Article pubs.acs.org/crt
Cite This: Chem. Res. Toxicol. 2018, 31, 556−569
Modeling of Human Hepatic and Gastrointestinal Ethanol Metabolism with Kinetic-Mechanism-Based Full-Rate Equations of the Component Alcohol Dehydrogenase Isozymes and Allozymes Yu-Chou Chi,† Shou-Lun Lee,‡ Yung-Ping Lee,† Ching-Long Lai,§ and Shih-Jiun Yin*,† †
Department of Biochemistry, National Defense Medical Center, 161 Minchuan East Road Section 6, Taipei 11490, Taiwan Department of Biological Science and Technology, China Medical University, 91 Hsueh-Shih Road, Taichung 40402, Taiwan § Department of Nursing, Chang Gung University of Science and Technology, 261 Wenhwa First Road, Taoyuan City 33303, Taiwan Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on July 17, 2018 at 06:58:00 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
‡
S Supporting Information *
ABSTRACT: Alcohol dehydrogenase (ADH) is the principal enzyme responsible for the metabolism of ethanol. Human ADH constitutes a complex family of isozymes and allozymes with striking variation in kinetic properties and tissue distribution. The liver and the gastrointestinal tract are the major sites for first-pass metabolism (FPM). The quantitative contributions of ADH isozymes and ethnically distinct allozymes to cellular ethanol metabolism remain poorly understood. To address this issue, kinetic mechanism and the steady-state full-rate equations for recombinant human class I ADH1A, ADH1B (including allozymes ADH1B1, ADH1B2, and ADH1B3), ADH1C (including allozymes ADH1C1 and ADH1C2), class II ADH2, and class IV ADH4 were determined by initial velocity, product inhibition, and dead-end inhibition experiments in 0.1 M sodium phosphate at pH 7.5 and 25 °C. Models of the hepatic and gastrointestinal metabolisms of ethanol were constructed by linear combination of the numerical full-rate equations of the component isozymes and allozymes in target organs. The organ simulations indicate that in homozygous ADH1B*1/*1 livers, a representative genotype among ethnically distinct populations due to high prevalence of the allele, major contributors at 1 to 10 mM ethanol are ADH1B1 (45% to 24%) and the ADH1C allozymes (54% to 40%). The simulated activities at 1 to 50 mM ethanol for the gastrointestinal tract (total mucosae of ADH1C*1/*1−ADH4 stomach and the ADH1C*1/*1−ADH2 duodenum and jejunum) account for 0.68%−0.76% of that for the ADH1B*1/*1−ADH1C*1/*1 liver, suggesting gastrointestinal tract plays a relatively minor role in the human FPM of ethanol. Based on the flow-limited sinusoidal app perfusion model, the simulated hepatic Kapp m , Vmax, and Ci at a 95% clearance of ethanol for ADH1B*1/*1−ADH1C*1/*1 livers are compatible to that documented in hepatic vein catheterization and pharmacokinetic studies with humans that controlled for the genotypes. The model simulations suggest that slightly higher or similar ethanol elimination rates for ADH1B*2/*2 and ADH1B*3/*3 individuals compared with those for ADH1B*1/*1 individuals may result from higher hepatocellular acetaldehyde. immunochemical and kinetic features.4−6 Class I ADH contains multiple isozymes: ADH1A (previously denoted αα), ADH1B (ββ), and ADH1C (γγ). The class II to IV ADHs contain a single isozyme each: ADH2 (ππ), ADH3 (χχ), and ADH4 (μμ or σσ), respectively. ADH1B and ADH1C exhibit allozymes arising from allelic variations of the corresponding genes.6,7 ADH1B*1 (encoding β1 subunit polypeptide) and ADH1B*2 (encoding β2 subunit) are predominant among Caucasians and East Asians,
1. INTRODUCTION The pharmacological and toxicological effects of ethanol depend on the duration of exposure and the concentrations of ethanol and its metabolite acetaldehyde in body fluids and tissue. Alcohol dehydrogenase (ADH) is the principal enzyme responsible for ethanol oxidation, a major rate-determining step in mammalian hepatic metabolism of ethanol.1−3 Cytochrome P450 CYP2E1, which is inducible with ethanol, plays a secondary role in removal of ingested alcohol at high concentrations.1,2 The human ADH family has been divided into five classes on the basis of protein sequence, genomic organization, electrophoretic mobility, and © 2018 American Chemical Society
Received: January 5, 2018 Published: May 31, 2018 556
DOI: 10.1021/acs.chemrestox.8b00003 Chem. Res. Toxicol. 2018, 31, 556−569
Article
Chemical Research in Toxicology respectively; ADH1B*3 (encoding β3 subunit) occurs exclusively in people of African ancestry. ADH1C*1 (encoding γ1 subunit) and ADH1C*2 (encoding γ2 subunit) are approximately equally distributed among Caucasians and American Indians, but the former is highly prevalent among the East Asian and African populations. Currently, class V ADH is the only isozyme unknown of catalytic functions due to its extreme instability.8 Acetaldehyde, the immediate metabolite of ethanol, is oxidized by aldehyde dehydrogenase (ALDH) to acetate. ALDH1A1 and ALDH2 are major isozymes in human ALDH superfamily responsible for the metabolism of acetaldehyde in vivo.9,10 About 40% of East Asians have a genetic deficiency in ALDH2 resulting from dominant-negative effect of the variant ALDH2*2 allele.7,10,11 In contrast, ALDH2 deficiency appears very rarely in Caucasian, African, and American Indian populations. Human ADH exhibits tissue-specific distribution: all three class I isozymes and ADH2 are detected in the liver,12 ADH4 and ADH1C in the gastric mucosa,13ADH2 and ADH1C in the mucosae of small intestine,14 whereas ADH3 appears ubiquitous.6 Both ALDH1A1 and ALDH2 are detected in human livers and gastrointestinal mucosae.12−14 ADH is localized in cytosol6 and ALDH1A1 and ALDH2 in the cytosol and mitochondria, respectively.9 Liver functions as the main organ for human ethanol elimination.12 The stomach and small intestine are principally responsible for absorption of ingested ethanol.15 Both liver and gastrointestinal tract are potential sites for firstpass metabolism (FPM) of ingested ethanol that influences the bioavailability and intoxicating effects of alcohol.16,17 Unlike the vast majority of drugs and other xenobiotics, the elimination of ethanol is not proportional to its concentration in body fluids (that is, first-order kinetics but with a pseudolinear phase of zeroorder saturation kinetics). This can be attributed to the unique nature of ethanol such as complete water miscibility, freely diffusible across the cell membrane, and low acute toxicity as well as social-drinking blood ethanol (∼3 mM) saturating low-Km (Michaelis constant) ADH forms. This Michaelis−Menten-type pharmacokinetics is of special interest in FPM, which is defined as the presystemic elimination of ingested ethanol before reaching peripheral circulation. Recently, the inlet ethanol concentration at 95% clearance in liver has been proposed as an estimate for potential near-maximum of the human hepatic FPM in accordance with Michaelis−Menten elimination kinetics because the hyperbolic kinetics appears no abrupt inflection point for 100% substrate saturation.18,19 It is well-documented that allelic variants of ADH1B*2 and ALDH2*2 protect against the development of alcoholism.6,7,20 The reduction of risks for alcohol dependence by ALDH2*2 and ADH1B*2 in East Asians appeared to be additive.21,22 ADH1B*3 is found to be protective against alcohol dependence in African populations.23,24 ADH1C*1 did not exhibit an independent protecting effect that may be due to the linkage disequilibrium between ADH1C*1 and ADH1B*2.21,25 The correlation of pharmacokinetic and pharmacodynamics studies, following a low- to moderate-alcohol challenge, has suggested that differential accumulations of blood acetaldehyde concomitantly with the corresponding dysphoric subjective feelings and cardiovascular responses may play a key role in the partial protection of ALDH2*1/*2 heterozygotes and the near-full protection of ALDH2*2/*2 homozygotes from developing alcoholism in East Asians.26−28 Because ADH1B2 and ADH1B3 allozymes exhibited 30−40-fold higher maximum velocities than that of ADH1B1,29,30 it was inferred that protection against alcoholism by ADH1B*2 and ADH1B*3 alleles also might be due to an
increase in blood acetaldehyde after alcohol consumption. 23−25,31 Pharmacokinetic studies with homozygous ALDH2*1/*1 East Asians revealed, however, barely detectable blood acetaldehyde as well as similar ethanol elimination rates in the three genotypes of ADH1B*1/*1, ADH1B*1/*2, and ADH1B*2/*2 following the ingestion of a moderate alcohol.32,33 It was also described that Jewish subjects carrying ADH1B*2 alleles (mostly ADH1B*1/*2) and African Americans with ADH1B*3 alleles showed, at most, 10 to 15% higher ethanol elimination rates compared to the ADH1B*1/*1 genotype.34,35 Thus, large differences in kinetic properties of ADH1B allozymes appeared to be associated with a relatively small distinction of alcohol elimination rates in individuals carrying different ADH1B alleles, suggesting that other cellular factors may regulate the activity of ADH isozymes and allozymes in liver such as potential inhibition by products acetaldehyde and NADH in the metabolism of ethanol.36 To address such a complex question involving myriad isozymes and ethnic-related functional genetic polymorphisms in human liver and gastrointestinal tract, simulation studies based on organ modeling would be an approach of choice. Previous work in this laboratory attempted modeling ethanol metabolism and ethanol−drug interactions with a simple Michaelis−Menten equation (a half-reaction in the absence of products NADH and acetaldehyde) for human ADH isozymes.18,19 In the current study, we have systematically characterized the kinetic properties of human ADH1A, ADH1B1-3, ADH1C1-2, ADH2, and ADH4 under a unified experimental condition and then constructed models for hepatic and gastrointestinal mucosal metabolism of ethanol using composite kinetic-mechanism-based full-rate equations of the component isozymes and allozymes. The model simulations of liver with the ADH1B*1/*1 genotype, which represents the most-prevalent ADH1B allele in world populations, appear reasonably compatible with results of the human pharmacokinetic and tissue activity studies. The simulations suggest that acetaldehyde could reduce alcohol elimination rates to varied extents in livers with ADH1B*2/*2 and ADH1B*3/*3. Future work will need to verify the steady-state acetaldehyde levels at varied ethanol in human livers containing high-activity ADH1B allozymes.
2. EXPERIMENTAL PROCEDURES 2.1. Chemicals. Sodium phosphate monobasic, acrylamide, isobutyramide, 4-methylprazole, sodium dodecyl sulfate (SDS), oxidized and reduced forms of β-nicotinamide adenine dinucleotide (NAD+ and NADH), Coomassie brilliant blue, trifluoroethanol, and bovine serum albumin were purchased from Sigma−Aldrich (Saint Louis, MO). Ethanol, acetaldehyde, and Folin−Ciocalteu’s phenol reagent were obtained from Merck (Darmstadt, Germany). All chemicals were of analytical grade. 2.2. Expression and Purification of Human ADH. The expression of recombinant enzymes in Escherichia coli and purification to apparent homogeneity for human ADH1A, ADH1B1, ADH1B2 ADH1B3, ADH1C1, ADH1C2, ADH2, and ADH4 were carried out as described previously.18,37,38 All of the isolated recombinant enzyme forms exhibited a single Coomassie blue-staining protein band with molecular mass of 40 kDa on SDS−polyacrylamide gel electrophoresis. Protein concentration was determined by the Lowry method39 using bovine serum albumin as the standard. 2.3. Kinetic Analysis. Kinetic studies were performed in 0.1 M sodium phosphate at pH 7.5 and 25 °C. Initial velocities were determined with a Varian Cary 3E UV−visible spectrophotometer (Mulgrave, Victoria, Australia) by monitoring the formation or oxidation of NADH at 340 nm using absorption coefficient of 6.22 mM−1 cm−1 or a PerkinElmer LS 50B luminescence spectrometer (Waltham, MA) with excitation at 557
DOI: 10.1021/acs.chemrestox.8b00003 Chem. Res. Toxicol. 2018, 31, 556−569
Article
Chemical Research in Toxicology
where F is the total hepatic blood flow, and Ci and Co are the ethanol concentrations in the inlet and outlet, respectively. The total blood flow from portal vein and hepatic artery equals the blood flow in the hepatic app vein (∼1.5 L/min).44 The Kapp m andVmax, apparent overall Km and Vmax for liver ethanol elimination, can be assessed by using composite numerical formulations consisting of the component rate equations of hepatic ADH isozymes and allozymes (see the Results section). Ĉ is the logarithmic average of Ci and Co, defined as:
340 nm and emission at 460 nm for the fluorescence of NADH. Acetaldehyde was redistilled before use. Steady-state kinetic data were analyzed by nonlinear least-squares regression using the statistical programs of Cleland:40 HYPER (eq 1) and SEQUEN (eq 2) results for a single substrate (substrate or coenzyme) and two substrates (both substrate and coenzyme) varied, respectively, and those of COMP (eq 3), UNCOMP (eq 4), and NONCOMP (eq 5) for competitive, uncompetitive, and noncompetitive product and dead-end inhibitions, respectively, were as follows:
Ĉ =
V S v = max Km + S
(1)
VmaxAB v= K iaKb + K aB + KbA + AB
(2)
v=
VmaxS K m(1 + I /K is) + S
(3)
v=
VmaxS K m + S(1 + I /K ii)
(4)
VmaxS v= K m(1 + I /K is) + S(1 + I /K ii)
v=
app V max × 0.317C i app = V max − 3FK mapp K mapp + 0.317C i
(8)
The estimated Ci, Co, Ĉ , and v at 95% clearance of ethanol in human app liver can be calculated from simulated hepatic Vapp max andKm and hepatic blood flow (1.5 L/min) using the above equations.19
3. RESULTS 3.1. Kinetic Mechanism and Steady-State Rate Equation. Initial velocity studies of coenzyme and substrate covaried for forward reaction for ADH1A, ADH1B1−3, ADH1C1−2, ADH2, and ADH4 showed linear intersecting patterns, consistent with a sequential Bi mechanism and excluding the Ping Pong Bi mechanism. The order of coenzyme and substrate binding was investigated by product and dead-end inhibition studies, and the results are summarized in Table 1. The coenzymes NADH and NAD+ were mutually competitive, and the substrates acetaldehyde and ethanol were mutually noncompetitive for all of ADH isozymes and allozymes studied, suggesting an ordered Bi Bi mechanism. The product inhibition patterns of acetaldehyde against varied substrate ethanol for the eight isozymes and allozymes are illustrated in Figure S1, which can distinguish an ordered Bi Bi system from those of the Theorell−Chance ordered Bi Bi and of the rapid equilibrium ordered or random Bi Bi.50 In line with this conclusion, trifluoroethanol (TFE), an inactive substrate analog, was a competitive inhibitor against ethanol and an uncompetitive inhibitor against NAD+; similar inhibition types were detected for carbonyl substrate analog isobutyramide (IBA) against acetaldehyde and NADH. 4-Methylpyrazole (4MP), a dead-end inhibitor of ethanol, was a competitive inhibitor against ethanol for most class I isozymes and allozymes but a noncompetitive inhibitor against ethanol for ADH1B3 allozyme and class II and IV isozymes. The dead-end inhibition patterns of 4MP against ethanol for 8 ADH isozymes and allozymes are illustrated in Figure S2, and it appears to be the most-potent inhibitor investigated. The noncompetitive type could be interpreted as that 4MP may reversibly bind to both the enzyme−NAD+ and enzyme−NADH binary complexes in a catalytic cycle. Thus, simplest kinetic mechanism consistent with most of best-fit patterns in the initial velocity, product, and deadend inhibition studies for human ADH isozymes and allozymes appeared to be an ordered sequential Bi Bi mechanism (that is, the binding of NAD+ first and NADH released last or vice versa). The reaction scheme is illustrated in Figure S3. This mechanism is in agreement with previous studies of horse ADH45,46 as a prototype and some of the human ADH forms.29,47−49 Table 2 shows key kinetic constants obtained from the steadystate initial velocity and product inhibition studies, revealing wide variations in human ADH isozymes and allozymes. For instance, the Km for ethanol (Kb) varied 2500-fold (0.013 to 32 mM), with ADH1B1 being the lowest and ADH4 the highest, suggesting
(5)
app ̂ V max C
K mapp + Ĉ
(7)
At 95% clearance (Ci − Co = 0.95 Ci), the following equations can be derived:19
where Vmax is the maximum velocity, S is substrate or coenzyme concentration, Km is the Michaelis constant; I, and Kis and Kii in eqs 3−5 are the inhibitor concentration, slope inhibition and intercept inhibition constants, respectively. In eq 2, A and B are coenzyme and substrate concentrations for a sequential Bi reaction, respectively; Ka and Kb are the Michaelis constants; and Kia is the dissociation constant for coenzyme. The type of product and dead-end inhibitions was determined by evaluating standard errors of the parameters and residual variance for the equation that best fit the data.40 The kinetic experiments were run in duplicate with four substrate concentrations usually ranging from 0.3 to 3Km and four (including one for control, I = 0) inhibitor concentrations ranging from 0.5 to 2 Ki when applicable. Values are expressed as mean ± standard error (SE) of the mean. The error of the fitted value for Vmax is ≤5.3%; all other errors are ≤18% of the fitted values, indicating that the fits are good.40 2.4. Protein Content of ADH Isozymes in Human Tissues. To simulate the human metabolism of ethanol, it will require protein contents of ADH isozymes in target organs as well as the component isozyme kinetic constants in full-rate equations. The isozyme protein amount in normal portions of the surgical liver specimens and mucosae of the duodenum and jejunum were determined by immunoblot using the corresponding affinity-purified class-specific antibodies and the respective purified recombinant ADH isozymes as the standard.5,12,14 All patients provided written informed consent, and the studies were approved by the Institutional Review Board of the National Defense Medical Center.12−14 It is also noted that sex and age did not significantly influence hepatic and gastrointestinal mucosal ADH activities with specified genetic phenotypes.12−14 The ratio of the total subunit contents of class I ADH1A, ADH1B and ADH1C in livers was estimated from CM-cellulose chromatography of the isolated mixture of class I isozymes.18 The protein content of gastric ADH isozymes was estimated from the isozyme activities in normal portions of surgical gastric mucosae and the specific activity of the purified isozymes.13,18 The liver mass for a 70 kg man was estimated as 2% of body weight (1400 g).18 The estimates of total mucosal mass of human stomach, duodenum, and jejunum were 23, 34, and 149 g, respectively.41 The protein contents and expression pattern of ADH isozymes in human liver and gastrointestinal mucosae are calculated and summarized in Table S1. 2.5. Flow-Limited Sinusoidal Perfusion Model. The model of hepatic drug metabolism was described previously.42,43 The rate (v) of drug (ethanol) elimination can be formulated as:
v = F × (C i − Co) =
C i − Co ln C i − ln Co
(6) 558
DOI: 10.1021/acs.chemrestox.8b00003 Chem. Res. Toxicol. 2018, 31, 556−569
Article
Chemical Research in Toxicology Table 1. Product and Dead-End Inhibition Patterns of Human Alcohol Dehydrogenase Isozymes and Allozymesa class I inhibitor acetaldehyde NADH ethanol NAD+ 4-methylpyrazole trifluoroethanol isobutyramide
class II
class IV
varied substrate
ADH1A
ADH1B1
ADH1B2
ADH1B3
ADH1C1
ADH1C2
ADH2
ADH4
ethanol NAD+ ethanol NAD+ acetaldehyde NADH acetaldehyde NADH ethanol ethanol NAD+ acetaldehyde NADH
N N N C N U U C C C U C U
N N N C N U U C C C U C U
N N N C N U U C C C U C U
N N N C N N U C N C U C U
N nd nd C N nd nd C C C U C U
N nd nd C N nd nd C C C U C U
N N N C N N U C N nd nd nd N
N N N C N U U C N C U C U
Steady-state kinetic studies were performed in 0.1 M sodium phosphate at pH 7.5 and 25 °C. C, competitive inhibition; U, uncompetitive inhibition; N, noncompetitive inhibition; nd, not determined.
a
kfcat(krcat-to-kfcatratios, 5.8 to 44) with ADH1B3 the lowest ratio and ADH1B1 the highest ratio. The ratio of overall catalytic efficiencies, which is the Haldane expression for the equilibrium constant (Keq),50 agreed well (less than 25% deviations except 50% deviation for ADH1C1) (Table 2) to the experimentally determined value (10 pM),51 indicating that the kinetic constants were self-consistent. The R values52 showed 1.2% to 14% (Table 2) of the total enzyme for ADH isozymes and allozymes present in the central ternary complexes, suggesting that chemical conversion may not be rate-limited in catalysis. The calculated rate constant (k4) for release of NADH from the enzyme were slightly (19−36%) higher than the corresponding experimentally determined kfcat for ADH isozymes and allozymes, suggesting that NADH release may involve in rate-limiting for the forward reaction. This appears compatible with the R analyses. Table 3 shows inhibition constants for dead-end inhibitors from the steady-state kinetic studies. It is noted that slope inhibition constants (Kis), the dissociation constant for inhibitor and the enzyme−coenzyme binary complex, for 4MP were nearly a thousandth of those for the corresponding TFE and IBA for class I ADH isozymes and allozymes, indicating the striking potency of 4MP as an ADH inhibitor. The steady-state kinetic equation for ordered sequential Bi Bi mechanism is as follows:50
that human ADH forms can effectively contribute to metabolizing a large range of physiological ethanol concentrations in social drinking settings. The Km for acetaldehyde (Kp) varied 260-fold (0.092−24 mM), with ADH1B1 being the lowest and ADH4 the highest. The Km for NAD+ (Ka) appeared to have two groupings: high-Ka forms (0.14−0.84 mM) for ADH1B2, ADH1B3, and ADH4, and low-Ka forms (0.0048−0.011 mM) for the rest of isozymes and allozymes. The Kmfor NADH (Kq) showed a similar pattern to that of Ka for the corresponding ADH isozymes and allozymes, including the high-Kq forms (0.096− 0.18 mM), and the low-Kq forms (0.0046−0.033 mM). Inhibition constants for NAD+ (Kia) and NADH (Kiq), dissociation constants for the binary complexes of enzyme−NAD+ and enzymes−NADH, respectively, showed a similar trend to those of the Ka and Kq. Inhibition constants for ethanol (Kib) and acetaldehyde (Kip) varied 76-fold (12−910 mM) and 140-fold (0.16−23 mM), respectively. kcat, the turnover number, for both the forward (kfcat) and the reverse (krcat) reactions clearly exhibited two groupings: the high-kcat group (300−870 min−1 for kfcat and 2300−26000 min−1 for krcat) for ADH1B2, ADH1B3, and ADH4 and the low-kcat group (5.9−30 min−1 for kfcat and 260−700 min−1 for krcat) for the rest of the isozymes and allozymes. It is noted that krcat values are considerably greater than that of the corresponding
v=
f r f r Vmax Vmax AB − Vmax Vmax PQ /Keq r r r r f f Vmax K iaKb + Vmax KbA + Vmax K aB + Vmax AB + Vmax KqP /Keq + Vmax K pQ /Keq f f r r + Vmax PQ /Keq + Vmax KqAP /K iaKeq + Vmax K aBQ /K iq + Vmax ABP /K ip f + Vmax BPQ /K ibKeq
(9)
3.2. Formulation of Composite Rate Equations for Ethanol Metabolism. The maximum ethanol-oxidizing (Vfmax) and maximum acetaldehyde-reducing (Vrmax) velocities of ADH isozymes and allozymes in human liver and gastrointestinal tract are summarized in Table S2, which were calculated from the turnover numbers (Table 2) and the corresponding protein contents and expression pattern of ADH isozymes in target organs (Table S1). Based on two maximum velocities (Vfmax and Vrmax) and the experimentally determined steady-state parameters, including four Michaelis constants (Ka, Kb, Kp, and Kq) and
four product inhibition constants (Kia, Kib, Kip, and Kiq) (Table 2) as well as the reported Keq51 for the reaction, composite numerical rate equations were constructed according to the kinetic rate equation (eq 9) for the steady-state metabolism of ethanol with the component ADH isozymes and allozymes in human livers and gastrointestinal (stomach, duodenum, and jejunum) mucosae carrying different genotypes (Tables S3 and S4). The isozyme and allozyme numerical rate equations contain four variables: the concentrations of NAD+ (A), ethanol (B), acetaldehyde (P), and NADH (Q). The units for the variables and the 559
DOI: 10.1021/acs.chemrestox.8b00003 Chem. Res. Toxicol. 2018, 31, 556−569
ADH1B2 0.21 ± 0.02 0.64 ± 0.03 0.40 ± 0.04 0.096 ± 0.003 0.95 ± 0.03 18 ± 1 0.44 ± 0.08 0.0065 ± 0.0002 300 ± 10 5500 ± 100 310 ± 20 640 ± 40 7.5 370 0.028
ADH1B3 0.84 ± 0.02 27 ± 1 2.1 ± 0.1 0.17 ± 0.01 2.2 ± 0.3 220 ± 10 3.0 ± 0.5 0.038 ± 0.002 400 ± 10 2300 ± 100 12 ± 1 5.1 ± 0.1 7.5 510 0.012
ADH1C1 0.0073 ± 0.0005 0.26 ± 0.02 0.21 ± 0.03 0.033 ± 0.006 0.13 ± 0.01 12 ± 1 0.16 ± 0.02 0.0017 ± 0.0002 30 ± 1 700 ± 10 150 ± 20 3300 ± 400 15 36 0.14
ADH1C2 0.0048 ± 0.0003 0.17 ± 0.01 0.21 ± 0.02 0.0046 ± 0.0002 0.042 ± 0.002 14 ± 1 0.21 ± 0.02 0.000 27 ± 0.000 02 19 ± 1 420 ± 10 67 ± 10 1300 ± 200 12 25 0.14
0.011 ± 0.001 13 ± 1 19 ± 1 0.019 ± 0.001 0.094 ± 0.004 910 ± 110 5.6 ± 1 0.000 72 ± 0.000 04 7.9 ± 0.1 260 ± 10 1.0 ± 0.1 13 ± 1 11 9.9 0.12
ADH2
class II
0.14 ± 0.01 32 ± 1 24 ± 1 0.18 ± 0.01 0.62 ± 0.08 580 ± 50 23 ± 3 0.0077 ± 0.0004 870 ± 30 26 000 ± 1000 23 ± 2 37 ± 2 10 1100 0.025
ADH4
class IV
a Ka, Kb, Kp, and Kq are Michaelis constants for NAD+, ethanol, acetaldehyde, and NADH, respectively; Kia, Kib, Kip, and Kiq are the corresponding product inhibition constants, respectively. kfcat and krcatare catalytic constants for the forward (ethanol oxidation) and reverse (acetaldehyde reduction) reactions, respectively. kfcat/Kb and kfcat/KbKia are bimolecular and termolecular rate constants for the forward reaction, respectively. Keq, equilibrium constants. k4, rate constant for dissociation of the enzyme−NADH binary complex (refer to the reaction scheme in Figure S3). R, fraction of the enzyme in the ternary complex. Values are expressed as the mean ± SE. bCalculated from the HYPER program (eq 1). For isozymes and allozymes with high Km for substrates or coenzymes, the data were corrected for sub-saturating fixed cosubstrates. cCalculated from the COMP program (eq 3) or NONCOMP program (eq 5) according to the corresponding product inhibition patterns shown in Table 1. Note that inhibition constants of the products are those against the corresponding substrates (rather than cosubstrates) in the reverse reaction; for example, Kia refers to product inhibitor NAD+ vs the varied substrate NADH (not cosubstrate acetaldehyde), and in this case, a competitive inhibition was detected for all of the ADH isozymes and allozymes studied. dkcat (min−1), turnover number, calculated from experimentally determined Vmax (μmol min−1 mg−1) of ADH isozymes and allozymes using a subunit molecular mass 40 kDa, i.e., kcat = Vmax × 40 mg μmol−1. Vmax was obtained from the HYPER program (eq 1) as that for Km and corrected for the subsaturating fixed cosubstrate. For the conversion of Vmax of ADH isozymes and allozymes to the Vmax of ADH isozymes and allozymes per total organ or tissue, see the Supporting Information eCalculated from the SEQUEN program (eq 2). fCalculated from Keq = (VfmaxKpKiq[H+])/(VrmaxKbKia), where [H+] = 10−7.5 M. gCalculated from k4 = krcatKiq/Kq. h Calculated from R = [(1 − Ka/Kia)/Vfmax + (1 − Kq/Kiq)/Vrmax]/(1/Vfmax + 1/Vrmax).52
ADH1B1 0.011 ± 0.001 0.013 ± 0.001 0.092 ± 0.003 0.0067 ± 0.0005 0.076 ± 0.011 14 ± 2 0.24 ± 0.03 0.000 18 ± 0.000 02 5.9 ± 0.1 260 ± 10 420 ± 40 5500 ± 300 12 7.0 0.033
ADH1A
0.012 ± 0.001 4.7 ± 0.7 2.7 ± 0.2 0.0050 ± 0.0002 0.046 ± 0.008 110 ± 10 0.86 ± 0.11 0.000 41 ± 0.000 04 22 ± 1 370 ± 10 2.5 ± 0.3 56 ± 5 9.7 30 0.066
constant
Ka (mM)b Kb (mM)b Kp (mM)b Kq (mM)b Kia (mM)c Kib (mM)c Kip (mM)c Kiq (mM)c kfcat(min−1)d krcat(min−1)d kfcat/Kb (mM−1min−1)e kfcat/KbKia (mM−2 min−1)e Keq (pM)f k4(min−1)g Rh
class I
Table 2. Kinetic Constants for Human Alcohol Dehydrogenase Isozymes and Allozymesa
Chemical Research in Toxicology Article
560
DOI: 10.1021/acs.chemrestox.8b00003 Chem. Res. Toxicol. 2018, 31, 556−569
Article
Chemical Research in Toxicology Table 3. Dead-End Inhibition Constants for Human Alcohol Dehydrogenase Isozymes and Allozymesa class I inhibition constants
ADH1A
ADH1B1
Kis, 4MP (μM) Kis, TFE (mM)
1.0 ± 0.1 2.1 ± 0.2
0.38 ± 0.04 0.14 ± 0.01
Kis, IBA (mM)
1.9 ± 0.2
7.9 ± 0.5
ADH1B2
ADH1B3
ethanol oxidation 0.52 ± 0.02 3.3 ± 0.2 0.85 ± 0.05 4.4 ± 0.1 acetaldehyde reduction 3.9 ± 0.4 50 ± 3
ADH1C1
ADH1C2
class II
class IV
ADH2
ADH4
0.069 ± 0.006 0.013 ± 0.001
0.061 ± 0.004 0.020 ± 0.001
1050 ± 110 nd
270 ± 10 6.5 ± 0.3
0.067 ± 0.004
0.079 ± 0.005
nd
56 ± 5
a
4MP (4-methylpyrazole) and TFE (2,2,2-trifluoroethanol) are dead-end inhibitors vs ethanol and IBA (isobutyramide) vs the substrate acetaldehyde. For corresponding inhibition patterns, see Table 1. Fixed coenzyme concentrations used: 2.4 mM NAD+ for ethanol oxidation (except 1 mM NAD+ for ADH1A, 2 mM NAD+ for ADH1B3, and 0.5 mM NAD+ for ADH4 in TFE inhibition experiments) and 0.2 mM NADH for acetaldehyde reduction. Kis, slope inhibition constants obtained by fitting data to the COMP program (eq 3) and (in a few cases) by fitting to the NONCOMP program (eq 5). Values are expressed as the mean ± SE; nd, not determined.
and ADH1B*3/*3−ADH1C*1/*1 livers are 7.5- to 10-fold and 83% to 3.9-fold, respectively, of those for the ADH1B*1/*1− ADH1C*1/*1 liver at 1 to 33 mM ethanol. The ethanolmetabolizing activities of ADH1C*1/*1−ADH4 gastric mucosa, at 2.5 to 50 mM ethanol, are 20% to 190% higher than that for the 1 mM ethanol, whereas the activities of ADH1C*1/*1−ADH2 duodenal mucosa and jejunal mucosa at the same range of ethanol are only 10% to 20% higher than that for the 1 mM ethanol. Table 5 shows relative activities of the component isozymes and allozymes in liver and gastrointestinal tract with different genotypes. At 1 to 33 mM ethanol, ADH1C1 appears to be the biggest contributing form (54% to 45%, decreasing as ethanol increasing) in ADH1B*1/*1−ADH1C*1/*1 liver, second by ADH1B1 (38% to 21%). ADH1A (7.4% to 28%) and ADH2 (0.79% to 6.4%), both relative activities are increasing as ethanol increasing from 1 to 33 mM. In ADH1B*2/ *2−ADH1C*1/*1 liver, ADH1B2 allozyme is a predominant contributor (≥92%), whereas in the ADH1B*3/*3−ADH1C*1/ *1 liver, ADH1B3 allozyme contributes from 26% to 80% as ethanol increases from 1 to 33 mM and ADH1C1 allozyme contributes decreasingly from 64% to 11%. In stomach, the relative activities of ADH1C allozymes and ADH4 appear in converse contribution at 1 to 50 mM ethanol. A similar converse contribution is detected for the ADH1C allozymes and the ADH2 in duodenum and jejunum, but the relative activities of ADH2 is quite small (≤3.5%). 3.4. Kinetic Parameters for Hepatic 95% Clearance of Ethanol. Based on the flow-limited sinusoidal perfusion model, the apparent Km and Vmax for ethanol elimination in human livers with different genotypes are assessed by the corresponding composite formulations for modeling of the steady-state hepatic metabolism (Table 6). It should be noted that the estimated values may vary depending on the ethanol range and the distribution of concentration points chosen for assessments because of widely varied kinetic constants, such as Kb, for the component isozymes and allozymes in liver. The hepatic ethanol saturation profiles at 0.10 to 25 mM ethanol for 6 combinatorial genotypes of the ADH1B and ADH1C variant alleles are illustrated in Figure S4. In homozygous ADH1B*1/*1− app ADH1C*1/*1 liver, the simulated Kapp m and Vmax are 0.12 ± 0.01 mM and 2.8 ± 0.1 mmol/min per liver, respectively. At 95% clearance of ethanol, the inlet ethanol (Ci), outlet ethanol (Co), and logarithmic average of the sinusoidal ethanol (Ĉ ) are calculated to be 1.6, 0.079, and 0.50 mM, respectively, and the corresponding hepatic ethanol elimination rate (v) is 2.3 mmol/ min. The ADH1B*1/*1−ADH1C*2/*2 liver shows a slightly lower values for the corresponding calculated kinetic parameters. ADH1B*1/*1−ADH1C*1/*1 is a representative combinatorial
individual isozyme and allozyme activities (v) are expressed as milllimolar and mmol/min per organ or total mucosae, respectively. The numerical formulations for total organ and tissue ethanolmetabolizing activity are expressed as linear combination of the corresponding component isozyme and allozyme rate equations. 3.3. Model Simulations of Steady-State Metabolism of Ethanol. Free cytosolic NAD+ concentration in rat hepatocytes is ca. 0.5 mM.53 Because the lactate dehydrogenase-catalyzed reaction reaches near-equilibrium in rat liver, free cytosolic NADH can be calculated from equilibrium constants of the reaction and cellular concentrations of the lactate and pyruvate.54,55 In the absence of ethanol, the ratio of hepatic free cytosolic NAD/NADH in vivo is ca. 1000,56 which means that the concentration of NADH is ca. 0.5 μM. When ethanol is introduced, the NAD/NADH immediately drops.57 Steady-state ethanol metabolism studies in perfused rat livers revealed that effluent acetaldehyde increased as the influent ethanol raised, whereas NADH rapidly reached a plateau.3,58 For instance, at constant infusing of 10 mM ethanol, steady-state hepatocyte acetaldehyde and NADH (calculated from the corresponding effluent lactate and pyruvate ratios) reached ∼20 and ∼5 μM, respectively. The findings appeared comparable to the studies in healthy Europeans (assuming ADH1B1*1/*1 and ALDH2*1/*1 genotypes) with hepatic vein catheterization, which showed that 25 min following ingestion of 0.8 g/kg ethanol, hepatic venous acetaldehyde reached 22 μM or higher (cf.
