Lack of phosphatidylinositol 3-kinase VPS34 in regulatory T cells leads to a fatal lymphoproliferative disorder without affecting their development

Abstract

Regulatory T (Treg) cells are essential for the maintenance of immunological tolerance, yet the molecular components required for their maintenance and effector functions remain incompletely defined. Inactivation of VPS34 in Treg cells led to an early, lethal phenotype, with massive effector T cell activation and inflammation, like mice lacking Treg cells completely. However, VPS34-deficient Treg cells developed normally, populated the peripheral lymphoid organs and effectively supressed conventional T cells in vitro. Our data suggest that VPS34 is required for the maintaining normal numbers of mature Treg. Functionally, we observed that lack of VPS34 activity impairs cargo processing upon transendocytosis, that defective autophagy may contribute to, but is not sufficient to explain this lethal phenotype, and that loss of VPS34 activity induces a state of heightened metabolic activity that may interfere with metabolic networks required for maintenance or suppressive functions of Treg cells.

Keywords: PI3K; Treg; VPS34; autophagy; endocytosis.

Figure 1

Deletion of VPS34 in Treg cells leads to a Scurfy-like phenotype. (A) Immunoprecipitation (IP), expression (WB) and in vitro kinase assay of tagged versions of the truncated (del21) and wild-type (WT) allele in transiently-transfected HEK293 cells. IPs were performed using anti-Myc antibodies. NT = non-transfected controls. In vitro lipid kinase assay was performed using PI as a substrate on 6xMyc-tagged VPS34 IPs. Representative data of 3 independent experiments is shown with 2 replicates per experiments (labelled as WT-1, WT-2, del21-1, del21-2). (B) Survival curves for Foxp3 ^YFP-Cre Pik3c3 ^flox (red) and wild-type Foxp3 ^YFP-Cre Pik3c3 ^WT mice (black). (C) Representative pictures of 32 day old wild-type Foxp3 ^YFP-Cre Pik3c3 ^WT mice (left) and Foxp3 ^YFP-Cre Pik3c3 ^flox (right). (D) Examples of enlarged spleen and lymph nodes from Foxp3 ^YFP-Cre Pik3c3 ^flox mice and wild-type (WT) littermate controls. (E) Absolute numbers of cells in the spleen, lymph nodes (inguinal, brachial, and axillary) and thymus from Foxp3 ^YFP-Cre Pik3c3 ^flox mice compared to wild-type mice. (F) Percentage and representative FACS plots of CD44^high CD62^low Tcon and CD8⁺ T cells in the spleen. (G) Percentage of IFN-γ⁺ CD4⁺ CD25^- T cells in the spleen. (H) Absolute numbers of Treg cells in spleen, lymph nodes (inguinal, brachial, and axillary), and thymus of Foxp3 ^YFP-Cre Pik3c3 ^flox and wild-type Foxp3 ^YFP-Cre Pik3c3 ^WT mice. (I) Tcon and Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox mice and control wild-type mice (WT) were co-cultured in the presence of dendritic cells and 1μg/ml anti-CD3 for 96 h. Wells were pulsed with Alamar Blue for 4 h and the measured fluorescence used to gauge proliferation. Error bars represent SEM of triplicate wells. Foxp3 ^YFP-Cre Pik3c3 ^flox mice and the respective control mice were between 4 and 5.5 weeks of age. Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and the respective control mice were between 8 and 12 weeks of age. n = 3-15 mice per group. Statistical significance was determined using an unpaired two-tailed Student’s t-test (C–E, I, H), paired two-tailed Student’s t-test (A, B, F, G, J), or a One-way ANOVA with Tukey’s post-test (I). Results are pooled from 2 to 4 independent experiments.

Figure 2

VPS34-deficient Treg cells are not intrinsically pathological, but have a competitive disadvantage. (A) Percentage of CD25^high cells from CD4⁺ in the spleen, the lymph nodes (inguinal, brachial, and axillary), and the thymus of Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and Foxp3 ^YFP-Cre/WT control mice (WT). (B) Percentage of YFP⁺ cells among CD25⁺ CD4⁺ cells in the lymph nodes (inguinal, brachial, and axillary) of Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and Foxp3 ^YFP-Cre/WT Pik3c3 ^WT control mice (WT). (C, D) Mean fluorescence intensity (MFI) of CTLA-4 (C) and CD25 (D) on splenic Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox mice and Foxp3 ^YFP-Cre control mice (WT). (E, F) Mean fluorescence intensity (MFI) of CTLA-4 (E) and CD25 (F) on splenic VPS34-deficient (Cre⁺) and VPS34-sufficient (Cre^-) Treg cells from Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice. (G, H) mean fluorescence intensity (MFI) of CD80 on antigen-presenting cells (APCs) from Foxp3 ^YFP-Cre Pik3c3 ^flox mice and Foxp3 ^YFP-Cre control mice (WT) (G) and Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice (H). (I) Representative flow cytometry plot for CD44 and CD62L expression and bar graph representing percentage of CD44 on splenic Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox mice and Foxp3 ^YFP-Cre control mice (WT). (J) Percentage of CD44 on splenic VPS34-deficient (Cre⁺) and VPS34-sufficient (Cre^-) Treg cells from Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice. Foxp3 ^YFP-Cre Pik3c3 ^flox mice and the respective control mice were between 4 and 5.5 weeks of age. Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and the respective control mice were between 8 and 12 weeks of age. n = 3-7 mice per group. Statistical significance was determined using an unpaired two-tailed Student’s t-test (C–E, I), paired two-tailed Student’s t-test (A, B, F–H, J). Results are pooled from 2 to 4 independent experiments.

Figure 3

Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox mice efficiently internalise IL-2 but might displayed impaired degradation of CD80. (A) Lymphocytes from Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mice were stimulated with 20 ng/ml rhIL-2 and stained for phospho-Stat5 for analysis by flow cytometry. The mean MFI of 3 mice is shown at each time point, and error bars indicate standard deviation. (B) Schematic representation of the IL-2 internalisation assay. Lymphocytes were incubated at 37°C for 4 h in the presence of biotinylated rhIL-2. After washing, surface-bound and internalised IL-2 were identified by flow cytometry using anti-biotin antibodies in different conjugations. (C) Median fluorescence intensity (MFI) of surface-bound and internalised IL-2 in Treg cells from mosaic Foxp3 ^YFP-Cre/WT Pik3c3 ^flox and wild-type Foxp3 ^YFP-Cre/WT Pik3c3 ^WT mice. (D) Schematic representation of the transendocytosis assay. Treg cells were enriched from Foxp3 ^YFP-Cre Pik3c3 ^flox and Foxp3 ^YFP-Cre Pik3c3 ^WT mice and co-cultured for 24 h with an equivalent number of Chinese Hamster Ovary (CHO) cells expressing GFP-tagged CD80 on their surface. Tregs should internalise and accumulate CD80-GFP. (E) Representative histograms showing GFP-tagged CD80 fluorescence associated on Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox and Foxp3 ^YFP-Cre Pik3c3 ^WT mice. (F) Schematic representation of the pulse-chase assay. Similar to in (D), Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^WT mice were enriched and co-cultured for 24 h with a 10:1 ratio of Chinese Hamster Ovary cells expressing GFP-tagged CD80 on their surface. After 24h co-culture, Treg cells were separated from the rest of the cells by fluorescence-activated cell sorting (FACS) and cultured in the presence of VPS34 IN1, a selective Vps34 inhibitor, for up to 48 h. (G) The percentage of the GFP signal was assessed by flow cytometry directly after separation from CHO cells (0 h), and then 24 and 48 h after separation. Foxp3 ^YFP-Cre Pik3c3 ^flox mice and wild-type control mice were between 4 and 5.5 weeks of age. Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and wild-type control mice were between 8 and 12 weeks of age. n = 3-4 mice per group. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Statistical significance between YFP-Cre^- and YFP-Cre⁺ Treg was determined for each time point using an unpaired t-test; no significant difference was found at any point. Results are pooled from two to three independent experiments.

Figure 4

Defective autophagy in VPS34-deficient Treg cells might contribute to, but is not sufficient to explain the phenotype of Foxp3 ^YFP-Cre Pik3c3 ^flox mice. (A, B) Densiometry of LC-3 I and LC-3II in ex vivo Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox (A) or Foxp3 ^YFP-Cre Atg7 ^flox mice (B) and their respective controls (WT) was determined by Western blot analysis. Actin was used as a loading control. Amounts of all proteins were normalised to actin and are relative to amounts in WT cells. (C) Percentage of Treg cells from the spleen and the lymph nodes of Foxp3 ^YFP-Cre Atg7 ^flox mice and wild-type control mice (WT). (D, F) Mean fluorescence intensity (MFI) of the MitoTracker Orange dye in splenic Treg cells from Foxp3 ^YFP-Cre Atg7 ^flox (D), Foxp3 ^YFP-Cre Pik3c3 ^flox (E), mosaic Foxp3 ^YFP-Cre/WT Pik3c3 ^flox (F) mice, and the respective wild-type mice (WT). (G) Quantitative PCR analysis of the mitonchondria DNA (mtDNA) levels of FACS-sorted Treg cells from Foxp3 ^YFP-Cre Pik3c3 ^flox and control mice (WT). (H) Survival curves for Foxp3 ^YFP-Cre Atg7 ^flox (green) and wild-type Foxp3 ^YFP-Cre Atg7^WT mice (grey). Foxp3 ^YFP-Cre Pik3c3 ^flox mice and wild-type control mice were between 4 and 5.5 weeks of age. Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice, Foxp3 ^YFP-Cre Atg7 ^flox, and the respective control mice were between 8 and 12 weeks of age. n = 3-11 mice per group. Statistical significance was determined using a two-tailed Student’s t-test. Results are pooled from 2 to 3 independent experiments.

Figure 5

Proteomic profiling revealed that loss of VPS34 kinase activity increases cellular respiration in Treg cells. (A) Workflow for the proteomic profiling of Treg cells from Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and xp3 ^YFP-Cre/WT Pik3c3 ^WT control mice. (B) Volcano plot representing the ratio of the protein copy numbers in VPS34-deficient Treg cells compared to VPS34-sufficient Treg cells. Proteins that exhibited statistically significant increase in abundance (p < 0.05 by one-sample Student’s t test) by greater than 1.5-fold are shown in red, while proteins that exhibit statistically significant reduction in abundance are shown in blue. (C) Volcano plot highlighting the binding partners of the VPS34 complex II. Bar graphs showing estimated copy numbers calculated using the proteomic data for RUBICON, ATG14L, VPS15, BECLIN-1, UVRAG, and VPS34. Individual data points from the three biological replicates performed for the proteomic analysis are shown, with the bar showing the means ± S.D. p values were calculated by two-tailed, one-sample Student’s t test. (D) Analysis of biological processes (BP), cellular components (CC), molecular function (MF) involving proteins upregulated in VPS34-deficient Treg cells identified in the proteomic profiling analysis. (E) Heat map presenting proteins upregulated in VPS34-deficient Treg cells identified in the proteomic profiling analysis. (F) Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) of VPS34-deficient Treg cells before and after addition of 0.5 µM FCCP and 10μM oligomycin following the Seahorse XF Cell Energy Phenotype Test Assay protocol. (G) Percentage increase of stressed OCR over baseline OCR, and stressed ECAR over baseline ECAR. The metabolic potential is the measure of cells’ ability to meet an energy demand via respiration and glycolysis. Foxp3 ^YFP-Cre/WT Pik3c3 ^flox mosaic mice and wild-type control mice were between 8 and 12 weeks of age. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Results are pooled from 2 to 3 independent experiments.