As many of the metabolic outcomes of insulin are achieved through recruitment of PI3K and the subsequent rise in PIP3 levels, PTEN may have a critical role in modulating sensitivity to insulin-stimulated glucose uptake. Recent studies in which Pten was ablated specifically in liver, adipose tissue and muscle in mice by Cre recombinase-based strategies have shown a role for Pten in the regulation of insulin sensitivity in those organs, and highlighted the contributions of those organs to glucose homeostasis of the whole animal [9, 10, 11]. However, the effect of global reduction of Pten levels on glucose uptake and metabolism has not yet been shown in an in vivo mouse model. Unfortunately, experimental genetic inactivation of both Pten alleles (Pten −/− ) results in early embryonic lethality [6, 12]. Thus mice harbouring inactivation of one Pten allele (haploinsufficiency, Pten +/−) are an important experimental model for studying the role of Pten in vivo. Haploinsufficiency for Pten has important consequences for cell proliferation and tumorigenesis in mice [6, 12, 13, 14]. Pten +/− mice exhibit defective apoptosis in T cells, B cells and macrophages and spontaneous neoplasms in various tissues including prostate, liver, endometrium and others [12, 13, 14]. Here we show that Pten haploinsufficiency results in insulin hypersensitivity and enhanced insulin-mediated glucose uptake.
Materials and methods
All laboratory animals were cared for and used according to guidelines of the Canadian Council on Animal Care. Pten +/− mice were generated by R. Parsons . These mice were backcrossed with C57BL6 mice for more than ten generations. The genotypes of the mice were determined as described .
Blood glucose determination
For determination of blood glucose levels, glucose levels in blood samples taken from tail vein were determined using a One Touch Ultra blood glucose monitor (LifeScan Inc., Milpitas, CA, USA). For the glucose tolerance test, mice were fasted overnight (15 h) prior to i.p. injection of glucose (2 mg/g body weight). Blood glucose levels were then determined at the indicated time-points following glucose injection. For the insulin challenge test, mice were fasted for 15 h prior to i.p. injection of 0.6 mU insulin/g body weight. Blood glucose levels were then determined at the indicated time-points following insulin injection.
Determination of insulin in plasma
Mice were fasted for 15 h, then approximately 100 μl of blood was collected from the tail vein using pipette tips precoated with heparin. The mice were then injected i.p. with 2 mg glucose/g body weight, and a second blood sample was obtained 15 min following injection. The blood samples were centrifuged at 805 g in a table-top centrifuge, and insulin in the plasma fraction was determined using an insulin ELISA system (ALPCO Diagnostics, Salem, NH, USA) according to the manufacturer’s instructions.
Isolation of pancreatic islets and glucose-stimulated insulin release
Islets were isolated from pancreata of mice as described . Pancreata from three mice were pooled. After 24 h in culture, islets were preincubated in Krebs–Ringer bicarbonate buffer plus 0.1% BSA (KRBB-BSA) and 1.67 mmol/l glucose. The islet cells were incubated for 1 h in 1 ml KRBB-BSA containing the desired concentration of glucose (1.67 or 16.7 mmol/l). The incubation medium was then collected, and centrifuged and the supernatant fractions were used for assay of insulin content by ELISA. Total extractable insulin in islets was determined by adding 0.5 ml lysis buffer (1 mol/l acetic acid with 0.1% BSA) to the islet pellet. Cell debris was pelleted by centrifugation, and the supernatant fraction was used for insulin assay.
Fluorescent immunohistochemistry of pancreatic islets
Immunocytochemistry has been described . Pancreata from wild-type and Pten +/− mice were embedded in paraffin. Five-micron sections were incubated for 1 h at room temperature with guinea-pig anti-insulin antibody (1:100 dilution; Dako, Mississagua, ON, Canada) and rabbit anti-glucagon antibody (Dako; 1:75 dilution). Sections were visualised by incubation with Alexa–Fluor 488 goat anti-guinea-pig (Invitrogen, Molecular Probes, Burlington, ON, Canada; 1:100 dilution) and Texas Red donkey anti-rabbit antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA; 1:100 dilution). Sections were examined by fluorescence microscopy.
Beta cell mass calculation
Following insulin immunostaining, pancreatic sections were captured under an FITC filter using a Zeiss Axioplan 2 microscope equipped for epifluorescence and Pathvysion imaging software. The insulin immunopositive area (as a proportion of pancreatic area) was determined in at least six fields per section using Zeiss AxioVision software, performed on eight pancreatic sections per mouse. Beta cell mass was calculated as pancreatic mass times the percent beta cell area for each mouse.
2-[18F]fluoro-2-deoxyglucose uptake and microPET scanning
Mice were fasted for 15 h prior to i.p. injection with 0.6 mU insulin/g body weight. Twenty minutes after insulin injection, mice were injected with 2-[18F]fluoro-2-deoxyglucose (18FDG) (4.1 MBq) via the tail vein, and whole-body distribution of 18FDG was monitored by positron emission tomography (PET) using a microPET R4 tomograph (Concorde Microsystems, Knoxville, TN, USA). The image data were attenuation-corrected using a 68Ge rod source. Wild-type and Pten +/− mice were positioned side-by-side and scanned simultaneously for 60 min after 18FDG injection. Data were analysed using the manufacturer’s software, ASIPro VM 184.108.40.206. For data analysis, regions of interest (ROIs) were specified in the coronal section of a hindlimb for each mouse. The 18FDG activities in each ROI were summed across all coronal planes, and expressed as a proportion of 18FDG in the whole mouse (total 18FDG activity was also taken as the sum of activity across coronal planes).
Primary myocyte culture
Primary myocytes were cultured from fore- and hindlimbs from wild-type and Pten +/− mice according to a protocol adapted from Springer et al. . Skeletal muscle was excised and minced and then incubated at 37°C for 2 h with 2.4 units/ml dispase II. Dispase digestion was halted by addition of 10 ml F-10-based myocyte medium (400 ml F-10 nutrient mixture, 100 ml fetal bovine serum, 25 μg basic fibroblast growth factor, penicillin/streptomycin). The tissue slurry was then strained through a fine metal mesh and centrifuged at 500 g for 5 min. The supernatant fraction was removed to a fresh centrifuge tube and the pellet was resuspended and plated to a 5-cm collagen-coated culture dish in a total of 4 ml F-10-based medium. The supernatant fraction was re-centrifuged and the pellet was pooled with the first. The cells were passaged when they reached 70–80% confluence. To enrich for myocytes, the old medium was removed by aspiration, the cells washed with PBS, then incubated in a film of this buffer for 15 min. Myocytes were loosened from the dish by striking the dish several times sharply on the bench top. Cells were washed from the dish with PBS. The PBS incubation was repeated and the cells were pooled. Cells kept in culture for more than 5–6 days began to exhibit the elongated morphology typical of myotubes . Cells in these older cultures also exhibited spontaneous twitching. All experiments were performed with cells on days 2–3 of culture.
Experiments were performed on wild-type or Pten +/− cells from passage two to eight, grown on 24-well (1-cm diameter wells) plates. Cells were preincubated in serum-free medium for 12 h prior to experiments. The cells were washed twice with HEPES-buffered saline (140 mmol/l NaCl, 20 mmol/l HEPES, 2.5 mmol/l MgSO4, 1 mmol/l CaCl2, 5 mmol/l KCl, pH 7.4), then incubated with 10 μmol/l 2-deoxy[3H]glucose (0.037 MBq/ml) for 5, 10, 20 or 30 min, in the absence or presence of 0.1 μmol/l insulin. At the end of the incubation period, the radioactive substrate was removed, and the cells were washed three times with ice-cold 20 mmol/l glucose. Cells were lysed with 0.2 ml 2% SDS, and the lysates were used for scintillation counting. Non-specific diffusion of 2-deoxy[3H]glucose was determined by incubating cells in the presence of 2-deoxy[3H]glucose and 10 μmol/l cytochalasin B, a specific inhibitor of glucose uptake .
Determination of cell-surface GLUT1 and GLUT4 levels
Primary myocytes were grown in 10-cm culture dishes. The cells were rinsed with PBS, and incubated with 0.5 mg/ml sulpho-NHS-LC-biotin (Pierce Biotechnology, Rockford, IL, USA) in PBS for 20 min. Cells are impermeable to this reagent and it labels the amine groups of cell-surface proteins. Following the incubation, the cells were washed three times with PBS, then lysed with 0.5 ml lysis buffer (1% NP-40, 150 mmol/l NaCl, 50 mmol/l Tris pH 7.5, including protease and phosphatase inhibitors). Protein was quantified with a BCA protein assay kit (Pierce). Streptavidin-agarose beads (Pierce) were pre-blocked by incubation with 2% BSA in lysis buffer. Each sample (1.2 mg protein) was incubated with the streptavidin beads (100 μl, 50% bead slurry) for 1 h at room temperature. The beads were washed five times, for 30–60 min each wash, using lysis buffer containing 0.1% SDS. After the final wash, 50 μl SDS sample-loading buffer were added, and heated at 95°C for 10 min, then vortexed to elute the biotinylated proteins. The proteins in the supernatant fractions were analysed by SDS-PAGE, then transferred to nitrocellulose. Proteins were probed using polyclonal antibodies against GLUT1 and GLUT4 (also known as SCLA2A1 and SCLA2A4, respectively).
Western immunoblotting for PKB, phospho-PKB and phospho-glycogen synthase kinase 3 (GSK3)-α/β
Total protein extracts of primary myocytes were prepared by lysing cells in NP-40-based lysis buffer. Thirty to fifty micrograms of protein per sample were analysed by SDS-PAGE, then transferred to nitrocellulose. Rabbit polyclonal anti-PKB, anti-phospho-PKB (Ser473) and anti-phospho-GSK3-α/β (Ser21/9) antibodies were obtained from Cell Signaling Technology (Beverly, MA, USA). Mouse monoclonal anti-vinculin antibody was obtained from Sigma (St Louis, MO, USA). Horseradish peroxidase-conjugated secondary antibodies were obtained from Dako. Protein bands were visualised using an enhanced chemiluminescence system and autoradiography by exposure to X-ray film.
Statistical analysis was performed using Student’s t test. Significance was assessed at the 95% confidence level.
Blood glucose levels in wild-type and Pten +/− mice during ad libitum feeding and after a 15-h fast
Mean blood glucose±SD (mmol/l)
p value (independent Student’s t test)
Ad libitum feeding
7.1 ± 0.4 (n = 5)
6.3 ± 0.2 (n = 4)
6.5 ± 0.3 (n = 5)
5.8 ± 0.3 (n = 5)
Plasma insulin levels in wild-type and Pten +/− mice after fasting, and 15 min after i.p. glucose administration (2 mg glucose/g body weight)
Mean plasma insulin±SD (ng/ml)
p value (independent Student’s t test)
Wild-type (n = 7)
Pten+/− (n = 6)
0.50 ± 0.04
0.36 ± 0.10
1.8 ± 0.5
2.4 ± 0.4
>0.05 (not significant)
Pancreatic beta cell mass of Pten +/+ and Pten +/− littermates
Mean beta cell mass±SE (mg)
1.24 ± 0.30 (n = 3)
1.29 ± 0.15 (n = 3)
Glucose transport into skeletal muscle occurs by a facilitated diffusion process mediated primarily by the glucose transporter proteins, GLUT1 and GLUT4. The level of GLUT4 at the plasma membrane is tightly regulated by insulin, whereas GLUT1 is found constitutively on the cell surface . Since 2-deoxy[3H]glucose uptake was elevated in Pten +/− cells even in the absence of insulin (Fig. 4a), we analysed cell-surface GLUT4 and GLUT1 by western immunoblot as described in Materials and methods (Fig. 4b). This analysis showed increased GLUT4 at the cell surface in Pten +/− cells, while similar GLUT1 levels were observed in the two cell types.
Recent studies utilising tissue-specific deletion of Pten in liver, adipose and muscle have shown that these organs make important contributions to the glucose homeostasis of the animal and clearly establish Pten as a negative regulator of insulin signalling in vivo [9, 10, 11]. In a separate report, in vivo antisense treatment resulted in 90% reduction of Pten levels in the liver of diabetic mice, normalising blood glucose levels and improving performance in an insulin tolerance test . While the above studies focused on the role of Pten in specific tissues, the current work defines the effects of partial systemic reduction of Pten on the regulation of insulin sensitivity in the context of the whole animal.
PTEN expression/activity is tightly regulated by a complex mechanism that is currently poorly understood. PTEN is transcriptionally regulated by transcription factors such as p53, Egr-1, NFκB and SMADs, while protein levels and activity are modulated by phosphorylation, oxidation, subcellular localisation, phospholipid binding and protein stability . Previous studies showed that Pten haploinsufficiency results in only a 15–50% reduction in Pten protein, leaving a substantial amount of Pten protein remaining in the tissues examined [14, 22, 23, 24, 30]. In our hands, western blot analysis indicated similar decreases in tissue Pten levels (data not shown). The present study provides the first evidence that Pten haploinsufficiency leads to insulin hypersensitivity in vivo, suggesting that even small changes in Pten levels or activity can lead to dramatic effects on insulin responses. Thus partial systemic inhibition of PTEN activity may represent a novel strategy to ameliorate the pathology of type 2 diabetes.
The PI3K/PKB signalling axis is central to maintenance of normal glucose homeostasis . Pten-deficient cells typically have elevated levels of intracellular PIP3 [31, 32]. The sustained phosphorylation of PKB/Akt and a downstream effector, GSK3β, observed in the current study is consistent with previous reports of elevated PKB/Akt activation as a result of impaired desensitisation of PI3K-dependent signals under conditions of Pten inactivation or inhibition [9, 10, 11, 28]. A subtle reduction of Pten in the Pten +/− mice appears sufficient to achieve a prolonged and robust insulin signal. The roles of PIP3 and PKB/Akt in insulin-mediated glucose uptake have been well-established; therefore, our data collectively suggest that sustained PI3K/PKB/Akt/GSK3 signalling may, in part, contribute to insulin hypersensitivity in Pten+/− mice. However, other PKB/Akt-independent signalling pathways, as well as additional signals delivered by the protein phosphatase and/or adaptor functions of Pten may also contribute to the insulin hypersensitivity phenotype manifested by Pten haploinsufficiency.
GLUT1 glucose transporters are constitutively present at the cell surface, while GLUT4 cycles between intracellular vesicles and the cell surface in a manner controlled by insulin-mediated signalling mechanisms . In unstimulated fat and muscle cells, the majority of GLUT4 exists in intracellular compartments, while the insulin signal stimulates the exocytosis of GLUT4 to the cell surface. Studies involving PKB/Akt-null animals and silencing of PKB/Akt by RNA interference have demonstrated that PKB/Akt is of primary importance in GLUT4 translocation [33, 34, 35, 36]. In the current study, we observed an enhancement of 2-deoxyglucose uptake in Pten +/− cells even in the absence of insulin stimulation, which was consistent with our observation of increased biotinylation of GLUT4 at the cell surface, while GLUT1 levels were unaffected. However, we detected no increase in phosphorylation of PKB/Akt in the absence of insulin in Pten +/− cells. Consistent with these observations, a recent study in which Pten was specifically knocked down by RNA interference reported slight but significant enhancement of deoxyglucose uptake in adipocytes, but no apparent increase in phosphorylation of PKB/Akt . These results could reflect a mechanism(s) affecting glucose transport involving Pten/PIP3 but independent of PKB/Akt. Such a mechanism could involve the atypical protein kinases C-λ and -ζ (PKC-λ and PKC-ζ), which, like PKB/Akt, are subject to control by PI3K via PIP3 and 3-phosphoinositide-dependent protein kinase-1 (PDK-1) [38, 39, 40]. A number of studies have implicated PKC-λ/ζ in insulin-stimulated glucose transport and GLUT4 translocation [38, 41, 42, 43]. These studies indicate a role for PKC-λ/ζ which appears to be complementary or parallel to that of PKB/Akt.
The importance of the PI3K pathway in insulin signalling has raised interest in developing pharmaceutical antagonists targeting PIP3 phosphatases such as PTEN or Sh2-containing inositol 5′-phosphatase 2 (SHIP2; also known as Inppl1) in an effort to enhance insulin responsiveness in type 2 diabetes . While a study on Ship2-null mice initially implicated Ship2 as a key regulator of glucose uptake , it was subsequently recognised that these mice also had deleted a neighbouring gene, Phox2a, which may have contributed to the observed phenotype (corrigendum appears in Nature vol. 431, p. 878). A new gene-targeted mouse has since been generated in which only the gene encoding Ship2 is deleted . Interestingly, this mouse does not exhibit insulin hypersensitivity but is protected against diet-induced obesity, [46, 47]. In the light of these new findings, our observations of dramatic enhancement of insulin sensitivity and glucose uptake resulting from Pten haploinsufficiency suggests that PTEN plays a role in the desensitisation of insulin signalling, and strengthens the hypothesis that PTEN is a key regulator of insulin-stimulated glucose uptake.
Type 2 diabetes is associated with insulin deficiency and resistance to insulin signalling [48, 49]. The current study provides further support for PTEN as a candidate for drug intervention in treating type 2 diabetes. However, enthusiasm for systemic PTEN inhibitors as a therapeutic approach for treating type 2 diabetes is tempered by the potential side-effects of such agents. Most notably, a modest reduction of Pten in Pten +/− mice promotes tumour formation in a wide variety of tissue, suggesting that long-term chronic administration of systemic PTEN inhibitors may lead to enhanced cancer development and progression. Furthermore, the PI3K pathway is an important regulator of numerous normal cellular processes and systemic PTEN inhibition may cause dysfunction of many cells and tissues, leading to alterations in liver, kidney, brain function, etc. For example, hepatocyte-specific Pten-deficient mice exhibit hepatomegaly and liver steatosis . Systemic inhibition of PTEN may therefore be unacceptable clinically because of these inherent potential toxicities . However, targeted delivery of PTEN antagonists to muscle cells or adipocytes may overcome these limitations and they hold promise as clinically viable agents in treatment of type 2 diabetes . This study shows that partial inhibition of Pten is sufficient to enhance insulin sensitivity. This observation is important from a therapeutic standpoint, as most drugs are unlikely to completely inhibit enzyme activity in vivo, nor would complete inhibition be desirable given the potential for cancer development resulting from chronic or overactivation of the PI3K/PKB pathway [12, 14, 23].
In conclusion, Pten haploinsufficiency results in insulin hypersensitivity, which underscores the importance of PTEN in attenuating insulin signals, and suggests that small-molecule drugs that antagonise PTEN activity may represent a novel class of insulin-sensitising agents and targeted delivery of these agents to muscle cells or adipocytes may have potential clinical utility in treating insulin resistance observed in type 2 diabetes.
Funding was provided by a grant from Health Canada. J. T. Wong is a recipient of a fellowship from the Canadian Institutes of Health Research. We gratefully acknowledge the assistance of S. McCormick and D. Yapp with microPET, G. Smirnova with immunohistochemistry and Y. Chen for technical assistance. Polyclonal GLUT1 and GLUT4 antibodies were gifts of A. Klip.
Duality of interest
The authors declare that no duality of interest exists relating to the subject matter of this manuscript.
- 17.Springer ML, Rando T, Blau HM (1997) Gene delivery to muscle. In: Boyle AL (ed) Current protocols in human genetics, vol. 13.4. Wiley, New York, NY, pp 1–19Google Scholar
- 39.Bandyopadhyay G, Standaert ML, Sajan MP et al (1999) Dependence of insulin-stimulated glucose transporter 4 translocation on 3-phosphoinositide-dependent protein kinase-1 and its target threonine-410 in the activation loop of protein kinase C-zeta. Mol Endocrinol 13:1766–1772PubMedCrossRefGoogle Scholar
- 40.Standaert ML, Bandyopadhyay G, Perez L et al (1999) Insulin activates protein kinases C-zeta and C-lambda by an autophosphorylation-dependent mechanism and stimulates their translocation to GLUT4 vesicles and other membrane fractions in rat adipocytes. J Biol Chem 274:25308–25316PubMedCrossRefGoogle Scholar
- 42.Bandyopadhyay G, Standaert ML, Kikkawa U, Ono Y, Moscat J, Farese RV (1999) Effects of transiently expressed atypical (zeta, lambda), conventional (alpha, beta) and novel (delta, epsilon) protein kinase C isoforms on insulin-stimulated translocation of epitope-tagged GLUT4 glucose transporters in rat adipocytes: specific interchangeable effects of protein kinases C-zeta and C-lambda. Biochem J 337:461–470PubMedCrossRefGoogle Scholar