Development of SCAR markers for rapid and specific detection of Pseudomonas syringae pv. morsprunorum races 1 and 2, using conventional and real-time PCR

Specific primers were developed to detect the causal agent of stone fruit bacterial canker using conventional and real-time polymerase chain reaction (PCR) methods. PCR melting profile (PCR MP) used for analysis of diversity of Pseudomonas syringae strains, allowed to pinpoint the amplified fragments specific for P. syringae pv. morsprunorum race 1 (Psm1) and race 2 (Psm2), which were sequenced. Using obtained data, specific sequence characterised amplified region (SCAR) primers were designed. Conventional and real-time PCRs, using genomic DNA isolated from different bacterial strains belonging to the Pseudomonas genus, confirmed the specificity of selected primers. Additionally, the specificity of the selected DNA regions for Psm1 and Psm2 was confirmed by dot blot hybridisation. Conventional and real-time PCR assays enabled accurate detection of Psm1 and Psm2 in pure cultures and in plant material. For conventional PCR, the detection limits were the order of magnitude ~100 cfu/reaction for Psm1 and 101 cfu/reaction for Psm2 in pure cultures, while in plant material were 100–101 cfu/reaction using primers for Psm1 and 3 × 102 cfu/reaction using primers for Psm2. Real-time PCR assays with SYBR Green I showed a higher limit of detection (LOD) − 100 cfu/reaction in both pure culture and in plant material for each primer pairs designed, which corresponds to 30–100 and 10–50 fg of DNA of Psm1 and Psm2, respectively. To our knowledge, this is the first PCR-based method for detection of the causal agents of bacterial canker of stone fruit trees.


Introduction
Bacterial canker of fruit trees occurs in stone fruit growing areas all over the world (Agrios 2005). In Poland, the disease incidence on stone fruit trees orchards is observed every year with different intensity and is becoming more economically significant. Moreover, in the last vegetative seasons, bacterial canker was dangerous not only to stone fruit trees, but also to apple and pear trees. The causal agents of the disease belong to the polyphagous Pseudomonas syringae species, able to infect more than 180 plant species, both annual and perennial, including fruit trees, ornamental plants and vegetables. P. syringae affects all organs of the aboveground parts of trees (i.e. the branches and main trunk as well as buds, blossoms, leaves and fruits), which causes reduction of yield and sometimes leads to death of the trees.
The diagnostics of bacterial canker are commonly based on isolation and phenotypic characterisation of the causal agent, including pathogenicity (Bultreys and Gheysen 1999;Vicente et al. 2004). The phenotypic tests LOPAT (Lelliott et al. 1966), GATTa and L-lactate utilisation (Lattore and Jones 1979) enable the determination of morphological, physiological and biochemical features of the bacteria. These features are used for identification of species and their discrimination into pathovars and races. However, this methodology requires the implementation of a high number of often laborious and time-consuming tests. Moreover, the obtained results can sometimes be ambiguous or difficult to interpret, and they are often not sufficient for proper strain classification (Vicente et al. 2004).
Concerning serological methods, the slide agglutination test, immunofluorescence and indirect-enzyme-linked immunosorbent assay (ELISA), with the antisera produced from live whole-cell antigens, were widely adopted for routine bacterial identification. However, nowadays these methods are less frequently used for the identification of bacteria that cause bacterial canker because of frequent cross-reactions with nonpathogenic bacteria. Furthermore, serological tests do not always provide a response in distinguishing isolates of P. syringae (Vicente et al. 2004).
Molecular methods are currently the most widely adapted and are considered very useful for the identification of bacterial canker causal agents and for studying their genetic diversity. For many years, the identification of the pathogen has been based on detection of genes encoding the toxins coronatine, syringomycin and the siderophore yersiniabactin (Bereswill et al. 1994;Sorensen et al. 1998;Bultreys and Gheysen 1999). However, it should be noted that the determination of presence of genes encoding for toxin production is not reliable for identification in itself and thus cannot be the only criterion for the classification of strains. In fact, strains of Psm1 and Pss, which do not have the ability to produce coronatine or syringomycin, respectively, are quite common (Ullrich et al. 1993;Renick et al. 2008;Kałużna et al. 2010a). On the other hand, although production of the siderophore yersiniabactin is now considered a stable feature of all Psm2 strains and could be a criterion for their identification, is should be mentioned that it is not an exclusive feature of strains of Psm2, since positive amplification with primers for the irp1 gene (encoding this siderophore) was also confirmed in other pathovars of P. syringae, including the following: antirrhini, apii, berberidis, delphinii, lachrymans, passiflorae, persicae, tomato, viburni, helianthi, tagetis and theae (Bultreys et al. 2006).
In recent years, fingerprinting methods have been widely applied for the identification and genotyping of P. syringae through the analysis of repetitive regions (i.e. Enterobacterial Repetitive Intergenic Consensus (ERIC), BOX, Repetitive Extragenic Palindromic Elements (REP) and Insertion Sequence (IS50) sequences) (Ullrich et al. 1993;Weingart and Völksch 1997) and through PCR MP (Kałużna et al. 2010b). However, it should be taken into account that all fingerprinting methods require inclusion of the reference strains for comparison of obtained amplification patterns (Vicente and Roberts 2007;Gilbert et al. 2009), and, in the case of heterogeneous strains of Pss (Vicente et al. 2004;Renick et al. 2008;Kałużna et al. 2010a, b), it is difficult to determine affiliation of analysed strains to this taxon.
Despite the availability of different approaches for characterisation and genotyping of P. syringae, they require timeconsuming and labour-intensive classical microbiological methods or complex analyses including comparison of amplification patterns and housekeeping gene sequencing. Therefore, there is still the need to develop a rapid and specific method of diagnosis that would allow the detection and identification of the causal agent of stone fruit bacterial canker (López et al. 2010). This specific, fast diagnostic system would be invaluable in the study on etiology of cankers on trunks and branches, which are similar to those caused by fungi of the genus Leucostoma (Valsa) and Monilinia, and also necrotic spots on leaves, which may be mistaken with those caused by Prunus necrotic ring spot virus or Clasterosporium carpophilum, especially late in the growing season. Moreover, the occurrence of gummosis on woody tissue often associated with bacterial infection may be related to the physiological response of the trees to damage caused by abiotic factors, such as frost, sunburn, periodic water flooding or mechanical damage, and is not due to biotic factors only (Saniewski et al. 2006).
Ideally, a novel diagnostic system would apply specific primers and the PCR technique, both conventional and real-time, making them more useful for a wide group of researchers according to available lab equipment, which allows for the detection and identification of the pathogen within a short amount of time. Additionally, such a system would undoubtedly be very useful in enforcing appropriate programmes to prevent and control disease occurrence in nurseries and orchards of stone fruit trees, especially sweet and sour cherry, where the damage is the most severe.
The aim of this study was to design and validate novel specific primers and to develop conventional and real-time PCR-based methodologies for rapid and specific detection of Psm1 and Psm2, with the aim of enhancing bacterial canker diagnostic procedures.

Bacterial strains
Species and pathovar identification of previously uncharacterised Pseudomonas strains from our collection, obtained from stone fruit trees in Poland, was determined on the basis of phenotypic tests (i.e. Gram reaction with 3 % KOH (Suslow et al. 1982), LOPAT (Lelliott et al. 1966), GATTa and L-lactate utilisation (Lattore and Jones 1979). A total of 168 isolates were analysed. The reference strains P. syringae pv. syringae-LMG 1247, P. syringae pv. morsprunorum race 1-LMG 2222 and P. syringae pv. morsprunorum race 2-CFBP 3800 were included in all tests (Table 1). Additionally, type and not-type strains of other P. syringae pathovars (79) and related species (three) were included in the analysis ( Table 2). The strains were kept at −75°C in a mixture of glycerol (200 μl/ml) and phosphate-buffered saline (PBS) and streaked on King's B medium (3.8 % Pseudomonas Agar F Difco, 1 % glycerol) (King et al. 1954) for routine culturing.

DNA isolation
Bacterial DNA was isolated using the method described by Aljanabi and Martinez (1997), with slight modifications described by Kałużna et al. (2012). DNA was diluted to a final concentration of 10 ng/μl and kept at −20°C for further analysis.

PCR melting profile
A slightly modified method of PCR MP described by Masny and Płucienniczak (2003) was used. An amount of 100 ng of DNA from 23 Pseudomonas strains (Figs. 1 and 2) was digested with PstI endonuclease (10 U/μl; Promega Corporation, Madison, WI, USA) or TaqI (10 U/μl; Thermo Scientific, Vilnius, Lithuania) according to the manufacturer's instructions. Digested DNA was ligated with two oligonucleotides forming an adaptor: DNA digested by PstI endonuclease with a PstI adaptor-5′-TGTACGCAGTCTAC-3′/5′-CTCGTAGACTGCGTACATGCA-3′ (Waugh et al. 1997) and DNA digested by TaqI endonuclease with a TaqI adap- -Marsan et al. 1997). PCR amplification was performed separately for PstI-or TaqIdigested DNA in a 25-μl reaction mixture containing the following: 1 μl of ligation mixture; 0.4 U of GoTaq DNA polymerase (Promega, Madison, WI, USA) for PstI and 0.4 U of Dream Taq Green DNA Polymerase (Thermo Scientific, Vilnius, Lithuania) for TaqI; and 1× of appropriate Taq polymerase buffer, 0.2 mM of dNTPs and 1 μM of each primer (PstI-0-5′-GACTGCGTACATGCAG-3′ for PstIdigested DNA (Waugh et al. 1997) or TaqI-0-5′-GACGATGAGTCCTGACCGA-3′ for TaqI-digested DNA (Ajmone-Marsan et al. 1997)). The amplification reactions were conducted in a Biometra T3000 thermocycler (Biometra, Göttingen, Germany) with the following conditions: initial step of 72°C for 5 min; 30 cycles at 86.5°C for PstI and 83°C for TaqI for 40 s, 55°C for 40 s and extension at 72°C for 90 s; and final extension at 72°C for 10 min. PCR products from each reaction and the O'GeneRuler 100-bp DNA Ladder Plus (Thermo Scientific, Vilnius, Lithuania) were separated on a 1.5 % agarose gel in 0.5× TBE buffer (0.045 M tris-boric acid, 0.001 M EDTA, pH 8.0) and electrophoresis was ran at 5-7 V/cm of gel. After staining with an ethidium bromide solution (0.5 μg/ml), the obtained amplification profiles were visualised under UV light. The same conditions were used in all subsequent electrophoresis.

Selection of specific fragments
Based on the results of genetic analyses using PCR MP, DNA fragments characteristic of Psm1 and Psm2 strains were selected. The fragments were excised from the gel, purified with the DNA AxyPrep Gel Extraction Kit (Axygen Scientific, Inc. Union City, CA, USA) and cloned into the pGEM T-Easy vector (Promega, Madison, WI, USA) according to the manufacturer's instructions. The resulting ligation mixture was used to transform Escherichia coli JM109 competent cells (Promega, Madison, WI, USA). The cloned fragments were sequenced with universal primers M13Rev 5′-CAGGAAACAGCTATGAC-3′ and M13 (−40) 5′-GTTTTCCCAGTCACGAC-3′ at Genomed S.A. (Warsaw, Poland). The sequences obtained were assembled using the SeqMan software package LASERGENE (DNASTAR, Madison, USA).

Design of SCAR primers
The sequences of specific fragments for Psm1 and Psm2 were used to design the SCAR primers, for both conventional and real-time PCR, with the PrimerSelect programme of the LASERGENE package (DNASTAR). Different primer pairs were designed for conventional PCR (five for Psm1 and 7 for Psm2) and real time PCR (four for each taxa). All primer sequences and their potential amplification reaction products were checked for homology (June 2015) to other sequences deposited in the GenBank database using the 'blastn' algorithm (Altschul et al. 1997). Selected primers were synthesised at Genomed S.A.

Dot blot hybridisation
High-throughput specificity assays were carried out using a dot blot platform, essentially as previously described (Albuquerque et al. 2011). PCR amplicons obtained using primers Psm1-6F/6R, with template DNA from strain Psm 28a (race 1), and primers Psm2-8F/8R, with Psm 77 (race 2), were purified using the GFX PCR and Gel Band Purification Kit (GE Healthcare, Buckinghamshire, UK) and labelled with digoxigenin, using the DIG-High Prime DNA labelling kit (Roche, Basel, Switzerland) in order to obtain the two tested hybridisation probes Psm1 and Psm2, respectively. Amounts of 100 ng of heat-denatured DNA from each bacterial strain were transferred to a nylon membrane using a Bio-Dot apparatus (Bio-Rad, Hercules, USA). Hybridisation was carried out overnight at 68°C with a final probe concentration of 100 ng/mL, and the washing and detection steps were carried out according to the DIG application manual (Roche). The chemiluminescent signal indicative of probetarget hybrids was detected using a Molecular Imager ChemiDoc XRS+ System (Bio-Rad), with all pixels below saturation point.

Conventional and real-time PCR amplifications
Amplification reactions with the two selected primer pairs, one specific for the strains of Psm1 and the second specific for Psm2, were performed in a Biometra T3000 thermocycler (Biometra, Göttingen, Germany). The reaction mixture in 15 μl of total reaction volume contained 10 ng of DNA, 0.4 U of Dream DNA Polymerase (Promega, Madison, WI, USA), 1× reaction Dream Taq Green buffer (Thermo Scientific, Vilnius, Lithuania), 0.15 mM dNTPs and 0.7 mM of each primer. The following     The resulting PCR products were separated by electrophoresis on 1.5 % agarose gels as described above.
Real-time PCR with SYBR Green I was conducted in the Bio-Rad CFX96 with SsoAdvanced SYBR Green Supermix (Bio-Rad, Hercules, USA). The reaction mixture in 20 μl of total volume contained 1× reaction SYBR Green Supermix and 0.5 mM of each of the following primers: Psm1-1F-RT/Psm1-1R-RT for Psm1 and Psm2-1F-RT/Psm2-1R-RT for Psm2. Bacterial DNA was used as a template (10 ng per PCR reaction). No-template reactions were used as negative controls. The PCR programme was started from one cycle of denaturation at 98°C for 130 s, followed by 35 cycles at 95°C for 10 s and then 60°C for 15 s, finished by a melting curve analysis for verification of the specificity of amplification in real-time PCR products. Progressive denaturation of products was carried out at a rising temperature, starting from 65°C and continuing to 95°C, with 0.5°C of increment for 5 s each.

Specificity of designed primers and their usefulness in detection in plant material
In the first stage of this part of the study, the specificity of the two designed primer pairs was determined with PCR using DNA from all strains of Psm1, Psm2 and Pss as well as strains of atypical taxa (Table 1). In the second stage, the primers were tested with DNA from other P. syringae pathovars and related species (Table 2). In order to assess the suitability of the designed primers for the detection of Psm1 and Psm2 strains in plant material, several leaves, shoots and fruits of sweet cherry, sour cherry and plum were collected. Amounts of 100 mg of crushed/cut plant tissue of each organ were placed in 1.9 ml of PBS buffer. For each type of tissue (organ) and host plant, two tubes were prepared (18 tubes in total). One hundred microlitres of bacterial suspension (10 5 cfu/ml) of the Psm1 reference strain (LMG 2222) or the Psm2 reference strain (CFBP 3800) were added to nine of the samples (one of each organ and of each plant). One hundred microlitres of sterile water were added to the remaining nine samples, which were tested to verify the purity of the plant material. After 1 h of shaking incubation at 26°C, 1 ml of washing liquid separate from each of all 18 samples was centrifuged; the resulting pellet was suspended in 100 μl of TE buffer, and the DNA was isolated using a Genomic Mini DNA Extraction Kit (A&A Biotechnology, Gdynia, Poland) according to the manufacturer's instructions.

PCR limit of detection
The limit of detection of PCR using the SCAR primers was evaluated using DNA extracted from pure bacterial cultures, DNA extracted from plant material that was mixed with suspensions of bacteria and bacterial genomic DNA (gDNA). A PCR assay was carried out with decimal dilutions of bacterial suspensions of strain LMG 2222 or CFBP 3800 (from~10 8 to 10 0 cfu/ml). DNA was isolated from 1 mL of each dilution using a Genomic Mini DNA Extraction Kit (A&A Biotechnology) according to the protocol supplied by the manufacturer. To determine the limit of detection of bacteria in the plant material, 100-mg portions of stems and leaves of sweet cherry (for Psm1 primers) or sour cherry (for Psm2 primers) and 100 μl of the previously prepared 10-fold serial dilutions of bacterial suspensions (from~10 8 to 10 0 cfu/ml) or 100 μl of sterile water, used as a control of material purity, were added to 1.9 mL of PBS buffer and shaken for 30 min at 26°C. After incubation, the washings were centrifuged (8,000 rpm, 5 min); the resulting pellet was suspended in 100 μl of TE buffer, and DNA was isolated using the Genomic Mini DNA Extraction Kit (A&A Biotechnology) according to the manufacturer's instructions. The sensitivity of gDNA detection was checked using 2-fold serial dilutions of gDNA isolated (11 ng to~11 fg per PCR reaction for Psm1 and 14 ng to~14 fg per PCR reaction for Psm2) using the method described by Aljanabi and Martinez (1997), with slight modifications described by Kałużna et al. (2012). The PCR efficiency was calculated from the slope of the standard curve generated for each run in the following equation E = 10 (−1/slope) where E = 2 and corresponds to 100 % efficiency (Ramakers et al. 2003).

Phenotypic characterisation
All 168 isolates have been classified into species P. syringae LOPAT group Ia. GATTa and L-lactate utilisation tests allowed further discrimination of pathovars and races: 49 isolates were identified as P. syringae pv. morsprunorum race 1 (Psm1), 10 as race 2 of this pathovar, 53 as pathovar syringae (Pss) and 56 as belonging to atypical taxa, having most of the features of Pss without, however, the ability of esculine hydrolysis (lack of β-glucosidase activity) (Table 1).

Design of SCAR primers
The nucleotide sequences obtained for the Psm1 and Psm2 fragments were used to design different SCAR primers. After validation, the most specific primers for conventional and real-time PCR were selected (Table 3). A BLAST analysis of selected primer sequences showed no similarity to any bacterial sequences in GenBank.

Dot blot hybridisation
The dot blot results confirmed the high specificity of the selected markers towards the target pathogens. Using probe Psm1, positive hybridisation results (dark dots) were observed with all tested Psm1 strains, and no unspecific hybridisation was observed with DNA from any non-Psm1 pseudomonads. Similarly, probe Psm2 was exclusively specific for the tested Psm2 strains. Additionally, the hybridisation results showed that the selected DNA regions were present in all their respective target strains, confirming their stability (Fig. 3).

Specificity of designed primers and usefulness in detection in plant material
The PCR assays using DNA from all tested P. syringae strains including reference strains (Table 1), as well as DNA from strains of other species within the Pseudomonas genus ( Table 2), showed that all the designed primers were specific for their respective taxa. PCR assays using primers Psm1-6F/ 6R and Psm1-1F-RT/1R-RT, specific for Psm1, successfully amplified the expected PCR products 793 bp (Fig. 4) and 101 bp (Fig. 5), respectively, using DNA from all strains of Psm1. No amplification was observed when DNA from strains identified as Psm2 or Pss and strains of atypical taxa were used. Amplification using primers Psm2-8F/8R and Psm2-1F-RT/1R-RT, designed for detection of Psm2, was achieved with DNA from all strains of Psm2, resulting in PCR products of expected lengths of 410 bp (Fig. 6) and 104 bp, respectively. No increase in fluorescence was observed with DNA from Psm1 or Pss and strains of atypical taxa. The melting curves of the reaction products obtained from real-time PCR revealed a single peak with a melting temperature of 80°C or 77°C for Psm1 and Psm2, respectively. Also, neither unexpected nor additional peaks in the product melting curves were observed, which clearly excluded possibilities or tendency of the primers to form dimers. Moreover, none of the four tested primer pairs amplified the DNA of 79 strains of other pathovars of P. syringae and other species ( Table 2).
The usefulness of the designed primers for detection of Psm1 and Psm2 strains in plant material was assessed with PCR assays using DNA extracted from a mixture of plant tissues and a suspension of target bacteria. The results confirmed the specificity of selected primer-pairs since positive amplification was achieved in mingled samples, while no nonspecific amplification was observed in samples without bacteria addition. Additionally, these assays showed that the proposed PCR detection methodology was not affected by potential inhibitors present in plant samples.
Limit of detection of P. syringae pv. morsprunorum for conventional and real-time PCR Both tested primer pairs designed for conventional PCR allowed for the detection of 10 0 cfu/reaction of Psm1 and 10 1 Psm2 in pure culture. Regarding the presence of bacteria in different organs of sweet and sour cherries, it was possible to detect 10 0 and 10 1 cfu/reaction for sweet cherry leaves and shoots, respectively, using the Psm1-specific primers and 10 2 cfu/reaction for sour cherry leaves and shoots using the Psm2- F forward primer, R reverse primer, RT real-time, Tm melting temperature specific primers. The sensitivity (LOD, limit of detection) of the detection in the conventional PCR assay was~4 pg for Psm1 strain 199 and~5 pg for Psm2 strain 745 when aliquots of serial 2-fold dilutions of purified DNA were used which corresponds to the order of magnitude~10 1 -10 2 cfu/reaction. Both tested primer pairs designed for Psm1 and Psm2 strains using real-time PCR allowed the detection of 10 0 cfu/reaction of Psm1 or Psm2 in pure culture and in plant material. Only the expected products and a single peak with melting temperature were obtained. Standard curves using template DNA from bacterial suspensions, DNA from plant material with additions of bacterial suspensions and bacterial gDNA showed high amplification efficiency and linearity of the data (Table  4). An exception occurred for the products obtained from shoots of sweet cherry with additions of bacterial suspensions of Psm1. Although linearity was quite good, the noted efficiency of 83 % was not in the range considered acceptable (90-110 %). Moreover, the efficiency obtained for the mixture of shoots of sour cherry and Psm2 suspension when testing with primers for Psm2 was also lower compared to DNA template from sour cherry leaves and bacterial suspension alone. The sensitivity (LOD) of the detection in the real-time PCR assay when using gDNA ranged from~30 to 100 fg for Psm1 strain 199 and~10 to 50 fg for Psm2 strain 745 when 1.0-μl aliquots of serial 2-fold dilutions of Fig. 3 Dot blot validation of probes Psm1 and Psm2. The probes were evaluated with total DNA from 167 strains; including P. syringae strains isolated from stone fruit trees, reference strains and others pathovars of P. syringae from the CFBP culture collection. The table grid above represents the coordinates of each strain tested in the dot blot, which are identified by their abbreviations further detailed in Tables 1 and 2. Positive hybridization signals are visualised as dark dots purified DNA were used which corresponds to the order of magnitude~10 0 cfu/reaction (Table 4).

Discussion
In this study, the methods and tools enabling the rapid and highly specific identification and detection of bacterial canker causal agent P. syringae pv. morsprunorum races 1 and 2 were developed. The methods based on the use of specific primers designed for conventional and real-time PCR allow in routine testing for omitting the application of often time-consuming methods of classical microbiology, fingerprinting methods or housekeeping gene sequence analysis used until now by other authors (Vicente and Roberts 2007;Gilbert et al. 2009). Of course in critical cases (i.e. first reports, claims, etc.) these other methods are still indispensable. Our newly developed Fig. 6 Evaluation of primers Psm2-8F and Psm2-8R for identification of P. syringae pv. morsprunorum race 2: M-O'GeneRuler 100-3000 bp (Thermo Scientific, Vilnius, Lithuania), strains Psm1: 2-25b,  Evaluation of primers Psm1-6F and Psm1-6R for identification of P. syringae pv. morsprunorum race 1: M-O'GeneRuler 100-3000 bp (Thermo Scientific, Vilnius, Lithuania), strains Psm1: strains Psm1: 2-LMG 2222, 3-28a, 4-29a, 5-38a, 6-175, 7-199, 8-201, 9-203, 10-205, 11-274, 12-755, strains Psm2:13-CFBP 3800, 14-732, strains Pss: 15-LMG 1247, 16-760, strains of atypical taxon: 17-58, 18-970a, 19-K-Appl Microbiol Biotechnol (2016 Fig 5 Real-time PCR with SYBR Green I (Bio-Rad, Hercules, USA) for specific detection of DNA from suspension of Psm1 strain LMG2222 (example). Fluorescence signal is related to the amount of template. Samples from 10 6 , 10 5 , 10 4 , 10 3 , 10 2 , 10 1 and 10 0 cfu/reaction methods and tools are very useful and invaluable in both epidemiological studies and in development of protection programmes for stone fruits against bacterial canker. Using the genetic fingerprinting PCR MP method, we demonstrated the diversity of P. syringae strains, which was very important in the selection of specific DNA fragments for two races of P. syringae pv. morsprunorum. Based on the obtained nucleotide sequences of these fragments, Psm1-and Psm2specific SCAR primers were designed. The specificity of the designed primers for Psm and amplified regions was confirmed by BLAST, since the fragments did not show (at present) any significant similarity hits within the NCBI database. Due to the high electrophoretic profile heterogeneity obtained for Pss strains arising from their high genetic diversity confirmed already by other authors (Vicente and Roberts 2007;Gilbert et al. 2009;Kałużna et al. 2010a, b), it was not possible to find a common DNA fragment for all strains belonging to this taxon.
Commonly used methods for designing SCAR primers include rep-PCR (repetitive PCR) (Sangdee et al. 2013), randomly amplified polymorphic DNA (RAPD) Cheng et al. 2015), amplified fragment length polymorphism (AFLP) (Zhang et al. 2012), PCR with universal rice primers (URP-PCR) (Lim et al. 2009) and intersimple sequence repeat (ISSR) (Giaj Merlera et al. 2015). Although the PCR MP method was described so far as helpful in the study of genetic diversity of bacteria and yeast (Leibner-Ciszak et al. 2010;Kałużna et al. 2010bKałużna et al. , 2014Zasada et al. 2014), it has not been previously reported to be used for the selection of SCAR markers. In this work, the PCR MP is for the first time used for the design of SCAR primers specific for detection of plant pathogenic bacteria.
The results obtained in this study showed that the designed SCAR primers can be applied for specific, direct detection of strains belonging to Psm1 or Psm2, both in pure culture and infected plant material. Their specificity was confirmed by PCR, using DNA from several Pseudomonas spp. strains, which showed that positive amplification occurred only with DNA of the targeted taxa strains. This is especially significant in the case of strains of atypical taxa and pathovars of P. syringae (i.e. pv. syringae and pv. avii, which also infect cherry (Ménard et al. 2003;Renick et al. 2008)) to exclude that symptoms are connected to another taxa/pathogen or to abiotic factors. Importantly, when testing the developed primers in conventional PCR, using DNA isolated from a mixture of plant material and bacteria of Psm1 or Psm2, the suppression of amplification by potential plant inhibitors like polyphenols and pesticide residues, as reported by Puławska et al. (1997), was not found. Additionally, for DNA from the asymptomatic plant material without addition of bacterial DNA, no positive amplification was observed. This means that the designed primers did not react with DNA of potential bacteria naturally inhabiting the plant material, which is essential to prevent false-positive diagnostic results. However, in the case of real-time PCR, which is the more sensitive method, some effects of plant material were noted. Although standard curves using different template DNA showed the high amplification efficiency and linearity of the data for the majority of DNA tested, for shoots of sweet cherry with additions of bacterial suspensions the efficiency was below the range considered acceptable, indicating higher dilution of those templates than expected. Also, a decrease of efficiency (Table 4) in the case of sour cherry shoots was observed. The results therefore may indicate the influence of shoots for more sensitive real-time PCR reactions. a E = PCR efficiency; ideally the efficiency should be 100 %, meaning that for each cycle the amount of product doubles; high/acceptable amplification efficiency (90-110 %). Efficiency = 10 (−1/slope) − 1 b R2 is a measure of data linearity amongst technical replicates of serial dilutions; indicates how good one value is in predicting another; R2 = 1 is perfect c The slope of the log-linear phase of the amplification reaction is a measure of reaction efficiency. To obtain accurate and reproducible results, reactions should have an efficiency as close to 100 % as possible, equivalent to a slope of −3.32 d Y = int represents the value of Ct where the curve crosses the y-axis The designing of primers for both systems, conventional and real-time PCR, makes the developed diagnosis system more accessible to a wider group of researchers, as many laboratories do not have access to special equipment or specialised personnel to perform the real-time PCR or have less funds. However, as described, the real-time PCR procedure is much faster (whole reaction with melting curve analysis is about 1 h from the beginning with SsoAdvanced SYBR Green Supermix); it allows the use of DNA quickly extracted from pure culture by the boiling method, without loss of detection resolution, and also excludes additional timeconsuming post-PCR processes (i.e. agarose gel electrophoresis). Therefore, using this technique, it is possible to obtain a very fast response about the causal agent of the disease. However, it should be noted that this system is highly sensitive and that false-positive results can occur. The risk of falsepositive results due to cross-contamination during preparation of the PCR can be minimised by using negative controls and high discipline during work (e.g., application of tips with filters during the DNA isolation step). Additionally, positive results obtained during those of the final PCR cycles should be treated as suspect only, for which additional, more detailed investigations should be conducted. Moreover, during all the assays the melting curve analysis is recommended to exclude nonspecific amplicons (as a consequence of which are visible in each run as the rest of the analysed specific ones). Dot blot hybridisation confirmed that the two selected DNA regions were highly specific for their target genomospecies and stable amongst all tested isolates of either Psm1 or Psm2, which is essential for preventing false-positive and false-negative results, respectively as much as possible.
In summary, when compared with so-far available methods for identification and differentiation of causal agents of stone fruit bacterial canker based on phenotypic characters, fingerprinting methods or MLST, the use of pathovar-specific primers allowed for greatly shortening the time required for diagnosis, while highly increasing assay accuracy and lowering detection limit. Moreover, this PCR-based method is relatively simple and inexpensive, and it does not require the timeconsuming step of pre-incubation on microbiological media (Schaad et al. 1995). Even in the presence of potential inhibitors present in plant material, which can affect the limit of detection, we could detect 1 and 3 × 10 2 cfu/reaction using primers specific for Psm1 and Psm2 in conventional PCR. A similar detection sensitivity in conventional PCR was obtained by other authors in their identification systems for other phytopathogens (Catara et al. 2000;Kerkoud et al. 2002;Biondi et al. 2013). The sensitivity of real-time PCR was higher than in the case of conventional ones, as 1 cfu/reaction was detected when different templates were used. This is especially important in the case of naturally infected material in the presence of a small amount of pathogen DNA, which be detected in a very short time. The limit of detection when using gDNA was in the range from~4-5 pg in conventional and~10-100 fg in real-time PCR for both taxa, which are similar to results obtained for P. syringae pv. actinidiae (Gallelli et al. 2014) and Clavibacter michiganensis subsp. sepedonicus (Cho et al. 2015). The high sensitivity of the developed assay (obtained in our hands) will be invaluable for detecting the target bacteria in the early latent period of the disease, allowing growers to undertake appropriate prevention or protection programmes.