A real time loop-mediated isothermal amplification (RealAmp) assay for rapid detection of Pleurostoma richardsiae in declining olive plants

Summary. Pleurostoma richardsiae is associated with host trunk diseases, known to cause dieback, cankers and wilting of woody trees, and human infections. This fungus was isolated from wood tissues of declining olive trees and grapevines showing esca disease symptoms, in the Apulia region of Italy. Fungus detection has been based on morphological and molecular features, which are time-consuming to identify and require well-trained personnel. Improvement of Pl. richardsiae detection in olive was achieved through development of real time loop-mediated isothermal amplification targeting the intergenic spacer (IGS) region of the fungus. Specificity of the assay was confirmed using ten Pl. richardsiae strains and 36 other fungus strains of species usually isolated from declining olive trees. The achieved limit of detection was 7.5 × 10 -2 ng μL -1 of Pl. richardsiae genomic DNA. A preliminary validation of RealAmp was also performed using material from infected olive plants artificially inoculated in a greenhouse.


INTRODUCTION
General decline of olive trees, similar to olive quick decline caused by Xylella fastidiosa (Saponari et al., 2013), was observed in olive orchards in Barletta-Andria-Trani (BAT) and Foggia provinces of the Apulia region of Southern Italy, which is not in the demarcated area infected by X. fastidiosa (Commission Implementing Decision EU 2017/2352 https://eur-lex.europa.eu/ legal-content/EN/TXT/HTML/?uri=CELEX:32017D2352&from=IT). Typical symptoms were wilting of apical shoots, dieback of twigs and branches, and brown streaking under the bark of the trunks of affected trees. According to Carlucci et al. (2013), three fungal species, Pleurostoma richardsiae, Neofusicoccum parvum and Phaeoacremonium minimum, were strongly associated with these symptoms, and were diagnosed through morphological, molecular and pathogenicity tests. Phaeoacremonium spp. and Botryosphaeriaceae spp. are known to cause dieback, cankers and wilting on olive and grapevine plants (Carlucci et al., 2015c;2015b). Pleurostoma richardsiae was isolated and associated for the first time in Italy from brown wood streaking and canker of olive (Carlucci et al., 2013) and grapevine plants (Carlucci et al., 2015a), confirming its pathogenicity. Further reports have associated this pathogen with olive and grapevine plants. Pintos Varela et al. (2016), Özben et al. (2017) and Canale et al. (2019) confirmed Pl. richardsiae as a grapevine pathogen in Spain, Turkey and Brazil, while Ivic et al. (2018), Canale et al. (2019), Agustì-Brisiach et al. (2021), Lawrence et al. (2021) and Van Dyk et al. (2021) confirmed this fungus as an olive pathogen in Croatia, Brazil, Spain, California and South Africa.
Pleurostoma richardsiae is also known as an uncommon cause of human infections, usually through traumatic skin inoculations with contaminated vegetable matter, causing subcutaneous Phaeohyphomycosis (Pitrak et al., 1988;Vijaykrishna et al., 2004).
Identification of Pl. richardsiae has been based on micromorphological and cultural features (Schol-Schwarz, 1970;Domsch et al., 1980;De Hoog et al., 2000;Vijaykrishna et al., 2004). Molecular tools targeting the 5.8S rDNA gene flanking internal transcribed spacers 1 and 2 (ITS) have more recently been used (Carlucci et al., 2015c;Vijaykrishna et al., 2015;Spies et al., 2020). The polymerase chain reaction is an efficient technique for the detection of Pl. richardsiae, although the DNA extraction phase and the thermal cycle requirements for the Taq polymerase function are constraints in terms of costs and time for the application of this technique at large scale. These limitations can be overcome by the loop-mediated isothermal amplification (LAMP) assay, which relies on partial DNA extraction and isothermal polymerase to amplify the DNA (Notomi et al., 2000). LAMP is a technique that uses six primers that are complementary to eight regions of genomic DNA, and therefore ensures high specificity and allows results to be obtained more rapidly than PCR or qPCR. In addition, the isothermal character of the reaction requires only a heat block or water bath at the target temperature to successfully amplify DNA. Bst polymerase used in LAMP assays allows a full reaction to take place in less than 1 h in an isothermal environment (ranging from 60 to 70°C) depending on the type of enzyme and the optimal temperature for the primer characteristics (Notomi et al., 2000).
LAMP reaction can be detected in end point, as resolving amplified products on agarose gel electrophoresis (Notomi et al., 2000), by eye observation of precipitate (Mori et al., 2001), or employing DNA-binding dyes (Le et al., 2012) or colorimetric indicators (Parida et al., 2008;Tomita et al., 2008). End point detection strategies do not allow real-time diagnoses. They may result in false positives detections due to contamination of samples (e.g., reaction tube opening procedure). In addition, determination of the LAMP reaction by visual observation is subjective, because it is based on individual perception of colour (Bista et al., 2007). These problems can be overcome by real-time LAMP reactions (RealAmp), that can also be performed in situ using Smart-DARTTM (Diagenetix), a portable device equipped with D&A software that can analyze eight samples each time. This is a major advantage of RealAmp for the rapid detection of pathogens in plant material, in the field and at entry points and nurseries, avoiding risks of pathogen spread through the movement of infected material (Yaseen et al., 2015). This technique has also been applied in the field and laboratory for diagnoses of human and animal diseases and food safety control (Abdulmawjood et al., 2014).
Considering the importance of Pl. richardsiae as a vascular pathogen of economically important woody plant species, and the high performance offered by the RealAmp assay, the present study aimed to develop a RealAmp assay specific for Pl. richardsiae.

Fungus strains
All fungus strains used in this study are reported in Table 1. The Pl. richardsiae strains were obtained from olive or grapevine plants showing symptoms of decline. Eight of the strains were from the north of Apulia region (Southern Italy), and two were of Algerian origin. One strain each of Pl. ootheca and Pl. repens from Algeria, were also included to assess the specificity of the RealAmp. Another 36 fungal strains of Phaeoacremonium and Neofusicoccum spp. from different countries and hosts were included in this study. All fungal strains were previously identified using morphological and molecular tools.

DNA extraction and IGS amplification
Genomic DNA of each strain reported in Table 1 was extracted from 200 mg of fresh mycelium, that was grown on potato dextrose agar plates (PDA, Oxoid Ltd) at 25°C for 14 d. according to the protocol optimized for vascular fungi by Carlucci et al. (2013). DNA quality and concentrations were determined with a Beckman Coulter DU-800 UV/Vis spectrophotometer.

Design of LAMP primers
As for the design of LAMP primers for Pl. richardsiae detection, the IGS regions of Pl. richardsiae and Phaeoacremonium sequences were aligned using BioEdit version 7.0.9 (http://www.mbio.ncsu.edu/BioEdit) to find conserved regions in the target fungi. Primers were designed using PrimerExplorer v. 5 (http://primerexplorer.jp) and were checked by BLAST in the GenBank to ensure their specificity. Three sets of LAMP primers were generated, with each set made of six LAMP primers (external primers F3 and B3, internal primers FIP and BIP, and loop primers LoopF and LoopB). All primers were analyzed to check for hairpin, self-dimer, and heterodimer production, using Integrated DNA Technologies (Coralville) OligoAnalyzer, software v. 3.1 (Owczarzy et al., 2008). The best performing LAMP primer set was selected, and primer sequences (Table 2) were synthesized by Eurofins MWG Operon.
The designed primers were the outer primers F3 and B3, the loop primers LoopF and LoopB, and the inner primers BIP and FIP that are complementary, respectively, to the sequences of the regions B1C and B2, F1C and F2 (Figure 1).

RealAmp assay
RealAmp reactions were each carried out in a final volume of 25 μL, containing; 15 μL of OptiGene master mix, 2.5 μL of primer mix (final concentrations of 1.6 μM FIP and BIP, 0.2 μM F3, B3 and 0.4 μM LoopF, LoopB), 2 μL of DNA template (30 ng) and 5.5 μL of water. RealAmp fluorescence signal (RFU) was plotted automatically using the CFX96 TM real-time PCR detection system (Bio-Rad) collecting fluorescence signals at  30 sec intervals at 70°C for 45 min. For all reactions, a result was considered positive when an amplification curve was obtained, while a linear or slightly oblique amplification curve indicated a negative result. All reactions were carried out three times, and the negative controls contained nuclease-free water.

Specificity and sensitivity of RealAmp
To assess the specificity of the LAMP primer dataset, assays were carried out using DNA extracted from ten strains of Pl. richardsiae, two strains phylogenetically related to Pleurostoma species, and other strains of Phaeoacremonium and Neofusicoccum commonly associated with olive trunk disease ( Table 1).
The limit of detection (LoD) of the RealAmp assay was determined through two successive experiments using pure genomic DNA obtained from Pl. richardsiae cultures. In the first experiment, a ten-fold dilution series was prepared, from 3 ng μL -1 to 3 × 10 -6 ng μL -1 (3 fg μL -1 ). Each dilution was analyzed in ten replicates to obtain an initial estimate of LoD. Results of the first experiment were used to narrow the range of target concentrations and make a more precise estimate. In the second experiment, a two-fold dilution series was prepared, from 3 × 10 -1 ng μL -1 to 7.5 × 10 -2 ng μL -1 . Each dilution was tested in 20 replicates. In both experiments, nuclease-free water was used as negative controls (NTC).

RealAmp detection of Pleurostoma richardsiae in inoculated olive plants
As part of an ongoing experiment on olive plants ('Ogliarola salentina') artificially inoculated in a greenhouse with the Pl. richardsiae strain Pl31 (Table 1), six symptomatic plants were used for a preliminary validation of the in vivo RealAmp assay. The tested plants showed clear symptoms of necrotic wood lesions and vascular discolourations after bark removal. To ascertain the presence of the fungus, isolations were made from the stem portions at the inoculation points, according to Carlucci et al. (2013), and each sample was also used for DNA extraction using the DNA secure Plant Kit (Tiangen), according to the manufacturer's protocol. DNA used as negative controls was extracted from three uninoculated olive plants, three olive plants inoculated with Neofusicoccum parvum and three olive plants inoculated with Phaeoacremonium italicum. Each assay was carried out using 30 ng of genomic DNA from olive plants, and reactions were carried out using RealAmp.

RESULTS AND DISCUSSION
This is the first reported application of RealAmp as a rapid, specific and sensitive method for detection of Pl. richardsiae using specific primers. To screen for optimal LAMP primers, the intergenic spacer (IGS) region was employed as targets for primer design. All tested Pl. richardsiae strains showed positive reactions in the RealAmp assays, as reported in Figure 2. Optimized conditions in this study, at 70°C for 45 min, enabled amplification of the target sequence giving clear curves.
The rapidity of this technique in pathogen detection was also reported by Besuschio et al. (2017). Unlike other nucleic acid amplification techniques, RealAmp can be carried out in 30 min, rather than at least 90 min (e.g. for PCR). For specificity of the RealAmp assay using DNAs of other relative fungi, the reaction curves indicated that the Pl. richardsiae primer dataset specifically amplified the target DNA sequence of this fungus (Figure 2), the DNA isolated from inoculated olive plants (Figure 3), but not DNA of the other fungi assayed (Figure 2; Table 1). The RealAmp products of Pleurostoma  were also observed in the agarose gel electrophoresis, which confirmed the specificity of the RealAmp assay ( Figure 4). Figure 5 shows the results of experiments conducted to determine the LoD assay. The analysis carried out on the ten-fold dilution series of Pl. richardsiae genomic DNA resulted in an initial estimate of the LoD. Positive reactions were obtained for target concentrations ranging from 3 ng μL -1 to 3 × 10 -2 ng μL -1 , with frequencies, respectively, of 100% and 50% (ten replicates for each dilution were analysed). No amplification signal was observed in subsequent dilutions from 3 × 10 -3 ng μL -1 to 3 × 10 -6 ng μL -1 , or in the NTC ( Figure 5A). To define the target concentration that produced at least 95% positive replicates (LoD at 95%) (Forootan et al., 2017), the assay was repeated over a narrower range of concentrations. On the basis of the results obtained, a two-fold dilution series was prepared starting from 3 × 10 -1 ng μL -1 , corresponding to the lowest concentration of targets with 100% positives replicates, to 7.5 × 10 -2 ng μL -1 .  of positive reactions were obtained for target concentrations of 3 ng μL -1 (red curves) and 3 × 10 -1 ng μL -1 (yellow curves); 50% of positive reactions for 3 × 10 -2 ng μL -1 (blue curves); no reaction was observed from 3 × 10 -3 ng μL -1 to 3 × 10 -6 ng μL -1 or in the NTC (green curves). (B and C). Results for the estimate of LoD at 95%: positive reactions with frequency of 100% were obtained analysing target concentration of 1.5 × 10 -1 ng μL -1 (purple curves), and 95% (orange curves) analysing 7.5 × 10 -2 ng μL -1 (C).
For the artificially inoculated olive plants, results from isolations indicated that all the plants were infected by Pl. richardsiae. These results were confirmed by RealAmp, which showed amplification for all inoculated olive plants and no amplification from the negative controls ( Figure 3). Similar RealAmp results were obtained by Lee et al. (2020) for detection of Peach latent mosaic viroid in peach pollen.
The developed RealAmp assay is specific, sensitive, and rapid, for detecting Pl. richardsiae from DNA extracted from pure strains of the fungus and from infected olive plants.
Three of the six species of Pleurostoma were used to validate the developed assay in this study: Pl. richardsiae (ten strains), the ex-type of Pl. repens, and the ex-type of Pl. ootheca. However, full RealAmp validation with the remaining three Pleurostoma species could not be accomplished, due to lack of data (for Pl. candollei) and funding limitations (for Pl. ochraceum and Pl. hongkongense). In addition, further studies are required to evaluate the field performance of this test for grapevine and other hosts of this pathogen. These promising results are in accordance with the successful application of this technique for reliable and rapid detection of several plant pathogens, including X. fastidiosa in olive trees (Yaseen et al., 2015). Among the several features that make RealAmp an excellent diagnostic tool, its use, both with a real time instrument for processing numerous samples and with portable devices for few samples, are strengths compared to other PCR techniques.
In the last two decades, infections caused by Pl. richardsiae have been widespread on olive groves, often with severe symptoms leading to general tree decline, such as browning and leaf drop, wilting and decay of twigs and branches, and dark streaks and necroses in the inner tissues of host wood (Hallen et al., 2007;Carlucci et al., 2013;Carlucci et al., 2015a;Canale et al., 2019). RealAmp could become a useful tool in largescale monitoring programmes for this pathogen, and in the phytosanitary controls with certified pathogen-free propagating material of olive. In addition, as Pl. richardsiae is an uncommon agent of human infection, it may also be of interest to further investigate the application of RealAmp assay in the medical field.

ACKNOWLEDGMENTS
The study was conducted under the PhD program of Biodiversity, Agriculture and Environment of the University of Bari (Italy), in collaboration with CIHEAM of Bari and the University of Foggia (Italy). The authors thank Dr M. Biche (Department of Agricultural and Forest Zoology, High National School of Agronomy, El-Harrach, Algiers, Algeria), Dr Gramaje (Instituto de Ciencias de la Vid y del Vino, ICVV, Logroño, Spain), and Dr Mostert (University of Stellenbosch, South Africa) for providing fungal DNA/isolates.