HLA II The emergence of the molecular model for the human major histocompatibility complex.код для вставкиСкачать
YEARBOOK OF PHYSICAL ANTHROPOLOGY 28:79-95 (1985) HLA II: The Emergence of the Molecular Model for the Human Major Histocompatibility Complex ROBERT C. WILLIAMS Department of Anthropology, Arizona State Uniuersity, Tempe, Arizona 85281 and Histocompatibility Laboratory, Blood Systems Incorporated Scottsdale, Arizona 85257 KEY WORDS Allele, Locus, Protein, Intron, Exon, Gene family, Antigen, Monoclonal, Polyclonal ABSTRACT The known complexity of the HLA system continues to grow. Simultaneously, investigators are constructing a molecular model for the human major histocompatibility complex that accounts for the extensive variation at the HLA loci. Three alleles were given new WHO nomenclature at the HLA-A locus, 7 at HLA-B, 7 at HLA-D, and 6 at HLA-DR. Two entirely new loci were given WHO status: HLA-DQ and HLA-DP. The HLA-A, -B, and -C loci (class I genes) code for a polypeptide chain containing cytoplasmic, transmembrane, and extracellular regions. In turn the extracellular region can be divided into three domains. The molecular organization of the class 1 gene is discussed. The class I1 genes in the HLA-D/DR complex can be grouped into three regions, DR, DC, and SB, with one WHO-defined locus associated with each: HLA-DR, HLA-DQ, and HLA-DP, respectively. Class I1 molecules have both an alpha- and a beta-chain, which can be divided into cytoplasmic, transmembrane, and extracellular regions. The extracellular region of both polypeptide chains has two domains. The molecular organization of the alpha- and beta-chain genes is discussed. Serological polymorphism for both the class I and I1 molecules has been located in the domains most distal to the cellular membrane. The domain closest to the membrane in both alpha- and beta-chains of class I1 molecules is conserved and exhibits homology with one another and with the constant regions of IgG molecules and with beta-2-microglobulin. Therefore, it is likely that the genes of the major histocompatibility complex have evolved from a primordial gene which has differentiated into at least four related families: HLA class I, HLA class 11, IgG immunoglobulin heavy and light chains, and beta-2-microglobulin, When discussing the major histocompatibility complex (HLA), arguably the most polymorphic and extensive human genetic system yet defined, one is confronted with a large array of baming nomenclature and, with the emergence of the new molecular model, with concepts that at first glance appear to have no relation. In a n earlier review, the basic serology, genetics, population genetics, and applications of the HLA system were described (Williams, 1982). The present review presents the 1984 WHO (World Health Organization) nomenclature that resulted from the 9th International HLA Workshop, a n introduction to the new HLA-DP and HLA-DQ loci, and a brief discussion of the molecular model that is emerging for the human histocompatibility complex. The IXth International HLA Workshop, which was held in Munich, West Germany, in May 1984, revealed two major trends, one familiar, one new (Albert et al., 1984). As in the many workshops of previous years, the HLA system grew in complexity and refinement. Many new alleles were added to the major loci while 0 1985 Alan R. Liss, Inc. YEARBOOK OF PHYSICAL ANTHROPOLOGY 80 [Vol. 28, 1985 not-a-few defined antigens lost their provisional workshop status. What characterized this gathering of HLA specialists as unique was the beginning of a molecular definition of the basis of the structure and polymorphism of the HLA products. The emergent recombinant DNA technology has permitted the analysis of the HLA genes a t the most primary level and is yielding a model for the organization and evolution of the major histocompatibility complex in humans. This essay presents the new serological and genetic observations in light of the model for the molecular structure of the HLA antigens. Before proceeding, a brief consideration of the general nomenclature and the organization of HLA genes is in order. Because of similar molecular configurations the genes in the major histocompatibility complex are divided into three classes: I, 11, and 111. Class I genes include the HLA-A, -B, and -C loci, while class I1 antigens include the products of genes of the HLA-D/DR region. Human complement components are the third class, about which nothing further will be said. The genes of the major histocompatibility complex are coded on human chromosome 6 (Fig. 1). Class I and I1 loci can be considered a gene family or cluster that has evolved from a common primordial locus. It is, along with the globin genes (Jeffreys, 1982), emerging as a model system for the evolution of gene families in humans. Class I molecules are heterodimers composed of a heavy chain of apparent molecular weight 44,000 associated with beta-2-microglobulin, which is not coded on chromosome 6. There might be as many as 30 or more different, related copies of these class I genes found on the sixth chromosome. The most extensively characterized are the HLA-A, -B, and -C loci. The organization and nomenclature of the class I1 genes are more complicated and require closer inspection. Part of the complication comes from the nomenclature Human ChrOmOIome 6 HLA DP SB Region >p -1ll11-1 2- 1111-11 Ip > D III-II111 HLA-DO HLA-DR DR-Region OC-Region OX DC 0 5 11111i1 111-1 0 5 1111-f1 111 18 28 11111-11111-11111 a D2D -11111- Fig. 1. Map of the human major histocompatibility complex on chromosome 6 (Bodmer, 1984; Dausset and Cohen, 1984). The class I genes are HLA-A, -B, and -C and are expressed as a series of proteins in noncovalent association with beta-2-microglobulin. Molecular analysis of the murine class I genes, a homologous protein, suggests that there might be as many as 30 copies of the human class I loci in this region which occur as discrete sets of exons and intervening sequences in the DNA. In addition to coding for the HLA-A, -B, and -C proteins, they might determine the expression of other class I molecules that have yet to be serologically defined. In addition there might occur among these, pseudogenes, that are no longer expressed because of the accumulation of point mutations. The class II proteins are coded by genes in the HLA-DIDR complex that can be subdivided into three loci with their associated regions: HLA-DR (DR region), HLA-DQ (DC), and HLA-DP (SB). There are at least two beta-chain and one alpha-chain genes in the DR region and two alpha and two beta genes in both DC and SB. These also occur as sets of exons and introns in the DNA of chromosome 6. Please see the text for details. HLA, THE MOLECULAR MODEL Williams] 81 itself, which does not yet have the necessary precision to account for all of the variation at the molecular level. For instance, four polymorphic genetic loci are now recognized by the WHO at the HLA-DLDR region: HLA-D, HLA-DR, HLA-DQ, and HLA-DP. Each class 11antigen is a heterodimer which is assembled from one alphaand one beta-chain, each of which is coded on chromosome 6. Yet traditional definitions of a locus have referred to “one gene, one protein.” Using this latter definition there are at least 12 loci in the HLA-DLDR region. Yet, not all appear to be polymorphic. In order to avoid confusion, three of the WHO loci will be discussed in terms of their associated regions, each region being a complex of alpha- and beta-genes. The HLA-DR locus is associated with the DR-region, the HLA-DQ locus with the DC region, and the HLA-DP locus with the SB region. HLA-D is probably in the DR region although the exact nature and location of this polymorphism is not yet determined. CLASS I GENES Nomenclature At the meeting of the HLA WHO nomenclature committee after the 1984 International Workshop (Bodmer et al., 19841, three new alleles were defined at the HLAA locus. HLA-Aw66 became the fourth split of the broad antigen HLA-A10. It had previously been called LN (Mesman et al., 1983) (Table 1). HLA-A28 was split into two antigens, HLA-Aw68 and HLA-Aw69. HLA-Aw69 is the first HLA specificity to be defined by a murine monoclonal antibody. Such serological reagents are mouse immunoglobulins that originate from one plasma cell that is fused with a mouse myeloma cell line (which confers immortality), and grown in culture. The daughter cells all produce a n immunoglobulin molecule that is uniform in structure and specificity. In contrast, antibodies in a normal alloantiserum are polyclonal. They have their origin from many different clones of cells, each producing an antibody with a slightly different, but related specificity that is directed toward a particular epitope of the sensitizing antigen. (An epitope is a part of a n antigen that is recognized by the antibody. It is determined by the chemical composition and the three-dimensional structure of the molecule. In proteins, such as the HLA antigens, it is not only the primary amino acid sequence that determines the epitope, but the secondary, tertiary, and quaternary structure as well.) In general, polyclonal antiTABLE 1. Alleles at the HLA-A locus, 1984 WHO nomenclature‘ Antigen 1. HLA-A1 2. HLA-A2 3. HLA-A3 4. HLA-A9 5. HLA-A10 6. HLA-A11 7. HLA-Awl9 8. HLA-A23(9) 9. HLA-A24(9) 10. HLA-A25(10) 11. HLA-A26(10) 12. HLA-A28 13. HLA-A29(w19) 14. HLA-A30(w19) 15. HLA-A3l(wl9) 16. HLA-A32(w19) 17. HLA-Aw33(w19) 18. HLA-Aw34(10) 19. HLA-Aw36 20. HLA-Aw43 21. HLA-Aw66(10) 22. HLA-Aw68(28) 23. HLA-Aw69(28) Previous equivalents LN A28 and negative to A2128 A2128 ‘The number in parenthesis is the “broad’ specificity. See text for details. 82 YEARBOOK OF PHYSICAL ANTHROPOLOGY [Vol.28, 1985 sera recognize more than one epitope while a monoclonal antiserum is very specific for only one epitope. The new antigen HLA-Aw69 is defined by a monoclonal antibody that reacts with a n epitope common to HLA-A2 and a subset of HLA-A28 antigens. HLA-Aw68 is defined as the complementary antigen to HLA-Aw69, that subset of HLA-A28 molecules that does not have the epitope in common with HLAA2 and therefore does not react with the monoclonal antibody. Five HLA-A antigens, A23, A24, A30, A31, and A32, lost their “w”, or workshop provisional status (Table 1). The nomenclature committee decided that a n antigen that had a correlation coefficient of .95 or better, for which antisera were readily available, would not be provisional. The correlation was done between local laboratories which had received a set of international workshop antisera. Each laboratory typed individuals for the HLA phenotype and sent the interpretations along with the raw data, the serological reactions. The data were pooled, analyzed, and interpreted again in Munich. If there was at least a .95 correlation between the interpretations of the local laboratories and those of the central analysis, the provisional status was dropped. Seven new antigens have been defined a t the HLA-B locus (Table 2). HLA-Bw64 and HLA-Bw65 are newly defined splits of the HLA-B14 antigen. HLA-Bw70 is a new broad antigen that is accompanied in the nomenclature by two splits, HLABw71 and HLA-Bw72. In addition to these, HLA-Bw67 and HLA-Bw73 also make their first appearance. Nine antigens a t this locus have lost their “w” provisional status: HLA-B16, B21, B35, B38, B39, B44, B45, B49, and B51. No new antigens were defined a t the HLA-C locus during the 1984 workshop (Table 3). However, the “w” designation has been retained for antigens HLA-Cwl, Cw2, Cw3, Cw4, Cw5, and Cw6 in spite of the fact that they are well characterized and that sera are readily available. The nomenclature of the HLA-C locus is similar to that of the complement system. The “w” was retained to avoid confusion. HLACw7 and HLA-Cw8 retain their provisional status. Molecular Organization and structure At the cell surface HLA-A, -B, and -C antigens are composed of two polypeptide chains, a glycosylated chain of 44,000 molecular weight associated noncovalently with a lighter chain of 12,000, beta-2-microglobulin. The heavy chain can be divided into three major regions, extracellular, transmembrane, and cytoplasmic (Fig. 2). The extracellular region can be divided into three domains, alpha-1, alpha-2, and alpha-3, that are numbered from the amino terminus of the protein to the cell membrane, alpha-1 being most distal to the membrane and alpha-3 being most proximal to it. There are two intrachain covalent disulfide bonds in this region that create loops in the heavy chain, one between the cysteine residues at positions 101 and 164 in the alpha-2 domain and the other between 203 and 259 in alpha-3 (Orr et al., 1979; Lopez de Castro et al., 1982). The molecular organization of the DNA that codes for the class I genes is reflective of the regions and domains of the heavy-chain protein (Malissen et al., 1982; Sood et al., 1984) and is represented in Figure 3. Exons are the portions of DNA that code for the primary structure (amino acid sequence) of the 44,000-dalton protein. There are seven exons for the class I heavy chain. The first codes for the leader sequence that is translated by messenger RNA (mRNA)but which is not part of the functional protein. Exons 2, 3, and 4 code for the alpha-1, alpha-2, and alpha-3 domains while exon 5 contains the information for the transmembrane and part of the cytoplasmic regions. Exons 6 and 7 code for the remaining portion of the cytoplasmic protein and the 3’ untranslated section of the messenger RNA. The exons are divided by sections of DNA called introns which are removed by splicing from the precursor RNA transcript to produce a processed mRNA that contains the information from the exons along with 5’ and 3’ untranslated regions (Darnell, 1983; Crick, 1979). This organization of DNA for class I genes in humans is very similar to that for the H-2 system in mice (Steinmetz and Hood, 1983; Hood et al., 1982). HLA, THE MOLECULAR MODEL Williams] TABLE 2. Alleles at the HLA-B locus, 1984 WHO nomenclature Antigen Previous equivalents 1. HLA-B5 2. HLA-B7 3. HLA-B8 4. HLA-B12 5. HLA-B13 6. HLA-€314 7. HLA-BE 8. HLA-B16 9. HLA-B17 10. HLA-B18 11. HLA-B21 12. HLA-Bw22 13. HLA-B27 14. HLA-B35 15. HLA-B37 16. HLA-B38(16) 17. HLA-B39(16) 18. HLA-B40 19. HLA-Bw41 20. HLA-Bw42 21. HLA-B44(12) 22. HLA-B45(12) 23. HLA-Bw46 24. HLA-Bw47 25. HLA-Bw48 26. HLA-B49(21) 27. HLA-Bw50(21) 28. HLA-B51(5) 29. HLA-Bw52(5) 30. HLA-Bw53 31. HLA-Bw54(w22) 32. HLA-Bw55(w22) 33. HLA-Bw56(w22) 34. HLA-Bw57(17) 35. HLA-Bw58(17) 36. HLA-Bw59 37. HLA-Bw60(40) 38. HLA-Bw61(40) 39. HLA-Bw62(15) 40. HLA-Bw63(15) 41. HLA-Bw64(14) 42. HLA-Bw65(14) 43. HLA-Bw67 44. HLA-Bw70 45. HLA-Bw7l(w70) 46. HLA-Bw72(w70) 47. HLA-Bw73 HLA-Bwl6 HLA-Bw21 HLA-Bw35 HLA-Bw44(12) HLA-Bw45(12) HLA-B14.1,(8w63) HLA-B14.2,(8~62) SN-2,Te90,(8~57),Te75 Bu+ SV,K5,Da(6),(8w59) Bu sv KA,IEH TABLE 3. Alleles at the HLA-C locus, 1984 WHO nomenclature 1. 2. 3. 4. 5. 6. 7. 8. HLA-Cwl HLA-Cw2 HLA-Cw3 HLA-Cw4 HLA-Cw5 HLA-Cw6 HLA-Cw7 HLA-Cw8 83 YEARBOOK OF PHYSICAL ANTHROPOLOGY 84 [Vol. 28, 1985 1 d ' Extracellular Region 'I Exon 3 Alpha-2 Domain 2 tExon 4 182 183 I Alpha-3 Domain 2 : 8-2 Microglobulin 274 275 Transmembrane Region (TM) Exon 5 Exon 6 313 314 324 325 Cytoplasmic Region [CT] Exon 7 COOH Fig. 2. This is a schematic representation of a class I protein for the HLA-A, -B, and -C loci. The heavy alpha-chain of the molecule penetrates the cytoplasmic membrane and can be divided into three regions: extracellular, transmembrane (TM), and cytoplasmic (CT). There are three extracellular protein domains, each coded by its own DNA exon. Domains alpha-2 and alpha-3 each have a disulfide bond and loop. The serological polymorphism at the class I loci is concentrated in the alpha-1 and alpha-2 domains. Alpha-3, the domain closest to the cell membrane, is most conserved and exhibits homology with beta-2-microglobulin, the alpha-2 and beta-2 domains of the class I1 molecules, and with the constant regions of human IgG immunoglobulins. Beta-2-microglobulin is noncovalently associated with the alpha-3 domain. The numbers are the positions of the amino acids that occur at the exons' boundaries and at the disulfide bonds in exons 3 and 4. Exon 1codes for the signal sequence which is cleaved from the primary translation product and does not form B part of the mature class I protein. See Figure 3 and the text for details. Polymorphism of class I molecules The extreme serological polymorphism characterizing the human class I genes has led to a search for a molecular correlate to this variation. A comparison of the amino acid sequences of the HLA-A2, HLA-A28, and HLA-B7 molecules has mapped regions of variability primarily to the alpha-1 and alpha-2 domains between residues 43 and 195 of the class I molecules (Orr et al., 1979; Lopez de Castro et al., 1982). Within this area there are three clusters that are most variable, amino acids 65-80, 105-116, and 177-194. One way to identify the regions of polymorphism is to compare the sequences of two antigens that are closely related to one another. HLAA2 and HLA-A28 are alleles at the A locus, whose products show a serological crossreaction; i.e., human antibodies that are made to an HLA-A28 antigen often have anti-HLA-A2 specificity. This suggests that the two antigens are very similar, and a comparison of their amino acid sequences confirms this (Lopez de Castro et al., 1982).In the extracellular region, 235 amino acids can be compared between the two molecules, and 225 (96%)are identical. The ten differences occur at positions 9, 66, 70, 71,72, 74,95, 107, 116, and 245. Seven of the ten fall within the clusters 6580 and 105-116, which implies that these two regions play a role in determining the antigenic, allelic, differences between the HLA-A2 and HLA-A28 molecules. HLA, T H E MOLECULAR MODEL Williams] E x o n i Exon? 5’ ss ai ----_-____-___-__ / Exon3 Exon4 a2 a3 Exon5 Yi ‘ 338 85 Exon6 CT CT Exon7 3’UT 3’ ___-_________ - lntrons ______ Flanking Regions Fig. 3. The exodintron organization of a class I molecule (adapted from Malissen et al., 1982).The exons contain the code for the primary structure of the protein. Introns are intervening sequences of DNA. The numbers represent the codons for the amino acids at the boundaries of the exons in the translated protein. At the 3’ (3-prime)end of the DNA is the 3’ untranslated region (3’UT)(shaded) which is transcribed into a processed mRNA molecule but which is not translated into protein. SS: signal sequence. TM: transmembrane region. CT: cytoplasmic region. See Figure 2 and the text for details. Molecular evidence from the H-2 system of the mouse, homologous to the HLA system in humans, indicates that there are multiple copies of the class I genes (Steinmetz et al., 1982; Hood et al., 1982; Steinmetz and Hood, 1983). Thirty-six class I genes, sets of exons and introns, were found in the BALB/c mou’se sperm DNA (Steinmetz et al., 1982). It is suggested that these clusters of genes are produced by homologous but unequal crossing-over. In addition, it is not known how many of these genes are expressed on the surface of murine cells. Point mutations have accumulated in this region which might have inactivated a number of them. When a gene is made nonfunctional by mutation, it is called a pseudogene. There is now a technique, called DNA-mediated gene transfer, that permits the isolation of one of these genes and its transfer to mouse cells in tissue culture which, if the gene is functional, will express the class I protein on their surface (Goodenow et al., 1982). In this manner the proportion of expressed to nonexpressed, nonfunctional, genes can be determined. Beta-2-microglobulin The lighter polypeptide of class I molecules is beta-2-microglobulin. It contrasts with the heavy, alpha-chain in a number of ways. First, it is the product of a structural gene on human chomosome 15(15q21-q22), not chromosome 6 (Goodfellow et al., 1975; Smith et al., 1975). Second, it is not polymorphic. The primary structure of human beta-2-microglobulin consists of a n invariant 99 amino acids listed in Table 4 (Suggs et al., 1981). A disulfide bond between the cysteine residues a t positions 25 and 80 gives the protein a looped configuration that is similar to the alpha-2 and alpha-3 regions of the class-I heavy chains. The exon-intron organization of mouse beta-2-microglobulin, a homologous protein, is shown in Figure 4 (Parnes and Seidman, 1982). It is also a protein of 99 amino acids and is determined by four exons. Exon 1 contains the information for the signal sequence and the first two amino acids of the mature protein. Exon 2 codes for 93 amino acids while exon 3 has the four remaining codons and the beginning of the 3’ untranslated region which finds its remaining expression in the fourth exon. Beta-2-microglobulin in mice and humans does not penetrate the cytoplasmic membrane and therefore does not have a transmembrane or cytoplasmic region. Instead it is noncovalently bound to the alpha-3 domain of the class I heavy chain (Yokoyama and Nathenson, 1983).It is a n essential part of the class-I antigens because, without it, the antigens of the H-2 system of the mouse and the HLA antigens in humans will not be expressed (Bodmer et al., 1978; Parnes and Seidman, 1982). YEARBOOK OF PHYSICAL ANTHROPOLOGY 86 [Vol. 28, 1985 TABLE 4. Primary structure of beta-2 microglobulin' Codon AA 1. ATC 4. ACT 7. ATT 10. TAC 13. CAT 16. GAG 19. AAG Ile Thr Ile TYr His Glu LYs Phe CYS Ser His ASP Val Leu GlY Ile Val Ser Ser LYS Ser Leu TYr Phe Thr ASP Ala Val Val Ser LY s LY s Are 22. mc 25. 28. 31. 34. 37. 40. 43. 46. 49. 52. 55. 58. 61. 64. 67. 70. 73. 76. 79. 82. 85. 88. 91. 94. 97. TGC TCT CAT GAC GTT CTG GGA ATT GTG TCA TCT AAG TCT CTC TAT TTC ACT GAT GCC GTG GTG TCA AAG AAG CGA Codon AA 2. 5. 8. 11. 14. 17. 20. 23. 26. 29. 32. 35. 38. 41. 44. 47. 50. 53. 56. 59. 62. 65. 68. 71. 74. CAG CCA CAG TCA CCA AAT TCA CTG TAT GGG CCA A'IT GAC AAG GAG GAA GAG GAC TTC GAC TTC 'ITG ACT ACC GAA 11. GAG 80. TGC 83. AAC 86. ACT 89. CAG 92. ATA 95. TGG 98. GAC Gln Pro Gln Ser Pro Asn Ser Leu TYr Gh Pro Ile ASP LYs Glu Glu Glu ASP Phe ASP Phe Leu Thr Thr Glu Glu CYS Asn Thr Gln Ile TrP ASD Codon 3. 6. 9. 12. 15. 18. 21. 24. 21. 30. 33. 36. 39. 42. 45. 48. 51. 54. 57. 60. 63. 66. 69. 12. 75. 78. 81. 84. 87. 90. 93. 96. 99. AA CGT AAG GTT CGT GCA GGA AAT AAT GTG Asn Val Phe Ser Glu Leu Asn Arg LYs His Leu Ser Trp Tyr TYr Glu l"I"l' TCC GAA TTA AAT AGA AAA CAT TTG AGC TGG TAT TAT GAA ccc Pro AAA TAT CGT CAC TTG LYs 5 r -4% His Leu Pro Val ASP Met ccc G'IT GAT ATG 'The primary structure presented in Table 4 was constructed from a cDNA clone isolated by Suggs et al. (1981).It differs in two positions from a sequence published by Cunningham et al. (1973): (1) Position 42 is aspartic acid (Asp) not asparagine (Asn);(2) there are 100 amino acids, not 99, an additional serine at position 67 of Cunningham et al.'s (1973) sequence. Exon 1 Exan2 Exon3 Exon 4 Fig. 4. The organization of the introns and exons of murine beta-2-microglobulin (adapted from Parnes and Seidman, 1982).Human and murine beta-2-microglobulin are homologous proteins each of which has 99 amino acids. In the mouse there are four exons and three intervening sequences. Exon 1 codes for the signal sequence (SS) and the first two amino acids, exon 2 for 93 amino acids, and exon 3 for the remaining 4 residues and the beginning of the 3' untranslated region (3'UT). Exon 4 codes for the remainder of the 3'UT (shaded). HLA, THE MOLECULAR MODEL Williams] 87 CLASS I1 GENES Nomenclature New loci in the HLA-D/DR region have been defined since the 1980 HLA Workshop, while the allelic variation a t the “old” loci, HLA-D and HLA-DR, has become greater. It is necessary to use the molecular organization of this region as a framework (Fig. 1)in order to understand it. There are four genetic loci recognized by the WHO: HLA-D, HLA-DR, HLA-DQ, and HLA-DP. HLA-DR is associated with the DR region, HLA-DQ with the DC region, and HLA-DP with the SB region. The nomenclature for the HLA-D locus is presented in Table 5. There are seven new alleles, HLA-Dw13 through HLA-Dwl9. Typing for specificities at this locus is done by the mixed lymphocyte culture assay (MLC) using homozygous typing cells as stimulators (Williams, 1982). These alleles have yet to be associated with a specific molecular product. However, HLA-D alleles are associated with certain HLA-DR antigens, which are determined by a different assay, the serological microlymphocytotoxicity test. The nature of this association between HLA-D and HLADR specificities and the molecular basis for it are yet to be elucidated. The HLA-DR locus has six new alleles, four from new serological specificities that were defined by the 9th workshop and two from the renaming of older specificities (Table 6). HLA-DRwll and DRwl2 are new splits of HLA-DR5, while HLA-DRwl3 and DRwl4 are splits of a very complex antigen that still has provisional status, HLA-DRw6. HLA-DRw52 and DRw53, in contrast, are the new nomenclature for MT2 and MT3, respectively, which were first defined in the 7th and 8th workshops and about which there was much discussion. It was not known whether these MT specificites were supertypic antigens analogous to HLA-Bw4 and HLA-Bw6 a t the HLA-B locus or whether they represented alleles at their own locus. Molecular characterization of the HLA-DR region demonstrates that these two specificities are found on a n HLA-DR-homologousbeta-protein and, strictly speaking, are alleles at a distinct locus. The other HLA-DR antigens are located on a separate beta-protein. However, because of the close linkage and homology of the two HLA-DR betaproteins, the MT2 and MT3 specificities were given their own HLA-DR designation. More will be said about the molecular basis of HLA-DR antigens in the next section. In addition to HLA-DR, molecular analysis of the HLA-D region of human chromosome 6 has led to the definition of two other clusters of genes that were originally called DC (or DS) and SB. Each of these clusters now has a new locus associated with it. TABLE 5. Alleles at the H L A D locus, 1984 WHO nomenclature Previous equivalents Antigen 1. HLA-Dwl 2. HLA-Dw2 3. HLA-Dw3 4. HLA-Dw4 5. HLA-Dw5 6. HLA-Dw6 7. HLA-Dw7 8. HLA-Dw8 9. HLA-Dw9 10. HLA-DwlO 11. HLA-Dwll(w7) 12. HLA-Dwl2‘ -~ 13. HLA-Dwl3 14. HLA-Dw14 15. HLA-Dwl5 16. HLA-Dwl6 17. HLA-Dwl7(w7) 18. HLA-Dwl8(w6) 19. HLA-DwlS(w6) HLA-DR associations HLA-DR1 HLA-DR2 HLA-DR3 HLA-DR4 HLA-DRwll(5) HLA-DRw13(w6) HLA-DR7 HLA-DRw8 HLA-DRwl4(w6) H L A -DR4 ~~ DB3 LD40 DYT,YT DB8.BS 7A,(i)w7A) 6A,(Dw6A) 6B,(Dw6B) HL A-DR7 HLA-DR2 HLA-DR4 HLA-DR4 HLA-DR4 HLA-DRw1 4 ( ~ 6 ) HLA-DR7 HLA-DRwl3(w6) HLA-DRwl3lw6) YEARBOOK OF PHYSICAL ANTHROPOLOGY 88 [Vol. 28, 1985 TABLE 6. Alleles at the HLA-DR locus, 1984 WHO nomenclature Antieen Previous eauivalents 1. HLA-DR1 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. HLA-DR2 HLA-DR3 HLA-DR4 HLA-DR5 HLA-DRw6 HLA-DR7 HLA-DRw8 HLA-DRw9 HLA-DRwlO HLA-DRwll(5) HLA-DRwl26) HLA-DRwl3(w6) HLA-DRwl4(w6) HLA-DRw52 HLA-DRw53 LB5 LB5X8,DR5 short,FT23 6.6,6.1,62 6.9,6.3,6X,901 MT2 MT3 TABLE Z Alleles at the new HLA-DQ locus, 1984 WHO nomenclature Antigen Previous equivalents 1. €€LA-DQwl 2. HLA-DQw2 3. IILA-DQw3 TABLE 8. Alleles at the new HLA-DP locus, 1984 WHO nomenclature - Antieen Previous eauivalents 1. 2. 3. 4. 5. 6. SBl,PL3A SB2 SB3 SB4,PL3B SB5 SB6 HLA-DPwl HLA-DPw2 HLA-DPw3 HLA-DPw4 HLA-DPw5 HLA-DPw6 The HLA-DQ locus encompasses those alleles which have their molecular correlates in the DC region (Table 7). There are three alleles, HLA-DQwl, DQw2, and DQw3. They incorporate a number of antigens that have been serologically defined in previous workshops including MT1, MT4, MB1, MB2, and MB3. The precise nature of the HLA-DQ polymorphism in this region is not yet understood. In contrast to the serological definition of the HLA-DQ antigens, the six alleles at the new HLA-DP locus are defined by a cellular assay, the primed lymphocyte typing test (PLT)(Table 8, Fig. 5) (Zier and Bach, 1975; Fradelizi and Dausset, 1975; Sheehy et al., 1975; Sheehy and Bach, 1976; Bach et al., 1977; Shaw et al., 1980,1981). The assay depends upon the principle of allogeneic stimulation in a mixed lymphocyte response (MLR) (Yunis and Amos, 1971; Amos and Bach, 1968; Bach et al., 1969). It has been found that antigens in the HLA-D/DR region can stimulate allogeneic responder T-cells in tissue culture. Briefly, when whole suspensions of peripheral lymphocytes, both B- and T-cells, are put into culture together from two individuals, say X and Y, the proliferative response of the T-cells of each person will depend on the differences a t the HLA-D/DR region. If a person X is HLA-Dwl, Dw2; HLA-DRl,DR2 and Y is HLA-Dw1,Dwl; HLA-DR1,DRl then the T-cells of Y HLA, THE MOLECULAR MODEL Williams] 89 A Culture 1 I QBg 1 Cell 1 Stimulator (Mitomycin Treated) HLA-Al. A2.67.68, Dwl. Dw2, DRl DR2, DPwl 2 Cell 2 Responder HLA-A1. A 2 . 6 7 . 6 8 . Dwl. Dw2,DRl, DR2. DPwl. DPwl Cell Zis'primed'tothe HLA-DPw2 antigenof cell 1 6. Culture 2 3. Cell 3 Stimulator (Mitomycin Treated) HLA-A3, A 2 4 . 6 ~ 6 0 6. ~ 6 2Dw3, : Dw4; DR3. DR4. QB2DPw3 4 Cell 2 ' - Responder Primed to HLA-DPw2 I HLA-A1. A2,67,68. Dwl,Dw2, DRl. DR2: DWI.DPwl Cell 2' will respond to the HLA-DPw2 antigen of cell 3 in an accelerated fashion Fig. 5. The production of a primed stimulator cell for HLA-DPtyping. Please see the text for details. would "recognize" the HLA-Dw2 and HLA-DR2 antigens as different, which would stimulate clones of T lymphocytes from person Y to divide and go into lymphoblast transformation. However, the T-cells from X would not recognize a difference in those of Y because X has both the HLA-Dwl and HLA-DR1 antigens. The response of cells in culture can occur in one or two ways. If, as is usual in MLR tests, only a one-way test is desired, then the mitotic apparatus of one of the cells is inactivated by X-irradiation or mitomycin. The inactivated cell is called the stimulator. The cell that is still able to divide is called the responder. When the two cells are placed together in a culture vessel, the proliferative response is measured by the incorporation of tritiated thymidine in the DNA of the dividing T-cells which is measured in counts per minute by a beta-liquid-scintillation counter. The primed lymphocyte typing assay depends upon the accelerated response of T-lymphocytes that are stimulated, primed, in culture (Sheehy and Bach, 1976). When a responder first divides as a result of a difference in the HLA-D/DR region, the response time for maximum lymphoblast proliferation is about 6-7 days. However, after a cell has been primed by one stimulation, the response from a second stimulation by the same antigen is only about 2 days. In this way panels of primed lymphocyte-typing responder cells can be assembled like panels of alloantisera. Figure 5 illustrates the technique. The ideal combination of responder and stimulator is a pair that exhibits a single HLA-DP antigen difference. In the example in Figure 5, cell 2 will respond to the HLA-DPw2 antigen on cell 1. Clones of T-cells in the responder, cell 2, will go into lymphoblast transformation which will reach its maximum in about 7 days. These clones will then have a n accelerated proliferative response maximum in 2 days, whenever they are stimulated with the HLA-DPw2 antigen. Notice that in cell 3 of Figure 5 there are many differences at the HLA- 90 YEARBOOK OF PHYSICAL ANTHROPOLOGY [Vol.28, 1985 D/DR region. The reason that the primed cell (cell 2') can be a specific typing cell for HLA-DPw2, even in the presence of other antigens on cell 3 different from those on cell 2', is the accelerated nature of the secondary response in the primed cell. HLADPw2 on cell 3 of Figure 5 elicits a secondary response in the primed cell, cell 2', before the other antigens on cell 3 could have a chance to effect a primary stimulation in the primed responder cell. Using this technique panels of primed lymphocyte typing cells can be made for HLA-DP testing. Each cell to be typed would be X-irradiated or treated with mitomycin and used as a stimulator against a panel of primed responder cells for all of the HLA-DP antigens. The PLT assay is not used exclusively for the HLA-DP system because it can be used for the other class I1 antigens and even for the class I specificities (Zeevi and Duquesnoy, 1983). However, it has found its greatest utility as a routine assay in defining the alleles at the HLA-DP locus. Molecular organization and structure The class I1 molecules of the human major histocompatibility complex are heterodimers, molecules composed of two different proteins, one a heavy alpha-chain of approximately 35,000-dalton molecular weight and a lighter beta-chain of approximately 29,000 daltons (Fig. 6 )(Steinmetz and Hood, 1983). In contrast with the class I molecules, both the alpha- and beta-chains of class I1 proteins penetrate the cell membrane of B-cells, activated T lymphocytes, and the macrophages on which they are expressed. Like the class I heavy chain, the alpha- and beta-chains of the class I1 molecules can be divided for analysis into extracellular, transmembrane, and cytoplasmic regions. In addition, the extracellular region of each protein can be divided into two domains. The alpha-chain has one disulfide bond in the second domain whereas the beta-chain has a disulfide bond and loop in both the beta-1 and beta-2 domains. Each of the three class I1 regions on chromosome six has its own set of alpha- and beta-chains which are closely linked to one another. The individual regions differ in particulars and will be discussed in turn: HLA-DR, HLA-DQ (DC,DX), and HLA-DP (SB). There are one DR-alpha and a t least two DR-beta genes. Lee et al. (1982a,b) and Korman et al. (1982) have published the intron-exon organization and amino acid sequence of the HLA-DR alpha-chains. There are 229 amino acids; the extracellular region has 191, the transmembrane 23, and the cytoplasmic 15. The gene for the DRalpha chain is divided into five exons and four introns (Fig. 7). Exon 1 codes for the signal sequence and the first three amino acids. Exons 2 and 3 code for the alpha-1 and alpha-2 domains while exon 4 has the nucleotides for the connecting peptide, that portion of the protein that is between the alpha-2 domain and the membrane, and the transmembrane and cytoplasmic regions. The 3' untranslated region of the messenger RNA is coded by part of exon 4 and all of exon 5. It is interesting to note that, as in the heavy chain of class I molecules, the exon organization of the DNA mirrors the molecular structure of the protein. In the 5' to 3' direction, the direction of mRNA transcription, the amino acids of the DR-alpha chain are coded in sequence from the amino terminal end of the protein most distal to the membrane to the carboxy terminus in the cytoplasm. In addition, exons 2 and 3 code for the two protein domains that were defined by biochemists before the organization of the DNA was determined. Long et al. (1983) have described the complete sequence of a DR-beta chain from a cDNA clone that they call HLA-DR beta I. (A cDNA clone is a length of DNA nucleotides that have been made from a messenger RNA molecule by the action of a n enzyme called reverse transcriptase.) The protein has 237 amino acids. The two disulphide loops are formed by covalent bonds between residues 17 and 79 for the beta-1 domain and 117 and 173 in the beta-2 domain. Published intron-exon organization of the DC-beta genes suggests that the HLA-DR DNA has six exons and five introns (Larhanmar et al., 1983; Boss and Strominger, 1984). The two extracellular domains would be coded by exons 2 and 3. HLA, THE MOLECULAR MODEL Williamsj 91 , TM Cell Membrane Exon 4 Exon 4 CT 225 226 Fig. 6. A schematic representation of a class I1 protein. The DR-alpha chain is adapted from reports by Lee et al. (1982a,b). The beta-chain represented in the figure is that for the HLA-DC molecule as reported by Larhammar et al. (1983) and Boss and Strominger (1984). It is supposed that the usual arrangement will be a heterodimer consisting of one alpha- and one betachain from each system: DR-alpha and DRbeta, DC-alpha and DC-beta, and so forth. However, the different class I1 alpha-chains are similar in their primary structure and in their molecular organization (Auffray et al., 1984). This is true as well for the class I1 beta-chains. Therefore Figure 6 uses two recently and well-described molecules, a DR-alpha and DC-beta chain, to illustrate the general conformation of class 11 heterodimers. Each chain has two extracellular regions. Each of these in the beta-chain has a disulfide bond and loop whereas only the alpha-2 domain shares this feature. As with class I molecules, the allelic variation has been located in the more distal domains, alpha-1 and beta-1, while the domains closest to the cell membrane are conserved and are found to be homologous with beta-2-microglobulin, the alpha-3 region of class I molecules, and the constant regions of human IgG heavy and light chains. The numbers define the amino acids at the boundaries of the exons and those that participate in the disulfide bonds. TM:transmembrane region. CT: cytoplasmic region. See Figures 7 and 8 and the text for further details. The primary structure of an HLA-DQ alpha-chain from a cDNA clone called pDCHl has been published by Auffray et al. (1982). The DC-alpha protein has three more amino acids than the HLA-DR heavy chain, 232. It c m be divided into alpha1(1-87) and alpha-2 (88-181) domains, a connecting peptide (182-194), a transmembrane portion (195-2171, and a cytoplasmic region of 15 amino acids (218-232). In a subsequent study Auffray et al. (1983)found that the DC-alpha chain is polymorphic and that the variation corresponds to what were once thought to be supertypic specificities of the HLA-DR locus (Table 7). A schematic diagram of a DC-beta chain is presented in Figure 6. The primary structure and molecular organization of this DC molecule have been reported by Larhammar et al. (1983) and Boss and Stromin- YEARBOOK OF PHYSICAL ANTHROPOLOGY 92 Exon Exon 2 1 ss 01 [Vol. 28,1985 Exon3 Exon4 Exon 5 0 2 TM.CT,YUT 3' Ul 5' \ Flanking Regions 229 Fig. 7. A schematic representation of the exodintron organization of a class I1 alpha-chain (adapted from Lee et al., 1982a,b; Kaufman et al., 1984). There are five exons and four intervening sequences in the DNA and 229 amino acids in the primary structure of the translated protein. Exons 1 and 2 code for the alpha-1 and alpha-2 domains of the extracellular region. The numbers are the amino acid codons that define the boundaries of the exons. T M transmembrane. C T cytoplasmic. 3'UT 3' untranslated (shaded). Please see Figure 6 and the text for further details. Exon 1 EXOP2 ss 81 5' -.-. Exon 3 Exon 4 TM.CY Exan 5 [CYI 3 UT 3' __. Fig. 8. Schematic representation of the introdexon organization of class I1 HLA-DC beta-chain gene (adapted from Larhammar et al., 1983; Boss and Strominger, 1984). There are five exons and four intervening sequences in the DNA and 229 amino acids in the primary structure of the translated protein. The filled-in exon between exons 4 and 5 ([CY]), which would code for eight amino acids, is not expressed in the HLA-DC protein because of a point mutation in a splicing signal whereas it is probably expressed in the HLA-DR beta-chain which has 237 amino acids (Long et al., 1983).The numbers are the amino acid codons that define the boundaries of the exons. Exons 2 and 3 code for the beta-1 and beta-2 protein domains of the molecule. SS: signal sequence. TM: transmembrane region. CY: cytoplasmic region. 3'UT 3' untranslated region (shaded). Please see Figure 6 and text for further details. ger (1984). It has 229 amino acids, eight residues shorter than the HLA-DR light chain. Like HLA-DR-beta it has two extracellular domains, each with a disulfide bond and loop. The molecular organization of the DC-beta chain is shown in Figure 8. There are five exons and four introns in the DNA. Exon 1 codes for the first four amino acids, followed by exons 2 and 3 that contain the information for the two extracellular domains, and exons 4 and 5 which specify the transmembrane, cytoplasmic, and 3' untranslated regions of the mRNA. An alpha-protein closely related to DC-alpha has been described and is called DXalpha (Auffray et al., 1984). Boss and Strominger (1984)report that there are at least two DC-beta chains, suggesting that the second might be the DX-beta gene and that it will have a similar primary structure and molecular organization to the DC-beta molecule that is represented in Figures 6 and 8. Auffray et al. (1984) have also reported the primary structure of the heavy-chain protein at the HLA-DP system, the SB-alpha chain. Like HLA-DR it has 229 amino Williams] HLA, T H E MOLECULAR MODEL 93 acids and the two extracellular domains are coded by separate exons. Recently, Roux-Dosseto et al. (1983), Gorski et al. (1984), and Trowsdale et al. (1984) have published partial cDNA clone sequences for the SB-beta chains of the HLA-DP system. Its primary structure is 229 residues long and is analogous to the betachains in the HLA-DR and HLA-DQ molecules. Mapping of the HLA regions Molecular characterization of the HLA-DDR genes is very new. Therefore, the exact number of alpha and beta chains in each region is not known. However, it appears a s if the genes have evolved in alpha-beta sets by duplication and the map in this region of chromosome 6 reflects this fact (Fig. 1). At the HLA-DR region at least two beta-chains are present (Long et al., 1983).One of these will have the dassical serological polymorphism of the HLA-DR locus whereas the other might be polymorphic for supertypic specificities. Only one alphachain has been identified in the HLA-DR region and does not appear to be polymorphic (Spielman et al., 1984; Lee et al., 1982a,b; Korman et al., 1982). At the HLA-DQ region it is likely that there are two sets of alpha-beta genes, those for the DC and DX proteins. Gorski et al. (1984) and Trowsdale et al. (1984) have recently published maps of the HLA-DP region SB genes. It is a cluster of two alpha and two beta genes. In addition, Spielman et al. (1984) report a sixth alpha-chain in the HLA-D/DR region, the so-called DZ-alpha. As the molecular characterization of these loci becomes more complete, the exact numbers of proteins and their relative positions to one another will become clear. Class II allotypic and isotypic variation As with the class I molecules earlier, the question naturally arises: what is the nature of the extensive polymorphism at the class I1 loci? There are two components, one isotypic and one allotypic. For instance, the human genome contains two DClike alpha-chains in the HLA-DQ region, DC-alpha and DX-alpha. In the mouse the homologous locus in the H-2 system on chromosome 17 is the I-A-alpha gene. Auffray et al. (1983) report that they can find only one alpha gene in the I-A region. Therefore, it appears that humans have two but mice have only one. This is a n example of isotypic variation. Allotypic variation is the more familiar allelism within members of the same species. Variation in the human class I1 molecules appears to be concentrated in the domain most distal t o the cell membrane, either in alpha-1 or beta-1. Auffray et al. (1984) report that the major differences between two DC-alpha alleles, DC-1-alpha and DC-/i-alpha, are found in the alpha-1 domain of the protein. They differ at 18 of 82 amino acids in alpha-1, but at only three of 94 residues in alpha-2. Likewise, Boss and Strominger (1984) report that allelism in the DC-beta chains is most likely the result of amino acid differences in the beta-1 domain of the molecule. Evolution of the human HLA region The fact that the allotypic variation in the class I and I1 genes has been found in the alpha-1 and alpha-2 regions of the HLA-A, -B, and -C molecules and the alpha-1 and beta-1 domains of the class I1 genes implies that the domains near the cell membrane have been conserved during evolution, which has been shown to be the case (Auffray et al., 1982; Larhammar et al., 1982; Korman et al., 1982). The alpha3 domain of class I genes and the alpha-2 and beta-2 domains of the class I1 show significant homology with one another and with the constant regions of human IgG heavy and light chains and with beta-2-microglobulin. Current speculation has the human histocompatibility molecule evolving from one primordial gene that differentiated into a t least four systems: HLA class I, HLA class 11, immunoglobulin, and beta-2-microglobulin. The mechanisms for such evolution are unknown, but it has been suggested that duplication by unequal crossing-over, gene conversion, and traditional point mutation each plays a role (Boss and Strominger, 1984). As the 94 YEARBOOK OF PHYSICAL ANTHROPOLOGY rvoi. 28,1985 complexity of the molecular organization unfolds for the human major histocompatibility system more and more sophisticated models for its origin and amplification will begin to emerge. Its historical association with Ig molecules leads to fascinating theories about the function of HLA molecules and their role in the immune response, both now and in the past. Molecular data are now beginning to shed light on many old problems and will, no doubt, present more new ones for evolutionists in the years ahead. LITERATURE CITED Albert, ED, Baur, MP, and Mayr, WR (eds) (1984) Histocompatibility Testing 1984. Berlin: Springer Verlag. Amos, DB, and Bach, FH (1968)Phenotypic expressions of the major histocompatibility locus in man (HL-A):Leukocyte antigens and mixed leukocyte culture reactivity. J. Exp. Med. 128t623-637. Auffray, C, Korman, AJ, Roux-Dosseto, M, Bona, R, and Strominger, JL. (1982) cDNA clone for the heavy chain of the human B cell alloantigen DC1: Strong sequence homology to the HLA-DR heavy chain. Proc. Natl. Acad. Sci. USA 79:6337-6341. Auffray, C, Ben-Nun, A, Roux-Dosseto, M, Germain, RN, Seidman, JG, and Strominger, JL (1983) Polymorphism and complexity of the human DC and murine I-A alpha-chain genes. EMBO 2t121124. Auffray, C, Lillie, JW,Arnot, D, Grossberger, D, Kappes, D, and Strominger, JL (1984)Isotypic and allotypic variation of human class I1 histocompatibility antigen alpha-chain genes. Nature 308:327-333. Bach, FH, Albertini, RJ, Amos, DB, Ceppellini, K, Mattiuz, PL, and Miggiano, VC (1969)Mixed leukocyte culture studies in families with known HLA genotypes. Transplant Pioc. I:339-341. Bach, FH, Jarrett-Toth, EK, Benike, CJ, Shih, CY, and Valentine, EA (1977) Primed LD typing: Reagent preparation and definition of the HLA-D region antigens. Scand. J. lmmunol. 6:469-475. Bodmer, WF (1984)The HLA system, 1984. In ED Albert, MP Baur, and WR Mayr (eds): Histocompatibility Testing 1984. Berlin: Springer Verlag, pp. 11-22. Bodmer, WF, Jones, EA, Barnstabie, CJ, and Bodmer, JG (1978) Genetics of HLA: The major human histocompatibility system. Proc. R. SOC.Lond. LBio1.J202r93-116. Bodmer, WF, Albert, ED, Bodmer, JG, Dausset, J, Kissmeyer-Nielsen, F, Mayr, W, Payne, R, van Rood, JJ, Tmka, 2, and Walford, RL (1984) Nomenclature for factors of (,he HLA system 1984. Tissue Antigens 24:73-80. Boss, JM, and Strominger, JL (1984) Cloning and sequence analysis of the human major histocompatibility complex gene QC-3-beta. Proc. Natl. Acad. Sci. USA 81:5199-5203. Crick, F (1979) Split genes and RNA splicing. Science 204:264-271. Cunningham, BA, Wang, JL, Berggard, I, and Peterson, PA (1973) The complete amino acid sequence of beta-2-microglobulin. Biochemistry 12:4811-4821. Darnell, J E (1983)The processing of RNA. Sci. Am 249(4):90-100. Dausset, J, and Cohen, D (1984) HLA at the gene level. In ED Albert, MP Baur, and WR Mayr (eds): Histocompatibility Testing 1984. Berlin: Springer Verlag, pp. 22-28. Fradelizi, D, and Dausset, J (1975) Mixed lymphocyte reactivity of human lymphocytes primed in vitro. I. Secondary response to allogeneic lymphocytes. Eur. J. Immunol. 5395-301. Goodfellow, PN, Jones, E, Van Heyingen, V, Solomon, E, and Bobrow, M (1975) The beta-2-microglobulin gene is on chromosome 15 and not in the HL-A region. Nature 254:267-269. Goodenow, RS, McMillan, M, Orn, A, Nicholson, M, Davidson, N, Frelinger, JA, and Hood, L (1982) Identification of a BALBk H-2Ld gene by DNAmediated gene transfer. Science 225.577-679. Goodfellow, PN, Jones, E, Van Heyingen, V, Solomon, E, and Bobrow, M (1975) The beta-2-microglobulin gene is on chromosome 15 and not in the HL-A region. Nature 254r267-269. Gorski, J, Rollini, P, Long, E, and Mach, B (1984) Molecular organization of the HLA-SB region of the human major histocompatibility complex and evidence for two SB beta-chain genes. Proc. Natl. Acad. Sci. USA 81:3934-3938. Hood, L, Steinmetz, M, Goodenow, R, Eakle, K, Fisher, D, Kobori, J, Malissen, B, Malissen, M, McMillan, M, McNicholas, J, Om, A, Pecht, M, Sher, BT, Smith, L, Stroynowski, I, Sun, H, Winoto, A, and Zuniga, M (1982) Genes of the major histocompatibility complex. Cold Spring Harbor Symp. Quant. Biol. 47~1051-1065. Jeffreys, AJ (1982) Evolution of globin genes. In GA Dover and RB Flavell (eds): Genome Evolution. New York: Academic Press, pp. 157-176. Kaufman, JF,Auffray, C, Korman, AJ, Shackelford, DA, and Stominger, J (1984) The class 11 molecules of the human and murine major histocompatibility complex. Cell 361-13. Korman, AJ, Auffray, C, Schamboeck, A, and Strominger, A.(19821 The amino acid sequence and gene oreanization of the heavv chain of the HLATDR antigen: Homology to immunoglobulins. Proc. Natl. Acad. Sci. USA 79:6013-6017. Larhammar, D, Schenning, L, Gustafsson, K, Wiman, K, Claesson, L, Rask, L, and Peterson, PA (1982) Complete amino acid sequence of an HLADR antigen-like beta chain as predicted from the nucleotide sequence: Similarities with immunoglobulins and HLA-A, -B, and -C antigens. Prw. Natl. Acad. Sci. USA 79:3687-3691. Larhammar, D, Hyldig-Nielsen, JJ, Servenius, B, Anderson, G, Rask, L, and Peterson, PA (1983) Exon-intron organization and complete nucleotide sequence of a human major histocompatibility antigen DC-beta gene. Proc. Natl. Acad. Sci. USA 80t7313-7317. Lee, JS, Trowsdale, J, and Bodmer, WF (1982a) cDNA clones coding for the heavy chain of human HLA-DR antigen. Proc. Natl. Acad. Sci. USA 79:545-549. Lee, JS, Trowsdale, J, Travers, PJ, Carey, J, Grosveld, F, Jenkins, J, and Bodmer, WF (1982M Sequence of a n HLA-DR alpha-chain cDNA clone and intron-exon organization of the corresponding gene. Nature 299:750-752. Long, EO, Wake, CT, Gorski, J, and Mach, B (1983) Williams1 HLA, THE MOLECULAR MODEL 95 Complete sequence of an HLA-DR beta-chain de- Smith, M, Gold, P, Freedman, SO, and Shuster, J duced from a cDNA clone and identification of (1975) Studies of the linkage relationship of betamultiple non-allelic DR-beta chain genes. EMBO 2-microglobulin in man-mouse somatic cell hy2389-394. brids. Ann. Hum. Genet. 39t21-31. Lopez de Castro, JA, Strominger, JL, Strong, DM, Sood, AK, Pan, J, Biro, PA, Pereira, D, Reddy, VB, and Orr, HT (1982) Structure of crossreactive hu- Das, HK, and Weissman, SM (1984) Cloning and man histocompatibility antigens HLA-A28 and structure analysis of histocompatibility class I and HLA-A2: Possible implications for the generation class II genes. In A Kuniar (ed): Eukaryotic Gene of HLA polymorphism. Proc. Natl. Acad. Sci. USA Expression. New York: Plenum, pp. 87-108. 79:3813-3817. Spielman, RS, Lee, J, Bodmer, WF, Bodmer, JG, Malissen, M, Malissen, B, and Jordan, BR (1982) and Trowsdale, J (1984) Six HLA-D region alphaExonhntron organization and complete nucleotide chain genes on human chromosome 6: Polymorsequence of an HLA gene. Proc. Natl. Acad. Sci. phisms and associations of DC alpha-related seUSA 79:893-897. quences with DR types. Proc. Natl. Acad. Sci. USA Mesnian, B, de Lange, G, and Engelfriet, CP (1983) 8113461-3465. A new HLA-A antigen, called LN, closely related Steinmetz, M, Winoto, A, Minard, K, and Hood, L to A25, A26, and Aw34. Tissae Antigens 21.192(1982) Clusters of genes encoding mouse trans197. plantation antigens. Cell 28:489-498. Orr, HT, Lopez de Castro, JA, Parham, P, Ploegh, Steinmetz, M, and Hood, L (1983) Genes of the HL, and Strominger, J L (1979) Comparison of major histocompatibility complex in mouse and amino acid sequences of two human histocom- man. Science 222727-733, patibility antigens, HLA-A2 and HLA-B7- Loca- Suggs, SV, Wallace, RB, Hirose, T, Kawashima, tion of the putative alloantigenic sites. Proc. Natl. EH, and Itakura, K (1981) Use of synthetic oligoAcad. Sci. USA 76:4395-4399. nucleotides as hybridization probes: Isolation of Parnes, JR,and Seidman, J G (1982) Structure of cloned cDNA sequences far human beta-2 microwild-type and mutant mouse beta-2 microgiobuiin globulin. Roc. Natl. Acad. Sci. USA 78t6613-6617. genes. Cell 29:661-669. Trowsdale, J, Kelly, A, Lee, J, Carson, S, Austin, P, Roux-Dosseto, M, Auffray, C, Lillie, JW,Boss, JM, and Travers, P (1984) Linkage map of two HLACohen, D, DeMars, R. Mawas, C, Seidman, JG, SB-beta and two HLA-SB-alpha-relatedgenes: An intron in one of the SB-beta genes contains a proand Strominger, JL (1983) Genetic mapping of a cessed pseudogene. Cell 38:241-249. human class I1 antigen beta-chain cDNA clone to Williams, RC (1982) HLA. Yearbk. Phys. Anthrothe SB region of the HLA complex. Proc. Natl. pol. 25t91-112. Acad. Sci. USA 805036-6040. Shaw, S, Johnson, AH, and Shearer, GM (1980) Yokoyama, K, and Nathenson, S (1983) IntramoEvidence for a new segregant series of B cell an- lecular organization of Class I MHC antigens: Lotigens that are encoded in the HLA-D region and calization of the alloantigenic determinants and that stimulate secondary allogenic proliferative the beta-2-microglobulinbinding site to different regions of the H-2Kb glycoprotein. J. Immunol. and cytotoxic responses. J. Exp. Med. 152:56513Ot1419-1425. 580. Shaw, S, Kavathas, P, Pollack, MS, Charmot, D, Yunis, EJ, and Amos, DB (1971) Three closely and Mawas, C (1981) Family studies define a new linked genetic systems relevant to transplantahistocompatibility locus, SB, between HLA-DR tion. Proc. Natl. Acad. Sci. USA 68t3031-3035. Zeevi, A, and Duquesnoy, Rd (1983) PLT specificity and GLO. Nature 293345747. of alloreactive lymphocyte clones for HLA-B locus Sheehy, MJ, Sondel, PM, Bach, ML, Wank, R, and Bach, FH (1975) HL-A LD (lymphocyte definedj determinants. Proc. Natl. Acad. Sci. USA 80t14401444. typing: A rapid assay with primed lymphocytes. Zier, KS, and Bach, FH (1975) Secondary responses Science 188t1308-1310. Sheehy, MJ, and Bach, FH (1976) Primed LD typ- of human lymphocytes to alloantigens in vitro. ing (PLT)-Technical considerations. Tissue Anti- Scand. J. Immunol. 4t607-611. gens 8t157-171.