Genetic Connection between Fatty Acid Metabolism and Sporulation in Aspergillus nidulans *

In the Ascomycete fungus Aspergillus nidulans , the ratio of conidia (asexual spores) to ascospores (sexual spores) is affected by linoleic acid moieties including endogenous sporogenic factors called psi factors. Deletion of odeA ( D odeA ) , encoding a D -12 desaturase that converts oleic acid to linoleic acid, resulted in a strain depleted of polyunsaturated fatty acids (18:2 and 18:3) but increased in oleic acid (18:1) and total percent fatty acid content. Linoleic acid-derived psi factors were absent in this strain but oleic acid-derived psi factors were increased relative to wild type. The D odeA strain was reduced in conidial production and mycelial growth; these effects were most noticeable when cultures were grown at 26 °C in the dark. Under these environmental conditions, the D odeA strain was delayed in ascospore production but produced more ascospores than wild type over time. This suggests a role for oleic acid-de-rived psi factors in affecting the asexual to sexual spore ratio in A. nidulans . Fatty acid composition and spore development were also affected by veA , a gene previously shown to control light driven conidial and ascospore development. Taken together our results indicate an interaction between veA and odeA alleles for fatty acid metabolism and spore development in A. nidulans . a for Identification and Cloning of the A. nidulans D -12 Desaturase Gene— An A. nidulans cosmid library (pWE15, Genetics City, and A. parasiticus cosmid library (pA1, pro- vided by J. Linz, Ref. 13) were screened using a putative D -12 desaturase gene from the yeast Candida albicans. The gene was contained in a 1.8-kb Eco RI fragment obtained from the plasmid pFAD4 (provided J. and MaGee). The screening of the A. parasiticus library yielded a cosmid, pAMC8, containing the putative D -12 desaturase gene. 20 m internal standard ethyl (Sigma) determina- tion of hydroxy fatty acid methyl esters (HFAME). were more selectively concentrated in the hexane layer, while HFAME were mainly extracted into the methanol layer. including the internal standard, were analyzed by ionization chromatography (FID-GC) by removing samples FID-GC was completed with a Spectra Physics Model SP-7100 gas chromatograph equipped with a ionization detector and a capillary column (0.25 film coated with a film of 007 CPS-2, J and S Scientific). The carrier flow was 1 ml/min, and the temperature programming was 100 to 180 °C at a rate of 3 °C/min.AfterFAME analysis, the methanol layer from the hexane/methanol partition, containing HFAME, was recovered, and evaporated to dry-ness. The residue was treated with trimethylchlorosilane/hexamethyl- disilazane/pyridine (3:2:2, v/v/v) to obtain trimethylsilyloxy (OTMSi) derivatives of the hydroxyl groups. After 15 min, the reagent was evaporated and 100 m l of hexane was added for analyses by FID-GC.

In the Ascomycete fungus Aspergillus nidulans, the ratio of conidia (asexual spores) to ascospores (sexual spores) is affected by linoleic acid moieties including endogenous sporogenic factors called psi factors. Deletion of odeA (⌬odeA), encoding a ⌬-12 desaturase that converts oleic acid to linoleic acid, resulted in a strain depleted of polyunsaturated fatty acids (18:2 and 18:3) but increased in oleic acid (18:1) and total percent fatty acid content. Linoleic acid-derived psi factors were absent in this strain but oleic acid-derived psi factors were increased relative to wild type. The ⌬odeA strain was reduced in conidial production and mycelial growth; these effects were most noticeable when cultures were grown at 26°C in the dark. Under these environmental conditions, the ⌬odeA strain was delayed in ascospore production but produced more ascospores than wild type over time. This suggests a role for oleic acid-derived psi factors in affecting the asexual to sexual spore ratio in A. nidulans. Fatty acid composition and spore development were also affected by veA, a gene previously shown to control light driven conidial and ascospore development. Taken together our results indicate an interaction between veA and odeA alleles for fatty acid metabolism and spore development in A. nidulans.
In addition to the effects of psi factor on Aspergillus development, recent studies have also shown that purified linoleic acid and hydroperoxy linoleic acids derived from seed also exhibit sporogenic activities toward several Aspergillus spp. including A. nidulans and the seed infecting fungi Aspergillus flavus and Aspergillus parasiticus (9). In all of these species, the primary effect of linoleic acid and hydroperoxy linoleic acids was to induce precocious and increased conidial development. Lower concentrations of linoleic acid and 9(S)-hydroperoxy linoleic acid stimulated sexual spore development rather than conidial development in A. nidulans (9). These data suggest a relationship between linoleic acid and/or its derivatives and Aspergillus developmental processes.
The sporogenic effects of linoleic acid, hydroperoxy linoleic acids, and psi factor were demonstrated on A. nidulans strains with an intact velvet (veA) locus (5,9). In A. nidulans veA strains, light delays and reduces sexual development and induces conidial production, while in the absence of light conidial production is repressed and the fungus develops cleistothecia (10). Strains with mutations in veA (veA1) exhibit light-independent development of conidia and ascospores (10). Furthermore, veA1 mutants do not develop spores in response to psi factor, linoleic acid, and other linoleic acid derivatives (5,9). veA locus was originally identified by Kä fer in 1965 (11), who also isolated the veA1 mutation. veA has been sequenced (accession number U95045). 1 Currently the cellular function of VeA is not known.
As a first step in understanding the molecular genetics of psi factor formation and psi factor effects on Aspergillus development, we deleted the odeA gene encoding an oleate ⌬-12 desaturase, which catalyzes the conversion of oleic acid into linoleic acid. The effects of the odeA deletion were examined in both veA and veA1 genetic backgrounds. The absence of odeA changed the fatty acid profile, including the composition of psi factor, and raised total percent fatty acids/fungal tissue. Interestingly, veA also affected fatty acid composition. Detailed examination of psi factor levels in the veA strains at 26°C indicated that, in comparison to the wild type, the ⌬odeA strain was delayed in psi factor biosynthesis; however, higher amounts of psi factor were found over time. This was paralleled by delayed but increased ascospore production in this strain compared with wild type. The ⌬odeA strain also displayed delayed and decreased conidial production compared with wild type in both veA and veA1 backgrounds. Interactions between odeA and veA affected spore development, fatty acid composition, and psi factor composition.

EXPERIMENTAL PROCEDURES
Fungal Strains and Growth Conditions-A. nidulans strains used in this study are listed in Table I. Cultures were grown on A. nidulans glucose minimal medium (GMM) 2 unless otherwise indicated. GMM consists of 10 g of glucose, 6 g of NaNO 3 , 0.52 g of KCl, 0.  4 EDTA in 100 ml of distilled H 2 O), in 1 liter of distilled H 2 O. pH was adjusted to 6.5 with a 10 N NaOH solution. Appropriate supplements corresponding to the auxotrophic markers were added to the medium (12). Agar (15 g/liter) was added to obtain solid medium. Temperature of incubation was 37°C, unless indicated otherwise. Cultures were grown in continuous white light or in the dark. Cultures requiring white light were grown in an incubator equipped with General Electric 15-W broad-spectrum fluorescent light bulbs (F15T12CW) positioned at a distance of 20 cm from the agar surface, with a light intensity of 66 mE/m 2 /s. Identification and Cloning of the A. nidulans ⌬-12 Desaturase Gene-An A. nidulans cosmid library (pWE15, Fungal Genetics Stock Center, Kansas City, KS) and A. parasiticus cosmid library (pA1, provided by J. Linz, Ref. 13) were screened using a putative ⌬-12 desaturase gene from the yeast Candida albicans. The gene was contained in a 1.8-kb EcoRI fragment obtained from the plasmid pFAD4 (provided by J. Beckerman and P. MaGee). The screening of the A. parasiticus library yielded a cosmid, pAMC8, containing the putative ⌬-12 desaturase gene. A 1.2-kb BamHI fragment from pAMC8 containing the carboxyl-terminal coding region of the A. parasiticus ⌬-12 desaturase gene was used as a probe to screen the A. nidulans pWE15 genomic cosmid library, and a single cosmid, pWEO2H5, was obtained. Fragments from these cosmids were subcloned to facilitate sequencing and construction of transformation vectors.
Deletion of A. nidulans odeA-The transformation vector utilized to delete odeA was called pAMC31.3. This plasmid included the argB marker gene and odeA flanking sequences without the odeA encoding region. pAMC31.3 was constructed as follows: first, the plasmid pAMC26.6 was obtained by insertion of a 7-kb SphI-XhoI fragment from pWEO2H5 into the SphI-SalI sites in pK19 (14). Another plasmid, pAMC29.1, was generated by digestion of pBlueScript SKϪ (Stratagene) with EcoRI and XhoI, followed by a blunt-end reaction and religation of the plasmid. A 4.3-kb SmaI fragment, containing the 1.3-kb odeA encoding region, was released from pAMC26.6 and ligated into the SmaI site in pAMC29.1 to create pAMC30.4. The entire odeA encoding region plus 389 base pairs downstream from the putative stop codon was then removed from pAMC30.4 by a SalI-NdeI double digest. This left two 1.3-kb genomic DNA fragments that were on either side of the excised odeA gene. The remaining linear vector was blunt-ended and ligated to a blunt-ended 1.8-kb fragment containing the A. nidulans argB gene to obtain the final transformation vector pAMC31.3. Using standard procedures (16), A. nidulans FGSC89 (biA1; argB2) ( Table I) was transformed with pAMC31.3 to create TAMC31.65 (biA1; veA1, ⌬odeA). Replacement of odeA by argB is denoted by the symbol ⌬odeA. The ⌬odeA allele was introduced in a veA background by sexual recombination of TAMC31.65 with WIM126 (pabaA1; veA). Ultimately, four isogenic strains differing only in veA and odeA alleles were created through transformation and sexual crosses ( Table I).
Southern Analysis-Approximately 5 g of genomic DNA from transformed A. nidulans strains were double digested with SalI/NdeI, and another 5 g were digested with BamHI. Genomic DNA samples were separated by electrophoresis in a 1% agarose gel and then transferred to Hybond membrane (Amersham Pharmacia Biotech) by capillary action. Gene replacement of odeA with argB was confirmed by probing with a 1.7-kb SalI-NdeI fragment from pAMC30.4 and a 4.3-kb SmaI fragment from pAMC26.6. DNA was labeled with 32 P by random primer extension as described by Sambrook (18).
mRNA Studies-Total RNA was isolated from mycelia by using Trizol as described by the supplier (Life Technologies, Inc.). Approximately 20 g of total RNA were used for RNA blot analysis. Temperature and light effects on expression of odeA and the A. nidulans ⌬-9 desaturase gene (sdeA) were analyzed by using A. nidulans veA1; veA1, ⌬odeA; veA; and veA, ⌬odeA strains. Inocula (10 6 conidia/ml) were grown in 3 ml of liquid GMM in 7-ml glass vials under stationary conditions for 72 h. Cultures were grown at 20, 26, 30, 37, and 40°C in the dark. Light effects were observed in cultures grown at 37°C. The 1.7-kb SalI-NdeI fragment containing the odeA encoding region and a sdeA cDNA fragment (an EST clone encoding the putative ⌬-9 desaturase of A. nidulans, provided by Drs. B. Roe and D. Kupfer) were used as probes.
Fatty acid feeding studies were carried out in the four near-isogenic strains (i.e. veA1; veA1, ⌬odeA; veA; and veA, ⌬odeA). Inocula (10 6 conidia/ml) were grown in 1-liter flasks containing 400 ml of liquid GMM at 37°C and 300 rpm. After 16 h of growth, equal amounts of mycelia were transferred into GMM plus sodium linoleate, GMM plus sodium oleate, GMM, control minimum medium without glucose (MM) plus sodium linoleate, MM plus sodium oleate and MM. Sodium linoleate and sodium oleate (Sigma) were added to a final concentration of 1 mM in 1% Tergitol Nonidet P-40 (19). Controls containing Tergitol, GMM plus Tergitol and MM plus Tergitol, were also included in the experiment. After 8 h the mycelia were harvested and processed for mRNA analysis.
Physiological Studies-Conidial production studies were performed on plates containing GMM agar plus appropriate supplements that were spread with 100 l of water containing 10 5 conidia of veA1; veA1, ⌬odeA; veA; and veA, ⌬odeA strains. The cultures were incubated in the dark at 26 and 37°C, and in the light at 37°C. After 72 h, a core of 12.5-mm diameter was removed from each plate and homogenized in 2 ml of water to release the spores. Conidia were counted using a hemacytometer. Colony growth was recorded as colony diameter. The experiments were carried out in triplicate.
The studies on sexual spore production were performed with the veA and veA, ⌬odeA strains. The strains were inoculated on YGT medium, since this medium has been used in previous research to promote sexual development in A. nidulans (5,6,9). Five ml of melted 0.7% agar-YGT containing 10 6 conidia were poured on 30 ml of solid 1.5% agar-YGT and incubated in the dark at 26 and 37°C, and in the light at 37°C. After 10 days, a core of 12.5-mm diameter was removed from each plate and homogenized in 2 ml of water to release the spores. Ascospores were counted using a hemacytometer. The experiments were performed with four replicates. Additionally, in order to study the early stages of sexual 2 The abbreviations used are: GMM, glucose minimal medium; kb, kilobase(s); FAME, fatty acid methyl esters; HFAME, hydroxy fatty acid methyl esters; FID-GC, flame ionization detection-gas chromatography; OTMSi, trimethylsilyloxy; PUFA, polyunsaturated fatty acids; HPLC, high performance liquid chromatography; GC-MS, gas chromatography-mass spectrometry; H NMR, proton nuclear magnetic reso- This study a Fungal Genetics Stock Center (Kansas City, KS). veA, velvet gene; veA1, mutation in veA; odeA, delta-12 desaturase gene: ⌬odeA, deletion of odeA; biA1, mutation in biA causing biotin-deficient auxotrophy; argB2, mutation in argB causing arginine-deficient auxotrophy; py-roA4, mutation in pyroA causing pyridoxine-deficient auxotrophy; pabaA1, mutation in pabaA causing 4-aminobenzoic acid-deficient auxotrophy. development in the absence of odeA, a time course was carried out taking microscopic observations at 42, 66, 90, 114, 138, and 162 h after inoculation. Free ascospores and mature asci of veA and veA, ⌬odeA strains were counted at the 162-h time point following the procedure described above.
Fatty Acid Methyl Esters (FAME) Analysis-veA1; veA1, ⌬odeA; veA; and veA, ⌬odeA strains were incubated in 15 ml of liquid GMM in 8-cm diameter plates under stationary conditions at 37°C in the light, 37°C in the dark, or 26°C in the dark. After 72 h, the mycelial mats were harvested and lyophilized to analyze FAME composition, including psi factors. Samples from veA and veA, ⌬odeA on YGT medium were also analyzed at 42, 66, 114, 162, and 240 h for psi factor and FAME composition. Mycelia (56 -134 mg) were sequentially extracted with two portions of 15 ml of chloroform/methanol (2:1, v/v) at Ϫ20°C for a total of 4 days followed by extraction with 20 ml of ethyl acetate/methanol (1:1, v/v) for 4 h at room temperature with agitation. A known quantity of methyl nonadecanoate (19:0, NuChek Prep) was added to the combined extract of fungal lipids (19:0 was previously shown to be absent in Aspergillus strains), and the extract was evaporated to dryness. The lipid residue was partitioned with 20 ml of chloroform/methanol/water (2:1:1, v/v/v), and the chloroform layer was collected and evaporated. The resultant lipid residue was saponified with 2 ml of 0.5 N NaOH in methanol at 65°C for 30 min, after which the solution was acidified with 1 M oxalic acid. Free fatty acids extractable with chloroform after partition in chloroform/methanol/water (2:1:1, v/v/v) were methyl esterified with an excess of diazomethane in diethyl ether/methanol (9:1, v/v) for 2 min at room temperature. FAME recovered after solvent removal were partitioned with 6 ml of hexane/methanol (2:1, v/v), and 20 g of internal standard ethyl ricinoleate (Sigma) was added for determination of hydroxy fatty acid methyl esters (HFAME). FAME were more selectively concentrated in the hexane layer, while HFAME were mainly extracted into the methanol layer. FAME, including the 19:0 internal standard, were analyzed by flame ionization detection-gas chromatography (FID-GC) by removing samples from the hexane layer. FID-GC was completed with a Spectra Physics Model SP-7100 gas chromatograph equipped with a flame ionization detector and a capillary column (0.25 mm ϫ 25 m; film thickness, 0.25 mm; coated with a film of 007 CPS-2, J and S Scientific). The carrier flow was 1 ml/min, and the temperature programming was 100 to 180°C at a rate of 3°C/min.
After FAME analysis, the methanol layer from the hexane/methanol partition, containing HFAME, was recovered, and evaporated to dryness. The residue was treated with trimethylchlorosilane/hexamethyldisilazane/pyridine (3:2:2, v/v/v) to obtain trimethylsilyloxy (OTMSi) derivatives of the hydroxyl groups. After 15 min, the reagent was evaporated and 100 l of hexane was added for analyses by FID-GC. OTMSi HFAME were analyzed with a Hewlett-Packard Model 5890 gas chromatograph equipped with flame ionization detection using the internal standard OTMSi ethyl ricinoleate. The capillary column used was a SPB-1 (dimethyl polysiloxane phase, 0.32 mm ϫ 30 m, film thickness 0.25 m, Supelco). Temperature programming was from 160 to 260°C at 5°C/min with a hold at 260°C for 5 min; the flow was 2 ml/min.
The identity of the FID-GC peaks were confirmed by GC-MS using a Hewlett-Packard Model 5890 gas chromatograph interfaced with a model 5971 mass-selective detector operating at 70 ev. The capillary column utilized was a Hewlett-Packard HP-5MS cross-linked 5% phenylmethylsilicone (0.25 mm ϫ 30 m, film thickness 0.25 m). FAME were analyzed by temperature programming from 140 to 260°C at a rate of 5°C/min with a flow rate of 0.67 ml/min. HFAME were analyzed identically, except the temperature was programmed from 160 to 260°C at a rate of 5°C/min with a hold at 260°C for 10 min.
Statistical Analysis-Chemical and sporulation data were evaluated by analysis of variance (ANOVA). To compensate for the fact that the experimental design was not fully factorial (i.e. all possible combina-tions of temperature and illumination were not tested) each condition (26°C dark, 37°C dark, 37°C light) was treated as an independent category within a single independent variable "environmental treatment." For cases in which the main effect of environmental treatment was significant, single degree of freedom contrasts to test the specific hypotheses of the effect of temperature (37°C dark versus 26°C dark) and the effect of illumination (37°C dark versus 37°C light) were performed.

Sequence Analysis of odeA and Complemention of the ⌬odeA Deletion with a Wild Type odeA Gene
Sequence analysis of A. nidulans odeA at both the nucleotide and amino acid level revealed high similarity with other ⌬-12 desaturases from plants and fungi. OdeA contained the conserved His-rich regions found in other ⌬-12 desaturases (data not shown) (20 -22). Deletion of odeA resulted in loss of polyunsaturated fatty acid biosynthesis and alterations in spore production (described below). Transformation of the ⌬odeA strain with the odeA gene recovered the wild type phenotype as revealed by fatty acid analysis and physiological studies (data not shown). This supports the conclusion that the defects in the ⌬odeA phenotype are solely due to loss of the odeA gene.

Effect of odeA, veA, and Environment on A. nidulans Fatty Acid Composition
Due to the extensive use of veA1 strains as research models throughout the international research community, the odeA deletion was placed in both a veA and veA1 genetic backgrounds, although our main interest was examining sporulation and psi factor composition in the veA strains. Also, because fatty acid composition has been shown to change with temperature (23,24), fatty acid composition was compared in the veA; veA1; veA, ⌬odeA; and veA1, ⌬odeA strains at both 26 and 37°C. Fatty acids were also examined under light and dark regimes at 37°C due to the importance of light in conidial versus ascospore development in veA strains (10).
Fatty acid composition was first examined under culture conditions known to promote asexual spore development (72 h growth and GMM medium). Table II shows the fatty acid composition of veA; veA1; veA, ⌬odeA; and veA1, ⌬odeA under the environmental conditions (treatments) tested. OdeA allele, veA allele, and treatment had a significant effect on fatty acid composition (p Ͻ 0.01). There was a near loss of polyunsaturated fatty acids (PUFA, 18:2 and 18:3) in ⌬odeA strains (Table  II). Their presence was, however, confirmed independently by silver-ion HPLC followed by GC-MS of the HPLC-purified FAME (not shown). To determine whether these low levels of PUFA were of fungal or exogenous origin in the ⌬odeA strains, GMM medium was examined for the presence of linoleic acid. The medium contained trace amounts of linoleic acid (0.8 g), but this amount was negligible in comparison to the total amount found in ⌬odeA mycelia (average of 23 g/culture). Levels of other fatty acids also changed in ⌬odeA mycelia (Table II). For example, the percentage of saturated fatty acids, palmitic acid (16:0), and stearic acid (18:0) was reduced. In addition, the total percentage of FAME per g of mycelium was approximately 2-3-fold higher in the ⌬odeA strains. These effects were conserved independently of veA or veA1 alleles, illumination regimens and temperature (Table II).
The veA allele also had a significant effect on the fatty acid profile (p Ͻ 0.01). The veA strain contained less linoleic acid and linolenic acid but more oleic and stearic acid than veA1 (Table II). An interaction between veA and odeA alleles on the relative percent of each fatty acid except stearic acid was illustrated by the fact that variations in these fatty acids between veA and veA1 strains were not maintained in the ⌬odeA back-ground (p Ͻ 0.01, Table II).
Treatment (temperature and light) also significantly (p Ͻ 0.01) affected fatty acid composition. Linolenic acid (18:3) percentage increased at 26°C, regardless of veA alleles. At this temperature, stearic acid (18:0) percentages also increased and palmitic acid (16:0) decreased, regardless of veA or odeA alleles. An interaction (p Ͻ 0.01) between environmental treatment and odeA was observed in the decrease in linoleic acid at 26°C in wild type odeA but not ⌬odeA strains. This was expected, as the ⌬odeA strain produced virtually no linoleic acid.  (8). We examined these HFAME as methyl ester/OTMSi ether derivatives whereas Mazur et al. (8) reported mass spectra of the methyl esters. The mass spectra of 8-HODE and 8-hydroxy-9,12,15-octadecatrienoic acid (as methyl esters/OTMSi ethers) were similar to those reported previously by Brodowsky and Oliw (25). The latter HFAME, derived from 18:3, was found only in cultures incubated at 26°C; furthermore, it was a minor component compared with the other HFAME (data not shown). This compound has not been previously identified in A. nidulans and we give it the term PsiB1␥. No PsiA1 was detected in any samples. The EI-MS of the other HFAME (methyl ester/OTMSi ether) were as follows in m/z, ion structure, and relative intensity:

Effect of odeA, veA, and Environment on A. nidulans Psi Factor Composition in GMM Medium
Psi factor composition was examined in 72-h GMM cultures (Table III). Under these conditions, odeA allele, veA allele, and treatment had significant effects (p Ͻ 0.01) on psi factor composition. odeA deletion resulted in the loss of detectable psiB1␣. PsiB1␤ levels were greatly increased in the ⌬odeA strains except in the veA, ⌬odeA strain at 26°C. An interaction between odeA and veA was also apparent by the high amount of psiC1␤ found only in the veA1, ⌬odeA strain (Table III).
Other observations included detection of 10-hydroxy-8,12octadecadienoic acid in veA1 and veA strains and 10-hydroxy-8-octadecenoic acid in all the strains (data not shown). This is the first report of their presence in A. nidulans although these PUFA have been detected in other fungi (25)(26)(27). Currently, it is not known if they also act as sporogenic elements.

Fatty Acid and Psi Factor Composition during Sexual Stage Development Is Affected by odeA
Because sexual development has typically been assessed by growth of veA strains on YGT medium and we are attempting to examine a possible role of psi factor composition in sexual development, we looked at psi factor and fatty acid composition in veA strains (wild type odeA and ⌬odeA backgrounds) grown on this medium. Both psi factor analysis and microscopic observations were performed at the same time points for these strains. Fig. 1 shows the changes in relative percent of FAME composition over time in veA (Fig. 1A) and the veA, ⌬odeA strains (Fig. 1B). In both strains, the most unsaturated fatty acid (linoleic in veA and oleic in veA, ⌬odeA) showed a similar trend of decreasing in percentage composition at 66 and 114 h but increasing in percentage at 162 and 240 h. The percent composition of the other detectable fatty acids showed an opposite pattern in both strains.
The FAME weight/mycelium weight peaked at 66 h in both the veA strain and the veA, ⌬odeA strain. As major components of the total FAME, both linoleic acid and oleic acid reflected the trend of the total FAME weight/mycelium weight (Fig. 1C). On a weight basis/g of mycelium, there was a correlation in linoleic acid versus PsiB1␣ plus PsiC1␣, as well as a correlation in oleic acid versus PsiB1␤ plus PsiC1␤. It was also noteworthy that compared with linoleic acid, oleic acid appeared to yield a much greater quantity of psi factors. This greater accumulation of oleic acid-derived psi factor proved to be valid when all values obtained in this study were examined. Linear plots of oleic acid versus PsiB1␤ plus PsiC1␤ (r ϭ 0.873) and linoleic acid versus PsiB1␣ plus PsiC1␣ (r ϭ 0.715) showed that oleic acid-derived psi factor accumulated 2.5-fold greater per quantity of oleic acid compared with linoleic acid-derived psi factor from linoleic acid. In general, about 5-5.8 g of PsiB1␤ plus PsiC1␤ were found per mg of oleic acid, and about 2.2 g of PsiB1␣ plus PsiC1␣ accumulated per mg of linoleic acid. Our examination also showed that although psi factor accumulation was delayed in the veA, ⌬odeA strain, more psi factor was found in this strain at time points later than 42 h (Fig. 1D).

Effect of odeA, veA, and Environment on A. nidulans Developmental Processes
Vegetative growth and asexual spore production were assessed in both veA and veA1 backgrounds. However, sexual development was assessed solely in veA strains because the veA allele is required for response to psi factor, and because sexual development has been better characterized in veA strains (5)(6)(7)(8)(9)(10).
Vegetative Growth- Fig. 2A shows that environmental treatment, veA allele and odeA allele had a significant effect (p Ͻ 0.01) on A. nidulans colony diameter. In wild type odeA backgrounds, the veA strain had a larger colony diameter than the veA1 strain. This was not true in ⌬odeA backgrounds, demonstrating an interaction between veA and odeA alleles (p Ͻ 0.01).  The lack of the odeA allele led to significant reduction in colony diameter. These effects were maintained over 5 days of incubation (data not shown).
Asexual Spore Development-Environmental treatment, veA allele, and odeA allele had a significant effect on conidial production (p Ͻ 0.01; Fig. 2B). There were also significant interactions between veA and odeA, veA, and environmental treatment, and odeA and environmental treatment (p Ͻ 0.01). There was a significant decrease in conidial production in the veA1, ⌬odeA and veA, ⌬odeA strains at 26°C (Fig. 2B). However, at 37°C only the veA1, ⌬odeA strain showed a reduction in conidial production in comparison to the wild type. Both veA and veA, ⌬odeA strains exhibited the light/dark response characteristic of a veA wild type allele which was to produce more conidia in the light than in the dark. No significant differences in conidial production were observed in veA1 strains, wild type odeA or ⌬odeA, with respect to illumination conditions (Fig.  2B). Lower temperature decreased conidial production in all veA1 strains, but only in the veA strain containing the ⌬odeA allele.
Sexual Spore Development-Our studies indicated that light and temperature and the odeA interaction with light and temperature significantly affected ascospore development in A. nidulans ( Fig. 3A and data not shown). As observed before (5,9,10), the wild type veA strain produced more ascospores in the dark than in the light at 37°C (p Ͻ 0.01). This effect was also observed in the ⌬odeA background (p Ͻ 0.05). The ⌬odeA strain showed an increase in ascospore production relative to the wild type odeA strain in 10-day-old cultures grown at 26°C (Fig. 3A, p Ͻ 0.01), but no significant differences between these strains were observed at 37°C in light or dark grown cultures.
To further investigate the effect of the odeA deletion on sexual development at 26°C, we made microscopic observations of sexual development from 42 to 162 h on the wild type and ⌬odeA strain. At 42 h, only hyphal growth was present in both cultures. Hü lle cells were observed at 66 and 90 h in both strains. At 114 and 138 h, the cleistothecial walls were forming. At 162 h asci (at different stages of maturity) and free ascospores were present. Quantitative analysis showed that the numbers of mature asci and free ascospores were higher in the wild type than in the ⌬odeA strain at 162 h (Fig. 3B). This was in contrast to the ⌬odeA 10-day-old culture, which showed an increase in ascospore production with respect to the wild type odeA strain (as mentioned above, Fig. 3A).

odeA and sdeA Are Temperature and Light Regulated
Positive regulation of desaturase genes by low temperatures and light has been reported in other organisms (24, 28 -32). We studied the effect of temperature and light on odeA and sdeA expression in A. nidulans. As previously mentioned, sdeA encodes a putative ⌬-9 desaturase in A. nidulans, that is responsible for the conversion of stearic acid into oleic acid. Fig. 4 shows that both odeA and sdeA transcript accumulation was induced by low temperatures (26 and 20°C) in both veA and veA1 strains and by light (only examined at 37°C) in veA1 strains. As expected, no odeA transcripts were observed in the ⌬odeA strains (Fig. 4, B and D). sdeA transcripts were elevated in the ⌬odeA strains (Fig. 4).

Polyunsaturated Fatty Acid-regulated Expression of odeA and sdeA
The increase in sdeA transcript in the ⌬odeA strains suggested a possible regulation of PUFA on sdeA expression. To further investigate this possibility, all four strains were grown in various carbon sources including unsaturated fatty acids. In the odeA wild type strains, odeA and sdeA expression was notably higher in the presence of glucose than in its absence, however, this was not observed in ⌬odeA strains (Fig. 5, B and  D). When the ⌬odeA strains were grown in linoleic acid as a sole carbon source there was a notable decrease in sdeA transcript accumulation in both veA and veA1 strains. This decrease was attenuated by the addition of glucose in the medium (Fig. 5, B and D). mRNA analysis also showed that the accumulation of odeA transcripts was higher in the veA strain than in veA1 strain (Fig. 5, A and C). In the veA strain, a slight reduction in odeA and sdeA transcript accumulation was also observed when exogenous linoleic acid was added (Fig. 5A). DISCUSSION The genus Aspergillus contains many industrially, medically, and agriculturally important species whose mode of reproduction depends primarily on the production of asexual spores called conidia and, for some species, sexual spores called ascospores. Factors contributing to spore development of this genus include linoleic acid (9) and various oxidized derivatives of linoleic acid. These include endogenous A. nidulans sporogenic molecules called psi factor (5-8) and plant defense metabolites, 9(S)-and 13(S)-hydroperoxy linoleic acid (9). This latter point is of significance as many Aspergillus spp. are seed infesting fungi that elicit hydroperoxy linoleic acid production in higher plants (33). The sporogenic response to linoleic acid moieties requires the presence of an intact veA gene. However, as veA1 mutant strains have been historically used by the research community due to their convenient trait of developing asexually in the dark, we have investigated the role of linoleic acid and psi factor on fungal development through character-ization of both A. nidulans veA and veA1 strains deficient in linoleic acid biosynthesis.
As expected, chemical analysis of ⌬odeA strains demonstrated the absolute requirement of OdeA for normal fatty acid metabolism in A. nidulans (Table II and Fig. 1, A and B). In contrast to odeA strains, where linoleic acid content was ϳ50% of FAME, the ⌬odeA strains presented only trace amounts of linoleic acid. The odeA deletion also resulted in a 2-3-fold increase in total fatty acids/weight of fungal biomass. Moreover, the chemical makeup of the fatty acid profile was altered in these strains: palmitic acid content was decreased and both stearic and oleic acid content increased compared with wild type strains. The extraordinarily high amount of oleic acid was likely due not only to the block in the fatty acid pathway, but also to the increase in sdeA transcript accumulation in the ⌬odeA strains (Figs. 4 and 5).
We also found that fatty acid composition was affected by veA. The differences observed between veA and veA1 strains with respect to the fatty acid profile were medium-dependent. Lower amounts of PUFA and higher amounts of monounsaturated fatty acids were found in veA strains compared with those found in veA1 strains in glucose minimum medium, a medium which promotes asexual spore development (Table II). The inverse was observed when the fungal strains were grown in YGT (data not shown), the medium used for promoting the sexual stage in A. nidulans (5)(6)(7)(8). Furthermore, there were significant interactions between veA and odeA alleles on fatty acid metabolism as detailed under "Results" (Tables II and III;  Figs. 4 and 5). This suggests a complex regulation of fatty acid metabolism involving veA and fatty acid biosynthetic genes.
Elimination of odeA also led to changes in both % psi factor/ weight of fungal biomass and psi factor composition. Both PsiB1␣ and PsiB1␤ were found in the wild type strain, but the ⌬odeA strain was crippled in its ability to synthesize PsiB1␣ (Table III). Instead, high levels of PsiB1␤ and PsiC1␤ were found in the mutant strain. Furthermore, an interaction between veA and odeA alleles was demonstrated by the fact that the PsiB1␤ and PsiC1␤ levels were statistically greater in the veA1, ⌬odeA strain versus the veA, ⌬odeA strain (Table III). Also, when grown at 26°C in YGT medium, psi factor was not detected in the ⌬odeA strain until 66 h, at which time the level of psi factor was 5-fold above that of wild type (Fig. 1D). Although there were differences in the amount of psi factor found dependent on experiment (Fig. 1D, Table III), in general the total amount of psi factor detected in ⌬odeA strains was several fold greater than that of wild type odeA strains.
Experiments by Champe et al. (5,6) led to the hypothesis that PsiB1␣ and PsiC1␣ play a prominent role in increasing the sexual to asexual spore ratio in A. nidulans but no experiments were conducted with PsiB1␤ and PsiC1␤ to determine if they also had a role in spore development. Although we did not directly assess the effect of PsiB1␤ or PsiC1␤ on Aspergillus development, our results suggest that these derivatives may also act as sexual sporogenic factors as the increase in ascospore numbers in the ⌬odeA, veA strain at 240 h and 26°C (Figs. 1D and 3A) was accompanied by an increase in Psi␤ level. Our data also suggested that the oleic acid:linoleic acid ratio may be playing a role in the relative development of conidia and ascospores. We note that in Neurospora crassa oleic acid is the predominant fatty acid found in developing asci and mature ascospores, whereas linoleic acid is the predominant fatty acid in asexual tissue in this fungus (34).
The ⌬odeA strains produced less conidia than the odeA wild type strains, especially at low temperatures (Fig. 2B). Aside from some possible role of the oleic acid:linoleic acid ratio on directing asexual to sexual spore development, this decrease could also be explained as a need for high PUFA content for conidial formation in cold environments. Temperature had a decided effect on odeA and sdeA transcript accumulation; both were more abundant when the fungus was grown at lower temperatures (Fig. 4). The adaptation of cells to maintain the membrane fluidity in response to a downward shift in temperature by desaturating fatty acids has been studied in higher plants (35)(36)(37), animals (38), and in cyanobacteria (39 -41). Low-temperature induction of desaturase genes has been reported in cyanobacterium species (24,30,31). Positive regulation of a fungal ⌬-9 desaturase gene by low temperature has been described previously in fungi in Mucor rouxii (32). Considering the increase in linolenic acid (18:3) in cultures grown at 26°C (Table II), it is also likely that A. nidulans contains an -3 desaturase positively regulated by low temperature in a similar manner as described in cyanobacteria (24,30,31).
A most interesting observation in this study was the response of the odeA and sdeA alleles to light. There was light induction of odeA and sdeA transcription but only in the strains containing the veA1 allele. These results also indicate another possible genetic link between veA and odeA. Perhaps this response is part of the reason that there are differences seen in the fatty acid profile between veA and veA1 strains. Lightinduced transcription of green algae and plant desaturases have been recorded (28,29) but this is the first report of a fungal desaturase that responds to light.
The increased expression of the sdeA gene in the ⌬odeA strain (Figs. 4 and 5) and the high levels of oleic acid in the ⌬odeA strain suggest a role of OdeA and/or linoleic acid in regulating fatty acid desaturation in A. nidulans. Additionally, we found that exogenous linoleic acid partially repressed sdeA expression (Fig. 5, A, B, and D). Feedback regulation of ⌬-9 desaturase activity has also been noted in mammals (42)(43)(44)(45) and yeast (19). The attenuation of the negative regulation of sdeA transcript in ⌬odeA strains when grown in medium containing glucose indicates interactions between carbon metabolism and PUFA metabolism. PUFA have also been shown to negatively regulate fatty acid synthase, the first committed step in fatty acid metabolism, in mammals (46). We suggest that depletion of PUFA in the ⌬odeA strain derepresses PUFA regulation of fatty acid metabolic genes leading to the observed 3-fold increase in total fatty acid content of this strain.
In conclusion, we have shown that fatty acid composition in A. nidulans varies during spore development and is influenced by odeA, veA, temperature, and light. Both odeA and veA alleles are required for normal asexual and sexual spore development. Although fatty acid and psi factor composition alters with mutations in both of these alleles, it is not yet possible to attribute asexual or sexual spore production to the presence of specific fatty acids as other aspects of fungal physiology also changed. We hope by characterizing sdeA mutants and genes involved in psi factor formation to further elucidate the role of oleic acid or oleic acid psi factors in spore development. We also found that odeA and sdeA expression, like that of other desaturase genes, is responsive to environmental factors including temperature and light and that OdeA and/or linoleic acid play a role in regulating fatty acid desaturation. In addition, it is important to note that even though most of the Aspergillus research community investigates on veA1 strains, the veA1 mutation leads to major changes in sporulation, fatty acid and psi factor profiles. These results open the possibility that previous findings in veA1 might not always apply in veA strains.