ALCOHOL-DEHYDROGENASE CLASS-III CONTRASTED TO CLASS-I - CHARACTERIZATION OF THE CYCLOSTOME ENZYME, THE EXISTENCE OF MULTIPLE FORMS AS FOR THE HUMAN ENZYME, AND DISTANT CROSS-SPECIES HYBRIDIZATION
O. Danielsson et al., ALCOHOL-DEHYDROGENASE CLASS-III CONTRASTED TO CLASS-I - CHARACTERIZATION OF THE CYCLOSTOME ENZYME, THE EXISTENCE OF MULTIPLE FORMS AS FOR THE HUMAN ENZYME, AND DISTANT CROSS-SPECIES HYBRIDIZATION, European journal of biochemistry, 225(3), 1994, pp. 1081-1088
Alcohol dehydrogenases of classes I (the classical liver enzyme) and I
II (formaldehyde dehydrogenase) constitute a pair of moderately relate
d enzymes (63% residue identity between the human forms) that differ f
undamentally in many respects. To elucidate the nature of the differen
ces, we have characterized alcohol dehydrogenase from the most primiti
ve vertebrate line (a cyclostome, Atlantic Hagfish), related that to t
he multiplicity of the human enzyme, and submitted the enzymes to in v
itro hybridization for evaluation of subunit interactions. Three findi
ngs illustrate important principles of the enzyme system. First, the a
lcohol dehydrogenase purified from cyclostomes is a class-III protein,
compatible with the facts that cyclostomes constitute the earliest ex
tant vertebrate line and that class III has a distant pre-vertebrate o
rigin. Second, the hagfish enzyme shows multiplicity, with acidic form
s in decreasing yield and with amino acid sequences identical between
two major isoforms, both aspects constituting properties similar to th
ose of the corresponding human forms. The chemically different subunit
s are present as homodimers and heterodimers of unmodified and modifie
d subunits, suggesting that the class-III multiplicity derives from mo
dification of a type common to lines as divergent as mammals and cyclo
stomes. Third, the human enzyme can form cross-species hybrid dimers i
n vitro with the cod and hagfish or Drosophila class-m enzymes (positi
onal identity with the human form of 82, 76 and 70%, respectively). He
nce, the results provide experimental evidence for little class-III di
vergence in the segments of subunit interactions. The extent of conser
vation of residues directly involved in the formation of the subunit i
nterface also reveals a clearly different pattern between classes I an
d III. This highlights separation of divergent forms in an enzyme syst
em, with the constant form (class III) resembling house-keeping enzyme
s, and exhibiting a correlation between subunit-interacting and substr
ate-interacting segments.