Site-specific DNA double-strand break generated by I-SceI endonuclease enhances ectopic homologous recombination in Pyricularia oryzae



To evaluate the contribution of DNA double-strand breaks (DSBs) to somatic homologous recombination (HR) in Pyricularia oryzae, we established a novel detection/selection system of DSBs-mediated ectopic HR. This system consists of donor and recipient nonfunctional yellow fluorescent protein (YFP)/blasticidin S deaminase (BSD) fusion genes and the yeast endonuclease I-SceI gene as a recipient-specific DSB inducer. The system enables to detect and select ectopic HR events by the restoration of YFP fluorescence and blasticidin S resistance. The transformed lines with donor and recipient showed low frequencies of endogenous ectopic HR (> 2.1%). Compared with spontaneous HR, c. 20-fold increases in HR and absolute frequency of HR as high as 40% were obtained by integration of I-SceI gene, indicating that I-SceI-mediated DSB was efficiently repaired via ectopic HR. Furthermore, to validate the impact of DSB on targeted gene replacement (TGR), the transformed lines with a recipient gene were transfected with an exogenous donor plasmid in combination with the DSB inducer. TGR events were not observed without the DSB inducer, whereas hundreds of colonies resulting from TGR events were obtained with the DSB inducer. These results clearly demonstrated that the introduction of site-specific DSB promotes ectopic HR repair in P. oryzae.


Rice blast caused by Pyricularia oryzae Sacc. [synonym, Magnaporthe oryzae (Hebert) Barr] is one of the most economically devastating fungal diseases of crops worldwide. Pyricularia oryzae can break developed resistant rice cultivars with resistance (R) genes within a very short time after the introduction of new resistance cultivars (Kiyosawa, 1982). The frequent appearance of new races (or pathotypes) has been attributed to the rapid adaptation of P. oryzae through the loss of the function of avirulence (AVR) genes corresponding to R genes. The loss of AVR function includes point mutations, insertions of transposable elements, and deletions of entire genes (Orbach et al., 2000; Kang et al., 2001; Zhou et al., 2007; De Wit et al., 2009). In addition, AVR gene deletions, duplications, and translocations have been found in the genome of field isolates of P. oryzae (Khang et al., 2008; Yoshida et al., 2009; Chuma et al., 2011). These genetic variations and the whole-genome study suggest that somatic homologous recombination (HR) is one of the mechanisms behind the loss of AVR function (Dean et al., 2005; Miki et al., 2009; Sone et al., 2013). Somatic HR can be mutagenic if the template is similar but not identical to the broken sequence: ectopic HR (Johnson & Jasin, 2000).

DNA double-strand breaks (DSBs) occur as a result of exogenous or endogenous DNA damage caused by DNA-damaging agents such as reactive oxygen species throughout the life cycle of organisms (Daley et al., 2005). In eukaryotic cells, DSBs can be repaired by two pathways: HR and nonhomologous DNA end joining (NHEJ). We previously reported a construction of a detection/selection system of ectopic HR at the somatic cell level of P. oryzae using two nonfunctional enhanced yellow fluorescence protein (YFP) and blasticidine S deaminase (BSD) fusion genes and have revealed that ectopic HR consistently occurs in mycelial growth, albeit at a low frequency (Arazoe et al., 2013). In connection with this observation, spontaneously generated DSB sites were visualized by EGFP-Rhm51 in the asexual life cycle of P. oryzae (Ndindeng et al., 2010). However, the role of genomic DSBs in HR and their repair mechanisms in filamentous fungi including P. oryzae has remained obscure.

The yeast mitochondrial I-SceI endonuclease, which has an 18-bp recognition site (Colleaux et al., 1988), has been applied to introduce specific DSBs in many organisms (Fairhead & Dujon, 1993; Rouet et al., 1994; Puchta et al., 1996). In this study, we introduced the I-SceI-mediated DSB into an ectopic HR detection/selection system and evaluated the effect of DSB in the frequency of ectopic HR repair in P. oryzae.

Materials and methods

Fungal strains, media, DNA manipulations, and molecular analyses

The blast fungus P. oryzae strain Hoku-1, a Japanese pathogenic strain maintained in our laboratory, was used in this study. Genomic DNA extraction and isolation of protoplasts were performed as previously described (Arazoe et al., 2013). For sporulation, mycelia were grown on oatmeal agar plates for 7 days. For PCR amplification, KOD-plus-neo polymerase (TOYOBO, Japan) was used following the manufacturer's instruction.

Construction of ectopic HR detection/selection substrates and transformation

The HR substrates were constructed from the YFP::BSD gene (Arazoe et al., 2013). To construct ISTG-YFP::BSD, an HR substrate vector, we introduced the I-SceI recognition site with two stop codons into the YFP::BSD gene and inserted the nonfunctional YFP::BSD gene into the backbone vector containing hph cassette (Fig. 1). RS-YFP::BSD, the other HR substrate vector, as the homologous template for HR repair, was used as described previously (Arazoe et al., 2013; Fig. 1). These HR substrates were cotransformed into the P. oryzae genome by PEG transformation as reported previously (Arazoe et al., 2013).

Figure 1.

Schematic representation of the types of DNA DSB repair in the HR detection system. Two different nonfunctional YFP and BSD fusion genes (YFP::BSD), ISTG-YFP::BSD and RS-YFP::BSD, were a target of DSB induced by I-SceI and a repair substrate, respectively. In ISTG-YFP::BSD, an I-SceI recognition site containing stop codons was inserted between nucleotide positions 327 and 328 of the YFP::BSD open reading frame (indicated by an underline). In RS-YFP::BSD, translation elongation factor (tef) promoter (Ptef) region and the 18 N-terminal residues were deleted from the functional YFP::BSD construct. These nonfunctional genes were stably integrated into the genome of Pyricularia oryzae by PEG transformation. By expression of the I-SceI gene as the DSB inducer, a DSB would be introduced into the I-SceI recognition site of the ISTG-YFP::BSD gene. The break can be repaired by two pathways: NHEJ, in which 2 DNA ends are joined together regardless of their DNA sequence homology; and HR, which requires regions of homology between the donor and recipient DNAs. (a) Imprecise NHEJ: the product repaired by the joining of two ends with the deletion of ISTG-YFP::BSD sequence; (b) ectopic HR: HR repair product using RS-YFP::BSD as the homologous template, which results in a functional YFP::BSD sequence. Tgla: glucoamylase (gla) terminator (Tgla), hph: hygromycin B phosphotransferase.

To construct a vector for Agrobacterium tumefaciens-mediated transformation (ATMT), a fragment of ISTG-YFP::BSD with the hph gene cassette was transferred into pCAMBIA-Lin, which resulted in pFAG-ISTG-YFP::BSD (Fig. 4a). We used A. tumefaciens strain AGL1 (CAMBIA, Australia) harboring pFAG-ISTG-YFP::BSD for transformation. ATMT was performed as previously reported (Rho et al., 2001). The details of the cloning procedure were described as supporting information.

Construction of I-SceI expression vector

On the basis of the coding sequence of Saccharomyces cerevisiae I-SceI (GenBank Accession No. V00684.1), a synthetic gene optimized for expression in P. oryzae was created (Operon Technologies). To construct I-SceI expression vector, the I-SceI gene was transferred into the pGEM-PTEF-Tgla, which resulted in pGEM-I-SceI. In addition, we added the nuclear localization signal (NLS) from Simian virus 40 large T antigen (Khang et al., 2010) to the C-terminus of I-SceI by linker primer insertion, and bialaphos resistance gene (bar) cassette was transferred into the upstream of I-SceI cassette (Fig. 3a). The details of the cloning procedure were described as supporting information.


Construction of HR detection/selection system harboring an I-SceI recognition site

Recently, we have reported experimental evidence of the occurrence of HR in P. oryzae using a somatic HR detection/selection system consisting of two nonfunctional YFP::BSD genes (Arazoe et al., 2013). We further attempted to develop a detection and selection system for ectopic HR induced by a site-specific DSB in one of two nonfunctional YFP::BSD genes using I-SceI endonuclease. In ISTG-YFP::BSD, a recipient gene, the I-SceI recognition site including two stop codons was inserted between nucleotide positions 327 and 328 of the YFP::BSD open reading frame (Fig. 1). In RS-YFP::BSD, a donor gene for HR repair template, promoter region and the 18 N-terminal residues of YFP::BSD were deleted from the functional YFP::BSD construct (Fig. 1). pISTG-YFP::BSD containing the hph gene cassette and pRS-YFP::BSD were simultaneously integrated into the P. oryzae genome by PEG transformation. Hygromycin B-resistant mycelia were further screened to confirm the integration of both ISTG-YFP::BSD and RS-YFP::BSD by PCR. As a result, four transformed lines harboring both of the two substrates were obtained [ISTR (ISTG-YFP::BSD and RS-YFP::BSD integrated) 6, 11, 25, and 30].

Because we have reported that ectopic HR arises naturally in the genome of P. oryzae during the hyphal growth (Arazoe et al., 2013), we examined whether ISTG-YFP::BSD and RS-YFP::BSD in each ISTR line act as a detecting system of endogenous HR events in active hyphae, protoplasts from mycelia, and conidia by YFP fluorescence. YFP fluorescence was detected in all stages of P. oryzae, indicating that our constructed HR substrates can detect ectopic HR by YFP fluorescence (Fig. 2a–c). Furthermore, to evaluate whether HR between two substrates confers BS resistance, each ISTR line cultured on the PSA plates for 7 days was transferred to the PSA plates with BS. BS-resistant mycelial growth and YFP fluorescence were observed in each ISTR line (Fig. 2d), indicating that our constructed HR substrates can detect and select the cells in which endogenous HR between ISTG and RS has occurred. These results demonstrated that introduction of an I-SceI recognition sequence into the recipient gene did not affect the capability for detecting and selecting HR-affected cells without DSB inducer.

Figure 2.

Detection of YFP fluorescence and blasticidin S (BS) resistance of the cells in which HR occurred in ISTR lines without DSB inducer. Various stages of ISTR lines were examined: DIC, differential interference contrast images (a–c, left panels) and YFP, epifluorescence microscopy images (a–c, right panels). (a) Active hyphae. (b) Protoplasts from active mycelium. (c) Harvested conidia. (d) Mycelium under BS selection. BF: bright-field image (left), and YFP: epifluorescence image (right).

Expression of I-SceI in P. oryzae

We constructed an I-SceI expression vector for introduction of site-specific DSB into ISTG-YFP::BSD (Fig. 3a). The pGEM-bar-I-SceI::SvNLS vector (I-SceI+bar cassette) was transformed into the P. oryzae genome to verify the detrimental effects of the expression of I-SceI on the life cycle of P. oryzae. Compared with the wild-type Hoku-1, the transformants exhibited no obvious phenotypic changes in vegetative growth, spore formation, spore germination, and pathogenicity to susceptible rice cultivar Nipponbare (data not shown).

Figure 3.

Transformation with I-SceI, ectopic HR selection, and its detection in ISTR lines. (a) Schematic representation of an I-SceI expression vector with a bar cassette (DSB inducer). A NLS from Simian virus 40 large T antigen was inserted into the 3′ end of the I-SceI gene (indicated by a black box). (b) Ectopic HR selection. The I-SceI expression vector was integrated into the ISTR line genome by PEG transformation, and the protoplasts were plated onto the plates containing bialaphos or bialaphos/BS 24 h after transformation. Bialaphos-(left) and bialaphos-/BS-resistant colonies (right) were photographed after 5 days. Arrows indicate bialaphos-/BS-resistant colonies. (c) Ectopic HR detection. Bialaphos/BS-selected mycelium was examined using bright-field imaging (BF, left) and epifluorescence imaging (YFP, right). bar: bialaphos resistance gene.

Ectopic HR was induced by expression of I-SceI

As the expression of I-SceI brought about no apparent phenotypic changes, we further investigated whether the expression of I-SceI can introduce site-specific DSB into ISTG-YFP::BSD, and the breakage site can be repaired by ectopic HR in ISTR lines.

Either I-SceI+bar cassette or pGEM-bar (bar cassette) was introduced into the genome of the ISTR lines (Fig. 3a). Twenty-four hours after transformation, transformed protoplasts were selected by PSA plates with bialaphos or bialaphos/BS. Transformation of the protoplasts from ISTR lines with the I-SceI+bar cassette yielded colonies on PSA plates with bialaphos (Fig. 3b, left). Meanwhile, selection of the transformed protoplasts with bialaphos/BS yielded decreased colony formation as compared with bialaphos selection (Fig. 3b, right). All of the bialaphos-/BS-resistant colonies showed YFP fluorescence (Fig. 3c), indicating that bialaphos-/BS-resistant colonies have a functional YFP::BSD gene restored by ectopic HR using the homologous template, RS-YFP::BSD (Fig. 1b). In all of the ISTR lines, transformation with I-SceI+bar cassette significantly increased the numbers of bialaphos-/BS-resistant colonies compared with control vector, bar cassette (Table 1). The frequencies of ectopic HR repair obtained by expression of the I-SceI gene were c. 20-fold higher than those by expression of control bar gene in every ISTR line (Table 1). These results suggested that the transformed I-SceI gene in ISTR lines produced sufficient restriction endonuclease activity for introducing DSBs at the recognition site in ISTG-YFP::BSD, and the introduced site-specific DSBs were efficiently repaired by ectopic HR. Transformation with bar cassettes also produced a few number of bialaphos-/BS-resistant colonies having YFP fluorescence (Table 1), suggesting that the colonies were derived from the cells in which functional YFP::BSD were produced by ectopic HR repair after spontaneous DSBs during protoplast isolation and/or transformation procedures, as was also observed in ISTR lines in Fig. 2a–c.

Table 1. The frequency of ectopic HR repair with or without I-SceI gene integrated into the genome of ISTR lines
LineCassetteTotalaHR repairbFrequency (%)cMagnificationd
  1. a

    The number of bialaphos-resistant colonies indicating the total number of colonies in which the I-SceI+ bar or bar cassette was integrated.

  2. b

    The number of bialaphos/BS colonies indicating the number of colonies in which YFP::BSD gene restored by ectopic HR.

  3. c

    The frequency of ectopic HR repair was estimated using the number of bialaphos-/BS-resistant colonies vs. the number of bialaphos-resistant colonies.

  4. d

    The magnification of HR repair was calculated from the frequency of HR repair obtained in I-SceI + bar cassette divided by that in bar cassette alone.

ISTR6I-SceI + bar1846736.420.2
bar 22241.8
ISTR11I-SceI + bar2117937.424.9
bar 19631.5
ISTR25I-SceI + bar2309240.019.0
bar 24352.1
ISTR30I-SceI + bar2569938.720.4
bar 25951.9

Ectopic HR induced by transfection of DSB-inducer and exogenous donor gene

On the basis of the results described above, we hypothesized that site-specific DSB and its repair by ectopic HR may be applicable for increasing the frequency of targeted gene replacement (TGR). To test this hypothesis, we tried to develop a simple TGR evaluation system using ISTG lines in which a single insertion of an ISTG-YFP::BSD in the genome as a TGR target gene (Fig. 4a). Conidia were transformed with ISTG-YFP::BSD by ATMT, and the transformants (ISTG5, 11, 13, and 19) having a single ISTG-YFP::BSD were selected by Southern hybridization (Fig. 4b). We used an ISTG-YFP::BSD in the genome of ISTG lines as a recipient (target gene), a RS-YFP::BSD cassette as a exogenous donor, and an I-SceI cassette as a DSB inducer for TGR evaluation, the latter two of which are for transient transfection (Fig. 4c).

Figure 4.

Southern blot analysis of ISTG lines and schematic representation of a system for TGR detection. (a) Schematic representation of Agrobacterium tumefaciens-mediated transformation (ATMT) vector, pFAG-ISTG-YFP::BSD. (b) Southern blot analysis. ISTG-YFP::BSD was integrated into the genome of Pyricularia oryzae by ATMT. Then, hygromycin B-resistant transformants (ISTG lines) were isolated. Genomic DNAs were digested with SphI/KpnI and hybridized with YFP::BSD probe. Lane 1, linearized pFAG-ISTG-YFP::BSD as a positive control; lane 2, untransformed P. oryzae; lane 3, ISTG5; lane 4, ISTG6; lane 5, ISTG7; lane 6, ISTG8. Fragment sizes are given in kbp. (c) Schematic representation of a system for TGR detection. ISTG lines were cultured in YG medium for the isolation of protoplasts. Plasmid-containing I-SceI gene and plasmid-containing RS-YFP::BSD were simultaneously transfected into protoplasts by PEG treatment. The BS-resistant colonies were derived from protoplasts in which TGR had occurred by ectopic HR.

To evaluate the TGR frequency, the protoplasts from each ISTG line were transfected with an RS-YFP::BSD cassette, an I-SceI cassette, or a combination of RS-YFP::BSD and I-SceI cassette (Fig. 4c and Fig. S1, Supporting information). Twenty-four hours after transfection, protoplasts were selected on PSA plates with BS. The transfection with RS-YFP::BSD or I-SceI/RS-YFP::BSD cassettes into ISTG lines is expected to generate BS-resistant colonies as a result of TGR between the genome-integrated ISTG-YFP::BSD and the introduced exogenous RS-YFP::BSD (Fig. 4c). The transfection of the RS-YFP::BSD cassette into each ISTG line resulted in no appearance of BS-resistant colonies on the plate (Fig. 5), suggesting that the frequency of TGR in P. oryzae genome is quite low without I-SceI expression. On the other hand, about 100–200 BS-resistant colonies appeared upon an addition of the I-SceI cassette to the RS-YFP::BSD cassette (Fig. 5). These results clearly showed that the introduction of DSB generated by I-SceI leads to the emergence of colonies via ectopic HR events that were not observed without I-SceI.

Figure 5.

TGR efficiency in ISTG lines by expression of I-SceI. The numbers of colonies resistant to bialaphos obtained by transformation with bar cassette, a gene integration control, and the numbers of colonies resistant to BS obtained by transfection with either I-SceI, RS-YFP::BSD or I-SceI plus RS-YFP::BSD cassettes were counted 5 days after plating on each selection medium. Each bar represents the mean value of three independent experiments, and error bars show standard deviations. ND: colonies not detected.

When an I-SceI cassette alone was introduced into ISTG lines as a control experiment, BS-resistant colonies were observed; however, the number of colonies was much lower than those of I-SceI/RS-YFP::BSD cassettes in all ISTG lines. As for the BS-resistant colonies appeared in I-SceI transient transfection, it is conceivable that DSB into ISTG-YFP::BSD was repaired by NHEJ that included a small deletion (Fig. 1a). Due to NHEJ repair, the BS resistance and YFP fluorescence might be restored by in-frame deletions of the inserted I-SceI site including stop codons in the YFP gene.

To characterize the novel formed DNA junction of HR, we further analyzed ectopic HR repair using RS-smYFP::BSD in which silent mutations were introduced (see supporting information). BS-resistant colonies with YFP fluorescence were obtained by transient cotransfection with I-SceI cassette and RS-smYFP::BSD into protoplasts of ISTG19. The thirty BS-resistant colonies were subjected to amplification of the HR-repaired functional YFP::BSD by PCR using specific primers (see supporting information and Fig. S2). The PCR products from BS-resistant colonies were all 1179 bp in length and were cloned into pGEM-T easy vector (see supporting information). From sequencing analysis of 30 individual functional YFP::BSD genes, they contained the silent mutations of RS-smYFP::BSD ranging from c. 100 to 150 bases in the vicinity of I-SceI site corresponding to ISTG-YFP::BSD (Fig. S2). Because ISTG lines contain an single copy of ISTG-YFP::BSD, this experiment strongly supported that ectopic HR occurred between ISTG-YFP::BSD and RS-YFP::BSD, and both DNA junctions of HR events might be located in the vicinity of the I-SceI site.

These results suggested that a highly effective system of TGR in filamentous fungi could be expected by cotransfection with a combination of donor and site-specific DSB-inducer genes.


It has been suggested by many researchers that ectopic HR is a mechanism by which genomic rearrangement including the loss of avirulence occurs in P. oryzae, but there are few reports dealing with DSB repair by HR at the molecular level. This study investigated the direct interaction between DSBs and ectopic HR repair in the genome of filamentous fungi for maintaining genome integrity using two ectopic HR substrates, one of which was cleaved by I-SceI.

It is generally accepted that NHEJ is the main DSBs repair pathway in many eukaryotic cells including filamentous fungi (Pastink et al., 2001). However, our results showed that ectopic HR reached up to c. 40% by the expression of I-SceI (Table 1). This result suggested that the incidence of HR repair in filamentous fungi might be higher than those in animals and plants (Smith et al., 2007). HR is generally considered to be a precise repair, because it can repair the original sequence of the sister chromatid or another identical sequence as a repair template (Fung & Weinstock, 2011). However, ectopic HR uses the similar but not original sequence as a repair template. In P. oryzae, more than 9.7% of the genome is made up of repetitive DNAs, a significant proportion of which is derived from transposable elements (TEs; Dean et al., 2005). AVR genes and effector genes are genetically linked to many types of TE in different isolates (Orbach et al., 2000; Khang et al., 2008; Yoshida et al., 2009; Chuma et al., 2011). In a recent study, convincing evidence was obtained that the TE, Occan, could lead to the deletion of the Avr gene in the genome of P. oryzae via ectopic HR (Sone et al., 2013). The authors hypothesized that this deletion was due to spontaneous DSBs in TE. The result obtained in this study fully supports this hypothesis.

In the ascomycete Neurospora crassa, disruption of NHEJ-related genes brought about a large increase in the frequency of TGR (70–100%), suggesting that the NHEJ and HR pathways compete each other for integration of exogenous DNA fragments (Ninomiya et al., 2004; Ishibashi et al., 2006). In P. oryzae, similar effect of disruption of NHEJ on the frequency of TGR (70–100%) has been reported (Villalba et al., 2008). The mutants of NHEJ-related gene had no differences in growth and morphology from the wild type. On the other hand, decreases in growth and pathogenicity have been observed by disruption of HR-related genes (T. Sone, pers. commun.). Taken together, HR may play more important role in integrity of the genome in P. oryzae than NHEJ, although NHEJ is the major pathway of DNA repair in this fungus.

Nuclease-mediated DSBs could be powerful tool to improve TGR frequencies in eukaryotes. In our TGR experiments in a filamentous fungus P. oryzae, the TGR frequencies were dramatically increased by I-SceI-mediated DSB (Fig. 5). The use of I-SceI is restricted because the recognition sequence is required to resided in the target gene. The new technologies that can introduce DSB into a specific target have been developed. These include zinc-finger nucleases and transcription activator-like effector nucleases, which can digest DNA as a tailored site-specific endonuclease that induce DSBs at the unique target sites of the genome (Urnov et al., 2010; Li et al., 2011). These site-specific endonucleases can provide new approaches to explore the DSBs-mediated gene evolution and improve DSBs-mediated TGR.

In conclusion, we established an ectopic HR detection/selection system and DSB introducing procedure by I-SceI. We demonstrated that an introduced DSB increases the frequency of ectopic HR in the genome of P. oryzae using the system. This system might be very useful for dissecting DSBs repair mechanisms. We expect that these data and future use of the system will provide essential information for understanding of gene evolution in plant pathogens and an important step toward the development of new genetic technologies for many filamentous fungi.


The authors thank Dr. Takehiko Shibata of RIKEN Advanced Science Institute for valuable suggestions. This work was supported in part by the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for JSPS Fellows (Grant No. 24.11224) and JSPS Grant-in-Aid for Scientific Research (C) (Grant No. 24580070).