Reduced Frequency of Regulatory T Cells in Peripheral Blood Stem ...

BRIEF ARTICLES

Reduced Frequency of Regulatory T Cells in Peripheral Blood Stem Cell Compared to Bone Marrow Transplantations Celine Blache,1 Joe-Marc Chauvin,1 Aude Marie-Cardine,2 Nathalie Contentin,3 Pascal Pommier,4 Ingrid Dedreux,1 Stephanie François,1 Serge Jacquot,1 Dominique Bastit,4 Olivier Boyer1 Peripheral blood stem cell transplantation (PBSCT) is an alternative to bone marrow transplantation (BMT). Although CD41CD251CD127lo regulatory T cells (Tregs) have been shown to play important roles in the control of T cell reactivity, the Treg contents of both graft types have not been analyzed comparatively to date. We report herein that Treg frequencies are significantly reduced in PBSC compared to BM transplants. Furthermore, most Tregs from PBSC transplants are CD62Llo, a phenotype reported to have poor suppressor activity. Both granulocyte-colony stimulating factor (G-CSF) administration and leukapheresis were found to contribute to the loss of CD62L1 Tregs. Although higher T cell numbers are infused in PBSCT than in BMT, it is possible that the reduced Treg content of PBSC transplants may represent 1 factor contributing to the higher risk of GVHD reported after PBSCT. Biol Blood Marrow Transplant 16: 430-434 (2010) Ó 2010 American Society for Blood and Marrow Transplantation

KEY WORDS: Regulatory T cells, Hematopoietic stem cell transplant, Human

INTRODUCTION Hematopoietic stem cell transplantation (HSCT) is a common therapeutic option for different hematologic diseases including leukemia. In addition to cord blood (CB), the HSCTs mostly consists of bone marrow (BM) cells or peripheral blood stem cells (PBSC) that are recovered by leukapheresis after granulocyte-colony stimulating factor (G-CSF) mobilization. Both BM and PBSC transplants (PBSCTs) contain high numbers of mature T cells that contribute to T cell reconstitution, favor engraftment, and may provide an antileukemic effect, but also cause graft-versus-host disease (GVHD). From the 1Inserm, U905, Rouen, France; University of Rouen, IFRMP, Institute for Biomedical Research, Rouen, France; Rouen University Hospital, Department of Immunology, Rouen, France; 2Rouen University Hospital, Pediatric Immuno-Hemato-Oncology Unit, Rouen, France; 3Centre Henri Becquerel, Department of Hematology, Rouen, France; and 4 Etablissement Français du Sang de Normandie, Cell Therapy Unit, Rouen, France. Financial disclosure: See Acknowledgments on page 433. Correspondence and reprint requests: Olivier Boyer, MD, PhD, Inserm U905, Faculte de Medecine, 22 Bd Gambetta, 76000 Rouen, France (e-mail: [email protected]). Received September 7, 2009; accepted October 23, 2009 Ó 2010 American Society for Blood and Marrow Transplantation 1083-8791/10/163-0016$36.00/0 doi:10.1016/j.bbmt.2009.10.027

430

CD41CD251 regulatory T cells (Treg) have recently emerged as key players in the regulation of immune responses, notably for preventing autoimmune diseases [1]. Quantitative and/or qualitative Treg defects have been reported in various inflammatory/autoimmune diseases since 2004 [2,3], and there is strong evidence that these cells also play a major role in the control of T cell alloreactivity, including in HSC transplantation. Although initial studies trying to correlate Treg numbers and the risk of GVHD brought controversial results [4], several authors have now reported a lower cumulative incidence of GVHD in patients receiving a PBSCT containing higher Treg numbers [5-7] or an inverse correlation between GVHD and Foxp3 gene expression in recipient peripheral blood mononuclear cells (PBMCs) [8]. This is consistent with experimental data in mice in which Treg depletion of HSC transplants exacerbates lethal GVHD, whereas Treg supplementation is protective [9,10]. Hence, ex vivo expanded Tregs have been proposed as a strategy to modulate GVHD after HSC transplantation [9,11,12] and the challenges and opportunities of developing human Treg therapy recently discussed [13]. Despite the growing importance of this subset in T cell immunobiology, there has been no systematic study to date comparing the Treg contents of the main 2 types of HSCTs, that is, BM and PBSC. We report herein that Treg frequencies are significantly

Biol Blood Marrow Transplant 16:430-434, 2010

Low Treg Frequency in PBSC Transplants

431

reduced in PBSC compared to BM transplants (BMT), and that most Tregs from PBSCTs are CD62Llo. MATERIAL AND METHODS Biologic Samples BMTs (n 5 29) or PBSCTs (n 5 55) were harvested consecutively by Etablissement Français du Sang de Normandie (Bois-Guillaume, France) from healthy donors who gave informed consent. All biologic samples used in this study were fresh (not frozen). Flow Cytometry FITC-, PC5-, PE-, ECD-, RD1-, or APC-Cy5conjugated monoclonal antibodies (mAbs) directed to CD3, CD4, CD8, CD25, CD19, HLA-DR, CD2, CD62L, CD45RO, CD45RA, and CD127, or FITC-labeled Annexin V were from Beckman Coulter (Fullerton, CA), BD PharMingen (San Diego, CA), or eBioscience (San Diego, CA). Intracellular FoxP3 staining was performed following the manufacturer’s protocol (eBioscience). Flow cytometry analyses were performed on an EPICS XL-MCL (Beckmann Coulter) using the System II software, except Annexin V analyses performed on a FACS Canto (Becton Dickinson, Franklin Lakes, NJ) using the FlowJo software (Tree Star, Ashland, OR). Statistical Analysis The Mann-Whitney test was used to compare data (Prism 4.00 software, Graph Pad, San Diego, CA), unless otherwise mentioned. RESULTS AND DISCUSSION Tregs, defined as CD41CD25hi, CD41FoxP31 or CD41CD251CD127lo can be readily detected in HSCTs (Figure 1A). When comparing BM and PBSC, Treg frequencies appeared significantly lower in the latter compared to the former (Figure 1B). For instance, the median proportion of CD25hi cells among transplants’ CD41 T cells dropped from 7.2% to 4.7% (P\.001). The frequency of PBSC CD41CD25hi cells correlated with that of CD41FoxP31 cells (P \ .0001, not shown). The reduction was also highly statistically significant when Tregs were identified as CD251 CD127lo cells because their median frequency decreased from 9.4% to 6.8% (P \ .001). The frequency of Tregs within HSCTs has a major impact on the risk of GVHD. In mice, manipulating the ratio of Tregs to conventional T cells (Tconv) by Treg depletion or addition significantly accelerates or reduces GVHD mortality, respectively [9]. Here, the ratio of Tregs to CD41CD25- Tconv dropped by

Figure 1. Reduced Treg frequencies in PBSC compared to BM transplants. (A) Flow cytometry analysis of a representative PBSC transplant. Cells were stained with RD1 anti-CD25 and ECD anti-CD4 (left), PE anti-FoxP3 and ECD anti-CD4 (intracellular staining, center), or RD1 anti-CD25 and FITC anti-CD127 (right). Data are gated on MNC (left and center) or CD41 T cells (right). Numbers indicate cell frequency in the depicted region or quadrant. (B) Treg frequencies in BM (n 5 29) and PBSC (n 5 55) transplants. Horizontal trait indicates median, box represents 25th and 75th percentiles, error bars depict 10th and 90th percentiles, and dots are individual values below or above the 10th and 90th percentiles, respectively. (C) Total CD31 T cell numbers in BM and PBSC transplants: (mean + SEM). (D) Box plots of the frequency of Annexin-V1 cells within CD41CD251CD127lo (Tregs) and CD41CD252 (Tconv) subsets, in the blood of donors after G-CSF administration (before leukapheresis) and in the PBSC transplant. Horizontal trait indicates median, box represents 25th and 75th percentiles, and error bars depicts range of values. Significant P-values #.05, #.01, and #.001 are indicated by *, **, and ***, respectively.

nearly 50% between BM and PBSC transplants (from an average ratio of 0.17 6 0.03 to 0.09 6 0.01, respectively). This Treg impoverishment may be an important

432

C. Blache et al.

factor governing the overall alloreactivity of the graft inasmuch as the total numbers of T cells infused is dramatically higher when using a PBSCT rather than a BMT (Figure 1C). One explanation for the lower PBSC Treg frequencies could be a preferential sensitivity to apoptosis. In support of this hypothesis, AnnexinV labeling was significantly higher in Tregs. This was not only the case when comparing Treg and Tconv within PBSC samples, but also before leukapheresis when analyzing donor blood after G-CSF administration (Figure 1D). It has been suggested that CD62L1 Tregs may have higher immunosuppressive properties in vitro than CD62L- Tregs [14], and that only the former might possess the capacity to control GVHD [15,16]. Nevertheless, this point remains controversial because the 2 Treg subpopulations seems equally capable of preventing gastritis and colitis despite possible different migration pattern [14]. When investigating the CD62L phenotype of Tregs, we found BM Tregs indistinguishable from donor blood Tregs on that basis (Figure 2A). One of the striking findings of this study was a dramatic loss of CD62L1 Tregs within PBSCTs. Indeed, whereas the vast majority of BM Tregs were CD62L1 (mean 6 SEM, 88.3% 6 1.5%), there was a 2-fold reduction in the frequency of CD62L1 cells among Tregs when turning to PBSCTs (45.8% 6 2.9 %). Tregs can be separated into a major CD45RO1 memory-type and minor CD45RA1 naive-type subset (Figure 2A), in accordance with previous observations [17]. The reduction in PBSC CD62L1 Tregs cannot be ascribed to a selective loss of naive-type CD45RA1 Tregs, because their frequency was not different between donor blood and PBSC samples (Figure 2A). There was even a slight increase in the frequency of naive-type Tregs in BM, but this subtype remained in minor proportion compared to memory-type Tregs (Figure 2A). Hence, Treg frequencies are diminished within PBSCTs compared to BMTs, and half of these Tregs display a CD62L2 phenotype. The loss of CD62L on T cells may possibly be due to the effect of G-CSF administration and/or to the process of leukapheresis. To further investigate this point, we performed a longitudinal analysis of CD62L expression on Tregs (1) in donor blood before G-CSF administration, (2) in donor blood after G-CSF administration but before leukapheresis, and (3) in the final PBSC product. The frequency of CD62L1CD25hi cells among CD41 T cells significantly decreased from 4.7% 6 0.4% to 3.9% 6 0.5% upon the administration of G-CSF (Figure 2B, P \ .05), consistent with a previous report showing that G-CSF in vivo may induce loss of CD62L on total CD31 cells [18]. Interestingly, loss of CD62L on Tregs could not be ascribed to the sole effect of G-CSF because leukapheresis itself further aggravated this phenomenon by reducing the CD62L1CD25hi


Figure 2. Decrease of CD62L1 Tregs during collect of PBSC transplants. (A) Flow cytometry analysis of Tregs from PBSC donor blood samples (before G-CSF, n 5 19) or from BM (n 5 26) or PBSC (n 5 53) transplants. Histograms (mean 1 SEM) indicate the frequency of CD62L expression on Tregs, the proportion of CD62L1 Treg and the frequency of several phenotypic markers in CD41CD25hi Tregs. Stainings were performed using RD1 anti-CD25, ECD anti-CD4 and 1 of the following FITC-labeled anti-CD62L, anti-CD45RO, anti-CD45RA, or anti-HLA-DR antibodies. (B) Flow cytometry analysis of CD62L expression at the surface of CD41CD25hi Tregs. Box plots indicate the percentage of CD25hiCD62L1 cells within the CD41 T cell gate (n 5 11), in the blood of donors before G-CSF administration, after G-CSF administration (before leukapheresis), and in the PBSC transplant. Horizontal trait indicates median, box represents 25th and 75th percentiles, and error bars depicts range of values. Data were compared by ANOVA. Significant P-values #.01 and #.001 are indicated by ** and ***, respectively.


cell frequency from 3.9% 6 0.5% to 2.2% 6 0.4% (Figure 2B, P \ .01). The mechanisms by which leukapheresis may influence CD62L shedding on Tregs may deserve further investigation. Yet, it has been reported that intracellular molecules released by cellular stress or cell death, such as adenosine triphosphate (ATP), may act on surface receptors notably the purinergic receptor P2X7 to induce shedding of CD62L on lymphocytes and even T cell death in mice and humans [19,20]. Because leukapheresis is a cause of cellular stress [21] and Tregs are more sensitive to ATP than Tconv [22], it is possible that this mechanism contributes to the low CD62L1 Treg content of PBSC transplants. Together, the results of this study show that Treg frequencies are lower in PBSCTs than in BMTs and that Tregs from PBSCTs are mostly CD62L2. Different reports have shown that G-CSF treatment may be associated with alterations in donor lymphocyte function and/or induction of potentially GVHD-protective lymphocytes [23-27]. From this biologic observation, it could be expected that PBSCT is endowed with a low risk of GVHD. However, several clinical studies including meta-analyses have shown that, for yet poorly understood reasons, the probability of developing chronic GVHD (cGVHD) is actually significantly increased in patients receiving PBSCTs compared to BMTs [28-30]. The present study was designed to compare the Treg content of both types of transplants but not to analyze the risk of GVHD. In addition to already published data [5-8], it would be interesting to further evaluate the correlation between GVHD occurrence and the type of graft (BMT versus PBSCT) with regard to the graft Treg content. One of the main factors that may augment the risk of GVHD after PBSCT compared to BMT is the considerably higher (10-fold) number of administered CD31 T cells in PBSCT. The reduced Treg content and possibly the CD62L2 phenotype of PBSCTs described herein may represent additional factors contributing to this higher risk of GVHD. ACKNOWLEDGMENTS The authors thank Sebastien Calbo, Florence Divay, Sophie Noir, Katia Nunes, Arnaud Roucheux, Michel Seman, and Sahil Adriouch for their help during the study. AUTHORSHIP STATEMENT Celine Blache performed research and wrote the paper. Joe-Marc Chauvin, Aude Marie-Cardine, Nathalie Contentin, Pascal Pommier, Ingrid Dedreux, Stephanie François, and Dominique Bastit collected samples and/or performed research. Serge Jacquot

Low Treg Frequency in PBSC Transplants

433

wrote the paper. Olivier Boyer designed research and wrote the paper. Financial disclosure: The authors have nothing to disclose.

REFERENCES 1. Sakaguchi S, Yamaguchi T, Nomura T, Ono M. Regulatory T cells and immune tolerance. Cell. 2008;133:775-787. 2. Boyer O, Saadoun D, Abriol J, et al. CD4 1 CD251 regulatory T-cell deficiency in patients with hepatitis C-mixed cryoglobulinemia vasculitis. Blood. 2004;103:3428-3430. 3. Viglietta V, Baecher-Allan C, Weiner HL, Hafler DA. Loss of functional suppression by CD4 1 CD251 regulatory T cells in patients with multiple sclerosis. J Exp Med. 2004;199:971-979. 4. Cohen JL, Boyer O. The role of CD4 1 CD25hi regulatory T cells in the physiopathogeny of graft-versus-host disease. Curr Opin Immunol. 2006;18:580-585. 5. Rezvani K, Mielke S, Ahmadzadeh M, et al. High donor FOXP3-positive regulatory T-cell (Treg) content is associated with a low risk of GVHD following HLA-matched allogeneic SCT. Blood. 2006;108:1291-1297. 6. Pabst C, Schirutschke H, Ehninger G, Bornhauser M, Platzbecker U. The graft content of donor T cells expressing gamma delta TCR1 and CD41foxp31 predicts the risk of acute graft versus host disease after transplantation of allogeneic peripheral blood stem cells from unrelated donors. Clin Cancer Res. 2007;13:2916-2922. 7. Wolf D, Wolf AM, Fong D, et al. Regulatory T-cells in the graft and the risk of acute graft-versus-host disease after allogeneic stem cell transplantation. Transplantation. 2007;83:1107-1113. 8. Miura Y, Thoburn CJ, Bright EC, et al. Association of Foxp3 regulatory gene expression with graft-versus-host disease. Blood. 2004;104:2187-2193. 9. Cohen JL, Trenado A, Vasey D, Klatzmann D, Salomon BL. CD4(1)CD25(1) immunoregulatory T Cells: new therapeutics for graft-versus-host disease. J Exp Med. 2002;196:401-406. 10. Hoffmann P, Ermann J, Edinger M, Fathman CG, Strober S. Donor-type CD4(1)CD25(1) regulatory T cells suppress lethal acute graft-versus-host disease after allogeneic bone marrow transplantation. J Exp Med. 2002;196:389-399. 11. Taylor PA, Lees CJ, Blazar BR. The infusion of ex vivo activated and expanded CD4(1)CD25(1) immune regulatory cells inhibits graft-versus-host disease lethality. Blood. 2002;99:3493-3499. 12. Trenado A, Charlotte F, Fisson S, et al. Recipient-type specific CD4 1 CD251 regulatory T cells favor immune reconstitution and control graft-versus-host disease while maintaining graftversus-leukemia. J Clin Invest. 2003;112:1688-1696. 13. Riley JL, June CH, Blazar BR. Human T regulatory cell therapy: take a billion or so and call me in the morning. Immunity. 2009; 30:656-665. 14. Fu S, Yopp AC, Mao X, et al. CD4 1 CD25 1 CD62 1 T-regulatory cell subset has optimal suppressive and proliferative potential. Am J Transplant. 2004;4:65-78. 15. Taylor PA, Panoskaltsis-Mortari A, Swedin JM, et al. L-Selectin(hi) but not the L-selectin(lo) CD4 1 25 1 T-regulatory cells are potent inhibitors of GVHD and BM graft rejection. Blood. 2004;104:3804-3812. 16. Ermann J, Hoffmann P, Edinger M, et al. Only the CD62L1 subpopulation of CD4 1 CD251 regulatory T cells protects from lethal acute GVHD. Blood. 2005;105:2220-2226. 17. Seddiki N, Santner-Nanan B, Tangye SG, et al. Persistence of naive CD45RA1 regulatory T cells in adult life. Blood. 2006; 107:2830-2838. 18. Sugimori N, Nakao S, Yachie A, et al. Administration of G-CSF to normal individuals diminishes L-selectin1 T cells in the peripheral blood that respond better to alloantigen stimulation than L-selectin- T cells. Bone Marrow Transplant. 1999;23: 119-124.

434

C. Blache et al.

19. Gu B, Bendall LJ, Wiley JS. Adenosine triphosphate-induced shedding of CD23 and L-selectin (CD62L) from lymphocytes is mediated by the same receptor but different metalloproteases. Blood. 1998;92:946-951. 20. Seman M, Adriouch S, Scheuplein F, et al. NAD-induced T cell death: ADP-ribosylation of cell surface proteins by ART2 activates the cytolytic P2X7 purinoceptor. Immunity. 2003;19: 571-582. 21. Moir S, Donoghue ET, Pickeral OK, et al. Continuous flow leukapheresis induces expression of stress genes in lymphocytes: impact on microarray analyses. Blood. 2003;102:3852-3853. 22. Aswad F, Kawamura H, Dennert G. High sensitivity of CD4 1 CD251 regulatory T cells to extracellular metabolites nicotinamide adenine dinucleotide and ATP: a role for P2X7 receptors. J Immunol. 2005;175:3075-3083. 23. Tayebi H, Kuttler F, Saas P, et al. Effect of granulocyte colonystimulating factor mobilization on phenotypical and functional properties of immune cells. Exp Hematol. 2001;29:458-470. 24. Rutella S, Pierelli L, Bonanno G, et al. Role for granulocyte colony-stimulating factor in the generation of human T regulatory type 1 cells. Blood. 2002;100:2562-2571. 25. Vela-Ojeda J, Esparza MA, Reyes-Maldonado E, et al. CD4 1 CD251 lymphocyte and dendritic cell mobilization with intermediate doses of recombinant human granulocyte colony-


26.

27.

28.

29.

30.

stimulating factor in healthy donors. Stem Cells Dev. 2005;14: 310-316. Morris ES, MacDonald KP, Hill GR. Stem cell mobilization with G-CSF analogs: a rational approach to separate GVHD and GVL? Blood. 2006;107:3430-3435. Noel G, Bruniquel D, DeGuibert S, et al. Regulatory CD4 1 CD25hi T cells conserve their function and phenotype after granulocyte colony-stimulating factor treatment in human hematopoietic stem cell transplantation. Hum Immunol. 2008;69: 329-337. Cutler C, Giri S, Jeyapalan S, Paniagua D, Viswanathan A, Antin JH. Acute and chronic graft-versus-host disease after allogeneic peripheral-blood stem-cell and bone marrow transplantation: a meta-analysis. J Clin Oncol. 2001;19:3685-3691. Schmitz N, Eapen M, Horowitz MM, et al. Long-term outcome of patients given transplants of mobilized blood or bone marrow: a report from the International Bone Marrow Transplant Registry and the European Group for Blood and Marrow Transplantation. Blood. 2006;108:4288-4290. Stem-Cell-Trialists’-Collaborative-Group. Allogeneic peripheral blood stem-cell compared with bone marrow transplantation in the management of hematologic malignancies: an individual patient data meta-analysis of nine randomized trials. J Clin Oncol. 2005;23:5074-5087.