Alterations in the adenosine metabolism and CD39/CD73 adenosinergic machinery cause loss of Treg cell function and autoimmunity in ADA-deficient SCID

Abstract

Adenosine acts as anti-inflammatory mediator on the immune system and has been described in regulatory T cell (Treg)-mediated suppression. In the absence of adenosine deaminase (ADA), adenosine and other purine metabolites accumulate, leading to severe immunodeficiency with recurrent infections (ADA-SCID). Particularly ADA-deficient patients with late-onset forms and after enzyme replacement therapy (PEG-ADA) are known to manifest immune dysregulation. Herein we provide evidence that alterations in the purine metabolism interfere with Treg function, thereby contributing to autoimmune manifestations in ADA deficiency. Tregs isolated from PEG-ADA-treated patients are reduced in number and show decreased suppressive activity, whereas they are corrected after gene therapy. Untreated murine ADA(-/-) Tregs show alterations in the plasma membrane CD39/CD73 ectonucleotidase machinery and limited suppressive activity via extracellular adenosine. PEG-ADA-treated mice developed multiple autoantibodies and hypothyroidism in contrast to mice treated with bone marrow transplantation or gene therapy. Tregs isolated from PEG-ADA-treated mice lacked suppressive activity, suggesting that this treatment interferes with Treg functionality. The alterations in the CD39/CD73 adenosinergic machinery and loss of function in ADA-deficient Tregs provide new insights into a predisposition to autoimmunity and the underlying mechanisms causing defective peripheral tolerance in ADA-SCID.

Trial registration: ClinicalTrials.gov NCT00598481 NCT00599781.

Figures

Figure 1

Phenotype of Tregs in ADA-SCID patients after different treatments. (A) Representative FACS staining and gating strategy of Tregs in a pediatric HD, GT, and PEG-ADA–treated patient. Total PBMCs were stained for CD4, CD25, CD127, and FOXP3. Gates were established on the basis of isotype-identical monoclonal antibody controls. Treg frequencies are indicated. (B) Percentage of CD25+FOXP3+ Tregs in the CD4+ gate of 22 HD, 11 GT, and 7 PEG-ADA–treated patients. Horizontal bars indicate mean values. *P < .05. **P < .005. (C) Percentage of Tregs identified by the low or absent expression of CD127 in the CD25+FOXP3+ gate. Horizontal bars indicate mean values. **P < .005.

Figure 2

Treg function in ADA-SCID patients after different treatments. (A) Representative autologous and allogeneic suppression assays for a PEG-ADA–treated patient compared with an HD. E* indicates effector cells stimulated with allogeneic accessory cells (APCs) and soluble anti-CD3. Percentages of suppression are indicated. Data are intraexperimental mean ± SD. (B) Representative autologous and allogeneic suppression assays for a GT-treated patient compared with an HD. Percentages of suppression are indicated. (C-D) Stimulated Tregs (Treg*) of the same PEG-ADA-treated (C) and GT-treated (D) patients with their respective HDs and reversion of Treg anergy by IL-2. (E-F) Summarized percentages for autologous (E) and allogeneic (F) suppression assays in PEG-ADA– and GT-treated patients compared with healthy controls. Data are interexperimental mean ± SD. ***P < .0005 (Mann-Whitney test). **P < .005 (Mann-Whitney test). *P < .05 (Mann-Whitney test).

Figure 3

Expansion of the Treg compartment in untreated ADA−/− mice. (A) Anti-CD3 and FoxP3 immunohistochemistry in thymi of 3-week-old ADA+/+ and ADA−/− mice (original magnification ×10). (B) TaqMan gene expression analysis for CD4 and FoxP3 on total thymi of ADA−/− (n = 6) mice compared with ADA+/+ (n = 6). ADA+/+ gene expression was arbitrarily considered 1 as represented by the dashed line. Data are mean ± SD. (C) Representative FACS analysis for CD4+CD25+FoxP3+CTLA4+CD45RBlow Tregs in ADA+/+ and ADA−/− PB. (D) Percentage of CD25+FoxP3+CTLA4+CD45RBlow Tregs in CD4+ compartment in PB, lymph nodes (LN), thymus, and spleen of ADA+/+ (n = 11) and ADA−/− (n = 11) mice. Data are mean ± SD. **P = .005-.0005. ***P ≤ .0005. (E) ADA enzymatic activity in ADA+/+ CD4+CD25− T effector (n = 3) and CD4+CD25+ T regulatory cells (n = 3). Data are mean ± SD. **P = .008.

Figure 4

Altered CD39 and CD73 expression and decreased suppressive function of ADA−/− Tregs. (A) TaqMan gene expression analysis for CD39 (ADA+/+: n = 15, ADA−/−: n = 16; **P = .0047) and CD73 (ADA+/+: n = 9, ADA−/−: n = 10. **P = .0045) in purified CD4+CD25+ Tregs; box and whiskers graphs. (B) Summarized MFI of CD39 and CD73 gated on CD4+CD25+FoxP3+ cells from ADA+/+ (n = 5) and ADA−/− (n = 5) mice; box and whiskers graphs. (C) Representative histogram plots for CD39 and CD73 on CD4+CD25+FoxP3+ splenocytes from untreated ADA+/+ and ADA−/− mice. Isotype-identical monoclonal antibodies served as controls (solid gray histogram). (D) AXP levels divided into ATP and AMP illustrate ectonucleotidase activity in serum samples from ADA−/− (n = 7) and age-matched ADA+/+ mice (n = 8) at 3 weeks of age. Data are mean ± SD. *P < .05. (E) Representative suppression assay using ADA+/+ effector cells cocultured with Tregs from untreated ADA+/+ (n = 3) or ADA−/− mice (n = 5). E indicates unstimulated effector cells; and E*, effector cells stimulated with anti-CD3 and anti-CD28. Data are interexperimental mean ± SD. (F) Serum adenosine levels in ADA−/− (n = 7) and age-matched ADA+/+ mice (n = 8) at 3 weeks of age. Data are mean ± SD. ***P < .0005. (G) Serum hypoxanthine levels in ADA−/− (n = 7) and age-matched ADA+/+ mice (n = 8) at 3 weeks of age. Data are mean ± SD. **P < .005. (H) Representative histogram plots for CD39 and CD73 on CD4+CD25+FoxP3+ splenocytes from ADA+/+ and ADA−/− mice treated with 50 mg/kg of caffeine (nonselective Adora antagonist). Isotype-identical monoclonal antibodies served as controls (solid gray histogram). (I) Representative suppression assay using ADA+/+ effector cells cocultured with Tregs from ADA+/+ (n = 4) or ADA−/− mice (n = 6) treated with 50 mg/kg of caffeine. Data are intraexperimental mean ± SD.

Figure 5

Immunologic reconstitution in ADA−/− mice treated with PEG-ADA, GT, and BMT. (A) After 18 weeks (wks), white blood cell counts (WBC) significantly increase in PEG-ADA–treated mice (n = 12) compared with GT (n = 8) or BMT (n = 15) treated ADA−/− mice and ADA+/+ mice (n = 17). Data are mean ± SD. *P < .05. (B) Compared with control groups, platelet (PLT) counts in PEG-ADA–treated mice significantly decrease after 18 weeks of treatment. Data are mean ± SD. **P < .005. Absolute numbers of PB lymphocyte populations measured by FACS analyses as percentage of WBC counts, at weeks 10, 14, and 18. Data are mean ± SEM. (C) CD4+ T cells. (D) CD8+ T cells. (E) B220+ B cells. *P = .05-.005. **P = .005-.0005. (B-E) PEG-ADA–treated ADA−/− (n = 12), GT-treated ADA−/− (n = 8), BMT-treated ADA−/− (n = 15), and ADA+/+ mice (n = 17). (F) IgG1 (ng/mL × 107) serum immunoglobulin levels in PEG-ADA (n = 7), GT (n = 6), or BMT treated ADA−/− mice (n = 10) compared with ADA+/+ (n = 7) and 3-week-old ADA−/− (n = 6) and ADA+/+ (n = 6) mice. Data are mean ± SD.

Figure 6

Increased antibody production in PEG-ADA–treated ADA−/− mice. (A) ELISA assessing antibody production against ADA in PEG-ADA (n = 7), GT (n = 6), and BMT (n = 10) treated ADA−/− mice compared with ADA+/+ mice (n = 7) and 3-week-old ADA−/− mice (n = 3); scatter plot; horizontal bars represent mean. (B) ELISA assessing autoantibody production against platelets in PEG-ADA (n = 10), GT (n = 6), and BMT (n = 10) treated ADA−/− mice compared with ADA+/+ mice (n = 7) and 3-week-old ADA−/− mice (n = 5); scatter plot; horizontal bars represent mean. (C) Autoantibody production directed against thyroid tissue in PEG-ADA but not GT- and BMT-treated ADA−/− or ADA+/+ mice. (D) Thyroid damage occurring in autoantibody-producing PEG-ADA–treated mice compared with a GT- or BMT-treated ADA−/− and WT control. (E) Percentage of apoptotic Ccl3-positive nuclei in the thyroid of PEG-ADA–treated ADA−/− (n = 7) compared with GT (n = 5) and BMT (n = 5) treated ADA−/− and ADA+/+ mice (n = 5). Data are mean ± SD. *P = .05. (F) Concentration of TSH in sera of PEG-ADA–treated ADA−/− mice (n = 14) compared with ADA+/+ mice (n = 8). Data are mean ± SD. *P = .05.

Figure 7

Inhibition of Treg function in the presence of PEG-ADA. (A) Percentage of CD25+FoxP3+ Tregs in ADA−/− rescued with PEG-ADA (n = 20), lentiviral GT (n = 8), and BMT (n = 11) compared with ADA+/+ mice (n = 23). Data are mean ± SEM. ***P = .0001. *P = .05. N.S. indicates not significant. (B) Absolute numbers of Tregs. Data are mean ± SEM. (C) Suppressive function versus ADA+/+ effector cells of ADA+/+ (n = 11) and ADA−/− (n = 8) Tregs compared with Tregs from ADA−/− mice treated with BMT (n = 3), GT (n = 3), and PEG-ADA after 10 (n = 3), 14 (n = 3), and 18 (n = 7) weeks. Data are interexperimental mean ± SD. ***P < .0005. **P < .005. *P < .05. (D) Summarized MFI of CD39 and CD73 gated on CD4+CD25+FoxP3+ cells from ADA+/+ (n = 6) mice compared with ADA−/− mice treated with PEG-ADA for 18 weeks (n = 4). (E) Representative histogram plots for CD39 and CD73 on CD4+CD25+FoxP3+ splenocytes from ADA+/+ and ADA−/− mice treated with PEG-ADA for 18 weeks. Isotype-identical monoclonal antibodies served as controls (solid gray histogram). (F) Measurement of ATP consumption and AMP formation (left panel) as well as adenosine formation (right panel) by purified CD4+CD25+ Treg cells from PEG-ADA–treated ADA−/− (n = 4) and ADA+/+ mice (n = 3). Data are interexperimental mean ± SD. *P < .05. (G) Serum adenosine levels in PEG-ADA–treated ADA−/− (n = 4) and age-matched ADA+/+ mice (n = 4) at 10, 14, and 18 weeks of follow-up. Data are interexperimental mean ± SD. *P < .05. (H) Serum hypoxanthine levels in PEG-ADA–treated ADA−/− (n = 4) and age-matched ADA+/+ mice (n = 4) at 10, 14, and 18 weeks of follow-up. Data are interexperimental mean ± SD. **P < .005.