After virus exposure, early bystander naïve CD8 T cell activation relies on NAD+ salvage metabolism

Abstract

CD8 T cells play a central role in antiviral immunity. Type I interferons are among the earliest responders after virus exposure and can cause extensive reprogramming and antigen-independent bystander activation of CD8 T cells. Although bystander activation of pre-existing memory CD8 T cells is known to play an important role in host defense and immunopathology, its impact on naïve CD8 T cells remains underappreciated. Here we report that exposure to reovirus, both in vitro or in vivo, promotes bystander activation of naïve CD8 T cells within 24 hours and that this distinct subtype of CD8 T cell displays an innate, antiviral, type I interferon sensitized signature. The induction of bystander naïve CD8 T cells is STAT1 dependent and regulated through nicotinamide phosphoribosyl transferase (NAMPT)-mediated enzymatic actions within NAD⁺ salvage metabolic biosynthesis. These findings identify a novel aspect of CD8 T cell activation following virus infection with implications for human health and physiology.

Keywords: CD8 T cells; NAD+ salvage metabolism; antiviral immunity; bystander activation; immunometabolism; metabolic reprogramming; naïve CD8 T cells; type I interferons.

Figure 1

Reovirus induces early bystander activated naïve CD8 T cells in vivo (A) Representative dot plots demonstrating the gating strategy for the identification of CD8 T cell subsets- T_N, T_CM, T_EM and CD44lowCD62Llow (n=3 mice per non-treated group). (B) UMAP showing CD8 T_N cells in non-treated (NT) and reovirus treated (Reo 1 d.p.i.) C57BL/6 splenocytes. (C) UMAP depicting bystander activation markers in CD8 T_N cells. (D) UMAP representing TCR activation markers in CD8 T_N cells. (n=9 mice; NT-3 mice, Reo 1 d.p.i.-6 mice in all UMAP plots). (E) Bar graphs for the induction of CD8 bT_N cells upon treatment of splenocytes from C57BL/6 mice with varying MOIs of reovirus (n=3 independent experiments). (F) Bar graphs for % of CD8+IFN-γ+ T cells in OT-1 splenocytes (n=3 independent experiments). Two-way ANOVA with 95% confidence interval was used for statistical analysis in bar graphs. Significance has been indicated only for CD8 bT_N cell populations induced within 24 hours in comparison with untreated control population levels. Not significant (ns) = p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; **** p <0.0001.

Figure 2

Quantitative in vivo proteomic analysis of CD8 bT_N cells. (A) Schematic for the workflow of quantitative in vivo proteomics with T cells isolated from spleens of C57BL/6 mice (n=5 mice pooled for isolation of each cell type in an independent experiment). (B) Bar graph for protein expression from quantitative proteomics of granzyme K expression in CD8 bT_N cells (1 d.p.i. and 7 d.p.i.). (C) Bar graphs for protein expression from quantitative proteomics of memory markers TCF7 and FOXO1 expression in T_EM cells (7 d.p.i.) and CD8 bT_N cells (7 d.p.i.). (D) Histogram overlay for memory marker CD127 in T_EM cells (7 d.p.i.) and CD8 bT_N cells (7 d.p.i.). Representative histogram shown from spleen of one mouse (n=3 mice) (E) Volcano plot compares all identified proteins across CD8 T_N and CD8 bT_N cells (1 d.p.i.). (F) Cluster 3 of GO TERM analysis of proteomics depicting differentially regulated pathways in CD8 bT_N cells compared to T_N cells. (G) Heatmap for the relative levels of interferon-stimulated proteins and interferon signalling proteins identified in T_N and CD8 bT_N cells (1 d.p.i.). Corresponding bar graphs show the levels of several type I interferon-associated proteins in T_N and CD8 bT_N cells. Two-tailed Student’s t-test with 95% confidence interval was used for statistical analysis. Not significant (ns) = p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3

Induction of CD8 bT_N cells occurs in a STAT1-dependent manner. Bar graphs showing induction of CD8 bT_N cells upon ex vivo treatment of splenocytes from C57BL/6 mice with (A) varying concentrations of IFN-α1 from 10 U/mL- 200 U/mL (n=3 independent experiments), (B) varying concentrations of IFN-β1 from 10 U/mL- 200 U/mL (n=3 independent experiments) and (C) reovirus (MOI = 10) + IFNAR1 antibody (10µg/mL, n=3 independent experiments; Significance shown as indicated in figure) or isotype control. (D) IRF9 protein intensity (Two-tailed Student’s t-test). (E, F) Bar graphs for induction of CD8 bT_N cells upon ex vivo treatment of splenocytes from STAT1 KO mice with reovirus MoI = 10 (n=3 independent experiments) (E) and, IFN-α1 and IFNβ1 (200 units (U)/ml each) (n=2 independent experiements) (F). Two-way ANOVA with Tukey’s multiple comparisons test and 95% confidence interval was used for statistical analysis unless otherwise indicated. Significance has been indicated for CD8 bT_N cells in treatment conditions versus non-treated conditions unless otherwise indicated. Not significant (ns) = p > 0.05; **p < 0.01; **** p <0.0001.

Figure 4

Induction of CD8 bT_N cells ex vivo upon exposure to different viruses. Bar graphs showing the induction of CD8 bT_N cells upon ex vivo exposure of C57BL/6 splenocytes to (A) two different strains of herpes simplex virus (ICP0 and 1716, n=3 independent experiments each) at MoI = 0.1, (B) vesicular stomatitis virus (Indiana strain, n=3 independent experiments) at MoI = 10, (C) interferon activating strain of measles (Edmonston, n=2 independent experiments) at MoI = 0.1 and (D) wild type strain of measles (IC323, n=2 independent experiments) at MoI = 0.1. Two-way ANOVA with Sidak’s multiple comparisons test and 95% confidence interval was used for statistical analysis in bar graphs. Significance has been indicated only for CD8 bT_N cell populations induced within 24 hours in comparison with untreated control population levels. Not significant (ns) = p > 0.05; **p < 0.01; ***p < 0.001; **** p <0.0001.

Figure 5

Semi-targeted metabolome analysis of CD8 bT_N cells. (A) Schematic for the workflow of metabolome analysis of T cells. (B) Whole cell metabolome heatmap. (C) List of significant top upregulated and downregulated metabolites in CD8 bT_N cells. (D) Significant metabolite Enrichment analysis of significantly changed metabolites. (E) Volcano plot depiciting metabolites that are significantly changed in CD8 bT_N cells versus CD8 T_N cells. NAD⁺ salvage metabolites are highlighted in red. (F) NAD⁺ salvage pathway. NAM- Nicotinamide, NMN- Nicotinamide mononucleotide and NAD⁺- Nicotinamide adenosine dinucleotide.

Figure 6

NAD⁺ salvage metabolism regulates induction of CD8 bT_N cells. (A) Bar graphs depicting the relative peak heights of NAD⁺ salvage metabolites - NAM, NMN and NAD⁺ - using targeted metabolomics (n=4 mice per group). (B) Bar graphs for NAMPT levels- proteomics (n=5 mice pooled per condition) and quantitative PCR analysis (n = 3 independent experiments). (C–E) Bar graphs show the induction of CD8 bT_N cells upon ex vivo treatment of splenocytes from C57BL/6 mice with reovirus (MoI = 10) (C) IFN-α1 (20 U/mL) (D), and IFN-β1 (20 U/mL) (E) in the presence of FK866 (5nM) and NMN (200µM) (n=3 independent experiments for each treatment, ethanol vehicle control for FK866). Two-tailed student t-test used for statistical analysis for (A, B) Two-way ANOVA with Tukey’s multiple comaprisons and 95% confidence interval was used for statistical analysis of (C–E) Not significant (ns) = p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; **** p <0.0001.