ARCHIVES
Vol.
OF BIOCHEMISTRY
AND
299, No. 1, November
BIOPHYSICS
15, pp. 23-29,
1992
Human Mitochondrial Aldehyde Dehydrogenase Substrate Specificity: Comparison of Esterase with Dehydrogenase Reaction’ Neeta Mukerjee Centrr
Received
of Alcohol
April
and Regina Studies,
Rutgers
1, 1992, and in revised
Pietruszko’ University,
form
July
Piscataway,
New Jersey
17, 1992
Substrate specificity of human mitochondrial low K, aldehyde dehydrogenase (EC 1.2.1.3) E2 isozyme has been investigated employingp-nitrophenyl esters of acyl groups of two to six carbon atoms and comparing with that of aldehydes of one to eight carbon atoms. The esterase reaction was studied under three conditions: in the absence of coenzyme, in the presence of NAD (1 mM), and in the presence of NADH (160 KM). The maximal velocity of the esterase reaction with p-nitrophenyl acetate and propionate as substrates in the presence of NAD was 3.9-4.7 times faster than that of the dehydrogenase reaction. Under all other conditions the velocities of dehydrogenase and esterase reactions were similar; the lowest k,,, was forp-nitrophenyl butyrate in the presence of NAD. Stimulation of esterase activity by coenzymes was confined to esters of short acyl chain length; with longer acyl chain lengths or increased bulkiness (p-nitrophenyl guanidinobenzoate) no effect or even inhibition was observed. Comparison of kinetic constants for esters demonstrates that p-nitrophenyl butyrate is the worst substrate of all esters tested, suggesting that the active site topography is uniquely unfavorable for p-nitrophenyl butyrate. This fact is, however, not reflected in kinetic constants for butyraldehyde, which is a good substrate. The substrate specificity profile as determined by comparison of k,,,/K, ratios was found to be quite different for aldehydes and esters. For aldehydes k,,JK, ratios increased with the increase of chain length; with esters under all three conditions, a V-shaped curve was produced with a minimum at p-nitrophenyl butyrate. 1. IS92
Academic
Press,
Inc.
-
’ This work was supported by R. Brinkley Smithers Institute for Alcoholism Prevention, Charles and Johanna Busch Memorial Fund, and Research Scientist Award K05 AA00046 from NIAAA. ’ To whom correspondence should be addressed. Fax: (908) 932.5944.
Copyright Al1 rights
‘C’ 1992 by Academic Press, of’ reproduction in any form
Inc. reserved.
0885,5-0969
Aldehyde dehydrogenase (EC 1.2.1.3) with high affinity for short chain aliphatic aldehydes catalyzes dehydrogenation of aldehydes as well as hydrolysis of esters (l-3). NAD (not NADP) is the coenzyme for aldehyde dehydrogenation; NAD and NADH have been also reported to stimulate ester hydrolysis (1, 3). Aldehyde substrates constitute a wide range of aldehydes and include naturally occurring substrates such as aldehyde metabolites of biogenie amines, corticosteroids, histamine, and putrescine (4). The substrate specificity of human mitochondrial E2 isozyme (or other similar enzymes) with ester substrates of increasing chain length has not been studied in detail. Investigation of the effect of coenzymes on the esterase reaction was confined to p-nitrophenyl acetate (1,3) and p-nitrophenyl propionate (1). More recently Senior and Tsai (5) described a more detailed substrate specificity of esterase reaction of a rat mitochondrial aldehyde dehydrogenase employingp-nitrophenyl esters of varying chain lengths. However, the effect of coenzymes on esterase reaction (5) was also confined top-nitrophenyl acetate. The enzyme employed by Senior and Tsai (5) is different from that employed in this investigation; it is also mitochondrial in origin but utilizes both NAD and NADP as coenzymes (6) and is characterized by high millimolar K,,, values for short chain aliphatic aldehydes. It is generally believed that the esterase reaction of aldehyde dehydrogenase is “unphysiological” and only important for investigators of the aldehyde dehydrogenase catalytic mechanism. More recent data, however, suggest that metabolites of disulfiram which are esters of diethyldithiocarbamic acid might inhibit aldehyde dehydrogenase via its esterase function in a mechanism-based manner (7). Our recent results also demonstrate that isosorbide dinitrate, commonly employed as an antianginal medication and known to produce alcohol aversion, inactivates aldehyde dehydrogenase via a suicide mechanism (8) involving nitrate ester hydrolysis. It was also reported recently that metabolism of ethyl carbamate was inhibited
24
MUKERJEE
AND
PIETRUSZKO
by acetaldehyde, suggesting involvement of aldehyde dehydrogenase in ethyl carbamate metabolism (9). Human liver aldehyde dehydrogenase with a high affinity for short chain aliphatic aldehydes consists of three isozymes, mitochondrial E2 isozyme and two cytoplasmic isozymes, El and E3. The mitochondrial isozyme, E2, is the enzyme involved in metabolism of ethanol-derived acetaldehyde (10). In this paper we examine the mitochondrial E2 isozyme substrate specificity with p-nitrophenyl esters with acyl groups of two to six carbon atoms in the absence of coenzymes, and in the presence of NAD or NADH and compare them with aldehydes of similar structure under identical experimental conditions. The results are also compared with those of previous investigators and with the recent data of Senior and Tsai (5).
following the conversion ofp-nitrophenyl esters to p-nitrophenol spectrophotometrically at 25”C, at a wavelength of 400 nm, at pH 7.0, in 50 mM sodium phosphate buffer containing 1 mM EDTA. Esters were dissolved in ethanol and added in 0.05 ml of ethanol/3-ml cuvette volume. In order to compare the effect of coenzymes on the kinetic constants for all esters, four different reaction rates were recorded simultaneously on a Gilford Model 252 four-channel recording spectrophotometer for a fixed concentration of ester (i) in the absence of coenzyme, (ii) in the presence of 1 mM NAD, (iii) in the presence of 160 pM NADH, and (iv) control with no enzyme to correct for spontaneous hydrolysis. This process was repeated for five or six different concentrations used for Lineweaver-Burk plots. This procedure allowed unequivocal determination if changes in reaction velocity were real or a result of experimental error. Stock solution concentrations of p-nitrophenyl esters were determined by measuring the increase in absorbance at 400 nm, following hydrolysis in 0.1 N sodium hydroxide using an extinction coefficient of anion (18). The k,,, are turnover 18 mM-’ cm-’ for thep-nitrophenolate numbers (pm01 NADH mini active site-‘) calculated on the basis of two active sites/molecule (19) of molecular weight of 217,000.
MATERIALS
Thin-layer chromatography. Thin-layer chromatography was performed on silica plates with fluorescent indicator (10 X 10 cm, Machery Nagel & Co.). Two different solvent systems were used. Solvent system I consisted of petroleum ether:diethyl ether:acetic acid in the ratio of 90:10:1, and solvent system II consisted of benzene:ethanol in the ratio of 95:5. The spots were visualized under ultraviolet light at 254 nm.
AND
METHODS
Materials. Formaldehyde (37% v/v aqueous solution) and all p-nitrophenyl esters were obtained from Sigma; all other aldehydes and gold label ethanol were from Aldrich. All aldehydes, except formaldehyde, were distilled prior to use. NAD and NADH (both 100% grade 1) were from Boehringer Mannheim. Enzymepreparation. The mitochondrial E2 isozyme of human liver aldehyde dehydrogenase was purified to homogeneity by following the procedure of Hempel et al. (11). Homogeneity was confirmed by isoelectric focusing and specific activity. The enzyme was stored at 4’C in nitrogen-saturated 30 mM sodium phosphate buffer, pH 6.0, containing 1 mM EDTA and 2% (v/v) 2-mercaptoethanol. Before use, the enzyme was dialyzed against eight changes of nitrogen-saturated 30 mM sodium phosphate, pH 6.0, buffer containing 1 mM EDTA. To achieve nitrogen saturation, buffers were first exhaustively evacuated on an aspirator at room temperature to remove dissolved air. Nitrogen was then continuously bubbled through the buffer while it was cooled to 4°C; the buffer was stored under nitrogen in a firmly capped container at 4°C. Determination of specific actiuity. Specific activity was determined employing a standard assay described previously (12). Protein concentrations were determined by both 280.nm absorption (13) and the procedure of Lowry et al. (14), with bovine serum albumin as a standard. The concentration of active sites of E2 isozyme was determined from specific activity and the molecular weight of 217,000, assuming maximal specific activity of 1.6 pmol/min/mg and two active sites per molecule. Determination of kinetic constants. Aldehyde dehydrogenase activity was determined spectrophotometrically at 25°C in l-cm light path cuvettes by monitoring production of NADH at 340 nm. An extinction coefficient of 6.22 mM-’ cm-l was used for NADH. Phosphate buffer, 50 mM, pH 7.0, containing 500 pM NAD and 1 mM EDTA was used. Ethanol (0.05 m1/3-ml cuvette volume) was added to obtain better approximation to conditions employed for esterase reaction, which necessitated use of ethanol because of the insolubility of esters in water. The kinetic constants for formaldehyde were determined by measuring steady-state velocity (the E2 isozyme exhibits hysteretic time progress curves) at five different formaldehyde concentrations and data were plotted according to the Lineweaver-Burk (15) procedure. Slopes and intercepts were calculated by the statistical procedure of Wilkinson (16). For all other aldehydes determination of Michaelis constants by the Lineweaver-Burk procedure was not possible due to extremely low K,,, values. Therefore, kinetic constants were determined by using a single reaction curve and applying the integrated Michaelis-Menton equation (17). The concentrations of aldehydes used in the reactions were ca. 25 X K,. The results of the integration were then plotted according to the Lineweaver-Burk procedure and subjected to least squares analysis. In the majority of cases, duplicate determinations were performed, and average values were calculated. Esterase activity was determined by
RESULTS
Effect of coenzymes on esters of different acyl group chain length. The Lineweaver-Burk plots for five p-nitrophenyl esters of increasing acyl chain length (acetate to caproate) in the presence and absence of NAD (1 mM) or NADH (160 PM) are shown in Fig. 1. It can be seen that the effect of coenzymes is greatly dependent upon the chain length of ester substrates and varies from activation to inhibition. Activation occurs only with p-nitrophenyl acetate and p-nitrophenyl propionate. At acyl chain lengths longer than three carbon atoms NADH has little effect: there appears to be slight inhibition with p-nitrophenyl butyrate, some activation with p-nitrophenyl valerate, and no effect with p-nitrophenyl caproate. NAD is clearly inhibitory at acyl chain lengths longer than three carbon atoms (see C, D, and E in Fig. 1). The LineweaverBurk plots for ester hydrolysis in the presence of NAD lie below controls in Figs. 1A and lB, but move above controls in Figs. lC, lD, and 1E. Both activation and inhibition are more pronounced in the presence of NAD than in the presence of NADH. Kinetic constants for esters. Table I lists kinetic constants for saturated p-nitrophenyl esters from carbon atoms 2 through 6 determined in the absence of coenzyme. It can be seen that the K,,, value forp-nitrophenyl butyrate is unusually large as compared with other ester substrates. The k,,,/K, ratios (but not Scatvalues) show a minimum atp-nitrophenyl butyrate. Table II lists kinetic constants obtained in the presence of NAD (1 mM). In the presence of NAD K,,, values of all ester substrates are larger than those in the absence of coenzyme. p-Nitrophenyl butyrate again has the largest K,,, value in the way analogous to that in the absence of coenzyme. The kCatvalues for p-
HUMAN
MITOCHONDRIAL
ALDEHYDE
25
DEHYDROGENASE
Table III. In the presence of NADH K, values for pnitrophenyl acetate andp-nitrophenyl propionate are also larger than those in the absence of coenzyme (compare Tables I and III), and kcat values for these two substrates are also larger; k,,, values and k,,,/K, ratios also show a minimum at p-nitrophenyl butyrate. Both K, and kcat values in the presence of NADH are smaller than those in the presence of NAD (compare Tables II and III) and are more similar to those in the absence of coenzyme. Purification and retesting of p-nitrophenyl butyrate. Since p-nitrophenyl butyrate was an unusually poor substrate, the possibility was considered that it was contaminated with some unknown inhibitor. For this reason the compound was subjected to thin-layer chromatography as described under Materials and Methods. In both solvent systems a single ultraviolet absorbing spot was observed. The ultraviolet absorbing band was eluted from silica plates following thin-layer chromatography in solvent system I, dissolved in ethanol, filtered, and used directly for kinetic constant determinations. The values for K, (33 PM), kcat (22 mini’), and kcat/Km (0.7) were obtained in the absence of coenzyme. These values are similar to the values obtained with commercially supplied p-nitrophenyl butyrate (compare with Table I).
us (&)
0.0
0.2
0.4 us ()IM-1)
0.6
plots in the FIG. 1. Effect of ester chain length on LineweaverBurk absence of coenzyme and in the presence of NAD or NADH. A, p-nitrophenyl acetate; B, p-nitrophenyl propionate; C, p-nitrophenyl hutyrate; D, p-nitrophenyl valerate; E, p-nitrophenyl caproate. In the absence of coenzyme (O), in the presence of NAD (1 mM) (O), and in the presence of NADH (160 pM) (A). V = turnover number.
nitrophenyl acetate and p-nitrophenyl propionate are much larger than those in the absence of coenzymes. The kcat values decrease from p-nitrophenyl acetate to p-nitrophenyl butyrate to a minimum at p-nitrophenyl butyrate and then increase again as do k,,,/K,,, ratios listed in the last column of Table II. Kinetic constants determined in the presence of NADH (160 PM) are listed in
TABLE
Substrate
Specificity
The effect of coenzymes on kinetic constants of esters. Comparison of data listed in Tables I-III shows that NAD alters K,,, values for esters to a greater extent than NADH. The effect of coenzymes on k,,, values is also more pronounced with NAD than NADH. With acetate and propionate NAD increases kcat values, as compared with those in the absence of coenzyme (Table I), it lowers the kcatwith butyrate, but appears to have no effect on k,,, values with valerate and caproate. The k,,, values of acetate and propionate are also increased by NADH but to a lesser extent than by NAD (see Table III). Although both coenzymes affect kcat values of esters, usually an increase in kcatis accompanied with the increase in K,,,. Thus, k,,,/K, ratios remain almost unchanged in the presence and absence of coenzymes (Tables I-III). Under all three conditions, p-nitrophenyl hutyrate has
I
for E&erase Reaction in the Absence of Coenzyme k,., (mini)
Km (PM) p-Nitrophenyl ester of Acetate Propionate Rutyrate Valerate Caproate”
Expt. 1
Expt. 2
Average
Expt. 1
Expt. 2
Average
LtlKm (mini PM-')
2.1 0.9 30 4 9
2.7 2.5 59 8
2.4 1.7 44.5 6 9
15 10 29 10 49
20 14 24 18
17.5 12 26.5 14 49
7.3 7.1 0.6 2.3 5.4
Note. The assay system contained 50 a Single determination of K,,, and k,,,.
mM
sodium phosphate buffer, pH 7.0, 1 mM EDTA,
and 284
mM
ethanol.
26
MUKERJEE
AND TABLE
PIETRUSZKO II
Substrate Specificity for Esterase Reaction in the Presence of 1 mM NAD k,,, (min-‘) p-Nitrophenyl ester of Acetate Propionate Butyrate Valerate Caproate”
Expt. 1
Expt. 2
Average
Expt. 1
Expt. 2
Average
13 21 46 15 17
6 19 72 16
9.5 20 59 15.5 17
86 58 12 14 42
102 78 8 24
94 68 10 19 42
Note. The assay system contained 50 ’ Single determination of K,,, and k,,,
sodium phosphate buffer, pH 7.0, 1 mM EDTA,
mM
the lowest h,,,/K, ratio; this is most apparent especially in the presence of NAD. p-Nitrophenyl acetate under all three conditions has the highest k,,,/K,,, ratio. Kinetic constants of aldehydes. K,,, and kcat values for saturated aldehydes of one to eight carbon atoms are shown in Table IV. Except for formaldehyde, for which the K, value is high, all aldehydes have similar and very low (less than micromolar) K, values; kcatvalues are also quite similar for all aldehydes. The k,,,/K, ratio for formaldehyde is low, indicating that formaldehyde is the poorest substrate of all aldehydes employed in this set; this ratio increases with the increase of aldehyde chain length. Comparison of the dehydrogenase with the esterase reaction. The results presented in Tables I-IV demonstrate that kcat values for esters (lo-94 min-‘) are the same or higher than those for aldehydes (14-49 min-l). p-Nitrophenyl acetate and propionate in the presence of NAD (Table II) have kcatvalues considerably higher than aldehydes of the same chain length. The k,,, value for pnitrophenyl acetate in the presence of NAD is 3.9 times that for acetaldehyde, while that for p-nitrophenyl propionate in the presence of NAD is 4.7 times that for propionaldehyde (compare Tables II and IV). The k,,, value for p-nitrophenyl butyrate in the presence of NAD is the
TABLE
km/Km (min-’ pM 9.9 3.4 0.17 1.2 2.5
1 mM NAD, and 284 mM ethanol.
lowest of all k,,, values. Under all other conditions (compare Tables I-III with data on Table IV) the kcat values of these two reactions are similar. Despite similar k,,, values, since aldehydes have much smaller K, values than esters, they are better substrates. This is illustrated in Fig. 2 where k,,JK, ratios are plotted vs aldehyde or ester chain length (please note that the scale for aldehydes is an order of magnitude larger than that for esters). The k,,,/K, ratios for aldehydes increase rapidly with the increase of aldehyde chain length to butyraldehyde and then remain constant or even decrease slightly at hexanaldehyde (Fig. 2). To find out if aldehyde dehydrogenation followed a bell-shaped curve, octanaldehyde (Table IV) was employed as substrate. The k,,J K,,, ratio for octanaldehyde was almost twice as high as that for butyraldehyde, suggesting that aldehyde substrates continue improving with the increase of chain length. The ratios obtained with esters in the absence of coenzymes and in the presence of NAD (1 mM) and NADH (160 PM) are plotted for comparison. The dependence of esterase reaction upon the chain length of the acyl group of esters is different from that of aldehydes; it shows a most definite minimum atp-nitrophenyl butyrate. With the exception of p-nitrophenyl acetate, the lowest
III
Substrate Specificity for Esterase Reaction in the Presence of 160 FM NADH k,,, (min-‘)
Km(FM) p-Nitrophenyl ester of Acetate Propionate Butyrate Valerate Caproate”
Expt. 1
Expt. 2
Average
Expt. 1
Expt. 2
Average
4.5 2 27 8.5 8
3.6 4.4 40 8.5
4.1 3.2 33.5 8.5 8
31 16 25 17 52
39 21 15 29
35 22 20 23 52
Note. The assay system contained 50 ’ Single determination of K, and k,,,.
mM
‘)
sodium phosphate buffer, pH 7.0, 1
mM
EDTA,
160
pM
NADH,
and 284
mM
k,atlKn (min-’ PM-‘) 8.5 6.9 0.6 2.7 6.5 ethanol.
HUMAN
MITOCHONDRIAL
ALDEHYDE
h,,,/K, ratios for esters were in the presence of NAD and the highest were in the presence of NADH; the h,,,/K, ratios in the absence of coenzyme resembled those in the presence of NADH more than those in the presence of NAD (Fig. 2). Differences between h,,JK, ratios under these three conditions were, however, small (see Fig. 2 and Tables I, II, and III). Substrate inhibition Substrate and product inhibition. of the esterase reaction was not observed for any of the ester substrates within the concentration range used for K,,, determination (see Fig. 1). With p-nitrophenyl butyrate (46.5 PM) as substrate p-nitrophenol (125 PM) inhibited the esterase reaction to about 80% of control activity. p-Nitrophenol also inhibited the dehydrogenase reaction when formaldehyde was used as a substrate. Binding ofp-nitrophenol could not be demonstrated with propanal (Km = less than 1 PM), which saturates the enzyme at low micromolar concentrations. For this reason formaldehyde or glycolaldehyde (both poor substrates) was employed in these experiments. With formaldehyde (620 PM) as substrate the dehydrogenase reaction at 60 and 120 PM p-nitrophenol was 47 and 17% of control activity, respectively. Under the same conditions sodium acetate (50 mM) had no effect on the dehydrogenase reaction. When p-nitrophenyl butyrate (56 PM) was used as a substrate to monitor the esterase reaction, glycolaldehyde (2.5 mM) caused a decrease of enzyme activity to 43% of control. When glycolaldehyde (50 PM) was used as a substrate for the dehydrogenase reaction, ethyleneglycol monoacetate (16 mM, used here for its high solubility) caused a decrease of enzyme activity to 72% of control. Thus, dehydrogenase substrates inhibit the esterase reaction and ester substrates inhibit the dehydrogenase reaction. acetate E&erase burst. The substrates p-nitrophenyl andp-nitrophenyl guanidinobenzoate (K,,, = 40 PM; k,,, = 0.003 pmol/min/mg) chosen to study esterase burst differed significantly with respect to coenzyme effect; while hydrolysis of p-nitrophenyl acetate was stimulated by co-
TABLE
27
DEHYDROGENASE 200
12
150 8
E 0 z w
100 4
2 -0 2. fi u z
50
0
0 12345678
Chain Length FIG. 2. Comparison of chain length dependence of k,,,/K,,, values of aldehydes and p-nitrophenyl esters. Aldehydes (0). p-Nitrophenyl esters in the absence of coenzyme ( ), in the presence of 1 mM NAD (a), and in the presence of 160 pM NADH (0).
enzymes, both coenzymes had no effect on p-nitrophenyl guanidinobenzoate. The measurement of esterase burst was attempted at 400 nm employing a Varian 635 double beam spectrophotometer at a 1:lO scale expansion. Sodium phosphate, 50 mM, pH 7.0, containing 1 mM EDTA, kNAD (500 PM) or NADH (160 PM) at 25 or 5°C was used, employing an extinction coefficient of 9.8 mMpl cm -’ for the nitrophenolate anion. The reaction was started by addition of E2 isozyme (4.6-4.9 PM) to the above buffer containing the ester added to 1 ml of buffer in 0.015 ml ethanol, and also to the control containing buffer and ethanol. The burst amplitude was measured by subtracting absorbance of the control from that obtained by extrapolation of steady state velocity to 0 time. Correction was also made for any change of absorbance due to addition of enzyme. With p-nitrophenyl guanidinobenzoate (75 pM) as a substrate, the presence or absence of burst could be established at 25°C because its steady state velocity was low. With p-nitrophenyl acetate (150 PM), however, the steady
IV
Substrate Specificity for Straight Chain Saturated Aldehydes k,,, bin ‘1
Km(PM) Aldehyde Formaldehyde Acetaldehyde Propionaldehyde Butyraldehyde Valeraldehyde Hexanaldehyde Octanaldehyde
Expt. 1
Expt. 2
Average
Expt. 1
665 0.56 0.23 0.27 0.22 0.26 0.27
0.56 0.21 0.32 0.36 0.32 0.34
0.56 0.22 0.30
22 15 21 24 22 50
Note The assay system contained
Expt. 2
Average
26 14
24 14.5 27.5 26.5 23
29
50
mM
0.29 0.29 0.31
sodium phosphate buffer, pH 7.0, 1
mM
28 29 24 48
EDTA,
0.5
mM
NAD, and 284
0.04 43 66
92 91 79 158
49 mM
ethanol
28
MUKERJEE
AND
state velocity was high. Incorporation of chloral (50 or 200 mM) into the reaction system was attempted. In some experiments the temperature was lowered to 5°C to eliminate use of chloral with p-nitrophenyl acetate. With both substrates, in the absence of coenzymes, no burst of p-nitrophenolate anion was observed. In the presence of coenzymes, p-nitrophenyl guanidinobenzoate also exhibited no burst. In the presence of NAD p-nitrophenyl acetate appeared to exhibit a burst, but even at 5”C, or in the presence of chloral, the steady-state velocity was too fast to permit accurate extrapolation and calculation of the burst amplitude.
PIETRUSZKO
p-nitrophenyl acetate and propionate as substrates is, however, compensated by the corresponding increase of Km. Thus, the k,,,/K,,, ratios remain almost unchanged under all three conditions (compare Tables I, II, and III) showing that stimulation of reaction velocity is achieved by some other mechanism than improvement of substrate efficiency. Only p-nitrophenyl acetate efficiency as substrate may be somewhat improved by coenzymes as demonstrated by the slight increase of its k,,,/K,,, ratio [from 7.3-9.9 (NAD)-8.5 (NADH)]. When the acyl chain length of the ester is increased coenzyme stimulation disappears (see data for NADH) or the presence of coenzyme inhibits the esterase reaction (see data for NAD with esters of acyl groups of four to six carbon atom chain length). Even DISCUSSION with p-nitrophenyl propionate, the k,,/K,,, ratio decreases The most striking result of this investigation is the fact in the presence of NAD to 3.4 (compare Tables I and II), that the substrate specificity profile of aldehyde dehydrosuggesting that inhibition by NAD is already starting at genation is totally distinct from that of ester hydrolysis. p-nitrophenyl propionate. There was also no stimulation Comparison of k&K, ratios (Fig. 2) shows that catalytic or inhibition with a more bulky substrate, p-nitrophenyl efficiency of aldehyde substrates increases with the inguanidinobenzoate. Inhibition of esterase reaction by cocrease of chain length of aldehydes while the catalytic enzymes was never previously reported. Since inhibition efficiency of ester substrates decreases with the chain appears to occur only at longer acyl group chain length, length to a minimum at p-nitrophenyl butyrate and then it appears likely that coenzyme and ester binding sites increases again, forming a V-shaped curve. This suggests may overlap, this overlap becoming apparent with larger molecules. that aldehydes and esters of similar structure are bound Considerable changes in kinetic constants (both K,,, and differently to the enzyme. Both reactions could occur at the same active site with k catvalues) of all ester substrates occur following binding of NAD and NADH. The enzyme form with bound NAD binding of esters occurring in a different orientation from that of aldehydes. This is consistent with the fact that p- is not as good as the other two enzyme forms for ester nitrophenol also binds to aldehyde dehydrogenase and hydrolysis as demonstrated by consistently lower k,,,/K, acetate appears to be inhibits both esterase and dehydrogenase reactions (see ratios (Fig. 2). While p-nitrophenyl the best ester substrate of all esters used during this inabove). Therefore, binding of p-nitrophenyl esters could vestigation, p-nitrophenyl butyrate is the worst substrate occur both via their acyl groups and through p-nitrophenol. The fact that chloral (an aldehyde substrate an- for all three enzyme forms. This suggests that the active alogue) also inhibits both dehydrogenase and esterase re- site topography of all enzyme forms is such that it fits pactions (1, 3) is also consistent with different binding of nitrophenyl butyrate in a uniquely unfavorable way. In butyrate is the worst aldehydes and esters at the same active site. It is also terms of k,,,/K, valuep-nitrophenyl consistent with the conclusions of Senior and Tsai (5) substrate of all for the enzyme form that binds NAD and is primarily involved in dehydrogenation of aldehydes. who studied the esterase reaction of a related enzyme. This fact is, however, not reflected in Km or kcat values However, the possibility that aldehyde dehydrogenation and ester hydrolysis may occur at different active sites for butyraldehyde, which is a good substrate (Table IV) and is another indication of the binding sites for aldehydes cannot be excluded by any of the above experiments. Thus, the distinct substrate specificity profile for the de- and esters being distinct. There is at present no feasible explanation as to why p-nitrophenyl butyrate is the worst hydrogenase and esterase reactions (Fig. 2) could be insubstrate. At first, the presence of some inhibitory imterpreted in terms of the same catalytic site or in terms purity in butyrate was suspected. For that reason p-niof distinct catalytic sites as proposed by other investitrophenyl butyrate was repurified and its Michaelis congators (20-24). Coenzyme stimulation of the esterase reaction was also stant redetermined. The results, however, showed that there was no impurity in p-nitrophenyl butyrate which previously described (1, 3) but only with acetate and procould possibly account for these results. pionate as substrates. The coenzyme stimulating effect is Substrate specificity for p-nitrophenyl esters of inalso apparent in this work when p-nitrophenyl acetate creasing chain length was previously studied employing andp-nitrophenyl propionate are used as substrates. NAD proteolytic enzyme from Bacillus subtilis (25, increases the velocity of hydrolysis of p-nitrophenyl ac- subtilisin-a 26). The chain length profile employing k,,+/K, ratio also etate and p-nitrophenyl propionate about five-fold while the increase is only two-fold with NADH (compare kcat showed a bell-shaped curve with a maximum at p-nitrophenyl butyrate; chain lengths longer than butyrate provalues in Tables I-III). The increase of k,,, values with
HUMAN
MITOCHONDRIAL
ALDEHYDE
duced reduction in k,,,/K,,, value. Thus, the bell-shaped curve of its k,,,/K, chain length profile aldehyde dehydrogenase resembles subtilisin, except that the profile shows a minimum rather than the maximum at p-nitrophenyl butyrate. Substrate specificity with ester substrates of increasing chain length was also recently described for a high K,,, aldehyde dehydrogenase from rat liver mitochondria (5, 6). In this case the substrate specificity profile also showed a maximum at p-nitrophenyl butyrate demonstrating that p-nitrophenyl butyrate was the best substrate. The magnitudes of the k,,, values of the esterase reaction (IO-94 mini’) relative to those of dehydrogenase reaction (14-49 mini’) are at variance with those previously reported by Feldman and Weiner (1) who, employing horse mitochondrial aldehyde dehydrogenase, reported that the esterase reaction velocity was three times less than that of dehydrogenase reaction. These results also disagree with those of Sidhu and Blair (3) who employed the same enzyme as the one used during this investigation for their kinetic study and also reported lower rates for the esterase than for the dehydrogenase reaction. The reason why this occurred is not clear, unless both authors compared the dehydrogenase reaction at pH 9.0 (where it is 5 times faster) with esterase at pH 7.0. During this investigation extreme care was taken to employ conditions for the dehydrogenase reaction resembling as closely as possible those employed for the esterase reaction. The results have also been checked several times to assure that calculations were correct. Examination of kcatvalues (Tables I-IV) shows that esterase velocity of the reaction catalyzed by the E2 isozyme is in some cases considerably higher than that of dehydrogenase reaction. This is especially apparent with p-nitrophenyl acetate and propionate in the presence of NAD (Table II) where the esterase velocity is 3.9-4.7 times faster than the dehydrogenase velocity. These results suggested that coenzymes may be involved in alteration of the rate-limiting step of the esterase reaction. For this reason detection of burst with p-nitrophenyl acetate was attempted. However, due to the inadequacy of the available instrumentation this could not be resolved at this time. Our most recent results (27) demonstrate that both aldehyde and ester substrates form a covalent intermediate with the same cysteine residue (cysteine 302) of human cytoplasmic El and human mitochondrial E2 isozymes. Aldehyde dehydrogenase, however, is an enzyme which displays extremely pronounced “half-of-the-sites” reactivity, more pronounced than that seen in glyceraldehyde3-phosphate dehydrogenase (28), with only two active sites located on an enzyme composed of four structurally identical subunits (19). This makes it likely that cysteine 302 residues located on some subunits mav catalvze al-
DEHYDROGENASE
29
dehyde dehydrogenation while other cysteine 302 residues may be concerned with ester hydrolysis. However, more information is necessary, possibly from X-ray crystallography, before any definite conclusions about the sites of aldehyde dehydrogenation and ester hydrolysis can be made. REFERENCES 1. Feldman, R. I., and Weiner, H. (1972) J. Biol. Chen. 247, 267272. 2. Bodley, F. H., and Blair, A. H. (1971) Can. J. Biochem. 49, l-5. 3. Sidhu, R. S., and Blair, A. H. (1975) J. Biol. Chem. 250, 78947898. 4. Pietruszko, R. (1989) in Biochemistry and Physiology of Substance Abuse (Watson, R., Ed.), pp. 89-127, CRC Press,Boca Raton, FL. 5. Senior, D. J., and Tsai, C. S. (1990) Biochem. Cell Biol. 68, 758763. 6. Senior, D. J., and Tsai, C. S. (1988) Arch. Biochem. Biophys. 262, 211-220. 7. Johansson, B. (1989) Pharmacol. Z’oxicol. 64, 471-474. 8. Mukerjee, N., Blatter, E. E., and Pietruszko, R. (1992) FASBB J. 6, A333. [Abstract 19171 9. Kurata, N., Kemper, R., Hurst, H. E., and Waddell, W. J. (1990) Drug Metab. Dispos. 18, 504-507. 10. Parilla, R., Okhawa, K., Lindros, K. O., Zimmerman, U.-I. P., Kobayashi, K., and Williamson, J. R. (1974) J. Biol. Chem. 249,49264933. 11. Hempel, J. D., Reed, D. M., and Pietruszko, R. (1982) Alcoholism: Clin. Exp. Res. 6, 417-425. 12. MacKerell, A. D., Jr., MacWright, R. S., and Pietruszko, R. (1986) Biochemistry 25, 5182-5189. 13. Greenfield, N. J., and Pietruszko, R. (1977) Biochim. Biophys. Acta 483,35-45. 14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 15. Lineweaver, H., and Burk, D. (1934) J. Am. Chem. Sot. 56, 65% 667. 16. Wilkinson, G. N. (1961) Biochem. J. 80, 324-332. 17. Yun, S.-L., and Suelter, C. H. (1977) Biochim. Biophys. Acta 480, l-13. 18. Kezdy, F. J., and Bender, M. L. (1962) Biochemistry 1, 109771106. 19. Ambroziak, W., Kosley, L. L., and Pietruszko, R. (1989) Biochemistry 28,5367-5373. 20. MacGibbon, A. K. H., Haylock, S. J., Buckley, P. D., and Blackwell, L. F. (1978) Biochem. J. 171, 533-538. 21. Motion, R. L., Blackwell, L. F., and Buckley, P. D. (1984) Biochemistry 23, 6851-6857. 22. Deady, L. W., Buckley, P. D., Bennett, A. F., and Blackwell, L. F. (1985) Arch. Biochem. Biophys. 243, 586-597. 23. Tu, G.-C., and Weiner, H. (1988) J. Biol. Chem. 263, 1212-1217. 24. Tu, G.-C., and Weiner, H. (1988) J. Biol. Chem. 263, 1218-1222. 25. Philipp, M., Tsai, I.-H., and Bender, M. L. (1979) Biochemistry 18,
3769-3773. 26. Graae, J. (1954) Acta Chem. &and. 8, 356-357. 27. Blatter, E. E., Abriola, D. P., and Pietruszko, R. (1992) Biochem. J. 282, 353-360. 28. Harris, J. I., and Waters, M. (1976) The Enzymes (Bayer, P. D., Ed.), Vol. 13, 3rd ed., pp. l-40, Academic Press, New York.
Vol.
OF BIOCHEMISTRY
AND
299, No. 1, November
BIOPHYSICS
15, pp. 23-29,
1992
Human Mitochondrial Aldehyde Dehydrogenase Substrate Specificity: Comparison of Esterase with Dehydrogenase Reaction’ Neeta Mukerjee Centrr
Received
of Alcohol
April
and Regina Studies,
Rutgers
1, 1992, and in revised
Pietruszko’ University,
form
July
Piscataway,
New Jersey
17, 1992
Substrate specificity of human mitochondrial low K, aldehyde dehydrogenase (EC 1.2.1.3) E2 isozyme has been investigated employingp-nitrophenyl esters of acyl groups of two to six carbon atoms and comparing with that of aldehydes of one to eight carbon atoms. The esterase reaction was studied under three conditions: in the absence of coenzyme, in the presence of NAD (1 mM), and in the presence of NADH (160 KM). The maximal velocity of the esterase reaction with p-nitrophenyl acetate and propionate as substrates in the presence of NAD was 3.9-4.7 times faster than that of the dehydrogenase reaction. Under all other conditions the velocities of dehydrogenase and esterase reactions were similar; the lowest k,,, was forp-nitrophenyl butyrate in the presence of NAD. Stimulation of esterase activity by coenzymes was confined to esters of short acyl chain length; with longer acyl chain lengths or increased bulkiness (p-nitrophenyl guanidinobenzoate) no effect or even inhibition was observed. Comparison of kinetic constants for esters demonstrates that p-nitrophenyl butyrate is the worst substrate of all esters tested, suggesting that the active site topography is uniquely unfavorable for p-nitrophenyl butyrate. This fact is, however, not reflected in kinetic constants for butyraldehyde, which is a good substrate. The substrate specificity profile as determined by comparison of k,,,/K, ratios was found to be quite different for aldehydes and esters. For aldehydes k,,JK, ratios increased with the increase of chain length; with esters under all three conditions, a V-shaped curve was produced with a minimum at p-nitrophenyl butyrate. 1. IS92
Academic
Press,
Inc.
-
’ This work was supported by R. Brinkley Smithers Institute for Alcoholism Prevention, Charles and Johanna Busch Memorial Fund, and Research Scientist Award K05 AA00046 from NIAAA. ’ To whom correspondence should be addressed. Fax: (908) 932.5944.
Copyright Al1 rights
‘C’ 1992 by Academic Press, of’ reproduction in any form
Inc. reserved.
0885,5-0969
Aldehyde dehydrogenase (EC 1.2.1.3) with high affinity for short chain aliphatic aldehydes catalyzes dehydrogenation of aldehydes as well as hydrolysis of esters (l-3). NAD (not NADP) is the coenzyme for aldehyde dehydrogenation; NAD and NADH have been also reported to stimulate ester hydrolysis (1, 3). Aldehyde substrates constitute a wide range of aldehydes and include naturally occurring substrates such as aldehyde metabolites of biogenie amines, corticosteroids, histamine, and putrescine (4). The substrate specificity of human mitochondrial E2 isozyme (or other similar enzymes) with ester substrates of increasing chain length has not been studied in detail. Investigation of the effect of coenzymes on the esterase reaction was confined to p-nitrophenyl acetate (1,3) and p-nitrophenyl propionate (1). More recently Senior and Tsai (5) described a more detailed substrate specificity of esterase reaction of a rat mitochondrial aldehyde dehydrogenase employingp-nitrophenyl esters of varying chain lengths. However, the effect of coenzymes on esterase reaction (5) was also confined top-nitrophenyl acetate. The enzyme employed by Senior and Tsai (5) is different from that employed in this investigation; it is also mitochondrial in origin but utilizes both NAD and NADP as coenzymes (6) and is characterized by high millimolar K,,, values for short chain aliphatic aldehydes. It is generally believed that the esterase reaction of aldehyde dehydrogenase is “unphysiological” and only important for investigators of the aldehyde dehydrogenase catalytic mechanism. More recent data, however, suggest that metabolites of disulfiram which are esters of diethyldithiocarbamic acid might inhibit aldehyde dehydrogenase via its esterase function in a mechanism-based manner (7). Our recent results also demonstrate that isosorbide dinitrate, commonly employed as an antianginal medication and known to produce alcohol aversion, inactivates aldehyde dehydrogenase via a suicide mechanism (8) involving nitrate ester hydrolysis. It was also reported recently that metabolism of ethyl carbamate was inhibited
24
MUKERJEE
AND
PIETRUSZKO
by acetaldehyde, suggesting involvement of aldehyde dehydrogenase in ethyl carbamate metabolism (9). Human liver aldehyde dehydrogenase with a high affinity for short chain aliphatic aldehydes consists of three isozymes, mitochondrial E2 isozyme and two cytoplasmic isozymes, El and E3. The mitochondrial isozyme, E2, is the enzyme involved in metabolism of ethanol-derived acetaldehyde (10). In this paper we examine the mitochondrial E2 isozyme substrate specificity with p-nitrophenyl esters with acyl groups of two to six carbon atoms in the absence of coenzymes, and in the presence of NAD or NADH and compare them with aldehydes of similar structure under identical experimental conditions. The results are also compared with those of previous investigators and with the recent data of Senior and Tsai (5).
following the conversion ofp-nitrophenyl esters to p-nitrophenol spectrophotometrically at 25”C, at a wavelength of 400 nm, at pH 7.0, in 50 mM sodium phosphate buffer containing 1 mM EDTA. Esters were dissolved in ethanol and added in 0.05 ml of ethanol/3-ml cuvette volume. In order to compare the effect of coenzymes on the kinetic constants for all esters, four different reaction rates were recorded simultaneously on a Gilford Model 252 four-channel recording spectrophotometer for a fixed concentration of ester (i) in the absence of coenzyme, (ii) in the presence of 1 mM NAD, (iii) in the presence of 160 pM NADH, and (iv) control with no enzyme to correct for spontaneous hydrolysis. This process was repeated for five or six different concentrations used for Lineweaver-Burk plots. This procedure allowed unequivocal determination if changes in reaction velocity were real or a result of experimental error. Stock solution concentrations of p-nitrophenyl esters were determined by measuring the increase in absorbance at 400 nm, following hydrolysis in 0.1 N sodium hydroxide using an extinction coefficient of anion (18). The k,,, are turnover 18 mM-’ cm-’ for thep-nitrophenolate numbers (pm01 NADH mini active site-‘) calculated on the basis of two active sites/molecule (19) of molecular weight of 217,000.
MATERIALS
Thin-layer chromatography. Thin-layer chromatography was performed on silica plates with fluorescent indicator (10 X 10 cm, Machery Nagel & Co.). Two different solvent systems were used. Solvent system I consisted of petroleum ether:diethyl ether:acetic acid in the ratio of 90:10:1, and solvent system II consisted of benzene:ethanol in the ratio of 95:5. The spots were visualized under ultraviolet light at 254 nm.
AND
METHODS
Materials. Formaldehyde (37% v/v aqueous solution) and all p-nitrophenyl esters were obtained from Sigma; all other aldehydes and gold label ethanol were from Aldrich. All aldehydes, except formaldehyde, were distilled prior to use. NAD and NADH (both 100% grade 1) were from Boehringer Mannheim. Enzymepreparation. The mitochondrial E2 isozyme of human liver aldehyde dehydrogenase was purified to homogeneity by following the procedure of Hempel et al. (11). Homogeneity was confirmed by isoelectric focusing and specific activity. The enzyme was stored at 4’C in nitrogen-saturated 30 mM sodium phosphate buffer, pH 6.0, containing 1 mM EDTA and 2% (v/v) 2-mercaptoethanol. Before use, the enzyme was dialyzed against eight changes of nitrogen-saturated 30 mM sodium phosphate, pH 6.0, buffer containing 1 mM EDTA. To achieve nitrogen saturation, buffers were first exhaustively evacuated on an aspirator at room temperature to remove dissolved air. Nitrogen was then continuously bubbled through the buffer while it was cooled to 4°C; the buffer was stored under nitrogen in a firmly capped container at 4°C. Determination of specific actiuity. Specific activity was determined employing a standard assay described previously (12). Protein concentrations were determined by both 280.nm absorption (13) and the procedure of Lowry et al. (14), with bovine serum albumin as a standard. The concentration of active sites of E2 isozyme was determined from specific activity and the molecular weight of 217,000, assuming maximal specific activity of 1.6 pmol/min/mg and two active sites per molecule. Determination of kinetic constants. Aldehyde dehydrogenase activity was determined spectrophotometrically at 25°C in l-cm light path cuvettes by monitoring production of NADH at 340 nm. An extinction coefficient of 6.22 mM-’ cm-l was used for NADH. Phosphate buffer, 50 mM, pH 7.0, containing 500 pM NAD and 1 mM EDTA was used. Ethanol (0.05 m1/3-ml cuvette volume) was added to obtain better approximation to conditions employed for esterase reaction, which necessitated use of ethanol because of the insolubility of esters in water. The kinetic constants for formaldehyde were determined by measuring steady-state velocity (the E2 isozyme exhibits hysteretic time progress curves) at five different formaldehyde concentrations and data were plotted according to the Lineweaver-Burk (15) procedure. Slopes and intercepts were calculated by the statistical procedure of Wilkinson (16). For all other aldehydes determination of Michaelis constants by the Lineweaver-Burk procedure was not possible due to extremely low K,,, values. Therefore, kinetic constants were determined by using a single reaction curve and applying the integrated Michaelis-Menton equation (17). The concentrations of aldehydes used in the reactions were ca. 25 X K,. The results of the integration were then plotted according to the Lineweaver-Burk procedure and subjected to least squares analysis. In the majority of cases, duplicate determinations were performed, and average values were calculated. Esterase activity was determined by
RESULTS
Effect of coenzymes on esters of different acyl group chain length. The Lineweaver-Burk plots for five p-nitrophenyl esters of increasing acyl chain length (acetate to caproate) in the presence and absence of NAD (1 mM) or NADH (160 PM) are shown in Fig. 1. It can be seen that the effect of coenzymes is greatly dependent upon the chain length of ester substrates and varies from activation to inhibition. Activation occurs only with p-nitrophenyl acetate and p-nitrophenyl propionate. At acyl chain lengths longer than three carbon atoms NADH has little effect: there appears to be slight inhibition with p-nitrophenyl butyrate, some activation with p-nitrophenyl valerate, and no effect with p-nitrophenyl caproate. NAD is clearly inhibitory at acyl chain lengths longer than three carbon atoms (see C, D, and E in Fig. 1). The LineweaverBurk plots for ester hydrolysis in the presence of NAD lie below controls in Figs. 1A and lB, but move above controls in Figs. lC, lD, and 1E. Both activation and inhibition are more pronounced in the presence of NAD than in the presence of NADH. Kinetic constants for esters. Table I lists kinetic constants for saturated p-nitrophenyl esters from carbon atoms 2 through 6 determined in the absence of coenzyme. It can be seen that the K,,, value forp-nitrophenyl butyrate is unusually large as compared with other ester substrates. The k,,,/K, ratios (but not Scatvalues) show a minimum atp-nitrophenyl butyrate. Table II lists kinetic constants obtained in the presence of NAD (1 mM). In the presence of NAD K,,, values of all ester substrates are larger than those in the absence of coenzyme. p-Nitrophenyl butyrate again has the largest K,,, value in the way analogous to that in the absence of coenzyme. The kCatvalues for p-
HUMAN
MITOCHONDRIAL
ALDEHYDE
25
DEHYDROGENASE
Table III. In the presence of NADH K, values for pnitrophenyl acetate andp-nitrophenyl propionate are also larger than those in the absence of coenzyme (compare Tables I and III), and kcat values for these two substrates are also larger; k,,, values and k,,,/K, ratios also show a minimum at p-nitrophenyl butyrate. Both K, and kcat values in the presence of NADH are smaller than those in the presence of NAD (compare Tables II and III) and are more similar to those in the absence of coenzyme. Purification and retesting of p-nitrophenyl butyrate. Since p-nitrophenyl butyrate was an unusually poor substrate, the possibility was considered that it was contaminated with some unknown inhibitor. For this reason the compound was subjected to thin-layer chromatography as described under Materials and Methods. In both solvent systems a single ultraviolet absorbing spot was observed. The ultraviolet absorbing band was eluted from silica plates following thin-layer chromatography in solvent system I, dissolved in ethanol, filtered, and used directly for kinetic constant determinations. The values for K, (33 PM), kcat (22 mini’), and kcat/Km (0.7) were obtained in the absence of coenzyme. These values are similar to the values obtained with commercially supplied p-nitrophenyl butyrate (compare with Table I).
us (&)
0.0
0.2
0.4 us ()IM-1)
0.6
plots in the FIG. 1. Effect of ester chain length on LineweaverBurk absence of coenzyme and in the presence of NAD or NADH. A, p-nitrophenyl acetate; B, p-nitrophenyl propionate; C, p-nitrophenyl hutyrate; D, p-nitrophenyl valerate; E, p-nitrophenyl caproate. In the absence of coenzyme (O), in the presence of NAD (1 mM) (O), and in the presence of NADH (160 pM) (A). V = turnover number.
nitrophenyl acetate and p-nitrophenyl propionate are much larger than those in the absence of coenzymes. The kcat values decrease from p-nitrophenyl acetate to p-nitrophenyl butyrate to a minimum at p-nitrophenyl butyrate and then increase again as do k,,,/K,,, ratios listed in the last column of Table II. Kinetic constants determined in the presence of NADH (160 PM) are listed in
TABLE
Substrate
Specificity
The effect of coenzymes on kinetic constants of esters. Comparison of data listed in Tables I-III shows that NAD alters K,,, values for esters to a greater extent than NADH. The effect of coenzymes on k,,, values is also more pronounced with NAD than NADH. With acetate and propionate NAD increases kcat values, as compared with those in the absence of coenzyme (Table I), it lowers the kcatwith butyrate, but appears to have no effect on k,,, values with valerate and caproate. The k,,, values of acetate and propionate are also increased by NADH but to a lesser extent than by NAD (see Table III). Although both coenzymes affect kcat values of esters, usually an increase in kcatis accompanied with the increase in K,,,. Thus, k,,,/K, ratios remain almost unchanged in the presence and absence of coenzymes (Tables I-III). Under all three conditions, p-nitrophenyl hutyrate has
I
for E&erase Reaction in the Absence of Coenzyme k,., (mini)
Km (PM) p-Nitrophenyl ester of Acetate Propionate Rutyrate Valerate Caproate”
Expt. 1
Expt. 2
Average
Expt. 1
Expt. 2
Average
LtlKm (mini PM-')
2.1 0.9 30 4 9
2.7 2.5 59 8
2.4 1.7 44.5 6 9
15 10 29 10 49
20 14 24 18
17.5 12 26.5 14 49
7.3 7.1 0.6 2.3 5.4
Note. The assay system contained 50 a Single determination of K,,, and k,,,.
mM
sodium phosphate buffer, pH 7.0, 1 mM EDTA,
and 284
mM
ethanol.
26
MUKERJEE
AND TABLE
PIETRUSZKO II
Substrate Specificity for Esterase Reaction in the Presence of 1 mM NAD k,,, (min-‘) p-Nitrophenyl ester of Acetate Propionate Butyrate Valerate Caproate”
Expt. 1
Expt. 2
Average
Expt. 1
Expt. 2
Average
13 21 46 15 17
6 19 72 16
9.5 20 59 15.5 17
86 58 12 14 42
102 78 8 24
94 68 10 19 42
Note. The assay system contained 50 ’ Single determination of K,,, and k,,,
sodium phosphate buffer, pH 7.0, 1 mM EDTA,
mM
the lowest h,,,/K, ratio; this is most apparent especially in the presence of NAD. p-Nitrophenyl acetate under all three conditions has the highest k,,,/K,,, ratio. Kinetic constants of aldehydes. K,,, and kcat values for saturated aldehydes of one to eight carbon atoms are shown in Table IV. Except for formaldehyde, for which the K, value is high, all aldehydes have similar and very low (less than micromolar) K, values; kcatvalues are also quite similar for all aldehydes. The k,,,/K, ratio for formaldehyde is low, indicating that formaldehyde is the poorest substrate of all aldehydes employed in this set; this ratio increases with the increase of aldehyde chain length. Comparison of the dehydrogenase with the esterase reaction. The results presented in Tables I-IV demonstrate that kcat values for esters (lo-94 min-‘) are the same or higher than those for aldehydes (14-49 min-l). p-Nitrophenyl acetate and propionate in the presence of NAD (Table II) have kcatvalues considerably higher than aldehydes of the same chain length. The k,,, value for pnitrophenyl acetate in the presence of NAD is 3.9 times that for acetaldehyde, while that for p-nitrophenyl propionate in the presence of NAD is 4.7 times that for propionaldehyde (compare Tables II and IV). The k,,, value for p-nitrophenyl butyrate in the presence of NAD is the
TABLE
km/Km (min-’ pM 9.9 3.4 0.17 1.2 2.5
1 mM NAD, and 284 mM ethanol.
lowest of all k,,, values. Under all other conditions (compare Tables I-III with data on Table IV) the kcat values of these two reactions are similar. Despite similar k,,, values, since aldehydes have much smaller K, values than esters, they are better substrates. This is illustrated in Fig. 2 where k,,JK, ratios are plotted vs aldehyde or ester chain length (please note that the scale for aldehydes is an order of magnitude larger than that for esters). The k,,,/K, ratios for aldehydes increase rapidly with the increase of aldehyde chain length to butyraldehyde and then remain constant or even decrease slightly at hexanaldehyde (Fig. 2). To find out if aldehyde dehydrogenation followed a bell-shaped curve, octanaldehyde (Table IV) was employed as substrate. The k,,J K,,, ratio for octanaldehyde was almost twice as high as that for butyraldehyde, suggesting that aldehyde substrates continue improving with the increase of chain length. The ratios obtained with esters in the absence of coenzymes and in the presence of NAD (1 mM) and NADH (160 PM) are plotted for comparison. The dependence of esterase reaction upon the chain length of the acyl group of esters is different from that of aldehydes; it shows a most definite minimum atp-nitrophenyl butyrate. With the exception of p-nitrophenyl acetate, the lowest
III
Substrate Specificity for Esterase Reaction in the Presence of 160 FM NADH k,,, (min-‘)
Km(FM) p-Nitrophenyl ester of Acetate Propionate Butyrate Valerate Caproate”
Expt. 1
Expt. 2
Average
Expt. 1
Expt. 2
Average
4.5 2 27 8.5 8
3.6 4.4 40 8.5
4.1 3.2 33.5 8.5 8
31 16 25 17 52
39 21 15 29
35 22 20 23 52
Note. The assay system contained 50 ’ Single determination of K, and k,,,.
mM
‘)
sodium phosphate buffer, pH 7.0, 1
mM
EDTA,
160
pM
NADH,
and 284
mM
k,atlKn (min-’ PM-‘) 8.5 6.9 0.6 2.7 6.5 ethanol.
HUMAN
MITOCHONDRIAL
ALDEHYDE
h,,,/K, ratios for esters were in the presence of NAD and the highest were in the presence of NADH; the h,,,/K, ratios in the absence of coenzyme resembled those in the presence of NADH more than those in the presence of NAD (Fig. 2). Differences between h,,JK, ratios under these three conditions were, however, small (see Fig. 2 and Tables I, II, and III). Substrate inhibition Substrate and product inhibition. of the esterase reaction was not observed for any of the ester substrates within the concentration range used for K,,, determination (see Fig. 1). With p-nitrophenyl butyrate (46.5 PM) as substrate p-nitrophenol (125 PM) inhibited the esterase reaction to about 80% of control activity. p-Nitrophenol also inhibited the dehydrogenase reaction when formaldehyde was used as a substrate. Binding ofp-nitrophenol could not be demonstrated with propanal (Km = less than 1 PM), which saturates the enzyme at low micromolar concentrations. For this reason formaldehyde or glycolaldehyde (both poor substrates) was employed in these experiments. With formaldehyde (620 PM) as substrate the dehydrogenase reaction at 60 and 120 PM p-nitrophenol was 47 and 17% of control activity, respectively. Under the same conditions sodium acetate (50 mM) had no effect on the dehydrogenase reaction. When p-nitrophenyl butyrate (56 PM) was used as a substrate to monitor the esterase reaction, glycolaldehyde (2.5 mM) caused a decrease of enzyme activity to 43% of control. When glycolaldehyde (50 PM) was used as a substrate for the dehydrogenase reaction, ethyleneglycol monoacetate (16 mM, used here for its high solubility) caused a decrease of enzyme activity to 72% of control. Thus, dehydrogenase substrates inhibit the esterase reaction and ester substrates inhibit the dehydrogenase reaction. acetate E&erase burst. The substrates p-nitrophenyl andp-nitrophenyl guanidinobenzoate (K,,, = 40 PM; k,,, = 0.003 pmol/min/mg) chosen to study esterase burst differed significantly with respect to coenzyme effect; while hydrolysis of p-nitrophenyl acetate was stimulated by co-
TABLE
27
DEHYDROGENASE 200
12
150 8
E 0 z w
100 4
2 -0 2. fi u z
50
0
0 12345678
Chain Length FIG. 2. Comparison of chain length dependence of k,,,/K,,, values of aldehydes and p-nitrophenyl esters. Aldehydes (0). p-Nitrophenyl esters in the absence of coenzyme ( ), in the presence of 1 mM NAD (a), and in the presence of 160 pM NADH (0).
enzymes, both coenzymes had no effect on p-nitrophenyl guanidinobenzoate. The measurement of esterase burst was attempted at 400 nm employing a Varian 635 double beam spectrophotometer at a 1:lO scale expansion. Sodium phosphate, 50 mM, pH 7.0, containing 1 mM EDTA, kNAD (500 PM) or NADH (160 PM) at 25 or 5°C was used, employing an extinction coefficient of 9.8 mMpl cm -’ for the nitrophenolate anion. The reaction was started by addition of E2 isozyme (4.6-4.9 PM) to the above buffer containing the ester added to 1 ml of buffer in 0.015 ml ethanol, and also to the control containing buffer and ethanol. The burst amplitude was measured by subtracting absorbance of the control from that obtained by extrapolation of steady state velocity to 0 time. Correction was also made for any change of absorbance due to addition of enzyme. With p-nitrophenyl guanidinobenzoate (75 pM) as a substrate, the presence or absence of burst could be established at 25°C because its steady state velocity was low. With p-nitrophenyl acetate (150 PM), however, the steady
IV
Substrate Specificity for Straight Chain Saturated Aldehydes k,,, bin ‘1
Km(PM) Aldehyde Formaldehyde Acetaldehyde Propionaldehyde Butyraldehyde Valeraldehyde Hexanaldehyde Octanaldehyde
Expt. 1
Expt. 2
Average
Expt. 1
665 0.56 0.23 0.27 0.22 0.26 0.27
0.56 0.21 0.32 0.36 0.32 0.34
0.56 0.22 0.30
22 15 21 24 22 50
Note The assay system contained
Expt. 2
Average
26 14
24 14.5 27.5 26.5 23
29
50
mM
0.29 0.29 0.31
sodium phosphate buffer, pH 7.0, 1
mM
28 29 24 48
EDTA,
0.5
mM
NAD, and 284
0.04 43 66
92 91 79 158
49 mM
ethanol
28
MUKERJEE
AND
state velocity was high. Incorporation of chloral (50 or 200 mM) into the reaction system was attempted. In some experiments the temperature was lowered to 5°C to eliminate use of chloral with p-nitrophenyl acetate. With both substrates, in the absence of coenzymes, no burst of p-nitrophenolate anion was observed. In the presence of coenzymes, p-nitrophenyl guanidinobenzoate also exhibited no burst. In the presence of NAD p-nitrophenyl acetate appeared to exhibit a burst, but even at 5”C, or in the presence of chloral, the steady-state velocity was too fast to permit accurate extrapolation and calculation of the burst amplitude.
PIETRUSZKO
p-nitrophenyl acetate and propionate as substrates is, however, compensated by the corresponding increase of Km. Thus, the k,,,/K,,, ratios remain almost unchanged under all three conditions (compare Tables I, II, and III) showing that stimulation of reaction velocity is achieved by some other mechanism than improvement of substrate efficiency. Only p-nitrophenyl acetate efficiency as substrate may be somewhat improved by coenzymes as demonstrated by the slight increase of its k,,,/K,,, ratio [from 7.3-9.9 (NAD)-8.5 (NADH)]. When the acyl chain length of the ester is increased coenzyme stimulation disappears (see data for NADH) or the presence of coenzyme inhibits the esterase reaction (see data for NAD with esters of acyl groups of four to six carbon atom chain length). Even DISCUSSION with p-nitrophenyl propionate, the k,,/K,,, ratio decreases The most striking result of this investigation is the fact in the presence of NAD to 3.4 (compare Tables I and II), that the substrate specificity profile of aldehyde dehydrosuggesting that inhibition by NAD is already starting at genation is totally distinct from that of ester hydrolysis. p-nitrophenyl propionate. There was also no stimulation Comparison of k&K, ratios (Fig. 2) shows that catalytic or inhibition with a more bulky substrate, p-nitrophenyl efficiency of aldehyde substrates increases with the inguanidinobenzoate. Inhibition of esterase reaction by cocrease of chain length of aldehydes while the catalytic enzymes was never previously reported. Since inhibition efficiency of ester substrates decreases with the chain appears to occur only at longer acyl group chain length, length to a minimum at p-nitrophenyl butyrate and then it appears likely that coenzyme and ester binding sites increases again, forming a V-shaped curve. This suggests may overlap, this overlap becoming apparent with larger molecules. that aldehydes and esters of similar structure are bound Considerable changes in kinetic constants (both K,,, and differently to the enzyme. Both reactions could occur at the same active site with k catvalues) of all ester substrates occur following binding of NAD and NADH. The enzyme form with bound NAD binding of esters occurring in a different orientation from that of aldehydes. This is consistent with the fact that p- is not as good as the other two enzyme forms for ester nitrophenol also binds to aldehyde dehydrogenase and hydrolysis as demonstrated by consistently lower k,,,/K, acetate appears to be inhibits both esterase and dehydrogenase reactions (see ratios (Fig. 2). While p-nitrophenyl the best ester substrate of all esters used during this inabove). Therefore, binding of p-nitrophenyl esters could vestigation, p-nitrophenyl butyrate is the worst substrate occur both via their acyl groups and through p-nitrophenol. The fact that chloral (an aldehyde substrate an- for all three enzyme forms. This suggests that the active alogue) also inhibits both dehydrogenase and esterase re- site topography of all enzyme forms is such that it fits pactions (1, 3) is also consistent with different binding of nitrophenyl butyrate in a uniquely unfavorable way. In butyrate is the worst aldehydes and esters at the same active site. It is also terms of k,,,/K, valuep-nitrophenyl consistent with the conclusions of Senior and Tsai (5) substrate of all for the enzyme form that binds NAD and is primarily involved in dehydrogenation of aldehydes. who studied the esterase reaction of a related enzyme. This fact is, however, not reflected in Km or kcat values However, the possibility that aldehyde dehydrogenation and ester hydrolysis may occur at different active sites for butyraldehyde, which is a good substrate (Table IV) and is another indication of the binding sites for aldehydes cannot be excluded by any of the above experiments. Thus, the distinct substrate specificity profile for the de- and esters being distinct. There is at present no feasible explanation as to why p-nitrophenyl butyrate is the worst hydrogenase and esterase reactions (Fig. 2) could be insubstrate. At first, the presence of some inhibitory imterpreted in terms of the same catalytic site or in terms purity in butyrate was suspected. For that reason p-niof distinct catalytic sites as proposed by other investitrophenyl butyrate was repurified and its Michaelis congators (20-24). Coenzyme stimulation of the esterase reaction was also stant redetermined. The results, however, showed that there was no impurity in p-nitrophenyl butyrate which previously described (1, 3) but only with acetate and procould possibly account for these results. pionate as substrates. The coenzyme stimulating effect is Substrate specificity for p-nitrophenyl esters of inalso apparent in this work when p-nitrophenyl acetate creasing chain length was previously studied employing andp-nitrophenyl propionate are used as substrates. NAD proteolytic enzyme from Bacillus subtilis (25, increases the velocity of hydrolysis of p-nitrophenyl ac- subtilisin-a 26). The chain length profile employing k,,+/K, ratio also etate and p-nitrophenyl propionate about five-fold while the increase is only two-fold with NADH (compare kcat showed a bell-shaped curve with a maximum at p-nitrophenyl butyrate; chain lengths longer than butyrate provalues in Tables I-III). The increase of k,,, values with
HUMAN
MITOCHONDRIAL
ALDEHYDE
duced reduction in k,,,/K,,, value. Thus, the bell-shaped curve of its k,,,/K, chain length profile aldehyde dehydrogenase resembles subtilisin, except that the profile shows a minimum rather than the maximum at p-nitrophenyl butyrate. Substrate specificity with ester substrates of increasing chain length was also recently described for a high K,,, aldehyde dehydrogenase from rat liver mitochondria (5, 6). In this case the substrate specificity profile also showed a maximum at p-nitrophenyl butyrate demonstrating that p-nitrophenyl butyrate was the best substrate. The magnitudes of the k,,, values of the esterase reaction (IO-94 mini’) relative to those of dehydrogenase reaction (14-49 mini’) are at variance with those previously reported by Feldman and Weiner (1) who, employing horse mitochondrial aldehyde dehydrogenase, reported that the esterase reaction velocity was three times less than that of dehydrogenase reaction. These results also disagree with those of Sidhu and Blair (3) who employed the same enzyme as the one used during this investigation for their kinetic study and also reported lower rates for the esterase than for the dehydrogenase reaction. The reason why this occurred is not clear, unless both authors compared the dehydrogenase reaction at pH 9.0 (where it is 5 times faster) with esterase at pH 7.0. During this investigation extreme care was taken to employ conditions for the dehydrogenase reaction resembling as closely as possible those employed for the esterase reaction. The results have also been checked several times to assure that calculations were correct. Examination of kcatvalues (Tables I-IV) shows that esterase velocity of the reaction catalyzed by the E2 isozyme is in some cases considerably higher than that of dehydrogenase reaction. This is especially apparent with p-nitrophenyl acetate and propionate in the presence of NAD (Table II) where the esterase velocity is 3.9-4.7 times faster than the dehydrogenase velocity. These results suggested that coenzymes may be involved in alteration of the rate-limiting step of the esterase reaction. For this reason detection of burst with p-nitrophenyl acetate was attempted. However, due to the inadequacy of the available instrumentation this could not be resolved at this time. Our most recent results (27) demonstrate that both aldehyde and ester substrates form a covalent intermediate with the same cysteine residue (cysteine 302) of human cytoplasmic El and human mitochondrial E2 isozymes. Aldehyde dehydrogenase, however, is an enzyme which displays extremely pronounced “half-of-the-sites” reactivity, more pronounced than that seen in glyceraldehyde3-phosphate dehydrogenase (28), with only two active sites located on an enzyme composed of four structurally identical subunits (19). This makes it likely that cysteine 302 residues located on some subunits mav catalvze al-
DEHYDROGENASE
29
dehyde dehydrogenation while other cysteine 302 residues may be concerned with ester hydrolysis. However, more information is necessary, possibly from X-ray crystallography, before any definite conclusions about the sites of aldehyde dehydrogenation and ester hydrolysis can be made. REFERENCES 1. Feldman, R. I., and Weiner, H. (1972) J. Biol. Chen. 247, 267272. 2. Bodley, F. H., and Blair, A. H. (1971) Can. J. Biochem. 49, l-5. 3. Sidhu, R. S., and Blair, A. H. (1975) J. Biol. Chem. 250, 78947898. 4. Pietruszko, R. (1989) in Biochemistry and Physiology of Substance Abuse (Watson, R., Ed.), pp. 89-127, CRC Press,Boca Raton, FL. 5. Senior, D. J., and Tsai, C. S. (1990) Biochem. Cell Biol. 68, 758763. 6. Senior, D. J., and Tsai, C. S. (1988) Arch. Biochem. Biophys. 262, 211-220. 7. Johansson, B. (1989) Pharmacol. Z’oxicol. 64, 471-474. 8. Mukerjee, N., Blatter, E. E., and Pietruszko, R. (1992) FASBB J. 6, A333. [Abstract 19171 9. Kurata, N., Kemper, R., Hurst, H. E., and Waddell, W. J. (1990) Drug Metab. Dispos. 18, 504-507. 10. Parilla, R., Okhawa, K., Lindros, K. O., Zimmerman, U.-I. P., Kobayashi, K., and Williamson, J. R. (1974) J. Biol. Chem. 249,49264933. 11. Hempel, J. D., Reed, D. M., and Pietruszko, R. (1982) Alcoholism: Clin. Exp. Res. 6, 417-425. 12. MacKerell, A. D., Jr., MacWright, R. S., and Pietruszko, R. (1986) Biochemistry 25, 5182-5189. 13. Greenfield, N. J., and Pietruszko, R. (1977) Biochim. Biophys. Acta 483,35-45. 14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 15. Lineweaver, H., and Burk, D. (1934) J. Am. Chem. Sot. 56, 65% 667. 16. Wilkinson, G. N. (1961) Biochem. J. 80, 324-332. 17. Yun, S.-L., and Suelter, C. H. (1977) Biochim. Biophys. Acta 480, l-13. 18. Kezdy, F. J., and Bender, M. L. (1962) Biochemistry 1, 109771106. 19. Ambroziak, W., Kosley, L. L., and Pietruszko, R. (1989) Biochemistry 28,5367-5373. 20. MacGibbon, A. K. H., Haylock, S. J., Buckley, P. D., and Blackwell, L. F. (1978) Biochem. J. 171, 533-538. 21. Motion, R. L., Blackwell, L. F., and Buckley, P. D. (1984) Biochemistry 23, 6851-6857. 22. Deady, L. W., Buckley, P. D., Bennett, A. F., and Blackwell, L. F. (1985) Arch. Biochem. Biophys. 243, 586-597. 23. Tu, G.-C., and Weiner, H. (1988) J. Biol. Chem. 263, 1212-1217. 24. Tu, G.-C., and Weiner, H. (1988) J. Biol. Chem. 263, 1218-1222. 25. Philipp, M., Tsai, I.-H., and Bender, M. L. (1979) Biochemistry 18,
3769-3773. 26. Graae, J. (1954) Acta Chem. &and. 8, 356-357. 27. Blatter, E. E., Abriola, D. P., and Pietruszko, R. (1992) Biochem. J. 282, 353-360. 28. Harris, J. I., and Waters, M. (1976) The Enzymes (Bayer, P. D., Ed.), Vol. 13, 3rd ed., pp. l-40, Academic Press, New York.
