Quantitative analysis of how Myc controls T cell proteomes and metabolic pathways during T cell activation

Abstract

T cell expansion and differentiation are critically dependent on the transcription factor c-Myc (Myc). Herein we use quantitative mass-spectrometry to reveal how Myc controls antigen receptor driven cell growth and proteome restructuring in murine T cells. Analysis of copy numbers per cell of >7000 proteins provides new understanding of the selective role of Myc in controlling the protein machinery that govern T cell fate. The data identify both Myc dependent and independent metabolic processes in immune activated T cells. We uncover that a primary function of Myc is to control expression of multiple amino acid transporters and that loss of a single Myc-controlled amino acid transporter effectively phenocopies the impact of Myc deletion. This study provides a comprehensive map of how Myc selectively shapes T cell phenotypes, revealing that Myc induction of amino acid transport is pivotal for subsequent bioenergetic and biosynthetic programs and licences T cell receptor driven proteome reprogramming.

Keywords: Myc; T cell activation; T lymphocyte; amino acid transport; immunology; inflammation; mouse; proteomics.

Figure 1.. Myc controls cell growth by selectively remodelling T cell proteomes.

(A) Forward scatter area (FSC-A) of IL-7 maintained or 24 hr anti-CD3 + anti-CD28 (TCR) activated Cd4Cre⁺ (Myc^WT) and Cd4Cre⁺Myc^fl/fl (Myc^cKO) T cells. (B–C, E–O) Quantitative proteomics data of ex vivo naïve WT and 24 hr TCR activated CD4⁺ and CD8⁺ T cells from Myc^WT and Myc^cKO mice. (B) Total protein content (µg/million cells). (C) Mean protein copy number per cell estimated using proteomic ruler (Wiśniewski et al., 2014) of Myc. (D) Myc expression measured by flow cytometry in 24 hr TCR activated Myc^WT and Myc^cKO CD4⁺ and CD8⁺ T cells. Proteins from 24 hr TCR activated Myc^WT (E) CD8⁺ and (F) CD4⁺ T cells were ranked by mass contribution and the mean cumulative protein mass was plotted against protein rank (left panel). Numbers in each quartile indicate total proteins summed with those in the quartiles below. Volcano plots show foldchange in protein copy number between TCR activated Myc^cKO and Myc^WT T cells, with proteins that contribute the top 75% of the T cell mass shown in red (right panel). (G) Heat maps of naïve and TCR activated Myc^WT and Myc^cKO CD8⁺ and CD4⁺ proteomes. Relative protein abundance is graded from low (blue) to high (yellow) per row. Input data for heatmaps is listed in Supplementary file 1. Mean protein copy number per cell for activation markers (H) IL7ra (J) CD69 and (K) CD44 and key transcription factors (I) Klf2, (L) Rel, (M) JunB, (N) Tbet, and (O) Irf4. Symbols on bar charts represent biological replicates: error bars show mean ± S.E.M. Quantitative proteomics was performed on biological triplicates. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, and TCR Myc^WT vs TCR Myc^cKO protein copy number per cell is listed in Supplementary file 1.

Figure 1—figure supplement 1.. Immune activated Myc-deficient T cells fail to proliferate.

(A) CFSE labelled lymph node cells from Cd4Cre⁺ (Myc^WT) and Cd4Cre⁺ Myc^fl/fl (Myc^cKO) mice were stimulated with anti-CD3 and anti-CD28 (both 0.5 µg/mL) for 48 hr and CFSE dilution was measured. Representative of technical duplicate plots. (B) Lymph node cells from GFP-MycKI mice were stimulated with anti-CD3 (1 µg/ml) and anti-CD28 (3 µg/ml) in the presence of IL-2 receptor-blocking antibody PC61 (2 µg/ml) or isotype control and GFP-Myc expression was measured over time. Symbols show biological replicates.

Figure 1—figure supplement 2.. Myc-deficiency has a larger quantitative effect in CD8⁺ T cells.

Naïve WT and 24 hr TCR activated Myc^WT and Myc^cKO proteomic data was generated as described in Figure 1 and Materials and methods. (A) Venn diagram showing that there were a larger number of proteins that were classified as differentially regulated in CD8⁺ T cells compared to CD4⁺ T cells, where differentially regulated is defined as proteins that were more than 2-fold regulated and p<0.05 between Myc^WT vs Myc^cKO TCR activated conditions. (B) Heat maps of proteins that were Myc-regulated in CD4⁺ T cells only, both CD4⁺ and CD8⁺ T cells or CD8⁺ T cells only show similar qualitative effects of Myc-deficiency for most proteins. Relative protein abundance is graded from low (blue) to high (yellow) per row. (C) Mean copy number per cell for examples of proteins classified as Myc-regulated in CD8⁺ T cells only. Protein expression was much higher in Myc^WT CD8 T cells compared Myc^WT CD4⁺ T cells, whereas expression levels in Myc^cKO CD4⁺ and CD8⁺ T cells are reduced to a similar extent thus giving a larger quantitative effect of Myc-deficiency in CD8⁺ T cells. Examples include important transcription factors such as Nfkb1 and Nfkb2, the integrin Icam1 and cytokine receptor components such as Il2rb. Symbols show biological replicates. Mean ± S.E.M. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, and TCR Myc^WT vs TCR Myc^cKO protein copy number per cell is listed in Supplementary file 1.

Figure 2.. Myc control of T cell metabolism is selective.

Naïve WT and 24 hr TCR activated Myc^WT and Myc^cKO CD4⁺ and CD8⁺ T cell proteomic data was generated as described in Figure 1 and Materials and methods. (A) schematic of nutrient transporters and enzymes involved in glycolysis. (B) Mean copy number per cell for glucose transporters Slc2a1 and Slc2a3. (C) Total protein content (µg/million cells) (left panel) and % contribution to total cellular protein mass (right panel) of total glycolytic enzymes. (D) Mean protein copy number per cell for lactate transporters Slc16a1 and Slc16a3. (E) Schematic of transporters and enzymes involved in Glutamine transport and metabolism. (F) Mean copy number per cell for major enzymes involved in glutamine metabolism. Symbols on bar charts represent biological replicates from biological triplicate data, error bars show mean ± S.E.M. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, and TCR Myc^WT vs TCR Myc^cKO protein copy number per cell is listed in Supplementary file 1.

Figure 3.. Myc induces amino acid transporter expression, a critical step for proteome remodelling.

Naïve WT and 24 hr TCR activated Myc^WT and Myc^cKO CD4⁺ and CD8⁺ T cell proteomic data was generated as described in Figure 1 and Materials and methods. (A) Mean copy number per cell of abundant amino acid transporters in T cells. (B) Fold-change in amino acid transporter protein copy number from naïve WT to 24 hr TCR activated Myc^WT and Myc^cKO, mean (min, max). Transport by system L amino acid transporters was measured by uptake of fluorescent (emission 450 nm when excited at 405 nm) Tryptophan metabolite, Kynurenine (Kyn) (Sinclair et al., 2018) in (C) 2 hr IL-7 maintained and (D-E) 4 hr IL-7 maintained or TCR activated splenic CD4⁺ and CD8⁺ WT, Myc^cKO and Slc7a5^cKO T cells or (F) 20 hr TCR activated Myc-GFP reporter CD4⁺ T cells. In (C,E) system-L uptake is represented as the ratio of BCH (a system L inhibitor) untreated: treated T cells. In (E) dotted line indicates Slc7a5^cKO uptake level. (G) Forward scatter and CD69 expression of IL-7 maintained or 24 hr TCR activated wild-type and Slc7a5^cKO (Cd4Cre⁺ Slc7a5^fl/fl) T cells. (H-K) Quantitative proteomics data of naïve WT and 24 hr TCR activated CD4⁺ and CD8⁺ T cells from Ly5.1 (Slc7a5^WT) and Slc7a5^cKO mice. Baseline naïve WT data is the same as used for the Myc^cKO dataset. (H) Total protein content (µg/million cells). (I) Heat map of naïve and TCR activated Slc7a5^WT and Slc7a5^cKO CD4⁺ T cell proteomes. Relative protein abundance is graded from low (blue) to high (yellow) per row. Input data for heatmaps is listed in Supplementary file 1. (J) Venn diagram showing the overlap in TCR regulated proteins that are more than 2-fold regulated and p<0.05 in Myc^WT vs Myc^cKO and Slc7a5^WT vs Slc7a5^cKO TCR activated CD4⁺ T cells. (K) Volcano plots of TCR regulated proteins comparing Slc7a5^WT and Slc7a5^cKO datatsets. Proteins > 2 fold different between Myc^WT and Myc^cKO TCR activated T cells are highlighted in red; proteins reduced in the Myc^cKO (left panel), proteins higher Myc^cKO (right panel). Symbols in bar charts represent biological replicates: error bars show mean ± S.E.M. Dot plot in (F) is representative of biological triplicate data. Quantitative proteomics was performed on biological triplicates. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, TCR Myc^WT vs TCR Myc^cKO, naïve WT vs TCR Slc7a5^WT, naïve WT vs TCR Slc7a5^cKO and TCR Slc7a5^WT vs TCR Slc7a5^cKO protein copy number per cell is listed in Supplementary file 1.

Figure 3—figure supplement 1.. Myc-deficient T cells fail to induce protein translation machinery.

Naïve WT and 24 hr TCR activated Myc^WT and Myc^cKO proteomic data was generated as described in Figure 1 and Materials and methods. Diagram and mean copy number per cell of (A) translation initiation complex and (B) 43S preintiaition complex components. (C) Protein content of 40S and 60S ribosome (µg/million cells). Symbols show biological replicates. Mean ± S.E.M. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, and TCR Myc^WT vs TCR Myc^cKO protein copy number per cell is listed in Supplementary file 1.

Figure 3—figure supplement 2.. Amino acid transport capacity corresponds with transporter number.

Transport by system L amino acid transporters was measured by uptake of fluorescent (emission 450 nm when excited at 405 nm) Tryptophan metabolite, Kynurenine (Kyn) in 6 or 24 hr IL-7 maintained or anti-CD3/anti-CD28 (TCR) activated CD4⁺ and CD8⁺ WT and Myc^cKO T cells. Histograms are representative of at least three biological replicates.

Figure 3—figure supplement 3.. Myc-deficient T cells fail to induce Branched-chain amino acid and Methionine metabolism.

Naïve WT and 24 hr TCR activated Myc^WT and Myc^cKO proteomic data was generated as described in Figure 1 and Materials and methods. Diagram and mean copy number per cell of proteins involved in (A) Methionine metabolism and (B) Branched-chain amino acid metabolism. Symbols show biological replicates. Mean ± S.E.M. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, and TCR Myc^WT vs TCR Myc^cKO protein copy number per cell is listed in Supplementary file 1.

Figure 3—figure supplement 4.. Ribosome expression is reduced in both Myc and Slc7a5 deficient T cells, but other proteins are differentially regulated between Myc and Slc7a5 deficient T cells.

(A) Protein content of ribosomes (excluding mitochondrial) from naïve WT and 24 hr TCR activated Myc^WT and Myc^cKO proteomics data (generated as described in Figure 1 and Materials and methods). (B) Protein content of ribosomes (excluding mitochondrial) and (C) Slc2a3 mean copy number per cell from naïve WT and 24 hr TCR activated Slc7a5^WT and Slc7a5^cKO proteomics data (generated as described in Figure 3 and Materials and methods). Symbols show biological replicates. Mean ± S.E.M. Fold-change calculations and statistical testing comparing naïve WT vs TCR Myc^WT, naïve WT vs TCR Myc^cKO, TCR Myc^WT vs TCR Myc^cKO, naïve WT vs TCR Slc7a5^WT, naïve WT vs TCR Slc7a5^cKO and TCR Slc7a5^WT vs TCR Slc7a5^cKO protein copy number per cell is listed in Supplementary file 1. Splenocytes from Myc^WT, Myc^cKO and Slc7a5^cKO were stimulated with anti-CD3 and anti-CD28 for 24 hr and; (D) Golgi plug with or without PdBU and ionomycin was added to the cultures for the last 4 hr prior to harvest and IFNgamma expression was measured in CD8⁺ T cells by flow cytometry; or (E) Granzyme B was measured in CD8⁺ T cells by flow cytometry. Histograms are representative of at least three biological replicates.

Figure 4.. Myc induces amino acid transport early after TCR activation, triggering a feedforward loop maintaining its own expression.

(A) Expression levels of Slc7a5, Slc7a1, Slc1a5 and Slc38a2 vs Myc mRNA from published single cell RNAseq dataset of OT1 T cells stimulated with SIINFEKL (N4) peptide for the indicated times (Richard et al., 2018). Dotted lines represent 95th percentile of 0 hr gene expression. Symbols represent individual cells (B) Quantitative proteomics of OT-I CD8⁺ T cells activated with N4 peptide for the indicated time. Mean copy number per cell of the proteins Myc, Slc7a5, Slc7a1, Slc1a5, Slc38a1 and Slc38a2. (C) Slc7a5, Slc1a5 and Slc7a1 mRNA measured by qPCR from ex vivo naïve or 4 hr TCR activated lymph node CD4⁺ and CD8⁺ T cells. mRNA levels are relative to naïve CD4⁺ T cells. (D) Histograms and geometric mean fluorescence intensity (gMFI) vs time of Myc protein measured with antibody by flow cytometry in IL-7 maintained or TCR activated Slc7a5^WT and Slc7a5^cKO lymph node CD4⁺ and CD8⁺ T cells. Representative of 4 biological replicates. (E) Myc-GFP gMFI in IL-7 maintained or GFP+ TCR activated CD4⁺ and CD8⁺ T cells ± rapamycin from lymph nodes of Myc-GFP reporter mouse. Myc-GFP reporter expression in CD4⁺ and CD8⁺ T cells maintained in IL-7 or TCR activated in (F) amino-acid free media (-aa, HBSS), vs RPMI (+aa) or (G) media deficient in a single amino acid. Data representative of at least three biological replicates. Symbols unless otherwise stated represent biological replicates. Mean ± S.E.M. Quantitative proteomics was performed on biological triplicates.