The innate cytokine IL-33 is increasingly recognised to possess biological effects on various immune cells. We have previously demonstrated elevated serum level of soluble ST2 in patients with active systemic lupus erythematosus suggesting involvement of IL-33 and its receptor in the lupus pathogenesis. This study sought to examine the effect of exogenous IL-33 on disease activity of pre-disease lupus-prone mice and the underlying cellular mechanisms. Recombinant IL-33 was administered to MRL/lpr mice for 6 weeks, whereas control group received phosphate-buffered saline. IL-33-treated mice displayed less proteinuria, renal histological inflammatory changes, and had lower serum levels of pro-inflammatory cytokines including IL-6 and TNF-α. Renal tissue and splenic CD11b+ extracts showed features of M2 polarisation with elevated mRNA expression of Arg1, FIZZI, and reduced iNOS. These mice also had increased IL-13, ST2, Gata3, and Foxp3 mRNA expression in renal and splenic tissues. Kidneys of these mice displayed less CD11b+ infiltration, had downregulated MCP-1, and increased infiltration of Foxp3-expressing cells. Splenic CD4+ T cells showed increased ST2-expressing CD4+Foxp3+ population and reduced IFN-γ+ population. There were no differences in serum anti-dsDNA antibodies and renal C3 and IgG2a deposit in these mice. Exogenous IL-33 was found to ameliorate disease activity in lupus-prone mice with induction of M2 polarisation, Th2 response, and expansion of regulatory T cells. IL-33 likely orchestrated autoregulation of these cells through upregulation of ST2 expression.

© 2023 The Author(s). Published by S. Karger AG, Basel

Introduction

Interleukin (IL)-33 is an innate cytokine that possesses dual pro- and anti-inflammatory functions [1]. It is expressed by multiple cell types including fibroblasts, epithelial cells, and endothelial cells. Bioactive full-length IL-33 is released during cellular necrosis in inflammatory conditions as an alarmin [2]. This cytokine binds and signals via its receptor ST2 [3], a member of the IL-1 receptor family [4], activating the mitogen-activated protein kinase and nuclear factor-κB pathway. IL-33 was first characterised on its Th2-promoting effect since it was identified as a new member of the IL-1 cytokine family [5] and the ligand for ST2 in 2005 [3].

IL-33 plays a dual role in disease pathologies and is protective against T helper 1 (Th1) dominant diseases but exacerbates Th2- and mast cell-mediated inflammatory conditions [1]. IL-33 was shown to promote airway inflammation [6] and collagen-induced arthritis [7]. On the other hand, it was found to protect against intestinal inflammation [8], cardiac failure [9], atherosclerosis [10], and obesity [11]; ameliorate infective conditions such as pulmonary helminthiasis [12], cerebral malaria [13], and sepsis [14]; and prevent allograft rejection [15] and graft-versus-host disease [16]. IL-33 has increasingly been shown to regulate a diversity of immune cells among which regulatory T cell (Treg) was found to mediate the biological effects of IL-33 in some of these conditions [17]. Tregs are effector T cells that possess immunoregulatory function that is crucial for the maintenance of immune tolerance [18].

Systemic lupus erythematosus (SLE) is a multi-systemic autoimmune disease that affects predominantly young women [19] and poses treatment challenges despite emergence of biologic agents in recent years [20]. This condition arises from dysregulated innate and adaptive immune responses resulting in activation of autoreactive T and B lymphocytes with production of autoantibodies that form immune complexes which deposit in internal organs causing inflammation and damage [21]. Breakdown of immune tolerance related to reduced number or deficient function of Treg has been reported in these patients [22].

We have previously shown that soluble ST2 (sST2), a truncated form of the extracellular domain of membrane ST2, was significantly elevated in serum of SLE patients and correlated with lupus disease activity, suggesting a role of IL-33/ST2 axis in the pathogenesis of SLE [23]. Based on the reported effect of IL-33 on Treg expansion, we hypothesised that IL-33 possessed ameliorating effect on lupus disease. In this study, we demonstrated amelioration of disease in exogenous IL-33-treated MRL/lpr lupus-prone mice, which are known to have deficient Treg function [24], and examined the cellular mechanisms mediating the biological effects of IL-33.

Materials and MethodsAnimals

The study was approved by the animal Ethics Committees of the University of Hong Kong and the City University of Hong Kong. Female MRL/lpr lupus-prone mice (MRL/MpJ-Faslpr/J) (Stock 006825) were used in the experimental groups. Wild-type mice of the same genetic background MRL/MpJ (Stock 000486) and Balb/c were used in some experiments. Breeders of the mouse strains were acquired from the Jackson Laboratory (USA), and the colonies were maintained under pathogen-free conditions at the university animal facilities.

Experimental Design

Lyophilised recombinant murine IL-33 (BioLegend, USA) was reconstituted in sterile phosphate-buffered saline (PBS) with 1% bovine serum albumin (BSA) as stock solution which was diluted by PBS to 1 μg/100 μL and administered intraperitoneally to MRL/lpr mice 3 times per week for 6 weeks at the age of 12 weeks. Control group received injection of 100 μL of PBS only. Mice were monitored weekly on body weight, skin changes, and urinary protein for 7 weeks from baseline and were sacrificed for organ harvest 1–2 weeks after the last dose of injection (i.e., at the age of week 19–20) [25]. Skin in the dorsal neck region and the ears was scored semi-quantitatively (0 – none; 1 – minimal hair loss; 2 – redness, scabbing, and hair loss; 3 – extensive area of involvement) [26]. Urinary protein was examined semi-quantitatively by dipstick analysis (Healgen Scientific, USA) (0 – none; 1–30 to 100 mg/dL; 2 – 100 to 300 mg/dL; 3 – 300 to 1,000 mg/dL; 4 – >1 g/dL) and quantitatively by BCA protein assay (Bio-Rad, USA).

Flow Cytometry

Immune cells were isolated from the kidneys digested with collagenase IV (Sigma, USA) followed by Ficoll-Paque density gradient centrifugation. Splenocytes were isolated as described earlier [27] from which CD11b+ monocytes were extracted by magnetic isolation (Miltenyi Biotec, USA). Cells were incubated with mouse Fc block (Miltenyi) at 4°C followed by fluorochrome-labelled monoclonal antibodies and corresponding isotypic controls: anti-mouse CD3-APC, CD4-PE, CD4-APC, CD4-FITC, CD8-APCCy7, CD11c-FITC, CD11b-PE, CD25-PE, ST2-APC, CD206-PECy7 (BioLegend), CD19-PECy7, inducible nitric oxide synthase (iNOS)-APC (eBioscience, USA). For intracellular staining for interferon (IFN)-γ and IL-17, splenocytes were stimulated ex vivo by phorbol myristate acetate (50 ng/mL, Sigma) and ionomycin (1 μg/mL, Sigma) for 4 h in the presence of GolgiStop (eBioscience). After surface staining, Cytofix/Cytoperm kit (BD Biosciences) was used to permeabilise cells before staining with anti-IFN-γ-FITC, anti-IL-17A-PECy7, anti-iNOS-APC monoclonal antibody, and isotypic controls. Anti-mouse Foxp3 staining set APC (eBioscience) was used for Foxp3 immunostaining. Data were captured by flow cytometry (Beckman, USA) and analysed by the FlowJo software (Tree Star, USA).

Enzyme-Linked Immunosorbent Assays

Sera were collected from peripheral blood bi-weekly and by intracardiac puncture at study endpoint. In some experiments, isolated splenocytes were stimulated ex vivo by anti-CD3 antibodies (1 μg/mL) or graded doses of IL-33 as indicated. Cytokines including IL-6, tumour necrosis factor (TNF)-α, IL-1β, IFN-γ, IL-17, IL-12, IL-10, IL-4, IL-5, IL-13, immunoglobulin G (IgG) (all Thermo, USA), and IgG anti-BSA antibodies (MyBioSource, USA) were measured by commercial enzyme-linked immunosorbent assay kits. Anti-double-stranded (ds) DNA antibodies were detected by in-house enzyme-linked immunosorbent assay pre-coated with poly-L-lysine (Sigma) followed by calf thymus dsDNA (50 ng/mL, Sigma).

Real-Time Quantitative Polymerase Chain Reaction

Total RNA was extracted from tissues and isolated cells using TRIzol (Life Technologies). First-strand cDNA was synthesised using Superscript® III Reverse Transcriptase Kit (Life Technologies). The quantitative polymerase chain reaction assay was performed using SyBr Green PCR Reagent Mix (Life Technologies) and forward and reverse primers (BGI Health, Hong Kong) in StepOnePlus Real-Time PCR Systems as described previously [28]. Relative expression levels of target genes were calculated using 2−ΔCt formula and normalised to GAPDH.

Histology

Kidney sections were fixed in 10% neutral buffered formalin, paraffin-embedded and stained with haematoxylin and eosin, and periodic acid-Schiff, and were examined by Leica light microscope. Histologic features including glomerular inflammation, proliferation, crescent formation, necrosis, inflammatory infiltrate, tubular necrosis, and glomerulosclerosis were graded semi-quantitatively by a blinded observer (KW Chan) where 0 represents no damage to 4 – severe, and were summed up to a composite score [26]. Immunofluorescent staining with FITC-conjugated anti-mouse C3 (Abcam, USA) and IgG2a (BioLegend) was performed on frozen kidney sections. Images on at least 20 glomeruli at random positions were captured by fluorescence microscopy (Nikon, USA) and analysed by the ImageJ software. Fluorescence intensity is expressed as integrated density corrected by the area selected multiplied by mean fluorescence of background readings.

Statistics

Statistical analysis was performed by SPSS 27.0 (Chicago, USA). Comparisons between groups were performed by Student’s t test or ANOVA. Paired t test was used for comparison between pre- and post-treatment conditions and repeated measures ANOVA for comparison over time. Bonferroni correction was applied for multiple comparisons. Results are expressed in diagrams as means ± standard error of the mean. p value <0.05 is considered statistically significant. Statistical significance is depicted in the figures as *p < 0.05, **p ≤ 0.01, ***p ≤ 0.001.

ResultsIL-33 Ameliorated Lupus Disease Activity

Pre-disease MRL/lpr mice aged week 12 were given recombinant murine IL-33 for 6 weeks, whereas the control group received PBS only. Both groups had comparable gain in body weight (Fig. 1a). Control mice developed progressive hair loss and skin rash (Fig. 1b), whereas IL-33-treated mice had lower skin score at week 5 (0.92 vs. 0.24, p = 0.03) and week 6 (1.09 vs. 0.13, p = 0.02) after the first dose of injection. IL-33-treated mice also had significantly less proteinuria as detected by urine albustix at week 4 (p = 0.02), week 6, and week 7 (both p < 0.001) (Fig. 1c). Quantitative protein assay showed progressive worsened proteinuria in the control group at week 6 (p = 0.04) and at week 7 (p = 0.007) compared to week 5. On the other hand, IL-33-treated mice developed significantly less proteinuria at week 5 (1.0 ± 0.1 vs. 5.4 ± 7.6, p = 0.04), week 6 (1.4 ± 0.8 vs. 6.6 ± 4.6, p = 0.01), and week 7 (1.4 ± 1.5 vs. 8.3 ± 6.6, p = 0.02) compared with control mice (Fig. 1d).

Fig. 1.

Administration of IL-33 ameliorated disease severity in MRL/lpr mice. a IL-33-treated and control mice showed comparable gain in body weight. IL-33-treated mice had less severe skin score (b), proteinuria by urine albustix (c), quantified urine proteinuria by protein assay (d), and renal histological scores (e). f Histology of renal tissue in control mice showed more marked glomerular hypercellularity, cellular crescents (hollow arrows), and inflammatory infiltrates compared to IL-33-treated mice (original magnification ×200) (data from 9 mice per group for 3 rounds of experiments).

Histology of renal tissue of IL-33-treated mice demonstrated significantly lower scores for proliferative changes (1.7 ± 1.1 vs. 2.9 ± 0.3, p = 0.006), hyaline deposit (0 vs. 1.1 ± 0.8, p = 0.001), crescent formation (0 vs. 4.1 ± 5.5, p = 0.03), and overall composite score (2.7 ± 1.1 vs. 10.2 ± 5.2, p = 0.004) compared to the control group (Fig. 1e). There were marked glomerular hypercellularity, cellular crescents, and interstitial inflammation around the glomeruli in the control group, whereas IL-33-treated mice displayed only focal and mild increase in mesangial matrix (Fig. 1f).

IL-33 Reduced Pro-Inflammatory Cytokine Expression

To delineate the cellular mechanisms underlying the disease-ameliorating effect of IL-33, a panel of pro- and anti-inflammatory cytokines was examined. Renal tissue from IL-33-treated mice expressed significantly lower mRNA transcripts of IL-6 (1.0 ± 1.1 vs. 6.1 ± 5.7, p = 0.02), TNF-α (32.1 ± 14.7 vs. 77.0 ± 27.8, p < 0.001), and IL-1β (31.1 ± 10.1 vs. 77.8 ± 24.6, p < 0.001) compared to control mice (Fig. 2a). Similarly, mRNA levels of IL-6 (2.2 ± 1.3 vs. 5.5 ± 3.1, p = 0.01) and IL-1β (0.6 ± 0.4 vs. 1.2 ± 0.5, p = 0.02) were significantly lower in splenic tissue of IL-33-treated mice (Fig. 2a). IL-17 mRNA expression in the kidneys and the spleen was not different between both groups. Compared to control mice, IL-33-treated mice also had significantly lower serum levels of IL-6 (33.7 ± 28.8 vs. 10.1 ± 10.6 pg/mL, p = 0.045) and TNF-α (19.8 ± 6.3 vs. 7.3 ± 2.9 pg/mL, p < 0.001) (Fig. 2b). There was a trend of lower serum IFN-γ (10.9 ± 12.7 vs. 1.9 ± 1.6 pg/mL, p = 0.07) and IL-10 (113.3 ± 110.6 vs. 30.2 ± 23.1 pg/mL, p = 0.055), but the differences did not reach statistical significance. Serum levels of IL-12, IL-1β, and IL-17 were comparable between both groups. These data suggested that IL-33 possessed immunomodulatory effect on cytokines released in innate and adaptive immune responses.

Fig. 2.

IL-33 reduced renal and systemic pro-inflammatory cytokine expression. IL-33-treated mice had lower expression of mRNA transcripts of pro-inflammatory cytokines in renal and splenic tissues (a), lower serum pro-inflammatory cytokine levels (b), lower CD11b+ infiltrates in the kidneys (c), predominant mRNA expression of M2 markers in renal tissue (d), increased M2 marker-expressing splenic CD11b+ monocytes (e), and skewed M2 mRNA expression in isolated splenic CD11b+ cells (f) (data from 9 mice per group for 3 rounds of experiments).

IL-33-Treated Mice Showed Features of M2 Polarisation

Next, we examined whether the reduced pro-inflammatory cytokines occurred as a result of polarisation of myeloid cells towards an anti-inflammatory M2 phenotype. Infiltration of CD11b+ monocytes in the kidneys of IL-33-treated mice was found to be significantly lower in percentage and absolute count compared with control mice (2.9 ± 1.3% vs. 5.1 ± 2.0%, p = 0.045) (Fig. 2c). There were no differences in renal infiltrating CD11c+ dendritic cells, CD3+ T cells, and CD19+ B cells between both groups. Renal tissue from IL-33-treated mice displayed prominent features of M2 macrophage polarisation. M1 marker mRNA expressions including MCP-1 (12.7 ± 6.5 vs. 35.1 ± 12.0, p < 0.001) and iNOS (329.9 ± 226.3 vs. 907.7 ± 450.6, p = 0.003) in the kidneys were significantly lower in IL-33-treated mice than control mice, whereas that of M2 markers including Arg1 (288.3 ± 303.5 vs. 52.9 ± 41.1, p = 0.04) and FIZZI (46.2 ± 50.6 vs. 3.0 ± 2.5, p = 0.02) were significantly higher (Fig. 2d). Expressions of Ym-1, CD206, and CCL17 transcripts were similar in both groups.

In the spleen, there were higher percentage and absolute count of CD11b+ monocytes (21.3 ± 11.4% vs. 10.6 ± 3.6%, p = 0.045) but not CD11c+ dendritic cells in IL-33-treated mice (Fig. 2e). The population of CD206-expressing CD11b+ monocytes was significantly higher in the IL-33-treated group (5.2 ± 0.9 vs. 2.9 ± 0.9, p < 0.001), but iNOS-expressing CD11b+ monocytes were not different from control mice. Isolated splenic CD11b+ monocytes in IL-33-treated mice displayed significantly elevated mRNA expression of Arg1 (48,580.0 ± 33,672.5 vs. 50.8 ± 52.9, p = 0.02), Ym1 (23,662.7 ± 18,221.2 vs. 2.3 ± 2.2, p = 0.03), and IL-10 (42,705.0 ± 33,675.7 vs. 32.7 ± 51.4, p = 0.04), whereas TNF-α (1,254.3 ± 2,975.5 vs. 22,097.6 ± 9,911.9, p = 0.001) and iNOS (215.1 ± 50.3 vs. 2,207.0 ± 1,130.2, p = 0.005) were significantly reduced compared to control mice (Fig. 2f). In sum, these data suggested that IL-33 possessed inducing effect on M2 polarisation of myeloid cells.

IL-33 Induced Th2 Response in Treated Mice

As M2 polarisation of myeloid cells may be induced by Th2 cytokines [29], we examined whether IL-33 induced Th2 response in MRL/lpr mice. Bioactivity of IL-33 towards Th2 response was shown by a dose-dependent induction of IL-5 and IL-13 production in ex vivo-stimulated splenocytes from wild-type MRL/MpJ and Balb/c mice although these responses were relatively blunted in MRL/MpJ mice (Fig. 3a). Kidneys of IL-33-treated MRL/lpr mice showed more apparent increased mRNA transcripts of IL-13 than the spleen and were compared to control mice (100.7 ± 36.3 vs. 38.9 ± 25.2, p = 0.005) (Fig. 3b). Ex vivo stimulation of splenocytes of IL-33-treated mice produced modestly higher IL-5 and IL-13 compared to control mice though the differences were not statistically significant (Fig. 3c). Serum levels of IL-4, IL-5, and IL-13 were similar in both groups (Fig. 3d). ST2 mRNA expression was found to be elevated in renal (8.2 ± 6.0 vs. 1.0 ± 0.6, p = 0.048) and splenic (0.5 ± 0.3 vs. 0.2 ± 0.1, p = 0.004) tissues of IL-33-treated mice (Fig. 3b) suggesting autoregulation of immune responses by IL-33 systemically and in the kidneys.

Fig. 3.

IL-33 induced Th2 response in treated lupus mice. a Bioactivity of IL-33 on the induction of IL-5 and IL-13 production by splenocytes in wild-type MpJ and Balb/c mice. b Renal and splenic tissue of IL-33-treated mice expressed higher mRNA transcripts of IL-13 and ST2. c IL-33-treated mice exhibited increased IL-5 production in ex vivo anti-CD3 stimulated splenocytes. d Comparable serum levels of IL-4, IL-5, and IL-13 compared to control mice (data from 9 mice per group for 3 rounds of experiments).

Increased ST2-Expressing Treg in IL-33-Treated Mice

Next, we examined whether the reduced pro-inflammatory cytokines occurred as a result of immunomodulation of effector T cell subsets by IL-33. Spleen weight adjusted to body weight and total splenocyte count were similar in both groups (Fig. 4a). Among splenic CD3+ T cells, IL-33-treated mice had increased percentages and absolute counts of CD4+ (17.7 ± 3.9% vs. 13.0 ± 2.5%, p = 0.02), CD8+ (13.6 ± 3.4% vs. 6.5 ± 4.0%, p = 0.003) but reduced CD4-CD8-double-negative (68.3 ± 6.3% vs. 77.0 ± 8.1%, p = 0.04) lymphocytes compared to control mice (Fig. 4b). There was significantly higher CD4+CD25+ population among CD3+ T cells in IL-33-treated mice compared with control mice (4.0 ± 1.2 vs. 2.2 ± 0.2, p = 0.003) (Fig. 4c). CD25-expressing cells were increased among CD4+ lymphocytes in IL-33-treated mice, but the difference did not reach statistical significance (25.2 ± 6.0 vs. 20.2 ± 3.4, p = 0.06) (representative flow diagrams in Fig. 4d and data summary in Fig. 4c). IL-33-treated mice also had higher Foxp3+ (18.0 ± 5.5 vs. 10.5 ± 4.2%, p = 0.02) and CD25+Foxp3+ populations among CD4+ T cells (15.1 ± 5.6 vs. 8.2 ± 3.3%, p = 0.02), and higher Foxp3+ cells among CD4+CD25+ Treg (76.0 ± 5.0 vs. 59.3 ± 12.6%, p = 0.003) compared to control mice. ST2-expressing cells were significantly increased among splenic CD4+ T cells (4.0 ± 2.1% vs. 1.2 ± 0.5%, p = 0.02) and CD4+Foxp3+ (12.8 ± 2.8% vs. 4.0 ± 1.4%, p = 0.004) but not CD4+CD25+ Treg, in IL-33-treated mice compared with control mice suggesting autoregulation of Foxp3-expressing Treg by IL-33 (Fig. 4e).

Fig. 4.

IL-33 expanded Foxp3-expressing regulatory T cells (Treg). a Spleen weight and splenocyte count. b Percentage of splenic CD3+ T cell subsets. c, d Splenic CD25+CD4+ cells in splenic CD3+ T cells, Foxp3+ and Foxp3+CD25+ cells in splenic CD4+ T cells. d Foxp3-expressing cells in splenic CD4+CD25+ T cells. e ST2-expressing cells in splenic CD4+, CD4+CD25+, and CD4+Foxp3+ T cells. f IFN-γ and IL-17-expressing CD4+ splenic T cells. g mRNA transcript levels of T effector transcription factors in splenic and renal tissue. h Immunohistochemical staining for Foxp3 in paraffin-embedded kidney tissue sections (original magnification ×200 and ×400 for upper and lower panels, respectively) and number of Foxp3-expressing cells per high power field (HPF) of IL-33-treated mice and control mice (data from 9 mice per group for 3 rounds of experiments).

There were significantly lower percentage and number of IFN-γ-expressing CD4+ T cells (28.3 ± 15.5% vs. 13.0 ± 7.1%, p = 0.02) but not IL-17-expressing CD4+ T cells in IL-33-treated mice compared to control mice (Fig. 4f). Profile of transcription factors of effector T cells in the spleen revealed increased mRNA expression of Foxp3 (Treg) (0.42 ± 0.16 vs. 0.17 ± 0.11, p = 0.001) and Gata3 (Th2) (0.37 ± 0.20 vs. 0.12 ± 0.09, p = 0.02) but not Tbx21 (Th1) and Rorc (Th17) in IL-33-treated mice compared with control mice (Fig. 4g).

IL-33-Treated Mice Expressed Higher Foxp3 in the Kidneys

Similar profile of transcription factors for effector T cells was found in the kidneys of IL-33-treated mice showing significantly higher mRNA levels of Foxp3 (56.3 ± 42.0 vs. 20.0 ± 6.1, p = 0.02) and Gata3 (1.7 ± 0.5 vs. 0.9 ± 0.5, p = 0.007) together with lower Tbx21 (12.6 ± 6.0 vs. 29.6 ± 13.7, p = 0.004) and Rorc (2.6 ± 1.0 vs. 3.9 ± 0.8, p = 0.009) compared with control mice (Fig. 4g). Immunohistochemical staining of renal tissue showed increased number of infiltrating cells with nuclear Foxp3 expression per high power field in IL-33-treated mice compared with control mice (19.1 ± 9.0 vs. 3.1 ± 2.8, p < 0.001) (Fig. 4h). In sum, these data suggested a predominant effect of IL-33 on upregulation of Foxp3+ Treg systemically and in the kidneys.

IL-33 Had Modest Effect on Anti-dsDNA Antibody Titre and Renal C3 Deposition

Next, we examined whether IL-33 possessed any effect on humoral immune response. Serum IgG level was similar in both groups (Fig. 5a). Serum IgG anti-dsDNA antibody-to-total IgG ratio was also not particularly different except for a trend of lower ratio at week 6 (0.015 ± 0.004 vs. 0.029 ± 0.015, p = 0.078) in IL-33-treated mice compared to control mice (Fig. 5b). Serum IgG anti-BSA antibody levels were similar in both groups (Fig. 5c). In regard to glomerular deposits of C3 and IgG2a in renal tissue, MRL/lpr mice displayed increased immunofluorescence staining than MRL/MpJ mice as expected (Fig. 5d). There was a trend of lower C3 deposits in IL-33-treated mice compared with control mice, but the difference was not statistically significant (integrated density 4.8 × 106 vs. 7.5 × 106, p = 0.05). No difference in IgG2a deposit was observed between both groups.

Fig. 5.

Effect of IL-33 on humoral immune response. a IL-33-treated mice had comparable serum levels of IgG. b Anti-dsDNA antibody relative to IgG. c Immunofluorescence staining of renal tissue for C3 and IgG2a deposits (original magnification ×200) (data from 9 mice per group for 3 rounds of experiments).

Discussion

In this study, we showed that administration of exogenous IL-33 to pre-disease lupus-prone mice ameliorated disease severity with significant reduction in skin rash, proteinuria, renal inflammatory infiltrates, and glomerular crescent formation. Renal infiltration by myeloid cells has been reported at the onset of lupus nephritis which secretes IL-1β [30] and induces TNF-α production by intrinsic renal cells leading to immune cell recruitment and glomerular crescent formation [31]. Indeed, we found that IL-33-treated mice had less CD11b+ monocytic renal infiltrates with diminished mRNA expression of IL-6, TNF-α, and IL-1β. Reduced serum levels and lower transcripts of these cytokines in splenic tissue reflected the systemic effect of IL-33. Renal tissue of IL-33-treated mice also displayed lower MCP-1 transcript, a chemokine produced by renal mesangial cells with chemotactic activity for myeloid cells [32]. MCP-1 is involved in the progression of lupus nephritis [33] and is regarded as a biomarker for active lupus nephritis in SLE patients [34]. Activated macrophages produce pro-inflammatory cytokines, nitric oxide, reactive oxygen species, and matrix metalloproteinases which contribute to renal inflammation and damage [35]. Poor renal outcome in human lupus nephritis was previously shown to be associated with persistence of glomerular and tubular macrophages after treatment [36].

Macrophages display plasticity and adopt different functional phenotypes involved in inflammation, immunoregulation, and tissue modelling [37]. M1 macrophages, classically activated by IFN-γ and lipopolysaccharide, produce pro-inflammatory cytokines and release nitric oxide from enhanced iNOS activity [38]. M2 or alternatively activated macrophages produce anti-inflammatory cytokines IL-10 [39] and TGF-β [40] and are characterised by increased expression of CD206 [41], ARGI, FIZZ1, and CCL17 [42]. We demonstrated features of M2 polarisation in IL-33-treated mice including increased CD206-expressing CD11b+ monocytes in the spleen with skewed expression of M2 over M1 markers in the kidneys and the spleen. Serum levels of TNF-α and IL-6 were also reduced in IL-33-treated mice.

Evidences of in vivo M2 polarisation by exogenous IL-33 have also been demonstrated in other mouse models including adipose inflammation [11], cerebral malaria [13], and polymicrobial peritonitis [43] with amelioration of the underlying conditions. On the other hand, exogenous IL-33-induced polarisation of alveolar macrophages to M2 phenotype led to aggravation of antigen-induced airway inflammation mediated by Th2 cells [6]. The in vivo effect of IL-33 on M2 polarisation was demonstrated to be dependent on IL-4 receptor α-chain (IL-4Rα) [6], a common subunit of the functional receptor complex for IL-4 and IL-13 [44], as well as a downstream transcription factor STAT-6 that mediates signalling of these cytokines [43]. IL-4 and IL-13 are Th2 cytokines known to polarise M2 macrophages [29]. IL-33 was shown to induce M2 polarisation by inducing IL-13 production from CD4+ T cells [45] and innate lymphoid cells [12–14, 46] in other murine models.

Indeed, IL-33 has first been described on its Th2-promoting function [5]. IL-33 induced IL-5 and IL-13 production by polarised Th2 cells in vitro and in vivo [5]. We found increased IL-13 transcript in renal extract and modestly higher IL-5 production in ex vivo-stimulated splenocytes in IL-33-treated mice suggestive of Th2 inducing effects. Serum levels of IL-5 and IL-13 were not increased which may be related to their genetic background as reflected by the blunted Th2 cytokine response in ex vivo-stimulated splenocytes in MRL/MpJ mice compared to the Balb/c strain.

Instead, we demonstrated a prominent effect of IL-33 on the expansion of Foxp3+ and Foxp3+CD25+-expressing CD4+ Treg in the spleen and accumulation of Foxp3-expressing cells among renal infiltrates with increased Foxp3 and Gata3 expression in both organs. Expansion of functionally suppressive Foxp3+ Treg by exogenous IL-33 has been reported in the prevention of mouse models of allograft rejection [15] and graft-versus-host disease [16]. Treating NZM2328 and MRL/lpr lupus models by IL233, a hybrid cytokine of IL-2 and IL-33, also showed increased Foxp3+ Treg in the spleen and renal draining lymph nodes with amelioration of lupus nephritis [47]. IL-33 was shown to be a cofactor promoting Foxp3+ expression in inducible Treg (iTreg) in differentiating culture [8].

The role of ST2 mediating the in vivo effects of IL-33 was previously demonstrated by experiments using ST2 knock-out mice [6, 48]. IL-33-induced IL-13 production was shown to upregulate ST2 expression on macrophages, providing a positive feedback mechanism potentiating macrophage responsiveness to IL-33 [6]. We also found increased ST2-expressing cells among Foxp3+CD4+ but not CD25+CD4+ Treg in IL-33-treated mice suggesting autoregulation of IL-33-expanded Foxp3-expressing Treg. IL-33 signalling via ST2 has been shown to activate Gata3 which binds to ST2 promotor and upregulates ST2 expression on Th2 cells [49] and differentiated iTreg [8]. Activated Gata3 also binds to putative binding sites on Foxp3 promoter in iTreg with concomitant promotion of Foxp3 upregulation [50]. Indeed, exogenous IL-33 expanded Treg with elevated ST2 and Foxp3 expression was shown in an inflammatory bowel model [8]. ST2 expression was demonstrated to be essential for the proliferative capacity and persistence of Foxp3+ expression of these Tregs in the in vivo chronic inflammatory environment [8]. Some studies found more potent immunosuppressive function of ST2+ Treg than non-ST2-expressing Treg [8, 51], but the finding was not consistently reported [15].

In addition to Th1 response, cytokines of the IL-23/IL-17 axis also mediate pathogenesis in human lupus [52]. We observed diminished proportion of IFN-γ-expressing splenic CD4+ T cells and a trend of reduced serum IFN-γ level in IL-33-treated mice compared to control mice though no differences were found in regard to serum IL-17 and IL-17-expressing splenic T cells. Downregulated Th1 and Th17 responses were more apparent at the tissue level with reduced renal expression of Tbx21 and Rorc transcripts. Reduced serum IFN-γ level [13, 53] and IL-17-expressing CD4+ T cells [53] were also reported in other IL-33 treated models. Furthermore, ST2+ Treg has previously been shown to suppress IFN-γ-producing CD8+ cells [54]. We observed a trend of reduction in serum IL-10 in IL-33-treated mice compatible with the finding reported in other models [13, 15]. IL-10 is an immunoregulatory cytokine that downregulates Th1 response [55]. In active SLE patients, overproduction of IL-10 by autoreactive extrafollicular B cells may serve as an ectopic source [56] whereas B cell-targeted therapy in MRL/lpr mice was associated with a reduction in serum IL-10 [57]. Thus, our finding supports a systemic lupus disease-ameliorating effect of IL-33.

We did not find any difference in serum IgG anti-dsDNA antibodies and glomerular C3 and IgG2a deposit in IL-33-treated mice compared to control mice although humoral effect of IL-33 under certain circumstances has been reported. Endogenous IL-33 has been found to activate B1 cells with IgM production through IL-5 induction [58] and to induce IL-10-producing regulatory B cells that ameliorated mucosal inflammation [59]. On the other hand, exogenous IL-33 was shown to enhance IgE production that triggers mast cell degranulation [60] and promote BAFF-dependent autoantibody production upon chronic exposure [61]. Consistent with our finding, the absence of effect on immune-complex renal deposition was also reported on the lupus mouse models treated by the hybrid cytokine IL233 despite improvement in proteinuria [47].

In this study, we showed that exogenous IL-33 facilitated M2 polarisation, Th2 induction, and Treg expansion with amelioration of disease in lupus-prone mice (summarised in Fig. 6). Although we have not delineated the source of IL-13, increased IL-13 mRNA expressed in the kidney supported a role of IL-13 mediating M2 polarisation induced by IL-33. In addition, IL-33-expanded Treg may also contribute to induction of M2 macrophages [12] through arginase, IL-10, and TGF-β [62]. Depletion of Foxp3+ Treg in vivo was found to favour M1 over M2 macrophages [63]. On the other hand, IL-33-polarised M2 macrophages were shown to induce Treg via the production of IL-10 in vivo [13, 14]. Furthermore, IL-33 was also reported to promote IL-2 production by dendritic cells [54] essential for the expansion and survival of Treg [64, 65].

Fig. 6.

Schematic diagram demonstrating the orchestrated effects of exogenous IL-33 on M2 polarisation of myeloid cells, induction of Th2 cells, and expansion of Treg mediated by upregulated ST2 expression enabling autoregulation in the amelioration of disease activity in MRL/lpr mice.

We have previously reported elevated serum sST2 that correlated with disease activity in SLE patients suggesting a role of IL-33/ST2 in the pathogenesis of SLE [23]. Since sST2 has been regarded as a decoy receptor restricting bioavailability of IL-33 [66], our earlier finding supported an antagonistic effect of sST2 on endogenous IL-33 activity in the chronic inflammatory environment in SLE. While we and others have shown that early administration of IL-33 ameliorated lupus disease [47, 67], one study has reported improvement in lupus condition by treating MRL/lpr mice with neutralising anti-IL-33 antibody at the age of week 14 when the mice presented with higher level of proteinuria compared to those in our study [68]. Although there exists a potential difference in the pharmacokinetics between recombinant protein [47, 67] and immunoglobulin-based therapy [68], it is more likely that serum and tissue levels of IL-33 from endogenous production and IL-33-based treatment in different phases of lupus disease can be associated with different disease outcomes as IL-33 possesses both pro- and anti-inflammatory functions [1]. Indeed, conflicting outcomes of exogenous IL-33 leading to disease amelioration or aggravation have also been reported in the experimental autoimmune encephalomyelitis model [69, 70]. These discrepant findings highlight a need to titrate the dose of any IL-33-based therapy such that the expression level of IL-33 favours its role as a regulatory cytokine in inducing the anti-inflammatory M2 macrophage and Treg responses.

In conclusion, our study showed that IL-33 orchestrated the effects of M2 macrophages, Th2 cells, and Treg mediated by the upregulated ST2 expression enabling autoregulation and potentiation of the responsiveness of these immune cells to IL-33 resulting in alleviation of lupus disease activity in MRL/lpr mice. The immunoregulatory relationship between IL-33 and these immune cells under the best timing, dose, and duration of treatment can be harnessed for the development of therapeutic strategies for SLE in the future.

Statement of Ethics

This study protocol was reviewed and approved by the committee on the use of live animals in teaching and research of the University of Hong Kong (approval number: 2890-12) and the animal research Ethics Committee of the City University of Hong Kong (approval number: A-0155) and followed the ARRIVE guidelines.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

The study was supported by the Research Grant Council, Hong Kong (Grant No. 775313).

Author Contributions

Mo Yin Mok, Fang Ping Huang, Chak Sing Lau, and Godfrey Chi Fung Chan contributed to the study design, supervision of experiments, data analysis, and manuscript preparation. Ka Sin Law contributed to supervision of experiments, and data collection and analysis. Kwok Wah Chan contributed to data interpretation of renal histology. Wing Yin Kong, Cai Yun Luo, and Endale T Asfaw contributed to data collection and analysis.

Data Availability Statement

All data generated or analysed during this study are included in this article. Further inquiries can be directed to the corresponding author.

This article is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC). Usage and distribution for commercial purposes requires written permission.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.