- Email: [email protected]
Kinetic characteristics of the enzymatic conversion in presence of cyclodextrins: study of the oxidation of polyunsaturated fatty acids by lipoxygenase
Biochimica et Biophysica Acta 1347 Ž1997. 140–150
Kinetic characteristics of the enzymatic conversion in presence of cyclodextrins: study of the oxidation of polyunsaturated fatty acids by lipoxygenase Jose´ M. Lopez-Nicolas, ´ ´ Roque Bru, Francisco Garcıa-Carmona ´
)
Departamento de Bioquimica y Biologıa ´ Molecular A, Facultad de Biologıa, ´ UniÕersidad de Murcia, Campus de Espinardo, E-30001 Murcia, Spain Received 6 February 1997; revised 8 April 1997; accepted 15 April 1997
Abstract The capability of cyclodextrins ŽCDs. to greatly enhance the solubility in water of poorly water-soluble substances makes them an ideal alternative system for studying the expression of enzyme activity with such substrates in aqueous solution. In order to evaluate the behaviour of the enzymes in presence of CDs a study of the lipoxygenase ŽLOX.-catalyzed oxidation of polyunsaturated fatty acids ŽPUFA. as model reaction has been carried out. This was done by using LOX from two different sources Žsoybean and potato tuber., at two pH values Ž6.3 and 9.0., with two substrates Žlinoleic acid and arachidonic acid. and three types of CD Ž b-CD, methyl-b-CD and monoglucosyl-b-CD.. PUFA have been shown to form inclusion complexes of 1:2 stoichiometry which are in equilibrium with free PUFA and free CD, thus complexation is governed by two equilibrium constants, K 1 and K 2 ŽJ.M. Lopez-Nicolas ´ ´ et al., Biochem. J. 308 Ž1995. 151–154; R. Bru et Ž . . al., Colloids Surf. 97 1995 263–269 . For the oxidation of PUFA by LOX in the presence of b-CD we propose a model in which free PUFA is the only effective substrate, thus the oxidation of the complexed substrate requires the previous dissociation of the complex. The equilibrium constants of complex formation are determined by both a physico-chemical and an enzymatic method. In spite of giving quite similar results, the second was proven to be more accurate so it was employed in further studies. CD was shown to slow down the reaction rate of LOX, specifically due to the increase of K m , Vmax remaining unchanged. That apparent inhibition is due to removal of effective Žfree. substrate in the form of inclusion complexes. This ‘sequestered’ substrate can, however, be converted since it is in equilibrium with the free. The feasibility of realizing a CD-mediated accurate control over the conversion rate is demonstrated in the experiment called ‘cyclodextrin assay’ in which the concentration of the free substrate is calculated by using the equilibrium constants of complex formation and setting the initial concentrations of total substrate and total CD. From the observation of the reaction progress curves in the conditions of the CD assay, we have studied some characteristic parameters of the oxidation of PUFA by LOX in this new medium, such as enzymatic activity, duration of linear product accumulation and the lag phase. q 1997 Elsevier Science B.V. Keywords: Lipoxygenase; Cyclodextrin; Polyunsaturated fatty acid; Cyclodextrin assay; ŽPotato.; ŽSoybean.
Abbreviations: FA, fatty acids; PUFA, polyunsaturated fatty acids; CD, cyclodextrin; LOX, lipoxygenase; LA, linoleic acid; AA, arachidonic acid; c.m.c., critical micellar concentration; DLPA, duration of linear product accumulation. ) Corresponding author. Fax: q34 68364147. E-mail: [email protected] 0005-2760r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 5 - 2 7 6 0 Ž 9 7 . 0 0 0 6 1 - 1
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
1. Introduction Cyclodextrins Ž CDs. are torus-shaped oligosaccharides made up of 6–8 glucopyranose units and originated by the enzymatic degradation of starch, through the action of CD-glucano-transferase w1x. The cavity is carpeted by hydrogen atoms and so has a rather hydrophobic nature unlike the outer surface of the molecule, in which the primary and secondary hydroxyl groups are exposed to the solvent, thus making the whole molecule highly water-soluble w2x. Poorly water-soluble compounds and hydrophobic moieties of amphiphilic molecules interact non-covalently with the CD cavity to form the so-called inclusion complexes, which are also highly watersoluble w3,4x. For instance, steroids w5–7x and longchain fatty acids Ž FA. w8–10x, which are examples of each type of molecule, increase their aqueous solubility by several orders of magnitude when complexed with CDs. As opposed to other solubility enhancers Ždetergents, water-miscible solvents, etc.., CDs produce individual molecules of the solute, thus preventing side effects due to aggregation phenomena. The ability of CDs to enhance the aqueous solubility of many different compounds w3,11x makes them an interesting alternative system for performing enzyme-catalyzed conversions of poorly water-soluble substrates w12–15x compared with organic media or two-phase systems w16,17x. The interest of using CDs in enzyme-catalyzed reactions is 3-fold: first, they are of technological interest as high loads of poorly water-soluble substrates can be achieved; second, amphiphilic fatty, waxy or even liquid apolar substrates can be handled easily as they become readily water-soluble amorphous or crystalline solids when forming inclusion complexes, and third, information on kinetic properties, which otherwise would be inaccessible or masked by the effect of additives Ždetergents, cosolvents, etc.. , becomes available. Since substrates become water-soluble, it is of particular interest to use soluble enzymes acting on hydrophobic substrates. A growing number of studies are devoted to performing bioconversions of commercially interesting CD-solubilized compounds but, in spite of general interest in the field of basic and applied enzymology, no attempt has been made to date to understand the fundamentals and to model the process of enzyme catalysis in the presence of CDs.
141
Our interest was to investigate how an enzymatic reaction proceeds in CD-containing aqueous medium and then to characterize the different environmental factors which may affect the reaction. Lipoxygenase Žlinoleate:O 2 oxidoreductase, EC 1.13.11.12. ŽLOX. was found to be a good model enzyme for our purpose. This is a group of dioxygenases, many of them cytosolic, which catalyze the hydroperoxydation of polyunsaturated fatty acids ŽPUFA. containing at least one 1,4-ŽZ,Z.pentadiene system in their structure, such as linoleic ŽLA., linolenic and arachidonic ŽAA. acids w18–20x. Depending on its origin, LOX displays a marked product specificity as regards the position attacked by O 2 in AA w21x; for example, soybean LOX-1 is a 15-LOX, potato tuber and mammalian leukocyte LOX are both 5-LOX, blood platelet LOX is a 12-LOX, etc. In addition, there are important structural differences among the different LOX w22,23x. Fatty acids ŽFA. in general do not form true solutions due to their amphiphilic character but they form colloidal aggregates above a certain critical concentration; depending on pH these can adopt different arrangements, including a separate oil phase, bilayers and micelles w24x. The tendency of FA to form these kind of aggregates makes them excellent candidates for solubilization in CDs as inclusion complexes. Most LOXs display optimum activity at pH values around neutrality w25x but in these conditions PUFAs are quite insoluble. Raising pH to 9.0 or 10.0 favours PUFA solubilization through the formation of soaps w24x, but kills LOX activity with a few exceptions, such as soybean LOX-1. For this reason the kinetic properties of LOX are well known at pH 9–10 w26x and practically unknown at neutral or slightly acid pH values. The aggregation of LA and acid AA in the presence of CDs has been investigated previously by us w27,28x, and it was demonstrated that one PUFA molecule may interact with up to two CDs molecules in solution. Such an interaction leads to a CD concentration-dependent increase of the PUFA c.m.c. since the complexes do not take part in the aggregation. The inclusion process of specific PUFA is characterized by two equilibrium constants, K 1 and K 2 , which can be affected by pH, temperature and CD type. In this paper we propose a model for enzyme catalysis in a CD medium and apply the theoretical
142
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
concepts to the LOX-catalyzed dioxygenation of PUFA with good agreement. The model was tested by using LOX from two sources Žpotato tubers and soybean., two substrates ŽLA and AA., at two pH values Ž6.3 and 9.0. and different CDs Ž b-CD, methyl-b-CD and monoglucosyl-b-CD..
2. Materials and methods LA and AA acid were purchased from Cayman Chemical Co. ŽParis, France.. b-CD, methyl-b-CD and LOX Žtype V. from soybean were obtained from Sigma ŽMadrid, Spain. . Monoglucosil-b-CD was purchased from Ensoiko Sugar Refining Co., Ltd. ŽYokohama, Japan.. LOX was purified from potato tubers Žvar. Desiree´. according to Mulliez et al. w29x Žspec. act. 27 m mol O 2rminrmg protein.. Diphenylhexatriene ŽDPHT. was a product from Fluka ŽMadrid. and tetrahydrofurane ŽTHF. was from Merck ŽDarmstadt, Germany.. All the other chemicals used were of the highest purity. 2.1. Preparation of FA-CD complexes FA-CD complex solution was prepared by dissolving the CD in 0.1 M potassium phosphate buffer pH 6.3 containing 1% vrv EtOH or potassium borate buffer pH 9 containing 1% vrv EtOH, followed by addition of the fatty acid prepared in the same buffer. The samples were flushed with N2 to prevent oxidation of FA during preparation. Just before using, samples were vigorously shaken in order to supply O 2 for the enzymatic reaction. 2.2. Determination of equilibrium constants by means of critical micellar concentration (c.m.c.) Although the term c.m.c. is only correct to describe the aggregation of micelle-forming amphiphiles, we will use this term throughout this work to refer also to the concentration at which any type of aggregates appears. The c.m.c. value of LA was determined by means of a fluorescence spectroscopic method originally developed to determine c.m.c. values of detergents w30x and adapted to determine the concentration at which aggregation of fatty acids occurs w27,31x. Briefly, 2-ml samples contained 0.1
Scheme 1. Scheme of an enzymatic reaction in the presence of CD forming 1:2 complexes with the substrate.
M potassium phosphate buffer ŽpH 6.3., 0.88 m M DPHT Žsupplied in 2 m l THF., 1% vrv ethanol and the indicated concentration of CD and PUFA. The samples were flushed with N2 and incubated for 30 min in the dark at the desired temperature for equilibration and in order to reverse the photoisomerisation of the fluorescent probe. Fluorescence intensity measurements were made at 258C at 430 nm Ž358 nm excitation wavelength. in a Kontron SFM-25 spectrofluorimeter equipped with thermostated cells. C.m.c. was determined graphically from a plot of fluorescence relative values versus PUFA concentration as the extrapolated intersection between the lines defining the fluorescence in the pre- and post-micellar regions. This determination makes it possible to set the conditions in which FA is molecularly dispersed and does not form aggregates. C.m.c data at different CD concentrations were fitted by non-linear regression to an equation as described by ourselves w27,28x to determine the equlibrium constants, K 1 and K 2 , of Scheme 1. 2.3. Enzyme assay Lipoxygenase activity was assayed by monitoring the increase in absorbance at 234 nm Ž 234 s 25 000 My1 cmy1 . of the forming hydroperoxides in a Kontron Uvikon 940 spectrophotometer at 258C equipped with thermostated cells. The reaction was started by adding 5 m l enzyme to 1 ml of complex FA-CD preparation. 2.4. Model of enzyme catalysis in the presence of CDs The proposed reaction scheme assumes that free S is the only form of substrate that LOX is able to bind
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
to in order to react. Accordingly, the reaction rate equation should be: Vs
Vmax P S
f
Kmq S
f
Ž1.
The concentration of free substrate can be deduced by analyzing the equilibrium of complex formation. The mass balance of the S and CD in an aqueous solution may be represented by Eqs. Ž 2. and Ž 3.:
the reciprocal form being the equation: 1 Km P K1 P K2 2 s P CD t V Vmax . S t Km P K1 q P CD t Vmax P S t Km 1 q q Vmax P S t Vmax
143
Ž9.
3. Results and discussion S t s S f q S y CD q S y CD 2
Ž2.
CD t s CD f q S y CD q 2 S y CD 2
Ž3.
where subscripts f and t stand for free and total respectively. K 1 and K 2 are the equilibrium constants, which are defined as follows: S y CD
K1 s
S
f
CD
Ž4. f
S y CD 2
K2s
S y CD CD
Ž5. f
If the concentration of the complex is negligible with respect to wCDx t , then wCDx f , wCDx t , and the amount of free, non-complexed substrate can be expressed after rearrangement of Eqs. Ž 2. , Ž 4. and Ž 5. as: S fs
S
t
1 q K 1 P CD t q K 1 P K 2 P CD
Ž6.
2 t
The above condition applies when wCDx t 4 w S x t or, with a wCDx t not much higher than w S x t , if the equilibrium constants are not very strong. Substituting Eq. Ž6. into Eq. Ž1. gives: Vs
Vmax . S K mapp q S
t
Ž7. t
where K mapp s K m P 1 q K 1 P CD t q K 1 P K 2 P CD
ž
2 t
/
Ž8.
The use of CDs in enzyme-catalyzed bioconversions of poorly-water-soluble compounds should be advantageous due to the hugely increased aqueous solubility of such compounds. This is the case with the cholesterol oxidase-catalyzed synthesis of 7a-hydroxycholest-4-en-3-one w12x in a high yield in the presence of hydroxypropyl-b-CD as compared to the chemical synthesis, or the phenoloxidase-catalyzed regiospecific hydroxylation of 17-b estradiol w15x. In spite of these promising results, in which CDs were used as substrate carriers, the role of CD in the process is still not well understood since in some cases CDs are described as inhibitors w32x and in others as enhancers of the enzyme activity w33x, or, at least, as not having a negative effect w34x. In this study we investigate the role of CD in the LOX-catalyzed dioxygenation of PUFA using as a basis the hypothetical reaction Scheme 1, which takes into account the previously demonstrated model of PUFA-CD interaction w27,28x. The reaction rate equation which is obtained ŽEq. Ž7.. indicates that CDs do not affect Vmax , while the observed K mapp ŽEq. Ž8.. is a function of CD concentration and the equilibrium constants for inclusion complex formation. Accordingly, if Vmax and K m in the absence of CDs as well as the equilibrium constants are known, the reaction rate can be predicted at any CD concentration. Only by knowing the kinetic constants, the equilibrium constants might be evaluated by n.l.r. fitting of experimental data to Eq. Ž 7. or Eq. Ž 9., and vice versa. Sometimes, and specially for substrates of low water solubility, the independent evaluation of the kinetic constants in the absence of CDs, and any other additive, may not always be possible because of the technical problem that low solubility represents.
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
144
In such cases, the equilibrium constants might be determined by different physico-chemical methods w3x and then used in the above equations to have access to the kinetic constants, specially to K m , by n.l.r. of data in the presence of CDs. In our experimental system all the parameters in Eqs. Ž7. – Ž9. can be obtained independently and thus be used to establish the corresponding comparisons. 3.1. Enzymatic method Õs. physico-chemical method for the determination of complex formation equilibrium constants So that Eq. Ž9. could be applied to the determination of the equilibrium constants, the Vmax and K m of soybean LOX using LA as substrate were determined by Lineweaver-Burk plots in the absence of CDs. The resulting values were 2.6 m Mrmin and 9.7 m M at pH 6.3 and 34.8 m Mrmin and 62 m M at pH 9.0. Then, an experiment was performed in which LOX activity was determined at different CD concentrations with the LA concentration kept constant. Soybean LOX activity decreased as CD concentration increased in each set of measurements made at a fixed LA concentration. This inhibitory behavior, as analyzed through the Dixon plot Ž Fig. 1. , cannot be explained easily since 1rv dependence on CD concentration is not linear. According to Eq. Ž9., such dependence is expected to be quadratic. Thus Eq. Ž 9. was used to obtain the equilibrium constants by n.l.r. fitting of these data, and using the previously determined kinetic parameters as constant coefficients. Results are shown in Table 1. Although the equilibrium constants are commonly determined by physico-chemical methods, the above method may be regarded as the enzymatic alternative for determining the equilibrium constants of complex formation. We have also used the method based on the CD-induced increase of PUFA c.m.c. w27,28x to
Fig. 1. Dixon plot of soybean-LOX catalyzed LA oxidation at pH 6.3 in the presence of b-CD at different LA concentrations: Žv . 7 m M; Ž%. 55.2 m M.
determine the equilibrium constants. As seen in Table 1, the values obtained by this latter method are quite similar to those obtained using the enzymatic method, and so both seem equally valid. Both methods agree in that LA-b-CD interaction is best described by two equilibrium constants as depicted in Scheme 1. Furthermore, the fact that both methods yield similar results reinforces some of the assumptions made, particularly that LOX only uses free PUFA as substrate and that the apparent inhibition of LOX activity by CD is due to reduction of the effective substrate. 3.2. CDs as the basis for a new type of assay: the CD assay On the basis of Scheme 1, it should be possible to finely control the concentration of free PUFA if the equilibrium constants are known, merely by setting
Table 1 Equilibrium constants of LA-b-CD inclusion complex determined by using physico-chemical and enzymatic methods Method Physico-chemical Enzymatic b a b
pH 6.3 a
According to Refs. w27,28x. See text for explanation.
pH 9
K 1 ŽmMy1 .
K 2 ŽmMy1 .
K 1 ŽmMy1 .
K 2 ŽmMy1 .
10.2 " 1.4 11.2 " 1.2
2.6 " 0.9 1.7 " 0.4
1.5 " 0.1 4.1 " 0.8
1.7 " 0.5 3.7 " 0.7
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
Fig. 2. ‘Cyclodextrin assay’: soybean LOX-catalyzed oxidation of LA reaction rates at 7 m M free LA concentration at pH 6.3. The b-CD and LA concentrations were calculated to yield 7 m M free LA using the set of constants determined enzymatically Žv . and the set of constants determined by physico-chemical method Ž`.. For a given b-CD concentration the total LA concentration was readily calculated by applying Eq. Ž6..
the appropriate concentrations of both total CD and total PUFA. In the so-called CD assay, the total CD and PUFA concentrations are set so that the free PUFA concentration remains constant. Consequently, the enzymatic reaction rate should be constant regardless of total CD and PUFA concentrations. By using each set of constants determined through either the enzymatic or the physico-chemical method, we have performed the corresponding CD assays. A concentration of free PUFA below its c.m.c. Ž10.6 m M at pH 6.3 and 170 m M at pH 9.0 for LA w27x. has to be chosen in order to avoid turbidity in spectrophotometric measurements. Fig. 2 shows the results obtained at pH 6.3. The set of constants determined enzymatically responded as expected in a CD assay Žthat is activity independent of total CD or PUFA concentration. but that obtained with the set of constants determined physico-chemically deviated from expectations. Thus, the small differences observed between the enzymatically and physico-chemically obtained set of constants ŽTable 1. seem to be very important as regards the actual effective PUFA
145
concentration. Indeed, if we assume the enzymatic set to be correct, the initial conditions set for the physico-chemical constants give lower concentrations of free PUFA than expected which would explain the decrease in LOX activity observed. At pH 9, the enzymatic set of constants also worked well, but the physico-chemical set failed once again to yield a LOX activity independent of total CD or PUFA concentration Žresults not shown.. Although such physico-chemical method, based on the fluorescence intensity break-point of a probe at the c.m.c. w30x, has been used to properly describe different features of the CD-PUFA interaction w27,28x, it does not seem to be accurate enough to provide equlibrium constants for the performance of the CD-assay. Indeed, it must be less accurate than the enzymatic method because in the former the n.l.r. to determine the equilibrium constants is applied to a secondary plot Žc.m.c. vs. CD plot proceeds from fluorescence vs. PUFA concentration plots., while in the enzymatic method it is applied to a primary plot, the enzyme activity vs. CD. In any case, the CD assay
Fig. 3. ‘Cyclodextrin assay’: potato LOX-catalyzed oxidation of LA reaction rates at 7 m M free LA concentration at pH 6.3. The b-CD and LA concentrations were calculated to yield 7 m M free LA using the set of constants determined enzymatically with soybean LOX Žv . and the set of constants determined by physico-chemical method Ž%.. For a given b-CD concentration the total LA concentration was readily calculated by applying Eq. Ž6..
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
146
Table 2 Kinetic parameters of potato-LOX in 0.1 M potassium phosphate buffer pH 6.3 containing 1% vrv ethanol in the absence or presence of b-CD w b-CDx ŽmM. 0 0.5 1 1.5
Vmax Ž m Mrmin. 4.6 4.5 4.7 4.4
K m Ž m M. Observed
Expected enzymatic.
Expected phys-chem.
3.5 20.6 100.0 211.0
3.5 39.9 108.0 212.0
3.5 45.1 135.0 269.0
The values of K m observed are experimental data and the values of K m enzymatic and phys-chem. are obtained applying Eq. Ž8. using equilibrium constants of LA-b-CD inclusion complex formation determined through either physico-chemical or enzymatic methods.
behaved better with the enzymatic constants due to interdependence between the enzyme and the constants. The accuracy of the enzymatic set of constants was further checked. The constants determined as above using soybean LOX at pH 6.3 were applied to perform the CD assay with potato LOX which has different enzymatic properties Žoptimum pH, product specificity, molecular weight, etc.. w29,35x. The physico-chemical constants were also applied for comparison purposes. As seen in Fig. 3, the results were qualitatively similar to those obtained with soybean LOX ŽFig. 2., thus confirming the accuracy of the enzymatic set of constants and the fact that they are independent of the enzyme being determined. Furthermore the kinetic parameters for potato LOX at pH 6.3 were determined experimentally in different CD conditions and the determined K m values were compared with those calculated from Eq. Ž8. for each set of equilibrium constants. As seen in Table 2, the results obtained by using the enzymatic set are more similar to those determined experimentally than those obtained using the physico-chemical set. Note that the experimental Vmax is independent of CD concentration, as predicted from Scheme 1. We conclude that a set of enzymatically determined substrate-CD equilibrium constants regardless of which enzyme is used, is sufficiently accurate to be used for the fine control of an enzyme-catalyzed reaction of the same CD-solubilized substrate, as demonstrated in a CD-assay. 3.3. Characteristics of the CD-based assay From a study of the reaction progress curves in the conditions of the CD assay some interesting observa-
Fig. 4. Time course of the soybean LOX-catalyzed oxidation of LA in the absence Ža. and presence Žb, c and d. of b-CD at pH 6.3 in the conditions of the CD assay. In Ža. 7 m M LA, Žb. the medium contained 0.25 mM of b-CD and 34.8 m M of LA, Žc. the conditions were 0.5 mM of b-CD and 79 m M of LA and in Žd. 1.25 mM of b-CD and 310 m M of LA. In all cases the free LA concentration was 7 m M.
tions can be made. Figs. 4 and 5 show the curves for the soybean and potato LOX-catalyzed dioxygenation of LA, respectively. 3.3.1. Duration of linear product accumulation (DLPA) This is defined as the longest time that product accumulation is linear. As seen in both figures, in the absence of CD the substrate which coincides with the free substrate Ž7 or 9 m M., is quickly exhausted. Increasing total substrate concentration can be in-
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
147
during the DLPA by replenishment form the PUFACD complex pool Ž Scheme 1.. Therefore, our results do not support a direct transformation of complexed PUFA as suggested by others authors w32x. As long as the initial conditions are not significantly distorted, for instance by depletion of the substrate or the presence of products, the product will accumulate at a constant rate. These two processes diminished at increasing CD concentrations.
Fig. 5. Time course of the potato LOX-catalyzed oxidation of LA in the absence Ža. and presence Žb, c and d. of b-CD at pH 6.3 in the conditions of the CD assay. In Ža. 9 m M LA, Žb. the medium contained 0.5 mM of b-CD and 102 m M of LA, Žc. the conditions were 0.75 mM of b-CD and 180 m M of LA and in Žd. 1 mM of b-CD and 279 m M of LA. In all cases the free LA concentration was 9 m M.
creased in the presence of CD although the free substrate concentration remaining invariable, leads to a longer DLPA Žfrom 248 s without b-CD up to 900 s in presence of 1.25 mM of b-CD for a LA free concentration of 7 m M. at the same constant rate. If the amount of substrate transformed in the presence of CD is greater than the amount of the free substrate but the reaction rate is that corresponding to the concentration of free substrate, then the reaction pathway would involve the binding of LOX alone to the free PUFA, whose concentration is kept constant
3.3.2. Lag phase The LOX-catalyzed dioxygenation of PUFA displays a characteristic kinetic lag phase w36x, which represents the activation of LOX by its products w37,38x. The lag phase is determined as the extrapolation of the tangent of maximal slope of the reaction progress curve to the ‘time axis’. Such a lag phase can be eliminated by supplementing the reaction medium with stoichiometric Žmicromolar. amounts of hydroperoxide product with respect to the enzyme. As seen in Fig. 5 the lag phase increases with the concentration of CD. Such an increase, which ranges from 17 s when no b-CD is present in the medium up to 109 s when the concentration of b-CD is 1.25 mM, may be due to the competition between the free CD and the enzyme to bind the reaction product, the hydroperoxide. Thus a higher total concentration of hydroperoxide has to be accumulated for the enzyme to be activated by the free hydroperoxide, assuming that this product species is the only that binds the enzyme, as happens in the case of free substrate. In the case of soybean LOX at pH 6.3, the lag phase was practically nil at the enzyme concentration used.
Table 3 ‘Cyclodextrin assay’ wLAx t Ž m M.
w b-CDx t Ž m M.
Activity Ž m Mrmin.
DLPA Žs.
Lag Žs.
82.27 102.24 160.78 295.12 485.29
0 50 150 300 450
22.00 23.01 21.96 20.93 22.00
55 83 120 195 232
0 6 14 31 55
a
Soybean-LOX catalyzed oxidation of LA reaction at pH 9.0. The b-CD and LA concentrations were calculated to yield 82.3 m M free LA using the set of constants determined enzymatically. Data of activity, DLPA and Lag were obtained from the curves of reaction progress. a Too short to be accurately measurable.
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
148
Table 4 Equlibrium constants of LA-CD inclusion complex at different type of CDs y1 .
K 1 ŽmM K 2 ŽmMy1 .
b-CD
methyl-b-CD
monoglucosyl-b-CD
11.16 " 1.20 1.68 " 0.13
14.79 " 1.41 0.18 " 0.01
7.28 " 0.82 1.16 " 0.23
The set of constants was determined by using enzymatic method with soybean-LOX at pH 6.3.
3.4. Factors that affect the CD-based assay The CD assay relies on the interaction of a particular substrate with a particular CD in a particular pH and temperature environment. Since all these factors affect the equilibrium constants, as we previously demonstrated w27,28x, a new set of constants has to be determined in every new condition we wish to run a CD-assay. Thus, the corresponding tests have been performed at another pH, with another type of CD and with another substrate. 3.4.1. pH All the above experiments previous CD assay experiments were made at pH 6.3, so another CD assay was performed using the new set of constants determined with soybean LOX at pH 9.0 Ž Table 1., a pH at which potato LOX is inactive w29x. As seen in Table 3, reaction rate was constant over the range of b-CD concentration tested and the DLPA and lag phase increased as b-CD concentration increased. Note that a quite high concentration of free LA was necessary for lag phases to be measured. 3.4.2. Type of CD All the above experiments were performed with natural b-CD. In further experiments two chemically
modified CDs, namely methyl-b-CD and monoglucosyl-b-CD, were tested for use in the CD-based assay with potato LOX at pH 6.3. The set of constants obtained using soybean LOX ŽTable 4. was used to carry out the respective CD assays. K 1 in the modified CDs is significantly larger than K 2 , as was the case happens with the natural b-CD, and so hydrogen binding still seems to be the most important contribution to the stability of the inclusion complexes for these news CDs w27,28x. In all cases, as can be seen in Table 5 and Fig. 6 the activity remained constant, and the DLPA and lag phase increased under CD-assay conditions. Therefore the CD assay holds with the corresponding sets of constants for both methyl-b-CD and monoglucosyl-b-CD and soybean and potato LOXs. The characteristic increase of the DLPA points to the degree of replenishment of free substrate also for these modified CDs, too. 3.4.3. Type of substrate In addition to LA we also tested AA, which is the natural substrate of mammalian LOX and whose products are involved in a variety of important physiological functions w39x. The procedure is same as at followed for LA. Firstly kinetic parameters in the absence of b-CD
Table 5 ‘Cyclodextrin assay’ wmethyl-b-CD x ŽmM.
Activity Ž m Mrmin.
DLPA Žs.
wmonoglucosyl-b-CDx ŽmM.
Activity Ž m Mrmin.
DLPA Žs.
0 0.15 0.25 0.35 0.5 0.75 1
2.80 2.73 2.74 2.50 2.80 2.97 2.87
34 80 126 190 230 266 295
0 0.15 0.25 0.35 0.5 0.75 1
2.80 2.60 2.96 3.05 2.79 3.01 2.74
25 54 66 72 107 136 211
Potato-LOX catalyzed oxidation of LA reaction with two types of CD at pH 6.3. The methyl-b-CD, monoglucosyl-b-CD and LA concentrations were calculated to yield 5 m M free LA using the set of constants determined enzymatically. Data of activity and DLPA were obtained from the curves of reaction progress.
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
149
were determined Ž K m s 13.31 m M and Vmax s 1.56 m Mrmin.. These were then used in Eq. Ž9. to fit the activity vs. b-CD curves and thereafter determine the equilibrium constants. In this case the following results were obtained K 1 s 16.7 " 3.4 mMy1, K 2 s 1.55 " 0.5 mMy1. Soybean LOX was used for these determinations. The result of the CD assay using this set of constants applied to both soybean and potato LOXs is shown in Fig. 7. These result agree qualitatively with those obtained for LA, indicating the similar behaviour of both substrates with respect to the proposed reaction pathway ŽScheme 1..
4. Concluding remarks Fig. 6. ‘Cyclodextrin assay’. Soybean LOX Že,I. and potato LOX ŽB,l.-catalyzed oxidation of LA reaction rates at 5 m M free LA concentration at pH 6.3 by using different types of CDs: ŽB,I. monoglucosyl-b-CD and Žl,e. methyl-b-CD. Different enzyme concentrations were used in order to avoid overlapping of data for presentation purposes.
We have demonstrated by several means that the reaction pathway of Scheme 1 adequately describes the LOX-catalyzed dioxygenation of PUFA in the presence of CD. This was done by using LOX from two different sources, at two pH values, with two substrates and three types of CD. As result of this study, we have designed the CD assay that makes it possible to control the course of an enzyme catalyzed reaction in the presence of CDs. The CD assay explains the enzyme apparent inhibition by CD and is of general interest for enzyme catalyzed bioconversions in the presence of CDs. Eventually, it should be possible that substrate-CD combinations other than PUFA-CD might be enzymatically transformed as a complex itself if bulky reactive parts of the substrate remain exposed. The model in Scheme 1 should then be adapted accordingly. This would open up new exciting possibilities of the CD-facilitated bioconversions for exploiting the enzymes’ capabilities such as its stereo and regiospecificity.
Acknowledgements
Fig. 7. ‘Cyclodextrin assay’. Soybean-LOX Ž%. and potato-LOX ŽB. catalyzed oxidation of AA reaction rates at 10 m M free AA concentration at pH 6.3. The b-CD and AA concentrations were calculated to yield 10 m M free AA using the set of constants determined enzymatically.
This work has been supported in part by research grants form CICYT ŽProyecto BIO94-0541., INTAS ŽGrant 93r2223 EXT. and Consejerıa ´ de Educacion ´ y Cultura. C.A. Murcia Žproyecto PIB95r03.. J.M.L.N. is a holder of a grant of CajaMurcia. R.B. holds a contract for Doctores Reincorporados.
150
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
References w1x K. Haga, K. Harata, A. Nakamura, K. Yamane, J. Mol. Biol. 237 Ž1994. 163–164. w2x W. Saenger, Angew. Chem. Int. Ed. Engl. 19 Ž1980. 344– 362. w3x Szejtli, J., Cyclodextrin Technology, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1988 w4x J. Szejtli, Med. Res. Rev. 14 Ž1994. 353–386. w5x F. Acarturk, ¨ T. Imai, H. Saito, M. Ishikawa, M. Otagiri, J. Pharm. Pharmacol. 45 Ž1993. 1028–1032. w6x K. Uekama, T. Fujinaga, F. Hirayama, M. Otagiri, M. Yamasaki, Int. J. Pharm. 10 Ž1982. 1–15. w7x K. Koizumi, Y. Okada, Y. Kubota, T. Utamura, Chem. Pharm. Bull. 35 Ž1987. 3413–3418. w8x H. Schlenk, D.M. Sand, J. Am. Chem. Soc. 83 Ž1961. 2312–2320. w9x N. Jyothirmayi, C. Ramadoss, S. Divakar, J. Agric. Food Chem. 39 Ž1991. 2123–2127. w10x R. Bergeron, Y. Machida, K. Bloch, J. Biol. Chem. 250 Ž1975. 1223–1270. w11x Duchene, ˆ D., Žed.. New Trends in Cyclodextrins and Their Derivaties. Editions de Sante´ ŽParis., 1991. w12x D.L. Alexander, J.F. Fisher, Steroids 60 Ž1995. 290–294. w13x J. Jadoun, R. Bar, Appl. Microbiol Biotechnol. 40 Ž1993. 477–482. w14x B. Pekic, Z. Lepojevic, Biotechnol. Lett. 13 Ž1991. 399–404. w15x H.J. Woerdenbag, N. Pras, H.W. Frijlink, C.F. Lerk, T.M. Malingre, ´ Phytochemistry 29 Ž1990. 1551–1554. w16x Tramper, J., Vermue, ¨ M.H., Beeftink, H.H. and von Stockar, U., Progress in Biotechnology, Vol. 8: Biocatalysis in NonConventional Media, Elsevier, 1992. w17x Laane, C., Tramper, J. and Lilly, M.D., Studies in Organic Chemistry, Vol. 29, Biocatalysis in Organic Media, Elsevier, 1986. w18x H. Kuhn, ¨ T. Schewe, S.M. Rapoport, Adv. Enzymol. Relat. Areas Mol. Biol. 58 Ž1986. 273–311. w19x T. Schewe, S.M. Rapoport, H. Kuhn, ¨ Adv. Enzymol. Relat. Areas Mol. Biol. 58 Ž1986. 191–272. w20x Veldkin, G.A. and Vliegenthart, J.F.G., in: Eichhorn, G.L.,
w21x w22x w23x w24x w25x
w26x w27x w28x w29x w30x w31x w32x w33x w34x w35x w36x w37x w38x w39x
Marzili, L.G. ŽEds.., Advances in Inorganic Biochemistry, Vol. VI, pp. 139-161, Elsevier, Amsterdam, 1984. P. Redanna, J. Whelan, K.R. Maddipati, C. Channa Reddy, Meth. Enzymol. 187 Ž1990. 268–277. J.C. Boyington, B.J. Gaffney, L.M. Amzel, Science 260 Ž1993. 1482–1486. W. Minor, J. Steczko, T.J. Bolin, Z. Otwinowski, B. Axelrod, Biochemistry 32 Ž1993. 6320–6323. D.P. Cistola, J.A. Hamilton, D. Jackson, D.M. Small, Biochemistry 27 Ž1988. 1881–1888. Vick, B.A., Zimmerman, D.C., in: P.K. Stumpf ŽEd.., Lipids: Structure and Function, Vol. 9. Academic Press, New York, 1987. M.J. Schilstra, G.A. Veldkin, J. Verhagen, J.F.G. Vliegenthart, Biochemistry 31 Ž1992. 7692–7699. J.M. Lopez-Nicolas, F. Garcıa´ ´ R. Bru, ´ A. Sanchez-Ferrer, ´ ´ Carmona, Biochem. J. 308 Ž1995. 151–154. R. Bru, J.M. Lopez-Nicolas, Colloids ´ ´ F. Garcıa-Carmona, ´ Surf. 97 Ž1995. 263–269. E. Mulliez, J.P. Leblanc, J.J. Girerd, M. Rigaud, J.C. Chottard, Biochim. Biophys. Acta 916 Ž1987. 13–23. A. Chattopadhyay, E. London, Anal. Biochem. 139 Ž1984. 408–412. J. Serth, A. Lautwein, M. Frech, A. Wittinghofer, A. Pingoud, EMBO J. 10 Ž1991. 1325–1330. N. Jyothirmayi, C.S. Ramadoss, Biochim. Biophys. Acta 1083 Ž1991. 193–200. I. Singh, Y. Kishimoto, J. Lipid Res. 24 Ž1983. 662–665. S. Laakso, Biochim. Biophys. Acta 792 Ž1984. 11–15. J. Sekiya, H. Aoshima, T. Kajiwara, T. Togo, A. Hatanka, Agric. Biol. Chem. 41 Ž1977. 827–832. W.L. Smith, W.E.M. Lands, J. Biol. Chem. 247 Ž1972. 1038–1047. P.K. Pistorius, B. Axelrod, J. Biol. Chem. 251 Ž1976. 7144–7148. M.J. Schilstra, G.A. Veldkin, J.F.G. Vliegenthart, Biochemistry 32 Ž1993. 7686–7691. S. Yamamoto, Biochim. Biophys. Acta 1128 Ž1992. 117– 131.
Kinetic characteristics of the enzymatic conversion in presence of cyclodextrins: study of the oxidation of polyunsaturated fatty acids by lipoxygenase Jose´ M. Lopez-Nicolas, ´ ´ Roque Bru, Francisco Garcıa-Carmona ´
)
Departamento de Bioquimica y Biologıa ´ Molecular A, Facultad de Biologıa, ´ UniÕersidad de Murcia, Campus de Espinardo, E-30001 Murcia, Spain Received 6 February 1997; revised 8 April 1997; accepted 15 April 1997
Abstract The capability of cyclodextrins ŽCDs. to greatly enhance the solubility in water of poorly water-soluble substances makes them an ideal alternative system for studying the expression of enzyme activity with such substrates in aqueous solution. In order to evaluate the behaviour of the enzymes in presence of CDs a study of the lipoxygenase ŽLOX.-catalyzed oxidation of polyunsaturated fatty acids ŽPUFA. as model reaction has been carried out. This was done by using LOX from two different sources Žsoybean and potato tuber., at two pH values Ž6.3 and 9.0., with two substrates Žlinoleic acid and arachidonic acid. and three types of CD Ž b-CD, methyl-b-CD and monoglucosyl-b-CD.. PUFA have been shown to form inclusion complexes of 1:2 stoichiometry which are in equilibrium with free PUFA and free CD, thus complexation is governed by two equilibrium constants, K 1 and K 2 ŽJ.M. Lopez-Nicolas ´ ´ et al., Biochem. J. 308 Ž1995. 151–154; R. Bru et Ž . . al., Colloids Surf. 97 1995 263–269 . For the oxidation of PUFA by LOX in the presence of b-CD we propose a model in which free PUFA is the only effective substrate, thus the oxidation of the complexed substrate requires the previous dissociation of the complex. The equilibrium constants of complex formation are determined by both a physico-chemical and an enzymatic method. In spite of giving quite similar results, the second was proven to be more accurate so it was employed in further studies. CD was shown to slow down the reaction rate of LOX, specifically due to the increase of K m , Vmax remaining unchanged. That apparent inhibition is due to removal of effective Žfree. substrate in the form of inclusion complexes. This ‘sequestered’ substrate can, however, be converted since it is in equilibrium with the free. The feasibility of realizing a CD-mediated accurate control over the conversion rate is demonstrated in the experiment called ‘cyclodextrin assay’ in which the concentration of the free substrate is calculated by using the equilibrium constants of complex formation and setting the initial concentrations of total substrate and total CD. From the observation of the reaction progress curves in the conditions of the CD assay, we have studied some characteristic parameters of the oxidation of PUFA by LOX in this new medium, such as enzymatic activity, duration of linear product accumulation and the lag phase. q 1997 Elsevier Science B.V. Keywords: Lipoxygenase; Cyclodextrin; Polyunsaturated fatty acid; Cyclodextrin assay; ŽPotato.; ŽSoybean.
Abbreviations: FA, fatty acids; PUFA, polyunsaturated fatty acids; CD, cyclodextrin; LOX, lipoxygenase; LA, linoleic acid; AA, arachidonic acid; c.m.c., critical micellar concentration; DLPA, duration of linear product accumulation. ) Corresponding author. Fax: q34 68364147. E-mail: [email protected] 0005-2760r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 5 - 2 7 6 0 Ž 9 7 . 0 0 0 6 1 - 1
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
1. Introduction Cyclodextrins Ž CDs. are torus-shaped oligosaccharides made up of 6–8 glucopyranose units and originated by the enzymatic degradation of starch, through the action of CD-glucano-transferase w1x. The cavity is carpeted by hydrogen atoms and so has a rather hydrophobic nature unlike the outer surface of the molecule, in which the primary and secondary hydroxyl groups are exposed to the solvent, thus making the whole molecule highly water-soluble w2x. Poorly water-soluble compounds and hydrophobic moieties of amphiphilic molecules interact non-covalently with the CD cavity to form the so-called inclusion complexes, which are also highly watersoluble w3,4x. For instance, steroids w5–7x and longchain fatty acids Ž FA. w8–10x, which are examples of each type of molecule, increase their aqueous solubility by several orders of magnitude when complexed with CDs. As opposed to other solubility enhancers Ždetergents, water-miscible solvents, etc.., CDs produce individual molecules of the solute, thus preventing side effects due to aggregation phenomena. The ability of CDs to enhance the aqueous solubility of many different compounds w3,11x makes them an interesting alternative system for performing enzyme-catalyzed conversions of poorly water-soluble substrates w12–15x compared with organic media or two-phase systems w16,17x. The interest of using CDs in enzyme-catalyzed reactions is 3-fold: first, they are of technological interest as high loads of poorly water-soluble substrates can be achieved; second, amphiphilic fatty, waxy or even liquid apolar substrates can be handled easily as they become readily water-soluble amorphous or crystalline solids when forming inclusion complexes, and third, information on kinetic properties, which otherwise would be inaccessible or masked by the effect of additives Ždetergents, cosolvents, etc.. , becomes available. Since substrates become water-soluble, it is of particular interest to use soluble enzymes acting on hydrophobic substrates. A growing number of studies are devoted to performing bioconversions of commercially interesting CD-solubilized compounds but, in spite of general interest in the field of basic and applied enzymology, no attempt has been made to date to understand the fundamentals and to model the process of enzyme catalysis in the presence of CDs.
141
Our interest was to investigate how an enzymatic reaction proceeds in CD-containing aqueous medium and then to characterize the different environmental factors which may affect the reaction. Lipoxygenase Žlinoleate:O 2 oxidoreductase, EC 1.13.11.12. ŽLOX. was found to be a good model enzyme for our purpose. This is a group of dioxygenases, many of them cytosolic, which catalyze the hydroperoxydation of polyunsaturated fatty acids ŽPUFA. containing at least one 1,4-ŽZ,Z.pentadiene system in their structure, such as linoleic ŽLA., linolenic and arachidonic ŽAA. acids w18–20x. Depending on its origin, LOX displays a marked product specificity as regards the position attacked by O 2 in AA w21x; for example, soybean LOX-1 is a 15-LOX, potato tuber and mammalian leukocyte LOX are both 5-LOX, blood platelet LOX is a 12-LOX, etc. In addition, there are important structural differences among the different LOX w22,23x. Fatty acids ŽFA. in general do not form true solutions due to their amphiphilic character but they form colloidal aggregates above a certain critical concentration; depending on pH these can adopt different arrangements, including a separate oil phase, bilayers and micelles w24x. The tendency of FA to form these kind of aggregates makes them excellent candidates for solubilization in CDs as inclusion complexes. Most LOXs display optimum activity at pH values around neutrality w25x but in these conditions PUFAs are quite insoluble. Raising pH to 9.0 or 10.0 favours PUFA solubilization through the formation of soaps w24x, but kills LOX activity with a few exceptions, such as soybean LOX-1. For this reason the kinetic properties of LOX are well known at pH 9–10 w26x and practically unknown at neutral or slightly acid pH values. The aggregation of LA and acid AA in the presence of CDs has been investigated previously by us w27,28x, and it was demonstrated that one PUFA molecule may interact with up to two CDs molecules in solution. Such an interaction leads to a CD concentration-dependent increase of the PUFA c.m.c. since the complexes do not take part in the aggregation. The inclusion process of specific PUFA is characterized by two equilibrium constants, K 1 and K 2 , which can be affected by pH, temperature and CD type. In this paper we propose a model for enzyme catalysis in a CD medium and apply the theoretical
142
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
concepts to the LOX-catalyzed dioxygenation of PUFA with good agreement. The model was tested by using LOX from two sources Žpotato tubers and soybean., two substrates ŽLA and AA., at two pH values Ž6.3 and 9.0. and different CDs Ž b-CD, methyl-b-CD and monoglucosyl-b-CD..
2. Materials and methods LA and AA acid were purchased from Cayman Chemical Co. ŽParis, France.. b-CD, methyl-b-CD and LOX Žtype V. from soybean were obtained from Sigma ŽMadrid, Spain. . Monoglucosil-b-CD was purchased from Ensoiko Sugar Refining Co., Ltd. ŽYokohama, Japan.. LOX was purified from potato tubers Žvar. Desiree´. according to Mulliez et al. w29x Žspec. act. 27 m mol O 2rminrmg protein.. Diphenylhexatriene ŽDPHT. was a product from Fluka ŽMadrid. and tetrahydrofurane ŽTHF. was from Merck ŽDarmstadt, Germany.. All the other chemicals used were of the highest purity. 2.1. Preparation of FA-CD complexes FA-CD complex solution was prepared by dissolving the CD in 0.1 M potassium phosphate buffer pH 6.3 containing 1% vrv EtOH or potassium borate buffer pH 9 containing 1% vrv EtOH, followed by addition of the fatty acid prepared in the same buffer. The samples were flushed with N2 to prevent oxidation of FA during preparation. Just before using, samples were vigorously shaken in order to supply O 2 for the enzymatic reaction. 2.2. Determination of equilibrium constants by means of critical micellar concentration (c.m.c.) Although the term c.m.c. is only correct to describe the aggregation of micelle-forming amphiphiles, we will use this term throughout this work to refer also to the concentration at which any type of aggregates appears. The c.m.c. value of LA was determined by means of a fluorescence spectroscopic method originally developed to determine c.m.c. values of detergents w30x and adapted to determine the concentration at which aggregation of fatty acids occurs w27,31x. Briefly, 2-ml samples contained 0.1
Scheme 1. Scheme of an enzymatic reaction in the presence of CD forming 1:2 complexes with the substrate.
M potassium phosphate buffer ŽpH 6.3., 0.88 m M DPHT Žsupplied in 2 m l THF., 1% vrv ethanol and the indicated concentration of CD and PUFA. The samples were flushed with N2 and incubated for 30 min in the dark at the desired temperature for equilibration and in order to reverse the photoisomerisation of the fluorescent probe. Fluorescence intensity measurements were made at 258C at 430 nm Ž358 nm excitation wavelength. in a Kontron SFM-25 spectrofluorimeter equipped with thermostated cells. C.m.c. was determined graphically from a plot of fluorescence relative values versus PUFA concentration as the extrapolated intersection between the lines defining the fluorescence in the pre- and post-micellar regions. This determination makes it possible to set the conditions in which FA is molecularly dispersed and does not form aggregates. C.m.c data at different CD concentrations were fitted by non-linear regression to an equation as described by ourselves w27,28x to determine the equlibrium constants, K 1 and K 2 , of Scheme 1. 2.3. Enzyme assay Lipoxygenase activity was assayed by monitoring the increase in absorbance at 234 nm Ž 234 s 25 000 My1 cmy1 . of the forming hydroperoxides in a Kontron Uvikon 940 spectrophotometer at 258C equipped with thermostated cells. The reaction was started by adding 5 m l enzyme to 1 ml of complex FA-CD preparation. 2.4. Model of enzyme catalysis in the presence of CDs The proposed reaction scheme assumes that free S is the only form of substrate that LOX is able to bind
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
to in order to react. Accordingly, the reaction rate equation should be: Vs
Vmax P S
f
Kmq S
f
Ž1.
The concentration of free substrate can be deduced by analyzing the equilibrium of complex formation. The mass balance of the S and CD in an aqueous solution may be represented by Eqs. Ž 2. and Ž 3.:
the reciprocal form being the equation: 1 Km P K1 P K2 2 s P CD t V Vmax . S t Km P K1 q P CD t Vmax P S t Km 1 q q Vmax P S t Vmax
143
Ž9.
3. Results and discussion S t s S f q S y CD q S y CD 2
Ž2.
CD t s CD f q S y CD q 2 S y CD 2
Ž3.
where subscripts f and t stand for free and total respectively. K 1 and K 2 are the equilibrium constants, which are defined as follows: S y CD
K1 s
S
f
CD
Ž4. f
S y CD 2
K2s
S y CD CD
Ž5. f
If the concentration of the complex is negligible with respect to wCDx t , then wCDx f , wCDx t , and the amount of free, non-complexed substrate can be expressed after rearrangement of Eqs. Ž 2. , Ž 4. and Ž 5. as: S fs
S
t
1 q K 1 P CD t q K 1 P K 2 P CD
Ž6.
2 t
The above condition applies when wCDx t 4 w S x t or, with a wCDx t not much higher than w S x t , if the equilibrium constants are not very strong. Substituting Eq. Ž6. into Eq. Ž1. gives: Vs
Vmax . S K mapp q S
t
Ž7. t
where K mapp s K m P 1 q K 1 P CD t q K 1 P K 2 P CD
ž
2 t
/
Ž8.
The use of CDs in enzyme-catalyzed bioconversions of poorly-water-soluble compounds should be advantageous due to the hugely increased aqueous solubility of such compounds. This is the case with the cholesterol oxidase-catalyzed synthesis of 7a-hydroxycholest-4-en-3-one w12x in a high yield in the presence of hydroxypropyl-b-CD as compared to the chemical synthesis, or the phenoloxidase-catalyzed regiospecific hydroxylation of 17-b estradiol w15x. In spite of these promising results, in which CDs were used as substrate carriers, the role of CD in the process is still not well understood since in some cases CDs are described as inhibitors w32x and in others as enhancers of the enzyme activity w33x, or, at least, as not having a negative effect w34x. In this study we investigate the role of CD in the LOX-catalyzed dioxygenation of PUFA using as a basis the hypothetical reaction Scheme 1, which takes into account the previously demonstrated model of PUFA-CD interaction w27,28x. The reaction rate equation which is obtained ŽEq. Ž7.. indicates that CDs do not affect Vmax , while the observed K mapp ŽEq. Ž8.. is a function of CD concentration and the equilibrium constants for inclusion complex formation. Accordingly, if Vmax and K m in the absence of CDs as well as the equilibrium constants are known, the reaction rate can be predicted at any CD concentration. Only by knowing the kinetic constants, the equilibrium constants might be evaluated by n.l.r. fitting of experimental data to Eq. Ž 7. or Eq. Ž 9., and vice versa. Sometimes, and specially for substrates of low water solubility, the independent evaluation of the kinetic constants in the absence of CDs, and any other additive, may not always be possible because of the technical problem that low solubility represents.
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
144
In such cases, the equilibrium constants might be determined by different physico-chemical methods w3x and then used in the above equations to have access to the kinetic constants, specially to K m , by n.l.r. of data in the presence of CDs. In our experimental system all the parameters in Eqs. Ž7. – Ž9. can be obtained independently and thus be used to establish the corresponding comparisons. 3.1. Enzymatic method Õs. physico-chemical method for the determination of complex formation equilibrium constants So that Eq. Ž9. could be applied to the determination of the equilibrium constants, the Vmax and K m of soybean LOX using LA as substrate were determined by Lineweaver-Burk plots in the absence of CDs. The resulting values were 2.6 m Mrmin and 9.7 m M at pH 6.3 and 34.8 m Mrmin and 62 m M at pH 9.0. Then, an experiment was performed in which LOX activity was determined at different CD concentrations with the LA concentration kept constant. Soybean LOX activity decreased as CD concentration increased in each set of measurements made at a fixed LA concentration. This inhibitory behavior, as analyzed through the Dixon plot Ž Fig. 1. , cannot be explained easily since 1rv dependence on CD concentration is not linear. According to Eq. Ž9., such dependence is expected to be quadratic. Thus Eq. Ž 9. was used to obtain the equilibrium constants by n.l.r. fitting of these data, and using the previously determined kinetic parameters as constant coefficients. Results are shown in Table 1. Although the equilibrium constants are commonly determined by physico-chemical methods, the above method may be regarded as the enzymatic alternative for determining the equilibrium constants of complex formation. We have also used the method based on the CD-induced increase of PUFA c.m.c. w27,28x to
Fig. 1. Dixon plot of soybean-LOX catalyzed LA oxidation at pH 6.3 in the presence of b-CD at different LA concentrations: Žv . 7 m M; Ž%. 55.2 m M.
determine the equilibrium constants. As seen in Table 1, the values obtained by this latter method are quite similar to those obtained using the enzymatic method, and so both seem equally valid. Both methods agree in that LA-b-CD interaction is best described by two equilibrium constants as depicted in Scheme 1. Furthermore, the fact that both methods yield similar results reinforces some of the assumptions made, particularly that LOX only uses free PUFA as substrate and that the apparent inhibition of LOX activity by CD is due to reduction of the effective substrate. 3.2. CDs as the basis for a new type of assay: the CD assay On the basis of Scheme 1, it should be possible to finely control the concentration of free PUFA if the equilibrium constants are known, merely by setting
Table 1 Equilibrium constants of LA-b-CD inclusion complex determined by using physico-chemical and enzymatic methods Method Physico-chemical Enzymatic b a b
pH 6.3 a
According to Refs. w27,28x. See text for explanation.
pH 9
K 1 ŽmMy1 .
K 2 ŽmMy1 .
K 1 ŽmMy1 .
K 2 ŽmMy1 .
10.2 " 1.4 11.2 " 1.2
2.6 " 0.9 1.7 " 0.4
1.5 " 0.1 4.1 " 0.8
1.7 " 0.5 3.7 " 0.7
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
Fig. 2. ‘Cyclodextrin assay’: soybean LOX-catalyzed oxidation of LA reaction rates at 7 m M free LA concentration at pH 6.3. The b-CD and LA concentrations were calculated to yield 7 m M free LA using the set of constants determined enzymatically Žv . and the set of constants determined by physico-chemical method Ž`.. For a given b-CD concentration the total LA concentration was readily calculated by applying Eq. Ž6..
the appropriate concentrations of both total CD and total PUFA. In the so-called CD assay, the total CD and PUFA concentrations are set so that the free PUFA concentration remains constant. Consequently, the enzymatic reaction rate should be constant regardless of total CD and PUFA concentrations. By using each set of constants determined through either the enzymatic or the physico-chemical method, we have performed the corresponding CD assays. A concentration of free PUFA below its c.m.c. Ž10.6 m M at pH 6.3 and 170 m M at pH 9.0 for LA w27x. has to be chosen in order to avoid turbidity in spectrophotometric measurements. Fig. 2 shows the results obtained at pH 6.3. The set of constants determined enzymatically responded as expected in a CD assay Žthat is activity independent of total CD or PUFA concentration. but that obtained with the set of constants determined physico-chemically deviated from expectations. Thus, the small differences observed between the enzymatically and physico-chemically obtained set of constants ŽTable 1. seem to be very important as regards the actual effective PUFA
145
concentration. Indeed, if we assume the enzymatic set to be correct, the initial conditions set for the physico-chemical constants give lower concentrations of free PUFA than expected which would explain the decrease in LOX activity observed. At pH 9, the enzymatic set of constants also worked well, but the physico-chemical set failed once again to yield a LOX activity independent of total CD or PUFA concentration Žresults not shown.. Although such physico-chemical method, based on the fluorescence intensity break-point of a probe at the c.m.c. w30x, has been used to properly describe different features of the CD-PUFA interaction w27,28x, it does not seem to be accurate enough to provide equlibrium constants for the performance of the CD-assay. Indeed, it must be less accurate than the enzymatic method because in the former the n.l.r. to determine the equilibrium constants is applied to a secondary plot Žc.m.c. vs. CD plot proceeds from fluorescence vs. PUFA concentration plots., while in the enzymatic method it is applied to a primary plot, the enzyme activity vs. CD. In any case, the CD assay
Fig. 3. ‘Cyclodextrin assay’: potato LOX-catalyzed oxidation of LA reaction rates at 7 m M free LA concentration at pH 6.3. The b-CD and LA concentrations were calculated to yield 7 m M free LA using the set of constants determined enzymatically with soybean LOX Žv . and the set of constants determined by physico-chemical method Ž%.. For a given b-CD concentration the total LA concentration was readily calculated by applying Eq. Ž6..
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
146
Table 2 Kinetic parameters of potato-LOX in 0.1 M potassium phosphate buffer pH 6.3 containing 1% vrv ethanol in the absence or presence of b-CD w b-CDx ŽmM. 0 0.5 1 1.5
Vmax Ž m Mrmin. 4.6 4.5 4.7 4.4
K m Ž m M. Observed
Expected enzymatic.
Expected phys-chem.
3.5 20.6 100.0 211.0
3.5 39.9 108.0 212.0
3.5 45.1 135.0 269.0
The values of K m observed are experimental data and the values of K m enzymatic and phys-chem. are obtained applying Eq. Ž8. using equilibrium constants of LA-b-CD inclusion complex formation determined through either physico-chemical or enzymatic methods.
behaved better with the enzymatic constants due to interdependence between the enzyme and the constants. The accuracy of the enzymatic set of constants was further checked. The constants determined as above using soybean LOX at pH 6.3 were applied to perform the CD assay with potato LOX which has different enzymatic properties Žoptimum pH, product specificity, molecular weight, etc.. w29,35x. The physico-chemical constants were also applied for comparison purposes. As seen in Fig. 3, the results were qualitatively similar to those obtained with soybean LOX ŽFig. 2., thus confirming the accuracy of the enzymatic set of constants and the fact that they are independent of the enzyme being determined. Furthermore the kinetic parameters for potato LOX at pH 6.3 were determined experimentally in different CD conditions and the determined K m values were compared with those calculated from Eq. Ž8. for each set of equilibrium constants. As seen in Table 2, the results obtained by using the enzymatic set are more similar to those determined experimentally than those obtained using the physico-chemical set. Note that the experimental Vmax is independent of CD concentration, as predicted from Scheme 1. We conclude that a set of enzymatically determined substrate-CD equilibrium constants regardless of which enzyme is used, is sufficiently accurate to be used for the fine control of an enzyme-catalyzed reaction of the same CD-solubilized substrate, as demonstrated in a CD-assay. 3.3. Characteristics of the CD-based assay From a study of the reaction progress curves in the conditions of the CD assay some interesting observa-
Fig. 4. Time course of the soybean LOX-catalyzed oxidation of LA in the absence Ža. and presence Žb, c and d. of b-CD at pH 6.3 in the conditions of the CD assay. In Ža. 7 m M LA, Žb. the medium contained 0.25 mM of b-CD and 34.8 m M of LA, Žc. the conditions were 0.5 mM of b-CD and 79 m M of LA and in Žd. 1.25 mM of b-CD and 310 m M of LA. In all cases the free LA concentration was 7 m M.
tions can be made. Figs. 4 and 5 show the curves for the soybean and potato LOX-catalyzed dioxygenation of LA, respectively. 3.3.1. Duration of linear product accumulation (DLPA) This is defined as the longest time that product accumulation is linear. As seen in both figures, in the absence of CD the substrate which coincides with the free substrate Ž7 or 9 m M., is quickly exhausted. Increasing total substrate concentration can be in-
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
147
during the DLPA by replenishment form the PUFACD complex pool Ž Scheme 1.. Therefore, our results do not support a direct transformation of complexed PUFA as suggested by others authors w32x. As long as the initial conditions are not significantly distorted, for instance by depletion of the substrate or the presence of products, the product will accumulate at a constant rate. These two processes diminished at increasing CD concentrations.
Fig. 5. Time course of the potato LOX-catalyzed oxidation of LA in the absence Ža. and presence Žb, c and d. of b-CD at pH 6.3 in the conditions of the CD assay. In Ža. 9 m M LA, Žb. the medium contained 0.5 mM of b-CD and 102 m M of LA, Žc. the conditions were 0.75 mM of b-CD and 180 m M of LA and in Žd. 1 mM of b-CD and 279 m M of LA. In all cases the free LA concentration was 9 m M.
creased in the presence of CD although the free substrate concentration remaining invariable, leads to a longer DLPA Žfrom 248 s without b-CD up to 900 s in presence of 1.25 mM of b-CD for a LA free concentration of 7 m M. at the same constant rate. If the amount of substrate transformed in the presence of CD is greater than the amount of the free substrate but the reaction rate is that corresponding to the concentration of free substrate, then the reaction pathway would involve the binding of LOX alone to the free PUFA, whose concentration is kept constant
3.3.2. Lag phase The LOX-catalyzed dioxygenation of PUFA displays a characteristic kinetic lag phase w36x, which represents the activation of LOX by its products w37,38x. The lag phase is determined as the extrapolation of the tangent of maximal slope of the reaction progress curve to the ‘time axis’. Such a lag phase can be eliminated by supplementing the reaction medium with stoichiometric Žmicromolar. amounts of hydroperoxide product with respect to the enzyme. As seen in Fig. 5 the lag phase increases with the concentration of CD. Such an increase, which ranges from 17 s when no b-CD is present in the medium up to 109 s when the concentration of b-CD is 1.25 mM, may be due to the competition between the free CD and the enzyme to bind the reaction product, the hydroperoxide. Thus a higher total concentration of hydroperoxide has to be accumulated for the enzyme to be activated by the free hydroperoxide, assuming that this product species is the only that binds the enzyme, as happens in the case of free substrate. In the case of soybean LOX at pH 6.3, the lag phase was practically nil at the enzyme concentration used.
Table 3 ‘Cyclodextrin assay’ wLAx t Ž m M.
w b-CDx t Ž m M.
Activity Ž m Mrmin.
DLPA Žs.
Lag Žs.
82.27 102.24 160.78 295.12 485.29
0 50 150 300 450
22.00 23.01 21.96 20.93 22.00
55 83 120 195 232
0 6 14 31 55
a
Soybean-LOX catalyzed oxidation of LA reaction at pH 9.0. The b-CD and LA concentrations were calculated to yield 82.3 m M free LA using the set of constants determined enzymatically. Data of activity, DLPA and Lag were obtained from the curves of reaction progress. a Too short to be accurately measurable.
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
148
Table 4 Equlibrium constants of LA-CD inclusion complex at different type of CDs y1 .
K 1 ŽmM K 2 ŽmMy1 .
b-CD
methyl-b-CD
monoglucosyl-b-CD
11.16 " 1.20 1.68 " 0.13
14.79 " 1.41 0.18 " 0.01
7.28 " 0.82 1.16 " 0.23
The set of constants was determined by using enzymatic method with soybean-LOX at pH 6.3.
3.4. Factors that affect the CD-based assay The CD assay relies on the interaction of a particular substrate with a particular CD in a particular pH and temperature environment. Since all these factors affect the equilibrium constants, as we previously demonstrated w27,28x, a new set of constants has to be determined in every new condition we wish to run a CD-assay. Thus, the corresponding tests have been performed at another pH, with another type of CD and with another substrate. 3.4.1. pH All the above experiments previous CD assay experiments were made at pH 6.3, so another CD assay was performed using the new set of constants determined with soybean LOX at pH 9.0 Ž Table 1., a pH at which potato LOX is inactive w29x. As seen in Table 3, reaction rate was constant over the range of b-CD concentration tested and the DLPA and lag phase increased as b-CD concentration increased. Note that a quite high concentration of free LA was necessary for lag phases to be measured. 3.4.2. Type of CD All the above experiments were performed with natural b-CD. In further experiments two chemically
modified CDs, namely methyl-b-CD and monoglucosyl-b-CD, were tested for use in the CD-based assay with potato LOX at pH 6.3. The set of constants obtained using soybean LOX ŽTable 4. was used to carry out the respective CD assays. K 1 in the modified CDs is significantly larger than K 2 , as was the case happens with the natural b-CD, and so hydrogen binding still seems to be the most important contribution to the stability of the inclusion complexes for these news CDs w27,28x. In all cases, as can be seen in Table 5 and Fig. 6 the activity remained constant, and the DLPA and lag phase increased under CD-assay conditions. Therefore the CD assay holds with the corresponding sets of constants for both methyl-b-CD and monoglucosyl-b-CD and soybean and potato LOXs. The characteristic increase of the DLPA points to the degree of replenishment of free substrate also for these modified CDs, too. 3.4.3. Type of substrate In addition to LA we also tested AA, which is the natural substrate of mammalian LOX and whose products are involved in a variety of important physiological functions w39x. The procedure is same as at followed for LA. Firstly kinetic parameters in the absence of b-CD
Table 5 ‘Cyclodextrin assay’ wmethyl-b-CD x ŽmM.
Activity Ž m Mrmin.
DLPA Žs.
wmonoglucosyl-b-CDx ŽmM.
Activity Ž m Mrmin.
DLPA Žs.
0 0.15 0.25 0.35 0.5 0.75 1
2.80 2.73 2.74 2.50 2.80 2.97 2.87
34 80 126 190 230 266 295
0 0.15 0.25 0.35 0.5 0.75 1
2.80 2.60 2.96 3.05 2.79 3.01 2.74
25 54 66 72 107 136 211
Potato-LOX catalyzed oxidation of LA reaction with two types of CD at pH 6.3. The methyl-b-CD, monoglucosyl-b-CD and LA concentrations were calculated to yield 5 m M free LA using the set of constants determined enzymatically. Data of activity and DLPA were obtained from the curves of reaction progress.
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
149
were determined Ž K m s 13.31 m M and Vmax s 1.56 m Mrmin.. These were then used in Eq. Ž9. to fit the activity vs. b-CD curves and thereafter determine the equilibrium constants. In this case the following results were obtained K 1 s 16.7 " 3.4 mMy1, K 2 s 1.55 " 0.5 mMy1. Soybean LOX was used for these determinations. The result of the CD assay using this set of constants applied to both soybean and potato LOXs is shown in Fig. 7. These result agree qualitatively with those obtained for LA, indicating the similar behaviour of both substrates with respect to the proposed reaction pathway ŽScheme 1..
4. Concluding remarks Fig. 6. ‘Cyclodextrin assay’. Soybean LOX Že,I. and potato LOX ŽB,l.-catalyzed oxidation of LA reaction rates at 5 m M free LA concentration at pH 6.3 by using different types of CDs: ŽB,I. monoglucosyl-b-CD and Žl,e. methyl-b-CD. Different enzyme concentrations were used in order to avoid overlapping of data for presentation purposes.
We have demonstrated by several means that the reaction pathway of Scheme 1 adequately describes the LOX-catalyzed dioxygenation of PUFA in the presence of CD. This was done by using LOX from two different sources, at two pH values, with two substrates and three types of CD. As result of this study, we have designed the CD assay that makes it possible to control the course of an enzyme catalyzed reaction in the presence of CDs. The CD assay explains the enzyme apparent inhibition by CD and is of general interest for enzyme catalyzed bioconversions in the presence of CDs. Eventually, it should be possible that substrate-CD combinations other than PUFA-CD might be enzymatically transformed as a complex itself if bulky reactive parts of the substrate remain exposed. The model in Scheme 1 should then be adapted accordingly. This would open up new exciting possibilities of the CD-facilitated bioconversions for exploiting the enzymes’ capabilities such as its stereo and regiospecificity.
Acknowledgements
Fig. 7. ‘Cyclodextrin assay’. Soybean-LOX Ž%. and potato-LOX ŽB. catalyzed oxidation of AA reaction rates at 10 m M free AA concentration at pH 6.3. The b-CD and AA concentrations were calculated to yield 10 m M free AA using the set of constants determined enzymatically.
This work has been supported in part by research grants form CICYT ŽProyecto BIO94-0541., INTAS ŽGrant 93r2223 EXT. and Consejerıa ´ de Educacion ´ y Cultura. C.A. Murcia Žproyecto PIB95r03.. J.M.L.N. is a holder of a grant of CajaMurcia. R.B. holds a contract for Doctores Reincorporados.
150
J.M. Lopez-Nicolas ´ ´ et al.r Biochimica et Biophysica Acta 1347 (1997) 140–150
References w1x K. Haga, K. Harata, A. Nakamura, K. Yamane, J. Mol. Biol. 237 Ž1994. 163–164. w2x W. Saenger, Angew. Chem. Int. Ed. Engl. 19 Ž1980. 344– 362. w3x Szejtli, J., Cyclodextrin Technology, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1988 w4x J. Szejtli, Med. Res. Rev. 14 Ž1994. 353–386. w5x F. Acarturk, ¨ T. Imai, H. Saito, M. Ishikawa, M. Otagiri, J. Pharm. Pharmacol. 45 Ž1993. 1028–1032. w6x K. Uekama, T. Fujinaga, F. Hirayama, M. Otagiri, M. Yamasaki, Int. J. Pharm. 10 Ž1982. 1–15. w7x K. Koizumi, Y. Okada, Y. Kubota, T. Utamura, Chem. Pharm. Bull. 35 Ž1987. 3413–3418. w8x H. Schlenk, D.M. Sand, J. Am. Chem. Soc. 83 Ž1961. 2312–2320. w9x N. Jyothirmayi, C. Ramadoss, S. Divakar, J. Agric. Food Chem. 39 Ž1991. 2123–2127. w10x R. Bergeron, Y. Machida, K. Bloch, J. Biol. Chem. 250 Ž1975. 1223–1270. w11x Duchene, ˆ D., Žed.. New Trends in Cyclodextrins and Their Derivaties. Editions de Sante´ ŽParis., 1991. w12x D.L. Alexander, J.F. Fisher, Steroids 60 Ž1995. 290–294. w13x J. Jadoun, R. Bar, Appl. Microbiol Biotechnol. 40 Ž1993. 477–482. w14x B. Pekic, Z. Lepojevic, Biotechnol. Lett. 13 Ž1991. 399–404. w15x H.J. Woerdenbag, N. Pras, H.W. Frijlink, C.F. Lerk, T.M. Malingre, ´ Phytochemistry 29 Ž1990. 1551–1554. w16x Tramper, J., Vermue, ¨ M.H., Beeftink, H.H. and von Stockar, U., Progress in Biotechnology, Vol. 8: Biocatalysis in NonConventional Media, Elsevier, 1992. w17x Laane, C., Tramper, J. and Lilly, M.D., Studies in Organic Chemistry, Vol. 29, Biocatalysis in Organic Media, Elsevier, 1986. w18x H. Kuhn, ¨ T. Schewe, S.M. Rapoport, Adv. Enzymol. Relat. Areas Mol. Biol. 58 Ž1986. 273–311. w19x T. Schewe, S.M. Rapoport, H. Kuhn, ¨ Adv. Enzymol. Relat. Areas Mol. Biol. 58 Ž1986. 191–272. w20x Veldkin, G.A. and Vliegenthart, J.F.G., in: Eichhorn, G.L.,
w21x w22x w23x w24x w25x
w26x w27x w28x w29x w30x w31x w32x w33x w34x w35x w36x w37x w38x w39x
Marzili, L.G. ŽEds.., Advances in Inorganic Biochemistry, Vol. VI, pp. 139-161, Elsevier, Amsterdam, 1984. P. Redanna, J. Whelan, K.R. Maddipati, C. Channa Reddy, Meth. Enzymol. 187 Ž1990. 268–277. J.C. Boyington, B.J. Gaffney, L.M. Amzel, Science 260 Ž1993. 1482–1486. W. Minor, J. Steczko, T.J. Bolin, Z. Otwinowski, B. Axelrod, Biochemistry 32 Ž1993. 6320–6323. D.P. Cistola, J.A. Hamilton, D. Jackson, D.M. Small, Biochemistry 27 Ž1988. 1881–1888. Vick, B.A., Zimmerman, D.C., in: P.K. Stumpf ŽEd.., Lipids: Structure and Function, Vol. 9. Academic Press, New York, 1987. M.J. Schilstra, G.A. Veldkin, J. Verhagen, J.F.G. Vliegenthart, Biochemistry 31 Ž1992. 7692–7699. J.M. Lopez-Nicolas, F. Garcıa´ ´ R. Bru, ´ A. Sanchez-Ferrer, ´ ´ Carmona, Biochem. J. 308 Ž1995. 151–154. R. Bru, J.M. Lopez-Nicolas, Colloids ´ ´ F. Garcıa-Carmona, ´ Surf. 97 Ž1995. 263–269. E. Mulliez, J.P. Leblanc, J.J. Girerd, M. Rigaud, J.C. Chottard, Biochim. Biophys. Acta 916 Ž1987. 13–23. A. Chattopadhyay, E. London, Anal. Biochem. 139 Ž1984. 408–412. J. Serth, A. Lautwein, M. Frech, A. Wittinghofer, A. Pingoud, EMBO J. 10 Ž1991. 1325–1330. N. Jyothirmayi, C.S. Ramadoss, Biochim. Biophys. Acta 1083 Ž1991. 193–200. I. Singh, Y. Kishimoto, J. Lipid Res. 24 Ž1983. 662–665. S. Laakso, Biochim. Biophys. Acta 792 Ž1984. 11–15. J. Sekiya, H. Aoshima, T. Kajiwara, T. Togo, A. Hatanka, Agric. Biol. Chem. 41 Ž1977. 827–832. W.L. Smith, W.E.M. Lands, J. Biol. Chem. 247 Ž1972. 1038–1047. P.K. Pistorius, B. Axelrod, J. Biol. Chem. 251 Ž1976. 7144–7148. M.J. Schilstra, G.A. Veldkin, J.F.G. Vliegenthart, Biochemistry 32 Ž1993. 7686–7691. S. Yamamoto, Biochim. Biophys. Acta 1128 Ž1992. 117– 131.