Functional analysis of an essential Ran-binding protein gene, CpRbp1, from the chestnut blight fungus Cryphonectria parasitica using heterokaryon rescue.

A Ran binding protein (RanBP) homolog, CpRbp1, from Cryphonectria parasitica, has been identified as a protein that is affected by hypovirus infection or tannic acid supplementation. In this study, functional analyses of CpRbp1 were performed by constructing a knockout mutant and analyzing the resulting heterokaryon. Transformation-mediated gene replacement resulted in two putative CpRbp1-null mutants and genotype analyses identified these two mutants as heterokaryotic transformants consisting of two types of nuclei, one with the wild-type CpRbp1 allele and another with the CpRbp1-null mutant allele. Although stable mycelial growth of the heterokaryotic transformant was observed on selective medium containing hygromycin B, neither germination nor growth of the resulting conidia, which were single-cell monokaryotic progeny, was observed on the medium. In trans complementation of heterokaryons using a full-length wild-type allele of the CpRbp1 gene resulted in complemented transformants. These transformants sporulated single-cell monokaryotic conidia that were able to grow on media selective for replacing and/or complementing markers. These results clearly indicate that CpRbp1 is an essential gene, and heterokaryons allowed the fungus to maintain lethal CpRbp1-null mutant nuclei. Moreover, in trans complementation of heterokaryons using chimeric structures of the CpRbp1 gene allowed for analysis of its functional domains, which was previously hampered due to the lethality of the gene. In addition, in trans complementation using heterologous RanBP genes from Aspergillus nidulans was successful, suggesting that the function of RanBP is conserved during evolution. Furthermore, in trans complementation allowed for functional analyses of lethal orthologs. This study demonstrates that our fungal heterokaryon system can be applied effectively to determine whether a gene of interest is essential, perform functional analyses of a lethal gene, and analyze corresponding heterologous genes.


Results
characterization of CpRbp1 gene in response to TA and hypoviral infection. Among the 30 identified protein spots showing changes in accumulation by hypoviral infection or TA treatment, a protein spot tentatively identified as a homolog of RanBP was selected for further analyses 6 . The corresponding gene was identified by exploration of the genome sequence of C. parasitica (http://genome.jgi-psf.org/Crypa2/Crypa2.home.html). A 5,069-bp PCR amplicon containing the complete RanBP gene fragment was cloned, and sequence analyses of the cloned fragment demonstrated that the deduced amino acid sequences comprised the determined amino acid sequences of the gene. Based upon in silico analyses, a near full-length cDNA clone was obtained using the RT-PCR primer pair CpRbp1-cF1 and CpRbp1-cR1 at the positions of nucleotide (nt) -12 to 6 and 1202 to 1225 (1 is the first nucleotide of the start codon), respectively. Comparative analyses of the cDNA and genomic DNA sequences showed that the cloned gene comprised five exons, with four introns of 52 to 289, 443 to 564, 824 to 899, and 1029 to 1084 that were different from the prediction by ORF finder. The nucleotide sequence around the start codon matched to Kozak's consensus sequence where the −3 position was the A in CAACATG. A putative polyadenylation signal sequence consisting of AAAAGA was found 10-bp downstream of the stop codon.
The deduced protein was composed of 238 amino acid residues, with an estimated molecular weight of 26.4 kDa. In addition, the protein was acidic, with a pI of 4.95 (GenBank No. MN053004). Sequence homology analysis using the deduced amino acids of the cloned gene revealed that it was similar to other fungal RanBPs from Fusarium graminearum (83.5% amino acid identity), Metarhizium rileyi (81.1%), Colletotrichum simmondsii (79.7%), C. nymphaeae (79.7%), Aspergillus nidulans (79.3%), C. fioriniae (79.3%), and Ustilaginoidea virens (78.9%). Multiple alignment of seven closely related RanBPs showing an E value of ~0.0, as well as a founding member of the RanBPs, Yrb1 from S. cerevisiae, revealed that the protein product of the cloned gene had a multi-domain structure consisting of an N-terminal nuclear localization signal (NLS), a single canonical RBD in the central region, and an extended C-terminal coiled-coil region. Phylogenetic analyses indicated that the protein product of the cloned gene clustered together with fungal RanBPs, and those from M. rileyi and U. virens shared the most similar evolutionary lineage. The high bootstrap value of the phylogram suggested a genuine evolutionary relationship ( Supplementary Fig. S1). Together with the presence of the hallmark RBD and significant homology to known fungal RanBPs, the cloned gene was referred as CpRbp1 for C. parasitica RanBP1. expression of CpRbp1. Since our previous proteomic analyses indicated that the protein product of the CpRbp1 gene was upregulated by CHV1 infection or TA supplementation 6 , we examined the accumulation of gene transcripts under the corresponding conditions using Quantitative real-time reverse transcription PCR (qRT-PCR) (Fig. 1a). Contrary to the proteomic analyses, no significant changes in the accumulation of CpRbp1 transcripts were observed due to CHV1 infection. However, the accumulation of CpRbp1 transcripts was significantly changed in both wild-type and CHV1-infected UEP1 strains by TA supplementation. Interestingly, the opposite results were observed compared to the proteomic analyses, that is, the CpRbp1 gene was downregulated instead of upregulatied in both wild-type and CHV1-infected UEP1 strains cultured on the TA-supplemented medium. Northern blot analyses of RNAs from both strains at 24 h after the transfer to either supplemented or non-supplemented media verified the qRT-PCR results, showing no changes due to CHV1 infection but downregulation of CpRbp1 transcription by TA supplementation (Fig. 1b). These results clearly indicate that the accumulation of CpRbp1 transcripts in C. parasitica was significantly affected in response to TA, a well-known and abundant host defense component in chestnut trees. However, as the transcriptional analyses were not consistent with the proteomic data, the regulation of CpRbp1 expression appears to be far more complex than expected, occurring at the levels of transcription, post-transcription, and post-translation.
Identification and characterization of putative CpRbp1-null mutants. For functional analyses of the CpRbp1 gene, we attempted to construct a CpRbp1-null mutant by integrative transformation-mediated gene replacement. A total of 92 stable transformants were screened by PCR as described previously, using two pairs of outer gene-specific primers and inner hygromycin B phosphotransferase gene cassette (hph) primers (Primers 1&12 and 2&13 in Fig. 2a) corresponding to nt −1,953 to −1,933 and 1,222 to 1,241, and nt 446 to 465 and 3,097 to 3,116, relative to the start codons of CpRbp1 and hph, respectively. Two transformants showed 3,559-bp-and 3,521-bp PCR amplicons using 5′-proximal and 3′-proximal primer pairs, respectively, which corresponded to the expected sizes of the disrupted alleles of the CpRbp1 gene (data not shown). In addition, these PCR amplicons were further confirmed by sequencing. However, Southern blot analyses of the two putative CpRbp1-null mutants using HindIII digestion and a probe prepared using the 1,004-bp PCR amplicon containing the 3′flanking region of the CpRbp1 gene showed two hybridizing bands at 2.4 kb and 4.0 kb, corresponding to the wild-type and expected CpRbp1-null mutant alleles, respectively (Fig. 2b). PCR amplification using the outer gene-specific primers, located outside of the 5′ and 3′flanking regions of the replacement vector (Primers 1&2 in Fig. 2a), yielded two bands of 6,284 bp and 5,069 bp (Fig. 2c). Restriction enzyme analyses of PCR amplicons followed by sequencing analysis indicated that these PCR amplicons corresponded to the wild-type and CpRbp1-null mutant alleles. These results clearly indicate that the replacing CpRbp1-null mutant allele existed at the site of the CpRbp1 gene, but the wild-type CpRbp1 allele remained simultaneously in the putative CpRbp1-null transformants.
We, then, questioned whether the putative CpRbp1-null transformants were a simple mixture of wild-type and CpRbp1-null mutant strains or heterokaryons consisting of two different types of nuclei (i.e., one with the wild-type CpRbp1 allele and the other with the CpRbp1-null mutant allele) in a single strain. Considering that our stable transformants were maintained by successive transfer of young hyphal tips to the selection media, it is highly unlikely that the putative CpRbp1-null transformants were simply a result of mixed cultures. Taking advantage of the fact that an asexual spore (conidium) of C. parasitica is a single cell containing a single nucleus 29 , a single-spore isolation is enough to resolve the mixed cultures and heterokaryon. Thus, a single-spore isolation, which can breakdown the heterokaryotic state during sporulation, can be applied as a common biological method to obtain a single cell monokaryotic progeny. Conidia from the putative transformants grown on the nonselective potato dextrose agar (PDA) plates supplemented with methionine and biotin (PDAmb) medium were harvested (Fig. 2d), microscopically inspected (Fig. 2e), diluted to the number of 10 2 and 10 3 per plate, and www.nature.com/scientificreports www.nature.com/scientificreports/ plated on the hygromycin B-containing selective PDAmb medium. No colonies were observed on the selective medium even after prolonged (>4 weeks) cultivation (Fig. 2f). However, when the same spore suspension was tested on PDAmb medium, numerous colonies, ranging from 20-30% of the expected number of colonies, were observed, and the resulting colonies were all hygromycin B-sensitive, that is, none were able to grow when transferred to the selective PDAmb medium containing hygromycin B. In addition, PCR amplification of the CpRbp1 locus from the resulting colonies revealed the presence of the wild-type CpRbp1 allele alone. These results suggest that the spore suspension consisted of a mixture of two types of spores: one that was viable but hygromycin B-sensitive, containing the wild-type CpRbp1 allele, and another that was non-viable. Single-spore analyses of the putative CpRbp1-null transformants clearly excluded the possibility of a simple mixed culture of wild-type and CpRbp1-null mutant strains, and strongly suggested that the putative CpRbp1-null transformants were heterokaryon with two different nuclei containing either the wild-type or the null-type CpRbp1 allele existing in a common cytoplasm.
To confirm that the non-viable spores contained the CpRbp1-null mutant allele, in trans complementation of the parental heterokaryotic strains were conducted with the wild-type allele of the CpRbp1 and www.nature.com/scientificreports www.nature.com/scientificreports/ the geneticin-resistance selection marker. Complemented strains showing stable geneticin resistance were single-spored and subsequent colonies showing both hygromycin B-and geneticin-resistance were selected. PCR analyses of progeny showing hygromycin B-and geneticin resistance revealed the presence of both the wild-type and CpRbp1-null mutant alleles. However, colonies showing resistance to geneticin alone or susceptibility to both hygromycin B and geneticin revealed the presence of only the wild-type CpRbp1 allele.
These results clearly indicate that the CpRbp1 gene is essential, that is, its absence is lethal, and confirm that the parental mutant strains are heterokaryotic.

Morphological and cultural characteristics of heterokaryons. The two heterokaryons were named
TdRBP1-Het1 and TdRBP1-Het2. There were no differences in culture characteristics, and both strains were similar to the wild-type on nonselective medium (Fig. 2d); thus, TdRBP1-Het1 was selected for further analyses. Because the genotype frequency in a heterokaryon can vary depending on growth conditions 28,30 , we examined changes in the genotype frequency of heterokaryon progeny. The genotype frequency was estimated by measuring the ratio of colony-forming units (CFUs) from counted conidia of heterokaryon progeny that were successively transferred every fifth day for up to 20 transfers on selective and nonselective media (Fig. 3a). As shown in Fig. 3b, the heterokaryon TdRBP1-Het1 contained a lot more CpRbp1-null mutant nuclei than wild-type nuclei based on the fact that approximately 20% of observed conidia produced colonies at the beginning. When this heterokaryon was successively cultured on nonselective medium by passage every fifth day transfer, the number of CFUs, that is, the ratio of wild-type nuclei to CpRbp1-null mutant nuclei, gradually increased to 60% and remained thereafter. However, when TdRBP1-Het1 was cultured on selective medium, the wild-type frequency persisted at ~20%, suggesting that continuous selection pressure caused the fungus to maintain a high frequency of CpRbp1-null mutant nuclei for the hygromycin B resistance. The colony morphology of successively transferred heterokaryon progeny did not significantly change, and did not differ from that of the wild-type (Fig. 3a). These results indicate that the wild-type nuclei played a dominant role in determining the phenotype in heterokaryon, and that although the genotype frequency may change, balanced heterokaryons were maintained.
One of advantages using the heterokaryon rescue technique is the possiblity to determine the function of the null-allele to the terminal stage. As shown in Fig. 3c, no signs of early germination, such as swelling or germ tube www.nature.com/scientificreports www.nature.com/scientificreports/ formation, were observed from spores containing the mutant allele, suggesting that CpRbp1 is crucial for the fundamental cellular function of this fungus and is required from the beginning of fungal metabolism.
We attempted to transfer CHV1 from the CHV1-infected hypovirulent UEP1 strain to the heterokaryon using hyphal fusion. After the successive transfer of hyphal tips on selective medium, the presence of CHV1 was confirmed and the resulting heterokaryon containing CHV1 was used for further analyses. The CHV1-containing heterokaryon displayed viral symptoms of reduced pigmentation and conidiation similar to UEP1 (Fig. 3d). Compared to the hypovirulent UEP1 strain, no significant changes in CHV1 accumulation were observed in the CHV1-infected heterokaryon. functional analyses of the lethal CpRbp1 gene using heterokaryons. Functional analyses of lethal genes in filamentous fungi has been hampered due to limited sources of conditional mutants and tightly regulated gene expression systems. Since heterokaryon is a unique feature of filamentous fungi, allowing the maintenance of lethal genotypic nuclei due to the dominance of wild-type nuclei, we applied various chimeric structures of the CpRbp1 gene to transform the heterokaryon and evaluate whether, and how efficiently, the mutated versions could complement the CpRbp1-null allele (Fig. 4a-h). Various complementing vectors with or without conserved representative domains such as the NLS, RBD, and coiled-coil region were used to transform heterokaryon and the resulting complementing progeny were analyzed (Table 1 and Fig. 4i). At least 20 geneticin-resistant transformants were selected for each construct. Single-spore analyses of each complementing transformant was conducted using PDAmb, PDAmb supplemented with hygromycin B, and PDAmb supplemented with geneticin. All but one mutated construct failed to complement the CpRbp1-null nuclei; conidia either did not grow on any media, or grew on PDAmb and/or PDAmb supplemented with geneticin but not on PDAmb supplemented with hygromycin B. In addition, no conidial progeny grew on both hygromycin B-and geneticin-supplemented media. These results clearly indicate that protoplasts for complementation were heterokaryotic, and the single-spored transformants showing geneticin resistance contained wild-type nuclei transformed by the corresponding complementing vectors. Thus, the presence of intact NLS, RBD, and coiled-coil regions was necessary for the function of the CpRbp1 gene. However, the mutated CpRbp1 gene construct missing only the 20 N-terminal amino acid residues prior to the NLS resulted in transformants whose conidia were able to grow on PDAmb supplemented with hygromycin B and PDAmb supplemented with both hygromycin B and geneticin. These results indicate that this CpRbp1 variant was able to complement the CpRbp1-null nuclei, suggesting that the missing amino acid residues were not necessary for function of the CpRbp1 gene. As a comparison, the RanBP1 ortholog from www.nature.com/scientificreports www.nature.com/scientificreports/ A. nidulans was found to complement CpRbp1-null nuclei, indicating the existence of functional conservation among fungal RanBP1s.

Discussion
Ran plays important roles in many fundamental cellular functions, which are achieved by various effectors. Among several RanBPs, RanBP1 functions as a crucial cofactor to stimulate RanGAP activities by dissociating RanGEF. Although significant progress has been achieved in clarifying the regulatory functions of RanBPs in animals and yeasts, very limited studies have been conducted on the function of RanBPs in filamentous fungi.
Analyses of the deduced amino acid sequence of a cloned CpRBP1 revealed the presence of conserved motifs within the RBD, including KXRAKLXRF, WKERGTGXXXXLXHK, and RXXMRRXKTLKXCANH (where X is any amino acid). However, RanBP1s from higher organisms show differences from their S. cerevisiae and Schizosaccharomyces pombe counterparts in both the sequence and the location of an extension outside of the conserved RBD. A classic monopartite NLS [K(K/R)X(K/R)] was found at residues 36 to 39, which is outside of the RBD and is also found in other filamentous fungi such as A. nidulans, C. fioriniae, C. nymphaeae, C. simmondsii, F. graminearum, M. rileyi, and U. virens. Interestingly, RanBP1s from C. parasitica and the evolutionarily related fungi mentioned above, excepting A. nidulans, showed amino acid extensions at their C-terminus, and CpRBP1 harbors a unique motif of a coiled-coil domain, which is known for a variety of important interactions 31 . Although the function of Ran may largely rely upon RanBP as its regulatory factor, RanBPs have been known to have some different functions depending on the organism, for example, RanBP1 is essential for viability in S. cerevisiae but not in mice 32 , and Nup2 is essential for viability in A. nidulans but not in S. cerevisiae 20 . In addition, different functions of RanBPs within the same organism have been demonstrated in S. cerevisiae, where Yrb1 was essential for cell viability, while Yrb2 and Nup2 had little or no apparent effects on growth. Moreover, new functions, independent of binding to Ran, extend the function of RanBP beyond its well-known role as a Ran effector [33][34][35] . Thus, differences in the functional domains of CpRBP1 strongly suggest that it may have several functions that differ from the founding member, Yrb1.
Sequence homology showed that the CpRbp1 gene is highly similar to essential Yrb1 [E-value, 3e-75; amino acid identity, 145/238 (67%)], although not as similar as to RanBP genes from filamentous fungi, and function analyses of CpRbp1 using heterokaryons proved that the CpRbp1 gene is essential for cell viability. In addition, compared to other lethal genes exhibiting the terminal phenotypes of mutated genes 28,30 , no sign of initial germinating processes such as swollen spores was observed in conidia containing CpRbp1-null nuclei, suggesting that the CpRbp1 gene product is involved in an initial fundamental cellular function that regulates many important processes for viability from the beginning.
Heterokaryosis, defined as the presence of two or more genetically different nuclei in a common cytoplasm, is a unique characteristic of a coenocytic fungus, in which cells contain multiple nuclei in a shared cytoplasm. The two most common ways of generating heterokaryon include mutation of any of the nuclei or mycelial fusion. Due to the coenocytic characteristics, mutated nuclei with recessive genotype can proliferate along with the wild-type nuclei. The heterokaryon characteristic of maintaining genetically different nuclei allows proliferation of nuclei with lethal genotypes, allowing further functional analyses. The heterokaryon rescue technique has been applied in A. nidulans to determine whether a gene of interest was essential for viability and to further analyze the functions of the essential gene by observing its terminal phenotypes 25,26 . However, this application is limited in other fungi and generalization of the technique to other fungal systems would require further studies of other organisms. Although the extent of heterokaryon of C. parasitica in natural environments is largely unknown, the presence of stable heterokaryon has been reported in nature 27 , and well-balanced maintenance of genetically engineered-lethal nuclei has been achieved 28,30 . Therefore, C. parasitica appears to be a good model fungus for studying essential genes based on the ease of detection and single-spore resolution, and more importantly, the balanced proliferation of genetically different nuclei, which is necessary for functional analyses using heterokaryon.
In eukaryotes, a larger number of small G proteins have been identified, and implicated in a wide range of cellular functions 14 . Numerous upstream regulators and downstream effectors of small G protein have gradually been characterized for their modes of actions 14 . Considering that signaling and biological functions of small G proteins with their effectors are conserved, the investigation of essential genes using fungal systems is intriguing. Not many studies on RanBP1 have been conducted in fungi other than yeast. Highly conserved motifs as www.nature.com/scientificreports www.nature.com/scientificreports/ well as unique features were identified by in silico analysis of the deduced amino acid sequence of the cloned CpRbp1 gene. Although some RanBPs have already been identified as essential for viability, the occurrence of forced heterokaryon in C. parasitica facilitated our analyses of the function of a possible essential gene without the uncertainty of obtaining the corresponding mutant. Thus, the presence of heterokaryon and the following breakdown of the heterokaryotic state verified that the CpRbp1 gene is essential. In addition, a complementation assay using heterokaryon confirmed that defined domain structures such as the N-terminal NLS, the RBD, and the C-terminal coiled-coil region of the CpRbp1 gene were necessary for proper function of the CpRbp1 gene. However, interestingly, although the coiled-coil region of the CpRbp1 gene might be necessary for complementation of the CpRbp1-null nuclei, the ortholog from A. nidulans, which does not contain the coiled-coil region at the C-terminus, also complemented the CpRbp1-null nuclei. Although further studies are required to understand the structure and function of the coiled-coil in the C-terminal extension, homologous coiled-coil regions are important for adopting distinct conformations in the GTP-and GDP-bound states 36,37 . These results suggest that the coiled-coil region might be important for the intramolecular conformation, but not all RanBP1s require the presence of this structural motif. In addition, a fragment containing only the RBD of RanBP1 was enough to functionally complement a defective Yrb1 in yeast 38 . However our heterokaryon analyses indicated that other functional flanking regions were required. These results suggest that, although the sequence conservation is known among all RBDs from yeasts, plants, mammals, and other vertebrates, there are differences in the function of RanBP1 depending on the organism and the gene itself.
The Rbp1 ortholog from A. nidulans complemented the deletion phenotype suggesting the functional conservation of RanBP1. This functional conservation of RanBP1 is promising and suggests that our heterokaryon analysis system of C. parasitica can be extended to analyze the biological functions of other heterologous genes as long as there are related fungal counterparts.
RanBP is involved in viral infection, including viral transport 39,40 and host defense against viral infection 41,42 . Although CpRBP1 was affected by the presence of CHV1, no significant difference in viral symptoms or CHV1 accumulation was observed in CHV1-infected heterokaryon. These results suggest that the extent of wild-type nuclei in the heterokaryon is sufficient to reproduce viral symptoms in the fungus and support viral replication.
In this study, we investigated the biological functions of the RanBP1 gene of the chestnut blight fungus C. parasitica. Characterization of the putative CpRbp1-null mutants indicated that the mutants were heterokaryons consisting of two different types of nuclei carrying either the wild-type or the CpRbp1-null mutant allele in the common cytoplasm. In addition, single-spore resolution of the heterokaryotic CpRbp1-null mutants confirmed that the CpRbp1 gene is essential. Moreover, complementation of heterokaryon using various constructs of the CpRbp1 gene allowed us to determine that the conserved structural motifs including NLS, RBD, and coiled-coil were necessary for the function of the protein product of the CpRbp1 gene. Functional conservation of the essential gene such as the CpRbp1 gene further potentiate the application of heterokaryon to analyze the structure-function relationship of a lethal gene using this fungus.

Methods
Fungal strains and growth. C. parasitica strain EP155/2 (ATCC 38755) was maintained on PDAmb under constant low light at 25 °C (ref 43 .). For liquid culture of C. parasitica, EP complete medium was used 29 . The culture conditions and methods used to prepare the primary inoculum for liquid cultures have previously been described 43 . The mycelium was collected and lyophilized until use, as described previously 44 . Cloning and characterization of a RanBP1-like gene, CpRbp1. Proteomic analyses of previous studies revealed the amino acid sequences of selected protein spots. Search for the C. parasitica genome database (http://genome.jgi-psf.org/Crypa2/Crypa2.home.html) with the determined amino acid sequence identified the corresponding gene encoding an ortholog of the yeast RanBP, Yrb1. PCR amplification of genomic DNA using the primers CpRbp1-gF1 (forward) and CpRbp1-gR1 (reverse) was conducted to obtain a near full-length gene. The resulting 5,069-bp PCR amplicon was cloned and sequenced.
The cDNA clone of CpRbp1 was obtained using PCR using reverse transcriptase (RT-PCR) with the primers CpRbp1-cF1 (forward) and CpRbp1-cR1 (reverse). The resulting 717-bp cDNA amplicon was cloned and sequenced.
The primers used to clone the CpRbp1 gene and the various vector constructs for functional analyses of the gene are listed in Supplementary Table. 1.

Southern and Northern blot analyses.
Genomic DNA from C. parasitica was extracted according to a previously described method 45 . DNA (10 μg) was used to digest with the appropriate restriction enzymes, immobilized on a nylon membrane, and hybridized with radioactive-labeled probes.
RNA extraction from liquid and plate cultures and Northern blot analyses were conducted as previously described 6,46 . The level of gpd transcript was used as an internal control for gene expression of C. parasitica 47 . Construction of a replacement vector and fungal transformation. The gene replacement cassettes, which were applied to examine the biological effects of the CpRbp1 gene, were constructed as described for split-marker deletion cassettes 48 . Two molecular DNA cassettes, each of which contained a part of the hph fused to either the 926-bp 5′ flanking region or 1,004-bp 3′ flanking region of CpRbp1, were prepared by overlap PCR 49 as follows: A 954-bp PCR amplicon containing the 926-bp 5′ flanking region of CpRbp1 was amplified using gene-specific primers Rbp1-F1 and Rbp1-R1 (Supplementary Table. 1). A 1,610-bp fragment of a selection marker gene containing the promoter and part of the N-terminus was amplified using primers Hph-F1 and Hph-R1 (Supplementary Table.  www.nature.com/scientificreports www.nature.com/scientificreports/ 1,594-bp fragment of the selection marker gene containing the terminator and part of the C-terminus were amplified and fused using primer pairs Rbp1-F2/Rbp1-R2, Hph-F2/ Hph-R2, and Hph-F2/Rbp1-R2 (Supplementary  Table. 1). The resulting molecular cassettes were then used simultaneously to transform protoplasts of C. parasitica EP155/2 strain.
Functional complementation of the CpRbp1-null mutant was performed using the wild-type allele. The complementing vector pCRbp1 was constructed by insertimg a 4,615-bp NotI fragment of pSilent-Dual1G (pSD1G) containing the geneticin resistance cassette 50 into NotI-digested pRbp1 carrying a 5,559-bp fragment with the full-length CpRbp1 gene. The resulting vector was then used to transform the putative CpRbp1-null mutant. For functional analyses of the CpRbp1 gene, various chimeric structures of the CpRbp1 gene were constructed using the cDNA clone of the CpRbp1 gene and the constitutive expression cassette 51 . In addition, cDNAs of heterologous Rbp1 genes from A. nidulans (GenBank accession No XP_657688.1) was cloned and used for complementation.
Protoplast preparation and transformation were performed as previously described 43,45 . For in trans complementation, protoplasts of the putative CpRbp1-null mutant were obtained from young mycelial cultures grown in liquid medium inoculated with mycelial fragments instead of spores. Transformants were selected from agar plates supplemented with 150 μg/mL hygromycin B (Calbiochem, San Diego, CA, USA) or 150 μg/mL geneticin (Invitrogen, Carlsbad, CA) for replacement or complementation of CpRbp1, respectively. Cultures were passaged three to four times on selective media, and single-spores were isolated when possible, as previously described 30 . PCR and Southern blot analyses were conducted to confirm replacement and in trans complementation of the CpRbp1 gene.
To analyze variation in the ratio of different types of nuclei in heterokaryons, the putative heterokaryotic transformant was cultured and successively transferred every fifth day for more than 3 months on nonselective and selective (containing hygromycin B) PDAmb media. Then, conidia were collected from plates representing every fifth transfer, and 100 conidia were spread on PDAmb plates with or without hygromycin B. The resulting CFUs were counted to determine the numbers of wild-type nuclei and CpRbp1-null mutant nuclei.

Quantitative analyses of transcript accumulation by real-time RT-PCR.
To examine the expression levels of target and internal control genes, quantitative RT-PCR (qRT-PCR) was performed using a GeneAmp 7500 sequence detection system (Applied Biosystems, Foster City, CA, USA) and a SYBR green mixture RT kit (Applied Biosystems) as previously described 52 . Analyses were conducted in triplicate for each transcript, from at least two independent RNA preparations of the same sample, using primers specific for gpd (RT-gpd-F1 and RT-gpd-R1) and CpRbp1 (RT-Rbp1-F1 and RT-Rbp1-R1) (Supplementary Table. 1). Transcript abundance relative to the amount of gpd was analyzed using the 2 −ΔΔCT method 53 .
Accumulation of CpRbp1 transcripts in the wild-type and hypovirulent UEP1 strains, with or without TA supplementation was analyzed by Student's t-test at p < 0.01.