Aldehyde dehydrogenase content and composition of human liver

Alcohol, Vol. 2, pp. 375--381, 1985. ©Ankho InternationalInc. Primed in the U.S.A.

0741-8329/85 $3.00 + .00

Aldehyde Dehydrogenase Content and Composition of Human Liver C L A I R E M. F O R T E - M c R O B B I E

AND REGINA PIETRUSZKO

Center o f Alcohol Studies Rutgers Univerisity, New Brunswick, NJ 08903

FORTE-McROBBIE, C. M. AND R. PIETRUSZKO. Aldehyde dehydrogenase content and composition of human liver. ALCOHOL 2(3) 375-381, 1985.--Human liver contains only four proteins which catalyze dehydrogenation of acetaldehyde; two of these are tetrameric with MW of 220,000 and are structurally related. These enzymes were purified previously to homogeneity and are now known as the cytoplasmic El and mitochondrial E2. The other two proteins do not appear to be structurally related to El and E2. The recently isolated E4 enzyme is a dimer of MW of ca. 175,000; the E3 may be a polymorphic enzyme. The Enzyme Commission classification for E 1 and E2 is EC 1.2.1.3, that for E4 is at present uncertain since its Michaelis constants for short chain aldehydes are high, making it unlikely that these would be its natural substrates. The relationship be.tween E3 and E4 is also uncertain. Employing a suitably designed assay, El and E2 are assayed as "low Kin" enzymes while E3 and E4 are assayed as "high Kin" enzymes. Therefore by employing such an assay, combined with electrofocusing procedure, an assessment of enzyme content and composition of aldehyde dehydrogenase in human liver can be made. Human aldehyde dehydrogenase Human liver "Low Kin" aldehyde dehydrogenases "High Kin" aldehyde dehydrogenases Cytoplasmic El enzyme Mitochondrial E2 enzyme

lar and one millimolar. Therefore while there was no problem with classification of E2 isozyme as "low K m " aldehyde dehydrogenase, difficulties have been sometimes encountered with the E 1 enzyme. All aldehyde dehydrogenases are frequently assigned the same Enzyme Commission number (EC 1.2.1.3) and it is often assumed that the " l o w K m " and " H i g h K i n " enzymes are related. Validation of this assumption by purification of the enzymes to homogeneity and careful delineation of their molecular and kinetic properties has not been available from any species. Recent purification and characterization of human "high K i n " aldehyde dehydrogenase E4 has made it possible to reevaluate the state of our knowledge about aldehyde dehydrogenase. We are now able to comment on the relationship of various aldehyde dehydrogenase enzymes, properties of the various enzymes, their relative proportions in the liver and the status of aldehyde dehydrogenase (EC 1.2.1.3) in the human liver. Knowledge of various human enzymes allows us to estimate the nature and extent of measurement of individual enzymes by various assay systems [8, 9, 10, 14, 18, 20] previously employed, if used with human liver homogenates. By employing a modified assay in conjunction with other procedures enzyme content and composition of 13 human livers in discussed.

BY combining tissue factionation procedure with a dual substrate concentration assay system for aldehyde dehydrogenase it was established in the rat [18] and later in other species [8, 9, 10, 14, 20] that there are two broadly defined groups of aldehyde dehydrogenase: a "low K i n " group with micromolar Michaelis constants and a ' " h i g h K m " group with millimolar constants for short chain aliphatic aldehydes and that both groups are distributed among subcellular compartments. Purification and characterization of two enzymes of aldehyde dehydrogenase (EC 1.2.1.3), one cytoplasmic, another mitochondrial, has been subsequently achieved from several mammalian species [1, 2, 5, 15]. Considerable degree of similarity was observed between correspondingly localized enzymes of different species which included kinetic properties and susceptibility to inhibition by disulfiram (Antabuse). In fact, differences between the mitochondrial and cytoplasmic enzymes from any one species were greater than those between the correspondingly localized enzymes from different species. Therefore, although subceilular fractionation of human liver could not be done because of the non-availability of fresh tissue, comparison with other mammalian enzymes allowed for the assignment of the subcellular localization of the two enzymes isolated from liver [5,13] with fair degree of certainty. Thus, the enzyme referred to as E1 (enzyme 1) was identified as cytoplasmic and E2 (enzyme 2) was identified as mitochondriai [13]; this has been since confirmed by subcellular fractionation [16,17]. Aldehyde dehydrogenases E 1 and E2 both have micromolar Michaelis constants for short chain aliphatic aldehydes [5] and these constants decrease with the increase of chain length of the aldehyde. The E l enzyme, however, was previously shown to deviate from Michaelis-Menten kinetics at pH 7, where two constants could be determined: one micromo-

METHOD

Materials N A D was purchased from Boehringer Mannheim, acetaldehyde and propionaldehyde were obtained from Baker Chemical Company and redistilled before use. All other chemicals were reagent grade. Antibodies to homogeneous

375

376 aldehyde dehydrogenases, E1 and E2, were developed in the rabbit by Grand Island Biological Company. All human liver autopsy specimens were maintained frozen at -70°C prior to use.

Activity Assays All aldehyde dehydrogenase assays were performed spectrophotometrically at 340 nm and 25°C in 3 ml volume employing a modified assay of Feldman and Weiner [3]: 0.1 M sodium pyrophosphate buffer, pH 9.0, containing 1 mM EDTA, 500/.¢M NAD and propionaldehyde as the substrate. Crude human liver extracts were assayed at two concentrations of propionaldehyde: 13.5 raM, measuring the total activity of the mixture, and 68/zM or 100 p,M, measuring only the activity of the "low K m " enzymes: the difference in activity between the two being a measure of "high Kin" enzyme activity.

Protein Assays The protein content was determined by the Lowry procedure [12] using bovine serum albumin (Sigma) as the standard, or by Coomassie Brilliant Blue protein dye binding assay (Bio-Rad) employing 7-globulin (Sigma) as the standard. The protein of the homogeneous El enzyme was determined by 280 nm absorption employing an extinction coefficient of 0.96 O.D./mg, I cm light path [5].

Purification of Aldehyde Dehydrogenases Enzyme 1 was purified to homogeneity as previously reported by Hempel et al. [7] while E4 was purified to homogeneity by the procedure of Forte-McRobbie and Pietruszko [4].

Kinetic Studies Kinetic studies were performed spectrophotometrically at 340 nm and 25°C with both acetaldehyde and propionaldehyde. All Michaelis constants were calculated from tangents of initial steady state velocity employing the procedure of Lineweaver-Burk [ I 1].

Liver Homogenation Homogenates from human livers obtained at autopsy within 6-8 hr following death were maintained frozen at -70°C until use, Homogenates were prepared both in the presence (A) and absence (B) of 1% Triton X-100 in the extraction buffer: 30 mM sodium phosphate buffer, pH 6.0 containing 1 mM EDTA and 0.1% 2-mercaptoethanol. The aldehyde dehydrogenase activities of the livers are reported in /zmoles NADH/min/g of liver tissue. The activity units "per milliliter of homogenate" were converted to "per gram of tissue" by taking the total volume of homogenate as the sum of the weight of liver and volume of extraction buffer employed in homogenation. Activity units were then multiplied by the total volume and subsequently divided by the weight of tissue. The data is also reported in nmoles NADH/min/mg of protein for some homogenates. Optical density units were converted to absolute units by using an extinction coefficient for N A D H of 6 . 2 2 × c m - l × m M -1 at 340 nm.

Antibody Interaction Liver (No. 12, Table 3, 2 g) extract was prepared by adding 3 ml of buffer as in procedure (B). The extract was

FORTE-McROBBIE AND P1ETRUSZKO isoelectrically focused on two agarose plates (114×225 cm) composed of 1% w/v agarose, I , ~ w/v sorbitol, and 0.063~ of Pharmalyte, pH 3-10 (Pharmacia Fine Chemicals). On the first plate, liver extract was applied to four sections of the agarose gel approximately 4 cm from the cathode. Following isoelectric focusing at 150 V for 15 hr, the gel was cut into 4 sections. On one of the sections (D in Fig. 1) aldehyde dehydrogenase enzymes were visualized by -activity stain" using 100 mM Tris/HCl buffer, pH 8.5 containing NAD (10 rag/30 ml), nitroblue tetrazolium (10 mg/30 ml), phenazine methosulfate (1 mg/30 ml) and 13 mM propionaldehyde. On the second section protein was visualized by staining with Coomassie Brilliant Blue (A in Fig. 1). The third section was placed in a humidity chamber, overlayed with a mixture of anti-El and anti-E2 antibodies, and then incubated overnight at room temperature (C in Fig. 1). The fourth section was directly submerged in a solution of saline (0.9% NaCI) and allowed to stand overnight at room temperature ( B in Fig. 1). Following overnight incubation, the antibody overlayed gel and control were soaked in copious volumes of saline for 3 days. Subsequently, both the experimental and control sections of gel were stained with Coomassie Brilliant Blue. On the second plate the extract was also applied four times: one sample to be developed with activity stain and three other samples, two to be overlayed with antibodies, one with anti-El, the other with anti-E2, and the third to be submerged directly in saline as a control. R E S U L T S AND DISCUSSION

Comparison of the kinetic and molecular properties of the four known human liver aldehyde dehydrogenases, El, E2, E3 and E4 is shown in Table 1, where diagramatic representation of their isoelectric focusing separation pattern is also included. Although all four enzymes are separable by isoelectric focusing not all are easily detectable on gels from crude liver homogenates. The El and E2 enzymes are clearly visualized regardless of the degree of purity of the enzyme sample, while the E3 and E4 enzymes are visualized with difficulty unless partially purified. Visualization of E3 and E4 in crude liver homogenates by isoelectric focusing and other electrophoretic techniques is usually hampered due to comigration with other enzymes which interfere with the formation of formazan color necessary to detect aldehyde dehydrogenases. Enzymes 1 and 2 have been available for some time in a homogeneous form, while E4 enzyme was only recently purified to homogeneity [4]. The E3 enzyme is not yet available homogeneous; its purification has been impeded since not all livers contain this enzyme (see Tables 1 and 3). During the E4 purification several attempts were made to find out whether there are in the liver any aldehyde dehydrogenases other than those shown in Table 1. This was done by careful assay of all column fractions and by applying buffers with high salt content onto chromatographic columns following the elution of the E4 enzyme and assaying the eluted fractions with 13.6 mM propionaldehyde. No aldehyde dehydrogenase other than those shown in Table 1 has been detected by this procedure. While E1 and E2 enzymes are homotetramers of MW ca. 220,000 with a subunit size of 54,000-55,000 [5], E4 is a dimer of MW 175,000 having a subunit size of 70,600 [4]. The molecular weight of E3 determined by our preliminary experiments appears to be similar to or smaller than that of E4. Therefore E3 may or may not be related to E4 enzyme which

HUMAN

pH

ALDEHYDE

I0

-

DEHYDROGENASE

377

-

-

-

E4 E3

-

- - - -

__El "---" E2

I

J

(A)

(B)

(C)

(D)

FIG. 1. Isoelectric focusing gel of a h u m a n liver h o m o g e n a t e developed by overlay with a mixture o f anti-E1 and anti-E2 antibodies. T h e conditions o f focusing and sample application are given in the Method section. (A) protein developed control, (B) saline control, (C) antibody overlayed experimental gel, (D) activity stained control.

TABLE 1 KINETIC AND MOLECULAR PROPERTIES OF HUMAN LIVER ALDEHYDE DEHYDROGENASES K m (~.M)

Diagram human liver ALDH isoelectric focusing pattern

Nomenclature Harada Ours

e t al. *

pH 7.0 AcetalPropiondehyde aldehyde

pH 8.5-9.5 Acetal- propiondehyde aldehyde

Molecular

Weight

Subunit Number Molecular of Weight Subunits

Enzyme Presence Commisin all sion Livers Number

pH 10

7--

5--

E4

ALDH IV

4950t

E3

ALDH II1

--

El E2

ALDH I1 ALDH I

33.6(1060) 3.05

9400t -5,(1000) 0.7,

2280? -98 1.75

18,0007 930* 5.9 1.2,

175,000t <175,000 220,000, 220,000,

70,600t -54,8005 54,2005

2

Yes

1.2.1.?

?

No

1.2.1.?

4 4

Yes Yes

1.2.1.3 1.2.1.3

3--

*Harada e t a l . (1980); buffer systems: pH 9.5, 0. I M sodium pyrophosphate, 2 mM NAD [14]. tForte-McRobbie and Pietrnszko (1984, manuscript submitted); buffer systems: pH 7.0, 0.1 M sodium phosphate, 1 mM EDTA, 500/~M NAD; pH 8.5, 0.1 M sodium pyrophosphote, 1 mM EDTA, 500 ~M NAD [18]. ~Greenfield and Pietruszko (1977); buffer systems: pH 7.0, 0. I M sodium phosphate, 1 mM EDTA, I mM NAD; pH 9.5, 0. i M sodium pyrophosphate, I mM EDTA, 2 mM NAD ll0l. The higher of E1 two Km values is shown in parentheses.

378

FORTE-McROBBIE AND PIETRUSZKO

(~

i /,

fly

(a) 30

20

I0 lilltlldlhylll, I

I0

mM

I

I

20

30

I

40

(acetaldehyde) "1 , rnM"1

'

4b

'

do

'

' 120

[prol~onoldeh]~e) -I, mM-I

FIG. 2. Double reciprocal plots with acetaldehyde and propionaldehyde and acetaldehyde saturation curves for the cytoplasmic El enzyme. (A)=a reciprocal plot with varied acetaldehyde (25--400/zM) in 0.1 M sodium pyrophosphate, pH 9.0 containing I mM EDTA and 0.5 mM NAD. Inset I--velocity vs. acetaldehyde concentration (0.003-13.2 mM) in the same conditions; inset II--veiocity vs. acetaldehyde concentration (0.003-25.6 raM) in 0.1 M phosphate buffer, pH 7.0 containing 1 mM EDTA and 1 mM NAD. (B)=a reciprocal plot with varied propionaldehyde (0.008--0.128 mM) at pH 9.0 in the same conditions as in (A).

it resembles both in molecular weight and in Michaelis constant for propionaldehyde (Table l), both of which clearly differ from those o f E l and E2 enzymes. Anti-E1 and anti-E2 antibodies do not cross react with homogeneous E4 enzyme, suggesting differences in primary structure. Thus it appears that enzyme 4 should be assigned a different Enzyme Commission classification than the E 1 and E2 enzymes. The first three digits will, however, probably be the same as those for E1 and E2 enzymes (Table 1). Whether E3 enzyme will have the same number as E4 will be determined by the results of purification and characterization. Its absence from some livers suggests that it is a polymorphic enzyme. When crude liver homogenates were isoelectrically focused and reacted with anti-El and anti-E2 antibodies on agarose gels the formation of the precipitin ring was found only in the anodal section of the gel. (PI=4.5-5.0) where E1 and E2 enzymes were visualized on the control gel stained for activity (Fig. 1). The absence of precipitin reaction outside this area o f the gel indicates the unlikelihood of the existence of aldehyde dehydrogenases (EC 1.2.1.3) other than E1 and E2 enzymes in human liver homogenates. The Michaelis constants for acetaldehyde and propionaidehyde for E l and E2 enzymes are in the micromolar range, while those for E3 and E4 enzymes are millimolar (Table 1). Because E1 enzyme was previously demonstrated to have a biphasic velocity with acetaldehyde and propionaldehyde [5] at pH 7.0, kinetic experiments were carried out both at pH 7.0 and 9.0 to asssess the appropriateness of the standard dual concentration assay system employed to measure the concentration of "low K m " and "high K m " aldehyde dehydrogenases in liver homogenates (see the Method Section). Using a new preparation of E1

enzyme the biphasic nature of saturation curve at pH 7.0 with acetaldehyde has been confirmed (see inset lI in Fig. 2A). The Michaelis constants calculated from this curve are shown in Table l; these values reconfirm those previously reported [5]. Therefore the E l enzyme has two Michaelis constants at pH 7.0; one in micromolar range and another in millimolar range (see Table 1). Since our assays are usually at pH 9.0 the saturation curve with acetaldehyde and propionaldehyde was also done at this pH. That for acetaldehyde is shown in inset I of Fig. 2A. It can be seen that a normal substrate saturation curve is obtained for E l at pH 9.0 giving a Km of ca. 100 ~M for acetaldehyde at both low (Fig. 2A) and high substrate concentrations (Table l). Thus, at pH 9.0 for the substrate concentration curve for E l is normal, giving only a single Michaelis constant which is micromolar. The Michaelis constant for propionaldehyde is 5.9 /~M at pH 9.0 (Fig. 2B). The assay systems which have been employed by different investigators to measure the total " l o w K m " and "high K m " aldehyde dehydrogenase activities of tissue homogehates are shown in Table 2. The approximate percent of enzymes that would be measured by each assay system if used with human liver homogenates was calculated using data previously reported [5] and those from Fig. 2. All assays would include human mitochondrial E2 enzyme in the "low K m " measurement (Table 2). With none of the assay systems listed would the total human liver aldehyde dehydrogenase activity be fully measured due to incomplete assay of the "high K m " enzymes which have millimolar Michaelis constants. This problem, however, is difficult to correct since aldehydes at high concentrations react chemically with coenzymes. The E l enzyme has a biphasic velocity at pH 7.0 and

HUMAN ALDEHYDE DEHYDROGENASE

379 TABLE 2

ASSAYS DESIGNED TO MEASURE "LOW" AND "HIGH" kin ALDEHYDE DEHYDROGENASE ACTIVITY IN TISSUE HOMOGENATES FROM VARIOUS SPECIES AS APPLIED TO HUMAN LIVER Human Liver Calculations~l

Aldehyde Dehydrosenase Assay Systems

Approximate percent of enzymes measured at concentrations of aldehyde used:

Substrate Concentration

(mM)

Reference, Species

Buffer, pH

[18] Tottmer et al. (1973), pyrophosphate, rat pH 8.8

For For NAD assay of assayof Concen"'low Kin" total Wation Aldehyde enzyme enzyme (raM)

0-0.07 mM

Other Components

El

5-18 mM

E2 E3 E4

El

E2

E3

E4

Acetaldehyde

0.05

5

0.5

pyrnzole, 0.1 mM rotenone, 2 ~M

31

96 - -

2

99 100

--

70

0.5

sodium amytal, 1.0 mM

18

91 - -

1

99 100

--

70

1.33

pyrazole, 1.67 mM <60 > 9 0 -

<2

100 100

--

<85

1 <83 100

--

50

[8] Horton and Barter (1975), rat

pyrophosphate, pH 8.8

Acetaldehyde

0.025

5

[9] Koivula (1975), man

pyrophosphate, pH 8.0

Acetaldehyde

0.06

18

[10] Lebsack et al. (1981), phosphate, baboon pH 7.4

Acetaldehyde

0.05

5

0.5

MsCI2, 1.15mM rotenone, 2 ~M &methyl pyrazole, 200 ttM

29 >90 - -

[14] Smolen et al. (1981), mouse

phosphate, pH 7.4

Propionaldehyde

0.05

5

1.0

pyrazole, I mM

46

100 - - < I

83 100

--

35

[20] Weiner and Ardelt (1984), rat

phosphate, pH 7.5

Propionaldehyde

0.07

5

1.0

47

100-

83 100

--

35

pyrophosphate, pH 9.0

Propionaldehyde

0.068

13.6

0.5

92

100

[5] Present assay, man

EDTA, I mM

m

<1 1

100 100 >90*

67

*Determined from Km value reported by Harada et al. (1980) [14]. ¶Based on assay excluding other components.

therefore as determined from kinetic data (inset II in Fig. 2A), only 50% of its total activity can be estimated by assays of Smolen et al. [14], Weiner and Ardelt [20] and Lebsack et al. [10] in Table 2 as "low Km" activity at this pH. Therefore E1 enzyme would appear with these assay systems as two enzymes: one "low Km" and one "high Km." The assay systems (Table 2) of Smolen et al. [14] and Weiner and Ardelt [20] employing propionaldehyde would give a slightly better estimation of El and E2 activity than that of Lebsack et ai. [10] employing acetaldehyde since the Michaelis constant of El for propionaldehyde is 5 /zM, while that for acetaldehyde is 34 p.M at pH 7.0. With use of these three assays over estimation would be made in the amount of E3 and E4 aldehyde dehydrogenase activity and an under estimation would be made of the E1 and E2 activity. If the pH of the assay is increased (see [18, 8, 9, 5], Table 2) some of these problems are eliminated. The most important of these is the fact that the E I substrate saturation curve is no longer biphasic at higher pH (see Fig. 2), thus simplifying the assay design. However it should be noted that the Michaelis constants of El are larger with increased pH (Table 1) and greater for acetaldehyde than propionaldehyde (Fig. 2A and B, Table 2). Therefore the assay employing propionaldehyde would measure more of El. The concentrations of 50/zM and 25/zM acetaldehyde which were used by Tottmar et al. and Horton and Barret (Table 2), respectively, would measure only part of El; the bulk of E1 activity would be measured at 5 mM acetaldehyde. These assays

would also make E 1 appear as two enzymes in a way analogous to the pH 7.0 assays [10, 14, 20]. While we have no data with the enzymes of aldehyde dehydrogenase at pH 8.0 employed by Koivula [9] the calculations shown in Table 2 were approximated from knowledge of km values at pH's 7.0 and 9.0. Because the km values of El with acetaldehyde at pH's 7.0 and 9.0 are 33.6 ~M and 100 ~tM, respectively, it is unlikely that Koivula measured a large proportion of El enzyme activity with the 60 IzM acetaldehyde employed. All the calculated values in Table 2 are presented without consideration of the other components in the assay mixtures. Although pyrazole is added in many of the assay systems used with tissue homogenates to inhibit alcohol dehydrogenase, human aldehyde dehydrogenases El, E2 and E4 are also inhibited by this compound. Thus all the estimated percent values in Table 2 where pyrazole is used may be lower than those calculated. In addition, we recently have found E4 to be inhibited by large concentrations of NAD. Therefore with the assay system of Koivula (Table 2), in which a concentration of 1.33 mM NAD is used, the "high Km" activity measured would again be lower than that calculated. The assay of Lebsack et al. (Table 2), unlike the others contains MgCI2, which has been shown to activate human E2 enzyme by two-fold and suppress three-fold the activity of El [19]. MgCI~ in the assay would therefore also cause an inaccurate determination of the "low Km" activity in human liver. With regard to the remaining components, rotenone

380

FORTE-McROBBIE AND PIETRUSZKO TABLE 3 ALDEHYDE

DEHYDROGENASE

PROFILE OF HUMAN CAUCASIAN

LIVER HOMOGENATES

Activity v.moles NADH/min/g liver (nmoles NADH/min/mg protein) Liver No. 1 2

Sex

Age

Medical History

Male

21

No pathology

__

m

3 4 4 4* 5*

Male Male Male Male Male

20 55 55 55 37

6 6* 7* 8*

Male Male Male Male

72 72 40 59

9* 10"

Female Male

88 67

11

Female

41

12?

Male

44

13

Female

64

__

Cause of Death

High km

5.1

4.3

0.7

14.7

+

1.1 5.1 2.7 3.5 2.1(16.3) 0.3 (5.2)

0.4 0.8 0.6 0.6 1.2 (9.3) 0.4 (7.0)

28.6 13.5 18.4 14.2 36.4 57.1

+

Multiple Trauma Multiple Trauma Multiple Trauma Alcoholism

1.5 5.8 3.3 4.1 3.3(25.6) 0.7(12.1)

+ + + +

Multiple Trauma Multiple Trauma ---

4.8(34~t) 4.1(36) 3.4(33.2) 3.1.(30.3)

3.1(22~t) 2.0(17.5) 2.8(27.3) 1.3(12.7)

1.7(12¢) 2.2(19) 0.6 (5.9) 1.9(18.6)

35.5 53.7 17.6 61.3

+ + + +

Pulmonary Embolus 3.7(33.5) -3.4(36.6)

2.3(20.8) 2.6(28.0)

1.4(12.7) 1.2(12.9)

37.8 35.3

+ +

Gastrointestinal Hemorrhage --

2.7

1.5

1.2

43.5

5.3

4.1

1.3

24.5

Heart Disease

4.5

2.6

1.8

40.9

--

No pathology No pathology No pathology Fatty/Cirrhotic Liver Fatty Liver Fatty Liver None Diabetes/Fatty Liver No pathology 13 Year Alcoholic No pathology Possible Alcoholic Hypertension

Presence of E3

Low km

Asphixia by Hanging --

Total

High Km Activity % of Total

+

+

*Livers homogenized in the presence of 1.0% Triton X-100; protein of these livers were determined by Lowry procedure [16]. ?Liver homogenized for antibody interaction study (see the Method section). ~tProtein determined by Coomassie Blue assay (see the Method section).

and sodium amytal, nothing is known at present of their effects on the human aldehyde dehydrogenases. In conclusion, none of the assays discussed provides a perfect means of determining the enzymes, but in general, use of higher pH and propionaldehyde as the substrate allows for more accurate determinations. Higher pH also diminishes some of the interference encountered by alcohol dehydrogenase catalyzed reduction of aldehyde by reducing the hydrogen ion concentration. It can be seen from Table 3 that the amount o f aldehyde dehydrogenase activity extractable from different human livers is variable (0.7-5.8 tzmoles/min/g) with " l o w K m " activity varying between 0.3--4.3/zmoles/min/g and "high K i n " activity varying between 0.4--2.2 ~,moles/min/g liver. As demonstrated in Table 2, (present assay, man), the "low K i n " assay includes both E1 and E2 enzymes, the E3 and E4 enzymes are included in the "high K m " enzymes, but, only ca. 60% of E4 enzyme is measured (Table 2). Although the content is variable, the isoelectric focusing separation patterns are relatively non-variable, with the exception of the E3 enzyme which is missing from some livers (Table 3). By employing Triton X-100 in the extraction buffer an increase was found in the amount of "high K i n " activity extracted (compare livers 4 and 6 in Table 3). Because Triton X-100 solubilizes membranes, the "high K i n " enzYme_s, E3 and E4, may actually be bound to the membranes or associ-

ated with them. The "high K m " activity constituted between 15-61% of total measured aldehyde dehydrogenase activity extracted; preliminary data suggesting that more of these enzymes may be present in fatty livers (Table 3). However, with the exception of great variability o f both " l o w and high K i n " activity in extracts o f human livers it is difficult at the present time to arrive at any generalized conclusion about the association o f aldehyde dehydrogenase enzyme content or total activity with any of the known variables such as age, sex or disease. SUMMARY

(1) It appears that there are only four aldehyde dehydrogenase enzymes in the human liver: E l , E2, E3 and E4; the E3 enzyme is absent from some livers suggesting that it is a polymorphic form. (2) The cytoplasmic El enzyme and mitochondrial E2 enzymes are structurally related; there are no other relatives of El and E2 in the human liver. The E4 enzyme does not appear to be related to El and E2 enzymes; its relationship to E3 enzyme is uncertain. (3) The cytoplasmic El enzyme and mitochondrial E2 enzyme are aldehyde dehydrogenases (EC 1.2.1.3), the E4 enzyme is not of this group, its E n z y m e Commission classification is to be determined.

HUMAN ALDEHYDE DEHYDROGENASE

381

(4) In our assay the E 1 and E2 enzymes are assayed as " l o w K i n " enzymes while E3 and E4 enzymes are assayed as "high K m " enzymes.

(5) Aldehyde dehydrogenase content of several human livers is described and found to be extremely variable in content but not in composition.

REFERENCES 1. Crow, K. E., T. M. Kitson, A. K. H. MacGibbon and R. D. Batt. Intracellular localization and properties of aldehyde dehydrogenases from sheep liver. Biochim Biophys Acta 350: 121128, 1974. 2. Eckfeldt, J. H. and T. Yonetani. Subceilular localization of the FI and F2 isozymes of horse liver aldehyde dehydrogenase. Arch Biochem Biophys 175: 717-722, 1976. 3. Feldman, R. I. and H. Weiner. Horse liver aldehyde dehydrogenase. I. Purification and characterization. J Biol Chem 247: 260-266, 1972. 4. Forte-McRobbie, C. and R. Pietruszko. Manuscript submitted for publication, 1985. 5. Greenfield, N. J. and R. Pietruszko. Two aldehyde dehydrogenases from human liver; isolation via affinity chromatography and characterization of the isozymes. Biochem Biophys Acta 483: 35--45, 1977. 6. Harada, S., D. P. Agarwal and W. H. Goedde. Electrophoretic and biochemical studies of human aldehyde dehydrogenase isozymes in various tissues. Life Sci 26: 1773-1780, 1980. 7. Hempel, J. D. H., D. M. Reed and R. Pietruszko. Human aldehyde dehydrogenase: Improved purification procedure and comparison of homogeneous isoenzymes El and E2. Alcoholism: Clin Exp Res 6: 417--425, 1982. 8. Horton, A. A. and M. Barrett. The subcellular localization of aldehyde dehydrogenase in rat liver. Arch Biochem Biophys 167: 426-436, 1975. 9. Koivula, T. Subcellular distribution and characterization of human liver aldehyde dehydrogenase fractions. Life Sci 16: 1563-1570, 1975. 10. Lebsack, M. E., E. R. Gordon and C. S. Lieber. Effect of chronic ethanol consumption on aldehyde dehydrogenase activity in the baboon. Biochem Pharmacol 30: 2273-2277, 1981.

I 1. Lineweaver, H. and J. D. Burk. The determination of enzyme constants. J Am Chem Soc 56: 658-667, 1934. 12. Lowry, O. H., N. J. Rosebrough, A. L. Farr and R. J. Randall. Protein measurement with the folin phenol reagent. J Biol Chem 193: 263-275, 1951. 13. Pietroszko, R. and T. Yonetani. Aldehyde dehydrogenases from liver. In: Methods in Enzymology, edited by S. P. Collowick and N. O. Kaplan. New York: Academic Press, 1980, pp. 772-781. 14. Smolen, A., A. L. Wayman, T. N. Smolen, D. R. Petersen and A. C. Collins. Subcellular distribution of hepatic aldehyde dehydrogenase activity in four inbred mouse strains. Comp Biochem Physiol 69c: 199-204, 1981. 15. Sugimoto, E., N. Takahashi, Y. Kitagawa and H. Chiba. Intracellular localization and characterization of beef liver aldehyde dehydrogenase isozymes. Agr Biol Chem 40: 2063-2070, 1976. 16. Takada, A., S. Takase, J. Nei and Y. Matsuda. Subcellular distribution of ALDH isozymes in the human liver. Alcoholism: CIin Exp Res 8: 123, Abstract 263, 1984. 17. Tipton, K. F. and G. T. M. Henehan. Distribution of aldehyde dehydrogenase activities in human liver. Alcoholism: CIin Exp Res 8: 131, Abstract 314, 1984. 18. Tottmar, S. O. C., H. Pettersson and K.-H. Kiessling. The subcellular distribution and properties of aldehyde dehydrogenases in rat liver. Biochem 3 135: 577-586, 1973. 19. Vallari, R. C. and R. Pietruszko. Interaction of M8++ with human liver aldehyde dehydrogenase. I. Species difference in the mitochondrial isozyme. J Biol Chem 259: 4922--4926, 1984. 20. Weiner, H. and B. Ardelt. Distribution and properties of aldehyde dehydrogenase in regions of rat brain. J Neurochem 42: 109-115, 1984.