AU7340000A - Transgenic mammal capable of facilitating production of donor-specific functional immunity - Google Patents

Description

WO 01/15521 PCT/USOO/23971 TRANSGENIC MAMMAL CAPABLE OF FACILITATING PRODUCTION OF DONOR-SPECIFIC FUNCTIONAL IMMUNITY 5 STATEMENT AS TO RIGHTS TO INVENTIONS MADE UNDER FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT This invention was made with Government support. The Government has certain rights in this invention. 10 BACKGROUND OF THE INVENTION Many human diseases remain incurable in large part due to the lack of an appropriate model system for preclinical studies. Since many diseases are specific to either human pathogens or dysfunctional human tissues, it is difficult to model the course of such afflictions outside of the human body. For example, the basis of allergic responses is deeply 15 rooted in the genetics of the host and cannot be completely studied in a different species. Infectious diseases, such as HIV, have species-specific virulence factors. And cancer cells that arise from a combination of genetic factors usually display altered properties when transplanted into immunodeficient animals. Unfortunately, there are few methods for directly studying the pathology of 20 human diseases. This in turn limits the development of new drugs and novel therapies. Given the practical and ethical restrictions of experimenting in both humans and higher primates, there is an urgent need to develop alternative models of human diseases. In models of human disease where an interaction between the disease causing agent and the immune system is suspected, either hematopoietic stem cells or mature 25 circulating lymphocytes are transferred into naturally occurring strains of immunodeficient mice. Although better than their forerunners in certain respects, these models fail to reproduce many of the functional properties of human cells that are critical for unraveling disease processes. On a more basic level, even attempts to transplant hematopoietic stem cells between individuals of the same species have produced allogeneic chimeras that are 30 functionally impaired. The reasons for this are unclear, but involve the inability of the donor stem cells to differentiate properly in the mature lymphoid tissues of the new host. In an attempt to overcome these problems, researchers have added IL-7, either exogenously or transgenically to mice before engraftment. However, this approach unexpectedly led to further immunological dysfunction. For example, see, Kapp, et al., Blood WO 01/15521 PCTIUSOO/23971 -2 92:2024 (1998) (exogenous IL-7 led to decrease in B cell development); Rich, et al., J. Exp. Med. 177:305 (1993) (transgenic IL-7 under the control of immunogloubulin heavy chain promoter and enhancer led to dermal lymphoid infiltration and T and B cell lymphomas); Valenzona, et al., Exp. Hematol. 24:1521 (1996) (IL-7 under the MHC class II promoter 5 induced B lymphoid tumors); Watanabe, et al., J. Exp. Med. 187:389 (1998) (IL-7 transgenic mice developed chronic colitis); Uehira, et al., J. Invest. Dermatol. 110:740 (1998) (IL-7 transgenic mice developed dermatitis);and Mertsching, et al., Eur. J. Immunol. 26:28 (1996) (IL-7 transgenic mice developed lymphoproliferative disease). Thus, there remains a need for a standard transgenic animal model system that 10 supports the functional properties of human hematopoietic cells. This invention meets this and other needs. SUMMARY OF THE INVENTION The present invention provides for a recipient mammal comprising exogenous 15 cytokines capable of maintaining an immune system; and cells derived from donor-specific cells with hematopoietic stem cell properties (HSC); wherein the recipient mammal is capable of facilitating production of donor-specific functional immunity. In a preferred embodiment, the mammal is a mouse. In one aspect of this embodiment, the cytokines are from the same species as 20 the cells derived from donor-specific cells with hematopoietic stem cell properties. The cytokines are selected from the group consisting of interleukin 3, (IL-3), interleukin-6 (IL-6), interleukin-7 (IL-7), macrophage-colony stimulating factor (M-CSF), granulocyte-colony stimulating factor (GM-CSF), stem cell factor (SCF), leukemia inhibitory factor (LIF) and oncostatin M (OM). In a preferred aspect of this embodiment, the cytokines comprise IL-3, 25 IL-6, IL-7, M-CSF, GM-CSF and SCF. In another preferred aspect of this embodiment, the cytokines are introduced into the mammal transgenically. In another preferred aspect of this embodiment, the cells derived from cells with hematopoietic stem cell properties are xenogeneic to the mammal. In a particularly preferred aspect, the cells are human cells. In a more preferred aspect, the cells are hematopoietic stem 30 cells. In a most preferred aspect, the hematopoietic stem cells are derived from umbilical cord blood. However, in alternative aspects, the cells are derived from bone marrow, mobilized peripheral blood, embryonic stem cells, or other source of HSC. In another alternative embodiment, the cells derived from cells with hematopoietic stem cell properties are WO 01/15521 PCTUSOO/23971 -3 differentiated. In particular, the differentiated cells are of lymphoid, myeloid, or of an erythroid lineage. In another embodiment of this invention, a method of making a mammal with a donor immune system is provided. This method comprises the steps of introducing transgenes 5 into an immunodeficient mammal, wherein the transgenes encode cytokines necessary for the maintenance and maturation of donor-derived hematopoietic cells; and introducing cells with hematopoietic stem cell properties into the mammal. In one aspect of this embodiment the donor immune system is a xenogeneic immune system. In a particularly preferred aspect of this embodiment, the donor immune 10 system is a human immune system. In one aspect of this embodiment, the introduction of transgenes is through transfection of embryonic stem cells. In a second aspect of this embodiment, the introduction of transgenes is through pronuclear transfer. In an alternative aspect of this embodiment, the introduction of the transgenes is through breeding the mammal with a mammal that comprises is the transgenes. In a preferred aspect of this embodiment, the mammal is a RAG-1 or a RAG-2 mutant mouse. In another aspect of the invention, the mammal is a RAG-1 or RAG-2 mutant mouse expressing human leukocyte antigen (HLA) Class I and/or Class II genes. In a further aspect of the invention, the mammal is a SCID mouse expressing HLA Class I and/or Class II 20 genes. In yet another aspect of the invention, the mammal is an immunocompetent mouse expressing HLA Class I and/or Class II genes and rendered immunodeficient by, e.g., irradiation conditioning. In another aspect of this invention the donor-derived hematopoietic cells are from a xenogeneic mammal. However, in a particularly preferred aspect, the donor-derived 25 hematopoietic cells are from a human. In yet another embodiment of this invention, a method of determining an immune response to an antigen is provided. Transgenic chimeric mammals are immunized with proteins, peptides, cells or other sources of antigens, to determine epitopes involved in donor cell-derived immune responses. These include, but are not limited to, antigen-specific 30 immunoglobulin production, Thelpr responses, Tcytotoxic responses, cellular proliferation responses, innate allogeneic or xenogeneic responses, and natural killer cell activity.

WO 01/15521 PCT/US00/23971 -4 BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 demonstrates that allogeneic bone marrow engrafted RAG mice are tolerant to donor and host MHC, but responsive to third party alloantigens. CD4' T cells were isolated by cytotoxic elimination of class II* and CD8* cells from the lymph nodes of a 5 syngeneic engrafted RAG mouse (RAG-(syn), (A); an allogeneic engrafted RAG mouse (RAG-(allo), B); a syngeneic engrafted SCID mouse (SCID-(syn), C); an allogeneic engrafted SCID mouse (SCID-(allo), D); and a RAG mouse, E) engrafted with the same bone marrow preparation as placed in the SCID-(allo) mouse shown in panel D, as a positive control for the bone marrow inoculum. 10 CD4* T cells were co-cultured with irradiated LPS-induced splenic blasts from Balb/C (diamonds); C57B1/6 (squares); CBA (circles panels A and B) or (C57Bl/6 x CBA)F1 mice (circles, panels C, D and E). Proliferation was assessed colorometrically (Celltiter; Promega) on day (d) =5, and is reported as OD 4 90 x 1000 on the y axis. Background absorption has been subtracted. 15 Figure 2 demonstrates antigen specific IgG responses by RAG-(allo) mice. In Figure 2A, control Balb/c mice (open triangles), RAG-(allo) mice (closed squares; mouse #30), RAG-(syn) mice (open squares; mouse RN003), and SCID-(allo) mice (closed circles; two animals, SN005 and SNO06, are shown) were immunized in the hind foot pads with a total of 50 gg hen egg white lysozyme (HEL) emulsified in complete Freund's adjuvant (CFA). Two 20 weeks later, animals were boosted i.p. with the same amount of HEL in incomplete Freund's adjuvant (IFA). One week after boosting, the animals were bled and the serum was tested for the presence of HEL-specific IgG by ELISA. The y axis represents OD 4 15 .4 9 0 x 1000. In Figure 2 B, RAG-(syn) mice (striped bars; mice #61 and 62) and RAG-(allo) mice (speckled bars; mice #46 and 51) were immunized in the hind foot pads with a total of 50gg of KLH 25 emulsified in CFA on d=0. The animals were boosted subcutaneously two weeks later with KLH in IFA. Serum samples were taken at d=0, 14 and 21 days, and then tested for KLH specific IgG by ELISA. The plain bars on the graph represent individual control mice: (C57B1/6x129) F1 mice are represented by the open bars, and Balb/c mice by the gray bars. The y axis indicated the OD415490 x 1000 for a 1:1000 dilution of serum. The x axis represents 30 days post-primary immunization. Specificity was tested by ELISA on HEL coated plates; no cross-reactivity was seen. Figure 3 represents that antigen specific T cell proliferative responses are restricted to both donor and host MHC in neonatally constructed RAG-(allo) chimeras. RAG- WO 01/15521 PCT/US00/23971 -5 (allo) mice #135 (A) and 136 (B) were immunized in the hind foot pad with KLH in CFA. Ten days later, draining lymph nodes were removed and depleted of B cells and macrophages by cytotoxic elimination. The resulting lymph node T (LNT) cells were co-cultured for three days with a 2:1 ratio of Mitomycin C-fixed, antigen-pulsed LPS-blasts from Balb/c and C57B11/6 5 mice. Proliferative responses were quantitated using a colorometric assay (CellTiter; Promega). Background responses were subtracted. LNT from immunized Balb/c mice gave

OD

490 x 1000 = 547 in response to antigen pulsed Balb/c blasts. DETAILED DESCRIPTION OF THE INVENTION 10 I. INTRODUCTION It has been well established using a variety of model systems that thymic cortex epithelial cells perform the majority of the positive selection events that occur during T cell differentiation (Paul, ed. FUNDAMENTAL IMMUNOLOGY FOURTH EDITION (1999), Lipincott Raven Press). Recently, attempts to further define the cell types involved in positive selection is have revealed a dichotomy in the ability of CD4* and CD8* single positive cells to be selected by bone marrow (BM)-derived cells. It has been conclusively demonstrated that CD8* T cells can be positively selected by hematopoietic cells using chimeric animals constructed on an MHC class I deficient background (Bix & Raulet, Nature 359:330-333 (1992)). However, the opposite result has been shown for CD4* T cells using a similar model system employing 20 irradiated MHC class II deficient mice (Markowitz, et al., Proc. Nat'l Acad. Sci. 90(7):2779-83 (1993)). These results suggest that selection events are more stringently controlled for CD4* than for CD8* T cells. The inventors have found that fully allogeneic chimeric animals generated either directly after birth, or in adult, non-irradiated antigen receptor recombination-deficient, 25 e.g., recombination activation gene-2 (RAG-2) mutant, mice possess CD4' T cells in the periphery that exhibit donor MHC restricted antigen-specific responses. These results have not been seen in either neonatally or adult constructed SCID chimeras. This suggests that hematopoietic cells are capable of positively selecting CD4* T cells in the thymus, and present antigen receptor recombination-deficient strains of mammals as a unique model system which 30 may support T cell development more closely resembling normal ontogeny. It appears that positive selection of CD4' T cells by hematopoietic cells has not been routinely detected in other systems due to the use of incompletely immunoincompetent mice, and/or due to secondary effects of irradiation.

WO 01/15521 PCT/USOO/23971 -6 The inventors have further discovered methods by which xenogeneic transgenes required for the growth and development of a xenogeneic hematopoietic stem cells are incorporated into the host mammal. After incorporation, cytokine transgenic (CTG) mammals engrafted with xenogeneic hematopoietic stem cells (HSC) develop a functional immune 5 system capable of donor MHC-restricted antigen-specific responses. This modification provides a pathway for donor lymphocyte development in the context of xenogeneic MHC molecules expressed on the MHC-expressing tissues of the host. These mammals can then be used as a model system for human or other mammalian diseases. 10 II. DEFINITIONS Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. All references are incorporated by reference for all purposes. Although any methods and materials similar or equivalent to those described herein can be used in the is practice or testing of the present invention, the preferred methods and materials are described. For purposes of the present invention, the following terms are defined below. The phrase "major histocompatibility complex" (MHC) refers to immune response genes that encode cell surface glycoproteins that regulate interactions among cells of the immune system. The genes were discovered as a result of their involvement in graft 20 rejection. There are two main classes of MHC genes, Class I and Class II. The phrase "human leukocyte antigen" (HLA) refers to the MHC complex of humans. The phrase "MHC restriction" refers to the recognition of peptides by T cells in the context of particular allelic forms of MHC molecules. For a more complete description of the MHC complex in humans, as well as in mice, see,, FUNDAMENTAL IMMUNOLOGY, 4TH ED., Paul (ed.) 1999. 25 Cells that are "allogeneic" to a mammal are cells that are from an individual of the same species as the mammal but, because of differences in expression of major and minor histocompatibility molecules between the cell donor and the host mammal, are recognized by the host mammal as non-self. Cells that are "xenogeneic" to a host mammal are cells that are from an 30 individual of a different species as the mammal. Due to significant genetic differences, they are recognized by the host mammal as non-self. The phrase "bone marrow" refers to the red marrow of the bones of the spine, sternum, ribs, clavicle, scapula, pelvis and skull. This marrow contains hematopoietic stem WO 01/15521 PCT/USOO/23971 -7 cells. The phrase "umbilical cord blood" refers to whole blood obtained from the umbilical cord of a newborn. This blood also contains hematopoietic stem cells. The phrase "mobilized peripheral blood" refers to peripheral blood isolated from individuals treated with recombinant growth factors, e.g., granulocyte colony stimulating factor (GM-CSF) and stem cell factor 5 (SCF), for the purpose of increasing the proportion of hematopoietic stem cells in the circulation. The term "cytokines" refers to proteins that are commonly referred to as cytokines as well as other proteins, such as growth factors, interleukins, immune system modulators, and other types of proteins necessary to maintain an immune system. For io example, cytokines encompasses the interleukins, stem cell factors, colony stimulating factors and other factors known to those of skill in the art. "Exogenous cytokines" refers to cytokines that are not naturally occurring in the recipient mammal. These cytokines can be species homologs of naturally occurring cytokines or cytokines that do not have naturally occurring homologs in the recipient mammal. 15 The term "immunodeficiency" refers to a lack of antigen-specific immunity in a mammal. In these mammals, B and T lymphocytes fail to mature properly and are unable to recognize and respond to antigens. The phrase "recombination activation genes" (RAG) refers to the RAG-1 and RAG-2 genes that are involved with initiating the rearrangement of B and T cell antigen 20 receptors. The genetic recombination at the V, D and/or J gene segments is necessary to produce B and T-cell receptors. Mutations in the RAG-I an RAG-2 genes prevent early steps in this process, and result in a blockade of B cell development in the bone marrow and thymocyte development in the thymus (Mombaerts, et al., Cell 68:869-77 (1992); Shinkai, et al., Cell 68:855-867 (1992)). 25 The phrase "donor-specific cells with hematopoietic stem cell properties" refer to cells from a donor species that exhibit hematopoietic stem cell properties. The most obvious candidates are hematopoietic stem cells. However, other cells are envisioned, including but not limited to, cells that differentiate into HSC, such as embryonic stem cells. The phrase "donor immune system" refers to complete or partial immune 30 function that is not naturally found in a host mammal. For example, in a recipient mammal of this invention, cytokines necessary for the maintenance of a functional immune system as well as donor-specific immune cells are introduced into an immunodeficient mammal, either through introduction of transgenes that encode the cytokines or, less preferably, through the WO 01/15521 PCT/US00/23971 -8 addition of cytokines to the animal. Donor cells are the source of the recipient mammal's immune system (and typically, but not necessarily, the cytokine). It is not necessary that the donor immune system be fully functional, i.e., exhibit all functions of a mammalian immune system found in nature. However, it is preferred that the s donor immune system at least comprise donor T and B lymphocytes, and antigen presenting cells such as macrophages and dendritic cells. The phrase "embryonic stem cells" refers to cells that will grow continuously in culture and retain the ability to differentiate to all cell lineages, including but not limited to, hematopoietic cells. The term "differentiate" or "differentiated" refers to the process of 10 becoming a more specialized cell type. For example, hematopoietic stem cells differentiate into cells of the "lymphoid", "erythroid" and "myeloid" lineages. Lymphoid cells are cells that mediate the specificity of immune responses. They are divided into two main groups, T and B lymphocytes, and include a small population of large granular lymphocytes, or natural killer cells. Erythroid cells are erythroblasts and erythrocytes. Cells of the myeloid lineage include 15 platelets, neutrophils, basophilic, eosinophils and monocytes. The phrase "facilitating production of donor-specific functional immunity" refers to the ability of the recipient mammal to develop and maintain a functional donor derived immune system. Typically, the immune system comprises hematopoietic cells that are specific to a donor as well as cytokines and other ancillary compounds that are necessary, or 20 even desired, to allow the hematopoietic cells to be functional, e.g., bind to antigen, recognize an antigen as foreign or self, communicate with other cells of the immune system so that other cells, e.g., monocytes and macrophages, are activated, or cytokines, as defined herein, are released. The term "human cells" for purposes of this invention are cells that are derived 25 from a human. The derivation can be direct, i.e., primary cultures obtained from a human, or indirect, e.g., culture lines derived from primary cultures obtained from a human. The term "introduction" or "introducing" for purposes of this invention refers to the addition of exogenous compounds, particularly cytokine genes, to the recipient mammals of this invention. The compounds can be introduced into the recipient mammals of this 30 invention in a variety of methods, including but not limited to, introduction of the genes that encode the compounds. Introduction of the genes that encode the compounds can be through gene transfer into a non-fetal mammal or transgenically into a gamete or an embryonic mammal. In addition to direct introduction of the genes that encode the compounds, the WO 01/15521 PCTIUSOO/23971 -9 genetic material can be introduced into a recipient mammal through breeding or cloning, e.g., the introduction of the genes that encode the compounds into the germline of an offspring from a transgenic parent. The phrase "maintaining an immune system" refers to the ability of exogenous 5 cytokines to support a donor-derived immune system in a recipient mammal that otherwise would not support such an immune system. Typically, however not necessarily, the exogenous cytokines are naturally found in the same species as the donor. Thus, required interactions between the cells of the donor-derived immune system and cytokines naturally found in the donor to maintain the immune system are supplied in the recipient mammal. 10 The phrase "maintenance and maturation of donor-derived hematopoietic cells" refers to providing cytokines necessary to allow hematopoietic stem cells and other immature cell types to mature into functional cells, e.g., of the immune system, and providing necessary cytokines so that the cells, once mature, survive to function. In addition to the cells of the immune system, the maintenance and maturation of other types of hematopoietic cells, e.g., is erythrocytes, platelets, other lymphoid tissue (for example, the gut-associated immune system which consists of Peyer's patches, villi containing intraepithelial lymphocytes, and lymphocytes scattered throughout the lamina propria, and the connective tissue beneath the surface epithelium. The terms "making a mammal," or "producing a mammal" refers to the 20 manipulation of the physical characteristics of a mammal so the mammal, after the manipulation, comprises a donor immune system. For purposes of this invention, a human mammal cannot be made or produced. The manipulation of the physical characteristics can be through transgenic technology, gene transfer, or other manipulations of the genotype of the mammal, or manipulation can be through the introduction of cells, proteins or other 25 compounds that affect the physical characteristics of the mammal. A "mammal" is a warm blooded vertebrate of the class Mammalia. For purposes of this invention, a "recipient mammal" is a mammal that, because of the presence of exogenous transgenes and donor cells with hematopoietic stem cell properties, is capable of facilitating production of a donor-specific immune system. For purposes of this invention, a 30 recipient mammal is not a human. A "mouse" refers to a mammal of the species Mus musculus. Mice encompassed by this invention include all mice, but in particular, mice of the, or derived from the RAG-I and RAG-2 mutant strains. Because of the severe immunodeficiency of the mice of WO 01/15521 PCTIUSOO/23971 -10 this invention, it is unlikely they are found in nature. Typically, the mice of this invention are laboratory strains of mice, such as the RAG-1, RAG-2 or SCID mice. The term "transgenically" refers to the introduction of exogenous genetic material, or transgenes, into the genome of a gamete or an embryo. Preferably, the genetic s material encodes proteins and the introduction is into a fetal mammal. "Transgenes" refer to nucleic acids that comprise a protein coding region and regulatory elements, e.g., promoters and termination sequences, if desired. A transgene can comprise a single copy of a coding sequence or multiple copies of a coding sequence. If multiple copies are present, the coding regions can be partial coding regions. The coding regions can be separated by regulatory 1o elements or may be arranged in a 5' to 3' or in a 5' to 5' or 3' to 3' orientation. Multiple transgenes, representing independent coding and regulatory sequences, may be present, most preferably, in a contiguous fashion in the same region, or dispersed elsewhere in the recipient mammal's genome. Cells that are "xenogeneic" to a host mammal are from a different species as the is host mammal. The host mammal normally recognizes xenogeneic cells as non-self. II. PRODUCTION OF THE MAMMALS OF THIS INVENTION. In a preferred embodiment, the recipient mammals of this invention are immunodeficient. To produce immunodeficient mammals, the naturally occurring immune 20 systems of the mammals should be inactivated. Inactivation can take place by removing or disrupting multiple immune system-related activities or by removing or disrupting just one activity. Although immune function can be disrupted by many different mechanisms, e.g., spontaneous mutation, irradiation and antisense technology, in a preferred embodiment, immune function is disrupted by knocking out by e.g., homologous recombination or 25 spontaneous mutation, one or more gene functions necessary for maturation and maintenance of the immune system. Generation of knock out mammals Homologous recombination may be employed for gene replacement, inactivation or alteration of genes. A number of papers describe the use of homologous 30 recombination in mammalian cells. See, for example, Thomas & Capecchi, Cell 51:503 (1987); Nandi, et al., Proc. Nat'l Acad. Sci. USA 85:3845 (1988); and Mansour, et al., Nature 336:348 (1988); Schweizer, et al., J. Biol. Chem. 274:20450 (1999); Hauser, et al., Proc. Nat'l WO 01/15521 PCT/USOO/23971 -11 Acad. Sci. USA 96:8120 (1999); Haber, Trends Biochem. Sci. 24:271 (1999); and Bonaventure, et al., Mol. Pharmacol. 56:54 (1999). Furthermore, various aspects of using homologous recombination to create specific genetic mutations in embryonic stem cells and to transfer these mutations to the 5 germline have been described (Thomas & Capecchi, Cell 51:503 (1987); Thompson, et al., Cell 56:316 (1989); Antoine, et al., J. Cell Sci. 112:2559 (1999); Molotkov, et al., Cancer Lett. 132:187 (1998); Bleich, et al., Pflugers Arch. 438:245 (1999); Struble, et al., Neurosci. Lett. 267:137 (1999); Schweizer, et al., J. Biol. Chem. 274:20450 (1999); Cuzzocrea, et al., Eur. Cytokine Netw. 10:191 (1999); and Mombaerts, et al. Cell 68:869-77 (1992); and Shinkai, et 10 al. Cell 68:855-867 (1992)). Thus, the recipient mammals of this invention, which lack necessary endogenous gene(s) necessary for the maturation of lymphocytes, can be made using homologous recombination to effect targeted gene replacement. In this technique, a specific DNA sequence of interest is replaced by an altered DNA. In a preferred embodiment, the 15 genome of an embryonic stem (ES) cell from a desired mammalian species is modified (Capecchi, Science 244:1288 (1989) U.S. Patent No. 5,487,992). As mentioned above, the gene to be replaced by homologous recombination is one that is activated early in lymphocyte development. Without being bound by any particular theory, it is believed the desired gene is activated while the thymocyte is in the CD4~ and CD8~ 20 state (double negative) or the CD44 1 0w and CD25' state, and the B lymphocyte is in the B220du"II/CD43+ state. Because at these states, T and B cell receptor rearrangement occurs, it is believed the genes that encode proteins that modulate the VDJ recombination are likely targets for replacement. Examples of these genes are the RAG-1 and RAG-2 genes, the T cell receptor (TCR) and immunoglobulin (Ig) genes, the CD3 genes, the pre-T cell receptor, and 25 the SCID gene. Additional types of genes that regulate the survival and differentiation of lymphocyte precursors are also potential targets, e.g., the ikarus transcription factor, the common gamma chain subunit, IL-7, and the IL-7 receptor, among others. The procedures employed for inactivating one or both copies of a gene coding for a particular protein that modulates early thymocyte development will be similar, differing 30 primarily in the choice of sequence, selectable marker used, and the method used to identify the absence of the modulating protein, although similar methods may be used to ensure the absence of expression of a particular protein. Since the procedures are analogous, the WO 01/15521 PCT/USOO/23971 -12 inactivation of the RAG-2 gene in mice will be used as an example. See, U.S. Patent 5,859,307, the entirety of which is incorporated by reference. The homologous sequence for targeting the construct may have one or more deletions, insertions, substitutions or combinations thereof. For example, the RAG-2 gene may 5 include a deletion at one site and/or an insertion at another site. The presence of an inserted positive marker gene will result in a defective inactive protein product insertion as well as a gene that can be used for selection. Preferably, deletions are employed. For an inserted gene, of particular interest is a gene which provides a marker, e.g., antibiotic resistance such as neomycin resistance, including G418 resistance. 10 The deletion should be at least about 50 base pairs, or more usually at least about 100 base pairs, and generally not more than about 20,000 base pairs, where the deletion will normally include at least a portion of the coding region including a portion of or one or more exons, a portion of one or more introns, and may or may not include a portion of the flanking non-coding regions, particularly the 5'-non-coding region (transcriptional regulatory is region). Thus, the homologous region may extend beyond the coding region into the 5'-non coding region or alternatively into the 3'-non-coding region. Insertions should generally not exceed 10,000 base pairs, usually not exceed 5,000 base pairs, generally being at least 50 base pairs, more usually at least 200 base pairs. The homologous sequence should include at least about 100 base pairs, 20 preferably at least about 150 base pairs, and more preferably at least about 300 base pairs of the target sequence and generally not exceeding 20,000 base pairs, usually not exceeding 10,000 base pairs, and preferably less than about a total of 5,000 base pairs, usually having at least about 50 base pairs on opposite sides of the insertion and/or the deletion in order to provide for double crossover recombination. 25 Upstream and/or downstream from the desired DNA may be a gene which provides for identification of whether a double crossover has occurred. For this purpose, the herpes simplex virus thymidine kinase gene may be employed, since the presence of the thymidine kinase gene may be detected by the use of nucleoside analogs, such as Acyclovir or Gancyclovir, for their cytotoxic effects on cells that contain a functional HSV-tk gene. The 30 absence of sensitivity to these nucleoside analogs indicates the absence of the thymidine kinase gene and, therefore, where homologous recombination has occurred, that a double crossover event has also occurred.

WO 01/15521 PCT/USOO/23971 -13 The presence of the marker gene inserted into the RAG-2 gene of interest establishes the integration of the targeting construct into the host genome. However, DNA analysis might be required in order to establish whether homologous or non-homologous recombination occurred. This can be determined by employing probes for the target DNA s sequence that hybridize to the 5' and 3' regions flanking the insert. The presence of an insert, deletion, or substitution in the targeted gene, can be determined using restriction endonucleases that distinguish the size of a targeted allele from a wild type allele. The polymerase chain reaction may also be used in detecting the presence of homologous recombination (Kim & Smithies, Nucleic Acid Res. 16:8887-8903 (1988); and 10 Joyner, et al., Nature 338:153-156 (1989)). Primers may be used which are complementary to a sequence within the construct and complementary to a sequence outside the construct and at the target locus. In this way, one can only obtain DNA duplexes having both of the primers present in the complementary chains if homologous recombination has occurred. By demonstrating the presence of the primer sequences or the expected size sequence, the 15 occurrence of homologous recombination is supported. The construct may further include an origin of replication which is functional in the mammalian host cell. For the most part, these replication systems will involve viral replication systems, such as Simian Virus 40, Epstein-Barr virus, papilloma virus, adenovirus and the like. 20 Where a marker gene is involved, as an insert, and/or flanking gene, depending upon the nature of the gene, it may have the wild-type transcriptional regulatory regions, particularly the transcriptional initiation regulatory region or a different transcriptional initiation region. Whenever a gene is from a host where the transcriptional initiation region is not recognized by the transcriptional machinery of the mammalian host cell, a different 25 transcriptional initiation region will be required. This region may be constitutive or inducible, preferably inducible. A wide variety of transcriptional initiation regions have been isolated and used with different genes. Of particular interest as promoters are the promoters of metallothionein-I and II from a mammalian host, thymidine kinase, beta-actin, immunoglobulin promoter, human cytomegalovirus promoters, phosphoglycerate kinase 30 (PGK) and SV40 promoters. In addition to the promoter, the wild-type enhancer may be present or an enhancer from a different gene may be joined to the promoter region. The construct may further include a replication system for prokaryotes, particularly E. coli, for use in preparing the construct, cloning after each manipulation, WO 01/15521 PCT/USOO/23971 -14 allowing for analysis, such as restriction mapping or sequencing, followed by expansion of a clone and isolation of the construct for further manipulation. When necessary, a different marker may be employed for detecting bacterial transformants. Once the construct has been prepared and manipulated and the undesired 5 sequences removed from the vector, e.g., the undesired bacterial sequences, the DNA construct is now ready to be introduced into the target stem cells. Methods of introducing the desired DNA into stem cells are well known in the art. Briefly, preferred methods include, but are not limited to calcium phosphate/DNA coprecipitates, microinjection of DNA into the nucleus, electroporation, bacterial protoplast fusion with intact cells, lipofection, or the like. The DNA 10 may be single or double stranded, linear or circular, relaxed or supercoiled DNA. For various techniques for transforming mammalian cells, see Keown, et al., Methods in Enzymology 185:527-537 (1990); Sambrook, et al., MOLECULAR CLONING: A LABORATORY MANUAL (2ND ED.), VOLS. 1-3, Cold Spring Harbor Laboratory, (1989) ("Sambrook") or CURRENT PROTOCOLS IN MOLECULAR BIOLOGY, F. Ausubel et al., ed. Greene Publishing and Wiley 15 Interscience, New York (1987) ("Ausubel"). After transformation of the target cells, many target cells are selected by means of positive and/or negative markers, as previously indicated, neomycin resistance and Acyclovir or Gancyclovir resistance. Those cells which show the desired phenotype may then be further analyzed by restriction analysis, electrophoresis, Southern analysis, polymerase 20 chain reaction or the like. By identifying fragments which show the presence of the mutations at the target gene site, one can identify cells in which homologous recombination has occurred to inactivate the target gene. Cells in which only one copy of the, e.g., RAG-2, gene have been inactivated still retain a single unmutated copy of the target gene. If desired, these cells can be expanded 25 and subjected to a second transformation with a vector containing the desired DNA. If desired, the mutation within the desired DNA may be the same or different from the first mutation. If a deletion, or replacement mutation is involved, a second mutation may overlap at least a portion of the mutation originally introduced. If desired, a different positive selection marker can be used in this transformation. If a different marker is used, cells with both mutations can be 30 selected in double selection media. Alternatively, to determine if the cells comprise mutations in both copies of the transformed cells, the cells can be screened for the complete absence of the functional protein of interest. The DNA of the cell may then be further screened to ensure the absence of a wild-type target gene.

WO 01/15521 PCTUSOO/23971 -15 In an alternative embodiment, chimeric mammals can be developed from transformed stem cells (see, infra) and animals with one mutated sequence can be bred to other mammals with one or two mutated sequences and offspring that contain mutations in both copies (homozygotes) selected as recipient mammals of this invention. Similarly, recipient 5 mammals developed from chimeric mammals from transformed cells with two mutated genes can be bred to produce more recipient mammals. After transformation, the stem cells containing either one or two copies of the replacement DNA are inserted into recipient mammal embryos to produce chimeric mammals. Typically, this is done by injecting stem cell clones into mammalian blastocysts. The 10 blastocysts are then implanted into pseudopregnant females. The offspring derived from the implanted blastocysts are test-mated to animals of the parental line to determine whether the offspring comprise a chimeric germ line. Chimeras with germ cells derived from the altered stem cells transmit the modified genome to the offspring of the test matings, yielding mammals heterozygous for the target DNA (contain one target DNA and one replacement DNA). The is heterozygotes are then bred with each other to create homozygotes for replacement DNA. Because the recipient mammals of this invention are immunodeficient, it may be necessary to maintain them in a germ free environment. Such environments are well known to those of skill in the art and techniques for maintaining immunodeficient mice can be found in Immunodeficient rodents : a guide to their immunobiology, husbandry, and use, Committee 20 on Immunologically Compromised Rodents, Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council. Washington, D.C. : National Academy Press, 1989. In addition to producing knock-out mammals, the immunodeficient mammals of this invention are commercially available. For example, mice with a RAG-2 mutation are 25 available from Taconic, RAG-I and TCRbeta/delta mutant mice from Jackson Laboratory, or SCID mice from Jackson and Taconic. In another embodiment, introduction of transcriptionally active transgenes, e.g., a truncated forms of rearranged antigen receptors or human CD3 epsilon, are examples of achieving lymphocyte deficiencies. 30 It is desireable to screen the recipient mammals for the presence of the knocked out gene. Screening can be done phenotypically or genotypically. Phenotypic screening includes, but is not limited to, the absence of mature T and B cells and other phenotypic changes that correlate with the absence of mature T and B cells, such as the absence of serum WO 01/15521 PCT/US00/23971 -16 immunoglobulins. However, if the mutated gene presents as a dominant phenotype, animals that are heterozygous at that gene will present with the same phenotypic characteristics as the desired homozygotes. Therefore, it is desirable to screen for homozygotes by genotypic screening. 5 DNA screening is well known to those of skill in the art and can be found in, for example, Ausubel and Sambrook. Briefly, cells containing DNA are removed from the test animals. In mice, this can be done by removing the tip of the tail and isolating cells. The genomic DNA is isolated from the cells and cut into manageable size by restriction endonucleases. The cut genomic DNA is electrophoresed in an agarose gel and then probed 10 with a labeled nucleic acid that can distinguish the wild type from the modified DNA fragment. Binding of the labeled probe to the genomic DNA depends on the ability of the probe to remain hybridized to the genomic DNA under the wash conditions used. An extensive guide to the hybridization of nucleic acids is found in Tijssen, LABORATORY TECHNIQUES IN BIOCHEMISTRY AND MOLECULAR BIOLOGY--HYBRIDIZATION WITH NUCLEIC 15 ACID PROBES, Elsevier, New York (1993). Generally, highly stringent hybridization and wash conditions are selected to be about 5*C lower than the thermal melting point (Tm) for the specific sequence at a defined ionic strength and pH. The Tm is the temperature (under defined ionic strength and pH) at which 50% of the target sequence hybridizes to a perfectly matched probe. Very stringent conditions are selected to be equal to the Tm for a particular probe. An 20 example of stringent hybridization conditions for hybridization of complementary nucleic acids which have more than 100 complementary residues on a filter in a Southern or northern blot is 50% formalin with 1 mg of heparin at between 40 and 50 0 C, preferably 42*C, with the hybridization being carried out overnight. An example of highly stringent wash conditions is 0. 15M NaCl at from 70 to 80 0 C with 72'C being preferable for about 15 minutes. An example 25 of stringent wash conditions is 0.2x SSC wash at about 60 to 70 0 C, preferably 65'C for 15 minutes (see, Sambrook, supra for a description of SSC buffer). Often, a high stringency wash is preceded by a low stringency wash to remove background probe signal. An example medium stringency wash for a duplex of, e.g., more than 100 nucleotides, is lx SSC at 40 to 50*C, preferably 45*C for 15 minutes. An example low stringency wash for a duplex of, e.g., 30 more than 100 nucleotides, is 4-6x SSC at 35 to 45 0 C, with 40'C being preferable, for 15 minutes. In general, a signal to noise ratio of 2x (or higher) than that observed for an unrelated probe in the particular hybridization assay indicates detection of a specific hybridization. After WO 01/15521 PCT/USOO/23971 -17 removal of unbound probe, the label is detected and the presence or absence of the desired DNA in the genome of the mammals determined. As the mammal matures, it may exhibit a "leaky phenotype." For purposes of this invention, a leaky phenotype is one where a few thymocytes and/or pro-B cells undergo 5 functional receptor rearrangement and mature into T and B cells, respectively. Thus, a SCID mouse exhibits a leaky phenotype. This phenotype can be detected by monitoring the development of host T and B cells and/or serum immunoglobulin in the recipient mammals throughout the life of the animal. Transgenic Mammals 10 The differentiation of hematopoietic cells is a highly regulated process that involves the coordinate expression of many factors, including cytokines, adhesion molecules, and chemokines, among others. Due to evolutionary changes, considerable divergence has occurred between a number of murine and human growth factors such that the murine factors do not always interact as efficiently, or in the same manner, as their human counterparts. A is major consideration when supplying exogenous cytokines is the dosage, combination, and pattern of delivery. Since cytokines are powerful signaling molecules that work in close proximity to their origin, in low concentrations, and synergistically with one another, systemic delivery of exogenous cytokines is unlikely to provide the physiological levels necessary for normal development. 20 The preferred method of providing human-specific factors to the host is via transgenesis, whereby copies of genomic DNA encoding the desired factors are incorporated into the genome of the host. The DNA should include tissue-specific regulatory sequences and any introns and exons required for normal RNA processing, including alternatively spiced variants. The latter may be particularly important when the proteins present themselves as 25 both membrane bound and soluble forms having different physiological effects. Thus, to maintain a donor species-specific functional immune system, it is necessary to introduce donor-specific cytokines into the germline of the recipient mammals of this invention. The recipient mammals of this invention are produced by introducing transgenes into the germline of a non-human animal. Embryonal target cells at various 30 developmental stages can be used to introduce transgenes. Different methods are used depending on the stage of development of the embryonal target cell. For example, the zygote is the best target for micro-injection. In the mouse, the male pronucleus of the zygote reaches approximately 20 micrometers in diameter. At this size, reproducible injections of 1-2 pL of WO 01/15521 PCT/USOO/23971 -18 DNA solution can be performed. The use of zygotes as a target for gene transfer has another major advantage in that, in most cases, the injected DNA will be incorporated into the host genome before the first cleavage (Brinster, et al. Proc. Natl. Acad. Sci. USA 82:4438-4442 (1985)). As a consequence, all cells of the recipient mammal will carry the incorporated 5 transgene. This is also reflected in the efficient transmission of the transgene to offspring of the parent transgenic mammal since 50% of the germ cells of the offspring will harbor the transgene. In another, alternative embodiment, intracytoplasmic sperm injection (ICSI) can be used to introduce transgenes into metaphase oocytes. See, Perry, et al., Science 284:1180 io (1999). Briefly, sperm heads and linearized DNA are incubated for a short period of time and co-injected into an oocyte. Improved rates of transgenesis are seen when the sperm heads have undergone membrane disruption prior to incubation with the DNA. Retroviral infection can also be used to introduce a transgene into a recipient mammal. The developing embryo can be cultured in vitro to the blastocyst stage. The 15 blastomeres are then targets for retroviral infection (Jaenisch, Proc. Nat'l Acad. Sci USA 73:1260-1264 (1976)). Efficient infection of the blastomeres is obtained by enzymatic treatment to remove the zona pellucida (Hogan, et al., MANIPULATING THE MOUSE EMBRYO, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. (1986)). The viral vector system used to introduce the transgene is typically a replication-defective retrovirus carrying 20 the transgene (Jahner, et al. Proc. Natl. Acad. Sci. USA 82: 6927-6931 (1985); Van der Putten, et al. Proc. Natl. Acad. Sci USA 82: 6148-6152 (1985)). Infection is easily and efficiently obtained by culturing the blastomeres on a monolayer of virus-producing cells (Van der Putten, supra; Stewart, et al. EMBO J. 6: 383-388 (1987)). Alternatively, infection can be performed at a later stage. Virus or virus-producing cells can be injected into the blastocoel (Jahner, D., 25 et al. Nature 298:623-628 (1982)). Most of the founders will be mosaic for the transgene since incorporation occurs only in a subset of the cells which form the recipient mammal. Further, the founder may contain various retroviral insertions of the transgene at different positions in the genome which generally will segregate in the offspring. In addition, it is also possible to introduce transgenes into the germ line, albeit with low efficiency, by intrauterine retroviral 30 infection of the midgestation embryo (Jahner, D. et al. supra). A fourth type of target cell for transgene introduction is the embryonal stem cell (ES). ES cells are obtained from pre-implantation embryos cultured in vitro and fused with embryos (Evans, et al. Nature 292:154-156 (1981); Bradley, et al. Nature 309:255-258 (1984); WO 01/15521 PCTUSOO/23971 -19 Gossler, et al. Proc. Natl. Acad. Sci USA 83:9065-9069 (1986); and Robertson, et al. Nature 322:445-448 (1986)). Transgenes can be efficiently introduced into the ES cells by DNA transfection or by retrovirus-mediated transduction. Such transformed ES cells can thereafter be combined with blastocysts from a nonhuman animal. The ES cells thereafter colonize the 5 embryo and contribute to the germ line of the resulting chimeric recipient mammal. For review see Jaenisch, Science 240:1468-1474 (1988); Bradley, et al. Biotechnology (N Y) 10(5):534-9 (1992); and Williams, Bone Marrow Transplant 5(3):141-4 (1990). The actual transgenes of this invention include the coding sequences for proteins necessary for the maturation and maintenance of a donor-specific functional immune 10 system. Those of skill will recognize the required cytokines will vary depending on the desired functionality. Such cytokines include but are not limited to, IL-6, IL-7, GM-CSF and SCF, LIF, M-CSF, and OM. In addition or alternatively, MHC genes from the same species and haplotype as that of the donor HSC may be introduced into the recipient mammal and expressed in tissues that endogenously express MHC molecules. In this case, donor 15 thymocytes become "restricted" during development in the recipient's tissues, particularly the thymus, via interaction with the transgenic MHC molecules. This leads to the maturation of T cells that can have cognate interactions with donor B lymphocytes displaying the same haplotype of transgenic MHC molecules as the host mammal. For example, human HLA -DR, -DQ, and/or -DP genes of the same haplotype as the HSC donor are expressed in the mouse 20 tissues. The expression of transgenic MHC expression is most beneficial in mouse strains other than those with mutations in the RAG-2 or RAG-1 genes. In a preferred embodiment, the cytokine genes are derived from a human being or a human cell line. However, other mammalian sources may be used such as pig, sheep or rat. In an alternative embodiment, mammals of the same species but allogeneic to the donor 25 are the source of the cytokines. There are numerous other methods of isolating the DNA sequences encoding the cytokines of this invention. For example, DNA may be isolated from a genomic or cDNA library using labeled oligonucleotide probes having sequences complementary to known cytokine sequences. For example, full-length cDNA probes may be used, or oligonucleotide 30 probes consisting of subsequences of the known sequences may be used. Such probes can be used directly in hybridization assays to isolate DNA encoding cytokines. Alternatively probes can be designed for use in amplification techniques such as PCR, and DNA encoding cytokines may be isolated by using methods such as PCR.

WO 01/15521 PCTIUSOO/23971 -20 To prepare a cDNA library, mRNA is isolated from a source tissue or cells. For example, 11-7 is expressed by bone marrow stroma, thus, these cells would be a suitable source of mRNA that encodes 11-7. cDNA is reverse transcribed from the mRNA according to procedures well known in the art and inserted into bacterial cloning vectors. The vectors are 5 transformed into a recombinant host for propagation, screening and cloning. Methods for making and screening cDNA libraries are well known. See, Gubler & Hoffman, Gene 25:263 269, (1983) and Sambrook, et al. From a genomic library, total DNA is extracted from the host tissue or cells and then cut into smaller pieces of DNA by mechanically shearing or by enzymatic digesting to 10 yield fragments of about e.g., 12-50kb. Fragments of a desired size are then separated by gradient centrifugation and are inserted into bacteriophage lambda vectors or other vectors. These vectors and phage are packaged in vitro, as described in Sambrook, et al. Recombinant phage can be analyzed for the presence of cytokine nucleic acids by plaque hybridization as described in Ausubel. 15 Libraries containing genomic DNA sequences greater than 50kb are prepared using various cloning vectors, e.g., YAC, BAC, P1, and PAC vectors. Techniques for generating these libraries are well known. See, Markie, ed. (for YACS) Methods in Molecular Biology 54 (1997), Ramsay (for YACS), Mol Biotechnol 1(2):181-201 (1994), Monaco et al., Trends Biotechnol. 12:280-6 (1994), and Shepherd et al., Genet Eng (NY) 16:213-28 (1994). 20 Hybridization probes useful in this invention include known sequences that encode for the cytokines of interest or sequences that encode homologous cytokines from another species, for example a probe derived from a murine sequence to probe a human cDNA library for a homologous sequence. One of skill will recognize that if homologous sequences are used as probes, the stringency of the wash conditions should be lowered. 25 In addition to generating coding sequences of the cytokines to be used as transgenes, many of the nucleic acids that encode the necessary cytokines of this invention are commercially available. Such resources include R&D Systems, Genetic Systems, and CEPH. The preferred method of making transgenic mammals that express the necessary cytokines is as follows. From both animal and in vitro studies, IL-3, IL-6, IL-7, M-CSF, GM 30 CSF, stem cell factor, LIF and oncostatin M appear to either play a role in hematopoiesis, are expressed in the bone marrow or thymus, or the murine proteins show specificity for murine vs. human cells, thus suggesting that human HSC engrafted in a recipient mammal may not recognize the native mammalian cytokine. Genomic clones for these human genes can be WO 01/15521 PCT/USOO/23971 -21 obtained and transfected into ES cells. The ES cells can be introduced into blastocysts to transfer the donor transgenes into the germline of a recipient mammal. Isolation of genomic clones. In a preferred method, PCR primer sets are designed against either the 5' or the 5 3' end of genomic sequences so that constructs containing the genes can be identified by PCR. In addition, these primer sets can be designed to distinguish between mouse and human genes so that the native mammalian genes are not mistakenly identified during a genomic or transcriptional screen of the transfected ES cells or suspected recipient mammals. Human genomic libraries are screened using gene specific primer sets (See 10 Example 2 and Ausubel for a general description of genomic library screening). If desired, positive clones can be further confirmed by either other primer sets for the same gene or by Southern blot analysis. Depending on the size of the gene, different types of cloning vectors and libraries are readily available. Genes up to approximately 15kb may be obtained from lambda libraries. Those up to 50kb may be identified from cosmid libraries. Larger genes 15 over 50kb can be isolated from BAC, PAC, P1, YAC, MAC, or other such libraries. Demonstration of in vitro transcription from the genomic constructs. Because the transgenes will be expressed in mammalian hosts, it may be desirable to determine the ability of the human sequences to be transcribed by mammalian cells other than human, preferably murine before transfection into possibly rare ES cells. 20 Undigested or digested genomic constructs can be transfected by lipofection into a murine cell line that expresses the endogenous form of each cytokine. Commercial lipofection reagents are widely available and may require optimization for a particular cell type to obtain adequate transfection efficiencies in a transient assay. The mRNA from the transfected cells is then analyzed for transcription of the contruct. Depending on the preference of one of skill, the 25 mRNA can be electrophoresed according to standard procedures and then probed in a northern blot, or first strand cDNA can be synthesized by standard reverse transcription methods. The resulting cDNA can then be analyzed by labeled probe and Southern blot or by PCR methods. Selection of ES clones that contain human cytokine genes. Two equally preferred embodiments can be used to combine all of the desired 30 genes into one strain. In the first method, groups of transgenes are co-transfected into ES cells along with a selectable marker for neomycin resistance. For example, one group of genes can contain IL-7, SCF, and LIF constructs, and the second contain GM-CSF, M-CSF, and IL6 constructs. In the second method, all desired genes are co-transfected together. If all of the WO 01/15521 PCT/USOO/23971 -22 necessary transgenes are not present in the germline of one transgenic mammal, that mammal can be mated to a mammal comprising, in its germline, the necessary transgene to create offspring with all necessary transgenes. To introduce the transgenes into the ES cells, the DNA constructs are digested 5 with a desired restriction endonuclease that linearizes the DNA, if circular. Within each group, it is preferred that DNA constructs are mixed in equal molar ratio. However, one of skill will recognize that if more copies of one gene is desired, DNA constructs of that gene should be over-represented. For positive selection, a marker-containing plasmid can be mixed with these DNAs at molar ratio of about 4:1. 10 The DNA can then be introduced ES cells by lipofection or another suitable technique. The preferred transfection protocol is similar to that provided by the manufacturer of the lipofection reagent and is described in detail in Example 2. After a suitable time in selection media (preferably 5-20 days and more preferably 10-14 days), individual transfected ES cell colonies are transferred into 96-well dishes for cloning and expansion. 15 Although one method is described here and in Example 2, those of skill in the art will realize that other methods of inserting transgenes in the germ line of mammals are known and also available. Some of these methods can are found in US Patents 4,873,191, 5,434,340, 4,464,764, 5,487,992, 5,814,318; PCT published patent applications WO 97/20043, WO 99/07829, WO 99/08511; and Perry, et al., Science 284:1180 (1999). 20 Depending on the transgene that has been inserted into the ES cell, different techniques can be used to detect the transgene. For example, PCR can be used to detect genomic DNA or cDNA made from RNA transcripts, ELISA and other antibody-based assays can be used to determine whether the gene product of the transgenes are synthesized in ES cells or are present extracellularly, and if such an assay is available, a functional assay can be 25 used to detect the gene product. Reconstitution of an Immune System A common method for studying the biology of human hematopoietic stem cells is to transfer them into immunocompromised mice. Inbred strains of mice carrying spontaneous mutations, such as the SCID, NOD/SCID and beige/nude/xid mice, have been the 30 most widely used hosts. The advent of gene targeting has led to a large variety of immunodeficient animals that are presently under evaluation for this purpose. The animals are usually conditioned by sublethal irradiation, and are often supplemented with injections of human growth factors to augment the level of bone marrow engraftment and expansion of WO 01/15521 PCT/USOO/23971 -23 progenitor and/or mature cells. In studies with mice with the SCID mutation, it appears that a dominant genotype present in the B6/129 mouse, if the mutation is placed in this mouse, serves to nullify some of the phenotypes observed with the mutation. Fewer effects are seen on a Balb/c background. Even fewer effects are seen on a NOD background. Thus, the preferred s murine background for the RAG mutation may be NOD/SCID. It is not well understood why particular genetic backgrounds perform better than others, or if supplemental growth factors are beneficial. For example, transgenic SCID mice that express human genes for stem cell factor, GM-CSF, and IL-3, engraft more efficiently than non-transgenic SCID mice, but not any better than NOD/SCID mice. The human genes were driven by a viral promoter rather 10 than the endogenous regulatory sequences, and most likely had abnormal patterns and levels of expression (Bock, et al. J. Exp. Med. 182:2037-2043). In addition to the SCID, RAG and NOD mutations, other mutations may be beneficial to the recipient mammals of this invention. For example, experiments with the SCID mutation have shown that a 7c knock out mutation or a p-2 microglobulin knockout 15 mutation may improve the engraftment of xenogeniec HSC. Transplanted Hematopoietic Stem Cells Sources of hematopoietic stem cells include, but are not limited to, umbilical cord blood (CB), bone marrow (BM) and mobilized peripheral blood (MPB). Human CB can be obtained, for example, from Advanced Bioscience Resources 20 Inc (ABR), Alameda, CA, or Purecell, San Mateo, CA. CB is collected by ABR from local hospitals within 24 hours of shipping and is processed on site. Alternatively, human BM, CB, and/or MPB cells are obtained from Purecell as either fresh or frozen cells, and fractionated or unfractionated cells. Before use, all samples are tested for Hepatitis B and C, and HIV. Any experimental materials involving samples found to be virus positive are discarded immediately 25 and animals removed, marked and disposed of in accordance with procedures for disposing contaminated animal carcasses. Throughout the course of the experiment, all samples should be treated under the assumption that they may be contaminated with human blood borne pathogens (Biosafety Level 2, BL-2). All personnel handling mice with human blood cells should receive the hepatitis B vaccine. 30 The preferred age for transplant is either shortly after birth (approximately 72 hours) or as adult animals (4-8 weeks of age) that have been conditioned with gamma irradiation, however, other types of conditioning are known and are envisioned. The haplotype of the donor hematopoietic stem cells may be selected on the basis of the test population of WO 01/15521 PCT/US00/23971 -24 interest in order to determine the responsiveness of particular individuals to specific antigens, e.g., for the purpose of determining the allergenicity of proteins or peptides within the population. To transplant, lymphocytes from a source of hematopoietic stem cells are 5 isolated, preferably by density separation and counted. As a guide, adult mice can receive from 0.1-10 x 10 7 total cells. Alternatively, animals may also be transplanted with CD34' cells purified from, for example, cord blood using, e.g., the Dynal Detach-a-Bead research grade separation system or by FACS sorting. Purified CD34* cells from cord blood are then injected into neonatal or adult animals. In mice, approximately10 5 cells are injected per pup, or about 5 10 x 105 cells per adult. HSC-populations can be injected intravenously (iv). If desired, adult mammals may be treated on day minus-I with rabbit polyclonal anti-asialo GM antibodies to reduce the activity of NK cells. Alternatively, recipient mammals, in particular mice, may be crossed onto a background deficient in NK cell function, such as perforin-deficient mice. Neonatal is animals have little to no NK activity and, therefore, it is not necessary to treat them. Typically, adult mice are the recipient mammals. They are injected into the tail vein by standard procedures. They may also be injected via the retro-orbital sinus, in which case animals are anesthetized first. Typically, a mixture of ketamine and xylazine delivered intraperitoneally or intramuscularly is used. Animals should be kept warm and eyes moistened 20 during the procedure and monitored until they recover. To inject cells into the retro-orbital sinus of neonatal mice, pups are held below a light source to better visualize the eye vein. The needle is held parallel to the vein and inserted by aiming just below the surface of the skin. Cells are slowly injected while watching to determine that blood is cleared downstream of the needle point, and that swelling around the vein does not occur. After injection the needle is 25 held in place for a short period to allow the cells to enter the circulation. If necessary, pups may be chilled briefly on ice to reduce their activity. In an alternative embodiment, the recipient mammals may receive gamma irradiation conditioning according to established experimental protocols (Bix, et al. Nature 349:329-331 (1991); Markowitz, et al. Proc. Natl. Acad. Sci. 90: 2779-2783 (1993)). Using 30 mice as an example, a Mark I Model 30 sealed Cs-irradiator or similar instrument is used to administer doses ranging from 100-1000 cGy/animal according to manufacturer's protocol. A lethal level of radiation may be administered as a split dose (e.g., 2 x 50OR for Balb/c mice) for WO 01/15521 PCT/USOO/23971 -25 improved reduction of natural killer cells. Radiation is performed from 2-24 hours prior to injecting human hematopoietic cells. Methods of determining whether engraftment has taken place are well known in the art and include Pflumio, et al. Blood 88:3731-3740 (1996); Dick, et al. Stem Cells 15 Suppl 5 1:199-203 (1997); Ramirez, et al. Exp. Hematol. 26:332-344 (1998); Hogan, et al. Biol. Blood Marrow Transplant 3:236-246 (1997); Hogan, et al. Blood 90:85-96 (1997); and Lapidot, et al. J. Mol. Med. 75: 664-673 (1997). One of skill in the art will realize there are many ways to determine if the transplanted HSC have engrafted. In a preferred method, lymphocytes from the recipient 10 mammals are removed and investigated for the presence of human cell surface antigens. CD45 is preferred. This detection it typically done by staining cells with a dye conjugated to an antibody that binds to human CD45, which is found on both B and T lymphocytes, and not murine CD45. See, Colas, et al., Transplantation 67:984 (1999). Other methods include PCR amplification of human DNA sequences from lymphocytes. 15 III. USES OF MAMMALS OF THIS INVENTION The mammals of this invention have a variety of uses. For example, the recipient animals of this invention cab be used as a model for determining the allergenicity of non-donor, e.g., non-human, macromolecules. These macromolecules include proteins, 20 carbohydrates lipids, and other compounds. Proteins include but are not limited to commercially important enzymes such as bacterial proteases, fungal cellulases and the like. With the model, as described above, the animals can be exposed to the test compounds and the immune response of the animals monitored. Based on the response in the animal model, the compounds can be modified so that the immune response is lessened, for example to decrease 25 allergenicity, or increased, for example, if the compound is to be used as a vaccine. Alternatively, the model of this invention can be used to determine the effect compounds have on a human immune system. For obvious ethical reasons, experimental compounds cannot be used on human subjects. A mouse with a functional human immune system allows this more relevant preclinical testing to be done. 30 In another use of this invention, the recipient mammals generate fully human polyclonal or monoclonal antibodies to specific antigens. Unlike human antibodies currently available in the art, no genetic manipulation of antibodies of other species, e.g., humanization, would be needed, and would not pose the same biological hazard as purified antibodies from WO 01/15521 PCT/US00/23971 -26 humans with accidental exposure to antigens possess. It is also contemplated the mouse of the instant invention can be used to determine whether humanized or other monoclonal antibodies will raise a response in a human immune system. In addition to a humoral response, the mouse of this invention can be used to 5 investigate the human cell mediated response to pathogens and other immunomodulatory compounds. For example, T cell epitopes present in proteins present in pathogens can be determined. In addition to pathogens, other proteins, such as proteins involved in an autoimmune response can be examined for the presence of T cell epitopes. In still another use of the invention, factors involved in regulating the 10 development and function of human hematopoietic cells can be determined. These factors can serve to both identify the biological properties of these factors and to test their effectiveness as therapeutic molecules in preclinical models. Of particular interest are factors that augment hematopoietic reconstitution (e.g., after bone marrow transplantation or other states of immunodeficiency), can direct the targeting of immune responses towards cellular vs. humoral 15 effectors, suppress harmful inflammatory or autoimmune responses, and stimulate immune mediated clearance of infectious agents or cancerous tissue. EXAMPLES The following examples are submitted for illustrative purposes only and should not be interpreted as limiting the invention in any way. 20 Example 1. Allogeneic reconstitution of RAG -2 mutant mice. To compare the ability of CB 17.SCID (SCID; H-2d) and RAG2* mutant (RAG; H-2b) mice to support development of bone marrow (BM)-derived lymphocyte precursors, animals were engrafted intravenously with 107 T-cell depleted BM cells approximately 72 hours after birth. RAG mice were engrafted with syngeneic (H-2b) C57B11/6 or 25 (129xC57B1/6)F1 BM (RAG-(syn); H-2b -> H-2b) or with fully allogeneic Balb/c BM (RAG (allo); H-2 -> H-2b). SCID mice were engrafted with syngeneic Balb/c BM (SCID-(syn); H 2 -> H-2d) or with fully allogeneic C57Bl/6 BM (SCID-(allo); H-2b -> H-2). Hind leg bones (femur and tibia) were taken from euthanized 4 to 8 week old healthy donor mice and flushed using a 25 gauge needle attached to a 3 ml syringe filled with 30 cold DPBS to obtain cells in a single cell suspension. Cells were washed with DPBS and pelleted. Since bone marrow preparations contain functionally mature T cells, T cells were depleted using 10 ptg/ml anti-Thy-I mAb (30H 12) at 2-5 x 10 7 cells/mi for 30 minutes on ice. Cells were then centrifuged, resuspended in anti-rat IgG (MAR 18.5) culture supernatant at WO 01/15521 PCT/USOO/23971 -27 approximately 2-5x10 7 cells/ml. Guinea pig and rabbit complement were then added to 1:20 v:v each and incubated at 370 for 45 minutes. Cells were centrifuged and resuspended in 3 ml room temperature DPBS, and underlayed with 3 ml room temperature Histopaque density = 1.119 and centrifuged for 10 minutes at 800 g. Viable cells were collected from the interface, 5 washed in 2% FCS/DPBS, counted and resuspended in DPBS to a concentration of 3x10 8 cells/ml with no FCS Mice were anesthetized using an injectable ketamine/xylazine (100 and 20 mg/kg, respectively) (-200 1d for average mouse) solution. 3x10 7 (100 pl) T-depleted bone marrow cells were injected intravenously via the retro-orbital sinus into the appropriate donor 10 (same sex, 5 to 6 weeks of age). Neonatal mice received no more than 50 pl of injectate containing 10 7 bone marrow cells via the retro-orbital vein with no injectable anesthetic, just reduction of body temperature with ice. Both syngeneic (same MHC haplotype) and allogeneic (MHC haplotype mismatched) mice were transplanted. Mice were left for 8 weeks to allow bone marrow to engraft at which time the mice (usually 3 in each group) were is euthanized and organs removed for study. A thorough cellular analysis (FACS ) of the thymus, spleen, mesentaeric lymph nodes (in some cases) and peripheral blood (PB) was performed. T cells, B cells, and granulocytes were assessed in these areas using fluorescence conjugated cell type-specific antibodies. See Table 1. Table 1: Cell numbers and percentages for spleen, lymph node and thymus 20 for representative 7 to 8 week old neonatally en2rafted animals. Spleen Lymph Nodes Thymus Cell # (%CD3/B220) Cell # (%CD3/B220) Cell # (CD4/CD8) SCID (syn) 5.4 x 10 (28/53) 7.8 x 107 (79/17) 1.0 x 108 (6/15) SCID (allo) 7.5 x 107 (30/58) 4.8 x 107 (65/31) 2.6 x 108 (3/10) RAG (syn) 6.8 x 10' (28/33) 2.2 x 107 (78/18) 1.0 x 108 (4/10) RAG (allo) 3.6 x 10 7 (49/44) 2.6 x 10' (83/15) 4.6 x 10' (5.15) (129xB6)a 9.9 x 10 (32/58) 2.7 x 107 (63/29) 11 x 108 (3/10) Balb/ca 1.1 x 108 (35/54) 2.2 x 10 (66/23) 6.6 x 10 (2/12) RAG 2.-/a 5.8 x 106 (0/5) 3.1 x 10 5 (0/3) 2.5 x 106 (0/0) a: averaged data As can be seen, SCID animals tended to engraft higher percentages of B cells than RAG animals. This was found to be true over a number of different time points. Thymuses of eight week old animals engrafted well, and exhibited normal percentages of 25 CD4* and CD8' single positive cells. Additionally, thymus flow cytometric profiles are comparable, including the percentage of CD3* cells. The RAG (allo) chimera depicted had WO 01/15521 PCT/USOO/23971 -28 slightly more immature double negative (DN) thymocytes; however, an enrichment for DN thymocytes was not a reproducible finding over the X number of RAG-(allo) mice examined during the course of this study. Remaining mice of successful chimera studies were then immunized with 50 pg 5 KLH in CFA intraperitoneally and boosted 2 weeks later with KLH in IFA. Immunized mice were bled 1 week later and sera samples were tested for total IgG as well as IgG 1 (in some cases) by ELISA. See Figure 2 and Table 2 below. Table 2: Characterization of the functional development of hematopoietic cells in novel allogeneic chimaeras 10 Donor strain Recipient strain Engraftmt? Ab prod? B6 (H-2b) 129 RAG-2 (H-2b) Yes Yes Balb/c (H-2d) 129 RAG-2 (H-2b) Yes Yes Balb/c (H-2d) Baib SCID (H-2d) Yes - N/A B6 (H-2 b) Bab SCID (H-2d) No N/A B6 (H-2b) CID/ 129 RAG-2 (H-2b) CD4's Yes Balb/c (H-2d) CiD/ 129 RAG-2 (H-2b) CD4s, a few Slight CD8s B6 (H-2b) C2D/ 129 RAG-2 (H-2b) Mediocre N/A Balb/c (H-2d) C2D/ 129 RAG-2 (H-2b) No N/A B6 (H-2b) Irrad CID/ 129 RAG-2 (H-2b) Yes Yes Balb/c (H-2d) Irrad CID/ 129 RAG-2 (H-2b) Yes Yes B6 (H-2b) Irrad C2D/ 129 RAG-2 (H-2b) Yes Yes Balb/c (H-2d) Irrad C2D/ 129 RAG-2 (H-2b) Yes Yes Balb/c (H-2 d) Bab RAG-2 (H-2d) Yes Yes B6 (H-2b) Baib RAG-2 (H-2d) Yes Yes B6 (H-2b) B6 SCID (H-2b) Yes N/A Balb/c (H-2 d) B6 SCID (H-2b) No N/A B6 (H-2b) (129XB6) RAG-i (H-2b) Yes N/A Balb/c (H-2 d) (129XB6) RAG-i (H-2b) No N/A B6 (H-2b) repeat (129I6) RAG1 (H-2b) Yes Yes Balb/c (H-2 ) repeat (129XB6) RAG-i (H-2) No N/A Balb/c (11-2) (129XB6) RAG-I (H-2 ) 5E7 No N/A Balb/c (H-2 d) (129XB6) RAG-i (H-2b) irrad Yes Yes Balb/c (H-2Cd) 129 RAG-2 (H-2b) Yes Yes AKR (H1-2') 129 RAG-2 (H-2b) Yes Yes WO 01/15521 PCTUSOO/23971 -29 AKR (H-2k) 129 RAG-2 (H-2b) Yes Unknow n TCRtgxAKR (H-2k/s) 129 RAG-2 (H-2b) Yes unknow n B6 (H-2b) TCR p/S ko (H-2b) (irrad) Yes Yes Balb/c (H-2d) TCR p/S ko (H-2b) (irrad) Yes Yes As has been reported, SCID-(allo) mice responded poorly, indicating a lack of antigen-specific cognate T-B interactions. This has been presumed to be due to positive selection of donor CD4* T cells by the host thymic epithelium, resulting in T cells unable to 5 interact with donor MHC-expressing B cells. In contrast, RAG-(allo) mice produced equivalent levels of antigen-specific IgG as RAG-(syn) animals (Figure 2 A). Prebleed serum routinely showed absorbances equal to background. In addition, serum IgG was not cross reactive when tested on ovalbumin (OVA) coated plates. Antigen specific IgM responses were also noted. This phenomenon was investigated for a second antigen, KLH. As is shown in 10 Figure 2 B for two independent RAG-(allo) animals, the antigen-specific serum IgG response was found to be on the order of control RAG-(syn) responses, and developed with similar kinetics. Serum from all mice was found not to cross-react on ELISA plates coated with an irrelevant antigen. This result suggests that donor derived BM cells are capable of positively selecting CD4* T cells which can then interact with donor derived antigen-specific B cells in is the periphery, resulting in an isotype switch to IgG. Functional analyses on peripheral lymphocytes from chimeric animals were also performed. CD4* T cells were isolated from the lymph nodes of engrafted animals, and tested for reactivity in a mixed lymphocyte culture (MLR). Figures 1A and C depict proliferative responses of RAG-(syn) (A) and SCID-(syn) (C) CD4* T cells to LPS-induced splenic blasts 20 from C57Bl/6, Balb/c and third party H-2k expressing mice. Both RAG-(syn) and SCID-(syn) were tolerant to self, but were responsive to alloantigens. RAG-(allo) mice were tolerant to both C57B1/6 and Balb/c and were responsive to third party H-2k alloantigens. However, SCID-(allo) was functionally compromised in that a small response was mounted to Balb/c derived stimulators, indicating incomplete tolerance to self MHC. Additionally, the response 25 to third party H-2k expressing stimulators was impaired. A control RAG-(syn) created with the same BM inoculum as injected into the SCID-(allo) is shown in panel D. Thus, RAG-(allo) mice were found to be tolerant to both donor and host MHC, and were responsive to third WO 01/15521 PCT/USOO/23971 -30 party, but SCID-(allo) was functionally impaired. In addition, the mitogen reactivity of splenocytes and lymph node cells was tested. RAG-(allo) splenocytes responded normally to the T cell mitogen PHA, while SCID-(allo) T cells were hyporesponsive. All engrafted animals' splenocytes showed control level responses to LPS. 5 To further investigate the apparent donor restricted T cell responses, RAG-(allo) animals were immunized in the hind foot pads, then CD4+ T cells were purified from draining lymph nodes. Primed CD4* T cells were co-cultured with antigen-pulsed LPS-induced splenic blasts, and proliferation was assessed. Figure 3 shows the proliferative responses of KLH primed draining CD4* T cells from two different RAG-(allo) animals. The response to KLH io pulsed C57Bl/6 stimulators was found to dominate, indicating preferential selection of CD4* T cells to recognize antigen in the context of the thymic epithelium MHC. However, an antigen specific response to KLH-pulsed Balb/c blasts represented approximately 30% of the control response. This experiment was repeated 5 times, with a similar level of donor restricted response noted (range = 20 to 50% of host response). Similar results were obtained with adult 15 engrafted mice. These results suggest that for the fully allogeneic H-2d -> H-2b combination created in unirradiated neonatal RAG hosts, donor derived BM cells positively selected CD4 T cells. It had been well accepted that thymic epithelium affected the majority of the positive selection occurring in the thymus. However, selection by other cell types and across MHC 20 barriers had been observed only for CD8* T cells. Positive selection of both MHC class I restricted (Pawlowsky, et al. Nature 364:642-5 (1993)) and class II restricted (Hugo, et al. Proc. Nat'l Acad. Sci. 90:10335-10339 (1993)) T cells had been demonstrated to occur on transfected fibroblasts injected intrathymically. Selection of MHC class II restricted cells was demonstrated when the thymocytes shared MHC haplotypes with both the thymic epithelium 25 and the injected fibroblasts. This constraint was not evident for the selection of MHC class I restricted cells. While thymic positive selection by BM cells for CD4' T cells did not occur in BM engrafted irradiated adult MHC class II-deficient mice, others have demonstrated functional restriction to donor MHC using parental into F1 bone marrow chimeras. On the other hand, functional restriction of CD4' T cells to donor MHC, indicating positive selection 30 by BM-derived cells, has not been demonstrated in fully allogeneic chimeras, although with a few exceptions (Longo, et al. Nature 287:44-47 (1980); Longo, et al. J. Immunol 130:2525 2528 (1983); and Longo, et al. Proc. Nat'l Acad. Sci. 82:5900-5904 (1985)). Taken together these data indicate that thymic epithelium and BM-derived cells must share MHC haplotypes WO 01/15521 PCT/USOO/23971 -31 to effect efficient positive selection (Zink, and Elliot), and the requirements for selection of CD4* T cells may be more strict than those for CD8* T cells. The results presented here suggest that the RAG2-'- mutant strain is unique because it provides an environment that allows for BM derived cell selection events to occur 5 efficiently and in the absence of haplotype sharing by the thymic epithelium. The uniqueness of the RAG mouse may be due to the non-leaky nature of the mutation. The SCID mouse is well known to occasionally develop cells with functional antigen receptors. The development of even a few antigen-receptor positive cells may be enough of a signal to the thymic microenvironment to induce functional changes which preclude the recruitment and/or 10 functionality of donor BM-derived cells capable of positive selection. Both neonatal and adult SCID mice were used in these experiments and neither exhibited the capacity to support donor allogeneic BM restriction. Therefore, RAG mice may represent a model whose lymphopoietic microenvironments are functionally frozen at a fetal developmental stage, as has been suggested by thymocyte phenotype. If RAG mice represent a "fetal" model, then selection 15 onto BM derived cells may be a normal event in the thymus, and this phenomenon may not have been routinely detected in other systems due to the use of the SCID mouse, or due to secondary effects of irradiation. Example 2. Development of transgenic mice expressing human cytokine genes. 20 Based on both animal and in vitro studies, the following set of transgenes either play a role in hematopoiesis, are expressed in the bone marrow or thymus, and/or the murine proteins show specificity for murine vs. human cells: IL-3, IL-6, IL-7, M-CSF, GM-CSF, stem cell factor, LIF and oncostatin M. Genomic clones for this set of human genes were obtained and used to select ES cell clones to derive transgenic mice. 25 Isolation of genomic clones. PCR primer sets were designed against either the 5' or the 3' end of genomic sequences so that constructs containing the genes could be readily identified by PCR. In addition, these primer sets were designed to distinguish between mouse and human genes. The following primers and conditions were used to identify the human clones: 30 human IL-7 3181-SP6-F2: 5' AAATCAAGCTTGAATGACAAACTCC 3' (SEQ ID NO:1) 3181-SP6-R2: 5' GGACAGCATGAAAGAGATTGGAGC 3' (SEQ ID NO:2) WO 01/15521 PCT/USOO/23971 -32 product size: 121bp annealing temperature: 60 0 C human SCF 5 20180-T7-F: 5' ATGCAAGCTTGATTCATCCTC 3' (SEQ ID NO:3) 20180-T7-R: 5' CGTGGTITJTATTCGAAATGC 3' (SEQ ID NO:4) product size: 176bp annealing temperature: 60 0 C 10 human LIF hLIF-3F: 5' TTCCTCTGGGTAAAGGTCTGTAAG 3' (SEQ ID NO:5) hLIF-3R: 5' TCCACTTGTAACATTGTCGACTTC 3' (SEQ ID NO:6) product size: 388bp annealing temperature: 60 0 C 15 human GM-CSF GMCSF2/3F: 5' CTCAGAAATGTTTGACCTCCAG 3' (SEQ ID NO:7) GMCSF2/3R: 5' GTCTGTAGGCAGGTCGGCTC 3' (SEQ ID NO:8) product size: 729 bp 20 Annealing temperature: 60 0 C human M-CSF 31HU-MCSF-F: 5' GAAGACAGACCATCCATCTGC 3' (SEQ ID NO:9) 31HU-MCSF-R: 5' TGTAGAACAAGAGGCCTCCG 3' (SEQ ID NO:10) 25 product size: 401 bp Annealing temperature: 60 0 C human IL-6 51-BSF2-F: 5' TGGTGAAGAGACTCAGTGGC 3' (SEQ ID NO:11) 30 51-BSF2-R: 5' TACTTCAAGGCGTCTCCAGG 3' (SEQ ID NO:12) product size: 225 bp annealing temperature: 60 0

C

WO 01/15521 PCT/USOO/23971 -33 Human genomic P1(for IL-6, M-CSF, and LIF), BAC(for IL-7 and SCF), and PAC(for GM-CSF) libraries (Genome Systems, Inc.) were screened using the gene specific primer sets, above. IL-3 and OM are closely lined to GM-CSF and LIF, respectively, and were not screened for in the first round. Positive clones were further confirmed by either other 5 primer sets for the same gene or by southern blot. The following primers were used for PCR confirmation: human IL-7 51 IL7F: 5' GGCGTTGAGAGATCATCTGG 3' (SEQ ID NO:13) 10 51 IL7R: 5' TGCAGCTGGTTCCTCTTACC 3' (SEQ ID NO:14) product size: 342 bp Annealing temperature: 60 0 C FIL7: 5' CATACAGCATTACAAATTGC 3' (SEQ ID NO:15) 15 RIL-7: 5' TGTAGATTCTGGCCTGC 3' (SEQ ID NO:16) product size: 322 bp annealing temperature: 60 0 C human SCF 20 SCF-DF: 5' CCAAACTTCTGGGGCATTTA 3' (SEQ ID NO:17) SCF-DR: 5' CTCTCCACTGTCCCTGCTTC 3' (SEQ ID NO:18) product size: 220 bp annealing temperature: 60 0 C 25 SCF-3F2: 5' GCATGGAGCAGGACTCTATTF 3' (SEQ ID NO:19) SCF-3R4: 5' AGTTTGTATCTGAAGAATAAAGCTAGG 3' (SEQ ID NO:20) product size: 160 bp annealing temperature: 60 0 C 30 human LIF hLIF-3F: 5' TTCCTCTGGGTAAAGGTCTGTAAG 3' (SEQ ID NO:21) hLIF-3R: 5' TCCACTTGTAACATTGTCGACTTC 3' (SEQ ID NO:22) WO 01/15521 PCTUSOO/23971 -34 product size: 388bp annealing temperature: 60 0 C human OM 5 OSM5F1: 5' CCTAAAGTGAGGTCACCCAGAC 3' (SEQ ID NO:23) OSM5R1: 5' CTCTGTGGATGAGAGGAACCAT 3' (SEQ ID NO:24) product size: 456 bp annealing temperature: 60 0 C 10 OSM3F1: 5' GAGATCCAGGGCTGTAGATGAC 3' (SEQ ID NO:25) OSM3R1: 5' GATGCTGAGAAGGGGAGAGAG 3' (SEQ ID NO:26) product size: 384 bp annealing temperature: 60 0 C 15 human GM-CSF GMCSF1/2F: 5' AGCCTGCTGCTCTFGGGCAC 3' (SEQ ID NO:27) GMCSF1/2R: 5' CTGGAGGTCAAACATTTCTGAG 3' (SEQ ID NO:28) product size: 282 bp annealing temperature: 60 0 C 20 GMCSF3/4F: 5' ATGGCCAGCCACTACAAGCAG 3' (SEQ ID NO:28A) GMCSF3/4R: 5' GGTGATAATCTGGGTFGCACAG 3' (SEQ ID NO:29) product size: 878 bp annealing temperature: 60 0 C 25 human IL-3 IL-3F: 5' CGTCTGTTGAGCCTGCGCAT 3' (SEQ ID NO:29A) IL-3R: 5' AAATCTCCTGCCATGTCTGCC 3' (SEQ ID NO:29B) product size: 298 bp 30 annealing temperature: 60 0 C human M-CSF WO 01/15521 PCT/USOO/23971 -35 HUM-CSF-5F1: 5' GAGGGAGCAAGTAACACTGGAC 3' (SEQ ID NO:30) HUM-CSF-5R1: 5' CGTCTTCCTAGTCACCCTCTGT 3' (SEQ ID NO:31) product size: 322 bp 5 annealing temperature: 60 0 C human IL-6 IL6-3F: 5' CTAGATGCAATAACCACCCCTG 3' (SEQ ID NO:32) IL6-3R: 5' CAGGTTCTGACCAGAAGAAGG 3' (SEQ ID NO:33) 10 product size: 217 bp annealing temperature: 60 0 C Plasmid DNA from P1, BAC, or PAC clones was prepared using the KB-100 Magnum columns (Genome Systems, Inc.). Detailed experimental procedures were described in detail in the user's manual supplied by the manufacturer. To quantify DNA concentrations, is DNA constructs were digested with EcoRI, followed by electrophoresis on 0.8% agarose gel along with DNA standards with known concentration. Plasmid DNA concentrations were determined by comparison with the standards. The following DNA constructs were identified as having the full structural sequences for the target genes based upon the presence of both 5'- and 3'- ends of the coding 20 regions: IL-7: BAC20854 (100kb), BAC2267C7 (110kb), PAC24404 (90kb) SCF: BAC21029 (145kb) LIF: P1-20872 (100kb), P1-20873 (100kb) GM-CSF: PAC21689 (150kb), PAC21691 (194kb) 25 M-CSF: P1-3882 (55kb) IL-6: P1-3877 (n/d), P1-3878 (65kb) The sizes of the clones (in parentheses) were determined by restriction digestion with NotI followed by pulse-field gel electrophoresis. The gel running conditions were set as followed: 30 initial switch time: 1 sec final switch time: 6 sec total run time: 12 hrs voltage: 6 v/cm WO 01/15521 PCT/USOO/23971 -36 angle: 1200 Demonstration of in vitro transcription from the genomic constructs To determine the ability of the human genomic clones to be transcribed by murine cells, undigested genomic constructs were transfected into MM54 cells (a murine cell 5 line that expresses the endogenous form of each cytokine, ATCC # 6434-CRL) by lipofection using Tfx5O (Promega) according to the manufacturer's instructions. The cells were harvested 48 hours later for mRNA analysis. Total RNA was prepared using the UltrspecTm RNA isolation system (Biotecx). 10mg of gelatin carrier protein was added prior to ethanol precipitation to enhance RNA yield. First strand cDNA was synthesized by standard reverse to transcription methods. Briefly, the RNA was resuspended in 29.5 ml of H 2 0, and mixed with 10 ml 5x first strand buffer, 2.5 ml 10mM dNTP, 5 ml 0.1M DTT, 1 ml 0.5mg/ml random primer, and 2 ml M-MLV reverse transcriptase (Life Technologies). The reaction was incubated at 37 0 C for 1 hour, and the cDNA was purified by Phenol/Chloroform extraction. The resulting cDNA was then resuspended in 20 ml of H 2 0. Human specific transcripts were 15 analyzed by nested-PCR methods. 1 ml of cDNA sample was first amplified with the first PCR primer-set for 30 cycles. After that, a 5 pl aliquot was taken from the reaction mixture and subjected to a second round of PCR with the nested PCR primer set for an additional 30 cycles. The DNA samples were resolved on a 1% agarose gel. Primer sets for nested-PCR: 20 human IL-7 1 4 round primers (Clontech): CTIL-7F: 5' ATGTTCCATGTTTCTTTTAGGTATATCT 3' (SEQ ID NO:35) CTIL-7R: 5' TGCATTTCTCAAATGCCCTAATCCG 3' (SEQ ID NO:36) 25 product size: 681 bp annealing temperature: 60 0 C 2 "d round primers: hIL-7F1: 5' GCATCGATCAATTATTGGACAGC 3' (SEQ ID NO:37) hIL-7R1: 5' CTCTTTGTTGGTTGGGCTTCAC 3' (SEQ ID NO:38) 30 product size: 280 bp annealing temperature: 60 0 C human SCF WO 01/15521 PCTIUSOO/23971 -37 1 S" round primers: hSCF5F3: 5' CACTGTTTGTGCTGGATCGCAG 3' (SEQ ID NO:39) hSCFB-R: 5' TGAGACACGTGCTTTCTCTTCC 3' (SEQ ID NO:40) product size: 1173 bp 5 annealing temperature: 60 0 C 2 nd round primers: hSCF3F1: 5' CAGCCAAGTCTTACAAGGGCAG 3' (SEQ ID NO:41) hSCFA-R: 5' AGACCCAAGTCCCGCAGTCC 3' (SEQ ID NO:42) product size: 364 bp 1o annealing temperature: 60 0 C human LIF: 1 0 round primers: hLIF-F1: 5' TAATGAAGGTCTTGGCGGCAGGAG 3' (SEQ ID NO:43) 15 hLIF-R2: 5' TCCTGAGATCCCTCGGTTCACAGC 3' (SEQ ID NO:44) product size: 652 bp annealing temperature: 60 0 C 2 " round primers: hLIF-F2: 5' AACAACCTCATGAACCAGATCAGGAGC 3' (SEQ ID 20 NO:45) hLIF-R1: 5' ATCCTFTACCCGAGGTGTCAGGGCCGTAGG 3' (SEQ ID NO:46) product size: 402 bp annealing temperature: 60 0 C 25 human GM-CSF: 1 " round primers (from Clontech): CT-hGMCSF-F: 5' ATGTGGCTGCAGAGCCTGCTGC 3' (SEQ ID NO:47) CT-HGMCSF-R: 5' CTGGCTCCCAGCAGTCAAAGGG 3' (SEQ ID 30 NO:48) product size: 424 bp annealing temperature: 60 0 C 2 "d round primers: WO 01/15521 PCT/USOO/23971 -38 hGMCSF-F1: 5' CGTCTCCTGAACCTGAGTAGAG 3' (SEQ ID NO:49) hGMCSF-R1: 5' CAAGCAGAAAGTCCTTCAGGTTC 3' (SEQ ID NO:50) product size: 276 bp annealing temperature: 60 0 C 5 human IL-6: 1 " round primers (from Clontech): CT-hIL6F: 5' ATGAACTCCTTCTCCACAAGCGC 3' (SEQ ID NO:51) CT-hIL6R: 5' GAAGAGCCCTCAGGCTGGACTG 3' (SEQ ID NO:52) 10 product size: 628 bp annealing temperature: 60 0 C 2 nd round primers: hIL6-F2: 5' TGGGGCTGCTCCTGGTGTTGC 3' (SEQ ID NO:53) hIL6-R2: 5' CAGGAACTCCTTAAAGCTGCG 3' (SEQ ID NO:54) 15 product size: 560 bp annealing temperature: 60 0 C human M-CSF: Is round primers: 20 hMCSF-F: 5' CTCTCCCAGGATCTCATCAGCG 3' (SEQ ID NO:55) hMCSF-R1: 5' CAGGATGGTGAGGGGTCTTAG 3' (SEQ ID NO:56) product size: 492 bp annealing temperature: 60 0 C 2 " round primers: 25 hMCSF-F: 5' CTCTCCCAGGATCTCATCAGCG 3' (SEQ ID NO:57) hMCSF-R2: 5' TTGCTCCAAGGGAGAATCCGCTC 3' (SEQ ID NO:58) product size: 410 bp annealing temperature: 60 0 C The following genomic clones produced human-specific transcripts and were 30 chosen for use in ES cell transfection: IL-7: BAC20854 SCF: BAC21029 LIF: P1-20872, P120873 WO 01/15521 PCT/USOO/23971 -39 GM-CSF: PAC21689, PAC21691 M-CSF: P1-3882 IL-6: P1-3878 Selection of ES clones that contain human cytokine genes. 5 Murine embryonic stem (ES) cells (either RAG-'- ES cells or 129Sv/J wild type ES cells) were transfected with 3 sets of genes of human hematopoietic growth factors. The liposome reagent, Tfx-50 (Promega) was used according to the manufacturer's instructions. Each set of genes contained equal molar concentration of 3 linearized growth factor DNA. The first DNA set (GM-CSF set) contained GM-CSF, M-CSF and IL-6. The second DNA set (IL-7 10 set) contained IL-7, SCF and LIF. The third DNA set contained all 6 transgenes. Plasmid DNA with a selectable marker, either PGK-Hyg (for RAG-/- ES cells) or PGK-Neo (for 129Sv/J wild type ES cells), was used for positive selection. Briefly, the growth factor DNA was linearized by digestion with Not1 . The growth factor DNA mixtures (2.18 gg) and linearized selectable marker DNA (0.42 pg) was is mixed in 1 ml serum free Opti-MEM media and incubated with 170.6 pg (97.5 l) Tfx-50 for 15 minutes at room temperature. The molar ratio of marker versus DNA mixture was 4:1 and the ratio of Tfx-50 versus total DNA (marker and growth factor DNA) was 25:1. Then, 6 - 9 x 106 ES cells in 5 ml of serum free Opti-MEM media were added to the DNA/liposome mixture and incubated for 1 hr at 370C. After 1 hr incubation, the cells were harvested and replated in 20 6 well plates at a concentration of 2.5 x 10 5 ES cells per well. Hygromycin (120 gg/ml) or G418 (400 pg/ml) selection was started 24 hrs post transfection. Drug resistant ES colonies were picked after 10 to 14 days of selection. Southern blot analysis of transgenic ES clones and determination of gene copy numbers. All DNA probes for the genes were generated by PCR from either human 25 genomic DNA or cDNA samples, then cloned into the pCRR2.1-TOPO vector. The PCR fragments were then recovered from the plasmid by EcoRI digestion and gel purification using, e.g., a Gel Extraction Kit (Qiagen). human IL-7: A 350bp genomic fragment was amplified from total human DNA with primer set 51 IL7F/51 IL7R (SEQ ID NOs: 13 and 14). 30 human SCF: A 1173bp cDNA fragment was amplified from cDNA extracted from human embryonic kidney cell line 293 with primer set hSCF5F3/hSCFB-R (SEQ ID NOs:39 and 40). This fragment was subsequently cloned into the pCRR2.1-TOPO vector.

WO 01/15521 PCT/USOO/23971 -40 After EcoRI digestion, an 808bp DNA fragment was purified from the gel and was used as the probe for identifying SCF. human LIF: The 388bp PCR fragment(hLIF-3F/hLIF-3R) was subcloned and used as the probe (hLIF-3F/hLIF-3R; SEQ ID NO:21 and 22). 5 human GM-CSF: The 424 cDNA fragment was generated by PCR with primer set CT-hGMCSF-F/CT-hGMCSF-R (SEQ ID NO:47 and 48) from human 293 cell cDNA samples. human M-CSF: The 400bp probe was generated with PCR primer set 31HU MCSF-F/31HU-MCSF-R (SEQ ID NO:9 and 10). 10 human IL-6: The 298bp probe was generated with primer set 51-BSF2-F/51 BSF2-R (SEQ ID NO:11 and 12). ES cell clones were analyzed by Southern blotting to confirm the presence of genomic sequences and to determine relative copy number in comparison to human DNA controls. 10 -g of DNA from each ES cell clones was digested with either EcoRI (for IL-7 and 15 SCF), BamHI (for LIF and IL-6), or HindIl (for GM-CSF and M-CSF) and resolved on 1% agarose gel. The DNA was transferred to a nylon membrane by alkaline transfer (user's manual, Genescreen Plus). The membranes were then prehybridized over night at 420C with standard formamide containing buffer (Ausubel). Each probe was labeled using the Prime-It II Kit (Stratagene) and then added to the membrane. The hybridizations were carried out 20 overnight with rotation at 420C. The membranes were washed two times at room temperature with the low stringency buffer (2xSSC, 0.1% SDS) for 10 min each, and two more times at 500C for 10 min each. The membranes were then dried by blotting in between two layers of Whatman paper, and exposed to phospho screens (Molecular Dynamics). The image was quantified by the STORM System (Molecular Dynamics). The copy number for each 25 transgene was derived by comparison with the human control. Over 3400 drug resistant colonies were picked, of these, 264 clones have been expanded. From these clones, 179 ES clones were found suitable for injection. Among these injectable ES clones, 2 had 6 genes, 18 had 5 genes and 95 clones had 3 transgenes. The copy number for the same gene varied among different clones. For 30 example, clone 6 had one copy of IL-6 gene whereas clone 15 had two copies of the same gene. On the other hand, within the same clone, the copy number of one gene varied from the other gene. For example, clone 18 had one copy of IL-7 gene, two copies of SCF gene, and three copies of LIF gene.

WO 01/15521 PCT/USOO/23971 -41 Generation of transgenic mice ES cell clones containing the human cytokine genes are used to derive transgenic mice as described in Robertson (ed), Teratocarcinomas and embryonic stem cells a practical approach (1987), IRL Press. ES cells are injected into 3.5 day p.c. C57BL/6 5 embryos and implanted into the uterus of pseudopregnant females and allowed to develop to birth. Male chimeras are mated with wild type C57B11/6 females to obtain germline transgenic lines. Identification of mice with human transgenes There are two ways to identify mice that have incorporated human transgenes: 10 Southern Blot and PCR analysis. It is preferable to use PCR to genotype the mice due to its speed and ease of experimental procedure. However, whenever there is concern about the validity of PCR results, Southern Blot should be carried out to confirm the results. Briefly, DNA samples were isolated from the tips of mice tails following standard protocols (Qiagen manual, DNeasy 96 Tissue Kit). PCR analysis using human-specific primer sets (see Isolation 15 of genomic clones) was performed for each transgene. A positive control sample containing human DNA and a negative control sample containing mouse DNA were also carried out at the same time to ensure the specificity of the PCR products. Only mice that contain the expected human transgenes were selected for further breedings and experiments. Seven independent lines of transgenic mice have been established so far. Clone 20 12 and clone 71 have IL-6, M-CSF, and GM-CSF. Clone 74 and clone 75 have IL-7, SCF, and LIF. Clone 182 and clone 185 have all six transgenes in the germline, whereas clone 201 has every gene except for LIF. The same procedure was done throughout the breeding process to ensure the genotypes of the mice. Demonstration of in vivo trascription from the genomic constructs 25 Total RNA samples were prepared (RNeasy Midi Kit, Qiagen) from nine tissues of each transgenic mouse, including spleen, thymus, liver, kidney, heart, muscle, lung brain, and bone marrow. Total cDNA was prepared as previously described (see Demonstration of in vitro trascription from the genomic constructs). Gene expression analysis for each transgene was carried out by nested-PCR (see above). To ensure the reproducibility of the 30 results, at least two mice from each genotype were analyzed by this method. Mice from clone 71, which have human IL-6, M-CSF, and GM-CSF, showed expression of all three transgenes in different tissues. Human IL-6 was mainly expressed in the spleen and thymus. Human GM-CSF expression was restricted in the thymus. On the other WO 01/15521 PCT/USOO/23971 -42 hand, human M-CSF has a much wider tissue distribution, with transcripts in the spleen, thymus, liver, kidney, heart, muscle, lung, and brain. Mice from clone 75, which have human IL-7, SCF, and LIF, also showed expression of all three transgenes in tissues. Human IL-7 and SCF seem to have a wide distribution pattern similar to M-CSF, whereas human LIF 5 expression was restricted to the brain. Mice from other ES clones that contain either IL-6, M-CSF, and GM-CSF, or IL-7, SCF, and LIF were also analyzed by the same method. Although the expression pattern vary in certain tissues, the overall pattern was similar. This variation may be attributed to the difference in the insertion site of the transgenes and the copy numbers for each gene. The 10 murine endogenous genes were also analyzed by nested-PCR. Despite differences in certain specific tissues, the expression pattern largely agrees with that of the human transgenes. Protein expression of human transgenes in serum and in the supernatant of bone marrow stromal cell cultures derived from the injected mice were examined by ELISA. It was found that transgene expression patterns and levels varied between is clones, litters, littermates, and even stromal cells from same mouse. The following table provides ELISA results from mice injected with 4 ES cell clones Table2: Transgene Expression Patterns GM-CSF set of transgenes IL-7 set of transgenes ES clone Media M-CSF IL-6 GM-CSF SCF IL-7 LIF clone 12 Serum X -- -- N/A N/A N/A Supernatant -- -- -- N/A N/A N/A clone 71 Serum X X -- N/A N/A N/A Supernatant X X -- N/A N/A N/A clone 74 Serum N/A N/A N/A X -- - Supernatant N/A N/A N/A -- -- clone 75 Serum N/A N/A N/A X X - Supernatant N/A N/A N/A -- X - In addition to transgene transcription and translation, the ability of the stromal 20 cells to support hematopoietesis was investigated.

WO 01/15521 PCT/USOO/23971 -43 To examine the effects of transgenic murine hematopoietic microenvironment on human hematopoiesis, long term bone marrow cultures derived from transgenic or wildtype littermates were set up in tissue culture flasks. After 2 weeks of culture, a monolayer of bone marrow stromal cells will form 5 and adhere to the bottom of flasks. Hematopoietic stem/progenitor cells then adhere to the stromal layer. As hematopoietic cells proliferate and differentiate, they become non-adherent and float freely in the supernatant of the culture. The cell number of differentiated cells can be counted and stained to determine the extent of proliferation and differentiation of the hematopoietic stem cells. 10 Once a stromal layer formed, the cultures were irradiated to eliminate murine hematopoiesis and to stop proliferation of stromal cells. Irradiation, however, maintained the ability of the stromal cells to support human hematopoiesis in vitro. After irradiation, human cord blood mononuclear cells were added to the culture. Cell counts were made weekly, as was a 50% change in media. Every week, the non-adherent cells were counted and analyzed is by FACS. Stromal cells from clone 71 transgenic mice supported human hematopoiesis in vitro better than clone 12 and clone 75. Non-adherent cells harvested from transgenic co cultures were greater in number than wild type co-cultures established from clone 71 littermates. Mixtures of stromal cells from clone 71 and 75 transgenic mice supported human 20 hematopoiesis in vitro better than stromal cells from either clone 71 or clone 75 alone, in terms of non-adherent cellularity. The effects of human transgenes on murine hematopoiesis were also examined. Expression of human transgenes increased bone marrow B cell progenitor production in transgenic littermates of clone 71 and 75 mice. 25 Example 3: Ability of irradiated H-2 -> H-2b C1D/RAG-2 and H-2b C2D/RAG-2 bone marrow chimaeras to support functional en2raftment. MHC class I deficient (C1D)/RAG-2 and class II deficient (C2D)/RAG-2 mice were tested to assess whether MHC was necessary to facilitate alloengraftment. Unirradiated 30 allogeneic C1D/RAG-2 chimeras produced antigen specific IgG antibody when chimeras contained greater than 10% donor B lymphocytes in the peripheral blood. In comparison, irradiation (800 rads) of C1D/RAG-2 hosts led to a relative increase in the levels of donor cell WO 01/15521 PCT/USOO/23971 -44 engraftment, with a higher percentage of B cells in peripheral blood. All of these chimeras produced good antigen specific IgG antibody to KLH. Although radiation conditioning was not found to be an absolute requirement for the functional engraftment of allogeneic C1D/RAG-2 chimaeras, irradiated hosts supported more extensive cellular and functional 5 engraftment. In contrast to allogeneic C1D/RAG-2 chimeras, unirradiated C2D/RAG-2 mice were unable to support cellular alloengraftment, therefore irradiation preconditioning was used. The level of thymocyte development in H-2d -> H-2b C2D/RAG-2 mice that received 800R irradiation was significantly better. The relative percentage of CD4+ cells was diminished 1o relative to C 1D/RAG-2 chimeras, which correlated with the absence of host-expressed MHC Class II molecules, but CD4 development was present. These chimaeras elicited an anti-KLH antibody response following immunization, demonstrating the functional engraftment of these mice. This suggests that neither class I nor class II are absolutely required in the recipient for functional engraftment of RAG-2 mice. 1s Evaluating the MHC haplotype dependence in supporting donor MHC-restricted immunity. The following experiments were performed and conclusions were drawn that the RAG-2 mutation confers a "universal" property to support the functional development of allogeneic HSC. The following allogeneic and syngeneic bone marrow chimaeras were prepared: 20 (i) Balb/c (H-2d) -> Balb/c RAG-2 (H-2d) hosts (ii) C57Bl/6 (H-2b) -> Balb/c RAG-2 (H-2d) hosts (iii) Balb/c (H-2d) -> 129 RAG-2 (H-2 ) hosts (iv) C57B1/6 (H-2b) -> 129 RAG-2 (H-2 ) hosts (v) Balb/c (H-2d) -> C57B1/6 SCID (H-2b) hosts 25 (vi) C57B1/6 (H-2b) -> C57Bl/6 SCID (H-2b) hosts The first set of chimeras (i-iv) were designed to determine whether RAG-2 mutant mice on an H-2d background have the ability to support allogeneic donor-specific immunity. All of these chimeras supported functional engraftment. The second set of chimeras (v-vi) were prepared to test the ability of SCID 30 mutant mice, on an H-2b background, to support allogeneic donor-specific immunity. The allogeneic chimeras engrafted very poorly relative to the syngeneic group (thymuses were too small to sample) which is very similar to the H-2 into H-2b SCID results reported. These results suggest there is a significant difference between the SCID and RAG-2 mutations to WO 01/15521 PCTUSOO/23971 -45 support the cellular development of T lymphocytes independent of MHC haplotype. To assess whether RAG-2 hosts could support functional engraftment from a donor with an unrelated haplotype, an H-2k AKR -> H-2' RAG-2 chimera was produced. These mice supported donor-derived immunity. 5 Taken together, these results indicate the RAG-2 mutation supports bone marrow alloengraftment from different donor strains, independent of haplotype. This suggests these mice could support hematopoiesis from any donor. Evaluating other mutations for donor-derived immunity RAG-2 mice appeared to be unique relative to other immunodeficient strains in 10 supporting donor-restricted immunity until transplantation studies in RAG-I and TCR P/6 (with irradiation to eliminate host B cells) mice were performed. Although unirradiated RAG 1 chimaeras did not support engraftment from allogeneic donors, irradiated (800 rads) RAG-1 mice supported functional engraftment. Both RAG-I and RAG-2 genes are required to initiate T and B lymphocyte receptor rearrangements. In addition, studies showed that irradiated TCR is p/6 mice also supported donor-derived immunity. Cytochrome-c specific TCR transgenic (H-2k class II-restricted)-> H-2b RAG-2 bone marrow chimaeras support cellular engraftment of donor CD4+ T cells in the absence of host expression of cognate MHC Class II molecules. In order to determine the mechanism of donor-derived immunity, SJL 20 TgN(TcrAND)53Hed mice were obtained from Jackson Laboratory and backcrossed onto AKR (H-2k) mice to provide the appropriate MHC Class II molecule (I-Ek) for positive selection of TCR-transgenic (TCR-tg) T cells (which recognize cyt-c in the context of I-Ek). Bone marrow from these mice was used to engraft H-2b RAG-2 mice which do not express the cognate MHC Class II receptor for the transgenic T cells. This created a host environment for 25 the transgenic bone marrow cells that is functionally equivalent to a Class II knockout background. Donor T cell development would therefore be dependent on donor-derived antigen presenting cells to positively select TCR-tg T cells. The percentages of thymocytes in TCRtgxAKR -> RAG-2 chimaeras were similar to that of wild type AKR->RAG-2 mice. Both of these chimeras have overall lower 30 percentages of T cells in comparison to TCRtgxAKR donor mice consistent with other haplotype combinations of allogeneic RAG-2 chimaeras. The level of B cell reconstitution was relatively high (greater than 10%). Both the TCRtgxAKR -> RAG-2 and AKR->RAG-2 were immunized with cytochrome c to determine their ability to produce antigen-specific IgG WO 01/15521 PCT/US00/23971 -46 antibody. Example 4. Development of transgenic mice expressing human HLA Class II genes. An alternative embodiment of the invention involves the expression of human 5 HLA Class II molecules in MHC Class II-bearing tissues of the mouse. In this example, donor HSC are introduced that express the same HLA haplotype(s) as the transgenic HLA Class II molecules. This combination provides cognate interactions between donor T lymphocytes, which develop in the context of the transgenic HLA Class II molecules expressed on the host tissues (in particular the thymus), and donor-derived B lymphocytes. Although the methods 10 for making transgenic mice that express human HLA Class II molecules of the DR3 haplotype are taught, these methods can be applied to any desired HLA haplotype, including those for Class I genes, for the purpose of evaluating responses representing other individuals in the population. Preparation of YA C DNA for lipofection: 15 YAC 4D1 spans approximately 550kb of the HLA Class II region (Ragoussis et al. Nucleic Acids Research, 20:3135-3138 (1992), and Ragoussis, et al. in Tsuji, et al. (eds.) HLA 1991, Oxford Univ. Press (1992)). It is bordered on one end by the RING3 gene, and the opposite end by DRa. It contains the DRa, DRb, DQa, and DQb chains of the DR3 haplotype. Yeast cultures containing the 4D1 YAC were grown in AHC media. Agarose 20 blocks were formed in 1% low melting temperature agarose containing approximately 3x10 9 cells/ml. The YAC was separated from yeast chromosomes by pulse-field gel electrophoresis in a 1% low melting temperature agarose gel. Running conditions were: 200V, 40 hours duration, with a 50 second switch time. After electrophoresis, the gel was cut lengthwise at the outer edges and in the middle. The three slices were stained with ethidium bromide to 25 visualize the position of the 4D1 YAC vis a vis the host chromosomes. The position of the 4D1 YAC was marked with notches and the marker pieces realigned with the unstained gel sections. A horizontal band containing the 4D 1 YAC was excised based on the position of the notches. The 4D 1 gel slices were equilibrated twice for one hour/each in 1X gelase 30 buffer (Epicentre Technologies) on a rotating platform. The buffer was changed after the second rinse and left at 4'C overnight. Based on the input amount of yeast DNA, the estimated amount of 4D1 DNA in the entire gel was approximately 8 mg. The following day, the gel slices were cut into 20 blocks weighing approximately one gram each, and placed into WO 01/15521 PCT/USOO/23971 -47 individual tubes. The gel fragments were melted at 70*C for 20 minutes and then equilibrated at 45"C for 15 minutes. Ten units of Gelase (Epicenter Technologies, lunit/ml) was added per tube and incubated at 45*C for 45 minutes. The gelase step was then repeated. Transfection of YAC DNA into ES cells 5 Each agarose block contained approximately 400 ng of YAC DNA, or 400 ng/ml. A neomycin resistance plasmid (PGKneo) was added at a molar ratio of approximately 4:1 (20 ng per ml of gel block). Transfectam (Promega, lot 318402) was added at a 50:1 weight:weight ratio (approximately 19 mg per ml of gel block) and the mixture allowed to sit at room temperature for one hour. ES cells had been split 1:2 the day before and 10 seeded onto 100mm plates. The cells were trypsinized on the day of transfection and resuspended at 3x106 cells/ml in serum-free ES media. One ml of the ES cells was placed into 60mm dishes with eight ml of serum-free ES media. One ml of DNA/lipid mixture was added and the cells were incubated at 37*C for 4 hours. Afterward, the lipofection/ES cell mixture was plated onto feeder cells at 1x10 6 ES cells per 100mm dish. G418 [400 mg/ml] was added 15 to the media the following day and changed every other day for 9-12 days until clones appeared. Individual clones were picked and grown in 96 wells. The cells were split 1:2 into duplicate 96 well plates. One plate was frozen in situ and the other was harvested for DNA analysis. Characterization of 4D1-positive clones: 20 The presence of the entire YAC was determined using PCR primers for six genes that span the entire 550kb: TAP-1, TAP-2, DQb, DQa, DRb, and DRa. The first screen involved the TAP-I and DRa primer sets. Clones that were double positive for these two end region genes were further screened with the remaining four primer sets. Tap 1 : 25 1069 F: CAC CCT GAG TGA TTC TCT (SEQ ID NO:59) 1069 R: ACT GAG TCT GCC AAG TCT (SEQ ID NO:60) Tap 2: 1231 F: GCG GAG AGA CCT GGA ACG (SEQ ID NO:61) 30 1231 R: TCA GCA TCA GCA TCT GCA (SEQ ID NO:62) DQa: GH26: GTG CTG CAG GTG TAA ACT TGT ACC AG (SEQ ID NO:63) WO 01/15521 PCTIUSOO/23971 -48 GH27: CAC GGA TCC GGT AGC AGC GGT AGA GTT G (SEQ ID NO:64) GH28: CTC GGA TCC GCA TGT GCT ACT TCA CCA ACG (SEQ ID NO:65) 5 GH29: GAG CTG CAG GTA GTT GTG TCT GCA CAC (SEQ ID NO:66) DRa: DRaF: CTT TGC AAG AAC CCT TCC C (SEQ ID NO:67) DRaR: ATA GCC CAT GAT TCC TGA GC (SEQ ID NO:68) 10 DR: GH46: CCG GAT CCT TCG TGT CCC CAC AGC ACG (SEQ ID NO:69) GH50: CTC CCC AAC CCC GTA GTT GTG TCT GCA (SEQ ID NO:70) is All product sizes are 300bp. Tap 1 PCR Program: 92C 15" 55C 30" 20 72C l' (30X) Tap 2 PCR Program: 96C 20" 65C 30" 25 72C 30" (30X) (Requires 2 rounds of PCR) DR and DQ PCR Program: 95C 15" 30 55C 30" 72C l' (30X) WO 01/15521 PCT/US00/23971 -49 Generation of transgenic mice Clone 4D 1.18 was used to derive transgenic mice as described in Robertson (ed), TERATOCARCINOMAS AND EMBRYONIC STEM CELLS - A PRACTICAL APPROACH (1987), IRL Press. ES cells were injected into 3.5 days p.c. C57BL/6 embryos and implanted into the 5 uterus of pseudopregnant females and allowed to develop to birth. Chimeric males were mated with wild type C57B/6 females to obtain germline transgenic lines. Antibody response of transgenic mice Four D1/C2D/RAG-2 (HLA-transgenic) mice were bled and their sera tested for I-Ealpha, I-Ebeta and DRalpha expression using FACS analysis. Mice that were confirmed to 10 express surface DR but not I-Ealpha were chosen for functional testing. These mice were immunized via the footpad with 50 jig/mouse. Three proteins were used as an immunogen. Two were fungal proteases and the third was a hybrid of the two proteins which has been found to be of reduced allergenicity in an in vitro human T cell epitope assay. The proteins were emulsed with CFA for total volume of 100 pl per footpad and boosted 2 weeks later with 15 the same concentration in IFA in the other footpad. Immunized mice were bled 1 week later and sera samples were tested for antibodies to the appropriate protein by ELISA. A second set of animals were immunized for antibody responses but using a different protocol which is as follows: mice were immunized intraperitoneally with 50 pg/mouse of the same three proteins emulsed with CFA for total volume of 100 il per mouse and boosted ip 2 weeks later with the 20 same concentration in IFA. Mice were bled 1 week later and sera samples were tested for antibodies. To assess whether the T cells in these transgenic mice were functioning normally, a positive control immunogenic peptide known to be a major T cells epitope (HSP65 1-20) was used. Mice were immunized according to a previously reported protocol by Geluk 25 et. al. Popliteal lymph nodes were taken and T cell proliferation assessed using a T cell proliferastion assay (also reported by Geluk et. al.) A summary of the array of experiments is below. Protein T Cell Response Antibody Response Protein 1 ND ++ Protein 2 ND ++ Hybrid protein ND + HSP65 epitope ++ ND WO 01/15521 PCTIUSOO/23971 -50 Tetanus Toxoid ND +++++ KLH ND +++++ The results suggest that the immune system in these transgenic mice is functionally intact and may be used to assess DR3-specific immune responses. 5 Example 5. Engraftment of transgenic immunodeficient mice with human hematopoietic stem cells. Transgenic immunodeficient mice that express human IL-3, IL7, SCF, LIF, IL6, M-CSF, OM, and GM-CSF are engrafted with a source of human hematopoietic stem cells. The age for engraftment is approximately 72 hours after birth. The preferred source of donor 10 cells is umbilical cord blood. For this example, the haplotype of the donor cells is selected on the basis of the test population of interest in order to determine the responsiveness of particular individuals to specific antigens, e.g., for the purpose of determining the allergenicity of proteins or peptides within the population. Source material for transplantation: 1s Human cord blood (CB) is obtained from Advanced Bioscience Resources Inc (ABR), Alameda, CA, or Purecell, San Mateo, CA. CB is collected by ABR from local hospitals within 24 hours of shipping and is processed on site. Alternatively, human bone marrow (BM), CB, and/or MPB cells is obtained from Purecell as either fresh or frozen cells, and fractionated or unfractionated cells. All samples are tested for Hepatitis B and C, and 20 HIV. Any experimental materials involving samples found to be virus positive are discarded immediately and animals removed, marked and disposed of in accordance with procedures for disposing contaminated animal carcasses. Throughout the course of the experiment, all samples are treated under the assumption that they may be contaminated with human blood borne pathogens (Biosafety Level 2, BL-2). All personnel handling mice with human blood 25 cells should receive the hepatitis B vaccine. CB-17/SCID and NOD/SCID mice are engrafted as experimental controls, as they are currently the strains of choice for investigators working in the field (Bhatia et al, Proc. Nat'l Acad. Sci. 94:5320-5325 (1997), Cashman et al. Blood 89: 4307-4316 (1997), Hogan, et al. Blood 90(1):85-96 (1997), Vormoor, et al Blood 83:2489-2497 (1994), and Wang, et al. 30 Blood 89:3919-3924 (1994)).

WO 01/15521 PCT/USOO/23971 -51 Finally, human HLA DR transgenic mice expressing human cytokine genes are engrafted. The expression of human major histocompatibility gene products in mouse tissues provides an alternate developmental pathway for thymocyte development. Procedures for injecting mice: 5 HSC-enriched populations are injected intravenously (i.v). Neonatal mice should receive no more than 50 gL of cells. Pups are held below a light source to better visualize the eye vein. The needle is held parallel to the vein and inserted by aiming just below the surface of the skin. Cells are slowly injected while watching to determine that blood is cleared downstream of the needle point, and that swelling around the vein does not occur. 10 After injection the needle is held in place for 10-30 seconds to allow the cells to enter the circulation. The animals are kept warm and the eye moistened until recovery. If necessary, pups may be chilled briefly on ice to reduce their activity. Analysis of mice At various time intervals ranging from two weeks to several months, mice are 15 analyzed for reconstitution of human immune cells. Peripheral blood is drawn from the orbital sinus or tail vein and analyzed by to flow cytometry using monoclonal antibodies specific for human hematopoietic cells. Animals are sacrificed and their peripheral immune system organs examined for the presence of human cells by flow cytometry (Pflumio, et al. Blood 88:3731 3740 (1996); Dick, et al. Stem Cells 15 Suppl 1:199-203 (1997); Ramirez, et al. Exp. Hematol. 20 26:332-344 (1998); Hogan, et al. Biol. Blood Marrow Transplant 3:236-246 (1997); Hogan, et al. Blood 90:85-96 (1997); and Lapidot, et al. J. Mol. Med. 75: 664-673 (1997)). Functional properties of mature human cells are determined, such as mitogen stimulation, mixed lymphocyte reaction (MLR) to allogeneic or xenogeneic cells, and the ability to produce different classes of immunoglobulins. Immunization with specific proteins, peptides, or 25 glycoproteins is performed to determine the ability to induce particular subclasses of antibody, and antigen-specific T cell responses. Example 6. Methods for obtaining functional human immune cells. Four types of human HSC chimeras are envisioned by the invention. The strains of host mice all contain human growth factor transgenes that provide the necessary 30 species-specific factors for supporting the human donor HSCs. The host strains vary in the nature of their immunodeficiency, and whether they include the transgenic expression of HLA molecules.

WO 01/15521 PCT/USOO/23971 -52 The first type of chimeras consists of RAG-I or RAG-2 mutant mice that contain human CTG. As shown in Example 1, allogeneic RAG chimeras display the ability to support donor-restricted immune responses. Thus, human CTG RAG mice are engrafted with human HSC from umbilical cord blood, and develop human T and B lymphocytes which 5 function to provide antigen-specific T and B cell responses. Engraftment of either neonatal or adult mice leads to mixed bone marrow chimeras in which hematopoietic-derived antigen presenting cells of both the host and donor are present. A major biological effect of mixed chimerism is the negative selection of "host"-reactive T cells, mediated by APCs of donor and host origin. This leads to the development of T cells that are tolerant to both donor and host 10 MHC haplotypes, but the potential to react to either allogeneic or xenogeneic antigens in a MLR. T cells that develop independently of the thymus are also supported by the CTG hosts. These primarily include intestinal epithelial lymphocytes of the small and large intestine, and other gut associated lymphocyte tissues (GALT). Although RAG mice support the development of donor haplotype-restricted T is cells, the major developmental pathway for thymocytes involves positive selection in the context of MHC molecules expressed on thymic epithelial cells. The second type of chimeras thus consists of human CTG RAG mice that also express human HLA genes in host tissues that endogenously express MHC molecules. This provides an additional pathway for the donor human HSC to develop in the RAG-mutant host, and results in some mature T cells that are 20 restricted to the HLA haplotype of the transgenes. It is preferable to match the HLA haplotype of the donor HSC to that of the transgenic HLA molecules such that the mature T cells can subsequently interact with donor-derived B lymphocytes through cognate T cell receptor:HLA interactions. The third type of chimeras are prepared in mice that are genetically 25 immunodeficient by means other than a mutation in the RAG genes, and express human HLA transgenes of the same haplotype as the source of donor HSC. For example, human GFTG SCID mice are mated with DR3 Class II transgenic mice and used to engraft HLA DR3 haplotype HSC. Some of the human T lymphocytes are positively selected in the context of HLA DR3 Class II molecules and are capable of interacting with DR3+ donor B lymphocytes 30 to produce antigen-specific antibodies. The fourth type of chimeras are prepared in immunocompetent human CTG mice that express human HLA transgenes of the same haplotype as the source of donor HSC. In order to engraft human HSC, these hosts must be depleted of their endogenous immune WO 01/15521 PCT/USOO/23971 -53 system, e.g., with a lethal dose of irradiation or other method. The pattern of lymphocyte development is then similar to that of the third type of chimeras, above.

Claims (29)

1. A recipient mammal comprising: 2 exogenous cytokines capable of maintaining an immune system; and 3 cells derived from donor-specific cells with hematopoietic stem cell 4 properties; wherein the recipient mammal is capable of facilitating production of donor s specific functional immunity.

2. The mammal of claim 1, wherein said mammal is a mouse.

3. The mammal of claim 1, wherein said cytokines are from the same 2 species as said cells derived from donor-specific cells with hematopoietic stem cell 3 properties. 1

4. The mammal of claim 1 wherein said cytokines are selected from 2 the group consisting of IL-3, 11-6, IL-7, M-CSF, GM-CSF, SCF, LIF and Oncostatin-M.

5. The mammal of claim 4, wherein said cytokines comprise IL-3, Il 2 6, IL-7. M-CSF, GM-CSF, SCF.

6. The mammal of claim 1, wherein said cytokines are introduced into 2 said mammal transgenically. 1

7. The mammal of claim 1, wherein said cells derived from cells with 2 hematopoietic stem cell properties are xenogeneic to the mammal. 1

8. The mammal of claim 7, wherein said cells are human cells. 1

9. The mammal of claim 1, wherein said cells are hematopoietic stem 2 cells.

10. The mammal of claim 9, wherein said hematopoietic stem cells are 2 derived from umbilical cord blood. 1

11. The mammal of claim 1, wherein said cells are embryonic stem 2 cells.

12. The mammal of claim 1, wherein the cells derived from cells with 2 hematopoietic stem cell properties are differentiated.

13. The mammal of claim 12, wherein the differentiated cells are of 2 myeloid lineage. 1

14. The mammal of claim 12, wherein the differentiated cells are of 2 lymphoid lineage. 1

15. The mammal of claim 12, wherein the differentiated cells are of 2 erythroid lineage. WO 01/15521 PCT/USOO/23971 -55

16. A method of making a mammal with a donor immune 2 system, comprising the steps of: 3 introducing transgenes into a mammal deficient in VDJ recombination, 4 wherein said transgenes encode cytokines necessary for the maintenance and maturation 5 of donor-derived hematopoietic cells; 6 introducing cells with hematopoietic stem cell properties into said 7 mammal.

17. The method of claim 16, wherein said donor immune system is a 2 xenogeneic immune system.

18. The method of claim 17, wherein said donor immune system is a 2 human immune system.

19. The method of claim 16, wherein the step of introducing the 2 transgenes is through pronuclear transfer.

20. The method of claim 16, wherein the transgenes are in an 2 embryonic stem cell.

21. The method of claim 16, wherein the step of introducing the 2 transgenes is through breeding said mammal with a mammal that comprises the 3 transgenes.

22. The method of claim 16, wherein the mammal is a RAG-I or a 2 RAG-2 mouse.

23. The method of claim 16 wherein said cytokines are selected from 2 the group consisting of IL-3, 11-6, IL-7. M-CSF, GM-CSF, SCF, LIF and Oncostatin-M.

24. The method of claim 23, wherein said cytokines comprise IL-3, Il 2 6, IL-7. M-CSF, GM-CSF, SCF.

25. The method of claim 16, wherein said donor-derived hematopoietic 2 cells are from a human..

26. The mammal of claim 1, wherein said mammal further comprises 2 MHC transgenes. 3

27. The mammal of claim 26, wherein said MHC transgene is a human 4 HLA transgene. WO 01/15521 PCT/USOO/23971 56 AMENDED CLAIMS [received by the International Bureau on 6 February 2001 (06.02.01); original claims 1-27 replaced by new claims 1-29 (3 pages)] 1. A recipient moise comprising: 2 a disruption in both alleles of a gene such that lymphocyte maturation does not 3 occur; and 4 donor specific transgenes that encode cytokines comprising IL-7, SCF and LIF. 2. A recipient mouse comprising: 2 a disruption in both alleles of a gene such that lymphocyte maturation does not 3 occur; and 4 donor specific transgenes that encode cytokines comprising GM-CSF, M-CSF and 5 IL-6. 3. A recipient mouse comprising: 2 a disruption in both alleles of a gene such that lymphocyte maturation does not 3 occur; and 4 donor specific transgenes that encode cytokines comprising I:-7, SCF, LIF, GM CSF, M-CSF and IL-6. 4. The recipient mouse of claims 1-3, wherein the disruption in a gene 2 that modulates VDJ recombination. 5. The recipient mouse of claim 4, wherein the mutation is in a RAG 2 gene. 6. The recipient mouse of claims 1-3, wherein the transgenes encode 2 human cytokines. 7. A method of making a mouse lacking in mature T and B cells and 2 comprising exogenous cytokines comprising the steps of: 3 inactivating VDJ recombination; and 4 introducing transgenes, wherein said transgenes encode human cytokines s necessary for support of human cells in the mouse. 9. The method of claim 7, wherein the step of introducing the 2 transgenes is through pronuclear transfer. 10. The method of claim 7, wherein the transgenes are in an embryonic stem cell. 11. The method of claim 7, wherein the step of introducing the 2 transgenes is through breeding said mouse with a mouse that comprises the transgenes. WO 01/15521 PCT/USOO/23971 57 1 12. The method of claim 7, wherein the mouse is a RAG-1 or a RAG-2 2 mouse. 13. The method of claim 7 wherein said cytokines comprise IL-7. SCF 2 and LIF. 1 14. The method of claim 7 wherein said cytokines comprise IL-6. GM 2 CSF and M-CSF. 1 15. The method of claim 7 wherein said cytokines comprise IL-7. SCF, 2 LIF, IL-6, GM-CSF and M-CSF. 1 16. The mouse of claim 1, wherein said mouse further comprises MHC 2 transgenes. 3 17. The mouse of claim 16, wherein said MHC transgene is a human 4 HLA transgene. 1 18. A recipient mouse comprising: 2 a disruption in both alleles of a gene such that lymphocyte maturation does not 3 occur; and 4 a human transgene comprising a nucleic acid sequence that encodes a MHC Class 5 II DR3 molecule, 6 wherein the transgene comprises naturally linked DRab and DQab alleles. 19. The mouse of claim 18, wherein the disruption is in a gene that 2 modulates VDJ recombination. 1 20. The recipient mouse of claim 19, wherein the mutation is in a RAG 2 gene. 21. The mouse of claim 20, wherein said mouse is deficient for murine 2 I-Ea. 1 22. The recipient mouse of claim 18, wherein the transgene further 2 comprises a human HLA DQ2 molecule. 23. A method of making a recipient mouse, said method comprising: 2 disrupting both alleles of a gene so that lymphocyte maturation does not occur; 3 inserting a transgene comprising nucleic acid that encodes MHC Class II DR3 and DQ2 4 molecules, wherein the DRab and DQab alleles are naturally linked; and 5 inactivating murine I-Ea. WO 01/15521 PCT/USOO/23971 58 1 24. The method of claim 23, wherein said disruption is in a gene that 2 modulates VDJ recombination. 1 25. The method of claim 24, wherein said gene is RAG-2. 1 26. The method of claim 25, wherein said transgene is in an artificial 2 Yeast chromosome. 1 27. The method of claim 26, wherein the transgene is about 550 kb in 2 length. 1

28. The method of claim 27, wherein the artificial yeast chromosome is 2 4D1. 1

29. A method of making a mouse, said method comprising: 2 preventing VDJ recombination by mutating both alleles of the RAG-2 gene; 3 inserting a transgene comprising the Drab and DQab alleles of the MHC Class II DR3 4 haplotype; and s inactivating murine I-Ea