Calcium Signaling Is Impaired in PTEN-Deficient T Cell Acute Lymphoblastic Leukemia

Abstract

PTEN (Phosphatase and TENsin homolog) is a well-known tumor suppressor involved in numerous types of cancer, including T-cell acute lymphoblastic leukemia (T-ALL). In human, loss-of-function mutations of PTEN are correlated to mature T-ALL expressing a T-cell receptor (TCR) at their cell surface. In accordance with human T-ALL, inactivation of Pten gene in mouse thymocytes induces TCRαβ⁺ T-ALL development. Herein, we explored the functional interaction between TCRαβ signaling and PTEN. First, we performed single-cell RNA sequencing (scRNAseq) of PTEN-deficient and PTEN-proficient thymocytes. Bioinformatic analysis of our scRNAseq data showed that pathological Pten^del thymocytes express, as expected, Myc transcript, whereas inference of pathway activity revealed that these Pten^del thymocytes display a lower calcium pathway activity score compared to their physiological counterparts. We confirmed this result using ex vivo calcium flux assay and showed that upon TCR activation tumor Pten^del blasts were unable to release calcium ions (Ca²⁺) from the endoplasmic reticulum to the cytosol. In order to understand such phenomena, we constructed a mathematical model centered on the mechanisms controlling the calcium flux, integrating TCR signal strength and PTEN interactions. This qualitative model displays a dynamical behavior coherent with the dynamics reported in the literature, it also predicts that PTEN affects positively IP3 (inositol 1,4,5-trisphosphate) receptors (ITPR). Hence, we analyzed Itpr expression and unraveled that ITPR proteins levels are reduced in PTEN-deficient tumor cells compared to physiological and leukemic PTEN-proficient cells. However, calcium flux and ITPR proteins expression are not defective in non-leukemic PTEN-deficient T cells indicating that beyond PTEN loss an additional alteration is required. Altogether, our study shows that ITPR/Calcium flux is a part of the oncogenic landscape shaped by PTEN loss and pinpoints a putative role of PTEN in the regulation of ITPR proteins in thymocytes, which remains to be characterized.

Keywords: PTEN; T-ALL; TCR signaling; calcium signaling; qualitative mathematical model; single-cell RNA-seq; thymocytes.

Figure 1

Single-cell RNAseq analysis reveals an impact of PTEN loss on calcium pathway. (A) Overview of scRNAseq workflow. The scRNAseq assay was performed with thymus and spleens from 4 weeks-old Control OT-II and Pten^del×OT-II mice. Thymocytes and splenic T cells were harvested, each sample was labelled with a distinct Ab-HTO, pooled and then processed for scRNAseq using the 10X Genomics approach. Following sequencing, reads were assembled, quantified, normalized and demultiplexed to identify the original sites and samples. (B, C) UMAP plots of cleaned scRNAseq dataset of thymus and splenic T cells (1686 cells in total) from Control OT-II and Pten^del×OT-II mice. The UMAP plot is colored according to mouse genotype (B) or to the 8 clusters (C). (D) Dot plots showing the expression level of established marker genes of T cell differentiation. Dot color represents the scaled average expression of the specified gene across the various clusters, and dot size indicates the percentage of cells expressing the specified gene. (E, F) Analysis of pathways activity. Genes lists ( Supplementary Table 2 ) corresponding to pathways of interest were retrieved from MSigDB database and were scored in DP clusters using AUCell. (E) Violin plots reporting AUCell score for PI3K/AKT and calcium pathways activity in clusters 2, 4, 7 and 8. (F) Heatmap of –log p-value calculated by the z-test of pathway activity, for two by two combinations of the four clusters indicated on the left. The analyzed pathways are indicated at the bottom of each column.

Figure 2

Inhibition of calcium signaling in Pten^del T-ALL. (A, B) Flow cytometry analysis of Calcium flux. Thymocytes from Control mice or T-ALL cells previously loaded with Indo-1 AM and stained with anti-CD4 and anti-CD8 antibodies, were incubated one minute with biotinylated anti-CD3 antibody and then analyzed by flow cytometry for 10 minutes. Fifty seconds following the start of acquisition, TCR stimulation was induced by addition of streptavidin, and at 8 min, ionomycin (a calcium ionophore) was added. The plots display the evolution over time of the ratio between two wavelengths, 410 nm and 475 nm corresponding to Indo-1 bound to Ca²⁺ and free Indo-1, respectively. The mouse model used is indicated on the top of each plot. (A) The assays were performed with OT-II mouse models in duplicate. CD4 SP thymocytes from Control OT-II or leukemic Pten^del × OT-II mice (aged between 10 and 12 weeks) were analyzed. (B) The assays were carried out with Control, Pten^del and Cdkn2a^-/- mouse models. Plots shown are representative of 3 independent experiments. For Control and Pten^del mice, thymic and splenic CD4 SP cells were analyzed. For Cdkn2a^-/- model, splenic T-ALL cells were analyzed. (C) Flow cytometry histograms showing CD3 expression in cells analyzed in panels (A, B, D) Analysis of early TCR signaling by immunoblotting. A representative case of Pten^del T-ALL (n=3), Cdkn2a^-/- T-ALL (n=3) and control thymocytes (n=3) are shown. Cells from total thymus (Pten^del and control mice) or from total spleen of NSG mice (Cdkn2a^-/- mouse model) were unstimulated (-) or stimulated (+) with anti-CD3/CD28 antibodies for 2 minutes and analyzed by immunoblotting with antibodies specific for phosphorylated AKT S473 (P-AKT), phosphorylated PLCγ1 at Y783 (P-PLCγ1), phosphorylated NFAT1 at S54 (P-NFAT1), PTEN and ACTIN. (B, D) Control and Pten^del mice were aged between 10 to 15 weeks.

Figure 3

Mathematical modelling of calcium flux and analysis of the PTEN impact. (A) Schematic representation of TCR-induced calcium flux in thymocytes. The flow of calcium between the three major calcium compartments, endoplasmic reticulum (ER), cytosol (CYT), and mitochondria (MT) is controlled by calcium channels, such as ITPR1, ITPR2, SERCA, ORAI1, ITPR3 and MERCs with the flow direction indicated by blue arrows. Upon stimulation, TCR signaling is amplified by the LAT signalosome that notably yields to the production of IP3 that binds to its receptors (ITPR) leading to the release of Ca²⁺ from ER to cytosol. The drop of Ca²⁺ in the ER induces a conformational change of STIM1 that promotes the binding of STIM1 to ORAI1, yielding to calcium fluxes from extracellular matrix to cytoplasm. Following Ca²⁺ release from ER to cytosol, the active transport system mediated by sarcoplasmic/ER Ca²⁺-ATPases (SERCA) can sequester calcium back into ER. The putative interactions of PTEN with ITPR are indicated by green. (B) Logical regulatory network representing the thymocyte specific TCR-activated calcium signaling. Nodes of the network represent calcium signaling components (ellipse nodes for Boolean node, rectangular for multilevel). Green edges stand for the activations, red ones for the inhibitions (logical functions are given in Supplementary Table 5 ). (C) Attractors of the logical model for each combinations of inputs: simulations with PTEN=1 and PTEN = 0 are represented in the left and right columns respectively, Ubq_x =0 (top) and Ubq_x = 1(bottom), and in each situation are displayed the 3 values of the TCR (0/1/2). Colors represent the activity levels, red for active (1), yellow inactive (0). Bicolored cases represent an oscillating node (cyclical attractor). (D) Comparison of cytosol state simulated with the model, and biological data extracted from literature [control, see references (–34)] and experiments (Pten^del: non-tumor and T-ALL; PTEN-proficient T-ALL that corresponds to Cdkn2a^-/- T-ALL) according to CD3 stimulation (see Figure 2 and Supplementary Figure 6 ). ND means not determined.

Figure 4

PTEN-deficiency correlates with a down regulation of IP3 receptors expression. (A) Quantification of Itpr1, Itpr2 and Itpr3 mRNA expression by RT-qPCR performed on cDNA obtained from total thymus of Control (Ctrl) mice (n = 8) and of leukemic Pten^del mice (n = 5), and total spleen from leukemic NSG mice engrafted with distinct Cdkn2a^-/- T-ALL (n = 3). Control and Pten^del mice were aged between 10 and 15 weeks. The assay was performed in duplicate. Transcripts levels were normalized to ABL. Error bars show means with SD. Statistical significant difference was assessed using Mann-Whitney test (**P< 0.01). (B) Analysis of ITPR proteins expression. Immunoblotting assays were performed with antibodies specific for ITPR1, ITPR2, ITPR3, PTEN and ACTIN as a loading control. Total thymic cells from disease-free Control (Ctrl) and leukemic Pten^del mice were analyzed. Leukemic cells from Cdkn2a^-/- model were harvested from spleen of NSG mice. The identification of analyzed mice or Cdkn2a^{-/ -} T-ALL is indicated (#number). (C) Quantification of ITPRs protein expression levels. The bands of interest in immunoblots shown in panel (B) were quantified and values of ITPR bands were first normalized to ACTIN. Then ITPR/ACTIN values of the Pten^del and Cdkn2a^-/- samples were normalized to the mean values of Control samples.