Fred Hutchinson Cancer Research Center, and the Department of Medicine, University of Washington School of Medicine, Seattle, Washington, USA

Transplantation of hematopoietic stem cells, usually in the form of marrow grafts, has been increasingly used over the last 21/2 decades to treat patients with various malignant and nonmalignant hematological diseases. The number of transplants is steadily increasing worldwide. Marrow for transplantation is usually obtained by aspiration from the iliac crests. It is placed into tissue culture medium and screened before infusion into the recipient. The marrow is composed of a mixture of cells including large numbers of differentiated and mature cells, many committed progenitors with defined function and limited proliferation potential, and rare pluripotent stem cells that have broad developmental and proliferative potential. The ability of the pluripotent stem cells not only to self-renew, but also to differentiate into the various committed hematopoietic precursors, has made marrow transplantation possible. After the marrow harvest and screening, the marrow is placed into a blood administration bag and infused intravenously into the recipient whose underlying disease has been eradicated and whose immune system suppressed by high doses of chemoradiotherapy .Infused stem cells circulate through the blood stream and, presumably with the help of homing receptors, enter the marrow cavity , attach to the stroma bed, and begin to divide under the influence of humoral and cellular stimuli. In most recipients of HLA-identical sibling transplants, the cell inoculum settles in uneventfully. With extraordinary speed, the progenitor cells produce progeny 0 The first mature cells begin to appear in the circulation within 2-3 weeks of transplantation, normal granulocyte and platelet counts may be reached as early as day 40 to 50 after transplant, and normal hematocrit values are seen between the second and third months after transplantation. In most patients, hematopoiesis is entirely of marrow donor origin, although a minority may show a mixture of host and donor cells. Hematopoiesis is stable. The longest survivors after marrow transplantation have passed the 25-year mark, and they display a completely normal hematopoietic and immune system of donor type. Recent studies using molecular probes for methylation site polymorphisms on the inactivated X-chromosome have shown that hematopoietic repopulation after transplantation is of polyclonal origin, although one group of investigators reported clonal origin in a small minority of patients (1,2). Factors governing the fate of the marrow transplant include the immunosuppressive effects of the conditioning programs, the degree of histoincompatibility between donor and recipient, the composition of the cells in the graft, and the quality of the marrow bed. Inadequacies with any of these factors may lead to a problem, graft failure. With HLA-identical sibling donors, most grafts are successful. Graft failure is seen in less than 2% of recipients. In contrast with HLA-haploidentical family members who are mismatched for 1, 2 or 3 HLA-Ioci on the nonshared haplotype, graft rejection is seen in 7-20% of cases (3,4). In almost 80% of patients with graft failure, residual host lymphocytes could be seen in the peripheral blood, which is suggestive of host-mediated immune rejection. Graft failure is accompanied by a case fatality rate of 97%. Studies in a canine model of marrow transplantation had predicted that engraftment problems might arise in the HLA-nonidentical transplant setting (5). While grafts with marrow from DLA-identical littermates were successful in 95% of the cases, transplants from DLA-haploidentical littermates or DLA-nonidentical unrelated dogs had graft failure rates of 92% and 95 % 0 , respectively. At the time of rejection, dogs showed an increase in peripheral blood large granular lymphocytes of host type which functioned as natural killer cells (NK cells) (6). In coculture experiments, these recipient cells suppressed marrow colony-forming unit granulocytemacrophage production, not only from donor marrow but also from marrow of unrelated dogs that were either phenotypically DLA-identical or DLA-nonidentical with the marrow donor, consistent with the notion of cytotoxicity which was not restricted by the major histocompatibility complex (MHC) (Raff et al., unpublished). In agreement with the notion that NK-Iike cells were involved in graft rejection was the finding that agents directed against macrophages and T lymphocytes were ineffective in enhancing marrow engraftmentacross MHC barriers in the dog. Accordingly, we developed monoclonal antibodies (MAb) that recognized those host cells which mediate rejection. To this purpose we harvested canine marrow cells 6 days after TBI and immunized mice. Among the antibodies produced, one, S5, had a broad hematopoietic distribution, binding to normal marrow cells, peripheral blood mononuclear cells, granulocytes and lymph node lymphocytes. The antibody was used in a controlled in vivo experiment in which dogs were given 9.2 Gy of TBI followed by DLA-nonidentical unrelated marrow grafts (7). In this setting, graft failure was seen in 33 of 36 control dogs not given additional therapy, in 6 of 7 dogs given the isotype matched irrelevant MAb 6.4 before TBI, and in only 4 of 19 dogs given antibody S5 at a dose of 0.2 mg/kg/day from day -7 to day -2 before TBI. Thus, antibody S5 resulted in a significant enhancement of engraftment over that seen in control dogs given either no antibody or an isotype matched irrelevant antibody. Subsequent work using techniques of sequential immunoprecipitation, tryptic digestion, and N-glyconase digestion revealed antibody 55 to detect the CD44 antigen (8). CD44 functions in cell-cell and extracellular matrix adhesions and has been implicated in lymphocyte traffic. It is also involved in T cell activation via CD2 and CD3. Binding of anti-CD44 monoclonal antibodies causes secretion of IL-l and TNF-alfa. Further studies using positive selection have shown CD44 to be expressed on hematopoietic stem cells that can repopulate dog marrow after a supralethal dose of TBI. The mechanism by which an antibody to CD44 enhances engraftment across a major histocompatibility barrier is as yet unexplained. Alternative modes of action include that MAb 55 affects traffic of host immune cells, thereby permitting marrow to home, or increases the radiosensitivity of host NK cells thought to be involved in graft resistance. Elucidation of these mechanisms is important for clinical application of antibody-mediated enhancement of marrow grafts. Previous studies in dogs indicated that hematopoietic stem cells express MHC class II molecules as recognized by the anti-class II MAb 7.2 (9,10). Treatment of canine marrow in vitro with antibody 7.2 and rabbit complement prevents autologous marrow engraftment after lethal TBI, whereas selected 7.2-positive marrow cells are capable of permanently repopulating the hematopoietic system of irradiated dogs. Class II molecules are differentially expressed on hematopoietic progenitor cells of man. Committed progenitor cells assayed as colony-forming unit granulocyte-macrophage, mixed colonies, and burst-forming unit erythroid express HLA-DR and DP antigens whereas colony-forming units erythroid are predominantly HLA-DR negative. Low HLA-DR expression has been seen on murine and human long-term marrow culture initiating cells. It is not clear whether truly pluripotent stem cells express class II MHC antigens. If present, the class II molecules on stem cells could have significant clinical implications, because investigators have described a possible role of HLA-D restricted cell interactions in hematopoiesis. Such restrictions could be important for interactions between transplantable cells and as yet undetected accessory cells in the hematopoietic microenvironment. We have established a model of canine marrow autografts after 9.2 ay TBI to study the role of class II antigens in hematopoietic stem cell growth and differentiation (11). Twenty dogs were given 9.2 ay TBI marrow and intravenous murine anti-class II MAb. Infusion of 0.6 mg/kg/day of antibody H81.98.21, an Iga2a MAb reactive with HLA-DR, on days 0 to 4 after TBI did not prevent initial engraftment, but all dogs died with late graft failure. The critical time for effect of the antibody is during the first 4 days after transplant. Results of extensive studies argued against several pathogenetic mechanisms, including removal of antibody coated stem cells by the reticuloendothelial system, canine complement-mediated cytotoxic effects on stem cells, antibody-dependent cellular cytotoxicity , and inactivation of antibody coated cells by dog anti-mouse antibody. To distinguish between antibody-induced damage to microenvironment/accessory cells and late graft failure from lack of pluripotent stem cells, three dogs were given TBI, a marrow autograft, and antibody H81.98.21 on days 0 to 4; one given thoracic duct lymphocytes on day 6 developed graft failure; the other two given marrow depleted of accessory cells by L-Ieucil L-Ieucine O-methyl ester had sustained grafts. Findings support the notion that originally transplanted pluripotent stem cells are no longer present on day 6 and that the marrow microenvironment is functional and able to support newly injected stem cells. These initial observations on the development of graft failure after injection of MAb to class II during the early period after transplant suggest a role for class II molecules in the engraftment and/or regulation of pluripotent stem cells required for sustained marrow function. The mechanism by which MAb to class II induce late marrow graft failure is currently under investigation (12). Hematopoietic engraftment of donor type is "permanent" in human patients as indicated by blood genetic marker studies in patients, many of whom have now been followed for 10-25 years after transplant. Given the permanence of the engraftment and the fact that recovery in most patients is polyclonal, gene transfer into hematopoietic stem cells has become an attractive possibility .Gene transfer might be used to inhibit or alter genes causing disease, for example, in patients with hemoglobinopathies or inborn errors of metabolism, to confer drug resistance to marrow cells in patients undergoing high-dose chemotherapy, and to transfer histocompatibility genes with the aim of inducing tolerance to MHC gene products. For gene therapy to be accomplished, a sample of the patient's marrow would be obtained, placed into a long-term culture system, and exposed to a retroviral vector containing the genes of interest. In the interim, the patient would be treated to eradicate the underlying disease. That accomplished, the transduced marrow would be injected in hopes that genes would be expressed in the progeny of hematopoietic progenitor cells. Our work was conducted in a preclinical canine model using two Moloney leukemia virus based vectors, LNCA and LASN, which contain both the neomycin phosphotransferase gene and a human adenosine deaminase gene (13,14). The system was helper virus free. Two approaches were taken. In one, prospective recipient dogs were treated with canine recombinant granulocyte colony-stimulating factor for 7 days. Then, marrow was harvested and placed into long-term culture which was regularly fed with supernatant from a virusproducing feeder layer. After the marrow was in culture for 4 days, the dog was given 9.2 ay of TBI and the transduced marrow infused. The dog was then treated with granulocyte colony-stimulating factor until hematopoietic engraftment. The other approach involved pretreatment of the recipient dog with a dose of 40 mg of cyclophosphamide per kg intravenously, a maneuver which led to profound pancytopenia. At the time of greatest pancytopenia, marrow was harvested and cocultivated for 48 hours with a viral-producing feeder layer. Then the dog was given 9.2 Gy of TBI and the marrow was infused. The dog then received granulocyte colony-stimulating factor until recovery of peripheral blood granulocyte counts. Six dogs were treated with long-term cultured marrow. Of these six, two died with intercurrent infection, on the day of TBI and marrow transplantation and on day 15, respectively, while four became long-term survivors. The four are surviving between 2 and more than 4 years after transplant. Of the four animals treated with cyclophosphamide, two died with pneumonia on days 1 and 12 after TBI, respectively, and two are surviving between 2 and more than 4 years after transplant. At regular intervals, marrow has been harvested from these dogs and placed into colony-forming unit granulocytemacrophage cultures with and without the chemical G418 to test for neomycin resistance. For now up to more than 4 years after transplant, between 1% and 10% G418-resistant colony-forming units granulocytemacrophage have been found in each of the six surviving animals, indicating successful gene transfer . Furthermore, peripheral blood granulocytes, peripheral blood lymphocytes, lymph node lymphocytes, and marrow cells have all been tested for both the presence of the neomycin phosphotransferase gene and the adenosine deaminase gene using a polymerase chain reaction based test. The genes were found to be present for up to more than 4 years. Dogs have remained healthy. Tests for helper virus have remained negative. These data indicate for the first time successful and persistent transduction of long-term repopulating marrow cells in a large random bred species. It is clear that much work still needs to be done to increase the relatively low levels of gene expression in vivo through improvements in the development of both retroviral vectors and of packaging cell lines so that the highest possible virus titers can be generated in the absence of any detectable helper virus. Perhaps results can be improved through the use of various hematopoietic growth factors. Much work needs to be done in regards to developing safe conditioning programs for the recipient which do not cause life threatening pancytopenias. As our understanding of gene transfer techniques increases and our knowledge of gene regulation in the course of differentiation improves, gene transfer for the treatment of hematopoietic disorders may one day become clinical reality .

ACKNOWLEDGMENTS

This work was supported by grants HL36444, CA18221, CA31787, CA18029, and CA15704 from the National Institutes of Health, DHHS.

REFERENCES

1. Nash R, Storb R, Neiman P. Polyclonal reconstitution of human marrow after allogeneic bone marrow transplantation. Blood 1988;72:2031-2037.
2. Turhan AG, Humphries RK, Phillips GL, Eaves AC, Eaves CJ. Clonal hematopoiesis demonstrated by X-Iinked DNA polymorphisms after allogeneic bone marrow transplantation. N Engl J Med 1989;320: 1655-1661.
3. Beatty PG, Clift RA, Mickelson EM et al. Marrow transplantation from related donors other than HLA-identical siblings. N Engl J Med 1985;313: 765-771.
4. Anasetti C, Amos D, Beatty PG et al. Effect ofHLA compatibility on engraftment of bone marrow transplants in patients with leukemia or lymphoma. N Engl J Med 1989;320: 197-204.
5. Storb R, Deeg HJ. Failure of allogeneic canine marrow grafts after total body irradiation: Allogeneic "resistance" vs transfusion induced sensitization. Transplantation 1986;42:571-580.
6. Raff RF, Deeg HJ, Loughran TP,Jr. et al. Characterization of host cells involved in resistance to marrow grafts in dogs transplanted from unrelated DLA-nonidentical donors. Blood 1986;68:861-868.
7. Schuening F, Storb R, Goehle Set al. Facilitation of engraftment ofDLA-nonidentical marrow by treatment of recipients with monoclonal antibody directed against marrow cells surviving radiation. Transplantation 1987;44:607-613.
8. Sandmaier BM, Storb R, Appelbaum FR et al. An antibody that facilitates hematopoietic engraftment recognizes CD44. Blood 1990;76:630-635.
9. Szer J, Deeg HJ, Appelbaum FR, Storb R. Failure of autologous marrow reconstitution after cytolytic treatment of marrow with anti-Ia monoclonal antibody. Blood 1985;65:819-822.
10. Schuening F, Storb R, Goehle Set al. Canine pluripotent hemopoietic stem cells and cFu-GM express la-Iike antigens as recognized by two different class-II specific monoclonal antibodies. Blood 1987;69:165-172.
11. areinix HT, Ladiges WC, Graham TC et al. Late failure of autologous marrow grafts in lethally irradiated dogs given anti-class-1I monoclonal antibody. Blood 1991;78:2131-2138.
12. areinix HT, Storb R, Bartelmez SH. Specific growth inhibition of primitive hematopoietic progenitor cells mediated through monoclonal antibody binding to major histocompatibility class II molecules. Blood 1992;80:1950-1956.
13. Schuening Fa, Storb R, Stead RB et al. Improved retroviral transfer of genes into canine hematopoietic progenitor cells kept in long-term marrow culture. Blood 1989;74:152-155.
14. Schuening Fa, Kawahara K, Miller DA et al. Retrovirus-mediated gene transduction into long-term repopulating marrow cells of dogs. Blood 1991;78:2568-2576.