- Email: [email protected]
Two forms of HLA class I molecules in human plasma
Abstracts
163
CTL clones by stimulating selected responders in vito with different B27 variants. Two informative CTL clones designated CTL 4-47 and CTL 4-55 were generated against a cell expressing B27a. CTL 4-47 is specific for only B27a and Bw67, and does not recognize B27b, B27c, B27d, B27e, B27f, or B27D. CTL 4-45 is specific for B27a and B27d and also recognizes some but not all cells expressing B27e and B27f. Both clones are nonreactive with other antigens in that B27 CREG (B7,B40,B47,Bw22). These data indicate that there are at least two functionally distinct T-cell defined epitopes on B27a, in addition to five distinct epitopes defined by MoAbs. Thus, the B27 class I molecule and presumably related variants, are antigenically complex structures expressing multiple serologically defined and T-cell defined recognition sites. PRIMARY STRUCTURAL DIFFERENCES IN HLA-B27 VARIANTS: IMPLICATION FOR MAPPING SEROLOGIC DETERMINANTS. S.Y. Choo, S. Rojo, J.A. Lopez de Castro, and J.A. Hansen; Puget Sound Blood Center, Fred Hutchinson Cancer Research Center; University of Washington, Seattle, WA; and Fundacion Jimenez Diaz HLA-B27 is a family of alloantigens consisting of at least seven different variants that share serological specificities. The primary structural analysis of the three B27 variants B27.1, B27.2, and B27.3 has revealed a limited number of amino acid (aa) differences involving two or three substitutions in the a l and a2 domains. Although these differences are minimal, they result in changes that can be distinguished by monoclonal antibodies (MoAbs) and alloantiserum (see table). Preliminary comparative tryptic peptide mapping studies of B27f suggest three aa substitutions in peptide fragments 71-83. Substitutions in this region have also been identified for B27.2 (positions 77, 80, and 81) and B27.3 (77 and 152). B27.2 does not react with MoAb B27M2 and B27.3 reacts weakly with B27M2, suggesting that residue 77 might be critical for recognition by B27M2. B27f reacts strongly with alloserum KC-MJA but does not react with MoAb B27M2 (B27.2 does not react with KC-MJA). In contrast, the variant B27d reacts strongly with MoAb B27M2 but does not react with KC-MJA. As additional sequence data are available for B27f and B27d, it should be possible to locate the primary structure recognized by these serologic reagents.
B27 variants
MoAb
MoAb
Alloserum
SEA ID a
B27M2
P56.1
KC-MJA
B27a B27b/c
(B27.1, B27W) b
+
(B27.3, B27C)
+
+
+ +
+
+
+
+
-
+
+
+
B27d
(-)
B27e
(B27.2, B27K)
-
+
+
+
B27f
(-)
-
+
+
+
(B27.4, B27D)
+
-
+
+/+
-
+
+
+
+
+
+
aB27 variants identified by isoelectric focusing, bEquivalent.
TWO FORMS OF HLA CLASS I MOLECULES IN HUMAN PLASMA. Michael S. Krangel; Dana-Farber Cancer Institute, Boston, MA Previous studies of plasma HLA molecules have indicated that they are primarily lipid associated. Further, it has been shown that HLA-A9 is present in plasma at
164
Abstracts levels higher than any other allele examined. More recently, it has been demonstrated that class I molecules are secreted at low levels from PBL in water soluble form. These molecules are 39 kd, as opposed to 43 kd, and arise by an alternative splicing mechanism which deletes exon five, encoding the transmembrane hydrophobic region, from a fraction of class I-encoding transcripts. Strikingly, this was observed to occur for HLA-A24 (a split of HLA-A9) at levels higher than any other allele examined. In order to reconcile these observations with previous studies, immunoaffinity chromatography was used to purify class I molecules from the plasma of an HLA-A24 positive individual. The concentration of class I molecules was determined to be 0.5 /~g/ml, based upon silver staining. Both 43 and 39 kd forms were detected, in a ratio of roughly 2 : 1. By the criteria of ultracentrifugal flotation and Triton X-114 binding, the former was lipid associated, whereas the latter was not. Thus, the 43 kd form has properties typical of shed, membrane molecules, whereas the 39 kd form has properties identical to the secreted forms generated by alternative splicing. Similar analysis of a panel of individuals indicated levels of shed and secreted molecules which varied from individual to individual. However, it was striking that both shed and secreted forms were highest in all HLA-A24 positive individuals tested. Since the two forms are presumed to arise by completely independent mechanisms, this result is difficult to understand. It may suggest an unusual selection for or stability of HLA-A24 in plasma. However, a physiological role for these molecules has yet to be revealed.
SEROLOGICAL AND BIOCHEMICAL ANALYSIS OF HLA B 15 AND B5 COMPLEXES. Bill Clark, Min Sik Park, Richard Tonai, and Paul I. Terasaki; UCLA Tissue Typing Laboratory, Los Angeles, CA Structural analysis of HLA-B15 and HLA-B5 complexes was conducted by onedimensional isoelectric focusing (IEF, method of Yang). Frozen peripheral blood lymphocytes were metabolically labeled with 35S methionine. The class I antigens were immunoprecipitated with monoclonal antibody 4E, which detects a determinant shared by HLA-B locus and Aw19-complex antigens. The desialated IEF banding patterns were correlated to microcytotoxicity data from a panel of donors from a variety of racial groups. Serological analysis of HLA-B 15 complex allosera indicated the presence of a number of B15 variants: BW62, BW63, BW46, BW70, TE78, and TE79. Analysis of the gels clearly showed polymorphism in the B 15-complex. Cells with BW70 and TE78 showed bands distinct from other B 15-typed cells. This study presents biochemical evidence for these serologically observed variants. Investigated antigens include HLA-B35, 51, 52, 62, 63, 46, W70, W71, W72, TE76, TE78, and TE79. RESTRICTED GENE POLYMORPHISM IN THE MHC OF THE RAT. Donald V. Cramer; Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, PA The extensive gene polymorphism of the major histocompatibility complex (MHC) has been considered to be a reflection of selective pressures on the function of the MHC during evolution. There is, however, substantial variation in the level of polymorphism among species, suggesting that selection may not be the mechanism maintaining the genetic variation. Our previous studies on the MHC of the rat (RTl) have demonstrated that the rat MHC is less polymorphic than the mouse. We have recently examined the relationship between this restricted level
163
CTL clones by stimulating selected responders in vito with different B27 variants. Two informative CTL clones designated CTL 4-47 and CTL 4-55 were generated against a cell expressing B27a. CTL 4-47 is specific for only B27a and Bw67, and does not recognize B27b, B27c, B27d, B27e, B27f, or B27D. CTL 4-45 is specific for B27a and B27d and also recognizes some but not all cells expressing B27e and B27f. Both clones are nonreactive with other antigens in that B27 CREG (B7,B40,B47,Bw22). These data indicate that there are at least two functionally distinct T-cell defined epitopes on B27a, in addition to five distinct epitopes defined by MoAbs. Thus, the B27 class I molecule and presumably related variants, are antigenically complex structures expressing multiple serologically defined and T-cell defined recognition sites. PRIMARY STRUCTURAL DIFFERENCES IN HLA-B27 VARIANTS: IMPLICATION FOR MAPPING SEROLOGIC DETERMINANTS. S.Y. Choo, S. Rojo, J.A. Lopez de Castro, and J.A. Hansen; Puget Sound Blood Center, Fred Hutchinson Cancer Research Center; University of Washington, Seattle, WA; and Fundacion Jimenez Diaz HLA-B27 is a family of alloantigens consisting of at least seven different variants that share serological specificities. The primary structural analysis of the three B27 variants B27.1, B27.2, and B27.3 has revealed a limited number of amino acid (aa) differences involving two or three substitutions in the a l and a2 domains. Although these differences are minimal, they result in changes that can be distinguished by monoclonal antibodies (MoAbs) and alloantiserum (see table). Preliminary comparative tryptic peptide mapping studies of B27f suggest three aa substitutions in peptide fragments 71-83. Substitutions in this region have also been identified for B27.2 (positions 77, 80, and 81) and B27.3 (77 and 152). B27.2 does not react with MoAb B27M2 and B27.3 reacts weakly with B27M2, suggesting that residue 77 might be critical for recognition by B27M2. B27f reacts strongly with alloserum KC-MJA but does not react with MoAb B27M2 (B27.2 does not react with KC-MJA). In contrast, the variant B27d reacts strongly with MoAb B27M2 but does not react with KC-MJA. As additional sequence data are available for B27f and B27d, it should be possible to locate the primary structure recognized by these serologic reagents.
B27 variants
MoAb
MoAb
Alloserum
SEA ID a
B27M2
P56.1
KC-MJA
B27a B27b/c
(B27.1, B27W) b
+
(B27.3, B27C)
+
+
+ +
+
+
+
+
-
+
+
+
B27d
(-)
B27e
(B27.2, B27K)
-
+
+
+
B27f
(-)
-
+
+
+
(B27.4, B27D)
+
-
+
+/+
-
+
+
+
+
+
+
aB27 variants identified by isoelectric focusing, bEquivalent.
TWO FORMS OF HLA CLASS I MOLECULES IN HUMAN PLASMA. Michael S. Krangel; Dana-Farber Cancer Institute, Boston, MA Previous studies of plasma HLA molecules have indicated that they are primarily lipid associated. Further, it has been shown that HLA-A9 is present in plasma at
164
Abstracts levels higher than any other allele examined. More recently, it has been demonstrated that class I molecules are secreted at low levels from PBL in water soluble form. These molecules are 39 kd, as opposed to 43 kd, and arise by an alternative splicing mechanism which deletes exon five, encoding the transmembrane hydrophobic region, from a fraction of class I-encoding transcripts. Strikingly, this was observed to occur for HLA-A24 (a split of HLA-A9) at levels higher than any other allele examined. In order to reconcile these observations with previous studies, immunoaffinity chromatography was used to purify class I molecules from the plasma of an HLA-A24 positive individual. The concentration of class I molecules was determined to be 0.5 /~g/ml, based upon silver staining. Both 43 and 39 kd forms were detected, in a ratio of roughly 2 : 1. By the criteria of ultracentrifugal flotation and Triton X-114 binding, the former was lipid associated, whereas the latter was not. Thus, the 43 kd form has properties typical of shed, membrane molecules, whereas the 39 kd form has properties identical to the secreted forms generated by alternative splicing. Similar analysis of a panel of individuals indicated levels of shed and secreted molecules which varied from individual to individual. However, it was striking that both shed and secreted forms were highest in all HLA-A24 positive individuals tested. Since the two forms are presumed to arise by completely independent mechanisms, this result is difficult to understand. It may suggest an unusual selection for or stability of HLA-A24 in plasma. However, a physiological role for these molecules has yet to be revealed.
SEROLOGICAL AND BIOCHEMICAL ANALYSIS OF HLA B 15 AND B5 COMPLEXES. Bill Clark, Min Sik Park, Richard Tonai, and Paul I. Terasaki; UCLA Tissue Typing Laboratory, Los Angeles, CA Structural analysis of HLA-B15 and HLA-B5 complexes was conducted by onedimensional isoelectric focusing (IEF, method of Yang). Frozen peripheral blood lymphocytes were metabolically labeled with 35S methionine. The class I antigens were immunoprecipitated with monoclonal antibody 4E, which detects a determinant shared by HLA-B locus and Aw19-complex antigens. The desialated IEF banding patterns were correlated to microcytotoxicity data from a panel of donors from a variety of racial groups. Serological analysis of HLA-B 15 complex allosera indicated the presence of a number of B15 variants: BW62, BW63, BW46, BW70, TE78, and TE79. Analysis of the gels clearly showed polymorphism in the B 15-complex. Cells with BW70 and TE78 showed bands distinct from other B 15-typed cells. This study presents biochemical evidence for these serologically observed variants. Investigated antigens include HLA-B35, 51, 52, 62, 63, 46, W70, W71, W72, TE76, TE78, and TE79. RESTRICTED GENE POLYMORPHISM IN THE MHC OF THE RAT. Donald V. Cramer; Department of Pathology, University of Pittsburgh School of Medicine, Pittsburgh, PA The extensive gene polymorphism of the major histocompatibility complex (MHC) has been considered to be a reflection of selective pressures on the function of the MHC during evolution. There is, however, substantial variation in the level of polymorphism among species, suggesting that selection may not be the mechanism maintaining the genetic variation. Our previous studies on the MHC of the rat (RTl) have demonstrated that the rat MHC is less polymorphic than the mouse. We have recently examined the relationship between this restricted level