Introduction ORIGINAL ARTICLE. J.H.F. Falkenburg 7 S.A.P. van Luxemburg-Heijs F.T.H. Lim 7 H.H.H. Kanhai 7 R. Willemze

Transcription

1 Ann Hematol (1996) 72: Q Springer-Verlag 1996 ORIGINAL ARTICLE J.H.F. Falkenburg 7 S.A.P. van Luxemburg-Heijs F.T.H. Lim 7 H.H.H. Kanhai 7 R. Willemze Umbilical cord blood contains normal frequencies of cytotoxic T-lymphocyte precursors (ctlp) and helper T-lymphocyte precursors against noninherited maternal antigens and noninherited paternal antigens Received: 29 November 1995 / Accepted: 4 December 1995 Abstract Umbilical cord blood (UCB) has been successfully used as an alternative source of hematopoietic stem cells for allogeneic transplantation. A relatively low incidence and severity of graft-versus-host disease (GVHD) following UCB transplants has been reported, and it has been suggested that this may be caused by a low frequency of alloreactive lymphocytes in UCB. Low frequencies of alloreactive T lymphocytes in UCB may allow transplantation across major MHC barriers with an acceptable risk of GVHD. We investigated cytotoxic T-lymphocyte precursor (CTLp) and helper T-lymphocyte precursor (HTLp) frequencies in UCB. Normal frequencies of CTLp and HTLp were measured against 1 2 HLA class I and 0 1 HLA-DR mismatched stimulator cells. Since it has been postulated that due to maternal fetal transfusion during pregnancy, fetal blood lymphocytes may become tolerant for noninherited maternal antigens (NIMA), allowing transplantation over certain HLA barriers, reactivity of 24 umbilical cord blood samples was analyzed against both parents. The median frequencies of CTLp against NIMA with 1 class-i mismatch was 79 per 10 6 nucleated cells (range ) and with 2 class I mismatches 121 (range ). CTLp frequencies against noninherited paternal antigens (NIPA) were not statistically different from those against NIMA, with a median of 115 (range 8 336) and 176 (range ) for 1 or 2 HLA class I mismatches, respectively. HTLp frequencies in UCB against parents with 0 or 1 HLA-DR antigen mismatches were similar with respect to NIMA (median J.H.F. Falkenburg (Y) 7 S.A.P. van Luxemburg-Heijs F.T.H. Lim 7 H.H.H. Kanhai 7 R. Willemze Leiden University Hospital, Department of Hematology, Building 1: C2-R, P.O. Box 9600, NL-2300 RC Leiden, The Netherlands This study was supported by grants from the J.A. Cohen Institute for Radiopathology and Radiation Protection and from the Praeventiefonds 74, range and 88, range , respectively) and NIPA (median 125, range , and 110, range , respectively). In four cases, UCB from two HLA-identical siblings was tested against both parents. A correlation between the frequencies of CTLp and HTLp from HLA-identical individuals was found, illustrating that these frequencies are genetically determined. These results illustrate that UCB contains normal frequencies of CTLp and HTLp against MHC alloantigens. Key words CTLp 7 HTLp 7 NIMA 7 NIPA 7 Cord blood transplantation Introduction Umbilical cord blood (UCB) can be used as an alternative source of bone marrow-repopulating hematopoietic stem cells for allogeneic transplantation [1]. Sustained donor engraftment has been demonstrated after transplantation with HLA-matched sibling UCB and with partially matched unrelated UCB [2 5]. Although almost all UCB transplants have been performed in children, in whom the incidence of graft-versus-host disease (GVHD) after bone marrow transplantation is generally low, the incidence and severity of GVHD after UCB transplantation appeared to be even less [3, 5]. Successful transplants in children and young adults have been performed using UCB with up to three HLA loci mismatched [5]. It has been reported that in unrelated bone marrow transplantation in particular the frequencies of anti-recipient cytotoxic T-lymphocyte precursor cells (CTLp) and helper T-lymphocyte precursor cells (HTLp) from the bone marrow donor graft correlate with the risk of developing GVHD [6 11], and it has been postulated that UCB may contain lower frequencies of alloreactive CTLp and HTLp. Whereas some authors found low or even absent alloreactive T- cell activity in UCB [12 14], others have reported normal CTLp or HTLp frequencies [15, 16]. It has been

2 261 shown that before birth maternal cells may enter the fetal circulation, as illustrated by the presence of small numbers of maternal cells in UCB [17, 18]. Based on these observations it has been hypothesized that fetal lymphocytes may have become tolerant to noninherited maternal antigens (NIMA) expressed on cells to which the fetus has been exposed in utero [19, 20]. This presumed absence of alloreactivity to NIMA might result in a decreased incidence of GVHD in recipients expressing HLA antigens identical to these NIMA. In this study we analyzed the CTLp and HTLp frequencies in 24 UCB samples against NIMA and NIPA. We demonstrate that normal frequencies of CTLp and HTLp are present in UCB, and that these frequencies are similar when tested against NIMA and NIPA. Materials and methods After informed consent had been obtained from both parents, UCB was harvested by puncture of the umbilical vein after vaginal deliveries of healthy infants. UCB was collected in Hanks balanced salt solution containing preservative-free heparin [21] and cryopreserved in 10% dimethylsulfoxide (DMSO) in liquid nitrogen after computer-controlled freezing. Peripheral blood (PBL) was harvested from both parents, or from unrelated healthy blood donors, using preservative-free heparin as anticoagulant. These PBL samples were centrifuged over Ficoll-Isopaque (density g/ml, g, 20 min). The mononuclear cells (MNC) were harvested from the interphase and cryopreserved similar to UCB in medium-containing 10% DMSO in liquid nitrogen. Immediately prior to use, UCB samples were thawed and washed in RPMIc20% FBS at 0 7C; this was followed by red blood cell lysis of the residual erythrocytes using NH 4 Cl as reported previously [21, 22]. PBL-MNC were thawed similarly, and washed in RPMIc20% FBS. After washing, all cell suspensions were resuspended in RPMI containing 15% prescreened human AB serum from healthy blood donors. HLA typing HLA typing of HLA A, B, C, DR, and DQ was performed using serologic typing and, if necessary, by polymerase chain reaction with sequence-specific oligonucleotides (PCR/SSO). Flow cytometric analysis Phenotypic analysis of UCB was performed by two-color flowcytometry using a Becton Dickinson FACScan (BD, Mountain View, Calif., USA), using monoclonal antibodies against CD2, CD3, CD4, CD8, CD14, CD19, CD45 and CD56 (all obtained from BD). CTLp and HTLp frequencies CTLp and HTLp frequencies were determined using a limitingdilution assay (LDA) as described elsewhere [23 26]. Briefly, graded numbers (from cells/well down to 625 cells/well in twofold dilutions) of UCB nucleated cells as responder cells were cultured in 24 replicates with irradiated (30 Gy) stimulator PBL-MNC in a total volume of 0.1 ml RMPI supplemented with 15% human serum. After 3 days of culture, 80 ml of supernatant was harvested from each well using a pipetting robot (Biomek, Beckman Instruments, Mijdrecht, The Netherlands), and fresh medium was added to each well, containing interleukin-2 (IL-2, Roussel Uclaf, Paris, France) at a final concentration of 20 IU/ml. After an additional 7 days of culture, split well analysis was performed using the pipetting robot, and each well was individually tested for cytolytic activity against europium-labeled target cells [23]. The LDA cultures were incubated for 4 h with phytohemagglutinin-stimulated lymphoblasts (PHA blasts) from the original stimulator cell donor as well as with the autologous PHA blasts. Each individual well was considered cytolytic if the observed europium release exceeded the spontaneous europium release from the 24 replicates containing only stimulator cells and target cells plus 3b the standard deviation. Frequencies of CTLp were calculated using the statistical program described by Strijbosch et al [27]. Autologous reactivity was less than 10% of reactivity against allogeneic targets. HTLp frequencies were determined by analysis of the production of IL-2 into the supernatant from each individual well using the IL-2-dependent murine CTLL-2 cell line [25, 26]. A quantity of 80 ml from each well was added to CTLL-2 cells deprived of IL-2 for 24 h and plated at a concentration of cells/well. After 72 h of incubation with the culture supernatant from each individual well, LSQM medium containing Triton X-100 (Fluka, Buchs, Switzerland) ink (Leitz, Wetzlar, Germany) and propidium iodide (Sigma, St. Louis, Mo., USA) in EDTA buffer was added. Using an automated fluorescence microscope (Leica-Patimed, Wetzlar, Germany), the number of propidium iodidestained nucleated cells was determined as a measurement of proliferation of CTLL-2 cells. Frequencies of HTLp were calculated similar to the CTLp frequencies. Statistical analysis of differences between reactivities against NIMA and NIPA was performed using a Wilcoxon matched-pairs signed-ranks test. Results Table 1 illustrates the distribution of lymphocyte subsets expressed as percentage of the CD45 c CD14 cells in the lymphocyte forward-sideward scatter profile. As reported previously, the percentage of CD3 c cells was lower in UCB than in adult peripheral blood [22, 28]. No significant differences were observed in the frequencies of lymphocyte subsets from both parents used as stimulator cells. Also, B cells and monocytes highly expressing class-ii molecules, and thereby relevant as antigen-presenting cells for T-helper cell responses, were similar in the stimulator cell populations: B cells (CD19) were 8B3% and 11B5%, and monocytes defined as CD45 c CD14 c cells with the characteristic Table 1 Lymphocyte subsets in UCB and peripheral blood from mother and father Phenotype UCB (%) Mother (%) Father (%) CD2 71B 9 a 85B4 78B10 a CD3 54B13 77B6 71B 8 CD3 c CD4 c 36B 9 51B7 42B10 CD3 c CD8 c 14B 8 22B9 22B 8 CD3 P CD8 c 8B 5 3B3 5B 3 CD56 16B 9 12B8 17B 6 CD2 c /CD56 c 10B 7 10B7 13B 5 CD2 P /CD56 c 6B 4 2B2 4B 3 CD19 17B 6 8B3 11B 5 a Percentage of lymphocytes defined as CD45 c CD14 P cells within the lymphocyte scatter profile (meansbs.d.)

3 262 Fig. 1 CTLp frequencies against NIMA and NIPA. CTLp frequencies expressed per 10 6 nucleated cells from umbilical cord (mater) and father (pater). CTLp frequencies against parental cells expressing one HLA-A or -B locus mismatch are presented in panel A, against two HLA-A or -B locus mismatches in panel B Fig. 2 CTLp frequencies against NIMA and NIPA. CTLp frequencies expressed per 10 6 CD3 c T cells from umbilical cord (mater) and father (pater). CTLp frequencies against parental cells expressing one HLA-A, or -B locus mismatch are presented in panel A, against two HLA-A or -B locus mismatches in panel B scatter profile were 11B6% and 12B5% in the maternal or paternal stimulator cell suspensions, respectively. CTLp and HTLp frequencies were determined in unseparated UCB after red blood cell lysis. As illustrated in Fig. 1, normal frequencies of CTLp were found in UCB directed against 1 or 2 HLA class-i (HLA-A or B) mismatched antigens [23 26]. The median frequencies of CTLp against NIMA were 79 per 10 6 nucleated cells (range ) for one HLA-A or -B mismatch and 121 per 10 6 nucleated cells (range ) for two HLA-A or -B mismatches. Similar CTLp frequencies were observed against NIPA, with a median frequency of 115 per 10 6 nucleated cells (range 8 336) for one class-i mismatch and a median frequency of 176 (range ) per 10 6 nucleated cells for two class-i mismatches, showing that no difference in cytotoxic potential against NIMA or NIPA was present in UCB. Since the frequency of CD3 c cells varied between individual UCB samples, CTLp frequencies were also expressed per 10 6 CD3 c cells. Again, no differences were observed between cytolytic activity against NIMA or NIPA (Fig. 2). Similarly, normal HTLp frequencies were observed in UCB, and again no differences were observed between the reactivity against 0 or 1 HLA- DR locus mismatches from maternal or paternal cells (Fig. 3). Median HTLp frequencies per 10 6 nucleated cells against NIMA were 74 (range ) and 88 (range ) for 0 or 1 HLA-DR mismatch and against NIPA 125 (range ) and 110 (range ) for 0 or 1 HLA-DR mismatches, respectively. Also, expressed as HTLp frequencies per CD3 positive T cells, similar frequencies were observed against NIMA and NIPA (Fig. 4). When CTLp and HTLp reactivities of all 24 individual UCB against maternal Fig. 3 HTLp frequencies against NIMA and NIPA. HTLp frequencies expressed per 10 6 nucleated cells from umbilical cord (mater) and father (pater). HTLp frequencies against parental cells expressing no HLA-DR locus mismatch are presented in panel A, against one HLA-DR locus mismatch in panel B cells as were compared with the reactivities against paternal cells, no statistical differences were found (pp0.13 for CTLp; pp0.85 for HTLp). In four cases, HLA-identical sibling UCB was obtained from two consecutive deliveries from the same mother. As illustrated in Fig. 5, a significant correlation was found between CTLp and HTLp frequencies from both siblings. These results illustrate the genetic background of the CTLp and HTLp frequencies found in UCB.

4 263 Fig. 4 HTLp frequencies against NIMA and NIPA. HTLp frequencies expressed per 10 6 CD3 c T cells from umbilical cord (mater) and father (pater). HTLp frequencies against parental cells expressing no HLA-DR locus mismatch are presented in panel A, against one HLA-DR locus mismatch in panel B Fig. 5 CTLp and HTLp frequencies in UCB from HLA-identical siblings as against parental cells. CTLp and HTLp frequencies were determined in four pairs of UCB samples from two consecutive donations from the same mother. A correlation was found between the frequency of CTLp and HTLp in the two siblings tested Discussion Although bone marrow transplants in childhood between HLA-matched siblings carry a low risk of GVHD after transplantation, hematopoietic stem cell transplantations using UCB from HLA-identical siblings have been associated with an even lower incidence and severity of GVHD [3]. In addition, preliminary experience with UCB from partially HLAmatched unrelated donors suggests that UCB transplants across HLA barriers can be performed with an acceptable morbidity and low mortality due to GVHD [5]. Although this is promising, larger prospective studies are required to allow statistical evaluation of the hypothesis that UCB transplants are associated with less GVHD. It has been suggested that this low incidence of GVHD may be caused by the inability of UCB to react with allogeneic targets, in particular maternal cells. In this study, we demonstrated that human UCB contains normal frequencies of allo-reactive CTLp and HTLp. Furthermore, no evidence of differences in reactivity against NIMA as compared with NIPA was found. Therefore, low frequencies of CTLp and HTLp are not likely to explain a presumed low frequency and severity of GVHD after UCB transplants. The correlation of CTLp and HTLp frequencies between two consecutive HLA-identical brothers or sisters, as illustrated in Fig. 5, suggests that these frequencies are genetically determined and are not greatly influenced by consecutive pregnancies. Low or absent alloreactivity of UCB against HLAmismatched individuals have been reported [12 14], whereas other investigators found normal or high alloreactivity. In part, this discrepancy may be due to experimental conditions, since a high NK reactivity is present in UCB. If, for instance, UCB cells have been stimulated by tumor cells, or if a relatively high concentration of IL-2 was added at the onset of the cultures, high NK activity may have been induced in UCB cells, leading to high CTLp activity against autologous cells which cannot be discriminated from the reactivity against allogeneic cells. Although our results show that there appeared to be no inability of UCB lymphocytes to react with allogeneic cells, other mechanisms may be hypothesized to explain a possible low incidence and severity of GVHD after UCB transplants. Several investigators have illustrated that T cells from UCB are of the naive (CD45RA) phenotype [28 30]. Therefore, T cell responses against recipient-associated antigens are primary T-cell responses, and may be less vigorous. In addition, phenotype analysis of UCB T cells have shown that these T cells appear to have a more immature phenotype [30, 31]. Furthermore, it has been suggested that a suppressor activity may be present in UCB, leading to a more protected immune response against recipient cells [32]. However, since it is not evident that these assays correlate with the occurrence or severity of GVHD after transplantation, they are at present of low predictive value. Experiments studying the kinetics of the immune response are required to determine whether an explanation can be found for a possible low incidence of GVHD after UCB transplants. Acknowledgements The authors thank Dr. E.A. van der Velde for statistical analysis and Clary Labee for secretarial assistance.

5 264 References 1. Broxmeyer HE, Douglas GW, Hangoc G, Cooper S, Bard J, English D, Arny M, Thomas L, Boyse EA (1989) Human umbilical cord blood as potential source of transplantable hematopoietic stem/progenitor cells. Proc Natl Acad Sci USA 86: Gluckman E, Broxmeyer HE, Auerbach AD, Friedman HS, Douglas GW, Devergie A, Esperou H, Thierry D, Socie G, Lehn P et al (1989) Hematopoietic reconstitution in a patient with Fanconi s anemia by means of umbilical cord blood from an HLA-identical sibling. N Engl J Med 321: Wagner JE, Kernan NA, Steinbuch M, Broxmeyer HE, Gluckman E (1995) Allogeneic sibling umbilical cord blood transplantation in children with malignant and non-malignant disease. Lancet 346: Rubinstein P, Rosenfield RE, Adamson JW, Stevens CE (1993) Stored placental blood for unrelated bone marrow reconstitution. Blood 81: Kurtzberg J, Laughlin M, Casey J, Miller S, Olsen J, Ciocci G, Graham ML, Stevens C, Rubinstein P (1995) Umbilical cord blood: an alternative source of hematopoietic stem cells for bone marrow reconstitution in unrelated donor transplantation. Exp Hematol 23 : 942 (Abstract) 6. Kaminski E, Hows J, Man S, Brookes P, Mackinnon S, Hughes T, Avakian O, Goldman JM, Batchelor JR (1989) Prediction of graft-versus-host disease by frequency analysis of cytotoxic T cells after unrelated donor bone marrow transplantation. Transplantation 48 : Irschick EU, Hladik F, Niederwieser D, Nussbaumer W, Hooler E, Kaminski E, Huber C (1992) Studies on the mechanism of tolerance of graft-versus-host disease in allogeneic bone marrow recipients at the level of cytotoxic T-cell precursor frequencies. Blood 79: Roosnek E, Hogendijk S, Zawadynski S, Speiser D, Tiercy JM, Helg C, Chapuis B, Gratwohl A, Gmur J, Seger R, Jeannet M (1993) The frequency of pretransplant donor cytotoxic T-cell precursors with anti-host specificity predicts survival of patients transplanted with bone marrow from donors other than HLA-identical siblings. Transplantation 56 : Theobald M, Nierle T, Bunjes D, Arnold R, Heimpel H (1992) Host-specific interleukin-2-secreting donor T-cell precursors as predictors of acute graft-versus-host disease in bone marrow transplantation between HLA-identical siblings. N Engl J Med 327: Schwarer AP, Jiang YZ, Deacock S, Brookes PA, Barrett AJ, Goldman JM, Batchelor JR, Lechler RI (1994) Comparison of helper and cytotoxic antirecipient T cell frequencies in unrelated bone marrow transplantation. Transplantation 58: Spencer A, Brookes PA, Kaminski E, Hows JM, Szydlo RM, van Rhee F, Goldman JM, Batchelor JR (1995) Cytotoxic T- lymphocyte precursor frequency analysis in bone marrow transplantation with volunteer unrelated donors. Value in donor selection. Transplantation 59: Risdon G, Gaddy J, Stehman FB, Broxmeyer HE (1994) Proliferative and cytotoxic responses of human cord blood T lymphocytes following allogeneic stimulation. Cell Immunol 154: Harris DT (1994) Cord blood transplantation: implications for graft vs host disease and graft vs. leukemia. Blood Cells 20: Harris DT, Locascio J, Besencon FJ (1994) Analysis of the alloreactive capacity of human umbilical cord blood: implications for graft-versus-host disease. Bone Marrow Transplant 14: Deacock SJ, Schwarer AP, Bridge J, Batchelor JR, Goldman JM, Lechler RI (1992) Evidence that umbilical cord blood contains a higher frequency of HLA class II specific alloreactive T cells than adult peripheral blood: a limiting dilution analysis. Transplantation 53 : Roncarolo MG, Bigler M, Ciuti E, Martino S, Tovo PA (1994) Immune responses by cord blood cells. Blood Cells 20: Socié G, Gluckman E, Carosella E, Brossard Y, Lafon C, Brison O (1994) Search for maternal cells in human umbilical cord blood by polymerase chain reaction by amplification of two minisatellite sequences. Blood 83: Hall J, Lingenfelter P, Adams S, Bean MA, Hansen J (1994) Detection of maternal T-cell in human umbilical cord blood using FISH. Blood 84: 97a 19. Claas FHJ, Gijbels Y, van der Velden-de Munch J, van Rood JJ (1988) Induction of B cell unresponsiveness to noninherited maternal HLA antigens during fetal life. Science 241: Van Rood JJ, Claas FHJ (1988) The influence of allogeneic cells on the human T and B cell repertoire. Science 248: Falkenburg JHF, van Luxemburg-Heijs SAP, Zijlmans JM, Fibbe WE, Kluin-Nelemans JC, Kanhai HHH, Willemze R (1993) Separation, enrichment, and characterization of human hematopoietic progenitor cells from umbilical cord blood. Ann Hematol 67: Lim FTH, van Winsen L, Willemze R, Kanhai HHH, Falkenburg JHF (1994) Influence of delivery on numbers of leukocytes, leukocyte subpopulations, and hematopoietic progenitor cells in human umbilical cord blood. Blood Cells 20: Bouma GJ, van der Meer-Prins PMW, Van Bree FPMJ, van Rood JJ, Claas FHJ (1992) Determination of cytotoxic T-lymphocyte precursor frequencies using europium labeling as a nonradioactive alternative to labeling with chromium-51. Hum Immunol 35: Deacock S, Schwarer A, Batchelor R, Goldman J, Lechler R (1992) A rapid limiting dilution assay for measuring frequencies of alloreactive, interleukin-2-producing T cells in humans. J Immunol Methods 147: Schanz U, Roelen DL, Bruning JW, Kadol MJ, Van Rood JJ, Claas FHJ (1994) The relative radioresistance of interleukin-2 production by human peripheral blood lymphocytes: consequences for the development of a new limiting dilution assay for the enumeration of helper T-lymphocyte precursor frequencies. J Immunol Methods 169: Bouma GJ, van der Meer-Prins PMW, van der Woude FJ, van Rood JJ, Claas FHJ (1995) Relevance of pretransplant donor-specific T cell allorepertoire for human kidney graft survival. Transplantation 59: Strybosch LWG, Buurman WA, Does RJM, Zinken PH, Groenewegen G (1987) Limiting dilution assays: experimental design and statistical analysis. J Immunol Methods 97: Beck R, Lam-Po-Tang PR (1994) Comparison of cord blood and adult blood lymphocyte normal ranges: a possible explanation for decreased severity of graft-versus-host disease after cord blood transplantation. Immunol Cell Biol 72: Racadot E, van Lemmens P (1993). Cord-blood lymphocyte populations. Immunol Today 14: Harris DT, Schumacher MJ, Locascio J, Besencon FJ, Olson GB, DeLuca D, Shenker L, Bard J, Boyse EA (1992) Phenotypic and functional immaturity of human umbilical cord blood T lymphocytes. Proc Natl Acad Sci USA 89: Han P, Hodge G, Story C, Xu X (1995) Phenotypic analysis of functional T lymphocyte subtypes and natural killer cells in human cord blood: relevance to umbilical cord blood transplantation. Br J Haematol 89: Risdon G, Gaddy J, Horie M, Broxmeyer HE (1995) Alloantigen priming induces a state of unresponsiveness in human umbilical cord blood T cells. Proc Natl Acad Sci USA 92: 2413