microemulsions (i.e. containing sodium dodecyl sulfate), the enzyme quickly lost ... trimethylammonium bromide), the enzyme activity was more stable and with ...
Eur. J. Biochem. 163,609-617 (1987) 8 FEBS 1987
Enzymes and microemulsions Activity and kinetic properties of liver alcohol dehydrogenase in ionic water-in-oil microemulsions Jean-Pierre SAMAMA, Kang Min LEE and Jean-Franqois BIELLMANN Laboratoire de Chimie Organique Biologique, associt au Centre National de la Recherche Scientifique, Institut de Chimie, Universitt Louis Pasteur, Strasbourg (Received September 30/November 12, 1986) - EJB 86 1039
The activity and the kinetic properties of horse liver alcohol dehydrogenase have been studied in water-in-oil microemulsions containing sodium dodecyl sulfate (SDS) or hexadecyl trimethylammonium bromide (CTAB), 1-butanol or I-pentanol or 1-hexanol or t-butanol, water and cyclohexane alone o r with octane. In the anionic microemulsions (i.e. containing sodium dodecyl sulfate), the enzyme quickly lost its activity, but was efficiently protected by the coenzyme and some adenine nucleotides. In the cationic microemulsions (i.e. containing hexadecyl trimethylammonium bromide), the enzyme activity was more stable and with higher alcohols was stable for at least 20 min. The Michaelis constant of NAD' calculated with respect to the water content was nearly constant and higher than in water. The maximum velocity in anionic microemulsions depends on the water content whereas in cationic microemulsions, the maximum velocity did not show a clear dependence on the water content and was close to the maximum velocity found in water. The pH dependence of K , and V,,, in these microemulsions was similar to that observed in water. The kinetic data for an hydrophobic substrate, cinnamyl alcohol, showed that this alcohol partitions between the pseudo-phases and thus the apparent Michaelis constant and the concentration at which substrate-excess inhibition appeared were increased. The catalytic properties of the enzyme in microemulsions were illustrated by the preparative reduction of cinnamaldehyde with cofactor recycling. The rate determination of NAD' reduction and of 1-butanol/cinnamaldehyde redox reaction showed that at low water content (2.8%), the NAD' reduction rate was close to zero whereas the redox reaction rate was about half of the rate at higher water content. Probably at low water content the coenzyme binding-dissociation rates are reduced much more than the binding-dissociation rates of the substrates and the rates of the ternary complex interconversion. The cationic microemulsions seemed to be a very favorable medium for enzyme activity, the tetraalkyl ammonium surfactant causing less denaturation than the anionic detergent dodecyl sulfate. The use of water-soluble enzymes for chemical synthesis suffers from several limitations. The most obvious are the low water solubility of many organic substances and the fact that the presence of water makes the reversal of some enzyme reactions for synthetic purposes difficult. In order to overcome these problems, much attention has been paid over the last few years to the solubilization of biological macromolecules in micellar solutions [I - 51. Most of the published work deals with three component systems: water, organic solvent and surfactants. Water-soluble enzymes like ribonuclease, peroxidase, a-chymotrypsin, trypsin, lactate dehydrogenase, pyruvate kinase, pyrophosphatase, lysozyme and alcohol dehydrogenase have been solubilized [l - 141 in swollen reverse micelles from water, octane or isooctane and di(2-ethylhexy1)sodium sulfosuccinate (AOT). In some instances, other surfactant molecules and hydrocarbon solvents were also tested [7,13,14]. The developments in this area have taken great advantage of the pre-existing knowledge in the general field of reverse micelles [I 5 - 181. The study of enzymes in three-component reverse micellar systems raised a number of problems [l] such as: the solubility Correspondence to J. F. Biellmann, Institut de Chimie, 1 Rue Blaise Pascal, F-67 008 Strasbourg Cedex, France Abbreviations. CTAB, hexadecyl trimethylammonium bromide; SDS, sodium dodecyl sulfate. Enzyme. Alcohol dehydrogenase (EC 1.1.1 .I).
and the stability of the enzyme, the minimum water content compatible with enzymic activity, the structure of the enzymecontaining micelles, the pH definition in the water core and the kinetics of the enzymic reaction in such media. Our aim was to investigate the behaviour of alcohol dehydrogenase in a four-component milieu corresponding to the definition of microemulsion according to Danielson and Lindman [18]. At first we studied the behaviour of the enzyme in anionic microemulsion : sodium dodecyl sulfate (SDS)/ 1-butanol/water/cyclohexane whose pseudo-ternary phase diagram is known [19]. A number of physico-chemical studies on this and related systems have been published [20 -251. As the work progressed, the SDS system was found to be inactivating the enzyme. Since cationic surfactants are in general less prone to denature proteins, we studied the activity in microemulsions made from a cationic surfactant: hexadecyl trimethyl ammonium bromide (CTAB) '. Much less physicochemical data are available on such systems than for microemulsions made from SDS. In buffer, enzyme activity was insensitive to CTAB at saturating concentration (10mM) of this salt. The nature of the polar head (tetraalkylammonium salt) should lead to weaker interaction with anionic groups on the enzyme surface than primary ammonium salts. Indeed we found that dodecyl ammonium was a denaturing agent for alcohol dehydrogenase but less efficient than SDS. But this may not be general and the denaturing effect of long-chain primary amines should be investigated.
610 Alcohol dehydrogenase from horse liver was chosen because its activity has been extensively studied in aqueous solution [26], in micellar solutions [8, 101 and in the crystalline state [27, 281. In colloidal solutions, the kinetic properties of alcohol dehydrogenase are altered [S]. The use of enzymes in microemulsions may lead to new technical applications and it also offers the possibility of studying aspects of enzyme mechanisms in heterogeneous media as well as the properties of microemulsions. For this we take advantage of the transparency of the microemulsions. To our knowledge, work in which a four-component micellar solution has been used as a medium for enzyme reaction has only appeared quite recently [29 - 311. We present here our results on horse liver alcohol dehydrogenase in microemulsions made of cyclohexane-buffered water/SDS or CTABjl -butanol or other primary alcohols or t-butanol.
The concentrations of the NAD' coenzyme and of the nucleotides were determined from the absorbance at 260 nm of the nucleotide-containing microemulsions assuming a molar absorption coefficient of 17600 M-' cm-' for NAD' and 13500 M - ' cm-' for adenine nucleotide. The kinetic data were plotted according to Lineweaver-Burk [32]. Enzyme stability in microemulsions Enzyme stability was determined for microemulsions Ab-2 to Ab-7, Cb-2 to Cb-8, Cp-1 and Ch-1. For each system, two microemulsions were prepared : the first contained buffer and enzyme (1 pl, 5 pg) and the second contained 1.3 mM NAD'. After incubation times ranging over 0.5 -20 min for the first microemulsion (175 pl), the second microemulsion (525 p1) was added to it for enzyme activity determination. Effect of nucleotides on the enzyme activity in microemulsions
MATERIALS AND METHODS General Alcohol dehydrogenase was purchased as a crystalline suspension from Boehringer. The crystals were centrifuged, dissolved in 50 mM Tes buffer pH 7.5,0.2 M sodium chloride. After centrifugation, the enzyme solution was extensively dialyzed against 50 mM Tes buffer pH 7.5 at 4°C. The enzyme concentration was determined spectrophotometricdlly at 280nm: a I-mg/ml solution has an absorbance of 0.455. NAD', NADH, AMP and dAMP were obtained from Boehringer at the highest purity available. The surfactants SDS (Sigma) and CTAB (Fluka) were recrystallized from absolute ethanol. The surfactant was dried for 24 h under vacuum at 60°C. Cyclohexane, octane, 1-butanol, 1-pentanol, I-hexanol and t-butanol were of the highest quality available. Cinnamyl alcohol was sublimed at 60°C under 0.1 mmHg (13 Pa). Cinnamaldehyde was distilled prior to use. All buffer solutions were made using doubly distilled water in a quartz apparatus. Buffer was 50 mM Tes pH 7.5 except when stated otherwise. All concentration values are computed, unless stated, with respect to the total volume. Microemulsions Microemulsions with the two surfactants and several cosurfactants were prepared. The microemulsions prepared with SDS are termed A and those with CTAB C, with alcohols as cosurfactant: 1-butanol b, t-butanol t, I-pentanol p and 1hexanol h. The compositions of the systems are given in Table 1. w, is defined as the ratio of the water molar concentration to the surfactant molar concentration. The microemulsions were prepared at 20 "C by adding the various constituents under constant stirring and were filtered prior to kinetic measurements. The buffer component was 50 mM Tes pH 7.5. Activity tests The experiments were done at 20°C. In control experiments conducted in water, the enzyme substrates were ethanol, I-butanol or cinnamyl alcohol. In microemulsion type b, 1butanol had two functions: cosurfactant and enzyme substrate. In microemulsion type t (t-butanol as cosurfactant), cinnamyl alcohol was used as the enzyme substrate.
For microemulsion Ab-3. The following additions were made to microemulsion Ab-3: (a) no nucleotide; (b) 9.8 mM AMP; (c) 8.4mM dAMP and (d) 1.3mM NAD'. To microemulsions a-c (350 pl) was added an enzyme solution (7 pl, 23 pg). After a given time (2, 5, 8, 14 and 20 min) microemulsion d (350 pl) was added and the activity recorded. For microemulsion At-2. Cinnamyl alcohol was used as the enzyme substrate at a final concentration of 0.43 mM. The following additions were made to microemulsion At-2: (a) no nucleotide; (b) 6.6 mM AMP; (c) 6.6 mM dAMP; (d) 1.1 mM NAD'. To microemulsions a-c (350 pl) an enzyme solution (7 pl, 23 pg) was added. After a given time (4, 8, 14 and 20 min) a 43 mM solution of cinnamyl alcohol in cyclohexane (7 pl) and microemulsion d (350 pl) were added and the activity was monitored at 340 nm through the appearance of NADH. For investigating the effect of NAD', an enzyme solution (7 pl, 23 pg) was added to microemulsion d (350 pl). After the same times, microemulsion a (350 pl) and a 43 mM solution of cinnamyl alcohol in cyclohexane (7 pl) were added and the activity recorded. For microemulsion Ct. Two microemulsions Ct were prepared (a) only buffer and (b) buffer containing NAD' (50 mM). To microemulsion a (500 pl) a 0.01 M solution of cinnamyl alcohol in cyclohexane (7 pl) and an enzyme solution (1 pl, 6 pg) were added. After a given time ( 5 , 10, 15, 20 min) microemulsion b (200 pl) was added and the activity recorded. The effect of NAD' was investigated in the following manner: to microemulsion a (500 pl) and microemulsion b (200 pl) an enzyme solution was added (1 pl, 6 pg). After a given time (5, 10, 15, 20 min) a 0.1 M solution of cinnamyl alcohol in cyclohexane (7 pl) was added and the activity recorded. Protective effect of various concentrations of AMP on the enzyme in At-2 microemulsions The following concentrations of AMP were added to microemulsion At-2: 0, 0.98 mM, 0.61 mM, 0.29 mM, 0.12 mM, 85 pM, 31 pM. Another microemulsion At-2 contained 1.1 mM NAD'. An enzyme solution ( 5 pl, 16 pg) and a 43 mM solution of cinnamyl alcohol in cyclohexane (7 pl) were added to all microemulsions At-2 (350 pl) except the last. After a given time (4,8,14 and 20 min) the last microemulsion At-2 (350 pl) was added and the appearance of NADH recorded. For each experiment, the activity was plotted versus the incubation time of the enzyme and the time taken for the
61 1 Table 1. Composition of the microemulsions Microemulsions Ab-1 to Ab-8 and Cb-I to Cb-5 are such that the mass ratio of cyclohexane to surfactant plus cosurfactant is constant. Octane and cyclohexane for microemulsions Cb-6, Cp and Ch were present in 1 :1 ratio in volume. Buffer was 50 m M Tes pH 7.5. wo = [water]i[surfactant] molar ratio Microemulsion
Mass of organic solvent (cyclohexane)
W"
water (buffer)
surfactant (SDS)
cosurfactant (I-butanol)
% total
molimol
Ab-1 Ab-2 Ab-3 Ab-4 Ab-5 Ab-6 Ab-7 Ab-8
62.3 59.3 58.2 56.8 55.5 54.1 52.5 34
5.2 10.0 11.5 13 14.8 17.2 20.2 34.3
10.8 10.2 10.1 10.0 9.9 9.6 9.1 10.6
21.7 20.5 20.2 20.1 19.8 19.1 28.2 21.1
At-1 At-2
59.5 55.5
9.4 14.9
10.4 9.9
20.7 19.8
(CTAB)
(I-butanol)
10.2 9.5 8.8 8.4 7.1
13.7 12.8 11.7 11.3 10.3
7.7 15.7 18.2 20.8 23.9 28.7 35.6 51 .8
(t-butanol)
Cb-1 Cb-2 Cb-3 Cb-4 Cb-5
2.8 8.3 13.9 21.0 26.7
73.3 69.4 65.6 59.3 55.3
14.5 24.1
5.5 17.6 33.4 50.4 70.1
(t-butanol) Ct
80.1
7.1
6.1
+
(octane cyclohexane) [l : 1, v/v] Cb-6
80.5
6.7
23.5
(1-butanol) 7.0
4.2
8.3
33.8
(1-pentanol) CP
80.5
7.0
4.2
8.3
33.8
(1-hexanol) Ch
80.5
7.0
enzyme to lose 50% of its activity in the absence of AMP was determined. The protective effect of AMP was then evaluated for each experiment by measuring the percentage of activity remaining after the same time. Effect of the surfactants on the enzyme activity in buffer A series of experiments was carried out to find the SDS concentration in buffered solution for which the decrease of enzyme activity was similar to that observed in type Ab microemulsions. To buffer/l2 mM ethanol (650 pl) containing various concentrations of SDS was added an enzyme solution (1 pl, 5 pg). After a time ranging over 0.5 - 5 min, an NAD' solution (50 pl, final concentration 1 mM) was added and the activity recorded. The protective effect of NAD+ and nucleotides against enzyme denaturation by SDS was then studied. In buffer/0.4mM SDS were added one of the following components (final concentration) : (a) 1 mM NAD' ; (b) 1 mM AMP; (c) 1 mM dAMP. The enzyme solution (1 pl, 5 pg) was added to the incubation medium (700 pl).
4.2
8.3
33.8
The procedure was identical to that above: the reaction was initiated for a by ethanol addition and for b and c by NAD' addition. The final concentrations were 16 mM ethanol and 1 mM NAD'. To buffer/3.2 mM I-butanol (600 pl) containing various concentrations of CTAB (0.1, 1, 5, 10mM) was added an enzyme solution (1 pl, 6 pg). After times ranging over 520 min, an NAD' solution (100 pl, final concentration 1.4 mM) was added and the activity recorded. Effect of the water content qf the microemulsions on the enzyme activity Ab-1 to Ab-8 microemulsions (700 pl) were used in these experiments. The NAD' concentration was 1.3- 5.1 mM. Enzyme solution (1 pl, 7.6 pg) was added and the activity recorded. In microemulsions Cb-I to Cb-3, the rate of I-butanol oxidation by NAD' and the exchange rate between I-butdnol and cinnamaldehyde was determined in conditions in which
612 the water pool concentration of NAD' was 0.1 mM and enzyme was added to the final concentration 35 pg/ml. For the oxidation rate, the absorption increase was followed at 340 nm. For the determination of the exchange rate, cinnamaldehyde was added to an overall concentration of 0.1 mM. The absorption decrease was followed at 280 nm after appropriate dilution.
Michaelis constant of NAD' and maximum velocity of NAD' reduction in microemulsions The Michaelis constant (K,) and maximum velocity ( V,,,) of NAD' were determined in five type-Ab microemulsions in which the water content was varied from 10% (w/w) (Ab-2, w, = 15.7) to 34.3% (w/w) (Ab-8, w, = 51.8). For each determination, two microemulsions were prepared: one contained buffer only and the other buffer containing 20 mM NAD' . Increasing amounts of the second microemulsion (20 - 100 pl, five points) were added to the first to give a final volume of 0.7 ml. The reaction was initiated by addition of the enzyme solution (1 pl, 7 pg). The Michaelis constant (K,) and the maximum velocity (V,,,) of NAD' were determined in four microemulsions Cb-2, Cb-3, Cb-4 and Cb-5 in which the water content varied from 8.3% (w/w) (Cb-2) to 26.7% (w/w) (Cb-5). For each determination, two microemulsions were prepared : one contained buffer only and the other buffer containing NAD ' (0.1 M). Increasing amounts of the latter (5-40 pl, four points) were added to the former to give a final volume of 700 pl. The reaction was initiated by addition of the enzyme solution (1 pl, 6 pg).
p H dependence of the Michaelis constant of NAD' and of the maximum velocity of NAD' reduction in microemulsions and in water The Michaelis constant (K,) and maximum velocity (V,,,) of NAD' were determined in microemulsions Ab-5 prepared with 50 mM buffers at pH 6.5 (phosphate), 7.0, 8.0 (Tris) and 9.0 (borate). The method for these determinations is described above. When conducted in water, the K , and V,,, determinations were made using 0.1 M and 10 mM 1-butanol as the substrate and NAD' concentrations of 0.11 - 1.4 mM. 50 mM buffers pH 6.0, 7.0 and 7.5 (phosphate), pH 8.5 (Tris) and pH 8.5 and 9.0 (borate) were used in the preparation of type Cb-3 microemulsions. For each pH value, two microemulsions were prepared : one contained buffer only and the other buffer containing NAD' (10 mM). Increasing amounts of the latter (25 -200 pl, four points) were added to the former to give a final volume of 700 pl. The reaction was initiated by addition of the enzyme solution (1 pl, 6 pg).
Michaelis constant of cinnamyl alcohol in microemulsions To a microemulsion At-I 1.3 mM NAD' was added a 161 mM or 14.6 mM or 1.46 mM solution ofcinnamyl alcohol in cyclohexane. The alcohol concentration was varied in order to screen a wide range of cinnamyl alcohol concentration in the assay medium. The catalytic reaction was initiated by the addition of an enzyme solution (3 pl, 14 pg) to a final volume of 0.8 ml and monitored through the formation of cinnamaldehyde at 283 nm. To microemulsion Ct 4 mM NAD' (100 pl) was added a 0.1 M solution of cinnamyl alcohol in cyclohexane (1 - 20 pl, seven points). The total volume was taken to 700 p1 and the
reaction was initiated by the addition of an enzyme solution (1 6 1.18). PI3
Preparative reduction of cinnamaldehyde In microemulsion Ab-2,9.4 mM cinnamaldehyde, 1.5 mM NAD' and 5 pM active site of enzyme were incubated. At given times, aliquots (1 pl) were diluted in microemulsion Ab-2 (700 pl) and the spectrum was recorded. After completion of the reaction (absorption spectrum stable) the organic constituants were isolated and analyzed by gas chromatography. Less than 5% cinnamaldehyde remained and cinnamyl alcohol was identified by comparison with an authentic sample. RESULTS
Enzyme activity in microemulsions The enzyme activity determination was based upon the appearance of the reduced coenzyme NADH. NADH in the microemulsions studied here exhibits absorption maxima at 259 nm and 340 nm. The molar absorption coefficients are identical to those in water and the absorption in the microemulsions studied here was stable for at least an hour. In all microemulsions containing a primary alcohol, acting as a substrate of alcohol dehydrogenase, enzyme activity was detected. For the cases where the enzyme substrate should not be available from the microemulsion, we used the system surfactant/t-butanollwaterlcyclohexane. In the microemulsions with t-butanol as cosurfactant, enzyme activity was detected in the presence of a substrate. We then studied the enzyme activity and determined the kinetic constants with alcohol dehydrogenase in microemulsions.
Enzyme stability in microemulsions, protection by nucleotides Alcohol dehydrogenase in solution is sensitive to SDS action [33]. But on incubation of the enzyme for several hours in buffer containing CTAB up to 10 mM, the enzyme activity did not change (results not shown). From these observations, we expect a greater stability of enzyme activity in cationic microemulsions than in anionic microemulsions. Indeed in microemulsions Ab-2 to Ab-7 and At-2, the enzyme activity decreased quite rapidly : in microemulsions Ab, the activity showed an exponential-like decrease with t I l 2in the range 1 2 min (Fig. 1) and in microemulsion At-2, the tli2 was about 5 min (Fig. 2). In buffered solution, a 0.4 mM concentration of SDS caused a similar time-dependent activity loss (Fig. 1). In cationic microemulsions, the enzyme was definitely more stable than in anionic microemulsions and depended on the water content. As the water content increased, the activity was lost more slowly and above 21% of water (Cb-4 and Cb-5), the activity remained constant over 20 min (Fig. 3). In the cationic microemulsion Ct made with t-butanol, the activity loss was similar to that observed in microemulsion Cb-2 (results not shown). In cationic microemulsions made with higher alcohols (Cp and Ch) as surfactants, the stability was higher than with 1-butanol (Fig. 3). In the cationic microemulsion Ct, the coenzyme NAD' had only a slight stabilizing effect (results not shown). In contrast, NAD and nucleotides protected alcohol dehydrogenase against SDS action in buffered solution, type Ab and type At microemulsions (Figs 1 and 2). Even at NAD' con+
61 3
5
50
x
t
c .-
.->
; ; 100 a
75 01 0
10
20
0
10 Time Irninl
Time Irnin)
Fig. 1. Alcohol dehydrogenase activity in type Ab-3 microemulsions and in buffered solutions containing S D S as afunction of time. Activity in Ab-3 (A);Ab-3/9.8 mM A M P ( 0 ) ; Ab-318.4 mMdAMP(0);activity in buffered solutions containing 0.4 mM SDS in presence of buffer alone (A);1 mM NAD+ (V);1 mM AMP ( 0 ) ;1 mM dAMP (m)
20
Fig. 3. Time dependence qf the stability of the enzyme activity for alcohol dehydrogenase in microemulsions. (A) ( 0 )Cb-2 (8% water); ( 0 ) Cb-3 (14% water); (U) Cb-4 (21% water); (A) Cb-5 (27% water); (B) ( 0 )Cb-6 with I-butanol; (0) Cp with I-pentanol; (0) Ch with I-hexanol. The buffer was 50 mM Tes pH 7.5. The activity was expressed relative to the initial activity (100%)
100
E
1
+ AMP
K
E-AMP (a)
Edenaturated
-
log 2
0
.-Cm
-"
x
(b)
log($)
i
.-0
s
[AMPI
50
-
..-> Q
0 0
10
Time lrninl
20
where A , is the initial activity and [AMP] the AMP concentration. The results fitted Eqn (b) and for K a value of 30 pM was determined. As the concentrations are expressed with respect to the total volume, the K, wp expressed with respect to the water content is 0.2 mM. This can be compared to the inhibition constant in buffer of 1.1 mM. dAMP did not show any inhibition at 6 mM concentration so that its Ki > 6 mM. However dAMP protected the enzyme from SDS inactivation in buffer and in type-At microemulsions, but not in Ab microemulsions.
Fig. 2. Alcohol dehydrogenase activity in type At-2 microemulsions as a function of incubation time. Concentrations in the microemulsion were: buffer alone ( A ) ; 6.6 mM AMP (0); 6.6 mM dAMP (0);Michaelis constant in microemulsions 1.1 mM NAD+ (V)
centrations large in respect to its dissociation constant, the protection was not complete in buffer and in microemulsions At, whereas with AMP a better protection was observed. AMP is a competitive inhibitor with respect to NAD' for this enzyme: Ki 1.1 mM. A quantitative study of the AMP protection of the enzyme activity was done in microemulsion At-2 at six different concentrations of AMP. If the following reaction scheme (a) applies, the activity A remaining at a defined ligand (AMP) concentration at the time where 50% of the activity is lost in absence of the ligand (AMP) is related to the ligand concentration by Eqn (b) :
Of
N A D + and maximum
The Michaelis constant of NAD' in type Ab and Cb microemulsions, presented as K, ov, is defined with respect to the total volume. Assuming that the coenzyme is present only in the water phase, a Michaelis constant computed with respect to the water volume, K, wp, could be defined [7,9] in the following general way: KmwpFw where F, is the water volume fraction. The K , ov varies with the water content. However the K, wp is nearly constant and is one order of magnitude larger than in buffer using 1-butanol as the enzyme substrate. In anionic microemulsions V,,, increased with w o(Table 2) and varied linearly as a function of w, (Table 3). For microemulsion Ab-1, the activity was small. For the cationic =
614 Table 2. Michaelis constant of NAD' andmaximum velocity of NAD' reduction with alcohol dehydrogenase in buffer (0.1 M I-hutranol) and microemulsions A b and Cb The Michaelis constant K,,, was expressed with respect to the total volume and K,,, wp expressed with respect to the buffer volume. The water content w, is expressed as the molar ratio of the water to the surfactant
1-butonol]
Kmwp
w,
V,,, nM min(mg enzyme)-'
molimol Ab-2 Ab-3 Ab-5 Ab-7 Ab-8
15.7 18.2 23.9 35.6 51.8
50 63 73 140 160
610 630 560 820 540
0.09 0.13 0.25 0.59 0.69
Cb-2 Cb-3 Cb-4 Cb-5
17.6 33.4 50.4 70.1
88 110 180 210
1200 910 990 920
0.78 0.8 0.88 0.89
50 200
2 1.2
Buffer 100 mM I-butanol 10 mM I-butanol
E-NAD*
(
Cinnamyl alcohol
)4
E-NADH
butonal
Dy
Microemulsion
(I
Cinnamaldehyde
Fig. 4. Redox reaction catalyzed by alcohol dehydrogenase
Table 4. N A D + reduction rate and I-butanollcinnamaldehyde exchange rate ar u function of the water content in cationic microemulsions Cb-1, Cb-2 and Cb-3 containing 0.1 mM NAD' and 35 pg enzyme/ml The reduction and exchange rate were determined at 340 nm and 280 nm respectively Microemulsion
Cb-I Cb-2 Cb-3
Water content
NAD+ reduction rate
I-butanol/ cinnamaldehyde exchange rate
% (wlw)
mM min-' (mg enzyme)-'
2.8 8.3 13.9
%50 0.1 1 0.1 5
0.09 0.15 0.16
.- --
Table 3. Dependence of the reduction rate of NAD' by I-butanol on w, [warer]/(surfactant] molar ratio in microemulsions Ah Microemulsion
NAD' concn
Measured rate
mol/mol
mM
mM min-' (mg enzyme) XO
Ab-3 Ab-4 Ab-5 Ab-6 Ab-7 Ab-8
7.7 15.7 18.2 20.8 23.9 28.7 35.6 51.8
Water
-
1.3 1.3 1.6 1.7 2.0 3.O 3.0 5.1 5.1
Ab-I
, Ab-2
wo
E
1
m
0 -
.. -3
kE m ~
0.047 0.12 0.17 0.28 0.39 0.56 0.86 2.2
microemulsions (Table 2), K,,, wp was larger than in buffer. The V,,, which was close to 50% of the V,, in buffer, did not seem to depend on the water content. In cationic microemulsion Cb-1, as in Ab-1, only slight activitv was observed. The activitv is determined bv the reduction rate of NAD' and this rateincludes the NAD' binding and NADH dissociation rates. We therefore determined the exchange rate between 1-butanol and cinnamaldehyde. As shown in Fig. 4, the coenzyme bound to the enzyme is reduced by I-butanol and then oxidized by cinnamaldehyde (281. The exchange rate does not involve any coenzyme binding-dissociation rate. Indeed, we found that the exchange rate was much less affected than the oxidation rate in microemulsions at low water content (Table 4). Effect of p H on the Michaelis constants and maximum velocity
In buffer, at 0.1 M I-butanol, the K , was nearly constant at pH 6.5 - 8.0 and increased at pH 9.0. A similar trend was
-0
-2 6.5
a
7
9
PH
Fig. 5. Semi-log plot of K, of NAD' und V,,. K, ( 0 )and Vmax( 0 ) and in microemulsion Ab-3 (- - -) were determined in water (-) at pH 6.5 (50 m M phosphate); 7.0 and 8.0 (50 mM TrislHCI) and 9.0 (50mM borate). The I-butanol concentration was 0.1 M for the experiments in water -
observed in both anionic (Fig. 5) and cationic (results not shown) microemulsions. The V,, dependency on pH was comparable in buffer and microemulsions. Kinetic constants ofcinnamyl alcohol in microemulsions Cinnamyl alcohol is a substrate of alcohol dehydrogenase in buffer with a Michaelis constant of 1.6 pM. It exhibits a substrate excess inhibition above 8 pM [34]. In microemulsions At-I and Ct with cinnamyl alcohol, a similar pattern is observed (Fig. 6). However the Kmov was 0.140.19 mM and substrate excess inhibition was observed above 1 mM. The maximum velocities were 0.05 mmol 1-' min-' (mg enzyme)-' for At-I and 0.25 mmol 1-' min-' (mg enzyme)-' for Ct, compared to 25 mmol I-' min-' (mg enzyme)-' in buffer. The reduction of the V,, in microemulsions At-1 and Ct could be related to the already discussed dependency of V,,, in microemulsion on the value of w,. Indeed these microemulsions contain small amounts of water.
61 5 hydrates the polar heads of the surfactant molecules and the counterions. The remaining free water molecules rapidly exchange with bound water molecules with characteristic times of s [39, 401. The free water structure in water/oil microemulsions is identical to that of bulk water. Thus, enzyme may be active as solute in the water-dispersed phase. For the cases where the enzyme substrate should not be available from the microemulsion itself, we studied the SDS/ t-butanol/water/cyclohexanesystem. The phase diagram is very similar to that obtained with 1-butanol except that the range of constituent proportions are more limited (unpublished results). We found that in the CTAB/l-butanol/water/cyclohexane phase diagram, there were large domains of water-in-oil microemulsions. This was not unexpected since this is also the 0 6 12 case with toluene as the oil component [41]. 1/1S1 (mM-') Liver alcohol dehydrogenase lost its activity quite rapidly Fig. 6. Lineweaver-Burk representation of the determination of the in anionic microemulsions. The activity loss of alcohol dehyMichaelis constant and the V,,, of cinnamyl alcohol in microemulsion drogenase could be related to the presence of SDS molecules At-2 in the water pseudo-phase. Indeed, it is known that SDS in solution caused subunit dissociation and thus inactivation of the enzyme [33]. Although the inactivation mechanism is Preparative reduction of cinnamaldehyde in microemulsion unknown, according to the protein structures determined by X-ray methods the most accessible areas of the enzyme are Taking advantage of the dual function of 1-butanol as the substrate and the coenzyme binding sites [27, 421. The enzyme substrate and as cosurfactant and of the partition coenzyme NAD' and the nucleotides AMP and dAMP phenomena, we designed an experiment according to Fig. 4. (dAMP only in type At microemulsion) efficiently protected We expected the NAD+/NADH pair and the hydrophilic the enzyme against this inactivation. protein to be located in the water core and the aromatic AMP was a competitive inhibitor to NAD' whereas substrate to have a transitory location in this phase. The dAMP had no kinetic effect under the present conditions. The concentration of cinnamaldehyde was now significant for predifference in kinetic behaviour of these two nucleotides in parative purposes. In the microemulsion inhibition by substrate excess was avoided and the recycling of the coenzyme solution was expected, since dAMP lacks the C-2' hydroxyl was done by the 1-butanol oxidation. An isosbestic point was group which is essential for coenzyme binding: its interaction with an aspartic side chain of the protein acts as an anchor detected while the reaction proceeded. point for adenosine binding to alcohol dehydrogenase [42]. Nevertheless, the similarity of these two nucleotides might be useful to differentiate a specific effect on alcohol dehydroDISCUSSION genase and an effect on the microemulsions. The protection The anionic microemulsions used in this study lie in an by dAMP in buffer and in type At microemulsion but not in area of the pseudophase diagram where it is generally type Ab microemulsion is puzzling, as is the fact that the assumed, although this is debated, that the dispersed phase is dissociation constant of AMP, determined with respect to the made of spherical droplets [35, 361. The water is thus water content in the protection experiments, was found to be surrounded by an interfacial film of surfactant and lower than the inhibition constant. In CTAB microemulsions, at a water content below 8%, cosurfactant molecules; the hydrodynamic radius is in the lo-' m range [23, 371. This is compatible with the protein the enzyme activity was hardly detectable. With 1-butanol, size (4.5 x 6.0 x 11 nm3) [2]. The droplet diameter and the the enzyme activity was stable above 21 % water, but decreased interfacial area have been shown, in a related system, to in- with time at lower water content. When 1-pentanol and 1crease linearly with the mass fraction of water [37]. M'icro- hexanol were used as cosurfactant, the enzyme activity was emulsions are thermodynamically stable systems charac- stable at 7% water. In buffer solutions enzyme activity was terized by fast exchange of the water and of the cosurfactant insensitive to CTAB at the saturating concentration (10 mM) components between the pseudophases. So all the hydrophilic of this salt. No protection by NAD' was observed in cationic NAD' and NADH coenzyme molecules present in the water microemulsions. droplets will be accessible to the enzyme. In water/oil At lower water content in the microemulsion, the water microemulsions, the local motion of SDS especially at its polar droplet size comes close to the size of the enzyme. It is not head is small and decreases as the water content increases [38]. unexpected that the coenzyme binding and dissociation and The exchange of alcohol molecules between the interfacial therefore the activity would be affected under these film and the continuous and/or the dispersed phases depends conditions. We then used a test where if the coenzyme remains on their solubility in these phases but in any case the process bound to the enzyme, the binding-dissociation of the substrate is very fast with a characteristic time of about lo-' s [39]. and the interconversion of the ternary complexes can be There are, to our knowledge, no published indication nor detected [28]. In the microemulsion containing 2.8% water, method for the determination of the concentration of there was very low activity and the exchange rate was 50% of constituents in the different phases. that at higher water content. So the lower water content The nature of water as a dispersed phase is evidently of perturbs the coenzyme binding-dissociation process much crucial importance with respect to enzyme activity. In water/ more than the interconversion of the ternary complexes and oil microemulsions, part of the water, called bound water, the binding-dissociation of the substrate. The interconversion
616 of the ternary complexes occurs inside the protein, so the solvent surrounding the protein should not influence this step. But the binding-dissociation of a polar ligand like the coenzyme used here will be slowed down in nonpolar solvents and in microemulsions at low water content. In contrast the binding-dissociation of a nonpolar ligand should be less changed in nonpolar solvents except that the binding tends to become less favorable since the unfavorable solvation of a nonpolar ligand and of a nonpolar binding site has been suppressed in going from water to nonpolar solvents. However, because of their microheterogenity, microemulsions are not very favorable mediums to test the proposals. The Michaelis constant of NAD' and the maximum velocity of NAD' reduction were determined in microemulsions Ab-2 to Ab-8 and Cb-2 to Cb-5 (Table 2). The discrepancy between K, in buffer and Kmwpin anionic and cationic microemulsions might reflect a different mechanism of alcohol dehydrogenase in these mediums and/or depend on the amount of 1-butanol available for catalysis. When the mass ratio of 1-butanol versus CTAB was varied over 0.68 2.0 (0.6 - 1.8 molar ratio) identical K,,, and V,,, were determined, which indicates that a constant amount of substrate seems to be available for catalysis. The maximum velocity in anionic microemulsions is sensitive to the water content (linear relation between reaction rate under saturating NAD' concentration versus w,)'. For cationic microemulsions the maximum velocity is higher and does not depend greatly on the water content. Further study of the kinetics of alcohol dehydrogenase in these media are needed in order to discuss fully these points. As mentioned, the nature of water in the aqueous pseudophase might be different from that of bulk water and pH might be an indicator for such a difference [43]. An indirect means of measuring the proton activity in microemulsions was to study the dependency of the enzyme reaction on pH. The variations of both K, and V,,, in water and in anionic and cationic microemulsions are very similar. Thus a marked difference of local pH at the enzyme active site in microemulsion is not detected. Replacement of 1-butanol by t-butanol in SDS and CTAB microemulsions was only possible in a narrow range of constituent proportions. Microemulsions At-1 and Ct were used to study the oxidation of cinnamyl alcohol by alcohol dehydrogenase. Substrate specificity of alcohol dehydrogenase has been studied in Aerosol OT (AOT) micelles. It was found that the apparent second-order rate constant (kCaJK,,, ov) for linear primary alcohols was maximal for octanol in buffer and for butanol in these micelles [8]. The cinnamyl alcohol partitioned between the different pseudo-phases and if we assume a partition coefficient of 100, the partition coefficient between octanol and water for this alcohol [44], the KmOvgave a calculated K , wp which is very close to the K,,, value in water. The interesting and very useful point was the increase of the overall concentration of cinnamyl alcohol where substrate excess inhibition started to be detectable: 1 mM (Fig. 6) compared to 8 pM in water [34]. The cinnamaldehyde reduction with coenzyme recycling was then done in microemulsion according to Fig. 4. Under the present experiment conditions, a close to quantitative reduction (96%) of cinnamaldehyde (10 mM)
* The kinetics in microemulsion Ab-8 matches with those of microemulsions Ab-2 to Ab-7 although the molecular arrangement of microemulsion Ab-8 may be represented as fluctuating layers of water and oil phase.
was achieved using 1 nM enzyme active site and 1 mM NAD'. Microemulsions may be a very useful reaction mediums for substrates of low water solubility [45, 461 and hydrophlic enzymes. The influence of the cosurfactant in cationic microemulsions was studied. We had to substitute cyclohexane by a 1 : l (v/v) mixture of cyclohexane and octane in order to obtain stable microemulsions with 1-pentanol and 1-hexanol. In the cationic microemulsions made with 1-pentanol and 1hexanol (Cp and Ch), alcohol dehydrogenase was definitely more stable than in the cationic microemulsions made with n-butanol. Such variation should be taken into account to determine the appropriate microemulsion composition for enzyme activity. The study of alcohol dehydrogenase activity in microemulsions revealed the possibility of achieving enzyme catalysis in macroscopic homogeneous solutions essentially made of organic solvent. Replacement of the anionic surfactant by a cationic one, which was a softer detergent with respect to protein denaturation, provided a better medium for the use of enzymes in microemulsions. In CTAB microemulsions, maximum velocities, enzyme efficiency and enzyme stability were better than in sodium dodecyl sulfate microemulsions. How general this is remains an open question. We thank Prof. R. Zana and Dr M. Bourrel for helpful discussions. This work was supported by grants from the Fondationpour la Recherche Mkdicale Francaise and from the Sociktk Nationale Elf Aquitaine (Production).
REFERENCES 1. Luisi, P. L. & Wolf, R. (1982) in Solution behavior of surfactants (L. K. Mittal, ed.) vol. 2, p. 887, Plenum Press, New York. 2. Luisi, P. L. (1985) Angew. Chem. Int. Ed. Engl. 24, 439-450. 3. Levashov, A. V., Khmelnitsky, Y. L., Klyachko, N. L., Chungak, V. Y. A. & Martinek, K. (1982) J . Colloid Interface Sci. 88, 444-457. 4. Martinek, K., Levashov, A. V., Klyachko, N., Khmelnitski, Y. L. & Berezin, I. V. (1986) Eur. J . Biochem. 155,453 -468. 5. Waks, M. (1986) Proteins 1,4-15. 6. Wolf, R. & Luisi, P. L. (1979) Biochem. Biophys. Res. Commun. 89,209-217. 7. Martinek, K., Levashov, A. V., Klyachko, N. L., Pantin, V. I. & Berezin, I. V. (1981) Biochim. Biophys. Acta 657,277-294. 8. Martinek, K., Levashov, A. V., Khmelnitsky, Y. L., Klyachko, N. L. & Berezin, I. V. (1 982) Science (Wash. D C ) 218, 889 891. 9. Bonner, F. J., Wolf, R. & Luisi, P. L. (1980) J. Solid Phase Biochem. 5,255-268. 10. Meier, P. & Luisi, P. L. (1980) J. Solid Phase Biochem. 5 , 269282. 11. Grandi, C., Smith, R. E. & Luisi, P. L. (1981) J . Bid. Chem. 256, 837-843. 12. Barbaric, S. & Luisi, P. L. (1981) J. Am. Chem. SOC.103,42394244. 13. Ohshima, A., Narita, H. & Kito, M. (1983) J . Biochem. (Tokyo) 93,1421 - 1425. 14. Luisi, P. L. (1985) Angew. Chem. h t . Ed. Engl. 24,439-450. 15. Fendler, J. H. (1976) Ace. Chem. Res. 9, 153-161. 16. Kitahara, A. (1980) Adv. Colloid Interface Sci. 12, 109- 140. 17. Eicke, H. F. (1980) Top. Curr. Chem. 87, 85- 145. 18. Danielsson, I. & Lindman, B. (1981) Colloids Surfaces 3, 391 392. 19. de Bourayne, C., Maurette, M . T., Oliveros, E., Riviere, M., de Savignac, A. & Lattes, A. (1982) J. Chim. Phys. 79, 139-147. 20. Biais, J., Clin, B., Lalanne, P. & Lemanceau, B. (1977) J . Chim. Phys. 74,1197 - 1202.
617 21. Cazabat, A. M., Lagues, M., Langevin, D., Ober, R. & Taupin, C. (1978) C. R. Acad. Sci. Paris 287 B, 25 - 28. 22. Sauterey, C., Taupin, C. & Dvolaitzky, M. (1978) C. R. Acad. Sci. Paris 287C, 77-80. 23. Dvolaitzky, M., Guyot, M., Lagues, M., Le Pesant, J. P., Ober, R., Sauterey, C. & Taupin, C. (1978) J . Chem. Phys. 69,32793288. 24. Lalanne, P., Biais, J., Clin, B., Bellocq, A. M. & Lemonceau, B. (1978) C. R. Acad. Sci. Paris 286C, 55-57. 25. Lalanne, P., Biais, J., Clin, B., Bellocq, A. M. & Lemonceau, B. (1978) J . Chim. Phys. 75,236-240. 26. Branden, C. I., Jomvall, H., Eklund, H. & Furugren, B. (1975) in The enzymes (P. D. Boyer, ed.) vol. 11, p. 103, Academic Press, New York. 27. Eklund, H., Samama, J. P., WallCn, L., Branden, C. I., Wkeson, A. & Jones, T. A. (1981) J . Mol. Biol. 146, 561 -587. 28. Lee, K. M., Blaghen, M., Samama, J. P. & Biellmann, J. F. (1986) BiOOrg. Chem. 14,202-210. 29. Hilhorst, R., Laane, C. & Veeger, C. (1983) FEBS Lett. 159,225 228. 30. Hilhorst, R., Spruijt, R., Laane, C. & Veeger, C. (1984) Eur. J . Biochem. 144,459-466. 31. Fletcher, P. D. I., Rees, G. D., Robinson, B. H. & Freedman, R. B. (1985) Biochim. Biophys. Acta 832,204-214. 32. Lineweaver, H. & Burk, D. (1934) J . Am. Chem. Soc. 56, 658666. 33. Blomquist, C. H., Smith, D. A. & Martinez, A. M. (1967) Arch. Biochem. Biophys. 122,248 - 250.
34. Abdallah, M., Biellmann, J. F., Samama, J. P. & Wrixon, A. D. (1977) Eur. J. Biochem. 64, 351 -360. 35. Holt, S. L. (1980) Dispersion Sci. Technol. I , 423-464. 36. Lindman, B., Ahlnas, T., Almgren, M., Fontell, K., Johnsson, B., Kahn, A,, Nilsson, P. G., Olofsson, G., Soderman, 0. & Wennestrom, H. (1982) Finn. Chem. L., 74-85. 37. Nguyen, T. & Ghaffaric, H. M. (1979) J. Chim. Phys. 76, 513520. 38. Hansen, J. R. (1974) J . Phys. Chem. 78,256-261. 39. Zana, R. & Lang, J. (1982) in Solution behavior of surfactants. Theoretical and applied aspects (K. L. Mittal & E. J. Fendler, eds) p. 1195, Plenum Press, New York. 40. Wong, M., Thomas, J. K. & Nowak, T. (1977) J. Am. Chem. SOC. 99,4730-4736. 41. Mackay, R. A. & Hermansky, C. (1981) J . Phys. Chem. 85,739744. 42. Eklund, H., Samama, J. P. & Jones, T. A. (1984) Biochemistry 23, 5982- 5996. 43. Nome, F., Chang, S . A. & Fendler, .I.H. (1976) J . Chem. Soc. Faraday Trans I 72,296 - 302. 44. Leo, A., Hansch, C. & Elkins, D. (1971) Chem. Rev. 71, 525616. 45. Lee, K. M. & Biellmann, J. F. (1986) Bioorg. Chem. 14, 262273. 46. Lee, K. M. & Biellmann, J. F. (1986) Nouv. J . Chim. 10, 675.
