- Email: [email protected]
Rhodotorulic acid production by Rhodotorula mucilaginosa
Mycol. Res. 107 (8): 949–956 (August 2003). f The British Mycological Society
949
DOI: 10.1017/S0953756203008220 Printed in the United Kingdom.
Rhodotorulic acid production by Rhodotorula mucilaginosa
Ditte ANDERSEN1, Joanna C. RENSHAW2 and Marilyn G. WIEBE1* 1
Department of Life Sciences, Sohngaardsholmsvej 49, Aalborg University, DK-9000 Aalborg, Denmark. Centre for Radiochemistry Research, Department of Chemistry, University of Manchester, Manchester M13 9PL, UK. E-mail : [email protected] 2
Received 8 January 2003; accepted 20 June 2003.
Rhodotorula mucilaginosa produces the siderophore rhodotorulic acid (RA) when grown in iron-limited conditions. R. mucilaginosa grew at rates between 0.10 and 0.19 hx1 in iron-restricted conditions, depending on the carbon source, and at 0.23 hx1 in iron-sufficient conditions. In bioreactors inoculated with iron-starved pre-cultures, initial specific growth rates in batch culture were dependent on the iron concentration. The critical dilution rate (Dcrit, at which steady state cultures cannot be sustained) in continuous cultures was also dependent on the iron concentration and was lower than mmax in batch culture. Sucrose was the best carbon source for RA production [287¡11 mmol (g biomass)x1] and production could be further increased by supplementing the medium with the precursors acetate [460¡13 mmol (g biomass)x1], ornithine [376¡6 mmol (g biomass)x1], or both [539¡15 mmol (g biomass)x1]. Citric acid was an effective suppresser of RA production. RA was produced in a growth rate dependent manner and was optimally produced at pH 6.5.
INTRODUCTION The yeast Rhodotorula mucilaginosa produces the siderophore rhodotorulic acid (RA) when grown in ironrestricted conditions (Atkin, Neilands & Phaff 1970). Whilst most fungi make more than one siderophore (Winkelmann 1985), R. mucilaginosa produces only RA and is thus a good test organism for studying the influence of environmental parameters on siderophore production. In addition, siderophore production has recently been demonstrated to be important for both R. mucilaginosa (syn. R. rubra) and R. glutinis in their use as biocontrol agents for post-harvest fungal diseases of fruits (Calvente, Benuzzi & Tosetti 1999, Calvente et al. 2001) and RA can be used as an effective treatment of iron chlorosis in plants (Miller 1989, Johnson et al. 2002). Siderophores of other micro-organisms are of interest because of their potential use in treating medical conditions such as iron and aluminium overloading (Tilbrook & Hider 1998) and in the decorporation of toxic metals such as plutonium (Stradling 1998). The use of siderophore-drug conjugates for targeted drug delivery (Ghosh & Miller 1995, Lu & Miller 1999), siderophores as anti-tumour chemotherapeutic agents (Vergne, Walz & Miller 2000, Planalp et al. 2002), and * Corresponding author.
as chelators to reduce the toxicity of other chemotherapeutic treatments (Tilbrook & Hider 1998) is also being investigated. Establishing the factors which affect siderophore production in fungi will facilitate their isolation from a range of other fungal species. Although the production of RA by R. mucilaginosa and related species has been studied since 1968 (Atkin & Neilands 1968) and it can easily be obtained from cultures starved of iron, the role of parameters other than iron concentration has received little attention. Both carbon and nitrogen sources affect siderophore production in Aspergillus species (Crueger & Za¨hner 1968, Kappner, Hasenbo¨hler & Za¨hner 1977), and recently Calvente et al. (2001) observed that amino acids in the medium and pH affected RA production by Rhodotorula species. As with production of siderophores by other fungi, RA is produced both during exponential growth and during stationary phase (Atkin, Neilands & Phaff 1970) and the relationship between growth and siderophore production is unclear. This paper presents data on the growth of R. mucilaginosa under iron-restricted (i.e. iron is the first nutrient to restrict biomass production) batch and ironlimited (i.e. iron supply in the medium is limited to a sub-optimal concentration, while all other nutrients are provided in excess) chemostat cultures and investigates the importance of the carbon source and the precursors
Siderophore production by Rhodotorula mucilaginosa acetate and ornithine for RA synthesis. In addition, the relationships between growth, pH and RA production are established. MATERIALS AND METHODS Organism and media Rhodotorula mucilaginosa was obtained from the American Type Culture Collection (ATCC 26423) and was maintained as cells on agar-solidified (15 g agar lx1) medium at 4 xC or stored in glycerol (20 % w/v) at x70 x. The defined medium of Vogel (1956) was modified to give an iron-deficient medium with low buffering capacity and contained : 3.0 g (NH4)2SO 4 lx1, 0.3 g KH2PO4 lx1, 0.2 g MgSO4 . 7H2O lx1, 0.1 g CaCl2 . 2H2O lx1, 2.25 mg citric acid lx1 (to keep zinc and copper in solution during medium preparation), 2.25 mg ZnSO4 . 7H2O lx1, 0.11 mg CuSO4 . 5H2O lx1, 0.02 mg MnSO4 . 4H2O lx1, 0.02 mg H3BO3 lx1, 0.02 mg Na2MoO4 . 2H2O, 0.02 mg biotin lx1 and 2 mg thiamine lx1, with no Na3Citrate or Fe(NH4)2(SO4)2 . 6H2O. Iron was provided as FeCl3.6H2O (0–26.6 mM added Fe3+). Sucrose, glucose, fructose, NaAcetate, xylose, raffinose or maltose were used as carbon sources (10 g lx1) and some media were supplemented with ornithine (0.25–4 g lx1), glutamate (1 g lx1) or aspartate (1 g lx1). Media for batch cultures in 250 ml flasks were buffered with 0.5 M 2-(N-morpholino)ethanesulfonic acid (MES), pH 6.0. The defined medium of Vogel (1956) was used without modification, other than the substitution of (NH4)2SO 4 for NH4NO3 and addition of 2 mg thiamine lx1, for iron-sufficient medium. Iron-sufficient medium thus differed from iron-deficient medium in that it contained 1.05 mg Fe(NH4)2(SO4)2 . 6H2O lx1, with 2.5 g Na3Citrate lx1 and 5 g KH2PO 4 lx1. Cultural conditions Small scale batch cultures were grown in 250 ml flasks containing 50 ml medium. Flasks were incubated at 25 x on rotary shakers at 200 rpm (throw=2.5 cm). Bioreactor cultures were grown in a Braun Biostat M bioreactor (2.1 l full working volume). Cultures were maintained at 25 x and aerated at 0.8 l air [l culture]x1 minx1 with 500 rpm agitation (three 4-blade disk turbine impellers, 48 mm diam). pH was maintained at values between 3.0 and 7.6¡0.1, by addition of 0.5 M NaOH. Polypropylene glycol 2000 (0.05–0.4 % v/v, final concentration ; JT Baker) was added to prevent foaming. Measurements of growth and biomass Growth was measured as increase in optical density, measured at 550 nm (Cecil CE 7200 spectrophotometer), using water as the blank. Specific growth rates were also determined from the decrease in optical
950 density of bioreactor cultures with dilution rates set above the critical dilution rate (Dcrit, the dilution rate above which biomass is removed from the bioreactor at a faster rate than it can be replaced by growth ; Esener et al. 1981). For dry weight analyses, samples were collected on pre-dried, pre-weighed glass fibre filters (GA55 Advantec), rinsed with de-ionised water and dried at 90 x. Dry weight analyses were complicated by the presence of extracellular material (e.g. polysaccharides and polypropylene glycol) which blocked the filters and made it difficult to rinse the cells. The average relationship between dry weight and optical density was found to be OD at 550 nm=3.7rdry weight (g lx1), r=0.82, which was used to estimate biomass yields from flask cultures. Yield of biomass on carbon was expressed as Cmol biomass (Cmol carbon)x1, in order to readily compare yields on different carbon sources. Cmol biomass was taken as 25.1 Cmol (g biomass)x1, based on an average microbial cell composition of CH1.81O0.52N0.21 (Stephanopoulos, Aristidou & Nielsen 1998). Yield expressed as Cmol biomass (Cmol carbon)x1 is thus determined as the yield expressed as (g biomass) (g carbon)x1 divided by 25.1 g biomass (Cmol biomass)x1 and multiplied by the number of Cmol per g carbon for the specified carbon source [e.g. 30 Cmol (g glucose)x1]. This gives a small overestimate of the yield (Cmol Cmolx1), since ash content was not measured or taken into account. Rhodotorulic acid assay Rhodotorulic acid (RA) was quantified by measuring the absorbance of the ferric–RA complex at 475 nm (Atkin & Neilands 1968). There was no absorbance from other supernatant components at this wavelength. Cells were removed by centrifugation and supernatant (160 ml) was mixed with 840 ml FeCl3 . 6H2O (5 mM in 0.1 M HCl) and the absorbance measured in an Ultrospec 3000 spectrophotometer, with de-ionised water replacing the supernatant for the blank. HPLC analysis of carbohydrates Culture supernatant was filtered through 0.22 mm (pore diam) filters and samples were analysed on an Aminex HPX-87H HPLC organic acid analysis column (300r7.8 mm, Bio-Rad), with H2SO4 (5 mM, flow 0.4 ml minx1) as the mobile phase. Carbohydrates were detected with a Knaver refractometer. RESULTS Specific growth rate in iron-sufficient and iron-restricted media Rhodotorula mucilaginosa grew at 0.23 hx1 in ironsufficient medium with sucrose as the carbon source. The specific growth rate in iron-restricted medium was
D. Andersen, J. C. Renshaw and M. G. Wiebe
951 not significantly (P>0.05 hx1, Scheffe’s multiple range test) affected when sucrose medium was supplemented with NaAcetate, ornithine, glutamate or aspartate (Table 2). R. mucilaginosa ATCC 26423 was unable to utilise maltose.
Specific growth rate (h–1)
0.25
0.20
0.15
The effect of substrate concentration on biomass yield and rhodotorulic acid production
0.10
0.05
0.00 0.0
0.5
1.0
1.5
2.0
2.5
3.0
Added iron (M)
Fig. 1. The initial specific growth of Rhodotorula mucilaginosa in iron-restricted batch cultures (%) and the calculated specific growth rate during washout conditions in iron-limited continuous flow culture (&), for cultures growing at pH 6.0¡0.1, 25 x, 500 rpm, with various concentrations of added iron.
dependent on the inoculum, the iron concentration (Fig. 1) and the carbon source (Tables 1–2). Preliminary experiments were inoculated with yeast cells which had been grown on iron-sufficient agar-solidified medium, whereas subsequent experiments were inoculated with cells which had been grown in iron-restricted liquid medium. Even for cells grown on iron-sufficient medium, initial specific growth rates in iron-restricted (0.63 mM added Fe3+) medium with sucrose as the carbon source were reduced to 0.15¡0.004 hx1. When iron-starved inoculum was used, the initial specific growth rate with 0.63 mM added Fe3+ was 0.11 hx1, and specific growth rate increased with increasing iron concentration (Fig. 1). In iron-restricted medium, the initial exponential growth phase was maintained until approximately 30 % of the final OD had been attained, after which the cultures entered a prolonged deceleration phase of approximately 20 h. During decelerating growth, specific growth rates declined from approx. 0.08 to 0.015 hx1. Dcrit during continuous medium flow was considerably lower than mmax measured during batch growth in the bioreactor, but was similar to the measurement of specific growth rate during washout conditions (Fig. 1). Dcrit was lower when 0.63 mM added Fe3+ was supplied to the culture (Dcrit<0.017 hx1), than when 1.9 mM added Fe3+ was supplied (Dcrit<0.10 hx1). Initial specific growth rates in iron-restricted conditions (0.63 mM added Fe3+) were significantly (P< 0.05, Scheffe’s multiple range test) higher when glucose (m=0.19¡0.01 hx1) or fructose (m=0.20¡0.01 hx1) were provided as the carbon source, rather than sucrose (m=0.15¡0.01 hx1), but were significantly (P<0.05, Scheffe’s multiple range test) reduced when raffinose (m=0.12¡0.004 hx1), NaAcetate (0.10¡0.004 hx1) or xylose (m=0.06¡0.001 hx1) were provided as carbon sources (Table 1). The initial specific growth rate was
For glucose concentrations between 2.5 and 10 g lx1 (0.63 mM added Fe3+), Rhodotorula mucilaginosa had a yield of biomass on glucose of 0.56¡0.02 Cmol biomass (Cmol substrate)x1 [0.48 g biomass (g substrate)x1]. Biomass yield was slightly reduced with 15 and 20 g glucose lx1 [0.52¡0.03 and 0.48¡0.02 Cmol biomass (Cmol substrate)x1, respectively ; significant at P<0.15, Scheffe’s multiple range test, for Yx/C with 20 g glucose lx1]. Specific RA production increased with increasing glucose concentration between 2.5 [100 mmol RA (g biomass)x1] and 10 g glucose lx1 [170 mmol RA (g biomass)x1] and remained constant at concentrations above 10 g glucose lx1 (Fig. 2). The yield of RA on glucose was greatest with 10 g glucose lx1 (2.5 mmol RA Cmolx1 ; Fig. 2). The effect of iron concentration and the presence of citric acid on rhodotorulic acid production RA production decreased in iron-deficient medium with increasing concentration of added iron, but was still detectable with 26.6 mM added iron (Fig. 3). No significant differences were observed in the molar concentration of RA produced when 0–1.9 mM iron was added. However, when 2.55 mM iron was added in the presence of 1.63 g citric acid lx1, RA production [0.01 mM, or 0.03 mmol RA (Cmol sucrose)x1] was almost completely suppressed (Fig. 3). Biomass yield in the presence of 2.55 mM added iron and 1.63 g citric acid lx1 was similar to that observed with 26.6 mM added iron with only 0.0023 mg citric acid lx1. The effect of carbon source on rhodotorulic acid production More RA was produced when Rhodotorula mucilaginosa was grown on sucrose than on any other carbon source tested (Table 1). The yields of biomass on total carbon for R. mucilaginosa growing on medium supplemented with 0.63 mM Fe3+, with various carbon sources, are given in Table 1. The biomass yields on raffinose and sodium acetate were significantly (P<0.05, Scheffe’s multiple range test) reduced compared to the yields on other carbon sources. Biomass yields on sucrose were measured during chemostat cultures with 0.63 or 1.9 mM added iron. Residual carbon was measured and the yield of biomass on consumed carbon was 0.48 and 0.58 Cmol
Siderophore production by Rhodotorula mucilaginosa
952
Table 1. Specific growth rates, yield coefficients (biomass on carbon, Yx/C, and RA on carbon, YRA/C), and specific RA production [mmol RA (g biomass)x1] for Rhodotorula mucilaginosa growing in iron-restricted, modified Vogel’s medium (0.63 mM Fe3+, pH 6) on various carbon sources (10 g lx1). Cultures were grown for at least 90 h in 250 ml flasks shaken at 200 rpm at 25 x. Yield coefficients were estimated from OD measurements (OD550=3.67rDW g lx1). Values in the same column with the same superscript, a to e, did not differ significantly (P>0.05, Scheffe’s multiple range test).
Carbon source
Specific growth rate (hx1)a
Yield (Yx/C) of biomass on carbon (Cmol Cmolx1)b
Specific RA production (mmol RA gx1)
RA (YRA/C, mmol Cmolx1)c
Sucrose Glucose Fructose Raffinose Xylose NaAcetate
0.15¡0.004c 0.19¡0.003d 0.20¡0.008d 0.12¡0.004b 0.06¡0.001a 0.10¡0.004b
0.56¡0.003b 0.58¡0.016bc 0.64¡0.005c 0.24¡0.002a 0.53¡0.003bc 0.30¡0.003a
287¡11 181¡7 184¡4 170¡3 111¡4 221¡9
4.0¡0.1d 2.6¡0.04c 3.0¡0.04c 1.0¡0.01a 1.5¡0.05ab 1.7¡0.05b
a b c
n=3–9 replicates. n=3–42 replicates. n=3–46 replicates.
Table 2. Specific growth rates, yield coefficients (biomass on carbon, Yx/C, and RA on total carbon, YRA/C), and specific RA production [mmol (g biomass)x1] for Rhodotorula mucilaginosa growing in iron-restricted, modified Vogel’s medium (0.63 mM Fe3+, pH 6) with sucrose as the primary carbon source, and supplemented with acetate, ornithine, glutamic acid or aspartic acid. Cultures were grown for at least 90 h in 250 ml flasks shaken at 200 rpm at 25 x. Yield coefficients were estimated from OD measurements (OD550=3.67rDW g lx1). Values in the same column with the same superscript, a to f, did not differ significantly (P>0.05, Scheffe’s multiple range test).
Sucrose (Cmol)
Supplement (Cmol)
Specific growth rate (hx1)a
Yx/C (Cmol Cmolx1)b
Specific RA production (mmol RA gx1)
RA (YRA/C, mmol Cmolx1)c
0.35 0 0.09 0.18 0.26 0.35 0.26
– (0) Acetate (0.24) Acetate (0.18) Acetate (0.12) Acetate (0.06) Ornithine (0.04) Ornithine (0.04) Acetate (0.06) Glutamate (0.03) Acetate (0.06) Aspartate (0.03) Acetate (0.06)
0.15¡0.004bc 0.10¡0.004a 0.16¡0.014bc 0.16¡0.004bc 0.17¡0.0054c 0.14¡0.001ab 0.14¡0.002b
0.56¡0.003e 0.30¡0.003a 0.45¡0.005b 0.49¡0.005bc 0.53¡0.005d 0.50¡0.004c 0.48¡0.004bc
287¡11 221¡9 287¡6 460¡13 428¡15 376¡6 539¡15
4.0¡0.13c 1.7¡0.05a 3.2¡0.03b 5.6¡0.11d 5.8¡0.15d 4.7¡0.03cd 6.6¡0.12e
0.16¡0.004bc
0.56¡0.006de
323¡13
4.5¡0.13c
0.17¡0.003c
0.60¡0.005f
217¡7
3.3¡0.08b
0.26 0.26 a b c
n=3–12 replicates. n=9–28 replicates. n=6–30 replicates.
biomass (Cmol substrate)x1 with 0.63 and 1.9 mM added iron, respectively. Biomass yield was constant with 1.9 mM added iron for dilution rates (D) between 0.034 and 0.74 hx1. With 0.63 mM added iron, biomass yield was only measured at D=0.013 hx1, since the culture washed out at higher dilution rates. The effect of supplementation with acetate, ornithine, aspartate or glutamate Specific growth rate and biomass yield were low when acetate was provided as the sole carbon source (Tables 1–2), but when both acetate and sucrose were provided, in varying proportions, yields improved and Rhodotorula mucilaginosa grew at the same (P>0.05) specific growth rate as when only sucrose was supplied (Table 2). When 20 or 40 % of the carbon was provided as acetate, RA production was significantly (P<0.05, Scheffe’s multiple range test) increased, compared to
that observed when R. mucilaginosa was grown on sucrose alone (Table 2). When medium containing 10 g sucrose lx1 was supplemented with 1 g ornithine lx1, the yield of biomass on carbon [0.50¡0.004 Cmol biomass (Cmol carbon)x1] was reduced and the production of RA increased [376¡6 mmol RA (g biomass)x1 and 4.7¡0.03 mmol RA (Cmol substrate)x1], compared to growth in medium containing only sucrose (Table 2). R. mucilaginosa was also grown on medium containing 7.5 g sucrose lx1 and 2.5 g NaAcetate lx1, supplemented with 1 g ornithine lx1. The biomass yield on carbon [0.48¡0.004 Cmol biomass (Cmol carbon)x1] was reduced, compared to the same medium lacking ornithine [0.53¡0.005 Cmol biomass (Cmol carbon)x1], but the production of RA [539¡15 mmol RA (g biomass)x1 and 6.6¡0.12 mmol RA (Cmol substrate)x1] was significantly (P<0.05) increased (Table 2).
D. Andersen, J. C. Renshaw and M. G. Wiebe
953
200
3
150
600
10
cd c
ab
2
100
a 1
50
Specific RA prouction ( , M RA [g biomass]–1)
d
Specific RA production ( , mol [g biomass]–1)
Yield of RA on glucose ( , mmol RA Cmol–1)
500
8 d d
400
cd bc 6
b a
300
4 200 2
100 0 0
5
10
Yield of RA on carbon ( , mmol RA Cmol–1)
4
0 20
15
0
Glucose (g l–1)
0 0
Fig. 2. Yield of RA on glucose (%) and on biomass (&, specific RA production) for cultures of Rhodotorula mucilaginosa growing in iron-restricted Vogel’s medium (0.63 mM added iron, buffered at pH 6) containing different concentrations of glucose. Error bars represent ¡SEM and where not shown were smaller than the size of the symbol. Yield (of RA on glucose) values with different letters, a to d, were significantly (P<0.06, Scheffe’s multiple range test) different. 30 e
1
2 Ornithine (g l–1)
3
4
Fig. 4. The effect of ornithine concentration on the yield of RA on biomass (&, specific RA production) and on total carbon (%) for cultures of Rhodotorula mucilaginosa growing in ironrestricted Vogel’s medium (%, 0.63 mM added iron, 7.5 g sucrose lx1, 2.5 g NaAcetate lx1, buffered at pH 6). Where shown, error bars represent ¡SEM. Where not shown, error bars were smaller than the size of the symbol. Data points with different letters, a to d, were significantly (P<0.05, Scheffe’s multiple range test) different.
Yield of RA on glucose (YRA/C , , mmol Cmol–1)
e de
increased, but the production of RA was either unaffected (glutamate) or reduced (aspartate ; Table 2).
d
20
The relationship between growth and siderophore production
c 10
The specific rate of RA production increased with increase in dilution rate at dilution rates between 0.034 and 0.074 hx1, with 1.9 mM added iron, at pH 6.0 (Fig. 5). Biomass concentration decreased with increase in dilution rate.
b ab
ab a
0 0
10
20 Added Fe3+ (M)
30
Fig. 3. Yield of RA on glucose for cultures of Rhodotorula mucilaginosa growing in iron-restricted Vogel’s medium (%, 10 g glucose lx1, buffered at pH 6) containing different concentrations of added iron. The yield of RA on glucose for cultures of R. mucilaginosa growing in iron-sufficient complete Vogel’s medium (#) is also shown. Where shown, error bars represent ¡SEM. Where not seen, error bars were less than the size of the symbol. Data points with different letters, a to e, were significantly (P<0.05, Scheffe’s multiple range test) different.
Specific RA production increased with increasing concentrations of added ornithine, for concentrations of up to 1 g ornithine lx1 [539¡15 mmol RA (g biomass)x1 ; Fig. 4]. The yield of RA on carbon was highest with 0.25 or 0.5 g ornithine lx1 [7.0¡ 0.04 mmol RA (Cmol substrate)x1 ; Fig. 4]. When sucrose (7.5 g lx1) and NaAcetate (2.5 g lx1) medium was supplemented with either glutamate or aspartate (1 g lx1), the biomass yield on carbon was
The relationship between pH and siderophore production The relationship between pH and RA production was determined for cultures with 0.95 mM added iron at D=0.046 hx1. RA production increased with increase in pH for pH values between 3 and 6.5, with maximum production observed at pH 6.5 (Fig. 6). Above pH 6.5 RA production was reduced by increase in pH. Rhodoorula mucilaginosa could not sustain a growth rate of 0.045 hx1 at pH 8 under these conditions. DISCUSSION Rhodotorula mucilaginosa had a specific growth rate of 0.23 hx1 at 25 x in iron-sufficient, defined medium, which is comparable to the initial specific growth rate (estimated as 0.22 hx1) of R. glutinis in glucose-yeast extract medium at 24–28 x (Oudin et al. 1999, Bhosale & Gadre 2001a, b) and of R. mucilaginosa grown in phosphate-restricted defined medium at 25 x
Siderophore production by Rhodotorula mucilaginosa
10
0.5
5 0.0
0
5
6
2
4
1
2
0.02
0.04 0.06 0.08 Dilution rate (h–1)
de
200 cd bc
150 b
100
a 0
8
3
de
50
10
4
e 250 Specific RA production (mol RA [g biomass]–1)
15
Specific RA production rate ( , mol RA [g biomass]–1 h–1)
1.0
Residual carbon ( , g l–1)
Rhodotorulic acid ( , mM)
20
Biomass ( , g l–1)
300
25
1.5
0 0.00
954
0 0.10
Fig. 5. The effect of dilution rate on biomass concentration ($, g lx1), residual sugar concentration (#, g lx1), RA production (&, mM) and the specific RA production rate [%, mmol RA (g biomass)x1 hx1] for Rhodotorula mucilaginosa growing in iron-limited (1.9 mM added iron) chemostat culture with sucrose (10 g lx1) as the carbon source. Cultures were maintained at pH 6.0¡0.1, 25 x, 500 rpm, 0.7 l air (l culture)x1 minx1. Data are the averages of at least three consecutive samples taken during steady-state-conditions, at least three residence times after the condition had been established. Error bars represent ¡SEM, and where not seen were smaller than the size of the symbol.
(m=0.17 hx1 ; Button, Dunker & Morse 1973). In ironrestricted medium the specific growth rate was reduced, dependent on the concentration of iron in the medium, as has also been observed in iron-restricted cultures of Neisseria meningitidis (Archibald & DeVoe 1978), Agrobacterium tumefaciens (Kurowski & Pirt 1971) and mouse LS cells (Birch & Pirt 1970). Carry-over of iron in the inoculum could affect specific growth rates. In a chemostat, biomass is expected to be constant over a wide range of dilution rates, being significantly reduced only as the dilution rate approaches Dcrit, which should be very near the mmax of the organism (Pirt 1975). However, in an iron-limited chemostat the steady-state biomass may decrease with increase in dilution rate for dilution rates below Dcrit (Pirt 1975, Weger 1999). This decrease in biomass with increasing dilution rate was observed for R. mucilaginosa (Fig. 5). Dcrit was also reduced in iron-limited chemostats and was dependent on the iron concentration (Fig. 1). Reduced Dcrit values in iron-limited chemostats have previously been reported for the filamentous fungus Fusarium venenatum (Wiebe 2002) and the bacterium Methanobacterium thermoautotrophicum (Liu et al.
3
4
5
6
7
8
pH
Fig. 6. The effect of pH on RA production for Rhodotorula mucilaginosa growing in iron-limited (0.95 mM added iron) chemostat culture with sucrose (10 g lx1) as the carbon source. Cultures were maintained at 25 x, 500 rpm, 0.7 l air (l culture)x1 minx1. Data are the averages of at least three consecutive samples taken during steady state-conditions, at least three residence times after the condition had been established. Data points with different letters, a to e, were significantly (P<0.05, Scheffe’s multiple range test) different. Error bars represent ¡SEM, and where not seen were smaller than the size of the symbol.
1999), but a decrease in Dcrit with decreasing iron concentration has not previously been reported. Because iron concentration affected Dcrit, the relationship between growth rate and RA production could only be assessed at the relatively high iron concentration of 1.9 mM added iron. RA was produced in a growth rate dependent manner, with the highest specific production rate being observed at the highest specific growth rate (Fig. 5). This agrees with the high concentrations of RA observed during exponential growth of batch cultures and the production of RA synthetase during the early log phase in batch cultures (Anke & Diekmann 1972), but does not explain the increasing concentrations of RA observed during stationary phase (Atkin et al. 1970). Production of siderophores by F. venenatum was also growth rate correlated (Wiebe 2002). Atkin & Neilands (1968) observed that production of RA increased from approx. 2 g lx1 to 3–4 g lx1, when the carbon concentration in the medium was increased from 10 to 20 g lx1 and other researchers (Atkin et al. 1970, Anke & Diekmann 1972, Carrano & Raymond 1978, Calvente et al. 2001) have frequently used media containing 20–40 g carbohydrate lx1 to produce RA. However, high carbon concentrations are also likely to lead to oxygen limited conditions. When glucose concentration was increased from 2.5 to 10.0 g glucose lx1 (0.63 mM added Fe3+), the biomass yield [0.57 Cmol biomass (Cmol glucose)x1] remained constant and specific RA production increased (Fig. 2). This may reflect a reduction in availability of iron with increase in glucose concentration, since the concentration of
D. Andersen, J. C. Renshaw and M. G. Wiebe added iron was constant but the concentration of biomass produced increased, and therefore the concentration of iron available per g biomass decreased. For concentrations of glucose above 10 g glucose lx1, the yield of both biomass (Yx/C was reduced 14 % for 20 g glucose lx1 compared to 10 g glucose lx1 ; P<0.15) and RA (YRA/C was reduced 17 % for 20 g glucose lx1 compared to 10 g glucose lx1 ; P<0.06) on carbon were reduced. The reduced biomass yield suggests the occurrence of oxygen limited conditions, since biomass yield was not affected by the concentration of added iron for concentrations between 0.16 and 19 mM added iron (data not shown). Atkin et al. (1970) reported that RA production was sensitive to oxygen limitation, and the reduction in yield of both biomass and RA on carbon at high glucose concentrations indicates that high carbon concentrations (e.g. >20 g carbohydrate lx1) should only be used for RA production when adequate aeration can be provided. Atkin et al. (1970) observed decreasing RA production by R. mucilaginosa for iron concentrations between 0.18 and 1.8 mM added iron and complete suppression of RA with only 3 mM added iron. In contrast, we observed that R. mucilaginosa was able to produce 140 mmol RA (g biomass)x1 (0.6¡0.01 mM RA) with 3.2 mM added iron and that RA production was not completely suppressed even with 26.6 mM added iron. Atkin et al. (1970) grew R. mucilaginosa on a medium which contained 1 g citric acid lx1. When 1.63 g citric acid lx1 was included in the medium used in this work, RA production was suppressed in the presence of only 2.55 mM added iron (Fig. 3). This demonstrates the importance of citric acid as an iron chelator for R. mucilaginosa and suggests that when looking for siderophore production from other organisms, citric acid should not be included in the medium. Although it is generally assumed that more siderophore will be produced at high than at low pH, because of the improved solubility of iron at low pH values, few determinations of the optimal pH for siderophore production have been published. Calvente et al. (2001) observed that more RA was produced at pH 8 than at pH 3 in batch flask cultures, in which it is, however, difficult to control pH. When the relationship between pH and RA was assessed in ironlimited chemostat cultures (Fig. 6), a pH optimum for RA production of 6.5 was observed, and production was higher at pH 7.6 compared to pH 3.0. Meiwes et al. (1990) found the optimum pH for desferrioxamine E production by Streptomyces olivaceus was 6.5 in batch culture, but Wiebe (2002) observed a much lower pH optimum of 4.7 for siderophore production by the filamentous fungus Fusarium venenatum. Glucose, maltose, glycerol or sucrose, with ammonium acetate, have been used as carbon sources for growing R. mucilaginosa and other yeasts for siderophore production and it has been suggested that siderophore production is poorer with glucose as carbon source (Atkin et al. 1970). Studies carried out with
955 the filamentous fungi Aspergillus melleus (Crueger & Za¨hner 1968) and Aspergillus viridi-nutans (Kappner et al. 1977) found that the same level of siderophore production occurred with either glucose or fructose as carbon sources and that siderophore production could be increased by growth on carbons such as acetate, succinate, asparagine, glutamate, ornithine, arginine (Crueger & Za¨hner 1968), mannose and sucrose (Kappner et al. 1977). No siderophore was produced in the presence of the amino acids aspartate, alanine, or serine, or in the presence of other organic acids (Crueger & Za¨hner 1968). As with the Aspergillus species (Crueger & Za¨hner 1968, Kappner et al. 1977), siderophore production by R. mucilaginosa was unaffected by the choice of either glucose or fructose as carbon source, but was significantly (P<0.05) improved when R. mucilaginosa was grown on sucrose (Table 1). When R. mucilaginosa was grown on a mixture of glucose and fructose (1 : 1), the concentration of RA produced was the same as on either glucose or fructose alone (results not shown), demonstrating that the high RA production was associated with invertase production. Sucrose was the only carbon source tested on which RA production was increased compared to production on glucose (Table 1). R. mucilaginosa did not grow well on acetate as the sole carbon source, but both growth rate and biomass yield improved when sucrose and acetate were both present in the medium (Table 2). Atkin et al. (1970) found that RA production was increased when the concentration of ammonium acetate in the medium was increased. The results presented here demonstrate that it is the addition of acetate, rather than ammonium, to the medium which increases RA production by R. mucilaginosa. Addition of acetate to glucose containing medium, similarly increased the production of siderophores by both A. melleus (Crueger & Za¨hner 1968) and A. viridi-nutans (Kappner et al. 1977). Ornithine is also a precursor of RA production and Calvente et al. (2001) recently demonstrated that addition of 1.69 g ornithine lx1 (10 mM) to sucrose medium increased production of RA by approximately 33 % in both R. rubra (synonymous with R. mucilaginosa) and R. glutinis, which is comparable to the 31 % increase in specific RA production observed here (Table 2) with addition of 1 g ornithine lx1. Crueger & Za¨hner (1968) observed increasing specific siderophore production by A. melleus with addition of up to 13 g ornithine lx1, but there was no further increase in specific RA production by R. mucilaginosa with addition of more than 1 g ornithine lx1 (Fig. 4). The effects of adding acetate and ornithine to the medium were additive (Table 2), as has been observed for A. melleus (Crueger & Za¨hner 1968). Based on the results of Crueger & Za¨hner (1968) and Calvente et al. (2001), it was expected that addition of glutamate to the medium would increase RA production, but that addition of aspartate would not. However, less RA was produced when either glutamate or
Siderophore production by Rhodotorula mucilaginosa aspartate were added to the medium than when only sucrose and acetate were present (Table 2). Differences in media composition may account for the difference in results. The data in Table 2 demonstrate that the increase in RA production with addition of an amino acid was specific for ornithine, the immediate precursor of RA. RA production was optimal in iron-restricted sucrose (15–22 mM) medium supplemented with NaAcetate (30–60 mM) and ornithine (3–6 mM). The medium should be buffered to pH 6.5, but should not contain significant concentrations of citric acid. RA could be produced in iron-limited chemostat cultures, but the range of growth rates which could be sustained were severely limited by the iron concentration. Iron-limited chemostats with 0.63 mM added iron and both sucrose and acetate present as carbon sources could not be maintained, but high concentrations of RA [3 mM RA or 1100 mmol RA (g biomass)x1] were produced while the culture washed out at D=0.006 hx1 (data not shown). This suggests that fed-batch cultures may be more suitable for high RA production than ironlimited chemostats. REFERENCES Anke, T. & Diekmann, H. (1972) Biosynthesis of sideramines in fungi. Rhodotorulic acid synthetase from extracts of Rhodotorula glutinis. FEBS Letters 27 : 259–262. Archibald, F. S. & DeVoe, I. W. (1978) Iron in Neisseria meningitidis: minimum requirements, effects of limitation, and characteristics of uptake. Journal of Bacteriology 136: 35–48. Atkin, C. L. & Neilands, J. B. (1968) Rhodotorulic acid, a diketopiperazine dihydroxamic acid with growth-factor activity. I. Isolation and characterisation. Biochemistry 7: 3734–3739. Atkin, C. L., Neilands, J. B. & Phaff, H. J. (1970) Rhodotorulic acid from species of Leucosporidium, Rhodosporidium, Rhodotorula, Sporidiobolus, and Sporobolomyces, and a new alanine-containing ferrichrome from Cryptococcus melibiosum. Journal of Bacteriology 103: 722–733. Bhosale, P. B. & Gadre, R. V. (2001a) Production of b-carotene by a mutant of Rhodotorula glutinis. Applied Microbiology and Biotechnology 55 : 423–427. Bhosale, P. & Gadre, R. V. (2001b) Optimisation of carotenoid production from hyper-producing Rhodotorula glutinis mutant 32 by a factorial approach. Letters in Applied Microbiology 33 : 12–16. Birch, J. R. & Pirt, S. J. (1970) Improvements in a chemically defined medium for the growth of mouse cells (strain LS) in suspension. Journal of Cell Science 7: 661–670. Button, D. K., Dunker, S. S. & Morse, M. L. (1973) Continuous culture of Rhodotorula rubra: kinetics of phosphate-arsenate uptake, inhibition, and phosphate-limited growth. Journal of Bacteriology 113: 599–611. Calvente, V., Benuzzi, D. & de Tosetti, M. I. S. (1999) Antagonistic action of siderophores from Rhodotorula glutinis upon the postharvest pathogen Penicillium expansum. International Biodeterioration and Biodegradation 43: 167–172. Calvente, V., de Orellano, M. E., Sansone, G., Benuzzi, D. & Sanz de Tosetti, M. I. (2001) Effect of nitrogen source and pH on siderophore production by Rhodotorula strains and their application to biocontrol of phytopathogenic moulds. Journal of Industrial Microbiology and Biotechnology 26: 226–229. Carrano, C. J. & Raymond, K. N. (1978) Coordination chemistry of microbial iron transport compounds: rhodotorulic acid and iron uptake in Rhodotorula pilimanae. Journal of Bacteriology 136: 69–74.
956 Crueger, W. & Za¨hner, H. (1968) Stoffwechselprodukte von Mikroorganismen 70. Mitteilung. U¨ber den Einfluss der Kohlenstoffquelle auf die Sideraminbildung von Aspergillus melleus Yukawa. Archiv fu¨r Mikrobiologie 63: 376–384. Esener, A. A., Roels, J. A., Kossen, N. W. F. & Roozenburg, J. W. H. (1981) Description of microbial growth behaviour during the wash-out phase; determination of the maximum specific growth rate. European Journal of Applied Microbiology and Biotechnology 13: 141–144. Ghosh, M. & Miller, M. J. (1995) Design, synthesis, and biological evaluation of isocyanurate-based antifungal and macrolide antibiotic conjugates : iron transport-mediated drug delivery. Bioorganic and Medicinal Chemistry 3: 1519–1525. Johnson, G. V., Lopez, A. & Foster, N. L. V. (2002) Reduction and transport of Fe from siderophores. Plant and Soil 241: 27–33. Kappner, M., Hasenbo¨hler, A. & Za¨hner, H. (1977) Stoffwechselprodukte von Mikroorganismen 166. Mitteilung. Optimierung der Desferri-Ferricrocinbildung bei Aspergillus viridi-nutans Ducker & Thrower. Archives of Microbiology 115: 323–331. Kurowski, W. M. & Pirt, S. J. (1971) The iron requirement of Agrobacterium tumefaciens for growth and 3-ketosucrose production. The removal of iron from solutions by seitz filters. Journal of General Microbiology 68 : 65–69. Liu, J.-S., Schill, N., van Gulik, W. M., Voisard, D., Marison, I. W. & van Stockar, U. (1999) The coupling between catabolism and anabolism of Methanobacterium thermoautotrophicum in H2- and iron-limited continuous cultures. Enzyme and Microbial Technology 25 : 784–794. Lu, Y. & Miller, M. J. (1999) Syntheses and studies of multiwarhead siderophore-5-fluorouridine conjugates. Bioorganic and Medicinal Chemistry 7: 3025–3038. Meiwes, J., Fiedler, H.-P., Za¨hner, H., Konetschny-Rapp, S. & Jung, G. (1990) Production of desferrioxamine E and new analogues by directed fermentation and feeding fermentation. Applied Microbiology and Biotechnology 32: 505–510. Miller, G. W. (1989) Treatment of plant chlorosis with rhodotorulic acid. US Patent US4872899. Oudin, P., Toth, J. A., Bonaly, R. & Balogh, C. (1999) Quantitation of intracellular polychlorobiphenyls during growth of Rhodotorula glutinis. Biotechnology Techniques 13: 101–105. Pirt, S. J. (1975) Principles of Microbe and Cell Cultivation. Blackwell, Oxford. Planalp, R. P., Przyborowska, A. M., Park, G., Ye, N., Lu, F. H., Rogers, R. D., Broker, G. A., Torti, S. V. & Brechbiel, M. W. (2002) Novel cytotoxic chelators that bind iron (II) selectively over zinc (II) under aqueous aerobic conditions. Biochemical Society Transactions 30: 758–762. Stephanopoulos, G. N., Aristidou, A. A. & Nielsen, J. (1998) Metabolic Engineering. Academic Press, London. Stradling, G. N. (1998) Decorporation of actinides: a review of recent research. Journal of Alloys and Compounds 271–273: 72–77. Tilbrook, G. S. & Hider, R. C. (1998) Iron chelators for clinical use. In Metal Ions in Biological Systems. Vol. 35. Iron Transport and Storage in Microorganisms, Plants, and Animals (A. Sigel & H. Sigel, eds): 691–730. Marcel Dekker, Basel. Vergne, A. F., Walz, A. J. & Miller, M. J. (2000) Iron chelators from mycobacteria (1954–1999) and potential therapeutic applications. Natural Product Reports 17 : 99–116. Vogel, H. J. (1956) A convenient growth medium for Neurospora (Medium N). Microbial Genetics Bulletin 243: 112–119. Weger, H. G. (1999) Ferric and cupric reductase activities in the green alga Chlamydomonas reinhardtii : experiments using iron-limited chemostats. Planta 207: 377–384. Wiebe, M. G. (2002) Siderophore production by Fusarium venenatum A3/5. Biochemical Society Transactions 30: 696–698. Winkelmann, G. (1985) Iron uptake systems in fungi. In Iron, Siderophores and Plant Diseases (T. R. Swinburne, ed.) : 7–14. Plenum Press, London. Corresponding Editor: N. P. Money
949
DOI: 10.1017/S0953756203008220 Printed in the United Kingdom.
Rhodotorulic acid production by Rhodotorula mucilaginosa
Ditte ANDERSEN1, Joanna C. RENSHAW2 and Marilyn G. WIEBE1* 1
Department of Life Sciences, Sohngaardsholmsvej 49, Aalborg University, DK-9000 Aalborg, Denmark. Centre for Radiochemistry Research, Department of Chemistry, University of Manchester, Manchester M13 9PL, UK. E-mail : [email protected] 2
Received 8 January 2003; accepted 20 June 2003.
Rhodotorula mucilaginosa produces the siderophore rhodotorulic acid (RA) when grown in iron-limited conditions. R. mucilaginosa grew at rates between 0.10 and 0.19 hx1 in iron-restricted conditions, depending on the carbon source, and at 0.23 hx1 in iron-sufficient conditions. In bioreactors inoculated with iron-starved pre-cultures, initial specific growth rates in batch culture were dependent on the iron concentration. The critical dilution rate (Dcrit, at which steady state cultures cannot be sustained) in continuous cultures was also dependent on the iron concentration and was lower than mmax in batch culture. Sucrose was the best carbon source for RA production [287¡11 mmol (g biomass)x1] and production could be further increased by supplementing the medium with the precursors acetate [460¡13 mmol (g biomass)x1], ornithine [376¡6 mmol (g biomass)x1], or both [539¡15 mmol (g biomass)x1]. Citric acid was an effective suppresser of RA production. RA was produced in a growth rate dependent manner and was optimally produced at pH 6.5.
INTRODUCTION The yeast Rhodotorula mucilaginosa produces the siderophore rhodotorulic acid (RA) when grown in ironrestricted conditions (Atkin, Neilands & Phaff 1970). Whilst most fungi make more than one siderophore (Winkelmann 1985), R. mucilaginosa produces only RA and is thus a good test organism for studying the influence of environmental parameters on siderophore production. In addition, siderophore production has recently been demonstrated to be important for both R. mucilaginosa (syn. R. rubra) and R. glutinis in their use as biocontrol agents for post-harvest fungal diseases of fruits (Calvente, Benuzzi & Tosetti 1999, Calvente et al. 2001) and RA can be used as an effective treatment of iron chlorosis in plants (Miller 1989, Johnson et al. 2002). Siderophores of other micro-organisms are of interest because of their potential use in treating medical conditions such as iron and aluminium overloading (Tilbrook & Hider 1998) and in the decorporation of toxic metals such as plutonium (Stradling 1998). The use of siderophore-drug conjugates for targeted drug delivery (Ghosh & Miller 1995, Lu & Miller 1999), siderophores as anti-tumour chemotherapeutic agents (Vergne, Walz & Miller 2000, Planalp et al. 2002), and * Corresponding author.
as chelators to reduce the toxicity of other chemotherapeutic treatments (Tilbrook & Hider 1998) is also being investigated. Establishing the factors which affect siderophore production in fungi will facilitate their isolation from a range of other fungal species. Although the production of RA by R. mucilaginosa and related species has been studied since 1968 (Atkin & Neilands 1968) and it can easily be obtained from cultures starved of iron, the role of parameters other than iron concentration has received little attention. Both carbon and nitrogen sources affect siderophore production in Aspergillus species (Crueger & Za¨hner 1968, Kappner, Hasenbo¨hler & Za¨hner 1977), and recently Calvente et al. (2001) observed that amino acids in the medium and pH affected RA production by Rhodotorula species. As with production of siderophores by other fungi, RA is produced both during exponential growth and during stationary phase (Atkin, Neilands & Phaff 1970) and the relationship between growth and siderophore production is unclear. This paper presents data on the growth of R. mucilaginosa under iron-restricted (i.e. iron is the first nutrient to restrict biomass production) batch and ironlimited (i.e. iron supply in the medium is limited to a sub-optimal concentration, while all other nutrients are provided in excess) chemostat cultures and investigates the importance of the carbon source and the precursors
Siderophore production by Rhodotorula mucilaginosa acetate and ornithine for RA synthesis. In addition, the relationships between growth, pH and RA production are established. MATERIALS AND METHODS Organism and media Rhodotorula mucilaginosa was obtained from the American Type Culture Collection (ATCC 26423) and was maintained as cells on agar-solidified (15 g agar lx1) medium at 4 xC or stored in glycerol (20 % w/v) at x70 x. The defined medium of Vogel (1956) was modified to give an iron-deficient medium with low buffering capacity and contained : 3.0 g (NH4)2SO 4 lx1, 0.3 g KH2PO4 lx1, 0.2 g MgSO4 . 7H2O lx1, 0.1 g CaCl2 . 2H2O lx1, 2.25 mg citric acid lx1 (to keep zinc and copper in solution during medium preparation), 2.25 mg ZnSO4 . 7H2O lx1, 0.11 mg CuSO4 . 5H2O lx1, 0.02 mg MnSO4 . 4H2O lx1, 0.02 mg H3BO3 lx1, 0.02 mg Na2MoO4 . 2H2O, 0.02 mg biotin lx1 and 2 mg thiamine lx1, with no Na3Citrate or Fe(NH4)2(SO4)2 . 6H2O. Iron was provided as FeCl3.6H2O (0–26.6 mM added Fe3+). Sucrose, glucose, fructose, NaAcetate, xylose, raffinose or maltose were used as carbon sources (10 g lx1) and some media were supplemented with ornithine (0.25–4 g lx1), glutamate (1 g lx1) or aspartate (1 g lx1). Media for batch cultures in 250 ml flasks were buffered with 0.5 M 2-(N-morpholino)ethanesulfonic acid (MES), pH 6.0. The defined medium of Vogel (1956) was used without modification, other than the substitution of (NH4)2SO 4 for NH4NO3 and addition of 2 mg thiamine lx1, for iron-sufficient medium. Iron-sufficient medium thus differed from iron-deficient medium in that it contained 1.05 mg Fe(NH4)2(SO4)2 . 6H2O lx1, with 2.5 g Na3Citrate lx1 and 5 g KH2PO 4 lx1. Cultural conditions Small scale batch cultures were grown in 250 ml flasks containing 50 ml medium. Flasks were incubated at 25 x on rotary shakers at 200 rpm (throw=2.5 cm). Bioreactor cultures were grown in a Braun Biostat M bioreactor (2.1 l full working volume). Cultures were maintained at 25 x and aerated at 0.8 l air [l culture]x1 minx1 with 500 rpm agitation (three 4-blade disk turbine impellers, 48 mm diam). pH was maintained at values between 3.0 and 7.6¡0.1, by addition of 0.5 M NaOH. Polypropylene glycol 2000 (0.05–0.4 % v/v, final concentration ; JT Baker) was added to prevent foaming. Measurements of growth and biomass Growth was measured as increase in optical density, measured at 550 nm (Cecil CE 7200 spectrophotometer), using water as the blank. Specific growth rates were also determined from the decrease in optical
950 density of bioreactor cultures with dilution rates set above the critical dilution rate (Dcrit, the dilution rate above which biomass is removed from the bioreactor at a faster rate than it can be replaced by growth ; Esener et al. 1981). For dry weight analyses, samples were collected on pre-dried, pre-weighed glass fibre filters (GA55 Advantec), rinsed with de-ionised water and dried at 90 x. Dry weight analyses were complicated by the presence of extracellular material (e.g. polysaccharides and polypropylene glycol) which blocked the filters and made it difficult to rinse the cells. The average relationship between dry weight and optical density was found to be OD at 550 nm=3.7rdry weight (g lx1), r=0.82, which was used to estimate biomass yields from flask cultures. Yield of biomass on carbon was expressed as Cmol biomass (Cmol carbon)x1, in order to readily compare yields on different carbon sources. Cmol biomass was taken as 25.1 Cmol (g biomass)x1, based on an average microbial cell composition of CH1.81O0.52N0.21 (Stephanopoulos, Aristidou & Nielsen 1998). Yield expressed as Cmol biomass (Cmol carbon)x1 is thus determined as the yield expressed as (g biomass) (g carbon)x1 divided by 25.1 g biomass (Cmol biomass)x1 and multiplied by the number of Cmol per g carbon for the specified carbon source [e.g. 30 Cmol (g glucose)x1]. This gives a small overestimate of the yield (Cmol Cmolx1), since ash content was not measured or taken into account. Rhodotorulic acid assay Rhodotorulic acid (RA) was quantified by measuring the absorbance of the ferric–RA complex at 475 nm (Atkin & Neilands 1968). There was no absorbance from other supernatant components at this wavelength. Cells were removed by centrifugation and supernatant (160 ml) was mixed with 840 ml FeCl3 . 6H2O (5 mM in 0.1 M HCl) and the absorbance measured in an Ultrospec 3000 spectrophotometer, with de-ionised water replacing the supernatant for the blank. HPLC analysis of carbohydrates Culture supernatant was filtered through 0.22 mm (pore diam) filters and samples were analysed on an Aminex HPX-87H HPLC organic acid analysis column (300r7.8 mm, Bio-Rad), with H2SO4 (5 mM, flow 0.4 ml minx1) as the mobile phase. Carbohydrates were detected with a Knaver refractometer. RESULTS Specific growth rate in iron-sufficient and iron-restricted media Rhodotorula mucilaginosa grew at 0.23 hx1 in ironsufficient medium with sucrose as the carbon source. The specific growth rate in iron-restricted medium was
D. Andersen, J. C. Renshaw and M. G. Wiebe
951 not significantly (P>0.05 hx1, Scheffe’s multiple range test) affected when sucrose medium was supplemented with NaAcetate, ornithine, glutamate or aspartate (Table 2). R. mucilaginosa ATCC 26423 was unable to utilise maltose.
Specific growth rate (h–1)
0.25
0.20
0.15
The effect of substrate concentration on biomass yield and rhodotorulic acid production
0.10
0.05
0.00 0.0
0.5
1.0
1.5
2.0
2.5
3.0
Added iron (M)
Fig. 1. The initial specific growth of Rhodotorula mucilaginosa in iron-restricted batch cultures (%) and the calculated specific growth rate during washout conditions in iron-limited continuous flow culture (&), for cultures growing at pH 6.0¡0.1, 25 x, 500 rpm, with various concentrations of added iron.
dependent on the inoculum, the iron concentration (Fig. 1) and the carbon source (Tables 1–2). Preliminary experiments were inoculated with yeast cells which had been grown on iron-sufficient agar-solidified medium, whereas subsequent experiments were inoculated with cells which had been grown in iron-restricted liquid medium. Even for cells grown on iron-sufficient medium, initial specific growth rates in iron-restricted (0.63 mM added Fe3+) medium with sucrose as the carbon source were reduced to 0.15¡0.004 hx1. When iron-starved inoculum was used, the initial specific growth rate with 0.63 mM added Fe3+ was 0.11 hx1, and specific growth rate increased with increasing iron concentration (Fig. 1). In iron-restricted medium, the initial exponential growth phase was maintained until approximately 30 % of the final OD had been attained, after which the cultures entered a prolonged deceleration phase of approximately 20 h. During decelerating growth, specific growth rates declined from approx. 0.08 to 0.015 hx1. Dcrit during continuous medium flow was considerably lower than mmax measured during batch growth in the bioreactor, but was similar to the measurement of specific growth rate during washout conditions (Fig. 1). Dcrit was lower when 0.63 mM added Fe3+ was supplied to the culture (Dcrit<0.017 hx1), than when 1.9 mM added Fe3+ was supplied (Dcrit<0.10 hx1). Initial specific growth rates in iron-restricted conditions (0.63 mM added Fe3+) were significantly (P< 0.05, Scheffe’s multiple range test) higher when glucose (m=0.19¡0.01 hx1) or fructose (m=0.20¡0.01 hx1) were provided as the carbon source, rather than sucrose (m=0.15¡0.01 hx1), but were significantly (P<0.05, Scheffe’s multiple range test) reduced when raffinose (m=0.12¡0.004 hx1), NaAcetate (0.10¡0.004 hx1) or xylose (m=0.06¡0.001 hx1) were provided as carbon sources (Table 1). The initial specific growth rate was
For glucose concentrations between 2.5 and 10 g lx1 (0.63 mM added Fe3+), Rhodotorula mucilaginosa had a yield of biomass on glucose of 0.56¡0.02 Cmol biomass (Cmol substrate)x1 [0.48 g biomass (g substrate)x1]. Biomass yield was slightly reduced with 15 and 20 g glucose lx1 [0.52¡0.03 and 0.48¡0.02 Cmol biomass (Cmol substrate)x1, respectively ; significant at P<0.15, Scheffe’s multiple range test, for Yx/C with 20 g glucose lx1]. Specific RA production increased with increasing glucose concentration between 2.5 [100 mmol RA (g biomass)x1] and 10 g glucose lx1 [170 mmol RA (g biomass)x1] and remained constant at concentrations above 10 g glucose lx1 (Fig. 2). The yield of RA on glucose was greatest with 10 g glucose lx1 (2.5 mmol RA Cmolx1 ; Fig. 2). The effect of iron concentration and the presence of citric acid on rhodotorulic acid production RA production decreased in iron-deficient medium with increasing concentration of added iron, but was still detectable with 26.6 mM added iron (Fig. 3). No significant differences were observed in the molar concentration of RA produced when 0–1.9 mM iron was added. However, when 2.55 mM iron was added in the presence of 1.63 g citric acid lx1, RA production [0.01 mM, or 0.03 mmol RA (Cmol sucrose)x1] was almost completely suppressed (Fig. 3). Biomass yield in the presence of 2.55 mM added iron and 1.63 g citric acid lx1 was similar to that observed with 26.6 mM added iron with only 0.0023 mg citric acid lx1. The effect of carbon source on rhodotorulic acid production More RA was produced when Rhodotorula mucilaginosa was grown on sucrose than on any other carbon source tested (Table 1). The yields of biomass on total carbon for R. mucilaginosa growing on medium supplemented with 0.63 mM Fe3+, with various carbon sources, are given in Table 1. The biomass yields on raffinose and sodium acetate were significantly (P<0.05, Scheffe’s multiple range test) reduced compared to the yields on other carbon sources. Biomass yields on sucrose were measured during chemostat cultures with 0.63 or 1.9 mM added iron. Residual carbon was measured and the yield of biomass on consumed carbon was 0.48 and 0.58 Cmol
Siderophore production by Rhodotorula mucilaginosa
952
Table 1. Specific growth rates, yield coefficients (biomass on carbon, Yx/C, and RA on carbon, YRA/C), and specific RA production [mmol RA (g biomass)x1] for Rhodotorula mucilaginosa growing in iron-restricted, modified Vogel’s medium (0.63 mM Fe3+, pH 6) on various carbon sources (10 g lx1). Cultures were grown for at least 90 h in 250 ml flasks shaken at 200 rpm at 25 x. Yield coefficients were estimated from OD measurements (OD550=3.67rDW g lx1). Values in the same column with the same superscript, a to e, did not differ significantly (P>0.05, Scheffe’s multiple range test).
Carbon source
Specific growth rate (hx1)a
Yield (Yx/C) of biomass on carbon (Cmol Cmolx1)b
Specific RA production (mmol RA gx1)
RA (YRA/C, mmol Cmolx1)c
Sucrose Glucose Fructose Raffinose Xylose NaAcetate
0.15¡0.004c 0.19¡0.003d 0.20¡0.008d 0.12¡0.004b 0.06¡0.001a 0.10¡0.004b
0.56¡0.003b 0.58¡0.016bc 0.64¡0.005c 0.24¡0.002a 0.53¡0.003bc 0.30¡0.003a
287¡11 181¡7 184¡4 170¡3 111¡4 221¡9
4.0¡0.1d 2.6¡0.04c 3.0¡0.04c 1.0¡0.01a 1.5¡0.05ab 1.7¡0.05b
a b c
n=3–9 replicates. n=3–42 replicates. n=3–46 replicates.
Table 2. Specific growth rates, yield coefficients (biomass on carbon, Yx/C, and RA on total carbon, YRA/C), and specific RA production [mmol (g biomass)x1] for Rhodotorula mucilaginosa growing in iron-restricted, modified Vogel’s medium (0.63 mM Fe3+, pH 6) with sucrose as the primary carbon source, and supplemented with acetate, ornithine, glutamic acid or aspartic acid. Cultures were grown for at least 90 h in 250 ml flasks shaken at 200 rpm at 25 x. Yield coefficients were estimated from OD measurements (OD550=3.67rDW g lx1). Values in the same column with the same superscript, a to f, did not differ significantly (P>0.05, Scheffe’s multiple range test).
Sucrose (Cmol)
Supplement (Cmol)
Specific growth rate (hx1)a
Yx/C (Cmol Cmolx1)b
Specific RA production (mmol RA gx1)
RA (YRA/C, mmol Cmolx1)c
0.35 0 0.09 0.18 0.26 0.35 0.26
– (0) Acetate (0.24) Acetate (0.18) Acetate (0.12) Acetate (0.06) Ornithine (0.04) Ornithine (0.04) Acetate (0.06) Glutamate (0.03) Acetate (0.06) Aspartate (0.03) Acetate (0.06)
0.15¡0.004bc 0.10¡0.004a 0.16¡0.014bc 0.16¡0.004bc 0.17¡0.0054c 0.14¡0.001ab 0.14¡0.002b
0.56¡0.003e 0.30¡0.003a 0.45¡0.005b 0.49¡0.005bc 0.53¡0.005d 0.50¡0.004c 0.48¡0.004bc
287¡11 221¡9 287¡6 460¡13 428¡15 376¡6 539¡15
4.0¡0.13c 1.7¡0.05a 3.2¡0.03b 5.6¡0.11d 5.8¡0.15d 4.7¡0.03cd 6.6¡0.12e
0.16¡0.004bc
0.56¡0.006de
323¡13
4.5¡0.13c
0.17¡0.003c
0.60¡0.005f
217¡7
3.3¡0.08b
0.26 0.26 a b c
n=3–12 replicates. n=9–28 replicates. n=6–30 replicates.
biomass (Cmol substrate)x1 with 0.63 and 1.9 mM added iron, respectively. Biomass yield was constant with 1.9 mM added iron for dilution rates (D) between 0.034 and 0.74 hx1. With 0.63 mM added iron, biomass yield was only measured at D=0.013 hx1, since the culture washed out at higher dilution rates. The effect of supplementation with acetate, ornithine, aspartate or glutamate Specific growth rate and biomass yield were low when acetate was provided as the sole carbon source (Tables 1–2), but when both acetate and sucrose were provided, in varying proportions, yields improved and Rhodotorula mucilaginosa grew at the same (P>0.05) specific growth rate as when only sucrose was supplied (Table 2). When 20 or 40 % of the carbon was provided as acetate, RA production was significantly (P<0.05, Scheffe’s multiple range test) increased, compared to
that observed when R. mucilaginosa was grown on sucrose alone (Table 2). When medium containing 10 g sucrose lx1 was supplemented with 1 g ornithine lx1, the yield of biomass on carbon [0.50¡0.004 Cmol biomass (Cmol carbon)x1] was reduced and the production of RA increased [376¡6 mmol RA (g biomass)x1 and 4.7¡0.03 mmol RA (Cmol substrate)x1], compared to growth in medium containing only sucrose (Table 2). R. mucilaginosa was also grown on medium containing 7.5 g sucrose lx1 and 2.5 g NaAcetate lx1, supplemented with 1 g ornithine lx1. The biomass yield on carbon [0.48¡0.004 Cmol biomass (Cmol carbon)x1] was reduced, compared to the same medium lacking ornithine [0.53¡0.005 Cmol biomass (Cmol carbon)x1], but the production of RA [539¡15 mmol RA (g biomass)x1 and 6.6¡0.12 mmol RA (Cmol substrate)x1] was significantly (P<0.05) increased (Table 2).
D. Andersen, J. C. Renshaw and M. G. Wiebe
953
200
3
150
600
10
cd c
ab
2
100
a 1
50
Specific RA prouction ( , M RA [g biomass]–1)
d
Specific RA production ( , mol [g biomass]–1)
Yield of RA on glucose ( , mmol RA Cmol–1)
500
8 d d
400
cd bc 6
b a
300
4 200 2
100 0 0
5
10
Yield of RA on carbon ( , mmol RA Cmol–1)
4
0 20
15
0
Glucose (g l–1)
0 0
Fig. 2. Yield of RA on glucose (%) and on biomass (&, specific RA production) for cultures of Rhodotorula mucilaginosa growing in iron-restricted Vogel’s medium (0.63 mM added iron, buffered at pH 6) containing different concentrations of glucose. Error bars represent ¡SEM and where not shown were smaller than the size of the symbol. Yield (of RA on glucose) values with different letters, a to d, were significantly (P<0.06, Scheffe’s multiple range test) different. 30 e
1
2 Ornithine (g l–1)
3
4
Fig. 4. The effect of ornithine concentration on the yield of RA on biomass (&, specific RA production) and on total carbon (%) for cultures of Rhodotorula mucilaginosa growing in ironrestricted Vogel’s medium (%, 0.63 mM added iron, 7.5 g sucrose lx1, 2.5 g NaAcetate lx1, buffered at pH 6). Where shown, error bars represent ¡SEM. Where not shown, error bars were smaller than the size of the symbol. Data points with different letters, a to d, were significantly (P<0.05, Scheffe’s multiple range test) different.
Yield of RA on glucose (YRA/C , , mmol Cmol–1)
e de
increased, but the production of RA was either unaffected (glutamate) or reduced (aspartate ; Table 2).
d
20
The relationship between growth and siderophore production
c 10
The specific rate of RA production increased with increase in dilution rate at dilution rates between 0.034 and 0.074 hx1, with 1.9 mM added iron, at pH 6.0 (Fig. 5). Biomass concentration decreased with increase in dilution rate.
b ab
ab a
0 0
10
20 Added Fe3+ (M)
30
Fig. 3. Yield of RA on glucose for cultures of Rhodotorula mucilaginosa growing in iron-restricted Vogel’s medium (%, 10 g glucose lx1, buffered at pH 6) containing different concentrations of added iron. The yield of RA on glucose for cultures of R. mucilaginosa growing in iron-sufficient complete Vogel’s medium (#) is also shown. Where shown, error bars represent ¡SEM. Where not seen, error bars were less than the size of the symbol. Data points with different letters, a to e, were significantly (P<0.05, Scheffe’s multiple range test) different.
Specific RA production increased with increasing concentrations of added ornithine, for concentrations of up to 1 g ornithine lx1 [539¡15 mmol RA (g biomass)x1 ; Fig. 4]. The yield of RA on carbon was highest with 0.25 or 0.5 g ornithine lx1 [7.0¡ 0.04 mmol RA (Cmol substrate)x1 ; Fig. 4]. When sucrose (7.5 g lx1) and NaAcetate (2.5 g lx1) medium was supplemented with either glutamate or aspartate (1 g lx1), the biomass yield on carbon was
The relationship between pH and siderophore production The relationship between pH and RA production was determined for cultures with 0.95 mM added iron at D=0.046 hx1. RA production increased with increase in pH for pH values between 3 and 6.5, with maximum production observed at pH 6.5 (Fig. 6). Above pH 6.5 RA production was reduced by increase in pH. Rhodoorula mucilaginosa could not sustain a growth rate of 0.045 hx1 at pH 8 under these conditions. DISCUSSION Rhodotorula mucilaginosa had a specific growth rate of 0.23 hx1 at 25 x in iron-sufficient, defined medium, which is comparable to the initial specific growth rate (estimated as 0.22 hx1) of R. glutinis in glucose-yeast extract medium at 24–28 x (Oudin et al. 1999, Bhosale & Gadre 2001a, b) and of R. mucilaginosa grown in phosphate-restricted defined medium at 25 x
Siderophore production by Rhodotorula mucilaginosa
10
0.5
5 0.0
0
5
6
2
4
1
2
0.02
0.04 0.06 0.08 Dilution rate (h–1)
de
200 cd bc
150 b
100
a 0
8
3
de
50
10
4
e 250 Specific RA production (mol RA [g biomass]–1)
15
Specific RA production rate ( , mol RA [g biomass]–1 h–1)
1.0
Residual carbon ( , g l–1)
Rhodotorulic acid ( , mM)
20
Biomass ( , g l–1)
300
25
1.5
0 0.00
954
0 0.10
Fig. 5. The effect of dilution rate on biomass concentration ($, g lx1), residual sugar concentration (#, g lx1), RA production (&, mM) and the specific RA production rate [%, mmol RA (g biomass)x1 hx1] for Rhodotorula mucilaginosa growing in iron-limited (1.9 mM added iron) chemostat culture with sucrose (10 g lx1) as the carbon source. Cultures were maintained at pH 6.0¡0.1, 25 x, 500 rpm, 0.7 l air (l culture)x1 minx1. Data are the averages of at least three consecutive samples taken during steady-state-conditions, at least three residence times after the condition had been established. Error bars represent ¡SEM, and where not seen were smaller than the size of the symbol.
(m=0.17 hx1 ; Button, Dunker & Morse 1973). In ironrestricted medium the specific growth rate was reduced, dependent on the concentration of iron in the medium, as has also been observed in iron-restricted cultures of Neisseria meningitidis (Archibald & DeVoe 1978), Agrobacterium tumefaciens (Kurowski & Pirt 1971) and mouse LS cells (Birch & Pirt 1970). Carry-over of iron in the inoculum could affect specific growth rates. In a chemostat, biomass is expected to be constant over a wide range of dilution rates, being significantly reduced only as the dilution rate approaches Dcrit, which should be very near the mmax of the organism (Pirt 1975). However, in an iron-limited chemostat the steady-state biomass may decrease with increase in dilution rate for dilution rates below Dcrit (Pirt 1975, Weger 1999). This decrease in biomass with increasing dilution rate was observed for R. mucilaginosa (Fig. 5). Dcrit was also reduced in iron-limited chemostats and was dependent on the iron concentration (Fig. 1). Reduced Dcrit values in iron-limited chemostats have previously been reported for the filamentous fungus Fusarium venenatum (Wiebe 2002) and the bacterium Methanobacterium thermoautotrophicum (Liu et al.
3
4
5
6
7
8
pH
Fig. 6. The effect of pH on RA production for Rhodotorula mucilaginosa growing in iron-limited (0.95 mM added iron) chemostat culture with sucrose (10 g lx1) as the carbon source. Cultures were maintained at 25 x, 500 rpm, 0.7 l air (l culture)x1 minx1. Data are the averages of at least three consecutive samples taken during steady state-conditions, at least three residence times after the condition had been established. Data points with different letters, a to e, were significantly (P<0.05, Scheffe’s multiple range test) different. Error bars represent ¡SEM, and where not seen were smaller than the size of the symbol.
1999), but a decrease in Dcrit with decreasing iron concentration has not previously been reported. Because iron concentration affected Dcrit, the relationship between growth rate and RA production could only be assessed at the relatively high iron concentration of 1.9 mM added iron. RA was produced in a growth rate dependent manner, with the highest specific production rate being observed at the highest specific growth rate (Fig. 5). This agrees with the high concentrations of RA observed during exponential growth of batch cultures and the production of RA synthetase during the early log phase in batch cultures (Anke & Diekmann 1972), but does not explain the increasing concentrations of RA observed during stationary phase (Atkin et al. 1970). Production of siderophores by F. venenatum was also growth rate correlated (Wiebe 2002). Atkin & Neilands (1968) observed that production of RA increased from approx. 2 g lx1 to 3–4 g lx1, when the carbon concentration in the medium was increased from 10 to 20 g lx1 and other researchers (Atkin et al. 1970, Anke & Diekmann 1972, Carrano & Raymond 1978, Calvente et al. 2001) have frequently used media containing 20–40 g carbohydrate lx1 to produce RA. However, high carbon concentrations are also likely to lead to oxygen limited conditions. When glucose concentration was increased from 2.5 to 10.0 g glucose lx1 (0.63 mM added Fe3+), the biomass yield [0.57 Cmol biomass (Cmol glucose)x1] remained constant and specific RA production increased (Fig. 2). This may reflect a reduction in availability of iron with increase in glucose concentration, since the concentration of
D. Andersen, J. C. Renshaw and M. G. Wiebe added iron was constant but the concentration of biomass produced increased, and therefore the concentration of iron available per g biomass decreased. For concentrations of glucose above 10 g glucose lx1, the yield of both biomass (Yx/C was reduced 14 % for 20 g glucose lx1 compared to 10 g glucose lx1 ; P<0.15) and RA (YRA/C was reduced 17 % for 20 g glucose lx1 compared to 10 g glucose lx1 ; P<0.06) on carbon were reduced. The reduced biomass yield suggests the occurrence of oxygen limited conditions, since biomass yield was not affected by the concentration of added iron for concentrations between 0.16 and 19 mM added iron (data not shown). Atkin et al. (1970) reported that RA production was sensitive to oxygen limitation, and the reduction in yield of both biomass and RA on carbon at high glucose concentrations indicates that high carbon concentrations (e.g. >20 g carbohydrate lx1) should only be used for RA production when adequate aeration can be provided. Atkin et al. (1970) observed decreasing RA production by R. mucilaginosa for iron concentrations between 0.18 and 1.8 mM added iron and complete suppression of RA with only 3 mM added iron. In contrast, we observed that R. mucilaginosa was able to produce 140 mmol RA (g biomass)x1 (0.6¡0.01 mM RA) with 3.2 mM added iron and that RA production was not completely suppressed even with 26.6 mM added iron. Atkin et al. (1970) grew R. mucilaginosa on a medium which contained 1 g citric acid lx1. When 1.63 g citric acid lx1 was included in the medium used in this work, RA production was suppressed in the presence of only 2.55 mM added iron (Fig. 3). This demonstrates the importance of citric acid as an iron chelator for R. mucilaginosa and suggests that when looking for siderophore production from other organisms, citric acid should not be included in the medium. Although it is generally assumed that more siderophore will be produced at high than at low pH, because of the improved solubility of iron at low pH values, few determinations of the optimal pH for siderophore production have been published. Calvente et al. (2001) observed that more RA was produced at pH 8 than at pH 3 in batch flask cultures, in which it is, however, difficult to control pH. When the relationship between pH and RA was assessed in ironlimited chemostat cultures (Fig. 6), a pH optimum for RA production of 6.5 was observed, and production was higher at pH 7.6 compared to pH 3.0. Meiwes et al. (1990) found the optimum pH for desferrioxamine E production by Streptomyces olivaceus was 6.5 in batch culture, but Wiebe (2002) observed a much lower pH optimum of 4.7 for siderophore production by the filamentous fungus Fusarium venenatum. Glucose, maltose, glycerol or sucrose, with ammonium acetate, have been used as carbon sources for growing R. mucilaginosa and other yeasts for siderophore production and it has been suggested that siderophore production is poorer with glucose as carbon source (Atkin et al. 1970). Studies carried out with
955 the filamentous fungi Aspergillus melleus (Crueger & Za¨hner 1968) and Aspergillus viridi-nutans (Kappner et al. 1977) found that the same level of siderophore production occurred with either glucose or fructose as carbon sources and that siderophore production could be increased by growth on carbons such as acetate, succinate, asparagine, glutamate, ornithine, arginine (Crueger & Za¨hner 1968), mannose and sucrose (Kappner et al. 1977). No siderophore was produced in the presence of the amino acids aspartate, alanine, or serine, or in the presence of other organic acids (Crueger & Za¨hner 1968). As with the Aspergillus species (Crueger & Za¨hner 1968, Kappner et al. 1977), siderophore production by R. mucilaginosa was unaffected by the choice of either glucose or fructose as carbon source, but was significantly (P<0.05) improved when R. mucilaginosa was grown on sucrose (Table 1). When R. mucilaginosa was grown on a mixture of glucose and fructose (1 : 1), the concentration of RA produced was the same as on either glucose or fructose alone (results not shown), demonstrating that the high RA production was associated with invertase production. Sucrose was the only carbon source tested on which RA production was increased compared to production on glucose (Table 1). R. mucilaginosa did not grow well on acetate as the sole carbon source, but both growth rate and biomass yield improved when sucrose and acetate were both present in the medium (Table 2). Atkin et al. (1970) found that RA production was increased when the concentration of ammonium acetate in the medium was increased. The results presented here demonstrate that it is the addition of acetate, rather than ammonium, to the medium which increases RA production by R. mucilaginosa. Addition of acetate to glucose containing medium, similarly increased the production of siderophores by both A. melleus (Crueger & Za¨hner 1968) and A. viridi-nutans (Kappner et al. 1977). Ornithine is also a precursor of RA production and Calvente et al. (2001) recently demonstrated that addition of 1.69 g ornithine lx1 (10 mM) to sucrose medium increased production of RA by approximately 33 % in both R. rubra (synonymous with R. mucilaginosa) and R. glutinis, which is comparable to the 31 % increase in specific RA production observed here (Table 2) with addition of 1 g ornithine lx1. Crueger & Za¨hner (1968) observed increasing specific siderophore production by A. melleus with addition of up to 13 g ornithine lx1, but there was no further increase in specific RA production by R. mucilaginosa with addition of more than 1 g ornithine lx1 (Fig. 4). The effects of adding acetate and ornithine to the medium were additive (Table 2), as has been observed for A. melleus (Crueger & Za¨hner 1968). Based on the results of Crueger & Za¨hner (1968) and Calvente et al. (2001), it was expected that addition of glutamate to the medium would increase RA production, but that addition of aspartate would not. However, less RA was produced when either glutamate or
Siderophore production by Rhodotorula mucilaginosa aspartate were added to the medium than when only sucrose and acetate were present (Table 2). Differences in media composition may account for the difference in results. The data in Table 2 demonstrate that the increase in RA production with addition of an amino acid was specific for ornithine, the immediate precursor of RA. RA production was optimal in iron-restricted sucrose (15–22 mM) medium supplemented with NaAcetate (30–60 mM) and ornithine (3–6 mM). The medium should be buffered to pH 6.5, but should not contain significant concentrations of citric acid. RA could be produced in iron-limited chemostat cultures, but the range of growth rates which could be sustained were severely limited by the iron concentration. Iron-limited chemostats with 0.63 mM added iron and both sucrose and acetate present as carbon sources could not be maintained, but high concentrations of RA [3 mM RA or 1100 mmol RA (g biomass)x1] were produced while the culture washed out at D=0.006 hx1 (data not shown). This suggests that fed-batch cultures may be more suitable for high RA production than ironlimited chemostats. REFERENCES Anke, T. & Diekmann, H. (1972) Biosynthesis of sideramines in fungi. Rhodotorulic acid synthetase from extracts of Rhodotorula glutinis. FEBS Letters 27 : 259–262. Archibald, F. S. & DeVoe, I. W. (1978) Iron in Neisseria meningitidis: minimum requirements, effects of limitation, and characteristics of uptake. Journal of Bacteriology 136: 35–48. Atkin, C. L. & Neilands, J. B. (1968) Rhodotorulic acid, a diketopiperazine dihydroxamic acid with growth-factor activity. I. Isolation and characterisation. Biochemistry 7: 3734–3739. Atkin, C. L., Neilands, J. B. & Phaff, H. J. (1970) Rhodotorulic acid from species of Leucosporidium, Rhodosporidium, Rhodotorula, Sporidiobolus, and Sporobolomyces, and a new alanine-containing ferrichrome from Cryptococcus melibiosum. Journal of Bacteriology 103: 722–733. Bhosale, P. B. & Gadre, R. V. (2001a) Production of b-carotene by a mutant of Rhodotorula glutinis. Applied Microbiology and Biotechnology 55 : 423–427. Bhosale, P. & Gadre, R. V. (2001b) Optimisation of carotenoid production from hyper-producing Rhodotorula glutinis mutant 32 by a factorial approach. Letters in Applied Microbiology 33 : 12–16. Birch, J. R. & Pirt, S. J. (1970) Improvements in a chemically defined medium for the growth of mouse cells (strain LS) in suspension. Journal of Cell Science 7: 661–670. Button, D. K., Dunker, S. S. & Morse, M. L. (1973) Continuous culture of Rhodotorula rubra: kinetics of phosphate-arsenate uptake, inhibition, and phosphate-limited growth. Journal of Bacteriology 113: 599–611. Calvente, V., Benuzzi, D. & de Tosetti, M. I. S. (1999) Antagonistic action of siderophores from Rhodotorula glutinis upon the postharvest pathogen Penicillium expansum. International Biodeterioration and Biodegradation 43: 167–172. Calvente, V., de Orellano, M. E., Sansone, G., Benuzzi, D. & Sanz de Tosetti, M. I. (2001) Effect of nitrogen source and pH on siderophore production by Rhodotorula strains and their application to biocontrol of phytopathogenic moulds. Journal of Industrial Microbiology and Biotechnology 26: 226–229. Carrano, C. J. & Raymond, K. N. (1978) Coordination chemistry of microbial iron transport compounds: rhodotorulic acid and iron uptake in Rhodotorula pilimanae. Journal of Bacteriology 136: 69–74.
956 Crueger, W. & Za¨hner, H. (1968) Stoffwechselprodukte von Mikroorganismen 70. Mitteilung. U¨ber den Einfluss der Kohlenstoffquelle auf die Sideraminbildung von Aspergillus melleus Yukawa. Archiv fu¨r Mikrobiologie 63: 376–384. Esener, A. A., Roels, J. A., Kossen, N. W. F. & Roozenburg, J. W. H. (1981) Description of microbial growth behaviour during the wash-out phase; determination of the maximum specific growth rate. European Journal of Applied Microbiology and Biotechnology 13: 141–144. Ghosh, M. & Miller, M. J. (1995) Design, synthesis, and biological evaluation of isocyanurate-based antifungal and macrolide antibiotic conjugates : iron transport-mediated drug delivery. Bioorganic and Medicinal Chemistry 3: 1519–1525. Johnson, G. V., Lopez, A. & Foster, N. L. V. (2002) Reduction and transport of Fe from siderophores. Plant and Soil 241: 27–33. Kappner, M., Hasenbo¨hler, A. & Za¨hner, H. (1977) Stoffwechselprodukte von Mikroorganismen 166. Mitteilung. Optimierung der Desferri-Ferricrocinbildung bei Aspergillus viridi-nutans Ducker & Thrower. Archives of Microbiology 115: 323–331. Kurowski, W. M. & Pirt, S. J. (1971) The iron requirement of Agrobacterium tumefaciens for growth and 3-ketosucrose production. The removal of iron from solutions by seitz filters. Journal of General Microbiology 68 : 65–69. Liu, J.-S., Schill, N., van Gulik, W. M., Voisard, D., Marison, I. W. & van Stockar, U. (1999) The coupling between catabolism and anabolism of Methanobacterium thermoautotrophicum in H2- and iron-limited continuous cultures. Enzyme and Microbial Technology 25 : 784–794. Lu, Y. & Miller, M. J. (1999) Syntheses and studies of multiwarhead siderophore-5-fluorouridine conjugates. Bioorganic and Medicinal Chemistry 7: 3025–3038. Meiwes, J., Fiedler, H.-P., Za¨hner, H., Konetschny-Rapp, S. & Jung, G. (1990) Production of desferrioxamine E and new analogues by directed fermentation and feeding fermentation. Applied Microbiology and Biotechnology 32: 505–510. Miller, G. W. (1989) Treatment of plant chlorosis with rhodotorulic acid. US Patent US4872899. Oudin, P., Toth, J. A., Bonaly, R. & Balogh, C. (1999) Quantitation of intracellular polychlorobiphenyls during growth of Rhodotorula glutinis. Biotechnology Techniques 13: 101–105. Pirt, S. J. (1975) Principles of Microbe and Cell Cultivation. Blackwell, Oxford. Planalp, R. P., Przyborowska, A. M., Park, G., Ye, N., Lu, F. H., Rogers, R. D., Broker, G. A., Torti, S. V. & Brechbiel, M. W. (2002) Novel cytotoxic chelators that bind iron (II) selectively over zinc (II) under aqueous aerobic conditions. Biochemical Society Transactions 30: 758–762. Stephanopoulos, G. N., Aristidou, A. A. & Nielsen, J. (1998) Metabolic Engineering. Academic Press, London. Stradling, G. N. (1998) Decorporation of actinides: a review of recent research. Journal of Alloys and Compounds 271–273: 72–77. Tilbrook, G. S. & Hider, R. C. (1998) Iron chelators for clinical use. In Metal Ions in Biological Systems. Vol. 35. Iron Transport and Storage in Microorganisms, Plants, and Animals (A. Sigel & H. Sigel, eds): 691–730. Marcel Dekker, Basel. Vergne, A. F., Walz, A. J. & Miller, M. J. (2000) Iron chelators from mycobacteria (1954–1999) and potential therapeutic applications. Natural Product Reports 17 : 99–116. Vogel, H. J. (1956) A convenient growth medium for Neurospora (Medium N). Microbial Genetics Bulletin 243: 112–119. Weger, H. G. (1999) Ferric and cupric reductase activities in the green alga Chlamydomonas reinhardtii : experiments using iron-limited chemostats. Planta 207: 377–384. Wiebe, M. G. (2002) Siderophore production by Fusarium venenatum A3/5. Biochemical Society Transactions 30: 696–698. Winkelmann, G. (1985) Iron uptake systems in fungi. In Iron, Siderophores and Plant Diseases (T. R. Swinburne, ed.) : 7–14. Plenum Press, London. Corresponding Editor: N. P. Money