The Trichoderma atroviride photolyase-encoding gene is transcriptionally regulated by non-canonical light response elements

Authors


Abstract

The BLR-1 and BLR-2 proteins of Trichoderma atroviride are the Neurospora crassa homologs of white collar-1 and -2, two transcription factors involved in the regulation of genes by blue light. BLR-1 and BLR-2 are essential for photoinduction of phr-1, a photolyase-encoding gene whose promoter exhibits sequences similar to well-characterized light regulatory elements of Neurospora, including the albino proximal element and the light response element (LRE). However, despite the fact that this gene has been extensively used as a blue light induction marker in Trichoderma, the function of these putative regulatory elements has not been proved. The described LRE core in N. crassa comprises two close but variably spaced GATA boxes to which a WC-1/-2 complex binds transiently upon application of a light stimulus. Using 5′ serial deletions of the phr-1 promoter, as well as point mutations of putative LREs, we were able to delimit an ~ 50 bp long region mediating the transcriptional response to blue light. The identified light-responsive region contained five CGATB motifs, three of them displaying opposite polarity to canonical WCC binding sites. Chromatin immunoprecipitation experiments showed that the BLR-2 protein binds along the phr-1 promoter in darkness, whereas the application of a blue light pulse results in decreased BLR-2 binding to the promoter. Our results suggest that BLR-2 and probably BLR-1 are located on the phr-1 promoter in darkness ready to perform their function as transcriptional complex in response to light.

Abbreviations
APE

albino proximal element

BLR

blue light regulator

BLRC

BLR complex

ChIP

chromatin immunoprecipitation

ELRE

early LRE

LLRE

late LRE

LOV

light, oxygen and voltage

LRE

light response element

PAS

PER-ARNT-SIM

PDA

potato dextrose agar

PLRR

potential light response region

TSS

transcription start site

UCR

upstream conserved region

WC

white collar

WCC

WC complex

Introduction

Light is an important environmental cue to which organisms respond in many different ways. Filamentous fungi are able to adapt their physiology to light signals to induce morphogenetic pathways [1]. Blue light regulates photoconidiation, phototropism, entrainment and resetting of circadian rhythms, carotenoid synthesis, sexual and asexual development, among other processes [1-3]. For decades, the filamentous fungus Neurospora crassa has been the classical model system to study blue light signal transduction. In this fungus, the white collar (WC) -1 and WC-2 proteins mediate all known responses to blue light [4-6]. Both proteins contain PER-ARNT-SIM (PAS) domains for protein–protein interactions, and GATA type Zn-finger DNA binding domains. WC-1 functions as a blue light photoreceptor by means of its specialized PAS domain called LOV (light, oxygen and voltage), which harbors a flavin as chromophore [4, 7]. WC-1 and WC-2 form the white collar complex (WCC) via their PAS domains [8]. Upon exposure to light, the WCC binds transiently to promoters of light-inducible genes to activate their transcription [9, 10]. In early studies, it was established that WCC binds to the light response element (LRE) core sequence 5′-GATNC–CGATN-3′, where N can be any nucleotide but the same nucleotide is used in both repeats and the space between the two GATA boxes is variable [9, 10]. Afterwards, a genome-wide microarray analysis revealed a hierarchical network of transcriptional light responses in Neurospora, and bioinformatic analyses identified early light response elements (ELREs) as well as a late light response element (LLRE) in the promoter of genes exhibiting different temporal light-responsiveness. The defined LLRE core was RTGAYRTCA, whereas the ELRE core was GATCB [11] which was subsequently expanded to GATCGA, a consensus derived from 29 light-induced genes identified in a whole genome chromatin immunoprecipitation and sequencing (ChIP-seq) analysis using a WC-2 antibody [12].

Trichoderma atroviride is used as a photomorphogenic model due to its ability to conidiate upon exposure to light [13]. In total darkness, T. atroviride grows indefinitely as a mycelium, whereas a short pulse of blue light induces conidiation forming a discrete ring in the colony perimeter where the light pulse was applied [13]. The corresponding wc orthologous genes in T. atroviride, blr-1 and blr-2, are essential for photoinduction of blue-light-responsive genes and photoconidiation in this fungus [14, 15]. Essentially, blue light regulator (BLR) proteins show similar structures to the WC proteins from N. crassa, except that the BLR proteins lack an evident transcription activation domain [14]. Based on this, it has been proposed that BLR-1 acts as a photoreceptor together with BLR-2 [14, 15]. Overexpression of blr-1 and -2 caused a diminished and an increased sensitivity to light, respectively, which led to the conclusion that BLR-2 is the limiting factor in the blue light response in T. atroviride [16]. In contrast to Neurospora, the putative BLR complex (BLRC) in T. atroviride is responsible for induction and repression of several genes. Analysis by cDNA microarrays (representing 1438 genes) revealed 30 upregulated genes and 10 downregulated genes by blue light [15]. Subsequently, quantitative genome-wide transcriptome analysis allowed the identification of 331 white-light-regulated genes and 204 specifically responsive to blue light, as well as blr-independent regulated genes, suggesting involvement of another blue light perception system [17]. BLR-1 and -2 also impact on the utilization of different carbon sources, the biosynthesis of secondary metabolites, sulfur metabolism and cellulase production [18].

A second photoreceptor, the ENV1 protein, orthologous to VIVID in N. crassa, was also described in Trichoderma reesei. The ENV1 and VIVID proteins contain a PAS/LOV domain and negatively regulate light responses in Trichoderma and Neurospora, respectively [19-22]. Although ENV1 and VIVID showed similar regulation and high sequence similarity at the amino acid level, the env1 gene was unable to complement a Δvvd strain [21], which suggests differences in light regulation in the two fungi.

The photolyase-encoding gene phr-1 was the first blue-light-induced gene identified in T. atroviride [23] and its expression is dependent on both BLR proteins [14]. The phr-1 transcript is absent in the dark, and it is strongly and rapidly upregulated upon exposure to blue light, reaching its maximum level at between 15 and 30 min and decreasing 60 min after the stimulus [23, 24]. Putative light regulatory elements have been described for the phr-1 promoter, including the albino proximal element (APE) [23] and an LRE sequence [15] similar to those defined for early light-responsive genes in N. crassa [10, 25]. Moreover, two putative protein-binding motifs, envoy upstream motifs 1 (EUM1: CTGTGC) and 2 (EUM2: ACCTTGAC), were identified on the env1 and other promoters of blue-light-regulated genes [21, 26]. APE and LRE consensus has also been found on the phr-1 gene promoter in T. reesei [26].

Although phr-1 has been widely used as a molecular marker for light responses in T. atroviride [14-16, 24, 27, 28], to date no experimental analysis exists to demonstrate the role of the predicted cis-acting elements in response to blue light or to prove the interaction of BLR proteins with the phr-1 promoter. In this work we identified the cis-acting elements responsive to blue light in the phr-1 promoter, and we analyzed the interaction of BLR-2 protein on this promoter. For this, we examined a set of 5′ phr-1 promoter serial deletions and Escherichia coli lacZ fusions harboring mutations in some putative regulatory elements, under different light conditions. Our results showed that a short promoter region including several CGATB boxes suffices to drive the blue light induction of phr-1, and that two repeats of inverted CGATB motifs play an important role in such regulation. Besides, we describe the presence of a putative repressor in darkness and demonstrate the binding of BLR-2 along the phr-1 promoter by ChIP.

Results

Delimitation of potential light-responsive elements on the phr-1 promoter

To determine the intergenic region between phr-1 (ID 302457) and its neighbor upstream gene, a bioinformatic analysis on the T. atroviride genome (http://genome.jgi.doe.gov/Triat2/Triat2.home.html) was performed. The phr-1 gene is organized on the genome in a divergent transcription configuration, with the xlf gene (Fig. 1A) that encodes for a 485 amino acid protein (ID 295995) orthologous to the human XRCC4-like factor. This factor belongs to the XLF family and interacts with the XRCC4-DNA ligase IV complex to promote DNA non-homologous end-joining during DNA double-strand break repair [29]. The intergenic region between phr-1 and xlf is 1326 bp in length (Fig. 1A).

Figure 1.

Promoter elements of the photoinducible gene phr-1 are conserved among Trichoderma spp. (A) Schematic representation of phr-1 gene promoter architecture which resulted from T. atroviride, T. virens, T. reesei, T. citrinoviride and T. longibrachiatum phr-1 promoter sequence alignment (Fig. S1). The UCR and the PLRR are depicted by dark grey rectangles. The PLRR contains three inverted CGATB boxes between the GATA-1 and GATA-2 boxes of the core LRE. The APE is shown as a light grey rectangle. The phr-1 ORF and the putative upstream xlf ORF are indicated as black solid arrows. The location and sequence of potential cis-acting elements on the promoter are indicated below rectangles for each. Bold numbers on top represent the location of the different elements along the promoter from the TSS (+1) (thin arrow). (B) RT-PCR analysis of T. atroviride, T. virens and T. reesei phr-1 gene expression in the dark (control) or 30 and 120 min after a blue light pulse. Elongation factor 1 (tef-1) gene was used as housekeeping control gene.

Sequence alignments of phr-1 promoter from five different Trichoderma species (T. atroviride, T. virens, T. reesei, T. citrinoviride and T. longibrachiatum) allowed us to identify two conserved regions along the promoters. The first one was located from −107 to −217 position [relative to the T. atroviride phr-1 transcription start site (TSS)] and contained sequences with significant similarity to previously described LREs [9]; consequently, it was termed the potential light response region (PLRR). The second conserved block of sequences was located from −767 to −938 relative to the phr-1 TSS and was named upstream conserved region (UCR) (Figs 1A and S1).

To verify that the phr-1 gene is regulated in a similar way by blue light in different Trichoderma species, cDNA from T. virens, T. reesei and T. atroviride mycelia exposed to a pulse of blue light was used for RT-PCR analysis of phr-1 gene. Indeed, T. reesei and T. virens phr-1 gene was blue light induced as in T. atroviride, suggesting that the cis-regulatory elements on this gene promoter have remained functional throughout evolution in this fungal genus (Fig. 1B).

To analyze more closely the function of the putative cis-regulatory elements, a set of 5′ serial deletions of the T. atroviride phr-1 promoter were PCR-amplified and translationally fused to the E. coli lacZ reporter gene by cloning the amplicons in the pCB-lacZ vector (see Materials and methods for details). T. atroviride wild-type strain was transformed with these constructs and stable transformants were selected for further analysis. The integrity of endogenous phr-1 gene (Fig. S2A) and integration of the construction into the genome were verified (Fig. S2B). Blue-light-induced gene expression on generated strains was analyzed by quantitative RT-PCR (qRT-PCR) (Fig. 2A,B). Control strain transformed with empty pCB-lacZ vector showed no lacZ transcript (Fig. S3). qRT-PCR analyses from −1326 to −196 promoter versions allowed us to determine that the blue light induction profile was similar to that of the phr-1 endogenous gene (Fig. 2A,B). Slightly higher expression levels of lacZ gene were detected in dark conditions for almost all promoter versions, independently of their transgene copy number (Table S1), compared with the expression level of phr-1 endogenous gene, which suggests the presence of a repressor (Fig. 2B). Furthermore, these results indicate that the UCR is dispensable for light transcriptional regulation of phr-1 under the tested conditions (Fig. 2A,B).

Figure 2.

A conserved region between −157 and −196 from the TSS is responsible for photoinduction of lacZ gene. (A) Schematic representation of 5′ serial deletions of the phr-1 promoter translationally fused to the lacZ reporter gene (white arrow). Putative APE, PLRR and UCR cis-acting elements are represented by dotted, diamond hatched and diagonal line filled rectangles, respectively. Numbers on top represent the distance from the TSS (+1) (thin arrow). The ability to photoinduce lacZ transcription is denoted by +, whereas − denotes lack of photoinduction of the reporter gene. (B) qRT-PCR analysis of phr-1 and lacZ transcripts in darkness (control) (black bars) or 30 min (light grey bars) and 120 min (dark grey bars) after a blue light pulse. The phr-1 and lacZ expressions were normalized to the expression level of act-1. The resulting values were normalized to the 30 min expression level after light treatment. Data represent the mean ± SD (n = 2 for each strain).

In subsequent downstream deletions, the reporter gene expression was constitutive, losing the typical light-regulated expression pattern detected for the phr-1 endogenous gene (Fig. 2B). Therefore, sequences critical for the transcriptional response to blue light reside within the −196 to −157 promoter segment. An additional observation is that the putative APE sequence located in the −87 truncated promoter does not seem to participate in blue light regulation of phr-1 gene (Fig. 2B), as previously suggested [23].

The GATA-2 motif is not necessary for transcriptional blue light regulation

According to the analysis of phr-1 promoter serial deletions crucial information required for phr-1 induction by light resides within the −196 to −157 promoter region. This segment includes a GATA motif (GATA-2 element) similar to those described for N. crassa LREs, which is 42 bp away from a conserved CGATC box (i.e. GATA-1) located at positions −141 to −137 (i.e. within the non-light-responsive −156 truncated promoter). Consequently we decided to examine the impact of mutations in the GATA-2 element on the regulatory qualities of the −196 phr-1 promoter. As can be appreciated in Fig. 3, mutations in the GATA-2 box had no effect on photoinduction of lacZ gene expression, indicating that sequences essential for transcriptional light-responsiveness of the −196 phr-1 promoter are located downstream of the mutated element.

Figure 3.

Effect of mutations of the GATA-2 box on photoinduction of lacZ transcription. (A) Schematic representation of wild-type (−196 bp) and GATA-2 box mutated (−196 mut) versions of PLRR. A grey arrow represents the lacZ reporter gene and the PLRR is represented by a black rectangle, whereas white rectangles inside the PLRR depict the GATA-1 and GATA-2 boxes. Numbers on top represent the distance from the TSS (+1) (thin arrow). The ability to photoinduce lacZ transcription is denoted by +, whereas − denotes lack of photoinduction of the reporter gene. (B) qRT-PCR analysis of phr-1 and lacZ transcripts after 30 (light grey bars) or 120 min (dark grey bars) of the application of a blue light pulse or maintained in the dark (black bars) as control. The phr-1 and lacZ expressions were normalized to the expression level of act-1 and the resulting values were normalized to the 30 min expression level after light treatment. Data represent the mean ± SD (n = 2 for each strain).

Since the 28 bp long promoter segment (−184 to 157) where cis-acting elements relevant for blue-light-responsiveness were mapped does not exhibit any DNA motifs related to the described fungal LREs, we performed a careful analysis of this regulatory region and its flanking sequences. As a result two 11 bp imperfect tandem repeats were identified at −178 to −157 nucleotides, which in the antisense strand displays a CGAT(t/g)TCAGTG sequence. The 7 bp core of these repeats, CGAT(t/g)TC, is very similar to the conserved GATA-1 element (i.e. CGATC) found 15 bp downstream (Figs S1 and 4). Therefore, a plausible interpretation of these data is that LREs on the phr-1 promoter of Trichoderma are actually composed of two or more CGATB boxes arranged in identical or alternative polarities. To search for indirect evidence supporting this interpretation, we carried out a thorough comparative analysis of phr-1 gene upstream non-coding sequences including five Trichoderma species (Fig. S1) and distant relatives in the fungal class Sordariomycetes (Fig. 4). As a result we delimited phr-1 promoter sequences of Fusarium oxysporum, Glomerella cingulata and N. crassa that are actually homologous (i.e. evolutionarily equivalent) to the T. atroviride phr-1 promoter region relevant for blue light transcriptional responses (i.e. the −190 to −106 region). Indeed, the PLRRs of these Sordariomycetes display an analogous arrangement, with several imperfect CGATB repeats in similar or opposite polarities (Fig. 4). In the case of N. crassa, the putative LRE also included imperfect tandem repeats CGATB, whereas in the G. cingulata LRE two CGATB motifs are part of a 10 bp palindromic element (Fig. 4). Experimental data have demonstrated that F. oxysporum and N. crassa phr-1 genes are also blue light regulated and their transcription is dependent on WC proteins [11, 30]. Our bioinformatic analysis indicates that PLRRs were conserved through evolution in Sordariomycetes and that their phr-1 genes could be regulated in the same way as phr-1 from Trichoderma spp.

Figure 4.

Sequence comparisons of the blue-light-responsive region of T. atroviride phr-1 promoter with homologous non-coding sequences of three species of Sordariomycetes. Sequence alignment of the region −106 to −190 of the phr-1 promoter of T. atroviride with the homologous regulatory regions of three distant relatives. The number on the right side of the T. atroviride sequence indicates the nucleotide position relative to the TSS, whereas numbers with an asterisk indicate the nucleotide positions relative to the start codon of Fusarium oxysporum, Neurospora crassa and Glomerella cingulata phr-1 genes [GenBank accession numbers AF500083.1 and X58713.1, and ID 1829160 (http://genome.jgi-psf.org/Gloci1/Gloci1.home.html) respectively]. Number 156 on top and the black arrow below it marks the 5′ border of the truncated version of the phr-1 promoter that did not activate the lacZ expression after a blue light pulse. Putative light-responsive motifs (CGATC or imperfect CGATB repeats) displaying the polarity of canonical WCC binding sites are highlighted in yellow, whereas CGATB motifs showing opposite polarity (indicated by brown arrows) are dark orange. Green arrows depict tandem repeats of a CGATB motif in N. crassa. A 10 bp perfect palindrome in the Glomerella sequence is underlined.

BLR-2 transiently binds to the PLRR and UCR on the phr-1 promoter

In N. crassa, the WCC binds to the ELREs located in the promoter of blue-light-regulated genes, in both darkness and after a light pulse [9-12]. Sequence analysis on several promoters of T. atroviride light-induced genes showed that some of them have proximal and distal LREs, whereas others have only one. It was also observed that the length of the sequence separating the GATA motifs of an LRE is variable among these functional elements (Fig. 5A). Since the light induction of phr-1 in T. atroviride is dependent on blr-1 and -2 products [14], we reasoned that BLR proteins could be interacting with the conserved regions of the phr-1 promoter. To investigate if the BLR-2 protein is able to bind to the phr-1 promoter and to map the possible binding sites of such protein, ChIP assays on crosslinked chromatin of T. atroviride wild-type strain under different light conditions were performed by using BLR-2 antiserum. BLR-2 binding was mapped through the phr-1 intergenic region using primers as indicated in Fig. 5B. In darkness, a discrete positioning of BLR-2 was observed on the UCR, the PLRR and the neighboring regions (Fig. 5C). Thirty minutes after a blue light pulse, a decrease in BLR-2 enrichment was observed. After 2 h of light induction, positioning of BLR-2 on phr-1 promoter was decreased significantly but still detectable (Fig. 5C). In contrast, there was no BLR-2 enrichment in the intermediate region 5 (around −600 bp) at any time.

Figure 5.

Binding of BLR-2 through the phr-1 promoter in T. atroviride. (A) Promoter sequence alignment of representative light-responsive genes in T. atroviride; putative LRE are highlighted in dashed boxes, and putative GATA boxes consensus is indicated in bold letters. Nucleotide positions are numbered relative to the TSS (+1). (B) Scheme of phr-1 promoter. Six pairs of primers were designed to scan the binding of BLR-2 to the phr-1 promoter by ChIP assays using a BLR-2 antiserum. The location of each primer is indicated along the promoter scheme (see Table S2 for details). The PLRR region is filled with diamonds whereas the UCR region is filled with diagonal lines. (C) A ChIP assay was carried out from wild-type strain mycelia grown in darkness (Dark) or exposed to a blue light pulse and collected 30 (light grey bars) and 120 min (dark grey bars) after the light stimulus. Upper panel, representative PCR amplification of purified DNA samples of BLR-2 immunoprecipitation (BLR-2), no antibody (No Ab) and input. Lower panel, histograms represent the difference between the values for BLR-2 and No Ab PCR products immunoprecipitated, divided by the PCR value with input DNA. Three independent ChIP experiments were performed.

Discussion

Species from the Hypocrea/Trichoderma genera have been employed as photomorphogenic models due to their ability to produce a ring of green conidia in response to a pulse of blue light given to a growing colony. After a light pulse, T. atroviride also shows a transient biphasic oscillation in intracellular cAMP levels, changes in membrane potential, as well as in ATP levels and in the induction and repression of a set of genes [15, 31]. Since Berrocal-Tito et al. reported phr-1 as a blue-light-induced gene in T. atroviride [23], it has been used as a molecular marker to study blue light responses in this fungus. Sequence analysis of promoters of blue-light-regulated genes of T. atroviride and T. reesei that were identified in studies with expression arrays [15] revealed the presence of elements similar to the LRE core reported in N. crassa [10]. We showed that those putative regulatory sequences are highly conserved on the phr-1 promoter of five Trichoderma species (Fig. S1). We also determined that sequences similar to described LREs are present in several light-induced genes of T. atroviride (Fig. 5A). Comparative genome analysis of T. atroviride, T. virens and T. reesei indicated that T. atroviride resembles the more ancient state of Trichoderma and that the other species evolved later. Besides, T. reesei has apparently been evolving faster than T. virens and T. atroviride since the time of divergence [32]. The molecular mechanism that controls the blue light transcriptional regulation appears to be conserved in these species, as demonstrated by the sequence (Fig. S1) and transcription analysis (Fig. 1B) of the phr-1 promoter genes and their transcription profiles.

The bioinformatic analysis performed for the upstream sequence of the phr-1 gene of T. atroviride revealed that it is organized on the genome in a divergent transcription configuration with the xlf gene that encodes for an ortholog to the human XRCC4-like factor. This suggests at first glance a co-regulation of both genes by the fact that xlf shares cis-acting elements with the phr-1 gene and because of the presumed function of the XRCC4-like factor on repairing DNA double-strand breaks. However, the xlf gene is not induced by white or blue light (Herrera-Estrella A, personal communication).

It is well established that the phr-1 transcript is absent in darkness and that after a blue light stimulus it undergoes upregulation [23]. However, in qRT-PCR analyses of the reporter gene expression we observed a low but significant transcription of lacZ in darkness with almost all the constructs tested (excluding the −600 bp construct), a fact that was in clear contrast with the transcription pattern of endogenous phr-1. To explain these odd results two alternative hypotheses can be proposed: (a) the native phr-1 promoter is actually active in the dark at very low level, although the phr-1 transcript is rapidly degraded in that condition; (b) there is a repressor present that downregulates phr-1 transcription in the dark but only in its native chromosomal context. If the latter suggestion is correct, then the high level lacZ expression in darkness could be attributed to an effect of chromatin organization on the integration site and/or to a chromosomal position effect. Although transformants were not examined in southern blots, qPCR showed that different constructs could be integrated randomly in the Trichoderma genome. In addition, no direct relationship was revealed between lacZ expression in the dark and the copy number of lacZ determined by qPCR. These statements led us to suggest that integration of almost all constructs was out of the phr-1 chromatin context, leading to loss of their repression in the dark.

The promoter of the T. atroviride phr-1 gene exhibits several sequences similar to the described light regulatory elements such as the LRE and APE of N. crassa [15, 23]. In this fungus it was demonstrated that an LRE conformed by two GATA boxes is required for induction of blue-light-responsive genes and binding of WCC to these sequences in response to light signals was also observed [9, 10]. However, in such studies the two GATA boxes comprising the LRE were not studied separately and no point mutations in those elements were generated and examined to prove their function in vivo. Derived from expression arrays, a tandemly repeated GATC motif related to the core sequence of the LRE [11] was identified in the promoter of the frequency (frq) gene, which is bound by WCC and has been shown to be necessary and sufficient to drive light induction of that gene in vivo [9]. This GATC consensus was extended to GATCGA by ChIP-seq analysis [12]. In the present work we demonstrated that a promoter region of T. atroviride harboring sequences similar to Neurospora LREs but displaying a different arrangement contains the necessary information for phr-1 blue light induction. The GATA-1 box of the presumed phr-1 light-responsive region and those GATA-1-like boxes found on the promoters of blue-light-regulated genes in T. atroviride (Fig. 5A) match the two described consensuses, taking into account that both of them display variability in the first and last bases [11, 12]. On the other hand, our mutagenesis analysis of the GATA-2 motif clearly showed that this element was dispensable for transcriptional light induction of the lacZ gene reporter in the context of the −196 truncated version of the phr-1 promoter.

Comparative analysis of the upstream sequences of photolyase-encoding genes from four different genera of Sordariomycetes (Fusarium, Glomerella, Neurospora and Trichoderma) revealed the conservation of an ~ 90 bp promoter region harboring four to six imperfect repeats with a CGATB core sequence. The arrangement of these repeats is variable among the examined fungal species, although the CGATB boxes with the same polarity as the typical LREs from Neurospora promoters predominate. In this sense, the light-responsive phr-1 promoter region of T. atroviride exhibits an atypical organization because three out of the five CGATB boxes have a different polarity to the canonical WCC binding sites (Fig. 4). In spite of this atypical arrangement of the CGATB repeats, the experimental evidence gathered in this study strongly suggests that the inverted elements are functionally equivalent to the direct CGATB repeats, because the integrity of the −184 to −157 promoter segment containing two inverted CGATB boxes was essential to preserve the blue-light-responsiveness of the truncated or mutated versions of the phr-1 promoter (Figs 2 and 3). Furthermore, there is evidence that the PHRB protein of the basidiomycetous mushroom Lentinula edodes, a homolog of WC-2/BLR-2, is able to bind to GATAWWC boxes in both the sense and antisense strand of a photoregulated gene promoter [33]. In this context, it should be pointed out that all known fungal homologs to WC-2 protein bind GATA-related sequences, hence indicating that those regulatory proteins have conserved their DNA binding specificity since times predating the Ascomycetes–Basidiomycetes evolutionary divergence.

With regard to the issue of the BLRC interaction with the phr-1 promoter we determined that the BLR-2 protein has the ability to bind to the UCR and PLRR, according to ChIP results. The in vivo kinetic enrichment in both regions in response to light suggests that BLR-2 (probably bound to BLR-1) is present in darkness associated with a transcriptional repressor that, upon light stimulus, releases an active BLRC, leading to activation of the phr-1 transcription immediately after the light pulse (Fig. 6). Taken together these results prompt the hypothesis that BLRC regulates the transcription of blue-light-responsive genes by binding to elements analogous to the GATA-1 box (CGATCTC) and the closely associated CGATt/gTC repeats in the antisense strand that are present in the photoresponsive segment of the phr-1 promoter of Tatroviride. Thus, the LREs in the latter regulatory region resemble the LREs described in N. crassa [10-12] because the composite elements are constituted by two or more imperfect CGATB repeats, but differ from the Neurospora regulatory elements because some repeats are in the sense DNA strand whereas others are in the antisense strand. These differences among fungal species could explain the differences in light-regulated protein complexes and molecular transduction pathways that have been described in other organisms like N. crassa and A. nidulans [1, 3].

Figure 6.

Model of phr-1 photoactivation by the BLR complex. (A) Our results showed that BLR-2 and probably BLR-1 proteins (BLRC) are positioned along the promoter of phr-1, including the PLRR (DNA strand blue colored) and the UCR (DNA strand red colored) in the dark, remaining repressed by a putative repressor (green rectangle). (B) Light treatment of Trichoderma leads to de-repression of the BLRC, enabling activation and rapid transcription of phr-1 by RNA polymerase II (Pol II). (C) As time elapses (in the dark conditions), the BLR proteins leave the promoter leading to a decrease of the phr-1 transcript, system desensitization, and repositioning of the BLR proteins on the promoter to be ready to activate the system upon light application. Solid lines indicate supporting evidence from experimental data, and dotted lines (including proteins) indicate hypothetical steps.

Regarding the UCR, we were unable to associate a function of such sequence to the phr-1 promoter under the tested conditions, even when BLR-2 binds to this region, probably to a GATA-like box in such region; however, we do not discard a role of this sequence in other processes regulated by the BLRC.

This is the first investigation in which blue-light-responsive elements in a Trichoderma gene promoter are delimited and characterized, and it is also the first work demonstrating that a promoter section harboring imperfect CGATB repeats in both the sense and antisense strands is able to drive photoinduction of the phr-1 and probably of other genes. However, much work still needs to be done to elucidate the regulatory mechanisms that underlie the photoinducible gene expression in Trichoderma.

Materials and methods

Strains, media and growth conditions

Trichoderma atroviride wild-type strain IMI 206040 (ATCC 32173), T. virens wild-type strain Tv 29.8 [34] and T. reesei wild-type strain QM9414 (ATCC 26921) were grown on potato dextrose agar (PDA) (Difco, Franklin Lakes, NJ, USA) at 25 ºC. E. coli strain TOP10F′ (Invitrogen, Carlsbad, CA, USA) was used for plasmid DNA transformation. The plasmid used was pCB-lacZ (see below).

Construction of pCB-lacZ and different phr-1 promoter versions

pCB-lacZ was generated by inserting a 3.0 kb SmaI-HindIII fragment from pDE1 plasmid [35], containing the E. coli lacZ gene fused to the Aspergillus nidulans trpC terminator, into the corresponding restriction sites of pCB1004 [36], which harbors the hygromycin B resistance cassette. pCB-lacZ was the parental vector for all phr-1 promoter versions and chimerical constructs (Fig. S4).

All the promoter versions of phr-1 gene were PCR amplified using genomic DNA of T. atroviride as a template (Fig. 2A). Forward and reverse universal primers used to amplify the different phr-1 promoter versions included SpeI and SmaI restriction sites, respectively, to facilitate their cloning into pCB-lacZ plasmid (Table S2). PCR conditions were 5 min at 94 °C, 25 cycles of 94 °C (30 s), 58 °C (30 s), 72 °C (30 s for < 500 bp products and 60 s for > 500 bp products) and a final extension of 72 °C for 5 min.

The pCB-lacZ constructs were used to transform T. atroviride protoplasts as described previously [34]. Transformed protoplasts were selected on PDA containing 100 μg·mL−1 hygromycin. Genomic DNA isolation from T. atroviride wild-type and transformants was done as described previously [34]. Candidate strains were analyzed by PCR to determine the presence of intact endogenous phr-1 gene, using −156 SpeI F primer on phr-1 promoter and phr1trR primer on phr-1 ORF. Insertions of translational fusions on the T. atroviride genome were verified by amplifying the phr-1 promoter and lacZ ORF using the primer pair −156 SpeI F and lacZ-R (see Table S2). The copy number of constructs in the transformants genome was calculated by qPCR using the 2−ΔΔCt method [37, 38] and DNA from the strain Δblr-1 was used as a calibrator since southern analysis has shown that it harbors one copy of the hph gene [14].

Photoinduction experiments

General growth conditions and manipulations have been described previously [23]. Briefly, Trichoderma spp. pre-inocula were grown on PDA plates in the dark at 25 ºC for 48 h. Mycelial plugs were cut and removed from the leading edge of the actively growing colony and placed on the center of PDA plates with a cellophane sheet overlying the solid medium. For DNA extraction, T. atroviride was grown for 48 h at 25 ºC and the mycelium was scraped off and immediately frozen in liquid nitrogen. For total RNA extraction, cultures were allowed to grow for 36 h in darkness, exposed to a 1200 μmol·m−2 pulse of blue light (LEE filter no. 183, fluence rate 5 μmol·m−2·s−1), placed back in the dark at 25 ºC, and collected at 30 and 120 min after the exposure to blue light. Mycelia grown in darkness were included as control. At the indicated times, mycelia were collected under red safelight (LEE filter no. 106, fluence rate 0.1 μmol·m−2·s−1) and immediately frozen in liquid nitrogen for further total RNA extraction.

RNA isolation and RT-PCR analysis

Total RNA was isolated using Trizol® Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. RNA quality was assessed using spectrophotometric methods and formaldehyde-agarose gel electrophoresis, taking into account the 28S/18S rRNA ratio. Five micrograms of total RNA were DNaseI (RNase-free) treated (Ambion, Austin, TX, USA). cDNA synthesis was performed using SuperScript II Reverse Transcriptase (Invitrogen), following the manufacturer's recommendations. cDNA was used as a template for quantitative real-time PCR reactions with gene-specific primers and the Fast SYBR Green Master Mix (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's recommendations. Data were measured with the ABI 7500 detection system (Applied Biosystems) and analyzed with the 7500 software V2.0. We amplified cDNA for lacZ, phr-1 and act-1 with specific primers (Table S2). Expression of individual genes was compared and normalized using the 2−ΔΔCt method [37] against the level of act-1 mRNA, which was found to be constant under different light/darkness conditions evaluated in T. atroviride [23]. RT-PCR was performed using GoTaq DNA polymerase (Promega, Fitchburg, WI, USA) with the following conditions: 5 min at 94 °C, 25 cycles of 94 °C (30 s), 57 °C (30 s) and 72 °C (30 s), and a final extension at 72 °C for 5 min.

Chromatin immunoprecipitation assays

Trichoderma atroviride wild-type was grown as described above for blue light induction experiments. Photoinduced mycelia from different times were fixed in crosslinking buffer (50 mm HEPES, pH 7.4, 137 mm KCl, 1 mm EDTA) containing 1% formaldehyde for 15 min at room temperature under red safelight. ChIPs were performed using a BLR-2 antiserum (synthesized by Biosynthesis, TX, USA), which recognizes the sequence CEVEEAQRQWAQSRDGRSDI of BLR-2 protein. After DNA extraction, the pellets were resuspended in 50 μL of TE (10 mm Tris-HCl, pH 7.5; 1 mm EDTA) and subjected to PCR with a set of primers along the phr-1 promoter (see Table S2 and Fig. 5B). The primer pairs used were ChIP R1 F and ChIP R1 R; ChIP R2 F and ChIP R2 R; ChIP R3 F and ChIP R3 R; ChIP R4 F and ChIP R4 R; ChIP R6 F and ChIP R6 R. The PCR program consisted of 5 min at 94 °C, 28–30 cycles of 94 °C (30 s), 58 °C (30 s), 72 °C (30 s), and a final extension at 72 °C for 5 min. To ensure that the amplified PCR products were in the linear range, the PCR conditions were calibrated with different amounts of input DNA (crosslinked chromatin before the immunoprecipitation). Band intensities were quantized by optical density analysis using quantity-one software (Bio-Rad Laboratories, Hercules, CA, USA). As negative controls, mock precipitations were performed in the absence of antibody. Three independent ChIPs were performed.

Acknowledgements

This work was supported by grant SEP-CONACYT 83798 to Alfredo Herrera-Estrella and Sergio Casas-Flores. MGCB, EEUR and TMC are indebted to CONACYT for PhD and Master of Science scholarships.

Ancillary