Blue Light Acts as a Double-Edged Sword in Regulating Sexual Development of Hypocrea jecorina (Trichoderma reesei)

The industrially important cellulolytic filamentous fungus Trichoderma reesei is the anamorph of the pantropical ascomycete Hypocrea jecorina. H. jecorina CBS999.97 strain undergoes a heterothallic reproductive cycle, and the mating yields fertilized perithecia imbedded in stromata. Asci in the perithecia contain 16 linearly arranged ascospores. Here, we investigated H. jecorina sexual development under different light regimes, and found that visible light was dispensable for sexual development (stroma formation and ascospore discharge). By contrast, constant illumination inhibited stroma formation, and an interruption of the darkness facilitated timely stroma formation in a 12 h/12 h light-dark photoperiod. The results of genetic analyses further revealed that H. jecorina blue-light photoreceptors (BLR1, BLR2) and the photoadaptation protein ENV1 were not essential for sexual development in general. BLR1, BLR2 and ENV1 are orthologues of the conserved Neurospora crassa WC-1, WC-2 and VVD, respectively. Moreover, BLR1 and BLR2 mediate both positive and negative light-dependent regulation on sexual development, whereas ENV1 is required for dampening the light-dependent inhibitory effect in response to changes in illumination. Comparative genome-wide microarray analysis demonstrated an overview of light-dependent gene expression versus sexual potency in CBS999.97 (MAT1–2) haploid cells. Constant illumination promotes abundant asexual conidiation and high levels of hpp1 transcripts. hpp1 encodes a h (hybrid)-type propheromone that exhibits features of both yeast a and a pheromone precursors. Deletion of hpp1 could rescue stroma formation but not ascospore generation under constant illumination. We inferred that the HPP1-dependent pheromone signaling system might directly prevent stroma formation or simply disallow the haploid cells to acquire sexual potency due to abundant asexual conidiation upon constant illumination.

Light, one of the important environmental cues, affects all the living organisms on Earth directly or indirectly. For example, in Neurospora crassa, light plays a major role in its development such as the induction of carotenoid synthesis [4], promotion of conidiation [5,6], direction of perithecial neck development [7] and entrainment of the circadian rhythm [8,9]. Most light-induced phenotypes in N. crassa were reported to be dependent on two photoreceptor proteins, white-collar-1 (WC-1) and white-collar-2 (WC-2). The WC-1 protein functions as a blue-light receptor via its LOV (light, oxygen, or voltage) domain and by binding to a flavin adenine dinucleotide (FAD) chromophore. WC-1 also physically interacts with WC-2 to form the heteromultimeric ''white-collar complex'' (WCC) [10]. N. crassa also has a light adaptation protein, Vivid (VVD), which is another member of the LOV domain family and acts as a photoreceptor [11]. VVD performs its function by both ligand binding and protein-protein interaction. VVD is localized in the nucleus upon illumination [12] and mediates photoadaptation by interaction with WCC [13,14,15].
The blue-light response is an evolutionarily conserved signaling pathway present in almost all asco-and basidiomycetous taxa [14,16,17]. In H. jecorina, the orthologs of conserved WC-1, WC-2, and VVD are BLR1, BLR2, and ENVOY (ENV1), respectively [18,19,20,21]. BLR1 and BLR2 mediate blue-light-induced mycelial growth and expression of cellulase genes. ENV1 functions in a negative regulatory feedback to tolerate continuous exposure to light and dampens the capacity of the fungus to perceive changes in light intensity [13]. This regulatory feedback involves functional interactions of ENV1 with the G-protein/cAMP/ protein kinase A (PKA) signaling pathway. ENV1 can either influence cAMP levels via its action on the transcription levels of Ga subunits [22,23] or dampen the outputs of Ga by its inhibitory effect on cAMP phosphodiesterase (PDE) [21].
As with other fungi, light is important for initiation of sexual development in H. jecorina [2,24]. Given the scarcity of information on the mechanisms and environmental regulations of sexual development in H. jecarina, the objective of the present study was to explore whether visible light regulates initiation of sexual development by monitoring stroma formation. Our results suggest that BLR1, BLR2 and ENV1 play regulatory roles during these processes, and that predominate light-induced asexual conidiation may disallow sexual development.

Sexual Development of H. jecorina is Sensitive to Light
To monitor sexual development, CBS999.97(1-1) and CBS999.97(1-2) were cultured on a malt extract agar (MEA) plate (see ''Materials and Methods''). Stromata were expected to form at the interaction zone ( Figure 1A). We found that stromata hardly developed under a 24 h photoperiod (24L) even after 30 days. Thus, constant illumination inhibits mating and stroma induction. By contrast, under a 12 h photoperiod (12L12D), stromata with dark brown pigmentation were observed at the interaction zone after 7-9 days. The diameter of stromata usually ranged from 2-5 mm ( Figure 1C). Stroma formation occurred much more slowly under a 0-h photoperiod (24D; constant darkness) than under a 12L12D photoperiod, explaining why it was reported that light was required for stroma formation [2]. The stromata that developed under a 24D photoperiod had pale brown pigmentation, and a diameter of up to 1-2 cm ( Figure 1D). Similar results were observed when CBS999.97(1-1) was crossed with QM6a under a 24L, 12L12D, or 24D photoperiod (data not shown). Haematoxylin and eosin staining for frozen sections of stromata revealed that perithecia were emerged and embedded into the upper surfaces of stromata generated under a 12L12D photoperiod ( Figure 2A) [2]. In contrast, under a 24D photoperiod, perithecia developed more slowly and located deep inside the interior of stromata, and eventually having larger volumes and longer necks toward to the upper surfaces ( Figure 2B-D). Together, our results indicate that different light regimes affect H. jecorina sexual development: constant illumination inhibited stroma formation, while unceasing darkness, compared to the 12L12D photoperiod, slowed down stroma formation. Therefore, an interruption of the darkness under a 12L12D photoperiod that mimics the natural photoperiod represented the optimal condition for sexual development.
The Dblr1 and Dblr2 Mutants are Blind to Light during Sexual Development H. jecorina BLR1 and BLR2 are required for induction of env1 transcript in response to blue or white light during vegetative growth. It is also known that white or blue light slows down vegetative growth of the Denv1 mutant [19,20,21,25]. Here we determined whether light influenced sexual development of Dblr1, Dblr2, Dblr1 Dblr2 (Dblr1,2), and Denv1 mutants in the CBS999.97 background. Each of these four mutants, like the wild-type parental strains ( Figure 1D), slowly developed large stromata under a 24D photoperiod ( Figure 1E). Thus, BLR1, BLR2 and ENV1 were not essential for H. jecorina sexual development under constant darkness. Each of the Dblr1, Dblr2, and Dblr1,2 mutants also slowly formed large stromata under either a 12L12D photoperiod ( Figure 1F) or a 24L photoperiod ( Figure 1G), just as under a 24D photoperiod ( Figure 1E). Therefore, BLRdependent blue-light perception not only inhibited stroma formation under a 24L photoperiod ( Figure 1B) but also accelerated stroma formation under a 12L12D photoperiod ( Figure 1C). All stromata of wild-type and mutant strains described here generated dechexads with 16 ascospores (data not shown), indicating that BLR-dependent blue-light perception was also not essential for meiosis and ascospore formation.

The Denv1 Mutant cannot Respond to Changes in Light Illumination
The CBS999.97 Denv1 mutant did not form stroma under a 12L12D photoperiod ( Figure 1F) or a 24L photoperiod ( Figure 1G). On the other hand, wild-type CBS999.97 could form stromata under a 12L12D photoperiod but not under a 24L photoperiod ( Figure 1C). Thus, Denv1 mutants failed to sense an interruption of darkness under a 12L12D photoperiod. We also showed that the Denv1 mutant could develop stromata and ascospores under a 24D photoperiod ( Figure 1E, right panel). Since transcription of the env1 gene is induced by light [19,20,21], the level of env1 transcript in the wild-type CBS999.97 was very low under a 24D photoperiod (see blow). It is not surprising that loss of the env1 gene does not affect sexual development under constant darkness.
To further confirmed if the function of ENV1 was dependent on illumination, we constructed the Dblr1 Denv1 double mutants. Fertilized Dblr1 Denv1 double mutant could form stromata under a 12L12D photoperiod (data not shown) or under a 24L photoperiod ( Figure 1H). Similar results were observed when a Denv1 single mutant was crossed with a Dblr1 Denv1 double mutant under a 12L12D photoperiod (data not shown) or under a 24L photoperiod ( Figure 1I). Because the Denv1 mutants could not form stroma under a 12L12D ( Figure 1F) or 24L photoperiod ( Figure 1G), we inferred that the Denv1 mutants might display a light-dependent defect in either female fertility or male fertility. The Conidia of CBS999.97 Denv1 are Functional Spermatia Sexual crosses could be performed by mixing the conidia from a CBS999.97 haploid male strain to the mycelia of a CBS999.97 haploid female strain ( Figure 3A). Next, we determined if ENV1 is required for male or female fertility. In brief, the conidia of a male strain were applied to a MEA plate where a female strain has grown into mycelia, and incubated for additional 14 days under a 12L12D photoperiod (Figure 3) or under a 24D photoperiod ( Figure 4). As shown in Figures 3A and 4A, both CBS999.97(1-1) and CBS999.97(1-1) Denv1 conidia induced a CBS999.97(1-2) recipient female strain to form stroma structures, whereas CBS999.97(1-2) and CBS999.97(1-2) Denv1 conidia also induced the CBS999.97(1-1) recipient female strain to form stroma structures. In contrast, conidia from donor strains did not induce stroma formation in female recipient strains with the same mating type. These results indicated that the Denv1 mutant, like wild-type strain, was male fertile under a 12L12D photoperiod. Thus, env1 deletion or blue-light illumination apparently does not affect male fertility.

The Denv1 Mutants become Female Sterile in Response to Light
Intriguingly, we found that CBS999.97(1-1) Denv1 and CBS999.97(1-2) Denv1 were female sterile under a 12L12D photoperiod ( Figure 3C) or a 24L photoperiod (data not shown). By contrast, they were not female sterile under a 24D photoperiod ( Figure 4C; also see Figure 1E, right panel). As the Denv1 mutants showed no defect in male fertility, our result indicated that the lack of env1 leads to perturbation of light-mediated regulatory mechanism responsible for stroma formation.
From these findings, we conclude that BLR proteins receive light and respond by inhibiting stroma formation with constant illumination (24L), but not under a 12L12D photoperiod, and that the photoadaptation protein ENV1 desensitizes this inhibitory effect under a 12L12D photoperiod. This BLR-mediated inhibition specifically affects female fertility. In addition, the effect of ENV1 on stroma development is regulated by an unknown pathway that is apparently independent of the cAMP signaling pathway.  Denv1 with constant illumination for 7.25 days. As internal controls, we first checked the env1 gene expression in these eight conditions by Northern blot hybridization ( Figure 5A) and qRT-PCR ( Figure 5B). Transcription of env1 was higher with illumination (e.g., W-24L and W-12L12D) than in darkness (e.g., W-12D12L and W-24DD). These results correlated with earlier data that transcription of env1 gene was dependent on light illumination [19,26]. The results of qRT-PCR further indicated that, with the level of env1 transcripts in W-24D set to 1, only background levels (,0.003) of env1 transcript in the four CBS999.97(1-2) Denv1 conditions (E-24L, E-12L12D, E-12D12L and E-24D) ( Figure 5B). Hence, we confirm that growth conditions and microarray methods were appropriate for our analysis.

Genome-wide Analysis of Light-dependent Transcription in H. jecorina
H. jecorina wild-type strain and the Denv1 mutant show considerable alternations in gene regulation in dependence on their sexual potency. For example, we found that 193 genes were at least two-fold down-regulated under 4 sexually potent conditions (i.e., W-12L12D, W-12D12L, W-24D and E-24D) in comparison with those under the four sexually impotent conditions (i.e., E-12D12L, E-12L12D, E-24L and W-24L) ( Figure 6A; Table S1). Among these down-regulated genes, we found several evolutionarily conserved genes, including 3 conidiation-specific genes (con-6, con-10, sand), a novel gene (Protein Id: TRIRE2_121135), a homolog of yeast AQY1 water channel gene, two DNA repair enzyme photolyase genes (e.g., phr1), the mata1 mating type gene, the h-type a propheromone gene (hpp1), and several genes involved in a propheromone processing and secretion (e.g., ste6, ste14, ste24, ram2, sir2) (Tables 1 and 2). con-6 and con-10 are two conidiation genes that were identified based on their preferential expression during N. crassa macroconidiophore development [27,28]. Both con-6 and con-10 showed a heighten response to photoinduction in the N. crassa Dvvd mutant [29]. N. crassa VVD is the orthologue of H. jecorina ENV1. The sand gene encodes a spindle pole body-associated protein that affects septation and conidiation in Apsergillus nidulans [30]. Yeast AQY1 gene encodes a spore-specific water channel that mediates the transport of water across cell membrane, and the Aqy1 protein was reported to be involved in spore maturation and freeze tolerance by allowing water outflow [31]. In T. harzianum and T. atroviride, phr1 gene transcripts were induced in conidia during development and spore formation [32,33]. hpp1 encodes H. jecorina h-type a propheromone protein that is not essential for sexual development in daylight [3]. Transcriptional co-induction of hpp1 with several conidiation-specific genes in the four sexually impotent conditions raises an intriguing possibility that hpp1 may have a role in promoting conidiation under constant illumination (see below). Alternatively, the pheromone system might get completely out of balance in the Denv1 mutants upon illumination and thus the mycelia suffered a loss of sexual identity [25].
The results of qRT-PCR ( Figure 5B-F) experiments further confirmed that microarray results could reflect actual expression patterns in the samples. Exceptionally high-levels of hpp1 ( Figure 5C), con-6 ( Figure 5D), con-10 ( Figure 5E), sand ( Figure 5F) and 121135 ( Figure 5G) transcripts were observed in three sexually impotent conditions (i.e., E-12D12L, E-12L12D, E-24L). The transcription levels of these five genes were also higher in CBS999.97(1-2) with constant illumination (W-24L) than in all four sexually potent conditions (W-12L12D, W-12D12L, W-24D, E24D). These results suggest that wild-type CBS999.97(1-2) and Denv1 mutant apparently undergo abundant photoconidiation in the four sexually impotent conditions. This supposition is consistent with earlier reports that Trichoderma species produce conidia in response to blue light illumination [34]. Because ENV1 dampens the light-dependent effect in response to changes in illumination, the Denv1 mutant undergoes abundant photoconidiation even under a 12h light/dark cycle (i.e., E-12D12L and E-12L12D).
Our genome-wide transcription analyses also revealed that 287 genes are at least two-fold up-regulated under 4 sexually potent conditions (i.e., W-12L12D, W-12D12L, W-24D and E-24D) in comparison with those under the three sexually impotent conditions (i.e., E-12D12L, E-12L12D, E-24D) ( Figure 6B; Table  S2). Among these up-regulated genes, we found significant enrichment of genes involved in cellulase and hemicellulase metobolism (hydrolyases, cellulose binding proteins and transporters), electron transport, redox regulation, protein folding, etc. Because these metabolic genes are needed for vegetative growth, we infer that CBS999.97(1-2) and CBS999.97(1-2) Denv1 preferentially undergo hyphal growth (rather than conidiation) in these four sexually potent conditions. We also found several evolutionarily conserved genes are up-regulated, including a homolog of mold-specific M46 gene in a dimorphic fungus Histoplasma capsulatum (Genbank: AAL12252), a G protein b subunit gene, 3 hypothetical G protein coupled receptor (GPCR) genes, a myosin heavy chain gene, and 4 fungus-specific transcriptional factor genes ( Table 3). The results of qRT-PCR experiments indicated that the G b gene (Genbank: EGR45759) ( Figure 5G) and the PTH11-type GPCR [Genbank: ERG51469] ( Figure 5F) were transcribed higher in the four sexually potent conditions (W-12L12D, W-12D12L, W-24D, E-24D) or when the wild-type strain was under constant illumination (i.e., W-24L). Further work will be needed to examine if these genes are indispensable for sexual development.
In summary, the results of genome-wide transcription analysis indicate that CBS999.97(1-2) or CBS999.97(1-2) Denv1 apparently elicit abundant photoconidiation in the four sexually impotent conditions. By contrast, these two strains undergo vegetative growth in the four sexually potent conditions. Constant illumination apparently induces photoconidiation in H. jecorina, and abundant conidiation may reduce its potency for sexual development.

Roles of hpp1 in H. jecorina Sexual Development Upon Constant Illumination
It was reported that deletion of hpp1 did not perturb CBS999.97 sexual development (stromata formation and ascospores discharge) in daylight [3]. We observed similar results when CBS999.97(1-1) Dhpp1 was crossed with CBS999.97(1-2) Dhpp1 under a 12L12D photoperiod (Figure 7, right panel). Because the wild-type CBS999.97(1-2) strain was sexually impotent and expressed a high level of hpp1 transcripts under a 24L photoperiod ( Figure 5), we then examined whether hpp1 is responsible for infertility upon constant illumination. We  Table 1. Annotation of 15 representative genes that are transcriptionally down-regulated in the four sexually potent conditions.

Trire2_121135
A novel gene.

Trire2_82321
Saccharomyces cerevisiae AQY1, an spore-specific water channel gene that mediates the transport of water across cell membranes, developmentally controlled; may play a role in spore maturation.

Trire2_107680
Class I DNA photolyase Phr1, rapidly regulated by blue light in Trichoderma Harzianum.

Trire2_31134
Saccharomyces cerevisiae Ste14 farnesyl cysteine-carboxyl methyltransferase, mediates the carboxyl methylation step during Cterminal CAAX motif processing of a-factor and RAS proteins in the endoplasmic reticulum.

Trire2_62693
Saccharomyces cerevisiae Ste6, a plasma membrane ATP-binding cassette (ABC) transporter required for the export of a-factor Trire2_124222 Saccharomyces cerevisiae Ste24, highly conserved zinc metalloprotease that functions in two steps of a-factor maturation, C-terminal CAAX proteolysis and the first step of N-terminal proteolytic processing.

Trire2_22093
Saccharomyces cerevisiae Ram2, the alpha subunit of both the farnesyltransferase and type I geranyl-geranyltransferase that catalyze prenylation of proteins containing a CAAX consensus motif; essential protein required for membrane localization of Ras proteins and a-factor.

Trire2_67057
Saccharomyces cerevisiae Sir2, a NAD+ dependent histone deacetylase of the Sirtuin family involved in regulation of lifespan; plays roles in silencing at HML, HMR, telomeres, and the rDNA locus; negatively regulates initiation of DNA replication. Sir2 interacts with two silencing and proteins, Sir3 (Ste8) and Sir4 (Ste9), which are required for production of mating pheromone. found that deletion of hpp1 could rescue stroma formation in a 24L photoperiod (Figure 7, left panel), however these stromata were smaller in size (diameter ,1-2 mm) and contained no ascospores after 14 days (data not shown). Thus, overexpression of hpp1 transcripts upon constant illumination is responsible for suppression of sexual mating or stroma formation.

Discussion
Blue light regulates many physiological and developmental processes in fungi [35,36]. N. crassa is the paradigm for studying the molecular mechanisms of blue-light responses. The two White Collar proteins (WC-1 and WC-2) and the photoadaptation protein VVD virtually mediate almost all light-induced phenotypes [15]. Upon the light induction, WC-1 and WC-2 form a WCC complex that binds to the promoters of light-responsive  Table 3. Annotation of representative genes that are transcriptionally up-regulated in the four sexually potent conditions.

Gene group or gene I.D. Annotation
Group I .30 Glycoside hydrolases or cellulose binding proteins (e.g., cel5a, man5a, cle12a, cel61b, cel3a, cip2, cle7b, cip1, etc) Group II .11 putative peptidases or proteinases Group III .25 genes involved in metabolic processes, including electron transport, oxidation-reduction reactions, etc. genes to trigger the transcription. Moreover, VVD mediate photoadaptation in which the phosphorylation of WC-1 results in the exclusion of WCC complex from the promoters of lightresponsive genes [10]. Intriguingly, recent molecular studies in various Trichoderma/Hypocrea species have revealed conserved mechanisms of blue-light perception in mycelia growth and conidiation through the White Collar orthologs BLR-1/BLR2 and the VVD ortholog ENV1 [13,20,21,33,37] and other fungi [16]. Both BLR proteins and ENV1 are required for conidiation.
In the present of light, ENV1 positively regulates the growth rate of mycelium. Gene expression analysis indicates that ENV1 expression depends on light, BLR1 and BLR2. ENV1 is essential for photoadaptation, and is able to alter the transcription pattern of light-dependent and light-independent genes [20,26]. Herein, we demonstrated the effects of the three light sensing components on sexual development of H. jecorina. We provided evidences that light, BLR proteins and ENV1 were not essential for wild-type H. jecorina to induce sexual development. In addition, with the light exposure, BLR-mediated blue-light responses act as a doubleedged sword for H. jecorina sexual development. The stroma development is blocked by constant illumination through the action of BLR proteins, but the perithecium formation is promoted under a 12L12D photoperiod compared with that in unceasing darkness. It is unknown how light facilitates perithecium formation, but it suggests that light may have a positive role in sexual development after stroma induction. Compared with other reports regarding the effects of WC protein homologs on fungal sexual development, WC-1 and WC-2 of the basidiomycetes Cryptococcus neoformans, similar to our results, negatively regulate the sexual filamentation [38]. In contrast, N. crassa WC proteins induce the formation of protoperithecia [39]. Aspergullus nidulans LreA (WC-1) and LreB (WC-2) protein complex acts as an activator of sexual development, and is able to interact with FphA, a phytochrome for red light sensing. Intriguingly, the activities of LreA and LreB are repressed by light through the activity of FphA [40]. Thus, although WC proteins are evolutionally conserved among fungal species, their activities in regulating the sexual development may vary to adapt individual environmental niche.
H. jecorina ENV1 apparently has similar function as VVD in N. crassa [29,41]. Combined with the effects of Denv1 or Dblr1 Denv1 under different light regimes (Figures 1), it appears that ENV1 is not directly involved in stroma development, but does play important roles in promoting sexual development and/or in inhibiting asexual conidiation by desensitizing BLR-mediated signaling responses. Based on the genome-wide transcription analyses described above, it is tempting to propose that H. jecorina BLR-mediated pathways induce both high levels of HPP1 pheromone and abundant asexual conidiation in CBS999.97(1-2) wild-type and Denv1 haploid strains under constant illumination. Further studies will be conducted to determine if and how the HPP1-dependent pheromone signaling system inhibits stroma formation upon constant illumination. Alternatively, HPP1 may promote abundant asexual conidiation and then prohibit the haploid cells to acquire sexual potency. Light also induces a high level of a-factor-like peptide pheromone PPG1 in CBS999.97(1-1) wild-type and Denv1 haploid strains [25], it is also of interest to examine the roles of PPG1 in sexual development under different light regimes.
Since the ENV1 homologs only present in ascomycetes, the interplay between BLR protein and ENV1 of H. jecorina may provide novel regulatory mechanisms on conidiation and sexual development in response to light. Consistently, our data further indicate that H. jecorina blue-light signaling machinery act via different mechanisms in regulating mycelial growth, asexual conidiation and sexual development. The blue light mediated mycelial growth in H. jecorina has been implicated by regulating the levels of cAMP through the inhibition of ENV1 on PDE [21,22]. Addition of caffeine, a PDE inhibitor, rescued the growth defect of Denv1, but did not restore the female fertility of Denv1. This indicated that ENV1-mediated stroma formation is through a distinct pathway, rather than cAMP signaling pathway. A deeper insight into the blue-light signaling networks in H. jecorina sexual development undoubtedly will allow us to understand how fungi succeed in their natural habitats, as well as provide new knowledge for optimizing their industrial applications.

Microbial Strains, Culture Conditions, Mutant Construction and Cytology
Maintenance, single spore isolation, and stock culture of QM6a and CBS999.97 strains were performed as described previously [2]. To monitor sexual development, CBS999.97(1-1) was crossed with CBS999.97(1-2) or QM6a on a 10-cm malt extract agar (MEA) plate as previously described [2]. The MEA plate was incubated in a growth chamber at 25uC under a 24L, 12L12D, or 24D photoperiod, respectively. A plant growth chamber with a light (either white light intensity of about 80 mmol/m 2 /s or blue light with the wavelength of 440-460 nm) was used in this study.

Fertility Assay
To test male or female fertility, sexual crosses were performed using conidia from the male tester strains as spermatia or male gamates. In brief, haploid male parent strain was inoculated on the center a 6-cm MEA plate and incubated at 25uC in a plant growth chamber under a 12L12D photoperiod for 5-7 days. To collect conidia, 2 mL double distilled water was added on the pheromone donor MEA plate for 2 h. This solution was transferred to a 2-mL microcentrifuge tube, centrifuged at 3,300 6 g for 6 min, and the top 500 mL of supernatant was harvested. 20 mL of supernatant were spotted onto a 10-cm MEA plate that had been inoculated with the haploid female parent strain. The MEA plate was then incubated at 25uC in a plant growth chamber under a 12L12D photoperiod ( Figure 3) or a 24D photoperiod (Figure 4) for 14 days.

Quantification of Gene Expression by Northern Blot and qRT-PCR
CBS999.97(1-2) or CBS999.97(1-2) Denv1 was cultured on a MEA plate covered by cellulose based cellophane sheet at 25uC. Only the mycelia that had not reached the edges of the plate were harvested and frozen by liquid nitrogen. The total RNA was extracted with the TRI Reagent Solution (Ambion, Carlsbad, CA) and the RNeasy Plant Mini Kit (Qiagen, Valencia, CA). Briefly, frozen mycelia were ground with a mortar and a pestle. Mycelial powder (100 mg) was mixed with one ml of TRI Reagent solution. Supernatant was centrifuged with 120006g and transferred to a new microcentrifuge tube, and then mixed with 0.2 ml chloroform. After centrifugation, aqueous phase was carefully transferred to a new microcentrifuge tube and mixed with equal volume of 100% ethanol to precipitate RNA. The solution containing the precipitations was loaded onto the RNeasy spin column. Subsequent procedures followed the manufacture's protocol, including on-column DNase digestion with the RNase-Free DNase set (Qiagen, Valencia, CA). The quality of extracted RNA was further analyzed with the RNA 6000 Nano kit by Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). For Northern blot analysis, DNA probes were synthesized by using a Nick Translation kit (GE Healthcare, United Kingdom). The total RNA samples were electrophoresed, transferred to Gene Screen Plus membrane, and autoradiographed.
For qRT-PCR, total RNAs were converted into cDNAs using the oligo(dT) 20 primers and the SuperScript III Fist-Strand Synthesis System (Invitrogen, Carlsbad, CA). Gene-specific primers (Table S3) were designed with the Primer Express 3.0 software. The cDNA samples were diluted, and mixed with genespecific primers and the Fast SYBR Green Master Mix (Applied Biosystems, Foster City, CA) according to manufacture's recommendations. The transcripts of the ribosome protein gene rpl6e were used for normalization of the qRT-PCR data, because this gene was shown to be a suitable reference for light/darkness transcription analysis in T. reesei [43]. The quantifications were performed by the Applied Biosystems 7500 fast Real-Time PCR system with default settings. The normalized expression levels of genes were analyzed with the 7500 Software V2.0.6.

Microarray Experiments and Data Analysis
Microarray experiments were performed using the gene expression full service provided by the Microarray facility and the Bioinformatics Core in the Institute of Molecular Biology, Academia Sinica. Custom-designed oligonucleotide array by Roche-NimbleGen based on Trichoderma reesei v2.0 genome sequence using the 4672000 formation [43]. Invitrogen Super-Script double-stranded cDNA synthesis kit was applied according to following the manufacturer's instructions. Labeling and hybridization were performed by using NimbleGen Systems (NG_Expression_UGuide_v6p0). NimbleGen expression arrays were scanned with a GenePix 4000B Scanner and associated software Genepix6.0. All the microarray experiment were performed following the vendor's standard operating protocol (NimbleGen Arrays User's Guide: Gene Expression Analysis v3.0). Image data were processed using NimbleScan software version 2.6.3 (Roche NimbleGen) to obtain the raw intensity data (.pair file) and converted to gene-leveled data(.calls). All experiments were done in technical triplicate with at least three different biological replicates. Data analysis and normalization were performed using Agilent Gene Spring GX 12.0. Raw intensity scales were transformed by quantile normalization, which used to correct array biases and to make all distributions the same. Both ttest and fold change criteria were employed simultaneously to identify differentially expressed genes with p-value , = 0.05 and fold-change . = 2. For the evaluation of results, the community annotation including the GO (Gene Ontology) classification is available at the T. reesei genome database v2.0 (http://genome.jgipsf.org/Trire2/Trire2.home.html). Table S1 193 genes were at least two-fold downregulated (p value ,0.05) under 4 sexually potent conditions (i.e., W-12L12D, W-12D12L, W-24D and E-24D) in comparison with those under the four sexually impotent conditions (i.e., E-12D12L, E-12L12D, E-24L and W-24L).