IL‐35Ig–expressing dendritic cells induce tolerance via Arginase 1

Abstract The cytokine interleukin IL‐35 is known to exert strong immunosuppressive functions. Indoleamine 2,3‐dioxygenase 1 (IDO1) and Arginase 1 (Arg1) are metabolic enzymes that, expressed by dendritic cells (DCs), contribute to immunoregulation. Here, we explored any possible link between IL‐35 and the activity of those enzymes. We transfected a single chain IL‐35Ig gene construct in murine splenic DCs (DC 35) and assessed any IDO1 and Arg1 activities as resulting from ectopic IL‐35Ig expression, both in vitro and in vivo. Unlike Ido1, Arg1 expression was induced in vitro in DC 35, and it conferred an immunosuppressive phenotype on those cells, as revealed by a delayed‐type hypersensitivity assay. Moreover, the in vivo onset of a tolerogenic phenotype in DC 35 was associated with the detection of CD25+ CD39+, rather than Foxp3+, regulatory T cells. Therefore, Arg1, but not Ido1, expression in DC 35 appears to be an early event in IL‐35Ig–mediated immunosuppression.


| INTRODUC TI ON
The control of immune response is operated by specialized cells, soluble molecules and membrane-bound signals, which modulate the intensity of immune reactivity and preside over the maintenance of homoeostasis. An imbalance between immunity and tolerance mechanisms can lead to pathological conditions, such as autoimmune diseases or neoplasia, characterized by excessive or deficient control of immune reactivity respectively.
Dendritic cells (DCs) are professional antigen-presenting cells with a key role in determining the outcome of the immune response, forcing naïve T cells into either activation or differentiation into regulatory T cells (Tregs). 1 The components of the local microenvironment critically take advantage of the plasticity of DCs, resulting in phenotype changes. The tolerogenic molecules CTLA-4, TGF-β and interleukin 35 (IL-35) are particularly effective in turning otherwise immunogenic CD8α − DCs into tolerogenic cells. 2,3 Reprogramming of a cell's phenotype involves an interplay between metabolic and immunological events known as cellular immunometabolism. 4 In CD8α − DCs, the increased metabolism of specific amino acids and the subsequent production of regulatory catabolites critically contribute to the acquisition of a newly expressed suppressive phenotype. The amino acid degrading enzymes indoleamine 2,3-dioxygenase 1 (IDO1) and arginase 1 (Arg1) are major components of immunometabolic pathways in DCs. 5 Interleukin-35 is a heterodimeric cytokine belonging in the IL-12 family. It powerfully dampens immune responses by suppressing T-cell proliferation and inducing the expansion of specific subsets of Tregs and regulatory B cells. Although elevated tissue and plasma levels of IL-35 are associated with a poor prognosis in many malignant tumours, 6 the cytokine has a protective role in the prevention of autoreactivity in several experimental autoimmune models and in human autoimmunity as well. 7 IDO1 and Arg1 control tryptophan and arginine metabolism, respectively. IDO1 degrades the essential amino acid l-tryptophan to l-kynurenine. In T lymphocytes, l-tryptophan depletion activates an integrated stress response triggered by GCN2, inhibiting cell proliferation and inducing anergy by down-regulating TCR's ζ chain. 8,9 Moreover, l-kynurenine is an endogenous agonist of the Aryl Hydrocarbon Receptor, thus promoting the expansion of Treg cells and acting to up-regulate Ido1 expression in a feedforward loop in DCs. 10 Arg1 hydrolyses l-arginine into urea and l-ornithine, which is a substrate for ornithine decarboxylase (ODC), to produce polyamine pathway catabolites. Although l-arginine consumption by Arg1 is a well-known immunoregulatory mechanism at work in M2 macrophages and in myeloid-derived suppressor cells in many tumour settings, 11 only recently has the immunosuppressive function of polyamines been unveiled in DCs. 12 In the current study, we investigated the possible role of IDO1 and Arg1 enzymes as potential immunometabolic effectors downstream of the tolerogenic action of IL-35Ig in splenic CD8α − DCs.

| Mice
Eight-to ten-week-old female C57BL/6 mice were purchased from Charles River Breeding Laboratories and Ido1 −/− mice from the Jackson Laboratory. All in vivo studies were in compliance with National and Perugia University Animal Care and Use Committee guidelines.

| Dendritic cell purification, transfection and treatment
Splenic DCs were fractionated using positive selection columns combined with CD11c and CD8 MicroBeads (Miltenyi Biotec, Germany). 13 Purified CD8α − DCs were transfected by DOTAP (Roche, USA) with IL-35Ig or control Ig gene constructs 3 and incubated overnight before in vitro analysis or in vivo administration.

| Real-time PCR and cytokine measurement
Real-time PCR analyses for mouse Ido1, Arg1 and Gapdh were carried out using previously reported specific primers. 12 Values were calculated as the ratio of the specific gene to Gapdh expression, as determined by the relative quantification method (ΔΔCT; means ± SD of triplicate determination). 12 Mouse TGF-β (Affymetrix, Santa Clara, USA), IFN-γ and IL-4 (Thermo Fisher Scientific, USA) ELISA kits were used to measure cytokines concentrations in culture supernatants.

| In vivo treatment, skin test assay and flow cytometry
The skin test assay has previously been described. 3,14 Briefly, puri-

| Statistical analysis
In vitro data were analysed by unpaired Student's t test. In the skin test assay, paired data were evaluated by paired Student's t test in each group of mice, using the vehicle-injected footpad of individual mice as an internal control.

| Ectopic IL-35Ig induces in vitro Arg1, but not Ido1, in DC 35
The ectopic expression of IL-35Ig, after transfection of the gene construct into murine splenic CD11c + CD8α − DCs, was previously demonstrated to confer powerful immunosuppressive properties on those cells. The presentation of diabetogenic autoantigen IGRP by DC 35 in prediabetic NOD mice protected animals from the occurrence of overt diabetes by a long-lasting antigen-specific tolerance. 3 To interrogate the effector mechanisms underlying the immunosuppressive outcome of IL-35Ig transfection in DCs responsible for the long-term tolerance observed in vivo, 3  transfected with Ig tag) were incubated for 6, 24 or 30 hours after transfection. Although Ido1 expression was similar in DC 35 and DC Ig over time, Arg1 was significantly increased in DC 35 relative to DC Ig at 24 hours (3.9-fold) and at 30 hours (2.2-fold) ( Figure 1A). IFN-γ, IL-4 and TGF-β, the most potent inducers of Ido1, Arg1 or both, respectively, 12 were not differentially secreted by DC 35 and DC Ig in culture supernatants at 24 hours post transfection ( Figure 1B). Therefore, besides the mere production of a tolerogenic cytokine, DC 35 seems to be endowed with an additional suppressive immunometabolic effector mechanism, namely, the expression of Arg1 induced by ectopic IL-35Ig.

| Arg1 is required for the tolerogenic effect of DC 35 in vivo
To confirm the selective involvement of Arg1 ( Figure 1A) (Figure 2A). Interestingly, in accordance with a previous study on the protective effect of DC 35 in autoimmune diabetes, 3 an increased percentage (8.6%) of CD25 + CD39 + T cells, rather than Foxp3 + T cells, was observed in DC 35 -sensitized group relative to DC Ig -sensitized group ( Figure 2D and E). These data confirm that DC 35 presented HY peptide in a tolerogenic manner and triggered a suppressive response mediated by Arg1 activation and involving CD25 + CD39 + , rather than Foxp3 + , regulatory T cells.

| D ISCUSS I ON
The immunosuppressive role of IL-35 has been observed and confirmed in many different studies, 6    The finding that Arg1 is a downstream effector of IL-35 has immunological relevance for several aspects. In IL-35-producing DCs (i.e., DC 35 , and most likely IL-35 + DCs 18 , as well) Arg1 induction might F I G U R E 2 Arg1 is the effector enzyme required for the immunosuppressive action of DC 35 . A, Schematic representation of in vivo treatments. Mice were sensitized on day 0 by intravenous (i.v.) injection of HY-loaded DC 35 or DC Ig . Two weeks later, they were challenged intrafootpad (i.f.p.) for skin test assay or intraperitoneally (i.p.) boosted for flow cytometry analysis, both with HY peptide. On day +15, DTH response was recorded, or MLNs were harvested for T-cell staining. B and C, Skin test reactivity of mice sensitized with DC 35 or DC Ig . Splenic HY-pulsed immunostimulatory CD11c + CD8α − DCs combined with a minority fraction (5%, indicated) of DC 35 or control DC Ig were i.v. transferred into syngeneic C57BL/6 recipient female mice to be assayed for skin reactivity to the eliciting peptide. The minority fractions were purified from either wild-type (wt DC) or Ido1 −/− mice (Ido1 −/− DC) (B), and wild-type DCs was either untreated or pretreated in vitro with the arginase inhibitor nor-NOHA 1.5 h before transfection (C). Skin reactivity of the recipient mice (n = 6 per group) to the eliciting peptide is represented as change in weight of treated footpads vs vehicle-receiving counterparts. Results are representative of two independent experiments (mean ± SD). Significance is referred to a two-tailed paired Student's t test (experimental vs control footpads) in each group of mice. **P < 0.01; ***P < 0.001. D and E, Flow cytometry analysis of Foxp3 + and CD25 + CD39 + cells among CD4 + T-cell population of MLN at day +15, after i.v. sensitization with HY-loaded DC 35 or DC Ig (day 0) and i.p. boost with HY peptide (day +14) (pools of five mice per group). Isotype controls were included in the analysis and number (upper right quadrant) indicates the percentage of doublepositive cells.