- Top of page
- Materials and Methods
In utero hematopoietic stem cell transplantation could potentially be used to treat many genetic diseases but rarely has been successful except in severe immunodeficiency syndromes. We explored two ways to potentially increase chimerism in a nonhuman primate model: (a) fetal immune suppression at the time of transplantation and (b) postnatal donor stem cell infusion. Fetal Macaca nemestrina treated with a combination of the corticosteroid betamethasone (0.9 mg/kg) and rabbit thymoglobulin (ATG; 50 mg/kg) were given haploidentical, marrow-derived, CD34+-enriched donor cells. Animals treated postnatally received either donor-derived T cell–depleted or CD34+-enriched marrow cells. Chimerism was determined by traditional and real-time polymerase chain reaction from marrow, marrow progenitors, peripheral blood, and mature peripheral blood progeny. After birth, the level of chimerism in the progenitor population was higher in the immune-suppressed animals relative to controls (11.3% ± 2.7% and 5.1% ± 1.5%, respectively; p = .057). Chimerism remained significantly elevated in both marrow (p = .02) and fluorescence-activated cell sorted and purified CD34+ cells (p = .01) relative to control animals at ≥ 14 months of age. Peripheral blood chimerism, both at birth and long term, was similar in immune-suppressed and control animals. In the animals receiving postnatal donor cell infusions, there was an initial increase in progenitor chimerism; however, at 6-month follow-up, the level of chimerism was unchanged from the preinfusion values. Although fetal immune suppression was associated with an increase in the level of progenitor and marrow chimerism, the total contribution to marrow and the levels of mature donor progeny in the peripheral blood remained low. The level of long-term chimerism also was not improved with postnatal donor cell infusion.
- Top of page
- Materials and Methods
Many diseases that affect normal fetal hematopoietic and immune function result in fetal death (α thalassemia), severe central nervous system developmental abnormalities at birth (leukodystrophies), or lifelong morbidity (sickle cell disease and β thalassemia) [1, 2]. Many of these disorders can be cured by postnatal bone marrow transplantation. Development of successful in utero hematopoietic stem-cell transplantation protocols would be advantageous, particularly for those disorders that result in fetal death or significant impairment at the time of birth. Naturally occurring models of in utero hematopoietic transplantation and preclinical animal models, primarily in fetal sheep, have suggested that clinical trials in the human fetus should be successful [1, 3, 4]. More than 40 attempts of human fetal transplantation have been reported for a variety of different diseases. The theoretical advantages to fetal transplantation (immature immune system and expanding hematopoietic environment) have been used to describe the fetus as the perfect recipient for hematopoietic transplantation . Unfortunately, evidence of engraftment and disease improvement has been demonstrated only in fetuses with severe immunologic deficiencies [6–9].
Why attempts of in utero stem cell transplantation for fetuses with normal immune development have failed has not been clearly determined. However, available data have demonstrated that, even in the fetus, there are significant barriers to donor cell engraftment. It is likely that the barriers to achieving clinically relevant levels of engraftment prenatally are similar to those for postnatal stem cell transplantation—that is, graft failure from an inadequate number of donor stem cells, immune rejection of donor cells by the fetus, inadequate hematopoietic space for donor cell engraftment, or inferior competitiveness of donor cells relative to the autologous fetal hematopoietic stem cells.
Our group  and others  have shown that fetal recipients of adult CD34+-enriched or T cell–depleted marrow produces engraftment in nonhuman primates. Tolerance has also been demonstrated in these animals by reduced mixed lymphocyte culture responses and prolonged solid organ graft survival. Unfortunately, the levels of chimerism achieved in both of these models have been low and would not be expected to correct any of the target diseases for in utero hematopoietic therapy. We explored two methods of potentially increasing the level of chimerism achieved after in utero hematopoietic transplantation: (a) fetal immune suppression as an adjunctive to fetal transplantation and (b) post-natal infusion of donor hematopoietic progenitor/stem cells.
- Top of page
- Materials and Methods
A total of five fetuses treated with the combination of ATG and betamethasone were transplanted with haploidentical, allogeneic CD34+ cells enriched from bone marrow of the sire. The total CD34+ (3.7 × 109/kg) and CD2+ (1.6 × 107/kg) donor cell number was similar to another group of animals similarly transplanted that did not receive immunosuppression (Table 1) .
Two of the five animals treated with immune suppression were electively delivered prior to viability, and three animals were delivered at full-term gestation by elective cesarean section. The first of the preterm fetuses was electively delivered at 120 days (0.70 gestation) after routine ultrasound monitoring of the fetus demonstrated oligohydramnios and large echogenic kidneys. These findings were consistent with the infantile polycystic kidney disease, a lethal disorder that has been previously described in the macaque species [23, 24]. Pathological examination of the kidneys confirmed these findings and the absence of GVHD. The second preterm animal was delivered at 77 days (0.45 gestation) when fetal ascites was noted at the time of the scheduled third intrauterine injection. This fetus did not receive the third cell infusion and was delivered 3 days later after persistence of the ascites was noted. Pathologic examination did not identify any abnormalities of the abdominal structures, and there was no evidence of GVHD. Hematopoietic tissue from both of these animals was obtained, and demonstrated high levels of chimerism in fetal liver of the younger animal and in the marrow and cord blood of the older animal. Of the three animals delivered at term by cesarean section, one was euthanized at 19 months of age due to chronic osteomyelitis involving the right femur. The other two animals are alive and well at 14 and 16 months of age.
Influence of Fetal Immunosuppression on Initial Engraftment of Hematopoietic Progenitors
The two animals that were electively delivered prior to viability demonstrated high levels of chimerism within fetal liver, marrow, and peripheral blood (Table 2). In the younger fetus (0.45 gestation) 17% of fetal liver progenitors (CFCs) and 2.7% of all fetal liver cells were of donor origin, suggesting a high level of initial engraftment in that organ. Cells obtained from the marrow plated at the same time yielded no CFCs. The older fetus (0.70 gestation) demonstrated donor cells in both peripheral blood (34% CFCs) and marrow (43% CFCs and 0.15% total cells), but not in the fetal liver.
Table Table 2.. Colony-forming cells (CFCs) from immune-suppressed preterm fetuses
|Animal no.||Days of gestation (% gestation)||PCR||CFC-BMa||Cord blood CFC||CFCaliver||Real-time PCR|
|099b||70 (0.45)||N/A||NG||N/A||23/132 (17%)||2.7% (FL)b|
|061c||120 (0.70)||10/10+c||35/80+ (43%)||20/58+ (34%)||NG||0.15% (BM)d|
In the three immunosuppressed animals delivered at term, initial evaluation of chimerism in the progenitor population (single CD34+ cell cultures) suggested that engraftment at birth was higher than that observed in the group of animals that did not receive immunosuppression (11.3% ± 2.7% and 5.1% ± 1.5%, respectively; p = .057) (Fig. 1, Table 3). Long-term follow-up of these animals, at 14, 16, and 19 months, respectively, demonstrated that the level of chimerism in both the marrow (p = .02) and FACS-purified CD34+ population (p = .01) was significantly higher in the immune-suppressed animals relative to controls. The proportion of donor-derived colonies formed from single CD34+ cells showed a trend toward higher levels in the immunosuppressed animals (12.0% ± 7.5% versus 5.0% ± 2.6%, respectively; p = .8) (Table 4). Nevertheless, the absolute number of donor cells in the marrow remained low.
Table Table 3.. Initial evaluation of chimerism in animals treated with immune suppression and delivered at terma
|Animal no.||BM (105)b||PBL (105)b||CD34+cell culturec|
Table Table 4.. Follow-up of chimerism (%) in marrow, progenitors, PBL, and FACS-purified PBL lineage cells
|Animal no.||Age (mo.)a||CFCb||CD34+c||BM||PBL||CD2+c||CD4+c||CD13+c||CD20+c|
Figure Figure 1.. Comparison of chimerism at birth in the hematopoietic progenitor population (single CD34+ cell cultures) in fetuses treated with (n = 3) and without (n = 7) immune suppression. The values are mean ± SEM (p = .06).
Download figure to PowerPoint
Influence of Fetal Immunosuppression on Peripheral Blood Chimerism
In addition to the higher levels of chimerism in both marrow and in marrow CD34+ progenitors in the immune-suppressed animals, there also was a trend toward higher levels of donor cells in the peripheral blood (p = .10) (Table 4). Even though the level of chimerism in peripheral blood was up to 10-fold higher in immune-suppressed than in control animals, the overall frequency of donor cells in peripheral blood was generally low (<1%) and would be unlikely to have clinical relevance. The one exception to this finding was in the CD13+ cell (5.2%) population in one animal (M01-088). This animal also had the highest level of progenitor cell chimerism. Unfortunately, at the time of that collection (19 months of age), this animal was euthanized for a bone lesion in one femur that was consistent with chronic osteomyelitis by pathological evaluation, and additional samples could not be obtained.
Influence of Postnatal Donor Cell Infusion on Chimerism
Three animals were treated postnatally with additional donor cell infusions to test the hypothesis that chimerism could be increased using this methodology [25–28]. The first animal (M00-025) received CD34+-enriched cells (2.2 × 109/kg and 3.6 × 107/kg CD3+ cells) in utero and was in the cohort of animals we have previously reported . The other two animals were from the group of animals that was treated with in utero immune suppression. Animal K00-025 had an initial level of chimerism in the progenitor population of 6.0% at birth, which subsequently declined to 1.3% at 8 months of age. Tolerance, demonstrated by the absence of a mixed lymphocyte culture (MLC) response to the sire, was noted at 1, 4, and 7 months of age. This animal received three monthly infusions of T cell–depleted marrow cells. The average CD34+ cell dose per infusion was 7.3 × 107/kg (total = 21.9 × 107/kg and 3.6 × 106/kg CD2+ cells). Chimerism in the progenitor population had increased to 11.5% by 1 month after the first donor cell infusion. However, 18 months after the third infusion, the level of chimerism decreased to a level that was similar to the preboost level (2.4%). In addition, MLC responses were similar for both the dam and sire (donor). The other two animals received two infusions from growth factor–stimulated CD34+-enriched marrow cells. The first and second infusion was separated by 1 month. The total cell dose for each animal was 2 × 107/kg CD34+ and 1.0 × 105/kg CD2+. Five-fluorouracil (50 mg/kg), although not at a level that would produce myeloablation, was given 7 days prior to donor cell infusions in an attempt to reduce endogenous hematopoiesis. Because enriched donor cells would be antibody-coated, the animals were pretreated with a single course of prednisone (2 mg/kg) 24 hours prior to donor cell infusions. Although there was an increase in the level of chimerism after the first donor cell infusion, by 6 months after the second donor cell infusion, chimerism in the progenitor compartment and the peripheral blood was similar to that noted before reinfusion therapy (Fig. 2).
Figure Figure 2.. Data from pre- and postnatal booster therapy in animals M01-157 (top) and M01-207 (bottom). The data are from preinfusion, 1 month postinfusion, and 6 months after the second (last) infusion.
Download figure to PowerPoint
- Top of page
- Materials and Methods
The major limitation of in utero hematopoietic stem cell transplantation in most species tested (mice, primates, and humans) has been low and nontherapeutic levels of chimerism. In the human fetus, the only conditions that have been successfully treated with in utero therapy have been X-linked severe combined immunodeficiency (SCID) and bare lymphocyte syndrome [7–9, 29]. Successful treatment of these disorders is likely related to the global immune defect, the lack of an effective immune response to allogeneic cells, and the proliferative advantage of transplanted genetically normal cells. In vitro evaluation of fetal alloimmune responses suggests that the fetus can mount an immune response very early in gestation. The human fetus produces a proliferative response to allogeneic stimuli by 11 weeks of age (0.28 gestation) [30–32]. Based on this information, it is reasonable to assume that the fetal immune system may be one of the factors limiting in vivo engraftment at the time of fetal transplantation (0.30 gestation). Although a variety of immunosuppressive approaches could have been selected, we chose the combination of thymoglobulin (ATG) and corticosteroid, based on the immunosuppressive effect seen postnatally. The immunosuppressive effects of ATG are through cell-mediated cytotoxicity, apoptosis, and complement-mediated mechanisms that result in rapid elimination of target cells from the circulation. Corticosteroids have well-established immunosuppressive effects.
The primary aim in these studies was to test the hypothesis that induced fetal immune suppression would have a positive influence on engraftment and chimerism. The preliminary studies presented in this report suggest that this hypothesis may be correct. It is important to note that we did not address the issue of different immunosuppressive regimens; neither did we attempt to determine if higher levels of immune suppression would produce more significant changes in chimerism, or if this approach would be toxic to the developing fetus. Despite these limitations, a number of important findings emerge from these experiments. First, in utero conditioning with immunosuppression resulted in high levels of fetal progenitor chimerism. Second, modest effects were noted in durable marrow engraftment, and modest nonsignificant changes were observed in peripheral blood chimerism. Third, chimerism in the CD34+ progenitor population, measured by both real-time PCR and single CD34+ cell culture, is significantly higher than the level of chimerism noted in the peripheral blood of the same animals. This last observation is similar to that reported in the fetal sheep model . It is also important to note that chimerism after in utero and postnatal treatment for X-SCID is almost exclusively limited to lymphoid compartments in which a selective advantage for donor cells exists, particularly within the T-cell compartment. These data would suggest that, in the absence of a competitive advantage, donor stem cell proliferation may be less efficient than that of recipient stem cells, resulting in limited expression of mature donor cell progeny.
The two animals that were delivered prior to viability demonstrated high levels of donor cell chimerism. The first animal delivered at 70 days (0.45 gestation) had evidence of donor cells that was restricted to the fetal liver, with no progenitors noted in samples from fetal marrow. This finding is consistent with results noted in fetal rhesus monkeys by Cowan et al. . In that species, donor cells were also confined to the fetal liver at 70 days (0.42 gestatation). At 70–80 days gestation, presumably with the transition from hepatic to marrow hematopoiesis, colony formation and donor cells were observed in both the fetal liver and marrow. Evaluation of chimerism in these two rhesus animals suggested that the upper estimate of donor cell engraftment in marrow progenitor cells was 6.4%–11% and 13% in progenitors from fetal liver (estimated ranges 0.64%–11% and 1.3%–13%, respectively). Direct comparisons to our model are limited by the difference in transplantation approach (T cell–depleted marrow versus CD34+-enriched cells) and species (M. mulata versus M. nemistrina). Even with these limitations, evaluation of the fetuses treated with immunosuppression in our study demonstrate high levels of donor cells within the progenitor population of fetal liver (17% at day 70) and marrow (34%–43% at day 120), suggesting that there may be an initial engraftment advantage with the use of fetal immunosuppression. However, in the absence of a continued growth or proliferative advantage, the contribution to marrow and peripheral blood in immune-suppressed animals, relative to controls, was very modest.
The second question we attempted to address in this report was the utility of postnatal donor cell infusion as a method to enhance chimerism. Several reports suggesting the utility of this therapy have been published in the mouse model of in utero transplantation. Enhanced tolerance and clinical effect after postnatal boost with donor cells have been demonstrated after in utero stem cell therapy in fetal β-thalassemic mice . Similar results have been reported in nondefective mice after donor hematopoietic cell infusions and after donor lymphocyte infusions [34, 35]. However, others have noted that changes in the level of chimerism after postnatal “booster” infusions did not produce durable changes in chimerism . This same group has also suggested that microchimerism after in utero therapy may produce sensitization to the donor and that postnatal donor cell infusions produce an immunological effect that is similar to booster immunization . In our model, the three animals treated with postnatal donor cell infusions demonstrated an initial enhancement in the level of donor cells in blood, marrow, and progenitors. However, these changes in the level of chimerism did not persist, and after 6 months the level of chimerism was not different from that observed prior to the first infusion. All three of the animals continue to show evidence of donor cells at levels similar to their pretreatment levels. In one of the animals we noted a change in the degree of MLC tolerance after postnatal therapy (from nonreactivity to donor to reactivity similar to that of the dam). Eighteen months after the third postnatal boost, the level of progenitor chimerism in this animal, determined from single CD34+ cell cultures, was the same as the preboost level (2.4%). If there was a change in the level of tolerance, it did not result in complete graft loss.
It has been suggested that the ability to enhance postnatal chimerism by donor cell infusions may be related to the level of chimerism. Hayashi et al.  noted that enhanced chimerism after donor lymphocyte infusions was only achieved when peripheral blood chimerism was above the microchimeric level. In this mouse model, nonreactivity in mixed lymphocyte culture was not seen unless peripheral blood chimerism was less than 1%–2%. The authors noted that transplantations (15 days gestation) occurred prior to the completion of thymic processing, the appearance of single positive CD4 and CD8 T cells, and, presumably, self-recognition. In our model, progenitor chimerism is almost always above 1%–2% when measured by both colony assays and real-time PCR of CD34+ cells. However, when the total donor cell contribution to marrow and peripheral blood cells is evaluated, the level of chimerism is low and consistent with microchimerism (<1.0%). Despite this overall low level of donor cells in peripheral blood and in the marrow, we consistently note nonreactive or markedly reduced mixed lymphocyte reactions. The lack of a high level of peripheral chimerism and the presence of tolerance after in utero transplantation has also been noted by others . It is unclear if higher levels of peripheral blood chimerism in our model would produce results that are more consistent with the data seen after postnatal reinfusion in the mouse model.
Based on the published experience from mice, sheep, primates, and humans, the greatest limitation of in utero therapy appears to be providing an environment that produces a distinct proliferative advantage to donor stem cells and their early progeny. In the human setting of fetal transplantation, it is clear that traditional conditioning regimens for postnatal stem cell transplantation will not be used due to unacceptable risk to the fetus. If advancement in the field of fetal hematopoietic transplantation is going to occur, transplantation protocols must be developed that produce an environment that provides a significant growth advantage to donor allogeneic cells without GVHD. At this time, the most promising approach for enhancing engraftment of donor cells in the fetus may be the use of other hematopoietic facilitating cells. Candidate cells include dendritic cells, ultraviolet (UV)–treated or irradiated T-cells, mesenchymal stromal cells (MSCs), and alloreactive natural killer (NK) cells.
In fetal mice, cotransplantation with dendritic cells and UV-treated T cells improved engraftment, but both were associated with unacceptable GVHD [39, 40]. In our nonhuman primate model, engraftment is enhanced by donor T cells, but unacceptable GVHD develops with T-cell dosages greater than 1 × 108/kg . Similar results have been noted in fetal sheep and in humans [41, 42]. In one animal we cotransplanted γ-irradiated T cells (1.3 × 1010/kg). The level of chimerism after birth was similar to our other animals and without GVHD (1.1%–1.2% in the progenitor population and less than 1% in both the marrow and peripheral blood; unpublished data). Based on available data, the use of large numbers of T cells or modified T cells does not appear to be a viable option for enhancing donor chimerism. More compelling data for the use of facilitating cells comes from the fetal sheep model in which cotransplantation of matched MSC and donor hematopoietic stem cells was associated with improved levels of chimerism and, more importantly, with higher levels of donor cells within the peripheral blood [33, 43]. The mechanism for this enhanced effect is not clear and may be related to MHC class I stromal stem cell interactions  or immunoregulatory effects of MSC [45, 46]. Alloreactive NK cells also appear to be a potential source of facilitating cells for in utero transplantation. In postnatal mice, alloreactive NK cells have been shown to enhance engraftment and, more important, reduce the risk of GVHD [47, 48]. If in utero transplantation ever becomes a clinical reality beyond the rare cases of severe immunodeficiency syndromes, it will require testing of new combinations of donor cell preparations and nontoxic forms of conditioning that enhance donor cell proliferation and expression of mature progeny in the peripheral blood.