- Email: [email protected]
Biotransformation of (R)-(+)- and (S)-(−)-limonene by fungi and the use of solid phase microextraction for screening
Phytochemistry 57 (2001) 199±208
www.elsevier.com/locate/phytochem
Biotransformation of (R)-(+)- and (S)-( )-limonene by fungi and the use of solid phase microextraction for screening Jan C.R. Demyttenaere *, Kristof Van Belleghem, Norbert De Kimpe Department of Organic Chemistry, Faculty of Agricultural and Applied Biological Sciences, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium Received 29 September 2000; received in revised form 1 December 2000
Abstract The biotransformation of (R)-(+)- and (S)-( )-limonene by fungi was investigated. More than 60 fungal cultures were screened for their ability to bioconvert the substrate, using solid phase microextraction as the monitoring technique. After screening, the best fungal strains were selected for further study and were grown as sporulated surface cultures in conical ¯asks and as submerged liquid cultures. It was found that (+)- and ( )-limonene were converted by Penicillium digitatum to a-terpineol (main metabolite), cis- and trans-p-menth-2-en-1-ol, neodihydrocarveol and limonene oxide (minor metabolites) using liquid cultures. The bioconversion of (R)-(+)- and (S)-( )-limonene by Corynespora cassiicola yielded (1S,2S,4R)- and (1R,2R,4S)-limonene-1,2-diol respectively. The bioconversions by liquid cultures were also monitored by solid phase microextraction as a function of time. The optimum conversion of limonene to a-terpineol by Penicillium digitatum was obtained after 8 hours (yield up to 100%). Since an important pH-decrease was noticed in some liquid broths, the stability of limonene under acidic conditions was investigated. No acid catalysed conversion products were recovered after 8 days from control ¯asks at pH 3.5 containing limonene. # 2001 Elsevier Science Ltd. All rights reserved. Keywords: Penicillium digitatum; Corynespora cassiicola; Fungi; Fungal spores; Biotransformation; Bioconversion; Limonene; a-Terpineol; (1S,2S,4R)-Limonene-1,2-diol; (1R,2R,4S)-Limonene-1,2-diol; SPME
1. Introduction Much work has been done on the biotransformation of the inexpensive hydrocarbon monoterpene limonene (Bowen, 1975), which is, besides a-pinene, the most widely distributed terpene in nature (Krasnobajew, 1984). The (+)-isomer is present in citrus peel oils at a concentration of over 90%; a low concentration of the ( )-isomer is found in oils from the Mentha species and conifers (Bauer et al., 1990). The microbial conversion of limonene is very well documented, the ®rst data date back to the sixties. A soil pseudomonad was isolated by an enrichment culture technique on limonene as the sole source of carbon (Dhavalikar and Bhattacharyya, 1966) and was able to convert the substrate to a large number of neutral and acidic products. Another group isolated a * Corresponding author. Tel.: +32-9-264-549-64; fax: +32-9-26462-43. E-mail address: [email protected] (J.C.R. Demyttenaere).
Pseudomonas incognita by an enrichment technique on the monoterpene alcohol linalool (Madyastha et al., 1977). The metabolism of limonene by this bacterium was investigated (Rama Devi and Bhattacharyya, 1977). After fermentation the medium yielded as the main product a crystallic acid, perillic acid, together with unmetabolised limonene and some oxygenated compounds, such as dihydrocarvone, carvone and carveol. The biotransformation of limonene by another Pseudomonas strain, P. gladioli was also reported (Cadwallader et al., 1989; Cadwallader and Braddock, 1992). Major conversion products from (+)-limonene were (+)-aterpineol and (+)-perillic acid. This was the ®rst time that the microbial conversion of limonene to a-terpineol was reported. a-Terpineol is widely distributed in nature and is one of the most commonly used perfume chemicals (Fenaroli, 1975). The ®rst data on fungal bioconversion of limonene date back to the late sixties. A Cladosporium sp. designated as T7 was isolated on limonene and was able to convert the substrate to translimonene-1,2-diol (Mukherjee et al., 1973). Another
0031-9422/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S0031-9422(00)00499-4
200
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
Cladosporium, designated T12, converted limonene into a-terpineol (Kraidman et al., 1969). Fungi, isolated from rotting orange rind, were also found to be able to convert limonene to a-terpineol (Mattison et al., 1971). Bowen (1975) isolated two common citrus moulds, Penicillium italicum and P. digitatum, responsible for the postharvest diseases of citrus fruits. Fermentation of P. italicum on limonene yielded cis- and trans-carveol (26%) as main products, together with (+)-carvone (6%) and minor metabolites. Conversion by P. digitatum yielded the same products in lower yields. The bioconversion of limonene by Aspergillus niger has also been described (Rama Devi and Bhattacharyya, 1978). This fungus was able to carry out three types of oxygenative rearrangements. More recently, a Japanese group studied the biotransformation of limonene and related compounds (1-methylcyclohexene and cyclohexene) by Aspergillus cellulosae M77 (Noma et al., 1992). (+)-Limonene was mainly converted to (+)-isopiperitenone (19%), (+)-limonene-1,2trans-diol (21%), (+)-cis-carveol (5%) and (+)-perillyl alcohol (12%). The production of isopiperitenone from limonene was published for the ®rst time. The production of glycols from limonene and other terpenes with a 1-menthene skeleton was reported (Abraham et al., 1984). The most appropriate strains were Corynespora cassiicola and Diplodia gossypina. The same group (Abraham et al., 1985) also investigated the biotransformation of (R)-limonene to (R)-a-terpineol by the fungus Penicillium digitatum. An extensive overview on the microbial transformations of terpenoids with a 1p-menthene skeleton was published by this group (Abraham et al., 1986). One of the most recent publications in this area reported the bioconversion of limonene to a-terpineol by immobilised P. digitatum (Tan and Day, 1998a) and studied the eect of organic cosolvents on the bioconversion of (R)-(+)-limonene to (R)-(+)-a-terpineol (Tan and Day, 1998b). In all these examples submerged liquid cultures were used. The bioconversion of limonene by fungal spores has not been published to date. This paper reports the biotransformation of (R)-(+)(1) and (S)-( )-limonene (19) by sporulated surface cultures and liquid cultures of fungi and the use of solid phase microextraction (SPME) for fast screening. 2. Results and discussion 2.1. Screening of sporulated surface cultures by fungi by SPME More than 60 fungal strains, grown as sporulated surface cultures, were screened for their ability to bioconvert the substrate (R)-(+)-limonene (1), using solid phase microextraction (SPME) as the monitoring
technique. Therefore, a method was developed to cultivate sporulated surface cultures of the fungi in small vials, and the SPME-parameters were optimised. Different SPME-®bers, extraction times and temperatures were compared. Extraction of volatiles (limonene and metabolites) was done by headspace SPME. It was found that the best SPME ®ber for extraction of limonene and its metabolites was 50/30 mm divinylbenzene/ carboxen on polydimethylsiloxane. The optimum adsorption time was 30 min. The best extraction temperature was room temperature, but because of temperature ¯uctuation, a constant adsorption temperature of 25 C was chosen as optimum. A desorption temperature of 250 C and desorption time of 2 min was sucient to desorb all volatiles into the gaschromatographic inlet. The most interesting conversion of (R)-(+)-limonene (1) was the formation of (R)-(+)-a-terpineol (2) by various Penicillium species. Other metabolites recovered were g-terpinene (3), terpinolene (4), a-phellandrene (5), endo-fenchol (11), menth-3-en-1-ol (6), p-cymene (9), neo-dihydrocarveol (12), cis-limonene oxide (7), translimonene oxide (8) and perillyl alcohol (10) (see Scheme 1). The most interesting strains were P. digitatum (both isolated (CMC) and obtained, DSM 62840) and Corynespora cassiicola DSM 62475. 2.2. Biotransformation of (R)-(+)-limonene by sporulated surface cultures After screening, the best fungal strains were selected for further study and were grown as sporulated surface cultures in conical ¯asks. The biotransformation was monitored by dynamic headspace, steam distillation solvent extraction and liquid/liquid extraction. In a ®rst experiment, the bioconversion of (R)-(+)limonene (1) by two strains was compared, namely Penicillium digitatum (CMC), selected as interesting strain, based on the SPME screening results, and Aspergillus niger (AND), since A. niger has been used successfully in many bioconversion reactions (Rama Devi and Bhattacharrya, 1978; Arfmann et al., 1987, 1988; Demyttenaere and Willemen, 1998; Demyttenaere and De Kimpe, 2000; Demyttenaere et al., 2000). To 10day old sporulated surface cultures, 100 ml of a 20% limonene in EtOH solution was sprayed. Two more substrate additions took place after 15 and 17 days respectively. It was found that Aspergillus niger was not able to convert limonene, whereas Penicillium digitatum produced (R)-(+)-a-terpineol (2) from limonene with low yields (0.7±1.4%). In a second experiment, the bioconversion of (R)-(+)-limonene (1) by two strains of Penicillium digitatum, namely strain CLE and strain PDD was compared in duplicate. The ®rst substrate addition (125 ml of a 20% limonene in EtOH solution) was performed after 7 days. Two more substrate
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
201
Scheme 1. Biotransformation of (R)-(+)-limonene by sporulated surface cultures of fungi Ð monitoring by SPME.
additions took place after 12 and 14 days resp. From the headspace extracts of the surface cultures, it could be seen that up to 44% of the substrate (R)-limonene (1) was recovered unchanged. The yield of conversion to (R)-a-terpineol (2) was 1.1 and 1.3% for the two P. digitatum cultures of strain CLE resp. and 2.7 and 2.9% for the two cultures of strain PDD resp. Minor metabolites were 1,3,5-p-menthatriene (13) (0.2±0.6%), cislimonene oxide (7) (0.1±0.2%) and exo-2-hydroxycineol (14) (0.1±0.2%) (see Scheme 2). It is important to note that the pH of the surface cultures dropped to pH 3 after the cultures were subjected to steam distillation/solvent extraction. Therefore, a control experiment was run to check the degradation of limonene in sterile agar medium and MEA-medium,
acidi®ed to pH 3.5 (at lower pH, the agar did not solidify), during headspace sampling and steam distillation/ solvent extraction. The substrate (R)-limonene happened to be very stable: only small traces (0.14%) of bmyrcene were recovered from the steam distillate. No acid catalysed conversion products from limonene were recovered from the headspace extracts. It can be concluded that sporulated surface cultures of P. digitatum are able to convert the substrate (R)(+)-limonene (1) to (R)-a-terpineol (2). Since the very high volatility of the substrate, limonene, a high amount of substrate loss due to evaporation during headspace extraction was noticed, resulting in high substrate recovery. Therefore, when dynamic headspace sampling is compared with headspace SPME-sampling, the latter
Scheme 2. Bioconversion of (R)-(+)-limonene by sporulated surface cultures of Penicillium digitatum.
202
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
method is more sensitive and ecient for fast screening. 2.3. Bioconversion of (R)-(+)- and (S)-( )-limonene by liquid cultures In a ®rst experiment the bioconversion of (R)-(+)limonene by submerged liquid cultures was studied, comparing ®ve fungal strains, namely Corynespora cassiicola (marked COC), Aspergillus niger (marked AND) and three dierent strains of Penicillium digitatum (marked CLE, CMC and PDD respectively). All strains were cultivated in duplicate, except strain CLE. As a control, the stability of the precursor was checked in acidi®ed broth (pH 3.5) that was not inoculated but kept under sterile conditions. Over a period of 8 days, 325 ml of (R)-(+)-limonene (each time 125 ml of a 20% solution of limonene in EtOH) was added to cultures of 50 ml each. The ®rst substrate addition was performed after 46 h of cultivation, the second after 5 days and the third after 7 days. Samples were taken 24 h after the ®rst substrate addition, and 22.5 h after the second and third substrate addition, respectively. Corynespora cassiicola was able to transform the substrate to (1S,2S,4R)-limonene-1,2diol (15) with a ®nal yield of 46.9±60.6%. This biotransformation con®rms previously published literature data (Abraham et al., 1986). All P. digitatum cultures converted (R)-(+)-limonene (1) to (R)-(+)-a-terpineol (2) with an initial yield up to 45.7% (after 24 h). These results are in agreement with literature data (Abraham et al., 1985). Aspergillus niger, however, was not able to convert the substrate. The substrate recovery was very low, namely 0.0% from the two C. cassiicola cultures and 0.11±0.14% from the A. niger cultures. From the control ¯asks, no acid catalysed conversion products were recovered. This means that all products recovered from the culture media are bioconversion metabolites and that all the substrate was either metabolised or evaporated. The substrate recovery and yields of biotransformation products obtained from limonene by the Penicillium digitatum cultures are displayed in Table 1. It is important to note that the conversion of (R)-(+)limonene to (R)-(+)-a-terpineol by P. digitatum took place during the ®rst part of the experiment (®rst 24-h period). The conversion of (R)-limonene (1) to (1S,2S,4R)-limonene-1,2-diol (15) by C. cassiicola on the other hand was slow. The highest yield was obtained at the end of the experiment (after 5 days of conversion). The ®nal pH of the C. cassiicola culture media was neutral, whilst the culture media of P. digitatum were slightly basic (pH 8.5) at the end of the experiment. In a second experiment the bioconversion of (R)-(+)limonene (1) and (S)-( )-limonene (19) by submerged liquid cultures was studied, comparing two fungal
strains, namely Corynespora cassiicola (COC) and Penicillium digitatum (PDD). Also the dierence between high (350 ml limonene per 50-ml culture) and low (325 ml limonene per 50-ml culture) precursor addition was compared (only for the conversion of (R)-limonene to limonene-1,2-diol by C. cassiicola). The ®rst substrate addition took place after 42 h of cultivation, the second after 50 h cultivation and the third after 66 h cultivation. Samples were taken 8 h after the ®rst substrate addition, 16 h after the second substrate addition and 8 and 24 h, respectively, after the third substrate addition. The percentage of substrate recovery and the yields of the bioconversion products are listed in Table 2 (bioconversion of (R)-limonene) and Table 3 (bioconversion of (S)-limonene). Again it was noticed that the bioconversion of both (R)-(+)- and (S)-( )-limonene to (R)-(+)- and (S)-( )a-terpineol, respectively, by P. digitatum was fast (67.2% yield of (+)-a-terpineol after 8 h when 50 ml substrate was administered to a 50-ml culture), whilst the conversion to limonene-1,2-diol by C. cassiicola was a slow process (max. yield only obtained after 3 substrate additions, i.e. at the end of the experiment (5 days)). No dierence was observed between high (350 ml) and low (325 ml) substrate addition in the conversion of limonene to limonene-1,2-diol by C. cassiicola. Substrate recovery was very low, especially for the cultures of C. cassiicola (<1%). When control experiments were run with 50 ml of (R)-limonene dissolved in 50 ml of sterile culture broth in the presence of 0.5% EtOH, it was shown that the concentration of limonene dropped to half of its original concentration after less than one hour and that all limonene was evaporated after six hours. The bioconversion of (R)-(+)- and (S)-( )-limonene by C. cassiicola and P. digitatum is given schematically in Schemes 3 and 4. 2.4. Monitoring of the bioconversion of limonene by liquid cultures during time course 2.4.1. Monitoring of the bioconversion by liquid cultures in conical ¯asks by liquid extraction The bioconversion of (R)-(+)-limonene by two Penicillium digitatum (CMC and PDD) strains and one Corynespora cassiicola strain grown as liquid cultures in conical ¯asks was monitored during time course by liquid extraction with Et2O. To the pregrown cultures of 50 ml, one substrate addition of 250 ml of a 20%-solution of (R)-limonene in EtOH (i.e. 50 ml limonene) was performed. At time t =0 min, 30 min and consequently every 90 min until 480 min, a 5-ml sample was taken from each culture and extracted with E2O. From the results displayed in Fig. 1 it can be seen that the maximum yield for the bioconversion of (R)-limonene to (R)-a-terpineol was obtained after 8 hours (almost
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
203
Table 1 Substrate recovery (%) and yield (%) of bioconversion products of R-(+)-limonene by fungi (experiment 1 with liquid cultures of Penicillium digitatum) Substrate and metabolites (R)-(+)-Limoneen (1) (R)-(+)-a-Terpineol (2) Unknown 1 exo-2-Hydroxycineol (14) Unknown 2 a b c
Recovery (%) Initial yield (%) Final yield (%) Final yield (%) Final yield (%)
CMC 1a
CMC 2
PDD 1b
PDD 2
CLEc
0.63 45.65 0.48 1.08 0.52
0.57 25.43 0.47 0.83 0.59
0.00 11.35 0.03 6.76 0.00
0.07 12.16 0.55 0.93 1.15
0.00 7.34 0.37 0.47 1.09
CMC=Penicillium digitatum isolated from spoiled mandarin. PDD=Penicillium digitatum DSM 62840. CLE=Penicillium digitatum ATCC 201167 isolated from spoiled tangerin.
Table 2 Substrate recovery (%) and yield (%) of bioconversion products after bioconversion of (R)-(+)-limonene by fungi (experiment 2 with liquid cultures) Substrate and metabolites (R)-(+)-Limonene (1) (1S,2S,4R)-Limonene-1,2-diol (15) (R)-(+)-a-Terpineol (2) cis-p-Menth-2-en-1-ol (17) trans-p-Menth-2-en-1-ol (18) neo-Dihydrocarveol (12) cis-Limonene oxide (7) Piperitone (16) a b c d
Recovery (%) Final yield (%) Initial yield (%) Final yield (%) Final yield (%) Initial yield (%) Final yield (%) Final yield (%)
COCa lowc
COC highd
PDDb high
0.85 46.7 0.00 0.35 0.00 1.70 0.00 0.27
0.07 48.6 0.00 0.35 0.10 1.04 0.14 0.27
3.16 0.00 67.2 1.27 0.68 1.70 0.29 0.00
COC=Corynespora cassiicola. PDD=Penicillium digitatum. Low=325 ml substrate addition. High=350 ml substrate addition/50 ml.
Table 3 Substrate recovery (%) and yield (%) of bioconversion products after bioconversion of (S)-( )-limonene by fungi (experiment 2 with liquid cultures) Substrate and metabolites (S)-( )-Limonene (19) (1R,2R,4S)-Limonene-1,2-diol (20) (S)-( )-a-Terpineol (22) cis-p-Menth-2-en-1-ol (21) trans-p-Menth-2-en-1-ol (23) trans-Dihydrocarvone (24) a b c d
Recovery (%) Final yield (%) Initial yield (%) Final yield (%) Final yield (%) Final yield (%)
COCa lowc
COC highd
PDDb low
0.00 50.49 0.00 1.18 0.00 0.00
0.23 49.11 0.00 0.00 0.35 0.00
5.05 0.00 16.97 1.47 0.50 0.10
COC=Corynespora cassiicola. PDD=Penicillium digitatum. Low=325 ml substrate addition. High=350 ml substrate addition/50 ml.
100% yield for strain CMC and approx. 75% yield for strain PDD). The conversion of (R)-limonene to limonene-1,2-diol by C. cassiicola however was only approx. 12% after 8 h, suggesting that this conversion is very slow. 2.4.2. Monitoring of the bioconversion by liquid cultures in SPME-vials by SPME The bioconversion of (R)-(+)-limonene by two strains of Penicillium digitatum (CMC and PDD), grown as liquid cultures in small SPME-vials, was
monitored by SPME. From 50-ml pregrown cultures of 48 h, 15 ml culture was transferred into 40-ml SPME vials and stirred (800 rpm). To each culture 10 ml of a 10%-solution of (R)-limonene in EtOH (i.e. 1 ml of limonene) was added. At time t=0 min and consequently every 90 min until 540 min, the cultures were sampled by liquid SPME. The conversion was monitored by plotting the peak area of (R)-limonene and (R)-a-terpineol of the chromatograms as a function of time (see Fig. 2). From this graph, it can be concluded that SPME is a very powerful technique to monitor the
204
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
Scheme 3. Biotransformation of (R)-(+)-limonene by liquid cultures of Penicillium digitatum and Corynespora cassiicola Ð for the names of the molecular structures, see Table 2.
Scheme 4. Biotransformation of (S)-( )-limonene by liquid cultures of Penicillium digitatum and Corynespora cassiicola Ð for the names of the molecular structures, see Table 3.
bioconversion of limonene by fungi as a function of time. However, since peak areas and not concentrations are plotted, no conversion yields can be calculated from SPME-extractions. It is clear that all (R)-limonene is either consumed or evaporated after 300 min and a maximum conversion is obtained after 360 min. 2.4.3. Monitoring of the bioconversion by spore suspensions in SPME-vials by SPME The bioconversion of (R)-(+)-limonene by spore suspensions in the course of time was monitored by SPME. Two strains of Penicillium digitatum were compared (CMC and PDD). For this experiment, spore suspen-
sions were prepared by harvesting spores from Petri dish cultures with a physiological solution (0.85% NaCl) containing 0.5% Tween 80. The spore suspensions thus obtained contained approx. 1108 spores/ml. A control experiment was also run, consisting of pure water. To each spore suspension and to the water control, 10 ml of a 10%-solution of (R)-limonene in EtOH (i.e. 1 ml of limonene) was added. At time t=0 min and consequently every 90 min until 450 min, the suspensions were sampled by liquid SPME. The conversion was monitored by plotting the peak area of (R)-limonene and (R)-a-terpineol of the chromatograms as a function of time (see Fig. 3). It can be concluded that
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
205
Fig. 1. Monitoring bioconversion of (R)-limonene by liquid fungal cultures in conical ¯asks by liquid extraction.
Fig. 2. Monitoring bioconversion of (R)-limonene by liquid cultures in small SPME vials by SPME.
the optimum conversion yield of spores of strain PDD was higher and appeared earlier (after 270 min) than the strain CMC (optimum later than 450 min). Since no conversion products were noticed in the control, it can be concluded that spores of Penicillium digitatum were able to convert (R)-(+)-limonene to (R)-(+)-a-terpineol and that SPME is a suitable screening technique for monitoring the biotranformation capacity of fungal spores.
3. Experimental 3.1. Microorganisms and cultivation More than 60 fungal strains were used in this study. They belonged to the following species: Penicillium digitatum, P. italicum, P. roquefortii, P. chrysogenum, P. lividum, Aspergillus niger, Aspergillus versicolor, Botryodiplodia malorum, Rhizopus oryzae, Beauveria
206
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
Fig. 3. Monitoring bioconversion of (R)-limonene by spore suspensions in small SPME vials by SPME.
bassiana, Cunninghamella elegans, C. blakesleeana, Hyphozima roseoniger, Corynespora cassiicola, Chaetomium cochliodes and Mortierella sp. The fungi were either isolated from spoiled fruit or contaminated media, or obtained from DSM (Deutsche Sammlung von Mikroorganismen). After screening experiments, some fungal strains were selected for further study and marked as follows: CMC (Penicillium digitatum isolated from spoiled mandarin), CLE (Penicillium digitatum Saccardo isolated from spoiled tangerine and assigned ATCC 201167), PDD (Penicillium digitatum DSM 62840), COC (Corynespora cassiicola DSM 62475), AND (Aspergillus niger DSM 821). The cultures were cultivated and conserved by periodic replications (every 2 weeks) on malt extract agar (MEA: malt extract 2%, bacteriological peptone 0.1%, glucose 2% and agar 2%±pH 5.4). 3.2. Screening of sporulated surface cultures by fungi by SPME The fungi were cultivated as small sporulated surface cultures in 40-ml SPME-vials (Supelco Inc., Bellafonte, USA) (Demyttenaere and De Kimpe, 2000; Demyttenaere et al., 2000). Therefore, the vials were ®lled with 10 ml medium (MEA), autoclaved, and inoculated with fresh spores. The vials were covered with cotton wool and the cultures were incubated at 30 C during 24 h and at room temp during 48 h, after which complete sporulation had taken place. To each sporulated surface culture, 5 ml of a solution of (R)-(+)-limonene (puriss., >99%, FLUKA) in EtOH (10%) was sprayed. The vials were covered with PTFE-Silicone Septa and Open Top Phenolic Closures (Supelco) and stored at room temperature until the start of the headspace SPME-
extraction. During extraction, the SPME ®bers were exposed to the headspace of the cultures for 5, 15 or 30 min at 25 C. To select the best sampling parameters, 40-ml SPMEvials were ®lled with 10 ml agar medium (2% agar/water) to which 5 ml of the same solution of limonene in EtOH (10%) or solutions containing limonene and a-terpineol in dierent ratios were sprayed. Dierent SPME ®bers were used, namely 100 mm polydimethylsiloxane (PDMS), 50/ 30 mm divinylbenzene/carboxen on polydimethylsiloxane, 65 mm carbowax/divinylbenzene (CWDB) and 85 mm polyacrylate (PA) (Supelco Inc., Bellafonte, USA). Dierent sampling times (5, 15, 30 and 60 min) and temperatures (10, 20, 30, 40, 50 C and room temperature) were compared. 3.3. Biotransformation of (R)-(+)-limonene by sporulated surface cultures For the experiments with sporulated surface cultures, 500-ml conical ¯asks were used, ®lled with 100 ml MEA-medium and inoculated with 0.5 ml of freshly obtained spore suspensions. For the ®rst experiment, spore suspensions of Penicillium digitatum (CMC) containing 1.8108 spores/ml and Aspergillus niger (AND) containing 4.5108 spores/ml were used as inoculum. For the second experiment with sporulated surface cultures, spore suspensions of Penicillium digitatum (CLE) containing 1.5108 spores/ml and P. digitatum (PDD) containing 9.9107 spores/ml (as counted with a Malassez cell) were used to inoculate the medium. After inoculation, the cultures were incubated at 30 C for 24 h and at room temp during the rest of the experiment. Substrate addition was carried out by spraying a
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
solution of (R)-(+)-limonene in EtOH (20%) onto the sporulated surface cultures. The biotransformation was monitored by dynamic headspace, steam distillation solvent extraction and liquid/liquid extraction as described earlier (Demyttenaere and Willemen, 1998). Control experiments with acidi®ed (pH 3.5) sterile agar medium and MEA-medium without fungi were also performed as described previously (Demyttenaere and De Pooter, 1998).
207
535 apparatus and the IR-spectrum was obtained with a Nicolet Impact 410 spectrometer. 3.7. Analysis of the samples with GC and GC±MS
The biotransformation by submerged liquid cultures was run during 5±8 days with various strains of Penicillium digitatum, Aspergillus niger and Corynespora cassiicola. The fungi were cultivated in 250-ml conical ¯asks, ®lled with 50 ml liquid medium (YMPG: yeast extract 0.5%, malt extract 1%, bacteriological peptone 0.5%, glucose 1%±pH 6.1) as described earlier (Demyttenaere and Willemen, 1998). Inoculation was done with 0.5 ml of freshly obtained spore suspensions containing 1.7108 spores/ml (AND), 9.7107 spores/ml (CMC), 9.9107 spores/ml (PDD) and 1.5108 spores/ml (CLE) as counted with a Malassez cell. The test substrate ((R)(+)- or (S)-( )-limonene) was added as a 20% (v/v) solution in absolute EtOH (125 or 250 ml of this solution were added to the cultures). At dierent time intervals, 5ml samples were taken and extracted with Et2O. This experiment was also run with control ¯asks, which contained sterile YMPG-broth which was acidi®ed to pH 3.5 by addition of acetic acid but no inoculation was executed.
GC- and GC±MS-analyses were performed as described earlier (Demyttenaere and Willemen, 1998). For the analysis of the SPME-extracts, a HP 6890 GC Plus coupled with a HP 5973 MSD (Mass Selective DetectorQuadrupole type), equipped with a CIS-4 PTV (Programmed Temperature Vaporisation) Injector (Gerstel), and a HP5-MS capillary column (30 m0.25 mm i.d.; coating thickness 0.25 mm) was used. Working conditions were: injector 250 , transfer line to MSD 250 , oven temperature: start 40 C, hold 2 min; programmed from 40 to 150 C at 5 C min 1, from 150 to 170 C at 10 C min 1 and from 170 to 250 C at 30 C min 1, hold 2 min; carrier gas (He) 1.2 ml min 1; split 1/10; ionisation: EI 70 eV; acquisition parameters: scanned m/z: 40± 180 (5±20 min), 40±250 (> 20 min). Substances were identi®ed by comparison of their mass spectra and retention indexes (KovaÂts Indexes) with those of reference substances (where possible) and by comparison with the NIST Mass Spectral Library (Version 1.6d, 1998). For elucidation of the chirality of the metabolite limonene-1,2-diol, the sample was analysed by chiral GC using a 25 m column (0.33 mm i.d.) with a permethylated CYDEX-B stationary phase (SGE) (coating thickness 0.25 mm) and it was co-injected with an authentic sample of (1S,2S,4R)-limonene-1,2-diol.
3.5. Chemical compounds
3.8. (1S,2S,4R)-Limonene-1,2-diol
The substrates used for the biotransformation experiments were (R)-(+)- and (S)-( )-limonene (puriss. 5 99%, Fluka, Belgium). As reference compounds a-terpineol (98%, Janssen Chimica, Belgium) and (1S,2S,4R)limonene-1,2-diol (kindly supplied by Dr. MarieÈt van der Werf, TNO-Voeding, Department of Applied Microbiology and Gene Technology, Zeist, The Netherlands) were used.
Yellow solid. Mp 53.9 C. 1H NMR (270 MHz, CDCl3): 1.18±1.97 (7H, m); 1.27 (3H, s); 1.73 (3H, s); 2.22±2.27 (1H, m); 3.50 (2H, s); 4.73 (2H, s). 13C NMR (68 MHz, CDCl3): 21.0 (q); 26.1 (t); 26.6 (q); 33.6 (t); 33.9 (t); 37.4 (d); 71.3 (s); 73.8 (d); 108.9 (t); 149.3 (s). IR (NaCl, cm 1): OH=3458. KovaÂts Retention Index: 1343. EIMS 70 eV m/z: (rel. int.): 152 (33), 109 (32), 108 (42), 93 (29), 82 (29), 71 (100), 69 (34), 67 (49), 43 (59), 41 (25).
3.6. Preparative separation and identi®cation of metabolites
Acknowledgements
3.4. Bioconversion by liquid cultures
For the isolation of the bioconversion product (1S,2S,4R)-limonene-1,2-diol, combined extracts were concentrated under a gentle N2-¯ow and injected in a Delsi Intersmat IGC 120 ML gas chromatograph equipped with a 3 m packed column (5% SE-30, Chromosorb W-AW 60±80, 6 mm o.d.) and a Thermal Conductivity Detector (TCD). The carrier gas was H2 (1 bar). The fraction was collected and the metabolite was identi®ed by 1H and 13C NMR (JEOL JNM-EX270, 270 MHz). The melting point was measured by a BuÈchi
Part of this work was supported by a grant form the European Community (FAIR-CT98-3559). The authors wish to express their gratitude to Dr. MarieÈt van der Werf, for supplying the reference compound.
References Abraham, W.-R., Homann, H.M.R., Kieslich, K., Reng, G., Stumpf, B., 1985. Microbial transformation of some monoterpenoids and
208
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
sesquiterpenoids. In: Porter, R., Clark, S. (Eds.), Enzymes Org. Synth. Ciba Found. Symp. 1985, Vol. 111. Pitman Press, London, pp. 146±160. Abraham, W.-R., Kieslich, K., Reng, H., Stumpf, B. 1984. Formation and production of 1,2-trans-glycols from various monoterpenes with 1-menthen-skeleton by microbial transformation with Diplodia gossypina In: Eur. Congr. Biotechnol., 3rd, Vol. 1. Verlag Chemie, Weinheim, Germany, pp. 245-248. Abraham, W.-R., Stumpf, B., Kieslich, K., 1986. Microbial transformations of terpenoids with 1-p-menthene skeleton. Applied Microbiology and Biotechnology 24, 24±30. Arfmann, H.-A., Abraham, W.-R., Kieslich, K., 1987. Microbial !hydroxylation of terpenoids. In: Neijssel et al. (Eds.), Proceedings 4th European Congress on Biotechnology, Vol. 2. Elsevier, Amsterdam, pp. 212±215. Arfmann, H.-A., Abraham, W.-R., Kieslich, K., 1988. Microbial !hydroxylation of trans-nerolidol and structurally related sesquiterpenoids. Biocatalysis 2, 59±67. Bauer, K., Garbe, D., Surburg, H. (Eds.), 1990. Common Fragrance and Flavor Materials. Preparation, Properties and Uses. 2nd Edition. VCH Publishers, New York. Bowen, E.R., 1975. Potential by-products from microbial transformation of D-limonene. In: Proceedings of the Florida State Horticultural Society, Miami Beach, pp. 304-308. Cadwallader, K.R., Braddock, R.J., 1992. Enzymatic hydration of (4R) -(+)-limonene to (4R)-(+)-a-terpineol. Dev. Food Sci. 29, 571-584. Cadwallader, K.R., Braddock, R.J., Parish, M.E., Higgins, D.P., 1989. Bioconversion of (+)-limonene by Pseudomonas gladioli. Journal of Food Science 54, 1241±1245. Demyttenaere, J., De Kimpe, N., 2000. Biotransformation of geraniol and nerol by sporulated surface cultures of Aspergillus niger. In: Schieberle, P., Engel, K.-H. (Eds.), Frontiers of Flavour Science, Proceedings of the Ninth Weurman Flavour Research Symposium 22±25 June 1999, Freising, Germany, pp. 400-404. Demyttenaere, J.C.R., De Pooter, H.L., 1998. Biotransformation of citral and nerol by spores of Penicillium digitatum. Flavour Fragr. J. 13, 173±176. Demyttenaere, J.C.R., Herrera, M., del, C., De Kimpe, N., 2000. Biotransformation of geraniol, nerol and citral by sporulated surface cultures of Aspergillus niger and Penicillium sp. Phytochemistry 55, 363±373.
Demyttenaere, J.C.R., Willemen, H.M., 1998. Biotransformation of linalool to furanoid and pyranoid linalool oxides by Aspergillus niger. Phytochemistry 47, 1029±1036. Dhavalikar, R.S., Bhattacharyya, P.K., 1966. Microbiological transformations of terpenes: part VIII Ð fermentation of limonene by a soil pseudomonad. Indian Journal of Biochemistry 3, 144± 157. Fenaroli, G., 1975. In: Furia, T.E., Bellanca, N. (Eds.), Handbook of Flavor Ingredients, 2nd Ed. CRC Press, Cleveland, Ohio. Kraidman, G., Mukherjee, B.B., Hill, I.D., 1969. Conversion of Dlimonene into an optically active isomer of a-terpineol by a Cladosporium species. Bacteriological Proceedings, 63. Krasnobajew, V., 1984. Terpenoids. In: Rehm, H.-J., Reed, G. (Eds.), Biotechnology, Vol. 6a. Verlag Chemie, Weinheim, pp. 97-125. Madyastha, K.M., Bhattacharyya, P.K., Vaidyanathan, C.S., 1977. Metabolism of a monoterpene alcohol, linalool, by a soil pseudomonad. Canadian Journal of Microbiology 23, 230±239. Mattison, J.E., McDowell, L.L., Baum, R.H., 1971. Cometabolism of selected monoterpenoids by fungi associated with monoterpenoidcontaining plants. Bacteriological Proceedings, 141. Mukherjee, B.B., Kraidman, G., Hill, I.D., 1973. Synthesis of glycols by microbial transformation of some monocyclic terpenes. Applied Microbiology 25, 447±453. Noma, Y., Yamasaki, S., Asakawa, Y., 1992. Biotransformation of limonene and related compounds by Aspergillus cellulosae. Phytochemistry 31, 2725±2727. Rama Devi, J., Bhattacharyya, P.K., 1977. Microbiological transformations of terpenes: part xxiii Ð fermentation of geraniol, nerol and limonene by a soil pseudomonad, Pseudomonas incognita (linalool strain). Indian Journal of Biochemistry and Biophysics 14, 288±291. Rama Devi, J., Bhattacharyya, P.K., 1978. Molecular rearrangements in the microbiological transformations of terpenes and the chemical logic of microbial processes. Journal of the Indian Chemistry Society 55, 1131±1137. Tan, Q., Day, D.F., 1998a. Organic co-solvent eects on the bioconversion of (R)-(+)-limonene to (R)-(+)-a-terpineol. Process Biochemistry 33, 755±761. Tan, Q., Day, D.F., 1998b. Bioconversion of limonene to a-terpineol by immobilized Penicillium digitatum. Applied Microbiology and Biotechnology 49, 96±101.
www.elsevier.com/locate/phytochem
Biotransformation of (R)-(+)- and (S)-( )-limonene by fungi and the use of solid phase microextraction for screening Jan C.R. Demyttenaere *, Kristof Van Belleghem, Norbert De Kimpe Department of Organic Chemistry, Faculty of Agricultural and Applied Biological Sciences, Ghent University, Coupure Links 653, B-9000 Ghent, Belgium Received 29 September 2000; received in revised form 1 December 2000
Abstract The biotransformation of (R)-(+)- and (S)-( )-limonene by fungi was investigated. More than 60 fungal cultures were screened for their ability to bioconvert the substrate, using solid phase microextraction as the monitoring technique. After screening, the best fungal strains were selected for further study and were grown as sporulated surface cultures in conical ¯asks and as submerged liquid cultures. It was found that (+)- and ( )-limonene were converted by Penicillium digitatum to a-terpineol (main metabolite), cis- and trans-p-menth-2-en-1-ol, neodihydrocarveol and limonene oxide (minor metabolites) using liquid cultures. The bioconversion of (R)-(+)- and (S)-( )-limonene by Corynespora cassiicola yielded (1S,2S,4R)- and (1R,2R,4S)-limonene-1,2-diol respectively. The bioconversions by liquid cultures were also monitored by solid phase microextraction as a function of time. The optimum conversion of limonene to a-terpineol by Penicillium digitatum was obtained after 8 hours (yield up to 100%). Since an important pH-decrease was noticed in some liquid broths, the stability of limonene under acidic conditions was investigated. No acid catalysed conversion products were recovered after 8 days from control ¯asks at pH 3.5 containing limonene. # 2001 Elsevier Science Ltd. All rights reserved. Keywords: Penicillium digitatum; Corynespora cassiicola; Fungi; Fungal spores; Biotransformation; Bioconversion; Limonene; a-Terpineol; (1S,2S,4R)-Limonene-1,2-diol; (1R,2R,4S)-Limonene-1,2-diol; SPME
1. Introduction Much work has been done on the biotransformation of the inexpensive hydrocarbon monoterpene limonene (Bowen, 1975), which is, besides a-pinene, the most widely distributed terpene in nature (Krasnobajew, 1984). The (+)-isomer is present in citrus peel oils at a concentration of over 90%; a low concentration of the ( )-isomer is found in oils from the Mentha species and conifers (Bauer et al., 1990). The microbial conversion of limonene is very well documented, the ®rst data date back to the sixties. A soil pseudomonad was isolated by an enrichment culture technique on limonene as the sole source of carbon (Dhavalikar and Bhattacharyya, 1966) and was able to convert the substrate to a large number of neutral and acidic products. Another group isolated a * Corresponding author. Tel.: +32-9-264-549-64; fax: +32-9-26462-43. E-mail address: [email protected] (J.C.R. Demyttenaere).
Pseudomonas incognita by an enrichment technique on the monoterpene alcohol linalool (Madyastha et al., 1977). The metabolism of limonene by this bacterium was investigated (Rama Devi and Bhattacharyya, 1977). After fermentation the medium yielded as the main product a crystallic acid, perillic acid, together with unmetabolised limonene and some oxygenated compounds, such as dihydrocarvone, carvone and carveol. The biotransformation of limonene by another Pseudomonas strain, P. gladioli was also reported (Cadwallader et al., 1989; Cadwallader and Braddock, 1992). Major conversion products from (+)-limonene were (+)-aterpineol and (+)-perillic acid. This was the ®rst time that the microbial conversion of limonene to a-terpineol was reported. a-Terpineol is widely distributed in nature and is one of the most commonly used perfume chemicals (Fenaroli, 1975). The ®rst data on fungal bioconversion of limonene date back to the late sixties. A Cladosporium sp. designated as T7 was isolated on limonene and was able to convert the substrate to translimonene-1,2-diol (Mukherjee et al., 1973). Another
0031-9422/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S0031-9422(00)00499-4
200
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
Cladosporium, designated T12, converted limonene into a-terpineol (Kraidman et al., 1969). Fungi, isolated from rotting orange rind, were also found to be able to convert limonene to a-terpineol (Mattison et al., 1971). Bowen (1975) isolated two common citrus moulds, Penicillium italicum and P. digitatum, responsible for the postharvest diseases of citrus fruits. Fermentation of P. italicum on limonene yielded cis- and trans-carveol (26%) as main products, together with (+)-carvone (6%) and minor metabolites. Conversion by P. digitatum yielded the same products in lower yields. The bioconversion of limonene by Aspergillus niger has also been described (Rama Devi and Bhattacharyya, 1978). This fungus was able to carry out three types of oxygenative rearrangements. More recently, a Japanese group studied the biotransformation of limonene and related compounds (1-methylcyclohexene and cyclohexene) by Aspergillus cellulosae M77 (Noma et al., 1992). (+)-Limonene was mainly converted to (+)-isopiperitenone (19%), (+)-limonene-1,2trans-diol (21%), (+)-cis-carveol (5%) and (+)-perillyl alcohol (12%). The production of isopiperitenone from limonene was published for the ®rst time. The production of glycols from limonene and other terpenes with a 1-menthene skeleton was reported (Abraham et al., 1984). The most appropriate strains were Corynespora cassiicola and Diplodia gossypina. The same group (Abraham et al., 1985) also investigated the biotransformation of (R)-limonene to (R)-a-terpineol by the fungus Penicillium digitatum. An extensive overview on the microbial transformations of terpenoids with a 1p-menthene skeleton was published by this group (Abraham et al., 1986). One of the most recent publications in this area reported the bioconversion of limonene to a-terpineol by immobilised P. digitatum (Tan and Day, 1998a) and studied the eect of organic cosolvents on the bioconversion of (R)-(+)-limonene to (R)-(+)-a-terpineol (Tan and Day, 1998b). In all these examples submerged liquid cultures were used. The bioconversion of limonene by fungal spores has not been published to date. This paper reports the biotransformation of (R)-(+)(1) and (S)-( )-limonene (19) by sporulated surface cultures and liquid cultures of fungi and the use of solid phase microextraction (SPME) for fast screening. 2. Results and discussion 2.1. Screening of sporulated surface cultures by fungi by SPME More than 60 fungal strains, grown as sporulated surface cultures, were screened for their ability to bioconvert the substrate (R)-(+)-limonene (1), using solid phase microextraction (SPME) as the monitoring
technique. Therefore, a method was developed to cultivate sporulated surface cultures of the fungi in small vials, and the SPME-parameters were optimised. Different SPME-®bers, extraction times and temperatures were compared. Extraction of volatiles (limonene and metabolites) was done by headspace SPME. It was found that the best SPME ®ber for extraction of limonene and its metabolites was 50/30 mm divinylbenzene/ carboxen on polydimethylsiloxane. The optimum adsorption time was 30 min. The best extraction temperature was room temperature, but because of temperature ¯uctuation, a constant adsorption temperature of 25 C was chosen as optimum. A desorption temperature of 250 C and desorption time of 2 min was sucient to desorb all volatiles into the gaschromatographic inlet. The most interesting conversion of (R)-(+)-limonene (1) was the formation of (R)-(+)-a-terpineol (2) by various Penicillium species. Other metabolites recovered were g-terpinene (3), terpinolene (4), a-phellandrene (5), endo-fenchol (11), menth-3-en-1-ol (6), p-cymene (9), neo-dihydrocarveol (12), cis-limonene oxide (7), translimonene oxide (8) and perillyl alcohol (10) (see Scheme 1). The most interesting strains were P. digitatum (both isolated (CMC) and obtained, DSM 62840) and Corynespora cassiicola DSM 62475. 2.2. Biotransformation of (R)-(+)-limonene by sporulated surface cultures After screening, the best fungal strains were selected for further study and were grown as sporulated surface cultures in conical ¯asks. The biotransformation was monitored by dynamic headspace, steam distillation solvent extraction and liquid/liquid extraction. In a ®rst experiment, the bioconversion of (R)-(+)limonene (1) by two strains was compared, namely Penicillium digitatum (CMC), selected as interesting strain, based on the SPME screening results, and Aspergillus niger (AND), since A. niger has been used successfully in many bioconversion reactions (Rama Devi and Bhattacharrya, 1978; Arfmann et al., 1987, 1988; Demyttenaere and Willemen, 1998; Demyttenaere and De Kimpe, 2000; Demyttenaere et al., 2000). To 10day old sporulated surface cultures, 100 ml of a 20% limonene in EtOH solution was sprayed. Two more substrate additions took place after 15 and 17 days respectively. It was found that Aspergillus niger was not able to convert limonene, whereas Penicillium digitatum produced (R)-(+)-a-terpineol (2) from limonene with low yields (0.7±1.4%). In a second experiment, the bioconversion of (R)-(+)-limonene (1) by two strains of Penicillium digitatum, namely strain CLE and strain PDD was compared in duplicate. The ®rst substrate addition (125 ml of a 20% limonene in EtOH solution) was performed after 7 days. Two more substrate
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
201
Scheme 1. Biotransformation of (R)-(+)-limonene by sporulated surface cultures of fungi Ð monitoring by SPME.
additions took place after 12 and 14 days resp. From the headspace extracts of the surface cultures, it could be seen that up to 44% of the substrate (R)-limonene (1) was recovered unchanged. The yield of conversion to (R)-a-terpineol (2) was 1.1 and 1.3% for the two P. digitatum cultures of strain CLE resp. and 2.7 and 2.9% for the two cultures of strain PDD resp. Minor metabolites were 1,3,5-p-menthatriene (13) (0.2±0.6%), cislimonene oxide (7) (0.1±0.2%) and exo-2-hydroxycineol (14) (0.1±0.2%) (see Scheme 2). It is important to note that the pH of the surface cultures dropped to pH 3 after the cultures were subjected to steam distillation/solvent extraction. Therefore, a control experiment was run to check the degradation of limonene in sterile agar medium and MEA-medium,
acidi®ed to pH 3.5 (at lower pH, the agar did not solidify), during headspace sampling and steam distillation/ solvent extraction. The substrate (R)-limonene happened to be very stable: only small traces (0.14%) of bmyrcene were recovered from the steam distillate. No acid catalysed conversion products from limonene were recovered from the headspace extracts. It can be concluded that sporulated surface cultures of P. digitatum are able to convert the substrate (R)(+)-limonene (1) to (R)-a-terpineol (2). Since the very high volatility of the substrate, limonene, a high amount of substrate loss due to evaporation during headspace extraction was noticed, resulting in high substrate recovery. Therefore, when dynamic headspace sampling is compared with headspace SPME-sampling, the latter
Scheme 2. Bioconversion of (R)-(+)-limonene by sporulated surface cultures of Penicillium digitatum.
202
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
method is more sensitive and ecient for fast screening. 2.3. Bioconversion of (R)-(+)- and (S)-( )-limonene by liquid cultures In a ®rst experiment the bioconversion of (R)-(+)limonene by submerged liquid cultures was studied, comparing ®ve fungal strains, namely Corynespora cassiicola (marked COC), Aspergillus niger (marked AND) and three dierent strains of Penicillium digitatum (marked CLE, CMC and PDD respectively). All strains were cultivated in duplicate, except strain CLE. As a control, the stability of the precursor was checked in acidi®ed broth (pH 3.5) that was not inoculated but kept under sterile conditions. Over a period of 8 days, 325 ml of (R)-(+)-limonene (each time 125 ml of a 20% solution of limonene in EtOH) was added to cultures of 50 ml each. The ®rst substrate addition was performed after 46 h of cultivation, the second after 5 days and the third after 7 days. Samples were taken 24 h after the ®rst substrate addition, and 22.5 h after the second and third substrate addition, respectively. Corynespora cassiicola was able to transform the substrate to (1S,2S,4R)-limonene-1,2diol (15) with a ®nal yield of 46.9±60.6%. This biotransformation con®rms previously published literature data (Abraham et al., 1986). All P. digitatum cultures converted (R)-(+)-limonene (1) to (R)-(+)-a-terpineol (2) with an initial yield up to 45.7% (after 24 h). These results are in agreement with literature data (Abraham et al., 1985). Aspergillus niger, however, was not able to convert the substrate. The substrate recovery was very low, namely 0.0% from the two C. cassiicola cultures and 0.11±0.14% from the A. niger cultures. From the control ¯asks, no acid catalysed conversion products were recovered. This means that all products recovered from the culture media are bioconversion metabolites and that all the substrate was either metabolised or evaporated. The substrate recovery and yields of biotransformation products obtained from limonene by the Penicillium digitatum cultures are displayed in Table 1. It is important to note that the conversion of (R)-(+)limonene to (R)-(+)-a-terpineol by P. digitatum took place during the ®rst part of the experiment (®rst 24-h period). The conversion of (R)-limonene (1) to (1S,2S,4R)-limonene-1,2-diol (15) by C. cassiicola on the other hand was slow. The highest yield was obtained at the end of the experiment (after 5 days of conversion). The ®nal pH of the C. cassiicola culture media was neutral, whilst the culture media of P. digitatum were slightly basic (pH 8.5) at the end of the experiment. In a second experiment the bioconversion of (R)-(+)limonene (1) and (S)-( )-limonene (19) by submerged liquid cultures was studied, comparing two fungal
strains, namely Corynespora cassiicola (COC) and Penicillium digitatum (PDD). Also the dierence between high (350 ml limonene per 50-ml culture) and low (325 ml limonene per 50-ml culture) precursor addition was compared (only for the conversion of (R)-limonene to limonene-1,2-diol by C. cassiicola). The ®rst substrate addition took place after 42 h of cultivation, the second after 50 h cultivation and the third after 66 h cultivation. Samples were taken 8 h after the ®rst substrate addition, 16 h after the second substrate addition and 8 and 24 h, respectively, after the third substrate addition. The percentage of substrate recovery and the yields of the bioconversion products are listed in Table 2 (bioconversion of (R)-limonene) and Table 3 (bioconversion of (S)-limonene). Again it was noticed that the bioconversion of both (R)-(+)- and (S)-( )-limonene to (R)-(+)- and (S)-( )a-terpineol, respectively, by P. digitatum was fast (67.2% yield of (+)-a-terpineol after 8 h when 50 ml substrate was administered to a 50-ml culture), whilst the conversion to limonene-1,2-diol by C. cassiicola was a slow process (max. yield only obtained after 3 substrate additions, i.e. at the end of the experiment (5 days)). No dierence was observed between high (350 ml) and low (325 ml) substrate addition in the conversion of limonene to limonene-1,2-diol by C. cassiicola. Substrate recovery was very low, especially for the cultures of C. cassiicola (<1%). When control experiments were run with 50 ml of (R)-limonene dissolved in 50 ml of sterile culture broth in the presence of 0.5% EtOH, it was shown that the concentration of limonene dropped to half of its original concentration after less than one hour and that all limonene was evaporated after six hours. The bioconversion of (R)-(+)- and (S)-( )-limonene by C. cassiicola and P. digitatum is given schematically in Schemes 3 and 4. 2.4. Monitoring of the bioconversion of limonene by liquid cultures during time course 2.4.1. Monitoring of the bioconversion by liquid cultures in conical ¯asks by liquid extraction The bioconversion of (R)-(+)-limonene by two Penicillium digitatum (CMC and PDD) strains and one Corynespora cassiicola strain grown as liquid cultures in conical ¯asks was monitored during time course by liquid extraction with Et2O. To the pregrown cultures of 50 ml, one substrate addition of 250 ml of a 20%-solution of (R)-limonene in EtOH (i.e. 50 ml limonene) was performed. At time t =0 min, 30 min and consequently every 90 min until 480 min, a 5-ml sample was taken from each culture and extracted with E2O. From the results displayed in Fig. 1 it can be seen that the maximum yield for the bioconversion of (R)-limonene to (R)-a-terpineol was obtained after 8 hours (almost
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
203
Table 1 Substrate recovery (%) and yield (%) of bioconversion products of R-(+)-limonene by fungi (experiment 1 with liquid cultures of Penicillium digitatum) Substrate and metabolites (R)-(+)-Limoneen (1) (R)-(+)-a-Terpineol (2) Unknown 1 exo-2-Hydroxycineol (14) Unknown 2 a b c
Recovery (%) Initial yield (%) Final yield (%) Final yield (%) Final yield (%)
CMC 1a
CMC 2
PDD 1b
PDD 2
CLEc
0.63 45.65 0.48 1.08 0.52
0.57 25.43 0.47 0.83 0.59
0.00 11.35 0.03 6.76 0.00
0.07 12.16 0.55 0.93 1.15
0.00 7.34 0.37 0.47 1.09
CMC=Penicillium digitatum isolated from spoiled mandarin. PDD=Penicillium digitatum DSM 62840. CLE=Penicillium digitatum ATCC 201167 isolated from spoiled tangerin.
Table 2 Substrate recovery (%) and yield (%) of bioconversion products after bioconversion of (R)-(+)-limonene by fungi (experiment 2 with liquid cultures) Substrate and metabolites (R)-(+)-Limonene (1) (1S,2S,4R)-Limonene-1,2-diol (15) (R)-(+)-a-Terpineol (2) cis-p-Menth-2-en-1-ol (17) trans-p-Menth-2-en-1-ol (18) neo-Dihydrocarveol (12) cis-Limonene oxide (7) Piperitone (16) a b c d
Recovery (%) Final yield (%) Initial yield (%) Final yield (%) Final yield (%) Initial yield (%) Final yield (%) Final yield (%)
COCa lowc
COC highd
PDDb high
0.85 46.7 0.00 0.35 0.00 1.70 0.00 0.27
0.07 48.6 0.00 0.35 0.10 1.04 0.14 0.27
3.16 0.00 67.2 1.27 0.68 1.70 0.29 0.00
COC=Corynespora cassiicola. PDD=Penicillium digitatum. Low=325 ml substrate addition. High=350 ml substrate addition/50 ml.
Table 3 Substrate recovery (%) and yield (%) of bioconversion products after bioconversion of (S)-( )-limonene by fungi (experiment 2 with liquid cultures) Substrate and metabolites (S)-( )-Limonene (19) (1R,2R,4S)-Limonene-1,2-diol (20) (S)-( )-a-Terpineol (22) cis-p-Menth-2-en-1-ol (21) trans-p-Menth-2-en-1-ol (23) trans-Dihydrocarvone (24) a b c d
Recovery (%) Final yield (%) Initial yield (%) Final yield (%) Final yield (%) Final yield (%)
COCa lowc
COC highd
PDDb low
0.00 50.49 0.00 1.18 0.00 0.00
0.23 49.11 0.00 0.00 0.35 0.00
5.05 0.00 16.97 1.47 0.50 0.10
COC=Corynespora cassiicola. PDD=Penicillium digitatum. Low=325 ml substrate addition. High=350 ml substrate addition/50 ml.
100% yield for strain CMC and approx. 75% yield for strain PDD). The conversion of (R)-limonene to limonene-1,2-diol by C. cassiicola however was only approx. 12% after 8 h, suggesting that this conversion is very slow. 2.4.2. Monitoring of the bioconversion by liquid cultures in SPME-vials by SPME The bioconversion of (R)-(+)-limonene by two strains of Penicillium digitatum (CMC and PDD), grown as liquid cultures in small SPME-vials, was
monitored by SPME. From 50-ml pregrown cultures of 48 h, 15 ml culture was transferred into 40-ml SPME vials and stirred (800 rpm). To each culture 10 ml of a 10%-solution of (R)-limonene in EtOH (i.e. 1 ml of limonene) was added. At time t=0 min and consequently every 90 min until 540 min, the cultures were sampled by liquid SPME. The conversion was monitored by plotting the peak area of (R)-limonene and (R)-a-terpineol of the chromatograms as a function of time (see Fig. 2). From this graph, it can be concluded that SPME is a very powerful technique to monitor the
204
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
Scheme 3. Biotransformation of (R)-(+)-limonene by liquid cultures of Penicillium digitatum and Corynespora cassiicola Ð for the names of the molecular structures, see Table 2.
Scheme 4. Biotransformation of (S)-( )-limonene by liquid cultures of Penicillium digitatum and Corynespora cassiicola Ð for the names of the molecular structures, see Table 3.
bioconversion of limonene by fungi as a function of time. However, since peak areas and not concentrations are plotted, no conversion yields can be calculated from SPME-extractions. It is clear that all (R)-limonene is either consumed or evaporated after 300 min and a maximum conversion is obtained after 360 min. 2.4.3. Monitoring of the bioconversion by spore suspensions in SPME-vials by SPME The bioconversion of (R)-(+)-limonene by spore suspensions in the course of time was monitored by SPME. Two strains of Penicillium digitatum were compared (CMC and PDD). For this experiment, spore suspen-
sions were prepared by harvesting spores from Petri dish cultures with a physiological solution (0.85% NaCl) containing 0.5% Tween 80. The spore suspensions thus obtained contained approx. 1108 spores/ml. A control experiment was also run, consisting of pure water. To each spore suspension and to the water control, 10 ml of a 10%-solution of (R)-limonene in EtOH (i.e. 1 ml of limonene) was added. At time t=0 min and consequently every 90 min until 450 min, the suspensions were sampled by liquid SPME. The conversion was monitored by plotting the peak area of (R)-limonene and (R)-a-terpineol of the chromatograms as a function of time (see Fig. 3). It can be concluded that
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
205
Fig. 1. Monitoring bioconversion of (R)-limonene by liquid fungal cultures in conical ¯asks by liquid extraction.
Fig. 2. Monitoring bioconversion of (R)-limonene by liquid cultures in small SPME vials by SPME.
the optimum conversion yield of spores of strain PDD was higher and appeared earlier (after 270 min) than the strain CMC (optimum later than 450 min). Since no conversion products were noticed in the control, it can be concluded that spores of Penicillium digitatum were able to convert (R)-(+)-limonene to (R)-(+)-a-terpineol and that SPME is a suitable screening technique for monitoring the biotranformation capacity of fungal spores.
3. Experimental 3.1. Microorganisms and cultivation More than 60 fungal strains were used in this study. They belonged to the following species: Penicillium digitatum, P. italicum, P. roquefortii, P. chrysogenum, P. lividum, Aspergillus niger, Aspergillus versicolor, Botryodiplodia malorum, Rhizopus oryzae, Beauveria
206
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
Fig. 3. Monitoring bioconversion of (R)-limonene by spore suspensions in small SPME vials by SPME.
bassiana, Cunninghamella elegans, C. blakesleeana, Hyphozima roseoniger, Corynespora cassiicola, Chaetomium cochliodes and Mortierella sp. The fungi were either isolated from spoiled fruit or contaminated media, or obtained from DSM (Deutsche Sammlung von Mikroorganismen). After screening experiments, some fungal strains were selected for further study and marked as follows: CMC (Penicillium digitatum isolated from spoiled mandarin), CLE (Penicillium digitatum Saccardo isolated from spoiled tangerine and assigned ATCC 201167), PDD (Penicillium digitatum DSM 62840), COC (Corynespora cassiicola DSM 62475), AND (Aspergillus niger DSM 821). The cultures were cultivated and conserved by periodic replications (every 2 weeks) on malt extract agar (MEA: malt extract 2%, bacteriological peptone 0.1%, glucose 2% and agar 2%±pH 5.4). 3.2. Screening of sporulated surface cultures by fungi by SPME The fungi were cultivated as small sporulated surface cultures in 40-ml SPME-vials (Supelco Inc., Bellafonte, USA) (Demyttenaere and De Kimpe, 2000; Demyttenaere et al., 2000). Therefore, the vials were ®lled with 10 ml medium (MEA), autoclaved, and inoculated with fresh spores. The vials were covered with cotton wool and the cultures were incubated at 30 C during 24 h and at room temp during 48 h, after which complete sporulation had taken place. To each sporulated surface culture, 5 ml of a solution of (R)-(+)-limonene (puriss., >99%, FLUKA) in EtOH (10%) was sprayed. The vials were covered with PTFE-Silicone Septa and Open Top Phenolic Closures (Supelco) and stored at room temperature until the start of the headspace SPME-
extraction. During extraction, the SPME ®bers were exposed to the headspace of the cultures for 5, 15 or 30 min at 25 C. To select the best sampling parameters, 40-ml SPMEvials were ®lled with 10 ml agar medium (2% agar/water) to which 5 ml of the same solution of limonene in EtOH (10%) or solutions containing limonene and a-terpineol in dierent ratios were sprayed. Dierent SPME ®bers were used, namely 100 mm polydimethylsiloxane (PDMS), 50/ 30 mm divinylbenzene/carboxen on polydimethylsiloxane, 65 mm carbowax/divinylbenzene (CWDB) and 85 mm polyacrylate (PA) (Supelco Inc., Bellafonte, USA). Dierent sampling times (5, 15, 30 and 60 min) and temperatures (10, 20, 30, 40, 50 C and room temperature) were compared. 3.3. Biotransformation of (R)-(+)-limonene by sporulated surface cultures For the experiments with sporulated surface cultures, 500-ml conical ¯asks were used, ®lled with 100 ml MEA-medium and inoculated with 0.5 ml of freshly obtained spore suspensions. For the ®rst experiment, spore suspensions of Penicillium digitatum (CMC) containing 1.8108 spores/ml and Aspergillus niger (AND) containing 4.5108 spores/ml were used as inoculum. For the second experiment with sporulated surface cultures, spore suspensions of Penicillium digitatum (CLE) containing 1.5108 spores/ml and P. digitatum (PDD) containing 9.9107 spores/ml (as counted with a Malassez cell) were used to inoculate the medium. After inoculation, the cultures were incubated at 30 C for 24 h and at room temp during the rest of the experiment. Substrate addition was carried out by spraying a
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
solution of (R)-(+)-limonene in EtOH (20%) onto the sporulated surface cultures. The biotransformation was monitored by dynamic headspace, steam distillation solvent extraction and liquid/liquid extraction as described earlier (Demyttenaere and Willemen, 1998). Control experiments with acidi®ed (pH 3.5) sterile agar medium and MEA-medium without fungi were also performed as described previously (Demyttenaere and De Pooter, 1998).
207
535 apparatus and the IR-spectrum was obtained with a Nicolet Impact 410 spectrometer. 3.7. Analysis of the samples with GC and GC±MS
The biotransformation by submerged liquid cultures was run during 5±8 days with various strains of Penicillium digitatum, Aspergillus niger and Corynespora cassiicola. The fungi were cultivated in 250-ml conical ¯asks, ®lled with 50 ml liquid medium (YMPG: yeast extract 0.5%, malt extract 1%, bacteriological peptone 0.5%, glucose 1%±pH 6.1) as described earlier (Demyttenaere and Willemen, 1998). Inoculation was done with 0.5 ml of freshly obtained spore suspensions containing 1.7108 spores/ml (AND), 9.7107 spores/ml (CMC), 9.9107 spores/ml (PDD) and 1.5108 spores/ml (CLE) as counted with a Malassez cell. The test substrate ((R)(+)- or (S)-( )-limonene) was added as a 20% (v/v) solution in absolute EtOH (125 or 250 ml of this solution were added to the cultures). At dierent time intervals, 5ml samples were taken and extracted with Et2O. This experiment was also run with control ¯asks, which contained sterile YMPG-broth which was acidi®ed to pH 3.5 by addition of acetic acid but no inoculation was executed.
GC- and GC±MS-analyses were performed as described earlier (Demyttenaere and Willemen, 1998). For the analysis of the SPME-extracts, a HP 6890 GC Plus coupled with a HP 5973 MSD (Mass Selective DetectorQuadrupole type), equipped with a CIS-4 PTV (Programmed Temperature Vaporisation) Injector (Gerstel), and a HP5-MS capillary column (30 m0.25 mm i.d.; coating thickness 0.25 mm) was used. Working conditions were: injector 250 , transfer line to MSD 250 , oven temperature: start 40 C, hold 2 min; programmed from 40 to 150 C at 5 C min 1, from 150 to 170 C at 10 C min 1 and from 170 to 250 C at 30 C min 1, hold 2 min; carrier gas (He) 1.2 ml min 1; split 1/10; ionisation: EI 70 eV; acquisition parameters: scanned m/z: 40± 180 (5±20 min), 40±250 (> 20 min). Substances were identi®ed by comparison of their mass spectra and retention indexes (KovaÂts Indexes) with those of reference substances (where possible) and by comparison with the NIST Mass Spectral Library (Version 1.6d, 1998). For elucidation of the chirality of the metabolite limonene-1,2-diol, the sample was analysed by chiral GC using a 25 m column (0.33 mm i.d.) with a permethylated CYDEX-B stationary phase (SGE) (coating thickness 0.25 mm) and it was co-injected with an authentic sample of (1S,2S,4R)-limonene-1,2-diol.
3.5. Chemical compounds
3.8. (1S,2S,4R)-Limonene-1,2-diol
The substrates used for the biotransformation experiments were (R)-(+)- and (S)-( )-limonene (puriss. 5 99%, Fluka, Belgium). As reference compounds a-terpineol (98%, Janssen Chimica, Belgium) and (1S,2S,4R)limonene-1,2-diol (kindly supplied by Dr. MarieÈt van der Werf, TNO-Voeding, Department of Applied Microbiology and Gene Technology, Zeist, The Netherlands) were used.
Yellow solid. Mp 53.9 C. 1H NMR (270 MHz, CDCl3): 1.18±1.97 (7H, m); 1.27 (3H, s); 1.73 (3H, s); 2.22±2.27 (1H, m); 3.50 (2H, s); 4.73 (2H, s). 13C NMR (68 MHz, CDCl3): 21.0 (q); 26.1 (t); 26.6 (q); 33.6 (t); 33.9 (t); 37.4 (d); 71.3 (s); 73.8 (d); 108.9 (t); 149.3 (s). IR (NaCl, cm 1): OH=3458. KovaÂts Retention Index: 1343. EIMS 70 eV m/z: (rel. int.): 152 (33), 109 (32), 108 (42), 93 (29), 82 (29), 71 (100), 69 (34), 67 (49), 43 (59), 41 (25).
3.6. Preparative separation and identi®cation of metabolites
Acknowledgements
3.4. Bioconversion by liquid cultures
For the isolation of the bioconversion product (1S,2S,4R)-limonene-1,2-diol, combined extracts were concentrated under a gentle N2-¯ow and injected in a Delsi Intersmat IGC 120 ML gas chromatograph equipped with a 3 m packed column (5% SE-30, Chromosorb W-AW 60±80, 6 mm o.d.) and a Thermal Conductivity Detector (TCD). The carrier gas was H2 (1 bar). The fraction was collected and the metabolite was identi®ed by 1H and 13C NMR (JEOL JNM-EX270, 270 MHz). The melting point was measured by a BuÈchi
Part of this work was supported by a grant form the European Community (FAIR-CT98-3559). The authors wish to express their gratitude to Dr. MarieÈt van der Werf, for supplying the reference compound.
References Abraham, W.-R., Homann, H.M.R., Kieslich, K., Reng, G., Stumpf, B., 1985. Microbial transformation of some monoterpenoids and
208
J.C.R. Demyttenaere et al. / Phytochemistry 57 (2001) 199±208
sesquiterpenoids. In: Porter, R., Clark, S. (Eds.), Enzymes Org. Synth. Ciba Found. Symp. 1985, Vol. 111. Pitman Press, London, pp. 146±160. Abraham, W.-R., Kieslich, K., Reng, H., Stumpf, B. 1984. Formation and production of 1,2-trans-glycols from various monoterpenes with 1-menthen-skeleton by microbial transformation with Diplodia gossypina In: Eur. Congr. Biotechnol., 3rd, Vol. 1. Verlag Chemie, Weinheim, Germany, pp. 245-248. Abraham, W.-R., Stumpf, B., Kieslich, K., 1986. Microbial transformations of terpenoids with 1-p-menthene skeleton. Applied Microbiology and Biotechnology 24, 24±30. Arfmann, H.-A., Abraham, W.-R., Kieslich, K., 1987. Microbial !hydroxylation of terpenoids. In: Neijssel et al. (Eds.), Proceedings 4th European Congress on Biotechnology, Vol. 2. Elsevier, Amsterdam, pp. 212±215. Arfmann, H.-A., Abraham, W.-R., Kieslich, K., 1988. Microbial !hydroxylation of trans-nerolidol and structurally related sesquiterpenoids. Biocatalysis 2, 59±67. Bauer, K., Garbe, D., Surburg, H. (Eds.), 1990. Common Fragrance and Flavor Materials. Preparation, Properties and Uses. 2nd Edition. VCH Publishers, New York. Bowen, E.R., 1975. Potential by-products from microbial transformation of D-limonene. In: Proceedings of the Florida State Horticultural Society, Miami Beach, pp. 304-308. Cadwallader, K.R., Braddock, R.J., 1992. Enzymatic hydration of (4R) -(+)-limonene to (4R)-(+)-a-terpineol. Dev. Food Sci. 29, 571-584. Cadwallader, K.R., Braddock, R.J., Parish, M.E., Higgins, D.P., 1989. Bioconversion of (+)-limonene by Pseudomonas gladioli. Journal of Food Science 54, 1241±1245. Demyttenaere, J., De Kimpe, N., 2000. Biotransformation of geraniol and nerol by sporulated surface cultures of Aspergillus niger. In: Schieberle, P., Engel, K.-H. (Eds.), Frontiers of Flavour Science, Proceedings of the Ninth Weurman Flavour Research Symposium 22±25 June 1999, Freising, Germany, pp. 400-404. Demyttenaere, J.C.R., De Pooter, H.L., 1998. Biotransformation of citral and nerol by spores of Penicillium digitatum. Flavour Fragr. J. 13, 173±176. Demyttenaere, J.C.R., Herrera, M., del, C., De Kimpe, N., 2000. Biotransformation of geraniol, nerol and citral by sporulated surface cultures of Aspergillus niger and Penicillium sp. Phytochemistry 55, 363±373.
Demyttenaere, J.C.R., Willemen, H.M., 1998. Biotransformation of linalool to furanoid and pyranoid linalool oxides by Aspergillus niger. Phytochemistry 47, 1029±1036. Dhavalikar, R.S., Bhattacharyya, P.K., 1966. Microbiological transformations of terpenes: part VIII Ð fermentation of limonene by a soil pseudomonad. Indian Journal of Biochemistry 3, 144± 157. Fenaroli, G., 1975. In: Furia, T.E., Bellanca, N. (Eds.), Handbook of Flavor Ingredients, 2nd Ed. CRC Press, Cleveland, Ohio. Kraidman, G., Mukherjee, B.B., Hill, I.D., 1969. Conversion of Dlimonene into an optically active isomer of a-terpineol by a Cladosporium species. Bacteriological Proceedings, 63. Krasnobajew, V., 1984. Terpenoids. In: Rehm, H.-J., Reed, G. (Eds.), Biotechnology, Vol. 6a. Verlag Chemie, Weinheim, pp. 97-125. Madyastha, K.M., Bhattacharyya, P.K., Vaidyanathan, C.S., 1977. Metabolism of a monoterpene alcohol, linalool, by a soil pseudomonad. Canadian Journal of Microbiology 23, 230±239. Mattison, J.E., McDowell, L.L., Baum, R.H., 1971. Cometabolism of selected monoterpenoids by fungi associated with monoterpenoidcontaining plants. Bacteriological Proceedings, 141. Mukherjee, B.B., Kraidman, G., Hill, I.D., 1973. Synthesis of glycols by microbial transformation of some monocyclic terpenes. Applied Microbiology 25, 447±453. Noma, Y., Yamasaki, S., Asakawa, Y., 1992. Biotransformation of limonene and related compounds by Aspergillus cellulosae. Phytochemistry 31, 2725±2727. Rama Devi, J., Bhattacharyya, P.K., 1977. Microbiological transformations of terpenes: part xxiii Ð fermentation of geraniol, nerol and limonene by a soil pseudomonad, Pseudomonas incognita (linalool strain). Indian Journal of Biochemistry and Biophysics 14, 288±291. Rama Devi, J., Bhattacharyya, P.K., 1978. Molecular rearrangements in the microbiological transformations of terpenes and the chemical logic of microbial processes. Journal of the Indian Chemistry Society 55, 1131±1137. Tan, Q., Day, D.F., 1998a. Organic co-solvent eects on the bioconversion of (R)-(+)-limonene to (R)-(+)-a-terpineol. Process Biochemistry 33, 755±761. Tan, Q., Day, D.F., 1998b. Bioconversion of limonene to a-terpineol by immobilized Penicillium digitatum. Applied Microbiology and Biotechnology 49, 96±101.