This invention was made with government support under NIH Grant HL39875- 04. The government has certain rights in the invention.

Field of the Invention

The invention relates to improved postnatal and in utero fetal hematopoietic stem cell (HSC) transplantation methods. The invention further relates to the ex vivo expansion of fetal HSC. The invention also relates to methods for confirming the presence of expanded fetal HSC. Finally, the invention relates to methods for transducing fetal HSC with retrovirus vectors.

Background of the Invention

Hematopoiesis, the production of mature blood cells, is a complex scheme of multilineage differentiation. Mature blood cells are derived from pluripotent HSC. The defining characteristics of HSC are the capacity for extensive self- renewal and retention of multilineage differentiation potential. HSC proliferate and differentiate to produce progenitor cells, which in turn form precursor cells, which differentiate to form mature blood cells.

During ontogeny, hematopoiesis moves consecutively from yolk sac to liver/spleen and then to the bone marrow (Tavassoli, M., Blood Cells 17:269 (1991)). During most of fetal life hematopoiesis occurs within the liver and spleen. In the latter part of gestation, bone marrow spaces begin to develop and expand. HSC then migrate from liver/spleen to the bone marrow occupying "niches" in the developing marrow (Zanjani et al., J. Clin. Invest. 59.1178 (1992)).

Postnatal allogeneic bone marrow transplantation (BMT) has been used successfully to treat a number of congenital hematopoietic and metabolic disorders (Sullivan, K.M., Transplant Proc. 21 (Suppl. 1):41 (1989)). However, conventional postnatal BMT has significant limitations. Only 35% of candidates for BMT have a human leukocyte antigen (HLA)-identical family member. This precludes an optimal donor for the majority of patients in need (Flake et al., Exp. Hemat. 79.1061(1991). In general, results using HLA-mismatched donors have been disappointing (O'Reilly et al., Immunodef. Rev. 7:23 (1990); Anasetti et al., N. Engl. J. Med. 320:191 (1989); Ferrara et ai, N. Engl. J. Med. 324:661 (1991)). Recipient tolerance is generally induced by marrow conditioning regimens such as radiation and/or cytoablative therapy (Sullivan, K.M., Transplant Proc. 21 (Suppl. 7):41-50 (1989)). These treatments suppress the immune system and deplete the endogenous stem cell pool thereby permitting expansion of donor HSC in absence of endogenous stem cell competition. However, the attendant risks of such conditioning regimens include immunosuppression and overwhelming infection. Also, rejection and graft vs host disease (GNHD) with bone marrow transplantation (BMT) remain a constant threat. The combination of preexisting disease and harsh conditioning regimens results in high morbidity and mortality for the pediatric recipient (Clark, J.G., Mayo Clin. Proc. 65:111 (1990); Parkman, R., Science 232:1313 (1986)). However, despite these complications, postnatal transplantation using HLA-mismatched HSC from adult marrow remains the sole option for treating numerous disorders.

HSC transplantation is also emerging as in utero therapy for the fetus with prenatally diagnosed disorders of hematopoiesis (Flake et al., Exp Hematol 79.1061 (1991); Touraine, J-L, Blood Cells 77:379-387 (1991)). This approach exploits immunogenical fetal "naivety" during the first and second trimester. Studies of fetal ontogeny suggest that human fetuses are tolerant of foreign antigen and allow engraftment of allogeneic stem cells without rejection and without immunosuppression (Brent et al. , in: Albertini A, Crosignani PG, (Eds.) Progress in Perinatal Medicine. Amsterdam: Excerpta Medica, p. 3 (1983)). In addition, first trimester fetal bone marrow is relatively empty, awaiting the migration of

HSCs from the fetal liver. Thus, reconstitution of the hematopoietic system could potentially be achieved without the morbidity of bone marrow ablation. In the case of certain glycogen storage diseases, neurologic injury occurs prior to birth (Krivit et al. , N. Engl. J. Med. 322:28 (1990)), making prenatal treatment not only advantageous but necessary to avoid devastating sequelae of the disease.

Currently, HSC used for clinical transplantation in the treatment of malignancy are usually harvested from adult bone marrow or peripheral blood. However, in addition to the shortage of compatible family donors, autologous marrow is frequently damaged from prior anti-cancer treatment or contaminated with malignant cells (Gale et al , Bone Marrow Transplant 7:153 (1991)). The shortage of suitable marrow HSC for clinical transplantation is even more problematic when multiple cycles of chemotherapy are required. Thus, there has been much interest in the expansion of bone marrow cells ex vivo prior to transplantation. Dexter-type long term bone marrow cultures (LTBMCs), first developed in the mid-1970s, produced stable ex vivo hematopoietic systems for several months (Dexter et al , J. Cell Physiol. 97:335 (1977); Dexter et al , in: Wright, D.G., Greenberger, J.S. (Eds.): Long-term Bone Marrow Culture. New York, NY, Liss, p. 57 (1984)). However, human LTBMCs have always exhibited exponentially decreasing numbers of total and progenitor cells with time, rendering the cultures unsuitable for cell expansion (Eaves et al. , J. Tissue Cult. Methods 13:55 (1991)). As enrichment protocols for primitive hematopoietic cells became available, many thought that enriched primitive cells could be expanded in cultures. Several groups reported incubating enriched cells in suspension culture with cytokines (Haylock et al , Blood SO: 1405 (1992); Brandt et al , Blood 79:634 (1992)). Large cell expansion numbers were often obtained. However, cellular differentiation, accompanied by depletion of primitive cells, occurs in these systems. In fact, when long-term culture-initiating cells (LTC-IC) have been measured, the numbers obtained after culture of enriched primitive hematopoietic cells have always been below the input value (Sutherland et al. , Blood 78:666 (1991); Verfaillie, CM., Blood 79:2821 (1992)). However, it has been stated that

hematopoietic cells must be enriched prior to culture for ex vivo expansion to succeed (Edgington, S.M., Biotechnology 10:1099 (1992)).

In contrast, mononuclear cell (MNC) populations obtained from adult marrow stroma without enrichment and cultured by continuous perfusion result in the expansion of progenitor cells (Koller et al, Biotechnology 77:358 (1993); Palsson et al, Biotechnology 77:368 (1993)). A large-scale perfusion bioreactor system has been used to expand cells from MNC populations obtained from adult marrow stroma donors (Koller et al, Blood 82:318 (1993)). However, this study is limited because the cells were characterized only by the LTC-IC assay, and not by phenotype analysis or an in vivo model of engraftment. Thus, it is uncertain whether Koller et al. expanded pluripotent stem cells. The LTC-IC, high proliferative potential-colony-forming cells (HPP-CFC), and blast colony-forming cells (CFU-B1) assays identify a primitive hematopoietic cell (Williams, D.A., Blood 81(12):3169 (1993)). However, the primitive cells identified by these assays include not only HSC but also more mature progenitor cells. Therefore, the extent of HSC identification may be overstated when these assays are used for characterization.

HSC used for clinical transplantation have also been harvested from fetal liver. Fetal liver HSC may have homing and proliferative advantages over adult derived cells for clinical transplantation (Fleischman et al, J. Exp. Med. 159:131 (1984)). In addition, the absence of post-thymic, mature T-cells in fetal liver may reduce the risk of GVHD (O'Reilly et al. in Seligman, M., Hitzig, W.H., (Eds.), Primary Immunodeficiencies, New York: Elsevier/North-Holland, p. 419 (1988)). However, the use of fetal liver HSC is limited by ethical objections, microbial contamination, and high cost of tissue processing (Rice et al. , Fetal Diag. Ther. 8:14 (1992)). Many of these difficulties could be overcome by the ex vivo expansion of fetal HSC. However, to date, methods for the ex vivo expansion of fetal HSC have not appeared in the literature.

As indicated, stem and progenitor cells (as characterized by the LTC-IC assay) have been expanded ex vivo without prior enrichment from adult marrow stroma (Koller et al. , Blood 82:318 (1993)). However, the microenvironments

of marrow stroma and fetal liver stroma are considerably different. Hematopoiesis in the bone marrow microenvironment proceeds under the control of many non- hematopoietic support cells (mainly marrow fibroblasts and endothelial cells) and the associated extra-cellular matrix. This complex arrangement of supportive tissue is collectively called the marrow stroma (Koller et al, Biotechnology 77:358 (1993)). In contrast, the fetal liver is predominantly comprised of cells committed to the erythroid lineage (Meunch et al, Exp. Hematol. 27:1013 (1993)). Therefore, it was, at best, speculative whether results obtained using adult marrow stroma would extrapolate to the fetal liver system.

Summary of the Invention

Fetal liver HSC have been used in transplantation therapies for patients suffering from a variety of disorders. However, these therapies are limited by the shortage of fetal liver HSC due to ethical objections to using fetal tissue, microbial contamination, and the high cost of fetal tissue processing. This limitation could be overcome if an immediately accessible source of fetal liver HSC was available.

This goal has been realized by the present invention which provides, for the first time, a method for expanding fetal liver HSC ex vivo in an artificial capillary system (ACS) cartridge. Moreover, by the invention, expansion of fetal HSC is achieved without HSC enrichment prior to culture. The method involves inoculating fetal liver cells into an ACS cartridge, perfusing the ACS cartridge with at least one hematopoietic growth factor, and culturing the fetal liver cells for a sufficient time to achieve expansion of fetal HSC.

The invention is also directed to improved postnatal and in utero fetal HSC transplantation methods. For postnatal transplants, the invention involves harvesting fetal liver cells which have been cultured as described above and transplanting the cells to a patient suffering from an immunodeficiency disease, an inborn error of metabolism, aplastic anemia, leukemia, or hemoglobinopathy. For

in utero transplants, the invention involves harvesting fetal liver cells which have been cultured as described above and transplanting the cells in utero into a fetal patient suffering from at least one of the above disorders.

The invention further provides a xenograft system for unambiguously confirming the presence of fetal HSC after ex vivo expansion. The xenograft system confirms the presence of expanded fetal HSC by measuring engraftment and multilineage reconstitution in an animal model. After expansion in an ACS cartridge and harvesting, donor fetal liver cells containing HSC are transplanted in utero to a non-human fetal mammalian recipient. The extent of donor cell engraftment in the recipient animal confirms the presence (or absence) of donor fetal HSC and can be assessed before or after birth. Methods for determining fetal HSC engraftment include: karyotypic analysis of the recipient's peripheral blood, bone marrow, or liver hematopoietic cells; cytometric analysis of differentiation antigens associated with hematopoietic subpopulations; and fluorescence in situ hybridization (FISH) with DNA probes specific for repeated satellite sequences on donor autosomes.

The invention also provides a method for transducing ex vivo expanding fetal HSC with a packaged recombinant retrovirus vector. The method involves inoculating an ACS cartridge with fetal liver cells and a suspension containing packaged retrovirus vectors, perfusing the cartridge with at least one hematopoietic growth factor, culturing the cells in the presence of the packaged recombinant retrovirus vector for a sufficient time to achieve expansion of fetal HSC, and harvesting cultured fetal liver cells containing expanded HSC which have been transduced with the recombinant retrovirus vector. The retrovirus vector can contain a heterologous gene encoding a therapeutically effective product. After expansion and harvesting, the transduced fetal HSC can be administered to a patient using the transplantation methods of the invention.

Brief Description of the Drawings

Figure 1. Growth curve of single culture of human fetal liver cells in artificial capillary system as measured by daily lactate production. Inset arrows represent time of harvest of cells from extracellular space of the culture cartridge. Figure 2: Flow cytometric analysis of human fetal liver cells after culture in artificial capillary system. Panel A-Forward vs. orthogonal scatter plot of human fetal liver cells to represent population of cells gated for analysis. Panel B-Dual color scatter plot analysis with FITC and PE isotype controls of human fetal liver after 10 days of culture in artificial capillary system. Panel C-Dual color scatter plot analysis with FITC-CD34 and PE-HLA-DR monoclonal antibodies of human fetal liver after 10 days of culture.

Figures 3(A) and 3(B): Immunohistochemical stains with anti-CD-34 monoclonal antibody staining of human fetal liver at baseline (Figure 3 A) and after 10 days of culture (Figure 3B) in an artificial capillary system. CD34 ⁺ cells stain red with this antibody system.

Figure 4: Giemsa stain of human fetal liver after 20 days of culture in an artificial capillary system. Note the maturation pattern which demonstrates immature and mature myeloid and lymphoid elements.

Figure 5: Summary of progenitor colony analysis of human fetal liver at baseline and at intervals after culture in an artificial capillary system. Data are expressed as mean ± standard deviation for five human fetal liver samples for CFU-E, CFU-GM, and BFU-E.

Figure 6: Flow cytometric analysis of fetal sheep liver. Panel A-Forward vs. orthogonal scatter plot of sheep liver to represent population of cells gated for analysis. Panel B-Frequency histogram of normal fetal sheep liver stained with FITC-conjugated anti-CD45. Panel C-Similar frequency histogram of liver from sheep 60 days after transplant with human fetal liver cells which had been expanded for 5 days in artificial capillary system.

Detailed Description of the Invention

The invention provides a method for the ex vivo expansion of fetal liver hematopoietic stem cells (HSC). The method involves inoculating fetal liver cells into an artificial capillary system (ACS) cartridge, perfusing the ACS cartridge with a medium containing at least one hematopoietic growth factor, and culturing the fetal liver cells for a sufficient amount of time to achieve expansion of fetal HSC. The inventors have discovered that fetal HSC can be expanded using an ACS without enrichment for HSC prior to culturing. This is surprising since others have suggested that hematopoietic cells must be enriched before culture for the expansion to succeed (Edgington, S.M., Biotechnology 10:1099 (1992)).

The invention encompasses expanding fetal HSC from mammals.

Preferably, fetal HSC are obtained from human fetal livers. More preferably, fetal

HSC are obtained from the livers of fetuses prior to 30 weeks' gestation; still more preferably prior to 20 weeks' gestation; and most preferably prior to 15 weeks' gestation.

Fetal liver cells are harvested from the livers of electively aborted fetuses according to conventional techniques. For example, livers can be dissected from fetuses following transvaginal extraction. A single-cell suspension of the fetal liver mononuclear cells can be created using a glass homogenizer. The single-cell suspension containing fetal liver cells is then ready for inoculation into an ACS cartridge.

Hematopoietic growth factors capable of stimulating expansion of fetal liver HSC are described in the literature (Moore et al. , Proc. Natl. Acad. Sci. USA 84:1134 (1987); Leary et al. , Blood 71:1159 (1988); Brandt et al , J. Clin. Invest. 56:932 (1990); Kobayashi et al , Blood 78:1941 (1991); Meunch et al , Blood 57:3463 (1993); Bernstein et al , Blood 77:2316 (1991); and Bodine et al , Blood 79:913 (1992)). In particular, the following growth factors can be used to stimulate expansion of fetal HSC: interleukin-1 (IL-1), interleukin-lα- (IL-lα), interleukin- 1 β (IL- 1 |S) , granulocyte colony-stimulating factor (G-CSF) , granulocy te- macrophage colony-stimulating factor (GM-CSF), interleukin-3 (IL-3),

interleukin-6 (IL-6), interleukin-11 (IL-11), erythropoietin, and stem cell factor (SCF). These hematopoietic growth factors are readily available and can be used alone or in combinations of two or more. Preferably, the growth factors used for stimulating expansion of fetal liver HSC are SCF and IL-3. The fetal liver cells are cultured in the ACS cartridge for a sufficient amount of time to achieve expansion of fetal HSC. The amount of time that is "sufficient" to achieve fetal HSC expansion can easily be determined empirically. For example, the phenotype analysis techniques described in Example 1 are useful for determining the amount of culturing time required to achieve fetal HSC expansion. The inventors have discovered that culturing for 5-10 days achieves significant fetal HSC expansion. However, it will be recognized that different time-frames can be used depending on the exigencies of each experiment.

Any culture medium recognized in the art as appropriate can be used to perfuse the ACS cartridge. In addition to one or more hematopoietic growth factors, the culture medium can be supplemented with serums such as fetal calf serum (FCS) and antibiotics such as Penicillin, Streptomycin, and Amphotericin B. A particularly suitable medium is Isocove's Modified Dulbecco's Medium (IMDM). Complete AIM-N is another example of a usable tissue culture medium. Complete AIM-N consists of the proprietary formula AIM-V (Gibco, Grand Island, New York) and also contains 10 μg/ml Gentamicin, 50 μg/ml Streptomycin, 50 μg/ml Penicillin, and 1.25 μg/ml Fungizone. IMDM and complete AIM-N are readily available as are other tissue culture media. As indicated, the culture medium will also contain at least one hematopoietic growth factor. Hematopoietic growth factors should be included in the culture medium at concentrations of approximately 0.1-500 ng/ml; preferably at concentrations of approximately 1-100 ng/ml; more preferably at concentrations of approximately 1-50 ng/ml; and most preferably at concentrations of approximately 5-25 ng/ml. However, as the skilled artisan will recognize, other concentrations can be used as required.

Expanded fetal HSC are easily harvested from an ACS cartridge simply by flushing the cartridge' s extra-capillary space (ECS) with fresh medium or by gently shaking the cells from the cartridge. A small number of cells remain behind which

can re-seed the bioreactor for further culture. By "extra-capillary space (ECS)" is intended the space wherein the fetal HSC grow within the shell of the ACS that is external to the semi-permeable fibers. After harvesting, the suspension containing expanded fetal HSC can be pelleted by centrifugation (400-600 g) and the cells resuspended at desired concentrations in culture medium for further use. An ACS cartridge consists of an outer shell casing that is biocompatible with the growth of mammalian cells, a plurality of semi-permeable hollow fibers encased within the shell that are also biocompatible with the growth of mammalian cells on or near them, and the ECS, which contains the cells and the ECS cell supernatant.

Tissue culture medium flows within the fiber lumens and is also included within the shell surrounding the fibers. The tissue culture medium, which may differ in these two compartments, contains diffusible components that are capable of expanding fetal HSC. The medium is provided in a reservoir from which it is pumped through the fibers. The flow rate can be controlled by varying the applied pressure.

ACS are described in Knazek et al , U.S. Patent Nos. 4,220,725, 4,206,015, 4,200,689, 3,883,393, and 3,821,087. A typical ACS consists of a standard glass media bottle, which serves as the reservoir, a pump, a hollow fiber bioreactor, which consists of the fibers and shell casing in which cells are cultured, and medical grade silicone rubber tubing, or other connecting means, which serves as a gas exchanger to maintain the appropriate pH and PO ₂ of the culture medium. All components are secured to a tray of sufficiently small dimensions to fit within a standard tissue culture incubator chamber. The pump speed is determined by an electronic control unit which is placed outside of the incubator and is connected to the pump motor via a flat ribbon cable which passes through the gasket of the incubator door. The pump motor can be magnetically coupled to the pump, being lifted from the system prior to steam autoclaving. The pump motor can also drive a cam shaft which moves a pin means which, in turn, compresses tubing and causes unidirectional flow through a series of one-way valves.

Tissue culture medium, which may, for example, include growth promoting substances, such as IL-3, and/or recombinant growth factors or vectors, is drawn from the reservoir, pumped through the gas exchange tubing in which it is reoxygenated and its pH readjusted and then through the lumina of the hollow fibers prior to returning to the reservoir for subsequent recirculation. The order of sequences may be altered without substantially changing the functionality.

The flow rate can be increased as the number of cells increases with time. Typically the initial flow rate of the medium is adjusted to about 4 ml/min. The entire system is sterilized prior to cell inoculation and is designed for operation in a standard air-CO ₂ tissue culture incubator. Upon inoculation through the ACS side ports, the cells settle onto the surface of the hollow fibers, through walls nutrients pass to feed the cells and through which metabolic waste products pass and are diluted into the large volume of the recirculating perfusate. The selected fiber should be semi-permeable to permit the passage of nutrients into the EFS and should be of a material on which or in the vicinity of which the cells are able to grow. The fibers are made of material, such as cellulose diacetate or polypropylene, that is semi-permeable or porous and suitable for the growth of mammalian cells. It may be necessary to treat the surface of certain types of fibers with reagents to enable some cells to adhere to the surface. For example, cellulose hollow fibers 15 cm in length, whose walls nominally restrict diffusion to substances having a molecular weight less than 80,000 Daltons are suitable for use in practicing this invention.

The ex vivo expansion of fetal liver HSC has a variety of uses. These include providing a rich source of transplantable cells, facilitating retro viral transduction of HSC, and in vitro maintenance of pluripotent stem cells during which time cells can be assayed for pathogenic contamination. Thus, maintenance of stem cells by the ex vivo system of the present invention avoids maintenance by potentially damaging "freezing for storage. " Also, the ex vivo system of the present invention mimics the in vivo hematopoietic system providing a valuable model for complex tissue formation, and for the discovery and evaluation of novel growth factors and chemotherapeutics.

The invention further provides a xenograft system for confirming the presence of expanded fetal HSC after ACS culture. The xenograft system of the invention confirms the presence of expanded fetal HSC by measuring engraftment and multilineage reconstitution in a recipient animal model. Multilineage reconstitution in an animal model is the only unambiguous assay for detecting the presence of "true" fetal HSC. Thus, the invention provides a method for unambiguously confirming the presence of fetal HSC after expansion in an ACS cartridge.

By the invention, engrafted fetal HSC cells are not only retained by the recipient animal, but are capable of multilineage differentiation into hematopoietic cells at differing degrees of maturity. For example, transplanting human fetal HSC in utero into a fetal lamb results in chimeric lambs with human HSC capable of multilineage differentiation into cells bearing human surface markers, human karyotype, and in vivo response to human-specific growth factors (Zanjani et al, J Clin Invest 89:1118 (1992)). Moreover, bone marrow from the chimeric lambs can be harvested and re-inoculated into a second sheep fetus.

HSC are the only hematopoietic cells capable of reconstituting the entire hematopoietic system. This is because HSC not only have the capacity for multilineage differentiation but also for extensive self-renewal. More mature hematopoietic progenitor and precursor cells are also considered "primitive". However, since progenitor and precursor cells are "committed" to become mature cells of a specific lineage, they are not capable of reconstituting the entire hematopoietic system. As indicated, previous attempts at ex vivo expansion of enriched HSC (isolated from adult bone marrow) resulted in cellular differentiation, accompanied by depletion of HSC. Thus, during expansion, HSC differentiated into the more mature cells. Koller et al. (Koller et al., Blood 82:373(1993)) attempted to overcome this problem by expanding stem and progenitor cells from adult marrow stroma without prior enrichment of the stem cells. Koller et al. used the LTC-IC assay for identification of expanded primitive cells. However, cells identified by the LTC-IC assay include not only HSC, but also the more mature progenitor cells. In contrast, the xenograft system of the invention provides a

method for unambiguously identifying "true" fetal HSC after expansion in an ACS cartridge.

After expansion in an ACS cartridge, harvesting, and resuspending as described above, fetal liver cells containing expanded HSC are transplanted in utero into a fetal non-human mammal recipient. Suitable transplantation techniques include the microbubble method described in Zanjani et al, J Clin Invest 59.1178(1992). The microbubble technique involves injecting the fetal liver cells into the peritoneal cavity of the fetal recipient. Other suitable in utero transplantation methods are known. For example, infusing the donor fetal HSC intravenously through the umbilical vein or placental vein can also be used.

The concentration of human fetal liver cells containing expanded HSC that is needed for injection is approximately Ixl0 ⁶ -lxl0 ^π cells/kg estimated fetal body weight in 0.5-1.0 ml. Preferably, a concentration of 2xl0 ⁹ -lxl0 ¹⁰ cells/kg fetal body weight is injected into the fetal lamb. However, other concentrations can be used as needed. Fetal body weight can be estimated using known techniques.

Donor fetal HSC engraftment in the recipient can be detected mixed with the recipient's hematopoietic cells after the transplant. Suitable techniques for determining the extent of HSC engraftment are known and include: karyotypic analysis of the recipient's peripheral blood, bone marrow, or liver hematopoietic cells; cytometric analysis of differentiation antigens associated with human hematopoietic subpopulations; and fluorescence in situ hybridization (FISH) with DNA probes specific for repeated satellite sequences on human (or other donor species) autosomes.

The extent of donor cell engraftment can be assessed in the non-human recipient both before and after birth. For engraftment analyses, mononuclear cells can be obtained by standard procedures (e.g. homogenization, flushing of long bones, density gradient separation) from liver, bone marrow, and blood of the recipient fetus at 2, 4, 6, and 8 weeks posttransplantation. Also, such cells can also be obtained for engraftment analyses after birth of the recipient. Of course, other time intervals can also be used as a matter of experimental discretion.

Suitable non-human mammal recipients for use in the invention include large mammals such as lambs and non-human primates. Other suitable non-human mammal recipients are rodents. Preferred non-human mammals are lambs. The transplant recipient should be at an early gestational stage. The gestational term of fetal lambs is 145 days. Thus, it is preferable that the transplant occur prior to 75 days gestation. More preferably, prior to reaching 50 days gestation.

Other large mammals have different gestational terms. Early gestation for these mammals will be readily apparent to the skilled artisan.

The fetal HSC to be detected in the xenograft system should be from a different species than that used as the transplant recipient. Preferably, the fetal HSC is from human fetal liver cells as discussed above.

For example, the extent of donor human HSC engraftment in a recipient mammal can be determined as follows. If engraftment is successful, the human donor fetal HSC will differentiate into the multiple cell lineages of the human hematopoietic system in the recipient animal.

The presence of human hematopoietic progenitors (multipotent colony- forming units [CFU-MIX]; colony forming unit-granulocyte, macrophage [CFU- GM]; and erythroid burst-forming units [BFU-E]) in a lamb (or sheep) recipient can be determined by establishing semi-solid cultures of lamb mononuclear cells from fetal liver, fetal bone marrow, or newborn or adult bone marrow with culture medium (IMDM) and erythropoietin (2 IU/ml). Supplementing the culture medium with 5 ng/ml each of recombinant human (rHu) IL-3 and GM-CSF will simulate colony formation of the human hematopoietic progenitors if they are present as subpopulations in the lamb (or sheep) mononuclear cells. During culture (e.g. days 9, 14, and 19), samples of the colonies can be taken and subjected to karyotypic analysis. Karyotypic analysis is performed by analyzing a colony sample for lymphoid elements expressing human karyotype after stimulation with PHA as described in Harrison et al, Lancet κ:1425 (1989).

The presence of cells of the human erythroid lineage in the sheep recipient can be demonstrated by flow cytometry. In particular, peripheral blood cells are isolated from the sheep recipient and the erythrocytes labeled with monoclonal

antibodies specific for the human form of the erythrocyte cell surface protein, glycophorin A (GPA). Using monoclonal antibodies (10F7, 6A, or BRIC157) specific for the human form of GPA to label this erythrocyte antigen is described in Langlois et al, Cytometry 11:513 (1990) and Langlois et al, Am. J. Immunol 734:4009 (1985). These human GPA specific monoclonal antibodies will bind subpopulations of human erythrocytes present in the sheep peripheral blood cells. However, the monoclonal antibodies will not bind subpopulations of sheep erythrocytes present in the sheep peripheral blood cells. Flow cytometric analysis of the erythrocytes can then be performed as described previously (Langlois et al, Cytometry 77:513 (1990)) to determine whether subpopulations of human erythrocytes bearing the GPA antigen are present in the peripheral blood cells isolated from the sheep.

The presence of human cells in the myeloid and lymphoid lineages in blood and bone marrow of transplanted sheep can be determined using flow cytometric analysis of subpopulations labeled with monoclonal antibodies specific for CD antigens present on human (but not sheep) T cells (CD3,4, and 5), natural killer and neutrophils (CD 16), B lymphocytes (CD20), leukocytes (CD45), and immature myeloid precursors and stem cells (CD34). The above human CD antigens can be labeled using commercially available monoclonal antibodies conjugated with either FITC or phycoerythrin (PE). Monoclonal antibodies specific for CD34, CD20, CD14, CD3, CD4, CD16, and CD8 can be obtained from Becton Dickinson Monoclonals, San Jose, CA. After labelling the sheep bone marrow and blood cell samples with the above monoclonal antibodies using conventional techniques, flow cytometric analysis can then be performed to determined whether subpopulations of cells bearing the human CD antigens are present in the samples.

The biotinylated DNA probe PC 190 will hybridize to repeated alpha satellite sequences on human chromosomes. Previous studies have shown that this probe is unreactive with sheep cells and allows detection of human cells present at frequencies of IO ^"4 in sheep background (Pallavicini et al, Cytometry 13:356 (1992)). Thus, fluorescence in situ hybridization (FISH) with biotinylated PC190 labeling repeated alpha satellite sequences on human chromosomes can be used to

detect human cells in recipient sheep blood and marrow. Appropriate hybridization conditions are described in Pallavicini et al, Mol Cell. Biol. 70:401 (1990). The biotinylated probe PC 190 is described in Weir et al, Chromosoma (Bed.) 100:311 (1991). Hybridized probe can be visualized using FITC-conjugated avidin and a fluorescence microscope.

All of the above described assays for assessing the extent of human fetal HSC engraftment in recipient lamb/ sheep are described in detail in Zanjani et al. , J. Clin. Invest. 59:1178 (1992).

The xenograft system of the invention has many uses. In addition to confirming the presence of "true" fetal liver HSC, the xenograft system can also be used for the in vivo storage and/or expansion of human fetal HSC. A certain degree of expansion of donor HSC occurs as the bone marrow tissue expands during the latter part of fetal development. Further expansion can be induced by the in vivo administration of human-specific recombinant growth factors. Subsequently, human cells can readily be separated and recovered from host sheep cells by their specific surface molecules such as CD antigens not expressed on sheep cells. Thus, the xenograft system of the present invention can serve as a reservoir of human HSC.

Human-specific growth factors are known in the art and include rHuIL- 3, GM-CSF, rHuIL-2, rHuIL-1, rHuIL-6, and SCF. These growth factors can be administered using conventional techniques to a lamb or sheep containing engrafted human fetal liver HSC either before or after birth. For example, human- specific growth factors can be administered to the fetal lamb by intraperitoneal injection or by infusion through the umbilical or placental vein. Human-specific growth factors can be administered to a sheep or lamb after birth by intraperitoneal injection or intravenous infusion. Concentrations of human specific growth factors that are sufficient to induce the in vivo expansion of the engrafted human fetal liver HSC are 5-10 μg/Kg body weight. However, other concentrations can be used as required. Preferably, these concentrations are administered twice daily for four days beginning 12 months after the original transplantation of human fetal HSC.

Successful in vivo expansion can be detected in sheep bone marrow and peripheral blood using the karyotyping techniques described above.

Bone marrow transplantation (BMT) from an HLA-identical sibling remains the best treatment of many disorders. However, only 35% of candidates for BMT have an HLA-identical family member. When no HLA-compatible donor is available, recipient tolerance must be induced using harsh conditioning regimens such as ionizing radiations and/or cytoablative therapy (Sullivan, K.M. , Transplant Proceedings 21 (Suppl. 1):41 (1989)). The attendant risks of such conditioning regimens include immunosuppression and overwhelming infection in the patient (Clark, J.G., Mayo Clin. Proc. 65:111 (1990); Parkman, R., Science 232:1313

(1986)). In general, results using HLA-mismatched donors have been disappointing (O'Reilly et al, Immunodef. Rev. 7:23 (1990); Anasetti et al, N.

Engl J. Med. 320:191 (1989); Ferrara et al, N. Engl. J. Med. 324:661(1991)).

When no HLA-compatible donor is available, transplants using donor fetal liver cells are an attractive option. This is because post-thymic mature T-cells are absent in the fetal liver during early gestation (O'Reilly et al. in Seligman, M., Hitzig, W.H., (eds.), Primary Immunodeficiencies, New York:Elsevier/North- Holland, p. 419 (1988)) and, consequently, GNHD has not been a problem in the extensive postnatal experiments with fetal liver transplantation (Gale, R.P., et al, Thymus 10:45 (1987); Royo et al, Thymus 10:5 (1987)).

In 1987, Touraine etal, Thymus 10:15 (1987), provided an extensive study of successful fetal tissue transplants resulting in therapeutical solutions for patients with a variety of diseases. Touraine et al. reported that, despite complete HLA incompatibility between transplanted liver stem cells and host cells, functional activities of donor lymphocytes were not restricted and GNHD was only a minimal problem. Patients who can be treated with liver transplants suffer from a vast array of maladies including immunodeficiency diseases such as severe combined immunodeficiency diseases (SCID), bare lymphocyte syndrome (BLS), adenosinedeaminase deficiency, chronic granulomatosis disease (CGD), and Wiscott-Aldrich Syndrome. Touraine et al. also performed fetal liver transplants on patients suffering from the following diseases of inborn errors of metabolism

(IEM): Gaucher, Fabry, Fucosidosis, Hurler, Metachromic leucodystrophy, Hunter, Glycogenesis, San Filippo B, Morquio B, Niemann-Pick A, Niemann-Pick B, and Niemann-Pick C. Finally, Touraine et al. suggested that fetal liver transplants are also very efficient treatments for leukemia, aplastic anemia and other hemoglobinopathies.

However, notwithstanding the above, transplants using HLA-mismatched donor HSC from adult bone marrow remain as the sole option in many instances. This is largely due to the shortage of fetal HSC. The use of HSC from fetal livers is limited by ethical objections, microbial contamination, and high cost of tissue processing (Rice et al, Fetal Diagn. Ther. 5:74 (1993)). The present invention overcomes many of these problems by providing a rich source of ex vivo expanded fetal HSC.

Thus, the invention also provides improved postnatal and in utero fetal HSC transplantation methods. For postnatal transplants, the invention involves inoculating fetal liver cells into an ACS cartridge, perfusing the cartridge with culture medium containing at least one hematopoietic growth factor, culturing the fetal liver cells in the cartridge for a sufficient amount of time to achieve expansion of the fetal HSC, harvesting cultured fetal liver cells containing expanded fetal HSC, and transplanting a therapeutically effective amount of the fetal liver cells containing expanded fetal HSC to a patient in need of a fetal HSC transplant.

Methods for expanding and harvesting fetal liver cells containing HSC using an ACS are discussed above. After harvesting, fetal liver cells containing the ex vivo expanded HSC can be transplanted into a patient using techniques described in Touraine et al, Thymus 10:15 (1987). Other transplantation techniques are known in the art. For example, the cells can be transplanted by intravenous infusion or by injecting the cells intraperitoneally. As indicated, preferably, the fetal liver cells are obtained from the deceased fetus prior to 30 weeks' gestation; preferably, prior to 20 weeks' gestation; and most preferably, prior to 15 weeks' gestation. The concentration of fetal liver cells containing the ex vivo expanded HSC that is therapeutically effective if transplanted to the patient should be approximately 10 ⁵ to 10 ¹⁰ cells per kg of patient body weight. Preferably, the

concentration of transplanted cells is 10 ⁷ -10 ⁸ cells per kg of patient body weight. However, other concentrations can be used as needed.

The immune reconstitution provided by ex vivo expanded fetal liver HSC may be enhanced by simultaneous transplantation of thymic epithelial cells, preferably from the same donor (Bortin et al, Science 164:316 (1969); Pahwa et al, Proc. Natl. Acad. Sci. USA 74:3002 (1979)). The thymic cells are believed to provide a syngeneic environment for T-lymphocyte differentiation of transplanted fetal HSC. Thus, optionally, thymic cells can be transplanted to a patient simultaneously with the cultured fetal liver cells containing expanded HSC. A suitable number of thymic cells for transplantation will be apparent to the skilled artisan (Touraine et al, Thymus 10:15 (1987)). For example, a concentration of IO ⁷ - 10 ⁸ thymic cells can be administered. Alternatively, the thymic cells can be mixed with the donor HSC in the ACS prior to administration to the recipient. The thymic cells are obtained from the deceased fetus using conventional dissection techniques.

Alternatively, after cultured fetal liver cells are harvested from the ACS cartridge, the expanded fetal HSC can be separated from other fetal liver cells using conventional cell sorting techniques. For example, the anti-CD34 ⁺ monoclonal antibody cell sorting techniques described in Brandt et al, J. Clin. Invest. 86:932 (1990); Edgington, S.M. , Biotechnology 10:1099 (1992); and Srour et al, Blood 81 (3) :661 (1993) can be used. Separation significantly increases the concentration of the fetal liver HSC. Thus, much lower total cell numbers can be used for the transplant. As indicated, if the cultured fetal liver cells are transplanted without separating HSC from non-HSC, it is preferable to inject about 10 ⁷ -10 ⁸ cells per kg of body weight. However, if the HSC are separated from other fetal liver cells after expansion, the amount needed to be therapeutically effective for transplantation is only about 10 ⁵ - 10° cells per kg of body weight. Thus, the present invention reduces the number of fetal liver cells required for transplantation. Patients that can be treated by present invention include those suffering from the disorders listed above. The method can be used to treat mammalian patients, including humans.

Recent advances in prenatal diagnosis now allow the diagnosis of most congenital hematopoietic and metabolic disorders during the first trimester of pregnancy (Flake et al, Exp. Hematol 79:1061 (1991)). These include immunodeficiency diseases, hemoglobinopathies, and inborn errors of metabolism. Specific diseases include α-thalassemia, β-thalassemia, bare lymphocyte syndrome (BLS), severe combined immunodeficiency disease (SCID), chronic granulomatosis disease (CGD), Wiscott-Aldrich Syndrome, Hurler's disease, and Hunter's disease. Allogeneic BMT has been used to treat a number of these diseases postnatally (Sullivan, K.M., Transplant Proc. [Suppl. 1] 21:41 (1989)). However, even when an HLA-compatible sibling is the donor, postnatal transplants are often performed too late to prevent damage already caused by disease. For example, in the case of certain glycogen storage diseases, neurologic injury occurs prior to birth (Krivit etal, N. Engl. J. Med. 322:28 (1990)), rendering postnatal treatment ineffective to prevent injury. In utero HSC transplantation has emerged as an attractive alternative to postnatal BMT. Studies of fetal ontogeny reveal that human fetuses are tolerant of foreign antigens and allow engraftment of allogeneic stem cells without rejection (Brent et al. In: Albertini, A., Crosignani, P.G. (eds.) "Progress in perinatal medicine" Amsterdam: Excerpta Medica, p. 3 (1983)). Also, first trimester fetal bone marrow is relatively empty. Thus, reconstitution of the hematopoietic system is achievable without the need for bone marrow ablation. In 1991, Jean-Louis Touraine (Touraine, J-L., Blood Cells 77:379(1991) reported transplanting fetal liver cells in utero to fetal patients diagnosed with bare lymphocyte syndrome (BLS), severe combined immunodeficiency disease (SCID), and thalassemia. Because of the immunological naivety of the fetal recipient, no HLA-matching was necessary. Thus, taken together, the advantages of a fetal donor and fetal recipient avoid practically all the complications of postnatal bone marrow transplantation.

The invention further provides an improved method for transplanting expanded fetal liver HSC in utero to fetal patients. The invention improves on Touraine's method by providing an immediately accessible source of expanded fetal liver HSC. The invention involves inoculating fetal liver cells into an ACS

cartridge, perfusing the cartridge with culture medium containing at least one hematopoietic growth factor, culturing the fetal liver cells for a sufficient amount of time to achieve expansion of fetal HSC, harvesting cultured fetal liver cells containing expanded fetal HSC from the ACS cartridge, and transplanting a therapeutically effective amount of the cells in utero to a fetal patient in need of an HSC transplant.

Methods for expanding and harvesting fetal liver cells containing HSC using an ACS are described above. After harvesting, fetal liver cells containing the ex vivo expanded HSC can be transplanted in utero to a fetal patient as described in Touraine, J.-L., Blood Cells 17:319 (1991). For example, the cells can be transplanted to the fetus by infusion into the umbilical or placental vein or by intraperitoneal injection. Preferably, the fetal liver cells are obtained from human fetal livers at early gestation (see above). To be therapeutically effective, the concentration of cells that are transplanted in utero should be approximately 10 ⁵ - 10 ¹⁰ cells per kg of estimated fetal body weight. Preferably, the concentration of transplanted cells is 10 ⁷ -10 ⁸ cells per kg of estimated fetal body weight. However, other concentrations can be used as needed. Fetal body weight can be estimated using known techniques.

Alternatively, after the cultured fetal liver cells are harvested from the ACS cartridge, the expanded fetal HSC can be separated from other fetal liver cells using conventional cell sorting techniques. For example, anti-CD34 ⁺ monoclonal antibody cell sorting techniques as described above can be used. Therapeutically effective concentrations of separated fetal liver HSC that can be transplanted in utero are the same as that described above for postnatal transplants. Thus, the invention also reduces the number of fetal liver cells required for transplantation in utero.

In utero transplantation to the fetal patient should occur during early gestation before maturation of the immune system. Preferably, in utero transplantation should occur prior to 30 weeks' gestation; more preferably, prior to 25 weeks' gestation; and most preferably, prior to 20 weeks' gestation. Fetal patients that can be treated by the invention include those suffering from the

disorders listed above. The method can be used to treat mammalian fetal patients, including humans.

Active target cell cycling is required for retroviral integration (Nolta et al. , Exp. Hematol. 20:1065 (1992)). Thus, the invention further provides a method for transducing ex vivo expanding fetal liver HSC with certain packaged viral vectors. An example of such vectors are retrovirus vectors. The method involves inoculating an ACS cartridge with fetal liver cells and a suspension containing packaged retrovirus vectors, perfusing the ACS cartridge with culture medium containing at least one hematopoietic growth factor capable of stimulating expansion of the fetal HSC, culturing the fetal liver cells in the presence of the packaged retrovirus vectors in the ACS cartridge for a sufficient amount of time to achieve expansion of the fetal HSC, and harvesting the cultured fetal liver cells containing expanded fetal HSC which have been transduced with the retrovirus vector. Retrovirus vectors are the prefeπed vectors for genetic therapy (Anderson et al, Science 226:401(1984)). This is because retrovirus infection is highly efficient and retrovirus vectors can be modified to stably integrate into the host cell's genome. Cis- and trans- acting elements control retrovirus replication. Trαns-acting elements include the viral proteins necessary for viral encapsidation, binding to the target cell surface, entering the target cell, reverse transcription, and integration of provirus into the target cell genome. Cw-acting elements interact with trans elements during viral replication (Coffin, J. (1985) in RNA Tumor Viruses, vol. 2, p. 17-74, R. Weiss et al , eds., Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.). Retrovirus vectors having traits elements deleted (e.g. env, gag, and pol gene deletions) are replication incompetent. Alone, these retrovirus vectors cannot replicate or infect target cells. Packaging cell lines have been constructed to include an env gene from a retrovirus having an amphotropic or ecotropic host range. These cell lines are capable of "packaging" the retrovirus vectors thereby rendering them infectious, i.e., capable of transducing target cells. Upon transduction into the target cell, the retrovirus vectors can stably integrate into the host cell genome because they are replication incompetent.

Miller (Miller, D., U.S. Patent No. 4,861,719) constructed various packaging cell lines by transfecting different cell lines (NIH/3T3 TK ^" , PA12, PSI- 2, or 208F) with packaging DNA constructs pPAM, pPAM2, pPAM3 or pPAM4. These packaging DNA constructs provide tra/w-acting factors capable of packaging retrovirus vectors. For, example Miller introduced the packaging DNA construct pPAM3 into NIH/3T3 TK ^" cells generating a packaging cell line he named PA317 (ATCC Accession No. CRL 9078). Miller then tested the ability of PA317 to package the following recombinant retrovirus vectors in monolayer culture: the Neo- virus N2 (a recombinant retrovirus containing the Neo resistance gene); the DHFR-virus SDHT (a recombinant retrovirus containing the DHFR gene); and the HPRT- virus LPL2 (a recombinant retrovirus containing he HPRT gene). The results showed that the packaging cell lines were capable of packaging high concentrations of the recombinant retrovirus vectors.

Similarly, Hock et al, Blood 74:816 (1989) packaged high concentrations of LASN (a retrovirus vector containing the ADA and Neo resistance genes) in a cell line derived from PA317. Moreover, Knazek et al, (abstract from presentation at BioEast 91 in Washington, DC, January 1991) showed that even more concentrated suspensions of packaged LASN are produced if the LASN- producing PA317 is grown to near solid tissue density with the CELLMAX™ 100 (Cellco, Inc.) artificial capillary system. This is important since highly concentrated suspensions of retroviral vectors are necessary to efficiently transduce target cells. Thus, from the above, it is clear that several packaging systems are available that can be used to package recombinant retrovirus vectors to yield highly concentrated suspensions. The retrovirus vector can be modified by inserting heterologous genes encoding therapeutically effective products. For example, LASN contains the ADA gene whose product is useful for treating a type of SCID. Other retrovirus vectors which can be modified by insertion of a heterologous gene encoding a therapeutically effective product are pN2 (Keller et al, Nature 375:149(1985)); pLHL (Miller et al, Cold Spring Harbor Symp. on Quant. Bio. , Vol. LI, Cold Spring Harbor Laboratory, p. 1013 (1986)); pSDHT (Miller et al, Somat. Cell.

Mol Genet. 12:115 (1986)); pLPL (Proc. Natl. Acad. Sci. USA 80:4109). These vectors are known and available to the skilled artisan.

In particular, genes, which encode the following therapeutically effective products, can be inserted as heterologous genes into the recombinant retrovirus vectors using conventional techniques: ADA, IL-2, IL-4, TNF, and interferon gamma. These genes are known and available to the skilled artisan.

When recombinant retrovirus vector-producing packaging cells lines are grown in culture as described above, high concentrations of packaged recombinant retrovirus particles are produced in the cell supernatant. Fetal liver cells containing HSC (which have been harvested from fetal livers as described above) can be added in suspension to the vector-containing supernatant. The suspension, containing packaged recombinant-retrovirus vector and target fetal HSC, can then be inoculated into an ACS cartridge. The fetal liver cells are then cultured in the ACS cartridge, in the presence of the packaged recombinant retrovirus vectors, for a sufficient amount of time to achieve expansion of the fetal HSC. Appropriate culturing time and conditions are described above. After culture, transduced fetal HSC can be harvested from the ACS cartridge by flushing the cartridge with culture medium or by gently shaking the cells from the cartridge.

To enhance infectivity of the retrovirus vectors, polycations (such as protamine), at concentrations of 5-10 μg/ml, can be added to the suspension containing target fetal HSC and retrovirus vector.

Alternatively, the fetal liver cells can be inoculated into the ACS cartridge prior to suspending with the recombinant retrovirus vectors which are then added later either before or after culture has been established. Also, the retrovirus vector-containing supernatant can be added to the ACS cartridge prior to inoculating the fetal liver cells. The order will depend on the experimenter's design. The target fetal HSC can be transduced either before or after inoculation into the ACS cartridge. However, at some point, the fetal HSC must undergo division to allow the retrovirus to integrate into the fetal HSC genome. To increase the concentration of fetal HSC that are transduced with retrovirus vector, supernatant containing packaged retrovirus vector can be added

at intervals during culture. For example, 5ml of the vector-containing supernatant can be added to the cartridge every two days. The culture can be terminated after eight days. Of course, different intervals and volumes of supernatant can be used if needed. The presence of the recombinant retrovirus vector in the fetal liver HSC genome after expansion can be confirmed using the polymerase chain reaction (PCR) on a sample of the expanded fetal HSC. For example, for detection of LASN in expanded fetal HSC, DNA primers flanking the Neo resistance gene can be used to amplify the Neo gene. The amplification product can then be loaded and run on a gel and probed with a Neo resistance gene specific probe. For other recombinant retrovirus vectors, presence of the vector could also be confirmed with PCR using primers flanking sequences specific to the vector (or heterologous gene insert) and not contained in the host HSC genome.

The transduced fetal HSC (fetal HSC containing a recombinant retrovirus vector) can then be introduced into a patient by transplantation postnatally or in utero using the transplantation methods described above.

Having generally described the invention, the same will be more readily understood through reference to the following examples which are provided by way of illustration and are not intended to be limiting.

Experimental

Example 1

Ex Vivo Expansion of Fetal Liver HSC Materials and Methods

Experimental Protocol

This study was designed to determine whether HSC harvested from human fetal livers could expand in an artificial capillary system (ACS). Single cell suspensions of the livers of five electively aborted fetuses were created. These suspensions contained hematopoietic and stromal cells. Each artificial capillary cartridge was seeded with 5-10 x 10° cells, and the ACS was perfused with medium containing stem cell factor and interleukin-3 for up to 50 days without additional supplementation.

At periodic intervals, various cell characteristics were measured. First the stem cell containing fraction was quantified based upon flow cytometric analysis of cell surface markers (CD34 and HLA-DR). Second, immunohistochemistry was performed using anti-CD34 monoclonal antibodies. Third, committed hematopoietic progenitor colonies were quantified using semisolid culture.

Finally, to confirm the presence of human stem cells after culture, the expanded cells were transplanted into three early gestation fetal lambs. The two surviving lambs were sacrificed near term gestation and fetal blood leukocytes, liver cells, and bone marrow were treated for human engraftment using flow cytometric analysis with a CD45 monoclonal antibody.

Human Fetal Liver Cell Preparation

Human fetal cells were harvested from the livers of five electively aborted fetuses of 14-20 week gestation. We dissected the livers from the fetuses following transvaginal extraction, placed the livers in Isocove's modified Dulbecco's medium (IMDM) with 10% fetal calf serum (FCS), penicillin (100 units/ml), streptomycin (100 μg/ml) and amphotericin B (2.5 μg/ml), and transported them to a sterile laminar flow hood. We washed the livers twice with fresh medium and created a single cell suspension with a hand-operated glass homogenizer. After lysing the erythrocytes with hypotonic solution, we washed the cell suspension twice in fresh medium, and then examined cell viability using a manual hemocytometer after staining the cells with ethidium bromide and acridine orange.

The study was approved the University of California Committee on Human Research. All fetal tissue was obtained with the assistance of Advanced Biosciences Resources, Inc., Alameda, CA, and performed in strict accordance with the guidelines of the UCSF Committee on Human Research.

Cell culture medium

For all cultures in the artificial capillary system, we used a medium of IMDM supplemented with 10% FCS, recombinant human interleukin 3 (IL-3, 10 ng/ml), recombinant human stem cell factor (SCF, 25 ng/ml), penicillin (100 units/ml), streptomycin (100 μg/ml), and amphotericin B (2.5 μg/ml) were added. Recombinant human IL-3 was obtained from Genzyme Corp. (Cambridge, MA). Recombinant human SCF was obtained from Amgen, Inc. (Thousand Oaks, CA).

Artificial capillary system

The artificial capillary system (CELLMAX Quad, Cellco, Inc., Germantown, MD) used in this study consisted of a 2 x 13 cm cylindrical cartridge containing hollow capillary fibers made of cellulose diacetate, silicone rubber tubing which serves as a gas exchanger to maintain the proper pH and pO ₂ , a 125 ml medium bottle which serves as a reservoir, and an external pump. Tissue culture medium was drawn from the reservoir, pumped through the gas exchange

tubing and lumen of the capillaries, and then returned to the reservoir for subsequent recirculation. Nutrients and oxygen in the medium diffused through the capillary walls into a 7 ml extra-capillary space (ECS), and cell metabolites diffused back from the ECS into the perfusate. We gradually increased the flow rate of the pump to approximately 50 ml/min over the duration of the culture and kept the system at 37 °C in 5% CO ₂ in a humidified incubator.

Capillaries with a 50% molecular weight cutoff of 80 kDa provided 0.07 m ² surface area for culture. We injected 5-10 x 10° cells directly into the ECS, and at periodic intervals partially harvested the cells by flushing the ECS with three separate 10 ml aliquots of fresh medium. The small number of cells that remained behind reseeded the bioreactor for further culture.

Cell proliferation

The rate of cell proliferation was determined by measuring the lactic acid concentration in the medium using an automated analyzer (Yellow Springs Instrument Co., Inc.). We calculated the daily lactate production by dividing the change in total lactate content by the days between measurements, plotted the rate of lactate production, and determined the doubling time of lactate production based on logarithmic growth rate.

Every 5-10 days, we harvested and quantified the mononuclear cells in suspension from the ECS. The rate of cell expansion was compared to the rate of daily lactate production.

Phenotype analysis

To determine the fraction of CD34 ⁺ and CD34 ⁺ HLA-DR ^" cells after culture in the ACS and the fraction of CD45 ⁺ cells after transplantation into fetal sheep, we performed single or two-color flow cytometry to detect cell surface markers. Briefly, we harvested the cells, washed them with cold staining medium (Hanks' buffered salt solution with 5% FCS and 0.02% sodium azide), and resuspended them at concentrations of 1-2 x IO ⁶ cells/ml. For characterization of the stem cell fraction, we used dual staining with fluorescein (FΙTC)-conjugated anti-CD34 and phycoery thrin (PE)-conjugated anti-HLA-DR monoclonal antibodies . For detection of human cells after transplantation into fetal sheep, we used a FITC-conjugated anti-CD45 monoclonal antibody. All monoclonal antibodies were obtained from Becton Dickenson, Inc. (San Jose, CA) and were used at 1:10 dilution. We stained 100 μl volumes of cell suspension (5%, vol./ vol.) with the antibodies and used isotype specific antibodies as negative controls. After incubating the cells for 30 min, we washed the cells twice with medium and analyzed them using a FACScan cytometer (Beckton-Dickenson, Inc., San Jose, CA) using Lysis II research software. Each measurement represented 10,000 cells and cell viability at the time of analysis was >95%.

Immunohistochemistry To confirm the presence of CD34 ⁺ cells after expansion, we harvested samples of cells for histologic and immunohistochemical analyses. Briefly, the cells were washed twice in phosphate-buffered saline (PBS) and resuspended at 1-5 x IO ⁶ cells/ml. We placed several drops of suspension onto glass slides, fixed the slides in acetone, and stored the slides at -80 °C for further processing. To examine gross histology, we incubated the slides with Giemsa stain for

30 min and then rinsed them with PBS. To examine immunohistochemistry, we stained the slides for 30 min with the following series of solutions and between each stain washed the slides twice with PBS. First, we stained the slides with mouse anti-human CD34 monoclonal antibody (1:20 dilution, Becton Dickenson

Inc. , San Jose, CA), then with alkaline phosphatase-conjugated goat anti-mouse IgG (1:200 dilution, Boehringer Mannheim, Inc., Indianapolis, IN), next with alkaline-phosphatase substrate (Vector Red, Vector, Inc., Burlingame, CA), and finally with a counterstain of 10% hematoxylin. Appropriate isotype specific monoclonal antibodies were used as controls. We confirmed the presence of human CD34 ⁺ cells by red staining under light microscopy.

Hematopoietic progenitor colony assays

At the time of inoculation and at periodic intervals after culture, we harvested cells from the ECS and plated them in semisolid medium to quantify hematopoietic progenitor colonies (CFU-GM, CFU-E, and BFU-E) as described previously (Zanjani et al, J. Clin. Invest. 59:1178 (1992)). To detect BFU-E and CFU-GM, we used a medium of 0.9% methylcellulose in IMDM with FCS (30%), phytohemagglutinin-leukocyte conditioned medium (5%, Fox Laboratories, Vancouver, B.C.), and erythropoietin (2 U/ml). To detect CFU-E, we used a plasma clot medium of IMDM with human AB plasma (10%), FCS (30%), α-Thio B (5 x 10 ⁷ M), thrombin (10%), L-asparagine (20 ng/ml), and erythropoietin (2 U/ml). We cultured the cells in triplicate (at three different concentrations, 2 x 10 ^s , 4 x IO ⁵ , and 6 x 10 ⁵ cells/ml). After 7-21 days in culture, we identified and quantified the hematopoietic progemtor colonies, and calculated the number of colonies per cells initially seeded. The assay was linear within this range of seeded concentrations.

In utero stem cell transplantation into fetal sheep

To confirm the presence of human stem cells, we harvested the cells after 5-10 days of ACS culture and transplanted them into three early gestation fetal sheep (55-65 days gestation, term = 145 days) as described previously (Zanjani et al, J. Clin. Invest. 59:1178 (1992)). Briefly, we performed a maternal laparotomy on time-dated pregnant ewes, exposed the uterus, and created a partial thickness hysterotomy. Under direct vision we injected the cells (2-20 x IO ⁶

cells/fetus) through this "amniotic window" into the peritoneal cavity of the fetal lambs.

Two of these three fetal lambs survived to near term ( > 120 days) and were then sacrificed. We analyzed the fetal peripheral blood leukocytes, liver cells, and bone marrow from these two lambs for human engraftment using flow cytometric analysis with FITC-conjugated anti-CD45 monoclonal antibody.

Results

Cell Proliferation

Human fetal liver cells demonstrated a logarithmic rate of expansion in the artificial capillary system, although the rate was highly variable between the different cartridges. These results are provided below in Table 1.

Table 1 Human Fetal Liver Cell Expansion in Artificial Capillary System

Cartridge Cell expansion Lactate doubling time

1 20x 3.1

2 16x 3.4

3 1.5x 6

4 3x 5.5

5 2.3x 1.7

Table 1 - Cell proliferation and lactate doubling time for five artificial capillary cartridges 10 days after seeding witti human fetal liver cells (cartridges were seeded with 5-10 x 10 ⁶ mononuclear cells). Cell expansion was calculated as ratio of total number of cells harvested from cartridge compared to seeded cell number. Lactate doubling time was determined based on logarithmic rate of growth over the 10 day period.

At maximum growth, there was a 20-fold cell expansion over the first 10 days of culture (Figure 1). After harvest, the remaining cells in the ECS resumed a rapid expansion rate. This cycle of harvest and regrowth was repeated up to five times in one of the cartridges. Although we did not characterize cell phenotype after 20 days of culture, we were able to maintain three of the cartridges for 50 days. The rate of daily lactate production exhibited a similar logarithmic increase over time, doubling every 1.6-6.0 days (doubling time 4.0 ± 1.6 days, mean + standard deviation) over the first 10 days of culture (Table 1).

Phenotype analysis

Using flow cytometric analysis, the percent of human cells that was CD34 ⁺ and the percent that was CD34 ⁺ HLA-DR ^" remained constant throughout the first 10 days of culture (Figure 2). Thus, this CD34 ⁺ fraction represents a 20-fold increase in the number of CD34 ⁺ cells over 10 days of culture. These results are provided below in Table 2. After 20 days in cultore, the percent of total cells that were CD34 ⁺ decreased.

Immunohistochemical staining with anti-CD34 monoclonal antibodies confirmed the presence of CD34 ⁺ cells at the time of initial ACS inoculation and after 10 days of culture (Figures 3 A and 3B).

Table 2

Flow Cytometric Analysis of Human Fetal Liver Cells After Culture

% CD34 ⁺ % CD34 ⁺ HLA-DR-

Baseline 5.8 ± 1.9 1.1 ±0.5

Day 5 6.5 ± 1.0 1.9 ± 0.1

Day 10 5.0 ± 0.6 1.9 ± 0.8

Day 20 1.4 ± 1.0 0.4 ± 0.2

Table 2-Summary of the percentage of the total cells diat are CD34 ⁺ and CD34 ⁺ HLA- DR ^" in human fetal liver cultures by single and dual color flow cytometric analysis at baseline and at various intervals after culture. Data are expressed as mean ± standard deviation for five liver cultures in artificial capillary system.

Histology

Giemsa staining demonstrated a gradual maturation pattern of hematopoietic cells after 20 days of culture (Figure 4). Initial examinations of the cell suspensions of fresh human fetal livers confirmed the presence of a heterogeneous cell population of immature appearing cells and stromal elements. However, stromal elements were not seen after cultore in the artificial capillary system. After 10 days of cultore, the cells developed a homogeneous blastic appearance. Finally, after 20 days of cultore, a predominance of mature and immature myeloid and lymphoid elements was noted.

Hematopoietic progenitor assays

The human fetal liver cells in the cultore retained hematopoietic progenitor colony potential at least to 5 days of cultore (Figure 5). The number of CFU-GM and BFU-E progenitors persisted until day 10, whereas CFU-E declined rapidly after 5 days of cultore.

In utero stem cell transplants

One of the two surviving fetal lambs that had undergone in utero injection of expanded cells demonstrated successful human cell engraftment. Flow cytometric analysis of this lamb demonstrated that 15.7% of liver cells and 4.9% of peripheral blood lymphocytes were CD45 ⁺ (Figure 6). In contrast, no CD45 ⁺ cells were detected in the bone marrow. The other surviving lamb had no detectable CD45 ⁺ cells in liver, bone marrow, or peripheral blood.

Discussion

Transplantation of bone marrow is now standard therapy for a variety of conditions, including hematopoietic reconstitotion following marrow ablative therapy, and disorders of hematopoiesis and metabolism. Similarly, in utero stem cell transplantation is emerging as a potential therapy for the fetus prenatally diagnosed with various hematopoietic disorders (Flake et al, Exp. Hematol 79:1061 (1991)). The ability to expand HSC ex vivo provides a rich source of transplantable cells.

Many recent studies have attempted to identify the requirements for ex vivo expansion of hematopoietic stem cells (Williams, D.A., Blood 81(12):3169 (1993); Muench et al, Blood 57:3463 (1993); Bernstein, et al, Blood 77:2316 (1991); Srour et al, Blood 57:661 (1993). Various assays have been used to characterize expanded progenitor cell populations, including phenotype analysis, long-term initiating cultore assay, and progenitor clonal growth assays. However, the ability to permanently repopulate an animal remains the gold standard for stem cell identification. Meunch et al. have successfully transplanted ex vivo expanded allogeneic stem cells into mice (Muench et al, Blood 57:3463 (1993). However, for the identification of human stem cells, we have developed a model of transplantation into fetal lambs (Zanjani et al, J. Clin. Invest. 59:1178 (1992)).

Cuπently, stem cells used in clinical transplantation are harvested from adult bone marrow or from peripheral blood. Stem cells harvested from fetal tissues may have homing and proliferative advantages over adult derived cells for

clinical transplantation (Fleischman et al, J. Exp. Med. 159:131 (1984); Emerson et al, Blood 74:49 (1989)). In addition, the absence of post-thymic, mature T- cells in fetal liver may reduce the risk of graft- versus-host disease (O'Reilly et al. in Seligman, M., Hitzig, W.H., ed. Primary Immunodeficiencies. New York: Elsevier/North-Holland p. 419 (1980)). However, the use of fetal HSC is limited by ethical objections, microbial contamination, and high cost of tissue processing (Rice et al, Fetal Diag. Ther. 5:74 (1992)). Although fetal liver is a rich source of hematopoietic cells, (Meunch et al, Exp. Hematol. 27:1013 (1993)), the small size limits the total cell yield for transplantation. Many of these difficulties could be overcome by the ex vivo expansion of fetal HSC.

Recently, artificial capillary systems (ACS) have been developed which allow the efficient expansion of large quantities of cells (Edgington, S.M., Biotechnol. 10:1099 (1992)). We wondered if hematopoietic stem cells harvested from human fetal livers could expand in an artificial capillary system. To test this question, we used an ACS to expand fetal liver cells, and then analyzed these cells using a number of techniques, including flow cytometry, immunohistochemistry, and hematopoietic progenitor colony forming assays. We then confirmed the expansion of HSC by in vivo transplantation into fetal sheep.

The availability of ex vivo expanded fetal stem cells would ease the short supply of fetal tissue. From a practical standpoint, expanded fetal stem cells would reduce the amount of abortus tissue necessary and would allow the storage of large numbers of cells for future transplantation. Furthermore, the ex vivo replication of stem cells improves their transduction by retro viruses, which would greatly facilitate their use for gene therapy. Finally, the use of cultured fetal HSC could provide the time required to assure sterility and absence of transmissible viral diseases prior to clinical transplantation.

Conventional cell cultore techniques are inefficient, cumbersome, and require large volumes of medium to grow significant numbers of cells. Moreover, since cells have a natural tendency to grow in three dimensions, traditional two- dimensional culture techniques restrict the diffusion of oxygen and nutrients necessary for optimal cell growth. In contrast, the artificial capillary system

allows growth of cells in a three dimensional array that more closely resembles biologic tissue. Nutrient and oxygen supply is enhanced, and the natural stromal microenvironment can grow on the fiber matrix.

Although we did not specifically study the stromal support network of the artificial capillary cartridges, it is quite likely that the fetal liver stroma greatly influenced the expansion of stem cells. Examination of the hematopoietic microenvironment in other ex vivo cultore systems has suggested a number of possible roles of the cellular microenvironment, including direct communication via cell-to-cell contact, local production or stabilization of growth factors via binding to membrane proteins, and co-localization of growth factors and hematopoietic cells in a local area network (Williams, D.A., Blood 81(12):3169 (1993); Trentin J.J., in Gordon, A.S., ed. Regulation of Hematopoiesis . New York: Appleton-Centory- Crofts, p. 161 (1970); Issaad et al, Blood 57:2916 (1993)). Recent work in our laboratory has suggested that fetal liver stroma supports multilineage hematopoiesis.

Previous studies have established the role of cytokines in the expansion of hematopoietic stem cells. IL-lα and G-CSF have synergistic activities on the expansion of hematopoietic cells with high and low proliferative capacity when used in vitro with marrow (Moore et al, Proc. Natl. Acad. Sci. U.S.A. 84:1134 (1987)). IL-6 and IL-3 may show similar synergy in supporting the proliferation of human blast cell colonies (Leary et al, Blood 77:1759 (1988)). Human bone maπow cells proliferate in response to IL-3, IL-1, and IL-6. (Brandt et al, J. Clin. Invest. 86:932 (1990); Kobayashi et al, Blood 75: 1947 (1991)). Finally, SCF may increase the proliferation of human stem cells when incubated with various combinations of cytokines (Muench et al, Blood 57:3463 (1993); Bernstein et al, Blood 77:2316 (1991); Bodine et al, Blood 79:913 (1992)). We chose to use SCF and IL-3 in our study, although the selection of other cytokines may allow more selective enrichment of fetal HSC in an ACS system.

Other approaches to cell cultore have been shown to enhance the growth of adult marrow HSC. Recently, Koller et al , by using a continuously perfusing cultore, expanded adult human bone marrow progenitor cells 10-fold over two

weeks of cultore (Koller et al, Blood 82:318 (1993)). This study is limited by the characterization of stem cells by only a long-term cultore initiating dilution assay, and not by phenotype analysis or an in vivo model of engraftment. The use of rapid medium exchange in standard cell culture results in a sixfold expansion of human bone marrow progenitor-colony forming cells over two weeks (Schwartz et al, Proc. Natl. Acad. Sci. U.S.A. 55:6760 (1991); Schwartz et al, Blood 78:3155 (1991)). Maintenance of low oxygen concentration may also facilitate expansion of human cord blood progenitor cells (Koller et al, Blood 80:403 (1992)). We have demonstrated more rapid cell proliferation than these studies by use of fetal hematopoietic cells and artificial capillary technology.

Multilineage reconstitotion in an animal model is the best assay to determine the presence of true HSC. For the assay of human HSC, various animal models of human hematopoiesis have been developed, such as the SCID-hu mouse (Kyoizumi et al, Blood 79:1704 (1992)). In previous studies, we have found that the persistence of human engraftment in fetal lambs more than 30 days after the in utero human HSC transplant correlates with permanent, multilineage engraftment and confirms the transplantation of stem cells (Zanjani et al, J. Clin. Invest. 59:1178 (1992)). Furthermore, in this model the subsequent transplantation of recipient sheep bone marrow containing donor cells into a second fetal lamb resulted in human engraftment, confirming the presence of long term repopulating pluripotent human HSC (Zanjani et al, Blood 80:244 (1992)). In our current study, we have demonstrated human engraftment in the sheep more than 60 days after the transplant, supporting the transplantation of human HSC.

There is a high rate of microbial contamination of human fetal livers that are obtained from aborted tissue (Rice et al, Fetal Diag. Ther. 5:74 (1992)). However, an ACS system could support and expand the fetal cells while the tissue is undergoing screening for microbial contaminants. Although fetal liver transplantation has resulted in therapeutical solutions for patients with a variety of diseases (Touraine et al, Thymus 10:15-81 (1987); Touraine, J-L., Blood Cells 17:379-387 (1991)), more careful control of infectious contaminants will be essential for successful reinfiision of ex vivo expanded fetal stem cells.

Our report describes the successful expansion of hematopoietic stem cells harvested from human fetal livers in an artificial capillary system and suggest several areas of future investigations. Identification of the ideal cytokine milieu may allow the selective expansion of HSC without differentiation. The careful construction of fetal stroma in an ACS may closely approximate the fetal hematopoietic support network and may improve the rate of stem cell expansion. This cultore system should facilitate the study of hematopoietic stem cell biology and support other areas of study, such as the use of fetal stem cells as targets for gene therapy.

Example 2

In utero Fetal HSC Transplantation

A. In utero fetal HSC transplantation into a fetus suffering from bare lymphocyte syndrome (BLS).

Prenatal Diagnosis of BLS Prenatal diagnosis of BLS is performed by HLA analysis of fetal blood lymphocytes as described in Durandy et al, Prenat Diagn 7:27-31(1987). Briefly, at about 15 weeks after fertilization, a fetal blood sample (1.5 ml) is obtained by direct umbilical cord puncture with a 20G spinal needle, under ultrasonic visualization. The purity of the fetal blood sample is determined by Ii typing. Fetal blood lymphocytes are analyzed for HLA antigen expression using fluorescent monoclonal antibodies and a cytofluorometer. BLS is diagnosed by a lack of expression of both class I and II HLA antigens at the cell surface when contrasted with findings in immunologically normal fetuses.

Ex vivo expansion of fetal liver HSC

Fetal liver HSC are harvested from fetal liver and expanded ex vivo in an ACS as described in Example 1.

Transplantation of expanded fetal liver HSC into the fetus Ex vivo expanded fetal liver HSC are transplanted into the fetus suffering from BLS using previously described techniques (Touraine, J-L., Blood Cells 77:379-387(1991); Touraine et al, Thymus 10:75-87(1987); Flake et al., Exp Hematol 19:1061(1991)). Briefly, at 15 weeks after fertilization, a fetal blood puncture is performed at the insertion of the umbilical vein on the placenta, under direct ultrasonic visualization. Fetal liver HSC transplantation is performed by infusion in the umbilical vein approximately 5 ml of cultore medium containing approximately IO ⁶ - IO ⁷ fetal liver cells containing the expanded HSC.

Alternatively, before transplantation, the expanded fetal HSC are separated from the other fetal liver cells using anti-CD34 ⁺ monoclonal antibody cell sorting techniques as described in Brandt et al, J Clin Invest 56:932(1990); Strour et al, Blood 57:661(1993); Issaad et al, Blood 57:2916(1993). The separated fetal liver HSC are then suspended in approximately 5 ml of culture. The 5 ml suspension, containing approximately IO ⁵ - 10 ⁶ fetal HSC, is then infused into the umbilical vein as described above. No attempt at HLA matching of the donor and recipient is needed. After the patient is born, graft "take" is easily documented by identification of cells with an HLA phenotype different from that of the host.

B. In utero fetal HSC transplantation into a fetus suffering from severe combined immunodeficiency disease (SCID).

Prenatal Diagnosis of SCID

Prenatal diagnosis of SCID is performed as described in Touraine, J-L., Blood Cells 17: 379-387 (1991). Briefly, at approximately 15 weeks after fertilization, a fetal blood sample (1.5 ml) is obtained by direct umbilical cord

punctore with a 20G spinal needle, under ultrasonic visualization. The purity of the fetal blood sample is determined by Ii typing. Fetal blood lymphocytes are analyzed for CD2 ⁺ , CD3 ⁺ , CD4 ⁺ , and CD8 ⁺ antigens. SCID is diagnosed by a lack of these antigens as compared with findings in immunologically normal fetuses.

Harvesting and ex vivo expansion of fetal liver HSC

Fetal liver HSC are harvested from fetal liver and expanded ex vivo in an ACS as described in Example 1.

Transplantation of fetal liver HSC into the fetus Ex vivo expanded fetal liver HSC are transplanted into the fetus suffering from SCID as described in Example 2(A) above. Briefly, at 15 weeks after fertilization, approximately IO ⁶ - IO ⁷ fetal liver cells containing expanded HSC are infused intravenously through the umbilical vein under ultrasonic control.

Alternatively, before transplantation, the expanded fetal HSC are purified from the other fetal liver cells as described above. At 15 weeks after fertilization, approximately 10 ⁵ - 10 ⁶ separated fetal liver HSC are infused intravenously into the umbilical vein under ultrasonic control.

No attempt at HLA matching of the donor and the recipient is needed. After the patient is born, graft "take" is easily documented by identification of cells with an HLA phenotype different from that of the host.

Example 3

Postnatal Fetal HSC Transplantation

Harvesting and ex vivo expansion of fetal liver HSC

Fetal liver HSC are harvested from fetal liver and expanded ex vivo in an ACS as described in Example 1.

Transplantation of expanded fetal liver HSC into a patient suffering from SCID

When no HLA-identical donor is available for BMT, transplantation with ex vivo expanded fetal liver HSC is useful for treatment of SCID patients. The age of the fetal donor is approximately 10-15 weeks post-fertilization. No attempt at HLA matching of the donor and recipient is needed. The transplant consists of intravenous infusion into a patient suffering from SCID a suspension of fetal liver cells containing the ex vivo expanded HSC. The number of fetal liver cells transplanted to the patient is about 10 X 10 ⁸ cells, i.e. about 1-2 X 10 ⁸ per kg of body weight. Alternatively, before transplantation, the expanded HSC are separated from the other fetal liver cells as described above. Then, approximately IO ⁵ - 10 ⁶ of purified fetal HSC are injected into the patient.

For optimal graft "take", this transplant is repeated five to ten times with each administration at 2-3 day intervals.

Example 4

Transduction of ex vivo Expanding Fetal Liver HSC with Packaged Retroviral Vectors

Preparation of Packaged LASN

LASN is a retrovirus vector containing the ADA and Neo resistance genes. High concentrations of packaged LASN retrovirus particles are produced in the supernatant of LASN-producing PA-317 packaging cells (ATCC Accession No. CRL 9078) as described in Hock et al, Blood 74:876-881(1989). About 5 ml of supernatant from the packaging cells is filtered through a POLYDISC™ AS filter (Whatman Ltd. , Maidstone, England) and collected for transduction.

Transduction of Target Fetal HSC with Packaged LASN

Active target cell cycling is required for retroviral integration (Nolta et al, Exp Hematol 20:1065(1992). Thus, target fetal HSC are harvested from fetal liver and inoculated into an ACS for ex vivo expansion as described in Example 1. After the cultore is established, the expanded fetal HSC are removed in a laminar flow hood by gently shaking the ACS cartridge. About 20% of the cells are separated to serve as non-transduced controls. The remaining cells are pelleted and resuspended in 45 ml of the LASN filtrate described above. The resuspension, containing packaged LASN and target fetal HSC, is reinoculated into the ACS cartridge. Expansion of the target fetal HSC, in the presence of packaged LASN, is then resumed in the ACS using the culturing conditions described in Example 1. After two days, the transduction procedure is repeated with a second volume of LASN filtrate. After two more days, the transduction procedure is again repeated. Before each transduction procedure, samples of expanded fetal HSC are removed from the ACS for analysis. The cultore is terminated after about eight days.

Detection of LASN in Target fetal HSC

The presence of LASN in transduced target fetal HSC after expansion in the ACS is confirmed using the polymerase chain reaction (PCR). PCR analysis is caπied out using GeneAmp ^R reagents and DNA Thermal Cycler (Perkin Elmer Cetos, Emeryville, CA). DNA is isolated from the transduced and non-transduced fetal HSC described above according to conventional techniques. PCR is initiated with 1-2 μg of genomic DNA using primers flanking the Neo resistance gene which is contained in LASN. The DNA sequences of the primers are as follows: CAAGATGGATTGCACGCAGG

CCCGCTCAGAAGAACTCGTC The reaction mixture is heated at 94°C for 2 min, annealed at 56°C for 2 min, and extended at 72°C for 3 min in the DNA Thermal Cycler for 30 cycles. The products of the reaction are loaded and run on a gel and probed with a Neo- resistance gene specific probe.

It will be appreciated to those skilled in the art that the invention can be performed within a wide range of equivalent parameters of composition, concentrations, modes of administration, and conditions without departing from the spirit or scope of the invention or any embodiment thereof. The disclosure of all references, patent applications and patents recited herein are hereby incorporated by reference.