and III differ by 35-40 % in their amino acid
sequences.
The structural and functional variety in the
protein family of the alcohol dehydrogenases may
be explained by three successive gene duplications. The first duplication separated the sorbitol
and alcohol dehydrogenases; the ADH classes
arose in the second phase, and the third phase
saw the appearance of the isoenzymes of class I
[135] .
Comparison of the human and baboon ~
sequences suggests that the a/~y gene duplication
occurred about 60 million years ago, at approximately the same time as the appearance of the first
primates [270]. The rabbit ADH I also has a complicated pattern of homo- and heterodimers of
three different subunits [174]. The tissues of the
rat contain three ADH isoenzymes, ADH-l, -2
and -3, which are comparable with the human
ADH II, III and I in their substrate specificities
and pyrazole sensitivities. ADH-l predominates
in the eye, stomach and lungs; ADH-2 is present
in all organs, and ADH-3 is found mainly in the
liver [139]. Despite its functional similarity with
ADH II, the ADH of rat stomach is so different in
its sequence to all other rat ADHs (32-40 %) that
it may be considered to represent a new class, IV,
of mammalian ADHs. In this case, the number of
gene duplications that occurred is larger than was
previously assumed [204]. In the mouse, there are
three ADH isoenzymes, A2, B2 and C2, encoded
in the genes adh-l to -3 on chromosome 3; allelic
variants of adh-l and -3 are found. The individual
isoenzymes are structurally and functionally very
different. A has a molecular mass of 43 kDa, B
39 kDa, and C 47 kDa. A and C contain two catalytic zinc atoms, but B has only one non-catalytic
zinc. A2, which is present in particular in the liver,
kidney and gut, is the only one that can convert
ethanol under physiological conditions; the C2
from the stomach wall and lungs and the similarly
distributed B2 have a very low affinity for ethanol.
The complete gene sequence of the A subunit is
known; the amino acid sequences of mouse A and
horse E differ at 58 (15.5 %) out of 374 positions
[74]. There are three dimeric ADH isoenzymes in
the horse, and these are made up of the subunits
E (ethanol active) and S (steroid and ethanol active). The complete sequence is known for both
isoenzymes. It may be concluded from comparisons of the horse E/S and human ~hl sequence
that the two duplications occurred independently
[205]. The liver ADHs of various avian species
and the frog Rana perezi are more similar to mammalian ADH I than to ADH II or III [46, 140].
13.3.4 Ethanol Metabolism
477
Almost the only invertebrate ADH investigated in any detail is that of Drosophila; among
the many hundreds of publications, the majority
deal with the particularly marked enzyme polymorphism, and leave open several important biochemical questions. Although there are always
many ADH bands visible in electropherograms,
there is only one ADH locus in Drosophila melanogaster, as in most Drosophila species. However, at least eight allelic variants of ADH have
been described, of which the alleloenzymes F
(fast) and S (slow) are present in all natural
popUlations. This polymorphism is superimposed
upon variability arising from post-translational
modifications, which give three to five electrophoretic sub-bands for each of the alleloenzymes
(p.145).
The Drosophila ADH is very similar to the
yeast and mammalian enzymes in functional
properties but is structurally very different. The
subunits have only 254 amino acids (27.4 kDa)
compared with 347 (35.3 kDa) in yeast and 374
(39.8 kDa) in horse liver. The fly enzyme has no
zinc and the amino acid sequence shows no
homology to the sequences of the yeast and horse
enzymes. Whereas the yeast and mammalian
enzymes are clearly related to each other and to
mammalian sorbitol dehydrogenase, the Drosophila enzyme belongs to a completely different
protein super-family. The individual regions ofthe
ADH gene show large differences in their rate of
evolution due to varying restrictions to evolutionary change. This is already clear from comparisons of the allelic variants ADH-F and -S in Drosophila melanogaster (see Fig. 4.6b, p. 138), but is
seen even better in comparisons of the ADH
sequences of D. melanogaster with those of
the closely related species D. mauritana and
D. orena. Amino acid substitutions at only 72 out
of 254 positions have so far been described in
Drosophila ADH [280]. The ADH genes of
almost all Drosophila species have two promoters
which are active at different times in development. The gene transcripts encode the same
amino acid sequence but differ in their 5' nontranslated (NT) regions. In D. melanogaster and
D. ajfinis, the proximal promotor is used in the
larvae and the distal promoter is used in the adult
animal; in D. formella, the larval situation is also
found in the imago [231, 232]. A 34-kDa protein
that has been isolated activates the Adh distal
promoter of D. melanogaster; it is known as the
Adh distal factor 1 (adf-l), and also binds to the
promoters of other genes [78]. The ADH gene of
D.lebanonensis contains three TATA boxes [138].
Précédent

- 492/799

Suivant