HLA II The emergence of the molecular model for the human major histocompatibility complex.

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.