Vol. 14. No. 4 July/August 1990

0 l45-6008/90/ 1404-0534$2.00/0 ALCOHOLISM: Cl.INICAL A N D EXPERIMENTAL RESEARCH

Isoenzymes of Aldehyde Dehydrogenase in Human Lymphocytes Lillian E. Dyck

The types of isozymes of aldehyde dehydrogenase (ALDH) present in human lymphocytes has been investigated using isoelectric focusing of polyacrylamide gels followed by substrate-specific staining. Lymphocytes obtained from most individuals were found to contain both types I and II ALDH. This group of 'typical' individuals reported that they did not develop marked facial flushing or rapid heart rate after drinking alcohol nor did they develop an erythema to cutaneously applied ethanol. Lymphocytes obtained from 'atypical' individuals who do suffer from alcohol-induced flushing and rapid heart rate and who developed erythema to cutaneous ethanol displayed type II, but not type I, ALDH. Lymphocytes thus appear to be an easily accessible and suitable tissue for determining type I ALDH phenotype. Key Words: Aldehyde Dehydrogenase, Human, Lymphocyte, lsoenzymes

LDEHYDE DEHYDROGENASE (ALDH, E.C. A 1.2.1.3) is present in many tissues of the human body such as liver, skin, kidney, erythrocytes, platelets, and lympho~ytes.l-~ There are four major isoenzyme sets of ALDH and different types are found in different tissues.',' For example, the liver contains mainly types I, 11, and IV, while erythrocytes contain only type 11. In the present study, attention is focused on type I ALDH because liver ALDH-I appears to be quantitatively the most important isoenzyme with respect to catabolizing acetaldehyde6; furthermore, in Orientals a structural mutation of the ALDHI allele has occurred in about half of that population which leads to the production of an inactive ALDH-I isoen~ y m e . ' .It~ appears that individuals who are heterozygous or homozygous for the atypical mutant allele possess a catalytically inactive ALDH-L9,I0 Type I ALDH is a mitochondria1 enzyme with a high affinity (low K,,,) for acetaldehyde. When it is inactivated in humans either deliberately by administration of the drug, disulfiram, or unintentionally by inheritance of the inactive, atypical variant, the metabolism of alcohol is impaired because acetaldehyde is degraded more slowly than usual and accumulates."-" The accumulation of acetaldehyde following alcohol consumption is thought to cause physiological responses unpleasant enough to serve as a deterrent to further drinking in both disulfiramFrom the Neiiropsrlcchiatric Research L'nir, Department o/'Psychiulrj.. University of Saskatchewan. Saskatoon, Saskatchewnn, Canada. Receivedfi,r piiblication Diwember 26, 1989; accepted March 20, I990 Reprint requesls: Lillian E. Dyck. Ni~irrops.vcliiutricRewurch Unit. Department O/'P.rvchiurry.University of SaskatcheMnn. Saskatoon. Saskalchewan. Canada S 7 N 0 W0. Copyrighl Q I990 by The Rewarch Socicyj. on Alcoholism. 534

treated patients and in individuals who have inherited the atypical mutant allele of ALDH-I. Because of the association of ALDH-I activity with an individual's responses to alcohol, techniques for screening ALDH-I activity have been developed. Isoelectric focusing techniques have shown that hair roots contain ALDH-1 and thus analysis of hair roots has been used in population studies to differentiate between those individuals who possess the active (typical) and those who possess the inactive (atypical) variant of ALDH-I.'.* Hair roots, however, due to technical difficulties, may not be a reliable indicator of which ALDH-I variant a person expresses.14 It would be useful, therefore, to find another easily obtainable tissue source suitable for screening ALDH-I activity. Recently, it has been shown that human lymphocytes exhibit ALDH activity, which on the basis of kinetic studies and drug inhibition studies, appears to be type I ALDH.3-5 Lymphocytes, therefore, would seem to be a suitable tissue for determining human ALDH-I phenotypes (typical or atypical) using isoelectric focusing techniques. In the following paper, this supposition is addressed. Material and Methods C%rmicul.c The drugs used and their suppliers were as follows: @-nicotinamide adenine dinucleotide (NAD+), sodium pyruvate, disulfiram, (Sigma Chemicals): propionaldehyde (Terochem Laboratories); Meldolablau (Boehringer Mannheim); 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) (Aldrich); agarose (Bio-Rad); Ficoll-Paque, Sephadex (3-25(Pharmacia). Fradiunation of Blood Approximately 30 ml of blood from healthy volunteers was collected into vacutainers containing EDTA (10.5mg per tube) as the anticoagulant. Blood was centrifuged in a cold room for 30 min at 650 x g. Erythrocytes (250 pl) were lysed by the addition of 10 mM sodium phosphate buffer (750 pl). pH 7.4, which had been degassed, bubbled with nitrogen, and contained 1 mM EDTA-1 mM dithiothreitol. The buffy coats were pooled and layered on 3 ml of Ficoll-Paque in a 13 x 100 mm tube. After centrifuging for 30 min at 650 X g, the lymphocyte layer was removed, washed with 0.32 M sucrose, and centrifuged for 10 min at 650 x g. The lymphocyte pellet was lysed by the addition of 100 pl of the buffer described above and sonication (30 sec at 100 watts). Lysates from erythrocytes and lymphocytes were centrifuged for 5 min at 12,000X g, and then desalted by passage through a 1 X 4 cm column of Sephadex G-25. The G-25 gel was swelled with the buffer described above. The columns were centrifuged at 1400 X s f o r 1 min; the lysates were applied .A/coho/ U r n E.vp

Re\, Vol 14. No 4. 1990: pp 534-538

LYMPHOCYTE ALDEHYDE DEHYDROGENASES

535

to the gel surface and then the columns were centrifuged at 1400 x g for I min. The eluate was stored at -70°C for 1 week or less.

Preparation of Liver and Hair Root ALDH Human liver obtained post-mortem from Caucasians was homogenized in 1 mM EDTA-I mM dithiothreitol ( 1 g in 5 ml), centrifuged at 12,000 x g for 10 min and the supernatant was stored at -70°C for 2 to 3 months. Forty hair roots in the anagen phase were epilated quickly from the scalps of volunteer subjects. One mM EDTA- 1 mM dithiothreitol (100 p l ) was added to the hair roots; then the samples were frozen and thawed three times and centrifuged in the same manner as for blood cells. The supernate was stored at -70°C for 1 week or less. lsoelectric Focusing Ultrathin layers (0.5 mm) of polyacrylamide of the following composition were prepared: 5% acrylamide, 0.15% bisacrylamide, 2% ampholine (pH 3-10), 1% ampholine (pH 4-6), 10% sucrose, 32 mg% ammonium persulfate and 60 pl tetramethylenediamine/lOO ml. Samples (5-40 p l ) were applied to the gel using sample application papers. The electrode solutions were I M phosphoric acid and I M sodium hydroxide. Isoelectric focusing was camed out on a cooling plate (4°C) using an LKB 21 17 Multiphor electrophoresis unit. The gel was prerun for 20 min at maximum voltage, 15 mA and I5 watts (W). The samples were applied and run for 30 min at maximum voltage, 50 mA and 25 W. The application papers were removed and the samples focused for an additional 15 min; then, the power was increased to 30 W for an additional 15 min. ALDH Activity Staining The polyacrylamide gels were stained for ALDH activity by preparing an agarose gel overlay, which was placed on top of the polyacrylamide gel and incubated at 37°C for 1 hr in the dark.'.'' A 1 % agarose gel was prepared in 60 m l 0 . I M Tris, pH 8.6, which contained 24 mg MTT, 60 mg NAD+, 60 mg sodium pyruvate, 1.5 mg Meldolablau and 480 pI propionaldehyde. Blanks were prepared by omitting the substrate. Gels were washed with 5% acetic acid in 30% methanol to remove background stain, and then they were air dried. Ethanol Patch Test and Questionnaire The effect of cutaneous application of 70% ethanol was tested as described by Muramatsu et al., 1989.15Briefly, distilled water and 70% ethanol were placed separately onto the skin of the upper arm for 7 min, and the skin examined for erythema 10 min later. A positive reaction was the development of erythema. In addition, test subjects were asked to fill out a questionnaire regarding their physiological responses to the consumption of one typical alcoholic beverage. Consistent facial flushing and rapid heart rate after drinking alcohol were considered to be positive (adverse) responses.

RESULTS

ALDH Isoenzymes in Erytlzrocvtes und Lymphocytes Fig. 1 shows the pattern of enzyme staining which developed from samples incubated in the absence (i.e., blanks) and in the presence of propionaldehyde (PrAL). Samples of human liver and erythrocytes were applied to the gels as references. Liver contains mainly types I, 11, and IV ALDH, while erythrocytes (Rbc) contain only type II.I-3 The positions of hemoglobin (Hb), ALDH-I (open triangles), ALDH-I1 (filled triangles), and ALDH-IV (filled

Fig. 1. ALDH isoenzymes in samples of human liver (10 PI),hair roots (40 pl), lymphocytes (Wbc) (40 PI), and erythrocytes(Rbc) (20 pl). The top panel is a blank [no propionaldehyde (PrAL) in the gel overlay]. The anode was at the top of the gels. Propionaldehyde was used in the bottom panel. The positions of hemoglobin (Hb). type I ALDH (open triangles),type I1 ALDH (filled triangles). type IV ALDH (filled circles) and an aldehyde-independent dehydrogenase (open circles) are indicated.

circles) are indicated. The intensity of staining of ALDHI, 11, and IV was considerably greater when the substrate was present than when it was absent. A band which was more anodal than ALDH-I (indicated by the open circle) which was present in lysates from hair roots and lymphocytes stained as well in the absence as in 1.tpx..~-%i%%& propionaldehyde. Sometimes staining developed at the edges of the sample application papers. Hair roots lysates were applied onto the polyacrylamide gel at the cathodal end; they did not focus as well when applied at the anodal end. Lymphocyte samples, however, ran better when applied at the anodal side. Hair roots contained type I ALDH and other bands of activity which seemed to correspond to type IV, and possibly to type I11 which is slightly more anodal than IV.l-3 Lymphocytes (Wbc) contained types I and I1 ALDH. Blood samples were obtained from five typical individuals who reported that they did not respond adversely to alcohol consumption (i.e., no facial flushing or rapid heart rate) and who were negative o n the ethanol patch test. The lysates from their lymphocytes were pooled to give a

c

DYCK

536

large volume sample that could be analyzed repeatedly. This pooled lymphocyte sample as well as samples of liver and erythrocytes were applied to the gel in triplicate in the presence and absence of disulfiram (Fig. 2). In the absence of disulfiram, the lymphocyte samples displayed both ALDH-I and ALDH-I1 activities. When disulfiram ( 100 pM) was added to the gel overlay, ALDH staining was decreased, especially type I1 ALDH. In Fig. 3, lymphocytes obtained from typical individuals (1, 2, and 4) were compared with those obtained from atypical individuals (3, 5, and 6) who stated that they experienced facial flushing and rapid heart rate after alcohol consumption and who were positive in the ethanol patch test. Lymphoctyes obtained from typical people contained both type I and type I1 ALDH; in contrast, lymphocyte samples obtained from atypical individuals showed only ALDH-I1 activity; the type I isozyme was inactive. In Fig. 4,the pattern of ALDH isozymes present in hair root samples from seven typical individuals is shown. It is clear that type I ALDH was present in all the hair root samples, though it was stained weakly in two samples. None of the hair root samples displayed distinct type I1 ALDH activity* Sometimes bands Of ALDH activity ‘loser to the cathode (at the bottom of the photograph) which seemed to correspond to types 111 and IV ALDH were visible, but the intensity of staining was quite variable. A total of 30 individuals were asked to complete a questionnaire regarding adverse physiological responses to alcohol consumption and were subjected to the ethanol patch test. Hair roots and lymphocytes were collected from these people and their ALDH isozyme patterns deter-

Fig. 3. ALDH isozymes in lymphocytes from typical (1, 2, 4) and atypical (3, 5, 6)individuals. Human liver was also applied as a reference. The positions of type I

ALDH (open triangles) and type I1 ALDH (filled triangles) are indicated.

Fig. 4. ALDH isozymes in hair roots (40 pl of lysate was applied to the gel) from seven typical individuals. Human liver was also applied as a reference. The positions of type I ALDH (open triangles) and type II ALDH (filled triangles) are indicated. AlcoholClin Exp Res. Vol4, No 4, 1990 pp 534-538

Table 1. Results of the Ethanol Patch Test and Demonstration of the Presence or Absence of Type I ALDH in Hair Roots or Lymphocytes

Fig. 2. The effect of disulfiram on ALDH isozymes in human liver (5 pl), lymphocytes(40 pl) (WBCs). and erythrocytes (RBCs) (20 pl). The positions of type I ALDH (open triangles) and type II ALDH (filled triangles) are indicated. In the lower panel, the gel was stained in the presence of 100 pt.4 disulfiram.

Group

Ethanol patch test

Hair r m t ALDH-I

Lymphocyte ALDH-I

Typical Atypical

24 negatives 6 positives

24 positives 6 negatives

24 positives 6 negatives

Subjects were questioned about their physiological responses to consumption of an alcoholic drink. The atypical group reported the consistent appearance of marked facial flushing and rapid heart rate after drinking alcohol. Typical individuals did not develop these responses. In the patch test, negative denotes the absence and positive, the presence of erythema to ethanol. In the other columns, negative denotes the absence and positive, the presence of type I ALDH activity.

LYMPHOCYTE ALDEHYDE DEHYDROGENASES

mined. As can be seen in Table 1, the presence or absence of ALDH-I activity in hair roots and lymphocytes was concordant. All of the 24 people who showed type I ALDH activity in their hair root samples also showed type I activity in their lymphocytes. Similarly, the six individuals who did not show type I ALDH activity in their hair root samples also did not have type I activity in their lymphocytes. Furthermore, the absence of ALDH-I activity corresponded to reports of the development of marked facial flushing and rapid heart beat after alcohol consumption and to a positive ethanol patch test result. DISCUSSION

The ALDH activity of intact lymphocytes has been measured in vitro.”’ From these previous studies, it appeared that lymphocytes contained type I ALDH, but it is evident from the present results that lymphocytes also possess type I1 ALDH. Furthermore, because type I1 activity was present in all lymphocyte samples tested, the appearance of ALDH-I1 serves as a marker showing that ALDH activity (I or 11) is in fact measurable. By contrast, the activity of all types of ALDH in hair root samples was highly (see also Fig. 3); thus, the lack of ALDH-I activity staining in gels from hair roots might be due to a poor recovery of ALDH activity (I or any other isozyme) from the hair root rather than due to the presence of the inactive form of ALDH-I.I4 Such false negatives are not a problem with lymphocytes. Hair roots do contain ALDH-11, but it can stain weakly or not at all in some case^.',^.'^ The consistent appearance of type I1 ALDH in lymphocytes from all individuals (typical and atypical) serves as a marker of sample viability. The demonstration of the absence of type I in the presence of type I1 activity indicates a true type I deficiency, as opposed to an apparent deficiency due to poor enzyme recovery or loss of activity due to technical problems. Lymphocytes are thus a reliable indicator of an individual’s ALDH phenotype; they are also an easily obtainable tissue source for screening large numbers of people for the presence of the atypical (inactive) ALDH-I isozyme. Furthermore, lymphocytes could be used to monitor the effects of disulfiram therapy on ALDH-I activity as has been ~ u g g e s t e d , ~but . ~ the method may need to be modified to quantify the activity due to each ALDH isoenzyme. The effects of disulfiram on the activity of ALDH isozymes differs when given in vivo than when it is added in vitro. When injected into rats, disulfiram inhibits ALDH-I more than it inhibits ALDH-II,I6 but it has the opposite effect in vitro, where it inhibits ALDH-I1 more than ALDH-I.”-’* The present data agree with these previous findings. Disulfiram added to the agarose gel overlay appeared to inhibit ALDH-I1 more readily than ALDH-I. It has been shown that disulfiram therapy reduces erythrocyte ALDH activity by 90% and lymphocyte ALDH activity by 20-45%’,4; in view of the present findings, the

531

reduction in lymphocyte ALDH activity cannot be attributed solely to inhibition of ALDH-I without further experimentation. A method which separates and quantities the activity of each ALDH isoenzyme present in lymphocytes would have to be employed. It may be that disulfiram administration completely inhibits lymphocyte ALDH-I, but only partially inhibits ALDH-11. This remains to be determined. In conclusion, the present data show, in agreement with previous data, that lymphocytes contain type I ALDH; furthermore, it was demonstrated that lymphocytes also contain type I1 ALDH. Type I ALDH activity was absent in lymphocytes obtained from so-called atypical individuals who developed erythema to the cutaneous application of ethanol and who reported marked facial flushing and rapid heart rate after consumption of an alcoholic drink. Lymphocytes appear to be an easily accessible tissue source and a reliable indicator of ALDH I phenotype (typical or atypical) suitable for population screening. ACKNOWLEDGMENTS The author wishes to thank Professor A. A. Boulton for his interest and encouragement, P. McFie for excellent technical assistance, and Saskatchewan Health and Correctional Services of Canada, Regional Psychiatric Centre for financial support.

REFERENCES 1. Harada S, Agarwal DP, Goedde HW: Isozyme variations in acetaldehyde dehydrogenase (E.C. I .2. I . 3 ) in human tissues. Hum Genet 44: 18 1- 185, 1978 2. Goedde HW, Agarwal DP, Harada S: The role of alcohol dehydrogenase and acetaldehyde dehydrogenase isozymes in alcohol metabolism, alcohol sensitivity and alcoholism, in Rattazzi MC, Scandalios JG, Whiff GS (eds): Isozymes, Current Topics in Biological and Medical Research, vol. 8. New York, Alan R. Liss, Inc, 1983, pp 175-193 3. Helander A, Tottmar 0: Metabolism of biogenic aldehydes in isolated human blood cells, platelets and in plasma. Biochem Pharmacol 3611077-1082, 1987 4. Helander A, Carlsson S, Tottmar 0: Effects of disulfiram therapy on aldehyde dehydrogenase activity in human leukocytes and erythrocytes. Biochem Pharmacol 37:3360-3363, 1988 5. Helander A. Johansson B: Inhibition of human erythrocyte and leukocyte aldehyde dehydrogenase activities by diethylthiocarbamic acid methyl ester. Biochem Pharmacol 38:2 195-2 198, 1989 6. Weiner H: Aldehyde dehydrogenase: mechanism of action and possible physiological roles, in Majchrowicz E, Noble EP (eds): Biochemistry and Pharmacology of Ethanol, vol. 1. New York, Plenum Press, 1979, pp 107-144 7. Hsu L, Tani K, Fujiyoshi T, Kurachi K. Yoshida A: Cloning of cDNAs for human aldehyde dehydrogenases I and 2. Proc Natl Acad Sci USA 82:3771-3775, 1985 8. Hempel J, Kaiser R. Jornvall H: Mitochondria1 aldehyde dehydrogenase from human liver primary structure, differences in relation to the cytosolic enzyme and functional correlations. Eur J Biochem 153: 1328, 1985 9. Goedde HW. Singh S, Agarwal DP, Fritze G, Stapel K, Paik YK: Genotyping of mitochondria1 aldehyde dehydrogenase in blood samples using allele-specific oligonucleotides comparison with phenotyping in hair roots. Hum Genet 81:305-307, 1989 10. Crabb DW. Edenberg HJ, Bosron WF, Li T-K: Genotypes for aldehyde dehydrogenase deficiency and alcohol sensitivity. J Clin Invest 83:314-316, 1989

538

1 I . Kitson TM: The disulfiram-ethanol reaction. J Stud Alcohol 38~96-113,1977 12. Mizoi Y, Ijiri I, Tatsuno Y, Kijima T, Fujiwara S, Adachi J, Hishida S: Relationship between facial flushing and blood acetaldehyde levels after alcohol intake. Pharmacol Biochem Behav 10:303-3 I I , 1979 13. lnoue K, Fukunaga M, Kiriyama T, Komura S: Accumulation of acetaldehyde in alcohol-sensitive Japanese relation to ethanol and acetaldehyde oxidizing capacity. Alcohol Clin Exp Res 8:319-322, 1984 14. Yoshida A: Determination of aldehyde dehydrogenase phenotypes using hair roots: re-examination. Hum Genet 66:296-299, 1984 15. Muramatsu T, Higuchi S, Shigemori K, Saito M, Sasao M, Harada

DYCK

S, Shigeta Y, Yamada K, Muraoka H, Takagi S, Marayama K, Kono H: Ethanol patch test-a simple and sensitive method for identifying ALDH phenotype. Alcohol Clin Exp Res 13:229-23I , 1989 16. Marchner H, Tottmar 0: A comparative study on the effects of disulfiram, cyanamide and 1-aminocyclopropanol on the acetaldehyde metabolism in rats. Acta Pharmacol Toxicol 43:219-232, 1978 17. Dietrich F U , Erwin VG: Mechanism of the inhibition of aldehyde dehydrogenase in vivo by disulfiram and diethylthiocarbamate. Mol Pharmacol7:301-317,1971 18. Kitson TM: The effect of disulfiram on the aldehyde dehydrogenase of sheep liver. Biochemical J 15 1 :407-412, I975