Biosci. Biotechnol. Biochem., 68 (12), 2571–2580, 2004

Biosynthesis of Abscisic Acid by the Direct Pathway via Ionylideneethane in a Fungus, Cercospora cruenta

y Masahiro INOMATA,1 Nobuhiro HIRAI,2; Ryuji YOSHIDA,3 and Hajime OHIGASHI1

1Division of Food Science and Biotechnology, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan 2International Innovation Center, Kyoto University, Kyoto 606-8501, Japan 3Department of Agriculture Technology, Toyama Prefectural University, Toyama 939-0311, Japan

Received August 11, 2004; Accepted September 12, 2004

We examined the biosynthetic pathway of abscisic Key words: Cercospora cruenta; abscisic acid; allofar- acid (ABA) after isopentenyl diphosphate in a fungus, nesene; -ionylideneethane; all-E-7,8-dihy- Cercospora cruenta. All oxygen atoms at C-1, -1, -10, and dro--carotene -40 of ABA produced by this fungus were labeled with 18 18 O from O2. The fungus did not produce the 9Z- A sesquiterpenoid, abscisic acid (ABA, 1), is a plant carotenoid possessing -ring that is likely a precursor hormone which regulates seed dormancy and induces for the carotenoid pathway, but produced new sesqui- dehydration tolerance by reducing the stomatal aper- terpenoids, 2E,4E- -ionylideneethane and 2Z,4E- -ion- ture.1) ABA is biosynthesized by some phytopathogenic ylideneethane, along with 2E,4E,6E-allofarnesene. The fungi in addition to plants,2) but the biosynthetic origin fungus converted these sesquiterpenoids labeled with of isopentenyl diphosphate (IDP) for fungal ABA is 13C to ABA, and the incorporation ratio of 2Z,4E- - different from that for plant ABA (Fig. 1). Fungi use ionylideneethane was higher than that of 2E,4E- - IDP derived from the mevalonate pathway for ABA, ionylideneethane. From these results, we concluded that while higher plants biosynthesize IDP for ABA by the C. cruenta biosynthesized ABA by the direct pathway non-mevalonate pathway.3) For the biosynthetic path- via oxidation of ionylideneethane with molecular oxygen way of ABA after IDP, the carotenoid pathway involv- following cyclization of allofarnesene. This direct path- ing cleavage of 9Z-xanthophyll has been elucidated for way via ionylideneethane in the fungus is consistent with plant ABA.4) In contrast to ABA of higher plants, two that in Botrytis cinerea, except for the positions of pathways, the direct and carotenoid pathways after IDP, double bonds in the rings of biosynthetic intermediates, have been proposed for fungal ABA. The direct pathway suggesting that the pathway is common among ABA- is supposed to contain oxidation of ionylideneethanol by producing fungi. oxidase after cyclization of farnesyl diphosphate

Fig. 1. Outline of ABA Biosynthetic Pathway.

y To whom correspondence should be addressed. Fax: +81-75-753-6284; E-mail: [email protected] Abbreviations: ABA, abscisic acid; FDP, farnesyl diphosphate; IDP, isopentenyl diphosphate 2572 M. INOMATA et al. Table 1. Relative IntensitiesaÞ and CompositionsbÞ of Molecular Ions of 4 and Methyl Esters of 5 and ABA Isolated from C. cruenta Cultured 18 under O2

Compound 4 Methyl ester of 5 Methyl ester of ABA rel. int. comp. rel. int. comp. rel. int. comp. m=z ion species m=z ion species m=z ion species (%) (%) (%) (%) (%) (%) 222 ½M þ 2þ 3.0 58 288 ½M þ 8þ 1.4 6 286 ½M þ 8þ 0.5 9 220 ½Mþ 2.2 42 286 ½M þ 6þ 4.3 19 284 ½M þ 6þ 0.9 17 284 ½M þ 4þ 7.7 34 282 ½M þ 4þ 1.7 32 282 ½M þ 2þ 6.8 30 280 ½M þ 2þ 1.4 26 280 ½Mþ 2.7 12 278 ½Mþ 0.8 15

a)Relative intensities were corrected by natural 13C-, 2H- and 17O-isotopic abundance. bÞComposition of relative intensities of each ion group.

(FDP),5–7) the classical direct pathway in Fig. 1, and the Results and Discussion carotenoid pathway might be similar to that of plant 8) 18 ABA. Sufficient evidence appears to be lacking for Labeling of 4, 5, and ABA with O2 these two pathways, since the intermediates between 18O-Labeling of the intermediates 4 and 5 in addition FDP and ionylideneethanol have not been found for the to ABA was examined. The period of accumulation and direct pathway,9–11) and the immediate precursor car- the location were different between compound 4 on the otenoids have not been found for the carotenoid path- one hand and compound 5 and ABA on the other. way. We found that a phytopathogenic fungus, Botrytis Compound 4 was isolated from the mycelia of C. cruen- 18 cinerea, biosynthesized ABA by the direct pathway via ta cultured under O2 for 7 d, and compound 5 and ionylideneethane.12) In this pathway, FDP is converted ABA were isolated from the medium of the fungus 18 to 2E,4E,6E-allofarnesene (2), 2Z,4E,6E-allofarnesene cultured under O2 for 16 d. Compound 6 was not (3), and (R)-2Z,4E--ionylideneethane successively, and detected in the mycelia of the fungus during the culture then ionylideneethane is oxidized to ABA with molecu- for 16 d. lar oxygen. Recently, the bcaba1 gene encoding P450 Compound 4 and methyl esters of 5 and ABA were monooxygenase of B. cinerea was identified as the ABA analyzed by EIMS to determine their labeled positions 13) 18 18 biosynthetic gene. This finding is consistent with with O from O2. The relative intensities and oxidation of ionylideneethane with molecular oxygen in compositions of molecular ions of 4 and methyl esters B. cinerea. It is not known whether the direct pathway of 5 and ABA are summarized in Table 1. Compound 4 via ionylideneethane is common among ABA-producing gave molecular ions corresponding to ½M þ 2þ and fungi. This situation led us to investigate the biosyn- ½Mþ at m=z 222 and 220 respectively, indicating that thesis of ABA in another ABA-producing fungus, the oxygen atom at C-1 of 4 was labeled with 18O from 18 Cercospora cruenta. O2. Methyl ester of 5 showed molecular ions corre- C. cruenta is a typical fungus in which the two sponding to ½M þ 8þ, ½M þ 6þ, ½M þ 4þ, ½M þ 2þ, pathways have been proposed for the biosynthesis of and ½Mþ at m=z 288, 286, 284, 282, and 280 ABA. This fungus produces 2Z,4E--ionylideneethanol respectively. This showed that all four oxygen atoms 0 0 18 18 (4) and 2Z,4E-1 ,4 -dihydroxy--ionylideneacetic acid of 5 at maximum were labeled with O from O2. ABA (5) as the biosynthetic intermediates of ABA.7,14) The methyl ester gave ½M þ 8þ, ½M þ 6þ, ½M þ 4þ, production of these compounds appears to support the ½M þ 2þ, and ½Mþ ions at m=z 286, 284, 282, 280, direct pathway, but the fungus produces -carotene and and 278 respectively. This showed that all four oxygen an aldehyde-intermediate, 2Z,4E--ionylideneacetalde- atoms of ABA at maximum were labeled with 18O from 18 18 0 18 18 18 hyde (6), and incorporates O from O2 into C-1, -1 , O2 also. In B. cinerea cultured under O2, O-label and -40 of ABA.8) These facts suggest that ABA of the at C-40 of ABA was lost by exchange with 16O from 16 fungus is biosynthesized by the carotenoid pathway, H2 O due to the acidity of the medium. The remaining although the 9Z-carotenoid possessing -ring that 18O-label at C-40 in C. cruenta is probably due to the pH should be a precursor of 6 has not been found in this of the medium, which was kept between 5.8 and 7.5 fungus. We re-examined the 18O-labeling of ABA with during the culture. Exchange of 40-18O with 16Oof 18 16 12) 18 O2, made precise analysis of ABA-related carotenoids H2 O was less than 7% in this pH range. The O and sesquiterpenoids, and did feeding experiments of labels incorporated into compounds 4, 5, and ABA 18 labeled compounds in C. cruenta to solve the confusion might be derived from H2 O that the fungus formed 18 over the biosynthesis of ABA. This paper reports the from O2 by respiration during culture. This possibility occurrence of the direct pathway via ionylideneethane in appears to be small, since no oxygen atom from water C. cruenta as well as in B. cinerea. was incorporated into ABA in the labeling experiment 18 12) with H2 OinB. cinerea. Thus all oxygen of 4, 5, and Biosynthetic Pathway of Abscisic Acid in Cercospora cruenta 2573 ABA produced by C. cruenta is derived from molecular oxygen, but not from water. The result of 18O-labeling of ABA was different from that reported by Yamamoto et al.8) This difference can be explained by the metabolic rate of 6 to ABA. Accumulation of compound 6 in their cultural condition suggests slow turnover of 6 to ABA. The 18O-label at C- 16 16 1of6 would exchange easily with O from H2 Oina medium via a hydrate of the aldehyde group at C-1. Slow turnover of 6 to ABA should allow this exchange, resulting in a decrease or loss of 1-18Oof6. The temperature (30 C) for the culture used by Yamamoto et al. was higher than the optimum temperature for fungal growth and might affect the biosynthetic rate of ABA. We also observed that all four oxygen atoms of 18 18 ABA were not always labeled with O from O2. When the fungus began to produce ABA at a late period, day 22 after inoculation, two oxygen atoms at C-1 and 0 18 18 -1 were labeled with O from O2, but two oxygen atoms at C-1 and -40 were not (data not shown). Our finding is consistent with the direct pathway via ionylideneethane. But incorporation of two 18O atoms 18 into C-1 of ABA from O2 might be possible for the carotenoid pathway in which oxidation at C-1 of a Fig. 2. Carotenoids Produced by C. cruenta. putative C15-aldehyde intermediate was catalyzed by monooxygenase with molecular oxygen. If the carote- noid pathway occurs in C. cruenta, the fungus could possessing characteristic absorption spectra for carote- produce the 9Z-carotenoid possessing -ring as well as noids were detected at tRs 9.0, 11.7, 14.3, 14.9, 16.9, and the C15-intermediates 4–6. We carefully searched 21.4 min respectively, along with compound 7. These mycelia extract of the fungus for ABA-related carote- compounds were identified as 15,150Z-phytoene (8), noids and sesquiterpenoids to distinguish between the 15,150Z-phytofluene (9), all-E--zeacarotene (10), all-E- carotenoid pathway and the direct pathway via ionyli- -carotene (11), all-E-7,8-dihydro--carotene (12), and deneethane. all-E-neurosporene (13) (Fig. 2). These carotenoids appeared to have been accumulated by the inhibition Isolation and identification of ABA-related com- of desaturation of acyclic carotene in the biosynthesis of pounds 7 under anaerobic conditions. Carotenoid 12 has been First, the extract from the mycelia of the fungus known as the product of carotenogenesis genes ex- cultured in a flask capped with a cotton plug was pressed in Escherichia coli,15) but was first isolated from analyzed by HPLC to detect carotenoids. One carote- a natural organism. The contents of 7–13 were 3, 4, 0.1, noid-like compound was eluted at tR 13.6 min, and 0.1, 0.4, 0.8, and 0.1 g/g fresh weight of mycelia identified as all-E--carotene (7). The content of 7 was respectively. 9Z-Carotenoids possessing -rings were 17 g/g fresh weight of mycelia. No other carotenoid not found even under anaerobic conditions, suggesting was detected. No detection of precursor carotenoids in the absence of precursor carotenoids for ABA. We tried the mycelia might be due to the rapid conversion of to detect C15 hydrocarbons for the intermediate for ABA them to ABA. The oxidative cleavage of precursor biosynthesis. carotenoids is inhibited under anaerobic conditions. We The HPLC analysis of extract from the mycelia at an cultured C. cruenta in a flask capped with a silicone early growth stage showed two peaks at tRs 10.2 min and plug, which passes air with more difficulty than a cotton 14.1 min the absorption spectra of which were similar to plug. The fungus produced 0.2 mg/l of ABA at those of 2Z,4E--ionylideneethane and 2 respectively. maximum at day 11 after inoculation, while the fungus The GC–MS analysis of the material eluted at tR cultured in a flask capped with a cotton plug produced 10.2 min showed that it consisted of two compounds, 14 4 mg/l of ABA at maximum at day 13, suggesting that and 15, with a ratio of 86:14, having molecular ions at precursor carotenoids had accumulated in the mycelia of m=z 204. 1H- and 13C-NMR spectra of the mixture of 14 the fungus cultured in the flask capped with a silicone and 15 suggested that compound 14 was 2Z,4E-- plug. The extract from the mycelia of the fungus ionylideneethane. Identification of 14 as 2Z,4E--ion- cultured in the flask capped with a silicone plug was ylideneethane was confirmed by comparison of the analyzed by HPLC with a photodiode array detector at a spectral data with those of the chemically synthesized detection limit of 0.3 ng for 7. Compounds 8–13 2Z,4E--ionylideneethane. Compound 15 was identified 2574 M. INOMATA et al.

Fig. 3. The Direct Pathway via Ionylideneethane for ABA Biosynthesis in C. cruenta. Bold arrows show the major pathway confirmed by the feeding experiments. Compounds in parentheses were not detected in the fungus in this experiment. The oxygen atoms with an asterisk are derived from molecular oxygen. The introduction of molecular oxygen into C-1 of 6 was reported by Yamamoto et al.8) as 2E,4E--ionylideneethane also by comparison of the Table 2. Incorporation Ratios (%) of 2-13C Labeled 2, 3, 14, and 15 tR in GC and mass spectrum with those of the chemi- to the Products, and 13C-Isotopic Abundance (%) of the Products cally synthesized 2E,4E--ionylideneethane. Com- (shown in parentheses) pounds 14 and 15 are new compounds. The material Substrates eluted at tR 14.1 min was identified as 2 by spectral Products analysis.12) The contents of 2, 14, and 15 were 0.5, 15, [2-13C]-2 [2-13C]-3 [2-13C]-15 [2-13C]-14 and 3 g/g fresh weight of mycelia respectively. 2 –aÞ 1.2 (11.6) – – The absence of precursor carotenoids and the occur- 3 0.4 (26.8) – – – 15 0.1 (1.3) 0.1 (1.5) – NDbÞ rence of C15 compounds 2, 14, and 15 strongly suggested that this fungus biosynthesized ABA by the 14 0.5 (1.8) 0.4 (1.7) 0.2 (1.6) – 5 0.1 (1.1) 0.1 (1.1) 0.3 (1.6) 6.9 (3.4) direct pathway via ionylideneethane as well as B. cin- ABA 0.1 (1.2) 0.2 (1.2) 0.2 (1.6) 3.3 (3.2) erea. This pathway contains two possible routes: FDP is aÞNot tested. reduced at C-1 and desaturated at C-4 to give 2, bÞNot detected. isomerization at C-2 of 2 before cyclization of 3 or cyclization of 2 before isomerization at C-2 of 15 to give 14, and oxidation at C-1, -1, -10, and -40 of that with before cyclization of 2 to 15. Failure to detect 3 and 6 molecular oxygen to form ABA (Fig. 3). A feeding might be due to the rapid conversion of 3 to 14 and of 6 experiment with [2-13C]-2, 3, 14, and 15 was performed to 5 respectively. Compounds 2 and 3 gradually to confirm the occurrence of the direct pathway via decompose upon exposure to oxygen, water, and acid, ionylideneethane, and to distinguish the two routes from so this instability in addition to volatility might cause 2 to 14. low incorporation of these compounds. It is unknown at present whether desaturation at C-4 forming 2 precedes Feeding experiment of [2-13C]-2, 3, 14, and 15 reduction at C-1. 2-13C-Labeled 2, 3, 14, and 15 were synthesized and The stereospecificity of oxygenase catalyzing hydrox- fed to C. cruenta. The incorporation ratios of these ylation at C-10 of 14 was evaluated by the enantiomeric compounds into 2, 3, 5, 14, 15, and ABA are summa- excess of ABA that was formed by the fungus fed with rized in Table 2. The conversion of [2-13C]-2, 3, 14, and ()-[2-13C]-14. The incorporation ratio of [2-13C]-14 15 to 5 and ABA indicated that these compounds were into the ABA was 3.3%. If both enantiomers of [2-13C]- biosynthetic precursors for ABA. The incorporation 14 are converted to ABA, the enantiomeric excess of the ratio of 14 to ABA was clearly higher than that of 15, ABA should be 97.87%, and if (S)-[2-13C]-14 is while that of 2 was almost the same as that of 3. This selectively converted to ABA, the enantiomeric excess result suggests that isomerization of 2 to 3 proceeds of the ABA should be 95.73%. HPLC analysis of the Biosynthetic Pathway of Abscisic Acid in Cercospora cruenta 2575 ABA with a chiral column showed that it was a natural carotenoids, HPLC was performed with a Waters 600E (S) isomer with 98:91 0:04% ee, indicating that 10- multisolvent delivery system using a photodiode array hydroxylase of 4 recognizes the chirality at C-10, detector Waters 991J with a detection range of 200– although the selectivity was not high. 600 nm. The HPLC columns used were a YMC-Pack These findings strongly suggest that C. cruenta bio- ODS-AQ 311 column (100 mm length 6.0 mm i.d., synthesizes ABA by the direct pathway via ionylidene- YMC Co., Ltd., Kyoto, Japan), a YMC Carotenoid ethane. In this pathway, dehydrogenation at C-4, 5 and column (250 mm length 4.6 mm i.d., YMC Inc., reduction at C-1 of FDP gives 2, cyclization of 3 after Wilmington, NC, U.S.A.), and a Chiralcel OD column isomerization at C-2 of 2 forms (R)-14, and then C-1, -1, (250 mm length 4.6 mm i.d., Daicel Chemical Indus- -10, and -40 of that is oxidized by monooxygenase to give tries Ltd., Niigata, Japan). Column chromatography (S)-ABA via 4, 6, and 5 (Fig. 3). This direct pathway in was carried out on Wakogel C-200 (silica gel; particle C. cruenta corresponds to that in B. cinerea except for size, 0.075–0.15 mm, Wako Pure Chemical Industries the positions of double bonds in the rings of their Ltd., Osaka, Japan) or aluminium oxide 90 (alumina; intermediates. The direct pathway via ionylideneethane particle size, 0.063–0.20 mm, Merck KGaA, Darmstadt, appears to be common among ABA-producing fungi. Germany). The gene encoding P450 monooxygenase corresponding to the bcaba1 of B. cinerea might be present in Materials. The phytopathogenic fungi Cercospora C. cruenta. The biosynthetic pathway of ABA not only cruenta, IFO6134, which was used by Yamamoto et before but also after IDP in fungi is different from that in al.,8) was kindly provided by Professor Emeritus higher plants. Fungi probably acquired the genes for Takayuki Oritani of Tohoku University, Japan. ()- ABA biosynthesis independently of higher plants. ABA and -carotene were purchased from Wako Pure 18 18 Chemical Industries Ltd., and O2 (99 atom % O) was Experimental purchased from ISOTEC, Inc., Miamisburg, OH, U.S.A. [1-13C]Bromoethane (99 atom % 13C) was purchased General experimental procedures. 1H- and 13C-NMR, from Aldrich Chemical, Co., Inc., Milwaukee, WI, NOESY, HMQC, and HMBC spectra were measured U.S.A. with a Bruker ARX500 instrument (500 MHz for 1H and 13 18 125 MHz for C). Chemical shifts were given with Labeling of 4, 5, and ABA with O2. Ten milliliters of TMS ( 0.00 ppm) as the internal standard for 1H- and a modified potato-dextrose medium containing 0.2 g 13 C-NMR analysis with CDCl3, and with a signal of (1.1 mmol) of D-glucose, 0.04 g of yeast extract and 1 benzene-d6 ( 7.27 ppm) as the standard for H-NMR 0.015 g of agar was added to a 300 ml-sized Erlenmeyer analysis with benzene-d6. Direct EIMS was carried out flask. For the labeling experiment of compound 5 and with a JEOL JMS-600H mass spectrometer, the temper- ABA, a flask with a side arm sealed with a silicone ature of the direct probe being increased from 50 Cto rubber stopper was used to collect the medium using a 450 C at a rate of 128 C/min. The instrument was syringe. After sterilization by autoclave and inoculation operated at a chamber temperature of 250 C, accelerat- of C. cruenta, the flask was sealed with a stopcock. The ing voltage of 3 kV, ionization voltage of 70 eV, and flask was immediately evacuated and then filled with N2 ionization current of 300 A in the positive ion mode, to purge the air. This procedure was repeated two more the resolution being 1,000 during the measurement. GC– times. An air bag (5-liter-size) containing 0.25-liter 18 EI mass spectra were recorded on the above mass (10 mmol, 25 C) of O2 and 1.0-liter of N2 was fitted to spectrometer equipped with a Hewlett Packard HP6890 the flask, and C. cruenta was cultured on a rotary shaker instrument, using an HP-5 column (30 m length (110 rpm) at 25 C under a fluorescent light. The 0.32 mm i.d., 5% diphenyl–95% dimethylpolysiloxane, atmosphere in the air bag was not replenished with 18 film thickness 0.25 m, Hewlett Packard Co., Wilming- O2 during the culture. The culture periods were 7 d for ton, DE, U.S.A), and a He flow of 1.0 ml/min. The 4, and 16 d for 5 and ABA. The medium was collected parameters of the mass spectrometer were the same as every two days from day 10 after inoculation, and those of direct EIMS. The oven temperature for GC was analyzed by HPLC with a YMC-Pack ODS-AQ 311 programmed from 120 C to 270 C at a rate of 5 C/min column (solvent, 0.1% AcOH–45% MeOH–55% H2O; for methyl esters of ABA and 5, and from 100 Cto flow rate, 1.0 ml/min; detection, 254 nm; tR of ABA, 190 C at a rate of 3 C/min for 2, 3, 4, 14, and 15. The 12.9 min). 18O contents were calculated after correction by natural 13C-, 2H-, and 17O-isotopic abundance. IR spectra were Isolation of 4. After culture of C. cruenta for 7 d in an 18 obtained with a Shimadzu FTIR-8100AI spectrometer. O2 atmosphere, the mycelia were recovered and UV and VIS spectra, and optical rotation were measured washed with 30 ml of distilled water. The mycelia with a Shimadzu UV 2200AI and a JASCO DIP-1000 (1.6 g) were homogenized in 10 ml of MeOH with sea polarimeter respectively. HPLC was performed with a sand and extracted with 60 ml of EtOAc–MeOH (1:1) Hitachi L-7100 pump, Hitachi L-7400 UV detector, and solution. The extract was filtered and the filtrate was Hitachi D-7500 chromato-integrator. For analysis of concentrated to give 10 ml of a yellow aqueous solution. 2576 M. INOMATA et al. After the addition of 40 ml of distilled water to the cultured on a reciprocal shaker (160 rpm) with six 2- solution, this aqueous solution was partitioned three liter-sized flasks capped with cotton plugs, each con- times with 20 ml of EtOAc. The organic layers were taining 700 ml of a modified potato-dextrose medium at combined, washed with distilled water, dehydrated with 25 C under a fluorescent light for 8 d. After filtration, Na2SO4, filtered, and concentrated to give a yellow oil the mycelia (400 g) were washed with distilled water, (18 mg). The oil was subjected to silica gel (5 g) column homogenized in 100 ml of MeOH with sea sand, and chromatography with mixtures of n-hexane and EtOAc extracted three times using 2-liter of EtOAc–MeOH as the eluent. The material (5 mg) eluted with 15–20% (1:1) solution. These extracts were combined and EtOAc was subjected to preparative HPLC with a YMC- filtered, and the filtrate was concentrated to give Pack ODS-AQ 311 column (solvent, 75% MeOH–25% 290 ml of orange aqueous solution. After the addition H2O; flow rate, 1.0 ml/min; detection, 254 nm). The of 1.7-liter of MeOH to the solution, this 85% MeOH material eluted at tR 19.3 min was collected and solution was partitioned five times with 500 ml of n- concentrated to give 4 (0.5 g). GC–EIMS (tR hexane. The n-hexane layers were combined, washed þ þ 15.0 min) m=z (rel. int.): 222 ½M þ 2 (3), 220 ½M with 100 ml of 1% NaHCO3 and distilled water þ þ (2), 207 ½M þ 2 CH3 (6), 205 ½M CH3 (3), 202 successively, dried over Na2SO4, filtered, and concen- (13), 189 (53), 187 (52), 159 (24), 133 (51), 121 (43), trated to give an orange oil (9 g). The oil was analyzed 105 (100), 91 (71), 81 (75), 69 (54), 55 (39). The total by HPLC with a photodiode array detector (200– 18O content was calculated from the relative intensities 600 nm): column, YMC Carotenoid column; solvent, of ½M þ 2þ and ½Mþ ions to be 58%. 15% to 85% tert-butyl methyl ether in MeOH over a period of 30 min; flow rate, 1.0 ml/min; tR of compound Isolation of 5 and ABA. After the culture of 7, 13.6 min. The content of 7 was calculated from the 18 C. cruenta for 16 d under O2 was filtered, the mycelia calibration curve between the peak area in HPLC and were washed with 40 ml of distilled water. The filtrate the amount of authentic 7. The orange oil was subjected and the washing (50 ml, in total) were combined, to silica gel (200 g) column chromatography using acidified with 25% H3PO4 to pH 3, and partitioned mixtures of n-hexane and toluene as the eluent. The three times with 20 ml of EtOAc. The organic layers materials (10 mg) eluted with 10–20% toluene were were combined, dehydrated with Na2SO4, filtered, and subjected to alumina (20 g, 6% water) column chroma- concentrated to give a yellow oil (5 mg). The oil was tography with mixtures of n-hexane and toluene as the subjected to silica gel (2 g) column chromatography eluent. The materials (6 mg) eluted with 1–10% toluene using mixtures of toluene and EtOAc containing 1% were subjected to preparative HPLC with a YMC AcOH as the eluent. The material (0.2 mg) eluted with Carotenoid column (solvent, 70% MeOH–30% tert- 30–40% EtOAc was subjected to preparative HPLC butyl methyl ether; flow rate 1.0 ml/min; detection, using a YMC-Pack ODS-AQ 311 column (solvent, 40% 450 nm). The material eluted at tR 17.1 min was to 100% MeOH in H2O containing 0.1% AcOH over a collected and concentrated to give 7 (4 mg, red solid). period of 30 min; flow rate, 1.0 ml/min; detection, For spectral data of 7, see Britton et al.16) and Englert.17) 254 nm). Materials eluted at tRs 7.9 min and 10.3 min were collected and concentrated to give 5 (19 g) and Quantification of ABA produced under the two culture ABA (10 g) respectively. To the ABA dissolved in conditions. C. cruenta was cultured on a reciprocal MeOH ethereal CH2N2 was added, and the solution was shaker (160 rpm) with 700 ml of a modified potato- left at room temperature for 1 h. The solution was dextrose medium in a 2-liter-sized flask capped with a concentrated to give a methyl ester of ABA (10 g). cotton plug at 25 C under a fluorescent light. After 4 d, Methyl ester of ABA. GC–EIMS (tR 14.7 min) m=z (rel. 20 ml of aliquot containing both mycelium and medium int.): 286 ½M þ 8þ (0.5), 284 ½M þ 6þ (0.9), 282 ½M þ was collected from the culture, homogenized in 10 ml of 4þ (2), 280 ½M þ 2þ (1), 278 ½Mþ (0.8), 266 (1), 264 MeOH with sea sand, and extracted with 50 ml of (4), 262 (5), 260 (3), 196 (23), 194 (76), 192 (100), 190 EtOAc–MeOH (1:1) solution. The extract was filtered, (53), 164 (35), 162 (26), 129 (15), 127 (35), 125 (26). and the filtrate was concentrated. Twenty-five milliliters 18 The total O content was calculated from the relative of 1% NaHCO3 was added to the residue, and the intensities of the molecular ions to be 44%. Compound 5 solution was partitioned three times with 30 ml of was methylated by the same method as that described EtOAc. The aqueous layer was acidified to pH 3 with for ABA to give a methyl ester of 5 (19 g). Methyl 25% H3PO4 and partitioned three times using 30 ml of ester of 5. GC–EIMS (tR 14.9 min) m=z (rel. int.): 288 EtOAc. The organic layers were combined, dehydrated þ þ þ ½M þ 8 (1), 286 ½M þ 6 (4), 284 ½M þ 4 (8), 282 with Na2SO4, filtered, and concentrated to give a yellow ½M þ 2þ (7), 280 ½Mþ (3), 268 (4), 266 (12), 264 (18), oil (33 mg). ABA in this oil was quantified by HPLC 262 (8), 129 (70), 127 (100), 125 (81). The total 18O using a YMC-Pack ODS-AQ 311 column (solvent, 0.1% content was calculated from the relative intensities of AcOH–55% H2O–45% MeOH; flow rate, 1.0 ml/min; the molecular ions to be 45%. detection 254 nm; tR of ABA, 12.6 min). The content of ABA was calculated from the calibration curve between Detection and identification of 7. C. cruenta was the peak area in HPLC and the amount of authentic Biosynthetic Pathway of Abscisic Acid in Cercospora cruenta 2577 ABA. ABA from 5 to 14 d was quantified every day. tert-butyl methyl ether in MeOH over a period of ABA produced by the fungus cultured in a flask capped 30 min; flow rate, 1.0 ml/min; detection 350 or 400 nm). with a silicone plug was quantified every day from 4 to The materials eluted at tRs 11.7 min and 14.3 min were 14 d by the methods described above. collected and concentrated to give 9 (12 g) and 10 (4 g) respectively. The material (0.1 mg) eluted with Detection and identification of 7–13. C. cruenta was 5% toluene was purified with HPLC under the same cultured on a reciprocal shaker (160 rpm) with six 2- condition as that described for the material eluted with liter-sized flasks capped with silicone plugs, each 0% toluene, except that the detection was 400 or containing 700 ml of a modified potato-dextrose medium 450 nm. The materials eluted at tRs 14.9 min and at 25 C under a fluorescent light. After 8 d, the mycelia 21.4 min were collected and concentrated to give 11 (270 g) were washed with distilled water, homogenized (50 g) and 13 (14 g) respectively. in 100 ml of MeOH with sea sand, and extracted three Compound 9. NMR H (benzene-d6): 1.68 (9H, br.s, times using 2-liter of EtOAc–MeOH (1:1) solution. The H-16, 18, 160), 1.72 (6H, s, H-19, 180), 1.80 (6H, s, H-17, combined extracts were filtered, and the filtrate was 170), 1.81 (3H, s, H-20), 1.84 (3H, s,H-190), 1.99 (3H, s, concentrated to give 240 ml of a yellow aqueous H-200), 2.29 (20H, m, H-3, 4, 7, 8, 11, 12, 30,40,70,80), solution. After the addition of 1.6-liter of MeOH to 5.38 (5H, m, H-2, 6, 10, 20,60), 6.25 (2H, m,H-100,150), the residue, this 85% MeOH solution was partitioned 6.42 (1H, m, H-15), 6.54 (1H, d, J ¼ 15:0 Hz, H-120), 0 0 five times with 300 ml of n-hexane. The n-hexane layers 6.72 (3H, m, H-14, 11 ,14). Compound 10. NMR H were combined, washed with 100 ml of 1% NaHCO3 (benzene-d6): 1.27 (6H, s, H-16, 17), 1.61 (2H, m, H-2), 0 0 and distilled water successively, dried over Na2SO4, 1.68 (3H, s, H-16 ), 1.72 (3H, s,H-18), 1.80 (3H, s,H- filtered, and concentrated to give an orange oil (1.5 g). 170), 1.85 (3H, s, H-190), 1.93 (3H, s, H-18), 1.97 (3H, s, The oil was analyzed by HPLC with a photodiode array H-19), 2.00 (3H, s, H-200), 2.05 (3H, s, H-20), 2.09 (2H, detector (200–600 nm): column, YMC Carotenoid col- m, H-4), 2.27 (8H, m, H-30,40,70,80), 5.35 (1H, m, H-20), umn; solvent, 15% to 85% tert-butyl methyl ether in 5.40 (1H, m H-60), 6.26 (1H, m, H-100), 6.46 (4H, m,H- MeOH over a period of 30 min; flow rate, 1.0 ml/min. 8, 10, 14, 140), 6.51 (1H, d, J ¼ 16:1 Hz, H-7), 6.54 (1H, 0 For tRs of compounds 7–13, see ‘‘Results and Discus- d, J ¼ 17:0 Hz, H-12 ), 6.60 (1H, d, J ¼ 14:4 Hz, H-12), sion’’ above. The contents of 7, 8, and 12 were 6.79 (3H, m, H-15, 110,150), 6.89 (1H, m, H-11). The calculated from the calibration curves between the peak signal at H-30 could not be detected by superimposition areas in HPLC and the amounts of these compounds with impurity peaks. Compound 11.NMRH (benzene- 0 0 isolated. For calculation of the contents of 13, 9, and 10 d6): 1.68 (6H, s, H-16, 16 ), 1.72 (6H, s, H-18, 18 ), 1.80 and 11, the calibration curves of 7, 8, and 12 (6H, s, H-17, 170), 1.85 (6H, s, H-19, 190), 1.99 (6H, s, respectively were applied. The orange oil was subjected H-20, 200), 2.28 (16H, m, H-3, 4, 7, 8, 30,40,70,80), 5.35 to silica gel (50 g) column chromatography with mix- (2H, m, H-2, 20), 5.40 (2H, m, H-6, 60), 6.26 (2H, d, tures of n-hexane and toluene as the eluent to give the J ¼ 11:0 Hz, H-10, 100), 6.43 (2H, br.d, J ¼ 9:8 Hz, H- materials eluted with 10% toluene (8 mg) and with 15% 14, 140), 6.54 (2H, d, J ¼ 15:0 Hz, H-12, 120), 6.77 (2H, toluene (0.6 mg). m, H-11, 110), 6.78 (2H, m, H-15, 150). Compound 12. The materials eluted with 10% toluene were applied UV and VIS max (n-hexane) nm ("): 405 (73,600), 427 on alumina (15 g, 6% water) column chromatography (111,000), 453 (101,000). NMR H (CDCl3): 1.01 (6H, using n-hexane as the eluent, and two fractions of 30 ml s, H-16, 17), 1.03 (6H, s, H-160,170), 1.42 (2H, m, H-2), each were collected. The materials (2 mg) in the first 1.47 (2H, m,H-20), 1.59 (4H, m, H-3, 30), 1.62 (3H, s,H- fraction were purified with preparative HPLC with a 18), 1.72 (3H, s,H-180), 1.86 (3H, s, H-19), 1.92 (2H, t, YMC carotenoid column (solvent, 70% MeOH–30% J ¼ 6:4 Hz, H-4), 1.95 (3H, s, H-20), 1.97 (6H, s, H-190, tert-butyl methyl ether; flow rate, 1.0 ml/min, detection 200), 2.02 (2H, t, J ¼ 6:0 Hz, H-40), 2.12 (4H, m, H-7, 8), 254 or 450 nm). The material eluted at tR 9.9 min was 5.98 (1H, d, J ¼ 10:9 Hz, H-10), 6.13 (1H, d, J ¼ collected and concentrated to give 8 (0.9 mg, colorless 14:6 Hz, H-80), 6.15 (1H, m, H-100), 6.17 (1H, m, H-70), oil). The materials (1 mg) in the second fraction were 6.20 (1H, d, J ¼ 10:7 Hz, H-14), 6.24 (1H, d, J ¼ subjected to HPLC under the same condition as that 7:4 Hz, H-140), 6.26 (1H, d, J ¼ 15:0 Hz, H-12), 6.35 described for the first fraction. The materials eluted at (1H, d, J ¼ 14:9 Hz, H-120), 6.50 (1H, dd, J ¼ 10:9, 0 tRs 17.7 min and 29.3 min were collected and concen- 15.0 Hz, H-11), 6.61 (2H, m, H-15, 15 ), 6.65 (1H, m,H- trated to give 7 (0.4 mg, red solid) and 12 (0.2 mg, 110). EIMS m=z (rel. int.): 538 ½Mþ (100), 446 [M-92]þ yellow solid) respectively. (3), 401 [M-137]þ (29), 309 (18), 256 (10), 211 (11), The materials eluted with 15% toluene were subjected 157 (13), 145 (13), 137 (27), 133 (21), 119 (25), 95 (30), þ to alumina (15 g, 6% water) column chromatography 69 (34). HR-EIMS m=z ½M : Calcd. for C40H58: with mixtures of n-hexane and toluene as the eluent, and 538.4538. Found 538.4535. This is the first report of the materials eluted with 0% toluene and with 5% the UV, EI mass, and 1H-NMR spectral data of 12. toluene were collected. The material (0.1 mg) eluted Compound 13. NMR H (benzene-d6): 1.68 (6H, s,H- with 0% toluene was subjected to preparative HPLC 16, 160), 1.76 (3H, s, H-18), 1.78 (3H, s, H-170), 1.80 with a YMC carotenoid column (solvent, 15% to 85% (3H, s, H-17), 1.85 (6H, s, H-19, 180), 1.97 (3H, s,H- 2578 M. INOMATA et al. 190), 2.00 (3H, s, H-20), 2.03 (3H, s, H-200), 2.29 (12H, 7:0 Hz, H-1,), 1.83 (3H, br.s, H-6,), 2.05 (1H, m, H-50), m, H-3, 4, 7, 8, 30,40), 5.35 (2H, m, H-2, 20), 5.41 (1H, 2.29 (1H, m,H-50), 2.51 (1H, d, J ¼ 9:6 Hz, H-10), 4.58 m, H-6), 6.28 (2H, m, H-10, 60), 6.44 (1H, m, H-14), 6.46 (1H, s,H-90), 4.74 (1H, s, H-90), 5.35 (1H, q, J ¼ 7:0 Hz, (2H, m,H-100,140), 6.55 (2H, d, J ¼ 14:9 Hz, H-12, 80), H-2), 5.78 (1H, dd, J ¼ 9:6, 15.5 Hz, H-5), 6.43 (1H, d, 0 6.60 (1H, d, J ¼ 14:5 Hz, H-12 ), 6.79 (4H, m, H-11, 15, J ¼ 15:5 Hz, H-4). NMR C (CDCl3): 13.0 (C-1), 20.8 70,150), 6.87 (1H, m,H-110). See Ebenezer and (C-6), 23.5 (C-40,70), 29.6 (C-80), 34.7 (C-50), 35.5 (C- Pattenden18) for the spectral data of 8, and see the 20), 39.2 (C-30), 58.3 (C-10), 108.3 (C-90), 123.0 (C-2), literature for the UV and EI mass spectral data of 9,19) 128.7 (C-4), 128.9 (C-5), 132.8 (C-3), 150.6 (C-60). GC– 10, 11, and 13.16) EIMS m=z (rel. int.): 204 ½Mþ (93), 189 (54), 175 (10), 161 (24), 148 (22), 135 (87), 119 (70), 107 (100), 94 Detection and identification of 2, 14, and 15. (76), 79 (40), 69 (57), 55 (35). Compound 15. GC–EIMS C. cruenta was cultured on a reciprocal shaker m=z (rel. int.): 204 ½Mþ (93), 189 (50), 175 (11), 161 (160 rpm) with six 2-liter-sized flasks capped with (24), 147 (21), 135 (84), 119 (69), 107 (100), 94 (73), 79 cotton plugs, each containing 700 ml of a modified (42), 69 (53), 55 (36). For the spectral data of 2, see potato-dextrose medium at 25 C under a fluorescent Inomata et al.12) light for 5 d. After filtration of the culture, the mycelia (310 g) were washed with distilled water, homogenized Synthesis of 14 and 15. At room temperature, 1.0 M in 100 ml of acetone with sea sand, and extracted three THF solution of potassium tert-butoxide (0.4 ml, 0.4 times with 1-liter of acetone. The extracts were mmol) was added dropwise to a stirred suspension of combined and filtered, and the filtrate was concentrated ethyl triphenylphosphonium bromide (148 mg, 0.4 to give 100 ml of an orange aqueous solution. After the mmol) in 5 ml of THF under N2. After 30 min stirring, addition of 500 ml of 1% NaHCO3 to the residue, the -ionone (39 mg, 0.2 mmol) was added to the orange solution was partitioned five times using 200 ml of n- solution, and the mixture was stirred for 30 min at room hexane. The n-hexane layers were combined, washed temperature under N2. Distilled water (0.5 ml) was with distilled water, dried over Na2SO4, filtered, and added to the mixture and the solution was stirred for 1 h concentrated to give an orange oil (3 g). The oil was at room temperature. After the addition of another 50 ml analyzed by HPLC with a photodiode array detector of distilled water to the mixture, the solution was (200–600 nm): column, YMC Carotenoid column; sol- partitioned three times with 20 ml of Et2O. The organic vent, 10% H2O–90% MeOH; flow rate, 1.0 ml/min, and layers were combined, washed with distilled water, dried by GC–MS, which was set at the condition described over Na2SO4, filtered, and concentrated to give a yellow above in General experimental procedures. tRsof2, 14, oil (230 mg). The oil was subjected to AgNO3–silica gel and 15 in GC were 12.6 min, 7.4 min, and 8.1 min (1:20, 10 g) column chromatography using mixtures of respectively. The content of 2 was calculated from the n-hexane and toluene as the eluent. The materials eluted calibration curve between the peak area in HPLC and with 5–10% and 10% toluene were concentrated to give the amount of authentic 2. The total content of compounds 14 (18 mg, 42% yield) and 15 (10 mg, compounds 14 and 15 was calculated from the calibra- 22% yield) respectively as a colorless oil. Compound tion curve between the peak area in HPLC and the 14 (2Z,4E-5-(20,20-dimethyl-60-methylenecyclohexyl)-3- amount of authentic 14. The content of 14 and that of 15 methyl-2,4-pentadiene). UV max (n-hexane) nm ("): 1 were calculated by multiplying the total content and the 237 (24,900). IR: max (KBr) cm : 3080, 3040, 2965, ratio of the peak areas in GC. Chemically synthesized 2 2930, 2865, 1645, 1455, 1435, 1385, 1365, 970, 890. þ and 14, described below, were used as the authentic GC–HR-EIMS m=z ½M : Calcd. for C15H24: 204.1878. samples. The orange oil was subjected to silica gel (50 g) Found: 204.1880. For other spectral data, see 14 isolated column chromatography with mixtures of n-hexane and from C. cruenta. Compound 15 (2E,4E-5-(20,20-dimeth- toluene as the eluent. The material (18 mg) eluted with yl-60 -methylenecyclohexyl)-3-methyl-2,4-pentadiene). 5% toluene was subjected to alumina (15 g, 0% water) UV max (n-hexane) nm ("): 233 (22,700). IR: max column chromatography using mixtures of n-hexane and (KBr) cm1: 3080, 3030, 2930, 2865, 1645, 1460, 1440, toluene as the eluent. The materials (16 mg) eluted with 1385, 1365, 970, 890. NMR H (CDCl3): 0.81 (3H, s,H- 0–5% toluene were subjected to preparative HPLC using 70), 0.89 (3H, s,H-80), 1.32 (1H, m, H-30), 1.49 (1H, m, 0 0 a YMC carotenoid column (solvent, 10% H2O–90% H-3 ), 1.57 (2H, m,H-4), 1.71 (3H, d, J ¼ 6:9 Hz, H-1), MeOH, flow rate, 1.0 ml/min; detection, 254 nm). The 1.77 (3H, s, H-6), 2.03 (1H, m,H-50), 2.28 (1H, m, H-50), 0 0 materials eluted at tRs 10.2 min and 14.1 min were 2.44 (1H, d, J ¼ 9:6 Hz, H-1 ), 4.57 (1H, s, H-9 ), 4.72 separately collected and partitioned three times with (1H, s, H-90), 5.47 (1H, q, J ¼ 6:9 Hz, H-2), 5.65 (1H, 10 ml of n-hexane, and the n-hexane layers were dd, J ¼ 9:6, 15.5 Hz, H-5), 6.04 (1H, d, J ¼ 15:5 Hz, H- 0 combined and concentrated to give a mixture of 14 4). NMR C (CDCl3): 12.3 (C-6), 13.7 (C-1), 23.5 (C-4 , and 15 (3 mg, colorless oil), and 2 (1 mg, colorless oil) 70), 29.6 (C-80), 34.6 (C-50), 35.5 (C-20), 39.2 (C-30), 0 0 respectively. Compound 14. NMR H (CDCl3): 0.81 57.9 (C-1 ), 108.2 (C-9 ), 125.0 (C-2), 125.8 (C-5), (3H, s,H-70), 0.90 (3H, s, H-80), 1.33 (1H, m,H-30), 1.50 134.5 (C-3), 136.5 (C-4), 150.7 (C-60). GC–HR-EIMS 0 0 þ (1H, m, H-3 ), 1.59 (2H, m, H-4 ), 1.71 (3H, d, J ¼ m=z ½M : Calcd. for C15H24: 204.1878. Found Biosynthetic Pathway of Abscisic Acid in Cercospora cruenta 2579 204.1880. For GC–EI mass spectral data, see 15 isolated 2, 3, 14, and 15 was analyzed by GC–MS. from C. cruenta. The culture that was shaken for 9 d was filtered and the black mycelia were washed with 700 ml of distilled Synthesis of [2-13C]-2, 3, 14, and 15. [1-13C]Ethyl water. The filtrate and the washing (2.2-liter, in total) triphenylphosphonium bromide, and [2-13C]-2 and [2- were combined and partitioned three times with 700 ml 13C]-3, were synthesized by the same methods as those of EtOAc at pH 7. The aqueous layer was acidified to 12) described in the literature. In the same manner as that pH 3 with 25% H3PO4 and partitioned three times with described for the synthesis of 14 and 15, -ionone 700 ml of EtOAc. The organic layers were combined, 13 (100 mg, 0.5 mmol) was reacted with [1- C]ethyl dried over Na2SO4, filtered, and concentrated to give a triphenylphosphonium bromide (390 mg, 1.0 mmol) yellow oil (106 mg). The oil was subjected to silica gel and potassium tert-butoxide (1.0 ml of 1.0 M THF (5 g) column chromatography using mixtures of toluene solution, 1.0 mmol) in 5 ml of THF, and the reaction and EtOAc containing 1% AcOH as the eluent. The mixture was purified to give [2-13C]-14 (37 mg, 36% material (32 mg) eluted 30–40% EtOAc was subjected to yield) and [2-13C]-15 (28 mg, 27% yield) respectively. preparative HPLC using a YMC-Pack ODS-AQ 311 13 The C contents of these compounds were calculated column (solvent, 40% to 100% MeOH in H2O contain- from the relative intensities of the molecular ions in their ing 0.1% AcOH over a period of 30 min; flow rate, 13 mass spectra to be 99%. [2- C]-14. GC–EIMS (tR 1.0 ml/min; detection, 254 nm). The materials eluted at þ 7.7 min) m=z (rel. int.): 205 ½M (100), 190 (56), 175 tRs 8.7 min and 11.4 min were collected and concen- (8), 162 (25), 148 (19), 136 (85), 120 (61), 108 (71), 95 trated to give 5 (6 mg) and ABA (4 mg) respectively. 13 (87), 80 (25), 69 (43), 55 (23). [2- C]-15. GC–EIMS (tR These compounds were separately methylated with þ 8.0 min) m=z (rel. int.): 205 ½M (100), 190 (52), 175 ethereal CH2N2 to give methyl esters of 5 (6 mg) and (9), 162 (27), 148 (20), 136 (86), 120 (68), 108 (82), 95 ABA (4 mg). Feeding of [2-13C]-3, 14, and 15 was (99), 80 (36), 69 (60), 55 (36). performed in the same scale and manner as that of [2- 13C]-2, and compounds 2, 3, 14, and 15, and compound Feeding of 2-13C labeled 2, 3, 14, and 15 to 5 and ABA were purified from the mycelia and the C. cruenta. C. cruenta was sub-cultured on a reciprocal media respectively. shaker (160 rpm) with three 2-liter-sized flasks capped The incorporation ratio was calculated using the with cotton plugs, each containing 500 ml of a modified following equation: Incorporation ratio (% of the potato-dextrose medium at 25 C under a fluorescent metabolite biosynthesized from the administered light for 6 d. The mycelia (435 g) were harvested by substrate) = PA/SB 100%; P = mole of the product centrifugation at 6000 g for 10 min and then suspend- isolated, A = % of 13C-isotopic abundance of the ed in sterile water. This suspension was centrifuged product 1.1 (% of natural abundance of 13C), again at 6000 g for 10 min, and the supernatant was S = mole of the substrate administered, and B = 99 removed by decantation. The mycelia were suspended in (% of 13C-isotopic abundance of the substrate). The 13C- 150 ml of Czapeck–Dox medium, and 3.0 mg of [2-13C]- isotopic abundance of 2, 3, 14, and 15 was calculated 2 dissolved in 300 l of EtOH was added to the from their GC–EI mass spectra, and that of 5 and ABA suspension. The mixture of mycelia and substrate was was evaluated by the 13C-NMR spectra of their methyl treated with Taitec VP-30S ultrasonic homogenizer esters.3) (20 kHz) at 10 C for 30 min, and then added in equal amounts to three 2-liter-sized flasks, each containing Optical purity of ABA produced by the fungus fed with 450 ml of Czapeck–Dox medium. This culture was ()-[2-13C]-14. ABA (5 mg) produced by C. cruenta incubated on a reciprocal shaker (160 rpm) at 25 C fed with ()-[2-13C]-14 was analyzed by HPLC with a under a fluorescent light. The mycelia and the media Daicel Chiralcel OD column (solvent, 0.1% AcOH–80% were recovered after shaking for 5 d and 9 d respectively. n-hexane–20% 2-PrOH; flow rate, 1.0 ml/min; detec- The mycelia (360 g) that were shaken for 5 d were tion, 254 nm; tRsof(þ)- and ()-ABA, 6.8 min and washed with distilled water, homogenized in 100 ml of 8.3 min respectively). The optical purity of ABA was MeOH with sea sand, and extracted three times with calculated from peak areas and shown as percent of 600 ml of EtOAc–MeOH (1:1) solution. The extracts enantiomeric excess. Data is the mean value SE of were filtered, and the filtrates were combined and three measurements. concentrated to give 200 ml of an orange aqueous solution. After the addition of 50 ml of distilled water to Acknowledgment the residue, the solution was partitioned three times with 60 ml of n-hexane. The organic layers were combined, The authors thank Dr. Hirotaka Yamamoto of Japan washed with distilled water, dried over Na2SO4, filtered, NUS Co., Ltd., for valuable discussion and advice and concentrated to give an orange oil (0.9 g). The oil throughout the research, Professor Emeritus Takayuki was subjected to silica gel (7 g) column chromatography Oritani of Tohoku University for donating C. cruenta with a mixture of n-hexane and toluene as the eluent. IFO6134, and Professor Takashi Maoka of Kyoto The material (9 mg) eluted with 5% toluene containing Pharmaceutical Science University for his suggestions 2580 M. INOMATA et al. regarding spectral analysis of carotenoids. This work Phytochemistry, 29, 3473–3477 (1990). was supported by Grant-in-Aid no. 14360065 from 10) Yamamoto, H., and Oritani, T., Incorporation of farnesyl the Ministry of Education, Culture, Sports, Science and pyrophosphate derivatives into abscisic acid and its Technology of Japan. biosynthetic intermediates in Cercospora cruenta. Bio- sci. Biotechnol. Biochem., 61, 821–824 (1997). 11) Yamamoto, H., Inomata, M., Uchiyama, T., and Oritani, References T., Identification of 2,3-dihydro--ionylideneethanol in Cercospora cruenta. Biosci. Biotechnol. Biochem., 65, 1) Leung, J., and Giraudat, J., Abscisic acid signal trans- 810–816 (2001). duction. Annu. Rev. Plant Physiol. Plant Mol. Biol., 49, 12) Inomata, M., Hirai, N., Yoshida, R., and Ohigashi, H., 199–222 (1998). The biosynthetic pathway to abscisic acid via ionylide- 2) Hartung, W., and Gimmler, H., VII. A stress physio- neethane in the fungus Botrytis cinerea. Phytochemistry, logical role for abscisic acid (ABA) in lower plants. 65, 2667–2678 (2004). Prog. Bot., 55, 157–173 (1994). 13) Siewers, V., Smedsgaard, J., and Tudzynski, P., The 3) Hirai, N., Yoshida, R., Todoroki, Y., and Ohigashi, H., P450 monooxygenase BcABA1 is essential for abscisic Biosynthesis of abscisic acid by the non-mevalonate acid biosynthesis in Botrytis cinerea. Appl. Environ. pathway in plants, and by the mevalonate pathway in Miocrobiol., 70, 3868–3876 (2004). fungi. Biosci. Biotechnol. Biochem., 64, 1448–1458 14) Oritani, T., and Yamashita, K., Biosynthesis of (þ)- (2000). abscisic acid in Cercospora cruenta. Agric. Biol. Chem., 4) Schwartz, S. H., Qin, X., and Zeevaart, J. A. D., 49, 245–249 (1985). Elucidation of the indirect pathway of abscisic acid 15) Takaichi, S., Sandmann, G., Schnurr, G., Satomi, Y., biosynthesis by mutants, genes, and enzymes. Plant Suzuki, A., and Misawa, N., The carotenoid 7,8-dihydro- Physiol., 131, 1591–1601 (2003). -end group can be cyclized by the lycopene cyclases 5) Bennett, R. D., Norman, S. M., and Maier, V. P., from the bacterium Erwinia uredovora and the higher Biosynthesis of abscisic acid from farnesol derivatives in plant Capsicum annum. Eur. J. Biochem., 241, 291–296 Cercospora rosicola. Phytochemistry, 23, 1913–1915 (1996). (1984). 16) Britton, G., Goodwin, T. W., Harriman, G. E., and 6) Neill, S. J., Horgan, R., Walton, D. C., and Mercer, C. A. Lockley, W. J. S., Carotenoids of the ladybird beetle, M., The metabolism of -ionylidene compounds by Coccinella septempunctata. Insect Biochem., 7, 337–345 Cercospora rosicola. Phytochemistry, 26, 2515–2519 (1977). (1987). 17) Englert, G., NMR spectroscopy. In ‘‘Carotenoids’’ 7) Kato, T., Oritani, T., and Yamashita, K., Metabolism of Vol. 1B, eds. Britton, G., Liaaen-Jensen, S., and Pfander, (2Z,4E)--ionylideneethanol and (2Z,4E)--ionylidene- H., Birkha¨user Verlag, Basel, pp. 147–260 (1995). acetic acid in Cercospora cruenta. Agric. Biol. Chem., 18) Ebenezer, W. J., and Pattenden, G., cis-Stereoisomer of 51, 2695–2699 (1987). -carotene and its congeners in the alga Dunaliella 8) Yamamoto, H., Inomata, M., Tsuchiya, S., Nakamura, bardawil, and their biogenetic interrelationships. J. M., Uchiyama, T., and Oritani, T., Early biosynthetic Chem. Soc., Perkin Trans. I, 1869–1873 (1993). pathway to abscisic acid in Cercospora cruenta. Biosci. 19) Kushwaha, S. C., Pugh, E. L., Kramer, J. K. G., and Biotechnol. Biochem., 64, 2075–2082 (2000). Kates, M., Isolation and identification of dehydrosquar- 9) Bennett, R. D., Norman, S. M., and Maier, V. P., ene and C40-carotenoid pigments in Harobacterium Intermediate steps in the biosynthesis of abscisic acid cutirubrum. Biochim. Biophys. Acta, 260, 492–506 from farnesyl pyrophosphate in Cercospora rosicola. (1972).