genes”; their current designation as histocompatibility-2
(H-2) genes was in reference to Gorer’s group II blood-group
antigens. Although Gorer died before his contributions were
recognized fully, Snell was awarded the Nobel prize in 1980
for this work.
The MHC Encodes Three Major
Classes of Molecules
The major histocompatibility complex is a collection of
genes arrayed within a long continuous stretch of DNA on
chromosome 6 in humans and on chromosome 17 in mice.
The MHC is referred to as the HLA complex in humans and
as the H-2 complex in mice. Although the arrangement of
genes is somewhat different, in both cases the MHC genes are
organized into regions encoding three classes of molecules
(Figure 7-1):
a73
Class I MHC genes encode glycoproteins expressed on
the surface of nearly all nucleated cells; the major
function of the class I gene products is presentation of
peptide antigens to T
C
cells.
chapter 7
a73 General Organization and Inheritance of the MHC
a73 MHC Molecules and Genes
a73 Detailed Genomic Map of MHC Genes
a73 Cellular Distribution of MHC Molecules
a73 Regulation of MHC Expression
a73 MHC and Immune Responsiveness
a73 MHC and Disease Susceptibility
Major
Histocompatibility
Complex
E
???? ????????? ??????? ??????? ?? ????
possesses a tightly linked cluster of genes, the ma-
jor histocompatibility complex (MHC), whose
products play roles in intercellular recognition and in dis-
crimination between self and nonself. The MHC partici-
pates in the development of both humoral and cell-
mediated immune responses. While antibodies may react
with antigens alone, most T cells recognize antigen only
when it is combined with an MHC molecule. Furthermore,
because MHC molecules act as antigen-presenting struc-
tures, the particular set of MHC molecules expressed by an
individual influences the repertoire of antigens to which that
individual’s T
H
and T
C
cells can respond. For this reason, the
MHC partly determines the response of an individual to
antigens of infectious organisms, and it has therefore been
implicated in the susceptibility to disease and in the devel-
opment of autoimmunity. The recent understanding that
natural killer cells express receptors for MHC class I antigens
and the fact that the receptor–MHC interaction may lead to
inhibition or activation expands the known role of this gene
family (see Chapter 14). The present chapter examines the
organization and inheritance of MHC genes, the structure of
the MHC molecules, and the central function that these
molecules play in producing an immune response.
General Organization and
Inheritance of the MHC
The concept that the rejection of foreign tissue is the result
of an immune response to cell-surface molecules, now called
histocompatibility antigens, originated from the work of
Peter Gorer in the mid-1930s. Gorer was using inbred strains
of mice to identify blood-group antigens. In the course of
these studies, he identified four groups of genes, designated
I through IV, that encoded blood-cell antigens. Work carried
out in the 1940s and 1950s by Gorer and George Snell estab-
lished that antigens encoded by the genes in the group desig-
nated II took part in the rejection of transplanted tumors
and other tissue. Snell called these genes “histocompatibility
Presentation of Vesicular Stomatitis Virus Peptide (top)
and Sendai Virus Nucleoprotein Peptide by Mouse MHC
Class I Molecule H-2K
b
8536d_ch07_161-184 8/15/02 8:41 PM Page 161 mac114 Mac 114:2nd shift:
a73
Class II MHC genes encode glycoproteins expressed
primarily on antigen-presenting cells (macrophages,
dendritic cells, and B cells), where they present processed
antigenic peptides to T
H
cells.
a73
Class III MHC genes encode, in addition to other
products, various secreted proteins that have immune
functions, including components of the complement
system and molecules involved in inflammation.
Class I MHC molecules encoded by the K and D regions in
mice and by the A, B, and C loci in humans were the first
discovered, and they are expressed in the widest range of
cell types. These are referred to as classical class I molecules.
Additional genes or groups of genes within the H-2 or HLA
complexes also encode class I molecules; these genes are
designated nonclassical class I genes. Expression of the non-
classical gene products is limited to certain specific cell
types. Although functions are not known for all of these
gene products, some may have highly specialized roles in
immunity. For example, the expression of the class I HLA-
G molecules on cytotrophoblasts at the fetal-maternal in-
terface has been implicated in protection of the fetus from
being recognized as foreign (this may occur when paternal
antigens begin to appear) and from being rejected by ma-
ternal T
C
cells.
The two chains of the class II MHC molecules are en-
coded by the IA and IE regions in mice and by the DP, DQ,
and DR regions in humans. The terminology is somewhat
confusing, since the D region in mice encodes class I MHC
molecules, whereas the D region (DR, DQ, DP) in humans
refers to genes encoding class II MHC molecules! Fortu-
nately, the designation D for the general chromosomal loca-
tion encoding the human class II molecules is seldom used
today; the sequence of the entire MHC region is available so
the more imprecise reference to region is seldom necessary.
As with the class I loci, additional class II molecules en-
coded within this region have specialized functions in the
immune process.
The class I and class II MHC molecules have common
structural features and both have roles in antigen processing.
By contrast, the class III MHC region, which is flanked by the
class I and II regions, encodes molecules that are critical to
immune function but have little in common with class I or II
molecules. Class III products include the complement com-
ponents C4, C2, BF (see Chapter 13), and inflammatory cy-
tokines, including tumor necrosis factor (TNF) and
heat-shock proteins (see Chapter 12).
162 PART II Generation of B-Cell and T-Cell Responses
VISUALIZING CONCEPTS
FIGURE 7-1 Simplified organization of the major histocompat-
ibility complex (MHC) in the mouse and human. The MHC is re-
ferred to as the H-2 complex in mice and as the HLA complex in
humans. In both species the MHC is organized into a number of
regions encoding class I (pink), class II (blue), and class III
(green) gene products. The class I and class II gene products
shown in this figure are considered to be the classical MHC mol-
ecules. The class III gene products include complement (CH11032) pro-
teins and the tumor necrosis factors (TNF-H9251 and TNF-H9252).
II III
Complex
MHC class
Region
Gene
products
IA
αβ
C′ proteinsH–2K H–2LH–2D
IE
αβ
TNF-α
TNF-β
TNF-α
TNF-β
H–2
DSIEIAK
III
Complex
MHC class
Region
Gene
products
DQ
αβ
C′ proteins HLA-AHLA-CHLA-B
DR
αβ
HLA
II I
ACBC4, C2, BFDRDQDP
Human HLA complex
Mouse H-2 complex
DP
αβ
8536d_ch07_161-184 8/16/02 12:09 PM Page 162 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
Allelic Forms of MHC Genes Are Inherited
in Linked Groups Called Haplotypes
As described in more detail later, the loci constituting the
MHC are highly polymorphic; that is, many alternative
forms of the gene, or alleles, exist at each locus among the
population. The genes of the MHC loci lie close together; for
example, the recombination frequency within the H-2 com-
plex (i.e., the frequency of chromosome crossover events
during mitosis, indicative of the distance between given gene
segments) is only 0.5%—crossover occurs only once in every
200 mitotic cycles. For this reason, most individuals inherit
the alleles encoded by these closely linked loci as two sets, one
from each parent. Each set of alleles is referred to as a haplo-
type. An individual inherits one haplotype from the mother
and one haplotype from the father. In outbred populations,
the offspring are generally heterozygous at many loci and will
express both maternal and paternal MHC alleles. The alleles
are codominantly expressed; that is, both maternal and pater-
nal gene products are expressed in the same cells. If mice are
inbred (that is, have identical alleles at all loci), each H-2 lo-
cus will be homozygous because the maternal and paternal
haplotypes are identical, and all offspring therefore express
identical haplotypes.
Certain inbred mouse strains have been designated as
prototype strains, and the MHC haplotype expressed by
these strains is designated by an arbitrary italic superscript
(e.g., H-2
a
, H-2
b
). These designations refer to the entire set of
inherited H-2 alleles within a strain without having to list
each allele individually (Table 7-1). Different inbred strains
may have the same set of alleles, that is the same MHC hap-
lotype, as the prototype strain. For example, the CBA, AKR,
and C3H strains all have the same MHC haplotype (H-2
k
).
The three strains differ, however, in genes outside the H-2
complex.
If two mice from inbred strains having different MHC
haplotypes are bred to one another, the F
1
generation inher-
its haplotypes from both parental strains and therefore ex-
presses both parental alleles at each MHC locus. For exam-
ple, if an H-2
b
strain is crossed with an H-2
k
, then the F
1
in-
herits both parental sets of alleles and is said to be H-2
b/k
(Figure 7-2a). Because such an F
1
expresses the MHC pro-
teins of both parental strains on its cells, it is histocompatible
with both strains and able to accept grafts from either
parental strain (see example in Figure 7-2b). However, nei-
ther of the inbred parental strains can accept a graft from the
F
1
mice because half of the MHC molecules will be foreign to
the parent.
The inheritance of HLA haplotypes from heterozygous
human parents is illustrated in Figure 7-2c. In an outbred
population, each individual is generally heterozygous at each
locus. The human HLA complex is highly polymorphic and
multiple alleles of each class I and class II gene exist. How-
ever, as with mice, the human MHC loci are closely linked
and usually inherited as a haplotype. When the father and
mother have different haplotypes, as in the example shown
(Figure 7-2c) there is a one-in-four chance that siblings will
inherit the same paternal and maternal haplotypes and
therefore be histocompatible with each other; none of the
offspring will be histocompatible with the parents.
Although the rate of recombination by crossover is low
within the HLA, it still contributes significantly to the diver-
sity of the loci in human populations. Genetic recombina-
tion generates new allelic combinations (Figure 7-2d), and
the high number of intervening generations since the ap-
pearance of humans as a species has allowed extensive re-
combination, so that it is rare for any two unrelated
individuals to have identical sets of HLA genes.
MHC Congenic Mouse Strains Are Identical
at All Loci Except the MHC
Detailed analysis of the H-2 complex in mice was made
possible by the development of congenic mouse strains. In-
bred mouse strains are syngeneic or identical at all genetic
loci. Two strains are congenic if they are genetically identical
Major Histocompatibility Complex CHAPTER 7 163
TABLE 7-1 H-2 Haplotypes of some mouse strains
H-2 ALLELES
Prototype strain Other strains with the same haplotype Haplotype KIAIE SD
CBA AKR, C3H, B10.BR, C57BR kkkkkk
DBA/2 BALB/c, NZB, SEA, YBR dddddd
C57BL/10 (B10)C57BL/6, C57L, C3H.SW, LP, 129 bbbbbb
A A/He, A/Sn, A/Wy, B10.A akkkdd
A.SW B10.S, SJL ssssss
A.TL t1 skk kd
DBA/1 STOLI, B10.Q, BDP qqqqqq
8536d_ch07_161-184 8/16/02 8:28 AM Page 163 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
(a) Mating of inbred mouse strains with different MHC haplotypes
b/b
b/b
b/b
b/bb/k
b/k
k/k
k/k
k/k
b/k
F
1
progeny (H-2
b/k
)
H-2
k
parentH-2
b
parent
Homologous chromosomes with MHC loci
(b) Skin transplantation between inbred mouse strains with same or
different MHC haplotypes
Parental recipient Skin graft donor
Parent
Progeny recipient
b/b
b/kb/k
k/k
Progeny
b/b
k/kb/k
k/k
Parent
(c) Inheritance of HLA haplotypes in a typical human family
Parents
Progeny
A/B C/D
B/RA/C A/D B/C B/D
(d) A new haplotype (R) arises from recombination
of maternal haplotypes
1 7 w3 2 1 1
ABC
HLA Alleles
DR DQ DP
2 8 w2 3 2 2
344w44 1 3
11 35 w1 7 3 4
3
A
Haplotypes
B
C
D
R44w47345e
2
FIGURE 7-2 (a) Illustration of inheritance
of MHC haplotypes in inbred mouse strains.
The letters b/b designate a mouse homozy-
gous for the H-2
b
MHC haplotype, k/k ho-
mozygous for the H-2
k
haplotype, and b/k a
heterozygote. Because the MHC loci are
closely linked and inherited as a set, the
MHC haplotype of F1 progeny from the mat-
ing of two different inbred strains can be pre-
dicted easily. (b) Acceptance or rejection of
skin grafts is controlled by the MHC type of
the inbred mice. The progeny of the cross be-
tween two inbred strains with different MHC
haplotypes (H-2
b
and H-2
k
) will express both
haplotypes (H-2
b/k
) and will accept grafts
from either parent and from one another.
Neither parent strain will accept grafts from
the offspring. (c) Inheritance of HLA haplo-
types in a hypothetical human family. In hu-
mans, the paternal HLA haplotypes are
arbitrarily designated A and B, maternal C
and D. Because humans are an outbred
species and there are many alleles at each
HLA locus, the alleles comprising the haplo-
types must be determined by typing parents
and progeny. (d) The genes that make up
each parental haplotype in the hypothetical
family in (c) are shown along with a new hap-
lotype that arose from recombination (R) of
maternal haplotypes.
8536d_ch07_161-184 8/16/02 12:09 PM Page 164 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
except at a single genetic locus or region. Any pheno-
typic differences that can be detected between congenic
strains are related to the genetic region that distinguishes
the strains. Congenic strains that are identical with each
other except at the MHC can be produced by a series of
crosses, backcrosses, and selections. Figure 7-3 outlines the
steps by which the H-2 complex of homozygous strain B
can be introduced into the background genes of homozy-
gous strain A to generate a congenic strain, denoted A.B.
The first letter in a congenic strain designation refers to the
strain providing the genetic background and the second
letter to the strain providing the genetically different MHC
region. Thus, strain A.B will be genetically identical to
strain A except for the MHC locus or loci contributed by
strain B.
During production of congenic mouse strains, a crossover
event sometimes occurs within the H-2 complex, yielding a
recombinant strain that differs from the parental strains or
the congenic strain at one or a few loci within the H-2
complex. Figure 7-4 depicts haplotypes present in several re-
combinant congenic strains that were obtained during pro-
duction of a B10.A congenic strain. Such recombinant
strains have been extremely useful in analyzing the MHC be-
cause they permit comparisons of functional differences
Major Histocompatibility Complex CHAPTER 7 165
F
2
a/ab/b
×
a/ba /b
×
Strain-A skin grafts
Cross
Interbreeding
Select for b/b
at H-2 complex
F
1
a/aa/ba/bb/b
Strain A
×
a/ba/b
×
Backcross
Interbreed, select, and
backcross for ≤ 10 cycles≤
Strain A
?
B
a/a
FIGURE 7-3 Production of congenic mouse
strain A.B, which has the genetic background of
parental strain A but the H-2 complex of strain B.
Crossing inbred strain A (H-2
a
) with strain B (H-2
b
)
generates F
1
progeny that are heterozygous (a/b)
at all H-2 loci. The F
1
progeny are interbred to pro-
duce an F
2
generation, which includes a/a, a/b,
and b/b individuals. The F
2
progeny homozygous
for the B-strain H-2 complex are selected by their
ability to reject a skin graft from strain A; any prog-
eny that accept an A-strain graft are eliminated
from future breeding. The selected b/b homozy-
gous mice are then backcrossed to strain A; the re-
sulting progeny are again interbred and their
offspring are again selected for b/b homozygosity
at the H-2 complex. This process of backcrossing
to strain A, intercrossing, and selection for ability to
reject an A-strain graft is repeated for at least 12
generations. In this way A-strain homozygosity is
restored at all loci except the H-2 locus, which is
homozygous for the B strain.
Strain
Parental
Congenic
Recombinant
congenic
A
B10
B10.A
B10.A (3R)
B10.A (2R)
B10.A (4R)
B10.A (18R)
H-2
haplotype
a
b
a
i3
h2
h4
i18
KA A E E S D
H-2 loci
ββα α
FIGURE 7-4 Examples of recombinant congenic mouse strains
generated during production of the B10.A strain from parental strain
B10 (H-2
b
) and parental strain A (H-2
a
). Crossover events within the
H-2 complex produce recombinant strains, which have a-haplotype
alleles (blue) at some H-2 loci and b-haplotype alleles (orange) at
other loci.
8536d_ch07_161-184 8/16/02 12:09 PM Page 165 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
between strains that differ in only a few genes within the
MHC. Furthermore, the generation of new H-2 haplotypes
under the experimental conditions of congenic strain devel-
opment provides an excellent illustration of the means by
which the MHC continues to maintain heterogeneity even in
populations with limited diversity.
MHC Molecules and Genes
Class I and class II MHC molecules are membrane-bound
glycoproteins that are closely related in both structure and
function. Both class I and class II MHC molecules have been
isolated and purified and the three-dimensional structures
of their extracellular domains have been determined by x-
ray crystallography. Both types of membrane glycoproteins
function as highly specialized antigen-presenting molecules
that form unusually stable complexes with antigenic pep-
tides, displaying them on the cell surface for recognition by
T cells. In contrast, class III MHC molecules are a group of
unrelated proteins that do not share structural similarity
and common function with class I and II molecules. The
class III molecules will be examined in more detail in later
chapters.
Class I Molecules Have a Glycoprotein Heavy
Chain and a Small Protein Light Chain
Class I MHC molecules contain a 45-kilodalton (kDa) H9251
chain associated noncovalently with a 12-kDa H9252
2
-microglob-
ulin molecule (see Figure 7-5). The H9251 chain is a transmem-
brane glycoprotein encoded by polymorphic genes within the
A, B, and C regions of the human HLA complex and within
the K and D/L regions of the mouse H-2 complex (see Figure
7-1). H9252
2
-Microglobulin is a protein encoded by a highly con-
served gene located on a different chromosome. Association
of the H9251 chain with H9252
2
-microglobulin is required for expres-
sion of class I molecules on cell membranes. The H9251 chain is
anchored in the plasma membrane by its hydrophobic trans-
membrane segment and hydrophilic cytoplasmic tail.
Structural analyses have revealed that the H9251 chain of class I
MHC molecules is organized into three external domains
(H92511, H92512, and H92513), each containing approximately 90 amino
acids; a transmembrane domain of about 25 hydrophobic
amino acids followed by a short stretch of charged (hy-
drophilic) amino acids; and a cytoplasmic anchor segment of
30 amino acids. The H9252
2
-microglobulin is similar in size and
organization to the H92513 domain; it does not contain a trans-
membrane region and is noncovalently bound to the class I
glycoprotein. Sequence data reveal homology between the H92513
166 PART II Generation of B-Cell and T-Cell Responses
α
1
α
2
β
1
β
2
β
2
-microglobulin
Transmembrane segment
Cytoplasmic tail
α
2 α
1
α
3
S
Class I molecule Class II molecule
S
S
S
S
S
S
S
SS
SS
Peptide-binding
cleft
Membrane-distal
domains
Membrane-proximal
domains
(Ig-fold structure)
FIGURE 7-5 Schematic diagrams of a class I and a class II MHC
molecule showing the external domains, transmembrane segment,
and cytoplasmic tail. The peptide-binding cleft is formed by the mem-
brane-distal domains in both class I and class II molecules. The
membrane-proximal domains possess the basic immunoglobulin-
fold structure; thus, class I and class II MHC molecules are classified
as members of the immunoglobulin superfamily.
8536d_ch07_161-184 8/16/02 12:09 PM Page 166 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
domain,H9252
2
-microglobulin, and the constant-region domains
in immunoglobulins. The enzyme papain cleaves the H9251 chain
just 13 residues proximal to its transmembrane domain, re-
leasing the extracellular portion of the molecule, consisting of
H92511, H92512, H92513, and H9252
2
-microglobulin. Purification and crystal-
lization of the extracellular portion revealed two pairs of in-
teracting domains: a membrane-distal pair made up of the H92511
and H92512 domains and a membrane-proximal pair composed of
the H92513 domain and H9252
2
-microglobulin (Figure 7-6a).
The H92511 and H92512 domains interact to form a platform of
eight antiparallel H9252 strands spanned by two long H9251-helical re-
gions. The structure forms a deep groove, or cleft, approxi-
mately 25 ? H11003 10 ? H11003 11 ?, with the long H9251 helices as sides
and the H9252 strands of the H9252 sheet as the bottom (Figure 7-6b).
This peptide-binding cleft is located on the top surface of the
class I MHC molecule, and it is large enough to bind a peptide
of 8–10 amino acids. The great surprise in the x-ray crystallo-
graphic analysis of class I molecules was the finding of small
peptides in the cleft that had cocrystallized with the protein.
These peptides are, in fact, processed antigen and self-pep-
tides bound to the H92511 and H92512 domains in this deep groove.
The H92513 domain and H9252
2
-microglobulin are organized into
two H9252 pleated sheets each formed by antiparallel H9252 strands of
amino acids. As described in Chapter 4, this structure, known
as the immunoglobulin fold, is characteristic of im-
munoglobulin domains. Because of this structural similarity,
which is not surprising given the considerable sequence sim-
ilarity with the immunoglobulin constant regions, class I
MHC molecules and H9252
2
-microglobulin are classified as
members of the immunoglobulin superfamily (see Figure
4-20). The H92513 domain appears to be highly conserved among
class I MHC molecules and contains a sequence that interacts
with the CD8 membrane molecule present on T
C
cells.
H9252
2
-Microglobulin interacts extensively with the H92513 do-
main and also interacts with amino acids of the H92511 and H92512
domains. The interaction of H9252
2
-microglobulin and a peptide
with a class I H9251 chain is essential for the class I molecule to
reach its fully folded conformation. As described in detail in
Chapter 8, assembly of class I molecules is believed to occur
by the initial interaction of H9252
2
-microglobulin with the fold-
ing class I H9251chain. This metastable “empty” dimer is then sta-
bilized by the binding of an appropriate peptide to form the
native trimeric class I structure consisting of the class I H9251
chain,H9252
2
-microglobulin, and a peptide. This complete mole-
cular complex is ultimately transported to the cell surface.
In the absence of H9252
2
-microglobulin, the class I MHC H9251
chain is not expressed on the cell membrane. This is illus-
trated by Daudi tumor cells, which are unable to synthesize
H9252
2
-microglobulin. These tumor cells produce class I MHC H9251
chains, but do not express them on the membrane. However,
if Daudi cells are transfected with a functional gene encoding
H9252
2
-microglobulin, class I molecules appear on the membrane.
Major Histocompatibility Complex CHAPTER 7 167
(b)
α1 domain
α2 domain
α3 domain
α2 domain
α1 domain
β
2
-microglobulin
α helix
β sheets
(a) Peptide-binding
cleft
FIGURE 7-6 Representations of the three-dimensional structure of
the external domains of a human class I MHC molecule based on x-
ray crystallographic analysis. (a) Side view in which the H9252 strands are
depicted as thick arrows and the H9251 helices as spiral ribbons. Disulfide
bonds are shown as two interconnected spheres. The H92511 and H92512 do-
mains interact to form the peptide-binding cleft. Note the im-
munoglobulin-fold structure of the H92513 domain and H9252
2
-microglobulin.
(b) The H92511 and H92512 domains as viewed from the top, showing the
peptide-binding cleft consisting of a base of antiparallel H9252 strands
and sides of H9251 helices. This cleft in class I molecules can accommo-
date peptides containing 8–10 residues.
8536d_ch07_161-184 8/16/02 12:09 PM Page 167 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
Class II Molecules Have Two Nonidentical
Glycoprotein Chains
Class II MHC molecules contain two different polypeptide
chains, a 33-kDa H9251 chain and a 28-kDa H9252 chain, which asso-
ciate by noncovalent interactions (see Figure 7-5b). Like class
I H9251 chains, class II MHC molecules are membrane-bound
glycoproteins that contain external domains, a transmem-
brane segment, and a cytoplasmic anchor segment. Each
chain in a class II molecule contains two external domains:
H92511 and H92512 domains in one chain and H92521 and H92522 domains in
the other. The membrane-proximal H92512 and H92522 domains, like
the membrane-proximal H92513/H9252
2
-microglobulin domains of
class I MHC molecules, bear sequence similarity to the im-
munoglobulin-fold structure; for this reason, class II MHC
molecules also are classified in the immunoglobulin super-
family. The membrane-distal portion of a class II molecule is
composed of the H92511 and H92521 domains and forms the antigen-
binding cleft for processed antigen.
X-ray crystallographic analysis reveals the similarity of
class II and class I molecules, strikingly apparent when the
molecules are surperimposed (Figure 7-7). The peptide-
binding cleft of HLA-DR1, like that in class I molecules, is
composed of a floor of eight antiparallel H9252 strands and sides
of antiparallel H9251 helices. However, the class II molecule lacks
the conserved residues that bind to the terminal residues of
short peptides and forms instead an open pocket; class I pre-
sents more of a socket, class II an open-ended groove. These
functional consequences of these differences in fine structure
will be explored below.
An unexpected difference between crystallized class I and
class II molecules was observed for human DR1 in that the
latter occurred as a dimer of H9251H9252 heterodimers, a “dimer of
dimers” (Figure 7-8). The dimer is oriented so that the two
peptide-binding clefts face in opposite directions. While it has
not yet been determined whether this dimeric form exists in
vivo, the presence of CD4 binding sites on opposite sides of
the class II molecule suggests that it does. These two sites on
the H92512 and H92522 domains are adjacent in the dimer form and a
CD4 molecule binding to them may stabilize class II dimers.
The Exon/Intron Arrangement of Class I and
II Genes Reflects Their Domain Structure
Separate exons encode each region of the class I and II pro-
teins (Figure 7-9). Each of the mouse and human class I
genes has a 5H11032 leader exon encoding a short signal peptide
168 PART II Generation of B-Cell and T-Cell Responses
FIGURE 7-7 The membrane-distal, peptide-binding cleft of a hu-
man class II MHC molecule, HLA-DR1 (blue), superimposed over
the corresponding regions of a human class I MHC molecule, HLA-
A2 (red). [From J. H. Brown et al., 1993, Nature 364:33.]
(a)
(b)
FIGURE 7-8 Antigen-binding cleft of dimeric class II DR1 molecule
in (a) top view and (b) side view. This molecule crystallized as a
dimer of the H9251H9252 heterodimer. The crystallized dimer is shown with
one DR1 molecule in red and the other DR1 molecule in blue. The
bound peptides are yellow. The two peptide-binding clefts in the
dimeric molecule face in opposite directions. [From J. H. Brown et al.,
1993, Nature 364:33.]
8536d_ch07_161-184 8/16/02 12:09 PM Page 168 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
followed by five or six exons encoding the H9251 chain of the class
I molecule (see Figure 7-9a). The signal peptide serves to fa-
cilitate insertion of the H9251 chain into the endoplasmic reticu-
lum and is removed by proteolytic enzymes in the
endoplasmic reticulum after translation is completed. The
next three exons encode the extracellular H92511, H92512, and H92513 do-
mains, and the following downstream exon encodes the
transmembrane (T
m
) region; finally, one or two 3H11032-terminal
exons encode the cytoplasmic domains (C).
Like class I MHC genes, the class II genes are organized
into a series of exons and introns mirroring the domain struc-
ture of the H9251 and H9252 chains (see Figure 7-9b). Both the H9251 and
the H9252 genes encoding mouse and human class II MHC mole-
cules have a leader exon, an H92511 or H92521 exon, an H92512 or H92522 exon,
a transmembrane exon, and one or more cytoplasmic exons.
Class I and II Molecules Exhibit
Polymorphism in the Region That
Binds to Peptides
Several hundred different allelic variants of class I and II MHC
molecules have been identified in humans. Any one individual,
however, expresses only a small number of these molecules—
up to 6 different class I molecules and up to 12 different class II
molecules. Yet this limited number of MHC molecules must be
able to present an enormous array of different antigenic pep-
tides to T cells, permitting the immune system to respond
specifically to a wide variety of antigenic challenges. Thus, pep-
tide binding by class I and II molecules does not exhibit the fine
specificity characteristic of antigen binding by antibodies and
T-cell receptors. Instead, a given MHC molecule can bind
Major Histocompatibility Complex CHAPTER 7 169
3′DNA 5′
α1 α2 α3 CCT
m
(a)
COOH
H
2
N
α chain
3′DNA 5′
β1 β2CCT
m+C
(b)
Class I MHC
molecule
mRNA
mRNA mRNA
(A)
n
(A)
n
(A)
n
3′DNA 5′
α1 α2
α1 α2
CT
m+C
SS SS
SS
L
α1
α
2
α
1
α
3
β
2
- microglobulin
α3α2 T
m
LCC β1 β2LCC
L
L
T
m+C
LCT
m+C
COOH
COOH
H
2
N
H
2
N
β chain
α chain
Class II MHC
molecule
SS SS
SS
β
1
β
2
α
1 α
2
FIGURE 7-9 Schematic diagram of (a) class I and (b) class II MHC
genes, mRNA transcripts, and protein molecules. There is corre-
spondence between exons and the domains in the gene products;
note that the mRNA transcripts are spliced to remove the intron se-
quences. Each exon, with the exception of the leader (L) exon, en-
codes a separate domain of the MHC molecule. The leader peptides
are removed in a post-translational reaction before the molecules are
expressed on the cell surface. The gene encoding H9252
2
-microglobulin is
located on a different chromosome. T
m
H11005 transmembrane; C H11005
cytoplasmic.
8536d_ch07_161-184 8/16/02 12:09 PM Page 169 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
numerous different peptides, and some peptides can bind to
several different MHC molecules. Because of this broad speci-
ficity, the binding between a peptide and an MHC molecule is
often referred to as “promiscuous.”
Given the similarities in the structure of the peptide-bind-
ing cleft in class I and II MHC molecules, it is not surprising
that they exhibit some common peptide-binding features
(Table 7-2). In both types of MHC molecules, peptide lig-
ands are held in a largely extended conformation that runs
the length of the cleft. The peptide-binding cleft in class I
molecules is blocked at both ends, whereas the cleft is open in
class II molecules (Figure 7-10). As a result of this difference,
class I molecules bind peptides that typically contain 8–10
amino acid residues, while the open groove of class II mole-
cules accommodates slightly longer peptides of 13–18 amino
acids. Another difference, explained in more detail below, is
that class I binding requires that the peptide contain specific
amino acid residues near the N and C termini; there is no
such requirement for class II peptide binding.
The peptide–MHC molecule association is very stable
(K
d
~ 10
H110026
) under physiologic conditions; thus, most of
170 PART II Generation of B-Cell and T-Cell Responses
TABLE 7-2 Peptide binding by class I and class II MHC molecules
Class I molecules Class II molecules
Peptide-binding domain H92511/H92512 H92511/H92521
Nature of peptide-binding cleft Closed at both ends Open at both ends
General size of bound peptides 8–10 amino acids 13–18 amino acids
Peptide motifs involved in Anchor residues at both ends of Anchor residues distributed along
binding to MHC molecule peptide; generally hydrophobic the length of the peptide
carboxyl-terminal anchor
Nature of bound peptide Extended structure in which both ends Extended structure that is held
interact with MHC cleft but middle at a constant elevation above
arches up away from MHC molecule the floor of MHC cleft
(a) Class I MHC (b) Class II MHC
FIGURE 7-10 MHC class I and class II molecules with bound pep-
tides. (a) Space-filling model of human class I molecule HLA-A2
(white) with peptide (red) from HIV reverse transcriptase (amino
acid residues 309–317) in the binding groove. H9252
2
-microglobulin is
shown in blue. Residues above the peptide are from the H92511 domain,
those below from H92512. (b) Space-filling model of human class II mol-
ecules HLA-DR1 with the DRH9251 chain shown in white and the DRH9252
chain in blue. The peptide (red) in the binding groove is from in-
fluenza hemagglutinin (amino acid residues 306–318). [From D. A.
Vignali and J. Strominger, 1994, The Immunologist 2:112.]
the MHC molecules expressed on the membrane of a cell
will be associated with a peptide of self or nonself origin.
CLASS I MHC–PEPTIDE INTERACTION
Class I MHC molecules bind peptides and present them to
CD8
H11001
T cells. In general, these peptides are derived from en-
dogenous intracellular proteins that are digested in the cy-
tosol. The peptides are then transported from the cytosol
into the cisternae of the endoplasmic reticulum, where they
interact with class I MHC molecules. This process, known as
the cytosolic or endogenous processing pathway, is discussed
in detail in the next chapter.
Each type of class I MHC molecule (K, D, and L in mice
or A, B, and C in humans) binds a unique set of peptides. In
addition, each allelic variant of a class I MHC molecule (e.g.,
H-2K
k
and H-2K
d
) also binds a distinct set of peptides. Be-
cause a single nucleated cell expresses about 10
5
copies of
each class I molecule, many different peptides will be ex-
pressed simultaneously on the surface of a nucleated cell by
class I MHC molecules.
8536d_ch07_161-184 8/16/02 1:49 PM Page 170 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
Major Histocompatibility Complex CHAPTER 7 171
H
3
N COO
?
COO
?
COO
?
COO
?
COO
?
COO
?
COO
?
H
3
N
H
3
N
H
3
N
VGPQKNENL
SGPRKAIAL
VGP SGKYF I
SGPERI LSL
123456789
Eluted
from
H-2D
d
H
3
N
H
3
N
H
3
N
SYFPE ITHI
TYQRTRALV
SYIGS INNI
Eluted
from
H-2K
d
A = alanine
E = glutamic acid
F = phenylalanine
G = glycine
H = histidine
I = isoleucine
K = lysine
L = leucine
N = asparagine
P = proline
Q = glutamine
R = arginine
S = serine
T = threonine
V = valine
Y = tyrosine
FIGURE 7-11 Examples of anchor residues (blue) in nonameric
peptides eluted from two class I MHC molecules. Anchor residues
that interact with the class I MHC molecule tend to be hydrophobic
amino acids. [Data from V. H. Engelhard, 1994, Curr. Opin. Immunol.
6:13.]
In a critical study of peptide binding by MHC molecules,
peptides bound by two allelic variants of a class I MHC mol-
ecule were released chemically and analyzed by HPLC mass
spectrometry. More than 2000 distinct peptides were found
among the peptide ligands released from these two class I
MHC molecules. Since there are approximately 10
5
copies of
each class I allelic variant per cell, it is estimated that each of
the 2000 distinct peptides is presented with a frequency of
100–4000 copies per cell. Evidence suggests that as few as
100 peptide-MHC complexes are sufficient to target a cell
for recognition and lysis by a cytotoxic T lymphocyte with a
receptor specific for this target structure.
The bound peptides isolated from different class I mole-
cules have been found to have two distinguishing features:
they are eight to ten amino acids in length, most commonly
nine, and they contain specific amino acid residues that ap-
pear to be essential for binding to a particular MHC mole-
cule. Binding studies have shown that nonameric peptides
bind to class I molecules with a 100- to 1000-fold higher
affinity than do peptides that are either longer or shorter,
suggesting that this peptide length is most compatible with
the closed-ended peptide-binding cleft in class I molecules.
The ability of an individual class I MHC molecule to bind to
a diverse spectrum of peptides is due to the presence of the
same or similar amino acid residues at several defined posi-
tions along the peptides (Figure 7-11). Because these amino
acid residues anchor the peptide into the groove of the
MHC molecule, they are called anchor residues. The side
chains of the anchor residues in the peptide are comple-
mentary with surface features of the binding cleft of the
class I MHC molecule. The amino acid residues lining the
binding sites vary among different class I allelic variants and
determine the identity of the anchor residues that can inter-
act with the molecule.
All peptides examined to date that bind to class I mole-
cules contain a carboxyl-terminal anchor. These anchors are
generally hydrophobic residues (e.g., leucine, isoleucine), al-
though a few charged amino acids have been reported. Be-
sides the anchor residue found at the carboxyl terminus,
another anchor is often found at the second or second and
third positions at the amino-terminal end of the peptide (see
Figure 7-11). In general, any peptide of correct length that
contains the same or similar anchor residues will bind to the
same class I MHC molecule. The discovery of conserved an-
chor residues in peptides that bind to various class I MHC
molecules may permit prediction of which peptides in a
complex antigen will bind to a particular MHC molecule,
based on the presence or absence of these motifs.
X-ray crystallographic analyses of peptide–class I MHC
complexes have revealed how the peptide-binding cleft in a
given MHC molecule can interact stably with a broad spec-
trum of different peptides. The anchor residues at both ends
of the peptide are buried within the binding cleft, thereby
holding the peptide firmly in place (Figure 7-12). As noted al-
ready, nonameric peptides are bound preferentially; the main
contacts between class I MHC molecules and peptides in-
volve residue 2 at the amino-terminal end and residue 9 at the
carboxyl terminus of the nonameric peptide. Between the an-
chors the peptide arches away from the floor of the cleft in the
middle (Figure 7-13), allowing peptides that are slightly
longer or shorter to be accommodated. Amino acids that arch
away from the MHC molecule are more exposed and pre-
sumably can interact more directly with the T-cell receptor.
CLASS II MHC–PEPTIDE INTERACTION
Class II MHC molecules bind peptides and present these
peptides to CD4
H11001
T cells. Like class I molecules, molecules of
class II can bind a variety of peptides. In general, these pep-
tides are derived from exogenous proteins (either self or
nonself), which are degraded within the endocytic process-
ing pathway (see Chapter 8). Most of the peptides associated
with class II MHC molecules are derived from membrane-
bound proteins or proteins associated with the vesicles of the
endocytic processing pathway. The membrane-bound pro-
teins presumably are internalized by phagocytosis or by
receptor-mediated endocytosis and enter the endocytic pro-
cessing pathway at this point. For instance, peptides derived
from digestion of membrane-bound class I MHC molecules
often are bound to class II MHC molecules.
Peptides recovered from class II MHC–peptide com-
plexes generally contain 13–18 amino acid residues, some-
what longer than the nonameric peptides that most
commonly bind to class I molecules. The peptide-binding
cleft in class II molecules is open at both ends (see Figure
7-10b), allowing longer peptides to extend beyond the ends,
like a long hot dog in a bun. Peptides bound to class II MHC
molecules maintain a roughly constant elevation on the
8536d_ch07_161-184 8/16/02 1:49 PM Page 171 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
In addition, over 30% of the peptides eluted from class II mol-
ecules contain a proline residue at position 2 and another
cluster of prolines at the carboxyl-terminal end.
Class I and Class II Molecules Exhibit
Diversity Within a Species and Multiple
Forms Occur in an Individual
An enormous diversity is exhibited by the MHC molecules
within a species and within individuals. This variability
echoes the diversity of antibodies and T-cell receptors, but
the source of diversity for MHC molecules is not the same.
Antibodies and T-cell receptors are generated by several so-
matic processes, including gene rearrangement and somatic
mutation of rearranged genes (see Table 5-2). Thus, the gen-
eration of T and B cell receptors is dynamic, changing over
time within an individual. By contrast, the MHC molecules
expressed by an individual are fixed in the genes and do not
change over time. The diversity of the MHC within a species
stems from polymorphism, the presence of multiple alleles at
a given genetic locus within the species. Diversity of MHC
molecules in an individual results not only from having dif-
ferent alleles of each gene but also from the presence of du-
plicated genes with similar or overlapping functions, not
unlike the isotypes of immunoglobulins. Because it includes
genes with similar, but not identical structure and function
(for example, HLA-A, -B, and -C), the MHC may be said to
be polygenic.
The MHC possesses an extraordinarily large number of
different alleles at each locus and is one of the most poly-
morphic genetic complexes known in higher vertebrates.
These alleles differ in their DNA sequences from one indi-
vidual to another by 5% to 10%. The number of amino acid
differences between MHC alleles can be quite significant,
with up to 20 amino acid residues contributing to the
unique structural nature of each allele. Analysis of human
HLA class I genes has revealed, as of early 2002, approxi-
mately 240 A alleles, 470 B alleles, and 110 C alleles. In mice,
the polymorphism is similarly enormous. The human class
II genes are also highly polymorphic and, in some cases,
there are different gene numbers in different individuals.
The number of HLA-DR beta-chain genes may vary from 2
to 9 in different haplotypes, and approximately 350 alleles of
DRB genes have been reported. Interestingly, the DRA chain
is highly conserved, with only 2 different alleles reported.
Current estimates of actual polymorphism in the human
MHC are probably on the low side because the most detailed
data were obtained from populations of European descent.
The fact that many non-European population groups can-
not be typed using the MHC serologic typing reagents avail-
able indicates that the worldwide diversity of the MHC
genes is far greater. Now that MHC genes can be sequenced
directly, it is expected that many additional alleles will be
detected.
This enormous polymorphism results in a tremendous
diversity of MHC molecules within a species. Using the num-
bers given above for the allelic forms of human HLA-A, -B,
172 PART II Generation of B-Cell and T-Cell Responses
floor of the binding cleft, another feature that distinguishes
peptide binding to class I and class II molecules.
Peptide binding studies and structural data for class II
molecules indicate that a central core of 13 amino acids deter-
mines the ability of a peptide to bind class II. Longer peptides
may be accommodated within the class II cleft, but the bind-
ing characteristics are determined by the central 13 residues.
The peptides that bind to a particular class II molecule often
have internal conserved “motifs,” but unlike class I–binding
peptides, they lack conserved anchor residues. Instead, hydro-
gen bonds between the backbone of the peptide and the class
II molecule are distributed throughout the binding site rather
than being clustered predominantly at the ends of the site as
for class I–bound peptides. Peptides that bind to class II MHC
molecules contain an internal sequence comprising 7–10
amino acids that provide the major contact points. Generally,
this sequence has an aromatic or hydrophobic residue at the
amino terminus and three additional hydrophobic residues in
the middle portion and carboxyl-terminal end of the peptide.
FIGURE 7-12 Model of the solvent-accessible area of class I H-2K
b
,
depicting the complex formed with a vesicular stomatitis virus (VSV-
8) peptide (left, yellow backbone) and Sendai virus (SEV-9) nucleo-
protein (right, blue backbone). Water molecules (blue spheres)
interact with the bound peptides. The majority of the surface of both
peptides is inaccessible for direct contact with T cells (VSV-8 is 83%
buried; SEV-9 is 75% buried). The H-2K
b
surface in the two com-
plexes exhibits a small, but potentially significant, conformational
variation, especially in the central region of the binding cleft on the
right side of the peptides, which corresponds to the H9251 helix in the H92512
domain (see Figure 7-6b). [From M. Matsumura et al., 1992, Science
257:927; photographs courtesy of D. H. Fremont, M. Matsumura,
M. Pique, and I. A. Watson.]
8536d_ch07_161-184 8/16/02 12:09 PM Page 172 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
and -C, we can calculate the theoretical number of combina-
tions that can exist by multiplying 240 H11003 470 H11003 110, yielding
upwards of 12 million different class I haplotypes possible in
the population. If class II loci are considered, the 5 DRB
genes B1 through B5 have 304, 1, 35, 11, and 15 alleles re-
spectively, DQA1 and B1 contribute 22 and 49 alleles, respec-
tively and, DPB1 96 alleles; this allows approximately 1.8 H11003
10
11
different class II combinations. Because each haplotype
contains both class I and class II genes, the numbers are mul-
tiplied to give a total of 2.25 H11003 10
18
possible combinations of
these class I and II alleles.
LINKAGE DISEQUILIBRIUM
The calculation of theoretical diversity in the previous para-
graph assumes completely random combinations of alleles.
The actual diversity is known to be less, because certain allelic
combinations occur more frequently in HLA haplotypes
than predicted by random combination, a state referred to as
linkage disequilibrium. Briefly, linkage disequilibrium is the
difference between the frequency observed for a particular
combination of alleles and that expected from the frequencies
of the individual alleles. The expected frequency for the com-
bination may be calculated by multiplying the frequencies of
the two alleles. For example, if HLA-A1 occurs in 16% of in-
dividuals in a population (frequency H11005 0.16) and HLA-B8 in
9% of that group (frequency H11005 0.09) it is expected that about
1.4% of the group should have both alleles (0.16 H11003 0.09 H11005
0.014). However, the data show that HLA-A1 and HLA-B8
are found together in 8.8% of individuals studied. This dif-
ference is a measure of the linkage disequilibrium between
these alleles of class I MHC genes.
Several explanations have been advanced to explain link-
age disequilibrium. The simplest is that too few generations
have elapsed to allow the number of crossovers necessary to
reach equilibrium among the alleles present in founders of
the population. The haplotypes that are over-represented in
the population today would then reflect the combinations of
alleles present in the founders. Alternatively, selective effects
could also result in the higher frequency of certain allelic
combinations. For example, certain combinations of alleles
might produce resistance to certain diseases, causing them to
be selected for and over-represented, or they might generate
harmful effects, such as susceptibility to autoimmune disor-
ders, and undergo negative selection. A third hypothesis is
that crossovers are more frequent in certain DNA sequence
regions, and the presence or absence of regions prone
to crossover (hotspots) between alleles can dictate the
Major Histocompatibility Complex CHAPTER 7 173
1
4
5
6
7
8
9
2
3
Bulge
N
C
Hydrogen bonds with
MHC molecule
(a)
1
2
3
4
5
6
7
8
9
(b)
(c)
FIGURE 7-13 Conformation of peptides bound to class I MHC
molecules. (a) Schematic diagram of conformational difference in
bound peptides of different lengths. Longer peptides bulge in the
middle, whereas shorter peptides are more extended. Contact with
the MHC molecule is by hydrogen bonds to anchor residues 1/2 and
8/9. (b) Molecular models based on crystal structure of an influenza
virus antigenic peptide (blue) and an endogenous peptide (purple)
bound to a class I MHC molecule. Residues are identified by small
numbers corresponding to those in part (a). (c) Representation of H92511
and H92512 domains of HLA-B27 and a bound antigenic peptide based
on x-ray crystallographic analysis of the cocrystallized peptide–HLA
molecule. The peptide (purple) arches up away from the H9252 strands
forming the floor of the binding cleft and interacts with twelve water
molecules (spheres). [Part (a) adapted from P. Parham, 1992, Nature
360:300, ? 1992 Macmillan Magazines Limited; part (b) adapted
from M. L. Silver et al., 1992, Nature 360:367, ? 1992 Macmillan
Magazines Limited; part (c) adapted from D. R. Madden et al., 1992,
Cell 70:1035, reprinted by permission of Cell Press.]
8536d_ch07_161-184 8/16/02 12:09 PM Page 173 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
174 PART II Generation of B-Cell and T-Cell Responses
frequency of allelic association. Data in support of this was
found in mouse breeding studies that generated new recom-
binant H-2 types. The points of crossover in the new MHC
haplotypes were not randomly distributed throughout the
complex. Instead, the same regions of crossover were found
in more than one recombinant haplotype. This suggests that
hotspots of recombination do exist that would influence
linkage disequilibrium in populations.
Despite linkage disequilibrium, there is still enormous poly-
morphism in the human MHC, and it remains very difficult to
match donor and acceptor MHC types for successful organ
transplants. The consequences of this major obstacle to the
therapeutic use of transplantation are described in Chapter 21.
FUNCTIONAL RELEVANCE OF MHC POLYMORPHISM
Sequence divergence among alleles of the MHC within a
species is very high, as great as the divergence observed for
the genes encoding some enzymes across species. Also of in-
terest is that the sequence variation among MHC molecules
is not randomly distributed along the entire polypeptide
chain but instead is clustered in short stretches, largely
within the membrane-distal H92511 and H92512 domains of class I
molecules (Figure 7-14a). Similar patterns of diversity are
observed in the H92511 and H92522 domains of class II molecules.
Progress has been made in locating the polymorphic
residues within the three-dimensional structure of the mem-
brane-distal domains in class I and class II MHC molecules
and in relating allelic differences to functional differences
(Figure 7-14b). For example, of 17 amino acids previously
shown to display significant polymorphism in the HLA-A2
molecule, 15 were shown by x-ray crystallographic analysis to
be in the peptide-binding cleft of this molecule. The location
of so many polymorphic amino acids within the binding site
for processed antigen strongly suggests that allelic differences
contribute to the observed differences in the ability of MHC
molecules to interact with a given antigenic peptide.
Detailed Genomic Map
of MHC Genes
The MHC spans some 2000 kb of mouse DNA and some
4000 kb of human DNA. The recently completed human
genome sequence shows this region to be densely packed
Variability
α1 α2 α3
20 40 60 80 100 120 140 180 200 220 240 260160
Residue number
(a)
(b)
45
12
62
63
66
70 74
9
95
97 116
114
156
105
107
N
FIGURE 7-14 (a) Plots of variability in the amino acid sequence of
allelic class I MHC molecules in humans versus residue position. In
the external domains, most of the variable residues are in the mem-
brane-distal H92511 and H92512 domains. (b) Location of polymorphic amino
acid residues (red) in the H92511/H92512 domain of a human class I MHC
molecule. [Part (a) adapted from R. Sodoyer et al., 1984, EMBO J.
3:879, reprinted by permission of Oxford University Press; part
(b) adapted, with permission, from P. Parham, 1989, Nature 342:617,
? 1989 Macmillan Magazines Limited.]
8536d_ch07_161-184 8/16/02 12:09 PM Page 174 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
Major Histocompatibility Complex CHAPTER 7 175
Complex
Class
Loci
Centromere
MOUSE CHROMOSOME 17
II
H-2
III I II
Complex
Class
Loci
Centromere
HUMAN CHROMOSOME 6
II
HLA
4000 kb
I
α α
III
I
50kb
Tla
1500kb
400kb
... ...
Qa Tla
K2 K1 IA IA IE IE IE
β β
β β
α
αα α
β
α β α β
β
β α
α β
ββ β
α
β β
α
α αβ
α β
2
β
2
β
1
P O M M M LMP2 TAP2 LMP7 TAP1 O CYP21 C4B CYP21P C4A Bf C2 HSP G7a/b TNF- TNF- DL
1000kb 1000kb 2000kb
Telomere
Telomere
DP 2 DP 2 DP 1 DP 1 DO DM DM LMP2 TAP1 LMP7 TAP2 DO DQ 2 DQ 2 DQ 3 DQ 1 DQ 1 DR DR CYP21 C4B CYP21P C4A Bf C2 HSP70 G7a/b TNF- TNF- HLA-B HLA-CMICB
MICA
HLA-X HLA-E MICC HLA-J HLA-A MICD HLA-H* HLA-G MICE HLA-F
C2, C4A, C4B, Bf Complement components
CYP21,CYP21P Steroid 21-hydroxylases
G7a/b Valyl-tRNA synthetase
HSP Heat-shock protein
LMP2, LMP7 Proteasome-like subunits
TAP1, TAP2 Peptide-transporter subunits
TNF- , TNF- Tumor necrosis factors and
KEY *Now designated HFE
Gene Encoded protein
FIGURE 7-15 Detailed genomic map of the mouse and human MHC, in-
cluding genes encoding classical and nonclassical MHC molecules. The class
I MHC genes are colored red, MHC II genes are colored blue, and genes in
MHC III are colored green. Classical class I genes are labeled in red, class II in
blue, and the nonclassical MHC genes are labeled in black. The concept of
classical and nonclassical does not apply to class III. The functions for certain
proteins encoded by the nonclassical class I genes are known. In the mouse,
there are nonclassical genes located downstream from Tla that are not shown.
with genes, most of which have known functions. Our cur-
rent understanding of the genomic organization of mouse
and human MHC genes is diagrammed in Figure 7-15.
The Human Class I Region Spans
about 2000 kb at the Telomeric End
of the HLA Complex
In humans, the class I MHC region is about 2000 kb long
and contains approximately 20 genes. In mice, the class I
MHC consists of two regions separated by the intervening
class II and class III regions. Included within the class I re-
gion are the genes encoding the well-characterized classical
class I MHC molecules designated HLA-A, HLA-B, and
HLA-C in humans and H-2K, H-2D, and H-2L in mice.
Many nonclassical class I genes, identified by molecular
mapping, also are present in both the mouse and human
MHC. In mice, the nonclassical class I genes are located in
three regions (H-2Q, T, and M) downstream from the H-2
complex (M is not shown in Figure 7-15). In humans, the
nonclassical class I genes include the HLA-E, HLA-F, HLA-G,
HFE, HLA-J, and HLA-X loci as well as a recently discovered
family of genes called MIC, which includes MICA through
MICE. Some of the nonclassical class I MHC genes are
pseudogenes and do not encode a protein product, but oth-
ers, such as HLA-G and HFE, encode class I–like products
with highly specialized functions. The MIC family of class I
genes has only 15%–30% sequence identity to classical class
I, and those designated as MICA are highly polymorphic.
The MIC gene products are expressed at low levels in epithe-
lial cells and are induced by heat or other stimuli that influ-
ence heat shock proteins.
8536d_ch07_161-184 8/16/02 12:09 PM Page 175 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
176 PART II Generation of B-Cell and T-Cell Responses
The functions of the nonclassical class I MHC molecules
remain largely unknown, although a few studies suggest that
some of these molecules, like the classical class I MHC mol-
ecules, may present peptides to T cells. One intriguing find-
ing is that the mouse molecule encoded by the H-2M locus is
able to bind a self-peptide derived from a subunit of NADH
dehydrogenase, an enzyme encoded by the mitochondrial
genome. This particular self-peptide contains an amino-
terminal formylated methionine. What is interesting about
this finding is that peptides derived from prokaryotic organ-
isms often have formylated amino-terminal methionine
residues. This H-2M–encoded class I molecule may thus be
uniquely suited to present peptides from prokaryotic organ-
isms that are able to grow intracellularly. Such organisms in-
clude Mycobacterium tuberculosis, Listeria monocytogenes,
Brucella abortus, and Salmonella typhimurium.
Up to this point, all description of antigen presentation
by class I and class II molecules has been confined to presen-
tation of peptide antigens. As will be seen in the description
of antigen presentation (Chapter 8), there are also molecules
with structural similarity to class I molecules that present
non-peptide antigens, such as glycolipids, to T cells. A major
family of such molecules, designated CD1, has been shown
to present lipid antigens derived from bacteria. The CD1
molecules are not encoded within the MHC but are located
on chromosome 1.
The Class II MHC Genes Are Located
at the Centromeric End of HLA
The class II MHC region contains the genes encoding the H9251
and H9252 chains of the classical class II MHC molecules desig-
nated HLA-DR, DP, and DQ in humans and H-2IA and -IE
in mice. Molecular mapping of the class II MHC has re-
vealed multiple H9252-chain genes in some regions in both mice
and humans, as well as multiple H9251-chain genes in humans
(see Figure 7-15). In the human DR region, for example,
there are three or four functional H9252-chain genes. All of the H9252-
chain gene products can be expressed together with the H9251-
chain gene product in a given cell, thereby increasing the
number of different antigen-presenting molecules on the
cell. Although the human DR region contains just one H9251-
chain gene, the DP and DQ regions each contains two.
Genes encoding nonclassical class II MHC molecules
have also been identified in both humans and mice. In mice,
several class II genes (OH9251,OH9252,MH9251, and MH9252) encode non-
classical MHC molecules that exhibit limited polymorphism
and a different pattern of expression than the classical IA
and IE class II molecules. In the human class II region, non-
classical genes designated DM and DO have been identified.
The DM genes encode a class II–like molecule (HLA-DM)
that facilitates the loading of antigenic peptides into the class
II MHC molecules. Class II DO molecules, which are ex-
pressed only in the thymus and mature B cells, have been
shown to serve as regulators of class II antigen processing.
The functions of HLA-DM and HLA-DO will be described
further in Chapter 8.
Human MHC Class III Genes
Are Between Class I and II
The class III region of the MHC in humans and mice con-
tains a heterogeneous collection of genes (see Figure 7-15).
These genes encode several complement components, two
steroid 21-hydroxylases, two heat-shock proteins, and two
cytokines (TNF-H9251and TNF-H9252). Some of these class III MHC
gene products play a role in certain diseases. For example,
mutations in the genes encoding 21-hydroxylase have been
linked to congenital adrenal hyperplasia. Interestingly, the
presence of a linked class III gene cluster is conserved in all
species with an MHC region.
Cellular Distribution
of MHC Molecules
In general, the classical class I MHC molecules are expressed
on most nucleated cells, but the level of expression differs
among different cell types. The highest levels of class I mole-
cules are expressed by lymphocytes, where they constitute
approximately 1% of the total plasma-membrane proteins,
or some 5 H11003 10
5
molecules per cell. In contrast, fibroblasts,
muscle cells, liver hepatocytes, and neural cells express very
low levels of class I MHC molecules. The low level on liver
cells may contribute to the considerable success of liver
transplants by reducing the likelihood of graft recognition
by T
c
of the recipient. A few cell types (e.g., neurons and
sperm cells at certain stages of differentiation) appear to lack
class I MHC molecules altogether.
As noted earlier, any particular MHC molecule can bind
many different peptides. Since the MHC alleles are codomi-
nantly expressed, a heterozygous individual expresses on its
cells the gene products encoded by both alleles at each MHC
locus. An F
1
mouse, for example, expresses the K, D, and L
from each parent (six different class I MHC molecules) on
each of its nucleated cells (Figure 7-16). A similar situation
occurs in humans; that is, a heterozygous individual ex-
presses the A, B, and C alleles from each parent (six different
class I MHC molecules) on the membrane of each nucleated
cell. The expression of so many class I MHC molecules al-
lows each cell to display a large number of peptides in the
peptide-binding clefts of its MHC molecules.
In normal, healthy cells, the class I molecules will display
self-peptides resulting from normal turnover of self pro-
teins. In cells infected by a virus, viral peptides, as well as self-
peptides, will be displayed. A single virus-infected cell
should be envisioned as having various class I molecules on
its membrane, each displaying different sets of viral pep-
tides. Because of individual allelic differences in the peptide-
binding clefts of the class I MHC molecules, different
individuals within a species will have the ability to bind dif-
ferent sets of viral peptides.
Unlike class I MHC molecules, class II molecules are ex-
pressed constitutively only by antigen-presenting cells, pri-
8536d_ch07_161-184 8/15/02 8:41 PM Page 176 mac114 Mac 114:2nd shift:
Major Histocompatibility Complex CHAPTER 7 177
marily macrophages, dendritic cells, and B cells; thymic
epithelial cells and some other cell types can be induced to
express class II molecules and to function as antigen-pre-
senting cells under certain conditions and under stimulation
of some cytokines (see Chapter 8). Among the various cell
types that express class II MHC molecules, marked differ-
ences in expression have been observed. In some cases, class
II expression depends on the cell’s differentiation stage. For
example, class II molecules cannot be detected on pre-B cells
but are expressed constitutively on the membrane of mature
B cells. Similarly, monocytes and macrophages express only
low levels of class II molecules until they are activated by in-
teraction with an antigen, after which the level of expression
increases significantly.
Because each of the classical class II MHC molecules is
composed of two different polypeptide chains, which are en-
coded by different loci, a heterozygous individual expresses
not only the parental class II molecules but also molecules
containing H9251 and H9252 chains from different chromosomes. For
example, an H-2
k
mouse expresses IA
k
and IE
k
class II mole-
cules; similarly, an H-2
d
mouse expresses IA
d
and IE
d
mole-
cules. The F
1
progeny resulting from crosses of mice with
these two haplotypes express four parental class II molecules
and four molecules containing one parent’s H9251 chain and the
other parent’s H9252 chain (as shown in Figure 7-16). Since the
human MHC contains three classical class II genes (DP, DQ,
and DR), a heterozygous individual expresses six parental
class II molecules and six molecules containing H9251and H9252chain
combinations from either parent. The number of different
class II molecules expressed by an individual is increased fur-
ther by the presence of multiple H9252-chain genes in mice and
humans, and in humans by multiple H9251-chain genes. The di-
versity generated by these mechanisms presumably increases
the number of different antigenic peptides that can be pre-
sented and thus is advantageous to the organism.
Regulation of MHC Expression
Research on the regulatory mechanisms that control the dif-
ferential expression of MHC genes in different cell types is
still in its infancy, but much has been learned. The publica-
tion of the complete genomic map of the MHC complex is
expected to greatly accelerate the identification and investi-
gation of coding and regulatory sequences, leading to new
directions in research on how the system is controlled.
Both class I and class II MHC genes are flanked by 5H11032 pro-
moter sequences, which bind sequence-specific transcrip-
tion factors. The promoter motifs and transcription factors
that bind to these motifs have been identified for a number
of MHC genes. Transcriptional regulation of the MHC is
mediated by both positive and negative elements. For exam-
ple, an MHC II transactivator, called CIITA, and another
transcription factor, called RFX, both have been shown to
bind to the promoter region of class II MHC genes. Defects
in these transcription factors cause one form of bare lym-
phocyte syndrome (see the Clinical Focus box in Chapter 8).
Patients with this disorder lack class II MHC molecules on
their cells and as a result suffer a severe immunodeficiency
due to the central role of class II MHC molecules in T-cell
maturation and activation.
The expression of MHC molecules is also regulated by
various cytokines. The interferons (alpha, beta, and gamma)
and tumor necrosis factor have each been shown to increase
expression of class I MHC molecules on cells. Interferon
gamma (IFN-H9253), for example, appears to induce the forma-
tion of a specific transcription factor that binds to the pro-
moter sequence flanking the class I MHC genes. Binding of
this transcription factor to the promoter sequence appears
to coordinate the up-regulation of transcription of the genes
encoding the class I H9251 chain, H9252
2
-microglobulin, the protea-
some subunits (LMP), and the transporter subunits (TAP).
IFN-H9253 also has been shown to induce expression of the class
II transactivator (CIITA), thereby indirectly increasing ex-
pression of class II MHC molecules on a variety of cells, in-
cluding non-antigen-presenting cells (e.g., skin keratin-
ocytes, intestinal epithelial cells, vascular endothelium, pla-
cental cells, and pancreatic beta cells). Other cytokines influ-
ence MHC expression only in certain cell types; for example,
IL-4 increases expression of class II molecules by resting B
cells. Expression of class II molecules by B cells is down-reg-
ulated by IFN-H9253; corticosteroids and prostaglandins also de-
crease expression of class II molecules.
MHC expression is decreased by infection with certain
viruses, including human cytomegalovirus (CMV), hepatitis
IE
k k
αβ
IA
k k
αβ
IE
d d
αβ
IA
d d
αβ
IE
k d
αβ
αβ
IA
k d
αβ
IE
d k
αβ IA
d k
αβ
K
k
K
d
D
k
D
d
L
k
L
d
Maternal MHC
Paternal MHC
IA
d d
αβIE
d d
K
d
D
d
L
d
αβIA
k k
αβIE
k k
K
k
D
k
L
k
Class II
molecules
Class I
molecules
FIGURE 7-16 Diagram illustrating various MHC molecules ex-
pressed on antigen-presenting cells of a heterozygous H-2
k/d
mouse.
Both the maternal and paternal MHC genes are expressed. Because
the class II molecules are heterodimers, heterologous molecules
containing one maternal-derived and one paternal-derived chain are
produced. The H9252
2
-microglobulin component of class I molecules
(pink) is encoded by a gene on a separate chromosome and may be
derived from either parent.
8536d_ch07_161-184 8/16/02 12:09 PM Page 177 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
178 PART II Generation of B-Cell and T-Cell Responses
B virus (HBV), and adenovirus 12 (Ad12). In some cases, re-
duced expression of class I MHC molecules on cell surfaces
is due to decreased levels of a component needed for peptide
transport or MHC class I assembly rather than in transcrip-
tion. In cytomegalovirus infection, for example, a viral pro-
tein binds to H9252
2
-microglobulin, preventing assembly of class
I MHC molecules and their transport to the plasma mem-
brane. Adenovirus 12 infection causes a pronounced de-
crease in transcription of the transporter genes (TAP1 and
TAP2). As the next chapter describes, the TAP gene products
play an important role in peptide transport from the cyto-
plasm into the rough endoplasmic reticulum. Blocking of
TAP gene expression inhibits peptide transport; as a result,
class I MHC molecules cannot assemble with H9252
2
-microglob-
ulin or be transported to the cell membrane. Decreased ex-
pression of class I MHC molecules, by whatever mechanism,
is likely to help viruses evade the immune response by re-
ducing the likelihood that virus-infected cells can display
MHC–viral peptide complexes and become targets for CTL-
mediated destruction.
MHC and Immune Responsiveness
Early studies by B. Benacerraf in which guinea pigs were im-
munized with simple synthetic antigens were the first to
show that the ability of an animal to mount an immune re-
sponse, as measured by the production of serum antibodies,
is determined by its MHC haplotype. Later experiments by
H. McDevitt, M. Sela, and their colleagues used congenic and
recombinant congenic mouse strains to map the control of
immune responsiveness to class II MHC genes. In early re-
ports, the genes responsible for this phenotype were desig-
nated Ir or immune response genes, and for this reason
mouse class II products are called IA and IE. We now know
that the dependence of immune responsiveness on the class
II MHC reflects the central role of class II MHC molecules in
presenting antigen to T
H
cells.
Two explanations have been proposed to account for the
variability in immune responsiveness observed among dif-
ferent haplotypes. According to the determinant-selection
model, different class II MHC molecules differ in their abil-
ity to bind processed antigen. According to the alternative
holes-in-the-repertoire model, T cells bearing receptors that
recognize foreign antigens closely resembling self-antigens
may be eliminated during thymic processing. Since the T-
cell response to an antigen involves a trimolecular complex
of the T cell’s receptor, an antigenic peptide, and an MHC
molecule (see Figure 3-8), both models may be correct.
That is, the absence of an MHC molecule that can bind and
present a given peptide, or the absence of T-cell receptors
that can recognize a given peptide–MHC molecule com-
plex, could result in the absence of immune responsiveness
and so account for the observed relationship between
TABLE 7-3
Differential binding of peptides to mouse class II MHC molecules and correlation
with MHC restriction
PERCENTAGE OF LABELED PEPTIDE BOUND TO
?
MHC restriction
Labeled peptide* of responders
?
IA
d
IE
d
IA
k
IE
k
Ovalbumin (323–339)IA
d
11.8 0.10.20.1
Influenza hemagglutinin (130–142)I
d
18.9 0.67.10.3
Hen egg-white lysozyme (46–61 A
k
0.00.0 35.2 0.5
Hen egg-white lysozyme (74–86)I
k
2.02.3 2.9 1.7
Hen egg-white lysozyme (81–96 E
k
0.40.20.7 1.1
Myoglobin (132–153)I
d
0.8 6.3 0.50.7
Pigeon cytochrome c (88–104)IE
k
0.61.21.7 8.7
H9261 repressor (12–26)
§
IA
d
H11001 IE
k
1.6 8.9 0.32.3
*
Amino acid residues included in each peptide are indicated by the numbers in parentheses.
?
Refers to class II molecule (IA or IE) and haplotype associated with a good response to the indicated peptides.
?
Binding determined by equilibrium dialysis. Bold-faced values indicate binding was significantly greater (p H11021 0.05) than that of the other three class II molecules
tested.
§
The H9261 repressor is an exception to the rule that high binding correlates with the MHC restriction of high-responder strains. In this case, the T
H
cell specific for the H9261
peptide–IE
d
complex has been deleted; this is an example of the hole-in-the-repertoire mechanism.
SOURCE: Adapted from S. Buus et al., 1987, Science 235:1353.
8536d_ch07_161-184 8/16/02 12:09 PM Page 178 mac100 mac 100: 1268_tm:8536d:Goldsby et al. / Immunology 5e-:
Major Histocompatibility Complex CHAPTER 7 179
MHC haplotype and immune responsiveness to exogenous
antigens.
According to the determinant-selection model, the MHC
polymorphism within a species will generate a diversity of
binding specificities, and thus different patterns of respon-
siveness to antigens. If this model is correct, then class II
MHC molecules from mouse strains that respond to a par-
ticular antigen and those that do not should show differen-
tial binding of that antigen. Table 7-3 presents data on the
binding of various radiolabeled peptides to class II IA and IE
molecules with the H-2
d
or H-2
k
haplotype. Each of the
listed peptides binds significantly to only one of the IA or IE
molecules. Furthermore, in all but one case, the haplotype of
the class II molecule showing the highest affinity for a par-
ticular peptide is the same as the haplotype of responder
strains for that peptide, as the determinant-selection model
predicts.
The single exception to the general pattern in Table 7-3
(residues 12–26 of the H9261 repressor protein) gives evidence
that the influence on immune responsiveness can also be
caused by absence of functional T cells (holes-in-the-reper-
toire model) capable of recognizing a given antigen–MHC
molecule complex. The H9261 repressor peptide binds best in
vitro to IE
d
, yet the MHC restriction for response to this pep-
tide is known to be associated not with IE
d
but instead with
IA
d
and IE
k
. This suggests that T cells recognizing this re-
pressor peptide in association with IE
d
may have been elim-
inated by negative selection in the thymus, leaving a hole in
the T-cell repertoire.
MHC and Disease Susceptibility
Some HLA alleles occur at a much higher frequency in those
suffering from certain diseases than in the general popula-
tion. The diseases associated with particular MHC alleles
include autoimmune disorders, certain viral diseases,
disorders of the complement system, some neurologic disor-
ders, and several different allergies. The association between
HLA alleles and a given disease may be quantified by deter-
mining the frequency of the HLA alleles expressed by indi-
viduals afflicted with the disease, then comparing these data
with the frequency of the same alleles in the general popula-
tion. Such a comparison allows calculation of relative risk
(see Table 7-4). A relative risk value of 1 means that the HLA
allele is expressed with the same frequency in the patient and
general populations, indicating that the allele confers no in-
creased risk for the disease. A relative risk value substantially
TABLE 7-4 Some significant associations of HLA alleles with increased risk for various diseases
Disease Associated HLA allele Relative risk*
Ankylosing spondylitis B27 90
Goodpasture’s syndrome DR2 16
Gluten-sensitive enteropathy DR3 12
Hereditary hemochromatosis A3 9.3
B14 2.3
A3/B14 90
Insulin-dependent diabetes mellitus DR4/DR3 20
Multiple sclerosis DR2 5
Myasthenia gravis DR3 10
Narcolepsy DR2 130
Reactive arthritis (Yersinia, Salmonella, Gonococcus) B27 18
Reiter’s syndrome B27 37
Rheumatoid arthritis DR4 10
Sjogren’s syndrome Dw3 6
Systemic lupus erythematosus DR3 5
*
Relative risk is calculated by dividing the frequency of the HLA allele in the patient population by the frequency in the general population:
(Ag
H11001
/Ag
H11002
) disease
RR H11005
(Ag
H11001
/Ag
H11002
) control
SOURCE: Data from SAM CD: A Comprehensive Knowledge Base of Internal Medicine, D. C. Dale and D. D. Federman, eds., 1997, Scientific American, New York.
8536d_ch07_161-184 8/15/02 8:41 PM Page 179 mac114 Mac 114:2nd shift:
180 PART II Generation of B-Cell and T-Cell Responses
above 1 indicates an association between the HLA allele and
the disease. As Table 7-4 shows, individuals with the HLA-
B27 allele have a 90 times greater likelihood (relative risk of
90) of developing the autoimmune disease ankylosing
spondylitis, an inflammatory disease of vertebral joints
characterized by destruction of cartilage, than do individu-
als with a different HLA-B allele.
The existence of an association between an MHC allele
and a disease should not be interpreted to imply that the ex-
pression of the allele has caused the disease—the relationship
between MHC alleles and development of disease is complex.
In the case of ankylosing spondylitis, for example, it has been
suggested that because of the close linkage of the TNF-H9251 and
TNF-H9252 genes with the HLA-B locus, these cytokines may be
involved in the destruction of cartilage. An association of
HLA class I genes with the disease hereditary hemochro-
matosis is discussed in the Clinical Focus box in this chapter.
When the associations between MHC alleles and disease
are weak, reflected by low relative risk values, it is likely that
multiple genes influence susceptibility, of which only one is
in the MHC. That these diseases are not inherited by simple
Mendelian segregation of MHC alleles can be seen in identi-
cal twins; both inherit the MHC risk factor, but it is by no
means certain that both will develop the disease. This find-
ing suggests that multiple genetic and environmental factors
have roles in the development of disease, especially autoim-
mune diseases, with the MHC playing an important but not
exclusive role. An additional difficulty in associating a par-
ticular MHC product with disease is the genetic phenome-
non of linkage disequilibrium, which was described above.
The fact that some of the class I MHC alleles are in linkage
disequilibrium with the class II MHC alleles makes their
contribution to disease susceptibility appear more pro-
nounced than it actually is. If, for example, DR4 contributes
to risk of a disease, and if it occurs frequently in combination
with A3 because of linkage disequilibrium, then A3 would
incorrectly appear to be associated with the disease. Im-
proved genomic mapping techniques make it possible to an-
alyze the linkage between the MHC and various diseases
more fully and to assess the contributions from other loci.
milligrams per day; of this, only 1 to 2 mg
is absorbed. The iron balance is main-
tained by control of its absorption from di-
gested food in the intestinal tract. The
primary defect in HH is increased gas-
trointestinal uptake of iron and, as a result
of this, patients with HH may throughout
their lives accumulate 15 to 35 grams of
iron instead of the normal 3 to 4 grams.
The iron overload results in pathologic ac-
cumulation of iron in cells of many or-
gans, including the heart and liver.
Although a severe form of HH may result
in heart disease in children, the clinical
manifestations of the disease are not usu-
ally seen until 40 to 50 years of age. Males
are affected eight times more frequently
than females. Early symptoms of HH are
rather nonspecific and include weakness,
lethargy, abdominal pain, diabetes, impo-
tence, and severe joint pain. Physical ex-
amination of HH sufferers reveals liver
damage, skin pigmentation, arthritis, en-
Hereditaryhemochro-
matosis (HH) is a disease in which
defective regulation of dietary iron ab-
sorption leads to increased levels of
iron. HH (which in earlier reports may be
referred to as idiopathic or primary he-
mochromatosis) is the most common
known autosomal recessive genetic dis-
order in North Americans of European
descent, with a frequency of 3–4 cases
per 1000 persons. Recent studies show
that this disease is associated with a mu-
tation in the nonclassical class I gene
HFE (formerly designated HLA-H),
which lies to the telomeric side of HLA-A.
The association of the HFE gene with
HH is an example of how potentially life-
saving clinical information can be ob-
tained by studying the connection of
HLA genes with disease.
The total iron content of a normal hu-
man adult is 3 to 4 grams; the average di-
etary intake of iron is about 10 to 20
CLINICAL FOCUS
HFE and Hereditary
Hemochromatosis
High-magnification iron stain of
liver cells from HH patient. The
stain confirms the presence of
iron in both parenchymal cells
(thick arrow) and bile duct cells
(thin arrow). This woman with
hemochromatosis required
removal of 72 units (about 36
liters or 9 gallons) of blood dur-
ing one and a half years to ren-
der her liver free of excess iron.
[SAM CD: A Comprehensive
Knowledge Base of Internal
Medicine, D. C. Dale and D. D.
Federman, eds., 1997, Scientific
American, New York.]
8536d_ch07_161-184 9/6/02 11:40 AM Page 180 mac48 Mac 48: 420_kec:
Major Histocompatibility Complex CHAPTER 7 181
association is well documented, but the
relatively high frequency of the HLA-A3 al-
lele (present in 20% of the North Ameri-
can population) makes this an
inadequate marker; the majority of indi-
viduals with HLA-A3 will not have HH.
Further studies showed a greatly in-
creased relative risk in individuals with
the combination of HLA-A3 and HLA-
B14; homozygotes for these two alleles
carried a relative risk for HH of 90. De-
tailed studies of several populations in
the US and France with high incidence of
HH revealed a mutation in the nonclassi-
cal HLA class I gene HFE in 83%–100%
of patients with HH. HFE, which lies
close to the HLA-A locus, was shown in
several independent studies to carry a
characteristic mutation at position 283 in
HH patients, with substitution of a tyro-
sine residue for the cysteine normally
found at this position. The substitution
precludes formation of the disulfide link
between cysteines in the H92513 domain,
which is necessary for association of the
MHC H9251 chain with H9252
2
-microglobulin and
for expression on the cell surface. HFE
molecules are normally expressed on the
surface of cells in the stomach, in-
testines, and liver. There is evidence
showing that HFE plays a role in the abil-
ity of these organs to regulate iron uptake
from the circulation. The mechanism by
which HFE functions involves binding to
the transferrin receptor, which reduces
the affinity of the receptor for iron-loaded
transferrin. This lowers the uptake of iron
by the cell. Mutations that interfere with
the ability of HFE to form a complex with
transferrin and its receptor can lead to in-
creased iron absorption and HH.
There are several possible reasons for
why this defect continues to be so com-
mon in our population. Factors that favor
the spread of the defective HFE gene
would include the fact that it is a reces-
sive trait, so only homozygotes are af-
fected; the gene is silent in carriers. In
addition, even in most homozygotes af-
fected with HH, the disease does not
manifest itself until later in life and so
may have minimal influence on the
breeding success of the HH sufferer.
Studies of knockout mice that lack
the gene for H9252
2
-microglobulin demon-
strate that MHC class I products on cell
surfaces are necessary for the mainte-
nance of normal iron metabolism. These
mice, which are unable to express any of
their class I molecules on the cell sur-
faces, suffer from iron overload with dis-
ease consequences similar to HH.
larged spleen, jaundice, and peripheral
edema. If untreated, HH results in hepatic
cancer, liver failure, severe diabetes, and
heart disease. Exactly how the increase in
iron content results in these diseases is
not known, but repeated phlebotomy (tak-
ing blood) is an effective treatment if the
disease is recognized before there is ex-
tensive damage to organs. Phlebotomy
does not reverse damage already done.
Phlebotomy (also called blood-letting)
was used as treatment for many condi-
tions in former times; HH may be one of
the rare instances in which the treatment
had a positive rather than a harmful effect
on the patient.
Prior to appearance of the recognized
signs of the disease, such as the charac-
teristic skin pigmentation or liver dys-
function, diagnosis is difficult unless for
some reason (such as family history of
the disease) HH is suspected and spe-
cific tests for iron metabolism are per-
formed. A reliable genetic test for HH
would allow treatment to commence
prior to disease manifestation and irre-
versible organ damage.
Because it is a common disease, the
association of HH with HLA was studied;
initially a significant association with the
HLA-A3 allele was found (RR of 9.3). This
FIGURE 7-17 Cheetah female with two nearly full grown cubs.
Polymorphism in MHC genes of the cheetah is very limited, presum-
ably because of a bottleneck in breeding that occurred in the not too
distant past. It is assumed that all cheetahs alive today are descen-
dants of a very small breeding pool. [Photograph taken in the Oka-
vango Delta, Botswana, by T. J. Kindt.]
A number of hypotheses have been offered to account for
the role of the MHC in disease susceptibility. As noted ear-
lier, allelic differences may yield differences in immune re-
sponsiveness arising from variation in the ability to present
processed antigen or the ability of T cells to recognize pre-
sented antigen. Allelic forms of MHC genes may also encode
molecules that are recognized as receptors by viruses or bac-
terial toxins. As will be explained in Chapter 16, the genetic
analysis of disease must consider the possibility that genes at
multiple loci may be involved and that complex interactions
among them may be needed to trigger disease.
Some evidence suggests that a reduction in MHC poly-
morphism within a species may predispose that species to
infectious disease. Cheetahs and certain other wild cats, such
as Florida panthers, that have been shown to be highly sus-
ceptible to viral disease have very limited MHC polymor-
phism. It is postulated that the present cheetah population
(Figure 7-17) arose from a limited breeding stock, causing a
loss of MHC diversity. The increased susceptibility of chee-
tahs to various viral diseases may result from a reduction in
8536d_ch07_161-184 9/6/02 11:40 AM Page 181 mac48 Mac 48: 420_kec:
182 PART II Generation of B-Cell and T-Cell Responses
the number of different MHC molecules available to the
species as a whole and a corresponding limitation on the
range of processed antigens with which these MHC mole-
cules can interact. Thus, the high level of MHC polymor-
phism that has been observed in various species may provide
the advantage of a broad range of antigen-presenting MHC
molecules. Although some individuals within a species
probably will not be able to develop an immune response to
any given pathogen and therefore will be susceptible to in-
fection by it, extreme polymorphism ensures that at least
some members of a species will be able to respond and will
be resistant. In this way, MHC diversity appears to protect a
species from a wide range of infectious diseases.
SUMMARY
a73
The major histocompatibility complex (MHC) comprises
a stretch of tightly linked genes that encode proteins asso-
ciated with intercellular recognition and antigen presenta-
tion to T lymphocytes.
a73
A group of linked MHC genes is generally inherited as a
unit from parents; these linked groups are called haplo-
types.
a73
MHC genes are polymorphic in that there are large num-
bers of alleles for each gene, and they are polygenic in that
there are a number of different MHC genes.
a73
Class I MHC molecules consist of a large glycoprotein
chain with 3 extracellular domains and a transmembrane
segment, and H9252
2
-microglobulin, a protein with a single
domain.
a73
Class II MHC molecules are composed of two noncova-
lently associated glycoproteins, the H9251 and H9252 chain, en-
coded by separate MHC genes.
a73
X-ray crystallographic analyses reveal peptide-binding
clefts in the membrane-distal regions of both class I and
class II MHC molecules.
a73
Both class I and class II MHC molecules present antigen to
T cells. Class I molecules present processed endogenous
antigen to CD8 T cells. Class II molecules present pro-
cessed exogenous antigen to CD4 T cells.
a73
Certain conserved motifs in peptides influence their abil-
ity to interact with the membrane-distal regions of class I
and class II MHC molecules.
a73
Class I molecules are expressed on most nucleated cells;
class II antigens are restricted to B cells, macrophages, and
dendritic cells.
a73
The class III region of the MHC encodes molecules that
include a diverse group of proteins that play no role in
antigen presentation.
a73
Detailed maps of the human and mouse MHC reveal the
presence of genes involved in antigen processing, includ-
ing proteasomes and transporters.
a73
Studies with mouse strains have shown that MHC haplo-
type influences immune responsiveness and the ability to
present antigen.
a73
Increased susceptibility to a number of diseases, predomi-
nantly, but not exclusively, of an autoimmune nature, has
been linked to certain MHC alleles.
References
Brown, J. H., et al. 1993. Three-dimensional structure of the hu-
man class II histocompatibility antigen HLA-DR1. Nature
364:33.
Drakesmith, H., and A. Townsend. 2000. The structure and
function of HFE. BioEssays. 22:595.
Fahrer, A. M., et al. 2001. A genomic view of immunology. Na-
ture 409:836.
International Human Genome Sequencing Consortium. 2001.
Initial sequencing and analysis of the human genome. Nature
409:860.
Madden, D. R. 1995. The three-dimensional structure of pep-
tide-MHC complexes. Annu. Rev. Immunol. 13:587.
Margulies, D. 1999. The major histocompatibility complex. in
Fundamental Immunology, 4th ed. W. E. Paul, ed. Lippincott
Raven, Philadelphia.
Meyer, D., and G. Thompson. 2001. How selection shapes vari-
ation of the human major histocompatibility complex: a re-
view. Ann. Hum. Genet. 65:1.
Natarajan, K., et al. 1999. MHC class I molecules, structure and
function. Revs. in Immunogenetics 1:32.
Parham, P. 1999. Virtual reality in the MHC. Immunol. Revs.
167:5.
Rothenberg, B. E., and J. R. Voland. 1996. Beta 2 knockout mice
develop parenchymal iron overload: A putative role for class I
genes of the major histocompatibility complex in iron metab-
olism. Proc. Natl. Acad. Sci. U.S.A. 93:1529.
Rouas-Freiss, N., et al. 1997. Direct evidence to support the role
of HLA-G in protecting the fetus from maternal uterine nat-
ural killer cytolysis. Proc. Natl. Acad. Sci. U.S.A. 94:11520.
Vyse, T. J., and J. A. Todd. 1996. Genetic analysis of autoimmune
disease. Cell 85:311.
Yung, Y. C., et al. 2000. The human and mouse class III region: a
parade of 21 genes at the centromeric segment. Immunol. To-
day 21:320.
USEFUL WEB SITES
http://www.bioscience.org/knockout/b2micrgl.htm
for beta-2 microglobulin KO
http://www.bioscience.org/knockout/mhci.htm
for MHC class I KO
Go to www.whfreeman.com/immunology Self-Test
Review and quiz of key terms
8536d_ch07_161-184 8/15/02 8:41 PM Page 182 mac114 Mac 114:2nd shift:
http://www.bioscience.org/knockout/mhcii.htm
for KO of an MHC class II chain
http://www.bioscience.org/knockout/mhc2inva.htm
for KO of the invariant chain
This series of destinations in the Bioscience Web site provides
updated information on studies of the consequences of tar-
geted disruption of MHC molecules and other component
molecules including H9252
2
microglobulin and the class II invari-
ant chain.
http://www.bshi.org.uk/
British Society for Histocompatibility and Immunogenetics
home page contains information on tissue typing, transplan-
tation, and links to worldwide sites concerned with MHC.
http://www.ebi.ac.uk/imgt/hla/
The International ImMunoGeneTics (IMGT) database sec-
tion contains links concerned with HLA gene structure and
genetics. It also contains up-to-date listings and sequences for
all HLA alleles officially recognized by the World Health Or-
ganization HLA nomenclature committee.
Study Questions
CLINICAL FOCUS QUESTION Almost 90% of Caucasians homozy-
gous for a mutation in position 283 of the HFE gene have clinical
signs of hemochromatosis. The fact that 10% of those with the
mutation are not affected causes a critic of the work to state that
the HFE is not involved with HH. She contends that this associa-
tion is just a result of linkage disequilibrium. How would you an-
swer her? Can you design an experiment to shed further light on
this association?
1. Indicate whether each of the following statements is true or
false. If you think a statement is false, explain why.
a. A monoclonal antibody specific for H9252
2
-microglobulin
can be used to detect both class I MHC K and D mole-
cules on the surface of cells.
b. Antigen-presenting cells express both class I and class II
MHC molecules on their membranes.
c. Class III MHC genes encode membrane-bound proteins.
d. In outbred populations, an individual is more likely to
be histocompatible with one of its parents than with its
siblings.
e. Class II MHC molecules typically bind to longer peptides
than do class I molecules.
f. All cells express class I MHC molecules.
g. The majority of the peptides displayed by class I and class
II MHC molecules on cells are derived from self-proteins.
2. You wish to produce a syngeneic and a congenic mouse
strain. Indicate whether each of the following characteristics
applies to production of syngeneic (S), congenic (C), or both
(S and C) mice.
a. Requires the greatest number of generations
b. Requires backcrosses
c. Yields mice that are genetically identical
d. Requires selection for homozygosity
e. Requires sibling crosses
f. Can be started with outbred mice
g. Yields progeny that are genetically identical to the parent
except for a single genetic region
3. You have generated a congenic A.B mouse strain that has
been selected for its MHC haplotype. The haplotype of
strain A was a/a and of strain B was b/b.
a. Which strain provides the genetic background of this
mouse?
b. Which strain provides the haplotype of the MHC of this
mouse?
c. To produce this congenic strain, the F1 progeny are al-
ways backcrossed to which strain?
d. Why was backcrossing to one of the parents performed?
e. Why was interbreeding of the F
1
and F
2
progeny per-
formed?
f. Why was selection necessary and what kind of selection
was performed?
4. You cross a BALB/c (H-2
d
) mouse with a CBA (H-2
k
)
mouse. What MHC molecules will the F
1
progeny express on
(a) its liver cells and (b) its macrophages?
5. To carry out studies on the structure and function of the
class I MHC molecule K
b
and the class II MHC molecule
IA
b
, you decide to transfect the genes encoding these pro-
teins into a mouse fibroblast cell line (L cell) derived from
the C3H strain (H-2
k
). L cells do not normally function as
antigen-presenting cells. In the following table, indicate
which of the listed MHC molecules will (H11001) or will not (H11002)
be expressed on the membrane of the transfected L cells.
Major Histocompatibility Complex CHAPTER 7 183
MHC molecules expressed
on the membrane of the
transfected L cells
Transfected gene D
k
D
b
K
k
K
b
IA
k
IA
b
None
K
b
IAH9251
b
IAH9252
b
IAH9251
b
and IAH9252
b
6. The SJL mouse strain, which has the H-2
k
haplotype, has a
deletion of the IEH9251 locus.
a. List the classical MHC molecules that are expressed on
the membrane of macrophages from SJL mice.
b. If the class II IEH9251 and IEH9252 genes from an H-2
s
strain are
transfected into SJL macrophages, what additional clas-
sical MHC molecules would be expressed on the trans-
fected macrophages?
7. Draw diagrams illustrating the general structure, including
the domains, of class I MHC molecules, class II MHC mole-
cules, and membrane-bound antibody on B cells. Label each
8536d_ch07_161-184 8/15/02 8:41 PM Page 183 mac114 Mac 114:2nd shift:
chain and the domains within it, the antigen-binding regions,
and regions that have the immunoglobulin-fold structure.
8. One of the characteristic features of the MHC is the large
number of different alleles at each locus.
a. Where are most of the polymorphic amino acid residues
located in MHC molecules? What is the significance of
this location?
b. How is MHC polymorphism thought to be generated?
9. As a student in an immunology laboratory class, you have
been given spleen cells from a mouse immunized with the
LCM virus.You determine the antigen-specific functional ac-
tivity of these cells with two different assays. In assay 1, the
spleen cells are incubated with macrophages that have been
briefly exposed to the LCM virus; the production of inter-
leukin 2 (IL-2) is a positive response. In assay 2, the spleen
cells are incubated with LCM-infected target cells; lysis of the
target cells represents a positive response in this assay. The re-
sults of the assays using macrophages and target cells of dif-
ferent haplotypes are presented in the table below. Note that
the experiment has been set up in a way to exclude alloreac-
tive responses (reactions against nonself MHC molecules).
a. The activity of which cell population is detected in each
of the two assays?
b. The functional activity of which MHC molecules is de-
tected in each of the two assays?
c. From the results of this experiment, which MHC mole-
cules are required, in addition to the LCM virus, for spe-
cific reactivity of the spleen cells in each of the two assays?
d. What additional experiments could you perform to un-
ambiguously confirm the MHC molecules required for
antigen-specific reactivity of the spleen cells?
e. Which of the mouse strains listed in the table below could
have been the source of the immunized spleen cells tested
in the functional assays? Give your reasons.
10. A T
C
-cell clone recognizes a particular measles virus peptide
when it is presented by H-2D
b
. Another MHC molecule has
a peptide-binding cleft identical to the one in H-2D
b
but dif-
fers from H-2D
b
at several other amino acids in the H92511H92521
domain. Predict whether the second MHC molecule could
present this measles virus peptide to the T
C
-cell clone.
Briefly explain your answer.
11. How can you determine if two different inbred mouse
strains have identical MHC haplotypes?
12. Human red blood cells are not nucleated and do not express
any MHC molecules. Why is this property fortuitous for
blood transfusions?
13. The hypothetical allelic combination HLA-A99 and HLA-
B276 carries a relative risk of 200 for a rare, and yet un-
named, disease that is fatal to pre-adolescent children.
a. Will every individual with A99/B276 contract the disease?
b. Will everyone with the disease have the A99/B276 combi-
nation?
c. How frequently will the A99/B276 allelic combination be
observed in the general population? Do you think that
this combination will be more or less frequent than pre-
dicted by the frequency of the two individual alleles?
Why?
184 PART II Generation of B-Cell and T-Cell Responses
Response of spleen cells
Mouse strain MHC haplotype of macrophages
used as source of and virus-infected target cells IL-2 production in Lysis of LCM-
macrophages and response to LCM-pulsed infected cells
target cells K IA IE D macrophages (assay 1) (assay 2)
C3H kkk k H11001H11002
BALB/c ddd d H11002H11001
(BALB/c H11003 B10.A)F
1
d/k d/k d/k d/d H11001H11001
A.TL skkd H11001H11001
B10.A (3R) bbb d H11002H11001
B10.A (4R) kkH11002 b H11001H11002
For use with Question 9.
8536d_ch07_161-184 8/15/02 8:41 PM Page 184 mac114 Mac 114:2nd shift: