Class I PI3K Provide Lipid Substrate in T Cell Autophagy Through Linked Activity of Inositol Phosphatases

Abstract

Autophagy, a highly conserved intracellular process, has been identified as a novel mechanism regulating T lymphocyte homeostasis. Herein, we demonstrate that both starvation- and T cell receptor-mediated autophagy induction requires class I phosphatidylinositol-3 kinases to produce PI(3)P. In contrast, common gamma chain cytokines are suppressors of autophagy despite their ability to activate the PI3K pathway. T cells lacking the PI3KI regulatory subunits, p85 and p55, were almost completely unable to activate TCR-mediated autophagy and had concurrent defects in PI(3)P production. Additionally, T lymphocytes upregulate polyinositol phosphatases in response to autophagic stimuli, and the activity of the inositol phosphatases Inpp4 and SHIP are required for TCR-mediated autophagy induction. Addition of exogenous PI(3,4)P₂ can supplement cellular PI(3)P and accelerate the outcome of activation-induced autophagy. TCR-mediated autophagy also requires internalization of the TCR complex, suggesting that this kinase/phosphatase activity is localized in internalized vesicles. Finally, HIV-induced bystander CD4⁺ T cell autophagy is dependent upon PI3KI. Overall, our data elucidate an important pathway linking TCR activation to autophagy, via induction of PI3KI activity and inositol phosphatase upregulation to produce PI(3)P.

Keywords: HIV; PI3K I; T cells; autophagy; cytokines; lipid kinase; lipid phosphatase.

FIGURE 1

Class 1 PI3K activity is necessary for starvation-induced autophagy. (A) AVO profiles gated on live CD4 T lymphocytes starved for 48 h in HBSS and treated with the indicated activators inhibitors of PI3K. A 5mM 3MA and higher doses of PIK75 drastically inhibited AVO formation, while 20 nM PIK75 and 10 ng/mL IL-7 had slightly less potent effects. Atg3^f/fLck-cre T cells were used as a negative control to assess background lysosomal contribution. Gates and numbers indicate the percentage of cells that have upregulated autophagy above basal levels. (B) Quantitation of (A), representative of at least 4 independent experiments per condition. (C) Fluorescence microscopy of endogenous LC3 punctate structures 48 h post starvation with the indicated inhibitors. (D) Quantitation of starvation-induced LC3 aggregates 48 h post starvation. Data are representative of at least 2 independent experiments per condition with at least 30 cells counted in each replicate.

FIGURE 2

TCR-induced autophagy is dependent upon class I PI3K and 5′ phosphatase activity. (A) AVO formation histograms gated on live CD4 T lymphocytes stimulated with 1ug/mL soluble anti-CD3 and anti-CD28 for 24 and 48 h with the indicated reagents. A 5 mM 3MA, 100 nM PIK75, and 2 μM PIK75 potently reduced AVO formation, especially at 48 h. Histograms are representative of at least 3 independent experiments under each condition. (B) Quantitation of AVO formation from (A) summarizing at least 3 independent experiments per condition. (C) AVO formation in p85^f/f and p85^f/fER-cre CD4 T cells demonstrates the requirement for class I PI3K in TCR-mediated autophagy induction. Splenocyte and lymphocyte mixtures were cultured for 4 days in 1 ng/mL IL-7 and either 500 nM 4OH Tamoxifen or EtOH, and stimulated for 2 days with 1 μg/mL soluble anti-CD3 and anti-CD28 for 48 h. (D) Quantification of (C). Ratios of AVO formation of CD4 T cells pre-treated with 4OH Tamoxifen or EtOH to those kept in complete media for 96 h. Data are compiled from 3 independent experiments. (E) p62 degradation and LC3 lipidation are impaired in p85-deficient CD4 T cells. Cells were treated as in (C). Numbers indicate band densities compared to those of β-actin. (F) PI(3)P levels measured 24 h after activation through soluble anti-CD3 and anti-CD28 as in (C).

FIGURE 3

Common gamma chain cytokines inhibit starvation-induced autophagy. (A) AVO profiles of naïve CD4 T lymphocytes starved in HBSS for 48 h with the indicated cytokines. Histograms are representative of 3 independent experiments. IL-4 and IL-7 had potent effects on starvation-induced AVO formation, while IL-15 had a less drastic effect. (B) Fluorescence microscopy of LC3 aggregates in naïve, starved CD4 T lymphocytes treated with the indicated γ_c cytokines for 48 h, permeabilized, and stained with anti-LC3. Data are representative of fields observed. (C) Quantitation of (B), with at least 30 cells counted for each condition in two independent experiments. The indicated cytokines significantly reduced LC3 punctate formation. P-values are from unpaired, two-tailed Student’s t-tests.

FIGURE 4

TCR internalization is required for TCR-mediated autophagy induction. (A) AVO formation in CD4 T cells pre-treated with 100 μM Dynasore or 5 mM 3MA for 1 h and stimulated for 24 h with 1 μg/mL soluble anti-CD3 and anti-CD28. Lower panel indicates the relative levels of TCR-β on the cell surface to demonstrate the efficacy of Dynasore versus other treatments. (B) Quantification of AVO formation in (A), data represents 3 independent experiments.

FIGURE 5

Inositol phosphatases respond to autophagic stimuli. (A) AVO formation in CD4 T lymphocytes pre-treated with 500 nM or 15 uM Anchor (AS1949490) for 1 h and stimulated with either 1 μg/mL soluble anti-CD3 alone or with anti-CD28 for 24 h. 500 nM Anchor inhibits SHIP2 activity and 15 μM inhibits both SHIP1 and SHIP2. (B) Quantification of AVO’s from (A). p-values are from paired, two tailed, Student’s t-tests from 6 replicates in 3 independent experiments. (C) Relative levels of inositol phosphatase mRNA in CD4 T cells 48 h after starvation in HBSS or with 10 ng/mL IL-7. QPCR results were normalized to that of GAPDH. Data are representative of 3 independent experiments with 3 replicates per experiment. (D) Histogram of Inpp4A protein level in naïve or 24 h stimulated CD4 T cells, quantified in the right panel as 4 replicates from 3 independent experiments. Gray histogram represents the isotype control, the dotted histogram is the naïve level, and black histogram is the level in TCR activated cells. (E) Intracellular stains for Inpp4A protein levels in WT or Vps34^f/fLck-cre CD4 T cells (Vps34-/-). Data are from 3 replicates in 2 independent experiments.

FIGURE 6

Exogenous PI(3,4)P₂ alters the autophagic profile and is converted to PI(3)P. (A, left panel) Histograms of relative levels of PI(3)P and PI(3,4)P₂ CD4 T cells were loaded with 10 μM PI(3,4)P₂ and stimulated with 1 μg/mL soluble anti-CD3 and anti-CD28 for 24 and 120 h. PI(3,4)P₂ is largely converted to PI(3)P only after TCR stimulation. (A, right panel) AVO formation of CD4 T cells loaded with PI(3,4)P₂. Exogenous PI(3,4)P₂ effects an accelerated autophagic profile that is resolved by 120 h compared to unloaded controls. (B) Quantification of phosphatidyl inositol levels from CD4 T cells left untreated or loaded with PI(3,4)P₂ and stimulated for 24 h as in (A). Data represent 4 replicates from 3 independent experiments. (C) Fluorescence microscopy of phosphatidylinositol species in T cells activated with plate-bound anti-CD3 and anti-CD28 for the indicated time periods. (D) Quantitation of colocalization of phosphatidylinositol species from (C).

FIGURE 7

Human T cell autophagy is dependent upon class I PI3K activity and inositol phosphatase activity. (A) AVO formation in human lymphocytes. Human CD4⁺ T cells were isolated from healthy donors, purified by Ficol gradient centrifugation, and stimulated with 0.5 μg/mL anti-CD3 and anti-CD28 for 24 h. AVO formation was inhibited by 100 nM PIK75, 3MA, and 15 μM Anchor. Data are representative of at least 3 independent experiments per condition. (B) LC3 punctate formation in stimulated human CD4⁺ T cells for 48 h as in (A), permeabilized, and stained for endogenous LC3. (C) Quantification of (B), representative of two replicates with at least 30 CD4 cells counted per condition. (D) AVO formation in siRNA mediated knockdown of Inpp4A, Inpp4B, and both SHIP1 and SHIP2. Human CD4⁺ T cells were purified as in (A) and electroporated to allow uptake of 5 μg/5 × 10⁶ cells siRNA cocktails, allowed to recover in complete media for 72 h and stimulated with 0.5 μg/mL anti-CD3 and anti-CD28 for 48 h. (E) Quantification of stimulated CD4 T cells in (D), comparing the percentage of cells upregulating AVO formation in the siRNA knockdown samples to a scrambled control siRNA. Data are representative of cells from 5 donors in 4 independent experiments. (F) AVO formation in human CD4⁺ T cells exposed to soluble HIV GP140 fusion protein at the indicated concentrations. Total PBMC’s were treated with the glycoprotein and 100 nM PIK75 for 48 h.