IL-33 increases the magnitude of the Trm response in intestinal tissues during local infection

Abstract

IL-33 plays an important role in the early programming of CD8 T cells; however, its contribution to the differentiation of tissue-resident memory (Trm) cells in vivo remains poorly defined. After infection of mice with Y.pseudotuberculosis, IL-33 expression was increased in the intestinal tissue and this coincided with the expression of ST2 on T cells infiltrating the intestinal epithelium and lamina propria. Blocking IL-33 signaling after T cells infiltration of the intestinal tissue did not significantly impact the number or phenotype of Trm cells generated. However, overexpression of ST2 on T cells was able to increase expression of TCF1 and T cell number in the intestine compared to the lymphoid organs during infection. We also observed enhanced accumulation and maintenance of ST2 overexpressing cells in the intestine after infection was resolved. This points to a role for IL-33 in increasing the number of T cells that commit to intestinal tissue residency in vivo.

Introduction

IL-33 is a member of the IL-1 family of cytokines that also includes IL-1b, IL-1a, and IL-18. Under homeostatic conditions, IL-33 is expressed but remains localized within the nucleus, and upon cellular damage or stress, this ‘alarmin’ is released from the cell (1). The receptor for IL-33 (ST2) is expressed on a variety of immune cells and is critical for type 2-associated immune function (2, 3). ST2 is also upregulated on CD4 and CD8 T cells during type 1 immune responses, albeit at a lower level and more transiently than what is observed on type 2 immune cells, and this requires the transcription factors T-bet and STAT4 (4–6). After IL-33 binding, ST2 pairs with IL1RAcP to trigger activation of NF-KB and MAPKs via the MyD88/TRAF6 signaling axis (3). While these signaling components are activated downstream of many other receptors expressed by T cells, IL-33 has been shown to play an important cell-intrinsic role in CD4 and CD8 T cell differentiation during acute viral infection, positively regulating T cell expansion and promoting both effector function and memory differentiation (4, 5, 7). In addition, more recent work has shown IL-33 signaling can promote CD8 T cell stemness and enhance the magnitude of the T cell response during chronic viral by promoting the expression of TCF1 (7, 8). This early requirement for IL-33 in CD8 T cell differentiation has made analysis of its involvement during later stages of infection difficult, and this includes differentiation of tissue-resident memory T cell (Trm) populations.

Trm cells are a subset of memory T lymphocytes that take up residence in nonlymphoid tissues and provide more rapid, localized responses against pathogen reencounter when compared to their circulating counterparts (9). IL-33 is thought to play an important part in the differentiation of Trm cells through its ability to downregulate expression of the transcription factor KLF2 and its downstream effector molecules S1PR1 and CD62L, preventing tissue egress (10). IL-33 stimulation can also enhance expression of CD69 (10, 11), which is implicated in Trm recruitment and maintenance in the skin, kidney, and female reproductive tract (12–15). TCF1 has been shown to restrict CD103 expression in the lung (16), suggesting IL-33 driven TCF1 expression may negatively impact Trm cell differentiation; however, Tcf7 remains high in some Trm subsets and may promote the fitness of these populations (17–19). There is some data indicating IL-33 alone is sufficient to impact T cell retention and differentiation in vivo. Blocking ST2 in influenza-primed mice reduced the frequency of T cells in the lung parenchyma (20), and exogenous IL-33 enhanced vaccine-induced accumulation of Tcm and Trm cells (8). Additionally, the expression of IL-33 in tumors drives ST2-dependent accumulation of CD103⁺ tumor-resident CD8⁺ T cells (21). However, the cell intrinsic role of IL-33 signaling in regulating the magnitude and function of intestinal Trm populations has not been investigated.

IL-33 is expressed in the intestine at steady state; however, its expression is increased upon overt intestinal tissue damage or type 2 immune insults like helminth parasite infection (22–24). Many tissue cells are capable of producing IL-33, including epithelial cells, fibroblasts, macrophages/DCs (23–26). Here we demonstrate that infection with an intestinal bacterial pathogen increased IL-33 production, particularly by intestinal myofibroblasts. Using IL-33 blockade and ST2 overexpression after CD8 T cell activation, we were able to determine the impact of IL-33 on Trm cells without altering the abundance of circulating effector and memory T cell populations. We found that forced expression of ST2 on T cells infiltrating the intestinal tissue had a positive impact on Trm accumulation and differentiation. These data support the critical role of IL-33 in enhancing Trm lodgment and differentiation.

Materials and Methods

Mice and Infections.

C57BL/6 and B6.129P2(Cg)-Cx3cr1^tm1Litt/J mice were purchased from Jackson Laboratories. YopE TCR transgenic CD8 T cells specific for Yersinia pseudotuberculosis epitope YopE_69–77 were generated and bred in house (18). Both male and female mice were used in each experiment. For T cell transfers, naïve CD8 T cells were isolated using a CD8 T cell enrichment kit (Miltenyi Biotec; 130–104-075) with the addition of anti-CD44-biotin (ThermoFisher Scientific; IM7). Mice received 1×10⁵ YopE T cells intravenously one day prior to infection with ΔyopM Yersinia pseudotuberculosis YPIII (Yptb). Prior to infection, Yptb was cultured in LB broth for 24 hours at 26ºC, bacterial cells were collected, and resuspended in PBS. Mice were given 2×10⁸ Yptb in 200ul of PBS by oral gavage. Where indicated, mice were treated intraperitoneally with 4μg of Ig-Fc or ST2-Fc (R&D Systems; 1004-MR).

Isolation of Cells.

To discriminate between intravascular and tissue-resident T cells, 3μg of anti-CD8α-Alexa Fluor 700 (ThermoFisher Scientific; 53–6.7) was administered intravenously 10 minutes before the mice were sacrificed. For isolation of the intestinal T cells, the Peyer’s patches were removed, the small intestine was cut longitudinally, cleaned to remove intestinal contents, and cut into 1 cm pieces. Intraepithelial lymphocytes were isolated using 1mM dithiothreitol, epithelium was removed with 1.3 mM EDTA, and LP cells were isolated by digestion with 150 U/ml collagenase type 2 (Worthington Biochemicals, CLS-2). Intestinal cells were further purified by gradient centrifugation with 44% and 67% Percoll (Cytiva; 17089101). Spleen and MLN cells were isolated by homogenizing tissue using a 100 μm screen and plunger and red blood cells were lysed to generate a single cell suspension.

Immunofluorescence microscopy.

The ileum was harvested and frozen in Tissue-Tek O.C.T. media (Sakura Finetek; 4583). When GFP-expressing cells were present, tissues were fixed in 2% paraformaldehyde and rehydrated in 20% sucrose prior to freezing in O.C.T. Tissue sections were cut at a thickness of 7–8μm and fixed in ice cold acetone before storage at −80°C. Tissue sections were treated with an Avidin/Biotin Blocking Kit (Vector Laboratories; SP2001) and stained with the following reagents: anti-Madcam-biotin (BioLegend; MECA-376), polyclonal goat anti-IL-33 C-terminal domain (R&D Systems; AF3626); rabbit anti-SMA (Abcam; ab5694), anti-goat AF647 (Thermo Fisher Scientific; A21447), anti-rabbit AF555 (ThermoFisher Scientific; A27039), streptavidin-AF555 (Thermo Fisher Scientific; S32355), and 1μg/mL DAPI (Millipore Sigma; D9542), and mounted with Prolong Gold antifade mounting media (ThermoFisher Scientific; P36934). Stained slides were imaged using a Keyence fluorescence BZ-X series microscope and analyzed using Adobe Photoshop software. Quantification of the number of IL-33⁺ cells in the intestinal tissue was performed using ImageJ software. software with the Fiji plugin package (27). Automated counting was performed using the thresholding algorithm ‘Moments’ (28) to determine cells staining positive for IL-33 and DAPI. For each mouse, 2–4 representative images were counted and the ratio of IL33⁺ cells to DAPI⁺ nuclei was calculated. The average of the ratios of each mouse was graphed.

Flow cytometry.

Single cell suspensions were stained with the following antibodies/stains, all from ThermoFisher Scientific unless otherwise indicated: CD8α (53–6.7), CD8β (H35–17.2), CD45.1 (A20), CD90.1 (HIS1), CD69 (H1.2F3), CD103 (2E7), KLRG1 (2F1), IL7Ra (eBioSB/199), ST2 Biotin (RMST2–2), streptavidin-AF647 (S32357), CCR9 (eBioCW-1.2), CD4 (RM4–5), FoxP3 (FJK-16s), and TCF1 (Cell Signaling; C63D9) in the presence of anti-mouse CD16/CD32 (Bio X Cell; 2.4G2). YopE_69–77 MHC class I tetramer was provided by the NIH Tetramer Core. Cells were also stained with LIVE/DEAD Fixable Near-IR stain (ThermoFisher Scientific; L10119). Cells were analyzed on an Attune Nxt (ThermoFisher Scientific) flow cytometer and analyzed with FlowJo software (BD Biosciences).

Retroviral transductions.

Platinum-E (Plat-E) cells were cultured in DMEM supplemented with 10% FBS, HEPES, L-glutamine, 2-mercaptoethanol, gentamicin sulfate, penicillin/streptomycin and cultured at 37 °C and 6% CO₂. Plat-E were transfected with control MigR1 plasmid (EV) or MigR1-ST2 (ST2^OE), which contained the coding sequence of ST2 containing the transmembrane domain, using Lipofectamine 3000 Transfection Reagent (ThermoFisher Scientific; L3000001) per the manufacturer’s instructions. Six hours after transfection, media was replaced with fresh media and 48 hours after transfection, supernatant was collected and supplemented with 6ug/ml polybrene (Millipore Sigma; TR-1003) and 25U/ml IL-2 (Peprotech; 212–12), and virus was introduced into T cells activated with plate bound anti-CD3 (2μg/ml, Biolegend; 2C11) and CD28 (Biolegend; 37.51) 24 hours prior to transduction. Transduction was done by spin infection at 32ºC for 1 hour and following the spin, media was replaced with RPMI supplemented with 10% FBS, HEPES, 2-mercaptoethanol, L-glutamine, and 25U/ml IL-2. Twenty-four hours after transduction, T cells were isolated and 3–5×10⁵ each GFP⁺ EV and ST2^OE YopE T cells were transferred intravenously into mice that were infected 3 days earlier with ΔyopM Yptb.

In vitro IL-33 stimulation.

Retrovirally transduced cells were stimulated with 10ng/ml IL-33 (Peprotech; 210–33) for 20 hours and CD69 expression was measured by flow cytometry or GFP⁺Dapi^– cells were collected by FACS for gene expression analysis. RNA was isolated using Nucleospin RNAplus kit (Macherey-Nagel; 740984) and cDNA was generated using EcoDry Premix (Double Primed) (Takara; 639549). Quantitative PCR was performed on cDNA using TB Green Advantage GC qPCR Premix (Takara; 639676) and primers for S1pr1, Klf2, and Actb (10, 29).

Statistics.

Statistical analysis was performed with Prism 10.0 software (GraphPad Software). Paired/unpaired two-tailed Student’s t-test or one-way/two-way ANOVA were used with corrections for multiple tests as appropriate. P values less than 0.05 were considered significant.

Results

Infection induces IL-33 production by intestinal myofibroblasts

IL-33 is constitutively expressed at low levels within the intestinal tissue, and we examined the baseline expression of IL-33 in the ileum of naïve mice by immunofluorescence microscopy. In line with previously published data, we found there was modest expression of IL-33 in uninfected mice, and IL-33⁺ cells were found primarily in in the LP (Fig. 1A). Infection or intestinal tissue damage can lead to upregulation and release of IL-33 (24, 30–32); therefore, we examined IL-33 expression in the intestinal tissue in response to Yersinia pseudotuberculosis (Yptb) infection. On day 7 post infection (p.i.), we found an increase in the frequency of cells expressing IL-33 within the tissue, and again these cells were localized in the lamina propria (LP) but were now more often found near the crypts and rarely observed in the villi (Fig. 1A). The number of IL-33⁺ cells was quantified over the course of infection and significantly increased at days 3 through 14 p.i. relative to uninfected mice (Fig. 1B).

Figure 1: Increased IL-33 expression by myofibroblasts during intestinal infection.

C57BL/6 mice received ΔyopM Yptb by oral gavage and ileal tissue was harvested for analysis at 3, 5, 7, 9, and 14 days p.i. and compared to uninfected mice. Representative images from day 7 p.i. are shown. (A) IL-33 (red) and nuclei (blue) in uninfected control (left) and Yptb-infected mice (right). (B) Quantification of the number of cells expressing IL-33 per 10⁶ nuclei from 3–4 mice/group. (C) Localization of IL-33 (green) and myofibroblasts (SMA⁺, red) in uninfected (left) and day 7 Yptb infected (right) mice. Inset shows LP with IL-33-expressing SMA⁺ cells indicated with arrowheads. Representative images of 5 mice. Statistical analysis was done using a one-way ANOVA with Holm-Šidak test for multiple comparisons. *p<0.05

IL-33 can be expressed by a variety of cell types in the intestine during homeostasis and disease (33), including epithelial cells, fibroblasts, and immune cells including macrophages and dendritic cells (23–26, 32). We utilized markers of these cell types to identify the sources of IL-33 in the LP. IL-33⁺ cells were primarily smooth muscle actin-expressing (SMA⁺) myofibroblasts in both infected and uninfected control mice, and little expression of IL-33 above baseline was observed in intestinal epithelial layer (Fig. 1C). We found that only a small number of CX3CR1⁺ macrophages expressed IL-33 during infection (Supplemental Fig. 1A). While murine blood endothelial cells do not constitutively express IL-33 (25), we examined whether IL-33 is upregulated on inflamed vasculature during infection and found that MadCAM-1⁺ cells did not express detectable levels of IL-33 (Supplemental Fig. 1B). These data indicated IL-33 is produced primarily by LP myofibroblasts during infection, and this coincides with T cell infiltration into the tissue and Trm differentiation in the Yptb infection model (34).

Transient expression of ST2 on T cells during intestinal infection

ST2 is upregulated on effector CD8⁺ T cells during LCMV infection and plays an important role in regulating the magnitude and quality of the T cell response (4); however, the expression of ST2 on tissue infiltrating CD8⁺ T cells and its role in Trm differentiation have not been well established. To examine the dynamics of ST2 expression on T cells during intestinal infection, mice received congenically marked TCR transgenic Yptb-specific YopE_69–77 CD8⁺ T cells (YopE T cells) and were then infected with Yptb (18). ST2 expression was analyzed at various timepoints p.i. on YopE T cells from the intestinal epithelium (IE), lamina propria (LP), mesenteric lymph nodes (MLN), and spleen (SP) and compared to CD44^lo CD8 T cells in the MLN and SP from infected mice (Fig. 2A). We found that the percentage of ST2⁺ cells was significantly increased among YopE T cells from the MLN and SP on day 5 p.i. when compared to CD44^lo T cells (0.083% ±0.025% ST2⁺) (Fig. 2B). In the MLN, ST2 expression was reduce by day 7 p.i. on YopE T cells and no longer significantly different than CD44^lo CD8 T cells, while the percentage of ST2⁺ SP YopE T cells remained elevated at day 7 p.i. and then returned to baseline on day 10 p.i. There are few if any YopE T cells found in the intestine on day 5; however, by day 7 p.i. YopE T cell had infiltrated the IE and LP and expressed significantly more ST2 than CD44^lo T cells on both day 7 and day 10 p.i. (Fig. 2B). ST2 expression was also transient on YopE T cells in the intestinal tissue, and by 14 days p.i. ST2 levels were also approaching that on CD44^lo T cells. We also analyzed expression of ST2 on endogenous YopE_69–77 tetramer⁺ T cells in the LP and found similar expression levels compared to transferred YopE TCR transgenic T cells (Supplemental Fig. 2). These data indicate ST2 expression on Yptb-specific T cells is increased after priming and expansion in the lymphoid organs, peaking at 5–7 days p.i. T cells then traffic to the intestinal tissue, where the percentage of ST2⁺ T cells is highest from days 7–10 p.i.

Figure 2: Transient expression of ST2 on T cells during intestinal infection.

C57BL/6 mice received 10⁵ CD45.1⁺ YopE T cells and were then infected with ΔyopM Yptb and tissues were harvested at 5, 7, 10, 14, and 40 days p.i. for analysis. (A) Representative flow cytometry plots showing expression of ST2 on transferred YopE T cells and endogenous CD44^lo CD8 T cells at 10 days p.i. (B) Percentage of ST2⁺ YopE T cells in intestine, MLN, and SP at indicated timepoints. (C) Representative flow cytometry plots of CD69 and CD103 expression on ST2⁺ and ST2^– LP YopE T cells. Percentage of CD69⁺ (D) and CD69⁺CD103⁺ (E) YopE T cells in the LP at day 10 p.i., with lines connecting data points from individual mice. All data are pooled from 2–3 experiments with 3–5 mice per group. Statistical analysis comparing the percent ST2⁺ YopE T cells within each tissue to endogenous CD44^lo MLN CD8 T cells was performed using a one-way ANOVA with Holm-Šidak test for multiple comparisons and statistical significance for each timepoint is shown in the color corresponding to each organ. Paired Students’ t-test was used in (D,E). ns: not significant; *p<0.05; **p<0.005; ***p<0.0005.

IL-33 stimulation can lead to increased CD69 expression in CD8 T cells (11), and we examined whether the expression of ST2 correlated with upregulation of CD69. Similar percentages of ST2⁺ and ST2^– intestinal T cells expressed CD69 (Fig. 2C and 2D), and further analysis of CD69⁺ subsets in the LP revealed no difference in the distribution within the CD103⁺ and CD103^– subsets (Fig. 2C and 2E). These data indicate ST2 is expressed transiently on tissue infiltrating lymphocytes during Yptb infection and coincides with IL-33 upregulation within the intestine. Additionally, ST2 expression does not predict acquisition of Trm markers CD69 or CD103.

ST2 blockade during intestinal T cell infiltration does not impact the magnitude of the Trm response

To address the role of ST2 signaling on the magnitude of the Trm response, we transiently blocked IL-33 signaling during infection by administering a soluble form of ST2 (ST2-Fc). Mice received congenically marked YopE T cells and were infected one day later with ΔyopM Yptb. Mice received ST2-Fc or Ig-Fc every other day from day 6 to day 12 p.i. During this window, YopE T cells enter the intestine and begin to differentiate into Trm cells (34), ST2 expression is at its peak on tissue infiltrating cells (Fig. 2B), and IL-33 levels are elevated in the tissue (Fig. 1B). Tissues were harvested at 30 days post infection for analysis (Fig. 3A). We found there was no significant reduction in the number of YopE T cells in the intestine or MLN of mice after ST2-Fc treatment compared to the Ig-Fc control group (Fig. 3B). The expression of CD69 and CD103 was similar when comparing LP T cells from both ST2-Fc and Ig-Fc treated groups (Fig. 3C), and the percentage and number of LP CD69⁺CD103^– and CD69⁺CD103⁺ Trm cells was not significantly altered by ST2-Fc treatment (Fig. 3D). These data indicate impaired IL-33 signaling during early Trm differentiation has minimal impact on the magnitude of the T cell response in the intestinal tissue during Yptb infection.

Figure 3: Blocking IL-33 signaling as T cells infiltrate the intestine does not impact the magnitude of the Trm response.

(A) C57BL/6 mice received 10⁵ CD45.1⁺ YopE T cells and were then infected with ΔyopM Yptb. 4μg of ST2-Fc or Ig-Fc was administered intraperitoneally on days 6, 8, 10, and 12 p.i. and tissues were harvest on day 30 p.i. (B) Number of YopE T cells in the intestinal tissue and MLN. (C) Expression of CD103 and CD69 on YopE T cells from ST2-Fc or Ig-Fc treated mice, quantified in (D). Number of CD69⁺CD103⁺ and CD69⁺CD103^– Trm cells in the LP (E). (F-J) C57BL/6 mice received 10⁵ CD45.1⁺CD90.1⁺ YopE T cells and were then infected with ΔyopM Yptb. 4μg of ST2-Fc or Ig-Fc was administered intraperitoneally on days 0, 2, and 4 p.i. and tissues were harvested on day 5 p.i. (G) Representative flow cytometry plots of the percentage of transferred YopE T cells in the MLN and SP. (H) Quantification of the number of YopE T cells in the MLN and SP. (I) Representative flow cytometry plots showing the percentage of FoxP3⁺ CD4 T cells, with quantification of the percentage and number of FoxP3⁺ T cells in (J,K). Data are representative of 3 experiments with 3–5 mice per group (A-E) or pooled from 2 experiments with 5 mice/group (F-J). P values calculated using unpaired Students’ t-test with Holm-Šidak test for multiple comparisons in (B,D,E,H). *p<0.05.

To confirm that administration of ST2-Fc was effective in vivo, we administered ST2-Fc or Ig-Fc every other day during acute Yptb infection. ST2-Fc or Ig-Fc was administered on days 0, 2, and 4 p.i. and tissues were harvested on day 5 p.i (Fig. 3F). During Yptb infection, limiting IL-33 availability was able to significantly reduce the expansion of YopE T cells in the spleen, while YopE T cell numbers in the MLN were only moderately impacted (Fig. 3G and 3H). The ability of ST2-Fc to reach the intestinal tissue was also addressed. An ST2⁺ population of intestinal LP FoxP3⁺ regulatory T cells is maintained by IL-33 produced by DCs (23, 35). Treatment with ST2-Fc was able to significantly reduce the percentage and number of FoxP3⁺ CD4 T cells in the LP compared to Ig-Fc treated mice (Fig. 3I–3K). These data confirm that ST2-Fc treatment is indeed effective in blocking IL-33 in Yptb-infected mice.

Overexpression of ST2 in T cells increases T cell lodgment and Trm cell number

ST2-Fc administration likely has many effects that could obscure the role of IL-33 signaling in intestinal T cells, and for this reason, we moved to an ST2 overexpression system to limit our analysis to the impact of IL-33 directly on antigen-specific T cells. We generated a full length ST2 overexpression vector (ST2^OE), and after transduction of in vitro activated YopE T cells, we observed increased expression of surface ST2 compared to T cells transduced with empty vector (EV) (Fig. 4A). We assessed the ability of ST2 overexpression to enhance responsiveness to IL-33 by stimulating EV and ST2^OE YopE T cells with IL-33 and examining CD69 expression 20 hours later. Basal levels of CD69 were comparable in unstimulated ST2^OE and EV cells (Fig. 4B and 4C). However, after IL-33 stimulation there was a significant increase in the expression of CD69 on ST2^OE T cells compared to EV cells (Fig. 4B and 4C). IL-33 stimulation was also able to inhibit the expression of tissue egress markers Klf2 and S1pr1 (Supplemental Fig. 3), as previously described (10).

Figure 4: Overexpression of ST2 enhances responsiveness to IL-33 and does not impact intestinal trafficking.

Activated CD45.1⁺CD90.1⁺ YopE T cells were transduced with a full length ST2 overexpression (ST2^OE) vector and CD45.1⁺ T cells with control empty vector (EV) and surface expression of ST2 on GFP⁺ transduced cells was examined by flow cytometry. Representative histograms are shown in (A). EV and ST2^OE T cells were stimulated with 10ng/ml IL-33 for 20 hours and the expression of CD69 was examined by flow cytometry (B) and quantified in (C). (D-I) C57BL/6 mice were infected with ΔyopM Yptb and 3 days later received ~3–5×10⁵ each CD45.1⁺CD90.1⁺ YopE T cells containing ST2^OE vector and CD45.1⁺ cells with control empty vector (EV). Flow cytometry gating strategy to identify transferred cells and representative flow cytometry plots from indicated organs (E). Ratio of ST2^OE to EV YopE T cells in the LP, MLN, and SP at 2 days post transfer with a dotted line showing the input ratio (F). (G) Representative histogram of CD69 expression on transferred ST2^OE to EV YopE T cells in the LP compared to endogenous LP CD8 T cell populations. (H) Quantification of the gMFI of CD69 expression in the LP with lines connected data points from the same mouse. Data are representative of 2 experiments. Statistical analyses were performed using an unpaired Student’s t-test with Holm-Šidak test for multiple comparisons in (C), paired Student’s t-test (H), or one-way ANOVA (F). *p<0.0001. ND: not detected

To examine the role of ST2 in vivo, EV and ST2^OE YopE T cells were cotransferred into mice that were infected 3 days prior with ΔyopM Yptb. To examine entry into the intestinal tissue, tissues were harvested 2 days after T cell transfer for analysis (Fig. 4D). Few T cells had moved into the IE at this timepoint, so they were excluded from this analysis. Previous reports indicate that ST2 signaling drives increased expansion of lymphoid CD8 T cells during priming (4), and data from earlier ST2-Fc treatment studies support that this also occurs in the Yptb infection model (Fig. 3G, 3H). However, the ratio of ST2^OE to EV YopE T cells in lymphoid tissues slightly favored the EV population (Fig. 4E, 4F). The ratio of ST2^OE to EV YopE T cells in the LP was similar to that observed in the lymphoid organs (Fig. 4E, 4F). The expression of CD69 was examined on YopE T cells that had infiltrated the intestinal tissue (Fig. 4G), and there was no significant difference in CD69 gMFI when comparing ST2^OE and EV YopE T cells in the LP (Fig. 4H). These data suggest in vitro activation and transfer of T cells in these studies potentially circumvented the impact of IL-33 on T cell expansion and indicated ST2^OE does not impact recruitment into the intestinal tissue.

To examine more long-term impacts of ST2^OE on Trm differentiation, tissues were harvested at 9 and 45 days after infection and the abundance of EV and ST2^OE YopE T cells was examined (Fig. 5A). We found that ST2^OE YopE T cells were significantly increased relative to EV YopE T cells in both the IE and LP on day 9 post infection by an average of 2–3-fold relative to the spleen (Fig. 5A, 5B). The number of ST2^OE YopE T cells in the mesenteric LN was also significantly elevated compared to EV controls (Fig. 5A, 5B). Data was normalized to the ratio found in the spleen for individual mice; however, analysis of data without normalization to the spleen yielded similar results (Supplemental Fig. 4A). We also examined the expression of CD103 and CD69 in the LP at day 9 post infection and saw no difference in the proportion of cells that were CD69⁺CD103^– and CD69⁺CD103⁺ between the ST2^OE and EV groups (Fig. 5C). The increased frequency of ST2^OE T cells in the intestine was maintained at 45 p.i., with an average of 3-fold more ST2^OE T cells (Fig. 5A, 5B, Supplemental Fig. 4A). At this late timepoint, ST2^OE and EV cells were present at similar numbers in the MLN and spleen. After infection was resolved, we also found a modest increase in the frequency of CD69⁺CD103⁺ Trm cells in the LP (Fig. 5D). These data suggest after T cell activation, ST2 plays an additional cell intrinsic role in regulating the size of the intestinal Trm pool.

Figure 5: Overexpression of ST2 increases Trm frequency in the intestinal tissue.

C57BL/6 mice were infected with ΔyopM Yptb and 3 days later received ~3–5×10⁵ each CD45.1⁺CD90.1⁺ YopE T cells containing ST2^OE vector and CD45.1⁺ cells with control empty vector (EV). (A) Representative flow cytometry plot with percentages of EV and ST2^OE of total transduced GFP⁺ YopE T cells in intestinal epithelium (IE) and spleen (SP) at 9 and 45 days p.i. (B) Ratio of ST2^OE to EV T cells on days 5, 9, and 45 p.i. normalized to the ratio in the spleen for each mouse. Dashed line indicates tissue to spleen ratio of 1.0. Expression of CD103 and CD69 on ST2^OE and EV YopE T cells in the LP at days 9 (C) and 45 p.i. (D), with percentages of CD69⁺CD103⁺ and CD69⁺CD103^– Trm subsets. (E) Expression of TCF1 in IE and LP ST2^OE and EV YopE T cells, quantified in (F). Statistical analyses were performed using one sample Wilcoxan rank sum test (C) and paired Students’ t-test with Holm-Šidak test for multiple comparisons (D, F). *p<0.05, **p<0.01.

ST2 signaling drives differentiation of short-lived effector T cells (KLRG1⁺IL7R^–) and results in a reduction in the memory precursor pool (KLRG1^–IL7R⁺) (4), and this could negatively impact the generation of Trm cells, which are thought to be derived from cells with a memory precursor phenotype (36, 37). However, direct analysis of the effector phenotype of YopE T cells at day 9 p.i. showed ST2^OE had fewer memory precursor KLRG1^–IL7R⁺ in cells when compared to EV T cells (Supplemental Fig. 4B, 4C). CCR9 expression also correlates with increased intestinal Trm potential (38), and there was comparable CCR9 expression on EV and ST2^OE T cells in the circulation of infected mice (Supplemental Fig. 4D, 4E). These data suggest that ST2^OE in lymphoid organs and circulation do not have a phenotype indicative of enhanced Trm potential. We also examined the ability of IL-33 to impact TCF1 expression, which has previously only been observed in the context of chronic infection (7). At day 9 after infection, we found significantly elevated levels of TCF1 in ST2^OE cells compared to EV cells in the intestinal tissue but not the MLN or spleen (Fig. 5E, 5F). Overall, these data indicate ST2 overexpression and exposure to IL-33 impacts tissue retention, leads to elevated levels of TCF1, and enhances the overall size of the intestinal Trm pool.

Discussion

IL-33 has been speculated to play positive role in the differentiation of Trm cells by suppressing expression of the transcription factor KLF2 and tissue egress markers and enhancing expression of tissue retention marker CD69 (10, 11). However, there are other cytokines including TGF-β and IL-12 that can also suppress KLF2 expression (10, 34), and the importance of IL-33 alone on the Trm response in vivo has remained unclear. Here we demonstrate that IL-33 is produced by intestinal myofibroblasts during local intestinal infection and T cells transiently express ST2 upon entry into the intestinal tissue. Blocking IL-33 by administration of ST2-Fc was able to prevent CD8 T cell expansion in the lymphoid tissue immediately after infection. The overexpression of ST2 on effector T cells improved their accumulation in the tissue, increased TCF1 expression by tissue-infiltrating T cells, resulting in a larger intestinal Trm pool and increased in the proportion of cells in the CD103⁺ Trm subset.

It was surprising that ST2-Fc treatment targeted to the window of T cell entry into the intestinal tissue and onset of gene expression changes that facilitate tissue residency (days 6–14 p.i.) was not able to significantly impact the size of the intestinal Trm pool. However, it is possible that the transient expression of ST2 on only a subset of responding YopE T cells is not a robust enough system to definitively examine the role of IL-33 in this infection model. Alternatively, the complex role of IL-33 in the intestine during disease may be obscuring the impact of IL-33 on Trm cells. IL-33 regulates intestinal epithelial barrier function and differentiation of goblet cells and paneth cells (32). Monocytes are infiltrating the intestine in large numbers at 5 days post Yptb infection (34), and their differentiation into resident macrophages could be impacted by IL-33 as well. Our studies demonstrated that FoxP3⁺ regulatory T cells in the LP are reduced after ST2-Fc treatment, and this could increase inflammation overall and/or directly influence Trm differentiation (39, 40). Therefore, our targeted approach to overexpress ST2 on effector T cells is likely a more reliable indicator of the impact of IL-33 on Trm cell differentiation.

The findings presented here are consistent with previous studies that have addressed the impact of IL-33 on Trm cells during infection. IL-33 has been shown to support continued seeding of T cells in the lung tissue after influenza infection, with no change in the phenotype of the resulting Trm cells (20). However, changes in lung Trm subset distribution were examined after 7 days of anti-ST2 treatment (20), and we only observed changes in Trm phenotype at later timepoints. IL-33-deficient mice infected with chronic MCMV had decreased Trm cell numbers in the salivary gland compared to wild-type mice, and in these studies the proportion of cells in the CD103⁺ Trm subset was also decreased (8). During MCMV infection, IL-33 also drove accumulation of memory CD8 T cells in the spleen (8), making it difficult to determine when IL-33 is acting to promote Trm accumulation or just increasing the number of T cells overall. Our data using ST2 overexpression indicated IL-33 plays a role in the tissue, as the accumulation of ST2^OE T cells is restricted to the intestinal tissue and not seen in the lymphoid organs. These data indicate targeting the IL-33/ST2 signaling axis to increase Trm numbers and CD103⁺ Trm cell frequency could provide improved protection from secondary infection, as intestinal CD103⁺ subset of Trm cells have elevated cytotoxic function relative to other Trm subsets (18, 19), and the number of CD8 Trm cells in the tissue correlates with protection from secondary infection in a variety of organs (41–45).

IL-33 has the potential to regulate Trm cells at several levels, both by increasing the magnitude of the effector response and, once T cells have entered the intestinal tissue, by enhancing tissue retention and Trm differentiation. However, recent work indicates commitment to the Trm lineage can begin before entry into the tissue, with signals like TGF-β and retinoic acid in the lymphoid tissue predisposing cells to become Trm cells (36, 38, 46). We attempted to address whether IL-33 influences Trm commitment before entry into the intestinal tissue by examining phenotypic markers indicative of Trm precursors on ST2^OE after transfer into Yptb-infected mice; however, we saw no shift in phenotypes that would indicate improved Trm potential of circulating T cells overexpressing ST2. While suggestive, we cannot definitively exclude that IL-33 stimulation prior to entry into the intestine influences Trm differentiation. The mechanisms underlying Trm commitment in the lymphoid tissue is a developing field, and the role of IL-33 and other inflammatory cytokines in regulating this process requires further study.

Localization within the intestinal compartment, whether it be in the IE vs. LP, duodenum vs. ileum, or small intestine vs. colon, can influence the number, phenotype, and function of Trm cells (34, 47). As Trm cells are thought to be relatively restricted in their movement, the local cues that are received upon T cell entry into the tissue are critical for informing the functional capabilities and size of the Trm pool in these local areas prior to secondary challenge. Our data suggest that the exposure of tissue-infiltrating Trm precursors to the IL-33 rich regions near the crypts can influence their retention/persistence within the intestinal tissue resulting in an increased density of Trm cells.

Key points.

• IL-33 producing cells are increased during intestinal bacterial infection

• Overexpression of ST2 on CD8 T cells enhances intestinal Trm differentiation