Circulating regulatory T cells in adult‐onset Still’s disease: Focusing on their plasticity and stability

Abstract

We investigated the characteristics of regulatory T cells in adult‐onset Still’s disease (AOSD) with a focus on their plasticity, stability and relationship to disease severity. The proportion of circulating CD4⁺CD25⁺forkhead box protein 3 (FoxP3⁺) cells (T_regs) and intracellular expression of effector cytokines, including interferon (IFN)‐γ, interleukin (IL)‐17 and IL‐4, was analysed in 27 untreated patients with AOSD (acute AOSD), 11 of the 27 patients after remission and 16 healthy controls (HC) using flow cytometry. The suppressive ability of T_regs was also evaluated. Regression analyses of the results were performed. The proportion of T_regs was significantly lower in patients with acute AOSD than in the HC. The expression levels of IFN‐γ, IL‐17 and IL‐4 in T_regs were significantly increased in patients with acute AOSD. IFN‐γ and IL‐4 expression levels were inversely correlated with the proportion of T_regs and positively correlated with serum ferritin levels. Decreased expression of FoxP3 in CD4⁺CD25⁺ cells, which was correlated with increased expression of IL‐17, and impaired suppressive function were observed in T_regs in acute AOSD. However, these aberrant findings in T_regs, including the reduced circulating proportion and functional ability and altered intracellular expression levels of cytokines and FoxP3, were significantly improved after remission. In acute AOSD, T_regs show plastic changes, including effector cytokine production and reductions in their proportion and functional activity. IFN‐γ and IL‐4 expression levels in T_regs may be associated with disease severity. Also, down‐regulation of FoxP3 may be related to IL‐17 expression in T_regs. Importantly, the stability of T_regs can be restored in remission.

This study investigated the plasticity and stability of regulatory T cells (Tregs) in adult‐onset Still’s disease. In the acute phase of disease, Tregs significantly showed plastic changes including overproduction of intracellular effector cytokines, as well as reductions in FoxP3 expression and functional ability. However, these aberrant findings in Tregs were significantly improved after remission.

INTRODUCTION

Adult‐onset Still’s disease (AOSD) is an autoinflammatory disease that is clinically characterized by spike fever, skin rash, sore throat and arthritis with elevated neutrophil counts and serum ferritin levels. Severe complications, such as macrophage activation syndrome (MAS), thrombotic thrombocytopenic purpura or other visceral impairments, sometimes develop [1, 2, 3]. In the disease mechanism of AOSD, activation of the innate immune system, including macrophages and neutrophils, is the hallmark of pathogenesis [3]. In the acute phase of AOSD, expression of proinflammatory cytokines is increased, and the relationship between serum proinflammatory cytokine levels and disease activity has been demonstrated [1, 2]. Among the proinflammatory cytokines in AOSD, interleukin (IL)‐18 was found to be a potential biomarker [4, 5, 6], and interferon (IFN)‐γ may be associated with the induction of macrophages and neutrophils during the development of AOSD [7, 8, 9]. Our recent study revealed the characteristics of the natural killer cells that produce IFN‐γ in AOSD [10]. Polarization of helper T (Th) type 1 cells, which also produce IFN‐γ, has also been demonstrated in AOSD [11], implicating acquired immune responses in the development of AOSD. Increased frequency of Th17 cells was found to be associated with disease activity in AOSD [12]. In the acquired immune system, regulatory T cells (T_regs), which express CD4, CD25 and the transcription factor forkhead box P3 (FoxP3), play a crucial role in mediating immune tolerance and regulating immune responses [13, 14]. To the best of our knowledge, only one study has examined the role of T_regs in AOSD, which showed a reduced frequency of CD4⁺CD25^high cells [15]. However, it has been reported that the plasticity of T_regs, i.e. a shift to Th‐like cells by effector cytokines, may be promoted in the inflammatory environments of autoimmune disease [16, 17]. Accordingly, it is necessary to assess the plasticity and stability of T_regs in AOSD, a representative systemic inflammatory disease.

In this study, we investigated the characteristics of circulating T_regs, including their proportion, intracellular expression of effector cytokines related to Th cells and suppressive ability in AOSD. In addition, the relationships between T_regs and clinical findings were also evaluated.

MATERIAL AND METHODS

Patients

Twenty‐seven patients with AOSD who had not yet received immunosuppressive therapy (acute AOSD), with a mean age [standard deviation (SD)] of 53.5 (16.9) years, were enrolled into this study. All patients fulfilled the diagnostic criteria for AOSD proposed by Yamaguchi et al. [18]. For the control group, 16 age‐matched healthy controls (HC), with a mean age (SD) of 51.5 (11.6) years (seven men and nine women), were included for comparison. Whole blood samples were obtained from all study participants. The clinical findings of patients with acute AOSD (Table 1) were recorded when whole blood samples were provided. The comprehensive disease score (Pouchot score) [19] and the development of macrophage activation syndrome (MAS) [20, 21] were also included in the analysis. Eleven of the 27 patients whose Pouchot score was zero after remission were also evaluated as the remission AOSD group. Whole blood samples were collected at a mean (SD) of 33.1 (34.3) months after initiating immunosuppressive therapy. Of the 11 patients in the remission AOSD group, nine were on maintenance therapy, including prednisolone [mean (SD) = 4.4 (4.3) mg daily) (n = 8)], methotrexate (n = 3), cyclosporin (n = 4) and/or tocilizumab (n = 2). The local ethics committee of Shinshu University approved this study (approval no.: 601/4294). All participants provided written informed consent.

TABLE 1.

Epidemiological and clinical findings of patients with AOSD

		Acute AOSD	Remission AOSD	HC	p value*
	Number	27	11	16
	Mean age, years	53.5	―	51.5	0.889
	Sex (M/F)	7/20	―	7/9	0.091
Physical findings, n (%)
	Spike fever (more than 38℃)	27 (100)	―	―
	Skin eruption	25 (93)	―	―
	Sore throat	20 (74)	―	―
	Lymphadenopathy	14 (52)	―	―
	Arthritis	24 (89)	―	―
	Myalgia	9 (33)	―	―
	Pleuritis	5 (19)	―	―
	Lung involvement	3 (11)	―	―
	Pericarditis	1 (4)	―	―
	Hepatomegaly	10 (37)	―	―
	Splenomegaly	13 (48)	―	―
	Abdominal pain	2 (7)	―	―
Laboratory data, mean ± SD					p value **
	White blood cells/µl	17 205 ± 11352	7345 ± 1787	―	0.001
	Neutrophils/µl	13 878 ± 10527	4988 ± 1547	―	<0.0001
	Lymphocytes/µl	971 ± 509	1706 ± 596	―	0.001
	C‐reactive protein, mg/dl	11.23 ± 8.16	0.04 ± 0.05	―	<0.0001
	ESR, mm/h	66 ± 36	5.1 ± 2.9	―	0.010
	Ferritin, ng/ml	9618 ± 9545	70 ± 48	―	<0.0001
	AST, U/l	82 ± 87	20 ± 6	―	<0.0001
	ALT, U/l	86 ± 103	20 ± 8	―	0.001
	LDH, U/l	622 ± 435	215 ± 43	―	<0.0001
Activity evaluation
	Pouchot score, mean ± SD	5.5 ± 1.7	0	―
	Fulfilled MAS criteria, n (%)	8 (30)	―	―

Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; ESR, erythrocyte sedimentation rate; HC, healthy controls; MAS, macrophage activation syndrome; SD, standard deviation.

^*Comparisons between 27 acute adult‐onset Still’s disease (AOSD) patients and 16 HC using the Mann–Whitney U‐test for age and Fisher’s exact probability test for sex.

^**Comparisons between 27 acute and 16 remission AOSD patients using the Mann–Whitney U‐test. A p value less than 0.05 was considered statistically significant.

Flow cytometry

Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood samples, which were collected into ethylene diamine tetraacetic acid (EDTA)‐coated tubes by gradient centrifugation with Ficoll‐Hypaque PLUS (GE Healthcare, Pittsburgh, Pennsylvania, USA). Surface and intracellular markers in PBMCs were detected by flow cytometric analysis. The population of CD4⁺CD25⁺FoxP3⁺ cells was defined as T_regs. PBMCs were stimulated with 0.5 μg/ml ionomycin, 0.04 μg/ml phorbol myristate acetate (both from Sigma‐Aldrich, St Louis, Missouri, USA) and 2 μM monensin (BD Biosciences, San Diego, California, USA) at 37°C for 4 h. Stimulated PBMCs were stained with phycoerythrin/cyanin 7 (PE/Cy7)‐conjugated anti‐CD4 (BioLegend, San Diego, California, USA) and PC5‐conjugated anti‐CD25 (Beckman Coulter, Brea, California, USA) with or without PE‐conjugated anti‐CD152 [cytotoxic T lymphocyte antigen 4 (CTLA‐4] (Beckman Coulter) antibodies. For detecting the population of CD4⁺CD25⁺CD127^{− /low}CD45RA⁺FoxP3⁺ cells, stimulated PBMCs were alternatively stained with fluorescein isothiocyanate (FITC)‐conjugated anti‐CD4 (Beckman Coulter), allophycocyanin (APC)‐conjugated anti‐CD25 (BioLegend), PE/Cy7‐conjugated anti‐CD127 (BD Biosciences) and PE/Cy5‐conjugated anti‐CD45RA (BioLegend) antibodies. The stained PBMCs were permeabilized with Cytofix/Cytoperm (BD Biosciences) and then stained with FITC‐conjugated anti‐IFN‐γ (Beckman Coulter), PE‐conjugated anti‐IL‐17 (BD Biosciences), PE‐conjugated anti‐IL‐4 (Beckman Coulter), FITC‐conjugated transforming growth factor (TGF)‐β1 (BioLegend) or PE‐conjugated IL‐10 (BioLegend) antibodies, as well as PE‐conjugated (BD Biosciences), FITC‐conjugated or Pacific blue‐conjugated anti‐FoxP3 (both BioLegend). The permeabilized PBMCs, which were stained with anti‐CD4, anti‐CD25, anti‐CD127 and anti‐CD45RA antibodies, were alternatively stained with PE‐conjugated anti‐T‐bet, anti‐RAR‐related orphan receptor gamma (RORγt) or anti‐GATA binding protein 3 (GATA3) antibody (all from BD Biosciences), as well as Pacific blue‐conjugated anti‐FoxP3 antibody. To assess the suppressive ability of T_regs, suppression assays were performed and evaluated as previously described [22, 23]. The CD4⁺CD25⁺ regulatory T cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) was used to isolate conventional T (con‐T) cells (CD4⁺CD25⁻ cells) and T_regs from unstimulated PBMCs. Allogenic con‐T cells labeled with carboxyfluorescein succinimidyl ester (CFSE, 2 μM; Invitrogen, Carlsbad, California, USA) and T_regs were co‐cultured with anti‐CD3/CD28 microbeads (at 1:1:1) at 37℃ for 4 days. The proliferation of con‐T cells was analysed by the CSFE dilution method, both with and without T_regs. Treated cells were acquired on a FACSCanto II flow cytometer (BD Bioscience), and the data were analysed using FlowJo software version 7.6.5 (Tree Star Inc., Ashland, Oregon, USA).

Enzyme‐linked immunosorbent assay (ELISA)

ELISA kits were used to measure serum levels of IL‐4, IL‐6, IL‐12, IL‐17, IL‐21, IFN‐γ (R&D Systems, Minneapolis, MN, USA) and IL‐18 (Medical and Biological Laboratories, Nagoya, Japan).

Statistical analysis

All data are presented as mean ± standard deviation (SD). The Mann–Whitney U‐test and Fisher’s exact probability test were used to compare two independent groups. Patient characteristics before and after treatment were compared by Wilcoxon’s signed‐rank test. The Kruskal–Wallis test was performed for comparisons between three independent groups and the Steel–Dwass test was used for multiple comparisons. A correlation coefficient test was performed to evaluate the significance of relationships. A p value less than 0.05 was defined statistically significant. All statistical analyses were performed using BellCurve for Excel (SSRI, Tokyo, Japan) and JMP (SAS Institute, Cary, North Carolina, USA).

RESULTS

Clinical characteristics of patients with AOSD

Of the 27 AOSD study patients, all (100%) had spike fever, 25 (93%) indicated skin eruption and eight (30%) fulfilled the criteria for MAS (Table 1). The mean Pouchot score was 5.5 ± 1.7 in the acute AOSD group. In the comparisons between three different clinical courses, which are classified into monocyclic, polycyclic and chronic pattern [1, 2], serum levels of C‐reactive protein at AOSD diagnosis were significantly higher in the polycyclic pattern than the chronic pattern (Supporting information, Table S1). Laboratory findings were significantly improved in the remission AOSD group and were within the normal ranges (Table 1).

Proportion of T_regs and intracellular expression of effector cytokines in acute AOSD

The proportion of circulating T_regs was significantly lower in the acute AOSD patients than in the HC (5.06 ± 3.92 versus 7.33 ± 3.47, p = 0.016) (Figure 1a). This value was significantly increased in the remission AOSD patients (8.90 ± 4.40, p = 0.004) (Figure 1b). Intracellular expression of IFN‐γ, IL‐17 and IL‐4 in T_regs was significantly higher in the acute AOSD patients than in the HC (p < 0.0001) (Figure 2a, Table 2). Expression levels of the aforementioned cytokines were significantly lower in the remission AOSD patients [median fluorescence index (MFI): p = 0.003, p = 0.003, p = 0.004, respectively (Figure 2b) (frequency: p = 0.004, p = 0.003, p = 0.003, respectively] (Table 2). Meanwhile, the proportion of circulating CD4⁺CD25⁺CD127^−/lowCD45RA⁺FoxP3⁺ cells was also significantly lower in the acute AOSD patients than in the HC (1.50 ± 0.51 versus 7.13 ± 3.27, p = 0.0001), and was significantly increased after remission (9.20 ± 2.50, p = 0.027) (Supporting information, Figure S1). Intracellular expression of IFN‐γ, IL‐17, and IL‐4 in CD4⁺CD25⁺CD127^−/lowCD45RA⁺FoxP3⁺ cells was also significantly higher in the acute AOSD patients than in the HC (p = 0.0006, p = 0.006, p = 0.0006, respectively), and their expression was significantly decreased after remission (p = 0.027) (Supporting information, Figure S2). In addition, intracellular expression of T‐bet, RORγt and GATA3, which are T helper type (Th)1, Th17 and Th2‐related transcription factors, respectively, in CD4⁺CD25⁺CD127^−/lowCD45RA⁺FoxP3⁺ cells, were also significantly higher in the acute AOSD patients than in the HC (p = 0.0001, p = 0.0001, p = 0.0006, respectively), and their expression was significantly decreased after remission (p = 0.027) (Supporting information, Figure S3).

FIGURE 1.

Proportions of circulating CD4⁺CD25⁺forkhead box protein 3 (FoxP3⁺) cells [regulatory T cells (T_regs)]. (a) Comparison of T_regs between patients with acute adult‐onset Still’s disease (AOSD) and healthy controls (HC). (b) Differences in the proportions of T_regs in 11 patients between the acute and remission phases of AOSD. The Mann–Whitney U‐test was used for comparison between the acute AOSD patients and HC. Wilcoxon’s signed‐rank test was used for comparison between acute and remission phases in 11 AOSD patients. *p < 0.05, **p < 0.005

FIGURE 2.

Intracellular expression levels of effector cytokines in CD4⁺CD25⁺forkhead box protein 3 (FoxP3⁺) cells [regulatory T cells (T_regs)]. (a): (Upper panels) Representative histograms of interferon (IFN)‐γ, interleukin (IL)‐17 and IL‐4 expression in T_regs (lower bar graphs). Comparisons of the median fluorescence index (MFI) of IFN‐γ, IL‐17 and IL‐4 in T_regs between patients with acute adult‐onset Still’s disease (AOSD) and HC. (b) Differences in the MFI of IFN‐γ, IL‐17 and IL‐4 in T_regs from 11 patients between acute and remission phases. The Mann–Whitney U‐test was used for comparisons between the acute AOSD patients and the HC. The Wilcoxon signed‐rank test was used for comparisons between acute and remission phases in 11 AOSD patients. **p < 0.005, ***p < 0.0001

TABLE 2.

Frequencies of intracellular cytokines in T_regs in patients with acute AOSD, remission AOSD and healthy controls

		Acute AOSD (n = 27)	Remission AOSD (n = 11)	HC (n = 16)	p value
					*/**
In CD4+CD25+FoxP3+ cells
	%IFN‐γ	24.18 ± 14.92	5.20 ± 9.61	4.33 ± 3.64	0.004/<0.0001
	%IL‐17	23.58 ± 21.15	1.20 ± 1.36	1.84 ± 2.19	0.003/<0.0001
	%IL‐4	26.97 ± 23.27	0.78 ± 1.41	2.04 ± 2.15	0.003/<0.0001

Data are presented as mean ± standard deviation (SD).

Abbreviations: FoxP3, forkhead box protein 3; HC, healthy controls; IFN‐γ, interferon‐γ; IL, interleukin; T_regs, regulatory T cells.

^*Comparisons between acute and remission phases in 11 acute adult‐onset Still’s disease (AOSD) patients using Wilcoxon’s signed‐rank test.

^**Comparisons between 27 acute AOSD patients and 16 HC using the Mann–Whitney U‐test. A p value less than 0.05 was considered statistically significant.

Serum levels of Th cell‐related cytokines and their relationships with T_reg plasticity in AOSD

Serum levels of IL‐4, IL‐6, IL‐12, IL‐17, IL‐18, IL‐21 and IFN‐γ were significantly higher in the acute AOSD patients than in the HC (Table 3). Comparisons of the acute and remission AOSD patients showed no significant differences in serum levels of IL‐4, IL‐12 or IL‐17, despite significant reductions in the levels of the other evaluated cytokines. However, serum levels of these cytokines were not significantly correlated with the proportion of circulating T_regs or the intracellular expression levels of IFN‐γ, IL‐17 and IL‐4 in T_regs (data not shown).

TABLE 3.

Serum cytokine levels in patients with acute and remission AOSD and healthy controls

	Acute AOSD	Remission AOSD	HC	p value
Cytokines	(n = 27)	(n = 11)	(n = 16)	*/**
IL‐4 (pg/ml)	13.1 ± 6.1	8.7 ± 2.1	8.5 ± 2.0	0.102/0.033
IL‐6 (pg/ml)	109.7 ± 139.4	11.7 ± 20.3	1.7 ± 0.6	0.005/<0.0001
IL‐12 (pg/ml)	1.5 ± 0.5	1.8 ± 1.0	0.7 ± 0.2	0.726/<0.0001
IL‐17 (pg/ml)	32.3 ± 17.6	23.3 ± 2.8	21.5 ± 4.1	0.368/0.025
IL‐18 (pg/ml)	2039.4 ± 779.6	351.6 ± 286.6	72.0 ± 11.1	<0.0001/<0.0001
IL‐21 (pg/ml)	26.6 ± 41.2	8.4 ± 1.5	9.2 ± 1.8	<0.0001/<0.0001
IFN‐γ (pg/ml)	75.3 ± 167.3	3.3 ± 2.0	2.3 ± 3.2	<0.0001/<0.0001

Data are presented as the mean ± standard deviation (SD).

Abbreviations: HC, healthy controls; IL, interleukin; IFN‐γ, interferon‐γ.

^*Comparisons between acute and remission phases in 11 acute adult‐onset Still’s disease (AOSD) patients using Wilcoxon’s signed‐rank test.

^**Comparisons between 27 acute AOSD patients and 16 HC using the Mann–Whitney U‐test. A p value less than 0.05 was considered statistically significant.

Correlations among circulating T_regs, intracellular effector cytokine expression and disease severity in acute AOSD

Correlation analyses of circulating T_regs showed that the proportion of T_regs was inversely correlated with the MFI of IFN‐γ and IL‐4 in T_regs in the acute AOSD patients, but was not correlated with that of IL‐17 (Table 4). Frequencies of IFN‐γ, IL‐17 and IL‐4 were inversely correlated with the proportion of T_regs. Moreover, the MFI and frequencies of IFN‐γ and IL‐4 in T_regs were correlated with serum ferritin levels in the acute AOSD patients, whereas those of IL‐17 showed no correlation. There was no significant correlation between the proportion of T_regs and serum ferritin levels (data not shown).

TABLE 4.

Correlations between intracellular cytokine levels and the proportions of T_regs or serum ferritin in patients with acute AOSD

		Versus %CD4+CD25+FoxP3+ cells		Versus serum ferritin levels
		Coefficient	p value	Coefficient	p value
MFI in CD4+CD25+FoxP3+ cells
	MFI−IFN‐γ	−0.421	0.029	0.433	0.027
	MFI−IL‐17	−0.324	0.099	0.156	0.425
	MFI−IL‐4	−0.449	0.019	0.413	0.035
% Frequency in CD4+CD25+FoxP3+ cells
	% IFN‐γ	−0.499	0.008	0.390	0.047
	% IL‐17	−0.547	0.003	0.184	0.349
	% IL‐4	−0.680	0.0005	0.435	0.030

A p value less than 0.05 was considered statistically significant.

Abbreviations: AOSD, acute adult‐onset Still’s disease; Coefficient, correlation coefficient; FoxP3, forkhead box protein 3; IL, interleukin; IFN‐γ, interferon‐γ; MFI, median fluorescence index.

Suppressive ability of T_regs in AOSD

To evaluate the suppressive ability of T_regs, the proliferation of con‐T target cells was evaluated with and without T_regs (Figure 3a,b). The proliferation of con‐T cells was significantly lower in the presence of T_regs from HC than in the absence of T_regs (p = 0.001). The proliferation of con‐T cells did not differ significantly in the presence of T_regs from acute AOSD patients or in the absence of T_regs (p = 0.193); however, proliferation was significantly higher in the presence of T_regs from acute AOSD patients than in the presence of T_regs from HC (p = 0.003), indicating abrogated suppressive function of T_regs in acute AOSD. The suppressive ability of T_regs from remission AOSD patients was significantly improved (p = 0.043) (Figure 3c). The expression levels of representative suppressive mediators, including TGF‐β1, IL‐10 and CTLA‐4 in T_regs from acute AOSD patients were not significantly different than those in T_regs from HC (p = 0.845, p = 0.220, p = 0.259, respectively) (Figure 4a). In contrast, the MFI and frequencies of FoxP3 in CD4⁺CD25⁺ cells were significantly lower in the acute AOSD patients than in the HC (p < 0.0001) (Figure 4b, Table 5) and were significantly increased after remission (p = 0.004, p = 0.013, respectively) (Figure 4c, Table 5). In CD4⁺CD25⁺CD127^−/lowCD45RA⁺ cells, the MFI and frequency of FoxP3 were also significantly lower in the acute AOSD patients than in the HC (p = 0.0002, p = 0.0001, respectively), and were significantly increased after remission (p = 0.027) (Supporting information, Figure S4). The MFI of FoxP3 was inversely correlated with the MFI and frequency of IL‐17 in T_regs (Figure 4d), but was not correlated with those of either IFN‐γ or IL‐4 (data not shown).

FIGURE 3.

Suppression assay of regulatory T cells [regulatory T cells (T_regs)]. (a) Representative histograms showing the proliferation of conventional T cells (con‐T cells) with and without T_regs. (b) Comparisons of proliferative frequency of con‐T cells without T_regs (n = 9) and with T_regs from either acute adult‐onset Still’s disease (AOSD) patients (n = 8) or healthy controls (HC; n = 9). (c) Differences in the proliferative frequency of con‐T cells in the presence of T_regs from five patients collected in the acute and remission phases (n = 5). CFSE = carboxyfluorescein diacetate succinimidyl ester; HC‐T_regs = T_regs from HC; acute AOSD‐T_regs = T_regs from patients with acute AOSD; a‐CD3/CD28 = anti‐CD3/CD28 microbeads; NS = not significant. The Kruskal–Wallis and Steel–Dwass tests were used for comparisons of proliferative frequency of con‐T cells between three independent groups. Wilcoxon’s signed‐rank test was used for comparisons between acute and remission phases in five AOSD patients. *p < 0.05, **p < 0.005

FIGURE 4.

Expression of functional mediators in regulatory T cells (T_regs). (a) Comparisons of median fluorescence index (MFI) of transforming growth factor (TGF)‐β, interleukin (IL)‐10, and cytotoxic T lymphocyte antigen (CTLA)‐4 in CD4⁺CD25⁺forkhead box protein 3 (FoxP3⁺) cells collected from patients with acute acute adult‐onset Still’s disease (AOSD) and healthy controls (HC). (b) (Left panel) Representative histograms of FoxP3 expression in CD4⁺CD25⁺ cells from patients with AOSD and HC (right bar graph). Comparison of the MFI for FoxP3 in CD4⁺CD25⁺ cells from patients with acute AOSD and HC. (c) Differences in the MFI of FoxP3 in CD4⁺CD25⁺ cells from 11 patients with AOSD collected in the acute and remission phases. (d) Regression analyses of MFI and frequency of IL‐17 in MFI⁻FoxP3 in CD4⁺CD25⁺ cells from patients with acute AOSD. The Mann–Whitney U‐test was used for comparisons between the acute AOSD patients and the HC. Wilcoxon’s signed‐rank test was used for comparisons between acute and remission phases in 11 AOSD patients. **p < 0.005, ***p < 0.0001; NS = not significant

TABLE 5.

Frequency of FoxP3 in CD4⁺CD25⁺ cells in patients with acute AOSD, remission AOSD and healthy controls

		Acute AOSD	Remission AOSD	HC	p value
		(n = 27)	(n = 11)	(n = 16)	*/**
In total lymphocytes
	% CD4+CD25+ cells	45.75 ± 15.48	43.40 ± 10.89	40.14 ± 18.02	0.374/0.218
In CD4+CD25+ cells
	% FoxP3	8.43 ± 7.83	18.73 ± 8.48	25.24 ± 9.62	0.013/<0.0001

Data are presented as the mean ± standard deviation (SD).

Abbreviations: HC, healthy controls; FoxP3, forkhead box protein 3.

^*Comparisons between acute and remission phases in 11 acute adult‐onset Still’s disease (AOSD) patients by using Wilcoxon’s signed‐rank test.

^**Comparisons between 27 acute AOSD patients and 16 HC using the Mann–Whitney U‐test. A p value less than 0.05 was considered statistically significant.

The relationship with clinical findings and clinical course

We investigated the relationship between the obtained results and clinical findings shown in Table 1. The expression of IL‐17 in T_regs was significantly higher in patients with sore throat than in those without (p = 0.015) (Supporting information, Figure S5). The expression of IL‐4 in T_regs was significantly higher in patients with pleuritis than in those without (p = 0.005). In accordance with previous investigation in which a cut‐off at 7.0 of the Pouchot score has an impact on the prognosis [24], we additionally compared the experimental results between patients showing more and less than this score, resulting in no significant differences (data not shown). Any experimental results had no significant differences between monocyclic, polycyclic and chronic pattern (data not shown).

DISCUSSION

In our first attempt at studying AOSD, we investigated the characteristics of circulating T_regs in AOSD, with a focus on their plasticity and stability. Previous investigations of T_regs in autoimmune diseases, such as rheumatoid arthritis, systemic lupus erythematous, systemic vasculitis and multiple sclerosis, inconsistently showed decreased and increased expression; however, defective T_reg function was largely a consistent result [17, 22, 25, 26]. Our study demonstrated a decreased proportion of T_regs and dysfunction in acute AOSD patients as significant results. However, an increase in the proportion of T_regs and improved suppressive activity were both observed in remission AOSD patients, suggesting that the T_reg disorder in AOSD may be reversible.

Intracellular expression of the effector cytokines IFN‐γ, IL‐17 and IL‐4 in the population of CD4⁺CD25⁺FoxP3⁺ cells was significantly higher in patients with acute AOSD. T_regs may include memory/effector phenotypes in the environment of chronic inflammation and/or in the process of immune‐aging [27, 28]. Meanwhile, subsets of CD127^−/low and CD45RA⁺ are found to be phenotypical markers for certifying the homeostasis of T_regs in healthy individuals [29, 30]. The proportion of CD4⁺CD25⁺CD127^−/lowCD45RA⁺FoxP3⁺ cells was significantly lower than that of CD4⁺CD25⁺FoxP3⁺ cells in the acute AOSD patients, suggesting that conversion to effector phenotypes might be shown even in the population expressing FoxP3. However, expression of IFN‐γ, IL‐17, IL‐4 and their transcription factors in CD4⁺CD25⁺CD127^−/lowCD45RA⁺FoxP3⁺ cells was significantly higher in the acute AOSD patients than the HC, demonstrating that the phenotypical changes may be the pivotal features of T_regs in the acute phase of AOSD. The plasticity of T_regs includes Th‐like changes, characterized by the expression of effector cytokines in the conventional phenotype of T_regs that express FoxP3 [16, 17, 31, 32], ultimately resulting in an imbalance in immune tolerance due to insufficient intervention by T_regs [16, 33]. We have two hypotheses to explain the impaired homeostasis of T_regs in AOSD. First, the decrease in the proportion of circulating T_regs might be attributed to intracellular induction of effector cytokines. Notably, IFN‐γ and IL‐4 expression levels in T_regs were associated with the reduction in circulating T_regs. Although no significant correlation was found between the MFI of IL‐17 in T_regs and the proportion of T_regs in this study, plastic alteration of T_regs into Th17‐like cells has also been shown in some autoimmune diseases [34, 35]. Secondly, the increased intracellular expression of IFN‐γ, IL‐17 and IL‐4 might impair the suppressive function of T_regs. In fact, both proportion and functional capacity of T_regs recovered after remission with a significant reduction in intracellular cytokine expression. However, in this study there were no significant deficiencies in TGF‐β1, IL‐10 or CTLA‐4, which are mediators of the suppressive activity of T_regs [14, 25]. Our result demonstrated significantly decreased FoxP3 expression in CD4⁺CD25⁺ cells in patients with acute AOSD. Moreover, this was related to increased intracellular IL‐17 expression in T_regs. Conventional T_regs express FoxP3 as a pivotal regulator preventing extreme immune responses [36, 37]; T_regs down‐regulate FoxP3 expression in response to inflammatory signals [38, 39, 40]. It has been shown that intracellular induction of IL‐17 in T_regs leads to the down‐regulation of FoxP3 expression [39, 41]. Therefore, we assumed that the T_regs in patients with acute AOSD could be converted into not only Th‐like T_regs that retain FoxP3 expression but also Th phenotypes that lose FoxP3 expression, resulting in a decreased proportion of circulating T_regs. In addition, intracellular expression of IL‐17 might contribute to the loss of FoxP3 in T_regs, leading to impaired function. In contrast, it has also been reported that Th‐like T_regs which retain FoxP3 have less suppressive activity than conventional T_regs, although their suppressive capacity is not abolished [16, 31]. Under our experimental conditions, CD4⁺CD25⁺ cells, which were isolated using a commercially available magnetic isolation kit, were defined as T_regs and their suppressive activity was evaluated. Therefore, the isolated T_regs might include cell populations with down‐regulated FoxP3 levels and Th‐like FoxP3⁺ populations, suggesting that both populations have defective suppressive activity.

Some studies have shown elevated serum levels of IFN‐γ or IL‐17 as well as an increase in the proportion of circulating Th1 or Th17 cells in AOSD [8, 11, 12, 42]. Increased serum IL‐4 levels were also shown to be significantly correlated with disease activity [43, 44]. Accordingly, Th1, Th2 and Th17‐related cytokines may be strongly implicated in the pathogenesis of AOSD. Based on the results of previous investigations, plastic alteration of T_regs is promoted by exposure to Th cell‐related inflammatory cytokines [16, 17, 32, 45, 46, 47, 48]. However, our results showed that serum levels of these cytokines were not significantly correlated with the intracellular expression levels of IFN‐γ, IL‐17 or IL‐4 in T_regs. In contrast, IFN‐γ and IL‐4 expression levels in T_regs were associated with serum levels of ferritin, which is a biomarker commonly used for evaluating disease severity in AOSD [2, 3, 49, 50]. Hence, it was suggested that the plastic change of T_regs especially expressing IFN‐γ or IL‐4 may be significantly associated with disease severity in AOSD.

There are some limitations in this study. Our results statistically demonstrated low correlation coefficients in the regression analyses of intracellular cytokines in T_regs, suggesting that their correlations with FoxP3 expression, the proportion of T_regs and serum ferritin levels are very weak. Therefore, Th‐like changes in T_regs may have limited implications for the variation of T_regs and disease severity in AOSD. Also, it has been shown that ethnic and genetic variations may affect the immunity of AOSD [1, 2, 3] as well as the development of T_regs [51, 52], although a small number of Japanese patients were employed in this study.

In conclusion, in acute AOSD, circulating T_regs, which showed increases in the intracellular expression levels of IFN‐γ, IL‐17 and IL‐4, were significantly reduced in both proportion and suppressive activity. The expression levels of IFN‐γ and IL‐4 in T_regs might be implicated in a decreased proportion of T_regs and elevated serum levels of ferritin. In addition, decreased expression of FoxP3 in CD4⁺CD25⁺ cells was significantly shown in acute AOSD, and might be associated with IL‐17 expression in T_regs. These findings were significantly improved in patients with remission AOSD, suggesting that the Th‐like shift and functional impairment of T_regs are reversible. Disease activity may affect the stability of T_regs in AOSD. However, numerous immunopathogenic mechanisms are involved in the development of AOSD, and the immune response of T_regs can be affected by broad immune system interactions. Thus, further investigations on the immune signals affecting the plasticity of T_regs in AOSD are required.

CONFLICTS OF INTEREST

The authors declare that they have no financial or personal conflicts of interest.

AUTHOR CONTRIBUTIONS

All authors designed this study, developed the structure and argument for this study. Y.‐S., T.‐I., D.‐K. and R.‐T. recruited blood samples and clinical data. Y.‐S. performed laboratory investigations and analyzed obtained data and prepared the draft of this manuscript. Y.‐S. and Y.‐S. contributed to revision of the manuscript. All authors revised and approved the final manuscript.

Supporting information

Fig S1‐S5

Table S1

Shimojima Y, Ichikawa T, Kishida D, Takamatsu R, Sekijima Y. Circulating regulatory T cells in adult‐onset Still’s disease: Focusing on their plasticity and stability. Clin Exp Immunol. 2021;206:184–195. 10.1111/cei.13648

DATA AVAILABILITY STATEMENT

The data for the analyses in this study are available from the corresponding author upon reasonable request.