Aiolos restricts the generation of antigen-inexperienced, virtual memory CD8+ T cells in mice Authors: Srijana Pokhrel, Gayathri Dileepan, Melissa R Leonard, Robert T Warren, Mariam Salem, Allison J Yacapraro, Qiaoke Gong, Devin M Jones, Jasmine A Tuazon, Kaitlin A Read, Jacob S Yount, Gang Xin, Adriana Forero, Hazem E Ghoneim, Patrick L Collins, Emily A Hemann, Kenneth J Oestreich Published: 2025 ABSTRACT AbstractCD8+ virtual memory T (TVM) cells rapidly respond to infection via antigen-independent bystander effector functions. While it is recognized that TVM cells arise independently of foreign antigen encounter, the mechanisms governing their development are not fully understood. Here, we identify the Ikaros transcription factor Aiolos as a negative regulator of TVM cell programming. We observe enhanced frequencies and numbers of TVM in the spleen, liver, and blood of unchallenged Aiolos-deficient (Ikzf3-/-) mice and in the lungs 1-day post-infection with influenza A virus (IAV). Furthermore, Ikzf3-/- TVM cells produce elevated IFN-γ and granzyme B in response to cytokine stimulation. Importantly, Aiolos-deficient mice control IAV more rapidly and exhibit reduced morbidity, indicating enhanced TVM cell functionality. Mechanistically, Aiolos represses the expression of the transcription factor Eomes and the IL-15R subunit CD122, known positive regulators of TVM gene program. Collectively, these findings establish Aiolos as a molecular repressor of TVM programming and responses.Subject terms: Cytotoxic T cells, Cellular immunity Introduction Host infection results in the generation of diverse, antigen-specific CD8+ memory T cell populations that persist long-term to safeguard against re-infection1–4. These include central memory T (TCM), effector memory T (TEM), and tissue-resident memory T (TRM) cell subsets, among others, which are distinguished by their functional properties, including the ability to recirculate or reside in peripheral tissues1,5,6. In contrast to these antigen-specific populations, a more recently described subset, termed virtual memory T (TVM) cells, arises independently of foreign antigen encounter7–9. As TVM and TCM cells both express L-selectin (CD62L) and the activation marker CD44, TVM were previously categorized as a part of the TCM repertoire. However, recent findings have established that low surface expression of the integrin receptor CD49d, an indicator of foreign antigen-encounter, distinguishes TVM (CD62L+CD44+CD49dLo) from these other T cell populations10,11. As TVM cell generation is not dependent upon foreign antigen exposure, they are found in significant numbers in uninfected and unchallenged hosts8. During infection, TVM cells rapidly infiltrate affected sites and provide bystander protection independent of TCR engagement7,10,12,13. As such, these cells comprise a critical component of the first line of defense against various pathogenic infections14–16. TVM effector responses are driven by signals from cytokines such as IL-12, IL-18, and IL-15 and include the production of additional inflammatory cytokines (e.g., IFN-γ) and NKG2D-mediated bystander cytotoxicity7,10,13. TVM cells can also differentiate into antigen-specific effector and memory T cell subsets, as their polyclonal repertoire allows for clonal expansion7,11,14,17. In addition to infection, TVM have also been shown to infiltrate tumors and target cancer cells independent of MHC-I18,19. Despite having a high avidity for self-antigen, which could be a concern for autoimmunity, these cells demonstrate high self-tolerance and promote homeostasis20,21. Together, their rapid, antigen-independent responses and ability to differentiate in an antigen-specific manner make TVM cells a uniquely versatile T cell population and notable therapeutic target for the treatment of a range of diseases. TVM development and function are dependent upon signals from the cytokine IL-15, as IL-15-deficient mice (IL-15-/-) lack TVM cells7,12. The response to IL-15 in TVM cells is mediated by signals received through the shared receptor subunit CD122 (IL-15Rβ/IL-2Rβ)7. Expression of CD122 is positively regulated by the T-box transcription factor Eomesodermin (Eomes), which is highly expressed in TVM cells and is essential for their differentiation and survival7,19. While Eomes promotes CD122 expression, IL-15 signaling in turn induces Eomes expression in CD8+ T cells12,22. However, the identity of the upstream regulators of this critical regulatory loop, as well as the comprehensive mechanisms underlying TVM effector functions, remain unknown. Here, we identify Aiolos as a repressor of TVM cell differentiation and function. We find that Aiolos-deficient (Ikzf3-/-) mice have significantly elevated frequencies and numbers of TVM cells both in secondary lymphoid organs at steady-state, and in the lungs 24 h following IAV infection. We further find that Aiolos-deficient TVM have elevated expression of Eomes and enhanced effector functions in response to cytokine stimulation. Importantly, this results in enhanced viral control and reduced morbidity of IAV-infected Aiolos-deficient mice relative to wild-type (WT). Collectively, these findings highlight Aiolos as a negative regulator of TVM cell formation and functional responses, suggesting that Aiolos could have therapeutic potential in enhancing antiviral immunity. Results Aiolos suppresses Eomes and CD122 expression in CD8+ T cells Eomes is a key regulator of CD8+ T cell differentiation and effector programs and is essential for TVM formation23–25. We previously established that loss of Aiolos in effector CD4+ T cells during IAV infection (A/PR/8/34 (H1N1), termed “PR8”) results in augmented expression of the transcription factor Eomes and induction of a cytotoxic gene program26. As such, we sought to determine whether the relationship between Aiolos and Eomes may be a conserved feature in CD8+ T cells and whether this may impact the generation of TVM. Initially, we examined Eomes protein expression in CD8+ T cells isolated from the spleens of unchallenged, specific-pathogen free (SPF) WT and Aiolos-deficient mice. Aiolos-deficient mice had elevated frequencies and numbers of Eomes+ T cells, and an overall increase in the abundance of Eomes protein expression compared to WT controls (Fig. 1a–c). Further analysis of CD8+ T cells revealed increased percentages of Eomes+ naive (CD62L+CD44-) and activated (CD44+) T cells in Aiolos-deficient mice compared to WT (Supplementary Fig. 1a–d). Aiolos deficiency similarly impacted the expression of the IL-15R subunit CD122, an Eomes target gene encoded by Il2rb (Fig. 1d, e)25. To further define the role of Aiolos in regulating Eomes and CD122 expression, we overexpressed Aiolos in Ikzf3-/- CD8+ T cells using retroviral transduction (Supplementary Fig. 2a, b). Overexpression of Aiolos resulted in a significant decrease in the expression of both Eomes and CD122 (Supplementary Fig. 2c, d). Collectively, these in vivo and in vitro data demonstrate that Aiolos represses Eomes and CD122 expression in primary CD8+ T cells. Aiolos expression inversely correlates with Eomes and CD122 in CD8+ T cells. Analysis of Eomes expression in CD8+ T cells isolated from the spleens of uninfected WT and Ikzf3-/- mice. a Representative flow contour plots depicting expression of Eomes in WT and Ikzf3-/- CD8+ T cells; and b respective bar graphs showing percent (%) of Eomes+ and numbers of Eomes-expressing CD8+ T cells (# counts, normalized to 6 × 105 events). c Representative histogram overlay for Eomes expression and associated data showing differences in median fluorescence intensity (MFI) fold change relative to WT control. For (a–c), data shown for 4 independent experiments, n = 14/group, mean ± SEM, two-sided, unpaired Student’s t-test, ****p ≤ 0.0001. d Representative flow plots and e associated bar graphs showing percent (%) of CD122+ and numbers of CD122-expressing CD8+ T cells (#: counts, normalized to 6 × 105 events), Data shown for 3 independent experiments, n = 11/group, mean ± SEM, two-sided, unpaired Student’s t-test, ****p ≤ 0.0001. f Analysis of publicly available Aiolos Chromatin Immunoprecipitation (ChIP)-Seq data (GSM5106065) and STAT5b ChIP-Seq data (GSM7887512) showing enrichment of Aiolos and STAT5b at the Eomes promoter region. Sequencing tracks were viewed using Integrative Genomics Viewer. The gene region cloned into a reporter vector for Eomes promoter activity is indicated. g, h EL4 T cells were transfected with an Eomes promoter-reporter construct in conjunction with vectors for Aiolos, Aiolos DNA binding mutant (AiolosDBM), constitutively active STAT5b (STAT5bCA), or empty vector control. As a control for transfection efficiency, cells were concurrently transfected with SV40-Renilla, and luciferase activity was used as a readout for promoter activity. Luciferase promoter-reporter values were normalized to SV40-Renilla control and presented as relative to the empty vector. Immunoblot analysis of the indicated proteins confirming the overexpression of respective vectors. β-actin was used as a loading control. Data are representative of 4 independent experiments, n = 4/condition, mean ± SEM, one-way ANOVA with Tukey’s multiple comparisons test, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001. Source data are provided as a Source data file. To determine the mechanism by which Aiolos suppresses Eomes and CD122, we analyzed publicly available Aiolos ChIP-Seq data from murine thymocytes [GSM510606527]. Analysis of Aiolos enrichment across the Eomes and Il2rb genes using Integrative Genomics Viewer (IGV) revealed peaks proximal to the Eomes and Il2rb promoter regions, suggesting that Aiolos may directly repress their expression (Fig. 1f, Supplementary Fig. 2e). To assess potential impacts of Aiolos, we generated plasmids to overexpress Aiolos or an Aiolos DNA-binding mutant (AiolosDBM), along with gene reporter constructs containing the identified Eomes or Il2rb regulatory regions. Aiolos overexpression resulted in suppression of both Eomes and Il2rb gene-reporter activity (Fig. 1g, Supplementary Fig. 2f). However, the repressive effect was lost with the AiolosDBM mutant, indicating that direct DNA binding is necessary for the repressive effect of Aiolos (Fig. 1g, Supplementary Fig. 2f). A known driver of Eomes expression in CD8+ T cells is the transcription factor STAT528. Interestingly, sites of STAT5b enrichment [GSM788751229] at the Eomes promoter region overlapped with the Aiolos enrichment sites (Fig. 1f). This is in line with previously published studies indicating that IkZF and STAT factor DNA-binding motifs overlap26,30,31. To determine whether Aiolos may repress Eomes expression by interfering with STAT5 mediated induction of Eomes, we overexpressed a constitutively active form of STAT5b (STAT5bCA) along with Aiolos in EL4 T cells. As expected, STAT5b overexpression alone significantly enhanced Eomes promoter activity. However, co-expression with Aiolos abrogated the STAT5-induced response (Fig. 1h). Together, these findings support Aiolos as a transcriptional repressor of Eomes in CD8+ T cells through direct interaction or via competition with STAT5 for promoter occupancy. Further, our data also support Aiolos-dependent suppression of CD122 in these cells, which could result in their decreased responsiveness to IL-15. Given the interplay between IL-15 signaling and Eomes, these findings are suggestive of a mechanism whereby Aiolos represses the IL-15/STAT5/Eomes regulatory axis and potentially TVM differentiation. Aiolos expression inversely correlates with that of Eomes and CD122 in TVM cells We next assessed the expression of Aiolos, Eomes, and CD122 in splenic CD8+ T naive, effector (TEFF, CD62L-CD44+), and TVM cells. To distinguish between circulating (CD45.2(i.v.Pos)) and non-circulating resident cells, we retro-orbitally injected mice with fluorophore-labeled anti-CD45.2 antibody prior to euthanasia (Supplementary Fig. 3a). Analysis of the resident (CD45.2(i.v. Neg)) CD8+ T cell populations revealed increased expression of Eomes and CD122 in TVM cells relative to TEFF cells, while Aiolos expression exhibited the opposite trend (Supplementary Fig. 3b, c). This inverse correlation between Aiolos expression and that of both Eomes and CD122 supports a role for Aiolos in negatively regulating TVM formation. Aiolos deficiency results in an expanded TVM population We next sought to determine the impact of Aiolos deficiency on TVM cell generation. We began by assessing T cell populations in the spleens of unchallenged, SPF WT, and Aiolos-deficient mice7,9. We observed no differences in CD3+ T cells, a slight reduction in CD4+ and CD8+ T cells, and a slight increase in double negative (DN) T cells in the spleen of Aiolos-deficient mice compared to WT, consistent with findings in human patients and mouse models harboring Aiolos variants and genetic mutations (Supplementary Fig. 4a, b)27,32–37. Further characterization of both CD4 and CD8 T cells revealed decreased naive cells with an increased effector population in the absence of Aiolos (Supplementary Fig. 4c, d), again in line with observations in human patients with Aiolos variants27,32–37. However, in the CD8+ T cell population, we observed a difference in the CD62L+CD44+ subset in which there was a substantial increase in Aiolos-deficient mice (Supplementary Fig. 4c, d). As it has been reported that TVM cells make up the majority of CD62L+CD44+ population in the uninfected host, we next used CD49d expression to distinguish TVM from TCM cells10,14,38. Consistent with elevated Eomes and CD122 expression (Fig. 1a–e), we observed a significant increase in both the frequency and numbers of CD8+ TVM cells in the spleens of Aiolos-deficient mice compared to WT (Fig. 2a, b). We next examined TVM populations in the blood, liver, and mediastinal lymph nodes and found similar increases in both frequency and numbers of TVM in Aiolos-deficient mice compared to WT (Fig. 2c, d). In contrast to the aforementioned tissues, few TVM were found in the lungs of either WT or Aiolos-deficient mice (Fig. 2d). We also noted that Aiolos-deficient TVM from these tissues exhibited elevated expression of both Eomes and CD122 relative to WT (Fig. 2e, f, Supplementary Fig. 5a, b). Aiolos deficiency results in an expanded TVM population. Flow cytometry analysis of CD8+ T cells from the spleens of uninfected WT and Ikzf3-/- mice. Mice were injected with fluorophore (PE) labeled anti-CD45.2 antibody via the retro-orbital route to distinguish resident (CD45.2(i.v. Neg)) from circulating (CD45.2(i.v. Pos)) TVM cells. a Representative flow contour plots depicting CD8+ TVM (CD62L+CD44+CD49dLo) cells in the spleen of WT and Ikzf3-/- mice and b corresponding data for percent (%) and counts (#, normalized to 6 × 105 events) of CD8+ TVM cells. Data shown for 4 independent experiments, n = 14/group, mean ± SEM, two-sided, unpaired Student’s t-test, **p ≤ 0.01 and ****p ≤ 0.0001. c Frequency (%) and d numbers (#) of TVM cells in the blood (BL), Liver (LV), mediastinal lymph node (mLN) and lung (LG) of WT vs Aiolos-deficient mice (#, normalized to 5 × 105 events each; 3 × 105 events for mLN); n = 11 for BL and LV, 2 independent experiments; n = 11 WT, 10 Ikzf3-/- for mLN, 3 independent experiments; n = 13 WT, 14 Ikzf3-/- for LG, 4 independent experiments. Data represented as mean ± SEM, two-sided, unpaired Student’s t-test, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001. Representative histogram overlays and corresponding median fluorescence intensity (MFI) fold change for (e) Eomes and (f) CD122 in the indicated populations, relative to WT naive CD8+ T cells. Data compiled from 4 independent experiments, n = 14 (for Eomes); and 2 independent experiments, n = 8/group (for CD122). Data presented as mean ± SEM; two-way ANOVA with Tukey’s multiple comparisons test, **p ≤ 0.01 and ****p ≤ 0.0001. Source data are provided as a Source data file. TVM cells form in the periphery; however, we considered the possibility that Aiolos may impact their cellular precursors in the thymus7,8,12,38. As such, we next examined thymic CD8+ T cells in WT and Aiolos-deficient mice. Consistent with published results, we observed no difference in the frequency of CD8+ T cells between WT and Aiolos-deficient samples (Supplementary Fig. 5c)36. Notably, frequencies of CD8+Eomes+ and CD8+CD122+ were relatively low in the thymus of both WT and Aiolos-deficient mice. There was no difference in the frequency and number of CD8+Eomes+ T cells, and only a slight increase in frequency, but not numbers, of CD8+CD122+ cells in the absence of Aiolos (Supplementary Fig. 5d, e). Taken together, these findings support a role for Aiolos in repressing TVM cell programming, at least in part, through modulation of Eomes and IL-15 responsiveness in the periphery. Aiolos deficiency promotes TVM gene programming To assess potential broader impacts of Aiolos on TVM programming, we performed RNA-sequencing analysis of TVM cells isolated from the spleens of unchallenged WT and Aiolos-deficient mice, in comparison to WT naive CD8+ T cells as a control. Principal Component Analysis (PCA) of transcripts for TVM cells and WT naive CD8+ T cells revealed distinct clusters between the naive and TVM samples, as has been previously reported7. Importantly, PCA analysis of the TVM transcripts revealed that Aiolos-deficient TVM samples formed a cluster distinct from WT TVM (Supplementary Fig. 6a). Hierarchical clustering of the top 200 differentially expressed genes (DEGs) identified unique gene expression patterns between WT and Aiolos-deficient TVM cells (Supplementary Fig. 6b). In Aiolos-deficient TVM cells, genes involved in TVM transcriptional regulation (Eomes, Tbx21) were upregulated along with cytokine receptors known to promote TVM homeostasis and responsiveness to IL-15 (ll15ra, Il2rb) (Fig. 3a, b)7,12. In addition, genes involved in TVM cell migration into infected tissues (Ccr2, Cxcr3, S1pr5) were upregulated in the absence of Aiolos, whereas genes encoding residency markers (Itgae, Cd69) were downregulated in Aiolos-deficient TVM samples (Fig. 3a, b)7,14. We subsequently validated the protein expression for T-bet (Tbx21) and the chemokine receptor CXCR3 (Cxcr3) in naive CD8+ T and TVM cells14,25. As expected, we observed increased expression of these factors in TVM cells compared to the naive counterparts. Importantly, their expression was significantly higher in Aiolos-deficient TVM cells when compared to WT TVM, corroborating our findings from the RNA-seq analysis (Fig. 3c, d). Aiolos represses genes underlying TVM differentiation and effector programs. CD8+ TVM cells were sorted from splenic CD8+ T cells of WT and Ikzf3-/- mice, and RNA-seq analysis was performed to assess differentially expressed genes (DEGs) between WT and Aiolos-deficient cells. Data representative of 4 independent experiments, n = 4/group. a Volcano plot displaying gene expression changes in WT vs. Aiolos-deficient cells. Genes were color-coded as follows: no significant changes in expression (gray), upregulated genes with >1.5-fold change in expression with a P < 0.05 (purple), downregulated genes with >1.5-fold change in expression with a P < 0.05 (green), and selected genes of interest (turquoise) (two-sided, Wald-test, DESeq2 analysis). b Representative heatmap of DEGs that are positively and negatively associated with the TVM gene program in WT vs. Aiolos-deficient cells. c Representative histogram overlays and d associated median fluorescence intensity (MFI) fold change for T-bet and CXCR3 expression in the indicated cell types, relative to WT naive CD8+ T cells. Data shown for 4 independent experiments, n = 14, mean ± SEM; two-way ANOVA with Tukey’s multiple comparisons test, **p ≤ 0.01 and ****p ≤ 0.0001. e, f Gene Set Enrichment Analysis (GSEA) analysis of pre-ranked (sign of fold change X −log10 (p-value)) genes using the Broad Institute GSEA software for comparison against immunological signature gene sets. Enrichment plots for the indicated gene sets are shown (pre-ranked GSEA analysis). Source data are provided as a Source data file. In addition to increased tissue infiltration, CCR2-expressing TVM cells are reported to exhibit enhanced effector responses compared to the CCR2-deficient TVM cells14. Indeed, analysis of the transcript data revealed that Aiolos-deficient TVM exhibited increased expression of natural killer cell-associated (Klrk1, Klrc1, Klra4, and Klra8), cytolytic (Gzma, Gzmb, Prf1, Fasl), and cytokine (Ifng) genes (Fig. 3a, b)9,39,40. Further, GSEA analysis of gene ontology and immunologic signature subsets suggested that genes involved in IL-15 signaling, innate immune responses, and effector and memory responses were upregulated in Aiolos-deficient TVM cells (Fig. 3e, f, Supplementary Fig. 6c–e). Collectively, these data suggest that TVM may have enhanced effector responses in the absence of Aiolos. Aiolos alters chromatin accessibility at regulatory regions for TVM genes To determine whether Aiolos impacts the TVM gene program through alterations in chromatin accessibility, we next performed Assay for Transposase-Accessible Chromatin (ATAC)-seq analysis on TVM cells isolated from the spleens of WT and Aiolos-deficient mice. Similar to RNA-Seq data, PCA analysis revealed distinct genotype clusters with broad changes in the chromatin accessibility between WT and Aiolos-deficient TVM samples, with the majority of gene regulatory regions displaying increased accessibility (Supplementary Fig. 7a, b). Consistent with the RNA-seq data, there was increased accessibility at the regulatory regions of key TVM transcription factors (Eomes, Tbx21), cytokine receptors (Il2rb, Il12rb2, Il18r1), chemokine receptors (Ccr2, Ccr5, Cxcr3), and effector molecules (Ifng, Gzmb, Prf1, Klrk1) in the absence of Aiolos (Fig. 4a–f, Supplementary Fig. 7c–f). In contrast, chromatin accessibility was reduced at the regulatory regions of other genes involved in cytokine responses (Ifngr1, Ifngr2, Tfgbr2), tissue-residency (Itgae, Cd69), and migration (Ccr9, Cxcr5) (Fig. 4a, b, Supplementary Fig. 7g, h). These findings suggest that Aiolos governs TVM gene programming directly and/or indirectly through alterations in chromatin accessibility at key regulatory and effector genes. Aiolos-deficiency alters chromatin accessibility at TVM-associated genes. Assay for Transposase-Accessible Chromatin (ATAC)-seq analysis of CD8+ TVM cells sorted from splenic CD8+ T cells of WT and Ikzf3-/- mice. Data representative of 3 independent experiments, n = 3/group. a Volcano plot displaying regions of altered accessibility in WT vs. Aiolos-deficient cells. Genes were color-coded as follows: no significant changes in accessibility (gray), increased accessibility with >1.5-fold change and a P < 0.05 (purple), decreased accessibility with >1.5-fold change and with a P < 0.05 (green), and selected genes of interest (turquoise) (two-sided, Wald-test, DESeq2 analysis). b Representative heatmap showing differentially accessible regions with distance from TSS indicated. c–f CPM-normalized representative IGV tracks. Gray boxes with asterisks represent regulatory regions of significant differential accessibility. Source data are provided as a Source data file. Aiolos deficiency enhances the functional capacity of TVM cells One of the functional characteristics of TVM cells is to provide rapid bystander protection through the production of molecules like IFN-γ, granzyme B, and NKG2D in response to cytokine stimulation (IL-12, IL-18, and IL-15)7,10. We hypothesized that the absence of Aiolos may augment responsiveness to these cytokines via alterations in cytokine receptor subunit expression. Indeed, genes encoding IL-15R subunits (Il15ra, Il2rb), IL-12R subunits (Il12rb1, Il12rb2), and IL-18R subunits (Il18r1 and Il18rap) were upregulated in TVM cells in the absence of Aiolos, suggesting that Aiolos deficiency may selectively augment TVM responsiveness to these cytokines (Figs. 3b and 5a). GSEA analysis of the immunologic signatures also revealed that genes related to IL-15/STAT5 and IL-12/STAT4 signaling were upregulated in Aiolos-deficient TVM cells (Fig. 3e, Supplementary Fig. 6e). To determine whether increased expression of cytokine receptor subunits resulted in a functional difference in cytokine signaling, we assessed the tyrosine phosphorylation status of STAT5 (pY-STAT5) and STAT4 (pY-STAT4) in TVM cells exposed to a cytokine cocktail (IL-12, IL-18, and IL-15C) ex vivo (Fig. 5b). Indeed, we observed increases in both pY-STAT5 and pY-STAT4 in the Aiolos-deficient TVM cells compared to WT, suggestive of an enhanced cytokine responsiveness in the absence of Aiolos (Fig. 5c, d, Supplementary Fig. 8a, b). To assess whether increased cytokine signaling had a functional consequence, we examined production of IFN-γ and GzmB and found that Aiolos-deficient TVM cells produced significantly elevated levels of IFN-γ and GzmB compared to WT (Fig. 5e, f, Supplementary Fig. 8c, d). We similarly examined the expression of NKG2D (encoded by Klrk1), which allows for the innate-like bystander cytotoxic functions of TVM cells38,41. Again, we observed increased frequency and expression of NKG2D in Aiolos-deficient TVM cells compared to WT in response to cytokine treatment (Fig. 5g, Supplementary Fig. 8e). We next performed a cytotoxicity assay and found increased killing of target YAC-1 cells by Aiolos-deficient TVM cells in the presence of stimulating cytokines compared to WT (Fig. 5h, Supplementary Fig. 8f)7,11,42. Notably, blocking NKG2D resulted in a significant reduction in target cell killing, specifically by Aiolos-deficient cells (Fig. 5h). Collectively, these data demonstrate that TVM have enhanced effector responses in the absence of Aiolos. Aiolos deficiency promotes cytokine responsiveness and effector functions of TVM cells. a Heatmap showing differences in different cytokine receptors between WT and Aiolos-deficient TVM samples. b Schematic of the in vitro treatment of WT and Aiolos-deficient splenic CD8+ T cells, created in Microsoft PowerPoint. Bulk CD8+ T cells were isolated from the spleens of WT and Aiolos-deficient mice and treated with rmIL-12, rmIL-18, and IL-15C (IL-15/IL-15Rα complex) for 30 min (pY-STAT5 and pY-STAT4) or 16 h (IFN-γ, granzyme B and NKG2D) with protein transport inhibitor (PTI) added 4 h prior to harvesting the cells for flow cytometry analysis. Analysis of frequency (%) and median fluorescence intensity (MFI) fold change for (c) pY-STAT5 and (d) pY-STAT4 between WT and Aiolos-deficient TVM cells, relative to WT. Data representative of 5 independent experiments, n = 5/group, mean ± SEM; two-sided, unpaired Student’s t-test, *p ≤ 0.05 and **p ≤ 0.01. Frequency (%) and MFI fold change for e IFN-γ, f granzyme B, and g NKG2D between WT and Aiolos-deficient TVM cells, relative to WT. Data represents 6 independent experiments, n = 6/group for e, f; and 5 independent experiments with n = 5/group for NKG2D. Data presented as mean ± SEM; two-sided, unpaired Student’s t-test, *p ≤ 0.05, ***p ≤ 0.001 and ****p ≤ 0.0001. h Percent (%) dead YAC-1 (target) cells when co-cultured with CD8+ T cells in the presence of cytokines (IL-12, Il-18, IL-15C), with or without anti-NKG2D antibody, data from 5 independent experiments with n = 5/condition. Data presented as mean ± SEM; two-way ANOVA with Tukey’s multiple comparisons test *p ≤ 0.05 and ****p ≤ 0.0001. Source data are provided as a Source data file. Aiolos deficiency results in increased control of IAV infection To determine the impact of Aiolos on TVM responses in vivo, we next infected WT and Aiolos-deficient mice with IAV (Supplementary Fig. 9a). Through the course of infection, animal body weights were measured daily as an assessment of infection progression and disease severity. Consistent with past observations, Aiolos-deficient mice exhibited significantly less weight loss compared to WT mice, with differences observed as early as 5 d.p.i. (Supplementary Fig. 9b)26. To determine whether the reduced weight loss was an outcome of reduced viral burden in the absence of Aiolos, we assessed viral load in the lung homogenates at 3 d.p.i. Indeed, both viral RNA and infectious virus titers were significantly reduced in Aiolos-deficient mice relative to WT, suggesting a level of increased antiviral response in the absence of Aiolos (Supplementary Fig. 9c). As Aiolos expression is largely restricted to cells of the lymphoid lineage, these early differences suggest an impact of Aiolos on an early-responding lymphocyte subset, perhaps TVM43. Aiolos deficiency enhances the response of CD8+ TVM cells to IAV infection Qi and colleagues demonstrated that TVM cells can be found in IAV-infected lungs as early as 1-day post-infection (1 d.p.i.)14. To determine whether the absence of Aiolos would augment this response, we infected WT and Aiolos-deficient mice with IAV and assessed TVM populations at 1 d.p.i. Similar to our observations in uninfected mice, the frequency and number of TVM cells were increased in the spleen of Aiolos-deficient mice at 1 d.p.i. compared to WT (Fig. 6a, b). Strikingly, there was also a substantial increase in the number of resident CD8+ TVM cells in the lungs of Aiolos-deficient mice compared to WT (Fig. 6c–e). Consistent with our previous findings, Aiolos-deficient TVM cells from infected spleens and lungs exhibited elevated expression of both Eomes and CD122 at 1 d.p.i. compared to WT (Supplementary Fig. 10a, b). Aiolos suppresses the response of CD8+ TVM cells to IAV infection. WT and Aiolos-deficient mice were intranasally (i.n.) infected with 100 plaque-forming units (PFU) of influenza A virus (IAV), and tissues were harvested 1-day post-infection (d.p.i). a Representative flow contour plot depicting resident (CD45.2(i.v. Neg)) CD8+ TVM cells in the spleens of WT and Ikzf3-/- mice 1 d.p.i. and b corresponding percent (%) positive and counts (#, normalized to 6 × 105 events) data for CD8+ TVM cells, n = 16 each from 5 independent experiments. c Representative flow plots and d corresponding % and counts (#, normalized to 5 × 105 events) data for resident CD8+ T cells in WT vs. Ikzf3-/- infected lungs 1 d.p.i. e Frequency (%) and numbers (#, normalized to 5 × 105 events) of CD8+ TVM cells in the lungs of WT and Aiolos-deficient mice 1 d.p.i., n = 16 for WT, and 14 for Ikzf3-/- from 5 independent experiments. For b, d, e, data presented as mean ± SEM; two-sided, unpaired Student’s t-test, ***p ≤ 0.001 and ****p ≤ 0.0001. f Fold change in the number of CD8+ TVM cells in the spleens and lungs of WT and Aiolos-deficient mice 1 d.p.i., relative to WT. Data presented as mean ± SEM; two-sided, unpaired Student’s t-test, **p ≤ 0.01. g Schematic of murine influenza A virus infection and concurrent treatment with Fingolimod (FTY720), created in Microsoft PowerPoint. WT and Ikzf3-/- mice were intranasally infected with 100 PFU of IAV. Mice were intraperitoneally (i.p.) injected with FTY720 (1 mg/kg/day) or saline control daily starting at 1 day prior (day -1) to infection to 2 days (day 2) post-infection. h Data showing differences in the number (#, normalized to 3 × 105 events) of resident CD8+ T cells in the lungs of WT and Ikzf3-/- mice in the control treated and FTY720 treated groups. Data shown for 3 independent experiments with n = 10 each, as mean ± SEM, Two-way ANOVA, Tukey’s multiple comparisons test, **p ≤ 0.01 and ****p ≤ 0.0001. Source data are provided as a Source data file. Comparative analysis of TVM cells in the spleens and lungs of WT and Ikzf3-/- mice revealed an average 2-fold increase in the number of TVM cells of Aiolos-deficient mice compared to WT. However, there was an approximate 12-fold average increase in the number of TVM cells in the lungs of these mice compared to WT, suggesting baseline differences in TVM numbers in Aiolos-deficient animals is not sufficient to account for this increase in the lungs (Fig. 6f). To determine whether the increased lung TVM population had trafficked from outside the lungs or rather had expanded from a lung-resident population, we treated WT and Ikzf3-/- mice with Fingolimod (FTY720) or saline control via the intraperitoneal route daily starting 1 day prior to IAV infection (Fig. 6g). FTY720 suppresses lymphocyte migration and has been widely used to study CD8+ T cell egress to infected tissues44,45. Following FTY720 administration, we infected mice with IAV and examined the presence of resident CD8+ T cells in the lungs at 3 d.p.i. As before, infected animals treated with saline control displayed a significant increase in resident CD8+ T cells in the lungs of Aiolos-deficient mice relative to WT. In contrast, FTY720-treated animals had diminished resident CD8+ T cell populations in the lungs of both WT and Aiolos-deficient mice, indicating that TVM cells migrated to the lungs in response to infection and that this migration was enhanced in the absence of Aiolos (Fig. 6h). Early lung-infiltrating Aiolos-deficient TVM correlate with enhanced control of IAV To determine whether Aiolos-deficient TVM harbored an enhanced capacity to combat infection, we initially infected WT and Ikzf3-/- mice with IAV, followed by FTY720 administration from 1 to 5 d.p.i to limit our analyses to the impact of early-responding TVM (Fig. 7a). Consistent with our previous findings (Supplementary Fig. 9b), Aiolos-deficient animals lost significantly less weight than their WT counterparts, with differences observed as early as 3 d.p.i (Fig. 7b). In fact, Aiolos-deficient mice lost very little weight throughout the 6-day time course. Analysis of the lungs at 3 and 6 d.p.i revealed increased resident CD8+ T cells in the lungs of Aiolos-deficient mice compared to WT, and an assessment of viral RNA at day 3 revealed significantly reduced levels in the lungs of FTY720-treated Aiolos-deficient mice compared to WT (Fig. 7c, d). Notably, there was no difference in viral RNA levels between saline control and FTY720-treated mice at day 3, despite reduced numbers of resident CD8+ T cells in FTY720-treated mice compared to controls (Supplementary Fig. 11a, b). By 6 d.p.i., viral RNA levels were comparable between WT and Aiolos-deficient mice (Fig. 7d). However, there was no difference between the viral RNA present in Aiolos-deficient mice at 3 and 6 d.p.i., again suggesting more rapid control of the virus in these animals relative to WT (Fig. 7d). Aiolos-deficient TVM cells provide enhanced control of IAV infection. a Schematic of murine influenza A virus infection and concurrent treatment with Fingolimod (FTY720). Mice were treated with FTY720 (1 mg/kg/day) intraperitoneally (i.p.) at days 1–5 following IAV (100 PFU) infection. Tissues were collected after euthanasia at day 3 or day 6 for flow cytometry and RNA analysis. b Weight of the mice was monitored daily up to 6 days post infection (d.p.i). Data representative of 3 independent experiments with n = 11 WT and 13 Ikzf3-/- mice, presented as mean ± SEM. *p ≤ 0.05, ****p ≤ 0.0001, two-sided, multiple unpaired Student’s t-tests. c Analysis of frequency (%) and numbers (#, normalized to 3 × 105 events) of CD45.2(i.v. Neg) CD8+ T cells in the lungs of WT and Ikzf3-/- mice at day 3 and 6. d qRT-PCR analysis for influenza virus nucleoprotein (NP) expression on RNA isolated from the lung homogenates on day 3 and 6. Data normalized to Rps18 and presented relative to WT. For (c, d), data are representative of 3 independent experiments, n = 9 each for day 3; n = 11 WT and 13 Ikzf3-/- for day 6, data presented as mean ± SEM; Two-way ANOVA, Tukey’s multiple comparisons test. **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001. e Schematic of the adoptive transfer of sorted WT and Ikzf3-/- TVM cells into recipient Cd8-/- mice and murine influenza virus infection for the determination of viral RNA levels as above. f qRT-PCR analysis for IAV NP on RNA isolated from the lung homogenates on day 3 post-infection. Data normalized to Rps18 and presented relative to WT. Data representative of 4 independent experiments with n = 17 WT and 16 Ikzf3-/- recipients. Data presented as mean ± SEM; two-sided, unpaired Student’s t-test. *p ≤ 0.05. Schematics in (a, e) were created in Microsoft PowerPoint. Source data are provided as a Source Data file. To further assess the enhanced protective capacity of Aiolos-deficient TVM cells, we adoptively transferred WT or Ikzf3-/- CD8+ TVM cells into recipient Cd8-/- mice and infected recipients with IAV 1-day post-transfer (Fig. 7e). We then measured viral RNA levels in the lungs of the recipient mice at 3 d.p.i. Consistent with our prior findings, we observed reduced viral RNA in the lungs of mice that received Aiolos-deficient TVM compared to mice that received WT TVM cells (Fig. 7f). Together, these findings demonstrate that Aiolos-deficient TVM have an enhanced capacity to control IAV infection relative to their WT counterparts. Discussion Although Eomes and IL-15 signaling are well-recognized drivers of TVM differentiation, the broader mechanisms governing CD8+ TVM cell generation remain poorly understood. Here, we demonstrate that Aiolos functions as a negative regulator of TVM cell generation and function. Aiolos deficiency results in an increase in TVM cell populations in multiple tissues under homeostatic conditions, and enhanced TVM cell recruitment to the lungs following IAV infection. Aiolos-deficient TVM cells display heightened sensitivity to signals from IL-12 and IL-15 and, consequently, elevated IFN-γ and granzyme B production, as well as NKG2D-dependent cytotoxicity, in response to cytokine stimulation. Importantly, both IFN-γ and NKG2D are critical to anti-viral responses with the former limiting viral replication and the latter a primary driver of bystander killing activity7,13. Aiolos-deficient TVM cells also have increased expression of Fasl, which could further contribute to the enhanced bystander killing properties of Aiolos-deficient TVM cells46. Ultimately, the combined increase in TVM cell numbers and heightened effector responses in the absence of Aiolos translates into the reduced viral burden and disease severity observed in Aiolos-deficient mice compared to WT. Our data support an underlying mechanism wherein Aiolos represses Eomes and CD122 expression, thus limiting cell responsiveness to IL-15 signaling and the formation of TVM cells. These data are consistent with identified roles for Aiolos in the suppression of STAT5-driven cytokine signaling pathways. For example, we previously demonstrated that Aiolos repressed CD25 expression in CD4+ T cells, which resulted in reduced IL-2 signaling and the promotion of T follicular helper cell differentiation26. In another study, it was determined that Aiolos repressed IL-15 signaling and cytotoxic programming of intestinal intraepithelial lymphocytes, and that Aiolos binding sites were proximal to those for STAT547. Similarly, data from our group also demonstrated a positive correlation between STAT5-binding sites and regulatory regions displaying increased chromatin accessibility in Aiolos-deficient CD4+ T cells26. In this study, we found that Aiolos-dependent repression of Eomes is due, at least in part, to antagonizing STAT5 activity at the Eomes promoter. Collectively, these findings are suggestive of a mechanism wherein the balance between Aiolos expression and STAT5 activation is critical, as competition for shared DNA-binding sequences at gene regulatory elements dictates IL-2 and IL-15 cytokine receptor expression and, consequently, sensitivity to these cytokines. While the necessity of Eomes in the promotion of TVM cell differentiation is clear, it should be noted that Eomes harbors several other described roles in CD8+ T cell biology23–25. For example, while Eomes is generally considered a regulator of cytotoxicity, it is also known to preferentially drive memory cell formation. This is in contrast to the related T-box factor T-bet, which promotes short-lived effector programming48. Within individual memory subsets, Eomes is known to promote circulating memory T cells while repressing tissue-resident cells24,49–51. As TVM cells contribute to the formation of antigen-specific memory populations, particularly those that are lung-resident in the context of pulmonary infections, it remains to be seen whether Aiolos deficiency may impact the development of long-term memory CD8+ T cells14. To this end, CCR2- TVM cells have been reported to preferentially differentiate into TRM cells following IAV infection14. Our RNA-seq data revealed increased Ccr2 transcript along with decreased Itgae and Cd69 transcripts in Aiolos-deficient samples, thus supporting the possibility that loss of Aiolos may limit TVM differentiation into TRM cells. Indeed, given the role of Aiolos in repressing Eomes, it is intriguing to consider that in the absence of Aiolos, unchecked Eomes expression may inhibit the formation of a robust TRM population, and perhaps, potentiate TCM or TEM subsets. Ultimately, future studies will be necessary to comprehensively assess the contribution of Aiolos to the generation of the full breadth of antigen-specific CD8+ T cell memory subsets. Finally, TVM cells have been implicated in various health conditions such as cancer and infections, where TVM infiltration into the affected sites has been shown to improve the outcome of the disease14–16,18. In that regard, the findings of this study on the role of Aiolos opens a new avenue that could be utilized to enhance TVM therapeutic capacity. In clinical settings, Aiolos, alongside Ikaros, is a known target of lenalidomide, a chemotherapeutic used for the treatment of multiple myeloma52–54. More recently, lenalidomide has been explored for lupus treatment and has demonstrated effectiveness in enhancing CAR T cell cytotoxicity in adoptive immunotherapy55–57. Our findings suggest that therapeutic targeting of Aiolos may have previously unappreciated impacts across CD8+ T cell subsets, with potential outcomes that could be either beneficial or detrimental, depending on the disease context. Ultimately, additional work is needed to comprehensively understand the contribution of Aiolos, as well as other Ikaros factors, to the differentiation and function of diverse effector and memory CD8+ T cell subsets. Methods Mouse strains and cell lines Wild-type CD45.2 C57BL/6 J (JAX stock #000664) mice were originally obtained from the Jackson Laboratory. Aiolos-deficient mice were originally obtained from Riken BRC (stock # RBRC05582) and backcrossed onto the CD45.2 C57BL/6 J background for more than 10 generations to obtain CD45.2 Aiolos-deficient (Ikzf3-/-) mice. Cd8-/- mice (JAX stock #002665) were generously provided by Dr. Ginny Bumgardner and Dr. Hazem Ghoneim at The Ohio State University. All mice were housed in a specific-pathogen free (SPF) facility. Mice were housed in 12/12 dark/light cycle, at an ambient temperature of 68–79 F with 29–33% humidity. Age- and sex-matched mice were used for both in vivo and in vitro studies to avoid any unintentional bias. Littermates were used wherever possible for experiments. The experimental and control animals were housed separately for the duration of the study. WT and Ikzf3-/- mice used in the study had similar starting weights for infection experiments. All studies performed utilizing mice were done with the approval of the Institutional Animal Care and Use Committee (IACUC) at The Ohio State University in Columbus, OH. The EL4 murine T cell lymphoma line (TIB-39) for transfection studies and YAC-1 cells (TIB-160) for cytotoxicity were acquired from the American Type Culture Collection (ATCC) and maintained in complete RPMI (RPMI media [catalog # 61870-036, Thermo Fisher Scientific] containing 10% FBS [catalog # 26140-079, Life Technologies] and 1% penicillin/streptomycin [catalog # 15140-122, Life Technologies]). For transduction studies, the Platinum-E (Plat-E) Retroviral Packaging cell Line (RV-101, Cell Biolabs, Inc.) was maintained in complete DMEM (DMEM [catalog # 11995-065, Thermo Fisher Scientific], 10% FBS, and 1% penicillin/streptomycin). MDCK (Madin–Darby canine kidney cells) cell line (ATCC, NBL-2) for viral titer determination was maintained in complete RPMI (RPMI, 10 mM HEPES [catalog # 15630080, Thermo Fisher Scientific], 1X L-glutamine [catalog # 25030081, Thermo Fisher Scientific], 1X sodium pyruvate [catalog # 11360070, Thermo Fisher Scientific], 1X MEM NEAA [catalog # 11140050, Thermo Fisher Scientific], 10% FBS and 1% penicillin/streptomycin). CD8+ T cell isolation and culture For overexpression studies, naive CD8+ T cells were isolated from the spleens and lymph nodes of 5–8-week-old Ikzf3-/- mice using the BioLegend MojoSort naive CD8+ T cell isolation kit (catalog # 480044) according to the manufacturer’s recommendations. Naive CD8+ T cells (6 × 105) in complete IMDM (IMDM [catalog # 12440061, Life Technologies] with 10% FBS, 1% penicillin/streptomycin, and 0.05% (50 μM) 2-mercaptoethanol [catalog # M3148, Sigma-Aldrich]) were plated in wells coated with anti-CD3 (5 μg/mL, clone 145-2C11, catalog # BDB553057, BD Biosciences) and anti-CD28 (10 μg/mL, clone 37.51, catalog # BDB553294, BD Biosciences) in the presence of IL-4-neutralizing antibody (5 μg/mL, clone 11B11, catalog # 504122, BioLegend) and IL-2-neutralizing antibody (5 μg/mL, JES6-1A12, catalog # 503706, BioLegend). IL-15 complex (IL-15C) was added to the samples at a concentration of 100 ng/mL of IL-15. The IL-15C was generated as previously described58. Briefly, soluble recombinant IL-15 (catalog # 447-ML-010, R&D) and recombinant mouse IL-15Rα Fc chimera (catalog # 551-MR-100, R&D) were mixed at a 1:2 ratio and incubated at 37 °C for 45 min immediately prior to use. Cells were cultured in a 37 °C incubator for the indicated times before downstream analyses. Primary T cell transduction Aiolos overexpression experiments in primary CD8+ T cells were performed as previously described26. Briefly, pMIG-Aiolos was generated by cloning the Aiolos (Ikzf3) coding sequence into the pMSCV-IRES–GFP II (pMIG II, catalog # 52107, Addgene) vector backbone. The cloning of Aiolos was confirmed by Sanger sequencing, and overexpression was validated via both qRT-PCR and immunoblot using anti-V5 antibody for Aiolos (clone SV5-Pk1; 1:20,000; catalog # R960-25, Thermo Fisher Scientific). Plat-E cells were used to package pMIG-Aiolos virus per the manufacturer’s instructions. Viral supernatants were collected 48 h post-transfection and snap frozen. Viral supernatants were thawed immediately before use and added to murine CD8+ T cells, as detailed above, in the presence of polybrene (8 μg/mL, catalog # 107689, Sigma Aldrich). Transduction was performed at 24 and 48 h after plating CD8+ T cells via spinfection for 2 h at 800 x g at room temperature. At 48 h, cells were treated with anti-IL-2 and IL-15C and harvested 24 h after the second round of transduction for analysis. Assessment of cytokine response and cytotoxic capacity For functional assessment of TVM cells, CD8+ T cells were isolated from the spleens of 5–8-week-old WT and Ikzf3-/- mice. Cells were purified using BioLegend MojoSort CD8+ T cell isolation kit (catalog # 480035) according to the manufacturer’s recommendations. Cells were then treated with rmIL-12 (10 ng/mL, catalog # 419-ML-010, R&D), rmIL-18 (10 ng/mL, catalog # 9139-IL-010, R&D), and IL-15C (100 ng/mL) for 30 min to determine phosphorylation of STAT5 and STAT4 or for 16 h to determine the production of IFN-γ and granzyme B, along with NKG2D expression. To preserve the phosphorylation of STAT proteins, cells were immediately fixed with 4% formaldehyde (37% formaldehyde, catalog # BP531, Fisher Scientific) post-stimulation and harvested for further analysis. For IFN-γ and granzyme B, cells were treated with protein transport inhibitors (PTI, catalog # 00-4980-93, Invitrogen) 4 h prior to harvest for downstream analysis. To determine TVM cytotoxic potential, cells were incubated with the cytokine mix as above and were co-cultured with YAC-1 cells (50,000) labeled with carboxyfluorescein succinimidyl ester (CFSE, 1 μM; catalog # C34554, ThermoFisher) at an effector-to-target (E:T) ratio of 10:1 for 6 h at 37 °C. Cells were then stained with the ghost viability dye (Red 780; 1:750; catalog # 13-0865-T100, Tonbo Biosciences) to determine % of dead target YAC-1 cells. The ratio of target cells killed/TVM was calculated as the number of dead cells in cytokine treated samples—untreated samples (background)/number of TVM cells. For NKG2D blocking, cells were treated with anti-NKG2D antibody (10 μg/mL, clone: HMG2D; catalog # BE0111, BioXCell) or isotype control (catalog # BE0091, BioXCell) for 30 min prior to YAC-1 cells addition. Promoter-reporter assay To generate the Eomes (pGL3-Eomes) and Il2rb (pGL3-Il2rb) promoter-reporter constructs, the regulatory region of the Eomes gene (~2 kb upstream of the transcriptional start site [TSS], Primers: Forward: 5′–GATCGGTACCGGCTTCGCTGCATCAAAAC –3′, Reverse: 5′– GATCGAGCTCGCTTTAGCGAATCGCAGACG –3′), and Il2rb gene (~805 bp proximal to the TSS, Primers: Forward: 5′–GATCGGTACCGGTGATGGGAAAAGAAGGGGCC –3′, Reverse: 5′–GATCACGCGTCACTCAGATCAACCACCACCTG –3′) were respectively cloned into the Promega pGL3-Promoter vector (catalog # PAE1761, Promega). Expression vectors for Aiolos, Aiolos DNA-binding mutant (AiolosDBM), and a constitutively active form of STAT5b (STAT5bCA) were generated as previously described26,59. Briefly, the coding sequence of Aiolos was cloned into the pcDNA3.1/V5-His-TOPO vector (catalog # K4800, Life Technologies). The coding sequence was confirmed by Sanger sequencing with T7 and BGH primers and then transferred into the pEF1/V5-His vector (catalog # V920-20, Life Technologies). A similar technique was utilized for the generation of AiolosDBM and STAT5bCA in conjunction with the Agilent QuikChange Site-Directed Mutagenesis kit (catalog # 200519). For AiolosDBM, point mutations were introduced into the first and second N-terminal zinc finger domains. For STAT5bCA, point mutations were introduced into STAT5b (H298R and S715F), resulting in constitutive phosphorylation of the Y699 residue. Using the Lonza 4D nucleofection system (program CM-120, buffer SF), EL4 T cells were nucleofected with the indicated vectors (V5-tagged Aiolos constructs, STAT5bCA, or empty vector control), along with pGL3-Eomes or pGL3-Il2rb and a SV40-Renilla vector as a control for transfection efficiency. Following 22–24 h of recovery, samples were harvested, and luciferase expression was analyzed using the Dual-Luciferase Reporter Assay System (catalog # E1910, Promega) according to the manufacturer’s instructions. Overexpression of the vectors was confirmed by Immunoblot analysis using an antibody against the V5 epitope tag (clone SV5-Pk1; 1:20,000; catalog # R960-25, Thermo Fisher Scientific) for Aiolos vectors and anti-STAT5b antibody (clone G-2; 1:10000; catalog # sc-1656, Santa Cruz Biotechnology) for STAT5bCA. RNA-Seq analysis CD8+ T cells were isolated from the spleens of WT and Ikzf3-/- mice as described above. Cells were then stained for CD44 (V450; 1:300; clone IM7; catalog # 560452, BD Biosciences), CD62L (APCe780; 1:300; clone MEL-14; catalog # 47-0621-82, ThermoFisher), CD49d (AF488; 1:100; clone R1-2; catalog # 103611, BioLegend) and viability dye (V510; 1:400; catalog # 13-0870-T100, Tonbo Biosciences) to identify TVM populations, and were sorted using the Sony MA900 cell sorter for RNA extraction. For naive RNA samples, CD8+ T cells from the spleens of WT mice were isolated using the BioLegend MojoSort naive CD8+ T cell isolation kit as mentioned above. RNA was isolated from the sorted WT and Aiolos-deficient TVM cells, as well as WT naive CD8+ T cell,s using the Machery-Nagel Nucleospin RNA kit (catalog # 740955.5) as per the manufacturer’s instructions. Purified RNA samples were submitted to Azenta Life Sciences for polyA selection, library preparation, and Illumina 2 × 150 bp sequencing. Data were pre-processed with Trimmomatic (0.36) and aligned with STAR (2.5.2b) to GRCm38. Transcript counts were generated with Subread (1.5.2) and quantified with DESeq2 (v 1.44.0) for analysis. Before analysis, expression data were filtered to exclude non-protein coding transcripts and lowly expressed genes (less than 100 counts). Genes with a DESeq2 adjusted p < 0.05 were considered differentially expressed. Principal Component Analysis (PCA) plots were generated with DESeq2 and exported to GraphPad Prism for visualization. Morpheus software (https://software.broadinstitute.org/morpheus) was used for heatmap generation and clustering by Euclidean distance using normalized log2 counts from DESeq2 analysis. Volcano plot was generated using −log10 (adjusted p value) and log2 fold change values from DESeq2 analysis in the VolcaNoseR software (https://huygens.science.uva.nl/VolcaNoseR/)60. For gene set enrichment analysis (GSEA), differentially expressed genes were pre-ranked by multiplying the sign of the fold-change by −log10 (adjusted p value) and analyzed using the Broad Institute GSEA software (v 4.3.3) for comparison against “gene ontology,” and “immunologic signature” gene sets. ATAC-seq analysis WT and Ikzf3-/- splenic TVM cells were isolated as described above. ATAC-seq analysis was performed as previously described61. Briefly, 5 × 104 cells at >90% viability were processed using the Illumina-compatible Tn5 Transposase (catalog # C01070012-30), followed by a cleanup using the Qiagen MinElute PCR Purification Kit (catalog # 28004) prior to amplification with NEBNext® High-Fidelity 2X PCR Master Mix (catalog # M0541S) and Illumina primers, according to the manufacturer’s instructions. Purified DNA were submitted to CD Genomics for sequencing by Illumina NovaseqX. Resultant sequences were trimmed and aligned to mm10 using Bowtie2 (v2.5.1), then processed with SAMtools (v1.21) and converted to bigwigs with deepTools (v3.5.6). Peaks were overlapped with ENCODE regulatory elements, and counts at peaks were determined using deepTools. Regions with adjusted P values < 0.05 were considered statistically significant. PCA, significant peaks, and data for heatmaps were generated with DESeq2. PCA values were exported to prism for visualization. Heatmap generation and hierarchical clustering of the regions with significant differential accessibility was performed using the Morpheus software (https://software.broadinstitute.org/morpheus). Volcano plot was generated using log10 (adjusted p value) and log2 fold change values in the VolcaNoseR software (https://huygens.science.uva.nl/VolcaNoseR/)60. CPM-normalized tracks were visualized using Integrative Genomics Viewer (IGV, v2.18.2). Influenza A virus infection and tissue preparation IAV strain A/PR/8/34 (H1N1, termed “PR8”) was propagated in 10-day-old specific pathogen free embryonated chicken eggs (Charles River Laboratories) and titered on MDCK cells (BEI Resources, NIAID, NIH: Kidney [Canine], NR-2628). WT and Aiolos-deficient mice between 8–12 weeks of age were infected intranasally with 30 or 100 plaque-forming units (PFU) of IAV. For the adoptive transfer studies, WT and Ikzf3-/- TVM cells were sorted as described above and retro-orbitally transferred into recipient Cd8-/- mice. The recipient mice were then infected via the intranasal route with 100 PFU of IAV. At designated times after infection, spleen and lung tissues were harvested following euthanasia with CO2 inhalation as previously described26. To distinguish circulating from resident cells, mice were retro-orbitally injected with fluorophore-labeled anti-CD45.2 antibody (PE-anti CD45.2, 2 μg, clone #104, catalog # 109808, BioLegend) 3 min prior to euthanasia14. For spleen, mediastinal lymph nodes, and thymus, single-cell suspensions were generated in tissue processing media (IMDM + 4% FBS) by passing the tissue through a 100 µm nylon mesh strainer followed by erythrocyte lysis in 0.84% NH4Cl. For lungs, the whole tissue was incubated in HBSS (catalog # 14175103, Gibco) with 1.3 mM EDTA. Lung as well as liver tissues were dissociated in Collagenase IV (catalog # LS004189, Worthington Biochemical) supplemented media (RPMI + 4% FBS) using a gentleMACS Dissociator (Miltenyi Biotech) according to the manufacturer’s instructions. Samples were then filtered through a 40 μm nylon mesh strainer and layered with a Percoll (catalog # 17-0891-01, Sigma) density gradient to isolate lymphocytes. Cells were centrifuged at 500 × g for 20 min with brakes off to allow for separation of mononuclear cells. Cells were then collected, and erythrocyte lysis was performed as previously described. Blood was collected into tubes containing 4% sodium citrate solution, and Ficoll (catalog # 17144003, Cytiva) was added for density separation. Cells were centrifuged as above with brakes off to allow mononuclear cell separation. Cells were resuspended in FACS buffer (PBS [catalog # 14190-136, Thermo Fisher Scientific] + 4% FBS) prior to staining for flow cytometry. FTY720 administration Mice were injected with saline control or Fingolimod (FTY720, catalog # S5002, Selleckchem) at 1 mg/kg in saline daily via the intraperitoneal route at the indicated time points. Tissues were then collected post-euthanasia at the indicated times for processing and analysis. Flow cytometry For flow cytometry analysis, cells were first incubated with anti-mouse CD16/32 Fc block (clone 93; catalog # 101320; BioLegend) for 5 min at 4 °C. Cells were then incubated for 30 min at 4 °C for extracellular staining with following antibodies: anti-CD8α (BV570, 1:300; clone 53-6.7; catalog # 100740, BioLegend), anti-CD3 (BV605; 1:100; clone 17A2; catalog # 100237, BioLegend), anti-CD44 (BV785; 1:300; clone IM7; catalog # 103059, BioLegend), anti-CD62L (BV650; 1:300; clone MEL-14; catalog # 104453, BioLegend), anti-CD49d (BUV395; 1:100; clone R1-2; catalog # 740219, BD Biosciences), anti-CD122 (FITC/APC; 1:00; clone 5H4; catalog # 554452, BD Biosciences/105912, BioLegend), anti-Cxcr3 (BV750; 1:100; clone CXCR3-173; catalog # 747298, BD Biosciences), anti-NKG2D (PE-Dazzle 594; 1:100; clone CX5; catalog # 130214, BioLegend) and Ghost viability dye (Red 780; 1:750; catalog # 13-0865-T100, Tonbo Biosciences). Cells were then washed twice with FACS buffer prior to intracellular staining. For intracellular staining, cells were fixed and permeabilized using the eBioscience Foxp3 transcription factor staining kit (catalog # 00-5523-00, Thermo Fisher Scientific) for 30 min or overnight at 4 °C. After fixation, samples were stained with the following antibodies in 1X permeabilization buffer (catalog # 00-5523-00, Thermo Fisher Scientific) for 30 min at room temperature: anti-Aiolos (PE/AF647; 1:100; clone S48-791; catalog # 564874/565265, BD Biosciences), anti-Eomes (PE-Cy7; 1:100; clone DAN11MAG; catalog # 25-4875-82, Thermo Fisher Scientific), anti-T-bet (PerCP-Cy5.5; 1:100; clone 4B10; catalog # 644806, BioLegend) anti-IFN-γ (APC; 1:500; clone XMG1.2; catalog # 505810, BioLegend) and anti-granzyme B (PE; 1:250; clone GB11; catalog # GRB04, Thermo Fisher Scientific). Cells were then washed with 1X permeabilization buffer and FACS buffer before resuspending in FACS buffer for analysis. To determine the phosphorylation levels of STAT proteins, cells were immediately fixed with 4% formaldehyde post-stimulation to preserve the phosphorylation of STAT proteins. Cells were then harvested and washed with 1X PBS prior to incubation with anti-CD49d (AF488) and anti-CD44 (V450) antibodies for 30 min at 4 °C. Following the extracellular staining, cells were permeabilized with cold methanol overnight at −20 °C. These cells were then incubated with antibodies against pY-STAT5 (Y694, APC; 1:12.5; clone: SRBCZX; catalog # 17901042, Fisher Scientific), pY-STAT4 (PY933, PerCP-Cy5.5; 1:12.5; clone: 38/p-Stat4; catalog # 561217, BD Biosciences), and CD8α (PE,1:300; clone 53-6.7; catalog # 100708, BioLegend), for 30 min before resuspending in FACS buffer for analysis. Samples were run on a BD FACS Canto II or BD FACS Symphony and analyzed using FlowJo software (version 10.8.1). Virus titers and viral RNA determination Virus titers and virus RNA were determined as previously described62–64. Briefly, mice were euthanized at designated time points post-infection, and lungs were harvested in 1 mL PBS + 0.5% BSA in Precellys tubes (catalog # 50-154-2932, Bertin Corp) for homogenization using the Precellys homogenizer. The lung supernatants were collected, snap frozen, and stored at −80 °C prior to use. For the determination of virus titers, MDCK cells were plated at a density of 7 × 105 cells/well in a 6-well plate. Virus supernatants were 10-fold serially diluted and added to MDCK cells in the presence of infection media (1X PBS, 35% BSA [catalog # ICN810061, Fisher Scientific], 1X Ca/Mg). Following 1 h incubation at 37 °C the the viral supernatant was removed, and cells were washed with 1X PBS. An overlay (2% Oxoid agar [catalog # LP0028, Thermo Fisher Scientific], 2X MEM [catalog # 12-684 F, Lonza], TPCK Trypsin [catalog # 20233, Thermo Fisher Scientific], DEAE-Dextran [catalog # MP 021951331, Thermo Fisher Scientific], 5% sodium bicarbonate, and H20) was added to the wells. After 48 h of incubation, the overlay was removed, and cells were fixed with 3% paraformaldehyde (PFA). Cells were blocked with 5% non-fat milk in 1X PBS and incubated with chicken anti-PR8 antibody (catalog # NR-3098, BEI Resources) and rabbit anti-chicken peroxidase (catalog # 303-035-003, Jackson ImmunoResearch). The number of plaques formed were quantified following the addition of TrueBlue Peroxidase (catalog # KPL 5510-0030, Thermo Fisher Scientific). For viral RNA determination, lungs were harvested in 1 mL Trizol (catalog # 15596026, Thermo Fisher Scientific) and processed as above. 1-bromo-3-chloropropane (catalog # B62404, Sigma Aldrich) was added to the clarified virus supernatant. The aqueous layer was collected post-centrifugation, ethanol was added, and the mixture was transferred onto a spin column from the Machery-Nagel Nucleospin RNA kit. RNA was extracted according to the manufacturer’s instructions. cDNA was prepared, and qRT-PCR analysis was then performed to determine virus RNA levels. RNA isolation and Quantitative Real-Time PCR RNA was isolated as described above utilizing the Machery-Nagel Nucleospin RNA kit per the manufacturer’s guidelines. cDNA was then generated using the Superscript IV First Strand Synthesis System (catalog # 18091050, Thermo Fisher Scientific). SYBR Select Mastermix for CFX (catalog # 4472942, ThermoFisher) was used for qRT-PCR reactions with 10 ng cDNA per reaction on the CFX Connect (BioRad). The following primers were used: (IAV NP primers- Forward: 5′ – GATTGGTGGAATTGGACGAT – 3′, Reverse: 5′ – AGAGCACCATTCTCTCTATT – 3′, Rps18 primers- Forward: 5′ –GGAGAACTCACGGAGGATGAG –3′, Reverse: 5′ – CGCAGCTTGTTGTCTAGACCG -3′). Outputs were normalized to Rps18 and are presented as relative to their respective control, as indicated. Immunoblot analysis For immunoblot analysis, cells were harvested, lysed in 1X SDS loading dye, (50 mM Tris [pH 6.8], 100 mM DTT, 2% SDS, 0.1% bromophenol blue, 10% glycerol), and boiled for 15 min. Lysates were separated via SDS-PAGE on 10% Bis-Tris Bolt gels (catalog # NW00102BOX, Thermo Fisher Scientific) and then transferred onto a 0.45 µm nitrocellulose membrane (catalog # 10600002, Cytiva). Membranes were blocked with 2% nonfat dry milk in 1X TBST (10 mM Tris [pH 8.0], 150 mM NaCl, 0.05% Tween-20). Membranes were then incubated with the following antibodies to detect proteins: α-V5 (1:20,000; clone SV5-Pk1, catalog # R960-25, Thermo Fisher Scientific), α-STAT5b antibody (1:10,000; clone G-2, catalog # sc-1656, Santa Cruz Biotechnology), goat α-mouse HRP (1:10,000; catalog # 115-035-174, Jackson Immunoresearch), α-β-actin-HRP (1:15,000; clone: 2D1D10, catalog # A00730, GenScript). Clarity Max Western ECL substrate (catalog # 1705062, BioRad) was used for the visualization of protein bands. Statistics and reproducibility All statistical analyses were performed using the GraphPad Prism software (version 10.3.1). Unpaired Student’s t-test was performed for single comparisons. For multiple comparisons, one-way or two-way ANOVA with Tukey’s multiple comparisons test were performed, as appropriate. Error bars indicate the standard error of the mean. P values ≤ 0.05 were considered statistically significant. DESeq2 was used for RNA- and ATAC-seq analysis. Software summary Data were collected utilizing the following open-source or commercially available software programs: BD FACSDiva (version 8.0.2), Sony MA900 cell sorter software (version 3.1.2), BioRad Image Lab (version 6.0.1, build 34), BioRad CFX Manager (version 3.1). Analyses and/or manuscript preparation were conducted using Microsoft Office Suite (including Microsoft Word, Microsoft Excel, and Microsoft PowerPoint), BD FlowJo (version 10.8.1), Integrative Genomics Viewer (version 2.18.2), Bowtie2 (version 2.5.1), SAMtools (version 1.21), deepTools (version 3.5.6), Morpheus (Broad Institute), Gene Set Enrichment Analysis (GSEA, version 4.3.3; Broad Institute) and VolcaNoseR (https://huygens.science.uva.nl/VolcaNoseR/)60. All statistical analyses were performed using the GraphPad Prism software (version 10.3.1). Data preparation for this manuscript did not require the use of custom code or software. Schematics in the manuscript were created in Microsoft PowerPoint. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Reporting Summary Transparent Peer Review file Source data Source Data Acknowledgements The authors would like to thank all members of the Oestreich Laboratory, as well as colleagues in the Department of Microbial Infection and Immunity, for critical feedback. The authors would also like to thank members of the Yount Laboratory for help with influenza virus preparation and members of the Hemann Laboratory (Abigail Solstad) and Forero Laboratory (Matthew McFadden) for their technical advice on influenza virus titer and RNA determination. The authors would also like to thank Dr. Ginny Bumgardner and Dr. Hazem Ghoneim for providing Cd8-/- mice, and Dr. Eugene Oltz and Matthew McCullen (Oltz Laboratory) for providing YAC-1 cells. This work was supported by grants from the National Institutes of Health (R01AI134972) and funds through The Ohio State University College of Medicine and The Ohio State University Comprehensive Cancer Center to K.J. Oestreich. S. Pokhrel was supported by funding through the American Heart Association Postdoctoral Fellowship (10.58275/AHA.24POST1196412.pc.gr.190821) and The Ohio State University Infectious Disease institute (IDI) Interdisciplinary Research Seed Grant (KJO). S. Pokhrel, M.R. Leonard, and J.A. Tuazon were supported by the National Institutes of Health T32: AI165391, Interdisciplinary Program in Microbe-Host Biology. G. Dileepan was supported by The Ohio State University Dean’s Distinguished University Fellowship. J.A. Tuazon was also supported by National Institutes of Health F30: AI172189-01A1 and The Ohio State University Susan Huntington Dean’s Distinguished University Fellowship. Author contributions S. Pokhrel assisted with the design of the study, performed experiments, analyzed data, and wrote the manuscript. G. Dileepan, M.R. Leonard, R.T. Warren, M. Salem, A.J. Yacapraro, Q. Gong, D.M. Jones, J.A. Tuazon, and K.A. Read performed experiments and analyzed data. J.S. Yount, G. Xin, A. Forero, H.E. Ghoneim, and E.A. Hemann assisted with the influenza virus infection experiments, data analysis, and reviewed the manuscript. P.L. Collins assisted with the RNA-and ATAC-Seq data analyses and reviewed the manuscript. K.J. Oestreich supervised the research, designed the study, analyzed data, and wrote the manuscript. Peer review Peer review information Nature Communications thanks anonymous reviewers for their contribution to the peer review of this work. A peer review file is available. Data availability All data are included in the article, Supplementary Information or available from the authors, as are unique reagents used in this Article. The raw numbers for charts and graphs are available in the Source Data file whenever possible. RNA- and ATAC-seq datasets generated in this study have been deposited in the GEO repository under accession number GSE305888. The following publicly available ChIP-seq data were used in this study: Aiolos ChIP-Seq: GSM510606527 and STAT5b ChIP-Seq: GSM788751229. Source data are provided with this paper. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary information The online version contains supplementary material available at 10.1038/s41467-025-67540-8. References Associated Data Supplementary Materials Supplementary Information Reporting Summary Transparent Peer Review file Source Data Data Availability Statement All data are included in the article, Supplementary Information or available from the authors, as are unique reagents used in this Article. The raw numbers for charts and graphs are available in the Source Data file whenever possible. RNA- and ATAC-seq datasets generated in this study have been deposited in the GEO repository under accession number GSE305888. The following publicly available ChIP-seq data were used in this study: Aiolos ChIP-Seq: GSM510606527 and STAT5b ChIP-Seq: GSM788751229. Source data are provided with this paper.