INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

In filamentous fungi many secondary metabolites with complex chemical structure are synthesized via the polyketide pathway (15, 33, 37). Lovastatin, monacolin J, monacolin L, and mevastatin can be produced by Monascus ruber (7), Penicillium brevicompactum, and Aspergillus terreus (1, 36). Lovastatin is an inhibitor of the enzyme hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase (mevalonate:NADP⁺ oxidoreductase [EC 1.1.1.34]) that catalyzes the reduction of HMG-CoA to mevalonate during synthesis of cholesterol (14, 23, 36). The biosynthetic pathway of lovastatin in A. terreus has been investigated by nuclear magnetic resonance and mass spectroscopy (5, 26, 38). These studies concluded that lovastatin is composed of two distinct polyketide chains joined through an ester linkage. Proof that these two polyketides are assembled by two discrete polyketide synthases came from the cloning and partial characterization of the lovastatin biosynthetic gene cluster from A. terreus (16, 18).

Despite the knowledge of the genes and the enzymes involved in the biosynthetic pathway, little is known about the regulation and the physiology of lovastatin biosynthesis. Reported growth and production conditions for lovastatin are from batch fermentations performed on media with glucose and a complex nitrogen source (1, 4, 13, 22, 28). Both carbon and nitrogen sources are thought to exert complex regulation on gene expression and enzyme activities for polyketide synthesis, possibly at the level of catabolite repression or signaling due to limitation in growth rate or substrate limitation.

The objective of this work was to investigate the influence of glucose and nitrogen source on the physiology of A. terreus and lovastatin production. Several carbon and nitrogen sources were tested in order to develop a chemically defined medium. The aim was to monitor biomass formation and lovastatin production in relation to the consumption pattern of carbon and nitrogen sources. We show here that although growth occurred on a large variety of substrates, the choice of nitrogen source had a major impact on lovastatin production. In addition, lovastatin biosynthesis was initiated only when assimilation of the carbon source was limited.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Microorganism and growth conditions. Stock culture of A. terreus Thom ATCC 74135 was maintained on potato dextrose agar (Oxoid). The slants were stored at 4°C after growth on potato dextrose agar for 8 days at 28°C. Conidiospores were harvested with sterile solution (0.9% NaCl, 0.05% Tween 80), washed twice with sterile buffer (20 mM KH₂PO₄, adjusted to pH 2.0 with HCl), and enumerated.

Culture on complex medium. A 250-ml Erlenmeyer flask containing 40 ml of medium A [10 g of glucose, 5 g of corn steep liquor (Sigma catalog no. C-4648), 40 g of tomato paste, 10 g of oatmeal, and 10 ml of trace elements1 g of FeSO₄ · 7H₂O, 1 g of MnSO₄ · 4H₂O, 200 mg of ZnSO₄ · 7H₂O, 100 mg of CaCl₂ · 2H₂O, 25 mg of CuCl₂ · 2H₂O, 56 mg of H₃BO₃, and 19 mg of (NH₄)₆Mo₇O₂₄ · 4H₂Oper liter of solution (1)] was inoculated with 2 · 10⁶ conidiospores. The flasks were shaken at 220 rpm for 1 day at 28°C. A second culture was prepared by inoculating 200 ml of medium B (containing [per liter] 45 g of glucose or lactose, 24 g of peptonized milk [Oxoid catalog no. LP032J], 2.5 g of yeast extract [Difco catalog no. 0127-17], 2.5 g of polyethylene glycol P₂₀₀₀ [Fluka Chemika catalog no. 81221] [1]) with 6 ml of the previous culture in a 1-liter Erlenmeyer flask. The flasks were shaken at 200 rpm and incubated at 28°C for 12 days. For cultivations in fermentors, a 1-liter Erlenmeyer flask containing 200 ml of medium A was inoculated with 4 ml (10⁷) of conidiospore suspension. The flask was shaken at 220 rpm for 1 day and then transferred to a fermentor. Batch cultivations were performed in duplicate at 28°C in a 7- or 15-liter bioreactor (New MBR, Zurich, Switzerland) with a culture volume of 4 or 6 liters of medium B, respectively. The reactor was inoculated with 3% inoculum. An overpressure of 0.3 · 10⁵ Pa was applied to the reactor. Dissolved oxygen was continuously monitored with an oxygen probe (Mettler Toledo, Greifensee, Switzerland). The stirrer speed was kept at 400 rpm, and the airflow rate was controlled at 0.4 or 1 vol/vol/min (vvm) use of a mass flow meter (Bronkhorst, Ruurlo, The Netherlands).

Culture on synthetic medium. The chemically defined fermentation medium developed in this study contained (per liter of distilled water) the following: 45 g of glucose (Fluka Chemika catalog no. 49159), 12.5 g of mono-hydrate sodium glutamate (Merck catalog no. 6445), 5 g of KH₂PO₄ (Merck catalog no. 104873), 5 g of K₂HPO₄ (Merck catalog no. 105101), 0.2 g of FeSO₄ · 7H₂O (Merck catalog no. 3965), 0.1 g of MnSO₄ · 4H₂O (Merck catalog no. 102786), 0.2 g of ZnSO₄ · 7H₂O (Merck catalog no. 8883), 0.1 g of MgSO₄ · 7H₂O (Fluka Chemika catalog no. 63138), 20 mg of CaCl₂ · 2H₂O (Fluka Chemika catalog no. 21097), 5 mg of CuCl₂ · 2H₂O (Merck catalog no. 2791), 11 mg of H₃BO₃ (Merck catalog no. 100165), and 5 mg of (NH₄)₆Mo₇O₂₄ · 4H₂O (Merck catalog no. 101182). The pH of the medium was adjusted to 6.5 with HCl (2 N) or KOH (2 N). A suspension of 10⁷ conidiospores was used to inoculate a 1-liter Erlenmeyer flask containing 200 ml of medium with 10 g of glucose liter¹. The inoculum was agitated at 200 rpm and incubated at 28°C for 1 day and then transferred to a fermentor containing 4.5 liters of synthetic medium. The cultivations were performed at a temperature of 28°C, an agitation rate of 400 rpm, and a volumetric aeration rate of 1 vvm.

Analytical methods. Biomass was determined by gravimetric analysis after filtration of cell samples through preweighed nylon filters (45-mm diameter; pore size, 0.8 µm) and dried at 95°C to a constant weight. Glucose, lactose, ethanol, and organic acid concentrations in the filtrate were determined by high-performance liquid chromatography (HPLC) (HPX-87H⁺ column; temperature, 35°C; mobile phase, 5 mM H₂SO₄; flow rate, 0.6 ml min¹; detection at 210 nm with a diode-array detector). The concentration of amino acids in the medium was determined by HPLC (Hewlett-Packard) on an AccQ Tag column (3.9 by 150 mm; Waters catalog no. WATO 52885), after derivatization of samples by 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (Waters AccQ-Fluor method). The two eluents were 10% AccQ Tag (Waters catalog no. WATO 52890) (eluent A) and 60% acetonitrile (eluent B), and the flow rate was 1 ml min¹. The following separation gradient was applied to the first solvent: 0% eluent A to 100% eluent B within 49 min and 100% eluent A to 0% eluent B within 11 min. The detector used was a HP 1046A fluorimetric detector (_EX = 250 nm and _EM = 395 nm).

1

1

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Batch fermentation of A. terreus in complex medium. A batch fermentation of A. terreus grown on complex medium containing (per liter) 45 g of glucose, 24 g of peptonized milk, and 2.5 g of yeast extract, is depicted in Fig. 1. Lactose (11 g liter¹) detected at the beginning of the cultivation was derived from peptonized milk. Glucose was exhausted after 45 h of cultivation, while 11 g of lactose liter¹ remained in the medium (Fig. 1A). pH decreased from 6.5 to 4.9 during the phase of glucose consumption; thereafter pH increased slowly and reached 6.9 by the end of the culture. At the time of glucose exhaustion the biomass concentration was 24 g liter¹, corresponding to a biomass yield on glucose of 0.55 g g¹. During the glucose consumption phase, the specific growth rate, µ, was 0.07 h¹, and the specific glucose consumption rate, q_glucose, was 0.13 g g¹ h¹. Lactose was consumed upon glucose exhaustion, and the biomass concentration reached 28 g liter¹. Ethanol and CO₂ were the main fermentation products, while acetate was produced at a very low level only. Ethanol accumulated in the medium during the period of rapid glucose consumption. The maximal ethanol concentration was 1.3 g liter¹ after 42 h of cultivation. Ethanol was consumed after glucose and was exhausted after 80 h of cultivation. The gas exchange rates (Fig. 1B) increased during biomass formation and reached a maximum value at the time of glucose exhaustion.

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Fermentation of A. terreus in complex medium with glucose (45 g liter1), peptonized milk (24 g liter1), and yeast extract (2.5 g liter1) at an aeration rate of 1 vvm and agitation speed of 400 rpm. (A) Symbols: , glucose; , biomass; , lactose; , lovastatin. (B) Symbols: , oxygen consumption rate (rO2); , carbon dioxide production rate (rCO2).

1

1

1

1

1

1

1

1

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Fermentation of A. terreus in complex medium with lactose (53 g liter1), peptonized milk (24 g liter1), and yeast extract (2.5 g liter1) at an aeration rate of 1 vvm and agitation speed of 400 rpm. , lactose; , biomass; , lovastatin.

Influence of nitrogen source in chemically defined medium. To investigate the effect of the substrates on lovastatin production, a defined medium with a sole carbon source and a sole nitrogen source was developed. Several inorganic and organic nitrogen sources were tested in shake flask cultures using a defined medium with 45 g of glucose liter¹ as the carbon source (Table 1). Ammonia, which plays a central role in nitrogen metabolism in filamentous fungi (2), was tested beside urea and nitrate, which can be consumed by some fungi (11). A. terreus grew on all nitrogen sources listed in Table 1, indicating their consumption. Glucose was exhausted after 140 h in all experiments except for that with ammonium acetate (Table 1). Although inorganic nitrogen sources like ammonium tartrate, ammonium nitrate, ammonium acetate, sodium nitrate, or urea were consumed for biomass formation, lovastatin production was very poor after 140 h of cultivation (Table 1 [first five rows of data]).

1

1

1

1

Influence of carbon source in chemically defined medium. In defined medium, the carbon and nitrogen sources play a critical role as a source of precursors and cofactors for synthesis of biomass building blocks and lovastatin production. In addition, the carbon source may exert complex regulation on gene expression and enzyme activities for polyketide synthesis. Lactose, glycerol, and ethanol as the sole C source or combined with glucose were tested in shake flasks cultures with sodium glutamate (12.5 g liter¹) as the nitrogen source. Lovastatin concentration, residual carbon source, and residual glutamate concentrations were measured 160 h after inoculation.

1

1

1

1

1

1

1

1

1

Batch cultures of A. terreus in chemically defined medium. A batch cultivation experiment was performed on synthetic medium containing 45 g of glucose liter¹ and 12.5 g of sodium glutamate liter¹ as unique carbon and nitrogen sources, respectively. Fig. 3A shows that glutamate was exhausted after about 60 h of cultivation, while glucose was exhausted after about 130 h. No ammonium was detected in the culture supernatant throughout the experiment. The biomass concentration reached 9.5 g liter¹ at the time of glutamate exhaustion. On synthetic medium the specific growth rate was lower than on complex medium (0.052 and 0.07 h¹, respectively), and biomass yield on glucose was lower (0.25 and 0.55 g g¹, respectively), resulting in a period of glucose consumption three times longer than that on complex medium. The glucose specific consumption rate, q_glucose, was 0.196 g g¹ h¹, and the glutamate specific consumption rate, q_glutamate, was 0.045 g g¹ h¹. Metabolism remained purely oxidative. Both the oxygen uptake rate and the carbon dioxide production rate increased during the fermentation (Fig. 3C). Maximal values were observed after 61 h of cultivation when glutamate became limiting; the oxygen uptake rate reached 18 mmol h¹, and the carbon dioxide production rate reached 24 mmol h¹. Gas exchange rates were lower than those during growth on complex medium because glucose was metabolized significantly slower.

View larger version (20K): [in this window] [in a new window]   FIG. 3.   Fermentation of A. terreus in chemically defined medium with glucose (40 g liter1) and glutamic acid (9.8 g liter1) at an aeration rate of 1 vvm and agitation speed of 400 rpm. (A) Symbols: , glucose; , biomass; , glutamic acid; , lovastatin. (B) Symbols: , oxygen consumption rate rO2; , carbon dioxide production rate rCO2.

1

1

1

1

1

1

1

1

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In filamentous fungi, biosynthesis of secondary metabolites is subject to complex regulations. This study focused on the influence of N and C sources on the regulation of lovastatin biosynthesis in A. terreus. Under all cultivation conditions, in complex and defined media, A. terreus produced neither monacolin J nor monacolin L nor mevastatin.

Since growth requirements of A. terreus are limited, a chemically defined medium could be developed. The buffer capacity of this medium allowed us to keep the pH between 6.2 and 7.0. In the defined medium, all nitrogen sources tested (Table 2) were consumed, which was confirmed by biomass formation. However, lovastatin biosynthesis proved dependent on the nitrogen source, as no lovastatin was detected during cultivation in presence of ammonium, nitrate, or urea (Table 1). Only glutamate and histidine, and to a lesser extent glycine, supported lovastatin biosynthesis. Study of NAD- and NADP-glutamate dehydrogenase showed that glutamate and histidine play a key role in generation of idiophase conditions by the formation of -ketoglutarate, which stimulates aflatoxin formation by inhibition of the tricarboxylic acid cycle (3). The influence of nitrogen sources on secondary metabolism has been illustrated for fumonisin B1 biosynthesis in Gibberella fujikuroi (32), for sterigmatocystin (10) and aflatoxin (6, 20) biosynthesis in Aspergillus nidulans and Aspergillus parasiticus, and for penicillin synthesis in Penicillium chrysogenum (9).

During growth on complex medium (glucose, peptonized milk, and yeast extract), glucose was rapidly metabolized under high oxygen consumption and biomass was formed (Fig. 1). Ethanol was also formed, probably as result of a glycolytic overflow. Lactose and ethanol were consumed only after glucose exhaustion. The onset of lovastatin biosynthesis after glucose consumption can be attributed either to relief from carbon catabolite repression or to carbon source limitation. Carbon catabolite repression in Aspergillus is mediated by the negative-acting creA gene product (30, 35). The binding of protein CreAp to specific promoter sites of structural or regulatory genes involved in the utilization of alternative carbon sources, prevents their expression in the presence of glucose or another repressing carbon source (e.g., for alc genes see reference 29 and for endoglucanase encoded by egl1 see reference 12). The utilization of ethanol in A. nidulans is repressed by double lock control by CreAp of activator alcR and structural gene alcA, encoding alcohol dehydrogenase I (19). The lovastatin biosynthetic gene cluster consists of 18 putative open reading frames (ORFs) (18), among which 2 were annotated to encode regulatory proteins, lovE and ORF 13. Analysis of the lovastatin biosynthetic gene cluster revealed that closely spaced SYGGRG consensus sequences, the motif of functional CreAp binding sites in vivo (25), are present in the 5'-proximal region of ORF 13 [SYGGRG(N)₁₅SYGGRG] and in the putative promoter of the divergently transcribed ORF 8 and lovE [SYGGRG(N)₇CYCCRW]. The presence of putative functional CreAp binding sites in two putative regulatory genes suggests that repression of lovastatin biosynthesis by glucose could be mediated by CreAp. Other carbon catabolite repression mechanisms might also be involved. Glucose represses penicillin synthesis in P. chrysogenum (9) by repressing pcbAB, pcbC, and penD, whereas in A. nidulans only pcbC is strongly repressed by glucose. In both species glucose repression of penicillin biosynthesis is not exclusively mediated by creA (8) but is mediated by another putative DNA-binding protein (24).

Batch experiments on lactose as the unique carbon source showed that lactose was consumed by A. terreus (Fig. 2) but consumption stopped when the residual concentration was 25 g liter¹. Lovastatin biosynthesis was low during ethanol and glycerol consumption in shake flasks (Table 2). As nonrepressing carbon sources, lactose, ethanol, and glycerol cannot activate CreAp. Interestingly, lovastatin biosynthesis started only when lactose consumption had stopped. This suggests that lovastatin synthesis was elicited at the cessation of substrate consumption or growth limitation, i.e., under starvation conditions. The role of starvation in secondary metabolism was illustrated in Trichoderma atroviride, in which chitinase gene ech42 is only expressed after glucose or glycerol starvation (21). An implication of starvation as an eliciting factor has been demonstrated for the induction of sterigmatocystin in A. nidulans (17, 34). The transcriptional regulator aflR is regulated by flbA, fluG, and aflR, showing that both asexual sporulation and sterigmatocystin require inactivation of proliferative growth through inhibition of fadA-dependent signaling (17). In A. nidulans, brlA plays a central role in the switch from vegetative growth to sporulation, and glucose starvation induces high expression levels of brlA (34).

A connection between secondary metabolism and sporulation was demonstrated in Streptomyces griseus, where a -butyrolactone-containing compound called A factor binds the ArpA DNA-binding protein, which in turn is no longer able to function as the negative regulator of genes required for sporulation and antibiotic production (31). These authors showed that addition of butyrolactone I (a -butyrolactone-containing compound inhibiting cyclin-dependent kinases which control cell cycle progression) to cultures of A. terreus resulted in submerged sporulation and increased production of lovastatin.

A threefold-higher specific productivity was found with the defined medium on glucose and glutamate as compared to complex medium with glucose, yeast extract, and peptonized milk. An explanation might be that stringent starvation conditions are required: in defined medium the only organic substrates are glucose and glutamate or histidine. Once they have been exhausted no alternative organic compound is available to the cells. By comparison, in a complex medium strict starvation conditions may not apply after glucose exhaustion.