Engraftment of Cultured Human Hematopoietic Cells in Sheep

IF IT WERE POSSIBLE to increase the number of hematopoietic stem cells in vitro, it would have a profound impact on future allogeneic and autologous stem cell transplantation. Therefore, the subject of in vitro (ex vivo) expansion of hematopoietic stem cells is currently one of the major emphases in hematology-oncology fields.1-3 Many investigators have documented that it is possible to increase in culture the number of hematopoietic progenitors such as colony-forming cells and long-term culture initiating cells (LTC-IC).1-3 Although some investigators attempted to expand the population of murine stem cells with in vivo reconstituting ability, the results to date remain controversial.4-9 In general, these investigators used the cytokines that have been shown to regulate cycling of primitive progenitors, such as c-kit ligand (KL), flt3/flk-2 ligand (FL), interleukin-1 (IL-1), IL-3, IL-6, IL-11, and granulocyte-macrophage colony-stimulating factor (GM-CSF).10 However, observations in our laboratory indicated that both IL-3 and IL-1 negatively affect in vitro expansion of stem cells with long-term engraftment capability.11 12 

We describe here our initial attempt to expand in culture human hematopoietic stem cells with in vivo engrafting capabilities. Because in vitro assays for human hematopoietic stem cells are not available, we used human/sheep xenogeneic transplantation as an assay for human stem cells.13 We earlier documented that human hematopoietic stem cells with long-term engrafting capabilities are in the c-kit^low population of the CD34⁺ marrow cells using the sheep/human xenograft model.14 Therefore, we cultured CD34⁺c-kit^low bone marrow cells in serum-free medium supplemented with KL, FL, and IL-6. We show here that human bone marrow cells cultured in the presence of combinations of cytokines can engraft sheep fetuses for at least 2 months posttransplantation. The sheep/human xenograft model is a useful assay for in vitro manipulation of human hematopoietic stem cells.

Recombinant human KL and FL were provided by Dr S. Lyman (Immunex, Seattle, WA). Recombinant human IL-3 was supplied by the Genetics Institute (Cambridge, MA). Recombinant human IL-6 was a gift from Dr M. Naruto (Toray Industries, Yokohama, Japan). Recombinant human erythropoietin (EP) was provided by Dr F.-K. Lin (Amgen Biologicals, Thousand Oaks, CA).

Aliquots of the human cells were cultured in 35-mm Falcon suspension culture dishes (Becton Dickinson Labware, Lincoln Park, NJ) using the methylcellulose culture technique described previously.14The culture medium consisted of α-medium, 1.2% 1,500-centipoise methylcellulose (Shinetsu Chemical, Tokyo, Japan), 30% fetal calf serum (FCS; Intergen Corp, Purchase, NY), 1% bovine serum albumin (BSA), 100 U/mL IL-3, 100 ng/mL IL-6, 100 ng/mL KL, and 2 U/mL EP. Dishes were incubated at 37°C in a humidified atmosphere of 5% CO₂/95% air (vol/vol). On day 14 of culture, colonies were scored on an inverted microscope.

Bone marrow CD34⁺c-kit^low cells were prepared as described previously.15 One day after cell separation, portions of these cells were injected into sheep fetuses as fresh cells. The remainders of the cells were suspended in α-medium (ICN, Irvine, CA) containing 2% deionized, crystallized BSA (Sigma, St Louis, MO), 300 μg/mL iron-saturated human transferrin (Sigma), 6 μg/mL cholesterol (Sigma), 10 μg/mL lecithin (Sigma), 1 × 10⁻⁷ mol/L sodium selenite (Sigma), 10 μg/mL insulin (Sigma), and a combination of 100 ng/mL KL, 100 ng/mL FL, 100 ng/mL IL-6, 100 U/mL IL-3, and 2 U/mL EP. The cells were incubated at 37°C in a humidified atmosphere of 5% O₂/5% CO₂/90% N₂ (vol/vol). The media was changed once on day 7. After 2 weeks of incubation, the cells were harvested and fractions were injected into sheep fetuses.

Donor cells were injected into preimmune fetal sheep recipients using the amniotic bubble procedure described previously.13,16 17Both fresh cells and cultured cells were resuspended in α-medium containing 20% FCS and 50 U/mL each of recombinant human IL-3 and GM-CSF and transferred to Reno, NV by overnight mail. Two thousand (experiments no. 1 and 2) to 3,000 (experiment no. 3) fresh cells or the whole cells expanded from 3,000 (experiments no. 1 and 3) to 6,000 (experiment no. 2) fresh cells were transplanted intraperitoneally into each fetus at the gestational ages of 58 to 63 days. Some of the recipient fetuses were killed on day 60 posttransplantation, and the bone marrow cells were analyzed for the presence of human cells. The remainders were allowed to be born and were examined monthly for signs of engraftment. In other experiments, CD45⁺ cells were collected by panning from pooled fetal samples that were positive for human cells and retransplanted into preimmune fetal sheep at 0.4 to 3.0 × 10⁶cells/fetus. Secondary recipient fetuses were killed on day 60 posttransplantation, and the bone marrow cells were analyzed for signs of human cell engraftment.

Bone marrow mononuclear cells from the fetal and newborn sheep transplanted with human cells were analyzed for the presence of human cells by flow cytometry (CD45⁺ cells), karyotype analysis, and clonal culture assays, as described previously.16 17CD45⁺ cells at 0.1% were considered positive.

Although there were significant differences in the magnitude of expansion among the three experiments, the total nucleated cell counts (TNCC) increased after 14 days of incubation with KL, FL, IL-6, and EP (Table 1). Progenitor populations also expanded except in one culture without IL-3. In all three experiments, IL-3 enhanced production of both TNCC and progenitors.

Table 2 shows engraftment of human cells in primary recipient fetuses on day 60 posttransplantation. In the three experiments, 8 of the total of 13 recipients injected with fresh cells became chimeric. The frequencies of CD45⁺ cells in the bone marrow ranged from 0.42% to 5.3%. Of the total of 22 fetuses receiving cultured cells, 15 fetuses became chimeric. In this group, CD45⁺ cells in the bone marrow ranged from 0.4% to 7.8%. All fetal samples that were positive for CD45⁺ cells contained multipotential progenitors (colony-forming units-mix [CFU-Mix]) and/or granulocyte-macrophage progenitors (colony-forming units–granulocyte-macrophage [CFU-GM]) of human origin. Both the frequencies of the chimeric fetuses and the levels of CD45⁺ were very similar between the fresh cells group and the cultured cells group. The presence of IL-3 in the suspension culture did not have consistent influence on the levels of engraftment. These data indicated that the ability of cells to support engraftment in vivo for 2 months is maintained by the cells in culture for 2 weeks with or without IL-3.

Some of the fetuses injected with fresh cells and those injected with cultured cells were allowed to be born. They were examined monthly for signs of engraftment. The results are presented in Table 3. In many animals injected with fresh cells, CD45⁺ cells were detected as late as 3 months after birth, which is approximately 6 months posttransplantation. In contrast, only 2 animals injected with the cultured cells showed signs of engraftment by human cells 1 week after birth. No animals showed the presence of CD45⁺ cells 1 month after the birth of the animals.

The analysis of the newborn animals presented in Table 3 may be interpreted to suggest that true stem cells with long-term engrafting capabilities are not maintained under current cell culture conditions. This hypothesis was tested in another transplantation model, namely transplantation to secondary recipients. The primary recipient fetuses were killed 60 days after the first transplantation and the secondary recipient fetuses were killed again 60 days later. Therefore, the total span of observation was 120 days after the first transplantation of the cells. The results of the positive samples are presented in Table 4. The frequencies of chimeric animals were significantly (P < .01 by the Student'st-test) higher in the fetuses transplanted with fresh cells than in those transplanted with cultured cells. Of the 13 secondary recipients tested for engraftment by fresh cells, a total of 10 recipients showed the presence of CD45⁺ cells. The levels of CD45⁺ cells in the bone marrow of these fetuses ranged from 1.1% to 7.2%. In contrast, of the total of 27 secondary recipient fetuses tested for engraftment by cultured cells, only 3 recipients showed signs of engraftment. The levels of the CD45⁺ cells in the bone marrow ranged from 0.3% to 1.3%. There were statistically significant differences in the levels of CD45⁺ cells between the fresh cells group and the cultured cells group in experiment no. 3. However, the number of cells injected did not seem to be a significant factor in whether the animals became chimeric. For instance, the number of CD45⁺ cells from donors receiving fresh cells and injected into secondary fetuses was 0.4 to 3 × 10⁶ cells/fetus, with an equal distribution of positives and negatives between the low and high doses. Ten of 13 of these animals were positive. The amount of CD45⁺ cells administered to secondary fetuses from animals that received cultured cells was 1.5 to 3.4 × 10⁶cells/fetus, and only 3 of 27 animals were positive. This study was consistent with the results presented in Table 3 and indicated that long-term engraftment capability of the stem cells is impaired during 2 weeks of incubation under current culture conditions.

Stem cell transplantation provides definitive therapy for a variety of malignant and inherited diseases. Recently, transplantation of stem cells harvested from the umbilical cord blood became an important therapy for leukemia and other malignancies of children. However, the limited quantity of cord blood harvest is thought to be a serious obstacle for use of cord blood stem cell transplantation in older children and adults. In vitro expansion of the population of stem cells would significantly extend the indication of cord blood stem cell transplantation. Furthermore, gene transduction into hematopoietic stem cells in gene therapy may be facilitated under similar culture conditions. For these reasons, in vitro expansion of human hematopoietic stem cells is an important current research interest. Already, several investigators have shown that it is possible to increase in culture the number of human hematopoietic progenitors, such as colony-forming cells and LTC-IC, but none studied engraftment potentials of the cultured human cells.18-24 

Recently, Brugger et al25 and Williams et al26reported transplantation of cultured autologous cells after high-dose chemotherapy of the patients. Because of the autologous nature of the transplantation, contributions by the cultured cells to the hematopoietic reconstitution could not be quantitated. We used the in utero human/sheep xenograft model to assay the engrafting capability of cultured cells. It appears that this system is sufficiently reproducible and quantitative and demonstrated repeatedly the engrafting capabilities of the freshly prepared cells. The study showed that short-term (2 months) engraftment capability of the human hematopoietic progenitors is maintained during the 2-week incubation period. We earlier documented the negative effects of IL-3 on the in vitro maintenance of long-term engrafting capabilities of murine hematopoietic stem cells.11 12 Therefore, we wished also to test if human IL-3 possesses negative effects on human hematopoietic stem cells. IL-3 did not affect negatively the short-term engraftment capability of the cultured cells. Unfortunately, only 1 of 6 animals transplanted with cultured cells showed chimerism at 4 months after cell transplantation and only 3 of the 27 secondary transplantation recipients of cultured cells became chimeric. In addition, the levels of chimerism were significantly lower than those of fresh cells group. Therefore, we are not certain of the long-term engrafting capability of cultured cells or the effects of IL-3 on human stem cell expansion.

After submission of this manuscript to Blood, investigators in two laboratories reported their studies of in vitro expansion of transplantable cord blood stem cells using nonobese diabetes/SCID (NOD/SCID) mice. Bhatia et al27 observed a twofold to fourfold increase in the SCID-repopulating cells after a 4-day suspension culture of CD34⁺CD38⁻ human cord blood cells. Their cytokines consisted of KL, FL, IL-3, and IL-6. Conneally et al28 performed a suspension culture of CD34⁺CD38⁻ human cord blood cells in serum-free culture containing KL, FL, IL-3, IL-6, and G-CSF for 5 to 8 days and observed the statistically significant twofold increase in the competitive repopulating units. Although the cytokine combinations used were similar, there are significant differences between their and our studies. We used adult human marrow cells rather than cord blood cells and our assay was in utero transplantation to sheep fetuses rather than transplantation to NOD/SCID mice. In addition, our suspension culture was significantly longer than that of the reported studies. Nonetheless, their results and our observations are in agreement in that the suspension culture with similar cytokine combinations can maintain cells that are capable of engrafting xenogeneic hosts for up to 2 months. However, our studies suggested that the current culture system may not be able to maintain longer term engrafting cells. Further studies are needed for the determination of precise combinations and concentrations of cytokines for the optimal expansion of the stem cells that are capable of long-term engraftment. Both the NOD/SCID murine model and in utero transplantation to sheep may complement each other and be useful in this endeavor.

The authors thank Dr Haiqun Zeng for assistance in cell sorting and Dr Pamela Pharr and Anne G. Leary for discussion and preparation of this manuscript.