Enhanced bioproduction of carvone in a two-liquid-phase partitioning bioreactor with a highly hydrophobic biocatalyst

The microbial biotransformation of (-)-trans-carveol to the flavor and fragrance compound (R)-(-)-carvone by Rhodococcus erythropolis DCL14 was carried out in a 3 L two phase partitioning bioreactor with an immiscible liquid second phase in an effort to improve upon the reactor performance achieved in a single aqueous phase system. The purpose of employing the liquid second phase is to minimize biotransformation rate inhibition due to the accumulation of the toxic substrate (cis-carveol) and product (carvone) in the aqueous phase. 1-Dodecene was chosen as the solvent for this application because it is biocompatible, non-biodegradable and has a superior affinity for the target product (carvone) relative to the other solvents tested. However, when 1-dodecene was used in the biotransformation, the extremely hydrophobic R. erythropolis DCL14 created an emulsion with the organic solvent with significant sequestering of the cells into the organic phase and negligible substrate conversion. To overcome these operational difficulties, silicone oil, which is considered a liquid polymer, was used with the aim of preventing emulsification and sequestration of cells in the non-aqueous phase. Although some emulsification of the water-silicone oil was again created by the cells, operability was improved and, in fed-batch mode, the system was able to convert approximately 2(1/2) times more carveol than a benchmark single aqueous phase system before substrate/product toxicity caused the biotransformation to stop. This study has demonstrated enhancement of a microbial biotransformation for the production of a high value nutraceutical compound via the use of a second partitioning phase, along with operational challenges arising from the use of a highly hydrophobic organism in such systems.     

Effects of R-(+)-limonene on submerged cultures of the terpene transforming basidiomycete Pleurotus sapidus.

The biotransformation of limonene by the basidiomycete Pleurotus sapidus yielded cis/trans-carveol and carvone as the main products. The transformation period was extended from 4 days after direct addition to 12 days by gas phase addition of the substrate. After 2 days of transformation, 97% of the substrate had accumulated in the mycelium, while only 3% were present in the culture medium. Substrate toxicity led to a decrease of dry matter. Adaptation of the precultures with small amounts of substrate doubled the concentration of carveol and increased the concentration of carvone by a factor of 3-4. Total product concentrations of > 100 mg l-1 were reached.     

Cell adaptation to solvent, substrate and product: a successful strategy to overcome product inhibition in a bioconversion system

Carvone has previously been found to highly inhibit its own production at concentrations above 50 mM during conversion of a diastereomeric mixture of (−)-carveol by whole cells of Rhodococcus erythropolis. Adaptation of the cells to the presence of increasing concentrations of carveol and carvone in n-dodecane prior to biotransformation proved successful in overcoming carvone inhibition. By adapting R. erythropolis cells for 197 h, an 8.3-fold increase in carvone production rate compared to non-adapted cells was achieved in an air-driven column reactor. After an incubation period of 268 h, a final carvone concentration of 1.03 M could be attained, together with high productivity [0.19 mg carvone h⁻¹ (ml organic phase)⁻¹] and high yield (0.96 g carvone g carveol⁻¹).     

Development of a reaction system for the selective conversion of (−)-trans-carveol to (−)-carvone with whole cells of Rhodococcus erythropolis DCL14

The present article addresses the development of a microbial reaction system for the transformation of carveol to carvone, using whole cells of Rhodococcus erythropolis DCL14. This strain contains a NAD-dependent carveol dehydrogenase (CDH) when grown on limonene or on cyclohexanol. When a mixture of (−)-cis and (−)-trans-carveol is supplied, only (−)-trans-carveol is converted. Thus, besides (−)-carvone, pure (−)-cis-carveol can be obtained as product.

Initial experiments were performed batchwise using an aqueous system. (−)-Trans-carveol conversion rate gradually decreased during successive reutilisation batches. After the third reutilisation, activity was completely lost. Cells grown on cyclohexanol showed a slightly higher activity as compared to cells grown on (+)-limonene. A production of 4.3 μmol (−)-carvone formed per mg protein was achieved. A significant improvement with respect to initial reaction rate and productivity was obtained with aqueous–organic two-phase systems. Using a 5 to 1 buffer/iso-octane system, a 40% increase in the initial rate and a 16-fold increase of the production was observed. A further improvement resulted from increasing the volume of solvent (1 to 1 buffer/dodecane ratio). An initial reaction rate of 26 nmol/(min*mg protein) was observed, while production increased to 208 μmol (−)-carvone formed per mg protein. As in the single-phase system, reaction rate gradually decreased along the successive cell reutilisation batches. Addition of co-substrates for the regeneration of NAD did not prevent this decay. A simple downstream process was developed for the recovery of carvone and cis-carveol.     

Improved reactor performance and operability in the biotransformation of carveol to carvone using a solid-liquid two-phase partitioning bioreactor

In an effort to improve reactor performance and process operability, the microbial biotransformation of (-)-trans-carveol to (R)-(-)-carvone by hydrophobic Rhodococcus erythropolis DCL14 was carried out in a two phase partitioning bioreactor (TPPB) with solid polymer beads acting as the partitioning phase. Previous work had demonstrated that the substrate and product become inhibitory to the organism at elevated aqueous concentrations and the use of an immiscible second phase in the bioreactor was intended to provide a reservoir for substrates to be delivered to the aqueous phase based on the metabolic rate of the cells, while also acting as a sink to uptake the product as it is produced. The biotransformation was previously undertaken in a two liquid phase TPPB with 1-dodecene and with silicone oil as the immiscible second phase and, although improvement in the reactor performance was obtained relative to a single phase system, the hydrophobic nature of the organism caused the formation of severe emulsions leading to significant operational challenges. In the present work, eight types of polymer beads were screened for their suitability for use in a solid-liquid TPPB for this biotransformation. The use of selected solid polymer beads as the second phase completely prevented emulsion formation and therefore improved overall operability of the reactor. Three modes of solid-liquid TPPB operation were considered: the use of a single polymer bead type (styrene/butadiene copolymer) in the reactor, the use of a mixture of polymer beads in the reactor (styrene/butadiene copolymer plus Hytrel(R) 8206), and the use of one type of polymer beads in the reactor (styrene/butadiene copolymer), and another bead type (Hytrel(R) 8206) in an external column through which fermentation medium was recirculated. This last configuration achieved the best reactor performance with 7 times more substrate being added throughout the biotransformation relative to a single aqueous phase benchmark reactor and 2.7 times more substrate being added relative to the best two liquid TPPB case. Carvone was quantitatively recovered from the polymer beads via single stage extraction into methanol, allowing for bead re-use.     

The synthesis of L-carvone and limonene derivatives with increased antiproliferative effect and activation of ERK pathway in prostate cancer cells

Thirty-one novel derivatives of carvone, carveol, and limonene were designed and synthesized using L-carvone as a starting material via chlorination, nucleophilic substitution, and reduction. The structures of these derivatives were characterized by MS and 1H NMR. The antiproliferative effect was evaluated in human prostate cancer LNCaP cells. L-carvone, L-carveol, and L-limonene were weak cell growth inhibitors and introduction of 4-(2-methoxyphenyl)piperazine to carvone, carveol or limonene significantly increased their antiproliferative effect. The antiproliferative effect was correlated with ERK activation and p21(waf1) induction.     

Principal Components Analysis as a Tool to Summarise Biotransformation Data: Influence on Cells of Solvent Type and Phase Ratio

The effect of some solvents, present in different amounts, upon whole cells of Rhodococcus erythropolis DCL14 carrying out the biotransformation of (-)-carveol to (-)-carvone was studied. The solvents tested were ethyl butyrate, n-hexane, cyclohexane, iso-octane, n-dodecane, dimethyl sulfoxide, bis(2-ethylhexyl) phthalate and FC-70. The volumes of each solvent corresponded to organic:aqueous phase ratios of 0.0005, 0.0025, 0.005, 0.025 and 0.2. To assess any potential solvent protection towards substrate toxicity, assays were carried out at two initial carveol concentrations (15 and 50 mM). Carvone accumulation was followed by gas chromatography. Cell viability, several aspects of cell morphology and the ability to form clusters were monitored by fluorescence microscopy. Principal components analysis (PCA) was used as a tool to explain the differences in the observations of the multidimensional data set obtained from the multiple conditions. PCA using the different volumes of each solvent as variable suggests that the variability of the observations can be summarised in six components which represent 79.4% of the variance of the data. Conversely, using cell and solvent data to perform the PCA, 97.1% of the variance of the data can be summarised in three components, the first two capturing 91.0% of the information. These components seem to represent solvent toxicity and a protective effect of the solvent from carveol toxicity.     

Biochemical and Histochemical Localization of Monoterpene Biosynthesis in the Glandular Trichomes of Spearmint (Mentha spicata)

The primary monoterpene accumulated in the glandular trichomes of spearmint (Mentha spicata) is the ketone (−)-carvone which is formed by cyclization of the C₁₀ isoprenoid intermediate geranyl pyrophosphate to the olefin (−)-limonene, hydroxylation to (−)-trans-carveol and subsequent dehydrogenation. Selective extraction of the contents of the glandular trichomes indicated that essentially all of the cyclase and hydroxylase activities resided in these structures, whereas only about 30% of the carveol dehydrogenase was located here with the remainder located in the rest of the leaf. This distribution of carveol dehydrogenase activity was confirmed by histochemical methods. Electrophoretic analysis of the partially purified carveol dehydrogenase from extracts of both the glands and the leaves following gland removal indicated the presence of a unique carveol dehydrogenase species in the glandular trichomes, suggesting that the other dehydrogenase found throughout the leaf probably utilizes carveol only as an adventitious substrate. These results demonstrate that carvone biosynthesis takes place exclusively in the glandular trichomes in which this natural product accumulates.     

Preventing biofilm formation: promoting cell separation with terpenes

Both carveol and carvone were effective in dispersing Rhodococcus erythropolis cells that were being stimulated to aggregate by the presence of organic solvents. The two terpenes influenced the fatty acid composition of the cell membrane, decreasing the percentage of fatty acids with more than 16 carbon atoms, and thus cell hydrophobicity, and also the degree of saturation of the fatty acids. In the presence of 250 micromol of terpene, the volume of biofilm was reduced by one third in comparison with biofilms in the absence of terpenes. The percentage of aggregated cells was also found to depend on carvone concentration during the bioconversion of carveol to carvone, in a membrane reactor. The extent of cell aggregation decreased from 90% to 10% when carvone concentration reached ca. 48 mM in the organic phase.     

Auxiliary phase guidelines for microbial biotransformations of toxic substrate into toxic product

When an industrial process is developed using the microbial transformation of a precursor into a desired chemical compound, high concentrations of substrate and product will be involved. These compounds may become toxic to the cells. In situ product removal (ISPR) may be carried out, using auxiliary phases such as extractants or adsorbents. Simultaneously, in situ substrate addition (ISSA) may be performed. It is shown that for uncharged substrates and products, the aqueous solubilities of substrate and product can be used to predict if ISPR might be required. When a particular auxiliary phase is selected and the distribution coefficients of substrate and product are known, it is possible to estimate a priori if this auxiliary phase might be good enough and how much of it might be needed for an efficient (fed-)batch biotransformation process. For biotransformation products of intermediate polarity (aqueous solubility of about 1-10 g/L) there seems to be a lack of extractants and adsorbents with the capacity to raise the product concentrations to commercially more interesting levels.     

Biotransformation of limonene by bacteria,fungi, yeasts,and plants

The past 5 years have seen significant progress in the field of limonene biotransformation, especially with regard to the regiospecificity of microbial biocatalysts. Whereas earlier only regiospecific biocatalysts for the 1,2 position (limonene-1,2-diol) and the 8-position (alpha-terpineol) were available, recent reports describe microbial biocatalysts specifically hydroxylating the 3-position (isopiperitenol), 6-position (carveol and carvone), and 7-position (perillyl alcohol, perillylaaldehyde, and perillic acid). The present review also includes the considerable progress made in the characterization of plant P-450 limonene hydroxylases and the cloning of the encoding genes.     

Monoterpene biosynthesis pathway construction in Escherichia coli

Four genes encoding sequential steps for the biosynthesis of the spearmint monoterpene ketone (-)-carvone from the C(5) isoprenoid presursors isopentenyl diphosphate and dimethylallyl diphosphate were installed in Escherichia coli. Inducible overexpression of these genes in the bacterial host allowed production of nearly 5 mg/l of the pathway intermediate (-)-limonene, which was mostly excreted to the medium such that products of the downstream steps, (-)-carveol and (-)-carvone, were not detected. Assay of pathway enzymes and intermediates indicated that flux through the initial steps catalyzed by geranyl diphosphate synthase and limonene synthase was severely limited by the availability of C(5) isoprenoid precursors in the host. Feeding studies with (-)-limonene, to overcome the flux deficiency, demonstrated the functional capability of limonene-6-hydroxylase and carveol dehydrogenase to produce the end-product carvone; however, uptake and trafficking restrictions greatly compromised the efficiency of these conversions.     

Evaluation of an electrochemical bioreactor system in the biotransformation of 6-bromo-2-tetralone to 6-bromo-2-tetralol

Biotransformation of 6-bromo-2-tetralone (Br-beta-tetralone) to 6-bromo-2-tetralol (Br-beta-tetralol) by yeast cells of Trichosporon capitatum (ATCC 74312) and its partially purified Br-beta-tetralone reductase was evaluated in an electrochemical bioreactor. The biotransformation rates and final product formation were significantly affected by substrate concentration, biomass and electric potential. At 2 g/l of substrate, the initial reaction rate and final product were increased by 35% and 15%, respectively, with -1.5 V of electric potential compared to without electric potential. Additional substrate (2 g/l) provided by pulse feeding to the reaction mixture at different intervals resulted in 2.1 g/l Br-beta-tetralol compared to a total of 1.2 g/l without feeding. However, the increased production was not proportionate to the amount of additionally fed substrate. Increased substrate availability by the addition of 5% (v/v) ethanol resulted in the highest reaction rate and product formation, but addition of ethanol at a concentration higher than 5% decreased the reaction rate. At low biomass, the initial reaction rates were enhanced significantly when electric potential was high, but a higher biomass was necessary to obtain a similar reaction rate when electric potential was reduced. The highest initial reaction rate (59.2 mg/l per min) was achieved with a two-fold biomass concentration of 15.6 g of dry cell weight/l, substrate at 4 g/l and electric potential at -6 V. The conversion of Br-beta-tetralone to Br-beta-tetralol with partially purified Br-beta-tetralone reductase was slow in the presence of electric potential.     

Nutrient broth/PEG200/TritonX114/Tween80/Chloroform microemulsion as a reservoir of solubilized sitosterol for biotransformation to androstenedione

Microemulsions (ME) can act as a reservoir of solubilized hydrophobic substrates. The biotransformation of hydrophobic sitosterol to androstenedione (AD) with MEs prepared from nutrient broth and PEG 200 (1:1) as aqueous phase, 40 g/l sitosterol dissolved in chloroform as organic phase, Triton X114 and Tween 80 (1:1) as surfactant phase, was investigated. The phase behavior of this system was studied for ten different ratios(w/w), 10:0, 9:1, 8:2, 7:3, 6:4, 5:5, 4:6, 3:7, 2:8, 1:9 and 0:10 of the organic phase and surfactant at 30 °C. A pseudoternary phase diagram was constructed to demarcate the region giving stable MEs. The maximum solubility of sitosterol in ME medium was observed to be 8 g/l, which is 3 orders of magnitude higher than the reported sitosterol solubility of 2–4 mg/l in aqueous medium. The ME medium was used for biotransformation studies and a comparative result has been reported. Transmission electron microscopy of cells grown in ME having oil, surfactant and aqueous phase in the ratio of 6:14:80 showed a weakened cell wall structure that permitted production of 465.86 mg/l AD.     

Synthesis of (<Emphasis Type="Italic">R</Emphasis>)-2-trimethylsilyl-2-hydroxyl-ethylcyanide catalyzed with (<Emphasis Type="Italic">R</Emphasis>)-oxynitrilase from loquat seed meal

The synthesis of optically active (R)-2-trimethylsilyl-2-hydroxyl-ethylcyanide by asymmetric trans-cyanation of acetyltrimethylsilane with acetone cyanohydrin in a biphasic system was achieved using (R)-oxynitrilase from loquat seed meal. Diisopropyl ether was the most suitable organic phase among the organic solvents examined. The optimal concentration of acetyltrimethylsilane, concentration of crude enzyme, volume ratio of the aqueous to the organic phase, temperature and the buffer pH value were 14 mM, 61.4 U ml^-1, 13% (v/v), 30 °C and 4, respectively. The substrate conversion and the product enantiomeric excess were 95% and 98% under the optimized conditions. Acetyltrimethylsilane was a better substrate of the enzyme than its carbon counterpart. Revisions requested 24 August 2004; Revisions received 12 November 2004     

Biotransformation of d-limonene to carvone by means of glucose oxidase and peroxidase

A novel method for enzymatic biotransformation of limonene to carvone has been developed. It involves addition glucose oxidase and peroxidase to the biotransformation medium. Some factors affecting biotransformation yield were investigated. Maximal yield of carvone occurred in the medium containing 1.5% substrate, at 50 degrees C and pH 7.0.     

Bioproduction of benzaldehyde in a solid–liquid two-phase partitioning bioreactor using <Emphasis Type="Italic">Pichia pastoris</Emphasis>

The bioproduction of benzaldehyde from benzyl alcohol using Pichia pastoris was examined in a solid–liquid two-phase partitioning bioreactor (TPPB) to reduce substrate and product inhibition. Rational polymer selection identified Elvax 40W as an effective sequestering phase, possessing partition coefficients for benzyl alcohol and benzaldehyde of 3.5 and 35.4, respectively. The use of Elvax 40W increased the overall mass of benzaldehyde produced by approx. 300% in a 5 l bioreactor, relative to a single phase biotransformation. The two-phase system had a molar yield of 0.99, indicating that only minor losses occurred. These results provide a promising starting point for solid–liquid TPPBs to enhance benzaldehyde production, and suggest that multiple, targeted polymers may provide relief for transformations characterized by multiple inhibitory substrates/product/by-products.     

Process modeling of the lipase-catalyzed dynamic kinetic resolution of (<Emphasis Type="Italic">R</Emphasis>, <Emphasis Type="Italic">S</Emphasis>)-suprofen 2,2,2-trifluoroethyl thioester in a hollow-fiber membrane

A Candida rugosa lipase immobilized on polypropylene powder was employed as the biocatalyst for the enantioselective hydrolysis of (R, S)-suprofen 2,2,2-trifluorothioester in cyclohexane, in which trioctylamine was added as the catalyst to perform in situ racemization of the remaining (R)-thioester. A hollow-fiber membrane was also integrated with the dynamic kinetic resolution process in order to continuously extract the desired (S)-suprofen into an aqueous solution containing NaOH. A kinetic model for the whole process (operating in batch and feed-batch modes) was developed, in which enzymatic hydrolysis and deactivation, lipase activation, racemization and non-enantioselective hydrolysis of the substrate by trioctylamine, and reactive extraction of (R)- and (S)-suprofen into the aqueous phase in the membrane were considered. Theoretical predictions from the model for the time-course variations of substrate and product concentrations in each phase were compared with experimental data.     

Process parameters and reusability of the free cell mass of Torulaspora delbrueckii for the production of L-phenylacetylcarbinol (L-PAC)

The effect of process parameters on the biotransformation of benzaldehyde to L-phenylacetylcarbinol (L-PAC) using a yeast isolate identified as Torulaspora delbrueckii was studied. The maximum yield of L-PAC obtained was (331 mg) per 100 ml biotransformation medium (glucose 3%, peptone 0.6% and at pH 4.5) from 600 mg of benzaldehyde with 8 h of reaction at 30 ± 2 °C. Growing the organism in presence of 3% glucose reduced the biotransformation time to 120 min. Addition of 0.6% acetaldehyde (30–35%) lead to an increase in L-PAC yield to 450 mg%. Semi-continuous feeding of benzaldehyde (200 mg) and acetaldehyde (200 l) four times at 30 min intervals could produce 683 mg of L-PAC/100 ml biotransformation medium. Chiral HPLC analysis of purified L-PAC and PAC-diol showed 99% enantiomeric purity. The cell mass was found to be reusable for biotransformation up to nine times when benzaldehyde and acetaldehyde levels were maintained at (350 mg and 350 l)–(400 mg and 400 l). At concentrations from 450 mg and 450 l to 600 mg and 600 l, however the cell mass could give efficient biotransformation only during one use.     

Use of <Emphasis Type="Italic">Pseudomonas mendocina</Emphasis>, or recombinant <Emphasis Type="Italic">Escherichia coli</Emphasis> cells expressing toluene-4-monooxygenase,and a cell-free tyrosinase for the synthesis of 4-fluorocatechol from fluorobenzene

The transformation of fluorobenzene (FB) by whole cell expressing toluene-4-monooxygenase (T4MO) resulted in the formation of various hydroxylated products. The predominant product was either 4-fluorophenol (4FP) or 4-fluorocatechol (4Fcat) depending on the ratio of biocatalyst to substrate concentration. The transformation of 1 mM FB by whole cells (1.5 mg CDW/ml) gave a 52% yield of 4Fcat as a single product. The yield of 4Fcat was improved 1.6-fold (80%) by adding 10 mM ascorbic acid to the biotransformations. A combination of two biocatalysts (whole cells expressing T4MO and cell free mushroom tyrosinase) also resulted in the transformation of FB (5 mM) to higher concentrations of 4Fcat (1.8 mM) compared to a whole cell biotransformation alone. However, mixed products were formed and the yield of 4Fcat from FB was lower using the two-step (tandem) method (27%) compared to the use of whole cells of P. mendocina KR1 alone (80%).