Long-term p110α PI3K inactivation exerts a beneficial effect on metabolism

The insulin/insulin-like growth factor-1 signalling (IIS) pathway regulates cellular and organismal metabolism and controls the rate of aging. Gain-of-function mutations in p110α, the principal mammalian IIS-responsive isoform of PI 3-kinase (PI3K), promote cancer. In contrast, loss-of-function mutations in p110α impair insulin signalling and cause insulin resistance, inducing a pre-diabetic state. It remains unknown if long-term p110α inactivation induces further metabolic deterioration over time, leading to overt unsustainable pathology. Surprisingly, we find that chronic p110α partial inactivation in mice protects from age-related reduction in insulin sensitivity, glucose tolerance and fat accumulation, and extends the lifespan of male mice. This beneficial effect of p110α inactivation derives in part from a suppressed down-regulation of insulin receptor substrate (IRS) protein levels induced by age-related hyperinsulinemia, and correlates with enhanced insulin-induced Akt signalling in aged p110α-deficient mice. This temporal metabolic plasticity upon p110α inactivation indicates that prolonged PI3K inhibition, as intended in human cancer treatment, might not negatively impact on organismal metabolism.


Table of contents: Supplemental Materials and Methods
 GFP-LC3 mice.  Immunoblotting and immunofluorescence assessment of autophagic activity.  T cell surface marker expression.

Supplemental References
Figure S1 Glucose homeostasis (at middle age) and life span in p110 D933A/WT female mice.

Figure S2
Weight gain in mice fed a high fat diet.

Figure S3
Class IA PI3K isoform expression in tissues from aged p110 D933A mice.

Figure S4
Effect of chronic p110 inhibition on free fatty acid-and high glucose-induced IRS downregulation in C2C12 myotubes.

Figure S5
Effect of chronic p110 inhibition on IRS-dependent signaling in 3T3-L1 adipocytes.

Figure S6
IRS-1 and -2 protein levels in white adipose tissue (WAT) and skeletal muscle of aged mice.

Figure S7
Insulin stimulated S6K Thr389 phosphorylation in white adipose tissue (WAT) and skeletal muscle of aged mice.

Figure S8
Distribution of T cells in naive and memory cell subsets between WT and p110 D933A mice.

Figure S9
Fasting plasma IGF-1 in young p110 D933A/WT and WT littermate mice Figure S10 Assessment of autophagy in MEFs derived from p110 D933A/WT embryos.

Table S2
Log-rank test comparisons and hazard ratios of p110 WT/D933A mice.
Immunoblotting and immunofluorescence assessment of autophagic activity. For starvation experiments, cells were washed three times with Earle's Balanced Salt Solution (EBSS, starvation medium) and incubated in EBSS medium for the indicated time in the presence or absence of 50 nM of Bafilomycin A1 (Baf A1). For GFP-LC3 lipidation assays, cells were lysed in ice-cold lysis buffer containing 1% NP-40, 20 mM Tris.HCl (pH 7.5), 1 mM EDTA, 150 mM NaCl, 1 mM DTT, 1 mM NaVO 4 , 100 M PMSF and protease inhibitor cocktail (Sigma-Aldrich). Equal amounts of total protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 15% polyacrylamide gels and transferred to polyvinylidene fluoride membranes. Immunodetection was performed with antibodies against phospho-S6 (Ser240/44) (Cell Signaling Technology), LC3 (Nanotool) or p62 (Abnova). For immunofluorescence studies, MEFs were seeded on pre-coated poly-Llysine coverslips the day before treatment. After treatment, cells were fixed in ice-cold methanol for 15 min followed by blocking in 5% BSA in PBS. Immunostaining for WIPI-II was performed with an antibody kindly provided by Sharon Tooze (London Research Institute-Cancer Research UK) as described elsewhere (3). Confocal images were acquired with an optical slice of 0.8 μm using a 63x oil immersion objective. Secondary antibodies used for this assay were anti-mouse IgG FITC and anti-rabbit IgG Cy5.
T cell surface marker expression. Blood collected from terminally-anaesthetized mice was analysed by flow cytometry. Naive and memory T cell subsets were defined by expression of the following cell surface markers: naive CD4/CD8 T cells = CD44 Low CD45RB High CD62L High ; memory CD4/CD8 T cells = CD44 High CD45RB Low CD62L Low .   The indicated insulin-sensitive tissues were isolated from approx. 80 week old male mice. Tissue homogenates were subjected to quantitative immunoblot analysis using the indicated PI3K isoform-specific antibodies. b-actin and GAPDH were used as loading controls for liver, white adipose tissue (WAT) and muscle (gastrocnemius), respectively. Intensities of protein signals were averaged and plotted in bar graphs (mean ± standard error) shown below the respective immunoblots. Each lane was loaded with tissue homogenate derived from an individual mouse (n=4 per genotype).  Monolayers were then rinsed with phosphate-buffered saline and lysed either directly for detection of IRS and S6K pT389 phosphorylation or after 1 h culture in serum-free medium and re-stimulation with insulin (10 nM for 10 min) for detection of Akt T308 phosphorylation. Lysates were processed for immunoblot analysis with the indicated antibodies. Panels have been rearranged for clarity, but they are from the same immunoblots. Signal intensity was quantified and normalized to that of vinculin, used as a loading control. (B) shows data (mean ± standard deviation) from three independent experiments. Statistical comparisons were performed with Student's t-test. (C) T308 Akt phosphorylation in cells pretreated with a combination of insulin, palmitate and high glucose and acutely re-stimulated with insulin, as above.  Figure S5. Effect of chronic p110a inhibition on IRS-dependent signaling in 3T3-L1 adipocytes. (A) 3T3-L1 adipocytes (cultured in medium with 5.5 mM glucose) were treated with insulin (10 nM), palmitate (500 M, nominal concentration) or glucose (25 mM) in the presence or absence of A66 (2 M) or TGX-221 (1 M) for 16 h. Monolayers were then rinsed with phosphate-buffered saline and lysed either directly for detection of IRS and S6K pT389 phosphorylation or after 1 h culture in serum-free medium and re-stimulation with insulin (100 nM for 10 min) for detection of Akt T308 phosphorylation. Lysates were processed for immunoblot analysis with the indicated antibodies. Panels have been rearranged for clarity, but they are from the same immunoblots. (B) Akt T308 phosphorylation in cells treated with a combination of insulin, palmitate and high glucose and stimulated with insulin, as above.    Figure S7. Insulin stimulated S6K Thr389 phosphorylation in white adipose tissue (WAT) and skeletal muscle of aged mice. Homogenates of epididymal fat pads and gastrocnemius muscles isolated from 80 week old p110a D933A/WT male mice and WT littermates following injection of 0.25 mU/g insulin through the inferior vena cava were subjected to immunoblot analysis using the indicated antibodies. Signal densities were quantified. Values of phosphorylated S6K (Thr389) normalised to total S6K and vinculin, as loading control, are presented in graphs. Bars represent the mean ± standard deviation of S6K phosphorylation in tissues from two mice (n=2) per genotype and treatment shown in the adjacent immunoblots. Plasma was isolated from overnight starved 8-9 week old mice. IGF-1 was measured by ELISA (Octeia, IDS). Bars represent mean ± standard error (n=9-11 mice per genotype and sex).   Figure S10. Assessment of autophagy in MEFs derived from p110a D933A/WT embryos. Autophagic activity was assessed using established markers i.e. conversion of LC3-I to LC3-II, levels of p62 and autophagosome density by WIPI-II staining. (A) WT or p110a D933A/WT MEFs expressing GFP-LC3 transgene were grown in normal condition (Fed). Each lane represents individual embryo-derived MEFs for each genotype (designated as #1-3). LC3-II and LC3-I forms were detected using an LC3 Ab and the levels of p62 expression were monitored. Quantification of the LC3-II/LC3-I ratio and p62 expression relative to the loading control (-tubulin) is shown in the graph. (B) Same as in (A) but in starved condition (EBSS medium). Each culture was treated for the indicated time with EBSS medium with or without 50 nM Bafilomycin A1 (BafA1). Quantification of the LC3-II/LC3-I ratio and p62 expression relative to the loading control (-tubulin) is shown in the graph. Bars represent mean ± standard error. Statistical comparisons were performed with Student's t test (C) LC3 and WIPI-II staining in fed vs starved condition. LC3 and WIPI-II punctate staining shown in the starved condition is characteristic of autophagy induction and autophagic vacuoles.