ITK deficiency protects against autoimmune pulmonary hemorrhage by promoting Amphiregulin-producing regulatory T cells

Reviewer #1 (Public review):

In this study, Hossain et al. investigated the role of Interleukin-2-inducible T cell kinase (ITK) in autoimmune lung injury, demonstrating that ITK-deficient (Itk-/-) mice are protected against pristane-induced pulmonary hemorrhage (PH). Authors suggest that this protection correlates with a significant remodeling of the T cell compartment in Itk-/- mice, including increased frequency of memory-like CD4+ and CD8+ T cells (CD44⁺CD62L⁺) as well as higher frequency of Treg populations. Furthermore, adoptive transfer of ITK-deficient Treg isolated from injured ITK-deficient mice confers protection against pulmonary hemorrhage in WT recipients.

Strengths:

The adoptive transfer of congenically marked wildtype and Itk-/- Treg populations clearly demonstrates that ITK-deficient Treg can actively rescue pre-existing lung injury and reverse systemic secondary metrics like proteinuria in wild-type recipients, providing proof-of-concept validation for the therapeutic utility of the ITK-Treg axis.

Weaknesses:

The molecular analysis of Tregs isolated from unchallenged wild-type ITK-deficient mice has limited value to understand how the ITK-deficient Tregs protect against autoimmune pulmonary hemorrhages in lungs. Detailed molecular analysis of Tregs isolated from the lung will be helpful in understanding the mechanistic regulation and identifying further therapeutic targets.

Reviewer #2 (Public review):

Summary:

In this manuscript, Hossaim and colleagues investigate the role of the ITK kinase in modulating inflammation in a pristane-induced lung hemorrhage model. Using a germline ITK KO mouse, they report that loss of ITK skews the T cell compartment toward a memory-like state, expanding Tregs, and conferring protection against alveolar hemorrhage, inflammatory monocyte recruitment, proteinuria, and systemic cytokine elevation. They further show that transfer of ITK-deficient Tregs into wild-type hosts with established disease attenuates injury and shifts the cytokine balance toward resolution, and that ITK-deficient Tregs carry a transcriptional signature enriched for OXPHOS, mTORC1, MYC, and cell-cycle programs. While these observations are interesting for the development of potential immunotherapies, there are several issues with the methodological approach that support the authors' claims, tempering my enthusiasm for this manuscript.

Strengths:

(1) The clinical motivation and potential targeted therapies are relevant.

(2) The murine phenotype seems robust.

Comments on revised version.

The authors have addressed my comments appropriately.

Reviewer #3 (Public review):

Summary:

Hossain et al. investigate the role of ITK as a central regulator of autoimmune lung injury. They used ITK-deficient mice and the pristane-induced pulmonary hemorrhage (PH) model to show that ITK deficiency confers protection against PH. The adoptive cell transfer experiment suggests a possible role for altered Treg cells in ITK-deficient mice in regulating the inflammatory response in the lungs of pristane-injected mice. This study shows that targeting the ITK axis may be beneficial by reducing systemic inflammatory injury that contributes to poor outcomes in PH.

Strengths:

This study highlights the importance of ITK in regulating pulmonary hemorrhage. The enrichment of Treg cells is known to confer protection in autoimmunity-mediated alveolar damage. However, the involvement of ITK in regulating the Treg cells' function is interesting and could be explored as a novel therapeutic approach for chronic inflammation.

Weakness:

The novelty of this study lies in the association between ITK-deficient Tregs and pulmonary hemorrhage in autoimmunity. One limitation of this study is that the authors performed all experiments in ITK global knockout mice, in which not only T cells but all other cell types are deficient in ITK. Therefore, further studies are required for a mechanistic rationale for targeting ITK in severe inflammatory lung disease.

Comments on revised version:

The authors have answered my comments by incorporating new results and clarification to their observation.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public review):

A primary limitation of this manuscript is its omission of foundational literature from the Schwartzberg and Littman laboratories, which originally established the indispensable role of IL-2-inducible Tcell kinase (ITK) in proximal T-cell receptor (TCR) signaling dynamics and thymic lineage commitment. Because classic studies demonstrate that ITK is a critical regulator of thymic T cell development and cellular proliferation (PMID: 8777721, 10213685), the authors' claim that "these findings indicate that ITK deficiency skews the T cell compartment toward a memory-like state, establishing a distinct immune baseline that may favor protective and regulatory responses over pathogenic inflammation" is not substantiated by evidence and requires more robust validation.

We thank the reviewer for highlighting the foundational studies defining the role of ITK in proximal TCR signaling, thymic T-cell development, and proliferation. We agree that these studies provide essential context for interpreting the altered T-cell phenotype observed in Itk^-/- mice. Accordingly, we have revised the Introduction and Discussion to incorporate this established literature, including PMID: 8777721 and 10213685, and to more clearly acknowledge the established role of ITK in T-cell development and differentiation.

We also agree that our original statement suggesting that the memory-associated phenotype of Itk^-/- mice may establish an immune baseline favoring protection from pulmonary inflammation was not directly supported by our experiments. Our data demonstrate altered relative distributions of naïve and memory-phenotype CD4⁺ and CD8⁺ T cells, together with increased expression of CD44, CD122, Eomes, and T-bet, but do not establish that these baseline differences mediate protection from pulmonary hemorrhage.

We have therefore removed this causal interpretation and now present these findings as baseline characterization consistent with established ITK biology.

The exclusive reliance on splenic immunophenotyping is a major limitation, as it fails to capture the local cellular dynamics within the primary organs of injury (the lung and kidney). Evaluating canonical and non-canonical Treg expansion solely in the spleen overlooks the distinct functional programming of tissue-resident subsets. The authors should extend their characterization of regulatory T cell compartments directly to the lungs and draining lymphoid structures.

We thank the reviewer for this important suggestion. We agree that splenic analysis alone does not adequately capture the local regulatory T-cell response during pulmonary injury. We therefore expanded the revised manuscript to directly examine Treg populations in the lungs and draining lymph nodes.

We now quantify canonical (CD4⁺CD25⁺Foxp3⁺) and non-canonical (CD4⁺CD25^-Foxp3⁺) Tregs in the lungs of WT and Itk^-/- mice, demonstrating that the altered regulatory phenotype associated with ITK deficiency is also evident locally in the lung (Figure 5E). In addition, using CD45.1 recipients and CD45.2 WT or Itk^-/- donor Tregs, we directly tracked transferred cells and observed significantly greater accumulation of Itk^-/- donor Tregs in both the lungs and draining lymph nodes (Figure 5F–I). Donor-derived CD4⁺Foxp3⁺CD45.2⁺ cells were also confirmed in these tissues by immunofluorescence microscopy (Figure 5J).

These new experiments extend our analysis beyond the spleen and provide direct evidence that the enhanced regulatory phenotype associated with ITK deficiency is present within tissues relevant to pulmonary inflammation.

More importantly, the authors overlook key historical publications that explicitly established ITK as a negative "rheostat" or gatekeeper for regulatory T cell (Treg) differentiation. Specifically, Huang et al. (PMID: 25063868) previously demonstrated that Treg abundance is inversely correlated with ITK expression, and that ITK activity serves as a vital negative tuner of IL-2-driven Foxp3⁺ Treg expansion. Since it is already well-established that suppressing or deleting ITK promotes Treg accumulation and function, and that these cells are intrinsically vital to suppressing systemic autoimmunity, it is unclear how these findings expand upon our existing mechanistic understanding of ITK regulatory biology.

We thank the reviewer for highlighting the important study by Huang et al. (PMID: 25063868), which established ITK as a negative regulator of IL-2-driven Foxp3⁺ Treg expansion. We agree that increased Treg abundance following loss of ITK is established biology and should not be presented as a novel finding. We have revised the Introduction and Discussion accordingly and incorporated this foundational literature. The present study extends this established ITK–Treg relationship into autoimmune pulmonary hemorrhage by demonstrating functional and tissue-associated differences between WT and Itk^-/- Tregs. Equaldose adoptive-transfer experiments showed that purified Itk^-/- Tregs provide substantially greater protection than WT Tregs in an otherwise ITK-sufficient host, including reduced pulmonary pathology and proteinuria and improved survival (Figures 5 and 6). Congenic CD45.1/CD45.2 tracking further demonstrated greater accumulation of Itk^-/- donor Tregs in the lungs and draining lymph nodes (Figure 5F–J).

In addition, lung-derived Itk^-/- Tregs expressed higher levels of Amphiregulin (AREG), identifying a tissue-repair-associated phenotype (Figure 5K), while RNA-seq demonstrated a distinct transcriptional state enriched for metabolic, proliferative, and regulatory programs (Figure 7). Thus, the advance of this study is not increased Treg abundance itself, but the demonstration that ITK-deficient Tregs exhibit enhanced protective activity and distinct tissue-associated properties during severe autoimmune pulmonary injury.

Reviewer #2 (Public review):

(1) All loss-of-function experiments are from a global ITK knockout. This is a major limitation and weakness of this study. The protection observed in the intact knockout, therefore, cannot be attributed to Tregs specifically. The Treg-intrinsic claim rests almost entirely on a single adoptive-transfer experiment. In order to show that this effect is Treg-specific, the authors would need to generate a Treg-specific ITK-deficient mouse.

We thank Reviewer #2 for highlighting this important limitation. We agree that experiments performed in global Itk^-/- mice cannot establish that the protective phenotype is specifically attributable to Treg-intrinsic loss of ITK, because other ITK-deficient immune populations may also contribute. We have revised the manuscript to make this limitation explicit and agree that a Treg-specific conditional model, such as Foxp3-Cre × Itk^fl/fl mice, would be required to establish Treg-specific necessity in the intact disease model. To directly assess the contribution of Tregs independently of the global knockout environment, we transferred equal numbers (1 × 10⁶) of FACS-purified WT or Itk^-/- CD4⁺CD25⁺Foxp3⁺ Tregs into pristane-challenged WT CD45.1 recipients. Thus, all endogenous immune and stromal compartments in the recipients remained ITK-sufficient. Under these conditions, Itk^-/- Tregs provided substantially greater protection than WT Tregs, including reduced pulmonary pathology and proteinuria and improved survival (Figures 5 and 6). Congenic tracking further demonstrated greater accumulation of Itk^-/- donor Tregs in the lungs and draining lymph nodes. We therefore conclude that purified Itk^-/- Tregs are sufficient to confer enhanced protective activity in an ITK-sufficient host, while recognizing that these experiments do not establish that Tregs are solely responsible for the protection observed in global Itk^-/- mice.

(2) In their sufficiency experiment (adoptive Treg cell transfer), donor and/or host cells are not congenially marked, so persistence, lung trafficking, and in vivo expansion of transferred Tregs are not demonstrated.

We thank Reviewer #2 for this important comment. To directly address this limitation, we performed new adoptive-transfer experiments using CD45.1 WT recipients and CD45.2 WT or Itk^-/- donor Tregs, allowing transferred cells to be clearly distinguished from endogenous host cells. Following transfer, flow cytometric analysis demonstrated significantly greater accumulation of donor-derived Itk^-/- Tregs in the lungs and draining lymph nodes compared with WT donor Tregs (Figure 5F–I). Within the CD45.2⁺ donor population, cells retained Foxp3 expression, and both CD25⁺Foxp3⁺ and CD25^-Foxp3⁺ regulatory populations were detectable at the experimental endpoint. The presence of donor-derived CD4⁺Foxp3⁺CD45.2⁺ cells within the lungs and draining lymph nodes was independently confirmed by immunofluorescence microscopy (Figure 5J).

These new experiments directly demonstrate localization and enhanced accumulation/recovery of Itk^-/- Tregs in disease-relevant tissues. However, because these are endpoint measurements, we have avoided attributing the increased donor-cell recovery specifically to proliferation or trafficking, as differences in survival, retention, or recruitment may also contribute.

(3) The authors claim that ITK-deficient Tregs possess enhanced metabolic fitness. This conclusion is based on transcriptional profiling of isolated splenic Tregs from unchallenged mice, yet it concerns lung protection during active disease narrative. Additional functional validation would be needed (Seahorse assay, mitochondrial mass/potential, etc). Some of these GSEA programs enriched in ITK-deficient Tregs could reflect a more general proliferative signature.

We thank Reviewer #2 for this thoughtful critique. We agree that enrichment of metabolic pathways by RNA-seq does not, by itself, demonstrate enhanced metabolic fitness and that direct functional assays, such as Seahorse analysis or measurements of mitochondrial mass and membrane potential, would be required to establish altered metabolic function.

We also acknowledge that our RNA-seq was performed on FACS-purified splenic Tregs from unchallenged mice and therefore does not directly define the transcriptional state of Tregs within the inflamed lung during active pulmonary hemorrhage. Our intent was to determine whether ITK deficiency is associated with baseline transcriptional differences prior to disease challenge. The analysis identified enrichment of oxidative phosphorylation, mTORC1, glycolysis, MYC, E2F, and cell-cycle-associated programs in Itk^-/- Tregs (Figure 7). We agree that some of these signatures may reflect proliferative or activation-associated transcription rather than enhanced metabolic function. We have revised the RNA-seq analysis, please updated Figure 8 and supplementary Tables. Accordingly, we have revised the Results and Discussion to describe these data as a distinct transcriptional state enriched for metabolic and proliferative programs, rather than direct evidence of enhanced metabolic fitness. We also note that disease-state lung Treg transcriptomics and direct metabolic assays will be required to determine the functional significance of these pathways during pulmonary injury.

Reviewer #3 (Public review):

The novelty of this study lies in the association between ITK-deficient Tregs and pulmonary hemorrhage in autoimmunity. The weakness of the manuscript is the lack of sufficient experiments to support the claim that ITK-deficient mice show protection specifically mediated by Treg cells, and to demonstrate that ITK-deficient Treg cells are more efficient than WT Treg cells in regulating other immune cells that drive pulmonary damage. The authors performed all the experiments in ITK global knockout mice, in which not only T cells but all other cell types are deficient in ITK. Furthermore, they have not performed any functional analysis to demonstrate the functional differences between WT Treg and ITK-deficient Treg cells, undermining the novelty of this study.

We sincerely thank Reviewer #3 for this constructive critique. We agree that experiments performed in global Itk^-/- mice alone cannot establish that protection is specifically mediated by Tregs, because other ITK-deficient immune populations may also contribute. We have revised the manuscript to clearly acknowledge this limitation.

To directly compare the functional protective capacity of WT and Itk^-/- Tregs independently of the global knockout environment, we transferred equal numbers of FACS-purified CD4⁺CD25⁺Foxp3⁺ Tregs into pristane-challenged WT recipients. Under these conditions, Itk^-/- Tregs provided substantially greater protection than WT Tregs, including reduced pulmonary pathology and proteinuria, a more regulatory systemic cytokine profile, and improved survival (Figures 5 and 6). These experiments provide direct in vivo evidence that purified Itk^-/- Tregs have enhanced protective activity in an otherwise ITK-sufficient host.

We further performed CD45.1/CD45.2 congenic tracking and observed significantly greater accumulation of donor-derived Itk^-/- Tregs in the lungs and draining lymph nodes compared with WT donor Tregs (Figure 5F– J). Donor-derived cells retained Foxp3 expression following transfer. In addition, canonical Tregs isolated directly from Itk^-/- lungs expressed substantially higher levels of Amphiregulin (AREG), identifying a tissue-repair-associated phenotype that accompanies their enhanced protective activity (Figure 5K). We have also incorporated our previous functional studies demonstrating enhanced suppressive activity of Itk^-/- regulatory T-cell populations compared with WT Tregs (Mammadli et al., Clinical and Translational Medicine, 2021).

Thus, the advance of the present study is not simply increased Treg abundance after loss of ITK, but the demonstration that purified Itk^-/- Tregs provide greater protection than WT Tregs in an ITK-sufficient host, accumulate more prominently in disease-relevant tissues, and exhibit increased AREG expression. We nevertheless acknowledge that a Treg-specific conditional Itk knockout will ultimately be required to establish Treg-specific necessity in the intact disease model.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

Major concerns

(1) The authors demonstrate that the frequency of memory-like T cells is elevated in the spleens of Itk-/- mice, but the absolute cellular numbers are notably absent. This proportional observation lacks conceptual novelty, as highly similar findings regarding the baseline expansion of memory-like/innate-like T cells under ITK-deficient conditions have been extensively established, most notably by the Schwartzberg laboratory (PMID: 16860760, 16860760). These early reports should be cited in the introduction and not presented as novel findings. More critically, the direct relevance of this baseline phenotype to the active autoimmune disease model remains unaddressed. The frequencies of these effector and memory-like compartments are evaluated exclusively in unperturbed, naïve animals and are never tracked or revisited during the subsequent kinetics of the pristane-induced lung injury model.

We thank Reviewer #1 for this important critique and for highlighting the foundational studies describing memory-like/innate-like T-cell development in ITK-deficient mice. We agree that this phenotype is well established and should not be presented as a novel finding. We have therefore incorporated the relevant literature, including PMID: 16860760, and now present Figures 1 and 2 as baseline characterization of the established Itk^-/- T-cell phenotype. We also agree that the frequency measurements in these figures do not demonstrate expansion in absolute cellular terms. Accordingly, we have removed the term “expansion” for this baseline naïve, central-memory, and effector-memory populations and now describe the findings as differences in their relative frequencies or distribution. Importantly, these analyses were performed in unchallenged mice, and we did not longitudinally track these memory-associated CD4⁺ and CD8⁺ populations during pristane-induced disease. We therefore no longer suggest that this baseline phenotype directly mediates protection from pulmonary hemorrhage. The revised manuscript now clearly separates this established baseline ITK phenotype from the disease-focused findings, with the mechanistic emphasis placed on the regulatory T-cell compartment, for which we provide direct functional and tissue-level analyses during pulmonary injury.

(2) This memory-like cell population expansion is observed strictly within naïve, SPF mice that have had no prior exposure to foreign antigens; it raises an important question of whether these are bona fide, antigen-experienced memory cells, or do they represent a "virtual memory" population (CD44+CD122+CD49d-) that typically arises from aberrant T cell development and homeostatic dysregulation? Prior work has already demonstrated that compromised TCR signal strength via ITK deficiency alters critical aspects of homeostatic proliferation, metabolic programming, and non-conventional effector differentiation in unexposed environments (Huang et al., 2018; bioRxiv, https://doi.org/10.1101/359604). Without providing absolute counts during active disease and ruling out a virtual memory phenotype, the authors' conclusion that this baseline skewing directly orchestrates active protection against pathogenic inflammation is currently unsupported.

We thank Reviewer #1 for this important comment. We agree that the CD44⁺CD122⁺ phenotype observed in naïve SPF Itk^-/- mice cannot be assumed to represent conventional antigen-experienced memory. Prior studies have shown that reduced ITK-dependent TCR signaling can promote innate-like or virtual-memory-associated T-cell populations under antigen-naïve conditions, and we have incorporated the relevant literature, including Huang et al., into the revised Introduction and Discussion. Because our experiments did not include CD49d or other markers required to distinguish conventional memory from virtual-memory cells, we have avoided assigning a definitive origin to these populations and now refer to them as memory-phenotype or memory-associated T cells. We also agree that these baseline measurements were obtained from unchallenged mice and that we did not determine absolute numbers or longitudinally track these populations during pristane-induced disease. We have therefore removed the interpretation that baseline memory-associated skewing directly promotes protection from pulmonary inflammation. Accordingly, Figures 1 and 2 are now presented as characterization of an established consequence of ITK deficiency, while our disease-focused mechanistic conclusions are based on the Treg experiments performed during pulmonary injury.

(3) The manuscript relies predominantly on qualitative histological representations to assess autoimmune tissue pathology. To substantiate the study's claims, the authors should provide standardized quantification of the histological changes. Additionally, direct measurement of autoimmune T cell responses, rather than descriptive tissue overviews, is required to definitively prove that ITK deficiency suppresses pathogenic autoimmunity.

We thank Reviewer #1 for these important comments. Regarding tissue pathology, the lung histology was evaluated using standardized semi-quantitative scoring performed by a board-certified pathologist blinded to genotype and treatment group. These quantitative scores are presented in Figure 3 and described in the Methods. We have revised the manuscript to make this analysis more explicit. The histopathologic findings are further supported by quantitative measurements of pulmonary CD11b⁺Ly6C⁺ inflammatory monocytes and CD11b⁺Ly6G⁺ neutrophils, as well as proteinuria as a measure of renal involvement (Figures 3 and 4). We also agree that direct analysis of antigen-specific pathogenic T-cell responses would provide additional mechanistic information. Such populations were not directly measured in the present study, and we have therefore revised the manuscript to avoid implying that ITK deficiency specifically suppresses autoreactive conventional T cells. Instead, our data support the conclusion that ITK deficiency is associated with reduced inflammatory tissue injury and that purified Itk^-/- Tregs are sufficient to confer enhanced protection in WT recipients. This is supported by reduced pulmonary pathology, decreased inflammatory-cell accumulation, improved proteinuria, and the adoptive-transfer experiments. We have revised the Discussion accordingly to distinguish these directly demonstrated effects from the broader question of how ITK regulates antigen-specific autoimmune T-cell responses.

(4) While the adoptive transfer experiments suggest that Tregs from injured ITK-deficient animals possess enhanced protective capacity, the interpretation of these data is heavily confounded by a lack of absolute cellular quantification. Prior literature establishes that while ITK deficiency increases Treg frequency, the absolute number of Tregs remains unchanged due to concurrent reductions in global T cell cellularity (PMID: 25063868). Consequently, the authors must provide the absolute numbers of Tregs infiltrating the lungs both at baseline and post-injury. Since the authors adoptively transfer matched numbers of WT and KO Tregs, they are testing an artificial cellular ratio that may completely bypass the actual physiological abundance, contraction, or recruitment dynamics operating within the lung tissue in vivo. Furthermore, the current experimental design is insufficient to establish whether this protection depends specifically on injury-primed effector Treg programming or whether naïve ITK-deficient Tregs are sufficient.

We thank Reviewer #1 for this important critique. We agree that Treg frequencies alone do not establish physiological abundance, particularly because previous studies have shown that increased Treg frequency in Itk^-/- mice may occur without a proportional increase in absolute Treg number.

To address this concern, we have added absolute quantification of canonical (CD4⁺CD25⁺Foxp3⁺) and non-canonical (CD4⁺CD25^-Foxp3⁺) Tregs in the lung at baseline and following adoptive transfer. Absolute numbers were determined from the total viable cellular recovery from each individual tissue together with the complete sequential flow-cytometric gating hierarchy. These analyses provide both frequency and absolute cellular abundance and demonstrate greater recovery of donor-derived Itk^-/- Tregs in the inflamed lung and draining lymph nodes following transfer (Figure 5E–I).

We also agree that equal-dose adoptive transfer does not reproduce the endogenous WT: Itk^-/- Treg ratio. Rather, this experiment was intentionally designed as a controlled functional comparison in which identical numbers of purified WT or Itk^-/- Tregs were introduced into the same ITK-sufficient recipient environment. The greater protection and tissue accumulation observed with Itk^-/- Tregs therefore reflect differences between the transferred populations under equivalent starting conditions, rather than differences in the number of cells administered. Importantly, the donor Tregs used for these experiments were isolated from unchallenged WT and Itk^-/- Foxp3-reporter mice. Thus, prior pristane-induced injury was not required for the enhanced protective activity of Itk^-/- Tregs. We have clarified this point in the revised manuscript.

Together, these analyses distinguish Treg frequency from physiological abundance, provide quantitative assessment of donor-cell recovery in disease-relevant tissues, and demonstrate that the enhanced protective capacity of Itk^-/- Tregs does not require prior injury-dependent priming.

(5) The transcriptomic analyses suggest metabolic and functional reprogramming of ITK-deficient Tregs. However, the interpretation of these findings currently remains speculative, as it relies on pathway analysis and no key genes involved in any of these programs were identified/highlighted as key differentially expressed genes. Lastly, there is no functional validation to suggest that these transcriptional changes are functionally relevant.

We thank Reviewer #1 for this important critique. We agree that pathway-level enrichment alone does not establish functional metabolic reprogramming and that the genes underlying these signatures should be provided more clearly. To address this concern, we have expanded the transcriptomic analysis to identify the individual differentially expressed genes contributing to the major pathways observed in Itk^-/- Tregs, including mTORC1, oxidative phosphorylation, MYC/E2F, glycolytic, and regulatory programs. We have revised Figure 8 and added supplementary tables providing the gene-level differential-expression data for the canonical and non-canonical Treg comparisons, allowing the underlying molecular changes to be evaluated directly rather than relying solely on pathway enrichment scores. We also agree that these transcriptional signatures do not demonstrate enhanced metabolic fitness. Direct functional assays, such as extracellular flux analysis, mitochondrial mass or membrane-potential measurements, would be required to establish altered metabolic function. We have therefore revised the Results and Discussion to describe Itk^-/- Tregs as exhibiting a distinct transcriptional state enriched for metabolic, proliferative, and regulatory programs, while acknowledging that some pathways may reflect proliferation or activation.

Separately, lung-derived Itk^-/- Tregs showed increased AREG protein expression (Figure 5K), which we interpret as a tissue-repair-associated phenotype rather than functional validation of the metabolic pathways.

Thus, we have added gene-level resolution to the RNA-seq analysis while narrowing our mechanistic interpretation to what is directly supported by the data. Accordingly, we have revised the Results and Discussion to describe these data as a distinct transcriptional state enriched for metabolic and proliferative programs, rather than direct evidence of enhanced metabolic fitness.

Minor concerns

(1) The introduction is structurally imbalanced, with three extensive paragraphs dedicated to describing and discussing the study's results. On the contrary, critical historical studies are missing from this section.

We thank Reviewer #1 for this constructive comment. We agree that the original Introduction devoted excessive space to summarizing our findings while providing insufficient historical context for ITK biology. We have therefore substantially shortened and reorganized the Introduction. The revised version now incorporates the foundational studies defining ITK function in TCR signaling and T-cell development, as well as the work establishing ITK as a regulator of Foxp3⁺ Treg differentiation and expansion. We also removed the extended discussion of our results and now conclude the Introduction with a concise statement of the knowledge gap and objective of the present study.

(2) In Figure 1A and 1E, a substantial CD44-CD62L- population is present in both WT and ITK KO mice, potentially reflecting gating-related issues or loss of CD62L staining.

We thank Reviewer #1 for this careful observation. We agree that the CD44^-CD62L^- population should be interpreted cautiously, as reduced CD62L staining may reflect biological variability or loss/shedding during tissue processing. To improve transparency, we have added the complete flow-cytometric gating strategy to the revised manuscript (Supplementary Figure 1; page 50, lines 845–855), including sequential exclusion of debris, doublets, and non-viable cells before identification of CD3⁺, CD4⁺, and CD8⁺ T-cell populations. WT and Itk^-/- samples were processed and analyzed under identical conditions. Importantly, we do not assign biological significance to the CD44^-CD62L^- population. Our conclusions are restricted to the conventionally defined naïve (CD44^lo^CD62L^hi^), central-memory (CD44^hi^CD62L^hi^), and effector memory (CD44^hi^CD62L^lo^) populations.

(3) The characterization of the myeloid compartment is limited given the highly complex nature of pulmonary inflammation in small vessel vasculitis models. To rigorously evaluate the local immune microenvironment, the authors must provide a detailed profiling of key populations within the affected tissues, including alveolar macrophages, interstitial macrophages, neutrophils, and eosinophils.

We thank Reviewer #1 for this important comment. We agree that pulmonary inflammation involves a heterogeneous myeloid compartment and that more comprehensive profiling of alveolar macrophages, interstitial macrophages, eosinophils, and other subsets would provide additional insight. In the present study, our myeloid analysis focused on two populations strongly associated with acute inflammatory lung injury: CD11b⁺Ly6C⁺ inflammatory monocytes and CD11b⁺Ly6G⁺ neutrophils. Quantitative flow cytometry demonstrated significantly reduced accumulation of both populations in the lungs of pristane-treated Itk^-/- mice compared with WT controls (Figure 3H–I). We acknowledge that this does not represent comprehensive profiling of the pulmonary myeloid compartment, and we have revised the manuscript to make this limitation explicit and to avoid broader claims of global myeloid remodeling. We also note in the Discussion that detailed analysis of macrophage and granulocyte subsets will be important for determining how ITK-dependent regulatory responses influence the broader pulmonary inflammatory environment.

(4) Although the study focuses on suppression of autoimmunity, chronic ITK deficiency may plausibly impair protective immunity against infection or malignancy. Some discussion of these potential tradeoffs would improve the translational interpretation of the findings.

We thank Reviewer #1 for raising this important translational consideration. We agree that chronic or systemic inhibition of ITK could impair protective immunity, particularly against infection and potentially malignancy. We have expanded the Discussion to address these potential trade-offs. In particular, we now note that inherited ITK deficiency in humans has been associated with impaired protective immune responses and increased susceptibility to infection, emphasizing that prolonged systemic ITK suppression cannot be assumed to be without consequence. At the same time, our previous studies in alloimmunity demonstrated that attenuation of ITK signaling can markedly reduce graft-versus-host disease while preserving substantial graft-versus-leukaemia activity (PMIDs: 33324410, 33851101, and 34919342). These findings suggest that pathogenic and protective immune responses may differ in their dependence on ITK signaling. We therefore emphasize that the translational effects of ITK modulation will likely depend on the magnitude, duration, and cellular context of inhibition. Future studies will be needed to determine whether partial, transient, or cell-selective ITK targeting can retain anti-inflammatory and Treg-associated benefits while preserving antimicrobial and antitumor immunity.

Reviewer #3 (Recommendations for the authors):

The manuscript by Hossain et al investigates the role of ITK as a central regulator of autoimmune lung injury, using the pristane-induced pulmonary hemorrhage (PH) model.

The authors claim that the deficiency of ITK confers complete protection primarily by marked remodeling of the T cell compartment, including expansion of CD44⁺, CD122⁺, Eomes⁺, T-bet⁺ memory-like subsets, and significant enrichment of Foxp3⁺ regulatory T cells (Tregs). However, they have used the ITK global KO mouse model, which undermines their claim. Adoptive transfer of Treg cells to suppress autoimmunity has been shown before in various studies. The authors have performed an adoptive transfer of ITK-deficient Tregs, which they claim are more efficient at suppressing autoimmunity, based solely on two cytokine levels in systemic circulation. Although the authors identified the ITK-deficient Treg as key mediators of tissue protection, they have not provided any functional characterization of ITK-deficient Treg cells as compared to WT Treg cells. There are a few comments that will help to strengthen the findings and establish the importance of ITK regarding autoimmune PH.

We sincerely thank Reviewer #3 for this rigorous and constructive critique. We agree that experiments performed in global Itk^-/- mice alone cannot establish that protection is specifically mediated by Tregs, because other ITK-deficient immune populations may also contribute. We have revised the manuscript to make this limitation explicit.

Importantly, the functional comparison between WT and Itk^-/- Tregs in the revised study extends beyond circulating cytokines. Equal numbers of FACS-purified WT or Itk^-/- CD4⁺CD25⁺Foxp3⁺ Tregs were transferred into pristane-challenged WT recipients, in which all endogenous immune and stromal compartments remained ITK-sufficient. Under these conditions, Itk^-/- Tregs provided substantially greater protection than WT Tregs, including reduced pulmonary pathology and proteinuria, a more regulatory systemic cytokine profile, and improved survival (Figures 5 and 6).

These experiments demonstrate enhanced protective activity of purified Itk^-/- Tregs in an ITK-sufficient host. We further performed CD45.1/CD45.2 congenic tracking and observed greater accumulation of donor-derived Itk^-/- Tregs in the lungs and draining lymph nodes compared with WT donor Tregs (Figure 5F–J). In addition, lung-derived Itk^-/- Tregs expressed substantially higher levels of Amphiregulin (AREG), identifying a tissue-repair-associated phenotype (Figure 5K). We have also more clearly incorporated our previous functional study demonstrating enhanced suppressive activity of Itk^-/- regulatory T-cell populations compared with WT Tregs (PMID: 34919342).

Together, these findings extend previous work by demonstrating that Itk^-/- Tregs exhibit enhanced protective activity and distinct tissue-associated properties during autoimmune pulmonary haemorrhage. We nevertheless acknowledge that a Treg-specific conditional Itk knockout will ultimately be required to establish Treg-specific necessity in the intact disease model.

Major comments:

(1) In Figure 1, the authors have shown the T cell profile in the spleen from WT vs ITK-KO mice under homeostatic conditions. But it will be important to show the immune cell population, both innate and adaptive, in the lung, as it is the organ of focus for their study.

We sincerely thank Reviewer #3 for this important suggestion. We agree that because the lung is the primary site of injury in this model, characterization of the local pulmonary immune compartment is important.

In the revised manuscript, we have expanded our lung-focused analyses. We now directly quantify canonical (CD4⁺CD25⁺Foxp3⁺) and non-canonical (CD4⁺CD25^-Foxp3⁺) Tregs in the lungs of WT and Itk^-/- mice (Figure 5E), demonstrating that the altered regulatory phenotype associated with ITK deficiency is also evident locally within the lung.

During pristane-induced disease, we additionally quantified key inflammatory innate populations and found significantly reduced accumulation of CD11b⁺Ly6C⁺ inflammatory monocytes and CD11b⁺Ly6G⁺ neutrophils in Itk^-/- lungs compared with WT controls (Figure 3H–I). Furthermore, using CD45.1/CD45.2 congenic adoptive transfer, we directly tracked WT and Itk^-/- donor Tregs and observed greater accumulation of Itk^-/- Tregs in the lungs and draining lymph nodes (Figure 5F–J).

We acknowledge that the present study does not provide comprehensive profiling of all pulmonary immune populations, including macrophage subsets, eosinophils, and conventional effector T-cell populations. We have therefore revised the manuscript to make the scope of our lung immune analysis explicit and identify broader pulmonary immune profiling as an important area for future study

(2) The authors need to show the status of Treg cells in the spleen and lung of ITK KO mice under homeostatic conditions.

We thank Reviewer #3 for this important suggestion. We agree that defining the homeostatic Treg compartment in both the spleen and lung is important for interpreting the disease-associated changes.

In the revised manuscript, we now provide baseline characterization of canonical (CD4⁺CD25⁺Foxp3⁺) and non-canonical (CD4⁺CD25^-Foxp3⁺) Tregs in naïve, unchallenged WT and Itk^-/- mice. Homeostatic splenic Tregs are shown in Supplementary Figure 7, demonstrating the established increase in the non-canonical Treg compartment in Itk^-/- mice. We also added direct analysis of lung Tregs under homeostatic conditions (Figure 5E), demonstrating that the altered regulatory phenotype associated with ITK deficiency is also present locally in the lung prior to pristane challenge.

These analyses establish the baseline Treg phenotype in both systemic and pulmonary compartments and provide an appropriate reference for interpreting the subsequent disease and adoptive-transfer experiments.

(3) The authors claim ITK deficiency shifts the memory T cell phenotype based on surface markers and transcription factors TBET and Eomes. It is well established that the memory T cell phenotype depends on a fine balance between TBET and Eomes (Cárdenas et al., Nature, 2024; Kar et al., Science Advances, 2026). Therefore, to support the claim, authors need to co-stain TBET and Eomes to show the relative balance of the TFs.

We thank Reviewer #3 for this important comment and for highlighting the literature emphasizing that memory T-cell differentiation is influenced by the relative balance of T-bet and Eomes. We agree that because T-bet and Eomes were measured independently in our experiments, our data do not establish their relative balance or co-expression at the single-cell level. Definitive analysis would require simultaneous co-staining of these transcription factors.

Accordingly, we have revised the manuscript to avoid interpreting T-bet and Eomes expression as evidence of a specific memory differentiation program. Figures 1 and 2 are now presented as baseline characterization of the established Itk^-/- T-cell phenotype. We describe the CD44/CD62L data as showing altered relative frequencies of naïve and memory-phenotype populations, while increased T-bet, Eomes, CD44, and CD122 expression is reported separately as a differentiation-associated phenotype.

We no longer infer that a specific T-bet/Eomes balance contributes to the memory phenotype or to protection from pulmonary hemorrhage, and we have incorporated the literature cited by the reviewer into the revised Discussion.

(4) A higher expression of TBET and Eomes could also lead to enhanced production of Th1 cytokines. Have the authors measured the CD4 T cell cytokines or CD8 T cell granzyme/perforin production in ITK KO mice?

We thank Reviewer #3 for this important point. We agree that increased T-bet and Eomes expression may be associated with altered CD4⁺ and CD8⁺ T-cell effector function and that direct functional measurements would be required to establish this.

In the present study, we did not directly measure cytokine production by conventional CD4⁺ or CD8⁺ T cells, nor did we quantify granzyme or perforin expression. Although systemic IFN-γ, TNF-α, IL-6, and IL-17 were measured during pristane-induced disease, these serum measurements do not identify the cellular source of the cytokines.

We have therefore revised the Results and Discussion to avoid inferring enhanced Th1 or cytotoxic function from increased T-bet and Eomes expression. Figures 1 and 2 are now presented as baseline phenotypic characterization of the conventional T-cell compartment, while the mechanistic conclusions of the study remain focused on the regulatory T-cell response during pulmonary hemorrhage.

We have also acknowledged that direct functional analysis of conventional CD4⁺ and CD8⁺ T cells will be required to define the consequences of altered T-bet and Eomes expression in this model.

(5) The authors have shown that ITK KO mice significantly restrict the pulmonary hemorrhage after 14 days post-pristane injection. Have the authors evaluated the findings at a longer time point, such as 6-8 weeks post-pristane, which is considered the optimal time point for disease progression? This will clarify whether ITK deficiency completely inhibits the autoimmunity or delays its progression.

We thank Reviewer #3 for raising this important point regarding disease kinetics. We agree that later time points, such as 6–8 weeks after pristane administration, are commonly used to evaluate chronic manifestations of pristane-induced systemic autoimmunity.

However, the pulmonary hemorrhage phenotype examined in our study follows a much more acute and severe course. We therefore extended our analysis to include survival and added these data to the revised manuscript (Figure 3—figure supplement 2). All pristane-treated WT mice died within approximately 20 days, whereas Itk^-/- mice exhibited significantly prolonged survival.

Because WT mice do not survive to the 6–8-week time point in this model, a direct WT versus Itk^-/- comparison at these later stages is not feasible without substantial survivor-selection bias. We have therefore clarified that the present study addresses the acute, severe phase of pristane-induced pulmonary hemorrhage.

Importantly, the survival data demonstrate that ITK deficiency protects against the early lethal phase of disease rather than simply delaying the day-14 pulmonary phenotype. However, these findings do not establish that ITK deficiency completely prevents later chronic manifestations of pristane-induced autoimmunity. We have added this limitation to the Discussion and identify longer-term analysis of surviving Itk^-/- mice as an important future direction.

(6) In Figure 3, the authors have shown that ITK KO mice show less monocyte infiltration in pristane-injected mice. However, they have not shown effector T cell percentages, such as Th1 and Th17, which are important drivers of disease progression. Given their focus on Treg cells, it will be important to show the effector T-cell population in the lung.

We thank Reviewer #3 for this important comment. We agree that direct quantification of pulmonary Th1 and Th17 populations would provide additional insight into the balance between pathogenic effector T cells and Tregs during pulmonary hemorrhage.

In the present study, we did not directly quantify lung Th1 or Th17 cells by intracellular cytokine or transcription-factor staining. We have therefore revised the manuscript to avoid implying that our experiments demonstrate direct suppression of these effector T-cell populations.

Our lung analysis instead demonstrated significantly reduced accumulation of CD11b⁺Ly6C⁺ inflammatory monocytes and CD11b⁺Ly6G⁺ neutrophils in Itk^-/- mice (Figure 3), together with increased pulmonary Treg populations (Figure 5). We also observed reduced systemic IFN-γ, TNF-α, IL-6, and IL-17 during disease (Figure 4); however, we recognize that these serum measurements do not establish the cellular source of these cytokines. Accordingly, we now distinguish the directly measured pulmonary cellular changes from broader Th1/Th17-associated inflammatory responses and identify direct profiling of lung effector T-cell populations as an important area for future investigation.

(7) In Figure 4, the authors have shown several plasma cytokine levels in KO vs WT mice to support their hypothesis. However, as their focus is on alveolar hemorrhage, they have also mentioned that these cytokines influence the pulmonary damage by altering neutrophil infiltration and vascular permeability. Showing data on neutrophil and vascular damage will strengthen their claim.

We thank Reviewer #3 for this constructive suggestion. We agree that direct assessment of pulmonary neutrophil accumulation and tissue injury provides important context for interpreting the systemic cytokine changes observed during pulmonary hemorrhage.

To address this concern, we added quantitative flow-cytometric analysis of pulmonary neutrophils. CD11b⁺Ly6G⁺ neutrophils were significantly reduced in the lungs of pristane-treated Itk^-/- mice compared with WT controls (Figure 3I). Together with the reduced CD11b⁺Ly6C⁺ inflammatory monocytes, these findings demonstrate decreased accumulation of major inflammatory innate-cell populations in the injured lung.

Pulmonary injury was also evaluated using blinded, standardized histopathologic scoring. WT mice exhibited extensive intra-alveolar hemorrhage, erythrocyte accumulation, and disruption of alveolar architecture, whereas Itk^-/- mice showed markedly reduced hemorrhage and preserved lung architecture (Figure 3B–E). We acknowledge that the present study does not include a direct assay of vascular permeability or endothelial barrier function. We have therefore revised the manuscript to describe these findings as evidence of reduced alveolarcapillary injury and pulmonary hemorrhage rather than as a direct demonstration of improved vascular permeability.

(8) The authors mentioned non-canonical Treg cells without proper characterization. These Treg cells are reported to lack CD25 and, in some cases, FOXP3, and to have reduced suppressive function. Therefore, if ITK KO mice show that both canonical and non-canonical Treg cells are increased, this raises the question of how that balance leads to less pulmonary hemorrhage. The authors need to clarify their findings and discuss the relevance.

We sincerely thank Reviewer #3 for raising this important point. In the present study, we define non-canonical Tregs specifically as CD3⁺CD4⁺CD25^-Foxp3⁺ cells. These cells were identified using Foxp3-RFP reporter mice, and therefore the population analyzed here is Foxp3⁺CD25 but lacks CD25.

We agree that CD25^-Foxp3⁺ Tregs may exhibit reduced stability or suppressive function in some settings. However, in our previous study, we performed dedicated phenotypic and functional analyses of this population and demonstrated that Itk^-/- CD4⁺CD25^-Foxp3⁺ Tregs possess substantial suppressive activity in both in vitro and in vivo assays (Mammadli et al., Clinical and Translational Medicine, 2021; PMID: 34919342). We have revised the manuscript to more clearly incorporate this prior work.

In the present study, both canonical CD25⁺Foxp3⁺ and non-canonical CD25^-Foxp3⁺ Treg populations were increased in Itk^-/- mice, including in the lung. However, we did not independently test the protective function of the non-canonical subset. Importantly, the adoptive-transfer experiments were initiated with purified canonical CD25⁺Foxp3⁺ Tregs.

We have therefore revised the manuscript to avoid implying that non-canonical Tregs independently mediate protection from pulmonary hemorrhage. Instead, we interpret their increase as an additional feature of the altered regulatory T-cell compartment associated with ITK deficiency, while restricting the direct functional conclusion to the enhanced protective activity of purified Itk^-/- canonical Tregs.

(9) Do the ITK KO mice have non-canonical Treg in the homeostatic lung?

We thank Reviewer #3 for this specific question. Yes, Itk^-/- mice contain noncanonical CD4⁺CD25^-Foxp3⁺ Tregs in the lung under homeostatic conditions.

To directly address this point, we added flow-cytometric analysis of canonical CD4⁺CD25⁺Foxp3⁺ and non-canonical CD4⁺CD25^-Foxp3⁺ Tregs in the lungs of naïve, unchallenged WT and Itk^-/- mice. These data demonstrate that non-canonical Tregs are present in the homeostatic lung and are increased in Itk^-/- mice compared with WT controls (Figure 5E).

Thus, the non-canonical Treg population observed during pulmonary hemorrhage is not exclusively induced by disease but is already present as part of the baseline regulatory T-cell phenotype associated with ITK deficiency

(10) The authors stated that ITK-/- Tregs confer enhanced protection against PH by shifting the cytokine balance from a proinflammatory toward a pro-resolving state. However, the adoptive transfer of ITK KO Treg results in differences only in IFNγ and IL10 levels compared with WT Treg. The authors need to clearly discuss this.

We thank Reviewer #3 for this careful observation. We agree that our original wording overstated the extent of the cytokine differences between recipients of WT and Itk^-/- Tregs. Upon direct comparison of the two Treg-treated groups, the clearest statistically significant differences were observed for IFN-γ and IL-10. Although Itk^-/- Treg treatment was associated with reductions in additional proinflammatory cytokines, these changes should not be interpreted as demonstrating significant superiority over WT Tregs for every cytokine measured.

We have therefore revised the Results and Discussion to emphasize that, compared with WT Tregs, Itk^-/- Tregs produced a greater reduction in IFN-γ together with increased IL-10, and we have removed the broader statement that they uniformly shift the entire systemic cytokine profile from a proinflammatory to a pro-resolving state.

Importantly, the enhanced protective activity of Itk^-/- Tregs is supported not only by cytokine measurements, but also by reduced pulmonary pathology and proteinuria, improved survival, and greater accumulation of donor-derived Itk^-/- Tregs in the lungs and draining lymph nodes.

(11) The claim that ITK-deficient Treg is better than WT Treg is the novelty of the study. Therefore, the functional characterization, such as cytokine profile and suppressive functions, could be shown to support their claim. Adoptive transfer of WT-Treg cells provides significant protection, as shown by Dall'Era et al. (2019, Arthritis Rheumatol) and by the authors in Figure 5. Providing supporting data on the advantages of ITK-KO Treg will be important for strengthening the study's novelty.

We sincerely thank Reviewer #3 for this important comment. We agree that the novelty of the present study is not simply that Tregs can protect against pulmonary hemorrhage, which has been previously demonstrated, but that Itk^-/- Tregs exhibit enhanced protective activity compared with WT Tregs in this disease setting.

To directly address this point, we compared equal numbers (1 × 10⁶ cells per recipient) of FACS-purified WT or Itk^-/- CD4⁺CD25⁺Foxp3⁺ Tregs following transfer into pristane-challenged WT recipients. WT Tregs provided measurable protection, consistent with previous studies. Importantly, however, Itk^-/- Tregs provided substantially greater protection, including reduced pulmonary pathology and proteinuria and improved survival (Figures 5 and 6). Thus, the functional advantage of Itk^-/- Tregs is supported by disease outcomes beyond systemic cytokine measurements.

We further performed CD45.1/CD45.2 congenic tracking and observed greater accumulation/recovery of donor-derived Itk^-/- Tregs in the lungs and draining lymph nodes compared with WT donor Tregs (Figure 5F–J). In addition, Itk^-/- Treg-treated recipients showed a greater reduction in IFN-γ and increased IL-10 compared with WT Treg-treated recipients. Canonical Tregs isolated directly from Itk^-/- lungs also expressed substantially higher levels of AREG, identifying a tissue-repair-associated phenotype (Figure 5K).

We have also more clearly incorporated our previous functional study demonstrating enhanced suppressive activity of Itk^-/- regulatory T-cell populations in in vitro and in vivo assays (Mammadli et al., Clinical and Translational Medicine, 2021; PMID: 34919342).

Together, these findings demonstrate that the advantage of Itk^-/- Tregs over WT Tregs is supported by greater protection after equal-dose transfer, enhanced accumulation in disease-relevant tissues, differential regulation of IFN-γ and IL-10, and increased AREG expression. We have clarified that AREG is associated with this protective phenotype but has not been shown here to be required for protection.

Minor comments

(1) In Figure 3, the authors need to provide important experimental control as H & E lung and spleen images under homeostatic conditions, without pristane.

We thank Reviewer #3 for this important suggestion. We agree that homeostatic histological controls are necessary to determine whether the differences observed after pristane challenge reflect pre-existing tissue abnormalities. In the revised manuscript, we have added H&E analyses of lungs and spleens from naïve, unchallenged WT and Itk^-/- mice. Both genotypes exhibited normal baseline tissue architecture, with no evidence of pulmonary hemorrhage or abnormal splenic pathology (Figure 3—figure supplement 1A–D and Figure 3—figure supplement 3A–C). These controls confirm that the pulmonary and splenic abnormalities observed after pristane administration arise during disease rather than from pre-existing histological differences between WT and Itk^-/- mice.

(2) In Supplementary Figure 6, the axis legend needs correction.

We sincerely thank Reviewer #3 for the careful review and for identifying this error. We have corrected the axis legend accordingly. The corrected graph is now presented as Supplementary Figure 7 (page 56) in the revised manuscript.