CXCR5⁺CD8⁺ T cell–mediated Suppression of Humoral Alloimmunity and AMR in Mice Is Optimized with mTOR and Impaired with Calcineurin Inhibition

Abstract

Background:

Adoptive cellular therapy (ACT) with antibody-suppressor CXCR5⁺CD8⁺ T cells (CD8⁺ T_Ab-supp) inhibits alloantibody production, AMR, and prolongs graft survival in multiple transplant mouse models. However, it is not known how conventional immunosuppressive agents impact the efficacy of CD8⁺ T_Ab-supp ACT.

Methods:

We investigated the efficacy of CD8⁺ T_Ab-supp cell ACT when combined with inhibitors of calcineurin (CNi) or mammalian target of rapamycin (mTORi) in a murine model of kidney transplant (KTx).

Results:

ACT-mediated decrease in germinal center (GC) B cells, posttransplant alloantibody titer and amelioration of AMR in high alloantibody-producing CCR5 KO KTx recipients were impaired when ACT was combined with CNi and enhanced when combined with mTORi. CNi (but not mTORi) reduced ACT-mediated in vivo cytotoxicity of IgG⁺ B cells and was associated with increased quantity of GC B cells. Neither CNi nor mTORi treatment impacted the expression of cytotoxic effector molecules (FasL, Lamp1, perforin, granzyme B) by CD8⁺ T_Ab-supp after ACT. Concurrent treatment with CNi (but not mTORi) reduced in vivo proliferation of CD8⁺ T_Ab-supp after ACT. The increase in quantity of splenic CD44⁺CXCR5⁺CD8⁺ T cells that occurs after ACT was reduced by concurrent treatment with CNi but not by concurrent treatment with mTORi (dose-dependent).

Conclusion:

Impaired efficacy of ACT by CNi is attributed to reduced persistence and/or expansion of CD8⁺ T_Ab-supp cells after ACT. In contrast, concurrent immunosuppression with mTORi preserves CD8⁺ T_Ab-supp cells quantity, in vivo proliferation, in vivo cytotoxic effector function after ACT and enhances suppression of humoral alloimmunity and AMR.

INTRODUCTION

Development of de novo DSA, which occurs in 13–27% of kidney transplant (KTx) recipients,¹ drives acute AMR^2–4 and correlates with reduced long-term graft survival.^5–7 Prevention of de novo DSA and treatment of AMR are limited by gaps in our understanding about the biology of transplant humoral alloimmunity and the lack of specific treatment options. Current treatment strategies include corticosteroids, plasmapheresis, intravenous immunoglobulin and adjunct therapies, resulting in variable approaches and outcomes with potential deleterious sequelae.^8–10 Thus, there is an unmet need to develop definitive and specific therapies that target alloimmunity while sparing the recipient’s protective humoral immunity.

Cellular-based immunotherapy is emerging as a promising approach for induction of immune tolerance, and/or immune regulation to circumvent rejection and to permit immunosuppression minimization or withdrawal in transplant recipients.^11–15 Capitalizing on the work done in the oncologic arena, there has been growing interest in exploiting a variety of autologous, ex vivo-expanded/modified, and engineered immunomodulatory cell types to minimize treatment toxicity associated with conventional therapies and to prolong long-term allograft survival.¹⁶ Phase 1/2A trials are ongoing with the use of regulatory T cells (Tregs), tolerogenic dendritic cells, myeloid regulatory cells (regulatory macrophages), and mesenchymal stromal cells in combination with maintenance immunosuppression.^12,13,17

We previously identified a novel regulatory CD8⁺ T cell subset and reported that the adoptive cell therapy (ACT) with alloprimed, antibody-suppressor CXCR5⁺CD8⁺ T cells (CD8⁺ T_Ab-supp) significantly inhibits alloantibody production and enhances allograft survival in mouse models of hepatocellular and kidney transplant.^18,19 This novel regulatory CD8⁺ T cell subset is phenotypically and functionally distinct from previously described CXCR5⁺CD8⁺ T cell subsets,²⁰ and the suppression of in vivo alloantibody production by 14-fold in an allogeneic hepatocellular transplant model is associated with the capacity to kill alloprimed IgG⁺ B cells.¹⁸

In the context of vascularized solid organ transplant, we reported that CCR5 KO KTx recipients (on day 7 posttransplant) have 3.6-fold fewer splenic and 5.1-fold fewer peripheral blood alloprimed CXCR5⁺CD8⁺ T cells compared to C57BL/6 (WT) recipients. In addition, alloprimed CXCR5⁺CD8⁺ T cells from CCR5 KO recipients have a 30% lower capacity to kill antibody-producing IgG⁺ B cells in vitro compared to WT CXCR5⁺CD8⁺ T cells.¹⁹ ACT with CD8⁺ T_Ab-supp cells into high alloantibody producing CCR5 KO KTx recipients significantly reduces alloantibody production (4.6-fold reduction of day 14 posttransplant alloantibody titer) and ameliorates AMR pathology (2.4-fold reduction in day 14 composite AMR pathology score). Furthermore, ACT-mediated reduction of alloantibody titer and improved pathology correlates with enhanced allograft survival in murine KTx recipients (MST= day 52 vs day 15 without ACT).¹⁹ CD8⁺ T_Ab-supp cell effector functions are allospecific as third party-primed CXCR5⁺CD8⁺ T cell do not mediate these in vivo antibody-suppressor functions.^18,19 The translational relevance of these findings is supported by the identification of CXCR5⁺IFNγ⁺CD8⁺ T cells in the peripheral blood of human KTx recipients. Furthermore, we reported that KTx recipients with an increased quantity of circulating CXCR5⁺IFNγ⁺CD8⁺ T cells had lower incidence of de novo DSA production, rejection, and graft loss at 1-year posttransplant that were not attributable to differences in immunosuppression or clinical characteristics.²¹ Considered collectively, CD8⁺ T_Ab-supp cells present an exciting potential biomarker and cellular therapeutic target to suppress alloantibody production and treat AMR.

The current studies investigate the efficacy of ACT with CD8⁺ T_Ab-supp cells under cover of conventional maintenance immunosuppression in a well-established murine model of vascularized KTx. Kidney transplant in CCR5 KO mice is a robust model to study AMR due to the development of high alloantibody titer and associated pathology that recapitulates AMR in human KTx recipients.^19,22–26 Thus, we used the CCR5 KO KTx model to determine how conventional immunosuppressants such as calcineurin inhibitors (CNi) and mammalian target of rapamycin inhibitors (mTORi) impact the efficacy of ACT with CD8⁺ T_Ab-supp cells to regulate alloantibody production and ameliorate AMR in KTx recipients. In addition, we examine the effect of CNi and mTORi on the in vivo cytotoxic function, persistence, and proliferation of CD8⁺ T_Ab-supp cells after ACT.

MATERIALS AND METHODS

Experimental animals.

CCR5 KO, C57BL/6 (wild-type; WT), CD8 KO, Green Fluorescent Protein transgenic (GFP Tg) (H-2^b), as well as A/J (H-2^a) mouse strains (male and female at 8–20 weeks of age; Jackson Labs) were used in this study. All experiments were performed in compliance with the guidelines of the IACUCs of The Ohio State University (Protocol 2019A00000124) and Nationwide Children’s Hospital (Protocol AR17–00045).

Kidney isolation and transplantation.

Murine KTx with ureteral reconstruction was performed as previously described.^19,27

CXCR5⁺CD8⁺ T cell preparation and isolation.

Allogeneic (A/J) or third-party (FVB/N) lysate were prepared, as previously published.²⁸ C57BL/6 or GFP Tg mice were primed with 2 mg lysate by intraperitoneal injection. On day 7 after lysate stimulation, spleens were excised and primed CXCR5⁺CD8⁺ T cells were isolated by fluorescence-activated cell sorting (FACS).

B cell isolation.

B cells (B220) and primed IgG⁺ B cells were purified (both >95%) from splenocytes using anti-mouse B220 (StemCell) or anti-IgG magnetic beads (StemCell) following the manufacturer instructions.

Adoptive cell therapy.

Primed CXCR5⁺CD8⁺ T cells (2×10⁶) were resuspended in serum free RPMI and adoptively transferred into CCR5 KO KTx recipients via tail vein injection. KTx recipients received ACT with primed CXCR5⁺CD8⁺ T cells on day 1 after KTx.

Donor-reactive alloantibody titer.

Alloantibody titer from recipient sera were quantitated using published methods.²²

Immunosuppression administration.

KTx recipients received tacrolimus (1 or 5 mg/kg/day) or rapamycin (0.1 or 1 mg/kg/day) by intraperitoneal (i.p.) injection, as described.^29–32

Tacrolimus trough level analysis.

Sera from tacrolimus treated KTx recipients was collected approximately 24 hours following the previous dose. Tacrolimus was analyzed by competitive enzyme-linked immune-sorbent ELISA using the manufacture’s instructions (Abbexa LLC, Houston, TX). Sera from untreated KTx recipients was used as a negative control.

In vivo cell proliferation assay.

FACS-sorted GFP⁺CXCR5⁺CD8⁺ T cells were stained with CellTrace Violet (Thermo Fisher, Waltham, MA) and adoptively transferred into CCR5 KO KTx recipients on day 1 posttransplant. On day 14 after KTx, splenic GFP⁺Violet⁺CXCR5⁺CD8⁺ T cells were assessed for in vivo proliferation by flow cytometry (CellTrace Violet dilution).

In vivo cytotoxicity.

Detection of in vivo cytotoxic clearance of alloprimed IgG⁺ B cells was performed as previously described.^33,34

Flow cytometric analysis of cell subsets.

Splenocytes and mononuclear cells (from peripheral blood and the kidney allograft) were isolated from A/J kidney transplant recipients on days 0, 7, and 14 posttransplant. Single cell suspensions were incubated for 4 hours with Leukocyte Activation Cocktail (PMA, ionomycin, and Brefeldin A; Becton Dickinson), followed by extracellular and intracellular (FIX&PERM cell permeabilization kit; Thermo Fisher) staining with antibodies for flow cytometry: CD8 (clone 53–6.7), CD44 (clone IM7), CXCR5 (clone 2G8), IFNγ (clone XMG1.2), B220 (clone RA3–6B2), GL-7 (clone GL-7), Fas (clone Jo2), FasL (MFL3), Lamp1 (ID4B), Perforin (eBioMAK-D), Gzmb (NGZB), CD4 (RM4–5), IL-21 (mhalx21), and PD-1 (J43). Fluorescence-minus-one (FMO) was used as a negative control to set the positive/negative boundaries for protein expression.³⁵ Samples were acquired on a Becton Dickinson LSRFortessa cytometer and data were analyzed using FlowJo.

Kidney Histology & Immunofluorescence Staining.

In order to grade AMR pathology in kidney allografts, kidney allografts were retrieved on day 14 posttransplant. Excised kidneys were sectioned and processed, as previously described.¹⁹ Briefly, half of the kidney was formalin-fixed and paraffin-embedded for H&E staining. The other half section of kidney allografts was immediately snap frozen for immunofluorescent analysis of C4d deposition.³⁶

Pathological analysis.

Following H&E and immunofluorescent C4d staining, kidney allografts were evaluated for severity of AMR pathology and scored in accordance to internationally accepted Banff criteria, as previously described.¹⁹ Individual scores [for microvascular inflammation/peritubular capillary (PTC) margination, arteritis, and PTC C4d deposition were graded 0–3 for each category^37,38] were summed to create a composite AMR score.

Statistical analysis.

Student’s t-tests were used to test differences in continuous outcomes between two experimental groups. When more than two experimental groups were assessed, continuous outcomes measured at one time point were compared using general linear models including experimental groups as an independent variable. Continuous outcomes measured at multiple time points were compared between relevant groups using general linear models including experimental groups, time, and their interaction as independent variables. As the measurements were not conducted on the same mice over time, measurements were assumed to be independent, and no additional correlation was considered. Assumptions of normally distributed residuals were assessed graphically using q-q plots and were not considered to be violated for any of the analyses. All analyses were conducted using SAS version 9.4 (SAS Institute, Inc., Cary, NC). Results are summarized as estimated mean ± standard error. Hypothesis testing was conducted at a 5% type I error rate (alpha=0.05) and p<0.05 was considered statistically significant.

RESULTS

CXCR5⁺CD8⁺ T cell-mediated suppression of alloantibody and AMR is impaired by CNi and enhanced with mTORi.

To examine the effect of conventional immunosuppressants on CXCR5⁺CD8⁺ T cell ACT, CCR5 KO KTx recipients underwent ACT while receiving concurrent CNi (tacrolimus; 1 mg/kg/day) or mTORi (rapamycin; 1 mg/kg/day). Alloantibody titer was determined on day 14 posttransplant. Whereas posttransplant alloantibody titer in recipients treated with ACT alone was reduced 4.5-fold compared to untreated controls (p<0.0001), ACT and CNi (1 mg/kg/day) combination therapy decreased posttransplant alloantibody by only 2-fold (p<0.0001 vs untreated controls) thus impairing the efficacy of ACT alone (p<0.0001; Figure 1A). Treatment of CCR5 KO KTx recipients with CNi alone did not reduce alloantibody titer after KTx when compared to untreated controls (p=NS). We considered the possibility that CNi at the 1 mg/kg/day dose was insufficient to mediate immunosuppressive effects in CCR5 KO hosts. However, the measured trough levels in our studies following CNi administration at 1 mg/kg/day (3.3±0.3 ng/mL) confirmed CNi administration in doses previously reported to mediate immunosuppressive effects in vivo following skin transplantation.³⁹

Figure 1. CNi impairs alloprimed CXCR5⁺CD8⁺ T cell–mediated suppression of alloantibody and AMR.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. Some CCR5 KO recipient groups underwent ACT with 2 × 10⁶ alloprimed CXCR5⁺CD8⁺ T cells via tail vein injection on day 1 after KTx. Cohorts of recipients were untreated or received ACT, tacrolimus (a CNi, low dose, 1 mg/kg/day or high dose, 5 mg/kg/day i.p.), or combination ACT + CNi on day 1–14 after KTx. A, CCR5 KO KTx recipient cohorts that received ACT produced significantly lower alloantibody titer on day 14 posttransplant (1306 ± 106, n=9; *p<0.0001) compared to untreated controls (5951 ± 630, n=7). Cohorts that received ACT and low-dose CNi also had lower alloantibody titer (3241 ± 408, n=8) than untreated CCR5 KO recipients (**p<0.0001) but significantly higher alloantibody titer than ACT alone (***p<0.0001). Treatment with low-dose CNi alone did not inhibit alloantibody titer (5867 ± 487, n=6; p=NS) and was significantly higher than ACT alone (*p<0.0001). B, Allokidneys were retrieved on day 14 for histopathologic analysis. Samples were scored according to Banff criteria, consisting of individual histologic scores for PTC margination, arteritis, and PTC C4d deposition and a histologic composite score. Composite AMR pathology was significantly ameliorated by ACT (3.8 ± 0.4, n=9; *p<0.0001) compared to untreated recipients (7.6 ± 0.5, n=7). Cohorts that received ACT and CNi had lower composite AMR pathology (5.7 ± 0.5, n=8; **p=0.003) compared to untreated CCR5 KO recipients but higher composite AMR pathology than ACT alone (***p=0.001). Treatment with CNi alone did not ameliorate AMR pathology (6.5 ± 0.7, n=6; p=NS) and composite AMR score in recipients treated with CNi alone was significantly higher than recipients treated with ACT alone (*p=0.0004). C, Cohorts that received ACT and high dose CNi (5 mg/kg/day) had high alloantibody titer (6079 ± 49, n=4) similar to untreated CCR5 KO recipient controls (p=NS) and significantly higher than in recipients treated with ACT alone (*p<0.0001). Treatment with high-dose CNi alone (6975 ± 156, n=4) increased alloantibody titer when compared to untreated recipients (**p=0.05) and was significantly higher than in recipients treated with ACT alone (***p<0.0001). D, Cohorts that received ACT and high-dose CNi had lower composite AMR pathology (4.8 ± 0.3, n=4; *p=0.0006) compared to untreated CCR5 KO recipients. AMR pathology was similar in combined ACT + high dose CNi treated recipients when compared to recipients treated with ACT alone or treated with high dose CNi alone (p=NS). Treatment with high dose CNi alone improved AMR pathology compared to untreated controls (5.0 ± 0.4, n=4, **p=0.002). Error bars designate standard error of combined data from independent experiments. ACT, adoptive cell therapy; AMR, antibody-mediated rejection; CNi, calcineurin inhibitor; i.p., intraperitoneal; KO, knockout; KTx, kidney transplant; MVI, microvascular inflammation; NS, not significant; PTC, peritubular capillary.

Similarly, while ACT alone significantly improved KTx pathology as reflected by a lower composite AMR score compared to untreated controls (p<0.0001), this beneficial effect of ACT was impaired in recipients that received ACT in combination with CNi (p=0.004 vs ACT alone; Figure 1B). However, the composite AMR score in recipients treated with ACT and CNi was lower than in untreated controls (p=0.003).

We also investigated if CNi administered at higher dosage (5 mg/kg/day) suppresses humoral alloimmunity when used alone or in combination with ACT in CCR5 KO KTx recipient mice. Treatment with high dose CNi alone (5 mg/kg/day) increased alloantibody titer (p=0.05 vs untreated controls; Figure 1C) and high dose CNi in combination with ACT, abolished ACT-mediated reduction of alloantibody titer (p<0.0001; Figure 1C). While treatment of CCR5 KO KTx recipients with CNi alone did not reduce alloantibody titer, treatment with CNi alone at high dose did reduce the severity of AMR (p=0.002; Figure 1D). Thus, CNi immunosuppression at both high and low doses failed to suppress alloantibody production after KTx in CCR5 KO mice and impaired the efficacy of ACT for suppression of posttransplant alloantibody and amelioration of AMR pathology.

Next, we examined posttransplant alloantibody titer and AMR pathology when ACT was administered in combination with mTORi. Treatment with mTORi alone (1 mg/kg/day) significantly reduced alloantibody titer compared to untreated controls (p<0.0001; Figure 2A). Combination therapy of ACT and mTORi (1 mg/kg/day) was associated with the greatest reduction in alloantibody titer (p<0.0001 vs untreated controls; p=0.01 compared to ACT alone; Figure 2A). The addition of mTORi (1 mg/kg/day) to ACT also improved the composite AMR score compared to ACT alone (p=0.006; Figure 2B).

Figure 2. mTORi enhances alloprimed CXCR5⁺CD8⁺ T cell–mediated suppression of alloantibody and AMR.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. Some CCR5 KO groups underwent ACT with 2 ×10⁶ alloprimed CXCR5⁺CD8⁺ T cells via tail vein injection on day 1 after KTx. Cohorts of recipients were untreated or received ACT, rapamycin (mTORi) (1mg/kg/day or 0.1 mg/kg/day i.p.), or combination therapy with ACT and mTORi day 1–14 after KTx. A, CCR5 KO KTx recipient cohorts that received ACT day 1 after KTx produced significantly lower alloantibody titer on day 14 posttransplant (1306 ± 106, n=9; *p<0.0001) compared to untreated controls (5951 ± 630, n=7). Cohorts that received ACT and high dose mTORi (1 mg/kg/day) had lower alloantibody titer (319 ± 44, n=8) than untreated CCR5 KO recipients (**p<0.0001) and recipients that received ACT alone (**p=0.01). Treatment with high-dose mTORi alone significantly inhibited alloantibody titer after KTx (412 ± 71, n=6; ***p<0.0001 compared to untreated controls). B, Allokidneys were retrieved on day 14 for histopathologic analysis. Samples were scored according to Banff criteria, consisting of individual histologic scores for PTC margination, arteritis, and PTC C4d deposition and a histologic composite score. Composite AMR pathology was significantly ameliorated by ACT (3.8 ± 0.4, n=9; *p<0.0001) compared to untreated recipients (7.6 ± 0.5, n=7). Cohorts that received ACT and high-dose mTORi had lower composite AMR pathology (1.7 ± 0.5, n=8) compared to untreated controls (**p<0.0001) and recipients that received ACT alone (**p=0.0006). Treatment with high-dose mTORi alone significantly ameliorated AMR pathology in KTx recipients (1.0±0.7, n=6; ***p<0.0001). C, Cohorts that received ACT and low-dose mTORi (0.1 mg/kg/day) had lower alloantibody titer (156 ± 15, n=4) when compared to untreated CCR5 KO recipients (*p<0.0001) or recipients that received ACT alone (*p=0.02) and similar alloantibody titer in comparison to recipients that received mTORi alone. Treatment with low-dose mTORi alone (0.1 mg/kg/day) significantly inhibited alloantibody titer after KTx compared to untreated controls (277 ± 35, n=5; **p<0.0001). D, Cohorts that received ACT and low-dose mTORi had lower composite AMR score (2.8 ± 0.3, n=4) compared to untreated controls (*p<0.0001) and compared to recipients treated with low-dose mTORi alone (**p=0.08). Treatment with low-dose mTORi alone significantly improved AMR score in KTx recipients compared to untreated controls (4.2 ± 0.2, n=5 ***p<0.0001). Error bars designate standard error of combined data from independent experiments. ACT, adoptive cell therapy; AMR, antibody-mediated rejection; i.p., intraperitoneal; KO, knockout; KTx, kidney transplant; mTORi, mammalian target of rapamycin inhibitor; MVI, microvascular inflammation; PTC, peritubular capillary.

In order to determine if the beneficial effect of mTORi upon ACT efficacy was dose-dependent, we tested a lower dose (0.1 mg/kg/day) of mTORi. Treatment with low dose mTORi alone significantly reduced alloantibody titer in CCR5 KO KTx recipients compared to untreated controls (p<0.0001; Figure 2C). Treatment with low dose mTORi in combination with ACT further reduced alloantibody titer (p<0.0001 vs untreated controls; p=0.02 compared to ACT alone; Figure 2C). The composite AMR score in KTx recipients treated with low dose mTORi in combination with ACT was lower than in recipients treated with mTORi alone (p=0.08; Figure 2D). It is notable that treatment with mTORi alone significantly improved KTx pathology compared to untreated controls (p<0.0001 for both doses) and the improvement was dose-dependent (0.1 mg/kg/day vs 1.0 mg/kg/day, p=0.0003; Figures 2B and D). Furthermore, the KTx AMR composite score was much lower in recipients treated with ACT and mTORi than in those treated with ACT and CNi (p<0.02 for all dose comparisons; Figures 1B and 2B).

CNi impairs ACT-mediated in vivo cytotoxicity targeting alloprimed, antibody-producing IgG⁺ B cells in KTx recipients.

We previously reported that the efficacy of ACT to suppress humoral alloimmunity in transplant recipients is associated with alloprimed CXCR5⁺CD8⁺ T cell-mediated cytotoxic effector function targeting alloprimed B cells in vitro and in vivo.^18,33 In the current studies, we tested the effect of CNi or mTORi treatment upon in vivo allocytotoxicity directed against target alloprimed B cells in KTx recipients receiving ACT. On day 7 following KTx, cohorts of CCR5 KO recipients were tested for in vivo cytotoxicity towards alloprimed, antibody-producing IgG⁺ B cells. We found that ACT-treated KTx recipients display marked in vivo IgG⁺ B cell-directed allocytotoxicity compared to untreated CCR5 KO KTx recipients (p=0.04; Figure 3). Furthermore, we observed that in vivo IgG⁺ B cell-directed cytotoxicity in ACT-treated KTx recipients is significantly reduced when ACT is combined with CNi treatment (p<0.0001). In contrast, the addition of mTORi immunosuppression did not impact alloprimed IgG⁺ B cell-directed cytotoxicity in ACT-treated KTx recipients (p=NS). CNi treatment (monotherapy) of KTx recipients did not impact in vivo cytotoxicity to alloprimed IgG⁺ B cell targets (p=NS). KTx recipients treated with mTORi monotherapy displayed a moderate increase in in vivo cytotoxicity towards alloprimed IgG⁺ B cell targets (p=0.013 vs untreated) but cytotoxicity was significantly less than observed in recipients that received combined ACT and mTORi (p=0.02).

Figure 3. CNi inhibits alloprimed CXCR5⁺CD8⁺ T cell-mediated in vivo cytotoxicity of alloprimed, antibody-producing IgG⁺ B cells.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. On day 1 after KTx, some CCR5 KO recipients received ACT with 2 × 10⁶ alloprimed CXCR5⁺CD8⁺ T cells via tail vein injection and some received concurrent immunosuppression with tacrolimus (CNi) or rapamycin (mTORi) dosed at 1mg/kg/d, i.p., on days 1–7. Cohorts of recipients were untreated or received ACT, ACT + CNi, ACT + mTORi or were treated with CNi or mTORi alone. On day 7, Carboxyfluorescein succinimidyl ester-high (CFSE^hi)-labeled alloprimed IgG⁺ B cells were adoptively transferred (along with control CFSE^lo naive B220⁺ B cells) into transplant recipients to test for in vivo cytotoxic clearance of IgG⁺ B cells. Naïve hosts served as negative controls. A, Representative flow plots show gating on lymphocytes, single cells, and analysis of CFSE^lo naive B cells and CFSE^hi alloprimed IgG⁺ B cells that were used to determine in vivo cytotoxicity. B, In vivo cell-mediated cytotoxicity (CMC) to IgG⁺ B cells was significantly increased in CCR5 KO recipients following CXCR5⁺CD8⁺ T cell ACT (40.2 ± 3.5%, n=6) compared to untreated CCR5 KO recipients (10.2 ± 3.6%, n=6) or to recipients treated with CNi (12.1 ± 2.8%, n=5) or mTORi alone (22.0 ± 4.2% n=5; *p<0.0003 for all). In vivo cytotoxicity to IgG⁺ B cells in CCR5 KO recipients that received ACT was significantly reduced by concurrent treatment with CNi (17.8 ± 1.3%, n=5; **p<0.0001) but was not reduced by concurrent treatment with mTORi (33.5 ± 2.0%, n=5; p=NS). In vivo cytotoxicity to IgG⁺ B cells was significantly less in CCR5 KO recipients that received ACT and concurrent treatment with CNi compared to concurrent treatment with mTORi (**p=0.002). Treatment of CCR5 KO KTx recipients with CNi alone had no impact on in vivo cytotoxicity to IgG⁺ B cells (12.1 ± 2.8%, p=NS vs untreated). Treatment of CCR5 KO KTx recipients with mTORi alone increased in vivo cytotoxicity to IgG⁺ B cells compared to untreated controls and CNi treated recipients (22.0 ± 4.2%; ***p<0.04). CCR5 KO KTx recipients that received ACT and mTORi had higher in vivo cytotoxicity to IgG⁺ B cells than recipients treated with mTORi alone (^†p=0.02). ACT, adoptive cell therapy; CFSE, carboxyfluorescein succinimidyl ester; CMC, cell-mediated cytotoxicity; CNi, calcineurin inhibitor; FSC-A, forward scatter area; FSC-W, forward scatter width; IgG, immunoglobulin G; i.p., intraperitoneal; KO, knockout; mTORi, mammalian target of rapamycin inhibitor; NS, not significant; SSC, side scatter.

To evaluate ACT allospecificity, we assessed the in vivo cytotoxicity of alloprimed CXCR5⁺CD8⁺ T cell therapy targeting 3^rd party-primed IgG⁺ B cell targets as well as 3^rd party-primed CXCR5⁺CD8⁺ T cell therapy targeting alloprimed IgG⁺ B cell targets in CCR5 KO recipients (Figure S1). In both of these experimental groups to test ACT allospecificity, no cytotoxicity was observed above baseline in vivo cytotoxic clearance of alloprimed IgG⁺ B cells or of 3^rd party primed IgG⁺ B cells in CCR5 KO KTx recipients (p=NS for both). These results complement our prior studies reporting the antigen-specificity of alloprimed CXCR5⁺CD8⁺ T cell effector function.^18,19

We next examined the expression of the cytotoxic effector molecules previously reported to be important for antibody-suppressor CXCR5⁺CD8⁺ T function³³. To do this, CXCR5⁺CD8⁺ T cells were retrieved from alloprimed GFP mice for adoptive transfer into CCR5 KO KTx recipients. On day 7 after KTx, splenic GFP⁺CXCR5⁺CD8⁺ T cells were retrieved from KTx recipients that received ACT alone or ACT in combination with CNi or mTORi and evaluated for expression of cytotoxic effector molecules by flow cytometry. On day 7 posttransplant, adoptively transferred GFP⁺CXCR5⁺CD8⁺ T cells express FasL, Lamp1, perforin, and granzyme B (Figure 4A and B), similar to baseline expression at the time of ACT (data not shown). The expression of these cytotoxic effector molecules was similar in KTx recipients that received ACT alone or ACT in combination with CNi or mTORi (p=NS for all). Thus, CNi-associated impairment of alloprimed CXCR5⁺CD8⁺ T cells’ in vivo cytotoxicity targeting IgG⁺ B cells is not due to interference with CD8⁺ T cell cytotoxic effector molecule expression.

Figure 4. CNi and mTORi do no inhibit the expression of cytotoxic molecules expressed by alloprimed CXCR5⁺CD8⁺ T cells.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. On day 1 after KTx, some CCR5 KO recipients received ACT of 2 × 10⁶ alloprimed GFP⁺CXCR5⁺CD8⁺ T cells via tail vein injection. Cohorts of recipients received ACT alone (no immunosuppression [IS]; n=4) or ACT with concurrent tacrolimus (CNi, n=4) or rapamycin (mTORi, n=4) dosed at 1mg/kg/day, i.p. on days 1–7. On day 7, GFP⁺CXCR5⁺CD8⁺ T cells were retrieved from recipient spleen and evaluated for expression of cytotoxic effector molecules by flow cytometry. A, Representative flow cytometry show gating on lymphocytes, single cells, GFP⁺CXCR5⁺CD8⁺ T cells. B, A significant proportion of adoptively transferred alloprimed GFP⁺CXCR5⁺CD8⁺ T cells express cytotoxic effector molecules including FasL (77.7 ± 2.5%), Lamp1 (88.2 ± 2.4%), perforin (68.6 ± 3.8%), and granzyme B (62.1 ± 3.6%). The expression of these cytotoxic molecules by transferred GFP⁺CXCR5⁺CD8⁺ T cells is similar in recipients that received ACT alone or ACT in combination with CNi or mTORi immunosuppression (p=NS for all groups). ACT, adoptive cell therapy; CNi, calcineurin inhibitor; FasL, Fas ligand; FMO, fluorescence-minus-one; FSC-A, forward scatter area; FSC-W, forward scatter width; GFP, Green Fluorescent Protein; i.p., intraperitoneal; IS, immunosuppression; KO, knockout; KTx, kidney transplant; Lamp1, Lysosomal-associated membrane protein 1; mTORi, mammalian target of rapamycin inhibitor; NS, not significant; SSC, side scatter.

Reduction of splenic germinal center (GC) B cells in ACT-treated KTx recipients is partially reversed by CNi but not mTORi treatment.

We have previously reported that the suppression of humoral alloimmunity in ACT-treated KTx recipients is associated with reduced quantity of GC B cells in KTx recipients.^18,19 In view of the observed impaired in vivo cytotoxic effector function to alloprimed B cells, we next investigated the effect of CNi and mTORi on quantity of splenic GC B cells on day 7 after KTx. ACT with alloprimed CXCR5⁺CD8⁺ T cells significantly reduces the quantity of splenic GC B cells after KTx compared to untreated CCR5 KO KTx recipients (p<0.0001; Figure 5A and B). The combination of ACT with mTORi (1 mg/kg/day) results in a further decrease in quantity of splenic GC B cells in KTx recipients (p<0.0001 vs untreated controls; p=0.07 vs ACT alone; p=0.08 vs mTORi alone; Figure 5A and B). Similarly, the combination of ACT with low dose mTORi (0.1 mg/kg/day) also decreased the quantity of splenic GC B cells in KTx recipients when compared to untreated recipients (p<0.0001), recipients that received ACT alone (p=0.008) or recipients that received mTORi alone (p=0.07; Figure S2A and B).

Figure 5. Reduction in quantity of germinal center B cells in recipients that received ACT with alloprimed CXCR5⁺CD8⁺ T cells is partially reversed by CNi but not mTORi.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. On day 1 after KTx, some CCR5 KO recipients received ACT with 2 × 10⁶ alloprimed CXCR5⁺CD8⁺ T cells via tail vein injection and some received concurrent immunosuppression with tacrolimus (CNi) or rapamycin (mTORi) dosed at 1mg/kg/d, i.p., on days 1–7. Cohorts of recipients were untreated or received ACT, ACT + CNi, ACT + mTORi or were treated with CNi or mTORi alone. On day 7, recipient splenocytes were retrieved and analyzed by flow cytometry for the proportion of GC B cells which was used to calculate the quantity of GC B cells per mm spleen tissue. A, Representative flow plots show gating on lymphocytes, single cells, B cells (B220⁺) and GC B cells (Fas⁺GL-7⁺). B, ACT significantly decreases the quantity of GC B cells detected on day 7 after KTx (3931 ± 244 cells/mm³; n=11, *p<0.0001) compared to untreated controls (8103 ± 482 cells/mm³; n=5). Combination ACT and mTORi further reduces GC B cell quantity (2838 ± 232 cells/mm³; n=6) compared to untreated controls (**p<0.0001), compared to ACT alone (**p=0.07), and compared to treatment with mTORi alone (^‡p=0.08). In contrast, concurrent treatment with CNi partially reverses ACT-mediated reduction in quantity of GC B cells (5811 ± 647 cells/mm³; n=5; ***p=0.005) though the quantity of GC B cells is still reduced in comparison to untreated recipients (^†p=0.003) or CNi alone (^†p=0.01). When compared to untreated recipients, treatment with mTORi alone significantly reduced the quantity of splenic GC B cell detected in CCR5 KO KTx recipients (4176 ± 587 cells/mm³; n=4, ^‡p<0.0001), while treatment with CNi did not (7788 ± 639 cells/mm³; n=5, p=NS). Treatment with mTORi alone was associated with a significant reduction in GC B cells when compared to recipients treated with CNi alone (^‡p<0.0001) or treated with ACT and CNi (^‡p=0.03). Treatment with CNi alone was associated with significantly greater quantity of GC B cells compared to recipients treated with ACT alone (^§p<0.0001). The reduction in GC B cells in recipients treated with mTORi alone was similar to recipients treated with ACT alone (p=NS). ACT, adoptive cell therapy; CNi, calcineurin inhibitor; FMO, fluorescence-minus-one; FSC-A, forward scatter area; FSC-W, forward scatter area width; GC, germinal center; i.p., intraperitoneal; KO, knockout; KTx, kidney transplant; mTORi, mammalian target of rapamycin inhibitor; SSC, side scatter.

In contrast, the addition of low dose CNi (1 mg/kg/day) to ACT is associated with an increase in splenic GC B cells when compared to KTx recipients receiving ACT alone (p<0.0001) but the quantity was still less than in untreated KTx recipients (p=0.003; Figure 5A and B). High dose CNi (5 mg/kg/day) abrogated ACT-mediated suppression of GC B cell quantity (p=NS compared to untreated recipients; p<0.0001 compared to ACT alone; Figure S2A and B).

KTx recipients that received mTORi monotherapy displayed a reduction in GC B cells (p<0.001 for both doses vs untreated controls), comparable to that observed in recipients that received ACT alone (p=NS). CNi monotherapy at low dose did not inhibit the quantity of splenic GC B cells in CCR5 KO KTx recipients (p=NS vs untreated controls; Figure 5A and B); and with high dose CNi, the quantity of splenic GC B cells was increased above the quantity detected in untreated KTx recipients (p=0.02; Figure S2A and B). The latter corresponds with increased alloantibody titer in high dose CNi treated KTx recipients (p=0.05; Figure 1C compared to untreated controls). These data demonstrate that the suppression of humoral alloimmunity by ACT is complemented by the inhibitory effects of mTORi on in vivo GC B cell development and/or survival. In contrast, CNi immunosuppression partially reverses ACT-mediated reduction of GC B cells in KTx recipients.

To determine how the quantity of CD4⁺ T follicular helper cells (CD4⁺ T_FH) correlated with the observed reduction in GC B cells, we analyzed day 7 splenocytes for CXCR5⁺PD1⁺CD4⁺ T cells and IL-21⁺CXCR5⁺PD1⁺CD4⁺ T cells. The quantities of CD4⁺ T_FH cells and IL-21⁺CD4⁺ T_FH cells were significantly decreased in KTx recipients that received ACT, ACT + CNi (1mg/kg/day), ACT + mTORi (1 mg/kg/day), CNi alone (1 mg/kg/day), or mTORi alone (1 mg/kg/day), compared to untreated control recipients (p<0.0001 for all groups vs. untreated; Figure 6A–C). Similar results were observed when CNi was tested at high dose (5 mg/kg/day) and mTORi was tested at low dose (0.1 mg/kg/day, p<0.0001; Figure S3A and B). There was no difference in quantity of total splenic CD4⁺ T cells among treated recipients compared to untreated controls (p=NS; Figure 6D). Thus, the decrease in CD4⁺ T_FH cells and IL-21⁺CD4⁺ T_FH cells in the treatment groups was not due to reduction in overall CD4⁺ T cells. Despite the similar inhibitory effects of ACT, CNi and mTORi upon the quantity of CD4⁺ T_FH cells and IL-21⁺CD4⁺ T_FH cells that develop after KTx in CCR5 KO mice, these same treatment groups had marked differences in the in vivo allocytotoxicity to IgG⁺ B cells targets, the quantity of splenic GC B cells, alloantibody production, and AMR pathology.

Figure 6. The in vivo development of CD4⁺ T_FH cells after KTx in CCR5 KO mice is inhibited by ACT with antibody-suppressor CXCR5⁺CD8⁺ T cells, ACT and CNi, ACT and mTORi, CNi alone, and mTORi alone.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. On day 1 after KTx, some CCR5 KO recipients received ACT with 2 × 10⁶ alloprimed CXCR5⁺CD8⁺ T cells via tail vein injection and some received concurrent immunosuppression with tacrolimus (CNi) or rapamycin (mTORi) dosed at 1mg/kg/d, i.p., on days 1–7. Cohorts of recipients were untreated or received ACT, ACT + CNi, ACT + mTORi or were treated with CNi or mTORi alone. On day 7, spleens were collected and CD4⁺ T cells were analyzed by flow cytometry. A, Representative flow plots show gating on lymphocytes, single cells, and CXCR5⁺PD1⁺CD4⁺ T cells. B and C, The quantities of CD4⁺ T_FH cells and IL-21⁺CD4⁺ T_FH cells were significantly decreased in KTx recipients that received ACT, ACT + CNi, ACT + mTORi, CNi alone or mTORi alone compared to untreated control recipients (*p<0.0001 for all; N= 4 to 9 mice per treatment group). The quantities of CD4⁺ T_FH cells and IL-21⁺CD4⁺ T_FH cells were unchanged in recipients that received ACT combined with CNi or mTORi, compared to ACT alone (p=NS). D, No differences were detected in quantities of total splenic CD4⁺ T cells among treated recipients compared to untreated controls (p=NS). ACT, adoptive cell therapy; CNi, calcineurin inhibitor; FMO, fluorescence-minus-one; FSC-A, forward scatter area; FSC-W, forward scatter width; IL-21, interleukin 21; i.p., intraperitoneal; IS, immunosuppression; KO, knockout; KTx, kidney transplant; mTORi, mammalian target of rapamycin inhibitor; NS, not significant; PD-1, programmed cell death protein 1; SSC, side scatter; T_FH, follicular helper CD4⁺ T cell.

CNi, but not mTORi, reduces the quantity and impairs proliferation of CXCR5⁺CD8⁺ T cells after ACT.

We next examined how CNi and mTORi impact the persistence and proliferation of adoptively transferred alloprimed CXCR5⁺CD8⁺ T cells. We first evaluated the quantity of alloprimed CD44⁺CXCR5⁺CD8⁺ T cells detected in the spleen on day 14 after KTx in recipients that received ACT. ACT was associated with a significant increase in overall quantity of CD44⁺CXCR5⁺CD8⁺ T cells detected in the spleen on day 14 after KTx (p=0.0007; Figure 7A and B). CCR5 KO KTx recipients that received combined ACT and CNi treatment had a 4.5 (1 mg/kg/day) to 6-fold (5 mg/kg/day) decrease in quantity of day 14 splenic CD44⁺CXCR5⁺CD8⁺ T cells compared to those receiving ACT alone (p<0.0001 for both CNi doses; Figure 7A and B and Figure S4A and B).

Figure 7. Concurrent treatment with CNi or mTORi reduces the quantity of CD44⁺CXCR5⁺CD8⁺ T cells detected in the spleen after ACT.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. On day 1 after KTx, CCR5 KO cohorts received ACT of 2 × 10⁶ alloprimed GFP⁺CXCR5⁺CD8⁺ T cells via tail vein injection, or ACT and concurrent treatment with either tacrolimus (CNi) or rapamycin (mTORi) dosed at 1mg/kg/d, i.p. on days 1–14. Recipient splenocytes were retrieved on day 14 after KTx and analyzed by flow cytometry for overall quantity of activated CD44⁺CXCR5⁺CD8⁺ T cells. A, Flow cytometric gating on lymphocytes, single cells, and CD44⁺CXCR5⁺CD8⁺ T cells. B, ACT significantly increases the quantity of splenic CD44⁺CXCR5⁺CD8⁺ T cells detected in the spleen on day 14 posttransplant (905 ± 148 cells/mm³; n=10, *p=0.0007) compared to untreated controls (354 ± 58 cells/mm³; n=5). However, this increase is negated when ACT is combined with concurrent CNi (201 ± 55 cells/mm³; n=4, **p<0.0001 compared to ACT alone) or with concurrent mTORi (521 ± 88 cells/mm³; n=5, ***p=0.01 compared to ACT alone). Treatment with CNi alone also reduced the quantity of endogenous CD44⁺CXCR5⁺CD8⁺ T cells (131 ± 55 cells/mm³; n=4, ^†p=0.07) while treatment with mTORi alone (377 ± 70 cells/mm³; n=4, p=NS) did not significantly impact the quantity of endogenous CD44⁺CXCR5⁺CD8⁺ T cells detected in the spleen on day 14 posttransplant compared to untreated controls. Error bars designate standard error of combined data from independent experiments. ACT, adoptive cell therapy; CNi, calcineurin inhibitor; FMO, fluorescence-minus-one; FSC-A, forward scatter area; FSC-W, forward scatter width; i.p., intraperitoneal; KO, knockout; KTx, kidney transplant; mTORi, mammalian target of rapamycin inhibitor; NS, not significant; SSC, side scatter.

CCR5 KO KTx recipients that received combined ACT and mTORi (1mg/kg/day) treatment had a 1.7-fold decrease in quantity of day 14 splenic CD44⁺CXCR5⁺CD8⁺ T cells compared to ACT alone (p=0.01; Figure 7A and B). However, combination therapy with low dose mTORi (0.1mg/kg/day) preserved the increase in quantity of day 14 splenic CD44⁺CXCR5⁺CD8⁺ T cells detected after ACT (p=NS; Figure S4A and B).

Treatment of KTx recipients with mTORi monotherapy at high or low doses did not impact the quantity of endogenous splenic CD44⁺CXCR5⁺CD8⁺ T cells that developed in untreated CCR5 KO KTx recipients (Figure 7A and B and Figure S4A and B). In contrast, CNi monotherapy reduced the quantity of endogenous splenic CD44⁺CXCR5⁺CD8⁺ T cells (p=0.07 for both doses; Figure 7A and B and Figure S4A and B). This may account for the higher quantity of GC B cells (Figure S2) and alloantibody titer (Figure 1C) in recipients treated with high dose CNi alone. Altogether, these data demonstrate that CNi (but not mTORi) reduces the quantity of endogenous splenic CD44⁺CXCR5⁺CD8⁺ T cells and impairs the persistence, survival or expansion of adoptively transferred alloprimed CD44⁺CXCR5⁺CD8⁺ T cells.

We next investigated the impact of CNi or mTORi treatment upon in vivo proliferation of adoptively transferred alloprimed GFP⁺CXCR5⁺CD8⁺ T cells labeled with CellTrace Violet. We found that treatment with CNi reduced in vivo proliferation of adoptively transferred CXCR5⁺CD8⁺ T cells compared to KTx recipients that received ACT alone (p=0.0003; Figure 8A and B). In contrast, we did not detect a significant impact of treatment with mTORi upon in vivo proliferation of adoptively transferred alloprimed CXCR5⁺CD8⁺ T cells compared to KTx recipients that received ACT alone (p=NS; Figure 8A and B).

Figure 8. In vivo proliferation of adoptively transferred alloprimed CXCR5⁺CD8⁺ T cells is significantly reduced in recipients treated with CNi but not mTORi.

CCR5 KO (H-2^b) mice were transplanted with allogeneic A/J (H-2^a) kidneys. Alloprimed CXCR5⁺CD8⁺ T cells were retrieved from alloprimed GFP mice. On day 1 after KTx, CCR5 KO cohorts received ACT of 2 × 10⁶ CellTrace Violet-stained alloprimed GFP⁺CXCR5⁺CD8⁺ T cells via tail vein injection, or ACT and concurrent treatment with either tacrolimus (CNi) or rapamycin (mTORi) dosed at 1mg/kg/d, i.p. on days 1–14. Recipient splenocytes were retrieved on day 14 after KTx and analyzed by flow cytometry for dilution of CellTrace Violet stain on GFP⁺CXCR5⁺CD8⁺ T cells. A, Flow cytometric gating on lymphocytes, single cells, and GFP⁺CD8⁺ T cells. Representative flow panels of splenic GFP⁺CD8⁺ T cells from CCR5 KO KTx recipients are shown. For reference, CellTrace Violet-stained GFP⁺CD8⁺ T cells prior to adoptive transfer (Pre ACT) are shown. B, In vivo proliferation of adoptively transferred GFP⁺Violet⁺CXCR5⁺CD8⁺ T cells was assessed by determining the proportion of GFP⁺CD8⁺ cells demonstrating dilution of CellTrace Violet. In vivo proliferation of adoptively transferred cells was significantly reduced in recipients that received ACT and concurrent treatment with CNi (26.1 ± 5.1%; n=4) when compared to recipients that received ACT alone (65.4 ± 6.3%, n=5; *p=0.0003) or when compared to ACT with concurrent treatment with mTORi (62.9 ± 2.0%; n=4, **p=0.0007). In vivo proliferation of adoptively transferred cells was similar in recipients treated with ACT and concurrent treatment with mTORi compared to recipients that received ACT alone (p=NS). Error bars designate standard error of combined data from independent experiments. ACT, adoptive cell therapy; CNi, calcineurin inhibitor; GFP, Green Fluorescent Protein; FSC-A, forward scatter area; FSC-W, forward scatter width; i.p., intraperitoneal; KO, knockout; KTx, kidney transplant; mTORi, mammalian target of rapamycin inhibitor; NS, not significant; SSC, side scatter.

DISCUSSION

Antibody-mediated rejection in human kidney transplant recipients is often resistant to treatment with standard immunosuppressive agents⁴⁰ and strategies to reduce the quantity of donor-specific antibody (e.g. intravenous immunoglobulin and plasmapheresis) are based on low quality of evidence⁹ and dramatically lose efficacy once AMR is established.⁴⁰ In addition, immunotherapeutic strategies targeting costimulatory blockade (CD28/CD80/CD86), B cells (CD20), plasma cells (CD38) or cytokines (IL-6) that are reported to initially decrease DSA are also associated with unintended consequences such as rebound increased quantity of DSA and/or development of acute TCMR.¹⁰ Thus, there is a vested interest in developing novel therapeutic approaches to treat AMR. An increasing body of work supports targeted cellular-based treatment to be viable immunotherapeutic strategies after transplantation. There are currently several regulatory cell subsets (T_reg, tolerogenic dendritic cells, and regulatory macrophages) being examined in early-stage clinical trials.^12,13,17 While regulatory T cells have historically been at the forefront of potential for translation to clinical application,^11–15,41 more recent studies have featured bone marrow-derived mesenchymal stem cells,^42,43 allogeneic stem cell therapy (FREEDOM-1 study, ongoing), and CAR T_regs.⁴⁴ The current studies expand upon our prior work investigating the translational potential of a novel regulatory CD8⁺ T_Ab-supp cell subset by determining the efficacy of ACT to suppress humoral alloimmunity and AMR when used in combination with conventional immunosuppressive agents.

Our group reported that ACT with a single dose of CD8⁺ T_Ab-supp cells into high alloantibody producing CCR5 KO KTx recipients significantly interrupts alloantibody production and reduces alloantibody titer, attenuates AMR pathology, and improves graft survival in the absence of maintenance immunosuppression.¹⁹ CD8⁺ T_Ab-supp cells mediate suppressor function through the killing of alloprimed IgG1⁺ splenic germinal center B cells,⁴⁵ mediated in part by the classic cytotoxic effector molecules FasL and perforin.³³ As CCR5 KO recipients produce detectable alloantibody as early as posttransplant day 5,⁴⁶ these data suggest that CD8⁺ T_Ab-supp cells arrest the AMR phenotype by interrupting active alloantibody production in a dose-dependent and allospecific manner. In addition to our studies in mice demonstrating powerful downregulation of humoral alloimmunity by CD8⁺ T_Ab-supp cells, we also published that CXCR5⁺CD8⁺ T cells and CXCR5⁺IFNγ⁺CD8⁺ T cell are detected in the peripheral blood of human KTx recipients and their quantity inversely correlates with the development of DSA in the first year posttransplant.²¹ This suggests that a relative deficit in CXCR5⁺IFNγ⁺CD8⁺ T cells is a significant independent risk factor to the development of DSA even with therapeutic maintenance immunosuppression. Thus, the translational focus of the current studies was to investigate the impact of conventional maintenance immunosuppressive agents on the quantity and efficacy of ACT with CD8⁺ T_Ab-supp cells.

Calcineurin inhibitors block T cell activation and downstream effector functions and are used as a first-line therapy at most transplant centers.^47,48 Our group has previously reported that CNi administered posttransplant does not reduce in vivo alloantibody production by alloprimed IgG1⁺ B cells in a hepatocellular transplant model.⁴⁹ Here, we find that ACT with CD8⁺ T_Ab-supp cells to suppress alloantibody production and AMR after KTx is less effective when administered in combination with CNi. ACT in combination with CNi reduces the in vivo proliferation, in vivo cytotoxicity and quantity of alloprimed CD44⁺CXCR5⁺CD8⁺ T cells detected after adoptive transfer and impairs or abrogates ACT therapeutic efficacy (summarized in Figure 9). CNi directly interferes with CXCR5⁺CD8⁺ T cell-mediated in vivo killing of alloprimed IgG⁺ B cells without inhibiting expression of cytotoxic effector molecules. Thus, we conclude that concurrent treatment with CNi is detrimental to the efficacy of ACT with CD8⁺ T_Ab-supp cells for suppression of humoral alloimmunity due to its inhibitory effects on CD8⁺ T_Ab-supp cell persistence, expansion, and function after transfer.

Figure 9. Differential impact of immunosuppressive agents, mTORi and CNi, on the downregulation of humoral immunity mediated by antibody-suppressor CXCR5⁺CD8⁺ T cells.

A, After allogeneic KTx in CCR5 KO mice, primed T_FH cells provide signals to GC B cells through IL-21/IL-21R and CD40L/CD40 interactions that drive the production of alloantibody. Recipients with a deficit in antibody-suppressor CXCR5⁺CD8⁺ T cells (such as CD8 deficient or CCR5 deficient hosts) have dysregulated humoral immunity and produce high levels of antibody. B, Antibody-suppressor CXCR5⁺CD8⁺ T ACT into KTx recipients inhibits antibody production by direct killing of IgG⁺ GC B cells through perforin/granzyme and FasL/Fas-mediated cytotoxicity. Antibody-suppressor CXCR5⁺CD8⁺ T cells also indirectly inhibit CD4⁺ T_FH cells. Treatment of KTx recipients with ACT suppresses alloantibody production. C, CNi impairs the efficacy of antibody-suppressor CXCR5⁺CD8⁺ T cell ACT to downregulate humoral immunity after KTx by reducing CXCR5⁺CD8⁺ T cell persistence, in vivo proliferation and in vivo cytotoxic effector function. KTx recipients that receive ACT in combination with CNi have higher alloantibody production than recipients that received ACT alone. D, mTORi independently suppresses humoral immunity through inhibition of antibody-producing B cells and direct/indirect inhibition of CD4⁺ T_FH cells resulting in reduced quantity of GC B cells. Concurrent treatment with mTORi preserves the quantity, in vivo proliferation and in vivo cytotoxic effector function of antibody suppressor CXCR5⁺CD8⁺ T cell after ACT. Both ACT and mTORi reduce the quantity of GC B cells by different mechanisms. Combination therapy of ACT and mTORi additively suppresses humoral immunity after KTx. ACT, adoptive cell therapy; CNi, calcineurin inhibitor; FasL, Fas ligand; GC, germinal center; IgG, immunoglobulin G; IL-21, interleukin 21; KO, knockout; KTx, kidney transplant; MHC, major histocompatibility complex; mTORi, mammalian target of rapamycin inhibitor; TCR, T cell receptor; T_FH, follicular helper CD4⁺ T cell.

In contrast, we found that the addition of mTORi to ACT did not impair the in vivo proliferation or in vivo cytotoxic effector function of adoptively transferred CD8⁺ T_Ab-supp cells and enhanced the suppression of humoral alloimmunity and AMR in CCR5 KO KTx recipients. Low dose mTORi preserved the quantity of alloprimed CD44⁺CXCR5⁺CD8⁺ T cells detected in the spleen after ACT whereas high dose mTORi partially reduced their quantity. These findings may reflect CNi and mTORi dose-dependent induction of apoptosis of alloprimed CXCR5⁺CD8⁺ T cells as suggested by in vitro co-culture studies (Figure S5).

In published studies mTORi enhances the efficacy of cell therapy with Treg cells as Tregs are expanded ex vivo in mTORi treated cultures and Treg cell therapy is optimal when combined with mTORi therapy.^50,51 Chen et al found that a subset of stem cells with a reduced mTOR complex (mTORC1) have reduced susceptibility to mTORi.⁵² It would be interesting to determine if mTORC1 is similarly reduced in Tregs and alloprimed CXCR5⁺CD8⁺ T cells. Our studies to date suggest that the optimal maintenance immunosuppression therapy to combine with alloprimed CXCR5⁺CD8⁺ T cell ACT would include low dose mTORi in a CNi-free regimen, and possibly other immunosuppression drugs.

It was interesting to note the differential effects of ACT, mTORi and CNi upon T_FH and GC B cells. While the quantities of T_FH and IL-21⁺T_FH cells were significantly reduced in all treatment groups, only KTx recipients treated with ACT, mTORi or ACT in combination with mTORi exhibited significantly reduced GC B cells, alloantibody titer, and the greatest amelioration in AMR pathology. Rat KTx studies report a CNi dose-dependent suppression of alloantibody production and AMR (high-dose cyclosporin) that was attributed to depletion of T_FH cells and decreased T_FH expression of IL-21 without impacting total B cell or IgG⁺ B cell frequency.⁵³ In the current studies using a high alloantibody-producing murine KTx model, we found that CNi monotherapy or CNi in combination with ACT both significantly inhibit the quantity of total T_FH and IL-21⁺T_FH cells detected posttransplant but did not significantly impact the quantity of GC B cells or alloantibody titer compared to untreated controls. The addition of CNi to ACT did not further reduce the quantity of T_FH or IL-21⁺T_FH cells compared to ACT alone, was associated with increased alloantibody titer compared to ACT alone, and impaired the overall efficacy of ACT with CD8⁺ T_Ab-supp cells to suppress humoral alloimmunity and AMR. While treatment of CCR5 KO KTx recipients with CNi alone did not reduce alloantibody titer, CNi at high dose did reduce the severity of AMR (Figure 1D). This might reflected CNi-mediated inhibition of innate immune cell development, proliferation, and/or function that participate in antibody-dependent cellular cytotoxicity (i.e. macrophages and NK cells).⁵⁴

Collectively, our studies demonstrate the opposing effects of CNi and mTORi upon the efficacy of ACT with CD8⁺ T_Ab-supp cells for suppression of humoral alloimmunity and AMR. The preferential benefit of mTORi for concurrent immunosuppression with ACT is associated with preserved quantity, in vivo proliferation and in vivo cytotoxic effector function of CD8⁺ T_Ab-supp after ACT and an independent inhibitory effect on GC B cell development and alloantibody production after KTx. In future research, it would be interesting to investigate the quantity of circulating CXCR5⁺CD8⁺ T cells detected in the peripheral blood of human KTx recipients receiving mTORi and/or CNi immunosuppression at various therapeutic dose ranges. Furthermore, in the subset of patients who are converted to a CNi-free immunosuppression regimen to minimize nephrotoxicity or other CNi-related side effects, it would be of interest to assess the quantity of peripheral blood CXCR5⁺CD8⁺ T cells before and after CNi withdrawal. Future investigation to explore the safety and successful application of ACT with CD8⁺ T_Ab-supp cells in humans requires significant additional research to identify optimal cell dosing strategies, strategies for ex vivo cell expansion, methods for in vivo tracking of transferred cells, determination of ACT safety profile, and long-term patient and graft outcomes. These studies also highlight the unintended consequences of systemic immunosuppression used for prevention of acute cellular rejection upon endogenous regulatory cells that suppress alloimmunity and have implications for the design of preclinical and clinical trials investigating cell-based therapies in transplant recipients.

Nonstandard Abbreviation

ACT

adoptive cell therapy

AMR

antibody-mediated rejection

GC

germinal center

i.p.

intraperitoneal

IS

immunosuppression

KTx

kidney transplant

T_Ab-supp

antibody-suppressor T cells

T_FH

follicular helper CD4⁺ T cell

T_regs

regulatory T cells

WT

wild-type