from individual to individual within an inbred
line. In the anurans, one finds at the most 40
spectrotypes and in carp there are only 23 spectrotypes. Different individuals of the shark Heterodontus francisci produce essentially identical
antibodies in response to the same hapten [147].
The significantly lower Ig variety, e.g. in the
clawed frog Xenopus, is particularly puzzling
because here the gene organization and the number ofVH, D and JH segments is sImilar to that of
the mammals [163].
6.2.4 Evolution of the Immunoglobulins
The direct determination of the amino acid
sequences of the immunoglobulins is hampered
by their diversity. Earlier, use was made of the
production of immunoglobulins with a uniform V
sequence by particular clonal tumours (myeloma
proteins); the resulting excess of L chains appears
in the urine (Bence-Jones proteins) [179]. Large
quantities of uniform (monoclonal) antibodies
can also be obtained by experimental selection of
lymph cell lines. More recently, the indirect determination of amino acid sequences by DNA analysis has made it possible to obtain a wealth of
immunoglobulin sequence data. However, the
sequences obtained are almost exclusively from
man, the mouse and a few other mammals;
sequence data for lower vertebrates, which would
be essential for the analysis of relationships and
for the construction of genealogical trees, are as
yet in short supply [101].
The multiple segments of the Ig genes arose by
duplications. This can be demonstrated, for
example, with the CH genes (Fig. 6.3b). Whereas
in the mouse the four Cy genes form a tandem
cluster, in man it appears that the cluster Cy-CyCE-Ca was duplicated as a whole [19]. The duplication of Ig genes in the germline is apparently a
rare evolutionary event; the frequency of 5 . 10- 7
per gene per year for the Ig genes is about 100fold less than for the rDNA. This reduced rate of
horizontal evolution helps maintain the genetic
variability of the Ig segments but, on the other
hand, it allows the formation of pseudo genes
[58]. Almost all the known Ig pseudogenes show
relatively few defects when compared with other
pseudogenes. Thus, they probably arose quite
recently from active genes and have been constantly corrected by horizontal evolution; they
can, therefore, be easily reactivated or used in
part for the construction of new gene sequences
[74]. Gene duplications or deletions are appar6.2.4 Evolution of the Immunoglobulins
227
ently very frequent somatic mutations; 1-3 % of
human individuals are said to be heterozygous for
the deletion of single CH genes. A consideration
of the degree of similarity indicates different subgroups within the multiple V sequences. The VH
subgroup VHIII, initially defined in man and the
mouse, was easy to sequence because of its free
N-terminus and it has been especially well characterized. VHIII-like sequences are already to be
found in the shark Ginglymostoma cirratum and
are apparently quite old. Such multi-gene subgroups can either expand or contract during
evolution and may differ in extent, as for example
between man and the mouse.
The gradient of similarity between the V and C
domains of the different Hand L chains suggests
the following model for immunoglobulin evolution. Duplication of an ancestral Ig gene led at
first to V and C and the joining system with the J
segment. This was followed by duplication of the
C gene to give the CL and CH genes. The original
CH region with four domains then arose by internal duplication. From this emerged the different
CH types, of which Cy, Cb , Ca lost all but the H
(hinge) section of the CH2 domain. In agreement
with this scheme, for example, is the greater similarity between the homologous domains of Il, yl
and y2b than between the different C domains of Il
in the mouse. Most of the H chains carry terminal
extensions of unknown biological function which
are not entirely homologous to each other. The
highest substitution rate is shown by the hypervariable sections of the V region. It is also clear
from sequence comparisons of homologous C
domains of different mammalian species that the
rate of evolution of the immunoglobulin C regions is consistently high, despite large differences
between the various chain types and individual C
domains (see Table 4.12, p.161). The rate of
evolution of A is higher than that of ' X.; in most of
the H chains investigated, the rate of evolution
decreases towards the C-terminal domains. The
evolution of the H region (hinge) appears to be
particularly rapid. The immunoglobulins (HMW
and LMW) of many anuran species show immunological cross-reactivity with those of Xenopus
laevis, despite the fact that some of these species
separated more than 150 million years ago. In
this case, the evolution of at least some parts of
the immunoglobulins appears to be progressing
more slowly than in the mammals.
line. In the anurans, one finds at the most 40
spectrotypes and in carp there are only 23 spectrotypes. Different individuals of the shark Heterodontus francisci produce essentially identical
antibodies in response to the same hapten [147].
The significantly lower Ig variety, e.g. in the
clawed frog Xenopus, is particularly puzzling
because here the gene organization and the number ofVH, D and JH segments is sImilar to that of
the mammals [163].
6.2.4 Evolution of the Immunoglobulins
The direct determination of the amino acid
sequences of the immunoglobulins is hampered
by their diversity. Earlier, use was made of the
production of immunoglobulins with a uniform V
sequence by particular clonal tumours (myeloma
proteins); the resulting excess of L chains appears
in the urine (Bence-Jones proteins) [179]. Large
quantities of uniform (monoclonal) antibodies
can also be obtained by experimental selection of
lymph cell lines. More recently, the indirect determination of amino acid sequences by DNA analysis has made it possible to obtain a wealth of
immunoglobulin sequence data. However, the
sequences obtained are almost exclusively from
man, the mouse and a few other mammals;
sequence data for lower vertebrates, which would
be essential for the analysis of relationships and
for the construction of genealogical trees, are as
yet in short supply [101].
The multiple segments of the Ig genes arose by
duplications. This can be demonstrated, for
example, with the CH genes (Fig. 6.3b). Whereas
in the mouse the four Cy genes form a tandem
cluster, in man it appears that the cluster Cy-CyCE-Ca was duplicated as a whole [19]. The duplication of Ig genes in the germline is apparently a
rare evolutionary event; the frequency of 5 . 10- 7
per gene per year for the Ig genes is about 100fold less than for the rDNA. This reduced rate of
horizontal evolution helps maintain the genetic
variability of the Ig segments but, on the other
hand, it allows the formation of pseudo genes
[58]. Almost all the known Ig pseudogenes show
relatively few defects when compared with other
pseudogenes. Thus, they probably arose quite
recently from active genes and have been constantly corrected by horizontal evolution; they
can, therefore, be easily reactivated or used in
part for the construction of new gene sequences
[74]. Gene duplications or deletions are appar6.2.4 Evolution of the Immunoglobulins
227
ently very frequent somatic mutations; 1-3 % of
human individuals are said to be heterozygous for
the deletion of single CH genes. A consideration
of the degree of similarity indicates different subgroups within the multiple V sequences. The VH
subgroup VHIII, initially defined in man and the
mouse, was easy to sequence because of its free
N-terminus and it has been especially well characterized. VHIII-like sequences are already to be
found in the shark Ginglymostoma cirratum and
are apparently quite old. Such multi-gene subgroups can either expand or contract during
evolution and may differ in extent, as for example
between man and the mouse.
The gradient of similarity between the V and C
domains of the different Hand L chains suggests
the following model for immunoglobulin evolution. Duplication of an ancestral Ig gene led at
first to V and C and the joining system with the J
segment. This was followed by duplication of the
C gene to give the CL and CH genes. The original
CH region with four domains then arose by internal duplication. From this emerged the different
CH types, of which Cy, Cb , Ca lost all but the H
(hinge) section of the CH2 domain. In agreement
with this scheme, for example, is the greater similarity between the homologous domains of Il, yl
and y2b than between the different C domains of Il
in the mouse. Most of the H chains carry terminal
extensions of unknown biological function which
are not entirely homologous to each other. The
highest substitution rate is shown by the hypervariable sections of the V region. It is also clear
from sequence comparisons of homologous C
domains of different mammalian species that the
rate of evolution of the immunoglobulin C regions is consistently high, despite large differences
between the various chain types and individual C
domains (see Table 4.12, p.161). The rate of
evolution of A is higher than that of ' X.; in most of
the H chains investigated, the rate of evolution
decreases towards the C-terminal domains. The
evolution of the H region (hinge) appears to be
particularly rapid. The immunoglobulins (HMW
and LMW) of many anuran species show immunological cross-reactivity with those of Xenopus
laevis, despite the fact that some of these species
separated more than 150 million years ago. In
this case, the evolution of at least some parts of
the immunoglobulins appears to be progressing
more slowly than in the mammals.