378
11 Extracellular Structural and Secretory Proteins
ponents C3 and C1q, C-reactive proteins, amyloid P, acetylcholinesterase, DNA, and bacterial
and animal cells [109, 116]. Furthermore, they
constitute the recognition and binding sites for
the Chagas disease agent Trypanosoma cruzi.
Each function has a specific binding site somewhere on the long polypeptide chain (Fig. 11.1).
Fibronectins have been found in all vertebrates
examined, and related proteins appear to be present in all Metazoa from the Porifera onwards.
An antibody against human fibronectin recognizes a glycoprotein of about 230 kDa in the haemolymph of the fly Drosophila, and this glycoprotein is apparently quite similar to the vertebrate
fibronectin [45].
The soluble plasma fibronectins of the rat and
other mammals can be clearly distinguished from
the insoluble cellular forms by means of monoclonal antibodies. Both show marked variability,
which is partly due to post-translational modifications, such as glycosylation or sulphation of tyrosine residues, but also partly due to differences in
the primary sequences [109]. Man and the rat
have only one fibronectin gene, from which, however, 12 different mRNAs are formed by alternative splicing at three positions (see Fig.2.12a,
p. 41). It is not known how far the different fibronectins exercise different functions [94, 109]. The
rat fibronectin gene extends for about 70 kb and
has almost 50 exons. Each repeated unit of the
type-land -II fibronectins is encoded by a specific
exon, whereas the repeats of type III arise from
two somewhat different exons. The primary
translation product contains a signal sequence of
20 amino acids and a pro-sequence of 5 amino
acids [47]. The gene structure confirms the participation of gene duplication and exon shuffling in
fibronectin evolution, but this is already suggested by the periodic amino acid sequence [96].
The blood plasma of various mammals and the
chicken contains a further cell-adhesive glycoprotein, vitronectin, which, with a length of 458
amino acids (75 kDa), is significantly smaller than
fibronectin and shows no sequence similarity [70].
11.2 Collagens
In the early 1970s only one collagen type was
recognised, and that is now known as collagen I
from the extracellular matrix of the vertebrate
connective tissues. Today, no fewer than 14 vertebrate collagen types are known, and these are
made up of 25 different polypeptide chains. Further collagen types are found in the invertebrates.
Collagens form fibrils in the extracellular matrix
and networks in the basement membranes; they
occur, for example, in the cuticula of the annelids
and nematodes, in the egg capsules of sharks,
skates and rays, in the cocoon silks of some insect
larvae, in the byssus threads of the edible mussels, and in the gemmules of the sponges.
Quantity-wise, they are amongst the most prevalent proteins; 25-30 % of total human protein
consists of collagen, and there are collagen-like
sequences in other proteins. For example, the
asymmetric, basement-membrane-bound form of
acetylcholinesterase has a 40-kDa segment with
the triple-helix structure typical of collagen, and
shows immunological cross-reactivity with collagen I. The three chains of the complement factor
C1q also have a short triple helix close to the Nterminus, which suggests the combination of a
collagen sequence with a non-collagen sequence
during evolution. The neurophysins and various
salivary proteins possess collagen-like fragments
[35, 150].
11.2.1 Vertebrate Collagens
The typical collagen fibre of the mammalian
connective-tissue matrix is a bundle of fibrils with
more than 1 million collagen molecules per centimetre, each of which is shifted 67 nm with respect
to the others by one so-called D unit, corresponding to 234 amino acids, giving the typical crossstriping of the fibrils (Fig. l1.2b). Each collagen
molecule is a stiff, rod-shaped structure of about
300 kDa with a length of 300 nm and a diameter
of 1.5 nm; it is made up of three a chains which
are, for the most part, twisted into a triple helix
(Fig. 11.2a). The central part of each chain consists predominantly of a repetitive sequence of
about 330 triplets with the structure -Gly-X-Y-,
in which X and Yare often proline and 4hydroxyproline, respectively. At both ends of
the a chains are so-called telopeptides, of
15-20 amino acids, which have no triple-helix
structure. As in the case of other secreted proteins, the newly synthesized a chains have an Nterminal signal sequence. After cleavage of the
latter in the cisternae of the endoplasmic reticulum, three chains associate to form the procollagen molecule which is then released to the
outside. The a chains of pro-collagen carry Nand C-terminal extensions which are cleaved off
by specific amino-and carboxypro-collagen peptidases during fibril formation. Further post-
11 Extracellular Structural and Secretory Proteins
ponents C3 and C1q, C-reactive proteins, amyloid P, acetylcholinesterase, DNA, and bacterial
and animal cells [109, 116]. Furthermore, they
constitute the recognition and binding sites for
the Chagas disease agent Trypanosoma cruzi.
Each function has a specific binding site somewhere on the long polypeptide chain (Fig. 11.1).
Fibronectins have been found in all vertebrates
examined, and related proteins appear to be present in all Metazoa from the Porifera onwards.
An antibody against human fibronectin recognizes a glycoprotein of about 230 kDa in the haemolymph of the fly Drosophila, and this glycoprotein is apparently quite similar to the vertebrate
fibronectin [45].
The soluble plasma fibronectins of the rat and
other mammals can be clearly distinguished from
the insoluble cellular forms by means of monoclonal antibodies. Both show marked variability,
which is partly due to post-translational modifications, such as glycosylation or sulphation of tyrosine residues, but also partly due to differences in
the primary sequences [109]. Man and the rat
have only one fibronectin gene, from which, however, 12 different mRNAs are formed by alternative splicing at three positions (see Fig.2.12a,
p. 41). It is not known how far the different fibronectins exercise different functions [94, 109]. The
rat fibronectin gene extends for about 70 kb and
has almost 50 exons. Each repeated unit of the
type-land -II fibronectins is encoded by a specific
exon, whereas the repeats of type III arise from
two somewhat different exons. The primary
translation product contains a signal sequence of
20 amino acids and a pro-sequence of 5 amino
acids [47]. The gene structure confirms the participation of gene duplication and exon shuffling in
fibronectin evolution, but this is already suggested by the periodic amino acid sequence [96].
The blood plasma of various mammals and the
chicken contains a further cell-adhesive glycoprotein, vitronectin, which, with a length of 458
amino acids (75 kDa), is significantly smaller than
fibronectin and shows no sequence similarity [70].
11.2 Collagens
In the early 1970s only one collagen type was
recognised, and that is now known as collagen I
from the extracellular matrix of the vertebrate
connective tissues. Today, no fewer than 14 vertebrate collagen types are known, and these are
made up of 25 different polypeptide chains. Further collagen types are found in the invertebrates.
Collagens form fibrils in the extracellular matrix
and networks in the basement membranes; they
occur, for example, in the cuticula of the annelids
and nematodes, in the egg capsules of sharks,
skates and rays, in the cocoon silks of some insect
larvae, in the byssus threads of the edible mussels, and in the gemmules of the sponges.
Quantity-wise, they are amongst the most prevalent proteins; 25-30 % of total human protein
consists of collagen, and there are collagen-like
sequences in other proteins. For example, the
asymmetric, basement-membrane-bound form of
acetylcholinesterase has a 40-kDa segment with
the triple-helix structure typical of collagen, and
shows immunological cross-reactivity with collagen I. The three chains of the complement factor
C1q also have a short triple helix close to the Nterminus, which suggests the combination of a
collagen sequence with a non-collagen sequence
during evolution. The neurophysins and various
salivary proteins possess collagen-like fragments
[35, 150].
11.2.1 Vertebrate Collagens
The typical collagen fibre of the mammalian
connective-tissue matrix is a bundle of fibrils with
more than 1 million collagen molecules per centimetre, each of which is shifted 67 nm with respect
to the others by one so-called D unit, corresponding to 234 amino acids, giving the typical crossstriping of the fibrils (Fig. l1.2b). Each collagen
molecule is a stiff, rod-shaped structure of about
300 kDa with a length of 300 nm and a diameter
of 1.5 nm; it is made up of three a chains which
are, for the most part, twisted into a triple helix
(Fig. 11.2a). The central part of each chain consists predominantly of a repetitive sequence of
about 330 triplets with the structure -Gly-X-Y-,
in which X and Yare often proline and 4hydroxyproline, respectively. At both ends of
the a chains are so-called telopeptides, of
15-20 amino acids, which have no triple-helix
structure. As in the case of other secreted proteins, the newly synthesized a chains have an Nterminal signal sequence. After cleavage of the
latter in the cisternae of the endoplasmic reticulum, three chains associate to form the procollagen molecule which is then released to the
outside. The a chains of pro-collagen carry Nand C-terminal extensions which are cleaved off
by specific amino-and carboxypro-collagen peptidases during fibril formation. Further post-