C57BL/6J (wild-type, WT) mice were purchased from the Jackson Laboratory (Bar Harbor, Maine). ST2 KO mice, a gift from Dr. A. McKenzie (Medical Research Council as part of UKRI, U.K), were bred for experimental use at the University of Pittsburgh. Mice were maintained under a 12h – 12h light-dark cycle with food and water provided at libitum. All animal procedures were approved by the University of Pittsburgh Institutional Animal Care and Use Committee and performed following the Guide for the Care and Use of Laboratory Animals. For Treg depletion, 300 μg anti-CD25 antibody (Thermo Fisher, Pittsburgh, PA) was diluted in PBS and injected ip 2 days prior to CCI. Control mice received the same amount of IgG2a isotype antibody.

TBI was induced by unilateral controlled cortical impact (CCI) as previously described (14). Briefly, mice were randomly assigned to sham and TBI groups with different treatments. All animals were anesthetized with 1.5% isoflurane in a 30% O₂/70% N₂O mixture under spontaneous breathing. Mice were drilled with a right parietal craniotomy (diameter of 3.5 mm; centered 0.5 mm anterior and 2.0 mm lateral to Bregma). The CCI was produced with a pneumatically driven CCI device (Precision Systems and Instrumentation) using a 3 mm flat-tipped impounder (velocity, 3.75 m/s; duration, 150 ms; depth, 1.5 mm). After CCI, the skin incision was sealed. Rectal temperature was maintained at 37°C ± 0.5°C during surgery and for up to 30 minutes after CCI using a heating pad. Sham animals were subjected to all aspects of the protocol (surgery, anesthesia, craniotomy, recovery) except for CCI. CCI animal models and all outcome assessments were performed by investigators who were blinded to the mouse genotype and the group assignments.

Animals were randomly assigned to intranasally receive recombinant IL-33 (Enzo, 2μg/30g body weight), or vehicle control treatment at 2h, 24h, and 48h after CCI. Mice were anesthetized as described above and placed in a supine position. The solution was applied into each nostril 3 times (~5 μl into each nostril for each infusion) with an interval of 5 minutes.

Animals were euthanized and perfused with saline followed by 4% paraformaldehyde (Sigma-Aldrich) in phosphate-buffered saline (PBS). Brains were cryoprotected in 30% sucrose in PBS. Coronal brain sections (25 μm) were sliced on a freezing microtome (Microm HM 450). Brain sections were subjected to crystal violet staining. Tissue loss was determined using Image J software by an observer blinded to group assignments. The actual tissue loss volumes were calculated as the volume of the contralateral hemisphere minus the non-injury volume of the ipsilateral hemisphere.

Brain sections were blocked with 5% donkey serum in PBS for 1 hour, followed by overnight incubation (4°C) with the primary antibodies. After washing, sections were incubated for 1 hour at 20°C with secondary antibodies conjugated with fluorophores (1:1000, Jackson ImmunoResearch Laboratories, Inc.). Fluorescence images were captured with an Olympus Fluoview FV1000 confocal microscope and FV10-ASW 2.0 software (Olympus America). Primary antibodies used in this study include: Goat anti-IL-33 (R&D System), mouse anti-APC (MilliporeSigma), rabbit anti-GFAP (Dako), mouse anti-NeuN (MilliporeSigma), and rabbit anti-Iba1 (Wako).

Peripheral blood was collected by cardiac puncture. ACK lysis buffer (Sigma- Aldrich) was used to lyse RBCs. Brains were dissected and the ipsilateral hemispheres were collected. Single-cell suspensions were prepared by using the Neural Tissue Dissociation Kit (Miltenyi Biotec), according to the manufacturer’s instructions. The suspension was passed through a 70-μm cell strainer (Thermo Fisher Scientific) and resuspended in 30% Percoll. Cells, myelin and debris were stratified on a 30–70% Percoll gradient (GE Healthcare BioSciences). Cells at the interface were collected and then washed with FACS buffer (1% penicillin/streptomycin antibiotic, 2% fetal bovine serum, 2 mM EDTA in HBSS) before staining. Cells were first incubated with antibodies to surface antigens for 30 minutes on ice at 4°C in the dark. After two washes, cells were permeabilized and fixed with the Intracellular Staining Kit (Thermo Fisher Scientific), and then stained with fluorophore-labeled antibodies for 30 minutes on ice in the dark. Fluorochrome compensation was performed with single-stained UltraComp eBeads (Thermo Fisher Scientific). Flow cytometry was performed on the BD LSRII flow cytometer (BD Biosciences). Data analyses were performed using FlowJo software. The following antibodies (Thermo Fisher Scientific) were used: anti-CD4 FITC, anti-ST2 APC, anti-CD25 PE, anti-Foxp3 Percp-cy5.5, anti-CD11b APC-cy7, anti-Ly6G FITC, anti-CD3 ef450, anti-CD11c Percp-cy5.5, anti-GITR PE Cy7, anti-CTLA4 BV421, and anti-F4/80 PE.

Blood plasma, brain lysates and cell culture media were collected. Protein concentrations were measured with commercial ELISA quantification kits for IL-10 and TGFβ (R&D Systems) according to the manufacturer’s instructions.

Spleens were harvested from WT or ST2 KO mice 5d after CCI to prepare single cell suspensions as we described (16, 17). CD4⁺CD25⁺ Tregs were isolated using a mouse Treg isolation kit (Miltenyi Biotec) according to the manufacturer’s instructions. The isolation was performed in a two-step procedure with a negative selection on CD4⁺ cells and a positive selection on CD25⁺ cells. Isolated Tregs were stimulated with soluble anti-CD3 (4 µg/ml), anti-CD28 (5 µg/ml) and IL2 (100 ng/ml) for 2d in Treg culture media (RPMI1640 containing 2mM L-Glutamine, 10%FBS, 1% penicillin/streptomycin, 1mM pyruvate sodium, and 55 μm β-mercaptoethanol) and then treated with IL-33 (R&D system, 50 ng/mL) or PBS for 24h.

T effector cells (Teff) were plated at 2 × 10⁵ per well in a U bottom 96-well plate in the presence of anti-CD3 and anti-CD28 to stimulate their proliferation. IL-33 treated or PBS-treated Tregs were added at a ratio of 1:1, 1:2, 1:4, 1:8, and 1:16 to the number of Teff. Cells were incubated for two days. Suppression of Teff proliferation was determined using the BrdU cell proliferation ELISA kit (Roche) according to the manufacturer’s instructions.

All statistical analyses were performed using GraphPad Prism software (version 9.0.0, La Jolla, CA). The Student’s t-test was used for comparison of two groups for continuous variables with normal distributions. The differences in means among multiple groups were analyzed using one-way analysis of variance (ANOVA). Differences in means across groups with repeated measurements over time were analyzed using the two-way repeated measures ANOVA. When the ANOVA showed significant differences, pairwise comparisons between means were tested by post hoc Bonferroni multiple-comparison tests. Results are presented as mean ± SEM. In all analyses, p<0.05 was considered statistically significant.

We evaluated IL-33 expression in the brain and the blood after CCI. Immunostaining showed that the number of IL-33⁺ cells increased significantly 5d after CCI (Figures 1A, B). ELISA results confirmed elevated IL-33 levels in the ipsilateral brain (Figure 1C) and in the plasma (Figure 1D) 1d and 3d after CCI. Co-labeling with cell specific markers demonstrated IL-33 protein expression mainly in APC⁺ oligodendrocytes in the lesioned area (Figure 1E). Some IL-33⁺GFAP⁺ astrocytes were noted (Figure 1E). There is no IL-33 staining in Iba1⁺ microglia/macrophages or NeuN⁺ neurons (Figure 1E).

We then assessed changes in immune cell composition in blood after TBI (Figure 1F). We did not see significant changes in the numbers of CD3⁺ total T lymphocytes (Figure 1G), CD19⁺ B lymphocytes (not shown), or CD11b⁺ myeloid cells (Figure 1H). The percentages of CD11b⁺CD11c⁺ dendritic cells, CD11b⁺Ly6G⁺ neutrophils or CD11b⁺F4/80⁺ macrophages remained unchanged among total CD11b⁺ myeloid cells 5d after TBI (Figure 1H). Meanwhile, the percentages of CD4⁺ T cells and CD4⁺CD25⁺Foxp3⁺ Tregs showed no change in blood (Figure 1G). Interestingly, the percentage of ST2⁺ Tregs significantly increased in the blood 5d after TBI (Figure 1G). There is no significant change in ST2 expression in total CD4⁺ T cells (not shown).

We then assessed the infiltration of Tregs in the TBI brain. Infiltrating Tregs were observed in the brain 3d after CCI, and further increased 7d after CCI. The number of ST2⁺ Tregs significantly increased in the brain 7d after CCI (Figures 2A, B). ST2 expression in infiltrating Tregs increased since 7d after CCI (Figures 2A, B).

Taken together, these results demonstrate that IL-33 is released in the brain and in the circulation after TBI, which is accompanied by ST2 upregulation in circulating and brain infiltrating Tregs.

We then used ST2 knockout mice to investigate the function of IL-33/ST2 axis after TBI. ST2 KO mice exhibited enlarged brain lesion size 5d after CCI (Figure 2C). ST2 deficiency also resulted in deteriorated sensorimotor deficits, as revealed by the increased time to remove an adhesive tape in the adhesive removal test (Figure 2D), increased error rate in the foot fault test (Figure 2E), lower score in hanging wire test (Figure 2F), reduced latency to fall off a rotating bar in the rotarod test (Figure 2G), and higher asymmetric rate in the cylinder test (Figure 2H). Flow cytometry results showed no significant differences in the numbers of circulating Tregs, as well as total T lymphocytes, CD4⁺ T lymphocytes, CD11b⁺ myeloid cells, or CD11b⁺Ly6G⁺ neutrophils between ST2 KO sham mice and WT sham mice (Figure S1). However, significantly reduced ST2 expression in circulating Tregs was observed, which was accompanied by a reduced number of Tregs in the blood in ST2 KO mice vs. wild-type (WT) mice 5d after CCI (Figure 2I). The percentages of CD11b⁺CD11c⁺ dendritic cells increased significantly in ST2 KO mice after CCI (Figure 2J). The numbers of other immune cells were similar between the two genotypes (Figure 2I, J). Consistent with the drop in Treg numbers, the expression of IL-10 (Figure 2K) and TGFβ (Figure 2L), two anti-inflammatory cytokines released by Tregs, was significantly reduced in ST2 KO mice compared to WT mice 5d after CCI. ST2 deficiency also reduced the numbers of brain infiltrating Tregs and ST2⁺ infiltrating Tregs 5d after CCI (Figure 2M).

We then treated mice with IL-33 (2 μg/30 g body weight) intranasally starting 2h after CCI and repeated for 2 consecutive days (Figure 3A). In contrast to ST2 deficiency, the brain lesion size significantly reduced in IL-33-treated mice compared to PBS-treated mice (Figure 3B). The supplementation of IL-33 improved sensorimotor functions in some behavioral tests including the adhesive removal test (Figures 3C, D), the hanging wire test (Figure 3F), and the cylinder test (Figure 3H) up to 5d after CCI. IL-33 did not significantly improve the performance in foot fault test (Figure 3E) or rotarod test (Figure 3G). IL-33 treatment did not change the number of circulating Tregs 5d after TBI (Figure 3I) but increased the percentage of ST2⁺ Tregs and the expression level of ST2 on Tregs in the blood (Figure 3I). The percentages of CD11b⁺CD11c⁺ dendritic cells, CD11b⁺Ly6G⁺ neutrophils and CD11b⁺F4/80⁺ macrophages remained the same among total CD11b⁺ myeloid cells in IL-33 and PBS-treated mice 5d after TBI (Figure 3J). IL-33 increased the levels of IL-10 (Figure 3K) and TGFβ (Figure 3L) in the blood 5d after TBI. IL-33 treatment increased the number of brain infiltrating Tregs (Figure 3M) and ST2⁺ infiltrating Tregs (Figure 3N). The expression levels of CD25, CTLA4 and GITR on infiltrating Tregs did not show significant differences 5d after TBI (Figure 3O).

These data suggest that IL-33 reduces brain lesion and sensorimotor deficits after CCI and enhances anti-inflammatory response, possibly by upregulating ST2 expression on Tregs.

Since ST2 is widely expressed in multiple types of immune cells, we next confirmed whether Tregs are essential for the protective effect of the IL-33/ST2 signaling early after TBI. Tregs were depleted by anti-CD25 injection 48 hours prior to TBI (Figure 4A). As expected, anti-CD25 injected mice maintained lower numbers of Tregs in the circulation regardless of IL-33 treatment 5d after CCI (Figure 4B). Treg depletion enlarged brain lesion size and exacerbated sensorimotor deficits, and IL-33 injection to Treg-depleted mice lost its protective effect (Figures 4C–I). The expression of IL-10 (Figure 4J) and TGFβ (Figure 4K) remained in low levels in Treg depleted mice with or without IL-33.

To elucidate the direct effect of IL-33 on Tregs, we isolated Tregs from the spleens of WT and ST2 KO mice at 5d after CCI and maintained in culture media with anti-CD3 and anti-CD28 for 2 days. Activated Tregs were then cultured for 24h with or without IL-33 treatment. T effector cells (Teff) were used as control. As shown in Figures 5A, B, there were higher levels of IL-10 and TGFβ in the conditioned media collected from Tregs compared to Teff cells. IL-33 treatment further increased the expression of these two cytokines. IL-33 failed to induce IL-10 or TGFβ production in ST2 KO Tregs (Figures 5A, B).

The capacity of Tregs to inhibit the proliferation of Teff cells was measured by the BrdU cell proliferation kit (Figure 5C). IL-33-treated WT Tregs demonstrated enhanced ability to inhibit the proliferation of Teff cells, as shown by significantly reduced BrdU incorporation (Figure 5C).

We also measured the expression of effector molecules CTLA4, GITR and CD25. IL-33 treatment enhanced the expression of CTLA4 and GITR in cultured Tregs but showed no effect on CD25 expression (Figure 5D).