effect of ex vivo expanded recipient regulatory t cells on ... · rds), the banting foundation (to...

OriginalBasicScienceçGeneral

Effect of Ex Vivo–Expanded Recipient RegulatoryT Cells on Hematopoietic Chimerism and KidneyAllograft Tolerance Across MHC Barriers inCynomolgus MacaquesRaimon Duran-Struuck, DVM, PhD,1,2 Hugo P. Sondermeijer, MD, MS,1 Leo Bühler, MD,3

Paula Alonso-Guallart, DVM,1 Jonah Zitsman, BS,1 Yojiro Kato, MD, PhD,1,2 Anette Wu, MD, MPH,1

Alicia N. McMurchy, PhD,4 David Woodland, MD,1,2 Adam Griesemer, MD,1,2 Mercedes Martinez, MD,1,5

Svetlan Boskovic, MD,6 Tatsuo Kawai, MD, PhD,6 A. Benedict Cosimi, MD,6 Cheng-Shie Wuu, PhD,7

Andrea Slate, DVM,8 Markus Y. Mapara, MD, PhD,1,9 Sam Baker, DVM,8 Rafal Tokarz, PhD,10

Vivette D'Agati, MD,11 Scott Hammer, MD,12 Marcus Pereira, MD,12 W. Ian Lipkin, MD,10 Thomas Wekerle, MD,13

Megan K. Levings, PhD,4 and Megan Sykes, MD1,2,9,14

Background. Infusion of recipient regulatory T (Treg) cells promotes durable mixed hematopoietic chimerism and allograft tol-erance in mice receiving allogeneic bone marrow transplant (BMT) with minimal conditioning. We applied this strategy in a Cyno-molgus macaque model. Methods. CD4+ CD25high Treg cells that were polyclonally expanded in culture were highlysuppressive in vitro andmaintained high expression of FoxP3. Eight monkeys underwent nonmyeloablative conditioning andmajorhistocompatibility complex mismatched BMTwith or without Treg cell infusion. Renal transplantation (from the same BMT donor)was performed 4 months post-BMTwithout immunosuppression to assess for robust donor-specific tolerance. Results. Tran-sient mixed chimerism, without significant T cell chimerism, was achieved in the animals that received BMT without Treg cells(N = 3). In contrast, 2 of 5 recipients of Treg cell+ BMT that were evaluable displayed chimerism in all lineages, including T cells,for up to 335 days post-BMT. Importantly, in the animal that survived long-term, greater than 90% of donor T cells wereCD45RA+ CD31+, suggesting they were new thymic emigrants. In this animal, the delayed (to 4 months) donor kidney graftwas accepted more than 294 days without immunosuppression, whereas non–Treg cell BMT recipients rejected delayed donorkidneys within 3 to 4 weeks. Early CMV reactivation and treatment was associated with early failure of chimerism, regardless ofTreg cell administration. Conclusions. Our studies provide proof-of-principle that, in the absence of early CMV reactivation(and BM-toxic antiviral therapy), cotransplantation of host Treg cell can promote prolonged and high levels of multilineage alloge-neic chimerism and robust tolerance to the donor.

(Transplantation 2017;101: 274–283)

Received 18 May 2016. Revision received 12 September 2016.

Accepted 30 September 2016.1 ColumbiaCenter for Translational Immunology, Department ofMedicine, ColumbiaUniversity Medical Center, Columbia University, New York, NY.2 Department of Surgery, Columbia University Medical Center, Columbia University,New York, NY.3 Transplantation Biology Research Center, Massachusetts General Hospital,Boston, MA.4 Department of Surgery, University of British Columbia and British Columbia Chil-dren’s Hospital Research Institute. Vancouver, Canada.5 Division of Gastroenterology, Department of Pediatrics, Columbia University Med-ical Center, Columbia University, New York, NY.6 Department of Surgery, Massachusetts General Hospital. Boston, MA.7 Department of Radiation Oncology, Columbia UniversityMedical Center, ColumbiaUniversity, New York, NY.8 Institute of Comparative Medicine, Columbia University. New York, NY.9 Division of Hematology/Oncology, Department of Medicine, Columbia UniversityMedical Center, Columbia University, New York, NY.

10 Center for Infection and Immunity, Mailman School of Public Health, ColumbiaUniversity. New York, NY.11 Department of Pathology, Columbia University Medical Center, Columbia Univer-sity, New York, NY.12 Division of Infectious Diseases, Columbia University Medical Center, ColumbiaUniversity, New York, NY.13 Section of Transplantation Immunology, Department of Surgery, Medical Univer-sity of Vienna, Vienna, Austria.14 Department of Microbiology and Immunology, Columbia University Medical Cen-ter, Columbia University, New York, NY.

Funding for these studies was provided by NIH grant RO1OD017949, by startupfunds from Columbia University Departments of Medicine and Surgery (to MS andRDS), the Banting Foundation (to MS), the Columbia University core award (to RDS),and the Irving Pilot Translational science award for new investigators (to RDS).Research reported in this publication was performed in the CCTI Flow CytometryCore, supported in part by the Office of the Director, National Institutes of Healthunder awards S10RR027050. The content is solely the responsibility of the authorsand does not necessarily represent the official views of the National Institutes of Health.

The authors declare no conflicts of interest.

274 www.transplantjournal.com Transplantation ■ February 2017 ■ Volume 101 ■ Number 2

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.

R.D.-S. and H.P.S. contributed equally.

R.D.S., H.S., M.S. performed research, analyzed data, and wrote the article.L.B., P.A.G., and J.Z. performed research and analyzed data. R.T., W.I.L.performed cytomegalovirus (CMV) testing. Y.K., B.C., S.V., T.K., and A.G.performed the surgeries and collected samples and participated in dataanalysis. C.S.W. provided the irradiation. S.H. and M.P. participated inresearch and CMV assessment. A.L.M. and M.L. provided Treg cell expertiseand artificial APCs. A.S. and S.B. participated in NHP care. M.M. participatedin research. M.M. participated in research as the bone marrow transplantconsultant. T.W. participated in research and article. V.A. performed pathology.

Correspondence: Megan Sykes, MD, Columbia Center for TranslationalImmunology, 630 W. 168th Street, Mailbox 127, NY NY 10032. ([email protected]).

Supplemental digital content (SDC) is available for this article. Direct URL citationsappear in the printed text, and links to the digital files are provided in the HTMLtext of this article on the journal’s Web site (www.transplantjournal.com).

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.

ISSN: 0041-1337/17/10102-274

DOI: 10.1097/TP.0000000000001559

© 2016 Wolters Kluwer Duran-Struuck et al 275

CD4+ FoxP3+regulatory T (Treg) cells modulate auto-immune and alloimmune responses.1-5 Induction of

kidney allograft tolerance, via transient mixed hematopoieticchimerism and nonmyeloablative conditioning, has beenachieved in large animal models,6 and humans.7 However,kidney allograft tolerance was achieved in only 60% to 70%of cynomolgus monkeys (cynos) and humans, and tolerancecould not be readily extended to islet, heart or lung allograftsin monkeys.8-10 Although durable mixed chimerism has beenachieved with total lymphoid radiation, antithymocyte globu-lin and donor kidney transplantation in the HLA-identicaltransplant setting, this approach has not yet succeeded inachieving durable chimerism or tolerance across HLA bar-riers.11-14 Another approach achieves renal allograft tolerancewith development of full donor chimerism across extensiveHLA barriers,15,16 but the full donor chimerism likely reflectsthe more rigorous and potentially toxic host conditioningand/or graft-versus-host reactivity of the infused donor Tcells, which eliminates recipient hematopoiesis, and highrates of opportunistic infection were observed.17 Mixed chi-merism, in contrast, provides a steady supply of recipient-derived antigen presenting cells (APCs), conferring superiorability to mount cytotoxic T cell responses that clear viral in-fections compared to full chimeras.17-20 Thus, the reliableachievement of durable mixed chimerism across HLA bar-riers, with its potential to induce tolerance to any type of do-nor organ and to cure congenital hematologic disorders,remains an important and elusive goal in humans.21,22

In mice, adoptive transfer of recipient blood-derived natu-ral Treg cells at the time of bone marrow transplant (BMT)with minimal conditioning regimen permitted the establish-ment of permanent hematopoietic mixed chimerism and skinallograft tolerance.23-26 We have adapted the use of Treg cellsfor the abovementioned cyno model that otherwise achievesonly transient mixed hematopoietic chimerism and whichhas been extensively characterized.8-10 We tested the hypoth-esis that the addition of expanded recipient Treg cells to the“standard” conditioning protocol would promote durablechimerism and allow acceptance of a donor kidney after amarked delay of 4months, when donor kidneys are uniformlyrejected by transient chimeras prepared with this protocol.27

MATERIALS AND METHODS

AnimalsMale adult cynos (Charles River Primates, Wilmington,

MA and Sanofi-Synthelabo, Bridgewater, NJ) were used. Allprocedures were approved by the IACUC of Columbia Uni-versity and Massachusetts General Hospital (MGH). Bothare AAALAC international accredited institutions.

Copyright © 2017 Wolters Kluwer H

Cynomolgus Major HistocompatibilityComplex Genotyping

Peripheral blood mononuclear cell (PBMCs) were geno-typed at theUniversity ofWisconsin PrimateResearchCenterLaboratory http://www.primate.wisc.edu/wprc/services/genetics.html.28-31

Conditioning RegimenRecipients of major histocompatibility complex (MHC)

mismatched donor BMT (Table 1 and Figure S1, SDC,http://links.lww.com/TP/B373) underwent the “standard”conditioning regimen as previously described6,32 +/− Tregcells (Figure 1A). Cyclosporine levels were maintained be-tween 200-400 ng/mL.

Treg Cell Sorting and ExpansionThe 1.0% of CD4+ T cells expressing the highest levels of

CD25 were sorted (FACSAria or Influx, BD Biosciences,Billerica, MA) and plated (1 � 105 cells/cm2) on fibroblasts(L929) (10 � 105 cells/cm2) expressing human CD32 (FcR),CD58 (LFA-3) and CD8033,34 (referred as artificial APCs[aAPC]) in combination with human recombinant IL-2(200 U/mL), anti-CD3 (SP-34) 100 ng/mL, and rapamycin100 μg/mL (Sigma-Aldrich, St Louis, MO) for 7 days.Growth medium consisted of RPMl-1640 (Gibco), fetal calfserum (Gibco), L-glutamine, penicillin/streptomycin, andnonessential amino acids. After 7 days, cells were replatedwith irradiated donor PBMCs (1 PBMC to 1 Treg cell) andIL-2 (200 U/mL), anti-CD3 (SP-34) 1 μg/mL or alternativelyin combination with aAPCs (3 � 105 to 5 � 105 cells percm2) (Table 2). Cells were cultured for another 7 days, thensplit and cultured for another 7 days. When irradiated aAPCswere used, Treg cells were cultured for an additional 5 days inthe presence of rapamycin 100 ng/mL (Table 2). After expan-sion, cells were cryopreserved in fetal calf serum (Gibco) with5% dimethyl sulfoxide for future use.

BMTBone marrow (BM) was harvested aseptically from donor

iliac bones by multiple percutaneous aspirations or surgicallyfrom the vertebrae. BM cells (1.3-3.0 � 108 mononuclearcells/kg) were infused intravenously. CD34+ content was1% (+/− 0.4%), as determined by flow cytometry.

Kidney TransplantationThe details of the kidney transplant procedure were re-

ported previously.35 Kidneyswere transplanted between days119 and 134 post-BMT. Recipients underwent unilateral na-tive nephrectomy and ligation of the contralateral ureter onthe day of transplant. The remaining native kidney was re-moved ~100 days after transplantation.

ealth, Inc. All rights reserved.

mailto:[email protected]

mailto:[email protected]

TABLE 1.

Donor: recipient MHC mismatches (refer to Figure S1, SDC,http://links.lww.com/TP/B373)

Recipient Treg cellsCMV

(serology)Mismatchin MHC-I

Mismatchin MHC-II

90-47 No + 2 haplotype 1 haplotype90-1 No - 2 haplotype 1 haplotype90-7 No + 2 haplotype 1 haplotypeM5210 Yes +* 2 haplotype 1 haplotype90-39 Yes - 1 haplotype 2 haplotype90-15 Yes + 1 haplotype 1 haplotype6c64 Yes + 1 haplotype 2 haplotype6c1 Yes + 1 haplotype 1 haplotype

All recipients had donors with 1 or both MHC-I and II alleles mismatched.* Animal M5210 was not tested but assumed to be positive.

276 Transplantation ■ February 2017 ■ Volume 101 ■ Number 2 www.transplantjournal.com

Flow Cytometric Analyses, Detection of Chimerism,and Cell Sorting

Whole bloodwas lysed and labeledwith a combination of thefollowingmAbs: CD3 PerCPCy5.5 (SP34.2), CD4-APC (L200),CD4-PE (L200), CD8-APC (SK1), CD11b-PE (ICRF44).CD20-PE (2H7), CD25-PE (BC96), CD31-PE (WM59),CD56-PE (MY31), pan-MHC A.B.C-PE (W6/32), FOXP3-PE(236A/E7). For chimerism analysis, we used H38 (anti-BW6;

FIGURE 1. Transplant scheme and Treg cell expansion. A, Transplant pThe average number of cells for each animal at each expansion timepointthe end of culture, with high levels of CD25 and of FOXP3.

Copyright © 2017 Wolters Kluwer

One Lambda, Inc., Canoga Park, CA). The recipient and donorpairs were chosen based on their MHC haplotypes and H38expression. The fluorescence of the stained samples was ana-lyzed using FACS Calibur and FlowJo software.

Mixed Lymphocyte Reactions and Treg CellSuppression Assays

Mixed lymphocyte reactions (MLRs) were performed aspreviously described.6 In addition, Treg cells were titratedfor their specificity in suppressing host antidonor versus thirdparty and donor antihost versus third party responses. Donoror host PBMC responders were stimulated with irradiatedhost, donor, or third-party PBMCs. Host nonirradiated Tregcells were added to the culture and pulsed with tritiated thy-midine 4 days after initiation of culture and read in a betacounter as previously described.6 Treg cells were also testedfor suppression of anti–CD2-, anti–CD3-, and anti–CD28-coated NHP activation bead-mediated activation (MiltenyiBiotec) at 1 bead to every 2 PBMCs.

RESULTS

Expansion, Phenotype, and Suppression ofCyno CD4+ CD25high

An average of 118907 ± 9588 CD4+ CD25high cells weresorted from each blood draw. Usually a 10- to 100-fold

rotocol. B, Expansion of Treg cell lines from 4 animals over 4 weeks.is graphed (SEM) (bars). C, A representative phenotype of Treg cells at

Health, Inc. All rights reserved.

TABLE 2.

Treg cell protocols

Reagents Day 0 Day 7 Day 14 Day 21

IL-2 ✓ ✓ ✓

Anti-CD3 ✓ ✓ ✓

Donor PBMCs - ✓ ✓ FREEZEArtificial APCs ✓ - -Rapamycin ✓ - -

Reagents Day 0 Day 7 Day 14 Day 21 Day 26

IL-2 ✓ ✓ ✓ ✓

Anti-CD3 ✓ ✓ ✓ ✓

Donor PBMCs - ✓ ✓ ✓ FREEZEArtificial APCs ✓ ✓ ✓ ✓

Rapamycin ✓ - - ✓


expansion was achieved within the first 7 (0.695 �106 ± 0.175 � 106) to 14 (22.47 � 106 ± 4.3 � 106) daysof culture (Figure 1B, representative lines). At the end of cul-ture, Treg cells were analyzed for phenotype (Figure 1C) andfunction (Figures 2A, B) before cryopreservation.

Infused Treg cell expressed high levels of FOXP3 andCD25(Figure 1C) (Table S1, SDC, http://links.lww.com/TP/B373).

FIGURE 2. Culture of cynomolgus regulatory T cells. A, Highest quality(anti-CD2CD3CD28) PBMCs at 1:32 Treg cell/PBMC ratio. B, All TregTreg cell/PBMC ratio. Microsuppression assays shown. (C-F) MLRs(E, F) PBMC responders were plated with either host, donor, or third-pindicated PBMC/Treg cell ratios (1:1, 1:2 and 1:4) and assessed for supbars indicate SE. Similar results were obtained in a repeat experiment (n


Inhibition of the proliferation of bead- (anti-CD2/CD3/CD28) stimulated autologous (cryopreserved pretransplant)PBMCs generally revealed greater than 95% suppression ata 1:1 ratio of PBMCs/Treg cell (Figures 2A and B). The in-fused Treg cells varied in suppressive potency but all achieved50% suppression at or above a 1:2 Treg cell/PBMC ratio(Figures 2A and B) (Table S1, SDC, http://links.lww.com/TP/B373).

We aimed to generate polyclonal, nonspecifically sup-pressive Treg cell lines with our expansion protocol. Whiledonor PBMCs were added during the expansion period as asource of APCs, specificity studies on 2 different Treg celllines (Figures 2C-F) revealed similar suppression of hostantidonor, host anti-third–party, antihost, and donoranti–third-party responses.

Proof of Concept that Treg Cell Infusion Can ProlongMultilineage Donor Cell Chimerism

We tested whether polyclonal Treg cells could prolong do-nor hematopoietic chimerism compared to controls, whichhistorically achieved transient (30-60 days) chimerism.6 Threecontrol animals were treated as previously described,6 exceptthey did not receive a donor kidney graft on Day 0. Five ani-mals received the same treatment plus Treg cell infusionsposttransplant. These 5 animals (M5210, 90-39, 6c64, 6c1,

Treg cell suppressed over 50% the proliferation of bead-stimulatedcell lines achieved at least 50% suppression of proliferation at a 1:2assessing the specificity of host Treg cell. Host (C, D) and donorarty stimulators. Host Treg cells were added to the cultures at thepressive activity. All data points represent means of triplicates. Errorot shown).


FIGURE 3. Summary of percent donor chimerism. Granulocyte (black circles), monocyte (blue triangles) and lymphocyte (red squares) line-ages of each animal are shown. Animals with boxed identification numbers (M5210, 90-39, and 90-1) received no antiviral treatments. Animalswhose figures have a cross died or were euthanized due to untreated or unmanageable CMV disease. Animal 90-47 was serologically CMV−pre-Tx, but developed CMVafter BMT from a CMV+ donor that had been serologically negative on initial screen. Animal 90-1, who was CMV−pre-Tx and received a BMT from a CMV− donor, never developed CMV. The 5 animals shown on the top part of the figure received BMT + Tregcells, whereas the 3 below the dotted line received BMTwithout Treg cells.


90-15) received expanded polyclonal autologous Tregcells (15-53 � 106 per infusion) during the first weekposttransplant (days 0, 2, 5, 7) and on day +50 (Table S1,SDC, http://links.lww.com/TP/B373). Total dose was88-96 � 106/kg. Two Treg cell recipients, M5210 and 90-39,developed significant multilineage chimerism (Figure 3, toprow). Although M5210 survived long-term, animal 90-39died of cytomegalovirus (CMV) disease on day 43 with sig-nificant donor chimerism in all lineages (Figure 3). The chi-merism in M5210 (Figure 3, top left panel) persisted longerthan ever observed in this model, remaining detectable inthe lymphoid, monocyte and granulocyte lineages until days292, 224, and 335, respectively.

We monitored CMV viremia, and when it exceeded10 000 copies/mL (initially) or 1000 copies/mL (after our ex-perience in the first few animals), we treated animals withGanciclovir and/or Foscarnet. Animal 6c64 was given antivi-ral prophylaxis to prevent CMV reactivation and developedonly low and short-lived chimerism (Figure 3) with pro-longed pancytopenia, suggesting that BM-toxic effects ofthe antiviral treatment may have impaired both donor andrecipient hematopoiesis. Two additional Treg cell recipients,6c1 and 90-15, experienced CMV reactivation with high vi-ral loads and required treatment with antivirals at high doseswithin the first week posttransplant. These animals developedonly short-lived and low levels of chimerism (6c64, 6c1,90-15 shown in Figure 3, middle row) in association with


protracted cytopenias, often requiring transfusions. These re-sults suggest that CMV reactivation and/or the bone marrowtoxic effects of early antiviral therapy may have potentiallyinterfered with initial engraftment of the donor marrow.

Of the 3 control animals (90-47,90-7,90-1) receiving BMTwithout Treg cell infusion, 1 (animal 90-47) died of CMVbe-fore the development and implementation of the CMV surveil-lance and treatment protocol, without showing any significantchimerism (Figure 3, bottom row, left panel). The 2 other con-trols survived long term. Animal 90-7 also developed low-level, short-lived chimerism. The third control animal, 90-1,was unique in that both it and the donor were CMV-negativeand chimerism lasted over 100 days before disappearing.6,32

In summary, recipients that reactivated CMV, regardless ofTreg cell infusion, succumbed to disease if not treatedpromptly with antivirals. Early CMV reactivation and itstreatment or prophylaxis were associated with very short-lived chimerism. Only 2 animals survived without antiviraltreatment (1 Treg cell recipient and 1 CMV-negative non-Treg cell control.) The Treg cell recipient, M5210, had only avery low level CMV viremia (<1000 copies/mL) and exhibitedthe longest documented donor chimerism ever seen with thisor related protocols over a period of more than twenty years.

Only Treg Cell Recipients Developed T Cell ChimerismThe 3 evaluable (ie, that were not treated early with antivirals)

animals that had measurable lymphoid chimerism included


FIGURE 4. Chimerism analysis of BMTrecipients. A, Representative flow cytometry of animalM5210 on day + 99post-BMT. Donor chimerismis measured with the Bw6+ (MHC-I) marker. B cell (CD20+), monocyte (CD11b+) and Tcell (CD3, CD4 and CD8) chimerism measured amongcells with low/medium forward and side scatter (not shown). Granulocytes (CD11b+) were analyzed among cells with high forward and sidescatter (not shown). B, B cell (black circle), CD4 (red triangle) and CD8 (blue square) Tcell chimerism in the 3 animals with the highest andmostprolonged chimerism. CMV− control animal 90-1 developed high levels B cell chimerism, but no Tcell chimerism and chimerism declined afterdiscontinuation of immunosuppression and was completely lost by day 110. Treg cell recipient M5210 developed high B cell chimerism anddelayed, prolonged Tcell chimerism in both CD4 and CD8 Tcell lineages. Chimerism lasted over 300 days post-BMT. Treg cell recipient 90-39was euthanized at day 43 due to CMV disease. At the time of euthanasia B cell and CD4 Tcell chimerism was detectable in the peripheral blood.


Treg cell recipients M5210 and 90-39 and CMV(−) controlrecipient 90-1 (Figure 4). However, in the non–Treg cell re-cipient 90-1 (the control animal in which donor and recipientwere CMV negative), lymphoid chimerism includedNK cells(data not shown) and B cells (eg, Figure 4B, left), but did notinclude significant donor T cell chimerism. In contrast, bothevaluable Treg cell recipients had not only B cell and NK cellchimerism, but also significant CD4 and CD8 T cell chime-rism (Figures 4B, center and C, right). In M5210, the long-lived Treg cell recipient, T cell chimerism first appeared45 days post-BMT (2.5 weeks after cyclosporine had beendiscontinued) and increased significantly on day +60 post-BMT. Similarly, Treg cell recipient 90-39 (which died ofCMVon day + 43) had a spike in donor T cell chimerism inthe peripheral blood 1 month after BMT (at the time immu-nosuppression was discontinued) (Figure 4B, right) and stillhad peripheral blood T cell chimerism (5%, mostly in CD4T cells) on the day of euthanasia. These results represent thefirst time that T cell chimerism has been observed using thisnonmyeloablative monkey BMT model and suggest thatTreg cells promote T cell chimerism.


We then investigated, in the only long-lived Treg cell recip-ient chimera, the phenotype of host and donor Tcells, includ-ing CD31, a marker expressed on new thymic emigrants36

and CD45RA, a marker of naïve T cells, among both donorand recipient T cells (Figure 5). Almost all donor CD4 andCD8 T cells in M5210 expressed CD31 throughout follow-up(Figures 5A and B). Consistent with de novo origin in the re-cipient thymus, the expression of CD45RA was also veryhigh on donor CD4+ T cells (Figure 5C), peaking close to90%. For recipient T cells, expression of CD31 in bothCD4 and CD8 cell populations was significantly less. How-ever, CD31 expression increased markedly in host T cells(80%) early after the transplant (Figures 5A-B), suggestingthat a wave of new host T cells was released from the thymusafter transplant. The percentage of host-derived CD31+ Tcellsslowly decreased from day+/−20 until day +50. CD45RA ex-pression on recipient CD4+ T cells peaked at 50% at about1 month posttransplant. The expression of CD31 andCD45RAwas lower in animals that did not develop donorT cell chimerism, as shown in Figures 5A to C (bottomrows) for animal 90-7, a control BMT recipient that did


FIGURE5. Immune reconstitution of animal M5210 Post-BMT (+Treg cell) and 90-7 (control). (A-C, top rowM5210 and bottom row 90-7 con-trol). Dotted lines indicate first detection of donor chimerism in CD4 or CD8 Tcells. (A, top) Total CD4 Tcells (black circle) expressed CD31 atincreased levels posttransplant. Donor CD4 Tcells (blue squares) maintained high CD31 expression. Recipient-derived CD4 Tcells (red trian-gles) expressed lower levels of CD31. (B, top) Total CD8 Tcells expressed high CD31 levels after transplant (black circle). CD31 expression inhost CD8 Tcells (red triangles) decreased, however donor-derived cells (blue squares) maintained high CD31 expression long after transplant.(C, top) Donor-derived CD4 Tcells (blue squares) expressed higher levels of CD45RA compared to recipient CD4 Tcells (red triangles). In con-trast, animal 90-7 never exhibited any donor chimerism (A-C, bottom row) and the levels of CD31 and CD45 were lower than those observed inM5210 donor (blue) Tcell populations. D, Average percentage of CD4 Tcells expressing FOXP3 after transplant in Treg cell recipients (red tri-angles) (n = 2 animals) compared to control animals (blue squares) (n = 3). (E) Animal M5210 absolute Treg cell numbers (red triangles) com-pared with the absolute number of CD8 Tcells (blue squares).


not receive Treg cells. In summary, donor T cells exhibitedhigh levels of CD45RA and CD31, suggesting de novo de-velopment from the thymus in animal M5210.

Kinetics of CD3+ CD4+ FOXP3+ Cells in the CirculationAfter Infusion

Peripheral Treg cell counts and percentages were similar inthe 3 controls and the 2 evaluable Treg cell recipients thatdeveloped high levels of chimerism (M5210 and 90-39)(Figure 5D). Treg cells were largely of recipient origin(Figure S2, SDC, http://links.lww.com/TP/B373). A peakwas observed on day +50 in M5210 after the infusion of28 million Treg/kg, which was given in an effort to reversea sudden increase in the absolute number of CD95+CD28−

effector CD8 T cells (Figure 5E) of mostly recipient origin(Figure 4B, middle panel) 20 days after the discontinuationof cyclosporine A (levels on day + 48 were subtherapeutic at118 ng/mL and low on day +53 at 35 ng/mL). CD8 T cellcounts declined after the infusion of Treg cells on day +50


(Figure 5E). Both the Tcell chimerism (Figure 4B) and the my-eloid chimerism (Figure 3, top left panel) increased shortly af-ter the Treg cell infusion and remained stable for an additional80 days (Figure 3, top left). Of note, on day 80 post-BMT,there was a second increase in the absolute CD8 T cell countthat was followed by a subsequent spontaneous Treg cell in-crease, after which CD8 counts normalized (Figure 5E).

In summary, infusion of Treg cells was associated withthe development of T cell chimerism, prolonged multilineagechimerism, and reversal of increasing recipient effector CD8+

T cell counts in animal M5210.

Proof of Principle: Persistent Chimerism Induced byTreg Cell Treatment Was Associated With In VitroDonor-Specific Hyporesponsiveness andAllograft Tolerance

M5210 showed donor-specific unresponsiveness in MLRat day 106, before kidney transplantation was performed(Figure 6B), whereas strong proliferative responses to the donor


FIGURE 6. Tolerance to the donor in M5210 Treg cell recipient but not in control. A, Pretransplant MLR in animal M5210 demonstrates strongproliferative responses to donor (diagonal stripes) and weaker response to third party (horizontal stripes). B, Day 106 post-Tx MLR (before kid-ney allograft from the same BM donor). Proliferation is maintained to third party but not to donor. C, Animals that lost chimerism never becametolerant nor showed decreased antidonor proliferation. An example is shown in (C) pretransplant and (D) posttransplant day 78. (E) After kidneytransplant, creatinine levels in M5210 (black circle) stayed in the normal range while that in 2 control (non-Treg cell) animals (dashed grey lines)showed increases 2 to 3weeks after kidney transplant. F, G, Kidney histopathology. Biopsies were taken from transplanted donor kidneys at the timeof euthanasia. Shown is day +28 and day +294 (day of euthanasia) post-kidney transplant in animals 90-1 (F) (control, left) andM5210 (G) (+Treg cell,right) respectively. 90-1 had extensive lymphocyte infiltrates scored as a Banff grade 3 rejection, while M5210 showed no signs of rejection.


were present pre-BMT (Figure 6A). Similar responses to thirdpartywere observed pretransplantation and posttransplantation.

Animals were challenged with a solid organ allograft(a kidney from the same BMTdonor) 4months after the orig-inal BMT, without immunosuppression. Only 1 Treg cell re-cipient (M5210) was evaluable at the 4-month timepoint. Atthe time of kidney transplant, M5210 remained chimeric inall lineages (Figures 3 and 4). The recipient's contralateralureter was ligated on day 0 and on day 100 postkidney trans-plant the recipient's contralateral kidney was removed. Serumcreatinine levels (Figure 6E) remained normal and stable un-til the day of euthanasia 293 days postkidney transplant, dem-onstrating tolerance to the donor kidney. Histopathology onday +294 postkidney transplant showed no evidence forrejection in M5210 (Figure 6F).

Two control animals that underwent the same protocol(without Treg cells) were also grafted with a kidney fromtheir same BM donor 4 months post-BMT. In contrast toM5210, the 2 controls rejected their donor kidneys within amonth (Figure 6E), in line with previous results.27 The donorkidneys in nonchimeric control animals showed Banff grade3 rejection (Figures 6F and G) at the time of euthanasia.Nonchimeric Treg cell recipients (ie, animals that received


Treg cells but had short-lived hematopoietic chimerism inassociation with early CMV reactivation and treatment)retained antidonor proliferative responses (Figures 6C andD) and rejected donor kidneys on day +120 post-BMT(n = 2), similar to controls (data not shown). These resultsare proof of concept and suggestive of the importance ofmixed chimerism in tolerance induction in this model. Noor minimal antidonor alloantibody was detected in animalsthat rejected their donor kidneys (Figure S3, SDC, http://links.lww.com/TP/B373).

In summary, control recipients rejected the donor kidneysuniformly, whereas the only evaluable long-term survivingTreg cell recipient M5210 maintained normal kidney func-tion until termination of the experiment. This result providesproof-of-principle that prolongation of chimerism using ex-panded Treg cells can promote more robust tolerance thanthat achieved in previous studies using this model.6,27,32

DISCUSSIONOur studies provide proof-of-concept that expanded

recipient-derived polyclonal Treg cells can increase and extenddonor hematopoietic chimerism and promote robust allograft



tolerance across MHC barriers in a NHP nonmyeloablativeBMT model without an increased risk of GVHD. Infusionof Treg cell is an attractive approach to overcoming HvG re-sponses, because it may further reduce the risk of GVHD37

rather than increasing this risk or the overall toxicity of theconditioning regimen like most other approaches.

Phase I clinical trials using Treg cells have shown safety,38

but efficacy remains to be proven. Both induced and naturalTreg cell promoted engraftment, stable mixed chimerism andtolerance in mice under a minimal conditioning protocol inwhich the BM is otherwise rejected.24,26 We demonstrate ina monkey model that host Treg cells improved the level andduration of chimerism, extending it to the T cells. Moreover,robust donor-specific tolerance was achieved in 1 evaluableanimal such that a donor kidney grafted at 4 monthsposttransplant was accepted without immunosuppression.Previous studies using this protocol without Treg cells, inwhich donor BM and kidney were co-transplanted on day 0,were associated with long-term kidney graft survival in about60% of animals6,39 and donor hematopoietic chimerism (insome animals reaching +/−85%) consistently disappearedby day 60 post-BMT.32 A delay in grafting a donor kidneyto more than 3 months post-BMT was always associatedwith rejection of the donor graft.27

Previous work in NHPs (rhesus macaques) using anonmyeloablative BMT regimen with co-stimulatory block-ade achieved prolonged levels of donor chimerism as longas basiliximab and belatacept were infused. Chimerism waslost after discontinuation of this treatment,40 and allografttolerance was not achieved. In contrast, an animal in ourstudy retained chimerism to 335 days and accepted a donorkidney grafted at 4 months, despite stopping immunosup-pressive monotherapy at 28 days post-BMT.

Inmice, we have shown that the presence of Tcell chimerismis associated with early and long-term deletional tolerance, be-cause thymic engraftment of donor T cell progenitors reflectssuccessful ablation of intrathymic alloreactivity, permittingintrathymic engraftment of both thymocyte progenitors anddonor APCs that contribute to negative selection of donor-reactive T cells.22,41–43 For the first time in the more than20 years using this monkey nonmyeloablative BMT regimen,we have obtained evidence of de novo donor thymopoiesis,with T cell chimerism consisting of recent thymic emigrants inthe peripheral blood. CD45RA and CD31 expressions sug-gested that almost all donor cells were newly-developed,whereas recipient T cells were a mix of new thymic emigrantsand naive or memory T cells that evaded host conditioning.

CMV reactivation presented a major impediment toachieving the goals of our studies. Although this complica-tion has not been described in previous studies using themodel we adopted, uniform CMV reactivation has been re-ported in cynomolgus monkeys receiving Thymoglobulin.44

CMV reactivation was not directly caused by the infusionof Treg cells because control animals had a similar rate of re-activation. Only 1 animal,M5210, was able to control CMVwithout antiviral treatment. Because CMV itself can directlyaffect BM function45-47 and antiviral treatments are knownto be BM toxic, modifications to the protocol are needed.The increased duration and level of chimerism observedin the 1 CMV-negative (non–Treg cell recipient) transplant(albeit without Tcell chimerism or tolerance) supports a directrole for CMV and/or its treatment in limiting hematopoietic


engraftment. We are currently exploring substitution ofthe mTOR inhibitor rapamycin for CSA to better controlCMV reactivation and enhance Treg cell function, expansion,and survival.24,48

In summary, we provide proof-of-concept that BM plusexpanded cryopreserved polyclonal recipient Treg cell canprolong donor chimerism, promote T cell chimerism and in-duce robust tolerance without an increase of toxic condition-ing intensity or GVHD risk in a preclinical monkey model.Successful refinement of this protocol has the potential tobe translated to the clinic.

ACKNOWLEDGMENTSThe authors thank Dr. Remi Creusot for the critical review ofthe article.

REFERENCES1. Hori S, Nomura T, Sakaguchi S. Control of regulatory T cell development

by the transcription factor foxp3. Science. 2003;299:1057–1061.2. Kang SM, Tang Q, Bluestone JA. CD4 + CD25+ regulatory Tcells in trans-

plantation: progress, challenges and prospects. Am J Transplant. 2007;7:1457–1463.

3. Sakaguchi S, Vignali DA, Rudensky AY, et al. The plasticity and stability ofregulatory Tcells. Nat Rev Immunol. 2013;13:461–467.

4. Krummey SM, Ford ML. Braking bad: novel mechanisms of CTLA-4 inhi-bition of T cell responses. Am J Transplant. 2014;14:2685–2690.

5. Amarnath S, Mangus CW, Wang JC, et al. The PDL1-PD1 axis convertshuman TH1 cells into regulatory Tcells.Sci Transl Med. 2011;3:111ra120.

6. Kawai T, Cosimi AB, Colvin RB, et al. Mixed allogeneic chimerism and re-nal allograft tolerance in cynomolgus monkeys. Transplantation. 1995;59:256–262.

7. Kawai T, Cosimi AB, Spitzer TR, et al. HLA-mismatched renal transplanta-tion without maintenance immunosuppression. N Engl J Med. 2008;358:353–361.

8. Kawai T, Sogawa H, Koulmanda M, et al. Long-term islet allograft functionin the absence of chronic immunosuppression: a case report of a nonhu-man primate previously made tolerant to a renal allograft from the samedonor. Transplantation. 2001;72:351–354.

9. Kawai T, Cosimi AB, Wee SL, et al. Effect of mixed hematopoietic chime-rism on cardiac allograft survival in cynomolgusmonkeys. Transplantation.2002;73:1757–1764.

10. Aoyama A, Ng CY, Millington TM, et al. Comparison of lung and kidney al-lografts in induction of tolerance by a mixed-chimerism approach in cyno-molgus monkeys. Transplant Proc. 2009;41:429–430.

11. Scandling JD, Busque S, Dejbakhsh-Jones S, et al. Tolerance and chime-rism after renal and hematopoietic-cell transplantation. N Engl J Med.2008;358:362–368.

12. Scandling JD, Busque S, Dejbakhsh-Jones S, et al. Tolerance and with-drawal of immunosuppressive drugs in patients given kidney and hemato-poietic cell transplants. Am J Transplant. 2012;12:1133–1145.

13. Scandling JD, Busque S, Shizuru JA, et al. Induced immune tolerance forkidney transplantation. N Engl J Med. 2011;365:1359–1360.

14. Millan MT, Shizuru JA, Hoffmann P, et al. Mixed chimerism and immu-nosuppressive drug withdrawal after HLA-mismatched kidney andhematopoietic progenitor transplantation. Transplantation. 2002;73:1386–1391.

15. Leventhal J, Abecassis M, Miller J, et al. Chimerism and tolerance withoutGVHD or engraftment syndrome in HLA-mismatched combined kidney andhematopoietic stem cell transplantation. Sci Transl Med. 2012;4:124ra128.

16. Leventhal J, Abecassis M, Miller J, et al. Tolerance induction in HLA dispa-rate living donor kidney transplantation by donor stem cell infusion: dura-ble chimerism predicts outcome. Transplantation. 2013;95:169–176.

17. Leventhal JR, Elliott MJ, Yolcu ES, et al. Immune reconstitution/immunocompetence in recipients of kidney plus hematopoietic stem/facilitating cell transplants. Transplantation. 2015;99:288–298.

18. Rüedi E, Sykes M, Ildstad ST, et al. Antiviral Tcell competence and restric-tion specificity of mixed allogeneic (P1 + P2——P1) irradiation chimeras.Cell Immunol. 1989;121:185–195.

19. Ildstad ST,Wren SM,Bluestone JA, et al. Characterization ofmixed alloge-neic chimeras. Immunocompetence, in vitro reactivity, and genetic speci-ficity of tolerance. J Exp Med. 1985;162:231–244.



20. Zinkernagel RM, Althage A, Callahan G, et al. On the immunocompetenceof H-2 incompatible irradiation bone marrow chimeras. J Immunol. 1980;124:2356–2365.

21. Sharabi Y, Sachs DH.Mixed chimerism and permanent specific transplan-tation tolerance induced by a non-lethal preparative regimen. J Exp Med.1989;169:493–502.

22. Tomita Y, Khan A, SykesM. Role of intrathymic clonal deletion and periph-eral anergy in transplantation tolerance induced by bone marrow trans-plantation in mice conditioned with a non-myeloablative regimen. JImmunol. 1994;153:1087–1098.

23. Pilat N, Wekerle T. Mechanistic and therapeutic role of regulatory Tcells intolerance through mixed chimerism. Curr Opin Organ Transplant. 2010;15:725–730.

24. Pilat N, Baranyi U, Klaus C, et al. Treg-Therapy allows mixed chimerismand transplantation tolerance without cytoreductive conditioning. Am JTransplant. 2010;10:751–762.

25. Pilat N, Farkas AM, Mahr B, et al. T-regulatory cell treatment preventschronic rejection of heart allografts in a murine mixed chimerism model.J Heart Lung Transplant. 2014;33:429–437.

26. Pilat N, Klaus C, Hock K, et al. Polyclonal recipient nTregs are superior todonor or third-party Tregs in the induction of transplantation tolerance. JImmunol Res. 2015;2015:562935.

27. Kawai T, Poncelet A, Sachs DH, et al. Long-term outcome and alloanti-body production in a non-myeloablative regimen for induction of renal al-lograft tolerance. Transplantation. 1999;68:1767–1775.

28. Campbell KJ, Detmer AM, Karl JA, et al. Characterization of 47MHC classI sequences in Filipino cynomolgus macaques. Immunogenetics. 2009;61:177–187.

29. O'Connor SL, Blasky AJ, Pendley CJ, et al. Comprehensive characteriza-tion of MHC class II haplotypes in Mauritian cynomolgus macaques. Im-munogenetics. 2007;59:449–462.

30. Pendley CJ, Becker EA, Karl JA, et al. MHC class I characterization of In-donesian cynomolgus macaques. Immunogenetics. 2008;60:339–351.

31. Wiseman RW, Karl JA, Bimber BN, et al. Major histocompatibility complexgenotyping with massively parallel pyrosequencing. Nat Med. 2009;15:1322–1326.

32. Kawai T, Sogawa H, Boskovic S, et al. CD154 blockade for induction ofmixed chimerism and prolonged renal allograft survival in nonhuman pri-mates. Am J Transplant. 2004;4:1391–1398.

33. de Waal Malefyt R, Verma S, Bejarano MT, et al. CD2/LFA-3 or LFA-1/ICAM-1 but not CD28/B7 interactions can augment cytotoxicity byvirus-specific CD8+ cytotoxic T lymphocytes. Eur J Immunol. 1993;23:418–424.

34. Levings MK, Sangregorio R, Galbiati F, et al. IFN-alpha and IL-10 inducethe differentiation of human type 1 T regulatory cells. J Immunol. 2001;166:5530–5539.


35. Cosimi AB, Delmonico FL, Wright JK, et al. Prolonged survival of nonhu-man primate renal allograft recipients treated only with anti-CD4monoclo-nal antibody. Surgery. 1990;108:406–413.

36. Kimmig S, Przybylski GK, Schmidt CA, et al. Two subsets of naive T helpercells with distinct Tcell receptor excision circle content in human adult pe-ripheral blood. J Exp Med. 2002;195:789–794.

37. Sawitzki B, Brunstein C, Meisel C, et al. Prevention of graft-versus-hostdisease by adoptive T regulatory therapy is associated with active repres-sion of peripheral blood Toll-like receptor 5 mRNA expression. Biol BloodMarrow Transplant. 2014;20:173–182.

38. TangQ, Bluestone JA. Regulatory T-cell therapy in transplantation: movingto the clinic. Cold Spring Harb Perspect Med. 2013;3.

39. Kimikawa M, Sachs DH, Colvin RB, et al. Modifications of the condition-ing regimen for achievingmixed chimerism and donor-specific tolerance incynomolgus monkeys. Transplantation. 1997;64:709–716.

40. Kean LS, Adams AB, Strobert E, et al. Induction of chimerism in rhesusmacaques through stem cell transplant and costimulation blockade-based immunosuppression. Am J Transplant. 2007;7:320–335.

41. Tomita Y, Khan A, Sykes M. Mechanism by which additional monoclonalantibody (mAB) injections overcome the requirement for thymic irradiationto achieve mixed chimerism in mice receiving bone marrow transplanta-tion after conditioning with anti-Tcell mAbs and 3-Gy whole body irradia-tion. Transplantation. 1996;61:477–485.

42. Tomita Y, Sachs DH, Khan A, et al. Additional monoclonal antibody (mAb)injections can replace thymic irradiation to allow induction of mixed chime-rism and tolerance in mice receiving bone marrow transplantation afterconditioningwith anti-Tcell mAbs and 3-Gywhole body irradiation. Trans-plantation. 1996;61:469–477.

43. Nikolic B, Khan A, Sykes M. Induction of tolerance by mixed chime-rism with nonmyeloblative host conditioning: the importance of overcom-ing intrathymic alloresistance. Biol Blood Marrow Transplant. 2001;7:144–153.

44. Han D, Berman DM, Willman M, et al. Choice of immunosuppression in-fluences cytomegalovirus DNAemia in cynomolgus monkey (Macacafascicularis) islet allograft recipients.Cell Transplant. 2010;19:1547–1561.

45. Steffens HP, Podlech J, Kurz S, et al. Cytomegalovirus inhibits the engraft-ment of donor bonemarrow cells by downregulation of hemopoietin geneexpression in recipient stroma. J Virol. 1998;72:5006–5015.

46. Paulin T, Ringdén O, Lönnqvist B. Faster immunological recovery afterbone marrow transplantation in patients without cytomegalovirus infec-tion. Transplantation. 1985;39:377–384.

47. Fries BC, Khaira D, PepeMS, et al. Declining lymphocyte counts followingcytomegalovirus (CMV) infection are associated with fatal CMV disease inbone marrow transplant patients. Exp Hematol. 1993;21:1387–1392.

48. Battaglia M, Stabilini A, Roncarolo MG. Rapamycin selectively expandsCD4 + CD25 + FoxP3+ regulatory T cells. Blood. 2005;105:4743–4748.


effect of ex vivo expanded recipient regulatory t cells on ... · rds), the banting foundation (to...

Documents