AMBRA1 controls the translation of immune-specific genes in T lymphocytes

Abstract

T cell receptor (TCR) engagement causes a global cellular response that entrains signaling pathways, cell cycle regulation, and cell death. The molecular regulation of mRNA translation in these processes is poorly understood. Using a whole-genome CRISPR screen for regulators of CD95 (FAS/APO-1)-mediated T cell death, we identified AMBRA1, a protein previously studied for its roles in autophagy, E3 ubiquitin ligase activity, and cyclin regulation. T cells lacking AMBRA1 resisted FAS-mediated cell death by down-regulating FAS expression at the translational level. We show that AMBRA1 is a vital regulator of ribosome protein biosynthesis and ribosome loading on select mRNAs, whereby it plays a key role in balancing TCR signaling with cell cycle regulation pathways. We also found that AMBRA1 itself is translationally controlled by TCR stimulation via the CD28-PI3K-mTORC1-EIF4F pathway. Together, these findings shed light on the molecular control of translation after T cell activation and implicate AMBRA1 as a translational regulator governing TCR signaling, cell cycle progression, and T cell death.

Keywords: AMBRA1; FAS signaling pathway; T cell activation; T cell death; protein translation.

Fig. 1.

CRISPR screen reveals that AMBRA1 facilitates FAS-induced T cell death. (A) Scheme of whole-genome CRISPR screen of FAS-induced cell death using Jurkat T cells. (B) Canonical FAS apoptosis pathway. Horizontal lines are cell membranes. Pathway genes found within the top 10 (blue) or lower ranked (>10) (green) hits of the CRISPR screen are highlighted. (C) Gene hit identification by related sgRNAs compared to the nontreatment group. P-values are corrected for multiple hypothesis testing. FAS (purple), other genes within the canonical FAS pathway (blue), the most significant gene hits within top 10 candidates (orange), and AMBRA1 (red) are highlighted. (D) Relative killing fold of indicated high ranking genes relative to control cells (single guide normal control sgNC-1 and sgNC-2; dark blue) after anti-FAS antibodies [Top, 10 ng mL⁻¹ anti-FAS with cross-linker (0.5 µg mL⁻¹ protein A)] or FASL [Bottom, 100 ng mL⁻¹ FAS ligand (FASL)] treatment. FAS (sgFAS-1 and sgFAS-2; purple), AMBRA1 (sgAMBRA1-1, sgAMBRA1-2, and sgAMBRA1-3; red), and other gene knockout (KO) cells that have a value higher than 2 (anti-FAS blue; FASL orange) are highlighted. Relative cell survival (Fold) was calculated by high-ranking gene sgRNA cell survival percentage/average of sgNC-1 and sgNC-2 cell survival percentage. Experiments (D) are representative of at least three independent repeats. Comparisons were calculated using a two-tailed, unpaired Student’s t test (D). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Bars (D) represent mean ± SEM.

Fig. 2.

AMBRA1 regulates FAS expression independently of protein degradation, mRNA levels, and autophagy. (A) FAS surface staining mean fluorescence intensity (MFI) fold change normalized to control Jurkat T cells of indicated genes from the CRISPR screen. Control (blue), FAS KO (purple), and AMBRA1 KO (red) cells are highlighted. (B) Western blot (WB) analysis of FAS, AMBRA1, and HSP90 (as a loading control), using lysates from control (NC) and AMBRA1 KO (KO1 and KO2 lines using two different guide RNAs) Jurkat T cells. (C) WB analysis of AMBRA1, BECN1, FAS, and GAPDH (as a loading control), in normal control (NC, NC1 and NC2 are two different scramble sgRNA), AMBRA1 KO, and BECN1 KO Jurkat T cells. (D) Flow cytometric analysis of FAS expression in NC, AMBRA1 KO, AMBRA1 overexpression (OE), and BECN1 KO Jurkat T cells. The right panel is quantification of FAS MFI. (E) Ratio of FAS mRNA determined by qPCR in NC and AMBRA1 KO (two different guide RNAs each) Jurkat T cells pulsed with Actinomycin D (1 μg mL⁻¹) and sampled at the indicated time points in hours (h). (F) FAS protein degradation ratio in control, AMBRA1 KO, and BECN1 KO Jurkat T cells after treatment with 50 µg mL⁻¹ cycloheximide (CHX) and sampled at the indicated time points in hours (h) determined by flow cytometry. (G) Flow cytometric analysis of FAS in control and AMBRA1 KO Jurkat T cells after 0.5 µM MG132 (Left) or 12.5 µM chloroquine (CQ; Right) treatments (tx). (H) Quantification of FAS MFI as in (G). Statistical comparison of AMBRA1 KO samples was nonsignificant (ns). Experiments are representative of at least three (A, B, D, and F–H) or two (C and E) independent repeats. Comparisons were calculated using a two-tailed, unpaired Student’s t test (A, D, and H). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Bars (A, D, F, and H) represent mean ± SEM.

Fig. 3.

AMBRA1 regulates FAS mRNA translation. (A) FAS protein ratio for normal control (NC), AMBRA1 KO, and BECN1 KO Jurkat T cells treated with cycloheximide overnight, washed, and cultured in fresh medium for the indicated time points in hours (h) determined by flow cytometry. FAS protein ratio was calculated by FAS MFI at the different time points normalized to 0 h in NC, AMBRA1 KO, and BECN1 KO. (B) Translation rate measurements experimental strategy. Jurkat T cells were incubated with the methionine mimic homopropargylglycine (HPG) in medium, followed by HPG conjugated to five-carboxytetramethylrhodamine (TAMRA) using Click-iT chemistry (Left). WB analysis of TAMRA after immunoprecipitation for FAS, HSP90, and GAPDH in the lysates from NC and AMBRA1 KO Jurkat T cells that were pulsed with Click-iT HPG for 2 h (Right). IP = immunoprecipitation. (C) Polysome profiles for NC and AMBRA1 KO Jurkat T cells after lysates were loaded onto a 10 to 50% sucrose gradient and centrifuged for 4 h (Top). WB analysis of large (RPL8) and small (RPS26) ribosomal subunits and AMBRA1 in the indicated fractions using control Jurkat T cell lysates (Bottom). RNP = ribonucleotide particles (soluble fractions). (D) qPCR analysis of FAS (two independent primers), CXCR4, and ACTB in RNA preparations from NC and AMBRA1 KO Jurkat T cells as in (C). (E) Flow cytometric analysis of purified human primary T cells at the indicated times in days (d) after TCR stimulation of FAS protein expression shown as mean fluorescence intensity (MFI), qPCR analysis of FAS mRNA expression after TCR stimulation at the indicated time points. FAS mRNA was normalized to ACTB. Results are representative of at least three (A, and C–E) or two (B) independent repeats. Comparisons were calculated using a two-way ANOVA (A) a two-tailed, unpaired Student’s t test (E). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Bars (A and E) represent mean ± SEM.

Fig. 4.

AMBRA1 is translationally induced by TCR/BCR stimulation through the CD28–PI3K–mTORC1–eIF4F axis. (A) Western blot (WB) analysis of AMBRA1 and GAPDH using lysates from purified human primary T cells incubated with anti-CD3/CD28 beads for the indicated time points in hours (h) or days (d). (B) qPCR analysis of AMBRA1 normalized to ACTB as in (A). (C) WB analysis of AMBRA1 and GAPDH using lysates from purified human primary B cells incubated with anti-IgM and CD40L for the indicated time points in hours (h). (D) qPCR analysis of AMBRA1 normalized to ACTB as in (C). (E) Comparison of protein amount determined by mass spectrometry (Y axis) and specific RNA by sequencing (X axis) from human T cells incubated with anti-CD3/CD28 beads for 24 h vs. unstimulated (0 h). Both the fold changes (FC) of mRNA and protein in 24 h were compared with the unstimulated group; the discordant genes/proteins (mRNA decreased and protein increased) are labeled in blue and AMBRA1 labeled in red. (F) The KEGG pathway analysis of the discordant genes as in (E). (G) Heat map of the protein and mRNA changes of top 30 mRNA discordant genes as in (E). (H) WB analysis of AMBRA1 and GAPDH using lysates from purified human primary T cells preincubated with the indicated inhibitors (LCK inhibitor A-770041 10 µM, MEK inhibitor U0126 10 µM, PI3K inhibitor 3-Methyladenine 10 µM, IKK inhibitor TPCA-1 10 µM, JNK inhibitor SP 600125 10 µM, and PKC inhibitor staurosporine 10 µM) for 30 min at 37 °C and then incubated with anti-CD3/CD28 beads for two days. (I) WB analysis of AMBRA1 and GAPDH using lysates from purified human primary T cells preincubated with the indicated inhibitors (Rapamycin (Rapa), 100 nM; Torin 1 µM) for 30 min at 37 °C and then incubated with anti-CD28 at the indicated concentration (µg mL⁻¹) for two days. (J) (Top) The eIF4F complex recognition element. (Bottom) The 5’-untranslated region (UTR) nucleotide sequence of AMBRA1 in different species. (K) WB analysis of AMBRA1, p-S6, and GAPDH using lysates from human pan-T cells preincubated with 4EGI-1 at the indicated concentrations for 30 min at 37 °C and then incubated with anti-CD3/CD28 beads for two days. (L) Jurkat T cells were transduced with the lentivirus plasmid with the GFP coding sequence, with or without the 5’UTR of AMBRA1 (Top), and then went stimulated with TCR second messenger mimetics (20 ng mL⁻¹ for phorbol myristate acetate (PMA) + 500 ng mL⁻¹ Ionomycin). GFP expression was analyzed by flow cytometry. The inhibitors Torin (1 µM) and 4EGI-1 (100 µM) were added before TCR stimulation, and the GFP expression level was analyzed after TCR stimulation. (M) Schematic of the mechanism by which AMBRA1 is translationally induced by TCR/BCR stimulation through the CD28–PI3K–mTORC1–eIF4F axis. Experiments are representative of at least three (A–H, and L) or two (I and K) independent repeats. Comparisons were calculated using a two-tailed, unpaired Student’s t test (B, D, and L). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Bars (B, D, and L) represent mean ± SEM.

Fig. 5.

Global translation is reduced in AMBRA1 KO cells. (A) Horizontal volcano plot showing the enrichment intensity versus intensity of proteins identified in AMBRA1-BioID experiments compared to the empty vector (EV). AMBRA1 (red) and ribosomal proteins (blue) are indicated. (B) STRING analysis of AMBRA1-associated proteins from BioID screen showing clusters involved in rRNA metabolic processing and ribosome biogenesis. (C) Scheme for mass spectrometry translation analysis, where Jurkat T cells were pulsed with heavy isotope-labeled amino acids for 2 h. (D) Volcano plot of relative protein levels analyzed by mePROD proteomics within a 2-h heavy amino acid pulsed period in AMBRA1 KO divided by normal control (NC) Jurkat T cells measured by mass spectrometry. Ribosomal proteins (red) are highlighted. (E) KEGG pathway analysis of down-regulated proteins in AMBRA1 KO Jurkat T cells compared to control cells. The ribosome pathway (red) is highlighted. (F) Schematic diagram illustrating protein synthesis measurement strategies using O-propargyl-puromycin (OPP) incorporation and puromycin antibody detection. (G) Flow cytometric analysis of OPP incorporation and puromycin (puro) detection in control and AMBRA1 KO Jurkat T cells after treatment with or without CHX for 2 h with anti-CD3/CD28 beads stimulation. (H) Quantification of the mean fluorescence intensity (MFI) of OPP incorporation and puromycin (puro) detection as in (G). (I) Volcano plot of amino acid fold change in AMBRA1 KO and control Jurkat T cells without (Left) or with (Right) TCR mimetic stimulation (20 ng mL⁻¹ PMA + 500 ng mL⁻¹ ionomycin). Experiments are representative of at least three (A, G, and H) or two (D and I) independent repeats. Comparisons were calculated using a two-tailed, unpaired Student’s t test (H). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Bars (H) represent mean ± SEM.

Fig. 6.

AMBRA1 regulates the mRNA translation of other immune genes. (A) Volcano plot of the fold change of proteins in AMBRA1 KO over normal control (NC) Jurkat T cells. (B) KEGG pathway analysis of down- and up-regulated proteins in AMBRA1 KO Jurkat T cells as in (A). TCR signaling (blue) and cell cycle (orange) pathways are highlighted. (C) Flow cytometric analysis of CD69, CD28, and CXCR4 in control and AMBRA1 KO Jurkat T cells incubated with 50 ng mL⁻¹ PMA and 1 µg mL⁻¹ ionomycin for the indicated time points in hours (h). (D) Quantification of CD69, CD28, and CXCR4 protein MFI (Top) and qPCR analysis of CD69, CD28, and CXCR4 mRNA (Bottom) as in (C). (E) WB analysis of AMBRA1, FAS, CD28, and HSP90 using lysates from normal control (NC) and AMBRA1 KO human primary T cells. (F) WB analysis of AMBRA1, FAS, and GAPDH using lysates from normal control (NC), AMBRA1 overexpression (OE), AMBRA1 KO, and AMBRA1 rescue (AMBRA1 KO + AMBRA1 OE) Jurkat T cells. (G) Flow cytometric analysis of FAS (Upper Middle) and CD28 (Upper Right) in Jurkat T cells gated for low (orange), middle (blue), and high (red) expression of AMBRA1 (Upper Left). Lower panels are FAS and CD28 intracellular and surface MFI statistics. (H) Proposed model of AMBRA1 as a TCR-inducible translational control circuit. Experiments are representative of at least three (C, D, F, and G) or two (A and E) independent repeats. Comparisons were calculated using a two-tailed, unpaired Student’s t test (G). *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001. Bars (D and G) represent mean ± SEM.