International Immunology, Vol. 4, No. 1, pp.
59-65
© 1992 Oxford University Press
Tolerance and MHC restriction in transgenic mice expressing a MHC class I gene in erythroid cells Helen Yeoman1 and Andrew L. Mellor National Insitute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK
Abstract Transgenic mice carrying a MHC class I structural gene (H-2Kb) linked to transcrlptional control elements from the human 0-globln gene, which direct erythroid lineage specific transcription, express H-2K" molecules In red blood cells but H-2K" expression cannot be detected in skin or lymphold cells. This limited pattern of self MHC expression Is sufficient to Induce tolerance to H-2Kb molecules and H-2K" restricted cytotoxic T cell responses can be generated in transgenic mice. Transgenic mice are unable to mount H-2K" specific cytotoxic T cell responses In vitro, even when exogenous IL-2 is provided. However, H-2Kb specific T cell proliferative responses are comparable with H-2Kb specific responses In non-transgenic mice, even In the absence of exogenous IL-2. Thus, expression of H-2Kb molecules under control of human 0-globln transcriptlonal control elements In transgenic mice Is tolerogenlc but does not result in elimination of all H-2K" reactive T cells from the mature repertoire. This suggests that tolerance In these mice may arise due to functional Inactivation of H-2Kb reactive T cells In vivo when they encounter H-2Kb molecules expressed on cells of erythroid cell lineages or on non-erythroid cells which express H-2Kb molecules at very low levels or in a developmentally regulated pattern. Furthermore, In spite of the failure to detect H-2Kb expression on non-erythrold cells In these mice, we conclude that H-2K" molecules participate In positive selection of the T cell repertoire since H-2Kb restricted T cell responses can be generated in these transgenic mice. Introduction During development in the thymus, T cells undergo two apparently paradoxical selection processes that determine the repertoire of peripheral T cells. Firstly, there is a selection of T cells whose receptors recognize self MHC molecules, so that mature T cells recognize foreign peptides only in association with self MHC molecules (positive selection which results in MHC restriction). Secondly, T cells that recognize self MHC molecules in the absence of foreign peptide are selected against (negative selection), resulting in a mature repertoire that is tolerant to self (1). Studies with chimeric mice indicate that positive selection is a function of the MHC haplotype of radioresistant thymic epithelial cells rather than that of the reconstituting bone marrow (2). Experiments with MHC class II transgenic mice that show a limited transgene expression indicate that cells of the thymus cortex, but not medulla, are able to positively select developing T cells (3). In contrast, tolerance induction appears to be mediated by cells of bone marrow origin (2) and can be a function of
'professional' antigen presenting cells (APCs) (4). The major mechanism of tolerance induction is the physical elimination of potentially autoreactive T cells. For example, T cell clones expressing receptors with certain Vfl chains are deleted intrathymically in mouse strains that express the corresponding ligand (5). Although the question of which cell types are able to affect deletion is not fully resolved, APCs, T cells and thymic medullary cells can cause clonal elimination of some thymocytes (1,6,7). However, clonal deletion is not the only mechanism of tolerance induction; antigen expression on cells of the thymic medulla results in specific tolerance in the periphery without physical elimination of potentially autoreactive clones which, instead, are rendered specifically unresponsive (anergic) to self antigen in the periphery (8). We are interested in defining cell lineages able to mediate positive and negative selection and the approach we have utilized is to put novel MHC class I genes under the control of specific
Correspondence to: A. L. Mellor Transmitting editor: E. Simpson
Received 11 July 1991, accepted 7 October 1991
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
Key words: MHC restriction, transgenic mice, tolerance, erythroid cells
60
T cell function in MHC class I transgenic mice
transcnptional regulators and introduce these recombinant genes into mouse oocyte DNA by transgenesis. This report describes the generation and immunological status of transgenic mice carrying a recombinant H-2Kb structural gene flanked by human /3-globin gene transcnptional control elements, which limits transgene expression to erythroid cells. Methods Mice CBA/Ca (H^*), BALB/c ( H - ^ and C57BL/10 (B10, H-2*) animals were bred at the National Institute for Medical Research; B10.BR (H^*) mice were obtained from Olac Ltd, Oxford, UK.
The H-2Kb-/3-globin (KjS) recombinant gene was constructed by ligating the transcriptional control elements from the human /3-globin gene to a promotorless H-2Kb gene (Fig. 1). A 5 kb Nott -H/ndlll fragment of the H-2Kb gene (9) was exchanged for the Hin6\\\-Kpn\ fragment of the /3-H-2 k -DCR microlocus cassette (10,11 and M. Antoniou, personal communication). Generation of transgenic mice JheXho\ -Nar\ (10.5 kb) fragment of the K/3 recombinant gene (Fig. 1) was used for microinjection. The DNA band was purified by Elutip elution (Schleicher & Schuell) resuspended in 10 mM Tris-HCI, 0.25 mM EDTA at a concentration of 1 figlm\ and injected into the pronudei of fertilized oocytes from CBA/Ca (y\-2*) mice. Two founder animals carrying integrated copies of
Flow cytometnc analysis of antigen expression on blood cells Peripheral Wood lymphocytes (PBLs) and red Wood cells (RBCs) were assayed for expression of H-2Kb molecules. Animals were tail bled and blood samples prepared with and without osmotic lysis to remove RBCs. Cells were stained with a mouse anti-mouse IgG mAb to H-2Kb (B8:24.3) (12) followed by an FITCconjugated goat anti-mouse y chain (Sigma) and analysed by flow cytometry (FACSTAR1*", Becton-Dickinson). Skin grafting Skin grafting was performed by transplanting tail skin from donor mice onto the back of recipient animals, following the method of Billingham etal. (13). Plaster bandages were removed at day 10 and the grafts inspected daily for signs of rejection. T cell proliferation assays Responder lymph node cells (5 x 104) were co-cultured with 5 x 105 irradiated (2000 rad) stimulator splenocytes in 200 /J of RPMI, containing 10% FCS, 10 mM glutamine, 100U/ml
Mouse H-2 K b Gene Na H
CCIal H Hind i n KKpnl NNot I NaNarl X Xhol f CAAT Box 9 TATA Box
DCR Region
Probe Fig. 1. Construction of the H-2Kb-/3-globin {KB) transgene. A 5 kb/Vofl-H/ndlll DNA fragment containing a promoterless H-2Kb structural gene (top line) was purified and ligated to the /3-H-2 - DCR microlocus cassette (10,11 and M. Antoniou, personal communication, bottom line) using standard procedures. Vertical arrows in the 5' flanking part of the microlocus cassette indicate DNase I hypersensitive sites which correlate with control of erythroid lineage specific transcription (10,11). DCR » Dominant control region. The location of the human C/al -H/ndlll probe used to detect the transgene in mouse tail biopsy DNA samples is shown.
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
Construction of the recombinant gene
the transgene were identified from a litter of eight, mated to CBA/Ca mice, and their progeny characterized for transgene expression. DNA from tail biopsies was digested with SamHI, and Southern blots hybridized with a /3-globin C/al -H/ndlll probe (Fig. 1) labelled by nick translation. Control transgenic CBA/Ca mice carrying a normal H-2Kb gene, complete with its own H-2 transcriptional promoter (subsequently referred to as CBK mice), were also generated. CBK mice express H-2Kb on most cells (unpublished data).
T cell function in MHC class I transgenic mice 61 penicillin, 100/»g/ml streptomycin, 10 mM HEPES, and 5 x 10" 4 M 2-mercaptoethano), in flat bottomed microtitre plates at 37°C in 5% 0 0 ^ 9 5 % air. On day 5 cultures were pulsed with 1 ^Ci of [3H]thymidine per well and [3H]thymidine incorporation determined 18 h later by scintillation counting. T cell cytotoxicity assays
Results Expression of the Kfi gene in transgenic mice Expression of H-2Kb on RBCs and PBLs, from both transgenic lineages, was determined by flow cytometric analysis using a mAb (B8:24:3) that recognizes H-2Kb. RBCs and lymphocytes were identified as discrete, non-overlapping populations from their forward and side scatter profiles (Fig. 2A) and gates were set appropriately. Fluorescence histograms of RBCs from control (C57BL/10, H-2b) and transgenic mice show that - 5 0 % of RBCs express H-2Kb polypeptides, which are not detectable on RBCs from non-transgenic littermates (equivalent to CBA mice) (Fig. 2B). In contrast PBLs from transgenic animals do not express H-2Kb (Fig. 2C) indicating that the recombinant gene is being expressed on erythroid but not lymphoid cells from transgenic mice. Further evidence that H-2Kb polypeptides are expressed only on erythroid cells in K/3 transgenic mice was obtained from skin grafting experiments. Thus, skin grafts from K/3 mice are not rejected by CBA recipients after a period of >150 days whereas skin grafts from control transgenic mice carrying a H-2Kb transgene complete with its own transcriptional promoter (CBK) are rejected with a mean survival time (MST) of 19 days (Table 1). Tolerance status of K/3 mice The tolerance status of K/3 transgenic mice gene was determined by skin grafting using control transgenic CBK mice as graft donors which differ from K/5 recipients only in the pattern of expression of the H-2Kb gene. CBK skin grafts persist on K/3 recipients for >150 days and show no evidence of rejection whereas their non-transgenic littermates (CBA mice) reject the H-2Kb disparate skin rapidly (MST = 19 days, Table 2). Therefore mice expressing H-2Kb under the control of human /3-globin transcriptional control elements are tolerant to H-2K" polypeptides.
H-2K" restricted T cell responses in K0 transgenic mice The ability of K0 transgenic mice (lineage 1) to make anti-minor histocompatibility (mH) antigen specific, H-2Kb restricted, CTL responses was investigated. B10.BR and CBA/Ca mice are MHC compatible (H-2*) but mH (B10 and CBA respectively) antigen incompatible. Furthermore, it is known that CTL responses raised against mH antigens of B10 origin are restricted, predominantly, by H-2Kb rather than by H-2* molecules (15, 16). Thus, K/3 transgenic mice were immunized against B10 mH antigens by injecting spleen cells from (B10.BR x CBK)F, mice, re-stimulated with irradiated spleen cells from the same source in vitro, and B10 mH-specific cytotoxicity determined in a 5 'Cr release assay using Con A blasts as target cells. H-2Kb restricted, anti-B10 mH cytotoxicity was assessed using B10 target cells. Table 5 shows that lymphocytes from K/3 transgenic mice lyzed B10, but not CBK, target cells indicating that the CTL response is specific for B10 mH antigens and is restricted by H-2Kb molecules. K/3 antiBALB/c and BALB/c anti-[B10.BR x CBK]F, CTL responses were included to demonstrate specificity and integrity of the CBK target cells.
Discussion In this study we describe the immunological status of transgenic mice which express a H-2Kb structural gene under the control of a transcriptional promoter and other control elements derived from the human /3-globin gene (17-19). The microlocus construct used to make the transgene has been shown to direct copy number dependent, but position independent, erythroid specific
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
Responder cells were generated in bulk one-way mixed lymphocyte reactions (MLRs): responder spleen cells (5 x 106/ml) were co-cultured with irradiated stimulator splenocytes (5 x lO^/ml) for 5 days, with or without recombinant IL-2 (4 U/ml, Sigma), and tested in a standard 51Cr release assay. Tests were performed (in triplicate) in round-bottomed microtitre plates at effectontarget ratios of 30:1, 10:1, 3:1 and 1:1. After 3 h, 20 /J of supernatant was collected from each well and counted in a gamma counter. Regression analysis was performed as described (14). EL-4 0-1-2"), L cells (H-2"), and P815 ( H - ^ targetcell lines were used. In experiments where Con A blasts were used as targets, osmotically-lysed spleen cells (1 x IC^/ml) were incubated with 4 /ig/ml of Con A (Sigma) for 5 days.
T cell responses to the H^hfi molecule Lymph node cells from K/3 transgenic mice (lineage 1) were cocuftured with irradiated splenic stimulator cells in the absence of exogenous IL-2 to investigate their ability to proliferate to H-2Kb molecules in vitro. Table 3 shows the outcome of a typical experiment when co-culturing K/3 responder cells with CBK stimulators results in an - 2-fold increase in pHJthymidine incorporation above background proliferation (i.e. proliferation of responder lymphocytes co-cultured with syngeneic stimulator cells). There is an ~ 4-fold increase in response to third party stimulator cells (BALB/c, H-2d) that differ at both MHC class I and class II loci. Responder cells from non-transgenic littermates (CBA mice) exhibit essentially identical proliferative responses to the K/3 transgenic mice (Table 3). In addition, the ability of lymphocytes from K/3 mice to make specific cytotoxic T lymphocyte (CTL) responses to H-2Kb polypeptides was determined. Spleen cells from transgenic mice were co-cultured with irradiated CBK stimulator cells, in the absence or presence of exogenous IL-2, for 5 days. Lymphocytes from K/3 transgenic mice show no cytotoxic activity against H-2Kb-positive target cells, although they can kill specifically third party (H^") targets (Table 4). In contrast, transgenenegative littermates can make specific cytotoxic responses against H-2Kb and H-2? target cells, demonstrating that failure to generate anti-H-2Kb cytotoxicity correlates with presence of the K/3 transgene. In addition, K0 recipients of CBK skin grafts do not exhibit anti-H-2Kb cytotoxicity (data not shown), indicating that the presence of H-2Kb on a skin graft does not immunize K0 mice for anti-H-2Kb cytotoxic responses in vitro.
64
T cell function in MHC class I transgenic mice
Notwithstanding the mechanism by which tolerance is acquired, it is also of interest to consider whether H-2Kb reactive T cells are tolerized in the thymus or in the extra-thymic compartment, and which cells are able to present H-2Kb to T cells to effect tolerance in K/3 mice. Although the transcriptional control elements contained in the /3 globin microlocus cassette have been shown to confer tightly regulated erythroid specific transcription of several genes in transgenic mice (10,11,19,20), we cannot formally exclude the possibility that the transgene is expressed in at least some thymic cells, either at levels too low to be detected by immunochemical staining, or at critical periods during development of the T cell repertoire. This problem cannot be resolved by resorting to tissue RNA analysis using PCR or S1 nuclease protection analysis to detect transcripts of the K/3 transgene since passenger erythroid cells are present in all tissues. To resolve the issue of whether erythroid of non-erythroid cells are responsible for influencing selection of the T cell repertoire in K/3 mice, we intend to carry out adoptive transfers of bone marrow stem cells or thymus grafts from K/3 mice into CBA recipients. We have also tested whether limiting H-2Kb expression to cells of the erythroid lineage affects the ability of the immune system to recognize and use H-2Kb molecules as restriction elements for CTL responses. Our results show that CTL responses to minor histocompatibility antigens from B10 mice are restricted by H-2Kb in K/3 responder mice. Since positive selection of immature thymocytes is thought to be the process which results in MHC restriction of T cell responses (2,27,28), this result implies that the T cell repertoire in K/3 mice is positively selected by H-2Kb molecules. However, positive selection is thought to take place as a result of interactions between thymocytes and thymic epithelial cells expressing self MHC molecules (2,27,28). Hence, the finding that H-2Kb restricted cytotoxic T cells are present in the peripheral repertoire of T cells in K/3 mice is rather surprising since human /3-globin transcriptional control elements are not thought to be active in epithelial cells. It is possible that red blood cells are able to act as H-2Kb presenting cells effecting positive selection of thymocytes or T cells in K/3 transgenic mice although, as discussed above, we cannot exclude the possibility that low levels of H-2Kb expression in certain thymic epithelial cells does take place either
constitutjvely or during critical periods in development, and is sufficient to bring about positive selection. Acknowledgements We thank Michael Antoniou for help with the K/3 construct, Chris Atkins for operating the FACSCAN, Marlene Bertagne for typing the manuscript, and Elizabeth Simpson and Sandra Husbands for advice with the restriction experiments.
Abbreviations APC CTL mH MLR MST PBL RBC
antigen presenting cell cytotoxic T lymphocyte minor histocompatibility mixed lymphocyte reaction mean survival time peripheral Wood lymphocyte red blood cell
References 1 Blackman, M., Kappler, J., and Ma/rack, P. 1990 The role of the T cell receptor in positive and negative selection of developing T cells. Science 2481335. 2 Bevan, M. J. and Fink, P. J. 1978. The influence of thymus H-2 antigens on the specificity of maturing kfller and helper cells. Immunol. Rev. 42:3. 3 Benoist, C and Malhis, D 1989. Positive selection of the T cell repertoire: where and when does it occur? Cell 58:1027. 4 Matzinger, P. and Guerder, S. 1989. Does T-cell tolerance require a dedicated antigen-presenting cell? Nature 338:74 5 Kappler, J W., Roehm, N., and Marrack, P. 1987. T cell tolerance by clonal elimination in the thymus. Cell 49:273. 6 Webb, S. R. and Sprent, J. 1990. Induction of neonatal tolerance to mis8 antigens by CD8 + T cells. Science 248:1643. 7 Swat, W., Ignatowicz, L., von Boehmer, H., and Kisielow, P. 1991. Clonal deletion of immature CD4 + CD8 + thymocytes in suspension culture by extrathymic antigen presenting cells. Nature 351.150. 8 Ramsdell, F., Lantz, T., and Fowlkes, B. J 1990. A nondeletional mechanism of thymic self tolerance. Science 246:1038. 9 Weiss, E. H., Golden, L, Fahrner, K., Mellor, A. L, Devlin, J. J , Bullman, H., Bud, H., and Flavell, R. A. 1984. Organisation and evolution of the class I gene family in the major histocompatibility complex of the C57BU10 mouse. Nature 310 650. 10 CoJIis, P., Antoniou, M , and Grosveld, F. 1990. Definition of the minimal requirements within the human /3-globin gene and the dominant control region for high level of expression. EMBO J. 9233 11 Antoniou, M. and Grosveld, F. 1990. 0-gtobin dominant control region interacts differently with distal and proximal promoter elements. Genes Dev. 4:1007 12 Kohler, G., Fischer-Lindahl, K., and Hensser, C. 1981. Characterization of a monoclonal anti-H-2Kb antibody. Immune System 2:202. 13 Billingham, R. E. and Medawar, P. B. 1951. The technique of free skin grafting in mammals. J. Exp. Biol. 28:385. 14 Simpson, E. and Chandler, P. 1986. Analysis of cytotoxic T cell responses. In Weir, D. M., e d , Handbook of Experimental Immunology, Volume 2, p. 68. Blackwell Scientific, Oxford. 15 Wettstein, P. J. and Frelinger, J. A. 1980. H-2 effects on cell-cell interactions in the response to single non-H-2 aBoantigens. III. Evidence for a second Ir-gene system mapping in the H-2K and H-2D regions. Immunogenetics 10211 16 Wettstein, P. J. 1982. H-2 effects on cell-cell interactions in the response to single non-H-2 alloantigens. V. Effects of H-2Kb mutations on presentation H-4 and H-3 alloantigens. J. Immunol. 128:2629. 17 Townes, T., Lingrel, J. B., Chen, H. Y , Brinster, R. L., and Palmier, R. D. 1985. ErythrcHd-specific expression of human 0-globin genes in transgenic mice. EMBO J. 4:1715. 18 Magram, J., Chada, K., and Constantini, F. 1985. Developmental regulation oJ a cloned adult /3-gk>bin gene in transgenic mice. Nature
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
grafts from donor mice expressing H-2Kb in skin tissues. However, the CD4/CD8 phenotype of the proliferating T cells has not been determined and, consequently, we cannot exclude that the C D 8 \ but not the C D 4 \ subset of H-2Kb reactive T cells are eliminated in K/3 mice. In this event, elimination of H-2Kb reactive thymocytes would be expected to occur at the CD8 + CD4" stage of thymocyte differentiation rather than at the less mature CD8 + CD4 + stage. Indeed, tolerance induction by deletion of CD8 + CD4" in the thymic medulla has been described previously in transgenic mice (26). Against this partial deletion model, there is almost no difference in anti H-2Kb proliferate T cell responses in K/3 and control CBA mice. This shows that the presence of the transgene has no marked effect on the ability of H-2Kb reactive T cells to proliferate in vitro compared to CBA mice. To resolve the issue of whether tolerance is acquired by a deletional or non-deletional mechanism in K/3 mice we intend to mate K/3 mice to transgenic mice carrying T cell receptor transgenes coding for an anti H-2Kb alloreactive clonotype and determine the fate of T cells in double transgenic mice.
T cell function in MHC class I transgenic mice 65 mechanism for extrathymic tolerance induction Cell 65293. 24 Ohashi, P. S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C. T., Odermatt, B., Mahssen, B., Zinkernagel, R. M., and Hengartner, H. 1991. Ablation of 'tolerance' and induction of diabetes by virus infection in viral antigen transgenic mice. Cell 65:305. 25 Oldstone, M. B. A., Nerenberg, M., Southern, P., Price, J., and Lewicki H. 1991. Virus infection triggers insulin-dependent diabetes mellitus in a transgenic model: Role of anti-setf (virus) immune response. Cell 65:319. 26 Pircher, H., BurW, K., Lang, R., Hengartner, H., and Zinkernagel, R. M. 1989. Tolerance induction in double specific T-cell receptor transgenic mice varies with antigen. Nature 342:559. 27 Spetser, D. E., Lees, R. K., Hengartner, H., Zinkernagel, R. M., and MacDonald, H. R. 1989. Positive and negative selection of T cell receptor V0 domains controlled by distinct cell populations in the thymus. J. Exp. Med. 170:2165. 28 Benoist, C. and Mathis, D. 1989. Positive selection of the T cefl repertoire: Where and when does it occur? Cell 58:1027
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
315338. 19 Talbot, D., Collis, P., Antoniou, M., VkJaJ, M., Grosveld, F., and Greaves D. 1989. A dominant control region from the human /3-gtobin locus conferring integration site-independent gene expression Nature 338:352. 20 Btom van Assendetft, G., Hanscombe, O., Grosveld, F., and Greaves, D. R. 1989. The /3-gk>bin dominant control region activates homologous and heterologous promoters in a tissue-specific manner. Cell 56:969. 21 Miller, J. F. A. P., Morahan, G.. and Allison, J. 1989. Immunological tolerance, new approaches using transgenic mice. Immunol. Today 10:53. 22 Miller, J. F. A. P., Morahan, G., Slattery, R , and Alison, J. 1990. Transgene models of T cell self tolerance and autoimmunity. Immunol. Rev. 118.21. 23 Schonrich, G., Kalinke, U., Momburg, M., Malissen, M., SchmittVerhulst, A.-M., Malissen, B., Hammerfing, G. J., and Arnold, B. 1991. Down-regulation of T cell receptors on self-reactive T cells as a novel
59-65
© 1992 Oxford University Press
Tolerance and MHC restriction in transgenic mice expressing a MHC class I gene in erythroid cells Helen Yeoman1 and Andrew L. Mellor National Insitute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK
Abstract Transgenic mice carrying a MHC class I structural gene (H-2Kb) linked to transcrlptional control elements from the human 0-globln gene, which direct erythroid lineage specific transcription, express H-2K" molecules In red blood cells but H-2K" expression cannot be detected in skin or lymphold cells. This limited pattern of self MHC expression Is sufficient to Induce tolerance to H-2Kb molecules and H-2K" restricted cytotoxic T cell responses can be generated in transgenic mice. Transgenic mice are unable to mount H-2K" specific cytotoxic T cell responses In vitro, even when exogenous IL-2 is provided. However, H-2Kb specific T cell proliferative responses are comparable with H-2Kb specific responses In non-transgenic mice, even In the absence of exogenous IL-2. Thus, expression of H-2Kb molecules under control of human 0-globln transcriptlonal control elements In transgenic mice Is tolerogenlc but does not result in elimination of all H-2K" reactive T cells from the mature repertoire. This suggests that tolerance In these mice may arise due to functional Inactivation of H-2Kb reactive T cells In vivo when they encounter H-2Kb molecules expressed on cells of erythroid cell lineages or on non-erythroid cells which express H-2Kb molecules at very low levels or in a developmentally regulated pattern. Furthermore, In spite of the failure to detect H-2Kb expression on non-erythrold cells In these mice, we conclude that H-2K" molecules participate In positive selection of the T cell repertoire since H-2Kb restricted T cell responses can be generated in these transgenic mice. Introduction During development in the thymus, T cells undergo two apparently paradoxical selection processes that determine the repertoire of peripheral T cells. Firstly, there is a selection of T cells whose receptors recognize self MHC molecules, so that mature T cells recognize foreign peptides only in association with self MHC molecules (positive selection which results in MHC restriction). Secondly, T cells that recognize self MHC molecules in the absence of foreign peptide are selected against (negative selection), resulting in a mature repertoire that is tolerant to self (1). Studies with chimeric mice indicate that positive selection is a function of the MHC haplotype of radioresistant thymic epithelial cells rather than that of the reconstituting bone marrow (2). Experiments with MHC class II transgenic mice that show a limited transgene expression indicate that cells of the thymus cortex, but not medulla, are able to positively select developing T cells (3). In contrast, tolerance induction appears to be mediated by cells of bone marrow origin (2) and can be a function of
'professional' antigen presenting cells (APCs) (4). The major mechanism of tolerance induction is the physical elimination of potentially autoreactive T cells. For example, T cell clones expressing receptors with certain Vfl chains are deleted intrathymically in mouse strains that express the corresponding ligand (5). Although the question of which cell types are able to affect deletion is not fully resolved, APCs, T cells and thymic medullary cells can cause clonal elimination of some thymocytes (1,6,7). However, clonal deletion is not the only mechanism of tolerance induction; antigen expression on cells of the thymic medulla results in specific tolerance in the periphery without physical elimination of potentially autoreactive clones which, instead, are rendered specifically unresponsive (anergic) to self antigen in the periphery (8). We are interested in defining cell lineages able to mediate positive and negative selection and the approach we have utilized is to put novel MHC class I genes under the control of specific
Correspondence to: A. L. Mellor Transmitting editor: E. Simpson
Received 11 July 1991, accepted 7 October 1991
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
Key words: MHC restriction, transgenic mice, tolerance, erythroid cells
60
T cell function in MHC class I transgenic mice
transcnptional regulators and introduce these recombinant genes into mouse oocyte DNA by transgenesis. This report describes the generation and immunological status of transgenic mice carrying a recombinant H-2Kb structural gene flanked by human /3-globin gene transcnptional control elements, which limits transgene expression to erythroid cells. Methods Mice CBA/Ca (H^*), BALB/c ( H - ^ and C57BL/10 (B10, H-2*) animals were bred at the National Institute for Medical Research; B10.BR (H^*) mice were obtained from Olac Ltd, Oxford, UK.
The H-2Kb-/3-globin (KjS) recombinant gene was constructed by ligating the transcriptional control elements from the human /3-globin gene to a promotorless H-2Kb gene (Fig. 1). A 5 kb Nott -H/ndlll fragment of the H-2Kb gene (9) was exchanged for the Hin6\\\-Kpn\ fragment of the /3-H-2 k -DCR microlocus cassette (10,11 and M. Antoniou, personal communication). Generation of transgenic mice JheXho\ -Nar\ (10.5 kb) fragment of the K/3 recombinant gene (Fig. 1) was used for microinjection. The DNA band was purified by Elutip elution (Schleicher & Schuell) resuspended in 10 mM Tris-HCI, 0.25 mM EDTA at a concentration of 1 figlm\ and injected into the pronudei of fertilized oocytes from CBA/Ca (y\-2*) mice. Two founder animals carrying integrated copies of
Flow cytometnc analysis of antigen expression on blood cells Peripheral Wood lymphocytes (PBLs) and red Wood cells (RBCs) were assayed for expression of H-2Kb molecules. Animals were tail bled and blood samples prepared with and without osmotic lysis to remove RBCs. Cells were stained with a mouse anti-mouse IgG mAb to H-2Kb (B8:24.3) (12) followed by an FITCconjugated goat anti-mouse y chain (Sigma) and analysed by flow cytometry (FACSTAR1*", Becton-Dickinson). Skin grafting Skin grafting was performed by transplanting tail skin from donor mice onto the back of recipient animals, following the method of Billingham etal. (13). Plaster bandages were removed at day 10 and the grafts inspected daily for signs of rejection. T cell proliferation assays Responder lymph node cells (5 x 104) were co-cultured with 5 x 105 irradiated (2000 rad) stimulator splenocytes in 200 /J of RPMI, containing 10% FCS, 10 mM glutamine, 100U/ml
Mouse H-2 K b Gene Na H
CCIal H Hind i n KKpnl NNot I NaNarl X Xhol f CAAT Box 9 TATA Box
DCR Region
Probe Fig. 1. Construction of the H-2Kb-/3-globin {KB) transgene. A 5 kb/Vofl-H/ndlll DNA fragment containing a promoterless H-2Kb structural gene (top line) was purified and ligated to the /3-H-2 - DCR microlocus cassette (10,11 and M. Antoniou, personal communication, bottom line) using standard procedures. Vertical arrows in the 5' flanking part of the microlocus cassette indicate DNase I hypersensitive sites which correlate with control of erythroid lineage specific transcription (10,11). DCR » Dominant control region. The location of the human C/al -H/ndlll probe used to detect the transgene in mouse tail biopsy DNA samples is shown.
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
Construction of the recombinant gene
the transgene were identified from a litter of eight, mated to CBA/Ca mice, and their progeny characterized for transgene expression. DNA from tail biopsies was digested with SamHI, and Southern blots hybridized with a /3-globin C/al -H/ndlll probe (Fig. 1) labelled by nick translation. Control transgenic CBA/Ca mice carrying a normal H-2Kb gene, complete with its own H-2 transcriptional promoter (subsequently referred to as CBK mice), were also generated. CBK mice express H-2Kb on most cells (unpublished data).
T cell function in MHC class I transgenic mice 61 penicillin, 100/»g/ml streptomycin, 10 mM HEPES, and 5 x 10" 4 M 2-mercaptoethano), in flat bottomed microtitre plates at 37°C in 5% 0 0 ^ 9 5 % air. On day 5 cultures were pulsed with 1 ^Ci of [3H]thymidine per well and [3H]thymidine incorporation determined 18 h later by scintillation counting. T cell cytotoxicity assays
Results Expression of the Kfi gene in transgenic mice Expression of H-2Kb on RBCs and PBLs, from both transgenic lineages, was determined by flow cytometric analysis using a mAb (B8:24:3) that recognizes H-2Kb. RBCs and lymphocytes were identified as discrete, non-overlapping populations from their forward and side scatter profiles (Fig. 2A) and gates were set appropriately. Fluorescence histograms of RBCs from control (C57BL/10, H-2b) and transgenic mice show that - 5 0 % of RBCs express H-2Kb polypeptides, which are not detectable on RBCs from non-transgenic littermates (equivalent to CBA mice) (Fig. 2B). In contrast PBLs from transgenic animals do not express H-2Kb (Fig. 2C) indicating that the recombinant gene is being expressed on erythroid but not lymphoid cells from transgenic mice. Further evidence that H-2Kb polypeptides are expressed only on erythroid cells in K/3 transgenic mice was obtained from skin grafting experiments. Thus, skin grafts from K/3 mice are not rejected by CBA recipients after a period of >150 days whereas skin grafts from control transgenic mice carrying a H-2Kb transgene complete with its own transcriptional promoter (CBK) are rejected with a mean survival time (MST) of 19 days (Table 1). Tolerance status of K/3 mice The tolerance status of K/3 transgenic mice gene was determined by skin grafting using control transgenic CBK mice as graft donors which differ from K/5 recipients only in the pattern of expression of the H-2Kb gene. CBK skin grafts persist on K/3 recipients for >150 days and show no evidence of rejection whereas their non-transgenic littermates (CBA mice) reject the H-2Kb disparate skin rapidly (MST = 19 days, Table 2). Therefore mice expressing H-2Kb under the control of human /3-globin transcriptional control elements are tolerant to H-2K" polypeptides.
H-2K" restricted T cell responses in K0 transgenic mice The ability of K0 transgenic mice (lineage 1) to make anti-minor histocompatibility (mH) antigen specific, H-2Kb restricted, CTL responses was investigated. B10.BR and CBA/Ca mice are MHC compatible (H-2*) but mH (B10 and CBA respectively) antigen incompatible. Furthermore, it is known that CTL responses raised against mH antigens of B10 origin are restricted, predominantly, by H-2Kb rather than by H-2* molecules (15, 16). Thus, K/3 transgenic mice were immunized against B10 mH antigens by injecting spleen cells from (B10.BR x CBK)F, mice, re-stimulated with irradiated spleen cells from the same source in vitro, and B10 mH-specific cytotoxicity determined in a 5 'Cr release assay using Con A blasts as target cells. H-2Kb restricted, anti-B10 mH cytotoxicity was assessed using B10 target cells. Table 5 shows that lymphocytes from K/3 transgenic mice lyzed B10, but not CBK, target cells indicating that the CTL response is specific for B10 mH antigens and is restricted by H-2Kb molecules. K/3 antiBALB/c and BALB/c anti-[B10.BR x CBK]F, CTL responses were included to demonstrate specificity and integrity of the CBK target cells.
Discussion In this study we describe the immunological status of transgenic mice which express a H-2Kb structural gene under the control of a transcriptional promoter and other control elements derived from the human /3-globin gene (17-19). The microlocus construct used to make the transgene has been shown to direct copy number dependent, but position independent, erythroid specific
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
Responder cells were generated in bulk one-way mixed lymphocyte reactions (MLRs): responder spleen cells (5 x 106/ml) were co-cultured with irradiated stimulator splenocytes (5 x lO^/ml) for 5 days, with or without recombinant IL-2 (4 U/ml, Sigma), and tested in a standard 51Cr release assay. Tests were performed (in triplicate) in round-bottomed microtitre plates at effectontarget ratios of 30:1, 10:1, 3:1 and 1:1. After 3 h, 20 /J of supernatant was collected from each well and counted in a gamma counter. Regression analysis was performed as described (14). EL-4 0-1-2"), L cells (H-2"), and P815 ( H - ^ targetcell lines were used. In experiments where Con A blasts were used as targets, osmotically-lysed spleen cells (1 x IC^/ml) were incubated with 4 /ig/ml of Con A (Sigma) for 5 days.
T cell responses to the H^hfi molecule Lymph node cells from K/3 transgenic mice (lineage 1) were cocuftured with irradiated splenic stimulator cells in the absence of exogenous IL-2 to investigate their ability to proliferate to H-2Kb molecules in vitro. Table 3 shows the outcome of a typical experiment when co-culturing K/3 responder cells with CBK stimulators results in an - 2-fold increase in pHJthymidine incorporation above background proliferation (i.e. proliferation of responder lymphocytes co-cultured with syngeneic stimulator cells). There is an ~ 4-fold increase in response to third party stimulator cells (BALB/c, H-2d) that differ at both MHC class I and class II loci. Responder cells from non-transgenic littermates (CBA mice) exhibit essentially identical proliferative responses to the K/3 transgenic mice (Table 3). In addition, the ability of lymphocytes from K/3 mice to make specific cytotoxic T lymphocyte (CTL) responses to H-2Kb polypeptides was determined. Spleen cells from transgenic mice were co-cultured with irradiated CBK stimulator cells, in the absence or presence of exogenous IL-2, for 5 days. Lymphocytes from K/3 transgenic mice show no cytotoxic activity against H-2Kb-positive target cells, although they can kill specifically third party (H^") targets (Table 4). In contrast, transgenenegative littermates can make specific cytotoxic responses against H-2Kb and H-2? target cells, demonstrating that failure to generate anti-H-2Kb cytotoxicity correlates with presence of the K/3 transgene. In addition, K0 recipients of CBK skin grafts do not exhibit anti-H-2Kb cytotoxicity (data not shown), indicating that the presence of H-2Kb on a skin graft does not immunize K0 mice for anti-H-2Kb cytotoxic responses in vitro.
64
T cell function in MHC class I transgenic mice
Notwithstanding the mechanism by which tolerance is acquired, it is also of interest to consider whether H-2Kb reactive T cells are tolerized in the thymus or in the extra-thymic compartment, and which cells are able to present H-2Kb to T cells to effect tolerance in K/3 mice. Although the transcriptional control elements contained in the /3 globin microlocus cassette have been shown to confer tightly regulated erythroid specific transcription of several genes in transgenic mice (10,11,19,20), we cannot formally exclude the possibility that the transgene is expressed in at least some thymic cells, either at levels too low to be detected by immunochemical staining, or at critical periods during development of the T cell repertoire. This problem cannot be resolved by resorting to tissue RNA analysis using PCR or S1 nuclease protection analysis to detect transcripts of the K/3 transgene since passenger erythroid cells are present in all tissues. To resolve the issue of whether erythroid of non-erythroid cells are responsible for influencing selection of the T cell repertoire in K/3 mice, we intend to carry out adoptive transfers of bone marrow stem cells or thymus grafts from K/3 mice into CBA recipients. We have also tested whether limiting H-2Kb expression to cells of the erythroid lineage affects the ability of the immune system to recognize and use H-2Kb molecules as restriction elements for CTL responses. Our results show that CTL responses to minor histocompatibility antigens from B10 mice are restricted by H-2Kb in K/3 responder mice. Since positive selection of immature thymocytes is thought to be the process which results in MHC restriction of T cell responses (2,27,28), this result implies that the T cell repertoire in K/3 mice is positively selected by H-2Kb molecules. However, positive selection is thought to take place as a result of interactions between thymocytes and thymic epithelial cells expressing self MHC molecules (2,27,28). Hence, the finding that H-2Kb restricted cytotoxic T cells are present in the peripheral repertoire of T cells in K/3 mice is rather surprising since human /3-globin transcriptional control elements are not thought to be active in epithelial cells. It is possible that red blood cells are able to act as H-2Kb presenting cells effecting positive selection of thymocytes or T cells in K/3 transgenic mice although, as discussed above, we cannot exclude the possibility that low levels of H-2Kb expression in certain thymic epithelial cells does take place either
constitutjvely or during critical periods in development, and is sufficient to bring about positive selection. Acknowledgements We thank Michael Antoniou for help with the K/3 construct, Chris Atkins for operating the FACSCAN, Marlene Bertagne for typing the manuscript, and Elizabeth Simpson and Sandra Husbands for advice with the restriction experiments.
Abbreviations APC CTL mH MLR MST PBL RBC
antigen presenting cell cytotoxic T lymphocyte minor histocompatibility mixed lymphocyte reaction mean survival time peripheral Wood lymphocyte red blood cell
References 1 Blackman, M., Kappler, J., and Ma/rack, P. 1990 The role of the T cell receptor in positive and negative selection of developing T cells. Science 2481335. 2 Bevan, M. J. and Fink, P. J. 1978. The influence of thymus H-2 antigens on the specificity of maturing kfller and helper cells. Immunol. Rev. 42:3. 3 Benoist, C and Malhis, D 1989. Positive selection of the T cell repertoire: where and when does it occur? Cell 58:1027. 4 Matzinger, P. and Guerder, S. 1989. Does T-cell tolerance require a dedicated antigen-presenting cell? Nature 338:74 5 Kappler, J W., Roehm, N., and Marrack, P. 1987. T cell tolerance by clonal elimination in the thymus. Cell 49:273. 6 Webb, S. R. and Sprent, J. 1990. Induction of neonatal tolerance to mis8 antigens by CD8 + T cells. Science 248:1643. 7 Swat, W., Ignatowicz, L., von Boehmer, H., and Kisielow, P. 1991. Clonal deletion of immature CD4 + CD8 + thymocytes in suspension culture by extrathymic antigen presenting cells. Nature 351.150. 8 Ramsdell, F., Lantz, T., and Fowlkes, B. J 1990. A nondeletional mechanism of thymic self tolerance. Science 246:1038. 9 Weiss, E. H., Golden, L, Fahrner, K., Mellor, A. L, Devlin, J. J , Bullman, H., Bud, H., and Flavell, R. A. 1984. Organisation and evolution of the class I gene family in the major histocompatibility complex of the C57BU10 mouse. Nature 310 650. 10 CoJIis, P., Antoniou, M , and Grosveld, F. 1990. Definition of the minimal requirements within the human /3-globin gene and the dominant control region for high level of expression. EMBO J. 9233 11 Antoniou, M. and Grosveld, F. 1990. 0-gtobin dominant control region interacts differently with distal and proximal promoter elements. Genes Dev. 4:1007 12 Kohler, G., Fischer-Lindahl, K., and Hensser, C. 1981. Characterization of a monoclonal anti-H-2Kb antibody. Immune System 2:202. 13 Billingham, R. E. and Medawar, P. B. 1951. The technique of free skin grafting in mammals. J. Exp. Biol. 28:385. 14 Simpson, E. and Chandler, P. 1986. Analysis of cytotoxic T cell responses. In Weir, D. M., e d , Handbook of Experimental Immunology, Volume 2, p. 68. Blackwell Scientific, Oxford. 15 Wettstein, P. J. and Frelinger, J. A. 1980. H-2 effects on cell-cell interactions in the response to single non-H-2 aBoantigens. III. Evidence for a second Ir-gene system mapping in the H-2K and H-2D regions. Immunogenetics 10211 16 Wettstein, P. J. 1982. H-2 effects on cell-cell interactions in the response to single non-H-2 alloantigens. V. Effects of H-2Kb mutations on presentation H-4 and H-3 alloantigens. J. Immunol. 128:2629. 17 Townes, T., Lingrel, J. B., Chen, H. Y , Brinster, R. L., and Palmier, R. D. 1985. ErythrcHd-specific expression of human 0-globin genes in transgenic mice. EMBO J. 4:1715. 18 Magram, J., Chada, K., and Constantini, F. 1985. Developmental regulation oJ a cloned adult /3-gk>bin gene in transgenic mice. Nature
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
grafts from donor mice expressing H-2Kb in skin tissues. However, the CD4/CD8 phenotype of the proliferating T cells has not been determined and, consequently, we cannot exclude that the C D 8 \ but not the C D 4 \ subset of H-2Kb reactive T cells are eliminated in K/3 mice. In this event, elimination of H-2Kb reactive thymocytes would be expected to occur at the CD8 + CD4" stage of thymocyte differentiation rather than at the less mature CD8 + CD4 + stage. Indeed, tolerance induction by deletion of CD8 + CD4" in the thymic medulla has been described previously in transgenic mice (26). Against this partial deletion model, there is almost no difference in anti H-2Kb proliferate T cell responses in K/3 and control CBA mice. This shows that the presence of the transgene has no marked effect on the ability of H-2Kb reactive T cells to proliferate in vitro compared to CBA mice. To resolve the issue of whether tolerance is acquired by a deletional or non-deletional mechanism in K/3 mice we intend to mate K/3 mice to transgenic mice carrying T cell receptor transgenes coding for an anti H-2Kb alloreactive clonotype and determine the fate of T cells in double transgenic mice.
T cell function in MHC class I transgenic mice 65 mechanism for extrathymic tolerance induction Cell 65293. 24 Ohashi, P. S., Oehen, S., Buerki, K., Pircher, H., Ohashi, C. T., Odermatt, B., Mahssen, B., Zinkernagel, R. M., and Hengartner, H. 1991. Ablation of 'tolerance' and induction of diabetes by virus infection in viral antigen transgenic mice. Cell 65:305. 25 Oldstone, M. B. A., Nerenberg, M., Southern, P., Price, J., and Lewicki H. 1991. Virus infection triggers insulin-dependent diabetes mellitus in a transgenic model: Role of anti-setf (virus) immune response. Cell 65:319. 26 Pircher, H., BurW, K., Lang, R., Hengartner, H., and Zinkernagel, R. M. 1989. Tolerance induction in double specific T-cell receptor transgenic mice varies with antigen. Nature 342:559. 27 Spetser, D. E., Lees, R. K., Hengartner, H., Zinkernagel, R. M., and MacDonald, H. R. 1989. Positive and negative selection of T cell receptor V0 domains controlled by distinct cell populations in the thymus. J. Exp. Med. 170:2165. 28 Benoist, C. and Mathis, D. 1989. Positive selection of the T cefl repertoire: Where and when does it occur? Cell 58:1027
Downloaded from http://intimm.oxfordjournals.org/ at University of California, San Fransisco on March 16, 2015
315338. 19 Talbot, D., Collis, P., Antoniou, M., VkJaJ, M., Grosveld, F., and Greaves D. 1989. A dominant control region from the human /3-gtobin locus conferring integration site-independent gene expression Nature 338:352. 20 Btom van Assendetft, G., Hanscombe, O., Grosveld, F., and Greaves, D. R. 1989. The /3-gk>bin dominant control region activates homologous and heterologous promoters in a tissue-specific manner. Cell 56:969. 21 Miller, J. F. A. P., Morahan, G.. and Allison, J. 1989. Immunological tolerance, new approaches using transgenic mice. Immunol. Today 10:53. 22 Miller, J. F. A. P., Morahan, G., Slattery, R , and Alison, J. 1990. Transgene models of T cell self tolerance and autoimmunity. Immunol. Rev. 118.21. 23 Schonrich, G., Kalinke, U., Momburg, M., Malissen, M., SchmittVerhulst, A.-M., Malissen, B., Hammerfing, G. J., and Arnold, B. 1991. Down-regulation of T cell receptors on self-reactive T cells as a novel
