Regulatory T cell features in chronic granulomatous disease

Summary

Chronic granulomatous disease (CGD) is a primary immunodeficiency caused by mutations in any of the genes encoding the phagocyte nicotinamide adenine dinucleotide phosphate (NADPH) oxidase system, responsible for the production of reactive oxygen species (ROS). CGD is marked by invasive bacterial and fungal infections and by autoinflammation/autoimmunity, of which the exact pathophysiology remains elusive. Contributing factors include decreased neutrophil apoptosis, impaired apoptotic neutrophil clearance, increased proinflammatory protein expression and reduced ROS‐mediated inflammasome dampening. We have explored a fundamentally different potential mechanism: it has been reported that macrophage‐mediated induction of regulatory T cells (T_regs) depends on ROS production. We have investigated whether numerical or functional deficiencies exist in T_regs of CGD patients. As the prevalence of autoinflammation/autoimmunity differs between CGD subtypes, we have also investigated T_regs from gp91^phox‐, p47^phox‐ and p40^phox‐deficient CGD patients separately. Results show that T_reg numbers and suppressive capacities are not different in CGD patients compared to healthy controls, with the exception that in gp91^phox‐deficiency effector T_reg (eT_reg) numbers are decreased. Expression of T_reg markers CD25, inducible T cell co‐stimulator (ICOS), Helios, cytotoxic T lymphocyte antigen 4 (CTLA‐4) and glucocorticoid‐induced tumor necrosis factor receptor (GITR) did not provide any clue for differences in T_reg functionality or activation state. No correlation was seen between eT_reg numbers and patients' clinical phenotype. To conclude, the only difference between T_regs from CGD patients and healthy controls is a decrease in circulating eT_regs in gp91^phox‐deficiency. In terms of autoinflammation/autoimmunity, this group is the most affected. However, upon culture, patient‐derived T_regs showed a normal phenotype and normal functional suppressor activity. No other findings pointed towards a role for T_regs in CGD‐related autoinflammation/autoimmunity.

Introduction

Chronic granulomatous disease (CGD) is a rare primary immunodeficiency (~1 : 250 000 patients) caused by inactivating mutations in the genes encoding the different nicotinamide adenine dinucleotide phosphate (NADPH) oxidase subunits. This results in impaired production of reactive oxygen species (ROS). The NADPH oxidase enzyme is a multi‐protein complex consisting of a membrane‐bound component [the gp91^phox (phox = phagocyte oxidase) and p22^phox proteins] as the catalytic core of the system, and a cytosolic component (the p67^phox, p47^phox and p40^phox proteins). After phagocytosis, the cytosolic component translocates towards the membrane‐bound component, activating the complex. Then, a burst of oxygen consumption takes place and ROS are formed 1, which are critical for the killing of pathogenic bacteria and fungi. CGD is therefore characterized by life‐threatening invasive bacterial and fungal infections 2. Another important characteristic of CGD is the presence of excessive inflammation, such as granuloma formation in among others the skin and the urinary tract, inflammatory bowel disease (IBD) and (discoid) lupus 3. Contributing factors implicated in the pathophysiology of this autoinflammation/autoimmunity include decreased neutrophil apoptosis, impaired apoptotic neutrophil clearance, increased proinflammatory protein expression and reduced ROS‐mediated inflammasome dampening 2, 4, 5, 6.

We here explored a fundamentally different potential mechanism: it has been reported that macrophage‐mediated induction of regulatory T cells (T_regs) depends on ROS production 7. Kraaij et al. showed that macrophage‐derived ROS of healthy people can induce T_regs, whereas macrophages from CGD patients were significantly impaired in T_reg induction. T_regs are CD4⁺CD25⁺forkhead box protein 3 (FoxP3)⁺ or CD127^– cells important for immunological tolerance. Naive, thymus‐derived T_regs (nT_regs) form the majority of T_regs in lymphoid tissues and express CD62L and CCR7. T_regs can also differentiate into effector T_regs (eT_regs), which is the most abundant T_reg fraction in non‐lymphoid tissues. These cells express high levels of CD44, inducible T cell co‐stimulator (ICOS) and glucocorticoid‐induced tumor necrosis factor‐related receptor (GITR) as well as B lymphocyte‐induced maturation protein‐1 (BLIMP)‐1 and IL‐10, consistent with a functional and suppressive phenotype 8.

It has also been suggested that asbestos‐derived ROS activate transforming growth factor (TGF)‐β 9 and that TGF‐β is important for T cell suppression by anti‐inflammatory (M2) macrophage‐induced T_regs in humans 10. Moreover, it has recently been shown that human CD8⁺ T_regs suppress CD4⁺ T cells by secreting NADPH oxidase 2 (NOX2) (gp91^phox)‐containing vesicles 11. This results in elevated ROS levels in the CD4⁺ T cell and consequently in reduced phosphorylation of the T cell receptor (TCR)‐associated zeta chain‐associated protein kinase of 70 kD (ZAP70), which inhibits the T cell receptor 12. Another role for ROS in T_reg functioning was seen when Efimova et al. compared the capacities of T_regs from wild‐type and p47^phox‐deficient mice in suppressing effector T cell proliferation. These authors showed that NCF1 (encoding p47^phox)‐deficient T_regs suppressed p47^phox‐deficient effector T cells very poorly compared to wild‐type T_regs 13. Kwon et al. also used a mouse model to show that Nox2 deficiency results in (among others) T cell hyperactivation and diminished T_reg differentiation 14.

Besides the fact that excessive inflammation is one of the characteristics of CGD, it is known that the prevalence of inflammatory complications differs between CGD subtypes 15. Therefore, it would be highly informative to further study the role of T_regs in CGD‐related inflammation. Hence, we have investigated whether numerical or functional T_reg deficiencies exist in different, well‐characterized and fully genotyped gp91^phox‐, p47^phox‐ and p40^phox‐deficient CGD patients.

Materials and methods

Blood samples and cell isolation

Blood samples from CGD patients were drawn after written informed consent had been obtained and after approval from the Medical Ethical Committee of the Academic Medical Center, University of Amsterdam, the Netherlands (2013_234/NL40331.078.12, ‘Primary immunodeficiencies: immunological and genetic background in relation to clinical complications’). Blood samples of healthy individuals were obtained from anonymized healthy donors recruited from an internal system at Sanquin, approved by the Medical Ethical Committee of the Academic Medical Center, University of Amsterdam. Written informed consent was also obtained. All procedures were conducted in accordance with the 1975 Declaration of Helsinki as revised in 2013. From all included patients, lymphocyte counts, CD4 and CD8 counts from the day of peripheral blood mononuclear cell (PBMC) isolation were collected from the medical chart, as well as the mutational details, age at blood sampling, medical history and current treatment.

PBMCs from both patients and controls were isolated by Percoll gradient (1·076 g/ml) centrifugation (1006 g rpm, 21°C, 20 min, Rotina 420 R; Hettich, Geldermalsen, the Netherlands). T_reg (subsets) from the peripheral blood were counted by flow cytometry. CD4⁺CD25⁺CD127^– and CD4⁺CD25⁺FoxP3⁺ were the markers used for T_reg counts. The markers CD4⁺CD25⁺CD127^–CD45RA⁺ and CD4⁺CD25⁺CD127^–CD45RA^– were used for naive and effector T_regs, respectively. For in‐vitro cell expansion viable cells were separated by flow cytometric sorting based on the expression of CD4, CD25, CD45RA and CD127 on a FACS Aria III (BD Biosciences, San Jose, CA, USA).

Antibodies

The following monoclonal antibodies (mAbs) were used: CD25 (43839), CD45RA (560675), ICOS (557802), CD8 (555369) from BD Biosciences. CD127 (351309) and CTLA‐4 (349906) from Biolegend (San Diego, CA, USA). FoxP3 (25‐4777‐42), Helios (48‐9883) and CD4 (12‐0047‐42) were from eBioscience (San Diego, CA, USA).

In‐vitro cell expansion

Naive conventional T responder cells (Tconvs) (CD4⁺, CD127⁺, CD25^‐, CD45RA⁺) and naive T_regs (CD4⁺, CD127^–, CD25⁺, CD45RA⁺) were isolated by fluorescence activated cell sorter (FACS) (FACS Aria III; BD Biosciences). The cells were then cultured in the presence of 0·1 µg/ml of anti‐CD3 mAb (M1654, clone 1XE, PeliCluster; Sanquin, Amsterdam, the Netherlands) and anti‐CD28 mAb (16‐0289‐85, clone CD28.2; eBioscience) for 14 days in Iscove's modified Dulbecco's medium (IMDM) containing 10% (v/v) fetal calf serum (FCS) and 300 U/ml IL‐2 (Proleukin).

Flow cytometry

Cells were labeled with fluorochrome‐conjugated antibodies in phosphate‐buffered saline (PBS), 0·5% (v/v) bovine serum albumin (BSA) for 30 min at 4°C. For intracellular and nuclear staining, cells were fixed and permeabilized with the FoxP3/transcription factor fixation/permeabilization buffers (eBioscience), according to the manufacturer's instructions. Near‐IR (L10119; ThermoFisher, Fremont, CA, USA) dye was used to stain dead cells. Cells were analyzed on LSR Fortessa or LSR II cytometers (BD Biosciences).

In‐vitro cell suppression assay

The suppressive capacity of the in‐vitro expanded T_regs was assessed by inhibition of proliferation of co‐cultured responder T cells (T_resp) as described previously 16. Briefly, PBMCs were purified, labeled with CellTrace™ Violet cell proliferation kit (C34557, ThermoFisher) and stimulated with 0·01 µg/ml of anti‐CD3 mAb (M1654, clone 1XE, PeliCluster) and anti‐CD28 mAb (16‐0289‐85, clone CD28.2; eBioscience) at different ratios of T_resp : T_reg in the absence of IL‐2. On day 5, cells were stained with antibodies against CD4 and CD8 and TOPRO‐3 (T3605; Invitrogen). Finally, cells were analyzed on LSR Fortessa or LSR II cytometers (BD Biosciences).

Statistics

Statistical analysis was performed with Prism version 6.0 (GraphPad Software). The variable distribution was assessed by the Kolmogorov–Smirnov test. A non‐parametric Kruskal–Wallis test was used, followed by a Dunn's multiple comparison test to assess intergroup differences when three or more groups were compared. A non‐parametric Mann–Whitney U‐test was used when two groups were compared. P < 0·05 was considered statistically significant (*) at a 95% confidence level.

Results

Seven patients with X‐linked CGD (gp91^phox deficiency), seven with autosomal recessive CGD (four with p47^phox deficiency and three with p40^phox deficiency) and seven healthy controls were included. Mutational and medical details are shown in Table 1.

Table 1.

Mutational and medical details from all included CGD patients.

Patient	CGD subtype	Mutation	Age at blood draw (years)	Gender (M/F)	Lympho‐ cytes × 109/l	CD4+ × 109/l	CD8+ × 09/l	% CD4+ CD25+ FoxP3+ T cells	% CD4+ CD25+ CD127–CD45RA– eTregs	Treatment at moment of sampling	Clinical phenotype
1	gp91phox	389Gly> Ala	31·8	M	1	0·2	0·37	1–4%	1–3%	Co‐trimoxazole, itraconazole, interferon γ	Recurrent pneumonia in childhood. Stable for the past 20 years
2	gp91phox	Frameshift	3·9	M	5·59	1·39	0·97	≥4%	≥3%	Co‐trimoxazole	Recurrent infections. Thymic abscesses, intra‐abdominal abcesses. No IBD
3	gp91phox	Promotor ‐52C>T	17·7	M	2·38	0·875	0·542	1–4%	≤1%	Co‐trimoxazole	Nocardia farcinica pneumonia during infancy
4	gp91phox	38Lys>stop	5·1	M	4·86	0·812	2·391	≥4%	1–3%	Co‐trimoxazole	Aspergillus spondylodiscitis, lymphadenitis colli
5	gp91phox	293Gln> stop	4·6	M	3·58	1·38	0·54	1–4%	≤1%	Co‐trimoxazole	Lymphadenopathy
6	gp91phox	Deletion exon 6‐13	14·9	M	1·74	0·72	0·35	≤1%	≤1%	Co‐trimoxazole, interferon γ	2 × otitis, 1 × dermatomycosis, celiac disease
7	gp91phox	c.877C>T; p.293 Gln> stop	NA	M	6·02	NA	NA	1–4%	≤1%	n.a.	n.a.
8	p47phox	Frameshift delGT + deletion cDNA exon 5	5·9	M	NA	1·49	4·45	1–4%	1–3%	Co‐trimoxazole, mesalazine	Fungal infection of the skin, tinea capitis, noduli in the neck. IBD
9	p47phox	Frameshift delGT homozygous	18	M	3·15	1·39	0·78	≥4%	≥3%	Co‐trimoxazole, interferon γ	Pneumonia, S. aureus bacteremia with hepatic abscesses, recurrent dermatomycosis
10	p47phox	Frameshift delGT Homozygous	11·3	F	2·03	0·79	0·59	1–4%	1–3%	Co‐trimoxazole, triamcinolone (topical)	Inguinal lymphadenitis (Aspergillus), sometimes apthous lesions. Recurrent otitis
11	p47phox	Frameshift delGT homozygous	15·9	F	2·09	1·14	0·9	1–4%	≤1%	Co‐trimoxazole, itraconazole, interferon γ	S. epidermis pyelonephritis, fungal infection of the skin, 2× pulmonary Aspergillosis
12	p40phox	Splice site c.118‐G>A Homozygous	14·5	F	1·27	n.a.	n.a.	1–4%	1–3%	Co‐trimoxazole	IBD
13	p40phox	Splice site c.118‐G>A Homozygous	3·1	F	2·3	n.a.	n.a.	≥4%	≤1%	Co‐trimoxazole, itraconazole	Recurrent pulmonary infections, lung and lymph granuloma
14	p40phox	Splice site c.118‐G>A Homozygous	10·6	M	0·64	n.a.	n.a.	≤1%		Co‐trimoxazole, itraconazole	IBD, S. aureus skin abscesses

IBD = inflammatory bowel disease; T_reg = regulatory T cell; CGD = chronic granulomatous disease; FoxP3 = forkhead box protein 3.

First, percentages of circulating T_regs among the T cell populations in these patients were calculated (direct measurements of absolute CD4⁺ and CD8⁺ T cell subset counts were not always available). Figure 1a shows the flow cytometry gating strategy used for these counts. Proportions of total CD4⁺CD25⁺CD127^– (P = 0·06) and CD4⁺CD25⁺FoxP3⁺ (P = 0·075) T_reg cell populations were not different between the cohort of all CGD patients and controls (Fig. 1b). However, the group of gp91^phox‐deficient CGD patients by itself did possess significantly lower proportions of CD4⁺CD25⁺FoxP3⁺ T_regs (Fig. 1b, right panel). Examination of T_reg subpopulations showed that both naive (CD4⁺CD25⁺CD127^–CD45RA⁺) and effector (CD4⁺CD25⁺CD127^–CD45RA^–) T_reg populations were reduced in CGD patients (Fig. 1c). These effects were again most prominent among gp91^phox‐deficient patients. As an essential surface protein for T_reg function and maintenance 17, we assessed IL‐2 receptor alpha‐subunit CD25 expression and observed reduced expression in T_regs of gp91^phox‐deficient patients (Fig. 1d, upper panel and 1E, left panel). In contrast, CGD patients had elevated percentages of T_regs expressing ICOS (Fig. 1d, lower panel and 1E, right panel), which is thought to be expressed in highly functional and suppressive cells 18.

Figure 1.

Numbers of regulatory T cells in healthy controls and in different chronic granulomatous disease (CGD) subtypes. (a) Flow cytometry gating strategy of the markers CD4, CD127, CD45RA, CD25 and forkhead box protein 3 (FoxP3). (b) Percentages of CD4⁺CD25⁺CD127^–regulatory T cells (T_regs) (left) and percentages of CD4⁺CD25⁺ FoxP3⁺ T_regs (right); (c) Percentages of CD4⁺CD25⁺CD127^–CD45RA⁺ nT_regs (left) and percentages of CD4⁺CD25⁺CD127^–CD45RA^– eT_regs (right); (d) expression patterns of CF25 and inducible T cell co‐stimulator (ICOS) on flow cytometry. (e) Expression of CD25 (left) and percentages of ICOS (right) on total isolated T cells from heparin‐derived peripheral blood mononuclear cells (PBMCs) in healthy controls and in different CGD subsets.

T_regs from p40^phox‐deficient patients expressed elevated levels of Helios and CTLA‐4 (not shown); these markers are associated with regulatory functionality 8, 19. However, this elevation was not seen in the other CGD subtypes, suggesting that it is not a general feature of CGD.

Finally, as a further phenotypical hallmark of T_regs 20, we assessed T_regs positive for glucocorticoid‐induced tumor necrosis factor‐related receptor (GITR) in the blood of CGD patients. No differences were seen between controls and patients, neither in percentage nor in expression level (not shown).

Collectively, there seemed to be no significant differences in T_reg numbers in the peripheral blood of CGD patients compared to healthy controls. No clues supporting an important role for T_regs in CGD‐related inflammation were observed, except for the reduced proportion of eT_regs in the peripheral blood of gp91^phox‐deficient patients. We subsequently isolated and cultured T_regs from the same CGD patients and controls. T_regs from CGD patients yielded similar cell expansion compared to controls after stimulation with antibodies to CD3 and CD28 in the presence of IL‐2 and expression levels of the T_reg markers CD25, FoxP3, Helios and CTLA‐4 were comparable to those of healthy controls after culture (not shown). The suppressive capacity of in‐vitro expanded T_regs was assessed by a well‐established in‐vitro CD4 and CD8 suppression assay 16. No differences were observed either between the CGD patients and controls nor between the CGD patients categorized according to the three different CGD subtypes (Fig. 2). As mentioned previously, CGD patients may suffer from very severe autoinflammatory/autoimmune manifestations, including Crohn‐like colitis or severe lung disease 2, 6, 15, 21 as we have also recently demonstrated in p40^phox deficiency 22. However, we did not observe any apparent association between individual patients' autoinflammatory/autoimmune manifestations and their T_reg numbers (Table 1, Fig. 1a,b), surface markers (Fig. 1c) or suppressive capacities (Fig. 2) at the moment of blood sampling.

Figure 2.

Effector functions of in‐vitro expanded regulatory T cells (T_regs). Functional capacities of regulatory T cells from chronic granulomatous disease (CGD) patients and from healthy controls are shown as their capability to suppress CD4⁺ and CD8⁺ T cells. Data are in median ± standard error of the mean (s.e.m.) when n > 3.

Discussion

Several reports suggest a role for ROS in T_reg function 7, 12, 13. Details on T_reg numbers and activity in CGD patients, who are defective in phagocytic ROS formation, are largely lacking and a role of T_regs in CGD‐related autoinflammation/autoimmunity remains to be examined. In this study we have investigated both T_reg numbers in CGD patients as well as their suppressive capacities upon culture ex vivo. Autoinflammatory/autoimmune manifestations may diverge extensively in the CGD patients 15, although in general the p47^phox‐deficient patients are the least affected in that sense. If any deficiency in T_regs would underlie the pathogenesis of their clinical manifestations, we were expecting large differences in T_reg numbers and/or function between CGD patients and controls. Our recent study of p40^phox deficiency has demonstrated extensive autoinflammation/autoimmunity, equal if not more prominent compared to gp91^phox‐deficient CGD 22. For this reason, we expected also differences in T_regs within the different CGD subgroups.

In our relatively small but well‐characterized patient cohort we observed a reduction in circulating CD4⁺CD25⁺CD127^–CD45RA^– eT_regs in patients diagnosed with X‐linked (gp91^phox‐deficient) CGD. These eT_regs are the cells with the most active and suppressive capacities within tissues 8, and considering the high prevalence of inflammation in gp91^phox‐deficient CGD patients compared to those deficient for p47^phox, this finding could indicate a role for eT_regs in the pathophysiology of CGD‐related inflammation. In line herewith, Kraaij et al. observed impaired numbers of FoxP3⁺ cells upon in‐vitro culture among CD4⁺CD25⁺ cells when comparing CGD‐macrophages from five patients to healthy control macrophages in the outgrowth of T_regs upon culture 7. However, it is hard to compare our findings with their results obtained in the past. First, these FoxP3⁺ cells were not further phenotyped or specified as CD4⁺CD25⁺CD127^–CD45RA^– eT_regs. Secondly, Kraaij et al. did also not distinguish between gp91^phox‐ and p47^phox‐deficient patients. Conversely, significant differences were not detected between controls and p40^phox‐deficient patients in our study, while p40^phox‐deficient CGD patients exhibit very prominent manifestations of autoinflammation/autoimmunity 22. Even though it is unfortunate that we were only able to examine a small number of p40^phox‐deficient patients, our data do not indicate a strong phenotypical difference with healthy controls.

Also, the expression patterns of surface markers associated with T_reg activity did not unequivocally support a specific hypothesis regarding the T_reg functionality and their possible role in CGD‐related autoinflammation/autoimmunity. For example, CD25 levels were significantly reduced on T_regs from gp91^phox‐deficient patients but this was not the case in the similarly affected p40^phox‐deficient patients. The general increase in expression of ICOS on T_regs from all CGD groups might reflect a need for these cells to be active in order to mitigate the inflammation present in these patients. When analyzing patients individually, no correlation was observed between expression of these surface markers and inflammatory complications (not shown). Moreover, in‐vitro‐expanded T_regs from CGD patients showed normal cell expansion, normal expression levels of CD25, FoxP3, Helios and CTLA‐4 upon subculture and a normal capacity to inhibit proliferation of co‐cultured responder T cells (T_resp), and in all these functions were similar to control T_regs.

Although it cannot be excluded that the in‐vivo function of T_regs in CGD may be disturbed, our data clearly suggest that CGD‐causing mutations do not appear to cause intrinsic defects in T_reg cell functionality. The reduction in eT_reg cell numbers in (gp91^phox‐deficient) CGD patients is consistent with diminished induction of such cells due to lack of ROS produced by macrophages or with the destabilization of T_regs as a consequence of the presence of inflammatory cytokines, as suggested previously 23. Apart from the reduction in eT_reg numbers in gp91^phox‐deficient CGD in the blood, no support for a clear role of T_regs in CGD‐related inflammation can be derived from our study.

Disclosures

The authors have declared that no conflicts of interest exist.

Author contributions

T. W. K. and D. A. designed the study. A. v. d. G, E. C and M. C. S performed the experiments. A. v. d. G and E. C. wrote the manuscript, which was critically read by the other authors.