218
Biochimica et Biophy~ca Acta, I 118 (1992) 218-222 1992 Elsevier Science Publishers B.V. All rights reseraed 0167-4838/92/$05.00
BBAPRO34072
Reaction rate with suspended lipase catalyst shows similar dependence on water activity in different organic solvents Rao H. Valivety i, Peter J. Hailing ! and Alasdair R. Macrae 2 t Department of Bioscience and Bioteclmolo~; Unit¢rsio"of Strathclyde, Glasgow (U.K ~and -"Unilerer Research. Colworth Laborator*.; Sharnbrook, Bedford (U.K) (Received 8 April 1991)
Key words: Water activity: Organic mlx'ent: Lipase actk-/t$."
We have studied the effect of thermodynamic ~later activity ( a . ) on the initial rate of esterilication catalysed by an immobilised lipose (Lipozyme) suspended in an organic rea..~.'on mixture. The catalyst and the organic phase were separately pre-equilibrated to the same aw value. The rate shows similar dependence on a,, in reaction mixtures based on five different organic solvents ranging in polarity from pentan-3-on¢ to hexane, and in a liquid reactant mixture. There is a maximum at a , about 0.5, with a decline to 30--70% at a,, of either 0.9 or less than 0.01. When the rates are presented in terms of water concentration in the organic phase (or total water content of the system), the maxima for the various solvents come at very different positions, reflecting the widely varying solubilities of water in the organic phase.
Introduction The activity of biocatal~ts in mainly organic reaction mixtures is usually greatly affected by the le:el of water remaining. The reaction rate is low or zero after exhaustive drying, but increases as slightly more water is present. Often it eventually reaches a maximum, and then falls again. The initial rise is considered to reflect a small quantity of water essential for enzyme activity, while the later decline is often due to catalyst particles becoming clumped together, reducing interfaciai area and limiting mass transfer. The variation in activity is usually present,.'d as a function of the water content of the whole reaction mixture. However, this is not a very fundamental parameter, as the water is always distributed between at least two distinct phases. Some is dissolved in the organic phase, as individual molecules or small oligomers, while some is present in a more polar phase around the biocatalyst (which may be a dilute aqueous solution). Without changing the compositions of the phases, alterations in their relative volumes will affect the total water content, but not the biocatalyst mi-
Correspondence: PJ. Hailing. Department of Bioscience and Biotechnology.. Royal College Building. University of Strathclyde. Glasgow G1 IXW. U.K.
croenvironment or behaviour. Hence it is not surprising that the critical water contents for optimal activity (or other properties) vary widely as other things are changed, such as the enzyme, the support or the solvent. The critical water contents are higher in more polar solvents, and the explanation stems from the higher solubility of water in these. The most common presentation is that these solvents have more tendency to 'strip essential water from the enzyme' [1]; an alternative way of looking at the same process is that higher water contents are required in these solvents to reach the same thermodynamic water actbAty (a,,) [2]. Further understanding came from the elegant studies of Zaks and Kl~anov [1], based on measurements of the water content of the enzyme particles (rather than that of the whoie system). They showed that the dependence of catalytic activity on the former was very similar in a range of solvents (though of coarse the critical values of the total water content varied widely). Subsequently it was argued that the water bound by the enzyme was likely to be a function of the a w of the system [2], and an analysis of published water adsorption isotherms supported this view [3]. As a consequence, the a,, of the system should be a good predictor of the reaction rate, at least as affected by changing the solvent [2]. This would add a further attraction to the use of a,, to characterise these mainly organic
219 reaction mixtures. The a,, of the system is relatively easily measured via an equilibrated gas phase (it is by definition equal in all phases at equilibrium). It also determines the mass action effect of water on the equilibrium position of hydrolytic side-reactions, and offers a guide to the nature of the aqueous/l:mlar phase around the biocatalyst, even when this is of very small volume. As well as our previous work (cited in Ref. 2), measurement and/or control of a,,, in mainly organic enzyme reaction mixtures has also been reported by Goderis et al. [4], Goldberg et al. [5-7], Kang and Rhee [81. We now report a direct experimental test of the relationship between reaction rate and a , as the water content is varied in reaction mixtures based on a range of different organic solvents. The activity of a suspended lipase (triacylglycerol acylhydrolase, E.C. 3.1.1.3) catalyst is indeed found to show a similar dependence on aw in the various solvents, even though the water concentrations vary widely.
Experimental procedures The organic phase of the reaction mixture consisted of 0.5 M dodecanol (Fisons SLR) and 0.5 M decanoic acid (Aldrich 99 + %) dissolved in the appropriate solvent; or an equimolar liquid mixture of the two reactants (2.4 M in each). The catalyst was Mucor miehei lipase immobilised to anion exchange resin beads, trade name Lipozyme IM 20 and generously given by Novo-NordisL The catalyst was dried over molecular sieve, then re-equilibrated as described below. Gas phase water vapour pressures were measured using the Philips LiCI humidity sensor, as described previously [9]. Both the organic phase and the catalyst were adjusted to the desired a,, before starting the reaction. Normally this was done by equilibration at 20°C through the valmur phase in separate sealed containers with saturated salt solutions: LiCI (a,,. 0.12), MgCI2- 6H20 (0.32), Mg(NO3)2" 6H 20 (0.55), KCI (0.86) or ZnSO47H 2° (0.90); or with the solid phases: molecular sieve (a,, < 0.01) or mixed Na,SO4" 10HzO + Na2SO~ (0.76). This method was first used for enzymic reaction mixtures by Goderis et al. [4]. Equilibration of organic phases was monitored by water analysis using a Metrohm 684 KF coulometric Karl Fischer apparatus. Care was taken to wash thoroughly the transfer syringe with the sample, until reproducible readings were obtained; otherwise water adsorption to or dissolution from the walls of the syringe can significantly change the water content. Equilibration normally took less than 1 day. For the organic phase without added solvent, however, at least 2 weeks was required. After this period
about 2.5% of ester had been produced by chemical reaction, which would have interfered with accurate measurement of the initial enzymic rate. Therefore, the reaction mixture was prepared by mixing samples of the acid and alcohol equilibrated separately to known water contents, so as to give a total water concentration in the mixture corresponding to the appropriate aw (taken from the solubility curve obtained in the normal way; the ester formed does not seem to significantly affect this, as measurement of head space a W above the prepared organic phase confirmed the correct value). Equilibration of the Lipozyme catalyst was monitored by weighing, and followed roughly exponential kinetics, with a half-time of about 0.5 day; between 7 and 10 days was allowed. The reaction was started by mixing the pre-equilibrated phases, 5 ml organic phase and about 0.1 g catalyst (dry basis), at 20°C. The mixture was stirred magnetically at about 250 ~m, which kept the catalyst fully suspended. Samples (50 ~1) of the organic phase were removed (stopping stirring for 5 s), 300 ~1 of internal standard (5 mM methyl palmitate (Aldrich 99 + %) in tetrahydrofuran) was added, ~nd they were analysed by GLC without derivatisation. A PerkinElmer 8600 with a flame ionisation detector and a 8300 autosampler was fitted with a 2 m × 3 mm internal diameter column, packed with 3% OV-I on Chromosorb WHP, and operated isothermally at 220°C with an N~ carrier gas flow rate of 30 ml/min. Under these conditions the acid and alcohol both had retention times of less than 1 rain, the internal standard came at about 2.1 rain, and the product ester at 6.8 rain.
Results and Discussion Proper measurement of catalyst activity as a function of aw requires that aw is known and constant during the period required for measurement of the initial reaction rate. At first sight, measurement of an esterification reaction, in which water is formed as a product, seems an unwise choice. However, with a lipase-catalysed transesterification, a small amount of hydrolysis is usually found as a side reaction; and if the progress is analysed in detail, a hydrolysis equilibrium is rapidly established during the initial stages, with consumption of a significant fraction of the total water content and hence presumably change in aw. Therefore, it is probably better to study an esterifieation, and measure the initial rate over a period in which the net production of water is not significant. In our experiments, the maximum water formed was 6% of the total dissolution capacity of the organic phase. pH changes are unlikely to be significant with the catalyst and reactants used [9].
220
v M
,z P_,
.
1.0
WATER ACTIVITY Fig. I. Solubilitycurves for water in organic reaction mixtures based on various solvents.The .solventspresent ~'ere hexane ( ~ ). toluene (×). trichlort, thyIene (
Biochimica et Biophy~ca Acta, I 118 (1992) 218-222 1992 Elsevier Science Publishers B.V. All rights reseraed 0167-4838/92/$05.00
BBAPRO34072
Reaction rate with suspended lipase catalyst shows similar dependence on water activity in different organic solvents Rao H. Valivety i, Peter J. Hailing ! and Alasdair R. Macrae 2 t Department of Bioscience and Bioteclmolo~; Unit¢rsio"of Strathclyde, Glasgow (U.K ~and -"Unilerer Research. Colworth Laborator*.; Sharnbrook, Bedford (U.K) (Received 8 April 1991)
Key words: Water activity: Organic mlx'ent: Lipase actk-/t$."
We have studied the effect of thermodynamic ~later activity ( a . ) on the initial rate of esterilication catalysed by an immobilised lipose (Lipozyme) suspended in an organic rea..~.'on mixture. The catalyst and the organic phase were separately pre-equilibrated to the same aw value. The rate shows similar dependence on a,, in reaction mixtures based on five different organic solvents ranging in polarity from pentan-3-on¢ to hexane, and in a liquid reactant mixture. There is a maximum at a , about 0.5, with a decline to 30--70% at a,, of either 0.9 or less than 0.01. When the rates are presented in terms of water concentration in the organic phase (or total water content of the system), the maxima for the various solvents come at very different positions, reflecting the widely varying solubilities of water in the organic phase.
Introduction The activity of biocatal~ts in mainly organic reaction mixtures is usually greatly affected by the le:el of water remaining. The reaction rate is low or zero after exhaustive drying, but increases as slightly more water is present. Often it eventually reaches a maximum, and then falls again. The initial rise is considered to reflect a small quantity of water essential for enzyme activity, while the later decline is often due to catalyst particles becoming clumped together, reducing interfaciai area and limiting mass transfer. The variation in activity is usually present,.'d as a function of the water content of the whole reaction mixture. However, this is not a very fundamental parameter, as the water is always distributed between at least two distinct phases. Some is dissolved in the organic phase, as individual molecules or small oligomers, while some is present in a more polar phase around the biocatalyst (which may be a dilute aqueous solution). Without changing the compositions of the phases, alterations in their relative volumes will affect the total water content, but not the biocatalyst mi-
Correspondence: PJ. Hailing. Department of Bioscience and Biotechnology.. Royal College Building. University of Strathclyde. Glasgow G1 IXW. U.K.
croenvironment or behaviour. Hence it is not surprising that the critical water contents for optimal activity (or other properties) vary widely as other things are changed, such as the enzyme, the support or the solvent. The critical water contents are higher in more polar solvents, and the explanation stems from the higher solubility of water in these. The most common presentation is that these solvents have more tendency to 'strip essential water from the enzyme' [1]; an alternative way of looking at the same process is that higher water contents are required in these solvents to reach the same thermodynamic water actbAty (a,,) [2]. Further understanding came from the elegant studies of Zaks and Kl~anov [1], based on measurements of the water content of the enzyme particles (rather than that of the whoie system). They showed that the dependence of catalytic activity on the former was very similar in a range of solvents (though of coarse the critical values of the total water content varied widely). Subsequently it was argued that the water bound by the enzyme was likely to be a function of the a w of the system [2], and an analysis of published water adsorption isotherms supported this view [3]. As a consequence, the a,, of the system should be a good predictor of the reaction rate, at least as affected by changing the solvent [2]. This would add a further attraction to the use of a,, to characterise these mainly organic
219 reaction mixtures. The a,, of the system is relatively easily measured via an equilibrated gas phase (it is by definition equal in all phases at equilibrium). It also determines the mass action effect of water on the equilibrium position of hydrolytic side-reactions, and offers a guide to the nature of the aqueous/l:mlar phase around the biocatalyst, even when this is of very small volume. As well as our previous work (cited in Ref. 2), measurement and/or control of a,,, in mainly organic enzyme reaction mixtures has also been reported by Goderis et al. [4], Goldberg et al. [5-7], Kang and Rhee [81. We now report a direct experimental test of the relationship between reaction rate and a , as the water content is varied in reaction mixtures based on a range of different organic solvents. The activity of a suspended lipase (triacylglycerol acylhydrolase, E.C. 3.1.1.3) catalyst is indeed found to show a similar dependence on aw in the various solvents, even though the water concentrations vary widely.
Experimental procedures The organic phase of the reaction mixture consisted of 0.5 M dodecanol (Fisons SLR) and 0.5 M decanoic acid (Aldrich 99 + %) dissolved in the appropriate solvent; or an equimolar liquid mixture of the two reactants (2.4 M in each). The catalyst was Mucor miehei lipase immobilised to anion exchange resin beads, trade name Lipozyme IM 20 and generously given by Novo-NordisL The catalyst was dried over molecular sieve, then re-equilibrated as described below. Gas phase water vapour pressures were measured using the Philips LiCI humidity sensor, as described previously [9]. Both the organic phase and the catalyst were adjusted to the desired a,, before starting the reaction. Normally this was done by equilibration at 20°C through the valmur phase in separate sealed containers with saturated salt solutions: LiCI (a,,. 0.12), MgCI2- 6H20 (0.32), Mg(NO3)2" 6H 20 (0.55), KCI (0.86) or ZnSO47H 2° (0.90); or with the solid phases: molecular sieve (a,, < 0.01) or mixed Na,SO4" 10HzO + Na2SO~ (0.76). This method was first used for enzymic reaction mixtures by Goderis et al. [4]. Equilibration of organic phases was monitored by water analysis using a Metrohm 684 KF coulometric Karl Fischer apparatus. Care was taken to wash thoroughly the transfer syringe with the sample, until reproducible readings were obtained; otherwise water adsorption to or dissolution from the walls of the syringe can significantly change the water content. Equilibration normally took less than 1 day. For the organic phase without added solvent, however, at least 2 weeks was required. After this period
about 2.5% of ester had been produced by chemical reaction, which would have interfered with accurate measurement of the initial enzymic rate. Therefore, the reaction mixture was prepared by mixing samples of the acid and alcohol equilibrated separately to known water contents, so as to give a total water concentration in the mixture corresponding to the appropriate aw (taken from the solubility curve obtained in the normal way; the ester formed does not seem to significantly affect this, as measurement of head space a W above the prepared organic phase confirmed the correct value). Equilibration of the Lipozyme catalyst was monitored by weighing, and followed roughly exponential kinetics, with a half-time of about 0.5 day; between 7 and 10 days was allowed. The reaction was started by mixing the pre-equilibrated phases, 5 ml organic phase and about 0.1 g catalyst (dry basis), at 20°C. The mixture was stirred magnetically at about 250 ~m, which kept the catalyst fully suspended. Samples (50 ~1) of the organic phase were removed (stopping stirring for 5 s), 300 ~1 of internal standard (5 mM methyl palmitate (Aldrich 99 + %) in tetrahydrofuran) was added, ~nd they were analysed by GLC without derivatisation. A PerkinElmer 8600 with a flame ionisation detector and a 8300 autosampler was fitted with a 2 m × 3 mm internal diameter column, packed with 3% OV-I on Chromosorb WHP, and operated isothermally at 220°C with an N~ carrier gas flow rate of 30 ml/min. Under these conditions the acid and alcohol both had retention times of less than 1 rain, the internal standard came at about 2.1 rain, and the product ester at 6.8 rain.
Results and Discussion Proper measurement of catalyst activity as a function of aw requires that aw is known and constant during the period required for measurement of the initial reaction rate. At first sight, measurement of an esterification reaction, in which water is formed as a product, seems an unwise choice. However, with a lipase-catalysed transesterification, a small amount of hydrolysis is usually found as a side reaction; and if the progress is analysed in detail, a hydrolysis equilibrium is rapidly established during the initial stages, with consumption of a significant fraction of the total water content and hence presumably change in aw. Therefore, it is probably better to study an esterifieation, and measure the initial rate over a period in which the net production of water is not significant. In our experiments, the maximum water formed was 6% of the total dissolution capacity of the organic phase. pH changes are unlikely to be significant with the catalyst and reactants used [9].
220
v M
,z P_,
.
1.0
WATER ACTIVITY Fig. I. Solubilitycurves for water in organic reaction mixtures based on various solvents.The .solventspresent ~'ere hexane ( ~ ). toluene (×). trichlort, thyIene (
