Journal o f Protein Chemistry, VoL I0, No. 1, 1991
Human Class III Alcohol Dehydrogenase/GlutathioneDependent Formaldehyde Dehydrogenase Rudolf Kaiser, 1 Barton Holmquist, 2 Bert L. Vallee, 2 and Hans J6rnvall m'4
Received October 15, 1990
The class III human liver alcohol dehydrogenase, identical to glutathione-dependent formaldehyde dehydrogenase, separates electrophoretically into a major anodic form (Xl) of known structure, and at least one minor, also anodic but a slightly faster migrating form (X2)- The primary structure of the minor form isolated by ion-exchange chromatography has now been determined. Results reveal an amino acid sequence identical to that of the major form, suggesting that the two derive from the same translation product, with the minor form modified chemically in a manner not detectable by sequence analysis. This pattern resembles that for the classical alcohol dehydrogenase (class I). Hence, the XI/X2 multiplicity does not add further primary forms to the complex alcohol dehydrogenase system but shows the presence of modified forms also in class III. KEY WORDS: Amino acid sequence; post translational modification; peptide analysis; isozymes; highperformance liquid chromatography.
1. I N T R O D U C T I O N
enzyme activities, showing the class III alcohol dehydrogenase to be identical to glutathione-dependent formaldehyde dehydrogenase (Koivusalo et aL, 1989). Tissue distributions are also distinct, as noted particularly for hydroxyleukotriene B4 dehydrogenase (Gotoh et aL, 1989) and other similar activities. In addition to this class multiplicity, the enzyme occurs in multiple forms within the classes. Thus, different class I forms derive both from the existence of different genetic loci (Smith et al., 1971) and from posttranslational modifications (J/Srnvall, 1973). The major human forms have been analyzed, c D N A cloned, and interpreted (JSrnvall et aL, 1989; Eklund et al., 1987; 1990; Sharma et aL, 1989; Girl et al., 1989). However, there are also electrophoretically detectable subforms of class III which can be resolved into two (Adinolfi et al., 1984; Par~s and Vallee, 1981) or three (Valkonen and Goldman, 1988) components. Their relationships are unknown but could be important in view of the established multiplicity of alcohol dehydrogenase (J6rnvall et al., 1987; Vallee and Bazzone, 1983), the separate activities of the classes (Val-
Alcohol dehydrogenase constitutes a complex enzyme system (J6rnvall et al., 1987). Three classes of liver enzyme have been defined (Vallee and Bazzone, 1983) differing by about 40% in primary structure (Kaiser et al., 1988) and known to originate from two separate gene duplications, the most recent one of which apparently occurred early in vertebrate evolution (Cederlund et al., 1990). Structural analysis of the class III type suggested that it constitutes an enzyme distinct from the classical liver alcohol dehydrogenase (Kaiser et al., 1989). This conclusion was confirmed by direct demonstration of different
Department of Chemistry I, Karolinska Institutet, S-104 01 Stockholm, Sweden. 2 Center for Biochemical and Biophysical Sciences and Medicine, Harvard Medical School, 250 Longwood Avenue, Boston, Massachusetts 02115. 3 Center for Biotechnology, Karolinska Institutet, S-141 86 Huddinge, Sweden. 4 To whom correspondence should be addressed.
69 0277-8033/91/0200-0069506.50/0 ((h 1991 PlenumPublishin~Corr,oration
70
Kaiser et al.
lee and Bazzone, 1983; Koivusalo et al., 1989), and the presence of class III pseudogenes (Matsuo and Yokoyama, 1990). In this study, we have determined the primary structure of the minor form of the class III human liver alcohol dehydrogenase and show it to be identical to that of the major form. This suggests that a chemical modification constitutes the molecular explanation for the occurrence of the minor form.
Hence, the structure of the X2 enzyme is identical to that already known for the major X~ form from peptide analysis (Kaiser et al., 1988) and one cDNA report (Sharma et aL, 1989). In particular, a possible alternative residue at position 166, suggested from another cDNA study (Giri et al., 1989), was not present in X2 and the microheterogeneities between the two cDNA reports do not correspond to the X~/X2 forms.
2. MATERIALS AND METHODS
4. DISCUSSION
The minor form, X2, of human class III alcohol dehydrogenase (Par6s and Vallee, 1981)/glutathionedependent formaldehyde dehydrogenase (Koivusalo et al., 1989), which exhibits higher anodal electrophoretic mobility than the major one (Adinolfi et al., 1984), was purified by DEAE-chromatography and AMP-chromatography in a manner similar to that for the major X1 form (Wagner et al., 1984). Final purification was by DEAE-HPLC (Holmquist, unpublished). The pure isozyme was ~4C-carboxymethylated with labeled iodoacetate as described (Kaiser et al., 1988), cleaved in separate batches with Lys-specific protease and CNBr, and the resulting peptides were purified by reverse-phase HPLC (Kaiser et al., 1988). Pure peptides were analyzed for total composition with a Beckman 121M analyzer after acid hydrolysis (24hr, ll0°C, 6M HC1/0.5% phenol), and for amino acid sequence by degradations on Applied Biosystems sequencers 470A and 477A with on-line 120A analyzers.
The present analysis proves that the minor X2 form of class III alcohol dehydrogenase/glutathionedependent formaldehyde dehydrogenase has the same amino acid sequence as the major form, although it has a greater electrophoretic anodic mobility. The minor class III enzyme type is c o n cluded to be derived from the same primary translation product as the major X1 form. Thus, there is no need to ascribe more than one gene to the class III locus, in agreement with results from genetic mappings (Smith, 1986). Peptides covering the entire structure were obtained and constitute overlapping segments proving that the two structures indeed have the same 373-residue amino acid sequence, when analyzed at the peptide level. The X2 form is therefore likely to be derived from X~ by a modification, not stable to reduction, carboxymethylation, or acid hydrolysis, and therefore not detected by sequence analysis. Attempts to differentiate X2 from X~ based on their enzymatic properties have not proven successful. They have identical specific activities toward ethanol, octanol, and 12-hydroxydodecanoate and show no activity toward methanol. 4-Methylpyrazole (10 -4 M) does not inhibit either enzyme, but 4-pentylpyrazole (10 -4 M) inhibits 75% of the activity of both (pH 10, 0.1M glycine, 0.5M ethanol, 2.4mM NAD+). One possibility is that the minor X2 form constitutes an adduct, perhaps with glutathione, considering the substrate specificity of the class III enzyme (Koivusalo et al., 1989). Another possibility is that the minor form is a deamidated derivative, an allelic variant, or an adduct with the coenzyme. All these possibilities could increase anodal mobility. Finally, partial oxidation by disulfide bridge formation might contribute to the difference. Notably, the "classical" liver alcohol dehydrogenase of class I is rich in subforms (Lutstorf et al., 1970) which, like X2, are more anodal than the major form. Similar subforms are also known for the short-chain alcohol
3. RESULTS The minor form X2 of human liver class III alcohol dehydrogenase was purified by a three-step chromatographic procedure (Wagner et al., 1984; Holmquist, unpublished), using DEAE-HPLC for separation of X2 from X~ (Fig. 1). X2 fractions from several preparations of different livers were pooled, ~4C-carboxymethylated, and submitted to proteolytic cleavages, peptides were purified by HPLC (Fig. 2) and analyzed for amino acid sequence to give the enzyme primary structure (Fig. 3). Total compositions of peptides are in agreement with the amino acid sequences determined. The same is true of the composition of the intact protein, which in addition is indistinguishable from that of the major X1 type (Kaiser et al., 1988). Peptides from all regions of the molecule were analyzed and showed the primary structure of the X2 form as in Fig. 3.
Class III Human Liver Alcohol Dehydrogenase
71
0.2
T Fig. 1. HPLC of class III alcohol dehydrogenase. Active fractions from an AMP-Sepharose step (Wagner et aL, 1984) were pooled, exchanged into 10 m M Tris-Cl, p H 8.0, 10 -4 M DT]? by repeated pressure filtration, and separated on a Waters D E A E - 5 P W column at 1 m l / m i n in the same buffer with a gradient of NaCI as shown. Active fractions (hatched) correspond to X1 at 20.5 and X2 at 34.2 min.
o
Human Class III Alcohol Dehydrogenase/GlutathioneDependent Formaldehyde Dehydrogenase Rudolf Kaiser, 1 Barton Holmquist, 2 Bert L. Vallee, 2 and Hans J6rnvall m'4
Received October 15, 1990
The class III human liver alcohol dehydrogenase, identical to glutathione-dependent formaldehyde dehydrogenase, separates electrophoretically into a major anodic form (Xl) of known structure, and at least one minor, also anodic but a slightly faster migrating form (X2)- The primary structure of the minor form isolated by ion-exchange chromatography has now been determined. Results reveal an amino acid sequence identical to that of the major form, suggesting that the two derive from the same translation product, with the minor form modified chemically in a manner not detectable by sequence analysis. This pattern resembles that for the classical alcohol dehydrogenase (class I). Hence, the XI/X2 multiplicity does not add further primary forms to the complex alcohol dehydrogenase system but shows the presence of modified forms also in class III. KEY WORDS: Amino acid sequence; post translational modification; peptide analysis; isozymes; highperformance liquid chromatography.
1. I N T R O D U C T I O N
enzyme activities, showing the class III alcohol dehydrogenase to be identical to glutathione-dependent formaldehyde dehydrogenase (Koivusalo et aL, 1989). Tissue distributions are also distinct, as noted particularly for hydroxyleukotriene B4 dehydrogenase (Gotoh et aL, 1989) and other similar activities. In addition to this class multiplicity, the enzyme occurs in multiple forms within the classes. Thus, different class I forms derive both from the existence of different genetic loci (Smith et al., 1971) and from posttranslational modifications (J/Srnvall, 1973). The major human forms have been analyzed, c D N A cloned, and interpreted (JSrnvall et aL, 1989; Eklund et al., 1987; 1990; Sharma et aL, 1989; Girl et al., 1989). However, there are also electrophoretically detectable subforms of class III which can be resolved into two (Adinolfi et al., 1984; Par~s and Vallee, 1981) or three (Valkonen and Goldman, 1988) components. Their relationships are unknown but could be important in view of the established multiplicity of alcohol dehydrogenase (J6rnvall et al., 1987; Vallee and Bazzone, 1983), the separate activities of the classes (Val-
Alcohol dehydrogenase constitutes a complex enzyme system (J6rnvall et al., 1987). Three classes of liver enzyme have been defined (Vallee and Bazzone, 1983) differing by about 40% in primary structure (Kaiser et al., 1988) and known to originate from two separate gene duplications, the most recent one of which apparently occurred early in vertebrate evolution (Cederlund et al., 1990). Structural analysis of the class III type suggested that it constitutes an enzyme distinct from the classical liver alcohol dehydrogenase (Kaiser et al., 1989). This conclusion was confirmed by direct demonstration of different
Department of Chemistry I, Karolinska Institutet, S-104 01 Stockholm, Sweden. 2 Center for Biochemical and Biophysical Sciences and Medicine, Harvard Medical School, 250 Longwood Avenue, Boston, Massachusetts 02115. 3 Center for Biotechnology, Karolinska Institutet, S-141 86 Huddinge, Sweden. 4 To whom correspondence should be addressed.
69 0277-8033/91/0200-0069506.50/0 ((h 1991 PlenumPublishin~Corr,oration
70
Kaiser et al.
lee and Bazzone, 1983; Koivusalo et al., 1989), and the presence of class III pseudogenes (Matsuo and Yokoyama, 1990). In this study, we have determined the primary structure of the minor form of the class III human liver alcohol dehydrogenase and show it to be identical to that of the major form. This suggests that a chemical modification constitutes the molecular explanation for the occurrence of the minor form.
Hence, the structure of the X2 enzyme is identical to that already known for the major X~ form from peptide analysis (Kaiser et al., 1988) and one cDNA report (Sharma et aL, 1989). In particular, a possible alternative residue at position 166, suggested from another cDNA study (Giri et al., 1989), was not present in X2 and the microheterogeneities between the two cDNA reports do not correspond to the X~/X2 forms.
2. MATERIALS AND METHODS
4. DISCUSSION
The minor form, X2, of human class III alcohol dehydrogenase (Par6s and Vallee, 1981)/glutathionedependent formaldehyde dehydrogenase (Koivusalo et al., 1989), which exhibits higher anodal electrophoretic mobility than the major one (Adinolfi et al., 1984), was purified by DEAE-chromatography and AMP-chromatography in a manner similar to that for the major X1 form (Wagner et al., 1984). Final purification was by DEAE-HPLC (Holmquist, unpublished). The pure isozyme was ~4C-carboxymethylated with labeled iodoacetate as described (Kaiser et al., 1988), cleaved in separate batches with Lys-specific protease and CNBr, and the resulting peptides were purified by reverse-phase HPLC (Kaiser et al., 1988). Pure peptides were analyzed for total composition with a Beckman 121M analyzer after acid hydrolysis (24hr, ll0°C, 6M HC1/0.5% phenol), and for amino acid sequence by degradations on Applied Biosystems sequencers 470A and 477A with on-line 120A analyzers.
The present analysis proves that the minor X2 form of class III alcohol dehydrogenase/glutathionedependent formaldehyde dehydrogenase has the same amino acid sequence as the major form, although it has a greater electrophoretic anodic mobility. The minor class III enzyme type is c o n cluded to be derived from the same primary translation product as the major X1 form. Thus, there is no need to ascribe more than one gene to the class III locus, in agreement with results from genetic mappings (Smith, 1986). Peptides covering the entire structure were obtained and constitute overlapping segments proving that the two structures indeed have the same 373-residue amino acid sequence, when analyzed at the peptide level. The X2 form is therefore likely to be derived from X~ by a modification, not stable to reduction, carboxymethylation, or acid hydrolysis, and therefore not detected by sequence analysis. Attempts to differentiate X2 from X~ based on their enzymatic properties have not proven successful. They have identical specific activities toward ethanol, octanol, and 12-hydroxydodecanoate and show no activity toward methanol. 4-Methylpyrazole (10 -4 M) does not inhibit either enzyme, but 4-pentylpyrazole (10 -4 M) inhibits 75% of the activity of both (pH 10, 0.1M glycine, 0.5M ethanol, 2.4mM NAD+). One possibility is that the minor X2 form constitutes an adduct, perhaps with glutathione, considering the substrate specificity of the class III enzyme (Koivusalo et al., 1989). Another possibility is that the minor form is a deamidated derivative, an allelic variant, or an adduct with the coenzyme. All these possibilities could increase anodal mobility. Finally, partial oxidation by disulfide bridge formation might contribute to the difference. Notably, the "classical" liver alcohol dehydrogenase of class I is rich in subforms (Lutstorf et al., 1970) which, like X2, are more anodal than the major form. Similar subforms are also known for the short-chain alcohol
3. RESULTS The minor form X2 of human liver class III alcohol dehydrogenase was purified by a three-step chromatographic procedure (Wagner et al., 1984; Holmquist, unpublished), using DEAE-HPLC for separation of X2 from X~ (Fig. 1). X2 fractions from several preparations of different livers were pooled, ~4C-carboxymethylated, and submitted to proteolytic cleavages, peptides were purified by HPLC (Fig. 2) and analyzed for amino acid sequence to give the enzyme primary structure (Fig. 3). Total compositions of peptides are in agreement with the amino acid sequences determined. The same is true of the composition of the intact protein, which in addition is indistinguishable from that of the major X1 type (Kaiser et al., 1988). Peptides from all regions of the molecule were analyzed and showed the primary structure of the X2 form as in Fig. 3.
Class III Human Liver Alcohol Dehydrogenase
71
0.2
T Fig. 1. HPLC of class III alcohol dehydrogenase. Active fractions from an AMP-Sepharose step (Wagner et aL, 1984) were pooled, exchanged into 10 m M Tris-Cl, p H 8.0, 10 -4 M DT]? by repeated pressure filtration, and separated on a Waters D E A E - 5 P W column at 1 m l / m i n in the same buffer with a gradient of NaCI as shown. Active fractions (hatched) correspond to X1 at 20.5 and X2 at 34.2 min.
o
