Altered I-J phenotype in Ea transgenic mice

One of the more intriguing puzzles in immunology is the genetic basis for control of murine T-cell I-J determinants. Molecules bearing I-J determinants (I-J molecules) play a role in information trafficking among immunocompetent cells, probably serving as self-recognition molecules that channel regulatory factors to their appropriate target cells. Although it is clear that I-J polymorphism is influenced by the major histocompatibility complex (MHC), molecular genetic studies provide evidence that an MHC gene does not encode I-J molecules. A possible explanation for this paradox is that I-J molecules are a set of non-MHC-encoded T cell receptors that are directly or indirectly selected for by self-MHC products. One key to resolving the genetic and molecular basis for control of I-J determinants is the identification of theMHC gene(s) involved. Herein, data are presented which show that Ea transgenic mice express an altered I-J phenotype, providing clear evidence that I region class II genes influence I-J polymorphism. Although further study is required to resolve how class II genes mediate this effect, this is a major piece to the I-J puzzle. The I region of the murine major histocompatibility complex (MHC) or H-2 gene complex on chromosome 17 contains polymorphic class II genes that encode glycoprotein molecules found on B cells and macrophages (1, 2). These molecules, Mr 28,000-35,000, are also referred to as I-regionassociated or Ia molecules. At present, four class II genes have been well characterized: Aa and An mapping in the I-A subregion, E, mapping in the I-E subregion, and E# spanning both subregions (linear order Ap, Aa, E1, Ea). Products of these genes are expressed as noncovalently associated cell surface dimers, designatedAAp and EaEp complexes, which ,are readily detected by antibodies produced in I-regionincompatible strains. Class II molecules are intimately involved in the presentation of foreign antigen to the immune system-i.e., helper T-cell activation and interaction is dependent on recognition of foreign antigen plus self class II molecules on macrophage and B-cell surfaces. Class II molecules are thus important self-recognition elements for helper T-cell function (3). Recent studies with transgenic mice have formally proven that class II molecules are the products of immune response or Ir genes, which determine high or low responsiveness to numerous foreign protein antigens (4-6). Evidence for additional genes mapping in the I region came from studies which showed that some antibodies produced in I-region-incompatible strains reacted with suppressor T cells and T-cell-derived suppressor factors but not with B cells or helper T cells (refs. 7-9, reviewed in refs. 10-12). These antibodies clearly recognized a determinant controlled by the MHC or by a closely linked gene on chromosome 17, as judged by analysis of numerous MHC congenic and recombinant mouse strains. On the basis of differential reactivity of these antibodies with intra-I region recombinant strains, we postulated that a distinct I region gene, Ia4, mapping between the I-A and I-E subregions, controlled determinants expressed on suppressor T cells, and that crossovers in the recombinant strains defined a previously unknown subregion, designated I-J (7, 9, 10).¶ We now refer to these determinants collectively as "I-J determinants," and we refer to the molecules that bear these determinants as 'I-J molecules." Subsequent studies have shown that I-J determinants are expressed on several T-cell subsets and factors involved in the generation of suppressor activity (11-15), a T-cell subset that augments helper T-cell activity (12), T-cell subsets and factors involved in the generation of contrasuppressor activity (11), and macrophages involved in the generation ofhelper and suppressor activity (11, 13, 16). Different I-J determinants are expressed on several of these cell types and factors (11, 12, 17), raising the possibility that distinct loci encode the determinants. I-J molecules in several systems determine the self Igh-V-region-restricted and/or MHC-restricted activity of suppressor factors (Igh-V, variable region of immunoglobulin heavy chain gene complex), providing evidence that these molecules play a role in information trafficking among immunoregulatory cells (11, 13, 15, 18). Studies with chimeras, showing that the self-restricted activity of suppressor factors or I-J molecules is determined by the environment in which stem cells mature, are compatible with the concept that I-J molecules are a set of T-cell receptors that channel regulatory signals to their appropriate target (11, 12, 15). Under this view, I-J molecules on macrophages are probably passively acquired. Several studies suggest that I-J molecules are polypeptides, Mr 25,000 (12, 15, 19). Problems with the original interpretation of the genetic basis for control of I-J determinants arose during molecular genetic analysis of the I region (reviewed in ref. 10). Crossovers in the strains utilized to map the Ia4 gene between the I-A and I-E subregions and define the I-J subregion were localized to a recombination hot spot within the Ed gene (20, Abbreviations: MHC, major histocompatibility complex; Ia, Iregion-associated; Igh-V, variable region of immunoglobulin heavy chain gene complex; kb, kilobase(s); PFC, plaque-forming cells; SRBC, sheep erythrocytes; TsiF, T suppressor-inducer factor; ABMi, antigen-binding molecule from TsiF; I-Ji, I-J molecule from TsiF. tTo whom reprint requests should be addressed. $Key recombinant strains in the mapping of the Ia-4 locus and the definition of the I-J subregion include BlO.A(3R), which types Ia-4b and was judged to be I-Al I-Jb/I-Ek, and BlO.A(5R), which types Ia-4k, and was judged to be I-Ab/IlJk I-Ek. Both strains appear to carry the same alleles at class II loci, and both express AaAb and EkEb complexes. Similar results were obtained with two additional intra-I region recombinant strains, B1O.HTT (Ia-4' and I-AS I-Js/I-Ek) and B1O.S(9R) (Ia-4k and I-As/IJk I-Ek). See refs. 7, 9, and 10 for assumptions and detail. 8308 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Proc. Natl. Acad. Sci. USA 83 (1986) 8309 21). In addition, no mRNA was detected in I-J+ T-cell hybridomas with DNA probes that overlap the Ed gene and span most of the I region (22). Thus, although it is still clear that an I-region gene influences polymorphism in I-J molecules, it would appear that a non-I-region gene encodes I-J molecules. For example, I-J molecules may be non-MHCencoded T-cell receptors that are directly or indirectly selected for by I-region products during ontogeny (10-13, 15, 20, 23, 24). Tentative support for this receptor-ligand model comes from chimeras which show that I-J phenotype is influenced by the environment in which T cells mature (25, 26). Further complicating the I-J story is the observation that a gene (Jt) mapping on chromosome 4 influences either cell surface expression of I-J molecules or the relative number of I-J+ T cells in the periphery (24, 27). There is no evidence that the Jt gene influences the qualitative nature of or encodes I-J molecules (28, 29). One key to understanding the genetic basis for control of I-J determinants is identification of the I-region gene(s) involved. We (10, 11) and others (12-15, 20, 23, 24) have speculated that class II genes influence I-J polymorphism. A direct test of this hypothesis was achieved by analyzing Ea transgenic mice for their I-J phenotype. As the data will show, these mice display an altered I-J phenotype, providing clear evidence that class II genes play a role in determining I-J polymorphism. This is an important advance in our understanding of the genetics and biology of I-J molecules, and it sets the stage for solving the I-J mystery. MATERIALS AND METHODS Mice. C57BL/6 (abbreviated B6), SJL, (C57BL/10 x A)F1 [abbreviated (B10 x A)F1], and (B6 x SJL)F1 mice utilized for suppressor factor production were bred and maintained at the Laboratoire de Gdndtique Moldculaire des Eucaryotes, Strasbourg, France. B1O.A(3R) mice were bred and maintained at Yale University School of Medicine, while B1O.A(5R) mice and B6/J mice utilized in the suppressor assay were purchased from the Jackson Laboratory, Bar Harbor, Maine. Construction and Analysis ofEk Transgenic Mice. Production of the E16 line of transgenic mice has been described (4). Briefly, an 8.2-kilobase (kb) Bgl I fragment containing the Ekgene was injected into (B6 x SJL)F2 embryos. The injected DNA derives from the genome of A/J mice and contains All of the Ek exons plus 2 kb ofDNA 5' to the cap site and 1.3 kb 3' to the polyadenylylation site. The founder Ea16 male mouse was backcrossed to B6 and one male offspring, Ea166 (figure 6 of ref. 4), was determined positive for Ek expression by Southern blotting of tail DNA. This mouse was backcrossed to several B6 female mice, and the progeny of this cross were used in these experiments. The presence of the Ea transgene was assessed by preparing liver DNA from the animals and analyzing on Southern blots BamHI and EcoRV double digests, using as a probe the appropriate BamHI-EcoRV fragment. This probe allows distinction of the Ek transgene from its endogenous counterpart, which is smaller due to a deletion in the Eb and El genes (figure 1 of ref. 4). Detection of cell surface expression of the Ek gene product was accomplished by cytofluorometric analysis of spleen cells stained with fluorescein isothiocyanate-labeled monoclonal antibody 14.4.4 (4). The MHC haplotype of test mice was determined on the basis of a Pst I restriction fragment length polymorphism in theAa gene, using as a probe theAacDNA clone pAAC6 (30). Antibodies and Antigens. Monoclonal anti-Lyt2.2 serum was generously supplied by F. W. Shen (Memorial SloanKettering Cancer Center, New York). Monoclonal anti-I-JI antibodies (WF8.C12.8) and anti-I-Jb antibodies (WF9.40.5) were generated as previously described (31) and were a generous gift of Carl Waltenbaugh (Northwestern University Medical School, Chicago). Monoclonal anti-En antibody (14.4.4) was a generous gift of H. 0. McDevitt (Stanford University, Palo Alto, CA) (32). Sheep erythrocytes (SRBC) were obtained from Colorado Serum Co. (Denver, CO). Preparation of T-Cell-Derived Suppressor-Inducer Factor (TsiF). Preparation of SRBC-specific TsiF has been previously described (18). Briefly, a suspension of spleen cells from mice hyperimmunized with SRBC was treated with anti-Lyt2.2 monoclonal antibody and complement and subsequently cultivated in vitro for 48 hr in RPMI 1640 medium plus 10% fetal calf serum at 107 cells per ml. After 48 hr, supernatant fluids were harvested and passed through Millipore filters before use. Depletion of cells bearing the Lyt-2 marker was achieved by incubating 1 x 107 cells per ml of antibody appropriately diluted in balanced salt solution for 45 mm at room temperature, washing, and incubating with complement for 45 min at 370C. Complement used in these experiments was serum from rabbits selected for low natural cytotoxicity to mouse spleen cells. Suppression Assay. B6 spleen cells were washed in balanced salt solution and were suspended in RPM1 1640 medium supplemented with glutamine, antibiotics, mycostatin, 20 ,uM 2-mercaptoethanol, and 10% fetal calf serum. All cells were suspended at 107 spleen cells in 1 ml and were cultured with 0.05 ml of a 1% (vol/vol) suspension of SRBC in Falcon 3008 plates (Falcon Labware, Becton Dickinson) in a 5% C02/95% air incubator at 370C for 5 days. Suppressive activity of TsiF or separated molecules was determined by adding these materials to cultures of unprimed spleen cells at a final dilution of 10%o on day 0 of culture. The number of plaque-forming cells (PFC) in control and test cultures on day 5 was determined by using the Cunningham modification (44) of the Jerne-Nordin plaque assay. Results are given as the mean of three independent calculations from each culture condition (33). Analysis of I-J Molecules Associated with TsiF (I-Js). TsiF is composed of two molecules: an I-J+ antigen-nonbinding molecule (I-Jsi molecule) and an I-Jantigen-binding molecule (ABMSi). Both molecules are required to interact with an Lyt-1,2' transducer T cell to induce suppressive activity (33). The phenotype of I-J~j molecules associated with TsiF was determined by absorption-elution analysis with anti-I-J immunoabsorbent columns (33). Briefly, TsiF was passed over immunosorbent columns made by coupling anti-IjJb or anti-I-Jk antibody to Sepharose 4B (Pharmacia). We have previously shown that TsiF passed over anti-I-J columns separates the I-JABMsi molecule (which passes through the column in the unbound filtrate) from the I-J~j molecule (which is retained by the anti-I-J column) (18). After extensive washing, the columns were eluted with 0.2 M sodium carbonate, pH 11.0, to obtain any bound I-J5, molecules, and immediately neutralized with 0.3 M sodium borate buffer, pH 8.3. The column eluates were then concentrated back to their original volume and dialyzed overnight, first against phosphate-buffered saline, then against RPMI 1640 medium. Eluates from these columns were then tested for the presence of I-Joi molecules by their ability to restore suppressor activity in cultures containing ABMsj and spleen cells. Filtrates obtained by passing TsiF over an anti-I-J column served as a source of ABM5i. TsiF from individual E,16-6derived backcross progeny and pooled control mice were typed for both I-Jk and I-Jb determinants. Suppressor factors prepared from control strains B1O.A(3R) (I-Jb) and B1O.A(SR) (I-Jk) were tested before and after testing experimental TsiF to ensure the I-J specificity ofthe typing immunosorbent (see Table 2). This general method of I-J typing is routinely used in several laboratories (12, 13, 15, 34-37). Ea,16-6 backcross progeny were typed blind for Ek expression and I-J phenotype. Immunology: Flood et aL Proc. Natl. Acad. Sci. USA 83 (1986)