WO1993005817A1 - Transgenic mhc class i and class ii antigen-deficient mammals - Google Patents

Abstract

Disclosed are transgenic non-human mammals, the surface of whose cells are deficient in all MHC class II antigens and one or more MHC class I antigens. Also disclosed are the use of mammals as a model system for MHC antigen immunodeficiencies and as a source of universal donor tissue for transplant or somatic cell therapy.

Description

TRANSGENIC MHC CLASS I AND CLASS II ANTIGEN-DEFICIENT MAMMALS Background of the Invention The invention relates to transgenic animals.

Major histocompatibility complex (MHC) class II antigens are polymeric cell surface glycoproteins which bind foreign antigens to form a complex that is recognized by T lymphocyte antigen receptors. This tripartite interaction triggers antigen-specific responses in the immune system to invading organisms such as bacteria and viruses. The class II antigens also act as immunogenic determinants, allowing an organism to distinguish self from non-self; it is this recognition which leads, e.g. , to graft rejection.

There are four urine MHC class II antigen genes: I-Aα, I-A3, I-Eα, and I-E3; and at least three human class II genes: HLA-DP, HLA-DR, and HLA-DQ. Expression of these genes is subject to an elaborate program of regulatory control, including tissue-specific and inducible expression.. Several functionally important promoter motifs located upstream of the coding sequences of mouse and human class II genes and nuclear proteins that bind to these sites have been identified. Defective MHC class II expression is associated with the disease, Bare Lymphocyte Syndrome (BLS) . This disease is a subset of severe combined immunodeficiency disease (SCID) and is characterized by a lack of all class II MHC expression because of an as yet undefined regulatory defect. Individuals afflicted with the disease generally develop recurrent, debilitating infections, and usually succumb, unless treated, within 2 years of birth, usually of lung abscesses, pneumocystic pneumonitis,^~or any of a variety of common viral infections, including chicken pox, measles, cyto egalovirus, and adenovirus. Currently, patients suffering from Bare Lymphocyte Syndrome receive bone marrow transplants at an early age.

Summary of the Invention In general, the invention features a transgenic non-human mammal, the surface of whose cells are deficient in all^' MHC class II antigens and one or more MHC class I antigens. In various preferred embodiments, a chromosome of the mammal includes a null mutation in an MHC class II antigen gene and a null mutation in the β2- microglobulin gene; the jS2-microglobulin gene and class II antigen genes including null mutations replace the endogenous jS2-microglobulin and MHC antigen genes; the mammal is homozygous for the mutant genes; and the mammal's cell surfaces are deficient in all MHC antigens. In a second aspect, the invention features a transgenic mouse, the surface of whose cells are deficient in MHC class I and class II antigens; a chromosome of the mouse includes a null mutation in the Aj0^b gene; a chromosome of the mouse includes a null mutation in the E_α ^b gene; the chromosome of the mouse includes a null mutation in the /3₂-microglobulin gene; and the mouse's cell surfaces are.deficient in all MHC- surface antigens. In a third aspect, the invention features a method of testing a substance for efficacy in the treatment of an MHC antigen deficiency. The method involves exposing a transgenic animal of the invention to the substance and then examining the animal's cell surfaces for the presence of the MHC antigen, the presence of the antigen being an indication that the substance is useful for the treatment of the deficiency.

In a fourth aspect, the invention features use of the isolated cells from the transgenic animal as a medicament of wound healing. In preferred embodiments. the cell used as a medicament is an epithelial cell, for example, a keratinocyte; the cell produces a recombinant growth factor; and the isolated cell is cultured in vitro prior to use for wound healing. In a fifth aspect, the invention features use of the transgenic tissue as a medicament for diseases requiring transplantation of tissue into a human. Preferably, the tissue used as a medicament is a whole organ, for example, a heart, a kidney, or a liver; or the method further involves introducing into the donor tissue of a gene which expresses a recombinant protein, for example, Factor VIII.

In a final aspect, the invention features a mammalian cell whose cell surface is deficient in all MHC class II antigens and an MHC class I antigen; the mammalian cell is isolated from a transgenic animal of the invention; the mammalian cell is a human cell; and the mammalian cell expresses a recombinant protein, e.g., Factor VIII. "Transgenic" as used herein means a mammal which includes a DNA sequence which is inserted by artifice into a cell and becomes a part of the genome of the animal which develops from that cell. Such a transgene may be partly or entirely heterologous to the transgenic animal. In the transgenic animals described herein, the DNA sequence is used to insertionally inactivate a gene encoding an MHC class II antigen. The inactivated class II antigen gene is homologously recombined into the mammal's chromosome, replacing all or most of the "endogenous" gene. The transgenic animals thus produced are "deficient" in MHC class II antigens, meaning that the animals exhibit fewer than wild-type numbers of such antigens on their cell surfaces; preferably, the animals • exhibit no class II antigens on their cell surfaces or exhibit so few class II antigens that they are undetectable by standard techniques of immunochemistry. Any non-human mammal which may be produced by transgenic technology is included in the invention; preferred mammals include, in addition to mice, cows, pigs, sheep, goats, rabbits, guinea pigs, hamsters, and horses.

By "MHC class II antigens" is meant any protein normally presented on the surface of immune cells which, in combination with a foreign antigen, form a complex which is recognized by T lymphocyte antigen receptors. In mice, MHC class II surface antigens include MHC class II I-A and MHC class II I-E molecules. In humans, MHC class II surface antigens include the products of the HLA-DP, HLA-DQ, and HLA-DR genes.

By "null mutation" is meant a mutation which renders a gene incapable of producing a functional protein product. As used herein, it includes mutations which destroy an MHC antigen gene's ability to express any product at all; MHC genes containing such a mutation are identified, e.g., using the immunohistochemical methods described herein. Null mutations also include those mutations which permit the production of some protein or protein fragment, but the protein or protein fragment does not function (as used herein) as an MHC antigen; MHC antigen functionality is tested, e.g., using methods described in Fundamental Immunology (ed. William E. Paul, Raven Press, New York, 1989) . A null mutation, as used herein, includes a mutation which leads to the production of a mutant protein or protein fragment which is degraded by the host cell before it is localized to the cell surface. One example of a null mutation is a functional null allele comprising a targeted alteration that interferes with the efficient expression of a functional gene product from the allele. By "cells deficient in MHC class I and MHC class II antigens" is meant cells of a mammal which are nucleated somatic cells and germ cells.

The invention provides useful model systems for MHC antigen immunodeficiencies, such as Bare Lymphocyte Syndrome (characterized by the absence of MHC class II surface antigens) . Further because MHC class II antigens are substantially absent, cells from such animals are less prone to rejection by recipient animals; the transgenic animals thus provide a universal source of donor tissue for transplant. Such tissue may be used to promote wound healing (e.g., using epithelial cells, e.g., keratinocytes) , or it may be used for whole organ transplantation. Finally, it can also be used to provide universal vehicles for somatic cell therapies (e.g., for delivery to a patient of such proteins as Factor VIII, α- galactosidase A or ornithine aminotransferase) .

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

Detailed Description The drawings will first briefly be described. Drawings

FIG. 1 illustrates gene targeting of the Ao^b gene and germ line transmission of the disrupted allele. In (A) , a schematic of the gene targeting construct used for disruption of the Ao^b gene is shown. Black boxes represent exons and restriction enzyme sites are as indicated. The neo^r gene from pMClneo-poly A was subcloned into the second exon of the A^^b gene and the tk gene from HSV-1 was positioned in the targeting vector as previously described in Johnson et al., (Science 245:1234. 1989). The probes used for Southern blot analysis are also indicated. In (B) , animals which were homozygous wild type (+/+) , heterozygous (+/-) , or homozygous for the disrupted Ao^b allele (-/-) were identified by Southern blot analysis of tail DNA. DNA was digested with EcoRI and hybridized with probe 2; this probe detected a 6.4 kb fragment of the endogenous Ao^b gene (^as seen in lane 1 with DNA from a wild type 129/SvJ mouse) and a 5.2 kb fragment of the disrupted allele.

FIG. 2 illustrates a flow cytometric analysis of MHC class II expression in the periphery of class II deficient animals. In (A) and (B) , single cell suspensions from spleens and lymph nodes of wild type

(+/+) mice were analyzed using antibodies specific for Ao^b (25-9-17, 34-5-3) arid A_α ^b (3JP, 1E9) ; control shows background staining obtained with the secondary reagent alone. In C and D, flow cytometric analysis of lymphocytes from mutant (-/-) animals was performed as described above. None of the class II antigen-specific antibodies revealed specific lymphocyte staining. Flow cytometric analysis was performed in triplicate with one wild type and one mutant animal each time; all trials yielded essentially identical results.

FIG. 3 is a tabular representation of the results of a two color flow cytometric analysis of lymphocyte subsets in wild-type and MHC class II deficient animals. The control animal for Experiment 1 was (+/+) and for Experiment 2 was (+/-) . The mutant animals were (-/-) for both Experiment 1 and Experiment 2.

FIG. 4 illustrates flow cytometric analysis of MHC class I and class II expression in the periphery of MHC- deficient animals. Single cell suspensions were prepared from spleens of control and MHC-deficient animals and 1 x 10⁶ cells were stained with hybridoma supernatants containing class I-specific (A) or class II-specific (B) antibodies, followed by a fluorescein-conjugated Ffab')₂ fragment of goat antibody to mouse IgG. MAbs were 3*-83P (H-2K^b) , 28-14-8S (H-2D^b) , 28-14-8S (H-2D^b) , 25-9-17S (I- A/3^b) , and 1E9 (I-Aα^b) . Control shows background staining with the secondary reagent alone.

FIG. 5 illustrates flow cytometric analysis of T cell subsets in the lymphoid organs of MHC-deficient animals. Single cell suspensions were prepared from thy i, spleens, and lymph nodes of control and MHC- deficient animals and analyzed as described in Fig. 4. Approximately 10,000 events were recorded for each analysis. FIG. 6 illustrates the phenotype and function B of cells in MHC-deficient animals. (A) Single cell suspensions were prepared from spleens of control and MHC-deficient animals and analyzed as described in Fig. 4. Approximately 10,000 events were recorded for each analysis. (B) TNP- specific antibody responses elicited in MHC-deficient animals following immunization with the T cell-independent antigen TNP-ficoll. Open circles indicate MHC-deficient mice and closed squares indicate results with the indicated control mouse. FIG. 7 depicts the results of MLR assays using MHC-deficient spleen cells as responder and stimulator.

(A) Proliferative response of MHC-deficient and control 129/Sv spleen cells to BALBcBy (H-2^d) stimulator cells.

(B) Proliferative response of BALBcBy spleen cells to MHC-deficient and control 129/Sv stimulator cells.

The responder populations were depleted of CD4⁺ cells where indicated.

Production of MHC Class II-Deficient Transgenic Mice Gene targeting is a powerful method by which specific genes can be altered in ES cells regardless of their expression, and subsequently passed through the germ line (Thompson et al., Cell 56:313, 1989; Johnson et al., Science 2_4_5:1234, 1989; Zijlstra et al., Nature 342:435. 1989; DeChiara et al.. Nature 345:78. 1990; McMahon et al.. Cell 62_.:1073, 1990; Soriano et al.,. Cell 64.:693, 1991; Chisaka et al. , Nature 3_50:473, 1991; Fung- Leung et al.. Cell ϋJ5:443, 1991; and Kita ura et al., Nature 350:423. 1991). We have used this technique to generate mice which are devoid of cell surface expression of MHC class II molecules by introducing a loss of function mutation into the Ao^b gene (i.e., one of the MHC class II genes) in ES-D3 cells. Disruption of the Aβ^b gene prevented the cell surface expression of I-A molecules on class II-expressing cells. The disruption in the E_α ^b gene (another of the MHC- class II genes) in the ES-D3 host cells (i.e., cells derived from mice of the H- 2^b ha lotype; Doetschman et al., J. Embryol . Exp. Morph . 81_: 27 , 1985) rendered such cells similarly devoid of I-E cell surface expression on class II-expressing cells (Mathis et al. , Proc. Natl . Acad. Sci . USA .80:273, 1983). To produce the transgenic animals of the instant invention, a 5.4 kb EcoRI-XhoI fragment of the Ao^b gene (encompassing the first four exons) was subcloned into a pBluescript SK(+) vector into which a 3.4 kb BamHI fragment of HSV-1 containing the tk gene was previously blunt cloned into the Nael site (Johnson, et al.. Science 245:1234-1236, 1989). The unique BstEII restriction enzyme site in the second exon of the Ao^b gene was changed to a Sail site with linkers, and the 1.1 kb Xhol-Sall fragment of pMClneo-poly A (Stratagene, La Jolla, CA) , containing the neo^r gene, was inserted into this site. 2 X 10⁷ ES-D3 cells (publicly available from the American Type Culture Collection, Rockville, MD; ATCC Accession No. CRL1934) in 0.8 ml phosphate buffered saline buffer were then transfected with 20 μg of this plasmid DNA (following digestion with NotI) using a Bio-Rad Gene Pulser (25 F/0.35 kV) . The entire transfection mixture was plated in ten 150 mm petri dishes in media (Dulbecco's Modified Eagle's Medium containing 20% fetal calf serum, 2 mM glutamine, 0.1 mM non-essential amino acids, and 0.1 mM 2-mercaptoethanol; Sigma, St. Louis, MO) supplemented with leukemia inhibitory factor- containing supernatants derived from CHO cell transfectants (e.g., ESGRO available from Gibco, Grand Island, NY) . Media containing 150 μg/ml Geneticin

(GIBCO, Grand Island, NY) and 2 μM gancyclovir (Syntex, Palo Alto, CA) was added 24 hr post-transfection, and drug selection continued for nine days. Individual colonies were trypsinized and expanded from 24-well to 6- well culture dishes. Genomic DNA was prepared from individual colonies and analyzed by Southern blot using the probes indicated in Fig. 1. Probe 1 is a 690 bp Bglll-EcoRI fragment which hybridizes to a 6.3 kb Ncol fragment of the endogenous A»^b gene and to a 3.8 kb fragment of the disrupted allele resulting from homologous recombination with the depicted targeting construct. Probe 2 is a 705 bp XhoI-EcoRI fragment which hybridizes to a 6.4 kb EcoRI fragment of the endogenous Ao^b gene and 5.2 kb fragment of the disrupted allele. The targeting vector (Fig. 1A) incorporates the neomycin resistance (neo^r) gene into the second exon of the Ao^b gene, contains 5.4 kb of homologous flanking sequence, and contains the Herpes simplex Virus (HSV-1) thymidine kinase (tk) gene, allowing positive-negative selection of transfectants by the method of Mansour et al. (Nature 136, 348-352, 1988). Of the 2 X 10⁷ ES-D3 cells transfected with this construct, 720 colonies were G418^r (as calculated from control plates) , while 143 colonies were resistant to both G418 and gancyclovir. Of these 143 colonies, 86 were screened by Southern blot analysis, and 4 clones were found to contain a disrupted Ao^b gene. One clone containing such a disrupted A«^b gene was injected into C57BL/6J blastocysts by standard techniques. Fifteen mice were born out of 39 embryos transferred to recipients that became pregnant. Of these 15 animals, seven males and one female were chimeric. Two males out of the three tested transmitted the ES cell genotype, one to <1% and the other to 28% of their offspring. Half of the progeny carried the disrupted A«^b allele as determined by Southern blot analysis of DNA obtained from tail biopsies. Heterozygous offspring were then mated to yield (129/Sv x C57BL/6J)F₂ animals, some of which were homozygous for the disrupted Ao^b allele (Fig. IB) . F₂ animals were produced under germ-free conditions and appeared healthy. The animals used for the experiments described herein were between 6-8 wks of age. Analysis of MHC Class II-Deficient Transgenic Mice

In order to assess the effect of this mutation on MHC class II expression, frozen sections of thymus from wild type (+/+) or mutant (-/-) animals were isolated and examined by immunohistochemistry as follows. Frozen sections (6/x) of thymus from wild type (+/+) and mutant (-/-) mice were hydrated in 0.05 M Tris-HCl (pH 7.5) for 5 min and then blocked with 3% horse serum for 15 min. Sections were then incubated with 50 _/vc-l.of purified Ao^b- specific monoclonal antibody 25-9-17 (30 μg/ml diluted in horse serum) for 60 min, followed by incubation with biotinylated horse anti-mouse IgG and avidin-horseradish peroxidase by the method of the manufacturer (Vectastain Elite ABC kit. Vector Laboratories, Burlingame, CA) . Stained sections were developed with 3,3'- diaminobenzidine as chromogen for 4 min.

A characteristic reticular pattern of staining in the cortex with a more uniform staining in the medulla was observed in wild type thymic sections. In contrast, thymic sections from animals homozygous for the disrupted Ao^b allele failed to reveal specific ≤taining either in the cortical or medullary regions. MHC class II expression in the periphery of class II deficient animals was also analyzed by flow cytometry. Single suspensions were prepared from spleens and lymph nodes of wild type (+/+) or mutant (-/-) animals, and 1.5 x 10⁶ cells were stained for 20 min at 4°C with 50/l of hybridoma supernatant containing class II specific antibodies; antibodies were 25-9-17 and 34-5-3 (A^) and 3JP and 1E9 AJ ) . Cells were washed once in FACS media (Hanks Buffered Saline Solution, 3% fetal calf serum, 0.1% NaN₃) and then incubated for 20 min at 4°C with fluorescein-conjugated F(ab')₂ fragment of goat anti- mouse IgG (γ-chain specific) (Cappel, Durham, NC) . Cells were then washed twice with FACS media, fixed for 20 min with 50 μl of 2% paraformaldehyde, and analyzed by flow cytometry on a Coulter 752 flow cyto eter (Coulter Electronics, Hialeah, FL) .

Flow cytometric analysis of lymphocytes from spleen and lymph nodes using antibodies specific for _β ^h (25-9-17, 34-5-3) and A_α ^b (3JP, 1E9) demonstrated an absence of MHC class II I-A antigens on lymphoid. cells in the periphery of animals homozygous for the disrupted A_β° alleles (Fig. 2C, D) . Although these H-2^b animals did not express class II I-E antigens due to a deletion in their E_α ^b gene (Kita ura et al., Nature 350:423-26. 1991), it was formally possible that expression of the functional E»^b and A_α ^b genes could result in the mixed isotypic molecule E_β ^bA_α ^b on the cell surface. Flow cytometric analysis of lymphoid cells using the E« -specific monoclonal antibody Y17 failed to reveal the presence of this mixed isotypic molecule. Thus, these results indicated that disruption of the Aa^b allele in these animals led to a null phenotype with respect to MHC class II antigens. Flow cytometric analysis using antibodies specific for the class I K^b molecule and the IgM demonstrated normal levels of these surface markers on cells from the spleen, thymus, and lymph nodes of class II deficient animals.

The composition of thymocyte subsets in class II deficient animals was also examined. Two color flow cytometric analysis of thymocyte subsets in mutant mice was carried out as described above using a FACScan flow cytometer (Becton-Dickinson, Lincoln Park, NJ) and fluorescein-conjugated anti-CD4 (Pharmingen, San Diego, CA) and phycoerythrin-conjugated anti-CD8 (Pharmingen) . Such an analysis revealed a striking absence of single positive CD4⁺ cells in the thymus [0.2%-0.3% vs 8.1% in homozygous (+/+) or heterozygous (+/-) control animals] (Fig. 3) . In contrast, the number of double positive CD4⁺CD8⁺ cells in mutant mice was identical to that seen in control animals. These results provided evidence that the expression of CD4 during T cell maturation in the thymus did not require the presence of I-A and I-E class II molecules, but that progression from the double positive CD4⁺CD8⁺ stage to the single positive CD4⁺ stage did.

Examination of single positive CD8⁺ T cells in the thymus and periphery of class II deficient animals showed elevated numbers of these cells relative to control littermates (Fig. 3) . It is possible that the CD8⁺ T cell population was expanded due to the deficit of CD4⁺ cells in the T cell compartment. However, β-2 icroglobulin (MHC class I)-deficient mice which lack CD8⁺ T cells do not demonstrate a comparable phenomenon, as these mice have normal levels of CD4⁺ cells (Zijlstra et al.. Nature, 344:742-746. 1990; Roller et al.. Science, 248:1227-1230, 1990). The number of single positive CD4⁺ cells in the periphery of class II deficient mice was markedly reduced in comparison to control littermates (FIG. 3). In contrast to the thymus, however, there were detectable numbers of CD4⁺ cells in the spleen and lymph nodes (l%-3%) . These cells may represent either class I restricted CD4⁺ T cells, CD4⁺ T cell receptor-bearing T cells, or CD4⁺ natural killer¬ like cells. Production and Analysis of MHC Class I and Class II Deficient Mice

An immuno-deficient strain of mice that lacks expression of both MHC class I and class II molecules was generated as follows. Mice harboring a gene disruption at the /3₂-microglobulin (j3₂m) locus (see, e.g., Zijlstra et al., Nature (London) 344:742-746. 1991, hereby incorporated by reference) were mated to animals carrying a similar mutation in the Aj9^b gene (e.g., those animals described above) . In particular an animal of each of these two strains, on a background of 129/Sv were mated and subsequently intercrossed, to produce mice homozygous for both the /3₂m and class II mutant alleles (MHC- deficient mice) . Animals homozygous for the disrupted _/3₂m allele and heterozygous for the disrupted A«^b allele were mated to mice heterozygous for the disrupted _/3₂m allele and homozygous for the disrupted Ao^b allele, thereby generating litters which include class I- deficient, class II-deficient, and MHC-deficient, as well as littermate control animals, which are singly mutant homozygotes and multiply mutant heterozygote combinations. All animals were maintained in autoclaved microisolate cages and provided with autoclaved food and water. When housed under these conditions, MHC-deficient animals and their littermates appeared healthy. BALB/cBy mice were purchased from The Jackson Laboratory (Bar

Harbor, Maine) . Southern blot analysis of DNA from mice born of an intercross between animals heterozygous for gene disruptions at both loci revealed that doubly— homozygous offspring were present at the expected frequency of 1 out of 16. The absence of MHC molecules in animals homozygous for gene disruptions at both loci was confirmed by flow cytometry performed on spleen cells using several class I and class II monoclonal antibodies (mAbs) . Said antibodies are publicly available from the American Type Culture Collection, deposit numbers HB20 (3-83P) , HB27 (28-1485) , and HB26 (25-9-175) . Flow cytometry was per ormed as described in Grusby et al. , Science 253:1417-1420. 1991. Approximately 1 x 10⁶ cells were stained with a hybridoma supernatant containing MHC class I- or class II-specific monoclonal antibodies (mAbs) . mAbs were 3-83P (H-2K^b) , 28-14-8S (H-2D^b) , and 25-9-17S (I-Ao:^b) . Cells were then washed once in Hanks balanced salt solution, 3% fetal calf serum, and 0.1% sodium azide and then incubated with a fluorescein-conjugated F(ab')₂ fragment of goat antibody to mouse IgG (r-chain- specific) . Cells were next washed two additional times, fixed in 2% paraformaldehyde, and analyzed by flow cytometry. For directly conjugated antibodies, cells were preincubated with purified anti-FcTlIR antibody (2,4G2; available from Pharmagen, San Diego, CA) for 5 min at 4°C. Fluorescein-conjugated anti-CD4 and anti- IgM, and phycoerythrin-conjugated anti-CD8 and anti-B220 were from Pharmingen (San Diego,. CA) . As shown in Fig. 4 , flow cytometry with mAbs for H-2K^b and H-2D^b revealed substantially undetectable staining on spleen cells from MHC-deficient animals. That low levels of class I heavy chain can be expressed at the cell surface in the absence of _/3₂m protein has been observed (Allen et al. , Proc. Natl . Acad. Sci . USA jJ3_: 447-7451, 1986) , and indeed, this phenomenon was also seen in the analysis of /3₂m-deficient mice (Zijlstra et al., Nature (London) 344:742-746. 1991; Bix et al. , J. Exp. Med. 176:829-834. 1992). Staining with mAbs to either Aα^b or A«^b failed to reveal specific staining. again consistent with that previously observed with class II-deficient mice (Grusby et al., Science 2J53.:1417-1420, 1990; Cosgrove et al., Cell 6:1051-1066, 1991)). Thus, mice homozygous for mutations of the ,3₂m and A«^b loci are essentially devoid of both MHC class I and class II molecules.

The phenotype of T cells subsets present in the lymphoid organs of MHC-deficient animals was examined. Flow cytometric analysis of thymocytes from MHC-deficient mice using mAbs specific for CD4 and CD8 revealed the presence of normal numbers of double-positive cells but a virtual absence of single positive CD4⁺ and CD8⁺ cells (Fig. 5) . Furthermore, these latter two populations were significantly depleted in the spleens and lymph nodes of MCH-deficient animals. The small number of single positive cells seen in these peripheral lymphoid organs (1-5%) is similar to that seen for CD8⁺ cells in _/3₂m- deficient mice (Zijlstra et al., Nature (London) 344:742- 746, 1990; Koller et al., Science 148:1227-1230, 1990) and CD4⁺ in class II-deficient animals (Grusby et al., Science 253:1417-1420, 1990; Cosgrove et al., Cell 66:1051-1066, 1991).

Although the spleens from MHC-deficient animals are essentially devoid of mature T cells defined by markers for CD4 and CD8 (Fig. 5) as well as CD3 and Aα_/3 T cell receptor, their cellarity is approximately twice that of littermate controls. When examined for the presence of cells displaying other surfaces markers, spleens from MHC-deficient animals have normal numbers of natural killer cells and τβ T cells, but slightly increased numbers of macrophages and granulocytes.

Class II-deficient mice have normal numbers of IgM⁺ B lymphocytes, but are unable to mount antibody responses to T-dependent antigens due to a lack of CD4⁺ T helper cells (Grusby et al., Science 253:1417-1420, 1990; Cosgrove et al., Cell .66.:1051-1066, 1991) as shown by the following experiments. Animals were immunized interperitoneally with 100 μg of trinitrophenol (TNP)- conjugated ficoll in phosphate buffered saline. Animals were bled 3 days prior to immunization, and 8 and 16 days post-immunization. Serum was prepared and stored at 4°C. TNP-specific antibody responses were measured by coating flat-well microtiter plates overnight at 4°C with 25 μg/ral of TNP-conjugated bovine serum albumin in Tris buffered saline (TBS) . After washing twice with TBS, wells were blocked with 2% goat serum in TBS, and then washed twice again with TBS. Serial five- old dilutions from 1:100 to 1:12500 were analyzed. Flow cytometric analysis of spleen cells from MHC-deficient animals shows that mature IgM⁺ B cells also develop normally in an environment devoid of MHC class I and class II molecules (Fig. 6A) . Furthermore, TNP-specific antibody responses of all isotypes can be elicited in MHC-deficient animals following immunization with the T-independent antigen TNP-ficoll (Fig. 6B) . Finally, B cells from MHC- deficient animals can be induced to proliferate following stimulation with the mitogen lipopolysaccharide. Thus, aside from their failure to express MHC molecules at the cell surface, the development and function of B cells in MHC-deficient animals appear normal.

The capacity of MHC-deficient cells to serve as both responder and stimulator populations in a mixed lymphocyte reaction (MLR) was also examined. Responder spleen cells (2 x 10⁵) and irradiated (2,000 rad) stimulator spleen cells (4 x 10⁵) were added to triplicate U-bottom wells in a final volume of 200 μl of RPMI 1640 supplemented with 20 mM Hepes, 2 mM L- glutamine, 0.1 mM nonessential amino acids, 50 μ,M-2- mercaptoethanol, 10% fetal calf serum, and 100 μg/ml gentarnicin. The cultures were pulsed with 1 ,Ci of [³H]thymidine (6.7 Ci/mmol; 1 Ci = 37 GBq) for the last 8-14 hr of a 3-4 day culture period. The samples were harvested onto glass fiber filters and [³H]thymidine uptake was measured by β scintillation counting (30 sec/sample) . Results are expressed as mean counts per min +/- SEM. In some experiments, spleen cells were depleted of CD4⁺ or CD8⁺ cells. Spleen cells (50 x 10⁶ cells/ml) were incubated for 30 min on ice with medium alone, or a 1:4 dilution of ascites containing mAb GK1.5 (anti-CD4) or 2.43 (anti-CD8) . Cells were washed once, resuspended at 40 x 10⁶ cells/ml, and then incubated with a 1:5 dilution of rabbit complement (C") (C-Six Diagnostic) for 30 min at 37°C. The cells were then washed twice and counted. As shown in Figure 7A, normal 129/Sv splenocytes are induced to proliferate when cultured with allogeneic stimulator cells. In contrast, splenocytes from MHC-deficient 129/Sv mice demonstrate only marginal levels of proliferation to the same allogeneic stimulator cells. This proliferation is completely abolished when the MHC-deficient responding population is pretreated with anti-CD4 mAb plus complement, suggesting that the responding population is contained within the small numbers of CD4⁺ cells found in the periphery of MHC-deficient animals. When MHC-deficient cells are used as stimulator cells in MLR (Fig. 7B) , they stimulate very low levels of proliferation of allogeneic responder cells relative to normal 129/Sv stimulator cells. This proliferation is also due to CD4⁺ cells, as pretreatment of the responder population with anti-CD4 mAb plus complement abrogates the response. This result was surprising since MHC- deficient cells do not express the class II antigens for which allogeneic CD4⁺ cells might. Use

Homologous replacement of MHC class II antigen genes makes possible the production of both animals and mammalian cells lacking MHC surface antigens; such animals and cells are useful for diagnostic and therapeutic purposes.

For example, the animals of the invention may be used as models of human disease characterized by MHC class II surface antigen deficiencies, and thus are useful to test potential therapeutics. In one particular example, the animals may be used to test compounds for those which are useful for the treatment of Bare Lymphocyte Syndrome (an immunodeficiency disease characterized by a lack of all MHC class II antigen expression resulting in an unusual susceptibility to infection and infant or early childhood mortality) .

To test for such useful therapeutics, an animal of the invention is contacted with a candidate material. Administration is by any known route, but is preferably intravenous and is preferably at a range of concentrations. Following an appropriate period of time, a sample of lymphoid cells is isolated from the animal (as described herein) and examined for the presence of MHC class II surface antigens (e.g., by the methods described herein) . A useful therapeutic is one which promotes an increase in the number of MHC class II surface antigens on any host cell (e.g., on lymphoid cells) .

The instant methods facilitate the development of other disease models as well. Because the animals described herein lack all MHC class II surface antigens, transgenic sequences encoding particular MHC class II or class I antigens which are either endogenous to the animal or heterologous (e.g., a human sequence in a murine animal) may be introduced into the animal, e.g.. by the methods described herein to create well-defined models for immune disorders which involve, e.g. , the loss of only one murine or one human class of surface antigen. In addition, availability of animals and mammalian cells lacking MHC surface antigens provides an abundant source of universal donor tissue for transplant. In one particular example, such tissue includes epithelial cells useful for the promotion of wound healing. The epithelial cells may be derived from a MHC surface antigen-deficient transgenic animal or may be derived from cultured epithelial cells (e.g., keratinocytes) which harbor inactivated MHC antigen genes (produced, e.g., by the general methods described herein; in Zijlstra et al., Nature 342:435. 1989; or in WO 91/01140, hereby incorporated by reference) . In addition, the cells may be engineered (by standard techniques) to produce and secrete a growth factor which further speeds the healing of a lesion. In another particular example, whole organs (e.g., hearts) isolated from MHC antigen-deficient transgenic animals may be transplanted into human, recipients. Donor organs are preferably isolated from larger transgenic animals such as pigs (also created by the general methods described herein) . Donor tissue is administered to the patient using standard transplantation techniques. The animal donors preferably include a null mutation in at least one MHC class II antigen gene (e.g., the murine genes: Att, A_/3, Eα, or E_/9; the human genes: HLA-DP, HLA-DQ, or HLA- DR; or the porcine genes: SLA class II) and a null mutation in at least one MHC class I antigen gene (e.g., the _/3₂-microglobulin gene from (B₂m) mice, humans, or pigs; see, e.g., Zijlstra et al., Nature 342:435, 1989 or WO 91/01140, hereby incorporated by reference), Such genes may be sequentially inactivated using a homologous replacement system (e.g., that described herein). Alternatively, animals lacking both MHC class I and MHC class II antigens may be produced by breeding animals deficient in one class of antigens with animals deficient in the other class. In one particular example, the MHC class II antigen-deficient mice described herein may be crossed with /?2-microglobulin-deficient, and thus MHC class I antigen-deficient, mice (e.g., those described in Zijlstra et al. , Nature 342:435. 1989) to produce offspring which are deficient in all MHC surface antigens as demonstrated above.

MHC antigen-deficient tissue may also be used as a general vehicle for somatic cell therapies. Any gene of interest (e.g. , the human Factor VIII gene) may be introduced into a transgenic cell of the invention such that the protein product is expressed by the cell. This is accomplished, e.g., by positioning the gene (e.g., the Factor VIII gene) downstream of a viral enhancer. General methods for insertion of genes and somatic cell expression are described in Mansour et al. (Nature 336:348. 1988) and U.S. Pat. 4,868,116, hereby incorporated by reference. Once administered (e.g., intravenously) , the cells supply the recombinant protein product to the patient. Because the transgenic cells lack MHC antigens, the problems normally associated with immune rejection are circumvented, and useful cells may therefore be derived from any organism. Alternatively, the gene targeting methods described herein may be used to specifically inactivate both alleles of the MHC class II genes (e.g. , using methods described herein combined with those described in Riele et al.. Nature 348:649 _r

1990) and, if desired, the MHC class I proteins (e.g., by inactivating the ,92-microglobulin gene) of the patient's own cells (e.g., the patient's isolated lymphocytes). The engineered cells can then be used as host cells for the production of a therapeutic protein (as described above) .

Finally, the MHC class II antigen-deficient animals of this invention can be used to examine the role of the class II antigens in any of several complicated immunological processes characterized by interaction between antigen presenting cells and T cells. The animals may also be used generally to examine the absence of class II expression .in vivo. This is not possible to do in humans, e.g., in BLS patients, since they receive bone marrow transplants at an early age.

We claim:

Claims

Claims 1. A transgenic non-human mammal, wherein the surfaces of cells of said mammal are MHC-deficient in an MHC class I antigen and class II antigens.

2. The mammal of claim 1, wherein said cell surfaces are deficient in all MHC antigens.

3. The mammal of claim 1, wherein said mammal is a transgenic mouse, the chromosomes of said mouse including null mutations in the Ajβ^b, E_α ^b and β2- microglobulin genes.

4. The mouse of claim 3, wherein said cell surfaces are deficient in all MHC surface antigens.

5. A method of testing a substance for efficacy in the treatment of an MHC antigen deficiency, said method comprising exposing a mammal of claim 1 or a mouse of claim 4 to said substance and then examining said mammal's or said mouse's cell surfaces for the presence of said MHC antigen, the presence of said antigen being an indication that said substance is useful for the treatment of said deficiency.

6. Use of cells lacking MHC class II antigens in the preparation of a medicament for wound healing.

7. Use of donor tissue isolated from a mammal deficient in MHC class II antigens for the preparation of a medicament for treating a human condition requiring tissue transplantation.

8. The use of claims 6 or 7, wherein said cells or said donor tissue are also deficient in an MHC class I antigen.

9. The use of claims 6 or 7, wherein said cells or the cells of said donor tissue further include a gene which expresses a recombinant protein.

10. The use of claim 9, wherein said recombinant protein is Factor VIII.

11. A mammalian cell whose cell surface is deficient in all MHC class II antigens and in an MHC class I antigen.

12. The mammalian cell of claim 11, wherein said cell expresses a recombinant protein.

13. The mammalian cell of claim 12, wherein said recombinant protein is Factor VIII.

14. A transgenic nonhuman animal or stem cell comprising a diploid genome comprising a 32-microglobulin gene null allele and a MHC class II null allele.

15. A transgenic nonhuman animal or stem cell according to claim 14, wherein the nonhuman animal or stem cell is homozygous for the MHC class II null allele.

16. A transgenic nonhuman animal or stem cell according to claim 15, wherein the nonhuman animal on stem cell is also homozygous for the MHC class I null allele.

17. A transgenic nonhuman animal or stem cell according to claim 16, wherein the MHC class II null allele is a disrupted A_β ^h allele.