238
6 Immunoproteins
the N-CAM of another cell. The N-terminal
extracellular portion of N-CAM contains five Ig
homology units and two segments corresponding
to type III fibrinogen repeats (Fig. 6.2). There is
a total of seven glycosylation sites on the third to
fifth Ig domains; the carbohydrate chains of the
fifth domain carry a(2-8)-linked polysialic acids
which make up 20 % of the N-CAM molecular
mass in embryonic brain but only 9 % in the adult
brain. The homophilic binding of the N-CAM
molecules is hindered by mutual repulsion of the
negatively charged sialic acids; consequently,
binding is stronger in the adult animal. Three NCAMs of different size are found in chicken and
mouse brain with deglycosylated molecular masses of 180 kDa (ld form), 140 kDa (sd form) and
120 kDa (ssd form). All three contain the same
N-terminal extracellular sequence of about 862
amino acids. In the ld form, the transmembrane
segment is followed by a large cytoplasmic
domain which is 261 amino acids smaller in the sd
form. There is no transmembrane or cytoplasmic
domain in the ssd form; in this case, the extracellular chain is bound to a short non-polar sequence
which is anchored to the membrane via phosphatidylinositol. The three forms arise by alternative
splicing of a primary transcript from a gene with
19 exons. It has recently been found that the
mRNA of mouse brain also has two sites in the
extracellular region in which 3, 18 or 30 nucleotides can be introduced by alternative splicing. At
least 27 alternatively spliced forms of the N-CAM
mRNA are expressed during rat heart development [144]. The N-CAM of Xenopus laevis brain,
which has been sequenced via the cDNA, corresponds approximately to the ld form of the higher
vertebrates, and much of the sequence is in common [93]. Further CAMs from the Ig super-family
have been described in vertebrates, e.g. the
myelin-associated glycoprotein (MAG) with five
Ig units and the neuronal L1 with six. Surprisingly, adhesion molecules from the Ig superfamily have been found in Drosophila melanogaster: amalgam with three Ig units, fascilin II
with five, and neuroglian with six [14].
The cadherins require Ca 2 + ions for their
adhesive activity; the Ca2+ -binding sites lie on the
extracellular domain but have not yet been identified. Cadherins are known from mammals, birds
and amphibians and are probably present in all
vertebrates; similar adhesion proteins have also
been found in Drosophila. The cadherins are
integral membrane proteins of 120-130 kDa with
723-748 amino acids and only one transmembrane segment. The sequence, which is partly
made up of repeats of about 110 amino acids,
agrees by at least 40 % in the bird and mammal
species examined. The first members of this
family were discovered in various tissues of the
mouse and chicken, and received corresponding
names. In spite of the multiplicity of names, all
vertebrate cadherins may be categorized into
three types: the E-cadherin found in epithelial
tissues is apparently identical to uvomorulin, LCAM and cell CAM; the neuronal N-cadherin is
identical to A-CAM and NcaICAM; and the final
class includes the placental P-cadherin. The three
classes show overlapping tissue distributions.
However, they separated from each other early in
vertebrate evolution; this may be deduced from
the fact that mouse N-cadherin is 92 % similar to
that of the chicken, but only 49 and 43 % similar
to mouse E- and P-cadherin, respectively [185].
The special binding structures between neighbouring cells (cell junctions) include specific junction proteins. The desmosomes on which the intermediary filaments are anchored contain desmoplakin I (250 kDa), and the different types of
intermediate junctions which are bound to microfilaments contain cell-specific proteins such as Ecadherin, vinculin and a-actinin. Plakoglobin
(82 kDa), which is found in both types of junction, has 744 amino acids and sequence similarities to other known proteins [51]. "Gap junctions" form channel-like connections between
cells and these allow the exchange of ions, secondary messengers and metabolites; furthermore,
their low electrical resistance facilitates the transmission of action potentials. Each cell-cell channel is composed of six identical molecules of the
protein connexin. A whole series of connexins
have been isolated from mammals, the chicken
and Xenopus laevis; they have chain lengths
between 283 and more than 400 amino acids and
molecular masses of 21-45 kDa. The connexins of
the frog and rat show 32-41 % sequence similarity, the resemblance being greatest in the four
transmembrane segments and the extracellular
domain [13, 45].
Of the SAMs, only the cellular receptors will
be considered here; their corresponding matrix
components are the subjects of Chapters 11 and
13. The majority of the receptors for matrix components belong to the family of integrins, which
are heterodimeric membrane proteins with an aand a ~-chain of 700-800 amino acids, including
56 invariant cysteine residues. Until recently, only
three types of ~-chain were known (~1, ~2 and
~3), and these combine with about ten types of a
subunit to form three integrin subfamilies: the
6 Immunoproteins
the N-CAM of another cell. The N-terminal
extracellular portion of N-CAM contains five Ig
homology units and two segments corresponding
to type III fibrinogen repeats (Fig. 6.2). There is
a total of seven glycosylation sites on the third to
fifth Ig domains; the carbohydrate chains of the
fifth domain carry a(2-8)-linked polysialic acids
which make up 20 % of the N-CAM molecular
mass in embryonic brain but only 9 % in the adult
brain. The homophilic binding of the N-CAM
molecules is hindered by mutual repulsion of the
negatively charged sialic acids; consequently,
binding is stronger in the adult animal. Three NCAMs of different size are found in chicken and
mouse brain with deglycosylated molecular masses of 180 kDa (ld form), 140 kDa (sd form) and
120 kDa (ssd form). All three contain the same
N-terminal extracellular sequence of about 862
amino acids. In the ld form, the transmembrane
segment is followed by a large cytoplasmic
domain which is 261 amino acids smaller in the sd
form. There is no transmembrane or cytoplasmic
domain in the ssd form; in this case, the extracellular chain is bound to a short non-polar sequence
which is anchored to the membrane via phosphatidylinositol. The three forms arise by alternative
splicing of a primary transcript from a gene with
19 exons. It has recently been found that the
mRNA of mouse brain also has two sites in the
extracellular region in which 3, 18 or 30 nucleotides can be introduced by alternative splicing. At
least 27 alternatively spliced forms of the N-CAM
mRNA are expressed during rat heart development [144]. The N-CAM of Xenopus laevis brain,
which has been sequenced via the cDNA, corresponds approximately to the ld form of the higher
vertebrates, and much of the sequence is in common [93]. Further CAMs from the Ig super-family
have been described in vertebrates, e.g. the
myelin-associated glycoprotein (MAG) with five
Ig units and the neuronal L1 with six. Surprisingly, adhesion molecules from the Ig superfamily have been found in Drosophila melanogaster: amalgam with three Ig units, fascilin II
with five, and neuroglian with six [14].
The cadherins require Ca 2 + ions for their
adhesive activity; the Ca2+ -binding sites lie on the
extracellular domain but have not yet been identified. Cadherins are known from mammals, birds
and amphibians and are probably present in all
vertebrates; similar adhesion proteins have also
been found in Drosophila. The cadherins are
integral membrane proteins of 120-130 kDa with
723-748 amino acids and only one transmembrane segment. The sequence, which is partly
made up of repeats of about 110 amino acids,
agrees by at least 40 % in the bird and mammal
species examined. The first members of this
family were discovered in various tissues of the
mouse and chicken, and received corresponding
names. In spite of the multiplicity of names, all
vertebrate cadherins may be categorized into
three types: the E-cadherin found in epithelial
tissues is apparently identical to uvomorulin, LCAM and cell CAM; the neuronal N-cadherin is
identical to A-CAM and NcaICAM; and the final
class includes the placental P-cadherin. The three
classes show overlapping tissue distributions.
However, they separated from each other early in
vertebrate evolution; this may be deduced from
the fact that mouse N-cadherin is 92 % similar to
that of the chicken, but only 49 and 43 % similar
to mouse E- and P-cadherin, respectively [185].
The special binding structures between neighbouring cells (cell junctions) include specific junction proteins. The desmosomes on which the intermediary filaments are anchored contain desmoplakin I (250 kDa), and the different types of
intermediate junctions which are bound to microfilaments contain cell-specific proteins such as Ecadherin, vinculin and a-actinin. Plakoglobin
(82 kDa), which is found in both types of junction, has 744 amino acids and sequence similarities to other known proteins [51]. "Gap junctions" form channel-like connections between
cells and these allow the exchange of ions, secondary messengers and metabolites; furthermore,
their low electrical resistance facilitates the transmission of action potentials. Each cell-cell channel is composed of six identical molecules of the
protein connexin. A whole series of connexins
have been isolated from mammals, the chicken
and Xenopus laevis; they have chain lengths
between 283 and more than 400 amino acids and
molecular masses of 21-45 kDa. The connexins of
the frog and rat show 32-41 % sequence similarity, the resemblance being greatest in the four
transmembrane segments and the extracellular
domain [13, 45].
Of the SAMs, only the cellular receptors will
be considered here; their corresponding matrix
components are the subjects of Chapters 11 and
13. The majority of the receptors for matrix components belong to the family of integrins, which
are heterodimeric membrane proteins with an aand a ~-chain of 700-800 amino acids, including
56 invariant cysteine residues. Until recently, only
three types of ~-chain were known (~1, ~2 and
~3), and these combine with about ten types of a
subunit to form three integrin subfamilies: the