IL-33 improves the suppressive capacity of human regulatory T cells

One of the main problems that immunologists have not solved yet involves the immunosuppression of ongoing exacerbated responses, as in transplant rejection and autoimmune diseases. To overcome these difficulties, different approaches have been used including the administration of immunosuppressive drugs, monoclonal antibodies, and more recently, the utilization of cellular therapy. In this last strategy, huTregs have been proposed as a safe tool to control transplant rejection or collateral immune reactions such as GvHD. Tregs are a subtype of CD4+ T cells characterized by the expression of markers such as CD25, CD127, Foxp3, among others, and their capacity to induce immune tolerance. For this reason, huTregs have become an attractive target for their manipulation and application in the clinic. In this work, we used an established protocol for the expansion of peripheral blood-derived huTregs, adding IL-33 as a putative enhancer of huTregs function, due to its previously identified capacity for driving transplant tolerance. Here, we characterized the phenotype, expansion rate and suppressive function of huTregs 33 . Although IL-33 did not modify the expression of Tregs canonical markers nor their proliferative activity, it did potentiate their suppressive function. This remarkable observation may be driven in part by the up-regulation of the immune regulatory-related genes, Foxp3, Amphiregulin and ST2. Thus, huTregs 33 should be considered for pre-clinical and clinical studies.


Introduction
Human regulatory T cells (huTregs) have become a target for cellular therapy. In contrast to murine Tregs, huTregs correspond to 5% of Peripheral blood mononuclear cells (PBMC) and express the surface markers CD4, CD25 and low/negative expression of CD127, and transcription factor Forkhead-box P3 (FoxP3), characteristic of murine and human Tregs [1].
Based on their pivotal role in the maintenance of immune tolerance, huTregs isolation and expansion have been improved recently with the purpose of being applied as cellular source of therapy. The most reported protocols include anti-CD3/anti-CD28-coated beads for T cell stimulation, along with administration of human IL-2 and the immunosuppressor drug Rapamycin [2,3], which favors Tregs growth instead of the proliferation of effector CD4+ T. To date, several clinical studies have been performed, and the data indicates that the administration of huTregs is safe even in high doses [4] and seems to ameliorate the symptoms of Graft versus Host Disease (GvHD) in some cases; however, this has not been accomplished in all patients, thus demonstrating the need for more knowledge and treatment alternatives [5][6][7][8]. One of the limitations of huTregs effectiveness is the potential lost of their functional stability (or plasticity); in other words, huTregs may convert from a suppressive to an inflammatory phenotype [9]. Among the molecules that could be considered to optimize the production of huTregs, we sought to study IL-33 based on its recent role in the generation of tolerance [10,11]. This cytokine belongs to the IL-1 superfamily of cytokines, and it was initially associated with inflammatory processes such as in asthma, and other CD4+ T helper 2 related diseases, but newer studies have described a role in tissue repair and suppression of organ rejection through the modification of Tregs biology [12,13]. With these antecedents, we decided to use IL-33 as a novel factor for obtaining better and more functional huTregs.

IL-33 does not modify either the phenotype nor cell expansion of human Tregs
With the aim to improve one of the established protocol for purification and expansion of huTregs, we tested the effects of IL-33 on the in-vitro amplification protocol. First, we isolated huTregs from PBMC and cultured them with human IL-2, Rapamycin and IL-33. Freshly isolated huTregs display a canonical phenotype given by high expression of CD25 (∼80% of cells), low levels of CD127 (∼90% of cells), in addition to high expression of Foxp3 (> 80% of cells), Figure  1A and 1B. In this analysis, we also included the surface marker Nrp1 (because it has been related to Tregs identity) and ST2 or IL-33R (due to the addition of IL-33 in the cultures), obtaining low expression for both of them (< 4% of cells), Figure 1B. In general, all huTregs samples obtained from 5 different donors showed a comparable phenotype with no significant differences in the expression of the molecules analyzed. Next, we cultured huTregs in the presence of human IL-2 and Rapamycin, stimulated with anti-CD3/anti-CD28-coated beads, in the presence (huTregs 33 ) or not of IL-33. huTregs and huTregs 33 cells were recovered every 12 days for re-stimulation and media replenishment. At these time points we assessed for cell number and viability, obtaining no differences in total cell number or expansion rate between huTregs and huTregs 33 cells, Figure 2A. At the end of the culture (day 36), the phenotype of huTregs and huTregs 33 was studied by flow cytometry, obtaining a high percentage of CD4+CD25+CD127 low T cells in both Our evidence shows that the adition of IL-33 to huTregs expansion protocol boost the generation of suppressive huTregs 33 . This observation establishes a new precedent in the improvement of huTregs expansion, which is very significant for use in cellular therapy.

Discussion
One of the main current problems in transplantation is the inability of assuring the acceptance and survival of the graft. For this purpose several therapies have been developed; however, all these treatments are associated with side effects that harm the patient´s life [19]. Based on what has been described in transplantation clinical trials, the function and application of huTregs are quite promising. Until now, it has been possible to generate protocols for the purification and expansion of huTregs from healthy donors and patients, with the following infusion for treating conditions as transplant rejection, GvHD or autoimmune diseases [1,2,4,20].
Here, we attempted to modify a huTreg expansion protocol with the aim to improve either the rate of huTregs proliferation or the function of the cells. For this intent, we selected IL-33, a cytokine with a wide range of functions, including the ability to induce Tregs 13 and enhance their suppressive function [21]. Our findings show that the established markers for huTregs phenotype (CD4+CD25+CD127 low Foxp3+) do not change upon IL-33 treatment, obtaining a phenotype for huTregs 33 similar to control cells. However, when we performed a suppression assay, huTregs 33 were more suppressive than control huTregs. This remarkable finding gives us the idea that even if IL-33 is not changing the phenotype of huTregs (at least not modifying the expression of CD25, CD127, Nrp1 or ST2), IL-33 may be targeting other mechanisms in the cell, resulting in huTregs 33 becoming better suppressors. In the process of Tregs differentiation, the concepts of stability and/or plasticity have been recognized since Tregs facing an inflammatory milieu may become unstable, down-regulating the expression of Foxp3 [22,23], or they may convert to inflammatory conditions (>90%), Figure 2B, which indicates that IL-33 does not change the phenotype of huTregs in the expansion process. Regarding Foxp3, a decrease in the initial level of expression is observed for huTregs and huTregs 33 (∼70% at day 0 versus ∼30% at day 26 for huTregs and ∼45% for huTregs 33 ), and for Nrp1 and ST2 expression we did not obtain variations among culture timing or presence of IL-33.

IL-33 boosts the suppressive function of huTregs
Since the use of IL-33 did not show an apparent effect in huTregs expansion, we sought to evaluate the suppressive function of huTregs. For this purpose, we co-cultured CFSE-labeled CD4+CD25-T cells (or responders) with different ratios of huTregs (or huTregs 33 ) in the presence of polyclonal stimulation by anti-CD3/anti-CD28 beads. After 3 days, cells were recovered and CFSE dilution was studied by flow cytometry. As shown in Figure 3A, ∼50% of responders cells proliferated upon anti-CD3/anti-CD28 stimulation, whereas ∼25% was observed for responders cells co-cultured with huTregs. Surprisingly, huTregs 33 cells inhibited even better than huTregs, bringing responders cells proliferation to only ∼15%. This potent inhibitory effect was observed for all ratios tested, Figure 3B. Considering this result, we inferred that IL-33 may influence huTregs function, instead of cell expansion or phenotype, possibly by boosting their cell stability. Thus, we decided to evaluate the expression of certain genes involved in Tregs activity/stability such as Foxp3, ST2 and Amphiregulin (Areg) using quantitative PCR (qPCR), Figure 3C. As mentioned before, Foxp3 has been established as the master transcription factor for Tregs and is involved in the maintenance of the regulatory phenotype and suppressive function [14,15]; ST2 is the receptor for IL-33 and it has been linked with Tregs function in gut mucosa [11,16]; Areg, on the other hand, is an epidermal growth factor that improves Tregs function in vitro and in-vivo mouse models of colitis and tumor, favoring Tregs stability, avoiding conversion toward an inflammatory phenotype [17,18]. For these three genes studied, we observed a clear trend in the up-regulation of their mRNA, although the variations were not significantly different. Based on the results described in this study, we state that the inclusion of IL-33 in the expansion culture for huTregs cells by gaining a Th17-like phenotype [24]. In this regard, molecules such as ST2 and Areg have been implicated in the maintenance of Tregs phenotype and function. For instance, the receptor for IL-33, or ST2, has served as a marker for highly activated murine Tregs with a Th2-like phenotype [25], and also for suppressive Tregs residing in the intestine [11]. Conversely, the treatment with IL-33 (for 24 hours) of freshly isolated huTregs from diabetic patients has been tested, resulting in an increased ST2 expression [26], no variation in Foxp3 levels, but greater capacity to block IFN-γ expression on effector T cells [20]. Comparing this last study with ours, it seems surprising that IL-33 only improves huTregs function from diabetic, and not healthy donors (as in this current report). Also, it is not reported whether IL-33-treated diabetic huTregs inhibit effector T cell proliferation. We believe that age of PBMC donors and in-vitro conditions (media, activation and duration of the cell culture) may affect the biology of the final huTreg population. Following the same line of what is reported for ST2, Areg seems to enhance the suppressive activity of tissue-resident Tregs as shown in mouse models of carcinoma and viral infection [27,28]. Supporting our findings, Carney et al. reported that anti-CD3 activated huTregs up-regulate Areg at the gene and protein levels [18], and even more ensuring are the observations stating that IL-33 is involved in Areg synthesis [29,30].
Taking all of this together, we think that the inclusion of IL-33 in the expansion protocol for human Tregs is a novel improvement for obtaining up-graded regulatory T cells, which may ease the manufacture and administration of these cells in the clinical setting.

Isolation and expansion of Human Regulatory T cells
Blood samples were obtained from healthy donors recruited at the Blood Bank of Hospital Militar de Santiago (La Reina, Chile) and Blood Bank of Clinica Santa Maria (Providencia, Chile). The blood was processed to obtain PBMC by Ficoll-Paque density gradient. Cells were recovered and washed twice with 1X PBS for further huTregs isolation using the CD4+CD25+ T regulatory cells isolation kit from Miltenyi (California, USA). For huTregs culture and expansion we followed a protocol from Safinia et al., 2016. Briefly, 1x10 6 cells were stimulated with anti-CD3/anti-CD28 beads (Invitrogen, USA) and cultured in X-VIVO 15 media (Lonza, Basel, Switzerland) complemented with 5% of AB+ plasma, 500 U/ml of human IL-2 (Miltenyi Biotech) and 100nM Rapamycin (SigmaAldrich). Addicionally, half of the sample was stimulated with 20 ng/ml of rhIL-33 (Peprotech). Media including fresh IL-2 and Rapamycin, was replaced every 2 days, antiCD3/anti-CD28 beads were replenished every 12 days. At day 36, expanded huTregs were recovered, counted and analyzed for phenotype and function.

Statistical analysis
Data were analyzed using an unpaired Student's t-test or a Mann-Whitney test (two-tailed). In all cases, P < 0.05 was considered with statistical significance. For data analysis, GraphPad Prism v5.0 (GraphPad Software, CA, USA) was used.