Regulatory T cells from patients with end-stage organ disease can be isolated, expanded and cryopreserved according good manufacturing practice improving their function

Background Here, we isolated, expanded and functionally characterized regulatory T cells (Tregs) from patients with end stage kidney and liver disease, waiting for kidney/liver transplantation (KT/LT), with the aim to establish a suitable method to obtain large numbers of immunomodulatory cells for adoptive immunotherapy post-transplantation. Methods We first established a preclinical protocol for expansion/isolation of Tregs from peripheral blood of LT/KT patients. We then scaled up and optimized such protocol according to good manufacturing practice (GMP) to obtain high numbers of purified Tregs which were phenotypically and functionally characterized in vitro and in vivo in a xenogeneic acute graft-versus-host disease (aGVHD) mouse model. Specifically, immunodepressed mice (NOD-SCID-gamma KO mice) received human effector T cells with or without GMP-produced Tregs to prevent the onset of xenogeneic GVHD. Results Our small scale Treg isolation/expansion protocol generated functional Tregs. Interestingly, cryopreservation/thawing did not impair phenotype/function and DNA methylation pattern of FOXP3 gene of the expanded Tregs. Fully functional Tregs were also isolated/expanded from KT and LT patients according to GMP. In the mouse model, GMP Tregs from LT or KT patient proved to be safe and show a trend toward reduced lethality of acute GVHD. Conclusions These data demonstrate that expanded/thawed GMP-Tregs from patients with end-stage organ disease are fully functional in vitro. Moreover, their infusion is safe and results in a trend toward reduced lethality of acute GVHD in vivo, further supporting Tregs-based adoptive immunotherapy in solid organ transplantation. Electronic supplementary material The online version of this article (10.1186/s12967-019-2004-2) contains supplementary material, which is available to authorized users.


Background
Organ transplantation is the only curative treatment for patients affected by end-stage liver or kidney disease. Although immunosuppressive (IS) drugs have reduced the incidence of acute rejection and transplantationassociated mortality, their administration is associated with major side effects, such as opportunistic infections, damage of transplanted organs and secondary cancer [1,2]. Therefore, new strategies to improve long-term graft survival while reducing/eliminating IS therapy and, ideally, inducing immunological tolerance are needed.
Cellular therapy can improve the outcome of solid organ transplantation through anti-inflammatory effects and the induction of immune tolerance. T cell-mediated immunomodulation is one of the main mechanisms to maintain operational tolerance (withdrawal of immunosuppressive drugs while maintaining normal graft function and histology) in vivo. It is now accepted that regulatory T cells (Tregs), a subpopulation of T helper lymphocytes, are responsible for this immunomodulatory activity, as a result of their suppressive effects directly on effector T cells and on antigen presenting cells [3,4]. Circulating Tregs constitutively express CD25 and FoxP3 and represent 5-10% of all peripheral CD4+ T cells [5,6]. Tregs, which are fundamental for the maintenance of immune homeostasis, demonstrated a key role for transplantation tolerance in animal models by impairing the function of CD8+ T cells [7][8][9][10]. In humans, cell therapy with human Tregs for the induction of transplantation tolerance represents a promising strategy [10,11]. Indeed, clinical trials using this approach have already demonstrated that expanded polyclonal and antigenspecific Tregs are safe and effective in the treatment of GVHD and Type 1 diabetes [12][13][14][15][16][17]. Conversely, very few results have been published on clinical trials testing the efficacy of Tregs in solid organ transplantation [18,19]. Moreover, isolated/ex vivo expanded Tregs have been only tested in animal models [20][21][22].
Here, we first established a preclinical protocol for expansion/isolation of Tregs from patients with endstage liver or kidney disease being in the waiting list for liver/kidney transplantation (LT/KT). We then scaled up and optimized such protocol according to good manufacturing practice (GMP) to obtain high numbers of purified Tregs which were tested in vitro and in a xenogeneic acute graft-versus-host disease (GVHD) mouse model.

Patients
Peripheral blood (20-60 mL) was obtained from 14 LT and 9 KT patients. Patients characteristics are outlined in Table 1. Patient selection for GMP Tregs isolation/ expansion was based on the following inclusion criteria: (1) age ≥ 18 years; (2) diagnosis of end-stage kidney disease in waiting list for living donor kidney transplantation or diagnosis of end-stage liver disease in waiting list for liver transplantation. Exclusion criteria included: (1) HIV, HBV, HCV positivity; (2) syphilis antibody positivity; (3) combined transplant; (4) concurrent uncontrolled infection. Patient selection for in vitro experiments were as above described except virus positivity. Peripheral blood (20-60 mL) and buffy-coat (30 mL) were also obtained from age and sex matched healthy controls. Human studies were conducted in accordance with the Declaration of Helsinki and approved by the local ethical Committee (232/2015/O/Tess by Investigation Drug Service, Azienda Ospedaliero-Universitaria di Bologna). Informed consent was obtained from all the subject prior to enrollment into the study.

Circulating Treg enumeration
Enumeration by flow cytometry of circulating Treg (CD4+CD25+CD127−FoxP3+) was carried out in the peripheral blood (PB) of selected KT and LT patients (n = 7 and n = 10, respectively) and of healthy controls (n = 9). The conjugated monoclonal antibodies used are shown in Additional file 1: Table S1. Surface marker staining was performed for 15 min at room temperature. For intracellular staining, anti-human FoxP3 (PCH101) Staining Set PE Kit was used (eBiosciences), according to the manufacturer's instructions. Isotype control rat IgG2 PE was used as a control. Briefly, cells were stained for surface markers CD4, CD25 and CD127, washed once in PBS and then fixed/permeabilized. After washing, cells were incubated with anti-human FoxP3 antibody for 30 min at 4 °C in the dark. A lysis buffer (Becton-Dickinson) was used in order to lysate red blood cells. The phenotype of Tregs was analyzed by flow cytometry FACSCantoII (Beckton Dickinson). Data were analyzed using the FACSDiva software (Becton-Dickinson). The percentage of positive cells was calculated by subtracting the value of the appropriate isotype controls. The absolute number of positive cells per µL was calculated as follows: percentage of positive cells × white blood cell count (WBC)/100.

Tregs isolation and expansion
EDTA-anticoagulated peripheral blood (60 mL) was collected from 4 LT patients, 2 KT patients and buffy-coat (30 mL) from 5 controls. Peripheral blood mononuclear cells (PBMC) were then isolated by Ficoll-Hystopaque density gradient centrifugation. Isolation: freshly isolated CD8−CD25+ T cells were purified from PBMC by negative selection of CD8+ T cells followed by positive selection of CD25+ T cells using specific Miltenyi-Biotec Beads (CD8 microbeads human and CD25 microbeads II human) with Midi-MACS separator and a purity (CD4+CD25+) of > 90%.
Expansion: freshly isolated cells were plated at 1 × 10 6 / mL cells and activated with anti-CD3/CD28 coated beads (Invitrogen, Paisley, UK; Miltenyi Biotech) at a 4:1 bead:cell ratio at day 0 and then 1:1 bead:cell ratio weekly. Cells were expanded in culture media (TECSMacs GMP medium, Miltenyi Biotech) 5% human AB plasma containing rapamycin (100 nM) (Rapamune ® , Wyeth, USA) for 21 days at 37 °C and 5% CO 2 . IL-2 (1000 IU/mL, Proleukin ® , Novartis, UK) was added at day 4 post-activation and replenished every 2 days. Cells were restimulated with beads every 7 days. After 21 days of culture, beads were magnetically removed and the cells washed in TECSMacs GMP medium. After washings, fresh beads, rapamycin and IL-2 were added. Expanded cells were used for further analysis at each time of re-stimulation up until day 21 of expansion.

FOXP3 promoter demethylation analysis
The highly conserved Treg-specific demethylation region (TSDR) within the human FOXP3 gene is demethylated exclusively on Tregs and not in any other blood-cell types. Specific DNA methylation of the Treg FOXP3 is referred to intron 1, as previously identified [23]. DNA methylation of FOXP3 gene (CNS2 region) of expanded cells (day + 21), either freshly isolated or after cryopreservation/thawing, was evaluated for a total of 6 subjects (3 patients: 1 KT and 2 LT patients; and 3 controls). Genomic DNA was extracted from purified Tregs using the AllPrep DNA/RNA Mini Kit (QIAGEN, Hilden, Germany). 500 ng of DNA was bisulphite-converted using the EZ DNA Methylation Kit (Zymo Research Corporation, Orange, CA) according to manufacturer's instructions, except for the thermal conditions of the conversion (21 cycles of 55 °C for 15 min and 95 °C for 30 s). Bisulphite-treated DNA was eluted in 100 μL of water. DNA methylation analysis of the genomic region chrX:49,117,049-49,117,467 within FOXP3 gene body was performed by EpiTYPER assay (Sequenom, San Diego, CA), a quantitative method based on mass spectrometry that allows to evaluate methylation level at single CpG sites/groups of adjacent CpG sites (CpG units). Ten ng of bisulphite-treated DNA were PCR-amplified and processed following manufacture's instructions. The bisulphite specific primers were: FOXP3 forward, AGG AAG AGA GAT TTG TTT GGG GGT AGA GGA TTTA; FOXP3 reverse, CAG TAA TAC GAC TCA CTA TAG GGA GAA GGC TCA AAA AAA ACC AAA TCT TCA AAA CT.
For each gene, CpG sites with missing values in more than the 20% of the samples were removed, as well as samples with missing values in more than the 20% of CpG sites. The R package massArray was used to test if bisulphite conversion reaction run to completion [24]. For all samples analysed bisulphite conversion was from 98.9 to 100%.

Mixed leukocyte reaction assay
The T cells were used as autologous responder cells for in vitro suppression assays. The suppressive activity of freshly isolated and expanded cells were tested by co-culturing CD8−CD25− T cells, labeled with 5 µM carboxyfluorescein succinimidyl ester (CFSE, Invitrogen), with serial dilutions of freshly isolated (CD8−CD25+ T cells) and ex vivo expanded autologous Tregs in the presence of CD3/CD28 GMP beads at ratio previously shown by us (in house experiments) to be effective in dose-response experiments (1:10 CD8−CD25− T cells to bead ratio) in RPMI-1640 medium containing 10% FBS, 1% penicilline/ streptomycine and l-glutamine (1%) for 5 days at 37 °C and 5% CO 2 . Proliferation was analyzed by flow cytometry. The percentage of proliferative CD8−CD25− T cells in the absence of Tregs was taken as 100% proliferation.

GMP Treg isolation and expansion for in vivo experiments
The GMP manufacturing was performed in the Cell Factory of Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, certified in compliance with European GMP regulations by the Italian Drug Agency (authorization number aM-51/2018). Steady state leukapheresis from 2 patients (1 KT and 1 LT patient) were processed using GMP-compliant devices and reagents (Additional file 1: Figure S1A). The subjects were evaluated for venous accesses suitability. Leukapheresis was carried out by the COM. TEC ® (Fresenius Kabi AG, Bad Homburg, Germany) cell separator. The treatment of two blood volumes was set up as the procedure end-point. ACD-A was used for anticoagulation at a ratio of 1:14-1:13. For prophylaxis of citrate-related hypocalcemia, calcium gluconate was administered intravenously during the leukapheresis.
CD8−CD25+ cells were purified using the Clini-MACS System (Miltenyi Biotec, Bergisch Gladbach, Germany) according to GMP procedures and following manufacturer's instructions. Briefly, we performed a sequential two-step process based on the magnetic depletion of CD8+ T cells, followed by a positive selection of CD25+ cells after an over night storage, using monoclonal antibodies anti-CD8 (Miltenyi Biotec) and anti-CD25 (Miltenyi Biotec), as described in the CliniMACS user manual for cell preparation, magnetic labelling, and selection.
Negative fraction (CD8−CD25− T cells) after GMP selection at day 0 and the final product after GMP expansion at day 21, were cryopreserved and thawed as above described.

Cryopreservation and thawing of freshly isolated and expanded cells
Freshly isolated fractions (CD8−CD25+ and CD8− CD25− T cells) and expanded cells (after 7, 14 and 21 days of expansion) were washed and cryopreserved with 10% DMSO (CRYOSERV, Mylan Institutionals, Canonsburg, PA, USA), 10% HSA (HAS; Kedrion, Lucca, Italy) in a saline solution (B. Braun, Melsungen, AG, Germany) using a controlled-rate freezer. The frozen units were transferred and stored immediately to vapor-phase liquid nitrogen into dedicated tanks. All the cellular fractions were thawed at 37 °C and the viability, phenotype and suppressive function were assessed as above described.

Xenogeneic GVHD mouse model
NOD scid gamma (NSG) female mice (6-7 weeks of age) were purchased from Charles River (Calco, Italy). All animals were housed under specific pathogen-free conditions in compliance with guidelines of the San Raffaele Institutional Animal Care and Use Committee (IACUC number: 632). Mice were maintained for at least 5 days in the animal facility for acclimatization before transplantation. To promote the engraftment of transplanted cells injected intraperitoneally, all mice were conditioned with total body irradiation (1.75 cGy) on day 1. Of note, day 1 of the study is considered the day of transplant. All groups of mice, transplanted as well as control mice, were monitored for 7 weeks. Clinical signs of GVHD (e.g., hunched back, fur loss, skin inflammation) were monitored daily. Body weight was monitored throughout the duration of the study. Animals showing marked clinical signs as of distress, loss of weight equal to or greater than 20% of their starting weight were immediately sacrificed.
To establish whether the expanded Tregs could counteract acute GVHD, a xeno-GvHD model was induced by intraperitoneally transfer of human autologous CD8−CD25− T cells from 1 LT and 1 KT patient into mice. Both CD8−CD25− T cells and expanded Tregs were injected within 30 min from thawing (Additional file 1: Figure S1B). Specifically, human CD8−CD25− T cells (6 × 10 6 ) from LT or KT patients were injected with (simultaneous injection) or without autologous Treg (6 × 10 6 ) at a Treg: CD8−CD25− T cells ratio of 1:1. Timing of Treg infusion is crucial. In our model Tregs were infused at the time of T effector transplantation.

In vivo engraftment monitoring
To assess the in vivo persistence of human cells, transplanted mice were bled 4 weeks after transplantation and sacrificed 7 weeks after transplantation. Phenotype analysis of injected cells and of peripheral blood at 4 weeks (± 3 days) after transplantation and of PB and spleen at 7 weeks (± 3 days) after transplantation was performed (Additional file 1: Figure S1C). Following a Fc blocking step, cell surface staining was performed with antimouse CD45 and anti-human CD4, CD3, CD25, FoxP3 mAbs at 1:100 dilution in staining buffer (PBS, 2% FCS, 0.1% NaN3) (list of mAbs is in Additional file 1: Table S1).
To detect FoxP3, cells were treated and stained with the FoxP3 fixation/permeabilization kit according to the manufacturer's instructions (eBioscience). Samples were acquired on a BD FACSCanto II flow cytometer. Manual analysis of flow cytometry data was performed with FCS Express V4 (DeNovo Software, Glendale, CA).

Statistical analysis
Statistical analyses were performed using GraphPad Prism (GraphPad Prism 5.0 soft-ware, SanDiego CA, USA), and data are presented as mean ± standard error of the mean (SEM). For all experiments involving multiple comparisons, analysis of variance (ANOVA) followed by a Dunnett's post hoc test was used. For those who involved comparisons between two groups, Student's t-test was used. The level of significance was set at p ≤ 0.05.
When cryopreserved/thawed 21 days expanded Tregs were analyzed, the mean percentage of viable cells was always more than 75% with no significant differences between patients and controls (Fig. 2d). The mean percentage of CD4+CD25+CD127−FoxP3+ cells was always > 80% (Fig. 2e).
When functional assays were performed, freshly isolated cells (CD8−CD25+ T cells) from the three groups exerted a limited (healthy controls and LT patients) or poor (KT patients) suppressive effect toward autologous CD8−CD25− T cells proliferation in vitro (Fig. 3a). After 21 days of in vitro expansion, flow cytometry analysis demonstrated that, according to the increased T regulatory phenotype, Tregs from patients/controls had significant suppressive activity (Fig. 3b), which was not modified by cryopreservation/thawing (Fig. 3c).
All together these data demonstrate that our small scale Treg isolation and expansion protocol from LT and KT patients lead to functional Tregs which are suitable for cell therapy. Moreover, cryopreservation/thawing does not impair phenotype/function of the expanded Tregs. FOXP3 methylation analysis Figure 4a shows that in males the locus, which is localized on X chromosome, displayed low DNA methylation values (CpG sites ranging from 0.08 to 0.44). As expected, in females CpG sites had higher methylation values (ranging from 0.24 to 0.73) due to X chromosome inactivation We evaluated DNA methylation differences at day + 21 in freshly isolated and cryopreserved/thawed expanded cells in patients and controls separately using paired t-test. The large majority of CpG sites was not significant, and intra-individual difference between the two groups were generally small, spanning between − 0.13 and 0.23. The only exception was the CpG_3 among controls (p-value: 0.002), but, also in this case, the difference between the freshly isolated and the cryopreserved/thawed cells was small, ranging from 0.06 to 0.07 (Fig. 4b).
Collectively, these data suggest that the DNA methylation pattern of FOXP3 gene is substantially conserved after cryopreservation/thawing of ex vivo expanded Tregs from LT and KT patients.

In vivo study
To investigate whether GMP expanded Tregs from 1 KT and 1 LT patient have inhibitory function in vivo after thawing, we used a xenograft model of lethal GVHD.
Specifically, irradiated NSG mice were injected with cryopreserved/thawed CD8−CD25− T cells either alone or in combination with autologous expanded cryopreserved/thawed Tregs at 1:1 ratio to assess their ability to ameliorate GVHD (Additional file 1: Figure S1). The rationale for this design is that early infusion of Tregs allows the cells to prevent T effector activation. After the immunomagnetic selection and expansion of 40 millions of freshly isolated CD8−CD25+ cells under GMP condition, we obtained 4.3 × 10 9 total cells (107.1-fold expansion) for the KT patient and 4.6 × 10 9 total cells (115.2-fold expansion) for the LT patient. Before expansion the percentage of CD45+/CD4+/CD25+ cells was 84.9% and 53.9% in KT and LT patient whereas, at the end of expansion, the percentage of CD45+/CD4+/CD25+ cells increased up to 98.3% and 89.2% respectively. At day 21 CD45+/CD4+/CD25+/CD127−/FoxP3+ cells were 79.2% and 84.4% respectively (data not shown).
Tregs were then thawed and characterized for the phenotype and in vitro function. As shown in Fig. 5a, b, the thawed Tregs were more than 80% CD4+CD25+CD127−FoxP3+ and highly suppressive.
Mice receiving KT CD8−CD25− T cells lost weight and all succumbed to GVHD by day 40. In contrast, mice receiving KT CD8−CD25− T cells and autologous Tregs showed delayed disease-associated weight loss and slightly increased survival rate (by day 48 17% of mice receiving CD8−CD25− T cells with Tregs from the KT patient survived versus 0% for mice receiving CD8−CD25− T cells only; p = NS) (Fig. 6a, b). Similar results were obtained with cells from the LT patient: by day 48 58% of mice receiving CD8−CD25− T cells with Tregs from the LT patient survived versus 12.5% for mice receiving CD8−CD25− T cells only (p = NS; Fig. 6c, d).
To assess human cells engraftment in PB, animals were bled after 4/7 weeks and the percentages of human CD45+ cells in blood was determined (Fig. 7). The gating strategy is shown in Fig. 7a. No human CD45+ cells were identified in the PB of mice after infusion of Tregs alone from KT or LT patients. Conversely, a wide distribution of human CD45+ cells was observed in the PB of mice after transplant of CD8−CD25− T cells alone or CD8− CD25− T cells plus Treg infusion (Fig. 7b, c).
To assess the biodistribution of infused human cells, spleen and PB cells of mice were stained with antihuman CD45, CD3, CD4, FoxP3 monoclonal antibodies after 7 weeks from transplant. The gating strategy is shown in Fig. 8a. By comparing PB and spleen, a higher mean percentage of human CD45+ cells was observed in the spleen versus PB of the mice after receiving CD8−CD25− T cells plus Treg of LT patients (p < 0.003; Fig. 8b). Of note, the majority of the human CD45+ cells were CD3+CD4+ in the blood/spleen of mice after receiving KT or LT cells (Fig. 8c). Human FoxP3+ Tregs were identified in the CD3+CD4+ fraction of PB and spleen of the two groups of mice without differences between KT and LT cells (Fig. 8d).

Discussion
The use of Tregs for the treatment of GVHD and for inducing operational tolerance after solid organ transplantation is a promising approach.
Here, we isolated, expanded and functionally characterized Tregs from patients with end stage kidney and liver disease, with the aim to establish a suitable method to obtain immunomodulatory cells for adoptive immunotherapy post-transplantation.
We demonstrated that fully functional Tregs from patients in waiting list for solid organ transplantation can be isolated/expanded for clinical studies. These cellular products show homogenous regulatory T cells characteristics with very low percentage of contaminating cells. Therefore, unwanted immune side effects are minimized. Of note, based on phenotype/function, the isolated/expanded Tregs from KT and LT patients are similar. Interestingly, our cryopreservation/thawing procedure allowed cell suspensions with phenotype/function and hypomethylation of the FOXP3 gene superimposable to that of the freshly isolated counterparts. This finding challenges the notion that cryopreservation of Tregs products is detrimental for their function [25,26].
Our freshly isolated cells from healthy donors and LT/ KT patients showed limited suppressive activity. This is in contrast with previous findings demonstrating 60-80% of suppressive activity by freshly isolated healthy donors cells, which were CD8−CD19−CD25+ cells [22] or CD8−CD25+ cells [27]. However, our results may be due (1) to the low purity of our freshly isolated cell suspension (CD8−CD25+ cells) and (2) to the fact that while our suppressive activity results were based on a 1:2 Treg/  CD8−CD25− T cells working ratio, their ratio was 1:1 Treg/Teff. Whether this results is also due to the fact that our freshly isolated healthy donors cells derive from 24 h aged buffy-coats remains a matter of speculation. Safinia et al. [10] firstly proposed a GMP production protocol to expand CD25+-enriched cells from PB in the presence of IL-2 and rapamycin to induce tolerance after liver transplantation. In their 36 days expansion protocol, multiple round of in vitro Treg stimulation were necessary to achieve clinically relevant Tregs number. In our protocol indeed, we obtained fully functional Tregs after 21 days of expansion containing less than 1% of contaminating CD8+ effector cells. Expanded, cryopreserved and thawed Tregs remain hypomethylated at intron 1 of the FOXP3 locus, confirming their epigenetic stability. Functionally, Tregs showed suppressive function against autologous CD8−CD25− T cells. Several human Treg products have been tested in animal models. However, previous data of Treg cell infusion in non human primates report conflicts regarding the ability of the infused cells to induce transplant tolerance [28][29][30]. Thus, we decided to use a mouse model of GVHD in immunodepressed mice [i.e. NOD-SCIDgamma KO (NSG) mice] receiving human effector T cells with or without GMP Tregs to prevent the onset of xenogeneic GVHD. In our mouse model, GMP Tregs from LT or KT patients proved to be safe and showed a trend toward reduced lethality of acute GVHD in vivo. Interestingly, GMP Tregs from LT and KT patients did not induce xenogenic GVHD and did not expand as documented by the absence of Tregs in the PB and spleen of mice after 4 and 7 weeks from transplantation. This is probably due to the lack of human IL-2.
Nevertheless, our data are consistent with previous models of xenogeneic GVHD and Treg infusion. The addition of CD25 expressing cells to human