Species-Specific Detection of Mycosphaerella polygoni-cuspidati as a Biological Control Agent for Fallopia japonica by PCR Assay

The ascomycete fungus Mycosphaerella polygoni-cuspidati has been undergoing evaluation as a potential classical biological control agent for the invasive weed Fallopia japonica (Japanese knotweed), which has become troublesome in Europe and North America. In advance of the potential release of a biocontrol agent into a new environment, it is crucial to develop an effective monitoring system to enable the evaluation of agent establishment and dispersal within the target host population, as well as any potential attacks on non-target species. Therefore, a primer pair was designed for direct, rapid, and specific detection of the Japanese knotweed pathogen M. polygoni-cuspidati based on the sequences of the internal transcribed spacer regions including the 5.8S rDNA. A PCR product of approximately 298 bp was obtained only when the DNA extracted from mycelial fragments of M. polygoni-cuspidati was used. The lower limit of detection of the PCR method was 100 fg of genomic DNA. Using the specific primer pair, M. polygoni-cuspidati could be detected from both naturally and artificially infected Japanese knotweed plants. No amplification was observed for other Mycosphaerella spp. or fungal endophytes isolated from F. japonica. The designed primer pair is thus effective for the specific detection of M. polygoni-cuspidati in planta.


Introduction
Japanese knotweed (Fallopia japonica) is a herbaceous perennial plant that originates from Japan and has become highly invasive and problematic in many parts of Europe and North America. This invasiveness often poses significant costs to biodiversity and to urban development [1]. Current control methods based on chemical applications and mechanical control are only effective in the short term as the plant has an extensive and resilient rhizome system [2]. In addition, the use of chemical application is restricted in riparian habitats predominantly invaded by Japanese knotweed, while mechanical control easily causes fragmentation of the rhizome, thereby further spreading through contaminated soil. In the UK, the cost of Japanese knotweed control has been estimated in excess of £1.5 billion ($2.1 billion) [3]. For those reasons, classical biological control, which is a more sustainable, economically viable and long-term strategy based on the use of highly specific and damaging natural enemies from the native range of the invasive species, is being considered part of an integrated management approach for this troublesome weed in its introduced range. Survey work has been previously conducted in Japan and selected arthropods and plant fungal pathogens associated with Japanese knotweed have been collected and assessed for their potential as classical biocontrol agents [4][5][6]. In 2011, the psyllid Aphalara itadori has been released in selected habitats as the first classical biological control agent of F. japonica following its extensive evaluation and the approval for release by the national authorities in 2010. The control efficacy of this agent against the weed in the field is currently under evaluation.
The hemibiotrophic leaf spot pathogen, Mycosphaerella polygoni-cuspidati, attacking Japanese knotweed in Japan, has been undergoing evaluation as an additional biological control agent for F. japonica, based on its apparent field host specificity and its severe impact on the weed in both the field as well as in greenhouse experiments [7].
From both a scientific and a regulatory point of view, it is crucial to monitor the fate and behaviour of any classical biological control agent once it has been released into the new environment. The same would apply to the leaf spot pathogen M. polygoni-cuspidati enabling an assessment of its establishment and impact on its invasive host Japanese knotweed as well as its dispersal within and between host populations. Furthermore, any potential attacks on nontarget species near release sites would need to be rapidly detected to prompt required intervention. Tracking of the leaf spot pathogen can be undertaken using classical microbiological methods of fungal isolation from Japanese knotweed and non-target plant species; however, molecular techniques developed for a specific pathogen would considerably aid the detection. Thus, there are a need and an opportunity to develop a rapid, sensitive and accurate monitoring method that can detect M. polygoni-cuspidati and distinguish the fungus from the general microbial flora, such as endophytes present in the leaves of F. japonica [8].
In addition, such a method can track the population dynamic of the pathogen over time.
Polymerase chain reaction (PCR) is a rapid and specific method, and its high sensitivity allows detection of target DNA in a complex mixture, offering an alternative to classical methods such as isolation or spore observation [9]. The development of molecular markers is required in order to detect and to exploit DNA polymorphism. The ribosomal internal transcribed spacer regions including the 5.8S gene of rDNA (rDNA-ITS) is suitable and extensively used for developing PCR primers specific to the target organism. The rDNA-ITS region has also a high copy number in the genome indicating the increase of the sensitivity in PCR amplifications. This region is present in conserved and variable regions of the genome [10,11]. Comparison of the nucleotide sequences of the rDNA-ITS region has proved useful in delimiting and differentiating species [12][13][14]. Differences of the rDNA-ITS region between species have been used to develop species-specific primers, and this has become a common approach in molecular identification strategies [15]. With respect to plant pathogenic fungi, species-specific primers for Mycosphaerella species have already been developed for the detection of M. fijiensis and M. musicola on banana [16][17][18], as well as M. cryptica, M. lateralis, M. marksii, M. nubilosa and M. parva on Eucalyptus globulus [19].
In order to enable future tracking of M. polygoni-cuspidati in its new environment and to differentiate the pathogen from other fungal species, a species-specific primer pair for the identification and detection of M. polygoni-cuspidati was developed, and the sensitivity and specificity of this primer pair were assessed using diseased F. japonica leaves.

Fungal Isolates
Eight different isolates of M. polygoni-cuspidati, which were previously isolated from lesions on diseased Japanese knotweed leaves sampled from various locations in Japan [6] and maintained on plugs of potato carrot agar (PCA; decoction of 20 g potato and 20 g carrot, and 20 g agar/l) under sterile distilled water, were selected (Table 1). For the specificity analysis of the primers designed in this research, M. shimabarensis isolated from F. japonica in Japan as an endophyte, and seven isolates of other Mycosphaerella species were investigated. These included M. berkeleyi, M. chrysanthemi, M. delegatensis, M. graminicola, M. macrospora, M. mori and M. praecox. These species isolated from various hosts worldwide were obtained from the Genetic Resources Collection, CABI (Egham, UK) (Table 1). Moreover, 23 fungal isolates belonging to 16 additional genera were included in the tests. These genera comprised Alternaria, Annulohypoxylon, Aureobasidium, Bionectria, Biscogniauxia, Botryosphaeria, Cladosporium, Colletotrichum, Cynanchum, Irpex, Nemania, Nigrospora, Pestalotiopsis, Phoma, Phomopsis and Xylaria, and were previously isolated as endophytes from F. japonica in Japan. All genera were identified based on morphological criteria and sequence analysis using the rDNA-ITS region ( Table 1). The isolates of all respective fungal endophytes have been maintained on PCA plugs under the sterile distilled water at room temperature in the Laboratory of Plant Pathology, Kyushu University, Fukuoka, Japan.

Genomic DNA Extraction, PCR Amplification and Sequencing
Total genomic DNA was extracted from 14-day-old cultures of the respective fungal isolates grown on PCA with the DNeasy Plant Mini Kit (Qiagen, Valencia, USA) according to the manufacturer's instructions. The DNA samples were diluted to the concentration of 10 ng/ll and stored at -20°C until use.
The rDNA-ITS region were amplified using the universal primers ITS1 (5 0 -TCCGTAGGTGAACCTGCGG-3 0 ) and ITS4 (5 0 -TCCTCCGCTTATTGATATGC-3 0 ) [20]. PCR reactions were performed in reaction volumes of 25 ll containing 10 ng of genomic DNA templates, 19 PCR Buffer (Toyobo, Osaka, Japan), 0.2 mM of dNTPs, 0.2 lM of each primer and 1.25 U Blend Taq (Toyobo). DNA amplification was performed in a MyCycler Thermal Cycler (Bio-Rad, Hercules, USA). The PCR profile consisted of denaturation at 96°C for 5 min, followed by 36 cycles of 94°C for 45 s, 50°C for 30 s, and 72°C for 90 s. The PCR-amplified DNA fragments were fractionated in 2.0 % (w/v) agarose gels using 0.5 % (v/v) Tris-Acetate-EDTA buffer, and visualised by ethidium bromide staining and UV illumination. PCR products were purified and sequenced as described previously [6]. A newly generated sequence was deposited in DDBJ/EMBL/ GenBank under the accession number LC146384.

Specificity of Primers and Sensitivity of PCR
The specificity of the primer pair was evaluated with genomic DNA extracted from five isolates of M. polygonicuspidati, eight isolates of other Mycosphaerella species and an additional 23 different fungal species shown in Table 1. Before the specificity of the primer pair was tested, all genomic DNA were amplified with the universal primer pair ITS1 and ITS4 in order to avoid false negative results with M. polygoni-cuspidati-specific primers. The lower limit of specificity for PCR amplification using the primer pair MP-F2 and MP-R1 was determined by testing a dilution series of DNA concentration from 10 ng/ml to 1 fg/ml. PCR reactions were conducted as previously described with M. polygoni-cuspidati-specific primers.

Detection of M. polygoni-cuspidati in Inoculated Diseased Leaves and Naturally Diseased Leaves
Inoculated leaves showing typical leaf spot symptoms (Fig. 1a) were harvested one month after inoculation. Approximately 20 mg of fresh leaf pieces cut from the lesions was surface sterilised by immersion in 70 % ethanol for 30 s, followed by immersion in 1.4 % sodium hypochlorite solution for 5 min. Leaf pieces were subsequently rinsed three times with sterile distilled water to avoid any DNA extraction from mycelial inoculum remaining on inoculated leaf surfaces. Leaf pieces from healthy Japanese knotweed plants grown in a glasshouse were used as negative control. Naturally diseased F. japonica leaves displaying leaf spot symptoms caused by M. polygoni-cuspidati were collected from a field site in Omura, Nagasaki Pref., Japan, in 2015 (Fig. 1b). After drying, approximately 10 mg of leaf pieces were segmented from asymptomatic and symptomatic areas of the diseased leaves. DNA was extracted from artificially and naturally infected leaves using DNeasy Plant Mini Kit according to the protocol. PCR amplification was performed with the designed primers following the protocol described above. DNA extracted from the selected M. polygoni-cuspidati isolate (IMI 395028) grown in vitro served as positive control.

Design of M. polygoni-cuspidati-specific Primers
The universal primers ITS1 and ITS4 amplified an approximately 500-bp fragment of genomic DNA on M.
polygoni-cuspidati isolates. The sequence identity was 100 % for all the assessed isolates. By contrast, these M. polygoni-cuspidati isolates showed less similarity with other fungal species. Out of these other species, the sequence of M. sumatrensis was the one most similar to that of the M. polygoni-cuspidati isolates with 97 % identity. On the basis of alignment of the sequences of the rDNA-ITS region with other fungal isolates of 57 Mycosphaerellaceae obtained from GenBank, the primer pair MP-F2/MP-R1 was designed for specific amplification. These primer sequences shared 100 % identity only with sequences of M. polygoni-cuspidati. Amplification of DNA from M. polygoni-cuspidati with the primer pair MP-F2/MP-R1 produced a PCR product of 298 bp. The PCR product was sequenced and had the same sequence as the original sequence.

Specificity and Sensitivity of the Designed Primers
All isolates tested had a positive PCR amplification using the universal primers ITS1 and ITS4 (Table 1). PCR products of 298 bp amplified with the specific primers MP-F2 and MP-R1 were obtained only when DNA was extracted from isolates of M. polygoni-cuspidati obtained from various field sites in Japan (Table 1, Fig. 2). No amplification was observed for other fungal species tested, including other Mycosphaerella spp. and fungal endophytes isolated from F. japonica. The lowest limit of detection of the PCR method determined for M. polygonicuspidati (IMI 395028) was established as 100 fg of total DNA (Fig. 3).

Detection of M. polygoni-cuspidati in Infected Japanese knotweed Leaves
The primer pair successfully detected M. polygoni-cuspidati from lesions of both artificially and naturally diseased leaves (Fig. 4). A weak band was also observed at 298 bp on some of the DNA samples extracted from asymptomatic areas of diseased leaves. The specific PCR product was sequenced to confirm that the designed primers have amplified the expected consensus region of the target organism. No PCR products were amplified with DNA from healthy leaves.

Discussion
In the present study, we developed species-specific PCR primers for the identification of M. polygoni-cuspidati, a potential biological control agent for F. japonica, based on sequence information of the fungus from the rDNA-ITS region. The primer pair MP-F2 and MP-R1 amplified a single fragment of approximately 298 bp in size only with DNA extracted from mycelia of M. polygoni-cuspidati cultured on PCA. This primer pair was also used to detect the presence of M. polygoni-cuspidati in infected Japanese knotweed leaves. These results confirm that this primer pair is suitable for the specific detection of M. polygoni-cuspidati and thus for monitoring the presence of this fungus in F. japonica field populations after a potential future release into a new environment.
To establish a species-specific PCR detection method with high accuracy, designing primers based on ubiquitously conserved known genes with sequence variation are of great importance. In this study, sequences of the rDNA-ITS region were selected as a target region because this region of the genome evolves fast, may vary among species within a genus or even among species populations [21][22][23][24][25] and is present in many copies in fungal genomes [26,27]. In addition, a large amount of the rDNA-ITS sequence data provided by public databases can be used for comparison of sequences of M. polygoni-cuspidati with other Mycosphaerella species, which facilitates the design of  pair also successfully detected the pathogen through the direct amplification of DNA from F. japonica leaves, both naturally and artificially infected with M. polygoni-cuspidati. Although naturally diseased Japanese knotweed leaves are also colonised by a range of endophytes, the designed primer pair MP-F2/MP-R1 amplified no PCR product from DNA extracted from any of these endophytic fungal species, which further demonstrates its specificity.
The detection sensitivity with the primer set MP-F2/MP-R1 was 100 fg of genomic DNA. This high sensitive detection is probably due to the high copy number of rDNA genes in any genome. Furthermore, touchdown PCR used in the present study resulted in higher sensitivity with the primer compared to normal PCR with consistent annealing temperature (data not shown). This is more sensitive than the previous reports including that the detection limit of M. graminicola or M. parva is as low as 10 pg [19,28]. In addition, the primers could detect M. polygoni-cuspidati from 1 pg of genomic DNA extracted from the lesions (data not shown), which showed sufficient and higher sensitivity than that shown in other report [19]. Amplicons were also successfully detected from asymptomatic areas of the diseased leaves. A difference in the detection sensitivity with the primer pair was observed between DNA extracted from symptomatic and asymptomatic areas of the diseased leaves (Fig. 4). The most likely reason for this is the higher proportion of target DNA in relation to the total DNA extracted from developed leaf lesions. Moreover, one of the samples extracted from asymptomatic areas of the diseased leaves showed higher sensitivity compared to the other two samples (Fig. 4). It suggests that the samples showing higher sensitivity may be cut more close to the symptomatic area compared to the other samples. Because the PCR assay is successful in detecting M. polygonicuspidati in infected leaves, it may also be suitable to research pathogen movement with the host plant. In order to track the disease development of the pathogen in the latent infection period, more information about the PCR detection from asymptomatic F. japonica leaves at the earlier stages after artificial infection would be required.
After a potential future release of M. polygoni-cuspidati into a new environment, it would be important to track the fungus in order to study its dispersal within and between populations of F. japonica. Using the primer pair MP-F2/ MP-R1 with fungal genomic DNA extracted from symptomatic and asymptomatic areas of leaves collected in the field as the template, M. polygoni-cuspidati was specifically detected, indicating that this primer pair can be useful for monitoring the presence of the fungus in the field. To our knowledge, there are no reports about developments of species-specific detection for foliar plant pathogens used as weed biological control agents. For the majority of such fungal pathogens, this detection method may not be applicable due to their easy identification in the field and unproblematic re-isolation from diseased plant material. In the case of M. polygoni-cuspidati, however, an early and quick confirmation of infection of the target weed or potentially any non-target species will be required to evaluate establishment and spread of the pathogen as well as to assess potential risks to non-target species. Re-isolation of the fungus from infected leaves and subsequent identification will be time consuming due to the slow growth rate of M. polygoni-cuspidati in vitro. Moreover, the lack of ascospore formation on artificial media and on artificially inoculated diseased leaves does not easily support the identification of the pathogen based on morphological characteristics. The rapid and accurate method of M. polygoni-cuspidati-specific PCR which we established will help overcome these limitations. In order to ensure this method is reliable in tracking the pathogen in a new environment, it will also be necessary to prove the specificity by using DNA extracted from leaves of F. japonica plant growing in the UK because the endophytic mycobiota differs from the one present in Japanese knotweed in Japan.
In conclusion, our research has successfully designed a rapid, sensitive and accurate method for the specific detection of M. polygoni-cuspidati from DNA extracted from pure cultures as well as diseased Japanese knotweed leaves by PCR assay. Should at any point in the future M. polygonicuspidati be released into a new environment as a classical biological control agent for F. japonica, this reliable detection assay will be a useful tool to monitor the establishment and spread of this pathogen in the introduced range.