Immunological tolerance is a state of unresponsiveness of the immune system which is specific for a particular antigen. An important aspect of tolerance is self-tolerance, which is an unreactiveness to self that allows the body to distinguish between self antigens presented to it on normal tissues and potentially dangerous situations such as infection. It is a generally held view that the immune system is unreactive to self antigens, but maintains a broad repertoire of receptors sufficient to recognise any non-self antigen presented to it (central tolerance). It is also generally accepted that there must be some fail-safe mechanisms that operate to minimise the effects of any mistakes that may lead to autoreactivity (peripheral tolerance). The possible mechanisms for tolerance, in particular for peripheral tolerance, are reviewed in Cobbold et al. 1996 Immunol. Rev. 149: 5-33.

Tolerance is clearly a matter of practical importance in treatments which require the introduction of foreign antigens into the body, such as tissue transplantation and vaccination, and in diseases which are characterised by immunosuppression such as AIDS or by a breakdown in self-tolerance resulting in autoimmunity. However, the techniques for studying tolerance and for investigating the effects of pharmaceutical compounds on tolerance have until now been limited.

Until relatively recently, neonatal tolerance induction was the primary technique for experimental study of tolerance. Specific tolerance to grafted skin was induced in mice by neonatal injection of spleen celis from a different strain. The resulting mice showed tolerance to skin grafts from the donor strain. Moving on from the neonatal induction models, transgenic methods for introducing foreign transgenes into mice have allowed the study of specific antigens in a defined genetic background which are treated as self by the immune system.

Techniques for inducing tolerance in the mature adult immune system have also been developed. Monoclonal antibodies against the murine CD4 and CD8 antigens are a powerful means to produce tolerance to a variety of different antigens in the adult mouse. Depletion of the CD4 or CD8 T cells is not necessary; a blockading with either (Fab') 2 fragments or a non-depleting isotype of anti-CD4 or anti-CD8 antibody is especially effective. Combinations of CD4 as well as CD8 antibodies can be used for the extended periods found necessary for inducing tolerance in some of the more difficult skin graft systems (Cobbold et al, 1990 Eur. J.

Immunol. 20: 2747-2755).

The discovery that tolerance could be induced to transplanted organs under the cover of non-depleting CD4 and CD8 antibodies given together was a major breakthrough (e. g. Cobbold et al. 1990). Because skin grafts can be used as both the tolerogen and later test challenge, the system gives a consistent read-out for rejection/tolerance in vivo that does not depend on making assumptions about the effector cells or the antigens, as is necessary for in vitro measurements. The main advantage of this approach over that of neonatal tolerance induction, or the introduction of foreign transgenes, is that it makes it possible to focus on events that take place without the thymus (by removing it) and without other previous exposure to the antigen during the development of the immune system. This allows more precise control of the timing of

tolerance induction and antigen challenge. There is also a wider choice of how the antigen is presented for tolerance by using different organs as grafts.

However, in all of the techniques described, there is a limited amount of specific information that can be gained relating to the mechanisms for rejection and tolerance, because of the complexity of the immune system and the immune response.

It is therefore an object of the invention to provide a simplified system for observing an immune response and/or tolerance to a particular antigen or to a limited number of antigens.

It is another object of the invention to provide a system for testing the effects of treatments on an immune response and/or tolerance to a particular antigen or to a limited number of antigens.

Surprisingly, it has been found that mice having only CD4 positive T cells, all of which have a single T cell receptor specific for the same transplantation antigen, can reject skin grafts. Previously, it had been thought that the role of CD4 positive T cell in graft rejection was mainly to provide T cell help to either CD8 positive T cells to become cytotoxic, or to B cells to make antibody. Thus, CD8 positive T celi receptor transgenic mice have been made (Hammerling et al. 1993 Immunol. Rev. 133: 93-104). Furthermore, it was not expected that such a CD4 T cell receptor"monoclonal"mouse once it had been found to reject grafts, could be tolerised to the transplantation antigen. Previously, it was often assumed that tolerance works either by eliminating the specific T cells (clonal deletion) or by inducing a separate population of suppressor cells. In fact, tolerance can be effectively induced by treatment with non- depleting anti-CD4 antibody.

The invention provides in one aspect a genetically modified non-human mammal having a population of CD4 positive T cells specific for one or a limited number of selected antigens, including at least one

transplantation antigen, the mammal being capable of an immune response against the transplantation antigen and of being tolerised to the transplantation antigen.

Preferably, the genetically modified animal according to the invention is capable of rejecting a tissue transplant containing the transplantation antigen and if applicable the other selected antigens.

A genetically modified animal according to the invention has T cell receptor genes which encode a T cell receptor specific for the transplantation antigen. The T cell receptor genes will generally be stably integrated into its genome. Most T cell receptors are made up of a and P chains which are encoded by separate genes. The T cell receptor genes <BR> <BR> <BR> <BR> employed in the invention may be derived for example from cloned a and ß chain genes for T cell receptors specific for the transplantation antigen, or from cc and ß chain genes prepared synthetically using sequence information from cloned genes. In another alternative, the genes are derived from a and ß chain genes for T cell receptors which are not necessarily specific for the transplantation antigen. In that case, well known methods such as site-directed mutagenesis may be used to alter the gene sequences to achieve transplantation antigen specificity.

Preferably, the genetically modified animal according to the invention lacks a normal population of CD8 positive T cells, or B cells, or both. More preferably, the animal has no functional CD8 positive T cells or B cells. An absence of lymphocytes other than CD4 positive T cells provides a far simpler system for studying the activity of the immune system in relation to a selected transplantation antigen.

An absence of CD8 positive T cells and B cells may be as a result of the animal having a deficiency in lymphocyte receptor recombination such that no CD8 T cell receptors or B cell receptors (antibody) are expressed. For example, mice which are deficient in RAG-1 and/or RAG-2 (recombinase activating gene no. 1 and/or no. 2) activity

have an inability to initiate VDJ recombination in the lymphocyte receptor gene and therefore cannot generate mature lymphocytes. RAG mutations can therefore be used in the invention to eliminate functional lymphocytes other than the specific CD4 lymphocytes directed to the transplantation antigen (and specific CD4 lymphocytes directed to any other selected antigens). RAG-1-'-mice are described in Alt et al. 1992 Ann. N. Y. Acad.

Sci. 651: 277-94 and Mombaerts et al. 1992 Cell 68 (5): 869-77.

The population of CD4 positive T cells which are specific for the transplantation antigen (or if appropriate, other selected antigens), will generally form the majority of the entire CD4 positive T cell population of the animal, for example 50-60% or more of the CD4 positive T cells. In the case of a recombination deficiency such as a RAG-deficiency, all functional CD4 positive T cells will be specific for the transplantation antigen (or other selected antigens).

The transplantation antigen to which the CD4 positive T cells are directed may be any transplantation antigen which is not expressed in the animal itself, or not expressed in the animal in a form recognisable by the CD4 positive T cells. A transplantation antigen is defined as a tissue antigen which can be introduced into the body in a tissue transplant in such a manner that it is recognised by the immune system. A tissue antigen is an antigen found on body tissues and organs. Tissue transplants for the purposes of the invention include foetuses in the maternal environment.

In a particular species there are characteristic transplantation antigens, each of which may be present or absent in any given individual, or any given group of individuals such as a strain, of the species. A transplantation antigen may occur in a variety of different forms such that it is recognisable as being foreign between different individuals or between different groups of individuals in the species. Transplantation antigens include major transplantation antigens such as the MHC (major histocompatability complex) antigens, or minor transplantation antigens

such as for example the murine male transplantation antigen H-Y, the murine ß2 microglobulin antigen H-1, or mitochondrial antigens.

For certain transplantation antigens, in particular some of the minor transplantation antigens, CD4 positive T cells specific for a single antigen may not be sufficient for rejection of a tissue graft. In the case of these minor transplantation antigens, further CD4 positive T cells having specificity for one or a few further antigens are provided, for example by stably integrating further specific T cell receptor (TCR) genes into the animal. The murine male transplantation antigen H-Y is an example of a minor transplantation antigen which, conveniently, is sufficient on its own for graft rejection in an animal which has only H-Y antigen specific TCRs.

In another aspect, the invention provides a method of screening for biologically active compounds, which method comprises administering the compounds to a genetically modified non-human mammal as described herein and observing the effect of the compounds on the ability of the mammal to reject or maintain a transplant containing the transplantation antigen.

In the case of a simple male/female transplantation antigen such as the male H-Y antigen, the method according to the invention can employ a female recipient mouse and graft tissue from a congenic male donor mouse. Where a male or female transplantation antigen is not used, the recipient and donor animals are preferably from congenic strains with the donor strain being transgenic for the transplantation antigen. This means that the system is kept as simple as possible with the only antigenic difference between the donor and recipient being the transplantation antigen (and other selected antigens if applicable).

The invention enables any specific changes in the T cell population, such as activation markers or cytokine production, to be easily monitored by conventional immunological techniques during transplantation or during tolerance induction or maintenance, and also

during and after pregnancy. A so-called"Toleramouse"is thus an ideal tool for investigating the mechanisms of immunosuppression and tolerance induction. Previously, such experiments had to be performed either in normal inbred or congenic mice in which the frequency of T cells against any single antigen is too low for direct observation using simple immunological techniques such as immunofluorescent staining or cytokine staining assays.

The study of tolerance mechanisms in pregnancy may be performed on pregnant animals, in particular pregnant animals as described herein near to and after the end of pregnancy term. Particularly useful are pregnant females carrying one or more male foetuses displaying a male transplantation antigen, where the pregnant female has TCRs directed against the male transplantation antigen. Pregnant animals carrying male and female foetuses can be compared. The studies performed usefully include looking at the T cell population and T cell markers, including cytokine profiles.

It is envisaged that the invention will be particularly useful for testing potential or existing therapeutic agents: i) for wanted or unwanted tolerogenic or immunosuppressive (side) effects; ii) for interference with the tolerance process, either once it has been established or during the induction of tolerance by appropriate therapeutic agents such as CD4 monoclonal antibody.

More specifically, the invention may be used to screen for agents which: i) have tolerising effects in a similar manner to CD4 monoclonal antibodies; ii) do not interact with tolerance-inducing drugs; iii) enhance the immune response in diseases which feature immunosuppression such as AIDS and cancer;

iv) interact in desired ways with other drugs or with vaccines.

The invention is now further described in the following examples, which are not intended to limit the scope of the invention in any way.

EXAMPLES Example 1-Construction of Transqenic Mice Methods Mice: CBA/Ca (Harlan/Olac, UK) mice were bred under specific pathogen free conditions and all experimental mice were maintained in the animal facility of the Sir William Dunn School of Pathology, Oxford, in a filtered cage system (Maximiser, Thorens Caging inc., Hazelton PA, USA).

RAG-1-'-mice that had been bred onto an H-2k background were obtained from Dr. B. Stockinger (NIMR, London, UK).

Generation of A1 (M) transgenic mice: To generate transgenic mice we used the TCR a and ß chain from the A1 CD4+ T cell clone isolated from CBA/Ca mice (Tonomari 1985. Ce// Immunol. 96: 147-62). TheA1 clone recognizes the minor histocompatibility antigen H-Y, present in male but absent in female mice, in the context of H2-Ek. The TCR expression was identified using primers specific for the Va and Vp gene families (Casanova et al. 1991. J. Exp.

Med. 174 (6): 1371-83), cloned and sequenced to check for productive rearrangement. The a chain was found to be encoded by LVa10-Ja30-Ca (EMBL accession no. AJ000157) and the ß chain is encoded by LVp5.1- Vß8.2-Dß2-Jß2.3-cep2 (EMBL accession no. AJ000158) (Figure 1). EcoRl- EcoRl fragments containing the productively rearranged a or ß chains were generated by RT-PCR. The oligonucleotides used for the amplification of the a chain were:

GCGAATCACAAGCACCATGAAGAGGCTG [SEQ ID NO: 1] and GCGAATTCCAGAC CTCAACTGGACCACAG [SEQ ID NO: 2]. The oligonucleotides used for the amplification of the ß chain were: GCGAATTCAGAGGAAGCATGTCTAACACT [SEQ ID NO: 3] and GCGAATTCAGGATGCATAAAAGT TTGTCTCAGG [SEQ ID NO: 4]. The full length cDNAs were cloned into the human CD2 minigene (VA) in pBluescript. Sall-Xbal fragments were used for microinjections into CBA/Ca oocytes. Transgenic A1 (M) founders were maintained on the CBA/Ca background and bred to homozygosity. Southern blot analysis indicated that the A1 (M) line carries a single copy per haplotype of each of the transgenic Va and V chains.

Skin grafting Pieces of tail skin approx. 0.5cm2 were grafted onto the lateral thoracic wall of anaesthetised recipient mice as previously described (Cobbold and Waldmann 1986. Transplantation 41: 634-639; and Qin et al.

1990. Eur. J. Immunol. 20: 2737-2745). Where two grafts were given simultaneously these were placed side by side in the same prepared graft bed. Plaster casts were removed on day 7 and the grafts observed daily, with rejection being defined as the day when no viable graft tissue could be seen. Statistical significance was determined using the LogRank method (Peto et al. 1977. Br. J. Cancer 35: 1-39). All procedures were carried out in accordance with the UK Home Office Animals (Scientific Procedures) Act of 1986.

Immunofluorescent analysis and antibodies Thymus, spleen or lymph nodes were removed, and erythrocytes lysed by isotonic shock. Cells were labelled in phosphate buffered saline (PBS) containing 0.1% (w/v) NaN3,1% (w/v) bovine serum albumin (BSA), and 5% (v/v) heat inactivated normal rabbit serum (HINRS:

to block Fc receptors) at 4°C. Antibodies used were: CD3e (KT3-FITC), V38 (KJ16-FITC), Vß8.2 (F23.2-FITC), CD4-PE (Sigma P2942), CD8a-QR (Sigma R3762), B220-QR (Sigma R4262), CD25 (PC61-biotin), CD44-QR (Sigma R5638), SA-APC (Pharmingen 13049A), and FITC goat anti-mouse IgG (Sigma F0257). After labelling and washing, cells were fixed in 1% formalin and stored in the dark at 4°C. Four colour analysis was performed using a FACSort (Becton Dickinson, Oxford, UK) with dual laser (488nm and 633nm) excitation together with data aquisition and cross-beam colour compensation using CeIlQuest 3.1 software. At least 50,000 events were stored in list mode for further analysis and gating on forward and side scatters.

Intracytoplasmic cytokine staining was performed using spleen cells given a 2 hr stimulation in vitro with 50 ng/ml PMA (Sigma, P8139) + 500 ng/ml ionomicin (Sigma 10634) in phenol red free RPMI 1640 medium + 10% fetal calf serum at 37°C) with the addition of 10ng/ml Brefeldin A (Sigma, B7651) for a further 2 hr (Ferrick et al. 1995. Nature 373: 255- 257). After washing, cells were fixed in 2% v/v formaldehyde in PBS (20min, 4°C), washed, and permeabilized with PBS + 0.5% saponin (Sigma S-2149). The following antibody conjugates were added in saponin buffer for 30 mins at 4°C, followed by extensive washing in saponin buffer followed by PBS + 0.1% azide + 1% BSA, + 5% HINRS: anti-IL-2 (S4B6- FITC; Pharmingen 18004A), anti-IL4 (11 B11-FITC; Pharmingen 18194A), anti-IFN-y (XMG1. 2-FITC; Pharmingen 18114A). Cells were finally labelled with antibodies to surface CD4 and CD44, fixed in 1% formalin, and analysed on a FACSort as above. The conditions of stimulation, staining and analysis were such that normal CBA/Ca CD4+ spleen cells were essentially negative for all cytokine stains.

Treatment with CD4 monoclonal antibody The non-depleting rat IgG2a anti-mouse CD4 (YTS 177.9 Qin et al. 1990) was made by growing the hybridoma in a hollow fibre bioreactor, and was purified under sterile and low endotoxin conditions by precipitation with 50% saturated ammonium sulfate. (These are standard techniques known to a person skilled in the art. An example of the technique is also available on http://www. molbiol. ox. ac. uk/www/pathology/tig/mprod. html). Grafted A1 (M) xRAG'-mice were given 5 x 1 mg intraperitoneally from the day of grafting over a two week period.

Results and Discussion Analysis of A1 (M) mice transgenic for TCR against H-Y+H2-Ek.

Thymus, spleens, and lymph nodes from A1 (M) mice were analysed by 3 colour immunofluorescence to determine whether the expression of the transgenic TCR led to the predicted functional modification of the T cell repertoire. The thymi of female A1 (M) mice were found to have a strong bias towards the generation of CD4+CD8-rather than CD8+CD4-T cells, as expected from an increased positive selection of the MHC- ! ! restricted anti-H-Y TCR. This led to a CD4/CD8 ratio in the peripheral lymphoid organs in excess of 10: 1, and expression of the VD8.2 transgenic receptor on more than 90% of CD3+ cells. In contrast, male A1 (M) mice had smaller thymi (a mean of 6X106 total thymocytes compared to 8x10'in females), with a mature CD4/CD8 ratio close to 1: 1, and a similar expression of Vd8.2 to non-transgenic CBA/Ca mice, suggesting clonal deletion of anti-H-Y transgenic T cells and escape of endogenous TCR rearrangements. These A1 (M) mice were then crossed onto a RAG- 1-'-background, to eliminate all B cells and T cells expressing other TCR molecules encoded by endogenous TCR rearrangements, so that any

ability of H-Y specific CD4+T cells to reject male skin grafts could be unambiguously identified.

Positive selection in female and negative selection in male A1 (M) xRAG-1-'- mice Immunofluorescent staining of A1 (M) xRAG-'-mice confirmed that the anti-H-Y TCR was functional, as positive selection and the generation of CD3+CD4+CD8-thymocytes was only observed in female <BR> <BR> <BR> <BR> thymi, while male thymi were much smaller, with very few CD4+CD8-cells.<BR> <BR> <BR> <BR> <BR> <BR> <P>When we looked in more detail at these few CD4+CD8-cells we found them present in similar numbers in both male A1xRAG-1-'-and RAG-1-'-controls, <BR> <BR> <BR> <BR> and that none of them expressed CD3 but were mostly CD11 c+, suggesting they may be related to CD4+ immature dendritic cells (Winkel et al. 1994 Immunol. Lett. 40: 93-9) rather than T celis that had somehow escaped <BR> <BR> <BR> <BR> deletion. Staining of lymph-nodes confirmed that only CD4+ and not CD8+ T cells were present in female A1 (M) xRAG-1-'-, and that clonal deletion in the male reduced the number of CD4+ cells down to that seen in RAG-1-'- <BR> <BR> <BR> <BR> <BR> mice (and these were again CD3-CD11c+). The expression of the TCR in<BR> <BR> <BR> <BR> <BR> <BR> <BR> female A1 (M) xRAG-1-'-, as measured by CD3 or Vß8.2 staining, was lower<BR> <BR> <BR> <BR> <BR> <BR> <BR> than in a normal CBA/Ca mouse (approx. 30% of the median fluorescence level), but similar to that of the A1 (M) founders, which may be a property of the CD2 expression system (G. Stockinger, personal communication).

Rejection of male skin by female A1 (M) xRAG-'-mice In initial experiments in two laboratories, a total of 8 female A1 (M) xRAG-1-'-mice were given single male skin grafts, four of which were rapidly rejected (within 16 days) and a further two were eventually rejected in a chronic fashion. Subsequently, a group of 5 female A1 (M) xRAG-1-'- mice were simultaneously grafted with male and female CBA/Ca skin in the same graft bed. All of the male grafts were rapidly rejected (within 14

days), while the female grafts remained perfectly healthy (Figure 2). A second group of five was grafted in an identical fashion but were also treated with saturating amounts of a monoclonal antibody that blocks CD4 function in vivo. All these grafts were accepted, proving that the rejection was both CD4 dependent and male specific. As illustrated in Table 1, non- depleting CD4 monoclonal antibodies produce an indefinite tolerant state (ability to accept fresh male skin grafts) even well after the antibodies are no longer in circulation.

Mechanism of CD4 dependent graft rejection The A1 (M) xRAG-1-'-mice should have no CD8+ T cells or antibody producing B cells that might be able to act as effectors of graft rejection, and this was checked by staining spleen cells from two female A1 (M) xRAG-1-'-mice that had been allowed to reject two sequential male grafts, and had been grafted with a third male skin 7 days previously, such that if there was any hypothetical expansion of, for example, a novel CD8+ population during graft rejection, this should become visible. However, it was confirmed that there was no CD3+CD8+ staining above background (Figure 3), and that CD25 expression was limited to the CD3+CD4+ subset.

Even if there had been some CD8+TCR+ cells that remained below the level of detection, these should have been unable to interact effectively with the H-2Ek presented male antigen, and would thus be extremely unlikely to contribute to graft rejection. Similarly, there were no surface tg" B cells that might have been able to contribute an antibody response.

There was clear evidence that the male graft was indeed being recognized by the transgenic anti-male TCR on CD4+ T cells, as the majority (approx.

70%) of these could be shown to be recently memory or recently activated cells (expressing CD44) as well as producing both IFN-y, and to a lesser extent, IL-4 (Figure 3). Interestingly, all T cells (both CD44+ and CD44-) appeared to be expressing IL-2 by this method of intracytoplasmic staining.

It is therefore clear that the transgenic TCR+CD4+ T cells in A1 (M) xRAG-1-'-females are sufficient to reject male skin. The mechanism by which the animals are made tolerant by non-depleting CD4 antibodies is neither deletion. nor a conversion of Thl to Th2 cells. The tolerisation may be due to a direct, CD4+ T cell mediated cytotoxicity or it may be due to an indirect, perhaps cytokine mediated, help for macrophages or another antigen non-specific effector cell. Recent data suggesting that neither Fas/FasL nor perforin (Selvaggi et al. 1996 Transplantation. 62 (12) : 1912- 5) are required for CD4 mediated rejection of MHC-I disparate skin would tend to favour the latter hypothesis. The MHC-I restricted CD8+ arm of the immune response in normal mice would therefore seem to be mostly dependent on (Antoniou et at. 1996 Eur. J. Immunol. 26 (5): 1094-1020), and an amplification of, the MHC-II restricted CD4+ response that we have shown here can itself provide all the necessary T cell functionality for graft rejection. <BR> <BR> <BR> <BR> <BR> <BR> <BR> <BR> <P> Example 2-Testing of reagents for mmunosuppressive or<BR> <BR> <BR> <BR> <BR> <BR> <BR> <BR> Toleroqenic effects in A1 (M) x RAG-'mice A variety of different agents was tested. The testing methods used were as follows and results are given in Table 1. <BR> <BR> <BR> <BR> a) Female A1xRAG-1-/-mice were given 10'mate A1xRAG-1-/-bone marrow or spleen cells intravenously. They were then grafted with male tail skin after a further 6 weeks to test for tolerance, which was accepted indefinitely. Immunofluorescent analysis of peripheral blood and spleen cells indicated substantial deletion of the CD4+ T cells in the tolerant mice, but not in non tolerant controls that had been given female marrow or spleen cells. b) Female A1xRAG-1-/-mice were given 5 x 1mg or a single injection of 1mg of non-depleting rat IgG2a anti mouse CD4 (YTS 177.9 Qin et al 1990 Eur. J Immunol 20: 2737-2745) at the time of grafting

male CBA/Ca tail skin. These grafts and subsequent male skin grafted at 42 and 84 days were accepted indefinitely. Analysis of blood, lymph-nodes and spleen cells from these mice demonstrated that CD4+ T cells had not been clonally deleted, and similar proportions of Th1 and Th2 cells could be identified by immunofluorescent staining for cytokines in both the tolerant mice and control animals that had not been treated with CD4 antibody but had received equivalent male skin grafts. c) Female A1xRAG-1-/-mice were given 5 x 1mg of monoclonal antibody to CD25 (PC61: Osawa et al., Immunol. lett. 1989 20: 205- 12) from the time of grafting with male CBA/Ca tail skin. d) Female A1xRAG-1-/-mice were given 5 x 1 mg of monoclonal antibody to CD40 ligand (MR1: Larsen et al. Transplantation 1996 61: 4-9) from the time of grafting with male CBA/Ca tail skin. e) Female A1xRAG-1-/-mice were given 5 x 1 mg of monoclonal antibody to CD28 (37.51: Sperling et al., J. Immunol. 1993 151: 6043-50) from the time of grafting with male CBA/Ca tail skin.

Table 1 Examples of testing reagents for immunosuppressive or tolerogenic effects in the Toleramouse Agent tested Source No. Effect Mechanism Comments Tested Male antigen iv. Bone marrow 6 Full tolerance Predominantly Model for blood to male skin clonal deletion transfusion effect Male antigen iv. Spleen cells 6 Full tolerance Predominantly Model for biood to male skin clonal deletion transfusion effect Non-depleting CD4 YTS 177.9 6 Full tolerance Not deletion and Accepted 2nd and monoclonal antibody to male skin not Th1-> Th2 3rd male skin grafts Anti-CD25 PC61 6 Immuno-Not done Rejection delayed monoclonal antibody suppression Anti-CD40 ligand MR1 6 None Not applicable Dose regimen may monoclonal antibody not have been optimal Anti-CD28 37.51 3 None Not applicable Antibody preparation monoclonal antibody may not have been active Example 3-Construction of A1 (M) xRAG-1-'-mouse from sequence information only Making TCR transgenic mice is a routine procedure where the appropriate TCR alpha and beta chain encoding DNA is micro-injected into fertilized eggs from the strain of mice in which it is desired to express the TCR (for example the standard Olac CBA/Ca strain). The injected eggs are then transferred into foster mothers, and the offspring typed by standard methods (usually by PCR typing of tail DNA) to select those which have stably incorporated the DNA into their genome. These are further checked to ensure the DNA is expressed as a functional TCR, using standard immunological techniques (staining with antibodies, T cell proliferation to antigen etc.). These A1 (M) transgenic mice are then crossed with RAG-1-'-deficient mice (Alt et al. 1992 and Mombaerts et al.

1992), and the offspring are backcrossed onto the A1 (M) parents for a

number of generations, selecting those that carry both the TCR and RAG- 1-'-genes.

The A1 TCR DNA can be made in a number of ways; including the following: 1) By cloning the TCR alpha and beta chain mRNAs as cDNA (standard method) from an original A1 CD4+ T cell clone such as the A1 CD4+ anti-H-Y T cell clone.

2) By cloning the TCR alpha and beta chain mRNAs as cDNA (standard method) from the A1 (M) or A1 (M) xRAG-1-'-female mice or CD4+ anti-H-Y T cell clones derived from them.

3) By using the DNA sequence information to sythesize the TCR alpha and beta chain genes. This can simply be performed by making a series of overlapping oligonucleotides of convenient length (eg. 30-50 bases each) such that the entire sequence on both strands is covered.

Simple annealing and ligation then generates the full length, double stranded DNA for each gene. This method requires only suitable TCR sequence information (for example the sequence in figure 1) to generate the A1 (M) mice, using techniques well known in the art of making transgenic mice.

4) By using the DNA sequence information to mutate (standard methods of site-directed mutagenesis) other (relate) TCR alpha and beta chain genes to a known desired sequence (either the entire sequence or just the complementarity determining regions (CDRs) that define the antigen binding site) such as the TCR gene sequences shown in figure 1.

Once the TCR alpha and beta chain genes have either been cloned or synthesized they are ligated into an appropriate expression vector or cassette that allows them to be expressed in T cells. The CD2

minigene cassette system described in Zhumabekov et al. 1995 J.

Immunol. Methods 185 (1): 133-40 is a suitable expression system.

Leqends to Fiqures Figure 1-A1 TCR a and P chains and nucleic acid sequences encoding them.

. Figure 2-Female A1 (M) xRAG-1-'-mice show CD4 dependent, specific rejection of male skin Female A1 (M) xRAGw mice were grafted with male and female CBA/Ca skin in the same graft bed. Survival plots are shown for the male skin that rejected with a median survival time (MST) of 14 days (w; n=5) compared to the accompanying female grafts that survived beyond day 30 (O; n=5). The P value for statistical significance was <0.003 (LogRank method). Also shown is the survival of male skin on similarly grafted A1 (M) xRAG-1-'-female mice treated with 5 x 1 mg non- depeting CD4 antibody (M; n=5; MST >30 days).

Figure 3-Phenotypic and functional immunofluorescent analysis of rejecting A1 (M) xRAG-'-mice Two female A1 (M) xRAG-1-'-mice, that had rejected two sequential male skin grafts were given a third graft, and their spleen cells stained for a number of surface and intracytopiasmic antigens.

Representative examples of four colour immunofluorescent analysis from one of the mice are shown. All samples were live gated on forward and side scatters, and the dot plot in the upper left panel shows CD4-PE versus CD8a-QR staining of the CD3-FITC positive lymphocytes. The upper right panel shows that there were no B cells expressing surface Ig in the A1 (M) xRAG-1-'- (filled histogram) compared to an A1 (M) control (broken

line histogram). The centre left panel shows the staining for CD44-QR of A1 (M) xRAG-1--lymphocytes, that was used as the basis for gating the remaining anti-cytokine stains (rat IgG1 anti-IFN-y, centre right; rat IgG1 anti-IL-2, lower left; rat IgG1 anti-IL-4, lower right), where the CD44+ cells are shown as filled histograms, the CD44-cells as open histograms, and the negative control histogram (based on the background staining of isotype matched, rat IgG1 anti-IL4-FITC in normal mice) is shown with a broken line.