Reactive oxygen species affect photomorphogenesis in Neurospora crassa.

In Neurospora crassa, several biological phenomena such as the synthesis of carotenoids in the mycelia and polarity of perithecia are regulated by light. We found that a sod-1 mutant, with a defective Cu,Zn-type superoxide dismutase (SOD), showed accelerated light-dependent induction of carotenoid accumulation in the mycelia compared with the wild type. The initial rate of light-induced carotenoid accumulation in the sod-1 mutant was faster than that in the vvd mutant known to accumulate high concentrations. This acceleration was suppressed by treatment with antioxidant reagents. Light-induced transcription of genes involved in carotenoid synthesis, al-1, -2, and -3, was sustained in the sod-1 mutant, whereas it was transient in the wild type. Moreover sod-1 was defective in terms of light-induced polarity of perithecia. By genetic analysis, the enhancement in light-inducible carotenoid synthesis in sod-1 was dependent on the wild type alleles of wc-1 and wc-2. However, the sod-1;vvd double mutant showed additive effects on the carotenoid accumulation in the mycelia. These results suggested that intracellular reactive oxygen species regulated by SOD-1 could affect the light-signal transduction pathway via WC proteins.

In Neurospora crassa, several biological phenomena such as the synthesis of carotenoids in the mycelia and polarity of perithecia are regulated by light. We found that a sod-1 mutant, with a defective Cu,Zn-type superoxide dismutase (SOD), showed accelerated lightdependent induction of carotenoid accumulation in the mycelia compared with the wild type. The initial rate of light-induced carotenoid accumulation in the sod-1 mutant was faster than that in the vvd mutant known to accumulate high concentrations. This acceleration was suppressed by treatment with antioxidant reagents. Light-induced transcription of genes involved in carotenoid synthesis, al-1, -2, and -3, was sustained in the sod-1 mutant, whereas it was transient in the wild type. Moreover sod-1 was defective in terms of light-induced polarity of perithecia. By genetic analysis, the enhancement in light-inducible carotenoid synthesis in sod-1 was dependent on the wild type alleles of wc-1 and wc-2. However, the sod-1;vvd double mutant showed additive effects on the carotenoid accumulation in the mycelia. These results suggested that intracellular reactive oxygen species regulated by SOD-1 could affect the lightsignal transduction pathway via WC proteins.
Light is one of the most important environmental factors being utilized for various biological phenomena by living organisms (1,2). Typical examples are the visual systems in animals, photosynthesis in plants, and photomorphogenesis in fungi. In Neurospora crassa, several biological phenomena are under the control of blue light. Light-induced processes are summarized as follows: (i) synthesis of carotenoids in the mycelia (3), (ii) induction of conidiation (3), (iii) formation of protoperithecia on nitrogen-deficient media (4), (vi) positive phototropism of perithecial beaks (5), (v) positioning of the beak formed on the perithecium (perithecial polarity) (6,7), and (vi) phase shift in the circadian rhythm of conidiation (8,9).
Much information on photoreceptors and light-inducible genes has been accumulated by genetic and molecular biological analyses to understand the molecular mechanism of these photobiological phenomena. In the search for the blue light photoreceptor in N. crassa, light-insensitive mutants, white collar (wc) 1 -1 and wc-2, deficient in most of these blue light responses, have been isolated (4,10,11). The products for these genes are photoreceptors that can activate transcription. Both possess a PAS (PER, ARNT, and SIM) domain, a zinc finger DNA binding motif, and a glutamine-rich putative transcription activation domain (12,13). WC proteins form heterodimers presumably through their PAS domains (14,15), having promoter binding activity of blue light-regulated genes in Neurospora crassa.
In light-inducible carotenogenesis in N. crassa, there is a phenomenon called adaptation (27). In the wild type, al-3 mRNA is transiently expressed in the mycelia under light. The accumulation of al-3 mRNA reached a peak at 30 min after illumination, and decreased to the level in darkness after 1 h even in the presence of light. For this light adaptation, VIVID (VVD) protein, a candidate for a photoreceptor possessing LOV domain has been suggested to function as a negative regulator of WC-1 protein (28,29). WC proteins play a critical role in the blue light signal transduction and light responses, such as carotenoid synthesis in the mycelia through transcriptional activation (30).
However, Dragovic et al. (31) showed that the wc-1, wc-2, and wc-1 wc-2 mutants showed self-sustained rhythmic banding of conidiation and resetting by light. These results indicated that there is a novel light receptor other than WC-1 and WC-2 in N. crassa.
In Fusarium aquaeductuum, the accumulation of carotenoid increased in the mycelia treated with methylene blue or toluidine blue under red light, although the fungus was thought to be insensitive to this spectrum of light (32). In Phaffia rhodozyma, singlet oxygen induced the carotenoid synthesis (33). These results suggested a relation between the intracel-lular redox state and light-induced carotenogenesis. Previously, we have isolated nucleoside diphosphate kinase-1 (NDK-1), as a novel candidate for a light signal-transducing factor other than WC proteins in N. crassa (7,34). The ndk-1 P72H mutant with reduced protein kinase activity showed a deficiency in the light response for perithecial polarity.
In the present study, we investigate reactive oxygen species (ROS) generated by light that could serve as factors to control redox state functioning as controlling factors for redundant pathways of light signal transduction. To test these possibilities, we tried to address the relationship between photomorphogenesis and ROS using a sod-1 mutant of N. crassa, defective in Cu,Zn-superoxide disumutase (Cu,Zn-SOD) (35), and found that light-induced phenomena, including carotenoid synthesis, in the mycelia of the sod-1 mutant are markedly affected. Moreover light-induced transcription of genes involved in carotenoid synthesis was sustained in the sod-1 mutant, under conditions where the expression of these genes was transient in the wild type.
Analysis of Perithecial Polarity-Perithecial polarity was analyzed as described by Oda and Hasunuma (6). The sod-1 mutant (FGSC number 7437) was incubated in darkness at 25°C on synthetic crossing medium for 5 days and fertilized with sod-1 (FGSC number 7438) conidia. After 3 days in darkness, the fertilized cultures were placed in a box with a slit 4 cm wide and illuminated directionally under 12-h light (20 E m Ϫ2 s Ϫ1 )/12-h dark regime for 2 weeks. Perithecia were observed and photographed through a dissecting microscope. In this article, perithecial polarity refers only those perithecia with the beak formed upward. Light-induced perithecial polarity is defined as the difference of those between under illumination and in darkness.
Quantification of Mycelial Carotenoid Accumulation-Conidia (1 ϫ 10 6 ) were inoculated into 20 ml of Vogel's minimal medium in 8.5-cm Petri dishes. The dishes were incubated at 30°C for 36 h in darkness, and then exposed to light (20 E m Ϫ2 s Ϫ1 ) for 0.5, 1.5, 3, 10, and 15 h. After the exposure, mycelia were harvested on filter paper by vacuum filtration. Carotenoid accumulation was determined essentially as described by Linden et al. (11). Dried mycelia (50 mg) were extracted in 3 ml of methanol for 30 min at 60°C. After filtration, the mycelia were re-extracted with 3 ml of acetone for 30 min at 50°C. The methanol/ acetone fraction was collected by centrifugation and the spectral absorption at 470 nm was determined. In the experiments with antioxidants after incubation at 30°C for 36 h in darkness, the mycelia were treated with 1 mM N-acetyl-L-cysteine (37), 1,3-dimethylurea (38), or n-propyl gallate (39) at 30°C for 1 h in darkness. After exposure to light for 2 h, mycelial carotenoids were extracted and the content was determined. N-Acetylglycine, an analogue of N-acetyl-L-cysteine with no antioxidation activity, was used as a control.
RT-PCR-The abundance of wc-1 mRNA was detected by RT-PCR analysis. Total RNA prepared from mycelia exposed to light (20 E m Ϫ2 s Ϫ1 ) for 30 min was treated with DNase I, RT-PCR grade (Invitrogen). Total RNA (5 g) and 1 l of oligo(dT) [12][13][14][15][16][17][18] primer (100 pmol) were denatured in 12 l of reaction solution for 5 min at 65°C and were annealed on ice. First strand cDNA was synthesized with the use of 1 l of SuperScript II reverse transcriptase (Invitrogen) for 50 min at 42°C to generate single stranded cDNA. The RT reaction mixture (1 l) was mixed with 50 pmol of wc-1/forward: 5Ј-cccgggttatccgcaacaacacagacaata-3Ј, wc-1/reverse: 5Ј-gagctctttgcgccaacaaaaacgtt-3Ј primer in 10 l of reverse transcriptase reaction buffer (50 mM KCl, 10 mM Tris-HCl, pH 8.3, 1.5 mM MgCl 2 , 0.2 mM of each dNTP and 0.2 units of Taq polymerase) and incubated for 2 min at 95°C. PCR was performed using the following program: 95°C (30 s), 60°C (30 s), and 72°C (2 min) for 20 cycles with a final 7 min of extension at 72°C. Amplified products were separated by 0.8% TAE-agarose gel electrophoresis and transferred to a piece of GeneScreen Plus membrane (PerkinElmer Life Sciences). Amplified wc-1 was detected by Southern hybridization using wc-1 probe. The relative amount of mRNA for each band was determined with the NIH Image program (National Institutes of Health, Bethesda, MD).
Enzyme Activity Assays for SOD-One ml of conidial suspension (10 6 conidia/ml) was inoculated into 50 ml of Vogel's minimal medium with 2% sucrose in 250-ml Erlenmeyer flasks. After the cultures were shaken (200 rpm) at 25°C for 24 h, they were harvested under room light (0.6 E m Ϫ2 s Ϫ1 ). The mycelia were disrupted in 2 volumes of ice-cold extraction buffer (25 mM Pipes/NaOH, pH 6.3, 0.25 mM EDTA, 0.1 M NaCl, 1 mM MgCl 2 , 0.25 M sucrose, 0.01 mM leupeptin, and 0.5 mM phenylmethylsulfonyl fluoride). The homogenized cell suspension was centrifuged at 10,000 ϫ g for 10 min at 4°C. The resulting supernatant (100 g of total proteins) were loaded onto 6% native polyacrylamide slab gels. After electrophoresis, SOD activity was determined by immersing the gels in 1 mM nitro blue tetrazolium for 20 min at room temperature, followed by an immersion in a solution containing 28 mM tetramethylethylenediamine, 28 M riboflavin, and 36 mM potassium phosphate buffer, pH 7.8, for 15 min under room light (42). SOD activity was visualized as a band where superoxide was decomposed by SOD.

Light-dependent Carotenoid Accumulation in the sod-1
Mutant-To investigate the interaction between the response of light and ROS in N. crassa, we examined light-dependent carotenoid accumulation in mycelia of the sod-1 mutant. In the experiment to quantify mycelial carotenoid, mycelia were harvested in a growing phase prior to conidiation, because carotenoid accumulation in conidia occurs independent of the light response. The light-insensitive mutant, wc-1, was used as a negative control of light response. The sod-1 mutant showed an accelerated light-dependent induction of carotenoid accumulation in the mycelia compared with the wild type (Fig. 1, A and  B). Levels of mycelial carotenoids in the sod-1 mutant were detectable after 1.5 h of illumination (Fig. 1A). The sod-1 mu-tant showed no significant difference in carotenoid accumulation after 15 h of illumination, whereas, the wc-1 mutant failed to produce mycelial carotenoid. A vvd mutant also showed the enhancement of light-inducible carotenoid accumulation in the mycelia (43). The rate of carotenoid accumulation in the sod-1 mutant was higher than that in the vvd mutant at the early period (3 h) of illumination (Fig. 1B). However, the vvd mutant continued to accumulate carotenoid ϳ1.5-fold that in the sod-1 mutant after 15 h of illumination (Fig. 1B).
Expression of Light-inducible Genes in Wild Type and the sod-1 Mutant-The sod-1 mutant showed an increase in the initial rate of light-inducible carotenoid accumulation (Fig. 1). This could result from expression of the carotenoid biosynthetic genes. To test whether ROS regulate the transcription of genes in carotenoid biosynthesis, we analyzed the accumulation of light-inducible gene transcripts after illumination for 0.25, 0.5, 1, 2, and 3 h by Northern blot analysis. In this study, we used al-1, al-2, and al-3, which are light-inducible genes that specify carotenoid synthetic enzymes. We also used bli-4, con-10, and wc-1, which are typical blue light-inducible genes in N. crassa. In the wild type, the expression of all six of these genes is transient, even after constant illumination (Fig. 2, WT). The abundance of al-1, -2, -3, bli-4, and wc-1 mRNA showed a rapid increase and reached a maximum after illumination for 30 min, and decreased rapidly after 1 h, forming a transient expression pattern. However, the expression of con-10 reached a maximum at 60 min, and decreased gradually. In the sod-1 mutant, the mRNAs of al-1, -2, -3, and bli-4 showed increases in response to light stimulation compared with those in the wild type and a more sustained expression pattern (Fig. 2, A and B). However, the accumulation of mRNA for con-10 showed no significant difference from that of the wild type. The vvd mutant showed the expression of the al-1 gene, which was sustained compared with that in the wild type (44). However, the peak of al-1 transcript accumulation in the vvd mutant was 1 h after illumination, and was different than that in the sod-1 mutant (Fig. 2, A and B).
We quantified the wc-1 accumulation by a RT-PCR method followed by Southern blot analysis in the sod-1 and vvd mutants. Significantly more of the light-inducible wc-1 mRNA accumulated in the sod-1 and vvd mutant than in wild type (Fig. 2C).
Effects of Antioxidants on Light-inducible Carotenoid Synthesis and Gene Expression-In the sod-1 mutant, the concentration of intracellular superoxide anion radical, O 2 . is maintained at a higher level than in wild type by the lack of Cu,Zn-SOD. To test whether the increase in carotenoid accumulation is caused by intracellular ROS, the effect of antioxidants on light-inducible carotenoid synthesis was examined. We used N-acetyl-L-cysteine (37), 1,3-dimethylurea (38), and n-propyl gallate (39) as antioxidants. N-Acetyl-L-glycine, which has no antioxidative effect, was used as a control for the N-acetyl-Lcysteine. Each antioxidant was used at a final concentration of 1 mM. This concentration had no inhibitory effect on the mycelial growth (data not shown). The mycelia were treated with each reagent for 1 h in darkness and then were exposed to light for 2 h, and mycelial carotenoids were measured. The accumulation of carotenoids after antioxidant treatment followed by illumination was suppressed by about 40 -50% compared with the control experiments in both wild type and the sod-1 mutant (Fig. 3A). Next, we examined the effect of antioxidant on light-inducible gene expression in the sod-1 mutant. After co-incubation with N-acetyl-L-cysteine for 1 h, the mycelia were exposed to light for 1 h. Light-inducible accumulation of al-1 mRNA was determined by Northern blot analysis. The accumulation of al-1 mRNA in the mycelia treated with N-acetyl-L-cysteine was suppressed by about 40% of that observed in the sample without it (Fig. 3B).
The increases in the accumulation of al-1, -2, -3, and bli-4 mRNAs in the sod-1 mutant may be dependent on the newly expressed wc-1 gene products after illumination (Fig. 2C). To test whether al-1 are under the control of de novo synthesized WC-1, we examined the effect of cycloheximide inhibiting de novo protein synthesis on the light-inducible accumulation of al-1 mRNA in wild type and the sod-1 mutant. After treatment with cycloheximide for 1 h, light-inducible accumulation of al-1 mRNA in the mycelia exposed to light for 15 min was assessed by Northern blot analysis. The increase in the accumulation of al-1 mRNA in the sod-1 mutant was not affected by the treatment with cycloheximide (Fig. 4). These results suggested that increases in the accumulation of light-inducible mRNAs in the sod-1 mutant do not require de novo synthesis of a protein component such as WC-1 and VVD.
Genetic Analysis of the Light-inducible Carotenoid Synthesis Using the sod-1;wc-1, sod-1 wc-2, and sod-1;vvd Double Mutants-The expression of the al-1, -2, and -3 genes is regulated downstream of the WC protein signaling cascade (12,13). The sod-1 mutant showed a higher level of accumulation of lightinducible genes (Fig. 2). The intracellular ROS may act to stimulate carotenoid synthesis via WC proteins. To test this possibility, we observed the carotenoid synthesis in sod-1;wc-1 and sod-1 wc-2 double mutants compared with the wc-1 and wc-2 single mutants. These double mutants series, the sod-1; wc-1, sod-1 wc-2, and sod-1;vvd mutants, showed the lack of Cu, Zn SOD (Fig. 5A). In continuous darkness, wild type and the sod-1 mutant accumulated low levels of the carotenoids in mycelia compared with that of the wc-1 and wc-2 mutants (Fig.  5C). The double mutants, sod-1;wc-1 and sod-1 wc-2, also failed to synthesize carotenoid after illumination (Fig. 5, B and C). These results indicate that the effects of ROS depended on the function of WC proteins.
Furthermore, a double mutant between sod-1 and vvd, sod-

FIG. 1. Light-induced mycelial carotenoid accumulation in wild type, wc-1, vvd, and sod-1 mutants.
A, photograph of mycelia exposed to white light (20 E m Ϫ2 s Ϫ1 ) for 0, 2, and 10 h. B, accumulation of carotenoids in mycelia exposed to white light for 0.5, 1, 3, 10, and 15 h. Carotenoid accumulation in the mycelia was quantified by A 470 /g dry weight of mycelia after extraction with methanol and acetone. Each value is the average of three independent experiments with standard errors.

FIG. 2. Accumulation of light-inducible transcripts after illumination.
A, Northern blot analysis of al-1, al-2, al-3, bli-4, and con-10 mRNA in wild type and the sod-1 mutant, and al-1 mRNA in the vvd mutant. After incubation of mycelia of the above strains for 36 h in darkness, total RNA was extracted from the mycelia after illumination (20 E m Ϫ2 s Ϫ1 ) for 0.25, 0.5, 1, 2, and 3 h. Total RNA was separated on agarose gel, transferred onto a piece of membrane, and hybridized with 32 P-labeled probe. ␤-tub and 25 S rRNA were used for monitoring equal loading. B, densitometric analysis of the Northern blot analysis. The abundance of al-1, bli-4, and con-10 mRNA was calculated relative to ␤-tub mRNA, and that of al-2 and al-3 was 25 S rRNA. Relative values were calculated using the values in darkness for the wild type. C, wc-1 mRNA in wild type, the sod-1 and vvd mutant after illumination for 30 min was quantified by RT-PCR. The amplified wc-1 was detected by Southern blot analysis. ␤-tub was used as a loading control. The fragment amplified by RT-PCR was separated on agarose gel and stained with ethidium bromide. Each value is the average of three independent experiments with standard errors.

FIG. 3. Light-inducible carotenoid and al-1 mRNA accumulation after treatment with antioxidants.
A, after mycelia were treated in the presence of various antioxidants (1 mM each of N-acetyl-L-cysteine, 1,3-dimethylurea, or n-propyl gallate), they were exposed to light (20 E m Ϫ2 s Ϫ1 ) for 2 h. Carotenoid accumulation per g dry weight of mycelia was quantified by A 470 nm . N-Acetylglycine was used as a negative control. B, the accumulation of al-1 mRNA in wild type and the sod-1 mutant treated with 1 mM N-acetyl-L-cysteine after illumination (20 E m Ϫ2 s Ϫ1 ) for 1 h. Each value is the average of three independent experiments with standard errors. 1;vvd, showed an additive effect on light-dependent carotenoid synthesis (Fig. 5B). The sod-1;vvd mutant showed 2-fold carotenoid accumulation compared with that of the vvd mutant after 2 h of illumination. This result suggests that there may be a parallel relationship between sod-1 and vvd in the pathway to induce the carotenoid synthesis.
Light-inducible Perithecial Polarity in the sod-1 Mutant-We next investigated a light-responsive phenotype other than carotenoid synthesis in the sod-1 mutant. Previously, we found that the perithecial polarity was controlled by illumination, because the perithecial beaks corresponding to 80 -90% of total perithecia are formed upward of perithecia under illumination in the wild type of N. crassa, while, in constant darkness, perithecial beaks formed upward was reduced to about 40% of total perithecia (6, 7) (Fig. 6). The effects of illumination on the probability of beaks formed at the top of the perithecia were determined in the mutants used above. In the sod-1 mutant, the upward positioning of the beaks was 20% of total perithecia both in darkness and under illumination, which was the same level as observed in the ndk-1 P72H mutant, whereas phototropism of perithecial beaks toward directional light was normal in both sod-1 and ndk-1 P72H mutants (Fig. 6A). In the photograph of perithecia under light in Fig. 6A, the positioning of beaks in wild type was upward. However, the beaks bent toward the directional light showing positive phototropism, because the light was provided from the left side in the photographs. On the other hand, the positioning of beaks in the sod-1 mutant was mostly sideward, whereas the beaks bent toward the left side. The upward positioning of the beaks in the double mutant between sod-1 and ndk-1 P72H , sod-1;ndk-1 P72H , was 10% of total perithecia both in darkness and under illumination, and an additive effect on perithecial polarity could be detected (Fig. 6B). This result suggests that there may be a parallel relationship between sod-1 and ndk-1 in the pathway to induce perithecial polarity irrespective of illumination. Furthermore, the sod-1;vvd double mutant displayed deficient light response to induce perithecial polarity, similar to that observed in the sod-1 mutant, although the perithecial polarity in the vvd mutant under light was the same level as that observed in the wild type (Fig. 6B). This result indicates that the phenotype of the sod-1 mutant was not affected by the vvd mutation.
To examine the relation of sod-1 and wc-1 or wc-2, we examined the light-inducible perithecial polarity in the wc-1 wc-2, sod-1;wc-1, and sod-1 wc-2 mutants. In the wc-1 and wc-2 mutants, perithecia formed upward were 5% of total perithecia, respectively, irrespective of illumination, whereas the phototropism of perithecial beaks was not detected (Fig. 6A). The sod-1;wc-1 and sod-1 wc-2 double mutants showed neither significant light-induced perithecial polarity nor positive photo-  wc-1,  wc-2, sod-1;wc-1, sod-1 wc-2, vvd, and  sod-1;vvd mutants. A, SOD activity of the crude extracts in wild type, and the sod-1, wc-1, sod-1;wc-1, wc-2, sod-1 wc-2, vvd, and sod-1;vvd mutants. The gel was stained by Coomassie Brilliant Blue as a quantitative control. B, photograph of mycelia of wild type, and the sod-1, wc-1, wc-2, sod-1;wc-1, sod-1 wc-2, vvd, and sod-1;vvd mutants exposed to white light (20 E m Ϫ2 s Ϫ1 ) for 10 h. C, accumulation of carotenoid in the mycelia in darkness and under light (20 E m Ϫ2 s Ϫ1 ) for 2 and 10 h. Carotenoids in the mycelia were extracted with methanol and acetone. Carotenoid accumulation was quantified by the absorbance at 470 nm/g dry weight of mycelia. Each value is the average of three independent experiments with standard error bars. tropism of perithecial beak, similar to the wc single mutants (Fig. 6B). These results indicate that perithecial polarity in the sod-1 mutant is under the control of WC proteins irrespective of illumination, which was essentially the same result as observed in the carotenoid synthesis.

Light-inducible Carotenogenesis Is Controlled by Intracellular ROS-
In this study, the sod-1 mutant showed accelerated light-induced carotenogenesis in the mycelia (Fig. 1). Although this phenotype in the sod-1 mutant was very similar to that of the vvd mutant, the sod-1 mutation was more effective on the initial rate (after 3 h of illumination) of the carotenogenesis than the vvd mutation (Fig. 1). This induction of carotenoid synthesis was suppressed by the exogenic application of antioxidants (Fig. 3). As a general role, SOD reduces levels of the superoxide anion radical O 2 . that is generated by aerobic environments and aerobic metabolism, catalyzing conversion of O 2 . to hydrogen peroxide, H 2 O 2 , and thus protecting the organism from oxidative toxicity (45). N. crassa has two types of SODs, Cu,Zn-SOD and Mn-SOD. The intracellular localization of Cu,Zn-SOD is cytosolic, whereas that of Mn-SOD is mitochondrial (35,46). The sod-1 mutant used in this study lacked the cytosolic type of Cu,Zn-SOD. In plant, it has been known that in coleoptile tips, the intracellular ROS concentration rises by blue light irradiation (47). The intracellular concentration of ROS generated by light may have been enhanced by the loss of SOD resulting in a high level of carotenoid accumulation in the early light induction of carotenoid synthesis. Recently, in N. crassa, catalase-3-deficient mutant showed enhanced production of carotenoid synthesis in colonies under illumination (48). These results suggested that intracellular ROS affect light inducible carotenoid synthesis.
In Northern blot analysis, the light-inducible mRNAs including those for enzymes involved in the synthesis of carotenoids showed a sustained expression in the sod-1 mutant, when compared with the transient expression in the wild type (Fig.  2). Interestingly, the expression of con-10 was not significantly different compared with that of wild type (Fig. 2). By microarray analysis, the expression of light-inducible genes was classified in several patterns (49). It is suggested that the effects of ROS on the gene expression could also be classified to the different patterns in light responses.
The accumulation of the al-1 transcript in the sod-1 mutant was reduced by treatment with antioxidants (Fig. 3B). These results suggest that intracellular ROS affect the expression of light-inducible genes.
In addition to al-1, -2, -3, and bli-4, the light-induced accumulation of the wc-1 transcript also showed an increase in the sod-1 mutant (Fig. 2). The sustained expression of these lightinducible genes in the sod-1 mutant may be caused by the increase in WC-1 protein after illumination, a critical blue light signal-transducing factor in the light-induced carotenoid synthesis in the mycelia (30). However, as the increase of al-1 mRNA in the sod-1 mutant was not affected even after treatment with cycloheximide (Fig. 4), light-inducible transcription in the sod-1 mutant must not require the de novo synthesis of WC-1 and VVD protein. Moreover, as the sod-1;wc-1 and sod-1 wc-2 double mutants failed to generate carotenoid after illumination (Fig. 5), suggesting that intracellular ROS could affect the initial stage in the light signal transduction pathway via WC proteins, independent from de novo protein synthesis.
The sod-1 and vvd mutants showed different kinetics in light-induced carotenoid synthesis (Fig. 1B). These results suggest that the effect in the process of ROS was different from FIG. 6. Light-induced perithecial polarity in wild type, the sod-1, vvd, ndk-1 P72H , sod-1;ndk-1 P72H , sod-1;vvd, wc-1, wc-2, sod-1;wc-1,  and sod-1 wc-2 mutants. A, photograph of perithecial polarity of wild type, the sod-1, wc-1, and wc-2 mutants. White light was provided parallel to the surface of the solid medium (left in the photographs). B, wild type, vvd, ndk-1 P72H , sod-1, sod-1,ndk-1 P72H , sod-1;vvd, wc-1, wc-2, sod-1;wc-1, and sod-1 wc-2 were used as protoperithecial mother strains. Three days after the crosses were fertilized, unilateral light was provided parallel to the solid medium with a 12-h light (20 E m Ϫ2 s Ϫ1 )/12-h dark regime at 25°C (open box). Control cultures were placed in continuous darkness (shaded box). The position of beaks, upward or sideward, of perithecia was observed using a dissecting microscope. The average values for the percentage of beaks formed upward against total beaks observed and their standard errors were calculated from three independent experiments. On the right side of the figure, the total number of perithecia scored after triplicate assays is presented under n. that of VVD. Furthermore, the vvd mutant also showed sustained expression of al-1 mRNA under the light (44) (Fig. 2, A  and B). However, the maximum accumulation of al-1 mRNA in the vvd mutant was observed at 1 h, and that in the sod-1 mutant was at 15 min after illumination (Fig. 2, A and B). WC-1/WC-2 promoted the vvd expression after illumination, and the vvd product suppressed the WC-1/WC-2 function (28,50). VVD may function about 1 h later after illumination. Hence, the sod-1 mutation caused al-1 accumulation at 15 min after illumination, ROS could function within 15 min and then VVD may function. The sod-1;vvd double mutant showed an additive effect on light-induced carotenoid synthesis (Fig. 5B). This additive phenotype could be caused by ROS induced by the sod-1 mutation in the initial stage and de novo synthesized VVD may function at around 1 h after illumination.
How does ROS control the expression of light-inducible genes? Two possibilities exist, regulation at the transcriptional level and regulation at the post-transcriptional level. If the regulation is at the transcriptional level, ROS would act directly on WC-1 and WC-2 proteins via some kind of modification such as phosphorylation of WC-1 and binding between WC-1 and WC-2. WC-1/WC-2 bound to the cis-acting element (light-responsive elements) in the frequency (frq) promotor. The mobility shift assay showed a decrease in the bound WC-1/WC-2 to the promoters by illumination (51). The WC-1/WC-2 binding to light-responsive elements may be controlled by ROS. The LOV domains in Phototropin are thought to function as internal sensors of oxygen, redox potential, and light (52,53). Flavin-binding proteins such as cryptochromes and Phototropins may be affected by the changes in redox state. For example, ROS may control the reversibility of the adduct formation between the chromophore and cysteine residue, which could be modified by ROS, within the LOV domain (54). Hence, intracellular ROS may change the redox state of the WC-1 protein and regulate enzymatic activities such as DNA binding and transcriptional activities.
As another possibility, light-inducible genes may be controlled by ROS at the post-transcriptional level. In general, the stability of transcripts is regulated by RNA-binding proteins or RNases (55). In Synechocystis, psbA mRNA stability is determined by the intracellular redox state (56,57). In humans, redox changes affect erythropoietin RNA-binding protein activity for stabilizing erythropoietin mRNA (58). It was indicated that the intracellular redox state controls the stability of mRNA. In N. crassa, ROS generated by light may act on the stability of mRNA for light-inducible genes. We indicated that intracellular ROS controls the transient expression of lightinducible carotenoid synthesis. In this study, it could thus be a key component in the mechanism of the initial stage in lightinducible carotenoid synthesis in the mycelia. This result suggested that intracellular ROS affect carotenoid synthesis.
ROS Control Light-induced Perithecial Polarity-In our previous reports (6,7,34), we have identified NDK-1 as a component of the light signal-transducing factor in N. crassa. We found that a ndk-1 P72H mutant had an undetectable level of autophosphorylation activity, resulting in loss of the lightinduced response of perithecial polarity (7). In the present study, the sod-1 mutant also showed a defective light-inducible perithecial polarity like the ndk-1 P72H mutant. It was suggested that an intracellular gradient of ROS is generated by light and controls the polarity of perithecia. Moreover, wc-1 and wc-2 mutants showed severer defects in perithecial polarity than the ndk-1 P72H and sod-1 mutants irrespective of illumination (Fig. 6). The sod-1;wc-1 and sod-1 wc-2 double mutants showed no significant difference in perithecial polarity from the wc mutants (Fig. 6B). These results suggested that wild type alleles of wc-1 and wc-2 are required to show the effect of sod-1 in the signaling cascade for perithecial polarity irrespective of illumination (Fig. 5).
The vvd mutant was normal in the light-induced perithecial polarity (Fig. 6B). However, the sod-1;vvd double mutant showed the loss of the light-induced perithecial polarity to the same level as that of the sod-1 mutant (Fig. 6B). These results indicated that VVD did not affect the light-inducible perithecial formation, and the change in the distribution of ROS caused by the sod-1 mutation was an important factor to cause loss of the light-induced perithecial polarity.
Several questions arising here are why about 40% of total perithecia showed that the positioning of the perithecial beak in wild type formed upward in continuous darkness (Fig. 6). The phenotype of the wc-1 and wc-2 mutants seem to be null phenotypes even in darkness, and the positioning of the beaks had random direction (5% of total perithecia; Fig. 6B). If the perithecial polarity is induced depending on illumination, which is accepted by WC-1/WC-2, the perithecial polarity of wild type in continuous darkness should correspond to that of the wc-1 and wc-2 mutants. In the case of light-dependent carotenogenesis, wild type also slightly accumulated the mycelial carotenoids in continuous darkness (Fig. 5C). In the circadian rhythm system in N. crassa, WC-1/WC-2 and FREQUENCY (FRQ) form the feedback loop (59). This WC-1/ WC-2 and FRQ feedback system can be reset not only by light but also temperature shift (60,61). These reports suggested that factors other than light have the capacity to affect the function of WC-1/WC-2. Hence ROS endogenously generated by mitochondria and intracellular metabolisms may affect the perithecial polarity under continuous darkness via WC-1/WC-2.
Another question was how the phenotypes of the sod-1 and ndk-1 P72H mutants can be additive in the double mutant irrespective of illumination (Fig. 6B)? Recently, we found that NDK-1 regulated catalase activity in N. crassa. 2 The primary function of catalase is to reduce hydrogen peroxide to water. Intracellular concentration of hydrogen peroxide could be increased in the ndk-1 P72H mutant. The ROS distribution may be disturbed by the increase of hydrogen peroxide. Consequently, light-inducible perithecial polarity in the ndk-1 P72H mutant may be disrupted. The increase of superoxide and hydrogen peroxide in the sod-1 and ndk-1 P72H mutant may lead to severer (additive) defects in light-inducible perithecial polarity irrespective of illumination than the single mutant of ndk-1 P72H or sod-1.
The ROS are suggested to function as signaling messengers in various biological phenomena in a wide range of organisms (62,63). In this study, we propose that intracellular ROS may function as a novel light signal transducer in carotenogenesis and perithecial polarity.