Genetic Influence upon Phenobarbital-induced Increase in Rat Liver Supernatant Aldehyde Dehydrogenase Activity*

Abstract Treatment with phenobarbital of rats which are genetically selected results in a greater than 10-fold increase in nicotinamide-adenine dinucleotide (NAD)-dependent aldehyde dehydrogenase present in the supernatant but not the mitochondrial fraction of liver. This effect represents one of only a few known cases where intrastrain differences in a drug-induced increase in enzyme activity is genetically controlled. The kinetic characteristics of the enzyme from phenobarbital-treated rats are the same, regardless of whether or not the enzyme activity in the liver is increased. It is also demonstrated that rat liver supernatant contains a second NAD-dependent aldehyde dehydrogenase which (a) is resistant to induction by phenobarbital; (b) oxidizes p-carboxybenzaldehyde; (c) is relatively uninhibited by disulfiram; (d) has a heat stability different from the inducible enzyme.


Treatment
with phenobarbital of rats which are genetically selected results in a greater than IO-fold increase in nicotinamide-adenine dinucleotide (NAD)-dependent aldehyde dehydrogenase present in the supernatant but not the mitochondrial fraction of liver. This effect represents one of only a few known cases where intrastrain differences in a drug-induced increase in enzyme activity is genetically controlled.
The kinetic characteristics of the enzyme from phenobarbital-treated rats are the same, regardless of whether or not the enzyme activity in the liver is increased.
It is also demonstrated that rat liver supernatant contains a second NAD-dependent aldehyde dehydrogenase which (a) is resistant to induction by phenobarbital; (b) oxidizes fi-carboxybenzaldehyde; (c) is relatively uninhibited by disulfirarn; (d) has a heat stability different from the inducible enzyme.
It has been found recently that the administration of phenobarbital to rats results in a IO-fold increase in NAD-linked aldehyde dehydrogenase in the supernatant fraction of the liver. This response is dependent upon the genotype of the animal and is due to the presence of a single autosomal co-dominant gene (2). Numerous examples of genetically controlled differences in enzyme levels are known (3). Likewise, the activity of some enzymes (e.g. the microsomal drug-metabolizing system) is increased by administration of various compounds (4). There are few known examples, however, of drug-induced increases in enzyme activity which are dependent upon the genotype of the individual.
Gielen el al. (5) have recently described a microsomal hydroxylase enzyme system that is inducible in genetically selected strains * This work was supported by Public Health Service Grants MH 15908, 18971, and 18948. A preliminary report of some of this work has appeared (1).
$ Recipient of Career L)evelopment Award CM 10475. 0 Recipient of Public Health Service Postdoctoral Fellowship GM 46810. of mice. The effect is inherited as a dominant trait and is demonstrable in a number of tissues besides the liver. The activit'y of &aminolevulinic acid synthetase is increased by barbiturates and this effect is dependent upon the genotype of the animal (6).
The present report further describes the characteristics of the increase in aldehyde dehydrogenase activity brought about by phenobarbital administration and presents evidence that similar enzymes in kidney and brain are unaffected by such treatment. Direct evidence is also presented for the presence of a second liver supernatant aldehyde dehydrogenase, the activity of which is also not altered by phenobarbital administration. Rat liver also contains an aldehyde dehydrogenase in the mitochondria; however, unlike the enzyme found in the supernatant, its activity was found to be unaltered by phenobarbital administration in any group of rats tested (2).

MATERIALS
AND METHODS il4aterials-All chemicals were of the highest quality available. Volatile aldehydes used as substrates were periodically distilled under nitrogen and stored until use at -20".
Solutions of aldehydes were assayed enzymatically as previously described (7). Phenobarbital for injection (20 mg per ml) was dissolved in NaOIl and adjusted to pH 9 with HCI. Other compounds were dissolved and administered as detailed in the tables.
Animals-Adult rats and mice of both sexes were obtained from commercial sources as noted in the tables. ,4 closed breeding colony of Long-Evans rats is maintained in this laboratory. Animals are designated RR for homozygous reactor animals, rr for homozygous nonreactor animals, and Rr for the heterozygous reactor animals as defined earlier (2). Unless otherwise noted, the quantity of phenobarbital injected intraperitoneally was 100 mg per kg once daily for 3 successive days.
Animals receiving injections of an equal volume of saline had aldehyde dehydrogenase activities no different from animals not receiving injections.
All controls reported in this report refer to animals not receiving injections.
The most convincing "control" is the animal receiving phenobarbital injections, which, because of its genetic constitution, does not demonstrate an increase in aldehyde dehydrogenase activity but does manifest all other observed responses to phenobarbital.
Tissue Preparation and Enzyme Assays-Whole organs were removed immediately after decapitation of the animal.
In order to perform liver biopsies the animals were anesthesized with ether and tissue samples were obtained through a small abdominal incision. Subcellular fractionation of various tissues was carried out as previously described (2). I3rains were homogenized as a 20% suspension in 0.25 M sucrose and centrifuged at 100,000 x g for 1 hour. The supernatant was decanted and adjusted to 80% saturation with (NHJzSOJ at 4". Mercaptoethanol was added to this solution to a final concentration of 0.01 M and it was centrifuged at 15,000 x g for 5 min. The resulting precipitate was dissolved in 0.25 M sucrose and centrifuged.
The supernatant from this procedure was used for assay of aldehyde dehydrogenase activity.
Aldehyde dehgdrogenase activity in all assays was determined by following the net rate of NA4DH formation spectrophotometrically.
Unless oGlerwise noted 1 inn< NAD and 0.33 m&l propionaldehyde were employed in each assay sample. Control samples contained no propionaldehyde.
I'yrazole (33 FM) was added in the assay of the liver and kidney supernatant enzyme in order to inhibit alcohol dehydrogenase, which interferes with the assay. hlcohol dehydrogenase activity was determined by the net rate of decrease in NSDH concentration in the presence of NADH and propionaldehyde, 0.33 mM each (9). Lactate dehydrogenase (lo), aniline hydroxylsse (1 I), cy-glycerophosphate dehydrogenase (la), and aminopyrine demet'hylase (13) were assayed by described methods. RESULTS We have shown that the ability of rats to respond to injected phenobarbital by an increase in aldehydc dehydrogenase in the liver supernatant is inherited as an autosomal co-dominant characteristic in a manner described by classic Mendelian genetics (2). The initial studies were carried out using animals with a phenotype determined by the assay of aldehydc dehydrogenase activity in a liver biopsy following phenobarbital treatment. Following the biopsy, the animals were allowed to heal and further experiments were carried out at least 4 weeks later.
More recent studies have employed animals of the F1 through the FS generation that, were found to breed true for an increase or no change in supernatant aldehyde dehydrogenase activity following phenobarbital administration.
The use of substrates other than propionaldehyde demonstrated that a second aldehyde dehydrogenase enzyme is present in the liver supernatant of rats treated with phenobarbital ( Table  I). The rate of NAD-dependent oxidation of p-carboxybenzaldchyde, as well as nn-glyccraldehyde, glycolaldehyde, and glyoxalic acid in the liver supernatant is not affected by phenobarbital treatment of the animal.
The presence of this second enzyme in the untreated animal is also evident from heat denaturation curves of the enzyme activities employing propionaldehyde and p-carboxybenzaldehyde as substrates (Fig. 1). Some progress has been made toward purification of the enzyme with activity increased by phenobarbital treatment as illustrated in Table II. It is evident from these data that activities toward propionaldehyde and p-carboxybenzaldehyde behave differently upon fractionation with ammonium sulfate. Also mixed substrate experiments with saturating amounts of propionaldehyde and p-carboxybenzaldehyde show additive or or nearly additive rather than competitive rates, again illustrating the presence of two enzyme activities.
Disul6ram at a dose of 600 mg per kg administered to phenobarbital-treated animals, 16 hours prior to removal of the liver, resulted in 95.8% inhibition of oxidative activity toward propionaldehyde, but only 19% inhibition of the activity toward p-carboqbenzaldehyde.

Kinetic
Characteristics-The apparent K, values for the enzymes, using propionaldehyde and NAD as substrates, are relatively constant regardless of the treatment of the animal prior to examination of the enzyme (Table III).
Likewise the K, value for p-carboxybenzaldehyde is unchanged by phenobarbital treatment.
Studies of the pH optimum for the enzyme from rats by guest on March 23, 2020 http://www.jbc.org/ Downloaded from receiving injections or control rats also revealed no differences. The use of acetaldehyde as a substrate is complicated by an apparent substrate activation at concentrations above 0.33 mu. Double reciprocal plots with propionaldehyde also show apparent, substrate activation at concentrations above 3.3 mM. Cnless otherwise noted, studies reported were carried out with a propionaldehyde concentration of 0.33 mM although the activity at 3.3 m&f was also routinely determined in order to detect possible effects evident at high concentrations of propionaldehyde. on other strains of rats and several strains of mice treated with phenobarbital in an attempt to increase the liver supernatant aldehyde dehydrogenase activity.
The "randomly bred" strains of rats (Sprague-Dawley, Long-Evans, Wistar and Charles-River) include animals of all three groups, RR, rr, and Rr. Inbred strains of rats (Fischer, Buffalo-Lewis, ACIF/Mai and BN/ F hlai) are homozygous as expected.
An insufficient number of mice in each strain has been tested to determine whether or not there are nonreacting animals present. However, the maximal stimulation achieved in these experiments was only 2-fold in any of the strains tested; a finding similar to that reported by Redmond and Cohen (14).
Since phenobarbital is metabolized in z&o (15) it seemed possible that the pathway of metabolism of phenobarbital might be different in the substrains of rats. Therefore, an inhibitor of the microsomal drug-metabolizing system 2-diethylaminoethyl-2,2-diphenylvalerate-HCl (SKF 525,4) was given to reactor     The effect is apparently uniquo to the liver. It is distinguished tested, the effect is confined to a genetically selected population for aniline hydroxylase is nanomoles of p-aminophenol formed X mg of protein-' X hour-'; for alcohol and lactic dehydrogenase is nanomoles of NAD formed X mg of protein-i X mini; and for cyglycerophosphate dehydrogenase is nanomoles of NADH formed animals that were either given phenobarbital or else were un treated.
Administration of SKF 5258 had no effect on the increase in aldehyde dehydrogenase activity produced by phenobarbital administration (Table V). Table VI shows that aniline hydroxylase, a microsomal enzyme known to be induced by phenobarbital (II), is increased in both reacting and nonreacting animals.
More recent studies in this laboratory of aminopyrine demethylase in the substrains (RR or rr) of Long-Evans rats also show that all animals respond to phenobarbital injection with a significant (p < 0.001) increase in the microsomal enzyme.
Several other supernatant, NAD-dependent enzymes were also examined for any change in activity following phenobarbital administration; no marked differences were found (Table VI).
The increased aldehyde dehydrogenase activity after phenobarbital treatment apparently is confined to the liver; neither the kidney nor brain supernatant enzymes are affected (Table  VII).
Although adrenalectomy is known to alter other enzyme systems (16), this procedure had no effect on the increase in aldehyde dehydrogenase elicited by phenobarbital. Administration of hydrocortisone 30 mg X kg+ X day-l for 3 days to normal reac-X mg of protein-i min-i.
Assay methods are given in t,he text. b p value for comparison of phenobarbital-treated group with control group.
In no case except that for aldehyde dehydrogenase is the difference between phenobarbital groups significant.  (6); (c) compounds such as 3-methylcholanthrene which induce microsomal enzymes or allylisopropylacetamide, which induces mitochondrial d-aminolevulinic acid synthetase, are relatively ineffective in increasing aldehyde dehydrogenase activity (17).
Since the kinetic characteristics of the enzymes from treated reactor (RR) and nonreactor (rr) animaIs are similar, we tentatively conclude that the enzymes are the same. Further purification and characterization of the enzymes from the livers of reactor and nonreactor animals is underway.
It is of interest that the apparent K, values for aldehydes obtained with this aldehyde dehydrogenase from rat liver are 2 to 3 orders of magnitude higher than are those of the corresponding enzymes from beef (7), human (19,20), or horse liver (21) or from beef brain (22). Although this property makes kinetic studies somewhat easier, the physiological significance of the rat liver enzyme which exhibits these higher K, values is as yet unknown.
There seems to be little doubt that there are two soluble NADdependent aldehyde dehydrogenase enzymes present in rat liver. The first line of evidence supporting this conclusion is the marked increase in activity of aldehyde dehydrogenase toward some substrates but not others when phenobarbital is administered to susceptible animals.
The second line of evidence is the additive character of the rates of NAD reduction observed when propionaldehyde and p-carboxybenzaldehyde are present in quantities which saturate each enzyme. The calculated and observed rate of NADH production in the presence of both substrates is nearly the same if one assumes the presence of two enzymes. This observation also affords presumptive evidence that p-carboxybenzaldehyde is not an effective inhibitor of the propionaldehyde enzyme nor is propionaldehyde an effective inhibitor of the p-carboxybenzaldehyde enzyme.
The differential sensitivity of these enzymes to disulfimm also illustrates that two enzyme activities are present.
Finally, the markedly different heat denaturation curves for aldehyde dehydrogenase when propionaldehyde and p-carboxybenzaldehyde are used as substrates provides strong evidence for the presence of two enzymes.
The enzyme which osidizes p-carboxybenzaldehyde may not be the one responsible for the oxidation of glyceraldehyde and glycoaldehyde.
Glyoxalie acid is a substrate for lactic dehydrogenase and this enzyme may account for the activity observed (23). The amount of each of these substrates oxidized by the inducible enzyme or other enzymes will have to await further purification of these enzymes. Recently, Shum and Blair (24) have presented evidence of two aldehyde dehydrogenases in rat liver supernatant.
The normal physiological role of the enzyme aldehyde dehydrogenase is presumably the oxidation of endogenous aldehydes such as those arising from deamination of various amines, e.g. norepinephrine, dopamine, serotonin, tryptamine, and histamine. This enzyme also functions in the oxidation of large quantities of exogenous aldehydes, such as acetaldehyde.
There is no evidence at present to indicate that this hepatic enzyme becomes rate-limiting in any physiologically normal situation, so that an increase in activity of the enzyme may not be reflected in a physiologically demonstrable way. Experiments are underway to determine whether or not the rate of acetaldehyde oxidation is altered in viva by phenobarbital treatment of reactor or nonreactor rats.
In this regard, Horton (25) has reported a slight stimulation of rat liver mitochondrial aldehyde dehydrogenase by chronic ethanol feeding.
It is unlikely that the small increase in activity can be considered important unless it can be shown that the enzyme is rate-limiting in the normal condition.
To date this has not been demonstrated.
It is true, of course, that if one inhibits the enzyme, e.g. with disulfiram, this enzyme does become ratelimiting for oxidation of aldehydes (26).
The findings reported in this paper demonstrate some of the properties of an NAD-dependent aldehyde dehydrogenase in the supernatant fraction of livers from genetically selected rats. This enzyme can be markedly increased by phenobarbital administration.
The availa.bility of these genetically selected strains of rats affords an unique opportunity to investigate the mechanism of this phenomena.
Whether or not this represents a net increase in enzyme protein must await further experiments. Attempts to inhibit the reaction with protein synthesis inhibitors has not been satisfactory because of the mortality rate after several days treatment with these compounds and also because of their interaction with phenobarbital.
The demonstration of a second enzyme in the supernatant fraction of rat liver which is resistant to induction by phenobarbital and which catalyzes the oxidation of p-carboxybenzaldehyde, and perhaps other aldehydes as well is of considerable interest since this is an indication that not all aldehydes are oxidized by a single nonspecific NAD-dependent enzyme.