Biology of Blood and Marrow Transplantation
Volume 13, Issue 10 , Pages 1135-1144, October 2007

Umbilical Cord Blood Xenografts in Immunodeficient Mice Reveal That T Cells Enhance Hematopoietic Engraftment Beyond Overcoming Immune Barriers by Stimulating Stem Cell Differentiation

  • Elizabeth O. Hexner

      Affiliations

    • Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • Gwenn-aël H. Danet-Desnoyers

      Affiliations

    • Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • Yi Zhang

      Affiliations

    • Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • Dale M. Frank

      Affiliations

    • Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • James L. Riley

      Affiliations

    • Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania
    • Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • Bruce L. Levine

      Affiliations

    • Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
    • Abramson Family Cancer Research Institute, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • David L. Porter

      Affiliations

    • Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • Carl H. June

      Affiliations

    • Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
    • Abramson Family Cancer Research Institute, University of Pennsylvania, Philadelphia, Pennsylvania
    • ,
  • Stephen G. Emerson

      Affiliations

    • Division of Hematology/Oncology, University of Pennsylvania, Philadelphia, Pennsylvania
    • Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
    • Abramson Family Cancer Research Institute, University of Pennsylvania, Philadelphia, Pennsylvania
    • Corresponding Author InformationCorrespondence and reprint requests: Stephen G. Emerson, MD, PhD, Pathology and Pediatrics, 510 Maloney Building, 3400 Spruce Street, Philadelphia, PA 19104.

Received 6 April 2007; accepted 22 June 2007.

Article Outline

Abstract 

Clinical experience and animal models have shown that donor T cell depletion (TCD) adversely affects engraftment of hematopoietic stem cells (HSCs). Although it is known that donor T cells are acting to overcome residual host immune barriers, they may also exert effects independent of host resistance via direct or indirect interactions with donor stem cells, their microenvironment, or key differentiation events. To more precisely define the effect of T cells on engraftment, we have performed human umbilical cord blood (UCB) transplantation into immunodeficient mice under limiting dilution conditions. UCB mononuclear cells (MNC) or TCD UCB were transplanted into NOD/LtSz-scid/scid B2mnull (NOD/SCID-β2m−/−) mice. Cohorts of mice received UCB MNC or TCD UCB at 5 dose levels between 5 × 104 and 5 × 106 cells. At dose levels at or above 105 cells, engraftment was higher in the MNC recipients (n = 32) than the TCD recipients (n = 31) in a dose-dependent manner. Despite this difference, limiting dilution analysis to determine functional stem cell frequency revealed that SCID repopulating cells in TCD UCB was not significantly less than in CB MNCs, suggesting that T cells may facilitate engraftment at stages beyond the stem cell. Add-back of CD3/CD28 costimulated T cells restored and appeared to enhance engraftment, both in NOD/SCID-β2m−/− as well as NOD/LtSz-scid IL2Rγnull (NOG) recipients. These results, in a model where there are minimal host immune barriers to overcome, suggest T cells possess additional graft-facilitating properties. CD3/CD28 costimulation of UCB T cells represents a potential strategy for enhancing the engraftment of UCB.

Key Words: Engraftment, T cells, Umbilical cord blood

 

Back to Article Outline

Introduction 

An attractive property of umbilical cord blood (UCB) as a graft for hematopoietic stem cell transplantation (HSCT) is the reduced incidence of acute graft-versus-host disease (aGVHD) when compared to blood or marrow stem cell sources, despite the use of UCB grafts that are frequently mismatched at 1 or more major histocompatibility (MHC) locus. This feature of UCB is thought, in part, to be mediated by uniquely tolerant naive T cells in UCB grafts [1, 2]. Although this increased tolerance to MHC mismatching makes UCB an attractive graft phenotype, UCB is slow to engraft, representing a major limitation to the broader application of UCB HSCT in Western countries. The slow kinetics of engraftment are known to be profoundly influenced by the absolute numbers of mononuclear cells (MNCs) in UCB grafts, but even above acceptable threshold numbers of cells, recipients of UCB grafts remain neutropenic and thrombocytopenic significantly longer than recipients of other unrelated HSC sources [3, 4, 5]. Three possible explanations for the empiric evidence are: (1) there are not enough hematopoietic stem cells (HSCs) in UCB grafts (intrinsic, quantitative defect in HSCs); (2) UCB HSCs are intrinsically less able and/or more slow to repopulate the HSC niche (intrinsic, qualitative defect in HSCs); or (3) nonhematopoietic, accessory cells in the UCB graft are less able to facilitate the engraftment and/or differentiation of UCB HSCs (defect extrinsic to UCB HSCs). In the last case, 1 candidate “defective” graft facilitating accessory cell is the UCB T cell. That is, the unique properties of UCB T cells that render UCB a better tolerated donor graft may at the same time be responsible for the delayed engraftment of UCB HSCs. If true, graft engineering of T cells in UCB could accelerate UCB engraftment.

From an experimental perspective, immunodeficient mouse models provide a useful system to dissect the relative importance of these 2 mechanisms: because host barriers are absent, if donor T cells fail to amplify engraftment, this suggests that their primary role in HSCT is to overcome barriers. On the other hand, if T cells maintain a graft-facilitating effect in these immunodeficient mice, this implies that T cells actively facilitate grafts beyond defending against host immune cells. Mouse models with progressively lower immune barriers have been developed over the last several decades. Scid/scid mice lacking T and B cells were the first immunodeficient mice shown to support lymphoid development of adoptively transferred human fetal liver cells following the implantation of human fetal thymus (the so-called SCID-hu mouse), although levels of engraftment were low, because of in large part to the presence of natural killer (NK) cells [6, 7]. NOD/LtSz-scid/scid (NOD/SCID) mice have reduced numbers of NK cells and represented a significant advance for in vivo functional stem cell assays because they support myeloid and lymphoid development of human hematopoietic cells [8, 9]. The low level of residual NK activity in NOD/SCID mice can be augmented by pyrogens like poly(I:C), and might still represent a significant barrier to transplantation following sublethal irradiation [6]. Engraftment was enhanced approximately 10-fold when these mice were crossed with mice null for the β2 microglobulin allele. The NOD/SCID-β2m−/− lacks cell surface class I molecules and NK cells, even in response to poly(I:C) 6, 10. Although NOD/SCID-β2m−/− mice have monocytes and macrophages that may yet represent a minimal barrier to xenotransplantation, they represent an improved strain for supporting human xenografts, excluding antibody pretreated animals [11]. Recent derivation of the NOD/LtSz-scid IL2Rγnull (NOG) mouse has shown outstanding multilineage engraftment of human cells, and may be the best current murine xenograft model [12].

If the predominantly naïve T cells in UCB underlie the engraftment delay in UCB HSCT, it may be possible to activate UCB T cells to accelerate engraftment. CD3 and CD28 stimulatory antibodies immobilized on magnetic beads deliver the required costimulatory signals for T cell activation and expansion, and can result in the large-scale expansion of polyclonal CD4+ and CD8+ T cells, with a preserved T cell receptor (TCR) repertoire [13]. The feasibility and clinical safety of delayed infusion of these cells has now been demonstrated in the autologous as well as the allogeneic setting; studies using CD3/CD28 costimulated UCB T cells at the time of HSCT or as delayed infusions have not been reported to date, but represent a potential approach to improving engraftment and/or graft-versus-tumor activity [14, 15, 16].

In this study we sought to directly test the specific role of mature UCB T cells on engraftment of UCB independent of their T cell effects overcoming immune barriers. Using NOD/SCID-β2m−/− and NOG mice as recipients of human UCB grafts, we have found that grafts depleted of T cells (TCD) have inferior levels of engraftment when compared to UCB MNCs transplanted into NOD/SCID-β2m−/− mice. Although the overall level of engraftment was significantly lower for NOD/SCID-β2m−/− recipients of TCD UCB, SCID repopulating cell (SRC) frequency was not significantly decreased per se. Engraftment capacity was restored and enhanced by add-back of ex vivo CD3/CD28 costimulated T cells from the same UCB unit. Our results suggest that T cells facilitate engraftment beyond simply overcoming host barriers, and that CD3/CD28 costimulated UCB T cells possess graft-promoting properties, representing a potential translational approach for facilitating engraftment in UCB HSCT.

Back to Article Outline

Materials and Methods 

Animals 

NOD/LtSz-scid/scid B2mnull mice and NOD/LtSz-scid IL2Rγnull (Jackson Laboratory, Bar Harbor, ME) were maintained in an animal facility at the University of Pennsylvania with approval of the Institutional Animal Care and Use Committee. Transplanted animals were maintained in microisolators with water supplemented with polymixin B sulfate (1000 U/mL) and neomycin sulfate (1.1 mg/mL).

Cell Preparations 

UCB was obtained from placentas from full-term scheduled Cesarean section births at the Hospital of the University of Pennsylvania under an institutional review board exempted protocol. Mononuclear cells were isolated on a Ficoll-Paque PLUS (Amersham Biosciences, Arlington Heights, IL) density gradient and cryopreserved until the day of xenotransplantation. A single UCB unit was used for limiting dilution analysis. Two additional UCB units were used in 2 separate experiments to test the engraftment properties of CD3/CD28 costimulated T cells (3 groups: TCD UCB, UCB MNC, and TCD UCB + CD3/CD28 costimulated T cells). HLA types of the UCB units were not known. Activated T cells from a UCB unit was always infused with TCD UCB from the same unit; and UCB MNC or TCD UCB groups from the same experiment received cells from the same unit as well. Prior to transplantation, UCB MNCs were thawed; the TCD preparations were depleted of T cells via CD2 immunomagnetic selection (Dynabeads M-450 CD2, Dynal Biotech, Bath, UK) according to the manufacturer’s protocol. Efficiency of depletion was determined by Fluorescence Activated Cell Sorting (FACS) analysis prior to infusion. For all infused cells, cell number and viability was determined by trypan blue exclusion immediately prior to infusion; all infused cells were suspended in PBS (250 μL/mouse).

To prepare CD3/CD28 costimulated T cells, cells were isolated from an aliquot of freshly collected UCB prior to the cryopreservation of the UCB unit. CD4+ and CD8+ T cells were enriched separately from 2.5 mL of UCB (RosetteSep, StemCell Technologies, Canada) according to the manufacturer’s protocol. Cells were plated and maintained at a cell density of 1 million/mL in RPMI 1640 supplemented with 2 mM L-Glutamine, 25 mM HEPES, 100 U/mL Penicillin, 100 μg/mL Streptomycin (Life Technologies, Baltimore, MD), and 10% fetal bovine serum (Gemini Bio-Products, Calabasas, CA). For the first 72 hours, cells were cocultured with Dynabeads coated with anti-CD3 and anti-CD28 monoclonal antibodies (mAb) at a ratio of 3 beads per cell [13]. Beads were removed magnetically after 72 hours of culture, and on days 7-9 cells were washed twice in PBS prior to infusion. Confirmation that T cells were activated as first described by Levine et al. [13] using this culture protocol was confirmed by increased cell surface expression of CD25, CD28, and CD69; by cyclic changes in cell volume, and by intracellular cytokine staining for interferon-γ.

Xenotransplantation 

Eight- to 10-week-old mice were exposed to sublethal doses (300 cGy) of radiation 24 hours prior to transplantation. Cell suspensions were infused via tail vein into recipient mice.

Histopathologic Evaluation 

Skin, liver, spleen, intestine, and lymph nodes were sampled immediately prior to engraftment analysis. Specimens were coded so intervention group was unknown. Tissues were fixed overnight in 10% buffered formalin, dehydrated, embedded in paraffin, sectioned, and stained with hematoxylin and eosin according to standard protocols. Slides were evaluated and photographed by an experienced pathologist.

Data Collection and Statistical Analysis 

Xenografted mice were sacrificed 7 to 8 weeks following transplantation (for limiting dilution analysis) or between days 23 and 28 (for recipients of TCD UCB with or without CD3/CD28 costimulated T cells along with a parallel MNC group). Stable engraftment of HSC in these mice is thought to be achieved at 6 to 8 weeks, and therefore that was felt to be the most accurate time to evaluate HSC contribution in the limiting dilution analysis [17]. Subsequent experiments were designed to analyze level of early engraftment, which may represent HSC as well as multipotent (and later) progenitor cells—so earlier time points were chosen for analysis. Bone marrow from iliac crests, femurs, and tibiae were collected from each animal and incubated with ammonium chloride solution (StemCell Technologies) prior to cell surface staining. Cells were washed and incubated with anti-human CD45-APC, CD19-PE, CD33-PerCP-Cy5.5, and CD 3-FITC. The analysis protocol for cells from NOG mice was changed to also include erythroid cells, analyzed with CD45-APC, Glycophorin A-PE, and CD36-FITC (BD Pharmingen, San Diego, CA, and Biolegend, San Diego, CA). FACS was performed on a FACSCalibur (Beckton Dickinson, Fullerton, CA); 100,000 events per sample were acquired. FACS data were analyzed using FloJo (Tree Star, Ashland, OR); the live cell gate was determined by forward scatter/side scatter properties and mice were considered engrafted if at least 0.5% of cells in this gate were CD45+ and demonstrated both human myeloid and B lymphoid engraftment. Human erythroid cells were defined by double staining with CD36 and Glycophorin A. Human CD3 staining was performed to enumerate and control for any surviving adoptively transferred CD3+ cells.

Level of engraftment data are presented as means ± standard error of the mean. Differences between level of engraftment were analyzed using a 2-sided Student’s t test or 1-way ANOVA when there were more than 2 groups to analyze; analysis was performed using JMP software (SAS Institute, Cary, NC). SRC frequency was determined using the single-hit Poisson model; a χ2 test was applied to confirm that the data were consistent with the model [17, 18]. Data were analyzed using L-Calc software (StemCell Technologies).

Back to Article Outline

Results 

TCD of UCB Impairs Level of Engraftment 

To test whether TCD adversely effects engraftment in this xenograft model, thawed UCB MNCs were depleted of T lymphocytes by immunomagnetic separation of CD2+ cells. Evaluation of cells following separation showed that the efficiency of depletion was 92% to 98%. In multiple experiments, TCD UCB was consistently more enriched for CD34+ cells when compared to UCB MNCs (Figure 2), although the difference was not significant (mean UCB MNC 0.38% CD34+ versus TCD UCB 0.53% CD34+; P = .62 for 3 UCB units used in all experiments). Equal numbers of viable TCD UCB or UCB MNCs were transplanted into sublethally irradiated 8- to 10-week-old NOD/SCID-β2m−/− mice, at 6 dose levels between 5 × 104 cells and 5 × 106 cells per mouse. There was no difference in survival of animals between groups. Overall, mean engraftment in the MNC recipients was 4.70% ± 0.85% (range: 0-27.3%) versus 1.90% ± 0.87% in the recipients of TCD UCB (range: 0-10.6%) (P = .02; Figure 3A) despite the slight enrichment for CD34+ cells in TCD UCB. The level of engraftment increased in a dose-dependent manner, and linear regression analysis by the least squares method revealed that engraftment on a per cell basis (slope) was greater for UCB MNC than that for TCD UCB (Figure 3B). Human CD3+ cells were only detectable (>0.1%) in 1 mouse; this was a recipient of 5 × 106 MNCs, and was the animal with the highest overall level of engraftment (27.3% CD45+, 6% CD45+CD3+).

  • View full-size image.
  • Figure 2. 

    Infused UCB MNCs. CB MNCs were thawed, and for TCD recipients, depleted of CD2+ cells immunomagnetically prior to transplantation. Percentages of CD34+ and CD3+ cells in the infused product were evaluated by flow cytometry, and a representative flow cytogram is shown. TCD efficiency was >97% and CD34+ cells were consistently enriched in the TCD infused product compared to unmanipulated CB MNC.

  • View full-size image.
  • Figure 3. 

    Despite absent host barriers and more CD34+ cells, the level of engraftment of TCD UCB is inferior to UCB MNC. Percent human leukocyte engraftment was evaluated in BM from recipient mice between days +22 and +27 or +39 and +45 following transplantation. A, Overall mean engraftment as assessed by human CD45 in UCB MNC recipients: 4.69% ± 0.85% (n = 32, range: 0-27.3%) and TCD UCB: 1.90% ± 0.87% (range: 0-10.6%, n = 31, P = .02). Diamonds represent mean (center) and 95% confidence intervals (top and bottom) for each distribution. B, Scatter plot of percent human CD45 by dose of cells transplanted. Linear regression lines are shown for UCB MNC (solid line; linear prediction human CD45 = 2.47 + 2.45 × cell dose; R2: 0.54) and TCD MNC (interrupted line, linear prediction human CD45 = 0.56 + 1.25 × cell dose; R2: 0.56). Limiting dilution analysis to determine functional stem cell numbers was performed using a single unit of UCB at 5 dose levels between 5 × 104 and 5 × 106 cells (2 to 8 mice/group; total mice transplanted: 58, 54 survived to analysis) Animals with >0.5% human CD45+ cells, and >0.1% human lymphoid (CD 19+) and myeloid (CD33+) cells in the FSC/SSC gate were considered to be engrafted. The difference in SRC frequency was not significant (P = .20).

SCID-Repopulating Activity in TCD UCB Is Not Significantly Lower than UCB MNCs 

We also asked whether functional engrafting stem cells were reduced in TCD UCB by performing a limiting dilution analysis into recipients using a single UCB unit (Table 1). Five dose levels between 5 × 104 and 5 × 106 cells were tested (4 to 8 mice/group; total mice transplanted: 58, 54 survived to analysis). Mice were considered engrafted if flow cytometry of flushed bone marrow demonstrated more than 0.5% human CD45+ cells in the forward scatter/side scatter gate with evidence for myeloid (CD33+) and B lymphoid (CD19+) engraftment (at least 0.1%) [19]. SRC frequency was 1 in 3.03 × 105 for UCB MNCs (95% confidence interval [CI] 1 in 1.65 × 105 − 1 in 5.58 × 105), a figure that was only slightly higher and not statistically different from the SRC frequency of 1 in 5.39 × 105 measured in TCD UCB (95% CI 1 in 2.82 × 105 − 1 in 1.03 × 106; P = .20). Thus, although TCD clearly reduces the level of engraftment (as assessed by the percent contribution of human hematopoietic cells in the mice) at all HSC doses, the actual number of human HSCs contributing to hematopoiesis is largely unchanged.

Table 1. SCID Repopulating Cell (SRC) Frequency in UCB MNC Is Not Significantly Different from TCD UCB
Live Cell DoseUCB MNCTCD UCB
5 × 1040/3(0%)0/4(0%)
1 × 1054/8(50%)1/6(16.7%)
5 × 1057/8(87.5%)6/7(85.7%)
1 × 1065/6(83.3%)4/6(66.7%)
5 × 1062/2(100%)4/4(100%)
SRC frequency (95% CI)1 in 3.03 × 105 (1/1.65 × 105–1/5.58 × 105)1 in 5.39 × 105 (1/2.82 × 105–1/1.03 × 106)

CI indicates confidence interval; UCB, umbilical cord blood; MNC, mononuclear cells; TCD, T cell depleted.

CD3/CD28 Costimulated UCB T cells Restore Engraftment Capacity 

To retain cell surface CD3 for subsequent activation steps, TCD was performed by CD2 negative selection. CD2 is expressed on thymocytes, T, and NK cells, so it was possible that graft-facilitating activity resided in any of these cellular compartments. To confirm that graft facilitating properties were specific to mature T cells, and to test whether CD3/CD28 costimulated T cells could facilitate engraftment, we cotransplanted limiting numbers (8 × 105 cells) of TCD UCB cells with CD3/CD28 costimulated T cells from the same UCB unit. T cells comprised 30% of the infused product, with a CD4+ to CD8+ ratio of 3 to 1, consistent with the original ratio for the particular UCB unit as well as empiric published data [1]. T cells were isolated and cultured in the presence of anti-CD3 and anti-CD28 coated magnetic beads for 72 hours without exogenous cytokines, and then maintained in culture for 7 days. Cell surface analysis of expanded UCB T cells demonstrated that 95% of cells coexpressed CD3 and CD28; 93% of cells were CD25bright; 9% expressed CD69; 2% expressed CD16; cells were negative for CD14 and CD19 (Figure 4A, left panels, and data not shown). UCB T cells could also be expanded directly from fresh or thawed UCB MNCs with >92% CD3+ purity; cell surface phenotype was similar with the exception that more expanded cells from MNCs expressed low levels of CD11b (12% versus 3% on expanded isolated T cells; Figure 4A, right panel).

  • View full-size image.
  • Figure 4. 

    Coinfusion of CD3/CD28 activated UCB T cells with TCD UCB restores engraftment. A, Immunophenotype of CD3/CD28 costimulated UCB T cells on day 7 of culture. Although transplanted CD3/CD28 costimulated T cells were derived from CD4+- and CD8+-enriched cells, T cells could also be expanded from unpurified MNCs (fresh or thawed) with a comparable purity. Left and middle panels demonstrate >93% purity of T cells in terms of CD3 expression (62.9% CD4+, 28.8% CD8+, not shown). There was bright expression of CD25 and CD28 in the entire population; CD11b was more highly expressed on T cells derived from MNC (gray line) compared to T cells expanded from CD2+ selected cells (black line); immunophenotype was otherwise similar. The T cell product that was adoptively transferred into mice was expanded from CD4+ and CD8+ isolated cells. B, Engraftment of CD45+ cells minus adoptively transferred CD45+CD3+ cells. Lines represent means and standard error. C, Representative histopathologic examination (liver, 20×) of recipients of CD3/CD28 costimulated UCB T cells showed no evidence of inflammation or significant pathology of sampled organs (4 of 7 recipient tissues examined, 0 of 4 affected).

Expanded T cells were infused along with TCD UCB and engraftment was assessed between day 23 and 27. Given that NOD/SCID-β2m−/− are known to support the expansion of adoptively transferred human T cells, it was possible that differences in human CD45+ expression represented only expanded mature T cells [6]. To eliminate this potential confounder, we measured engraftment by enumerating total percentage of human CD45+ cells minus total percentage of human CD45+CD3+ cells. Engraftment was significantly higher in recipients of CD3/CD28 costimulated UCB T cells when compared to mice receiving equal numbers of TCD UCB cells (CD45−CD3: 5.94% ± 1.65% versus 0.29% ± 0.79%, P = .01), and remained so even when compared to pooled data from groups of mice receiving as many as 5 million TCD UCB cells and UCB MNCs (Figure 4B). Two mice receiving CD3/CD28 costimulated T cells were evaluated at a later time point (56 days); mean engraftment was not significantly different when compared to the mice evaluated at the earlier time point (CD45-CD3 = 6.82% versus 5.59%; P = .33).

Recipient mice of CD3/CD28 costimulated T cells appeared clinically healthy. We examined liver, spleen, skin, intestine, and remnant lymph nodes from 4 of the 7 recipients of CD3/CD28 costimulated T cells, and there was no evidence of inflammation, suggestive of xeno GVHD or other significant pathology in any of the sampled tissues (Figure 4C). It therefore appears that engraftment ability is restored by add-back of CD3/CD28 costimulated T cells, and that xenoGVHD is uncommon.

Facilitation of Early Multilineage Engraftment by CD3/CD28 Costimulated UCB T Cells Is Maintained in NOG Mice 

We next asked whether early engraftment, representing in large part multilineage progenitor cells in the graft, was specifically affected by supraphysiologic numbers of CD3/CD28 costimulated T cells in NOG mice. Mice received 1 million CB MNCs, TCD UCB cells, or TCD UCB cells with an additional 1 million CD3/CD28 costimulated T cells from the same UCB graft on day 9 of ex vivo culture. Lymphomyeloid as well as erythroid engraftment (human CD36+Glycophorin A+CD45− BM cells) was analyzed. Mean engraftment of leukocytes (CD45−CD3+) was again higher in the recipients of TCD UCB + CD3/CD28 costimulated T cells (7.54 ± 1.73) compared to UCB MNC and TCD CB alone (4.94 ± 1.41 and 4.88 ± 1.41, respectively) although the difference was not statistically significant in this relatively small cohort of animals (P = .44). There was a striking increase in mean percent erythroid engraftment in recipients of CD3/CD28 costimulated T cells (36.42 ± 8.56; Figures 5A and B) compared to the other groups (TCD UCB: 9.30 ± 6.99; UCB MNC 2.52 ± 6.99; P = .02), which accounted in large part for the overall multilineage engraftment advantage seen in the recipients of CD3/CD28 costimulated T cells (sum of CD45+CD3− and CD45−GlyA+CD36+: 43.95 ± 9.95 versus 14.81 ± 8.12 in TCD UCB alone and 7.46 ± 8.12 for UCB MNC; P = .04; Figure 5B). The majority of human CD45+ cells in bone marrow of recipients of UCB MNC were CD3+ T cells, which were not present in significant numbers in recipients of TCD UCB with or without CD3/CD28 costimulated T cells (Figure 5A), demonstrating a survival advantage for adoptively transferred unmanipulated UCB T cells over the cultured T cells, despite the presence of fewer T cells in the original xenograft. Histopathologic examination of liver and spleen of all mice analyzed (n = 16) showed extramedullary hematopoiesis in the spleens of all recipients, and evidence for liver xenoGVHD in 4 of 6 recipients of UCB MNCs, 0 of 6 recipients of TCD UCB, and 1 of 4 recipients of TCD UCB with CD3/CD28 costimulated T cells (Figure 5C).

  • View full-size image.
  • Figure 5. 

    Facilitation of early multilineage engraftment by CD3/CD28 costimulated UCB T cells is maintained in NOG mice. A, Marked erythroid engraftment in recipients of TCD UCB with costimulated CD3/CD28 UCB T cells accounted, in large part, for the overall engraftment advantage in this cohort of mice (n = 6 per group; 16 of 18 mice survived to analysis). Contour plots (A) of bone marrow cells from representative mice gated on live cells: x-axes: human CD45; y-axes: Glycophorin A (A, top panels) and CD3 (A, lower panels). The majority of human CD45+ cells in recipients of UCB MNC were expanded adoptively transferred CD3+ cells. B, Engraftment of the composite sum of CD45+CD3− cells together with the GlyA+CD45− cells. C, Photomicrographs of spleen (H&E: left, 10×, middle, 40×) of a recipient of TCD UCB with costimulated CD3/CD28 UCB T cells demonstrating extramedullary hematopoiesis. Right panel: liver (H&E; 10×) showing mild periportal and intrahepatic inflammatory infiltrates.

Back to Article Outline

Discussion 

In this study we sought to test the specific role of mature UCB T cells on engraftment of UCB into NOD/SCID-β2m−/− mice. We were interested in the overall level of engrafted human cells as a continuous variable, as well as a quantitative assessment of functional stem cell numbers as assessed by a limiting dilution analysis (Figure 1). We found that UCB grafts depleted of T cells have inferior levels of engraftment when compared to UCB MNCs transplanted into NOD/SCID-β2m−/− mice. Although the overall level of engraftment was significantly lower for recipients of TCD UCB, SRC frequency was not significantly different, suggesting that T cells augment hematopoiesis at stages beyond the stem cell. Finally, we found that engraftment capacity could be restored by add-back of ex vivo CD3/CD28 costimulated T cells from the same UCB unit; this property was maintained in NOG xenografts as well.

Because depletion of T lymphocytes adversely influences engraftment of UCB into NOD/SCID-β2m−/− recipients who lack T, B, and NK cells, our results suggest that T cells have a role beyond overcoming immune barriers in HSCT. Transplanting limited numbers of cells, we have demonstrated that T cell replete UCB results in a more than 2-fold increase in the level of engraftment when compared to TCD UCB. However, although the level of engraftment of TCD UCB into NOD/SCID-β2m−/− mice is significantly impaired, quantitative assessment of SRC frequency by limiting dilution analysis did not reveal significant differences between TCD UCB and UCB MNCs. This suggests that the differences in the level of engraftment are explained by accessory cell influences at the progenitor stage or beyond. Of course, it is possible that we have failed to identify a modest difference in SRC frequency that a larger sample size might have detected, but our data clearly show a positive effect on the level of engraftment, beyond the level of the SRC. Preservation of total cell numbers took priority in our experimental design over normalization for CD34+ cells, but it is worth noting that CD34+ cells were slightly but consistently enriched in the TCD UCB grafts, although the difference was not statistically significant. Thus, it is possible that on a per SRC basis, our experimental design is minimizing the true difference between TCD UCB and UCB MNCs, both in terms of level of engraftment and SRC frequency.

Most importantly, we found in this preclinical model that CD3/CD28 costimulated UCB T cells infused at the time of transplantation possess graft-facilitating properties. TCD UCB coinfused with physiologic numbers of CD3/CD28 costimulated T cells restored and enhanced engraftment. When supraphysiologic numbers of CD3/CD28 costimulated T cells were adoptively transferred into NOG mice, graft facilitation was maintained in terms of early engraftment. Effects were particularly striking in the erythroid compartment, which might reflect the erythroid stimulating (burst promoting) activities of granulocyte-monocyte colony stimulating factor (GM-CSF) and interleukin-3 (IL-3), whose secretion is markedly increased in costimulated T cells [20, 21, 22]. Because GM-CSF and IL-3 do not act at the stem cell level itself, it may be that T cells enhance engraftment by acting only on multipotent or later stage progenitor cells.

Somewhat surprising in the NOG recipients of UCB MNC was the predominance of adoptively transferred human CD3+ cells, suggesting that this model is particularly permissive for T cell expansion, homeostatic proliferation, or otherwise, and that the predominantly naïve T cells in UCB MNC have an in vivo survival advantage over CD3/CD28 costimulated T cells. Although only a rigorous clinical phase 1 safety study can truly address the true toxicity of these cells in terms of GVHD, it is nonetheless reassuring that evidence for xenoGVHD was rare in recipients of CD3/CD28 costimulated T cells.

Goodman et al. [23, 24] first showed in the 1960s that thymocytes enhanced hematopoiesis in a rat HSCT model, and, somewhat surprisingly, that this property was present and uniquely potent in tolerant thymocytes. In 1978, it was shown that T cells were required for the proliferation of primitive red blood cell precursors in vitro, and that a soluble factor from the T cell culture was responsible for this effect [25]. The starkest evidence for a critical role of T cells in hematopoiesis in vivo came in the 1980s, with clinical outcomes from TCD bone marrow allografts, which were found to have a disappointingly high rate of graft failure: 10%-30% in recipients of HLA identical TCD grafts and 50%-70% in recipients of HLA mismatched TCD grafts [26, 27]. In the same decade it was shown that adoptive transfer of buffy coat cells could overcome graft rejection and increase overall survival (OS) in patients undergoing HSCT for severe aplastic anemia [28].

Mature immune cells have been implicated more recently by Kim et al. [29] in preclinical models exploring the fate of transplanting 2 UCB grafts into NOD/SCID mice. When equal numbers of total MNCs from 2 UCB units are transplanted, 1 unit predominates. Interestingly, if the UCB grafts are first depleted of lineage committed cells (including T cells, NK cells, B cells, and mature myeloid cells), there is balanced coengraftment of both UCB units. Similarly, cotransfer of bone marrow derived mesenchymal stem cells—cells thought to possess immunosuppressive properties—from a separate, third-party source will also result in balanced coengraftment, even when total MNCs from 2 UCB units are transplanted. Taken together, these observations suggest that the level of engraftment of UCB HSCs is a function of mature accessory cells in the graft, and that the predominance of 1 graft over another is likely immune mediated. Recent clinical studies by Barker et al. [30] at the University of Minnesota are consistent with these models: when 23 subjects underwent HSCT as recipients of 2 partially mismatched unrelated UCB units, in all cases 1 UCB unit ultimately predominated with apparent loss of the other graft. Neither the number of CD34+ cells nor total cell number in the graft predicted which graft would predominate, although there was a direct association between the number of T cells and the UCB unit with durable engraftment. In addition, the median time to engraftment (23 days) was meaningfully shorter than that in most published reports of UCB HSCT. Although the mechanisms underlying this observation have not been identified, one interesting possibility is that an alloresponse between UCB units mediates not only rejection of 1 graft, but also actually accelerates engraftment of the graft fated to survive, either directly through specific T cell interactions or via nonspecific cytokine-mediated mechanisms. The dual UCB experience is encouraging, and represents an important potential new approach to reducing transplant related mortality (TRM) of UCB HSCT. That only 1 UCB unit ultimately engrafts raises the possibility that functional stem cell activity is modifiable by accessory cells in the graft, and a more precise understanding of these mechanisms might also enhance engraftment of single UCB units.

Several lines of evidence suggest that both cytokine and contact dependent pathways are mechanistically involved in the establishment of hematopoiesis in xenograft as well as mouse HSCT models [31]. Our observation suggests that T cells are key accessory cells in this model. Mechanistically, Adams et al. [32] found that UCB CD8+ cells could facilitate engraftment of UCB-derived progenitors in a contact-dependent manner, and suggest that they influence the response of CD34+ cells to chemotactic gradients. Mouse HSCT models suggest that although CD8+ T cells lacking cytotoxic effector mechanisms can support the early establishment of progenitor cells, long-term engraftment of mature peripheral blood cells requires intact perforin [33, 34]. Whether graft facilitation is inherent to canonical CD8+ T lymphocytes or to other CD8+ populations in this model system is unknown: CD8lo cells lacking a T cell receptor (TCR−) have been termed facilitating cells and enhance engraftment of murine HSCs in mouse models; these cells were more recently further defined by Fugier-Vivier et al. [35, 36, 37] as plasmacytoid precursor dendritic cells, and also posess tolerogenic properties. The functional importance of these cells in clinical HSCT has not yet been determined, and in fact, UCB grafts have proportionally more CD8lo TCR− cells compared to blood or marrow graft sources, despite the impaired engraftment properties of UCB [38].

T cell depletion did not significantly decrease SRC frequency, which may help to pinpoint the early progenitor period as the critical window for T cell help. In (nontransplant) studies of normal hematopoiesis, Monteiro et al. [39] showed that T cell-deficient mice are neutropenic but have an accumulation of immature myeloid cells in bone marrow, and found that T cell infusions can restore normal hematopoiesis. Interestingly (despite the intense interest and substantial data in the transplant literature on the graft-facilitating effect of the CD8+ compartment), in this model only CD4+ cells could be shown to rescue these mice, whereas CD8+ cells could not [32, 35, 36, 37, 40].

We found that CD3/CD28 costimulated UCB T cells were potent graft-facilitating cells, establishing proof of principle for testing the intervention in clinical UCB transplantation. This model will also be useful for evaluating dose responses to supraphysiologic numbers of CD3/CD28 costimulated T cells. Here we only tested engraftment of UCB, so it is not possible to address whether unmanipulated UCB T cells are indeed “defective” with respect to graft facilitation. The interpretation that CD3/CD28 activated UCB T cells facilitate engraftment is consistent with the observation by Fugier-Vivier et al. that their CD8loTCR− facilitating cells are most effective in an activated state. It is possible that microenvironmental paracrine factors might explain such findings: to date, there is no cytokine, alone or in combination, that can support in vitro hematopoietic colony formation as well as lymphocyte-conditioned medium, suggesting that there are yet to be identified soluble factors produced by activated T cells that augment hematopoiesis [41]. IL-17, a product of activated T cells, deserves further study: it is known to expand hematopoietic progenitors and enhance granulopoiesis; more recently, an IL-17 receptor-deficient mouse was found to have impaired hematopoietic recovery following irradiation [42, 43, 44].

In summary, we have shown that T cells augment hematopoiesis in immunodeficient UCB xenograft models. Engraftment is increased in the presence of T cells, whereas functional numbers of stem cells are unchanged, suggesting that T cells support later stages of hematopoiesis. Our results suggest that T cells facilitate engraftment beyond simply overcoming host barriers, and that CD3/CD28 costimulated UCB T cells possess graft-promoting properties, representing a potential translational approach for facilitating engraftment in UCB HSCT.

Back to Article Outline

Acknowledgments 

E.O.H. is supported by K12 RR017625 in Patient Oriented Research. We are grateful for support from the Leukemia and Lymphoma Society, and for help from Jennifer Luongo, Anthony Secreto, and Ella Ofori in the xenograft experiments, as well as the staff at the John Morgan 3 animal facility.

Back to Article Outline

References 

  1. Szabolcs P, Park KD, Reese M, et al. Coexistent naive phenotype and higher cycling rate of cord blood T cells as compared to adult peripheral blood. Exp Hematol. 2003;31:708–714
  2. Wagner JE, Kernan NA, Steinbuch M, et al. Allogeneic sibling umbilical-cord-blood transplantation in children with malignant and non-malignant disease. Lancet. 1995;346:214–219
  3. Gluckman E, Rocha V, Arcese W, et al. Factors associated with outcomes of unrelated cord blood transplant: guidelines for donor choice. Exp Hematol. 2004;32:397–407
  4. Laughlin MJ, Eapen M, Rubinstein P, et al. Outcomes after transplantation of cord blood or bone marrow from unrelated donors in adults with leukemia. N Engl J Med. 2004;351:2265–2275
  5. Rocha V, Labopin M, Sanz G, et al. Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. N Engl J Med. 2004;351:2276–2285
  6. Christianson SW, Greiner DL, Hesselton RA, et al. Enhanced human CD4+ T cell engraftment in beta2-microglobulin-deficient NOD-scid mice. J Immunol. 1997;158:3578–3586
  7. McCune JM, Namikawa R, Kaneshima H, et al. The SCID-hu mouse: murine model for the analysis of human hematolymphoid differentiation and function. Science. 1988;241:1632–1639
  8. Larochelle A, Vormoor J, Hanenberg H, et al. Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: implications for gene therapy. Nat Med. 1996;2:1329–1337
  9. Shultz LD, Schweitzer PA, Christianson SW, et al. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J Immunol. 1995;154:180–191
  10. Kollet O, Peled A, Byk T, et al. Beta2 microglobulin-deficient (B2m(null)) NOD/SCID mice are excellent recipients for studying human stem cell function. Blood. 2000;95:3102–3105
  11. McKenzie JL, Gan OI, Doedens M, Dick JE. Human short-term repopulating stem cells are efficiently detected following intrafemoral transplantation into NOD/SCID recipients depleted of CD122+ cells. Blood. 2005;106:1259–1261
  12. Shultz LD, Lyons BL, Burzenski LM, et al. Human lymphoid and myeloid cell development in NOD/LtSz-scid IL2R gamma null mice engrafted with mobilized human hemopoietic stem cells. J Immunol. 2005;174:6477–6489
  13. Levine BL, Bernstein WB, Connors M, et al. Effects of CD28 costimulation on long-term proliferation of CD4+ T cells in the absence of exogenous feeder cells. J Immunol. 1997;159:5921–5930
  14. Porter DL, Levine BL, Bunin N, et al. A phase 1 trial of donor lymphocyte infusions expanded and activated ex vivo via CD3/CD28 costimulation. Blood. 2006;107:1325–1331
  15. Rapoport AP, Stadtmauer EA, Aqui N, et al. Restoration of immunity in lymphopenic individuals with cancer by vaccination and adoptive T-cell transfer. Nat Med. 2005;11:1230–1237
  16. Laport GG, Levine BL, Stadtmauer EA, et al. Adoptive transfer of costimulated T cells induces lymphocytosis in patients with relapsed/refractory non-Hodgkin lymphoma following CD34+-selected hematopoietic cell transplantation. Blood. 2003;102:2004–2013
  17. Wang JC, Doedens M, Dick JE. Primitive human hematopoietic cells are enriched in cord blood compared with adult bone marrow or mobilized peripheral blood as measured by the quantitative in vivo SCID-repopulating cell assay. Blood. 1997;89:3919–3924
  18. Porter EH, Berry RJ. The efficient design of transplantable tumour assays. Br J Cancer. 1963;17:583–595
  19. Danet GH, Pan Y, Luongo JL, et al. Expansion of human SCID-repopulating cells under hypoxic conditions. J Clin Invest. 2003;112:126–135
  20. Thompson CB, Lindsten T, Ledbetter JA, et al. CD28 activation pathway regulates the production of multiple T-cell-derived lymphokines/cytokines. Proc Natl Acad Sci USA. 1989;86:1333–1337
  21. Guba SC, Stella G, Turka LA, et al. Regulation of interleukin 3 gene induction in normal human T cells. J Clin Invest. 1989;84:1701–1706
  22. Donahue RE, Wang EA, Stone DK, et al. Stimulation of haematopoiesis in primates by continuous infusion of recombinant human GM-CSF. Nature. 1986;321:872–875
  23. Goodman JW, Burch KT, Basford NL. Graft-vs.-host activity of thymocytes: relationship to the role of thymocytes in hemopoiesis. Blood. 1972;39:850–861
  24. Goodman JW, Shinpock SG. Influence of thymus cells on erythropoiesis of parental marrow in irradiated hybrid mice. Proc Soc Exp Biol Med. 1968;129:417–422
  25. Nathan DG, Chess L, Hillman DG, et al. Human erythroid burst-forming unit: T-cell requirement for proliferation in vitro. J Exp Med. 1978;147:324–339
  26. Ho VT, Soiffer RJ. The history and future of T-cell depletion as graft-versus-host disease prophylaxis for allogeneic hematopoietic stem cell transplantation. Blood. 2001;98:3192–3204
  27. Martin PJ, Hansen JA, Torok-Storb B, et al. Graft failure in patients receiving T cell-depleted HLA-identical allogeneic marrow transplants. Bone Marrow Transplant. 1988;3:445–456
  28. Storb R, Doney KC, Thomas ED, et al. Marrow transplantation with or without donor buffy coat cells for 65 transfused aplastic anemia patients. Blood. 1982;59:236–246
  29. Kim DW, Chung YJ, Kim TG, et al. Cotransplantation of third-party mesenchymal stromal cells can alleviate single-donor predominance and increase engraftment from double cord transplantation. Blood. 2004;103:1941–1948
  30. Barker JN, Weisdorf DJ, DeFor TE, et al. Transplantation of 2 partially HLA-matched umbilical cord blood units to enhance engraftment in adults with hematologic malignancy. Blood. 2005;105:1343–1347
  31. Bonnet D, Bhatia M, Wang JC, et al. Cytokine treatment or accessory cells are required to initiate engraftment of purified primitive human hematopoietic cells transplanted at limiting doses into NOD/SCID mice. Bone Marrow Transplant. 1999;23:203–209
  32. Adams GB, Chabner KT, Foxall RB, et al. Heterologous cells cooperate to augment stem cell migration, homing, and engraftment. Blood. 2003;101:45–51
  33. Jiang Z, Adams GB, Hanash AM, et al. The contribution of cytotoxic and noncytotoxic function by donor T-cells that support engraftment after allogeneic bone marrow transplantation. Biol Blood Marrow Transplant. 2002;8:588–596
  34. Martin PJ, Akatsuka Y, Hahne M, Sale G. Involvement of donor T-cell cytotoxic effector mechanisms in preventing allogeneic marrow graft rejection. Blood. 1998;92:2177–2181
  35. Gandy KL, Domen J, Aguila H, Weissman IL. CD8+TCR+ and CD8+TCR− cells in whole bone marrow facilitate the engraftment of hematopoietic stem cells across allogeneic barriers. Immunity. 1999;11:579–590
  36. Kaufman CL, Colson YL, Wren SM, et al. Phenotypic characterization of a novel bone marrow-derived cell that facilitates engraftment of allogeneic bone marrow stem cells. Blood. 1994;84:2436–2446
  37. Fugier-Vivier IJ, Rezzoug F, Huang Y, et al. Plasmacytoid precursor dendritic cells facilitate allogeneic hematopoietic stem cell engraftment. J Exp Med. 2005;201:373–383
  38. Theilgaard-Monch K, Raaschou-Jensen K, Palm H, et al. Flow cytometric assessment of lymphocyte subsets, lymphoid progenitors, and hematopoietic stem cells in allogeneic stem cell grafts. Bone Marrow Transplant. 2001;28:1073–1082
  39. Monteiro JP, Benjamin A, Costa ES, et al. Normal hematopoiesis is maintained by activated bone marrow CD4+ T cells. Blood. 2005;105:1484–1491
  40. Martin PJ, Rowley SD, Anasetti C, et al. A phase I-II clinical trial to evaluate removal of CD4 cells and partial depletion of CD8 cells from donor marrow for HLA-mismatched unrelated recipients. Blood. 1999;94:2192–2199
  41. Nissen-Druey C, Tichelli A, Meyer-Monard S. Human hematopoietic colonies in health and disease. Acta Haematol. 2005;113:5–96
  42. Tan W, Huang W, Zhong Q, Schwarzenberger P. IL-17 receptor knockout mice have enhanced myelotoxicity and impaired hemopoietic recovery following gamma irradiation. J Immunol. 2006;176:6186–6193
  43. Schwarzenberger P, Huang W, Ye P, et al. Requirement of endogenous stem cell factor and granulocyte-colony-stimulating factor for IL-17-mediated granulopoiesis. J Immunol. 2000;164:4783–4789
  44. Schwarzenberger P, La Russa V, Miller A, et al. IL-17 stimulates granulopoiesis in mice: use of an alternate, novel gene therapy-derived method for in vivo evaluation of cytokines. J Immunol. 1998;161:6383–6389

PII: S1083-8791(07)00332-1

doi:10.1016/j.bbmt.2007.06.010

Biology of Blood and Marrow Transplantation
Volume 13, Issue 10 , Pages 1135-1144, October 2007

Article Tools

* Image Usage & Resolution