A hypomorphic Il2rb mutant mouse model recapitulates and reveals mechanisms of human T cell immune dysregulation in IL-2Rβ deficiency

SUMMARY

Here, we describe the use of a homologous knockin mouse model to further decipher the mechanism(s) of a novel human homozygous IL2RB hypomorphic mutation. Our model recapitulates the human immune dysregulation phenotype, showing decreased mutant interleukin-2Rβ (IL-2Rβ) cell-surface expression, impaired IL-2/15-dependent STAT5 signaling, elevated serum IL-2/15 levels, expanded effector memory CD8⁺ T cells, and severely reduced regulatory T cells (Tregs). Using mixed bone marrow chimeras (BMCs) and wild-type (WT) Treg transfers, we distinguish receptor-intrinsic from receptor-extrinsic immunopathogenesis. Both approaches suppress abnormal serum cytokine levels and autoimmunity without affecting endogenous mutant Tregs. Mutant animals receiving WT Tregs neonatally exhibit almost complete restoration of conventional T cell distribution, IL-2Rβ receptor surface expression, and STAT5 signal transduction, while BMC animals exhibit only partial restoration. Our findings demonstrate that CD8⁺ T cells and Tregs have distinct IL-2/15 ligand/receptor ratios and signaling thresholds required for proper development/function, revealing mechanistic insights applicable to immunotherapy for autoimmunity.

In brief

Cabrera-Martinez et al. investigate a hypomorphic Il2rb mouse model that recapitulates human immune dysregulation, revealing distinct IL-2/15 signaling thresholds for Tregs and CD8⁺ T cells. This translational human-to-mouse-back-to-human framework distinguishes receptor-intrinsic from cytokine-milieu-extrinsic mechanisms due to the mutant IL-2Rβ and supports early Treg-based or cytokine milieu immunomodulation as therapeutic strategies for cytokine-driven autoimmunity.

Graphical Abstract

INTRODUCTION

The discovery and study of inborn errors of immunity (IEIs) illuminates mechanisms of human immunity, which guides therapy for these disorders. IEIs caused by mutations in the interleukin-2 receptor (IL-2R), as well as its signaling intermediaries Janus kinase 3 (JAK3) and signal transducer and activator of transcription 5b (STAT5b), highlight the crucial role of IL-2 and IL-2R functions in maintaining immunity.^1–6 High-affinity IL-2R encompasses three subunits: α (CD25), β (CD122), and the common γ chain (γ_c; CD132). CD132 is shared between IL-15, IL-4, IL-7, IL-9, and IL-21, with each of these cytokines exhibiting some degree of biological redundancy and/or complementary function.^7,8 Notably, the receptors for IL-2 and IL-15 also share their β chain, with β and γ together acting as the signal transducers. The IL-2α chain acts to engage IL-2 with high affinity and stabilizes the complex. IL-2 and IL-15 are important pro-inflammatory cytokines functioning in proliferation and maintenance of T, B, and natural killer (NK) cells.^9–12 IL-2 plays an essential role in promoting T cell activation and memory formation, while IL-15 regulates T cell memory and NK survival, proliferation, and effector functions.^4,11,13 Both IL-2 and IL-15 play pivotal roles in the body’s defense against viral infections.^14–18 To balance opposing cytokine functions, different IL-2R subunits vary in their (1) protein expression levels on lymphocyte subsets, (2) ligand binding affinities, and (3) downstream signaling pathways that are shared with other cytokines.⁶

In humans, null mutations in IL2RG result in severe combined immunodeficiency (SCID), while hypomorphic IL2RG mutations lead to mild immunodeficiency, with severe immune dysregulation.^19,20 The Il2rb^−/− mouse model revealed the importance of IL-2Rβ signaling in T cell differentiation and control of autoimmunity through its effect on regulatory T cells (Tregs).^3,21–24 Malek and co-workers demonstrated that T-cell-specific expression of STAT5 is required to support Treg development, STAT5 activation is sufficient to restore Treg cells in the thymus of Il2rb^−/− mice,^25–27 and pSTAT5 binds to conserved sites in the foxp3 promoter, creating a direct link between IL-2R complex signaling and foxp3 expression.³ The absence of IL-2Rβ expression/function, however, prevented exploration of IL-2/15R-dependent signaling on Treg, CD8⁺ T, and NK cell differentiation/function. Thus, while genetic knockout mouse models have shed light on fundamental molecular mechanisms, they do not accurately represent human hypomorphic disorders. These models fail to replicate the tolerogenic-immunogenic imbalance seen in patients with genetic mutations that allow some level of protein expression and/or function. Animal models that faithfully recapitulate the pathogenesis of human IEI defects are rare but extremely useful to decipher mechanisms of immune dysregulation in human disorders. For example, the hypomorphic R229Q Rag2 mouse model revealed mechanisms of disrupted central and peripheral tolerance.²⁸

In 2019, we published a study describing two siblings with a novel homozygous hypomorphic mutation in IL2RB. These patients clinically presented with failure to thrive, hepatosplenomegaly, chronic diarrhea and colitis, interstitial lung disease, anemia, and susceptibility to cytomegalovirus (CMV) infection.¹ The three-amino-acid homozygous IL2RB deletion p.Pro222_Gln225del¹ resulted in decreased IL-2Rβ cell-surface expression, increased serum IL-2/15 levels, and dysregulated STAT5 signal transduction, showing increased baseline pSTAT5 refractory to upregulation of phosphorylation in response to exogenous IL-2/15 stimulation.¹ Counterintuitively, these defects led to the expansion of memory CD8⁺ T and CD56^bright immature NK cells, which require IL-2/15 downstream signaling, suggesting that both intrinsic (hypomorphic receptor) and extrinsic (serum cytokine milieu) mechanisms were at play. Understanding how both receptor-intrinsic and extrinsic mechanisms contribute to the clinical phenotype is critical to choosing therapy—replacement of mutant receptors with wild-type (WT)/healthy receptors requires hematopoietic stem cell (HSC) transplantation (HSCT; receptor intrinsic), while normalization of the cytokine environment requires immunomodulation (receptor extrinsic). Understanding the minimal number of WT cells required for normal immune function in mixed chimerism is a critical therapeutic question in HSCT, as decisions about additional interventions such as more aggressive immunosuppression and/or secondary transplant/boost depend on establishing this threshold.

Given the difficulty in addressing mechanistic questions from patient-derived samples, we created an animal model with the IL2RB p.Pro222_Gln225 deletion, which recapitulates the pathogenesis observed in the patients. This mouse model has allowed not only assessment of mutant IL-2Rβ expression and signaling capability but also evaluation of both receptor-intrinsic (mutant IL-2Rβ) and extrinsic (pro-inflammatory cytokine milieu) mechanisms of immune dysregulation. Here, we test the hypothesis that Tregs and CD8⁺ T effector memory (TEM) cells require different IL-2Rβ-dependent signaling thresholds. Our findings confirm this difference, demonstrating that normalization of the pro-inflammatory cytokine environment differentially rescues these cell populations from their mutant receptor phenotype while expanding our understanding of human IL-2 and IL-15 receptor biology and its potential applications for autoimmune immunotherapies.

RESULTS

The Il2rb^mut/mut mouse model recapitulates the immune dysregulation phenotype observed in human patients

We used CRISPR-Cas9 genome editing to generate a knockin mouse bearing the mouse homolog of the human IL2RB p.Pro222_Gln225del mutation (Il2rb^mut/mut; Il2rb Mut). This mutation is predicted to disrupt the highly conserved TrpSer-XaaTrpSer (WSXWS)²⁹ extracellular motif, a common motif important for type I cytokine receptor folding, trafficking, binding, and signaling^30–32 and highly conserved across species.^1,33 Founder animals were screened for off-target effects and propagated via crossing with WT and heterozygous (HET) animals. Homozygous Il2rb Mut mice (like affected patients) exhibit features of immune dysregulation, including failure to thrive (decreased body weight; Figure 1A), splenomegaly, lymphadenopathy (Figures 1A and S1A), intestinal inflammation with chronic diarrhea, and anemia. Given these manifestations, Il2rb Mut mice survive only to ~10 weeks of age. Importantly, in addition to pro-inflammatory cytokines such as IL-6, tumor necrosis factor-α (TNF-α), and interferon-γ (IFN-γ) (Figure S1B), peripheral blood from Il2rb Mut mice shows elevated serum IL-2 and IL-15 cytokine/IL-15Rα receptor complex (IL-15c) as early as weaning age (Figure 1B). Like patients with the homozygous mutation, Il2rb Mut mice exhibit decreased Tregs (Figure 1C) and increased CD4⁺ and CD8⁺ TEM cells (Figure 1D). Notably, these changes in T cell population dynamics are observed in both absolute counts and relative frequencies (Figure 1E). Unlike the increase in immature CD56^bright NK cells observed in humans, we observed decreased mature stage E (CD11b⁺CD27⁺) NK cells and an excess of mature/potent cytotoxic (CD11b⁺CD27⁻) cells³⁴ (Figure S2) in Il2rb Mut mice. The proportion of mature/cytotoxic NK cells increases with age, as circulating cytokines continue to accumulate (Figures S2 and 1B). These data suggest that NK cell maturation and cytotoxicity are linked to circulating cytokine levels.^34,35 Further assessment of NK cell development and function was not pursued due to the substantial differences observed in canonical cell marker expression and transcriptional programming between humans and mice.^34,36,37 Overall, Il2rb Mut mice largely recapitulate the immune phenotype of hypomorphic p.Pro222_Gln225del IL2RB deletion patients, enabling dissection of both receptor-intrinsic and extrinsic (cytokine milieu) mechanisms driving immune dysregulation.

Figure 1. Il2rb^mut/mut mice recapitulate the immune dysregulation phenotype observed in human patients.

Genotypes shown include WT (black) and Il2rb Mut (red). (A)–(E) each represent three individual experiments.

(A) Total body or spleen weight of 6- to 8-week littermates, n = 15 per genotype.

(B) Serum cytokine levels of IL-2 and IL-15c (IL-15/15Rα) comparing mice at age 4 weeks (left, WT, n = 5; Il2rb Mut, n = 6) and 7 weeks (right, IL-2, n = 19 per genotype; IL-15c, n = 17 for WT and n = 18 for Il2rb Mut).

(C) Representative dot plot depicting Tregs defined as CD4⁺CD25⁺Foxp3⁺ for genotypes indicated (left). Percentage of Tregs from CD4⁺ T cells, n = 11 per genotype (right).

(D) Representative dot plot depicting subpopulations of CD4⁺ or CD8⁺ T cells: naive, lower right quad; central memory (TCM), upper right quad; and effector memory (TEM), upper left quad, for genotypes indicated (left). Stacked graphs showing T cell distribution, n = 11 per genotype (right).

(E) Absolute numbers of Tregs–CD4⁺CD25⁺Foxp3⁺ (left), n = 11 per genotype. Absolute numbers of TEM–CD44⁺CD62L⁻ cells (right), n = 7 per genotype. Data are shown as mean ± SEM. All p values were calculated by non-parametric Mann-Whitney test for a single factor; displayed as ****p ≤ 0.0001, ***p ≤ 0.001, **p ≤ 0.01, and *p ≤ 0.05 or ns, non-significant.

Il2rb Mut mice exhibit chronic T cell activation in the periphery and an inflamed bone marrow niche

Using flow cytometry, we evaluated total (surface and intracellular) IL-2Rβ (CD122) protein expression in CD4⁺ and CD8⁺ T cells. In Il2rb Mut mice these subsets express decreased levels of CD122, failing to upregulate CD122 expression upon differentiation from naive (red circles) to TEM cells (red diamonds) as compared to WT cells (black diamonds) (Figure 2A). Despite this decrease, cells from Il2rb Mut mice maintained a WT hierarchical CD122 expression pattern (NK > CD8⁺ TEM > Tregs and other non-conventional CD4⁺ T; Figure S3). To measure CD122 cell-surface expression only (as opposed to total expression) we utilized imaging flow cytometry, corroborating that CD4⁺ and CD8⁺ T cells express decreased cell-surface CD122 in Il2rb Mut mice as compared to WT (Figures 2B, 2C, and S3). Expression of IL-2Rα (CD25) on these cells parallels the expression of total CD122, indicating the availability of both the α and the β chains for signal transduction (Figure 2D).

Figure 2. T cells from Il2rb Mut mice display reduced levels of IL-2Rβ (CD122) protein, dysregulated autocrine IL-2 production, and markers of chronic activation.

Genotypes shown include WT (black) and Il2rb Mut (red). (A)–(H) each represent 2 individual experiments.

(A) Total CD122 expression depicted as median fluorescence intensity (MFI) in subpopulations of splenic CD4⁺ and CD8⁺ T from both genotypes; naive (circles) and TEM cells (diamonds), n = 12 for WT, n = 14 for Il2rb Mut; isotype control antibody MFI is indicated by the gray dotted line.

(B) CD122 cell-surface expression (purple) measured by ImageStream. Membrane mask was applied to the analysis of cells stained with CD122 clone 5H4 (cell surface + intracellular). Images compare CD4⁺ (green) to CD8⁺ (yellow) splenocytes from WT (top) and Il2rb Mut (bottom).

(C) ImageStream quantification, with membrane intensity/area.

(D) CD25 expression depicted as in (A), n = 8 for WT and n = 9 for Il2rb Mut.

(E) Frequency of IL-2-, IFN-γ-, granzyme B-, and perforin-producing CD4⁺ and CD8⁺ splenic T cells, n = 7 per genotype.

(F and G) Activation and exhaustion markers PD-1 (F) and KLRG1 (G) measured in CD4⁺ and CD8⁺ splenocytes from both genotypes; naive (circles) and TEM cells (diamonds), n = 12 for WT and n = 14 for Il2rb Mut.

(H) Representative dot plot distinguishing BM subsets, specifically LSK (Sca-1⁺c-Kit⁺) cells (top right quad) and myeloid progenitors (MPs; top left quad). Graphs show percentage of LSK cells (left) and MP cells (right), n = 6 per genotype.

Data are shown as mean ± SEM. All p values were calculated by non-parametric Mann-Whitney test for a single factor. ImageStream statistics calculated by unpaired parametric Welch’s correction are displayed as ****p ≤ 0.0001, ***p ≤ 0.001, **p ≤ 0.01, and *p ≤ 0.05 or ns.

The regulation of CD122 production and cell-surface expression in T cells, particularly Treg and CD8⁺ TEM cells, is dictated by the availability of circulating ligand, with excess ligand leading to receptor downregulation.^13,38 Likewise, production of IL-2 ligand is regulated by the ability of T cells to induce IL-2 signal transduction, which, in turn, is dependent on cell-surface expression of CD122, spawning an autocrine-negative feedback loop.⁴ Consistent with this concept, splenic T cells from Il2rb Mut mice express reduced levels of CD122 at the cell surface (Figure 2B) yet produce increased amounts of IL-2 as compared to WT animals (Figure 2E, left), suggesting that decreased CD122 surface expression further limits the already deficient IL-2R signaling (receptor intrinsic). This limited amount of signaling is enough to drive autocrine production of IL-2 but not to induce homeostasis, resulting in an excess of circulating IL-2 (receptor extrinsic). Finally, IL-2-producing T cells express higher levels of CD122 compared to non-IL2-producing T cells in both Il2rb Mut and WT mice, indicating that residual signaling in mutant cells maintains this aspect of IL-2 regulation despite failing to establish proper immune homeostasis (Figure S4).

Not surprisingly, the pro-inflammatory cytokine environment found in Il2rb Mut mice leads to persistent splenic T cell activation and differentiation as evidenced by the excess production of IFN-γ, cytotoxic granules with granzyme B and perforin, and IL-2 (Figure 2E), consistent with the elevated frequency of effector/terminally differentiated T cells (Figure 1D). Increased expression of activation/exhaustion markers PD1 and KLRG1 (Figures 2F and 2G), for both naive and TEM cells, further substantiates the conclusion that T cells from mutant animals exist in a chronically activated state. Notably, the thymus in mutant animals shows normal population dynamics, lacking effector T cells with activation/exhaustion markers (Figure S5). This distinction between central (thymus) and peripheral (spleen) T cell profiles suggests that the observed peripheral skewing toward effector activated/exhausted T cell phenotypes is likely secondary to the circulating cytokine milieu (receptor extrinsic), rather than the cell intrinsic mutant IL-2Rβ receptor subunit. Finally, congruent with prior observations,^39,40 the peripheral pro-inflammatory cytokine environment in the Il2rb Mut mice negatively impacts the distribution of hematopoietic progenitor stem cells (HPSCs), with the Lin⁻Sca-1⁺c-Kit⁺ (LSK) population significantly overrepresented in bone marrow (BM), while myeloid progenitors (MPs) are decreased (Figure 2H). It is well established that in a pro-inflammatory cytokine milieu, chronic infections and autoimmune diseases can influence HPSC homeostasis,^39,40 implying that the pro-inflammatory BM environment in Il2rb Mut mice may drive the expansion of LSK cells at the expense of the MP pool.

Il2rb Mut mice mirror human pathology with persistent systemic inflammation resulting from poor expression of mutant CD122 (receptor intrinsic) amid excess environmental IL-2 (receptor extrinsic). This chronic ligand exposure alters IL-2R cell-surface expression and signaling, preventing direct WT-to-mutant comparisons due to drastically different ligand environments (Figure 1B). Further, differences in circulating cytokines also drive divergent cell differentiation (Figures 1 and 2), additionally complicating functional analyses. To accurately assess the effects of the IL2RB p.Pro222_Gln225del mutation, we must distinguish receptor-intrinsic from receptor-extrinsic defects by comparing the development/proliferation of WT and mutant cells in the same (host) environment.

Healthy IL-2/15c circulating cytokine levels in Mut:WT-mixed BMCs partially restore mutant conventional T cell differentiation and activation profiles but not Tregs

To differentiate receptor-intrinsic from extrinsic defects in Il2rb Mut mice, we generated mixed BM chimeras (BMCs) by combining congenically distinct BM cells (97% WT or mutant with 3% congenic WT) and transferring them into WT hosts (WT:WT and Mut:WT; see STAR Methods). Twelve weeks post-reconstitution, Mut:WT recipients appeared healthy, with size and weight comparable to those of WT:WT controls (Figure S6A). Flow cytometric analyses revealed that both congenic-WT donor (3%) and residual WT host splenic T cells expanded in Mut:WT-mixed BMCs (Figures S6B and S6D), demonstrating their competitive advantage over mutant cells specifically in the T cell compartment. This advantage was evident across all CD4⁺ and CD8⁺ T cell subsets, but most pronounced in CD8 TEM and Treg populations (Figures S6C and S6D), highlighting the critical importance of IL-2Rβ-mediated signaling. In contrast, B cells, where IL-2Rβ plays a minimal role, showed similar donor/host distributions in both mixed BMC types (Figure S6E). These data demonstrate that IL-2Rβ is necessary for normal T cell, but not B cell, development and that WT T cells outcompete mutant T cells in a non-inflammatory cytokine environment.

Serum IL-2 and IL-15c (Figure 3A) and other pro-inflammatory cytokines (IL-6, TNF-α, and IFN-γ) that are elevated in Il2rb Mut mice (Figure S7A) are normalized in the Mut:WT-mixed BMCs. Despite this normalization, mutant CD122 still fails to upregulate to the same degree as WT CD122 on TEM cells, though its expression is improved compared to mutant CD122 on T cells in Il2rb Mut mice (Figures 2A and 3B). Consistent with our previous analyses, imaging flow cytometry revealed similar levels of CD122 cell-surface expression between WT (opened black circles) and mutant (opened red circles) cells harvested from the same BMC hosts (Figures 3D and 3E). Expression of CD25 largely parallels CD122 expression in chimeric animals (Figure 3C), indicating, as observed in Figures 2A–2C, the availability of both α and β chains for signaling in both mutant and WT T cells.

Figure 3. A healthy WT-equivalent environment partially restores mutant T cell CD122 cell-surface expression, differentiation, function, and activation state.

Mixed (WT:WT or Mut:WT at 97:3, into WT) bone marrow chimera (BMC) mouse cells and sera were harvested and analyzed 12–14 weeks post-transplant. Black open shapes indicate WT cells from Mut:WT BMCs (WT/Mut:WT). Red open shapes indicate Il2rb Mut cells from Mut:WT BMCs (Mut/Mut:WT). Data depict one representative experiment in each of (A)–(I).

(A) Serum cytokine levels of IL-2, n = 3 for WT:WT and n = 4 for Mut:WT, and of IL-15c, n = 2 for WT:WT and n = 4 for Mut:WT, measured as in Figure 1.

(B) Total CD122 expression depicted as median fluorescence intensity (MFI) in subpopulations of splenic CD4⁺ and CD8⁺ T from both genotypes; naive (circles) and TEM cells (diamonds), n = 6 per genotype; isotype control antibody MFI is indicated by the gray dotted line.

(C) CD25 expression depicted as in (B), n = 6 per genotype.

(D) CD122 cell-surface expression (purple) measured by ImageStream. Membrane mask was applied to the analysis of cells stained with CD122 clone 5H4 (cell surface + intracellular). Images compare CD4⁺ (green) to CD8⁺ (yellow) splenocytes from WT (CD45.1) and Il2rb Mut (CD45.2) chimeric mice.

(E) ImageStream quantification, with membrane intensity/area, for WT (black) and Il2rb Mut (red).

(F) Frequency of IL-2-, IFN-γ-, granzyme B-, and perforin-producing CD4⁺ and CD8⁺ splenic T cells, n = 6 per genotype.

(G and H) Activation and exhaustion markers PD-1 (G) and KLRG1 (H) measured in CD4⁺ and CD8⁺ splenocytes from both genotypes; naive (opened circles) and TEM cells (opened diamonds), n = 6 per genotype.

(I) T cell distribution, n = 7 per genotype, over three experiments. Comparison of TEM cell numbers between genotypes, n = 4 per genotype (right).

Data show one representative experiment. Data are shown as mean ± SEM. All p values were calculated by non-parametric Mann-Whitney test for a single factor. ImageStream statistics calculated by unpaired parametric Welch’s correction are displayed as ****p ≤ 0.0001, ***p ≤ 0.001, **p ≤ 0.01, and *p ≤ 0.05 or ns.

Unlike intact Il2rb Mut mice, mutant splenic T cells from mixed BMCs produce significantly less IL-2, IFN-γ, granzyme B, and perforin (Figure 3F); exhibit reduced expression of activation/exhaustion markers, PD-1 and KLRG1 (Figures 3G and 3H); and display near-normal splenic T cell subset distribution (Figure 3I). The increase in the proportion of mutant naive T cells, relative to TEM cells, likely explains the decrease in pro-inflammatory cytokines (Figure 3F). In both Mut:WT- and WT:WT-mixed BMCs, CD4⁺ and CD8⁺ naive T cells originate mostly from the 97% donor BM cells (Mut or WT, respectively), while TEM cells in Mut: WT-mixed BMCs originate mostly from the 3% congenic-WT donor and residual WT host BM cells (Figures S6B and S6D). These findings demonstrate a key environmental effect—among mutant T cells that develop in an environment with excess ligand (intact Il2rb Mut mice), terminally differentiated, functionally activated cells predominate (Figures 1D and 2E–2G). In contrast, mutant T cells that develop in a WT environment (mixed BMCs) predominantly remain naive, neither displaying activation markers nor producing cytokines (Figures 3F–3H, S6B, and S6D). Additionally, in Mut:WT-mixed BMCs, WT Tregs (either congenic or host derived) constitute the majority of total peripheral Tregs, even though only 19.31% of the splenic CD4⁺ T cells were derived from congenic WT donor cells (Figures S6B and S6C). These data align with previous studies demonstrating that differentiation of Treg and CD8⁺ TEM cells, but not conventional CD4⁺ T cells, requires fully functional IL-2Rβ signaling. In our Il2rb Mut model, suboptimal signal transduction via the hypomorphic mutant receptor appears to be sufficient for CD8⁺ T, but not Treg, differentiation. Further, in an environment of excess ligand, the mutant receptor is constitutively active and drives development of TEM cells and production of pro-inflammatory cytokines, while in the absence of such constant stimuli, the mutant receptor still contributes to CD8⁺ TEM differentiation but to a lesser extent and with diminished activity/cytokine production. Similarly, the BM environment in Mut:WT-mixed BMCs expresses WT-like cytokine levels and near-normal ratios of LSK cells to MP cells (Figure S7B).

Taken together, these data indicate that the immunological abnormalities exhibited by Il2rb Mut mice, including dysregulation of serum cytokines, BM development, and peripheral T cell subset differentiation/activation, are largely normalized in the presence of a WT BM niche (host) and WT lymphocytes (3% WT donor cells). This parallels the post-HSCT immunologic recovery observed in human patients,¹ indicating that dysregulated cytokine levels represent a critical T-cell-extrinsic factor driving T lymphocyte dysfunction.

Transfer of WT Tregs to neonatal ll2rb Mut mice restores IL-2/15c-dependent STAT5 signaling in conventional T cells but not Tregs

The Mut:WT-mixed BMCs enable direct comparison between cells expressing either mutant or congenic-WT receptors as they develop and differentiate within the same WT (host) environment. In this experimental system, the WT-receptor-expressing congenic donor cells and the WT (host) BM niche provide both a cytokine sink (WT receptor)—absorbing excess IL-2/15c produced by the mutant cells—and a regulatory function via WT Tregs. This combination establishes a healthy cytokine milieu in which mutant cells acquire a WT-like profile. Nonetheless, it remains unclear whether the developmental influence of the WT BM environment or the homeostatic effects of WT receptor expression/regulatory function or both drive mutant T cell phenotypic normalization. To further dissect the contribution of these factors, we transferred WT Foxp3⁻GFP⁺ Treg cells into mildly irradiated Il2rb Mut or WT mice at neonatal age (see STAR Methods). This approach provides T cells with a WT receptor and regulatory function while maintaining mutant cells that developed within a mutant niche. Thirteen weeks post-transfer, the number of GFP⁺ Tregs in the spleen averaged 10% in Mut as compared to 1% in WT (control) hosts (Figure S8A). These data again demonstrate the preferential expansion/persistence, i.e., competitive advantage, of WT Tregs vs. Mut Tregs in Mut, but not WT, hosts. Like the Mut:WT-mixed BMCs, Il2rb Mut mice that received Tregs at neonatal age (WT Treg-Il2rb Mut) exhibited healthy IL-2/15c cytokine levels (Figure 4A), body weight (Figure 4B), and spleen size (Figure 4C) and lacked autoimmune pathology, e.g., persistent diarrhea/gastrointestinal (GI) disease, allowing them to reach an age beyond 10 weeks (Figure 4B).

Figure 4. Transfer of WT Tregs into neonatal Il2rb Mut mice restores WT-like cytokine levels, restores conventional T cell phenotypes, and curtails autoimmunity.

Analyses of splenocytes and serum from mice age 13 weeks. Genotypes include WT, n = 6 (black), and Il2rb Mut, n = 2 (red). Transferred Tregs are defined as CD4⁺CD25⁺GFP⁺.

(A) Serum cytokine levels of IL-2 and IL-15c (IL-15/15Rα) were measured by ELISA.

(B) Body weight of neonatal WT or Il2rb Mut mice with (right) and without (left) WT Treg transfer.

(C) Spleen weight.

(D) Representative dot plots depicting T cell subsets, CD4⁺ (left) and CD8⁺ (right). Naive, lower right quad; central memory (TCM), upper right quad; and effector memory (TEM), upper left quad, for both genotypes.

(E) Average T cell distribution for all populations and genotypes.

(F) Total CD122 expression depicted as median fluorescence intensity (MFI) in subsets of naive (circles) and TEM (diamonds) cells; isotype control antibody MFI is indicated by the gray dotted line.

(G and H) Activation and exhaustion markers PD-1 (G) and KLRG1 (H) measured in naive (circles) and TEM (diamonds) cells.

Data are shown as mean ± SEM. All p values were calculated by non-parametric Mann-Whitney test for a single factor, displayed as ****p ≤ 0.0001, ***p ≤ 0.001, **p ≤ 0.01, and *p ≤ 0.05 or ns.

Mutant T cells in WT Treg-Il2rb Mut animals approached WT-like conventional T cell profiles in terms of differentiation (Figures 4D and 4E), CD122 expression (Figure 4F), and activation status (Figures 4G and 4H). In contrast, these parameters were only partially restored in Mut:WT-mixed BMCs (Figures 3F–3I). Additionally, we observed partial restoration of a WT-like phenotype in mutant NK cells in WT Treg-Il2rb Mut animals (Figure S2C), demonstrating that NK cell abnormalities are likewise a consequence of defective IL-2Rβ signal transduction. Crucially, this “WT-like” normalization of the cytokine environment, T cell phenotype, and prolonged life span occurred only when WT Treg transfer was performed at neonatal age, prior to manifestation of disease—functioning essentially as a preventive intervention. WT Treg transfer failed to rescue the diseased phenotype (Figure S8B) when conducted at 5–6 weeks of age, a point when circulating IL-2 and IL-15c levels are already elevated and autoimmunity had already manifested with diarrhea and weight loss (Figures S8B and S8C). In this instance, transferred WT GFP⁺ Tregs were barely detectible in the mutant host 5–6 weeks post-transfer (Figure S8D), suggesting that the pro-inflammatory cytokine environment/abnormally large number of TEM cells impaired survival/proliferation of the transferred cells (Figures S8B–S8D). Conventional T cells in WT Treg-Il2rb Mut animals showed phenotypes closer to that of WT than those of Mut:WT BMCs, despite both having normalized cytokine environments. This difference implies enhanced restoration of mutant receptor expression and signal transduction in WT Treg-Il2rb Mut animals.

To evaluate this idea, we assessed IL-2Rβ signal transduction by measuring STAT5 phosphorylation (pSTAT5) in response to increasing concentrations of exogenous IL-2 and IL-15c (see STAR Methods). As a control for specificity and technical performance, we also measured pSTAT5 in response to IL-7, as IL-2Rβ is not required for this signal transduction pathway. We determined pSTAT5 levels in our three experimental systems: (1) intact WT vs. Mut mice, representing human healthy controls and patients (Figures 5 and 6A); (2) Mut:WT-mixed BMCs, where congenic-WT and mutant cells develop within the same healthy WT host complete with normal cytokine environment and T cells (Figures 5 and 6B); and (3) WT Treg-Il2rb WT vs. WT Treg-Il2rb Mut, where T cells develop in either a WT or a mutant host but are influenced by the presence of WT Tregs with normal receptor expression and regulatory function from the neonatal stage onward (Figures 5 and 6C).

Figure 5. Transfer of WT Tregs into neonatal Il2rb Mut mice significantly improves IL-2Rβ-dependent signal transduction in mutant CD8⁺ T cells but transfer into mixed BMCs does not.

Splenocytes were stimulated with IL-7 and increasing doses of IL-2 (top) and IL-15 (bottom). Genotypes shown include WT (black) and Il2rb Mut (red). Transferred Tregs are marked by CD4⁺CD25⁺GFP⁺. Bar graphs (left) quantify pSTAT5 response as a percentage of CD8⁺ T naive and TEM cells, based on the threshold set on unstimulated condition and isotype control for pSTAT5 antibody MFI per representative animal per genotype.

(A) Intact animals, n = 3 per genotype.

(B) Mut:WT-mixed BMCs, n = 6, with congenic WT (black) and Il2rb Mut (red) donors.

(C) Foxp3⁻GFP⁺ Tregs transferred into neonatal WT (n = 6) and Il2rb Mut (n = 2). The bivariate (right), representing each subpopulation from both genotypes across three experiments, shows unstimulated vs. maximum IL-2 and IL-15c concentrations.

See STAR Methods for p value calculations. The p values are indicated by letters (a)–(h) for each comparison performed per cytokine.

Figure 6. Transfer of WT Tregs into neonatal Il2rb Mut mice significantly improves IL-2Rβ-dependent signal transduction in mutant CD4⁺ T cells but transfer into mixed BMCs does not.

Splenocytes were stimulated with IL-7 and increasing doses of IL-2 (top) and IL-15 (bottom). Genotypes shown include WT (black) and Il2rb Mut (red). Transferred Tregs are marked by CD4⁺ CD25⁺GFP⁺. Bar graphs (left) quantify pSTAT5 response as a percentage of CD8⁺ T naive and TEM cells, based on the threshold set on unstimulated condition and isotype control for pSTAT5 antibody MFI per representative animal per genotype.

(A) Intact animals, n = 3 per genotype.

(B) Mut:WT-mixed BMCs, n = 6, with congenic WT (black) and Il2rb Mut (red) donors.

(C) Foxp3⁻GFP⁺ Tregs transferred into neonatal WT (n = 6) and Il2rb Mut (n = 2).

See STAR Methods for p value calculations. The p values are indicated by letters (a)–(h) for each comparison performed per cytokine.

As shown in Figures 5 and 6, WT T cells from intact WT animals show a graded pSTAT5 response to increasing concentrations of IL-2 and IL-15c, while mutant cells from Il2rb Mut animals lack an apparent pSTAT5 shift in response to IL-2 or IL-15c (Figures 5A and 6A). In WT CD8⁺ and CD4⁺ T cells, where CD122 expression is upregulated in the transition from naive to TEM phenotype (Figure 2A), the pSTAT5 dose response to IL-15c is significantly different between the two (p = 0.0496 [e] for Figure 5A; p = 0.00035 [e] for Figure 6A). In contrast, mutant T cells show no difference in response to IL-2 or IL-15c (p values [b] and [f]; Figures 5A and 6A), due to decreased CD122 expression and lack of transitional upregulation (Figure 2A). As expected, pSTAT5 dose responses comparing WT to Mut, for both naive and TEM T cells, are also statistically different (p values [c] and [d] for IL-2 and [g] and [h] for IL-15c; Figures 5A and 6A). WT and mutant T cells respond equivalently to IL-7 for all cell populations measured and in all experimental setups (p > 0.05; Figures 5 and 6), indicating that the signaling impairment is specific to IL-2R subunit β. In Mut:WT-mixed BMCs, congenic WT vs. Mut T cells maintain significant differences in pSTAT5 response across all subsets (p values [c] and[d] for IL-2 and [g] and [h] for IL-15c; Figures 5B and 6B). However, mutant CD4 T⁺ cells show significant changes in pSTAT5 response comparing naive to TEM (p values [b] and [f]; Figure 6B), reflecting partial restoration of CD122 upregulation (Figure 3). As expected, congenic WT T cells demonstrate significant differences in pSTAT5 response to IL-2 and IL-15c in naive to TEM comparisons (p values [a] and [e]; Figures 5B and 6B). In neonatal WT Treg-Il2rb Mut animals, mutant CD4⁺ and CD8⁺ TEM cells approach WT-like pSTAT5 levels across all IL-2 and IL-15c doses, with p values that are not statistically significant (p values [c] and [d] for IL-2 and [g] and[h] for IL-15c; Figures 5C and 6C). Furthermore, naive vs. TEM pSTAT5 responses for either cytokine show statistically significant p values (p values [b] and [f]; Figures 5C and 6C), consistent with CD122 upregulation from naive to TEM that enhances pSTAT5 signaling (Figures 4F, 5C, and 6C). Last, comparing the naive vs. TEM pSTAT5 differential response in WT vs. Mut cells ([a] vs. [b] for IL-2 and [e] vs. [f] for IL-15c across the experimental systems [A]–[C] in Figures 5 and 6) shows progressively decreasing statistical significance—intact animals > mixed BMCs > neonatal Treg transfer—indicating gradual restoration of CD122 expression and function with each intervention.

Similarly, in the transplanted IL2RB p.Pro222_Gln225 deletion patient, autoimmunity has remained controlled for 5 years post-HSCT despite waning donor chimerism and declining numbers of healthy (WT) Treg cells. This control is likely maintained by healthy peripheral IL-2 levels. Donor chimerism has decreased from 100% at transplant to the current 55% in CD3⁺ T cells (55% in Treg and CD8⁺ T), 35% in CD33⁺ cells, and 94% in NK cells, yet circulating IL-2 has consistently remained within reference range. The patient is currently off all immunosuppression, growing and developing well, though he received mycophenolate mofetil and sirolimus for graft-vs.-host prophylaxis until 1 year ago. Details regarding donor and induction regimens can be found in Fernandez et al., where we show that 1 year post-HSCT, CD122 expression and pSTAT5 responses to IL-2 and IL-15c were comparable between the patient and a healthy donor.¹

DISCUSSION

IEIs are rare genetic disorders that variously affect the immune system, often resulting in infection susceptibility and immune dysregulation, e.g., autoimmunity, autoinflammation, atopy, premature immune senescence, lymphoproliferation, and malignancy. Given the young age and high morbidity/mortality of these patients, dissection of the mechanistic underpinnings of novel IEIs is challenging at best. These challenges are compounded by the immunomodulatory therapies used to manage the patients’ immune dysregulation complications. Such therapies alter immune cell phenotypes and circulating cytokine levels and composition, preventing studies of the “intact” disease process. Current animal models fail to fully represent the genetic, molecular, and clinical phenotypes associated with IEI disorders. Thus, rather than addressing the underlying mechanisms driving immune dysregulation, treatment choices remain focused on symptom management. True understanding of IEIs and the ability to provide informed, targeted treatment requires rigorous study of molecular and cellular pathways. Only in this way will we reveal candidate targets for personalized therapy and advance our knowledge of human immunology.

Overall, we demonstrate that hypomorphic Il2rb Mut mice recapitulate human clinical and immunological phenotypes (Figures 1 and 2), serving as a useful tool for modeling human disease and investigating the immunological consequences of dysregulated, but not absent, IL-2/15-dependent signaling. Knockout mouse models of cytokine signaling have illuminated basic molecular mechanisms⁴¹ but inadequately represent common human hypomorphic defects⁴² and the tolerogenic-immunogenic imbalance pervasive in partial defects. The il2rb^−/− mouse model established IL-2Rβ’s role in T cell programming and control of autoimmunity, clarifying how IL-2R/STAT5 signaling influences Treg biology.^3,21–25 However, Il2rb^−/− mice have limitations for studying immune dysregulation in human disease. The complete knockout, unlike our hypomorphic model, fails to recapitulate the partially functional receptor dynamics seen in patients, where altered ligand/receptor ratios critically affect downstream signaling. Complete IL2RB deficiency causes human fetal demise,⁴³ further highlighting the need for models that better reflect clinical realities.

As in human patients, the excess of IL-2 and IL-15c circulating in the peripheral blood of Il2rb Mut mice (Figure 1B) creates a cell-extrinsic pro-inflammatory environment, which likely drives the observed T cell lymphoproliferation, activation/exhaustion, terminal differentiation, and infiltration into the organs with consequent inflammatory complications, e.g., intestinal, splenomegaly, and hematological (Figures 1A and S1). We hypothesize that this excess ligand constantly engages the mutant hypofunctional IL-2R trimer, leading to persistent impaired IL-2/15 signaling. Further, this cell-extrinsic pro-inflammatory drive is exacerbated by abnormally low numbers of Treg cells (Figure 1C). Without Tregs, conventional T cell proliferation and cytokine production cannot be adequately regulated. The impact of cell-extrinsic (excess ligand/pro-inflammatory cytokine milieu) vs. cell-intrinsic (inadequate receptor expression/function) factors on T cell phenotype/function and clinical manifestations is unclear. Ethical restraints prevent mechanistic studies of partial IL-2Rβ defects using human primary cells, while in vitro cell lines are impractical, as they lack one or more necessary IL-2R subunits, cytokine receptors, and co-stimulatory receptors. Further, cell lines cannot replicate organism-level systems or the effects of T cell organ infiltration. Our unique mouse model enables exploration of how hypomorphic IL2RB asymmetrically impacts Treg and CD8⁺ T cell differentiation and function—processes critical for the development of inflammation/autoimmunity.

Il2rb Mut mice exhibit the following features of immune dysregulation: (1) symptomatic sequelae of persistent inflammation (poor growth and splenomegaly, Figure 1A), (2) increased serum IL-2/15c (Figure 1B), (3) dysregulated T cell IL-2 and cytotoxic granule production (Figure 2D), (4) decreased Tregs (Figures 1C and 1E), and (4) skewed CD8⁺ T cell memory/activation (Figures 1D, 1E, and 3). Unlike Il2rb-null mice, the expression of mutant CD122 protein in the Il2rb Mut mouse allows development of T cells, especially those populations known to depend on IL-2/15 signaling (Figures 1C–1E). Treg numbers remain significantly depleted in mixed BMCs, implying a receptor-intrinsic defect. In contrast, a WT-like cytokine environment provided via mixed BMCs or neonatally transferred WT Tregs significantly improves both phenotype and function of CD4⁺ and CD8⁺ T cells that bear mutant IL-2Rβ (Figures 2, 3, and 4). Our experiments using the Il2rb Mut model reveal that specific IL-2/IL-15-dependent signaling thresholds differentially impact T cell homeostasis and susceptibility to autoimmunity/immune dysregulation, with the threshold needed for Treg development substantially higher than that required for CD8⁺ T cell effector memory and NK cell development/maturation (Figures 3, 4, 5, and 6). Using mixed BMCs, we dissected receptor-intrinsic vs. receptor-extrinsic mechanisms driving immunopathology, showing how a hyperinflammatory cytokine environment directly impacts both mutant and WT cell phenotypes and functions (Figure 3). We also deciphered the molecular basis for how hypomorphic (rather than absent) IL-2R signaling leads to a paradoxical accumulation of IL-2/IL-15 in circulation, creating a feedforward inflammatory loop that amplifies pathology beyond the primary receptor defect (Figures 1, 2, and 3).

Our mouse model and post-HSCT human data indicate that the hypomorphic receptor primarily impairs Treg development, contributing significantly to disease pathology (Figures 3 and 4). These findings support a model where reduced (but not absent) CD122 receptor expression and signal transduction fails to adequately clear IL-2 and IL-15 from circulation. These circulating cytokines accumulate in the periphery to levels sufficient to chronically engage the mutant receptor. In the absence of functional Tregs, chronic excess ligand promotes production of other pro-inflammatory cytokines, leading to abnormal T cell differentiation/activation. In both our mouse model and human patient data, functional WT Tregs, whether in mixed BMCs, infused into neonatal mutant animals, or transplanted into human patients following induction therapy, effectively reduce circulating pro-inflammatory cytokines, creating an environment that permits normal differentiation of WT/healthy cells.

Our pSTAT5 dose-response assessments demonstrate the complex relationship between ligand/receptor ratios and signal transduction. These requirements vary significantly across different T cell subsets, influencing both differentiation and activation profiles. As expected, dose-response data from intact animals clearly demonstrate a hierarchical pSTAT5 response in WT cells (naive T < TEM CD4 and CD8 T), contrasting with the overall poor response to both IL-2 and IL-15c in mutant animals (Figures 5A and 6A). This pattern is consistent with the hierarchical expression of CD122 in WT T cell subsets, particularly in the transition from naive T to TEM. In mutant animals, despite overall reduced CD122 expression and lack of upregulation during the naive-to-TEM transition, the hierarchical expression pattern of CD122 persists, with NK > CD8⁺ TEM > Treg > conventional CD4⁺ T cells. Our ability to evaluate the functionality of the mutant receptor, however, is complicated by the pathogenic IL-2/15 cytokine environment in which the mutant cells exist. In mixed BMCs, mutant receptors experience a WT-like cytokine environment yet still perform suboptimally compared to WT cells within the same animal (Figures 5B and 6B). Nonetheless, the pSTAT5 response differential between WT and mutant cells decreases significantly when comparing intact animals to mixed BMCs, suggesting that the WT-like cytokine environment partially improves mutant receptor function (Figures 5A, 5B, 6A, and 6B). Importantly, mixed BMCs were generated using HSCs from 7-week-old mutant animals transplanted into irradiated WT hosts. HSCs from these mutant animals display abnormal population dynamics and phenotypic marker expression, likely due to the pathogenic, pro-inflammatory cytokine BM environment (Figure 2H), which hampers normal differentiation and function even after transfer to a WT environment (Figure 3). Neonatal WT Treg transfer provides more comprehensive correction than mixed BMCs. In these animals, pSTAT5 dose responses in CD8⁺ TEM, but not other cell types, are similar to those of WT animals, with statistical analyses confirming this selective improvement (Figures 5C and 6C). WT Treg-Il2rb Mut animals also exhibit near-normal levels of circulating IL-2 and IL-15, WT-like T cell phenotypes, and normal spleen size, gastrointestinal function, and growth (Figure 4). Comparing pSTAT5 responses between naive and TEM cells shows the smallest differential in neonatally treated animals, consistent with improved CD122 upregulation and signal transduction in this experimental setting (Figure 4F). Together, these findings suggest that neonatal WT Treg administration serves two crucial functions: (1) providing a “sink” for excess IL-2/15 produced by mutant cells and (2) curtailing autoimmunity by restoring regulatory function. While neonatal Treg transfer rescues phenotypes in both Il2rb-knockout and Il2rb Mut mice, IL-2Rβ signal transduction returns only in Mut animals, clearly distinguishing our hypomorphic receptor model from the complete knockout and demonstrating the importance of both receptor-intrinsic and extrinsic factors in immune dysregulation.

These data collectively demonstrate the critical importance of both WT receptor expression and regulatory function, features present in both the mixed BMC and the neonatal Treg transfer experimental systems. However, notable differences exist—in the mixed BMCs, a WT BM niche is provided along with fully functional CD122 at early developmental stages, whereas in the WT Treg transfer into neonatal mutant animals, immunomodulation provides only the cells. While this distinction between the systems reveals insights regarding the requirements for CD122-driven regulatory function and its effect on the cytokine milieu, it does not determine whether CD122 expression alone is sufficient to curtail the pathogenic cytokine environment and consequent autoimmunity. To dissect expression from regulatory function, we transferred WT CD8⁺ T cells into Il2rb Mut mice at 4–7 weeks of age—when pathogenic serum cytokine levels and T cell phenotypes are fully established—and observed several key outcomes: (1) extended animal survival to ~13 weeks of age vs. the typical 10 weeks (data not shown), (2) normalization of circulating cytokines to WT-like levels (Figure S9A), and (3) improved overall health, even though T cell subset distribution/activation profiles do not reflect WT-like patterns (Figures S9B and S9C). These findings reveal that the presence of a WT receptor without regulatory function (WT CD8 T cells) can still act as a “sink” for excess cytokines, normalizing the cytokine environment and supporting animal survival/growth, even though the intervention occurred after the onset of autoimmunity/immune dysregulation. In contrast, WT Treg transfer post-disease-onset failed to alleviate pathological symptoms or correct T cell abnormalities (Figure 5), even though this treatment was effective when administered at neonatal age. Thus, receptor-extrinsic factors may contribute more significantly to autoimmunity/immune dysregulation than previously appreciated, highlighting the fundamental importance of cell-type-specific ligand/receptor ratios. In the high-IL-2/15c cytokine environment of 4-week-old Il2rb Mut mice, transferred WT Tregs fail to survive long enough to curtail excess ligand, whereas transferred WT CD8⁺ TEM cells both survive and successfully normalize the cytokine environment. This demonstrates that, although Treg deficiency initiates the process, pathology ultimately stems from a self-reinforcing cycle where intrinsic receptor dysfunction and the resulting dysregulated cytokine environment amplify each other.

Our findings carry significant therapeutic implications. The clinical consequences of cell-intrinsic defects can be ameliorated only by HSCT of WT donor cells, while the clinical consequences of a hyperinflammatory cytokine environment can be treated with immunomodulation, which carries significantly less morbidity/mortality than HSCT. Additionally, evidence of BM-niche inflammation (Figure 2) also implies the need for aggressive immunosuppression prior to HSCT to allow for engraftment of donor HSCs—a procedure with even higher morbidity/mortality. In IEIs with immune dysregulation, genetic mutations that disrupt regulatory cytokines, their receptors, or signaling intermediaries lead to secondary accumulation of inflammatory cytokines that often contribute to disease severity. For example, in IL-2Rα-deficient patients, peripheral Th1, Th2, and Th17 cell numbers all exceeded age-matched controls despite pharmacological immunosuppression,⁴⁴ with a pathogenic cytokine environment further driving T cell activation/proliferation and perpetuating inflammatory complications.⁴⁵ Similarly, STAT3 gain-of-function mutants significantly increase circulating IL-6 levels,⁴⁶ just as our published Il2rb hypomorphic mutant elevates IL-2/15c.¹

Treatment options in IEIs with immune dysregulation often require aggressive immunomodulation balanced against infection risk. A recent study showed that although HSCT resolved symptoms in 55% of IEI patients within 1 year, long-term survival remained unsatisfactory (67% at 5 year).⁴⁷ This poor outcome is closely linked to resurgence of inflammatory complications when donor engraftment is incomplete, creating chimerism where residual recipient cells reinitiate the establishment of a pathogenic cytokine environment. Thus, mechanistic understanding of receptor-intrinsic vs. extrinsic immunopathology necessarily informs strategies for aggressive control of hyperinflammation pre- and post-HSCT to optimize survival outcomes.

Data from this article were used to inform the management of the index patient from Fernandez et al.,¹ who lost donor chimerism soon after HSCT. At 3 years post-HSCT, the patient showed 66% mixed donor chimerism. Standard approach would have been to administer a boost from the same donor to increase WT donor cells. However, because we could demonstrate that the patient maintained a healthy circulating cytokine environment; a healthy immune profile, including donor Tregs; and functional STAT5 signaling, we chose conservative/supportive management instead. Looking ahead, mechanistic studies of IEIs with hypomorphic defects will enhance patient/pathway-specific therapy, illuminate the precise regulation of signal transduction, and expand our understanding of human immunology. This knowledge will accelerate development of immunotherapies for both common autoimmune diseases and malignancies.

Limitations

Our mechanistic study of a hypomorphic IL2RB mutation has the following limitations. First, questions regarding the threshold requirement for NK cell development and function remain and are not easily explored using the Il2rb Mut model system. Second, while our mixed BMC and WT Treg transfer approaches helped dissect the required elements (receptor signaling, regulatory function, and cytokine modulation) driving immune dysregulation, the timing of intervention proved critical. Interventions before disease onset (neonatal Treg transfer) were more effective than those after symptom establishment, suggesting different mechanistic requirements for prevention vs. treatment. Third, our BMC experiments achieved only partial restoration of mutant cell phenotypes, likely because we used HSCs from 7-week-old mutant animals. These HSCs, having developed in a pathogenic pro-inflammatory environment, already displayed abnormalities before transplantation into WT hosts, potentially compromising their ability to differentiate normally even in a healthy environment. Future studies using T-cell-depleted HSCs from mutant BM at different ages (before and after disease onset) would help determine how the accumulating inflammatory cytokine environment in the BM niche progressively damages HSC function and affects immune reconstitution potential. Fourth, although mutant neonatal mice receiving WT Treg cells exhibited normalization approaching WT in terms of cytokine environment, conventional T cell phenotypes, and T cell IL-2/15 signaling, we did not fully characterize whether host Tregs recovered in number, phenotype, and function. Additional studies on this topic, along with WT CD8⁺ T cell transfers into mutant neonates, will provide insights into the quantity of WT receptor expression/function required for mutant Treg development. Last, while our model recapitulates many features of human IL2RB deficiency, the complex interplay between receptor dysfunction and dysregulated cytokine environments in human patients remains challenging to study due to variable disease presentations and treatment regimens.

RESOURCE AVAILABILITY

Lead contact

Requests for further information, resources, and reagents should be directed to and will be fulfilled by the lead contact, Elena W.Y. Hsieh, elena.hsieh@cuanschutz.edu.

Materials availability

This study generated a novel mouse model via CRISPR-Cas, which is available upon request.

Data and code availability

Data regarding cell frequencies, absolute numbers, and geometric median fluorescence intensities (gMFIs) are publicly available (as of publication date) at Mendeley https://data.mendeley.com, accession no. https://doi.org/10.17632/tkwmmd27jh.1.

STAR★METHODS

EXPERIMENTAL MODEL

Mice

All mice were bred and maintained in a pathogen free facility at University of Colorado Anschutz Medical Campus. C57BL/6J (JAX #000664) and CD45.1 B6.SJL-Ptprc^a Pepc^b /BoyJ (JAX#002014) strains were purchased from The Jackson Laboratory, B6.Cg-Foxp3tm1Kuch/J (JAX #035864) strain was generously donated by Rebecca McCullough lab. CD45.1/2 mice were obtained by the cross of 00664 and 002014 strains. All the experiments were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC), protocol number 000914. Best practices were used to minimize mouse pain, distress and suffering.

IL2RB p.Pro222_Gln225del knock-in mice were generated using CRISPR/Cas9 editing technology at the National Jewish Mouse Genetics Core Facility. The guide sequence was 5’-AGCCCCTGACCTTTCGGACAAGG-3’. The HDR sequence used as a template was 5’-TCCCCTTACCCAGCTTCTCCCACTCCCACCCCCAACCTGTGCCACCCACTGTACCTGCTGGCCTAGTCCGAAATGTCAGAGGCTGACTCCAGGTACCGGTATTGTTTCGTTGAGCTTTGACCCTCACCTGGACCTCATATGAGGTACTAGGGATCAGC-3’. Guide design was done using CRISPR and the Broad Institute sgRNA Design Software. Guide activity was verified by incubating guide RNA and Cas9 protein with a PCR product containing the target sequence and comparing the ratio of cut to uncut PCR product. Zygotes were injected with guide RNA(s), Cas9, and DNA template if appropriate. Zygotes were then transferred into pseudopregnant recipients. F0 pups were genotyped by PCR using primers outside the region to be modified followed by restriction enzymedigest to identify putative positive founders. Primers used for genotyping include forward mCD122 F5 5’-CATCTCTGCCCAGGATGCATC-3’, and reverse mCD122R5 5’-CTTGGCAAGTGACATGATGAG-3’. Digestion using KPN1 restriction enzyme released the following PCR products indicating mouse genotype: wild type = 560 bp; heterozygous mutant = 560 bp + 430 bp; homozygous mutant = 430 bp. No influence of gender was established in the conducted experiments.

METHOD DETAILS

Mixed bone marrow chimera (BMC)

Mixed bone marrow chimeras were generated using T and B cell depleted bone marrow from either CD45.2 Il2rb Mut (Mut/Mut:WT) or WT donors (WT/Mut:WT), mixed with CD45.1 or CD45.1/2 bone marrow cells in a ratio of 97:3 into lethally irradiated (2× 500 rads, 35 cm distance) CD45.1 or CD45.1/2 WT recipients. Il2rb mice donor cells were typically harvested at 6–8 weeks of age whereas WT donor and recipient cells at 7–9 weeks of age. Both genders were included subject to availability. Transplant was performed via tail vein intravenous (i.v.) injection with mice kept on antibiotic treatment for two weeks post-transplant. Immune reconstitution of donor cells was tracked via peripheral blood screen at 8- and 10-weeks post-transplant. Mice were euthanized at ~12–14 weeks post-transplant and cells harvested for immunophenotyping, intracellular cytokine staining, or pSTAT5 measurements. Antibodies used for T/B depletion are listed in Table S1 and the key resources table. The Mojo sort Anti-APC Beads (BioLegend cat#480072) were used for the magnetic separation.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
17A2 (APC) [anti-CD3]	BioLegend	Cat# 100236; RRID:AB_2561456
RM4–4 (APC) [anti-CD4]	BioLegend	Cat# 116014; RRID:AB_2563025
53–6.7 (APC) [anti-CD8a]	BioLegend	Cat# 100712; RRID:AB_312751
30-H12 (APC) [anti-90.2]	BioLegend	Cat# 105312; RRID:AB_313183
17A2 (BUV563) [anti-CD3]	Becton Dickinson	Cat# 741319; RRID:AB_2870837
RM4–5 (BV510) [anti-CD4]	Becton Dickinson	Cat# 563106; RRID:AB_2687550
RM4–5 (BV785) [anti-CD4]	BioLegend	Cat# 100551; RRID:AB_11218992
GK1.5 (BV605) [anti-CD4]	BioLegend	Cat# 100451; RRID:AB_2564591
GK1.5 (AF488) [anti-CD4]	BioLegend	Cat# 100423; RRID:AB_389302
53–6.7 (AF700) [anti-CD8a]	Becton Dickinson	Cat# 557959; RRID:AB_396959
53–6.7 (BUV737) [anti-CD8a]	Becton Dickinson	Cat# 612759; RRID:AB_2870090
53–6.7 (BV510) [anti-CD8a]	BioLegend	Cat# 100751; RRID:AB_2561389
53–6.7 (PE-Dazzle 594) [anti-CD8a]	BioLegend	Cat# 100761; RRID:AB_2564026
IM7 (PerCP/Cy5.5) [anti-CD44]	Becton Dickinson	Cat# 560570; RRID:AB_1727486
IM7 (BV510) [anti-CD44]	BioLegend	Cat# 103044; RRID:AB_2650923
MEL-14 (Pe/Cy7) [anti-CD62L]	Invitrogen	Cat# 25-0621-82; RRID:AB_469633
MEL-14 (FITC) [anti-CD62L]	Invitrogen	Cat# 11-0621-82; RRID:AB_465109
MF23 (AF647) [anti-Foxp3]	Becton Dickinson	Cat# 560401; RRID:AB_1645201
J43 (PE/CF594) [anti-PD-1]	Becton Dickinson	Cat# 562523; RRID:AB_2737634
5H4 (BV421) [anti-CD122]	Becton Dickinson	Cat# 564925; RRID:AB_2739010
Tmb-1 (BV421) [anti-CD122]	Becton Dickinson	Cat# 562960; RRID:AB_2737918
A20 (AF488) [anti-CD45.1]	BioLegend	Cat# 110718; RRID:AB_492862
A20 (PE/Cy7) [anti-CD45.1]	BioLegend	Cat# 110730; RRID:AB_1134168
A20 (APC) [anti-CD45.1]	BioLegend	Cat# 110714; RRID:AB_313503
A20 (PE) [anti-CD45.1]	BioLegend	Cat# 110708; RRID:AB_313497
A20 (APC/Cy7) [anti-CD45.1]	BioLegend	Cat# 110716; RRID:AB_313505
104 (BV711) [anti-CD45.2]	BioLegend	Cat# 109847; RRID:AB_2616859
104 (BV605) [anti-CD45.2]	BioLegend	Cat# 109841; RRID:AB_2563485
2F1 (BV786) [anti-KLRG1]	Becton Dickinson	Cat# 565477; RRID:AB_2739256
TUGm2 (PE) [anti-CD132]	BioLegend	Cat# 132305; RRID:AB_2123703
PC61 (APC/Cy7) [anti-CD25]	Becton Dickinson	Cat# 557658; RRID:AB_396773
RA3–6B2 (APC) [anti-CD45R/B220]	BioLegend	Cat# 103212; RRID:AB_312997
PK136 (BUV395) [anti-NK1.1]	Becton Dickinson	Cat# 564144; RRID:AB_2738618
PK136 (BV510) [anti-NK1.1]	Becton Dickinson	Cat# 563096; RRID:AB_2738002
1D4B (AF488) [anti-CD107a]	BioLegend	Cat# 121607; RRID:AB_571982
MP6-XT22 (eFluor450) [anti-TNF-α]	Invitrogen	Cat# 48-7321-82; RRID:AB_1548825
XMG1.2 (PerCP/Cy5.5) [anti-IFN-γ]	Becton Dickinson	Cat# 560660; RRID:AB_1727533
JES6–5H4 (PE) [anti-IL-2]	Invitrogen	Cat# 12-7021-82; RRID:AB_466150
eBio17B7 (APC) [anti-IL-17A]	Invitrogen	Cat# 17-7177-81; RRID:AB_763580
QA16A02 (PE/Dazzle 594) [anti-Granzyme B]	BioLegend	Cat# 372216; RRID:AB_2728383
S16009A (APC) [anti-Perforin]	BioLegend	Cat# 154303; RRID:AB_2721462
47/Stat5 (pY694) (PE) [anti-STAT5]	Becton Dickinson	Cat# 612567; RRID:AB_399858
pY694 (AF647) [anti-STAT5]	Becton Dickinson	Cat# 612599; RRID:AB_399882
Cupk43k (PE/Cy7) [anti-pS6]	Invitrogen	Cat# 25-9007-42; RRID:AB_2637099
Live/dead (UV) [Zombie]	BioLegend	Cat# 423107
Live/dead (Red) [Ghost dye Red 780]	Cytek	Cat# 13-0865
Chemicals, peptides, and recombinant proteins
Phosphate Buffered Saline (PBS)	GIBCO	Cat# 14190-144
Roswell Park Memorial Institute media (RPMI 1640) 1X	CORNING	Cat# 15-040-CV
Fetal Bovine Serum (FBS)	PEAK Serum	Cat# PS-FB3
2mM L-Glutamine, 100 U/mL Penicillin, and 100 mg/mL Streptomycin	GIBCO	Cat# 14190-144
Brefeldin A	Sigma-Aldrich	Cat# B6542
IL-2	Peprotech	Cat# 212-12
IL-15	Peprotech	Cat# 210-15
IL-15Ra	R&D	Cat# 551-MR-100
IL-7	Peprotech	Cat# 217-17
Paraformaldehyde	Electron Microscopy Sciences	Cat# 15710
Methanol	BDH	Cat# 67-56-1
Critical commercial assays
MojoSort Anti-APC beads	BioLegend	Cat# 480072
MojoSort Mouse CD4 T cell Isolation Kit	BioLegend	Cat# 480033
MojoSort Mouse CD8 T cell Isolation Kit	BioLegend	Cat# 480035
CD4+CD25+ Regulatory T Cell Isolation Kit, mouse	Miltenyi Biotec	Cat# 130-091-041
V-PLEX Proinflammatory Panel 1 Mouse Kit	Meso Scale Diagnostics (MSD)	MSD# K15048D
IL-2 ELISA Kit	Invitrogen	Cat# BMS601
IL-15/IL15R ELISA Kit	Invitrogen	Cat# BMS6023
Foxp3/Transcription Factor Buffer Set	eBioscience	Cat# 00-5523-00
Deposited data
Il2rb Mut mice T Cell Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Mixed Bone Marrow Chimera T Cell Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Il2rb Mut mice young cohort T Cell Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
CD8 T cell transfer T Cell Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Treg Transfer T Cell Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Neonatal Treg Transfer T Cell Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Il2rb Mut mice pSTAT5 dose response Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Mixed Bone Marrow Chimera pSTAT5 dose response Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Neonatal Treg Transfer pSTAT5 dose response Analysis	Mendeley Data	https://doi.org/10.17632/tkwmmd27jh.1
Experimental models: Organisms/strains
Mouse: C57BL/6J	The Jackson Laboratory	JAX #000664
Mouse: CD45.1: B6.SJL-Ptprca Pepcb /BoyJ	The Jackson Laboratory	JAX #002014
Mouse: GFP-Foxp3: B6.Cg-Foxp3tm1Kuch/J	Donated by Rebecca McCullough	JAX #035864
Mouse: Il2rb Mut: IL2RB p.Pro222_Gln225del knock-in	National Jewish Mouse Genetics Core	N/A
Oligonucleotides
mCD122 F5 5’-CATCTCTGCCCAGGATGCATC-3’	Integrated DNA Technology	N/A
mCD122R5 5’-CTTGGCAAGTGACATGATGAG-3’	Integrated DNA Technology	N/A
Software and algorithms
CellEngine	CellCarta	https://docs.cellengine.com CellEngine (RRID:SCR_022484)
GraphPad Prism 10	GraphPad Prism	http://www.graphpad.com GraphPad Prism (RRID:SCR_002798)
IDEAS 6.2	Cytek Bioscience
sg RNA Design Software (CRISPR)	Broad Institute	https://www.broadinstitute.org

T regulatory cells adoptive transfer

Splenocytes from Foxp3-GFP mice (B6.Cg-Foxp3tm1Kuch/J JAX #035864) between 9–12 weeks of age, were harvested and enriched for CD4⁺ T cells (BioLegend cat#480033). The enriched fraction was FACS sorted for GFP expression. The highly pure (>95%) fraction of CD4⁺GFP⁺ Tregs was resuspended in Phosphate Buffered Saline (PBS) (GIBCO, cat#14190–144) and injected into Il2rb Mutant mice at 6 wk of age (2.5 × 10⁵ cells/mouse, i.v.). Both genders were included subject to availability. Recipient mice were euthanized at ~4 weeks post-transfer.

Neonatal T regulatory cell adoptive transfer

Splenocytes from Foxp3-GFP mice (B6.Cg-Foxp3tm1Kuch/J JAX #035864) preferably females between 9–12 weeks of age were harvested and enriched for CD4⁺CD25⁺ Regulatory T cells (Miltenyi Biotec cat#130–091-041). The enriched fraction of CD4⁺ CD25⁺GFP⁺ was resuspended in Phosphate Buffered Saline (PBS) (GIBCO, cat#14190–144) and injected into Il2rb Mutant neonates previously sublethal irradiated with one dose of 300 rads within 72 hours after birth. The neonate mice were injected ~1×10⁶ cell split into i.v. (facial vein) and i.h. (intrahepatically), a total volume of 50–70 uL were injected per mice. Recipient mice were euthanized 13 weeks post-transfer.

CD8⁺ T cell adoptive transfer

Splenocytes from CD45.1 mice (B6.SJL-Ptprc^a Pepc^b /BoyJ JAX #002014) were harvested and enriched for CD8⁺T cells (BioLegend cat#480035). WT donor cells were generally harvested at 7–9 weeks of age. Both genders were included depending on availability. The enriched fraction of CD8⁺ T cells was resuspended in Phosphate Buffered Saline (PBS) (GIBCO, cat#14190–144) and injected into Il2rb Mutant mice at 4–7 wk of age (2.5–3×10⁶ cells/mouse, i.v). Recipient mice were euthanized 6 weeks post-transfer.

Flow cytometry

Immunophenotyping and detection of intracellular cytokines

For each mouse, spleen and thymus were mechanically dissociated in Roswell Park Memorial Institute media (RPMI 1640) 1X (CORNING, cat#15–040-CV) supplemented with 10% Fetal Bovine Serum (FBS) (PEAK Serum, cat#PS-FB3), 2mM L-Glutamine, 100 U/mL Penicillin, and 100 mg/mL Streptomycin (Gibco, cat#10378016), red blood cells-lysed, and remaining cells washed in Phosphate Buffered Saline (PBS) (GIBCO, cat#14190–144). Cell samples were then filtered and counted using a Bio-Rad cell counter (TC20 Automated cell counter, cat#1450102). Cells were plated at a concentration of 4×10⁶ per well and incubated for 10 min at room temperature with fixable viability dye and monoclonal antibodies to block Fc receptors (2.4G2, in-house hybridoma), followed by a single wash in Fluorescence Activated Cell Sorting buffer (FACS buffer PBS + 3%FBS). Cells were surface stained with fluorochrome-conjugated antibodies for 30 min in the dark at 4°C, washed in FACS buffer, fixed, and permeabilized (eBioscience^™ Foxp3/Transcription Factor Staining Buffer Set, cat#00–5523-00). Intracellular staining with fluorochrome-conjugated antibodies was conducted overnight at 4°C. When intracellular cytokine and cytotoxic in vivo measurements were performed, mice were treated i.v. (300 uL/mouse) with Brefeldin A (10mg/mL, Sigma-Aldrich cat#B6542) 4 hours prior to euthanasia. All data acquisition was performed on a Cytek Aurora Spectral Flow Cytometer and analyzed using CellEngine software (CellCarta). Fixable viability dyes used include Zombie UV (BioLegend cat#423107) and Ghost Dye Red 780 (Cytek cat#13–0865). Antibodies used for staining are listed in Table S1 and the key resources table.

Measurement of STAT5 phosphorylation

Spleens were harvested and cells prepared as above. Cells were then labeled with fixable viability dye Ghost Dye Red 780 (Cytek cat#13–0865) and anti-CD62L-FITC (MEL-14, Invitrogen cat#11–0621-82) or anti-CD62L-PECy7 (MEL-14, Invitrogen 25–0621-82) for 30 min 4°C, washed in RPMI, and let rest for 2h. Rested cells were stimulated for 20 min at 37°C with either IL-2 (Peprotech cat#212–12), at 100 ng/mL, 1000 ng/mL and 3000 ng/mL doses, IL-15 complex composed by IL-15 (Peprotech cat#210–15), IL-15Ra (R&D cat#551-MR-100), at 10 ng/mL, 100 ng/mL and 1000 ng/mL doses or IL-7 (Peprotech cat#217–17) at 100 ng/mL single dose. Activated cells were fixed with 1.6% paraformaldehyde (Electron Microscopy Sciences, cat#15710) for 10 min at RT, permeabilized with cold methanol (BDH cat#67–56-1) 30 min on ice, washed with Fluorescence Activated Cell Sorting buffer (FACS buffer PBS + 3%FBS), and incubated with appropriate flurochrome-labeled antibodies overnight at 4°C. Data acquisition was performed on a Cytek Aurora Spectral Flow Cytometer and analyzed using CellEngine (CellCarta). Antibodies used for this assay are listed in Table S1 and The key resources table.

ImageStream analyses

Single cell suspensions from splenocytes were harvested, plated at 6×10⁶ cells per well, and prepared for antibody staining as above. Cells were interrogated using the ImageStream^X MK II flow cytometer (Cytek Biosciences Fremont, CA). Measurements of 100,000 cells or more were taken alternately using the 405nm, 488nm, 561nm, 642nm and 785nm lasers with the 60x objective. Laser power was set by determining the raw max pixels for each channel to eliminate saturation. The data was acquired using the INSPIRE software (Cytek Bioscience) and analyzed with IDEAS 6.2 software (Cytek Bioscience). Antibodies used for this assay are listed in Table S1 and the key resources table.

Cytokine quantification

Blood samples were collected from mice by cardiac puncture, and the blood was allowed to coagulate at room temperature. Following centrifugation, serum supernatants were collected and stored at −20C. Quantitation of IFNγ, IL-10, IL12-p70, IL-1β, IL-2, IL-4, IL-5, IL-6, KC-GRO, and TNFα was performed using Meso Scale Diagnostics (MSD) V-PLEX Proinflammatory Panel 1 Mouse Kit (MSD #K15048D), and data acquired via the MESO QuickPlex SQ 120 instrument. Measurement of IL-15/IL-15R was performed according to manufacturer’s instructions using the mouse IL-15/IL-15R ELISA Kit (Invitrogen cat#BMS6023), and a VERSA max microplate reader. Measurement of IL-2 for the Treg, CD8 T cell transfer assays and young mice cohort was performed according to manufacturer’s instructions using the mouse IL-2 ELISA Kit (Invitrogen cat#BMS601), and a VERSA max microplate reader.

QUANTIFICATION AND STATISTICAL ANALYSIS

Data analyses were done in CellEngine by CellCarta and IDEAS 6.2 by Cytek Bioscience. Statistical tests used are indicated in the figure legends. All analyses were calculated using GraphPad Prism 10 unless otherwise stated, and all plots show the median ± 95% CI. When data did not meet assumptions for parametric statistical testing (normality, equal variance, independence, no extreme outliers), the non-parametric alternative was used.

Statistical analysis for pSTAT5 experiments

To quantify the stimulation-induced response for each sample, we computed the percentage of stimulated pSTAT5 values exceeding the 95th percentile of a matched unstimulated (baseline) distribution. Specifically, for each condition, we derived a top 5% response metric by identifying the 95th-percentile threshold in the baseline and then calculating the fraction of stimulated events above that threshold, multiplied by 100: to calculate the percentage. This percentage-based measure facilitated direct comparisons of pSTAT5 responses across genotypes (WT vs. Mut), subpopulations (Naive vs. TEM), experimental groups (intact animals – FA, mixed BMC – CH, WT Treg neonatally transferred Mut – Treg), per cytokine as indicated in the methods section. Then, data from each of the three groups: (FA, CH, Treg) were combined to evaluate pSTAT5 expression across groups, subpopulations (Naive vs. TEM), genotypes (WT vs. Mut), and corresponding dose conditions for cytokines IL-2, IL-7 and IL-15c, respectively (see methods for experimental details). We then computed pairwise Euclidean distances on these scaled data, capturing the overall dissimilarity in pSTAT5 profiles among the relevant comparisons (e.g., Naive vs. TEM, WT vs. Mut, across different doses).

For multi-factorial analyses involving T cell subpopulations, genotypes, and type of animal groups (FA, CH, Treg), cytokine and dose, we used the Scheirer–Ray–Hare (SRH) test⁴⁸ -a non-parametric extension of the two-way ANOVA/Kruskal–Wallis approach: to detect main effects and interactions without assuming normality. When the SRH revealed significant effects for dose, we conducted pairwise Dunn’s tests^49,50 with Bonferroni corrections⁵¹ to identify which specific doses/dose pair differed. Where only a single factor (e.g., genotype or subpopulation) needed comparison at a single dose, we employed Kruskal–Wallis tests accordingly and again used Dunn’s test with Bonferroni adjustment to control Type I error across multiple comparisons. A threshold of p < 0.05 was considered statistically significant. Statistical analyses were conducted in R using packages including flowCore, rcompanion, FSA, and dplyr.

Supplementary Material

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2025.115902.

Highlights.

• A hypomorphic IL2RB mutation drives immune dysregulation in vivo

• A hypomorphic IL-2Rβ drives both receptor-intrinsic and extrinsic mechanisms of disease

• Distinct IL-2 and IL-15 signaling thresholds regulate Treg and CD8 TEM cell fate

• Early regulatory T cell therapy restores immune balance and prevents pathology