IL-2 receptor engineering enhances regulatory T cell function suppressed by calcineurin inhibitor

Abstract

Clinical trials utilizing regulatory T cell (Treg) therapy in organ transplantation have shown promising results, however, the choice of a standard immunosuppressive regimen is still controversial. Calcineurin inhibitors (CNIs) are one of the most common immunosuppressants for organ transplantation, although they may negatively affect Tregs by inhibiting IL-2 production by conventional T cells. As a strategy to replace IL-2 signaling selectively in Tregs, we have introduced an engineered orthogonal IL-2 (ortho IL-2) cytokine/cytokine receptor (R) pair that specifically binds with each other but does not bind with their wild-type counterparts. Murine Tregs were isolated from recipients and retrovirally transduced with ortho IL-2Rβ during ex vivo expansion. Transduced Tregs (ortho Tregs) were transferred into recipient mice in a mixed hematopoietic chimerism model with tacrolimus administration. Ortho IL-2 treatment significantly increased the ortho IL-2Rβ(+) Treg population in the presence of tacrolimus without stimulating other T cell subsets. All the mice treated with tacrolimus plus ortho IL-2 achieved heart allograft tolerance, even after tacrolimus cessation, whereas those receiving tacrolimus treatment alone did not. These data demonstrate that Treg therapy can be adopted into a CNI-based regimen by utilizing cytokine receptor engineering.

1 ∣. INTRODUCTION

Current clinical trials have opened the possibility of utilizing regulatory T cells (Treg) for cell therapy in the field of organ transplantation,^1-4 though the selection of concomitant immunosuppression remains unclear. Several reports demonstrate that transplant recipients treated with calcineurin inhibitors (CNIs) exhibited a reduced number of Tregs,^5,6 suggesting these drugs are incompatible with Treg therapy. CNIs exhibit their immune-suppressive potential against T cells via inhibiting nuclear translocation of nuclear factor of activated T cells (NFAT) resulting in reduced IL-2 production that is fundamental for T cell activation and proliferation.⁷ In contrast to conventional T cells (Tcons), NFAT in Tregs constitutively localizes in the nucleus and activity is independent of calcineurins, thereby Tregs are capable of continuous proliferation in response to TCR-stimulation after CNI treatment in the presence of IL-2.⁸ However, since Tregs do not produce IL-2, CNIs deprive Tregs of IL-2 signaling provided from activated Tcons. We sought to address if IL-2 replacement can restore Treg function altered by CNI treatment. Indeed, it was reported that low dose IL-2 treatment restored Treg number and function that was decreased by CNI treatment.⁹ However, reports of IL-2 mediated NK cell activity, even at low IL-2 doses in the clinic,¹⁰ poses a risk in the transplantation setting. We introduce here an innovative approach that may resolve this issue by cytokine receptor engineering. In the previous study, we generated a mutant IL-2 cytokine/cytokine receptor pair named orthogonal (ortho) IL-2/ortho IL-2 receptor β (ortho IL-2Rβ) which binds with each other but not to their wild-type counterpart.¹¹ By transducing ortho IL-2Rβ into the target cells before in vivo transfer, ortho IL-2 cytokine can selectively stimulate the target cells that are expressing ortho IL-2Rβ. We tested this cytokine engineering strategy on Treg cell therapy and evaluated whether ortho IL-2 stimulation can restore the dysfunction of Tregs deprived of IL-2 by CNIs.

2 ∣. MATERIAL AND METHODS

2.1 ∣. Animals

Eight- to 12-week-old BALB/cJ mice were purchased from the Jackson Laboratory. CD45.1⁺ BALB/c and C57BL/6 mice were inbred for more than 20 generations. We purchased and bred BALB/c mice with a green fluorescent protein (GFP) reporter, expressed in conjunction with transcription factor Foxp3 (C.Cg-Foxp3^tm2tch/J; Foxp3^GFP+ BALB/c), to facilitate enrichment of highly pure Tregs. Animal protocols were approved by the Administrative Panel on Laboratory Animal Care at Stanford University (APLAC #9216).

2.2 ∣. Treg isolation and transduction

Foxp3^GFP+ Tregs were isolated and expanded as previously described.¹² Isolated cells were inoculated together with CD3/CD28 T cell activation beads (Invitrogen) at a 1:1 ratio in the presence of 1000 IU/ml recombinant human IL-2 (hIL-2, teceleukin, Roche) on 96 well flat plates. hIL-2 was replenished on day 2. Retroviral supernatant produced from HEK293T cell culture was placed on 24 well plates coated with RetroNectin (Takara). The supernatant was removed after 2-3 h of centrifugation at 4000 rpm at 32°C. Preactivated Tregs were recovered and inoculated into the virus-loaded plates together with refreshed hIL-2 (1000 IU/ml) and CD3/CD28 beads (1: 3 cell to beads ratio), and incubated for 72h at 37°C. hIL-2 was replenished on day 2. The transduced cells were identified as TurboRFP (tRFP)⁺ cells and the efficiency of transduction was about 10%. The manipulated Foxp3^GFP+ Tregs were treated as a mixture of tRFP⁺ ortho IL-2Rβ transduced- and tRFP⁻ untransduced-fraction (ortho Tregs).

2.3 ∣. Proliferation assay

Ortho Tregs were co-cultured with naïve T cells isolated from CD45.1⁺ BALB/c mice and incubated with CD3/CD28 T cell stimulation beads at ortho Tregs:Tcons:Beads 1:1:2 ratio for 4days. Wt and ortho IL-2 were added and replenished every 2days. Based on the previous data that showed affinity of ortho IL-2/ortho IL-2R was 100 times lower than that of wt IL-2/wt IL-2R,¹¹ we adjusted the cytokine concentration to 100 000 IU /ml for ortho IL-2 and 1000 IU/ml for wt IL-2. Tacrolimus (Prograf©, Astellas) was diluted in PBS and added to the well at 100 ng/ml. Four days later, the number of Foxp3^GFP+ ortho Tregs and that of Foxp3^GFP− Tcons were analyzed by LSRII (BD). Cell number in each well was calculated from the count of Precision counting beads (BioLegend) according to the manufacturer's instructions.

2.4 ∣. Blood sample analysis

Blood samples were obtained from the tail vein into heparin-PBS. After lysing red blood cells, cells were incubated with Fc-block followed by staining with the following antibodies purchased from BioLegend; H2K^b, H2K^d, CD45.1, CD45.2, TCRβ, CD4, CD8a, CD19, CD25, Gr1, and CD11b. Cells were acquired by LSRII (BD) and data was analyzed with FlowJo software (BD).

2.5 ∣. Bone marrow and ear pinna heart transplantation

Bone marrow and heterotopic heart transplantation were performed as described previously.^12-14 Briefly, bone marrow cell suspension collected from femurs and vertebraes were passed through a 70 μm mesh. After blood cell lysis, 15 × 10⁶ bone marrow cells were co-injected with 3 × 10⁶ ortho Tregs into recipient mice prepared with 3.3-Gy total body irradiation (TBI). Anti-CD40L (MR1; BioX-cell) was diluted with PBS and injected i.p. at 0.3 mg/mouse right after BMC transfer. The newborn heart was excised and placed into the recipient's ear pinna through a tunnel made from a small incision in the occipital skin. Heart graft viability was assessed with visual observation. The rejection of the graft was confirmed by histopathology at the end of the experiment.

2.6 ∣. Statistics

Statistical differences were calculated with unpaired t-test for two group comparison, two-way ANOVA with Bonferroni's multiple comparison test for multiple comparisons, and Log-rank test for heart graft survival by GraphPad Prism version 8.3.0 (GraphPad Software, LCC).

3 ∣. RESULTS

3.1 ∣. In vitro proliferation assay with and without tacrolimus

Flow cytometry-sorted Foxp3^GFP+ Tregs were transduced with ortho IL-2Rβ via retroviral vector (ortho Tregs). To evaluate the effect of CNIs on ortho Tregs in vitro, ortho Tregs were co-cultured together with naïve Tcons in the presence or absence of tacrolimu (Figure 1A). Addition of tacrolimus reduced not only the number of Tcons (Figure 1B gray bar, 25% reduction, p < 0.001) but also the number of ortho Tregs (Figure 1C gray bar, 32% reduction, p < 0.001), demonstrating the suppressive effect of tacrolimus on T cells. In both cultures with and without tacrolimus, wt IL-2 significantly increased the number of ortho Tregs (Figure 1C, orange bar). However, wt IL-2 increased the number of Tcons more (Figure 1B, orange bar), thereby significantly reducing the percentage of ortho Tregs (Figure 1D, orange bar). In contrast, ortho IL-2 stimulation did not increase Tcon number (Figure 1B, red bars) while ortho Treg numbers increased 1.8-fold higher without tacrolimus, and 2.9-fold higher with tacrolimus compared to the PBS control (Figure 1C, red bars). The increase was observed especially in the ortho IL-2Rβ high fraction, indicating the selectivity of ortho IL-2 (Figure 2). As a result, the ortho IL-2 plus tacrolimus condition showed the highest percentage of ortho Tregs (71.6% ± 1.9%, Figure 1D red bars). These data collectively demonstrate that ortho IL-2 and tacrolimus can function synergistically in the enrichment of ortho Tregs.

FIGURE 1.

Ortho IL-2 and tacrolimus synergistically increase %ortho Tregs. Foxp3^GFP+ Tregs transduced with ortho IL-2Rβ (ortho Tregs) and naïve Tcons were incubated with CD3/CD28 beads. Wt IL-2 (1000 IU/ml), ortho IL-2 (100 000 IU/ml), and tacrolimus (100 ng/ml) were added as indicated and replenished every 2 days. (A) Representative pseudocolor plots of co-cultured cells. (B-D) Box plots indicating the number of Tcons (B), ortho Tregs (C), and the proportion of ortho Tregs among total cells (D) on day 4. Quantification of triplicate wells in 1 representative experiment of 2 independent experiments. *<0.05, **<0.01, ***<0.001; p-values calculated by Dunnett test comparing each column to PBS-control (gray bar) among tacrolimus (−) or tacrolimus (+) group. ns, not significant.

FIGURE 2.

Ortho IL-2 increase ortho IL-2Rβ hi Tregs. (A) Representative pseudo color plot for tRFP (ortho IL-2Rβ) hi and low gate. The absolute number of ortho IL-2Rβ hi (B) and low (C) fractions in each treatment group are shown. Quantification of triplicate wells in 1 representative experiment of 2 independent experiments. *<0.05, **<0.01, ***<0.001; p-values calculated by Dunnett test comparing each column to PBS-control (gray bar) among tacrolimus (−) or tacrolimus (+) group.

3.2 ∣. Mixed chimerism model with tacrolimus

We previously reported that ortho IL-2 administration selectively increased ortho Tregs in vivo and facilitated tolerance induction in a mixed chimerism model.¹² Here, we added daily tacrolimus administration in this model to determine whether ortho IL-2 increases ortho Tregs in the presence of CNIs in vivo (Figure 3A). Consistent with our previous report without tacrolimus administration, ortho IL-2 administration significantly increased the percentage of Foxp3^GFP+ ortho Treg population in the presence of tacrolimus on d14 (Figure 3B, p = 0.03). The proportion of ortho IL-2Rβ hi significantly increased compared to that before the transfer in the ortho IL-2-treated group (37.9% ± 29.3%), whereas that was even reduced in the PBS group (6.6% ± 4.1%; Figure 4A,B). It seems ortho IL-2 selectively affects ortho IL-2 hi Treg fraction because the increase in ortho Tregs was observed only in the ortho IL-2Rβ hi fraction but not in the ortho IL-2Rβ lo fraction (Figure 4C). Ortho IL-2-treated animals trended to have a higher proportion of donor cells on d14 (Figure 3C; 61.7% ± 9.1% in the ortho IL-2 group compared to 42.3% ± 31 0.4% in the PBS group, p = 0.14). The positive correlation between ortho Treg expansion and donor cell chimerism on d14 suggests the potential of ortho Tregs to facilitate donor cell engraftment (Figure 5). We surmise that tacrolimus administration enhanced the early phase donor cell chimerism because the percentage of donor cells declined after tacrolimus cessation in both ortho IL-2 and PBS-treated groups (Figure 3C). Two out of 5 chimeric mice in the PBS group eventually lost donor-specific chimerism by d180. Although the percentage of donor cells declined in the ortho IL-2-treated group as well, all of the mice sustained donor cell chimerism >1% until d180 (Figure 3C). At the experimental endpoint (d180), ortho IL-2-treated animals showed statistically significant higher proportion of donor cells compared to PBS-treated mice (28.9% ± 17.4% in ortho IL-2 group compared to 7.59% ± 10.4% in PBS group, p = 0.025). Consistently, only 2 of 7 mice (28.6%) sustained the heart allograft in the PBS group whereas all of the mice (7 of 7) in the ortho IL-2 group sustained transplanted heart allograft by d180 (Figure 6, p = 0.007). Taken together, we concluded that the orthogonal IL-2 can be incorporated with tacrolimus treatment to promote transplantation tolerance.

FIGURE 3.

Ortho IL-2 injection together with tacrolimus facilitates ortho Treg expansion and the establishment of heart transplant tolerance. (A) Schema of in vivo experiment. Anti-CD40L (MR1); 0.5 mg i.p. on d0. ortho IL-2; 25 000 IU, tacrolimus; 5 mg/kg, both i.p. from d0 to d14. (B) Left; representative pseudocolor plots for CD4⁺ gated cells on d14. Right; mean ± SD of %Foxp3^GFP+/CD4⁺ cells at indicated time points. (C) Mean ± SD of %donor (CD45.1+ cells/CD45⁺ PBMCs) at indicated time points, dashed line: 1%. Pooled data from two independent experiments with at least three mice per group. *p-values <0.05 calculated by unpaired t-test between indicated two groups.

FIGURE 4.

Ortho IL-2 increased the expression of ortho IL-2Rβ. (A) Representative pseudocolor plot of CD4⁺Foxp3^GFP+ gated cells in the peripheral blood on d14 showing gating strategy for ortho IL-2Rβ (tRFP) hi and low fraction. Mean ± SD with individual plots showing the proportion of ortho IL-2Rβ hi in Foxp3^GFP+ cells (B) and the proportion of ortho IL-2Rβ hi and low cells among CD45.2 (C). Pooled data from two independent experiments with at least three mice per group. *p-values <0.05 calculated by unpaired t-test between indicated two groups.

FIGURE 5.

The proportion of ortho Treg and donor cell chimerism shows positive correlation. Pearson's correlation for %Foxp3^GFP+/CD4⁺ cells (x-axis) and %donor (y-axis) on d14. Each dots represent individual mouse in ortho IL-2-treated (red) and PBS-treated mouse (blue). R = 0.6, p = 0.023.

FIGURE 6.

Combination of ortho IL-2 plus tacrolimus treatment promotes heart allograft tolerance. Heart allograft survival curve. **p-value <0.01 calculated by Log-rank test. Red, ortho IL-2-treated, N = 7; blue, PBS-treated, N = 7. Pooled data from two independent experiments.

4 ∣. DISCUSSION

In most clinical trials for organ transplantation tolerance, standard immunosuppressants including CNIs are initiated following the introduction of the transplanted organ and are tapered during the maintenance phase.^4,15 However, since CNIs can affect both effector T cells (Teffs) and Tregs, CNI induction may impede the establishment of operational tolerance induced by Tregs.¹⁶ Indeed, our preliminary experiment revealed trend of decrease in ortho Treg proportion after 7-day-tacrolimus administration (Figure S1). Interestingly, although the proportion was not fully recovered, ortho IL-2 administration increased ortho Treg proportion compared to PBS-control even under tacrolimus administration. In the current study, we administered tacrolimus for 14 days after the tolerance induction. Although both PBS and ortho IL-2-treated groups showed mixed chimerism at the induction phase, 71.4% of the PBS group eventually rejected chimerism and failed to establish heart transplant tolerance. This result is consistent with the previous report that showed tacrolimus administration reduced the proportion and the suppressive function of Tregs and resulted in the failure of transplantation tolerance.¹⁷ In contrast, the mice treated with ortho IL-2 sustained donor cell chimerism and tolerance by d180, indicating that ortho IL-2 treatment can reverse the negative impact of CNIs on Tregs. Consistent with our data, Whitehouse et al. previously demonstrated that low dose IL-2/anti-IL-2 antibody complex therapy improved Treg function deteriorated by tacrolimus and prolonged skin allograft survival.⁹ Although the direct comparison between low dose IL-2 and ortho IL-2 in vivo was not performed in the current experiment, we demonstrated here that the cytokine receptor engineering approach is capable of selective IL-2/STAT5 signaling in ortho Tregs and does not interfere with the immunosuppressive effect of tacrolimus (Figure 7).

FIGURE 7.

Ortho IL-2 replaces IL-2 signaling in ortho IL-2Rβ⁺ Tregs that was deprived by calcineurin inhibitor treatment. (Left) TCR stimulation activates NFAT and promotes transcription of IL-2 in Teffs. Wt IL-2 secreted from activated Teffs stimulate both Teffs and Tregs. (Mid) CNIs inhibit NFAT activation and result in IL-2 deprivation for both Teffs and Tregs. (Right) ortho IL-2 restores IL-2 signaling through transduced ortho IL-2R in Tregs but does not affect wt IL-2R on Teffs. Created with BioRender.com.

A limitation of this study is the low transduction efficiency of the Treg with the ortho IL-2R. In our previous report, we transferred 1 × 10⁶ ortho Tregs with ~30% transduction efficiency. Since the efficiency was only ~10% in the current study, we increased the total number of transferred cells from 1 × 10⁶ to 3 × 10⁶/mouse. The rate of mixed chimerism in the PBS group was higher in the current study (70%) compared to the previous study (20%). This is probably due to the higher number of ortho Treg infused or simply due to tacrolimus used in the induction phase. Despite a lower transduction efficiency in the current study, ortho IL-2 administration resulted in an increased number of ortho Tregs and enhanced donor cell chimerism. Furthermore, ortho IL-2-treated mice did not show delayed donor chimerism loss or allograft rejection that was observed in the PBS-treated control mice. These data collectively suggest that ortho IL-2 and tacrolimus synergistically function for enriching ortho IL-2Rβ⁺ Tregs in vivo as in the same manner as that shown in our in vitro experiments.

The efficiency of gene transduction can be improved by new genome editing methods.¹⁸ We used gammaretroviral vectors that are generally utilized for laboratory research for murine cell transduction. Since gamma retroviruses require host cell mitosis to integrate the target genome, target cells must be stimulated during transduction. This approach can result in cell exhaustion or activation-induced cell death. In contrast, lentiviral vectors can access chromosomes by active transport through the nuclear pore and possess the ability to deliver target genomes into non-dividing T cells.¹⁹ Safety and feasibility of lentiviral transduction have been proven in several clinical utilities such as manufacturing chimeric antigen receptor (CAR) T cells.²⁰ Future development of the CRISPR-Cas9 system in the clinic may also address some of the drawbacks of viral transduction such as a random integration of the gene of interest, or non-physiological gene regulations.^21,22 Those new technologies would promise cheap, efficient, and the most scalable genome editing.

Recently, human ortho IL-2/IL-2R pairs has shown the potential to selectively activate human T cells.²³ This product has a very low number of amino acid substitutions and is expected to have low immunogenicity. For clinical application, it is possible to implement strategies experienced with recombinant hIL-2 production, such as PEGylation to extend the half-life of the cytokine. In conclusion, the cell engineering approach could minimize the adverse effect of conventional immunosuppressants on Treg therapy and accelerate the development of clinical transplantation tolerance.

Abbreviations:

CNI

calcineurin inhibitor

IL-2R

IL-2 receptor

NFAT

nuclear factor of activated T cells

NK cell

natural killer cell

Ortho

orthogonal

TBI

total body irradiation

Tcons

conventional T cell

Treg

regulatory T cell

wt

wild-type

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.