Pim1 and Pim2 are strongly TCR-induced, but dispensable for T cell activation

Estimated copies per cell of PIM1 and PIM2 protein from quantitative proteomics analysis of (A) OT1 CD8 T cells stimulated with SIINFEKL peptide for indicated times from published dataset (Marchingo et al., 2020) or (B) naïve ex vivo and 24 hour αCD3/αCD28 (TCR) activated WT CD8 T cells (see Fig 1G-H) for further details. (C) Fragments per kilobase million (FPKM) of Pim1, Pim2 and Pim3 mRNA from published (Spinelli et al., 2021) bulk RNAseq analysis of naïve and 24 hour gp33-41 peptide stimulated P14 CD8 T cells. Lymph node cell suspensions from C57BL/6 (WT) and Pim1KO/Pim2KO (Pim dKO) mice were activated for 24 hours with αCD3/αCD28 (both 0.5 µg/mL) and CD4 and CD8 T cell (D) FSC-A SSC-A profiles, (E) expression of surface activation markers (CD25, CD44, CD71) or CD8 T cell intracellular IFNγ were measured by flow cytometry. (F) Lymph node single cell suspensions from WT and Pim dKO mice were labelled with CellTrace Violet (CTV), activated with αCD3/αCD28 (both 0.5 µg/mL) and CD4 and CD8 T cell CTV proliferation profiles were measured at indicated time points. (G-H) Lymph node cell suspensions from WT and Pim dKO mice were stimulated for 24 hours with αCD3/αCD28 (both 0.5 µg/mL) and activated CD4 and CD8 T cells were sorted for analysis by quantitative proteomics. Data was analysed using proteomic ruler method (Wisniewski et al., 2014) to estimate protein copy number per cell. (G) Total protein content (µg/million cells) (one-way ANOVA), (H) Volcano plots of p-value (-log10) versus fold-change (log2) in protein copy number between Pim dKO and WT. Horizontal dotted line represents multi-test correction cut-off of q=0.05, vertical dotted line shows 1.5-fold change. Lymph node single cell suspensions from WT and Pim dKO mice were labelled with CellTrace Violet (CTV) and (I) cells were cultured in IL-7 (5ng/mL) +/-rapamycin (20 nM) and CD8 T cell numbers measured over time or (J) cells were activated with αCD3/αCD28 (both 0.5 µg/mL) +/-rapamycin (20 nM) and CD8 T cell mean division number was calculated over time (two-way ANOVA). Symbols in bar charts represent biological replicates, symbols in (I) represent the mean. Error bars show mean ± S.D. Flow cytometry dot plots and histograms are representative of (D-E) n = 3, except for IFNγ staining which is n=2 or show pooled data from (I) n=3-4 and (F, J) n=5 biological replicates, with data collected over at least 2 independent experiments. Quantitative proteomics was performed on biological triplicates.

Pim1/Pim2 deficiency reduces IL-15-driven CD8 T cell proliferation but does not prevent memory differentiation.

(A) OT1 lymph node cell suspensions were SIINFEKL peptide activated for 36 hours, washed then cultured with no cytokine, IL-15 (20 ng/mL) or IL-2 (20 ng/mL) for 4 or 24 hours. Western blots of PIM1, PIM2 or pSTAT5 Y694 expression. (B) Schematic of cytokine driven memory and effector CD8 T cell expansion and differentiation cultures. Lymph node or spleen cell suspensions were activated for 2 days with TCR stimulus + cytokine, washed, then split daily into fresh media + cytokine. (C) WT (Ly5.1) and Pim dKO LN suspensions were mixed at a 50:50 ratio for T cells and cultured as outlined in (B) with TCR stimulus αCD3/αCD28 (both 0.5 µg/mL) + cytokine IL-15 (20 ng/mL), and CD8 T cell number was measured daily. (D) WT and Pim dKO T cells were expanded with IL-15 in separate cultures as per (B-C) and % live cells (PI-ve) were assessed on days 4 and 6 (two-way ANOVA). (E-J) WT and Pim dKO CD8 T cells were activated with TCR stimulus αCD3/αCD28 (both 0.5 µg/mL) + cytokine IL-15 (20 ng/mL), expanded with IL-15 as per (B), with an additional CD4 T cell magnetic depletion step on day 3 of culture. CD8 T cells were harvested on day 6 for parallel RNAseq and proteomic analysis. (E) Fold-change in mRNA expression between Pim dKO and WT versus average mRNA expression (TPM). mRNA expression (Transcripts per million, TPM) of (F) secondary lymphoid homing receptors Sell, Ccr7, S1pr1 and (G) key transcription factors involved in CD8 T cell memory differentiation and maintenance Tcf7, Klf2, Foxo1, Foxo3, Id3. (H) WT vs Pim dKO protein copy numbers, differentially expression proteins (FC>1.5, q<0.05) are highlighted in red (I) Protein copy numbers per cell for key mitochondrial proteins DRP1, OPA1 and CPT1A. (J) WT vs Pim dKO protein copy numbers, mitochondrial proteins (as defined in MitoCarta 3.0) are highlight in pink. Symbols in bar charts represent biological replicates, symbols in (C,H,J) represent the mean. Error bars show mean ± S.D. Data are representative of (A) n=3 or show pooled data from (C) n = 4, and (D) n=5 biological replicates with data collected over at least 2 independent experiments. Quantitative proteomics and RNAseq was performed on biological triplicates. ** q:≤0.01, fold-change (FC) shown on bar graphs when q<0.05.

Pim dKO IL-2 differentiated effector T cells have reduced cell size and sustained expression of CD62L

(A) Estimated copies per cell of PIM1 and PIM2 protein from published quantitative proteomics analysis (Howden et al., 2019; Brenes et al., 2023) of CD8 T cells expanded in IL-2 or IL-15 as outlined in Fig 2B. (B-D, F-G) WT (Ly5.1) and Pim dKO lymph node or spleen single cell suspensions were mixed at a 50:50 ratio of T cells, activated for 2 days with αCD3/αCD28 (both 0.5 µg/mL) and IL-2 (20 ng/mL), washed then split into fresh medium containing IL-2 (20 ng/mL) daily (as per Fig 2B). Some of the mixed cell suspensions were also cultured in IL-7 (5 ng/mL) to sustain a naïve T cell reference. (B) WT and Pim dKO CTL were treated 1 hour +/-Jak1/3 inhibitor Tofacitinib (100 nM) (negative control) before pSTAT5 Y694 expression was measured on day 3 and 6 of culture, (C) surface CD25 expression was measured on days 3 and 6 or culture, (D) CD8 T cell number vs time was calculated, (F) CD8 T cell FSC-A, SSC-A and surface activation markers (CD44, CD71) were measured on days 3 and 6 of culture (G) expression of adhesion molecule CD62L was measured daily. (E) WT and Pim dKO T cells were activated and expanded with IL-2 as per (Fig 3B-D, F-G) except in separate cultures and % live cells (PI-ve) was assessed on days 4 and 6 (two-way ANOVA). Symbols in bar charts represent biological replicates, symbols in (D) represent the mean. Error bars show mean ± S.D. Data are representative of (B, G) n=4, (C,F) n=6 or show pooled data from (D) n=4, (E) n=6 biological replicates with data collected over at least 2 independent experiments

Major glucose transporters and effector proteins are reduced in Pim dKO IL-2 expanded CTL

WT and Pim dKO CD8 T cells were activated for 2 days with αCD3/αCD28 (both 0.5 µg/mL) and IL-2 (20 ng/mL), washed then split into fresh medium containing IL-2 (20 ng/mL) daily (as per Fig 2B), with an additional CD4 T cell magnetic depletion step on day 3 of culture. CD8 T cells were harvested on day 6 of culture for high resolution mass spectrometry. (A) Estimated total protein content per cell (student t-test). (B) Volcano plots Pim dKO vs WT protein copy numbers, differentially expressed proteins (FC>1.5, q<0.05) are highlighted in red. Estimated protein copy number per cell of (C) transcription factor TBX21 and TCF1 (D) glucose transporters SLC2A1 and SLC2A3. (E) Volcano plots Pim dKO vs WT protein copy numbers. Proteins with KEGG term = “terpenoid backbone biosynthesis”, “biosynthesis of unsaturated fatty acids” or “steroid biosynthesis” are highlighted with proteins with FC>1.5, q<0.05 shown in red and proteins with FC<1.5 and/or q>0.05 shown in pink. (F) Heatmap of protein copy numbers for granzymes, perforin and effector cytokines. (G) Granzyme B and IFNγ expression was measured by flow cytometry in day 6 IL-2 expanded WT and Pim dKO CTL. Symbols in bar charts show biological replicates. Error bars show mean ± S.D. Data are representative of (G) n = 3-4, with data collected over at least 2 independent experiment Quantitative proteomics was performed on biological triplicates. * indicates q<0.05, fold-change (FC) shown on graph when q<0.05

Disconnect between protein and mRNA expression in Pim1/Pim2-deficient effector CTL corresponds with a reduction in protein synthesis

RNAseq analysis was performed in day 6 IL-2 expanded WT and Pim dKO CD8 T cells which were collected in parallel with proteomics analysis described in Fig 4A. (A) Volcano plot of RNAseq data, differentially expressed mRNA (FC>1.5, q<0.05) are highlighted in red. (B) Volcano plot of RNAseq data, Granzymes C-K, perforin, Pdcd4 and Sell are highlighted in red. (C) Heatmap of mRNA expression (TPM) for granzymes, perforin and effector cytokines. Bar chart of mRNA expression (TPM) of (D) Granzymes A and B (E) Glucose transporters Slc2a1 and Slc2a3. (F-G) Fold change of PimdKO/WT protein (copies) from proteomics analysis described in Fig 4 vs mRNA (TPM) (F) highlighting in red proteins that are differentially expressed (FC>1.5, q<0.05) where mRNA is not substantially different (FC<1.2) and (G) highlighting in red protein and mRNA that are both differentially expressed (FC>1.5, q<0.05). (H) Estimated cytosolic ribosome content per cell (left), % ribosome of total cellular protein content (right). (I) Estimated protein copy number per cell of translation repressor PDCD4. (J) PDCD4 expression measured by flow cytometry on day 3 and 6 in IL-2 expanded WT vs Pim dKO CD8 T cells. (K) Estimated protein copy number per cell of EIF4A1. (L) adjusted ratio of PDCD4 : EIF4A1 (assuming 1 PDCD4 binds 2x EIF4A1) in WT and Pim dKO proteomes. (M) Protein synthesis measured by OPP incorporation in day 6 IL-2-expanded WT CTL treated for 24 hours with pan PIM kinase inhibitors PIM447 (5 µM) or AZD1208 (10 µM). 30 min cycloheximide (100 µg/mL) treatment gives no protein synthesis background control. Symbols in bar charts show biological replicates: error bars show mean ± S.D. Data are representative of (M) n=2 biological replicates collected over two independent experiments, (J) n = 2 biological replicates. Quantitative proteomics and RNAseq were performed on biological triplicates. * indicates q<0.05, fold-change (FC) shown on graph when q<0.05.

PIM kinases regulate mTORc1 activity and lymphoid homing in effector CTL

(A) Western blot of pS6K T389 and pan S6K from day 6 WT and PimdKO IL-2 CTL (paired student t-test, *= p<0.05). (B) WT and Pim dKO T cells were mixed in a 50:50 ratio, activated with αCD3/αCD28 (both 0.5 µg/mL) and IL-2 (20 ng/mL) and expanded in IL-2 as per Fig 2B and pS6 (Ser235/236) measured after 1 hr +/-PdBU and ionomycin (positive control), +/-rapamycin (mTORc1 inhibitor, negative control) or no additional treatment. mRNA expression (TPM) from RNAseq analysis of IL-2 expanded WT and Pim dKO CTL described in Fig 5A for (C) cell homing receptors S1pr1, Ccr7 and Sell and (D) transcription factor Klf2. (E) WT and Pim dKO T cells were activated and expanded with IL-2 in separate cultures as per Fig 2B. On day 6 of culture WT and Pim dKO CTL were labelled with CFSE or CTV, mixed at a 50:50 ratio and transferred into C57Bl/6 recipient mice. Values indicate percentage of transferred cells detected in blood, lymph node or spleen 4 or 23 hours post-transfer that were WT or Pim dKO (one-way ANOVA). Symbols show biological replicates. Error bars show mean ± S.D. Data are representative of (A) n= 6, (B) n=2-4 collected across at least 2 independent experiments and (E) n= 6 recipient mice, from n=2 biological donor replicates. RNAseq was performed on biological triplicates. * q<0.05, ** q<0.01, *** q<0.001, fold-change (FC) shown on graph when q<0.05 but FC<1.5.

T cell counts in WT vs Pim dKO spleen

Comparison of Pim dKO and age/sex matched WT control mice for (A) total number of CD4 and CD8 T cells in spleens (one-way ANOVA), (B) total number of splenocytes (student t-test), (C) proportion of live splenocytes that were CD4 and CD8 (one-way ANOVA). Symbols in bar charts represent biological replicates: error bars show mean ± S.D. (A-C) n = 8 collected over 4 independent experiments. ** q<0.01, ***q<0.001.

Pim dKO IL2 expanded CD8 T cells exhibit an effector-like mitochondrial proteome profile

Heatmap showing protein copy number of mitochondrial proteins (as defined by inclusion on Mitocarta 3.0 list) from IL-2 expanded WT (effector), Pim dKO and IL-15 expanded WT (memory) proteomics experiments described in Fig 2 and 4. Proteins are ranked based on average expression in IL2 WT condition. Proteomics was performed on biological triplicates.

24 hour treatment of IL-2 CTL with pan-PIM kinase inhibitors PIM447 or AZD1208 recapitulates many features of Pim1/Pim2-deficiency

Single cell suspension from P14 TCR-transgenic mouse lymph nodes were activated with gp33 peptide (100 ng/mL), IL-2 (20 ng/mL) and IL-12 (2 ng/mL) for 2 days, then split daily into fresh media containing IL-2 (20 ng/mL). On day 5 of culture IL-2 expanded CTL were treated with pan-PIM kinase inhibitors PIM447 (5 µM) or AZD1208 (10 µM) or DMSO vehicle control for 24 hours harvested on day 6 of culture to measure (A) Cell number (B) FSC-A SSC-A. Proteome analysis was also performed on inhibitor treated CTL to measure (C) protein content, (D) protein expression of SCD1-3, SLC2A1, SLC2A3, GZMB or (F) PDCD4. (E) Day 6 IL-2 expanded CTL from WT (C57BL/6) mice were treated for 24 hours with PIM447 or AZD1208 (both 1 µM) and GZMB expression was measured by flow cytometry. Symbols show biological replicates. Error bars show mean ± S.D. Data are representative of (E) n=2 biological replicates collected across at least 2 independent experiments. Proteomics analysis was performed on biological triplicates, with (A-B) collected in parallel with proteomics analysis. * q<0.05.