Applied Biochemistry and Microbiology, Vol. 40, No. 4, 2004, pp. 392397.

Translated from Prikladnaya Biokhimiya i Mikrobiologiya, Vol. 40, No. 4, 2004, pp. 460465. Original Russian Text Copyright 2004 by Davoli, Mierau, Weber.

Carotenoids and Fatty Acids in Red Yeasts Sporobolomyces roseus and Rhodotorula glutinis
P. Davoli1, V. Mierau2, and R. W. S. Weber2,*
1 Dipartimento

di Chimica, Universita di Modena e Reggio Emilia, Via Campi 183, 41100 Modena, Italy .2 Lehrbereich Biotechnologie, Universitat Kaiserslautern, Paul-Ehrlich-Str. 23, 67663 Kaiserslautern, Germany
Received July 27, 2003

AbstractRhodotorula glutinis and Sporobolomyces roseus, grown under different aeration regimes, showed differential responses in their carotenoid content. At higher aeration, the concentration of total carotenoids increased relative to the biomass and total fatty acids in R. glutinis, but the composition of carotenoids (torulene > -carotene > -carotene > torularhodin) remained unaltered. In contrast, S. roseus responded to enhanced aeration by a shift from the predominant -carotene to torulene and torularhodin, indicating a biosynthetic switch at the -carotene branch point of carotenoid biosynthesis. The overall levels of total carotenoids in highly aerated asks were 0.55 mol-percent and 0.50 mol-percent relative to the total fatty acids in R. glutinis and S. roseus (respectively), and 206 and 412 g g1 dry weight (respectively).

Red yeasts are a diverse assemblage of unrelated organisms (mostly Basidiomycota) that grow on the phylloplane or on decaying plant material. The astaxanthin producer Phaffia rhodozyma M.W. Miller et al. 1976 belongs to the Heterobasidiomycetes, whereas Sporobolomyces roseus Kluyver & van Niel 1924 and Rhodotorula glutinis (Fresenius 1863) F.C. Harrison 1928 belong to the Urediniomycetes, i.e. the rust fungi and their allies [1]. A third discrete group of red yeasts is allied with the Ustilaginomycetes (smut fungi). Carotenoids can be produced in abundance by red yeasts, with typical concentrations ranging from 50 to 350 g g1 dry weight e.g. in wild-type strains of Rhodotorula spp. [2, 3]. Carotenoids are generally localized in lipid droplets in red yeasts and also in other fungi [4, 5]. Although the physiological role of carotenoids is difcult to prove conclusively, several lines of evidence point at an antioxidant function. Ecologically, the typical phylloplane habitat of red yeasts involves exposure to high UV irradiation. Studies with mutants or carotenoid biosynthesis inhibitors have shown that carotenoid-decient yeast strains are sensitive to free oxygen radicals or oxidizing situations, and that this sensitivity can be relieved by the addition of exogenous carotenoids [6]. Further, carotenoid synthesis responds to free radical generators [7], aeration [8], and illumination [9]. Carotenoids are thought to exert their antioxidant activity by two mechanisms [10]. Singlet oxygen is inactivated by the dissipation of excess energy as heat, whereby the carotenoid molecule remains structurally intact. In contrast, the quenching
* Corresponding author (tel.: +49631205-4657, fax: +496312052999; e-mail: rwsweber@rhrk.uni-kl.de

of free radicals such as the peroxyl radical destroys the carotenoid molecule. Previous studies on the effect of aeration on Rhodotorula spp. have revealed minor effects on the composition of carotenoids, although the overall synthesis was often stimulated by aeration higher than that required for optimal growth [8, 11]. Since no equivalent attention has been given to Sporobolomyces, a ubiquitous component of the mycota on the leaves of most plant species and in the air, the present study was carried out. Our initial results [4] indicated a strong stimulation of carotenoid synthesis by enhanced aeration. Here, we report on the composition of carotenoids under different aeration regimes and in relation to the fatty acid prole of S. roseus and R. glutinis. These two phylogenetically related yeasts showed very different responses to the experimental conditions employed. The use of HPLC and HPLCMS methods for separating and identifying carotenoids allowed us to distinguish -carotene from -carotene and quantify them separately. MATERIALS AND METHODS Rhodotorula glutinis var. dairenensis Hasegawa & Banno 1958 (ATCC 26085) was obtained from the American Type Culture Collection and S. roseus (D99040) from the Culture Collection, Dept. of Biotechnology, University of Kaiserslautern. Both yeasts were maintained on 2% malt extract agar plates and grown in liquid yeast extract-dextrose (YED) medium containing 30 g glucose, 4 g yeast extract, 1 g KH2PO4 and 0.5 g MgSO4 7H2O per 1 l tap water. Incubation was as 150 ml aliquots in 500 ml conical asks. The cotton bungs were loosely made to improve aeration. In

0003-6838/04/4004-0392 2004 MAIK Nauka /Interperiodica

CAROTENOIDS AND FATTY ACIDS IN RED YEASTS

393

order to test the effect of aeration, conical asks with one indentation were used in addition to the standard asks. Cultures were incubated at 24C on an orbital shaker (120 rpm, 5 cm amplitude) and harvested after 108 h when no further biomass growth or increase in carotenoid production was observed. Biomass was determined by pelleting 1.5 ml culture aliquots in preweighed Eppendorf tubes, and drying the pellets at 80C for 24 h followed by re-weighing. Each experimental treatment was carried out in triplicate, with the results expressed as arithmetic means standard deviation. Cultures were harvested by centrifugation (10 min, 12000 rpm) followed by washing in an equal volume of dist. water. The pellets were frozen at 20C. Carotenoids and lipids were extracted by suspending thawed pellets in 5 ml dimethyl sulfoxide (DMSO) and incubating them for 12 h in the dark at room temperature [12]. This was followed by centrifugation as above, collection of the DMSO phase, and resuspension of the pellet in 20 l of acetone. Centrifugation resulted in a colored acetone phase. Acetone extraction and, if necessary, DMSO incubation were repeated until the yeast pellet had become entirely discolored. Carotenoids were recovered from the pooled DMSO and acetone extracts (about 100 ml) by extraction with an equal volume of light petroleum (boiling point range, 4575C), adding ~50 ml distilled water and 510 ml saturated NaCl solution to achieve an improved phase separation. The pigmented upper phase was washed three times with dist. water, and rotary-evaporated to dryness. This crude extract was dissolved in 0.5 or 2 ml acetone for HPLC analysis of carotenoids; for lipid analysis, n-hexane (0.5, 1 or 2 ml) was used as the solvent. HPLC analysis was carried out with an HP 1090 Series II liquid chromatograph (Hewlett Packard, Waldbronn, Germany) equipped with a diode array detector and a LiChrospher(r) 100 RP18 column (5 m particle size; 250 4 mm; Merck, Darmstadt, Germany). The injection volume was 10 l, and the ow rate 1 ml min1. A water/acetone (A/B) gradient [13] was used, running from 70% B to 100% B in 15 min followed by 2 min at 100% B. Carotenoids were quantied against peak area calibrations obtained from standard curves. -Carotene was supplied by Sigma (St. Louis, USA) whereas torularhodin was puried from Rhodotorula mucilaginosa (A. Jrgensen) F.C. Harrison 1928 as follows. A crude extract prepared as described above was chromatographed on a silica gel column and eluted with cyclohexane/ethyl acetate (95 : 5) in order to remove non-polar carotenes; the red band which had remained adsorbed on the top of the column was recovered by elution with cyclohexane/ethyl acetate (60 : 40) and was found to contain torularhodin as the sole pigment by HPLC analysis. After the actual concentration of torularhodin in the stock solution had been spectrophotometrically assessed, this sample was used for preparing standard solutions for the HPLC calibration curve at 500 nm. The calibration curves for -carotene and torulene were
APPLIED BIOCHEMISTRY AND MICROBIOLOGY

calculated relative to that of -carotene at 450 nm by applying the appropriate correction factors on the basis of their published absorption values [14, 15]. The identication of these four pigments in the yeast extracts was conrmed by mass spectrometry using a Hewlett-Packard Series 1100LC-MSD liquid chromatograph-mass spectrometer in the atmospheric pressure chemical ionization (APCI) mode and under the same conditions as described previously [15], except that a water-acetone gradient was used as described above for HPLC. For gas chromatography, the crude extracts (200 l) were transesteried in a sealed vial by treatment with 100 l of a fresh solution of sodium methoxide in methanol [15]. Methyl 10-undecenoate (100 l of a 10 mg ml1 stock solution in n-hexane) was used as the internal standard for quantication. The composition and concentration of the resulting fatty acid methyl esters was determined by GC-MS with a Hewlett Packard 5890 Series II chromatograph coupled to an HP5989A mass spectrometer [15]. RESULTS Cultures of both yeasts grown in indented asks showed a more intense red pigmentation than those in straight asks, indicating an effect on carotenoid composition and/or concentration in both yeast species. A moderate increase in biomass in indented asks was also observed for both species, the values being 6.31 0.47 mg ml1 in straight asks and 9.01 0.34 mg ml1 in indented ask cultures of R. glutinis, and 4.82 0.34 and 9.3 1.5 mg ml1 (respectively) for S. roseus. In order to quantify these changes, HPLC analyses were performed, the results of which are presented in Fig. 1 and Table 1. When extracts from an equivalent volume of culture were measured, a substantial increase in the concentration of all four pigments was observed in R. glutinis (Fig. 1a), but no changes in pigment composition were evident. Calculated relative to biomass, the levels of all four pigments increased by a factor of about 1.32.8 (Table 1). In contrast, extracts of S. roseus obtained from indented asks showed a strong and selective increase in the torulene, torularhodin, and -carotene peaks (5.710.2 fold relative to biomass) whereas only a slight increase (1.4 fold) in -carotene was observed (Fig. 1b; Table 1). Expressed as percentages of total carotenoid content, torulene, torularhodin, and -carotene increased at the expense of -carotene in S. roseus whereas no significant changes in pigment proportions were seen in R. glutinis (Fig. 2). The identity of all four pigments was conrmed by their mass spectra in the APCI-negative mode, with the molecular ions at m/z 536 (- and -carotene), 534 (torulene), and 564 (torularhodin). The fatty acid composition was similar in both yeast species and did not change in response to aeration (Table 2). The dominant fatty acids were oleic and
No. 4 2004

Vol. 40

394 A450 500 Absorbance at 450 nm (mAU) 400 300 II 200 100 1 0 500 Absorbance at 450 nm (mAU) (b) 400 4 300 200 100 0 10 1 3 I II I 3 4 2

DAVOLI et al.

()

tion increased relative to the biomass (from 113 to 206 g g1 dry weight) whereas the fatty acid level remained unchanged around 19.6 mg g1. Overall, therefore, the mol-percentage of carotenoids relative to fatty acids increased from 0.33 to 0.55%. In contrast, S. roseus responded to enhanced aeration by a strong increase in both the fatty acid (from 14.4 to 42.2 mg g1) and carotenoid levels (from 109 to 412 g g1) relative to the biomass, effecting only a small increase in the molar ratio from 0.42 to 0.50%. DISCUSSION R. glutinis and S. roseus responded differently to the enhanced aeration and, possibly, shear stress imposed by the use of indented asks as compared to standard (straight) asks. For R. glutinis, our data on the fatty acid composition are equivalent to those reported by Perrier et al. [2], and the -carotene concentration was also very similar. However, in [2], -carotene was the major pigment, whereas torulene prevailed in our experiments. A prevalence of torulene was also observed in a wild-type strain of R. glutinis [3], although it was possible to generate mutants with a strong shift towards -carotene. Like us, Zalashko et al. [11] observed that enhanced aeration did not affect the ratio between -carotene, torulene and torularhodin in R. glutinis. In contrast, other induces of oxidative stress such as illumination or free radical generators had a signicant effect on the carotenoid composition, selectively promoting torularhodin synthesis [16, 17]. It is possible that different signalling cascades interact in integrating a range of environmental stimuli such as light, free radicals, and abundance of oxygen, to regulate the production of carotenoids. In S. roseus, the carotenoid composition was dramatically altered in indented asks due to a muchincreased synthesis of torulene, torularhodin and -carotene. The biosynthetic pathway of these carotenoids is relatively well understood (Fig. 3), -carotene representing a precursor and branch-point in the synthesis of the other three pigments [18, 19]. Cyclization of the acyclic end group of -carotene leads to -carotene, whereas desaturation at the 3',4' position removes two hydrogen atoms to give torulene (3',4'-didehydro-,carotene). Subsequent oxidation of the methyl group at

Retention time (min) 2

12

14

16 18 Retention time (min)

Fig. 1. HPLC chromatograms of the pigment proles of R. glutinis (a) and S. roseus (b). 10 l aliquots equivalent to extracts obtained from 1.5 ml original culture were injected in both cases. Straight line = straight ask culture; dotted line = indented ask culture. The four pigment peaks are labeled in order of their elution, i.e. torularhodin (1), torulene (2), -carotene (3) and -carotene (4).

linoleic (C18 : 1 and C18 : 2; 60.272.1%) followed by palmitic (C16 : 0; 13.420.4%) and stearic (C18 : 0; 5.5 16.6%). All other fatty acids were present only as traces (<5%). A summary of the carotenoid and fatty acid concentrations relative to the fungal biomass is given in Table 3. In R. glutinis, the total carotenoid concentra-

Table 1. Quantification of carotenoids (g carotenoids g1 dry weight) in shaken flask cultures of R. glutinis and S. roseus. The factor by which carotenoid production increased in indented relative to straight flasks is also given Pigments -Carotene -Carotene Torulene Torularhodin R. glutinis flask type 44.3 6.5 15.2 1.9 51.5 7.1 2.13 0.28 straight 74 19 19.2 4.5 107 46 6.0 2.3 indented 1.7 1.3 2.1 2.8 flask type 72 12 11.8 3.7 20.8 4.3 4.1 1.6 S. roseus straight 101 30 67 16 203 50 41.2 8.8
Vol. 40

indented 1.4 5.7 9.8 10.2

No. 4 2004

APPLIED BIOCHEMISTRY AND MICROBIOLOGY

CAROTENOIDS AND FATTY ACIDS IN RED YEASTS % 70 60 50 40 30 20 10 0 1 2 3 Rhodotorula glutinis 4 1 2 3 Sporobolomyces roseus 4 I II

395

()

(b)

Fig. 2. The carotenoid composition in response to changes in culture conditions. The left-hand columns of each pair show extracts from straight ask cultures, right-hand columns show indented asks. -Car = -carotene, -Car = -carotene, Tor = torulene, Trh = torularhodin.

C-16' by a mixed function oxidase yields torularhodin (3',4'-didehydro-,-caroten-16'-oic acid) via the corresponding alcohol and aldehyde, whose isolation has been reported [19, 20]. S. roseus therefore responded to enhanced oxidative stress by increasing the production of -carotene and switching biosynthesis at that branch point from -carotene to the red carotenoids (torulene and torularhodin). It is interesting to note in this context that one atom of atmospheric molecular oxygen is known to be incorporated by R. rubra (R. mucilaginosa) into the carboxylic acid group of torularhodin [22]. A similar biosynthetic switch was found in R. glutinis in response to temperature [21], channeling -car-

otene to -carotene at low temperature (5C), but torulene and torularhodin at 25C [18]. The antioxidant effect of torularhodin is superior to that of -carotene [16], and enhanced torularhodin synthesis has been observed in R. glutinis treated by peroxyl radicals or irradiated by blue light [16, 17]. The enhanced synthesis of torularhodin by S. roseus may well be a stress response of potential signicance in the survival of this organism in its exposed phylloplane habitat in nature and it is tempting to speculate that oxygen can affect carotenogenesis in yeast by acting on specic biochemical switches and thus channeling biosynthesis towards the production of particular caro-

Table 2. Fatty acid composition (% of total) of R. glutinis and S. roseus grown in straight and shaken flasks. All fatty acids were determined by GC-MS as their methyl esters Fatty acids C14:0 tetradecanoic acid C15:0 pentadecanoic acid C16:1 (Z)9-hexadecenoic acid C16:0 hexadecanoic acid C17:1 15-methyl11-hexadecanoic acid C17:0 heptadecanoic acid C18:2 (Z,Z)9.12-octadecadienoic acid C18:1 (Z)9-octadecenoic acid C18:0 octadecanoic acid C20:1 11-eicosenoic acid C20:0 eicosanoic acid C22:0 docosanoic acid C24:0 tetracosanoic acid R. glutinis flask type 1.21 0.45 0.52 0.14 2.08 0.53 15.52 0.77 1.28 0.26 0.38 0.06 72.1 1.6 5.72 0.52 0.45 0.19 0.11 0.05 0.36 0.17 0.24 0.16
Vol. 40

S. roseus straight flask type 1.12 0.19 0.39 0.19 1.75 0.06 13.38 0.11 0.56 0.07 0.59 0.04 60.2 2.1 16.58 0.71 1.47 0.42 1.05 0.26 1.55 0.48 1.39 1.10 straight 2.13 0.17 0.33 0.08 3.93 0.46 20.4 1.8 1.16 0.05 0.28 0.02 64.1 1.9 5.45 0.74 0.70 0.35 0.31 0.21 0.64 0.36 0.53 0.53

1.53 0.45 0.38 0.04 1.23 0.22 20.0 1.9 0.36 0.03 0.27 0.03 69.1 2.3 6.07 0.80 0.19 0.16 0.09 0.08 0.55 0.29 0.26 0.02
No. 4 2004

APPLIED BIOCHEMISTRY AND MICROBIOLOGY

396

DAVOLI et al.

Table 3. Total carotenoid and fatty acid balances for R. glutinis and S. roseus grown in straight and indented shaken flasks R. glutinis Parameter flask type Total carotenoids g g1 dry weight nmol g1 dry weight Total fatty acids (FAs) g g1 dry weight nmol g1 dry weight Carotenoids/FAs, mol % 19.7 6.2 70 21 0.33 0.14 19.5 7.1 70 21 0.55 0.04 14.3 4.9 51 14 0.42 0.15 42 23 160 68 0.50 0.11 113 14 211 27 206 72 385 134 109 21 203 40 412 103 766 192 straight flask type straight S. roseus

tenoids. We also note that the total amount of carotenoid produced by S. roseus in well-aerated asks (412 g g1 dry weight) is rather high for a wild-type yeast. Nonetheless, the total carotenoid values relative to fatty acids are lower for both yeasts than the 1 mol-percent concentration of -carotene effective in protecting phosphatidylcholine liposomes against oxidation by peroxyl radicals in vitro [23]. The values are also much lower than the combined and -carotene concentrations of approx. 6 mol-percent in aeciospores of the rust fungus Puccinia distincta [15]. In contrast, endogenous carotenoids are reported to protect R. mucilaginosa against experimentally induced oxidative injury by

reactive oxygen species [6], and it is noteworthy that this red yeast contains torularhodin as the main pigment. Whether the carotenoid concentrations necessary for antioxidant activity in living organisms are equivalent to those in articial membranes, or whether non-carotenoid antioxidants are also present in S. roseus and R. glutinis, are topics for further research. Finally, it would be very interesting to test the antioxidant properties of torulene and torularhodin in vitro, in comparison with those of -carotene and ketocarotenoids such as astaxanthin. Such data might provide an additional stimulus towards the commercial development of these two abundant yeast pigments as food dyes or nutraceuticals [24].

-carotene

3' 4'

16'

Torulene

-carotene

COOH Torularhodin
Fig. 3. The biosynthetic relationships between -carotene, -carotene, torulene and torularhodin. Based on the results described in [18] and [19]. APPLIED BIOCHEMISTRY AND MICROBIOLOGY Vol. 40 No. 4 2004

CAROTENOIDS AND FATTY ACIDS IN RED YEASTS

397

ACKNOWLEDGMENTS We are grateful to Prof. Timm Anke and Mrs. Anja Meffert (Dept. of Biotechnology, University of Kaiserslautern) for the use of experimental facilities and for recording the carotenoid mass spectra, respectively. REFERENCES
1. Fell, J.W., Boekhout, T., Fonseca, A., Scorzetti, G., and Statzell-Tallman, A., Int. J. Syst. Evol. Microbiol., 2000, vol. 50, pp. 13511371. 2. Perrier, V., Dubreucq, E., and Galzy, P., Arch. Microbiol., 1995, vol. 164, pp. 173179. 3. Bhosale, P.B. and Gadre, R.V., Appl. Microbiol. Biotechnol., 2001, vol. 55, pp. 423427. 4. Davoli, P. and Weber, R.W.S., Mycologist., 2002, vol. 16, pp. 102109. 5. Weber, R.W.S. and Davoli, P., New Phytol., 2002, vol. 154, pp. 471479. 6. Moore, M.M., Breedveld, M.W., and Autor, A.P., Arch. Biochem. and Biophys., 1989, vol. 270, pp. 419431. 7. Sakaki, H., Nochide, H., Komemushi, S., and Miki, W., J. Biosci. Bioeng., 2002, vol. 93, pp. 338340. 8. Simova, E.D., Frengova, G.I., and Beshkova, D.M., Z. Naturforsch. C, 2003, vol. 58, pp. 225229. 9. Sakaki, H., Nakanishi, T., Satonaka, K.Y., Miki, W., Fujita, T., and Komemushi, S., J. Biosci. Bioeng., 2000, vol. 89, pp. 203205. 10. Edge, R., McGarvey, D.J., and Truscott, T.G., J. Photochem. Photobiol. B, 1997, vol. 41, pp. 189200. 11. Zalashko, M.V., Salokhina, G.A., and Koroleva, I.F., Appl. Biochem. Microbiol., 2000, vol. 36, pp. 2932

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

(Translated from Prikl. Biokhim. Mikrobiol., vol. 36, pp. 3740). Martin, A.M., Chun, L., and Thakor, R.P., J. Ferment. Bioeng., 1993, vol. 76, pp. 321325. Steel, C.C. and Keller, M., Biochem. Soc. Trans., 2000, vol. 28, pp. 883885. Britton, G., UV/visible spectroscopy, Carotenoids 1B: Spectroscopy, Basel: Birkhauser, 1995, pp. 1362. Davoli, P. and Weber, R.W.S., Phytochem., 2002, vol. 60, pp. 309313. Sakaki, H., Nakanishi, T., Komemushi, S., Namikawa, K., and Miki, W., J. Clin. Biochem. Nutr., 2001, vol. 30, pp. 110. Sakaki, H., Nakanishi, T., Tada, A., Miki, W., and Komemushi, S., J. Biosci. Bioeng., 2001, vol. 92, pp. 294297. Simpson, K.L., Nakayama, T.O.M., and Chichester, C.O., J. Bacteriol., 1964, vol. 88, pp. 16881694. Bonaly, R. and Malenge, J.P., Biochim. Biophys. Acta, 1968, vol. 164, pp. 306316. Goodwin, T.W., The Biochemistry of the Carotenoids 1: Plants, second edition, London: Chapman and Hall, 1980. Nakayama, T., Mackinney, G., and Phaff, H.J., Antonie van Leeuwenhoek, 1954, vol. 20, pp. 217228. Simpson, K.L., Nakayama, T.O.M., and Chichester, C.O., Biochem. J., 1964, vol. 92, pp. 508510. Woodall, A.A., Britton, G., and Jackson, M.J., Biochim. Biophys. Acta, 1997, vol. 1336, pp. 575586. Ershov, Yu.V., Dmitrovsky, A.A., Polulyakh, O.V., Podoprigora, O.I., and Bykhovsky, V.Ya., Prikl. Biokhim. Mikrobiol., 1992, vol. 28, pp. 680684.

APPLIED BIOCHEMISTRY AND MICROBIOLOGY

Vol. 40

No. 4

2004