the agnathans to the mammals. Greater differences between species are found in the C peptide
than in the A and B chains; this is related to the
fewer functional specifications of the former
(Table 4.12; p. 161). The amino acids responsible
for hormone activity, zinc binding and chain
association are highly conserved; 22 of the 51 insulin amino acids are invariant in all vertebrates
[35]. The insulins of most mammals are very similar to human insulin, differences being found at
only eight positions of the two chains (Fig. 8.4)
[72, 131]. The insulin ofthe South American douroucouli Aotes trivirgatus differs markedly from
those of other primates; compared with that of
man, Aotes prepro-insulin has three amino acid
substitutions in the signal peptide, two in the
B chain, seven in the C peptide and three in the
A chain. The receptor binding affinity of Aotes
insulin is fivefold lower than that of human insulin [198].
The most marked differences to other mammalian insulins are shown by the insulins of the Hystricomorpha, e.g. the African porcupine Hystrix
cristata, the guinea-pig Cavia porcel/us, the casiragua Proechimys quarirea, the chinchilla Chinchilla /aniger, and the nutria Myocastor coypus.
The extremely aberrant nutria insulin has only
5 % of the receptor affinity and 3 % of the effect
on lipogenesis shown by porcine insulin in a test
system involving rat fat cells. Due to amino acid
substitutions at the zinc-binding site BlO-His and
positions responsible for molecular contacts, the
hystricomorphan insulins are always monomers.
Compared with bovine insulin, that of casiragua
has 17 differences, that of the guinea-pig 18, and
that of nutria as many as 22; the spatial structures, however, remain similar [6]. Despite large
sequence differences, the insulins of the H ystricomorpha fit into the genealogical tree of the mammalian insulins and are therefore orthologous
according to the definition on p. 155.
As expected, the rlSh differ most of all from the
mammals [12,40, 108, 186]. The prepro-insulin
of the agnathan Myxine g/utinosa, which with
115 amino acids is somewhat longer than that of
the mammals, has the signal sequence (26 amino
acids), B chain (31 amino acids), C peptide (37
amino acids) and A chain (21 amino acids)
arranged consecutively. However, the mature
insulin differs at 16 positions from that of the carp
and at 21 positions from human insulin. It cannot
bind zinc and can only form dimers, whereas
insulin from the shark Scyliorrhinus canicu/a can
already form zinc-stabilized hexamers. The insulin of the lamprey Petromyzon marin us differs
8.4.2 The Insulin Family
299
from that of the common hagfish Myxine (17
amino acid differences) more than from porcine
insulin or from that of the salmon (both with
14 differences) [168]; this underlines the large
evolutionary distance between the two agnathan
groups Myxinoidea and Petromyzontoidea. There
are 17 differences between the insulin of the carp
Cyprinus carpio and human insulin, including two
extra N-terminal amino acids on the B chain;
however, the zinc-binding B10-His and the amino
acids responsible for aggregation and biological
activity are all present. Carp insulin in the fat-cell
test has only threefold weaker binding and has
the same maximum activity as porcine insulin.
Amino acid substitutions in the region of the BI
C cleavage site lead to 37- to 38-amino-acid
B chains in the holocephalans Chimaera monstruosa and Hydro/agus colliei, whereas that of the
related species Callorhynchus milii has only the
normal 31 amino acids [12].
The gene sequences of human prepro-insulin
and those of a range of mammals, the chicken,
various teleosts and Myxine are now known. All
consist of three exons: the first exon codes for a
part of the non-translated (NT) 5' sequence; the
second for the remaining 5'-NT, the signal
sequence, B chain and a part of the C peptide;
and the third for the rest of the C peptide, the
A chain and the 3'-NT [209]. The sequence coding for the signal peptide occurs once again in a
slightly modified form in the C-coding region. It
is possible that two genes with identical signal
peptide regions arose initially by gene duplication, and were then later joined by a large intron
to give one gene. There are two prepro-insulin
genes in the mouse and rat which lie only 100 kb
apart on chromosome 1 of the rat but are found
separately on chromosomes 6 and 7 of the mouse.
In both rodent species, gene I is very similar to
gene II in the coding sequence but lacks the second intron. As in the chicken insulin gene, the
presence of two introns is probably the ancestral
state; the insulin genes rat-I and mouse-I apparently arose by the reinsertion of partly processed
transcripts. It is interesting that the retrotransposons of the two species differ much more than the
original insulin genes rat-II and mouse-II [232].
Insulin-like substances have been detected
by both immunological and pharmacological
methods in a wide spectrum of very different
organisms: tunicates, insects, crustaceans, gastropods and cephalopods, and even in the ciliate Tetrahymena, the lower fungi Aspergillus and Neurospora, and the prokaryote E. coli [68, 118].
Insulin-receptor proteins have been identified
than in the A and B chains; this is related to the
fewer functional specifications of the former
(Table 4.12; p. 161). The amino acids responsible
for hormone activity, zinc binding and chain
association are highly conserved; 22 of the 51 insulin amino acids are invariant in all vertebrates
[35]. The insulins of most mammals are very similar to human insulin, differences being found at
only eight positions of the two chains (Fig. 8.4)
[72, 131]. The insulin ofthe South American douroucouli Aotes trivirgatus differs markedly from
those of other primates; compared with that of
man, Aotes prepro-insulin has three amino acid
substitutions in the signal peptide, two in the
B chain, seven in the C peptide and three in the
A chain. The receptor binding affinity of Aotes
insulin is fivefold lower than that of human insulin [198].
The most marked differences to other mammalian insulins are shown by the insulins of the Hystricomorpha, e.g. the African porcupine Hystrix
cristata, the guinea-pig Cavia porcel/us, the casiragua Proechimys quarirea, the chinchilla Chinchilla /aniger, and the nutria Myocastor coypus.
The extremely aberrant nutria insulin has only
5 % of the receptor affinity and 3 % of the effect
on lipogenesis shown by porcine insulin in a test
system involving rat fat cells. Due to amino acid
substitutions at the zinc-binding site BlO-His and
positions responsible for molecular contacts, the
hystricomorphan insulins are always monomers.
Compared with bovine insulin, that of casiragua
has 17 differences, that of the guinea-pig 18, and
that of nutria as many as 22; the spatial structures, however, remain similar [6]. Despite large
sequence differences, the insulins of the H ystricomorpha fit into the genealogical tree of the mammalian insulins and are therefore orthologous
according to the definition on p. 155.
As expected, the rlSh differ most of all from the
mammals [12,40, 108, 186]. The prepro-insulin
of the agnathan Myxine g/utinosa, which with
115 amino acids is somewhat longer than that of
the mammals, has the signal sequence (26 amino
acids), B chain (31 amino acids), C peptide (37
amino acids) and A chain (21 amino acids)
arranged consecutively. However, the mature
insulin differs at 16 positions from that of the carp
and at 21 positions from human insulin. It cannot
bind zinc and can only form dimers, whereas
insulin from the shark Scyliorrhinus canicu/a can
already form zinc-stabilized hexamers. The insulin of the lamprey Petromyzon marin us differs
8.4.2 The Insulin Family
299
from that of the common hagfish Myxine (17
amino acid differences) more than from porcine
insulin or from that of the salmon (both with
14 differences) [168]; this underlines the large
evolutionary distance between the two agnathan
groups Myxinoidea and Petromyzontoidea. There
are 17 differences between the insulin of the carp
Cyprinus carpio and human insulin, including two
extra N-terminal amino acids on the B chain;
however, the zinc-binding B10-His and the amino
acids responsible for aggregation and biological
activity are all present. Carp insulin in the fat-cell
test has only threefold weaker binding and has
the same maximum activity as porcine insulin.
Amino acid substitutions in the region of the BI
C cleavage site lead to 37- to 38-amino-acid
B chains in the holocephalans Chimaera monstruosa and Hydro/agus colliei, whereas that of the
related species Callorhynchus milii has only the
normal 31 amino acids [12].
The gene sequences of human prepro-insulin
and those of a range of mammals, the chicken,
various teleosts and Myxine are now known. All
consist of three exons: the first exon codes for a
part of the non-translated (NT) 5' sequence; the
second for the remaining 5'-NT, the signal
sequence, B chain and a part of the C peptide;
and the third for the rest of the C peptide, the
A chain and the 3'-NT [209]. The sequence coding for the signal peptide occurs once again in a
slightly modified form in the C-coding region. It
is possible that two genes with identical signal
peptide regions arose initially by gene duplication, and were then later joined by a large intron
to give one gene. There are two prepro-insulin
genes in the mouse and rat which lie only 100 kb
apart on chromosome 1 of the rat but are found
separately on chromosomes 6 and 7 of the mouse.
In both rodent species, gene I is very similar to
gene II in the coding sequence but lacks the second intron. As in the chicken insulin gene, the
presence of two introns is probably the ancestral
state; the insulin genes rat-I and mouse-I apparently arose by the reinsertion of partly processed
transcripts. It is interesting that the retrotransposons of the two species differ much more than the
original insulin genes rat-II and mouse-II [232].
Insulin-like substances have been detected
by both immunological and pharmacological
methods in a wide spectrum of very different
organisms: tunicates, insects, crustaceans, gastropods and cephalopods, and even in the ciliate Tetrahymena, the lower fungi Aspergillus and Neurospora, and the prokaryote E. coli [68, 118].
Insulin-receptor proteins have been identified