Rela-Floxed (B6.129S1-Rela^tm1Ukl/J) mice were previously described (16). CD8^cre (C57BL/6-Tg(Cd8a-cre)^1Itan/J) were a gift from Ichiro Taniuchi (RIKEN Center for Integrative Medical Sciences, Japan) (17). CD4^cre (Tg(CD4-cre)^1Cw1) on a C57Bl/6 J background were purchased from the Jackson Laboratory. Rag2^–/– and C57Bl/6 CD45.1 (Ptprc^a Pepc^b/BoyJ) mice were purchased from Charles River Laboratories France. Mice were bred and used in specific-pathogen-free (SPF) conditions at the CRCL animal facility (P-PAC) or the Ecole Normale Supérieure de Lyon BSL-2 facility (PBES, for LCVM experiments only). Animals were housed in individually ventilated cages with temperature-controlled conditions under a 12-h-light/dark cycle with free access to drinking water and food. Adult (6- to 30-week-old) male or female mice were used for all experiments. Studies were conducted in accordance with the animal care guidelines of the European Union and French laws. Protocols were validated by the local Animal Ethics Evaluation Committee (C2A15). Project references were #16772, 30346 and 43605.

Blood samples from healthy volunteers were obtained through the Etablissement Français du Sang (EFS) (French Blood Transfusion Society).

MC-38 cells were a gift from Benoit Salomon (Paris, France). B16-F10 cells were purchased from the American Tissue Culture Collection (ATCC, catalog ##6475). Braf^V600EPten^-/- (Braf-Pten) cells were a gift from Julie Caramel (CRCL, Lyon, France). Cell lines were maintained in DMEM (Gibco, catalog #61965059) supplemented with 10% Fetal Bovine Serum (FBS, Thermo Fisher Scientific, catalog # 10437028) and Penicillin/Streptomycin (Gibco, catalog #15140-122). 2×10⁵ MC38 or B16-F10 or 4×10⁵ Braf-Pten cells diluted in 50 µL sterile PBS1X were injected subcutaneously into the shaved flank back of each mouse. After seven days, tumor size was monitored every two or three days with a Caliper. Tumor volume was obtained by using the formula: Width² x Length. Anti-PD-L1 (Clone: 10F.9G2, catalog #BP0101) and isotype control mAbs (catalog #BP0090) were obtained from BioXCell. Mice received intraperitoneal injections of 200 µg of mAb diluted in PBS1X at D7, 9 and 11.

Bone marrow cells were retrieved from tibia and femur of donor mice. Red blood cells were lysed with Ammonium-Chloride-Potassium lysis Buffer and enumerated. 1/3 WT CD45.1 bone marrow and 2/3 CD45.2 bone marrow of interest (CD4^cre or CD4^cre Rela^flox/flox) were mixed to prepare bone marrow solutions. After sub-lethal irradiation (7 Gy), recipient mice were transplanted intravenously (retro orbital sinus) with 10×10⁶ bone marrow cells. Mice were given Neomycin (200 ng/mL, Sigma catalog #N6386) in drinking water for ten days. 8 weeks after reconstitution tissues were harvested for subsequent analyses.

Mice were infected via intraperitoneal injection of 2x10⁵ PFU LCMV Armstrong (a gift from Julien Marie, CRCL, France). Virus stocks were prepared using BHK21 cells, and titrated on Vero.E6 cells, following published protocols (18). Animals were weighed every other day and euthanized 10 days post infection. Spleens and livers were harvested and snap frozen in liquid nitrogen. Tissues were dissociated using ceramic beads (CK14, Ozyme catalog #OZYME003-100) and a PreCellys homogenizer, directly in lysis buffer from the Nucleospin RNA kit (Macherey-Nagel, catalog #740955.250). RNA was subsequently extracted following the manufacturer’s protocol. cDNA was synthesized using the iScript cDNA synthesis kit (BioRad, catalog #1708891). qPCR was done using a SybrGreen iTAQ kit (BioRad, catalog #1725124) on a CFX96 instrument (Biorad). The following primers were used: LCMV NP-F: CAGAAATGTTGATGCTGGACTGC; LCMV NP-R: CAGACCTTGGCTTGCTTTACACAG; LCMV GP-F: CAGACCTTGGCTTGCTTTACACAG, LCMV GP-R: GCAACTGCTGTGTTCCCGAAAC; RPLPO-F: GGACCCGAGAAGACCTCCTT; RPLPO-R: GCACATCACTCAGAATTTCAATGG (19).

Peripheral blood mononuclear cells were separated using Ficoll density gradient centrifugation (Eurobio, catalog #CMSMSL0101), and red blood cells were eliminated using Ammonium-Chloride-Potassium lysis buffer. Naive CD8⁺ T cells were isolated with the EasySep Human Naive CD8 T cell Isolation kit II (Stemcell, catalog #17968), following the manufacturer’s instructions. After enumeration, cells were cultured for 3 days at a concentration of 1x10⁶ cells/mL in complete RPMI 1640 W/HEPES W/GLUTAMAX-I (supplemented with 10% FBS, Penicillin/Streptomycin, Non-Essential Amino Acids, Sodium Pyruvate, and β-Mercaptoethanol, all from Gibco, catalog #72400054, 15140-122, 11140035, 11360039, 31350010, respectively), along with human IL-2 (50U/mL, Proleukin, Novartis Pharma) and Dynabeads CD3/CD28 (Thermo Fisher, catalog #11131D, at a ratio of 1 bead for 2 T cells). Ribonucleoprotein (RNP) complexes were prepared by combining crRNA (IDT, catalog #1072544), ATTO550-tracrRNA (IDT, catalog #1075928), CAS9 (TrueCut v2 cas9, Thermo Fisher, catalog # A36499), and electroporation enhancers (IDT, catalog #1075916) at equimolar concentrations in a final solution of 7.5 nmol/mL. crRNA sequences: RELA#1: TGCCAGAGTTTCGGTTCACT, RELA#2: AGCTGATGTGCACCGACAAG. The crRNA and tracrRNA were incubated at 95°C for 5 min followed by 15 min at 37°C before adding CAS9. The solutions were then incubated for 15 min at 37°C. Finally, an electroporation enhancer was included in the mix. Dynabeads were removed, cells were washed, and then suspended at 14x10⁶ cells/mL in Buffer T (Neon transfection system, Thermo Fisher, catalog #MPK10096). Cells were combined with RNP complexes so that 1.3x10⁶ cells could be electroporated in a 100 μL Neon Tip with 50 pmol of RNP. Electroporation was performed using the Neon transfection system with the following parameters: 1600V, 10 ms, 3 pulses. Subsequently, cells were transferred into 1.9 mL complete RPMI with IL-2 (20 U/mL) and Dynabeads CD3/CD28 (1 bead for 4 T cells). Cells were cultured for 3 days, washed, and stained with DAPI (Cell Signaling Technologies, catalog #4083S) in FACS buffer. DAPI^- ATTO550⁺ cells were sorted using a FACS ARIA II cytometer and used for subsequent assays.

Single cell suspensions from LN, thymus and spleens were obtained by mechanical dilaceration in FACS Buffer (PBS 1X + 2% FBS, 2 mM EDTA) with glass slides, strained and washed in complete RPMI.

After being sliced in small pieces tumors were digested in RPMI 1640 (Gibco) supplemented with 1 mg/mL collagenase type IV (Sigma-Aldrich, catalog #C2674) and 250 µg/mL DNase I (Sigma-Aldrich catalog #DN25) for 25 min at 37°C followed by mechanical dissociation. Reaction was stopped by the addition of 15 mL PBS1X containing 5 mM EDTA. The solution was filtered through a 70 μm cell strainer, and any remaining solid pieces were mechanically disrupted. After centrifugation, cell pellets were resuspended in 8 mL of Percoll 40% (Sigma-Aldrich, catalog #17-08-91-01) and then carefully layered onto 4 mL of Percoll 80% in a 15 mL polypropylene tube. Tubes were centrifuged at 2,500 rpm for 20 min at RT. Mononuclear cells were retrieved from the interface of the 40:80% Percoll gradient and washed in complete RPMI.

Cells were washed in PBS1X and incubated with purified anti-CD16/CD32 (Biolegend, catalog #101302) and a viability marker for 10 min at RT in the dark. After a wash in PBS1X cells were incubated with the surface marker antibody mix in FACS Buffer (PBS1X, 2% FBS, 2 mM EDTA) for 20 min at 4°C in the dark. Cells were then washed in FACS buffer and fixed and permeabilized using the eBioscience Foxp3/Transcription Factor Staining Buffer Set (Thermo Fisher Scientific, catalog #00-5523-00) according to the manufacturer’s instructions. Cells were washed in permeabilization buffer and incubated with the intracellular marker antibody mix for 20 min at 4°C in the dark. Cells were then washed in permeabilization buffer and resuspended in FACS buffer. At times, biotin-coupled antibodies were employed. In such instances, an additional stage of staining with fluorochrome-coupled streptavidin was required (in FACS buffer for cell surface labeling or wash buffer for intracellular labeling).

For intracellular cytokine analyses, cell suspensions were incubated 3 h with 50 ng/mL PMA (Sigma, catalog #P8139), 1 μg/mL ionoymycin (Sigma, catalog #I0634) in the presence of 1X Protein Transport Inhibitor containing Brefeldin A (BD GolgiPlug, catalog #555029) prior to staining, as mentioned above.

NP396-404 PE and GP33-41 APC class I tetramers were obtained through the NIH tetramer facility. The complete list of antibodies can be found in Supplementary Table 1. Acquisition was performed on a LSR Fortessa (BD Biosciences) or an Aurora spectral cytometer (Cytek Bioscience). Data were analyzed with FlowJo software v10.9.0.

RNA from 0.25 to 1x10⁶ CD8⁺ T cells was isolated with Nucleospin RNA extraction kits (Macherey Nagel, catalog #740955.250); libraries were prepared using an Illumina TruSeq Library Kit and sequenced with an Illumina NovaSeq instrument. Reads were aligned on reference genomes (mm10 for mouse data, GRCh38 for human data) using the STAR universal RNA-seq aligner; DEGs were calculated with DESeq2. Heatmaps were created with Morpheus (Morpheus (broadinstitute.org)). For functional enrichment analyses, we used the enricher function (default parameters) from the clusterProfiler package v4.2.2 to perform hypergeometric tests for functional enrichment analysis. Only down-regulated genes were used as the input, and the universe/background was defined as all detected genes in our RNA-Seq. The human hallmark, C2 and GOBP (C5:BP) gene sets were retrieved from the Molecular Signatures Database [MSigDB (20)] using the msigdbr function and package v7.4.1 Finally, we applied the Benjamini-Hochberg method to control false discoveries in multiple hypothesis testing.

ELISA were performed using “Mouse IL-2 ELISA MAX Deluxe Set”, “Mouse IFN-g ELISA MAX Deluxe Set”, “Mouse TNF-a ELISA MAX Deluxe Set”, “Human IL-2 ELISA MAX Deluxe Set”, “Human IFN-g ELISA MAX Deluxe Set”, “Human TNF-a ELISA MAX Deluxe Set” (Biolegend, catalog #431004, 430804, 430904, 431815, 430104 and 430204) and “mouse GZM-B DuoSet ELISA”, “human GZM-B DuoSet ELISA” (R&D Systems, catalog # DY1865-05 and DY2906-05) following the manufacturer’s instructions.

Total lysates were extracted using RIPA buffer (Invitrogen, catalog #89900) supplemented with protease and phosphatase inhibitors. Proteins were denatured for 5min at 95°C in Laemmli buffer (containing 9% SDS and 9% β-mercaptoéthanol), loaded on 10% polyacrylamide gel (Biorad, catalog #456-1035) and transferred onto PVDF membranes (Biorad, catalog #1704157) using a TransBlot Turbo apparatus (BioRad, catalog #1704150). Membranes were blocked with TBS1x-Tween 0.1% - milk 5% for 1 h at room temperature (RT) and then incubated overnight at 4°C with primary antibodies (see Supplementary Table 1). Membranes were washed 3 times for 10 min in TBS 1x-Tween 0.1% and incubated with corresponding HRP-coupled secondaries antibodies for 1 h at RT. Finally, membranes were washed 3 times for 10 min in TBS 1x-Tween 0.1% and detection was performed using the Immobilon Classico Western HRP substrate or the Immobilon Forte Western HRP substrate (Merck, catalog #WBLUC0500 and WBLUF0500).

Statistical analyses were performed using GraphPad Prism Software v9 (https://www.graphpad.com/scientific-software/prism/). For FACS data and tumor weights, two-tailed Mann-Whitney tests or paired T-tests (when 2 groups) and Kruskall-Wallis followed by Dunn’s post-test (when more than 2 groups) were used to calculate statistical significance. For tumor volume 2-way ANOVA followed by Bonferroni-Dunn’s post-test (when more than 2 groups), and two-tailed Mann-Whitney test (when only 2 groups) were used.

To investigate the T-cell autonomous functions of RelA, we used mice carrying floxed alleles of Rela, which we crossed with CD4^cre mice, resulting in the deletion of the gene across all T-cell subsets and the concomitant expression of green fluorescent protein (GFP) (16). In order to avoid indirect perturbations in CD8+ T cells that may rely on a role of RelA in other T-cell subsets, and thus specifically study the intrinsic functions of Rela within CD8⁺ T cells, we conducted mixed bone marrow (BM) transfer experiments of WT CD45.1⁺ cells and CD4^cre (control) or CD4^creRela^F/F (Rela-cKO^T) CD45.2⁺ cells (Figure 1A). Flow cytometry analyses performed 8 weeks after BM transfer, showed that the distribution of thymocyte subsets was similar regardless of the genotype (Figure 1B). Interestingly, analysis of secondary lymphoid organs unveiled a slight competitive disadvantage for Rela-cKO CD8⁺ T cells (Figure 1C). Among CD8⁺ T cells, a decline in CD44^high activated cells was detected in the absence of RelA compared to controls; however, proliferation, illustrated by Ki67 expression, remained unchanged (Figure 1D). As NF-κB signaling has been linked to the production of inflammatory cytokines by immune cells, we explored cytokine expression upon PMA-ionomycin restimulation in vitro. We observed a dramatic decrease in the proportion of IFNγ, TNFα and IL-2-expressing Rela-deficient CD8⁺ T cells, in the spleen and, to a lesser extent, in LN compared to controls. In contrast, Granzyme B expression was similar between groups (Figure 1E). These data were corroborated by measuring cytokine levels in culture supernatants of sorted CD8⁺ T cells showing decreased concentrations of IFNγ, TNFα and IL-2 (Figure 1F).

Seeking deeper insights into the role of RelA in regulating global gene expression, we conducted RNA-sequencing analyses on both control and mutant CD8⁺ T cells that were stimulated with anti-CD3/CD28 and IL-2 for 4 h. Differential gene expression analyses highlighted substantial changes in the transcriptome of Rela-cKO^T CD8⁺ T cells, with 293 genes significantly up- or down-regulated when applying a fold change cut-off of 2. This number increased to 493 genes when the cutoff was reduced to 1.5 (Figure 1G, Supplementary Table 2). Of note, most differentially expressed genes (DEGs) were underrepresented in Rela-deficient conditions, supporting a transcription-promoting function of this NF-κB subunit. Functional enrichment analysis of down-regulated genes with Gene Ontology Biological Processes (GOBP), Hallmarks and canonical pathways revealed a number of differentially enriched pathways, including NF-κB-related pathways as well as T-cell activation or response to virus (Figure 1H). When looking at specific gene expression, we found massive down-regulation in NF-κB pathway-related transcripts, including negative regulators (Nfkbia, Tnfaip3), but also NF-κB subunits themselves (with the exception of Rel), establishing the apex function of RelA in the regulation of NF-κB-driven genes. Furthermore, genes involved in T-cell function (Ifng, Il2) or maturation (Tbx21, Eomes) were down-regulated in Rela-deficient cells. Surprisingly, the top enriched pathways were related to IFN signaling, as the expression of many Interferon-Stimulated Genes (ISGs) was dampened in the absence of Rela (Figure 1I). Although it cannot be excluded that this phenotype stems from the reduced expression of Ifng itself, this confirms that, as proposed in other cell types, the NF-κB and IFN pathways are strongly interconnected.

Taken together, these observations establish RelA as a critical regulator of the CD8⁺ T-cell phenotype and transcriptome following polyclonal stimulation.

Next, to explore with greater specificity the CD8⁺ T-cell-autonomous functions of RelA, we crossed mice carrying Rela-foxed alleles with mice expressing Cre recombinase driven by a combination of the core E8I enhancer and the Cd8α promoter (called CD8^cre in the manuscript) (17). This allowed the conditional ablation of Rela and concomitant expression of GFP in peripheral CD8⁺ T cells (hereafter named Rela-cKO^CD8 for conditional knock-out mice) (Figure 2A). Flow cytometric analyses performed on adult control and cKO animals revealed that the distribution of CD8⁺ T cells in the thymus, spleen and peripheral lymph nodes was similar between genotypes (Figure 2B). Moreover, levels of in vivo activation and proliferation, assessed by the expression of Ki67, CD44 and CD62L, were unaltered upon Rela ablation (Figures 2C, D). Following polyclonal stimulation with PMA and ionomycin, the percentage of IFNγ-producing CD8⁺ T-cells (but not TNFα, GzmB or IL-2) was reduced in Rela-cKO^CD8 mice (Figure 2E). Hence, CD8⁺ T-cell-restricted Rela ablation did not strongly impact steady-state homeostasis and function of T cells, suggesting that the strong impairment in T-cell homeostasis detected in Figure 1 likely relied on the competitive environment of mixed BM chimeras.

Next, we assessed the impact of Rela ablation on CD8⁺ T cell responses in vitro. Naïve CD8⁺ T cells isolated from the spleen and LN of control and Rela-cKO^CD8 mice, displayed similar levels of proliferation following 4 days of CD3/CD28 stimulation, in the presence or absence of IL-2 (Figures 3A, B). In contrast, we observed a dramatic impact of Rela ablation on the ability of CD8⁺ T cells to produce inflammatory cytokines, as illustrated by the reduction in the percentage of IFNγ- and TNFα-expressing cells as well as the quantity of secreted IFNγ, TNFα and GzmB in Rela-deficient cells compared to control (Figures 3C, D). RelA thus appears to exert critical functions in cytokine expression upon long-term TCR/CD28 engagement.

Although a few patients with RELA loss-of-function mutations have been reported, its cell-autonomous roles within human CD8⁺ T cells are unknown. We established a CRISPR-Cas9 Ribonucleoprotein electroporation protocol to ablate RELA in in vitro stimulated primary human CD8⁺ T cells from healthy donors (Figure 4A). RELA (encoded by RELA) ablation was verified by Western blotting and exceeded 80% (Figure 4B). We first analyzed the gene expression profiles of both normal and KO cells through RNA-sequencing after 4 h of re-stimulation with anti-CD3/CD28 and IL-2. Loss of RELA resulted in significant changes in the expression of 322 genes (Fold change >1.5, q-value <0.005) (Figure 4C, Supplementary Table 2). Consistent with mouse cells, enrichment analyses on down-regulated genes showed that RELA governed, in human CD8⁺ T cells, the expression of genes associated with the NF-κB pathway (NF-κB subunits and negative regulators of the pathway), markers of function, and cytokines, especially the response to type I and type III IFNs (Figures 4D, E). Thus, RELA controlled different aspects of human CD8⁺ T cell biology, suggesting a CD8⁺ T-cell-intrinsic role for RELA in the immunodeficiency features detected in patients with RELA LOF (13–15). After 4 days of stimulation with anti-CD3/28, a similar level of proliferation was observed in control and RELA-deficient cells, in accordance with our data from mouse experiments (Figure 4F). Flow cytometry and ELISA analyses showed a reduction in the expression of IFNγ-producing cells in RELA-deficient cells compared to control, while the expressions of TNF-α- or GZMB remained unaltered (Figures 4G, H).

These results reinforce the role of RELA in shaping the transcriptome and functions of CD8⁺ T cells both in humans and mice.

Our RNAseq data suggested altered expression of genes related to the response to viral infections in both mouse and human Rela-deficient CD8⁺ T-cells. To directly assess the contribution of RelA to CD8⁺ T-cell-directed antiviral responses, we infected control and Rela-cKO^CD8 mice with the Armstrong strain of lymphocytic choriomeningitis virus (LCMV) that induces a strong, acute and well-defined T-cell response. Ten days post infection, the viral load, assessed by qPCR quantification of LCMV-glycoprotein (GP) and nucleoprotein (NP) encoding mRNAs, in the spleen and liver of infected mice was comparable between genotypes (Figure 5A). We also assessed the accumulation and phenotype of LCMV-specific CD8⁺ T cells in the spleen, using tetramers against both GP_33-41 (GP33) and NP_396-404 (NP396), by flow cytometry. The proportion of CD8⁺ T cells with either specificity, was unaltered in the absence of Rela (Figure 5B). Accordingly, the gross distribution of naïve/TCM/TEM compartments was similar between groups, as well as their proliferation levels (Figure 5C). Rather counter-intuitively, the proportion of GzmB and PD-1-expressing GP33-specific T cells was slightly increased in Rela-cKO^CD8 spleens. Nevertheless, the proportion of TNFα- and IFNγ-expressing cells was not altered (Figures 5D, E). Altogether, these data suggest that Rela is dispensable for the establishment of optimal CD8⁺ T-cell responses during acute LCMV infection.

As CD8⁺ T cells are known to be critical actors of anti-tumor immunity, we wondered whether Rela was required in this context. To address this, control (CD8^cre) and Rela-cKO^CD8 mice were inoculated with B16-F10 melanoma cells. Intriguingly, Rela deletion had no impact on tumor growth (Figures 6A, B). These results were confirmed in the MC38 colon carcinoma cells (Figures 6C, D). Using spectral cytometry 19 days after MC38 transfer, we observed that the proportion of CD8⁺ T cells in the tumors was similar between strains (Figure 6E). Furthermore, the activation level of tumor-infiltrating CD8⁺ T cells was unchanged in the absence of Rela (Figure 6F), leaving the proportion of cytokine-producing cells following PMA-ionomycin restimulation unaltered, with the exception of a slight increase in GzmB⁺ CD8⁺ T cells in Rela-cKO^CD8 mice (Figure 6G).

Given the role of CD8⁺ T cells in the response to checkpoint-blockade cancer immunotherapies, we subsequently investigated whether Rela activity might be a requisite for optimal response. Control (CD8^cre) and Rela-cKO^CD8 mice were transplanted with Braf^V600E-Pten^-/- melanoma cells and injected with anti-PD-L1 or isotype control mAb at D7, 9 and 11. As with the B16-F10 and MC38 models, the ablation of Rela did not affect tumor growth compared to control littermates treated with an isotype antibody (Figures 6H, I). Furthermore, PD-L1 blockade was equally efficient at inducing tumor regression in control and Rela-cKO mice (Figures 6H, I). Thus, our data indicate that Rela does not support antitumor immunity in CD8⁺ T cells or response to immunotherapy.