a, Map of the genomic insert carried by BAC RP23-215F15 clone (RPCI library, C57BL/6J) used to generate mice with increased gene dosage of PTEN (“Super-PTEN” mice). The genomic insert is 218.50 Kb long and contains the entire Pten locus (red boxes represent PTEN coding sequence). Gray lines indicate the BAC vector (pBACe3.6) sequence. AscI was used to linearize the BAC clone. Primers used for detection of the transgene are shown (T7 side: F1,R1; SP6 side: F2,R2). b, Quantification of PTEN levels in the different BAC-PTEN-transgenic lines generated. MEFs were obtained from each line and total protein lysates were probed with antibodies towards PTEN and β-actin. The graph shows the fold increase in PTEN levels in each transgenic line relative to wild-type littermates. c, Representative immunoblotting of MEFs derived from lines with no expression (line 1), moderate expression (line 2) or high expression of the tg (line 3). d, Representative images of wt and tg embryos harvested at 13.5 days post-coitum (dpc). Note reduced body size in tg embryo compared to wt. e, Western blot showing PTEN expression during embryogenesis and in tissues from adult mice (wt and tg from line 3). *in the case of heart, coomassie is shown as loading control. f, Western blot showing PTEN levels in cytoplasmic (C) and nuclear (N) extracts from wt and tg MEFs (line 3). g, PTEN expression by immunohistochemistry in tissues from wt and tg mice (line 3).

a, Growth curves of wt and tg MEFs (n=3 per genotype). b, serial 3T3 cultivation of primary wt and tg MEFs (n=3). The figure shows the accumulated population doublings (PDL) at each passage. c, Colony-formation efficiency of wt and tg MEFs (n=4). d, Transformation susceptibility of wt and tg MEFs (n=4). The picture shows the numbers of neoplastic foci formed after transfection with a combination of oncogenic Ras and E1A. e, Susceptibility to 3MC-induced fibrosarcomas. Mice of the indicated genotypes were injected intramuscularly with 3MC in one of the rear legs and the latency for the development of fibrosarcomas was scored (line 3, n=5 per genotype). Super-PTEN mice developed tumors with a significantly longer latency than wild-type mice, as determined by Gehan-Breslow-Wilcoxon test (*p<0.05). Western blot shows PTEN expression levels in tumors derived from wild-type and Super-PTEN mice. f, Soft-agar assay in primary MEFs transformed by E1A+Ras (top) or cMyc+E1A+Ras (bottom). The graph shows the number of colonies per well (six-well plates, triplicate samples). A representative picture from the assay is showed. g, Immunoblotting of protein lysates from wt and tg MEFs and tissues. Total cell extracts were probed with antibodies towards c-Myc, PTEN and β-actin. Error bars in a, b, c, d and f denote s.d. See also Figure S2.

a, OCR (oxygen consumption rate) was measured (Seahorse XF24 analyzer) in primary wt and tg MEFs (n=3 per genotype) under basal conditions and after addition of oligomycin, FCCP and Rotenone. b, Mitochondrial ATP production in wt and tg MEFs after agonist stimulation as described in the methods section. Where indicated, cells were treated with 100 µM ATP. Data are expressed as a percentage of the initial value. wt: 162 ± 8%; tg: 251 ± 18%. n=15 from three independent experiments and p<0.05 (mean ± s.e.m). c, Mitochondrial Ca²⁺ homeostasis measurements after agonist stimulation as described in the methods section. Where indicated, cells were treated with 100 µM ATP. wt: [Ca²⁺]_m peak 155 ± 8 µM. tg: [Ca²⁺]_m peak 131 ± 5 µM. n=12 from three independent experiments (mean ± s.e.m). d, Analysis of mitochondrial membrane potential (ΔΨ_m) in wt and tg MEFs. Cells where loaded with TMRM as described in the methods section. Where indicated cells were treated with FCCP to collapse completely the ΔΨ_m. The traces are representative of single cell responses (wt n=28; tg n= 33). e, Analysis of total and single mitochondrial volume and mitochondrial numbers as described in the methods section (wt n=43; tg n= 40 from three independent experiments and p<0.05). Mitochondrial morphology in wt and tg MEFs as revealed by mitochondrial targeted GFP visualization. Mitochondrial fragmentation index was calculated as described in the methods section (wt n=43; tg n= 40 from three independent experiments). Ordinates for the graphs of Network volume, Average mitochondrial volume and Mitochondrial number are Voxel/Cell, Voxel/Object and N° Object/Cell respectively. Error bars in a denote s.d.; error bars in d, e denote s.e.m. See also Figure S4.

Model by which PTEN elevation negatively impacts two of the most noticeable metabolic features of tumor cells: glutaminolysis and Warburg effect.

a, Growth curves indicate that PTEN overexpression results in reduced body mass. Note that the effect is more severe as PTEN dose increases. b, Western blot showing PTEN levels in different tissues (spleen, liver) from different PTEN transgenic lines (line 1: 1.2-fold; line 2: 2-fold; line 3: 3.5-fold above the endogenous PTEN level). c, Effect of PTEN overexpression on body size (line 3). d, Cell number and cell size in organs isolated from Super-PTEN and wild-type mice (line 3, n=3 per genotype). FSC=Forward Scatter. CD4/8 cells: CD4⁺/Cd8⁺, KSL cells: Lin⁻/c-Kit⁺/Sca-1⁺, AT2 cells: Sca1⁻/CD45⁻/PECAM⁻/Autofl^hi, B cells: B220^mid+/IgM^mid+. Error bars in a and d denote s.d. See also Figure S1.

a, Percentage of fat and lean mass was determined in young (left panel: wt n=7; tg n=7) and old mice (right panel: wt n=12; tg n=10) by EchoMRI (line 3). b, Indirect calorimetry of Super-PTEN (red line) and wild-type mice (black line). Oxygen consumption (VO2), carbon dioxide release (VCO2), respiratory exchange ratio (RER; VCO₂/VO₂) and energy expenditure per kg of body weight were determined in Super-PTEN and wild-type mice in metabolic chambers (line 3, n=4 per genotype). c, Serum lactate levels in wt (n=10) and tg (n=9) mice (line 3). Error bars in a denote s.d.; error bars in b denote s.e.m. See also Figure S3.

a, Glucose, lactate and glutamine levels in culture media were measured in wt and tg cells (n=2) and normalized to cell number. b, Effects of pharmacological inhibition of mTORC1 and PI3K/Akt pathway on glycolytic and glutaminolytic enzymes. Cell lysates from wt and tg MEFs treated with DMSO (Cont.), 20 nM Rapamycin (Rapa.) for 24 h, or 100 nM Wortmannin (Wort.) for 8 h were subjected to immunoblotting with antibodies against PKM2, PFKFB3, GLS, p-S6, S6, p-Akt, Akt, PTEN and β-actin. c, Effects of Tsc2 depletion-mediated mTORC1 activation on PKM2 protein level. Cell lysates from GL3-shRNA (control shRNA against the luciferase gene) and Tsc2-shRNA infected wt and tg MEFs were subjected to immunoblotting with antibodies against PKM2, Tsc2, PTEN and ²-actin. d, Fold change ratio (tg versus wt) in PTEN, FKFB3 and GLS mRNA levels determined by qRT-PCR of total RNA from wt and tg MEFs (left). Proteasome-mediated degradation of GLS and PFKFB3. Cell lysates from wt and tg MEFs treated with 10 µM of the proteasome inhibitor MG132 for 4 h were subjected to immunoblotting with antibodies against GLS, PFKFB3, PTEN and β-actin (right). e, Impact of PFKFB3 overexpression in cell growth (48 hours after plating) in primary MEFs transformed by E1A+Ras. Percentage of growth relative to wt MEFs+E1A+Ras is shown. f, PTEN enhances APC/C-Cdh1-mediated ubiquitination of GLS. PTEN-deficient PC-3 cells were co-transfected with GLS, His-ubiquitin (Ub), CDH1 and PTEN, and treated with MG132 (10 µM) for 4 hr before harvesting. His-Ub-conjugated GLS was purified from cell lysates using Ni²⁺-NTA spin column under denaturing conditions. g, Cdh1-silencing recovers the level of GLS and PFKFB3 in tg MEFs. Cell lysates from wt and tg MEFs transfected with siRNAs for Renilla luciferase (siCont.) or Cdh1 were subjected to immunoblotting with antibodies against GLS, PFKFB3, PTEN, Cdh1 and ²-actin. h, Number of colonies (fold change ratio: tg versus wt MEFs+Ras) formed in soft-agar assay after Tsc2 or Cdh1 knockdown in immortalized MEFs transformed by oncogenic Ras. The graph shows the number of colonies per well (six-well plates, triplicate samples). Error bars in a, d, e and h denote s.d. See also Figure S5.