P. M. Hogarth

Demonstration of carbohydrate- and protein- determined ia antigens by monoclonal antibodies.

Ten monoclonal alloantibodies were examined by submitting each antibody to five independent tests in order to determine whether they reacted primarily with the glycoprotein or glycolipid class of Ia antigens. The tests employed were as follows: (1) the ability to precipitate an la-like protein from the cell surface as detected by SDS-PAGE; (2) inhibition by protein-la extracts free of CHO-Ia; (3) inhibition by CHO-Ia extracts free of protein-la; (4) neuraminidase sensitivity of the antigen and (5) inhibition by simple sugars. Using these tests, three of the ten monoclonal antibodies were shown to recognize a CHO-Ia antigen while seven recognized the protein class of Ia antigens. The three CHO-Ia-specific monoclonal antibodies recognized Ia specificities 2,9 and 17. Monoclonal antibodies recognizing protein-defined Ia.2 and 17 specificities were also characterized. These results imply that some Ia specificities, as defined by genetic testing, can occur both as carbohydrate-defined and protein-defined determinants.— Sugar inhibition studies showed that CHO-Ia.2 has D-glucosamine as its immunodominant sugar while CHO-Ia. 17 shows preference for aβ- linked galactose. Furthermore, studies with neuraminidase demonstrated that sialic acid plays a role in the antigenic determinants of CHO-Ia.9 and CHO-Ia.17. Finally, it is noteworthy that CHO-Ia.2, the private specificity of thek haplotype, appears to be expressed only on cells and not in serum. These studies clearly demonstrate the existence of the two Ia antigen classes and emphasize the complexity of the murineI region.

Chromosomal mapping of the high affinity Fcλ receptor gene

In this study, the high affinity Fc3, receptor gene, Fcgrl, is mapped to mouse chromosome 3. Fcgrl maps within a linkage group that is conserved with a portion of human chromosome 1 that spans the centromere. The arrangement of genes in this linkage group is compared to that of related Fc-receptor genes on mouse chromosome 1 and human chromosome 1. The results suggest that the duplication of a chromosome segment during evolution, in part explains the current relationship among these genes in both species. Receptors for IgG immune complexes (Fc3, receptors) are expressed on most haemopoietic cells. In the mouse there are three classes of Fc3, receptors. Fcgrl, a high affinity receptor and Fcgr2 and Fcgr3, both low affinity receptors. In the human, there is a high affinity receptor and multiple subtypes of low affinity receptors corresponding to Fcgr2 and Fcgr3 (Qiu et al. 1990). Human and mouse mononuclear phagocytes express a high affinity receptor for IgG that plays a role in macrophage and antibody dependant cellular cytotoxicity and clearance of immune complexes (Allen and Seed 1989). Fc3, receptors are sequence related to one another but have functional and structural diversity, cDNAs from mouse macrophage and T-cell lines that encode Fc-y receptors have been isolated to characterize structural and functional heterogeneity (Ravetch et al. 1986). The mouse Fc3, receptors show conserved extracellular domains and different transmembrane and intracytoplasmic domains. To map Fcgrl, the segregation of restriction fragment length variants (RFLVs) was examined in an interspecific backcross that has been analysed for 300 markers. These markers provide linkage maps for each mouse chromosome, except the Y (Saunders and Seldin 1990).

Mapping of the mouse Ly-6, Xp-14, and Gdc-1 loci to chromosome 15.

TheLy-6 locus is now regarded as a gene complex consisting of at least five closely linked loci (Ly-6A-Ly-6E) whose polymorphic products are identified by monoclonal antibodies and distinguished by different tissue distributions.Ly-6 has been assigned by other investigators to chromosome (Chr) 9 (linked toThy-1 or to Chr 2. We report that theLy-6 gene complex, together with theXp-14 andGdc -1 loci, is situated on Chr 15 linked toGpt1. These new linkage data are derived from four sources: (1) three separate crosses that failed to demonstrate linkage ofLy-6 to eitherThy-4 on Chr 9 or to any of five genes present on Chr 2; (2) the NXSM recombinant inbred strains, which suggested the linkage ofLy-6 andXp-14 toGpt-1 on Chr 15; (3) severalGpt-1 andGdc-1 congenic strains that confirmed the assignment ofLy-6 andXp-14 to Chr 15; and (4) backcrosses that further confirmed the linkage ofLy-6, Gpt-1, Gdc-4, andXp-14, the probable gene order beingGpt-11/Ly-6 Xp-14-Gdc-1.

The mouse Ly-12.1 specificity: genetic and biochemical relationship to Ly-1

The Ly-12 locus was originally defined in a study in which six loci (Ly-9, Ly-l l , Ly-12, Ly-13, Ly-14, and Ea-7) were detected using conventional alloantisera made by immunizing between strains C57L and 129 (Potter and McKenzie 1981). These antisera were difficult to analyze as they were complex and required extensive absorption to define each specificity and consequently were of low and variable titer. Thus, although the Ly-12 locus was defined, nothing was known of its chromosomal location, or chemistry, and its tissue distribution was also not completely defined. We now describe a more complete analysis of the Ly-12 locus using a monoclonal antibody, 5084-4.1, produced after immunizing 129/ReJ mice weeny with 5 X 107 C57L/J thymocytes for 5 weeks. The 129/ReJ spleen cells were prepared and fused with P3-NSI-1-Ag4 cells as described (Hogarth et al. 1980). The 5084-4.1 cell line was isolated and secreted on IgG2a Ly-12.1-specific antibody that was cytotoxic in the presence of rabbit complement. Only three inbred mouse strains expressed the Ly12.1 antigen-C57L/J, C57BR/cd, and SWR/J-while all other 34 inbred and congenic strains tested were Ly-12.1(Table 1). This pattern is identical with that previously reported for Ly-12.1 using conventional alloantisera (Potter and McKenzie 1981). The tissue distribution of the Ly-12.1 antigen was determined by rosetting with sheep-mouse Ig (SAMIg)-coated crythrocytes (Parish and McKenzie 1978) following titration of antiLy-12.1 ascites fluid. The occurrence of Ly-12.1 + cells in C57L lymphoid tissues suggested a Trather than a Bcell distribution (Table 2): 90% of thymocytes were Ly-12.1 + (titer 10-5), as were 30% of spleen cells and 64% of lymph node cells (titers 10-6). When T-cell rich populations were examined (prepared as described by Parish et al. 1974), the number of Ly-12. l + cells was similar to the number of Thy-1 + cells: Iglymph node cells (>90% Thy-1 +) contained 88% Ly-12.1 + cells and Igspleen cells (82% Thy-1 +) were 80% Ly-12.1 +. Ly-12.1 was absent from the majority of Thy-1Ig +

GENETIC AND BIOCHEMICAL CHARACTERIZATION OF ANTIGENS ENCODED BY THE LY‐24 (Pgp‐1) LOCUS1

Three allele‐specific monoclonal antibodies to Pgp‐1 (Ly‐24) were used to biochemically characterize the cell surface structures with which they reacted and to map the gene(s) coding for these antigens. The targets of these three monoclonal antibodies (mAb) were shown to be encoded by a gene situated on chromosome 2 close to β2m [gene order (Pgp‐1‐β2m‐a)] and no recombination between the loci detected by the three antibodies was revealed by genetic analysis. The genetic mapping of loci and tissue distribution of these antigens suggested that they might all correspond to a particular allelic form of the mouse phagocyte glycoprotein‐1 (Pgp‐1) antigen. Biochemical and serological analysis confirmed that this was indeed the case and revealed that all three mAbs were directed to one epitope. It is surprising that the tissue distribution defined by one mAb (Ly‐24A) was different from that for the two other (Ly‐24B) antibodies, despite the serological and biochemical identity of their respective targets. The possible reason for this unusual finding is discussed.

Expression of Qa alloantigens on peripheral T cells: the relationship of the Qa-m2, 7, 8, 9 specificities

The Qa loci are located between the H-2D and Tla loci, and they code for cell-surface alloantigens expressed on some lymphoid and myeloid cells (Stanton and Boyse 1976, Flaherty 1976, Sullivan and Flaherty 1979, Kincade et al. 1980). The protein products of the Qa-1 and Qa-2 loci resemble H-2K and H-2D antigens, having approximately the same molecular size, and by their association with/32m (Kincade et al. 1980, Michaelson et al. 1981). Serological investigations using several monoclonal antibodies directed to products of Qa loci have revealed that the mouse strain and tissue distributions of their antigenic targets are different; this suggests that each monoclonal antibody may identify the product of a different gene. Several such monoclonal antibodies, each defining distinct but partially overlapping cell types, have been described (H/immerling et al. 1979, Forman et al. 1982), including four from this laboratory (Hogarth et al. 1982, Sandrin et al. 1983, Sutton et al. 1983). These latter antibodies (Qa-m2, m7, m8, and m9) all react with peripheral T cells and some bone marrow cells, and on this basis they appear to be monoclonal representatives of the alloantisera used to define the Qa-2 locus (Flaherty 1976) and several new loci. Certain strains of mice were shown to exhibit selective reactivity when tested with these monoclonal Qa antibodies (Sandrin et al. 1983). Thus, A/HySn strain mice failed to react with Qa-m8 but they did react with Qa-m2 and Qa-m7. Furthermore, even within a single mouse strain capable of expressing all four Qa specificities, markedly different numbers of peripheral T cells were shown to react with each monoclonal antibody. These results suggested that the epitopes defined by Qa-m2, Qam7, and Qa-m8 could possibly reside on biochemically distinct molecules. Such a subdivision of the Qa loci has al-

Cross-linking of Ly 6-linked alloantigens: association between ThB and Ly 5

The bifunctional cross‐linking reagent dithiobis (succinimidyl propionate) (DSP) was used to cross‐link 125I surface‐labelled glycoproteins from viable thymocytes. The cells were solubilized, and the cross‐linked material immunoprecipitated and analysed by SDS‐PAGE. When DSP cross‐linked thymocyte material was immunoprecipitated with either anti‐ThB or anti‐Ly 5 monoclonal antibodies, and then cleaved, molecules with masses identical to Ly 5 (Mr180 kD) and ThB (Mr 16–18 kD) were obtained. However, if the cross‐linker was not cleaved, the intact product had a molecular mass of > 200 kD. The identity of these co‐precipitated, cross‐linked moieties was formally proved by limited proteolysis peptide map analysis. The data indicated that the ThB and Ly 5 antigens were associated on the thymocyte cell surface but no such association could be found on peripheral lymphocytes. The ThB‐Ly 5 interaction may indicate an association relevant to the differentiation of thymocytes.

THE EXPRESSION OF MURINE ALLOANTIGENS ON BLOOD LYMPHOCYTES

Single‐ and two‐colour immunofluorescence was used to analyse and compare the expression of 12 antigens on the surface of Ig− mouse blood lymphocytes (BL) and Ig− lymph node (LN) cells. Studies in different strains of mice showed that: (i) there were fewer Thy‐1+, Ly 1+, L3T4+ cells in BL compared to LN; (ii) Ly 2+ BL showed a unique fluorescence profile with a temporal variation in antigen density not evident in LN; (iii) Thy‐1− Ly 1−cells were more common in BL than LN; (iv) L3T4 and Ly 2 were present on mutually exclusive subpopulations in BL; (v) Ly 6A, (Ly 6.2), Ly 6C (Ly 28.2) Ly 28.6C and Ly 12.1 antigenic determinants were expressed on the same proportion of BL and LN cells and to the same level; (vi) Ly 24 (Pgp‐1) was the only alloantigen examined where the number of positive cells was increased in BL (65%) compared to LN (40%); (vii) Ly 5 and Ly 15 (LFA‐1) showed significant differences in antigen density distribution between BL and LN; (viii)Ly 21.2 was similar to Ly 15.2 expression; (ix) 20% of Ig−LN cells were Ia+, but la was absent from Ig−BL. Thus, BL differ in antigen distribution and density from lymphocytes in LN and other tissues and should be considered as a unique population of lymphocytes.