Biotechnol. ROQ. 1994, io. 187-192
187
Solvent Selection and Productivity in Multiphase Biotransformation Systems Harish K. Santhanam and Gina S. Shreve*J Department of Chemical Engineering, The University of Illinois at Chicago, Chicago, Illinois 60607 The ll-@ hydroxylation of the steroid precursor Reichstein’s substance S (cortexolone) to hydrocortisone by Curvularia lunata cells is used as a model system for investigating the basis of the effect of solvent choice on multiphase biotransformation reaction kinetics. The effect of solvent choice and phase ratio on cellular toxicity and productivity is examined in batch suspension cultures. In general, the greater the miscibility of the solvent with the aqueous phase, the more cellular toxicity it exhibits. However, the presence of solvent a t concentrations above the solubility limit showed no increased toxic effect over that achieved with saturation concentrations. Toxicity experiments indicate that of the solvents examined octane is the least toxic to the Curuularia lunata substance S biotransformation system under consideration. The production of hydrocortisone is determined to be approximately 20 times greater during the growth phase than during the stationary phase. The phase ratio of the organic solvent added has a significant effect on the productivity of the system. It was found that there is an increase in productivity with an increase in the phase ratio of the organic solvent added. Reaction kinetics are examined, and the increase in productivity with increasing amounts of organic solvent is attributed to the increase in concentration of the steroid reactant in the system.
Introduction Multiphase biotransformation systems have been demonstrated to enhance the reaction kinetics and yield of numerous reactions involving reactants of low aqueous solubility, such as steroid compounds (Buckland et al., 1975; Fukui and Tanaka, 1981). The main advantages of an aqueous/organic two-phase biotransformation system over a single aqueous phase system are that the organic phase serves as reservoir for reactants that are soluble to a very small extent in the aqueous phase and that the differing solubilities of the reactant and the product in the aqueous or organic phase may be exploited to achieve product separation. Increases in productivity have been noted for partially miscible systems and attributed to an increase in reactant concentration in the aqueous phase of the system (Fukui and Tanaka, 1981). This increase in reactant concentration in the aqueous phase is generally explained as being due to the cosolvencyeffect of the added organic phase on solubilization of the steroid substrate (Buckland et al., 1975). Hence, such multiphase systems may be useful for numerous applications involving reactants of low aqueous phase solubility, such as the biosynthesis of certain pharmaceutics, dyes, and flavorings. Since purification of such molecules from fermentation broths typically involves extraction of the fermentation broth with organic solvent, multiphase reaction schemes also offer the opportunity to achieve simultaneous reaction and product separation. Differential reactant and product solubilities may also play an important role in such biotransformation processes by making it possible to increase reaction rates and productivities by continuous product removal from reversible reaction systems (Holst et al., 1987; Lilly et al., 1987). This is especially true in Corresponding author. + Current address: Dept. of Chemical Engineering, Wayne State University, Detroit, MI 48202. 875&7938/94/3010-0187$04.50/0
the case of reaction systems where the product acts as an inhibitor. Previous work by Brink and Tramper (1985) on the effect of solvent toxicity on two-phase biocatalysis demonstrated that the presence of an organic phase offers certain other advantages, including a reduced incidence of microbial contamination and a reduction in substrate and/or product inhibition as a consequence of lower inhibitor concentration in the aqueous environment of the cells. The primary criteria influencing the selection of multiphase reaction systems are differential reactant and product solubilities and the effect of the solvent on microbial viability and productivity. Brink and Tramper (1985) found that the main drawback of introducing the organic solvent in a suspension cell culture was the reduction of microbial activity by the solvent. They then demonstrated that immobilization of the cells reduced the solvent toxicity. Laane et al. (1985)observed the same toxic effect of the solvent on microbial activity and related the logarithm of the solvent partition coefficient to its toxicity. They concluded that solvents possessing an octanol-water partition coefficient of greater than 4 were less toxic to microbial cells in suspension culture. Solvents with octanol-water partition coefficients under 2-4 were found to show increasing toxicity as the value of the octanol-water partition coefficient decreased. Bar et al. (1987) describe two types of toxicity: molecular toxicity and phase toxicity. They identify molecular toxicity aa the toxicity caused by the dissolved solvent molecules in the aqueous phase, and phase toxicity as that produced by the presence of a separate organic phase. Pinherio and Cabral (1990) studied the effect of solvent molecular toxicityon the activity of Arthrobacter simplex cells. They found that n-decan-1-01 was the least toxic to the Arthrobacter simplex cells, because it had a high octanolwater partition coefficient. This is consistent with the results described here, which demonstrate that octane is
0 1994 American Chemical Society and American Institute of Chemical Engineers
188
CHzOH
I 18 C E O
Biotechnoi. ~ o g . 1994, , VOI. IO,
-
CHZOH
I
c-0
&--OH 10
&--OH
NO. 2
0
1s
Curvularia lunata 0
5’ 4
NRRL-2380
3
6
4
5
6
7
Hydrocortisone
Substance S
Figure 1. Substance S to hydrocortisone biotransformation reaction. the least toxic of the solvents examined. This is not surprisingsince octane possessesthe highest octanol-water partition coefficient of the solvents examined. The percentage of organic solvent added has an effect on the toxicity up to the saturation concentration; because of molecular toxicity, however, beyond the saturation point when there are two phases, the effect is generally the same for all quantities of organic solvent added. The organic solvent concentration in the multiphase system is often described in terms of the phase ratio (6). The phase ratio is defined as the fraction of the total reaction volume occupied by the organic phase. The selection of a particular solvent for a biotransformation system plays an important role, not only because of the solvent’s effect on microbial activity but because the choice of solvent has an effect on the differential solubility of the reactant and product, which may be subsequently exploited in separation schemes. The 11-8hydroxylation of Reichstein’s substance S (cortexolone)to hydrocortisone by Curvularia Lunata cells was chosen as the model sytem for investigatingthe bases of the effect of solvent choice on multiphase biotransformation reaction kinetics (Figure 1). The biological conversionof substance Sto hydrocortisone by Curuularia lunata is a substrate-inducible reaction. Chen and Wey (1988)determined the effect of induction and the change in characteristics of mycelial growth of Curuularia lunata associated with steroid hydroxylations. They showed that the substance S to hydrocortisone reaction was induced by substance S, and the best yield was obtained when the time of induction was about 8 h. Other previous studies of 11-0hydroxylations performed by Curuularia Lunata have examined the 11-8hydroxylation of norethisterone acetate to produce the novel 11-8hydroxylation product, hydroxynorethisterone acetate (Chosson et al., 1991). Sukhodol’skaya et al. (1986)studied the physiological properties of Curvularia lunata during the 11-8hydroxylation of substance S and showed that the presence of substance S is necessary for induction of the 11-8hydroxylase. Despite limited previous investigationsof the selection of the organic solvents in biocatalysis and some separate examinations of multiphase steroid biotransformations, there has been no systematic identification of the general criteria involved in choosing a multiphase biotraneformation system and the effects these criteria have on productivity. Our work describes the effect of solvent choice on cellular toxicity and of the differential product and reactant solubility on productivity for a batch suspension culture in ther steroid biotransformation reaction of substance S to hydrocortisone. This work aims to identify the relevant criteria involved in choosing the biotransformation systemof solvent,reactant and product,
microorganism, and reactor type and to outliie procedures by which these criteria can be applied to the selection of the solvent and the mode of reactor operation for steroid biotransformation reactions. Materials and Methods Materials. Reichstein’s substance S (cortexolone)and hydrocortisone were obtained from Sigma Chemical Company. Curvularialunata (ATCC 12017)was obtained from the American Type Culture Collection (ATCC). Organic solvents were obtained from Aldrich Chemical and Sigma. The materials for the microbial growth medium were obtained from Aldrich and Difco Laboratories. HPLCgrade methanol was obtained from Aldrich. Growth Conditions. The medium used for the growth of Curvularia lunata consisted of the following (grams/ liter of distilled water): soya flour, 5.0;glucose, 20.0;yeast extract, 5.0;KzHPOs, 5.0; NaC1, 5.0. The final pH was adjusted to 7.0. The microorganism was grown on a rotary shaker at 200 rpm in 250-mL Erlemmeyer flasks a t room temperature. Curuularia lunata cultures were stored in 15% glycerolstock solutions or on agar plates of the growth medium. Inoculum was prepared by the addition of 5 mL of overnight culture to 50 mL of medium in an Erlenmeyer flask. The inoculum was grown for approximately 24 h h and was used as the inoculum for subsequent productivity experiments. Cell viability experiments were conducted on agar plates with the same medium and composition. HPLC Methods. Substance S and hydrocortisone concentrations were measured using the LC Module 1 HPLC instrument of Waters with the Millenium 2010 software. The mobile phase consisted of 60% methanol and 50% water at a flow rate of 1 mL/min. The column used was a C-18,Novopak column. A UV detector was used to measure the absorbances of substance S and hydrocortisone at a wavelength of 254 nm. Partition Coefficients. The experimental procedure for finding the partition coefficiients for substance S and hydrocortisone between the various organic solvents and water is described below. The solvent wes fist saturated with water, steroid was added, and the system was equilibrated until the steroid reached its solubility limit in the water-saturated solvent. The water-saturated solvent and steroid were then allowed to equilibrate with an equal amount of water in a partition vessel for 12 h. Samples were taken from the aqueous and organic phases and injected onto the HPLC for steroid quantification. Toxicity Experiments. Cellular toxicity experiments were performed with the addition of the organic solvents during both the growth and stationary phases. These experiments were conducted with varying phase ratios of solvent. To determine the toxicity of the solvent to the cells during the growth phase, solvent was added initially
mtechnol.Rog., 1994, Vol. 10, No. 2
188
Table 1. Partition Coefficients of Substance S and Hydrocortirone solvent log P P, pll log P 0.69 35.2 9.9 0.55 ethyl acetate 65.6 37.9 0.24 butanol 0.8 1.3 280.5 66.2 0.62 pentanol 1.7 77.5 14.1 0.74 butyl acetate 8.67 1.46 0.77 octane 4.5
to freshly inoculated cultures in 250-mL Erlenmeyer flasks. Samples were removed periodically and plated on agar plates. The number of viable cells within the culture was monitored by counting the number of colonies on the agar plates using a colony counter. To determine the cell viability at stationary phase, the organic solvent was added after the Curvularia lunata cultures had reached the stationary phase. The two-phase cultures were then mixed well for 6 h, after which time samples were taken and plated on agar plates. Colonies were then counted to determine the concentration of viable cells. Productivity Experiments. Productivity experiments were conducted with octane to evaluate the effect of the presence of solvent on the cellular biotransformation rate of substance S to hydrocortisone. Curuularia lunata innoculuum was grown as described earlier. Fifty milliliters of medium was innoculated with the Curvularia lunata cells, and these innoculated cultures were then induced with 0.4 mg/mL of substance S. Varying phase ratios of octane were then added to each of the flasks. The flasks were shaken at room temperature at 200 rpm in an orbital shaker. Two-milliliter samples were withdrawn periodically from each flask and extracted with 0.4 mL of butyl acetate. A sample from the butyl acetate phase containing the extracted steroid was then diluted 150times with water and injected onto the HPLC for steroid analysis. Results and Discussion The differential solubility of reactant and product in the two phases was established by measuring the partitioning of the reactant and the product between the two phases. The partition coefficient was defined as the ratio of the concentration of the reactant or product in the organic phase to its concentration in the aqueous phase. The partition coefficient of substance S is designated as P,, and the partition coefficient of hydrocortisone is designated as Ph. The differential partitioning coefficient (Pr)was then calculated as the ratio of P,to Ph. Table 1 gives the partition and the relative partitioning coefficients of substance S and hydrocortisone in the solvents examined and the relative partitioning coefficient for substance Sand hydrocortisone in the solvent. The greater the value of the logarithm of the differential partitioning coefficient (log Pr),the higher the differential solubility of the reactant and product between the organic and aqueous phases. A number of solvents were examined, including butanol, pentanol, butyl acetate, ethyl acetate, and octane. The log P value for butanol was low, while the log P values for octane and butyl acetate were very high and they were moderately high for ethyl acetate and pentanol. These results are consistent with the trend that, for a solvent with a particular functional group, the higher the molecular weight of the homologue, the higher the log PC (Table 1). This trend is observed for the alcohols and esters examined where butyl acetate and pentanol have higher differential partitioning values than their corresponding lower homologues ethyl acetate and butanol. This trend follows that published by Carrea et al. (1988)for the progesterone to androstenediol reaction system. They determined the partition coefficient of progesbrone and
" - 8
0
20
10
30
Pcrcrnt Orpnh Solvent
Figure 2. Organic solvent toxicity on stationary-phase Cur, butyl acetate; ethyl acetate; vularia lunata NRRL-2380 . e,pentanob A, octane.
+,
androstenediol in butyl acetate and ethyl acetate. The partition coefficients of both the product and the reactant were much higher in butyl acetate then in ethyl acetate, which is consistent with the results for the substance S to hydrocortisone reaction system examined here (Table 1). The solvents which gave the best differential solubility of the reactant an the product were screened for their effect on the microbial activity of Curuularia lunata cells. The effect of the different solvents on microbial activity was tested both at exponential and stationary growth phases of the Curvularia lunata cultures. Desirable solvents for multiphase processes involving simultaneous reaction and product separation are those possessing moderate or high differential partitioning values for the reactant and product. Therefore, the stationary-phase toxicity test was examined first for solvents possessing high differential partitioning coefficient values, such as butyl acetate, ethyl acetate, pentanol, and octane (Table 1). The percentage of cells that remained viable 6 h after the addition of the organic solvent during the stationary phase was determined from plate count data. Cellular toxicity results (Figure 2) indicate that octane and butyl acetate are less toxic to the cells than the other solvents examined. In the case of octane, 100% of the cells were viable, while for butyl acetate about 83% of the cells remained viable. Thirty to forty percent of the cells remained viable after 6 h of exposure to ethyl acetate. No viable cells remained after 6 h in the presence of pentanol. These results indicate that the greater the miscibility of the organic solvent in the aqueous phase, the greater the toxicity of the solvent to the Curuularia lunata cells. Octane and butyl acetate were less damaging to the cells than ethyl acetate or pentanol because they are less soluble in the aqueous phase than ethyl acetate and pentanol. Therefore, butyl acetate and octane were chosen to examined solventtoxicity on cells during the growth phase. Growth-phasetoxicity experimentswere first conducted for varying phase ratios of butyl acetate. Samples were taken and plated at regular time intervals, and cellular viability was monitored as a function of time for varying concentrations of butyl acetate added during the growth phase (Figure 3). After about 6 h of growth, the percentage of viable cells is only about 20% for all phase ratios of butyl acetate, indicating that above saturation conditions the solvent effect on cell viability is constant for all phase ratios. The saturation concentration of butyl acetate in water is about 0.83%. In the case of the toxicity test conductedwith butyl acetate during the stationary phase, the percentage of cells viable after 6 h was about 83%
Bbtechnol. Prog., 1994, Vd. I O ,
100
No. 2
’*1 J
i 0
0
0.0
0
5
10
20
10
15
T l m (HI)
Figure 3. Curuularia lunata NRRL-2380growth inhibition by organic solvent: m, 6% butyl acetate; e, 10% butyl acetate, A 20% butyl acetate, 0 , 30% butyl acetate; A 5% ethyl acetate.
(Figure 2). This demonstrates that, for organic solvents which exhibit molecular toxicity, the solvent is more toxic when it is added during the growth phase than when it is added the stationary phase. This is also apparent from the growth-phase toxicity test results obtained with ethyl acetate (Figure 3). The number of viable cells remaining when the solvent is added during the growth phase is only 2 % when 5% ethyl acetate is added (Figure31, as compared to 3&40% when the test was conducted during the stationary phase (Figure 2). The results in Figure 4 show the increasein cell density with time for the control culture, as well as for the cultures having 0.2 7% and 30 7% octane added during the growth phase. It is apparent that the number of viable cells remaining in the cultures containing 0.2% and 30% octane is almost the same as the number of viable cells in the control cultures with no added solvent. Therefore, the number of cells that remained viable in the presence of octane was the same as the number in the absence of octane. Thus, it can be inferred that octane is not toxic to the cella; therefore, its molecular toxicity is zero. This is most likely due to the immiscibility of octane with the aqueous phase. Solvent toxicity tests conducted at both the growth and stationary phases indicate that the solvents with a higher octanol-water partition coefficient are least toxic of those examined. That is, the less miscible the organic solvent is in the aqueous phase, the greater the retention of microbial activity of the Curvularia lunata cells. This is consistent with the observations of Pinheiro et al. (1990) and Lame et al. (1985). They observed that for biocompatibilitythe logarithmof the solvent’s partition coefficient in the octanol-water standard system should be at least 4 with solvents possessing decreasing values, demonstrating greater toxicity. Results obtained with the Curvularia lunata substance S to hydrocortisone reaction system showed that solvents which exhibit molecular toxicity were more toxic when present during the growth phase than during the stationary phase. For our experimentalsystem, the percentage of microbial activity retained when the solvent was added during the growth phase was much less than the activity retained when the solvent was added during the stationary phase. This was true for each of the solvents examined except octane, which did not exhibit molecular toxicity. Reactant and product differential partitioning results and cellular toxicity results indicate that octane is the best solvent of those examined for the Curvularia lunata NRRL-2380 catalyzed substance S to hydrocortisone
30
,
40
Timr (Hr)
Figure 4. Curuularia lunata NRRL-2380growth inhibition by octane: 4 0 % ; e, 0.2%; A, 30%.
reation, since it demonstrated the least toxicity (Figures 2 and 3) and possessed the highest differentialpartitioning coefficient value (Table 1). The cellular productivity of hydrocortisone was measured as a function of time for different phase ratios of octane added at the start of the experiment (Figure 5). Cellular productivity is defined as millimoles of hydrocortisone produced/cell/hour. The cellular productivity achieved in the control culture with no octane present indicates that hydrocortisone is produced during both the growth and stationary phases. Since the growth phase is determined to be between 5 and 16 h and the stationary phase begins after 16 h, these results indicatethat the cellular productivity is significantlyhigher during the growth phase than during the stationary phase. Therefore, it can be concluded that the biotransformation of substance S to hydrocortisone takes place during both the growth and stationary phases and is more properly described as a growth-associated process than a nongrowth-associated or secondary fermentativeprocess. This finding has important implications for the operational mode of the two-phase biotransformation reactor system. An increase in productivity with the two-phase system over the single phase aqueous system was also noted (Figure 5). This increase in cellular productivity was greatest for the highest phase ratios and could be attributed to the higher concentration of substance S in the reactor as the phase ratio of the organic solvent increased. However, since octane is essentially insoluble in water, this increase in productivity is unlikely to be due to a cosolvency effect. This increase in productivity may be explained by the following simple model. If we assume first-order nonlinear kinetics for the system, the productivity of the single aqueous phase system is given by
k [SI = K+[Sl where [SIis the saturation concentration of substance S in the aqueous phase. For the two-phase system, the productivity is given by vaq
where [SI,, represents the overall concentration of substance S in the two-phase system. It is assumed that the maximum rate of substance S transformation (k)and the substance S half-velocity constant ( K ) are unchanged in the presence of organic solvent. This assumption is presumably valid since the cellular activity and viability are not diminished in the presence of octane (Figures 1
Bbtechnd. Rog., 1994, Vol. 10, No. 2
10
20
191
so
40
Time (HI)
Figure 5. Productivity of hydrocortisone with phase ratios of octane: m, 0%; 5%; A, 10%; 0 , 15%; B, 20%, A, 25%; X, 30%.
*,
and 4). Initial rate kinetics experiments at varying concentrations of substance S indicate that the biotransformation kinetics for the single-phase system follows the first-order nonlinear rate law shown by eq 1,with a halfvelocity constant (K)of approximately 5.5 X mM. The maximum substance S biotransformation rate (k)is estimated to be 3 X 10-14 mmol/cell-h. From eqs 1 and 2, the increase in productivity due to the added solvent can be predicted. The predicted increase in productivity with increasing phase ratios, and hence substance S concentration in the system, was estimated and compared to the experimental results. Experimentally measured productivity values are within approximately 9 % of the productivity predicted using eq 2. This deviation can, for the most part, be attributed to an approximately 8 % error in cell counting and an approximately 1% error in the HPLC measurements of hydrocortisone. Thus, this increase in cellular productivity with increasing phase ratios is predicted by eq 2. This agreement between the experimentally observed rate and the rate predicted using substance S in both phases indicates that the cell is exposed to substance Sin the organic phase. The observed increase in productivity is not likely to be due to an effective increase in the aqueous-phase concentration, since octane is essentially insoluble in the aqueous phase and thus no cosolvency effect is expected to occur. Conclusions The most important criterion for the selection of a multiphase biotransformation system is the toxicity of the solvent to the cells. Toxicity experiments indicated that, of the solvents examined, octane was the least toxic to the Curvularia lunata catalyzed substance S biotransformation system under consideration. The more miscible the solvent is in the aqueous phase, the more toxic it is to the microorganism. However, the presence of solvent a t low mixing rates and in increasing phase ratios had no additional effect on toxicity over the solubility limit. Another important criterion is the differential solubility of the product and reactant in the two phases. This is especially important in systems where product inhibition occurs. In such systems, product removal may markedly enhance the reaction productivity. A high degree of differential partitioning of the reactant and product into separate immiscible phases can greatly simplify product separation and purification. Since many steroid products are extracted into organic solvent as a first step in purification, a high differential solubility of product and reactant enhances the recovery of unreacted substrate and
allows for continuous operation schemes, which may include reactant recycling. For the substance S to hydrocortisone model reaction, the less polar the organic solvent added, the greater the solubility of substance S and the lower the solubility of hydrocortisone in the solvent. Therefore, the less miscible solvents with high octanol-water partition coefficients, such as butyl acetate and octane, and also the solvents of choice for maximizing the differential partitioning of product and reactant in the aQueous-phase system. The relationship between productivity and the growth phase of the culture has important implications for the choice of reactor operating mode. For a non-growthassociated biotransformation reaction, a continuous reactor would be a less logical choice than a traditional batch culture since it would be necessary to operate the reactor at a high cell concentration and a very low flow rate to maintain stationary-phase conditions within the reactor. Since the cellular productivity for the substance S to hydrocortisone biotransformation reaction is highest during the growth phase, a continuous reactor is the logical choice for enhancing the overall productivity of the process; however, it was also found that those solvents which exhibited molecular toxicity were more toxic when present during the growth phase than during the stationary phase. Since the production of hydrocortisone is approximately 20 times greater during the growth phase than during the stationary phase, this has important implications for solvent choice. The presence of solvents exhibiting toxicity should be avoided in a continuous reactor operating at a dilution rate that maintains an exponentially growing cell population. An alternative strategy, which would be preferable in cases where nonbiocompatible solvents must be used, would be to run the process in batch, adding solvent only after the cells have entered the stationary phase. The phase ratio of the organic solvent added has a significant effect on the productivity of the system. It was found that there is an increase in productivity with an increase in the phase ratio of the organic solvent added. This increase in productivity with the organic solvent is mainly due to the increase in concentration of the reactant in the system. This increase in productivity is predicted by assuming that all dissolved substance S, in the aqueous as well as the organic phase, is available for reaction. The basic assumption of this model is that the cells come in contact with the substance S present in the organic solvent. Thus, it is assumed that most of the reaction takes place at the interface between the aqueous and organic phases. The interfacial area in the culture is presumably high due to the vigorous agitation of the system in the rotary shaker. Therefore, in the absence of cellular toxicity effects, addition of a water-immiscible organic phase may enhance the productivity of biotransformation reactions involving reactants of low aqueous solubility, such as steroids.
Literature Cited Bar, R. In Biocatalysis in Organic Media; Laane, C., Tramper, J., Lilly, M. D., Eds.; Elsevier: Amsterdam, 1987; p 147. Brink, L. E. S.;Tramper,J. Optimization of Organic Solvent in MultiphaseBiocatalysis. Biotechnol.Bioeng. 1985,27,125& 1269.
Buckland, B. C.; Dunnill, P.; Lilly, M. D. The Enzymatic TransportationOf Water Insoluble Reactants In Non Aqueous Solvents. Conversion Of Cholesterol To Cholest-4-ene-3-one by a Nocardia sp. Biotechnol. Bioeng. 1976, 17,815-826. Carrea, G.; Riva, S.;Bovara, R.; Pasta, P. Enzymatic Oxidoreductionof Steroidsin Two Phase Systems. Enzyme Microb. Technol. 1988, 10, 333-340.
192
Chen, C. K.; Wey, C. H. llg-hydroxylation of cortexolone by CurvulorM lunata. Enzyme Microb. Technol. 1990,12,30& 307. Choeeon, P.; Vidal, H.; Aumeias, A.; Couderc, F. llg-hydroxylation of norethiaterone acetate by Curuularia lunata. Microbiol. Lett. 1991,17-22. Fakai, S.; Tanakn, A. Bioconversion Of Lipophilic Compounds By Immobilized Microbial Cells In Organic Solvents. Acta Biotechnol. 1981,l (41, 339-360. Holst, 0.;Kaul, R.;Lareson, M.; Mattiasson, B. Integration Of Bioconversions And Downstream Proceesing. Some Model Studies. Anal. N.Y.Acad. Sci. 1987,506,488-477. Laane, C.; Boeren, S.; Vos, K. Optimizing Organic Solvents in Multi-liquid-phaseBiocatalysis. Trends Biotechnol. 1985,3, 251. Lilly, M. D.; et al. Biological Conversions Involving Water Insoluble Organic Compounds. In Biocatalysis in Organic Media; Laane, C., Tramper, J., Ldy, M. D., Eds.; Elsevier Science Publishing Company: New York, 1987; pp 3-17. Mukataka, S.; Haynes, C. A.; Prausnitz, J. M.; Blanch, H. W. Extractive Bioconversions in Aqueous Two Phase Systems. Biotechnol. Bioeng. 1992,40, 196-206.
Blotechd. Prog.., 1994, Vol. 10, No. 2
Pinheiro, H. M.; Cabral, J. M. 5.Effects of Solvent Molecular Toxicity and Microenvironment Composition on the G'Dehydrogenation Activity of Arthrobacter simplex Cells. Biotechnol. Bioeng. 1991,37,97-102. Smith,L. L. InBiotechnology; Rehm,H. J., Reed,R.,Eds.; Verlag Chemie: Weinheim, Germany, 1984, pp 64-65. Sukhodol'skaya, G. V.; Angelova, B. A.; Koshcheenko, K. A. Physiological and Biochemical Properties of a Culture of Curvularia lunata VKM-644with the Ability to Achieve 11sHydroxylation of Steroid Substrates. Prikl. Biokhim. Mikrobiol. 1986,22, 226-236. Sukhodol'akaya, G. V.; Angelova, B. A,; Koshcheenko, K. A,; Skryabin, G. K. Synthesis of Steroid llg-hydroxylase Characteristics of the Transformation of Reichstein's Crystalline Substancesby Freeand Immobilized Curuularialunata VKM644Mycelium. Prikl. Biokhim. Mikrobiol. 1991,27 (5),701710. Accepted November 4,1993.. a Abstract published in Advance ACS Abstracts, December 15, 1993.
