TOLERANCE OF ALLOGENEIC HEART GRAFTS IN MICE SIMULTANEOUSLY RECONSTITUTED WITH PURIFIED ALLOGENEIC HEMATOPOIETIC STEM CELLS¹

* Abbreviations: APC, antigen-presenting cell; EHT, ear-heart transplant; FACS, fluorescence-activated cell sorter; GVHD, graft-versus-host disease; HSC, hematopoietic stem cell; PBL, peripheral blood lymphocyte; TCD, T cell-depleted; WBM, whole bone marrow; KTLS HSC, c-Kit⁺ Thy-1.1^loLin^-/loSca-1⁺ hematopoietic stem cell.

Allogeneic transplantation tolerance can be defined in several ways, but, in its most general immunological sense, it refers to a lack of immunological reactivity to a defined set of donor transplant antigens during maintenance of immunological reactivity to third-party transplantation antigens. Animals administered allogeneic hematopoietic cells at fetal^(1-4), neonatal^(5-9), and adult^(10-19) stages of development may become tolerant of donor-specific solid organ grafts. The seminal observations and analyses that provided the impetus for investigation in this field were made over 40 years ago. Owen⁽²⁰⁾ noted that the frequency of identical blood types in twin pairs of cattle could not be explained by chance, monozygotic twinning, or a combination thereof. He hypothesized that this phenomenon resulted from the exchange of hematopoietic cells through vascular anastomoses between twins in utero. These observations were extended to the realm of transplantation immunity when Anderson et al. ⁽²¹⁾ noted that a certain percentage of skin grafts exchanged between such twins were accepted, even when the twins were not monozygotic. The stage was thus set for the evaluation of a possible tolerance-inducing role of hematopoietic chimerism in mammalian transplantation. The mechanism(s) of such tolerance induction and maintenance is not clear, but much analysis has focused on the need for the persistence of foreign antigen ^{(6, 22-26)}. Mitchison ⁽²²⁾ studied a model in which tolerance was induced in embryonic and newborn chicks after allogeneic red cell infusion. He noted that tolerance disappeared within 29 days when the transfusions were terminated. Simonsen ⁽²³⁾ noted a similar phenomenon in embryonic chicks; partial tolerance to human blood was lost when transfusions were not continued. Weissman ⁽²⁴⁾ noted that the level of transplantation (and tolerance inducing) male antigen was persistent but low in adult mice which had been rendered tolerant to the H-Y antigen by injection of male spleen cells at birth, a finding which suggested that michrochimerism had altered host immunoreactivity to H-Y ⁽²⁴⁾. Lubaroff and Sharabi ^{(25, 26)} independently found that tolerance to allospecific antigens was lost when the cells bearing the tolerizing antigen were ablated, clearly suggesting the need for the persistence of the hematopoietic innoculum. Abundant correlative data lend further support to these observations; when multiple lineages are reconstituted long-term (long-term arbitrarily defined as >5 months), tolerance is long-term and reproducible^{(6, 18, 27-30)}. Analysis of different models reveals that high levels of hematopoietic reconstitution assure reproducible and long-term tolerance, although such levels are not always necessary for tolerance induction or maintenance.

There is a need to evaluate the placement of the solid organ graft both simultaneously and subsequently to the administration of the hematopoietic graft; different clinical scenarios necessitate different induction protocols. In cadaveric solid organ transplantation, it will not be possible to administer the hematopoietic graft a significant period before placement of the organ graft; therefore, simultaneous hematopoietic/solid organ graft transplantation is necessary and was evaluated in this study. In circumstances where a living related donor is available, however, it will be possible to administer the hematopoietic graft before placement of the solid organ graft. Therefore, solid organ graft placement after hematopoietic graft administration was similarly evaluated.

The production of tolerance to donor-specific antigens by hematopoietic reconstitution has important implications for solid organ transplantation. Clinical allogeneic hematopoietic reconstitution has predominantly been performed with whole bone marrow (WBM*) or partially purified populations of WBM which have the potential to produce graft-versus-host disease (GVHD). Removal of T cells has resulted in less severe GVHD. Unfortunately, however, removal of these T cell populations also results in lower rates of allogeneic hematopoietic engraftment and decreased survival ^{(31, 32)}.

Our laboratory has identified a population of cells in WBM that has the ability to reconstitute all hematopoietic lineages ⁽³³⁾ and does not cause GVHD in a fully allogeneic host ⁽³⁴⁾. This population of hematopoietic stem cells (HSCs), can be identified by the cell surface markers c-Kit⁺Thy-1.1^loLin^-/ loSca-1⁺(KTLS) on the fluorescence-activated cell sorter (FACS). It has been previously demonstrated that such highly purified HSCs are sufficient to reconstitute mice across allogeneic barriers and that T lymphocytes from chimeric mice do not show GVHD against donor or host antigens⁽³⁴⁾. This study was designed to test the hypothesis that hematopoietic reconstitution with HSCs will result in solid organ tolerance without GVHD.

A mouse model was designed to evaluate the potential of HSCs in tolerance-inducing regimens. BALB/c mice were preconditioned for hematopoietic transplant with lethal irradiation and were reconstituted with C57BL/Ka, Thy-1.1 HSCs or WBM. At the same time or 35-70 days later, recipients were then given allogeneic or third-party solid organ grafts. The solid organ graft involved placing a neonatal heart in the subcutaneous tissue of the ear(ear-heart transplant, EHT), which resumed cardiac contraction after graft neovascularization by invading vessels. Hematopoietic reconstitution was monitored by evaluation of the presence of donor-derived cells in the blood. The heart graft was visually monitored for evidence of myocardial contraction; cessation of contraction indicated the graft had been rejected.

In some clinical circumstances, HSCs may be limited, and it may be advantageous to utilize cellular populations that enhance HSC engraftment. Subpopulations of WBM have been identified that facilitate the ability of HSCs to engraft across allogeneic barriers and would potentiate engraftment when limiting numbers of HSCs are available. The CD8 molecule, a heterodimeric protein on the surface of a subset of T cells, has been identified on the surface of such cells and can be used to isolate these populations on the FACS^(35-37). Accordingly, ear-heart survival was also evaluated in mice reconstituted with minimal numbers of allogeneic HSCs and sorted CD8⁺ cells.

MATERIALS AND METHODS

Mice. Adult male C3H (H2^k), BALB/c (H2^d), and C57BL/Ka, Thy-1.1, Ly-5.1 (H2^b) mice were maintained in the animal care facilities at Stanford. Hearts were obtained from mice within 24 hr of birth for EHT. Mice were given acidic water for at least 1 week before irradiation and antibiotic water (1.1 g/L neomycin sulfate and 10⁶ U/L polymixin B sulfate) for at least 8 weeks after receiving the transplant to reduce the chance of infection. Hematopoietic donors were 4-6 weeks of age; recipients were 7-12 weeks of age.

Antibodies for cell sorting and analysis. The rat antibodies 53-7.3 (anti-CD5), 53-6.7 (anti-CD8), Ter-119 (anti-erythro), GK 1.5(anti-CD4) KT 31.1 (anti-CD3), 6B2 (anti-B220), M1/70 (anti-Mac-1), 8C5(anti-Gr-1), 19XE5 (anti-Thy-1.1), and E13-161 (anti-Sca-1) were prepared from the respective clones. Antibodies were conjugated with fluorescein isothiocyanate (FITC), phycoerythrin (PE), or allophycocyanin: (KT 31.1-PE, 6B2-FITC, M1/70-FITC, 8C5-PE, 19XE5-FITC, A201.7-allophycocyanin (mouse anti-Ly5.1), and 2B8-allophycocyanin (anti-c-Kit). Goat anti-rat IgG-PE was obtained from Caltag (South San Francisco, CA, catalog no. R40004-3). Avidin Texas Red (avidin-TR, catalog no. 55894) was obtained from Cappel (West Chester, PA).

Stem cell sorting. HSCs were prepared as previously described⁽³⁸⁾. Briefly, WBM was obtained by flushing the tibias and femurs of adult mice with staining media (1% Hanks' buffered saline solution with 3% fetal calf serum and 10 mM Hepes, pH 7.0). Cells were incubated with rat antibodies specific for mouse lineage markers (53-7.3, 53-6.7, Ter-119, GK 1.5, KT31.1, 6B2, M1/70, and 8C5) for 20 min, washed with staining media, and incubated with goat antirat IgG-PE for 30 min. The washed cells were incubated with rat IgG to occupy any nonspecific binding sites. Cells were washed, incubated with anti-Sca-1 for 20 min, washed, and incubated with avidin-MiniMacs beads, anti-Thy-1.1-FITC, avidin-TR, with or without anti-c-Kit-antigen-presenting cells (APC). Cells were washed and passed twice through MiniMacs columns (Miltenyi Biotec, Auburn, CA). Cells retained on the MiniMacs columns (5-10-fold enriched for HSC activity) were eluted and resuspended in propidium iodide (PI)-containing media before sorting.

CD8⁺cell sorting. WBM was incubated with 53-6.7-biotin for 20 min. Cells were then washed and incubated with avidin-MiniMacs beads for 20 min. Cells were washed and passed through a MiniMacs column as described. Cells retained in the column by the magnetic beads were eluted and incubated with AV-PE (Caltag) for 20 min. Cells were washed, resuspended in PI, and sorted on the FACS.

Strains of mice used for reconstitution, solid organ grafts, and analysis. Allogeneic reconstitution was performed between strains that differed at both major and minor histocompatibility loci. HSCs and WBM donors were of the H2^b haplotype and could also be identified by the allelic Ly5.1 marker. Markers for differentiation of cellular origin are indicated in bold type (Table 1).

Reconstitution. For reconstitution, cells to be injected into an animal were mixed, and resuspended in 100 μl of phosphate-buffered saline before retroorbital injection.

Analysis of reconstitution. Reconstituted animals were bled 6-36 weeks after reconstitution for analysis of donor-derived hematopoietic cells. Blood was collected in ethylenediaminetetraacetate and separated in a dextran gradient. Mononuclear cells were collected and residual red cells were lysed with 0.15 M ammonium chloride/0.01 M potassium bicarbonate. Cells were divided into two or three aliquots and stained with the following antibody panels: A20-1.7-APC, 6B2-FITC, KT-31.1-PE; A20-1.7-APC, M1/70-FITC, 8C5-PE; A20-1.7-APC, GK-1.5-PE, 53.6.7-FITC; A20-1.7-APC, NK1.1-PE, 6B2-FITC; A20-1.7-APC, H2K^b-Bio, GK-1.5-PE, 53-6.7-FITC. Cells were resuspended in PI-containing staining media for FACS analysis. Antibodies were titered such that positive and negative haplotype controls had less than 1% nonspecific staining. Percent donor chimerism is reported without correction for such nonspecific staining.

EHT. The method of Fulmer et al. ⁽³⁹⁾, as modified separately by Hunter and Hollman ⁽⁴⁰⁾. and Babany ⁽⁴¹⁾, was used for the EHT. Briefly, the pinna of an adult mouse was prepped with betadine and a subcutaneous tunnel was prepared. The heart of a neonatal mouse was placed in the tunnel after residual blood has been removed with cold saline lavage. Residual fluid was cleared from the tunnel with a cotton swab after transplantation. Every 2-4 days during the first 4 weeks and weekly thereafter, grafts were observed under 10× magnification for evidence of cardiac activity. Failure of a graft to beat by day 9 after transplant was termed technical failure, and those data were excluded (n=2 of 81). Grafts that were deemed rejected by 7 days had shown evidence of cardiac activity before rejection. Rejection is defined as the point of cessation of visual cardiac activity. All but 15 allogeneic transplants were performed concurrently with the HSC reconstitution; third-party grafts were placed over 2 months after hematopoietic reconstitution.

RESULTS

Determination of limit dilution radioprotective dose of allogeneic HSCs. HSCs were characterized by the cell surface staining profile of c-Kit⁺Thy-1.1^loLin^-/loSca-1⁺ and represent 0.04-0.06% of WBM from 4- to 6-week-old donors. The left panel ofFigure 1A depicts the staining profile of WBM for Thy-1.1 and the lineage cocktail (anti-CD3, anti-CD4, anti-CD5, anti-CD8, anti-erythro, anti-Mac-1, anti-GR-1, anti-B220). The right panel shows that a clear population of c-Kit⁺/Sca-1⁺ cells are delineated by gating on the Thy-1.1^loLin^-/lo population of WBM. This is the population used in this study. To assure that sorted cells were highly purified for stem cell activity, the cells were sorted and resorted for the stem cell surface phenotype (Fig. 1B). The top panels show the probability plots for the four selected parameters and the purity after a single sort; the bottom panels show the improved purity after a second sort, the population now being >98% pure for the desired surface phenotype. Previous data indicates that 100 Thy-1.1^loLin^-/loSca-1⁺ HSCs can save 90-100% of lethally irradiated Ly5 congenic mice⁽³³⁾. We used 2 times this limiting dose for congeneic reconstitution, and, as expected, achieved 100% survival. The limiting dose for reconstitution across allogeneic barriers was then determined. BALB/c mice receiving no cellular infusion died by 21 days (range, 11-21 days) whether they did (n=5) or did not (n=10) receive a heart transplant. C57BL/Ka mice died by 17 days (n=10, range, 12-17) if they did not receive hematopoietic cells (data not shown). Figure 2 depicts the survival curve for BALB/c mice that were given 250, 500, 1000, and 3000 C57BL/Ka, Thy-1.1 HSCs across allogeneic barriers. All animals survived when 3000 HSCs(n=10) of the KTLS surface phenotype were injected into lethally irradiated hosts. Eighty-eight percent of animals survived when 1000 cells are used. Therefore, the limit dilution at which 100% of animals can be saved from lethal irradiation was between 1000 and 3000 allogeneic HSCs. We chose 3000 HSCs as the dose for which allogeneic ear heart transplants should be performed in this study, a dose that would assure that survival did not limit the evaluation of tolerance in the group given HSCs. A dose of whole bone marrow, 6×10⁶, that contains 3000 HSCs was chosen for comparison. Animals receiving 3000 HSCs and 6×10⁶ WBM survived and were donor reconstituted long-term, >7 months, indicating that long-term reconstitution had been achieved.

Mice reconstituted with allogeneic HSCs are reconstituted in multilineage fashion. Ten lethally irradiated BALB/c mice (H2^d) were given 3000 C57BL/Ka, Thy-1.1 KTLS HSCs (H2^b). All of these animals survived more than 250 days. Analysis of peripheral blood leukocytes (PBLs) reveals that the animals were reconstituted with donor type cells: myeloid, B-lymphoid, and T-lymphoid cells. Mice reconstituted with allogeneic WBM were similarly reconstituted with donor type cells in a multilineage fashion. By 6-9 months after reconstitution, myeloid and B lymphoid cells reached their normal portion of the total mononuclear populations of peripheral blood in animals reconstituted with HSCs and WBM, and each of these lineages was reconstituted with predominantly donor-derived cells (Table 2). Analysis of the T cell lineage at multiple time points, however, revealed significant differences in donor-recipient chimerism as previously reported ⁽³⁴⁾(Table 2). At 8-9.5 months, 15% of the CD3⁺ PBLs were host-derived in animals reconstituted with KTLS HSCs, whereas all CD3⁺ cells were donor-derived in animals reconstituted with WBM. At 4 months, such differences were analyzed with respect to select T lymphocyte subpopulations. The percentages of CD4⁺ and CD8⁺ cells in PBLs are reduced in mice reconstituted with allogeneic or syngeneic KTLS. Host-derived cells are detected in both of these subpopulations in KTLS-reconstituted animals. The difference in host-derived populations detected between animals reconstituted with KTLS and WBM is statistically significant (Fig. 3).

Mice reconstituted with allogeneic HSCs or WBM are tolerant of donor-specific allografts. Unmanipulated BALB/c mice reject C57BL/Ka, Thy-1.1 (n=10) EHT grafts by 9 days (Fig. 4). BALB/c mice simultaneously receiving allogeneic C57BL/Ka, Thy-1.1, Ly5.1 KTLS (n=8) or WBM(n=8) and C57BL/Ka, Thy-1.1, Ly5.2 allogeneic ear heart grafts did not reject these heart grafts. Cardiac contractions were still detectable at 250 days(Fig. 4C). C57BL/Ka, Thy-1.1 mice simultaneously given syngeneic C57BL/Ka, Thy-1.1 KTLS HSCs (n=20) and a BALB/c allogeneic heart graft rejected grafts by day 35.

Mice reconstituted with minimal numbers of HSCs and facilitator cells are tolerant of solid organ grafts. CD8⁺ populations enhance the engraftment of bone marrow populations that include HSCs^{(35, 36)} and purified KTLS HSCs⁽³⁷⁾ across allogeneic barriers. We found that survival is increased from 40% in BALB/c mice given 250 C57BLKa,Thy1.1 KTLS HSCs to 90% in animals given 250 C57BLKa,Thy1.1 KTLS HSCs and 20,000 C57BLKa,Thy1.1 CD8⁺ cells (Fig. 5B)⁽⁴²⁾. Such survival correlates with donor-derived reconstitution. Five surviving animals that had been reconstituted with CD8⁺ cells and HSCs were given EHT 35 days after hematopoietic reconstitution. Five surviving animals given 250 HSCs alone (a pool of two separate experiments) were given EHT at 35 days. All animals were tolerant of grafts at 125 days (155 days after hematopoietic graft; Fig. 5C), and no gross evidence of GVHD was noted.

Mice reconstituted with allogeneic HSCs reject third-party grafts. Unmanipulated BALB/c mice reject C3H (n=10) EHT grafts by 9 days(Table 3). Unmanipulated C57BL/Ka, Thy-1.1 mice reject C3H EHT grafts by 9 days (Table 3). BALB/c mice receiving C57BL/Ka, Thy-1.1 HSCs (n=5) or WBM (n=5) and third-party (C3H) EHT more than 2 months after hematopoietic reconstitution rejected heart grafts by day 12. The difference in rejection kinetics between reconstituted and unmanipulated mice is statistically significant.

DISCUSSION

The population of cells used as stem cells in this study represents 0.04-0.06% of 4- to 6-week-old WBM and is highly purified for stem cell activity. These HSCs radioprotect (injection of 100 HSCs saves 90-100% of lethally irradiated congenic mice) ⁽³³⁾, provide multilineage reconstitution (a single cell can give rise to myeloid, B lymphoid, and T lymphoid progeny) ^{(43, 44)}, and self-renew (serial transplantation studies demonstrate that HSCs can be harvested from a reconstituted mouse) ^{(43, 45)}. These cells can reconstitute lethally irradiated mice across major and minor histocompatibility barriers.

Kaufman and colleagues ⁽³⁶⁾ reported previously that as many as 10,000 cells that are enriched for HSCs cannot reconstitute allogeneic mice. However, the population of cells used by Kaufman et al.(c-Kit⁺Sca-1⁺Lin^-) were phenotypically different from the HSCs used in this study (c-Kit⁺, Thy-1.1^loLin^-/loSca-1⁺), and were not as highly enriched for HSC activity, as reflected by the fact that a large number of these cells were needed for syngeneic radioprotection and reconstitution.

On a similar note, a previous study demonstrated that 6000 HSCs Thy-1.1^loLin^-/loSca-1⁺ were required for reconstitution of all mice across allogeneic barriers. In the present study, 2-6-fold less HSCs are required to obtain the same effects. Several factors may explain this difference. Cellular populations in this study were obtained after double sorting and with the use of the additional c-Kit marker, i.e., double-sorted c-Kit⁺Thy-1.1^loLin^-/loSca-1⁺ populations were used as opposed to single-sorted Thy-1.1^loLin^-/loSca^-1⁺ populations. In 6- to 8-week-old mice, after a single sort, up to 40% of the sorted population may be c-Kit^-. In 4- to 6-week-old mice, only 10% of the Thy-1.1^loLin^-/loSca-1⁺ pool is found to be c-Kit^-. Donors were 4 weeks old in the current study. We have previously shown that no long-term progenitors are found in the c-Kit^- population ^{(44, 46)}. After double-sorting and with the use of the additional c-Kit marker, the c-Kit^- contaminants are usually reduced to less than 3%. The combination of double sorting and the c-Kit marker can improve purity up to 50%, probably accounting for the described differences. One final issue, and one that frequently complicates comparative data within the field and may explain some of the noted differences, involves the present condition of the mouse colony. The reconstitution assays used in this study are noncompetitive reconstitutions. We used these types of assays to allow the evaluation of tolerance in the most defined system possible. Such survival assays are very sensitive to minor variations in actual or opportunistic pathogens in the mouse colony. Indeed, in subsequent studies, after improvements in the housing conditions of our mice, we find survival of up to 71% of mice (n=7) after the administration of as few as 250 HSCs. Organ transplantation was performed in mice reconstituted with HSCs in numbers equal to or greater than the limiting dilution, or in mice given HSCs and facilitators, reducing the potential for variability.

Graft survival is not affected by the time of transplantation in animals reconstituted across allogeneic barriers with neonatal blood (a substance containing hematopoietic progenitors) or whole bone marrow⁽⁴⁷⁾. As a result, most of the ear heart grafts in these experiments were performed simultaneously with the hematopoietic graft. In cadaveric transplants, this scenario closely approximates that of the clinic in that the stem cell harvest and organ harvest need to be performed simultaneously. However, in certain situations it may be possible to perform the solid organ transplant after hematopoietic reconstitution, i.e., a kidney transplant from a relative. Experiments were performed in which mice were given EHT both simultaneously and subsequentlyly to the KTLS HSC administration. All recipients were tolerant of donor-specific EHT. We know from the rejection of the third-party grafts that some degree of immunocompetence has been achieved by 35 days. Therefore, these grafts are not accepted simply for lack of maturity in the immune response and confirm that the solid organ graft was accepted due to tolerance of allospecific antigens. On a similar note, mice were given EHT 35 days after HSCs and CD8⁺ cells. These mice were similarly tolerant of solid organ grafts, indicating that such cells can be used to enhance engraftment of reconstituting cells without a detrimental effect on tolerance induction.

Hematopoietic microchimerism has long been investigated for its role in tolerance induction and maintenance to solid organ grafts. The mechanisms of such tolerance, however, are unclear. It is known that donor-derived elements are persistent in many models where donor-specific tolerance is reproducible and long-term ^{(6, 18, 25, 29, 30, 48)}. Similarly, a large body of clinical data highlights a strong correlation between persistent donor hematopoietic chimerism and successful solid organ transplantation^(49-54). It is not known, however, which of the hematopoietic subsets are necessary for this tolerance induction or how long they must persist. Because the population of cells used for hematopoietic reconstitution in our studies self-renews and gives rise to long-term multilineage reconstitution of donor-type origin, any elements necessary for solid organ tolerance should be present after allogeneic reconstitution.

All irradiated mice given hematopoietic grafts in this study were reconstituted in a multilineage fashion with donor-derived cells. We determined the dose of cells that would save 100% of mice. Three thousand HSCs resulted in 100% survival and a predominantly donor-derived reconstitution profile. Similarly, 6×10⁶ WBM, an amount of WBM which contains about 3000 HSCs, resulted in predominantly donor-derived multilineage reconstitution. A small percentage of host-derived CD4 and CD8 cells persisted at 4-5 months after reconstitution with HSCs alone, but these do not visibly effect graft function. Differences in host derived T lymphocyte profiles were previously noted in comparisons between profiles of animals reconstituted with TCD marrow and WBM ⁽⁵⁵⁾. The data in this study are consistent with previous data from the laboratory in which the entire hematolymphoid system tested in animals reconstituted with HSCs is donor-derived, except for the T lymphocyte populations; and in which all hematolymphoid cells are donor-derived if allogeneic WBM is transplanted⁽³⁴⁾. Present data further define the surface phenotype of residual host T cells. Both CD4⁺ and CD8⁺ populations of host origin continue to persist. It is currently not known whether these populations are unresponsive to donor antigens or whether they are not present in sufficient quantity to mediate a rejection response. Previous data from a similar model, however, indicated that radioresistant T cell populations showed reactivity to donor-derived alloantigens ⁽⁵⁶⁾. It is not known whether the difference between the WBM and HSC reconstitution profiles is indicative of the potential of cells within WBM to delete residual host cells, the potential of cells within WBM to selectively enhance hematopoiesis of HSCs, the ability of populations of WBM to occupy niches that would otherwise be occupied by endogenous cells, a combination of these factors, or other as yet undetermined factors.

The ear-heart transplant model has been a reliable screening tool for decades in transplant pharmacology. The clinical relevance of EHT might be questioned because the graft is neovascularized as opposed to primarily revascularized. However, because the antigens are presented subcutaneously in the ear-heart model, a mode Jenkins et al. demonstrate to be one of the most immunogenic ⁽⁵⁷⁾, we believe the EHT is a stringent model for evaluating tolerance to allogeneic tissues and is appropriate for evaluation of clinical relevance.

HSCs did not give rise to an immune system that recognized third-party H2 disparate antigens as self in this model; H2-disparate EHTs given with concurrent congenic hematopoietic reconstitution were rejected. The antigenic load, location, and/or mode of presentation of the EHT may not be sufficient for tolerance induction. The observation that mice syngeneically reconstituted and simultaneously given allogeneic ear heart grafts were not tolerant of the ear heart grafts lends support to this hypothesis.

The potential of T cells to become immune-competent when developing in an allogeneic host has been a controversial issue. Many studies have demonstrated recovery of immune competence ^(58-61) whereas others have emphasized the deficiencies^(62-64). Results vary depending on the reconstitution profiles, the strains evaluated, the time of evaluation of immune competence, and the assay used to determine immune competence. In this study, we evaluated immune competence of the reconstituted mice by testing for rejection of third-party H2 disparate grafts. The correlation between the rejection of third-party grafts and a response to pertinent pathogens is difficult to determine, but previous data suggest that alloreactivity and other restricted T cell activity may develop in parallel under certain circumstances ⁽⁵⁸⁾. Untreated C57BL/Ka or BALB/c mice rejected C3H grafts between day 6 and 9. BALB/c mice that were reconstituted with C57BL/Ka, Thy-1.1 allogeneic HSCs 2-3 months before transplantation with C3H grafts rejected the grafts by day 12. Although these animals were immunocompetent, the response to third-party antigens was slightly delayed, and this delay was statistically significant. The attenuation could be the result of one or a combination of the following: a reduction in the total T-lymphocyte count, a reduction in a specific subpopulation of effector lymphocytes, the inability of lymphocytes to mature appropriately in an allogeneic thymus, and/or the possibility that a number of TCR's that are allospecific for BALB/c H2 antigens are also reactive to C3H antigens, being depleted during or after thymic development.

The T lymphocyte populations and the respective subpopulations contained therein are not fully recovered even by 4-5.5 months after reconstitution, and the third-party grafts were given at 8 weeks. In mice reconstituted with KTLS HSCs, the percentage of total CD3⁺ cells in peripheral blood has only reached 70-90% of normal levels by 6-9 months after reconstitution, and up to 15% of these CD3⁺ cells continue to be of recipient origin with an undetermined responsive capacity. Although experiments are being performed in which grafts are placed at 4-5 months after reconstitution, interest will continue to focus on periods immediately after reconstitution as the kinetics of the recovery of immune competence after hematopoietic reconstitution is critical to clinical application of hematopoietic reconstitution in tolerance induction.

The restoration of immune competence in allogeneically reconstituted mice probably involves not only thymic T cell maturation but also extrathymic T cell maturation. The importance of extrathymic T cell maturation was suggested and evaluated in bone marrow chimeras of nu/nu recipients many years ago⁽⁶⁵⁾. Recent studies demonstrated that lethally irradiated C57BL/Ka, Ly5.2 thymectomized recipients transplanted with 200-500 C57BL/Ka, Ly5.1 HSCs could develop T cells in vivo. Similarly, TCD BM from these thymectomized mice could give rise to T cells in vitro in the absence of thymic elements ⁽⁶⁶⁾. Several experiments have demonstrated that thymic education of T cells requires thymic elements that bear identical MHC molecules and can participate in positive or negative selection. HSCs may give rise to thymic dendritic APC⁽⁶⁷⁾, B cells, and myeloid cells, all of which have antigen-presenting potential. The emergence of these cells requires time, but donor-derived thymic APC capable of negative selection have been observed within 2 months after F1 to parent bone marrow transplantation⁽⁶⁸⁾. On a similarl note, myeloid and B cells emerge within 1-2 weeks after allogeneic and syngeneic reconstitution and may have attained appropriate antigen-presenting capacity by the time T cell precursors are mature enough to undergo self/non-self education. It shall be important to determine in HSC transplants whether donor-derived APC can participate in positive and negative selection. It may be that donor-restricted thymic education proceeds with delayed kinetics restricted by the development of donor-derived APC, that the role of thymic and extrathymic maturation is determined by the emergence of APC in the designated sites, and that donor-restricted lymphocyte maturation occurs efficiently, but with delayed kinetics. The issues are complex, but the kinetics, specificity, and relative contribution of thymic and extrathymic selection and maturation are under current investigation in the allogeneic reconstitution setting.

These studies demonstrate that purified HSCs are sufficient for tolerance induction in lethally irradiated mice. Human HSCs have been identified and can be similarly isolated by FACS analysis ⁽⁶⁹⁾, and data from clinical trials confirm that this population has multilineage reconstitution potential and that autologous reconstitution proceeds with rapid kinetics ⁽⁷⁰⁾. In the healthy young mice used in the experiments reported in this study, immune competence is only slightly deficient at 8 weeks as evaluated by the rejection of third-party H2 disparate grafts. The potential effects of this temporal deficiency in terms of responses to potentially pathogenic organisms is undetermined. Obviously, a more relevant set of situation-specific preclinical models will be required to evaluate the recovery of immune competence in old and/or diseased individuals. Perhaps, the most appropriate first-order comparison is between the described experimental animals and subjects undergoing other immunosuppressive protocols. The immune “incompetence” demonstrated in this model is little different from that shown in the least deleterious of other such models. Current efforts include attempts to decrease the period of immune incompetence by the injection of subpopulations of WBM, which have been shown to enhance HSC engraftment without producing GVHD, as some data suggest the possibility that such populations may have a role in enhancing immune competence.

Acknowledgments. The authors acknowledge Jos Domen and Motonari Kondo for many helpful discussions during these studies; Libuse Jerabek for excellent laboratory management; Tim Knaack for his patience, persistence, and indispensable expertise and skill in cell sorting; Veronica Braunstein for antibody preparation; Annette Schlageter for careful reading of the manuscript; and Lu Hidalgo and Bert Lavarro for their work in animal care and management.