A Powerful LAMP Weapon against the Threat of the Quarantine Plant Pathogen Curtobacterium flaccumfaciens pv. flaccumfaciens

Curtobacterium flaccumfaciens pv. flaccumfaciens (Cff) is a Gram-positive phytopathogenic bacterium attacking leguminous crops and causing systemic diseases such as the bacterial wilt of beans and bacterial spot of soybeans. Since the early 20th century, Cff is reported to be present in North America, where it still causes high economic losses. Currently, Cff is an emerging plant pathogen, rapidly spreading worldwide and occurring in many bean-producing countries. Infected seeds are the main dissemination pathway for Cff, both over short and long distances. Cff remains viable in the seeds for long times, even in field conditions. According to the most recent EU legislation, Cff is included among the quarantine pests not known to occur in the Union territory, and for which the phytosanitary inspection consists mainly of the visual examination of imported bean seeds. The seedborne nature of Cff combined with the globalization of trades urgently call for the implementation of a highly specific diagnostic test for Cff, to be routinely and easily used at the official ports of entry and into the fields. This paper reports the development of a LAMP (Loop-Mediated Isothermal Amplification) specific for Cff, that allows the detection of Cff in infected seeds, both by fluorescence and visual monitoring, after 30 min of reaction and with a detection limit at around 4 fg/μL of pure Cff genomic DNA.

The bacterial strains used in this study are listed in Table 1. The Cff strains include several different colony variants of this pathogen, that are yellow-, red-and orange-pigmented strains [21]. In addition to Cff strains isolated from leguminous plants and from several alternative hosts, C. flaccumfaciens isolates that are not pathogenic on beans, and the type strains of the four other C. flaccumfaciens pathovars betae, ilicis, oortii, and poinsettiae, were also used. Furthermore, closely related species (i.e., Clavibacter michiganensis subsp. michiganensis (Cmm)), as well as other important bean phytopathogens (i.e., Pseudomonas savastanoi pv. phaseolicola (Psp) and Xanthomonas axonopodis pv. phaseoli (Xap)), were also tested (Table 1). Bacterial strains were routinely grown at 26 • C on Luria Bertani (LB) [29] medium both as liquid and solid cultures, while Cff strains were specifically plated on nutrient broth yeast extract agar medium (NBY) [30]. Bacteria were preserved for long-term storage at −80 • C in LB broth, supplemented with 40% glycerol (w/v), and subcultured when required.

Bacterial DNA Extraction and Thermal Lysis
Bacterial DNA extraction and purification were carried out using the Puregene ® DNA Isolation Kit (Qiagen GmbH, Hilden, Germany), as recommended by the manufacturer. The yield and quality of the extracted DNA were evaluated both spectrophotometrically by using Nanodrop ND-100 (Nanodrop Technologies, Waltham, MA, USA), and visually by standard agarose gel electrophoresis (1% agarose (w/v) in TBE 1X) [31], respectively. The DNA was then stored at −20 • C until needed. Bacterial DNA was also obtained by thermal lysis of single colonies, each picked up from fresh agar plates with a sterile loop and diluted in sterile bidistilled water (100 µL/pellet), incubated at 95 • C for 15 min, and then immediately cooled on ice. After a quick spin in a microcentrifuge, 5 µL lysate was directly used in amplification reactions as a template.

Primer Design
Several primer sets for the specific LAMP amplification of Cff DNA were designed by using the Primer Explorer V5 software (Eiken Chemicals, Tokyo, Japan) (http://primerexplorer.jp/lampv5e), targeting the same 306 bp nucleotide sequence on which the conventional PCR-based protocol used for Cff identification was also based [16,17]. A manual check for the best default program parameters and a BLAST search for potential aspecific homologies (http://www.ncbi.nlm.nih.gov/blast) were also carried out. For comparison, the primer pair Cff FOR2/Cff REV4 was also used [16]. Primers were synthesized and HPLC-purified at Eurofins (Hamburg, Germany).

LAMP Reaction
The 25 µL LAMP reaction mixture contained 15 µL of GspSSD Isothermal Mastermix ISO-001 (Optigene, Horsham, UK), the Cff F3 and Cff B3 outer primers (200 nM each), the Cff FIP and Cff BIP inner primers (800 nM each), and, as a template, a 5 µL solution of purified genomic DNA from Cff strains reported in Table 1, at a concentration depending on the experimental purposes. In negative controls, the template consisted of sterilized molecular grade bidistilled water. Cross-reactivity with non-target species was tested, by using purified DNA from non-pathogenic C. flaccumfaciens strains, or closely related bacteria (i.e., Cmm), or other plant pathogenic bacteria occurring on beans (i.e., Psp and Xap). The optimized reaction was run at 63 • C for 30 min in the CFX96™ real-time fluorometer (Bio-Rad, Hercules, CA, USA), with fluorescence data recorded at 60 s time intervals. The LAMP generated amplicons (5 µL/reaction) were also endpoint analyzed on agarose gel (2% agarose (w/v) in TBE 1X) [31], with 5 µL reaction/well, then stained with ethidium bromide (0.5 µg/mL), and UV-visualized by using the Molecular Imager ® Gel Doc™ XR System (Bio-Rad).
A colorimetric LAMP assay was also performed, by using the WarmStart Colorimetric LAMP 2× master mix (New England BioLabs, Ipswich, MA, USA), according to the manufacturer's recommendations, and carried out at 63 • C for 30 min in a 25 µL final reaction. Sterilized molecular grade bidistilled water was included in each analysis as a negative control. Six independent experiments were conducted with three replicates for each sample.

Detection of Cff in Artificially and Naturally Infected Bean Plant Materials
The in planta performances of the newly developed Cff -specific LAMP assay have been assessed both on bean plants artificially inoculated with Cff and on naturally Cff -infected bean seeds of the Iranian cultivar (cv.) Pak.
Artificial inoculations on certified commercial bean seeds of the Italian cv. Cannellino (Linea Mediterranea srl, Pomezia-Roma, Italy) were performed by using the hilum injury method [32]. Each seed was injured by piercing the hilum with a sterile needle. Then, bean seeds were soaked for 1 h in a fresh Cff bacterial suspension (OD 600 = 0.5, corresponding approximately to 10 8 colony-forming unit/mL (CFU/mL)). Seeds used as negative controls were soaked in sterile physiological solution (SPS; 0.85% NaCl in bidistilled water). Inoculated and uninoculated bean seeds were then separately sown in 15 cm diameter plastic pots (1 seed/pot) with Cornell peat-lite mix, and incubated in a growth chamber with a 16 h photoperiod and temperature of 22 ± 2 • C. Plants were monitored daily for the appearance of symptoms, and at 14 days post-infection (d.p.i.), three true leaves from each Cff -inoculated and control plant were detached and used for DNA extraction, which was carried out with NucleoSpin ® Plant II (Macherey-Nagel, Düren, Germany), according to the manufacturer's recommendations. The resulting total genomic material contained both the host plant (bean) DNA and the bacterial DNA (Cff ). This DNA was then used as a template for both conventional PCR and LAMP assays. Three independent experiments were performed, with six replicates for each experimental condition for each run.
Samples (10 seeds each, corresponding to about 4.225 ± 0.350 g) of healthy-looking and symptomatic Cff -naturally infected bean seeds cv. Pak were separately washed three times in sterile distilled water. These seeds were derived from a Cff -infested area in Iran. Cff -free certified seeds of the same cultivar were used as a negative control. Each sample was then finely grounded, and transferred into 20 mL of SPS, under shaking conditions for 12-14 h. Any seed particle and residue were then eliminated by filtering the suspension on a sterile gauze, and the filtrate was then centrifugated at 6000× g, for 20 min at 4 • C. The supernatant was discharged, while the pellet was resuspended in 1 mL SPS, and then directly used for the amplification tests. Three independent experiments were performed, and six samples for each condition (i.e., healthy-looking and symptomatic Cff -infected seeds, and certified Cff free seeds) at each run were used.

Design and Selection of LAMP Primers
As a target sequence for LAMP primer design, the highly conserved 306 pb DNA fragment amplified by the Cff specific endpoint PCR-based test [4,16,17] (GenBank Accession Numbers AJ307048, AJ307049, and AJ307051) was selected. Among the primers generated by Primer Explorer V5, a set of four LAMP primers was chosen as the most appropriate candidate, according to several key parameters, such as GC content, melting temperature (Tm), distances between primers, and stability of primer ends expressed as free energy (∆G). These primers were also in silico analyzed by BLAST searches, and no homology hits were found but the expected Cff target sequence (data not shown).
The selected LAMP primers recognize six distinct regions on the target sequence, and consist of the two outer primers named Cff F3 and Cff B3, and the two inner hybrid primers named Cff FIP and Cff BIP (Cff FIP = sequences F1c + F2; Cff BIP = sequences B1c + B2). The primer sequences and their main features are reported in Table 2. Their annealing sites on the target sequence are shown in Figure S1. The stability of the ends of the selected LAMP primers, expressed as ∆G, appeared to be high, in particular to those ends essential as starting points for DNA synthesis. In particular, the 3 ends of the outer primers Cff F3 and Cff B3, and of the sequences F2 and B2 of the internal hybrid primers, showed ∆G values definitely lower than −4 Kcal/mol (Table 2), to guarantee a high degree of stability [33]. Similarly, the ∆G values for the 5 ends of the F1c and B1c sequences of the internal primer pair were −5.17 and −5.69, respectively. Finally, primer design is the main challenging step in the development of a new LAMP assay as the probability of secondary structures formation is increased by the high number of primers required (four as a minimum). According to their ∆G values, the LAMP primers here designed and selected were predicted to be poorly prone towards the formation of both cross-and self-dimers as well as of hairpins (Table S1).
No additional suitable loop primers were identified using Primer Explorer V5, with the short size of the template as the only limiting factor. However, loop primers have been demonstrated to be not essential for a successful LAMP reaction, although a reduction in amplification times can sometimes be achieved [34][35][36].

Optimization of LAMP Assay for Cff Detection
The optimal temperature and reaction time of the LAMP assay for Cff detection have been established by using pure DNA of the Cff type strain ICMP 2584 (40 ng/reaction) as template, and accordingly to the assessment of both gel electrophoresis and real-time fluorescence data.
Pure Cff ICMP 2584 T DNA was tested at six different amplification temperatures, within the range 62-67 • C, and the reaction time was 30 min as recommended by the manufacturer. The typical ladder-like DNA multiple bands of LAMP reaction have been detected on an agarose gel in positive samples for all the tested amplification temperatures ( Figure 1A).
No unspecific amplifications or primers cross-annealing were observed, regardless of the reaction temperature. Similarly, the real-time fluorescence monitoring showed successful reactions to take place within the range 62 • C-67 • C on positive samples, and the plateau was reached after 15 min ( Figure 1B). The reaction threshold times ranged from 5 to 9 min for amplification temperatures between 62 • C and 65 • C, while higher times (i.e., 12 and 15 min) were needed at higher temperatures (i.e., 66 • C-67 • C) ( Figure 1). Accordingly, the optimal conditions established for this Cff -specific LAMP assay were 63 • C with a 30-min running time (corresponding to 30 cycles of 1 min/each), then used for all the following applications. established by using pure DNA of the Cff type strain ICMP 2584 (40 ng/reaction) as template, and accordingly to the assessment of both gel electrophoresis and real-time fluorescence data.
Pure Cff ICMP 2584 T DNA was tested at six different amplification temperatures, within the range 62-67 °C, and the reaction time was 30 min as recommended by the manufacturer. The typical ladder-like DNA multiple bands of LAMP reaction have been detected on an agarose gel in positive samples for all the tested amplification temperatures ( Figure 1A). No unspecific amplifications or primers cross-annealing were observed, regardless of the reaction temperature. Similarly, the real-time fluorescence monitoring showed successful reactions to take place within the range 62 °C-67 °C on positive samples, and the plateau was reached after 15 min ( Figure 1B). The reaction threshold times ranged from 5 to 9 min for amplification temperatures between 62 °C and 65 °C, while higher times (i.e., 12 and 15 min) were needed at higher temperatures (i.e., 66 °C-67 °C) (Figure 1). Accordingly, the optimal conditions established for this Cff-specific LAMP assay were 63 °C with a 30-min running time (corresponding to 30 cycles of 1 min/each), then used for all the following applications.

Specificity and Sensitivity of the LAMP Assay for Cff Detection
The specificity of the LAMP primers designed here was assessed by testing, as a template, the purified genomic DNA (40 ng/reaction) from 35 Cff strains, having different geographical origin and isolated from several different host plants, including P. vulgaris L., V. unguiculata L., Capsicum annum L., Solanum lycopersicum L., and S. melongena L. Moreover, DNAs from non-target phytopathogenic bacteria were also included, such as several C. flaccumfaciens strains, or the type strains from the C. flaccumfaciens pathovars betae, ilicis, ooorti, and poinsettiae, or bacteria taxonomically related to Cff (i.e., Cmm), or bacteria that are pathogenic on common bean and seedborne (i.e., Psp and Xap) ( Table 1). All the Cff strains tested here gave positive results, as assessed by both real-time fluorescence monitoring and gel electrophoresis analysis. No cross-reactivity or aspecific amplification were found to occur when DNA from non-target bacteria species was used as a template (Table 1). Therefore, this LAMP assay for Cff detection has a specificity fully comparable with that of the existing conventional PCR test for this quarantine phytopathogen [16,17] (Table 1).
In order to evaluate the analytical sensitivity of the LAMP assay developed here, the smallest known amount of target DNA detected in each test sample was assessed. Several ten-fold serial dilutions of Cff ICMP 2584 T pure genomic DNA were used as a template, prepared in sterilized molecular grade bidistilled water and ranging from 100 fg/reaction to 100 ng/reaction, in a 25 µL final volume. The fluorescence-based real-time monitoring showed that this LAMP reaction is able to detect Cff genomic DNA down to 0.1 pg/reaction after 20 min. As expected, the amplification time was shorter as the DNA template concentration increased, and positive signals were obtained just after 15 min with 1 pg/reaction Cff DNA as a template (Figure 2A). The agarose electrophoresis analysis of LAMP products confirmed these data. The characteristic ladder-like DNA bands were observed when Cff DNA was used as a template in the range of 100 ng-100 fg per reaction ( Figure 2B).
detect Cff genomic DNA down to 0.1 pg/reaction after 20 min. As expected, the amplification time was shorter as the DNA template concentration increased, and positive signals were obtained just after 15 min with 1 pg/reaction Cff DNA as a template (Figure 2A). The agarose electrophoresis analysis of LAMP products confirmed these data. The characteristic ladder-like DNA bands were observed when Cff DNA was used as a template in the range of 100 ng-100 fg per reaction ( Figure  2B).

LAMP Detection of Cff on Artificially and Naturally Infected Bean Samples
Firstly, bean plants cv. "Cannellino" artificially inoculated with Cff ICMP 2584 T by hilum injured method were used to validate the LAMP assay for Cff detection. Initially, the DNA extracted from both Cff inoculated and uninoculated bean plants were subjected to conventional PCR, carried out using the primer pair CffFOR2-CffREV4 [16,17]. Positive results were obtained only for Cff inoculated

LAMP Detection of Cff on Artificially and Naturally Infected Bean Samples
Firstly, bean plants cv. "Cannellino" artificially inoculated with Cff ICMP 2584 T by hilum injured method were used to validate the LAMP assay for Cff detection. Initially, the DNA extracted from both Cff inoculated and uninoculated bean plants were subjected to conventional PCR, carried out using the primer pair Cff FOR2-Cff REV4 [16,17]. Positive results were obtained only for Cff inoculated plant samples, where the characteristic single 306 bp amplicon was specifically visualized on agarose gel ( Figure 3A).  When the LAMP reaction was monitored by fluorescent real-time analysis, the DNA extracted from every artificially Cff-inoculated plant showed a positive amplification curve, just after 5 min. No amplification signals were recorded using, as a template, the DNA extracted from control uninoculated plants ( Figure 3B). Data were further confirmed by agarose gel electrophoresis analysis of the LAMP products (data not shown).
This LAMP assay has been evaluated also on naturally Cff-infected bean seeds, where the presence of this phytopathogen is often asymptomatic. To this aim, bean seeds belonging to the cv. "Pak", collected in Iran from fields previously assessed as Cff-infected, were used for DNA extraction When the LAMP reaction was monitored by fluorescent real-time analysis, the DNA extracted from every artificially Cff -inoculated plant showed a positive amplification curve, just after 5 min. No amplification signals were recorded using, as a template, the DNA extracted from control uninoculated plants ( Figure 3B). Data were further confirmed by agarose gel electrophoresis analysis of the LAMP products (data not shown).
This LAMP assay has been evaluated also on naturally Cff -infected bean seeds, where the presence of this phytopathogen is often asymptomatic. To this aim, bean seeds belonging to the cv. "Pak", collected in Iran from fields previously assessed as Cff -infected, were used for DNA extraction [17], and then tested by conventional PCR with the primer pair Cff FOR2/Cff REV4 and by LAMP fluorescence real-time monitoring (data not shown). The infection level was about 72%. Seeds to be considered as Cff infected were those testing positive both by PCR and LAMP assays, regardless of the presence of any symptoms.
A protocol for the direct visual detection of LAMP results was also developed. The pH indicator dye used here turned from pink to yellow, as a consequence of the pH value decrease upon DNA amplification [37]. This protocol for visual LAMP was applied and tested on naturally Cff -infected bean seeds belonging to the cv. "Pak", and highly reproducible results were obtained. As shown in Figure 4, the colorimetric visual readouts of LAMP results were positive within a time period of 15 to 30 min after the start of the incubation at 63 • C, when DNA extracted from the Cff -naturally infected bean seeds cv. "Pak" was used as a template, regardless of whether or not these seeds showed or not any symptoms.  The detection rate was similar to that previously obtained by conventional PCR and LAMP fluorescence monitoring. Samples containing, as a template, DNA from Cff-free certified seeds cv. "Pak" always scored as negative. Controls with DNA template replaced with sterile PCR-grade water also scored as negative. At last, no false positive or false negative results were detected.

Discussion
Cff is an important seedborne bacterial phytopathogen, with a high negative impact on the yields of beans as well as of other legumes [2,4,38,39]. The increasing worldwide spreading of Cff in beanproducing areas is strongly driven by the global expansion of international trades of plant materials, as well as by the multifaceted impact of climate change on plant-pathogen interaction and on weather conducive conditions [40]. It is important to point out that leguminous crops are essential for food security, as they are a staple food and help fighting hunger in less developed countries, and contribute to healthy eating habits worldwide. Furthermore, legumes have a pivotal role in a sustainable agriculture, and in climate change mitigation. Their symbiotic relationship with nitrogenfixing rhizobia bacteria allows these crops to strongly improve soil health, and to reduce the expensive and polluting use of synthetic nitrogen fertilizers in agriculture [41].
Therefore, more research is urgently needed on these crops, as well as on their most important diseases and pathogens, including Cff. As it occurs in other seedborne phytopathogens, the availability of accurate and highly specific seed health tests is among the most important and essential means of control for Cff. Indeed, dramatic outbreaks can easily be caused by inadequate surveillance Figure 4. Visual detection of LAMP products obtained with bean seed samples cv. "Pak" for Cff detection. Visual detection of LAMP products from bean seed samples cv. "Pak" for Cff detection. Positive samples show a color change, from pink (negative samples) to yellow. Uninfected (samples 1 and 2) and naturally Cff -infected (samples 3, 4, 5, 6, 7). Molecular grade water was used as a template in negative controls (sample 8).
The detection rate was similar to that previously obtained by conventional PCR and LAMP fluorescence monitoring. Samples containing, as a template, DNA from Cff -free certified seeds cv. "Pak" always scored as negative. Controls with DNA template replaced with sterile PCR-grade water also scored as negative. At last, no false positive or false negative results were detected.

Discussion
Cff is an important seedborne bacterial phytopathogen, with a high negative impact on the yields of beans as well as of other legumes [2,4,38,39]. The increasing worldwide spreading of Cff in bean-producing areas is strongly driven by the global expansion of international trades of plant materials, as well as by the multifaceted impact of climate change on plant-pathogen interaction and on weather conducive conditions [40]. It is important to point out that leguminous crops are essential for food security, as they are a staple food and help fighting hunger in less developed countries, and contribute to healthy eating habits worldwide. Furthermore, legumes have a pivotal role in a sustainable agriculture, and in climate change mitigation. Their symbiotic relationship with nitrogen-fixing rhizobia bacteria allows these crops to strongly improve soil health, and to reduce the expensive and polluting use of synthetic nitrogen fertilizers in agriculture [41]. Therefore, more research is urgently needed on these crops, as well as on their most important diseases and pathogens, including Cff. As it occurs in other seedborne phytopathogens, the availability of accurate and highly specific seed health tests is among the most important and essential means of control for Cff. Indeed, dramatic outbreaks can easily be caused by inadequate surveillance and quarantine regulatory procedures, following the entry and establishment of a new and emerging plant pathogen. Indeed, the interception of Cff during a port of entry inspection of plant material is extremely difficult, because Cff -infected seeds are mostly asymptomatic [1,38]. Furthermore, the existing assay developed so far for Cff detection on plant material [16,17] requires several specialized equipment and instruments and takes longer times in comparison to other most recent molecular diagnostic approaches such as LAMP [23]. LAMP has also several advantages in comparison to traditional PCR-based approaches, such as its low cost, ease in use, and suitability for in-field/on-site testing.
The only challenging step of the LAMP approach is the correct design of effective primers. Due to the high number of primers needed for each primer set, and the limited options as far as target sequences are concerned, the risk related to the formation of secondary structures can be high. In this paper, we present the development of a LAMP assay for the specific detection of Cff. The newly developed assay targets the unique and highly conserved 306 bp sequence, that was successfully used for the design of the PCR-based assay for Cff and widely applied for in planta testing [16,17].
The optimized conditions established for the Cff -specific LAMP detection are 63 • C as amplification temperature, and 30 min as running time. The analytical sensitivity of the LAMP assay for Cff detection is 100 fg/reaction, which was detected by real-time fluorescence monitoring after 20 min, with 5 min needed to detect 100 ng/reaction. The sensitivity was further confirmed by tests carried out on Cff -infected plant samples. The four LAMP primer set here designed was thus demonstrated to be highly performant, without loop primers, that sometimes have been reported to cause unwanted instability in amplification [42,43].
Lastly, a procedure was developed to visualize LAMP amplification easily and directly through naked eyes, 30 min after isothermal incubation of samples. The colorimetric pH-sensitive indicator used here allowed the avoidance of the usual drawbacks related to other colorimetric indicators, often producing false-positive results [44][45][46]. To the best of our knowledge, this is the first time that a LAMP assay was developed for the specific quali-quantitative detection of Cff. This LAMP assay provides a reliable, specific, and sensitive testing procedure, able to reveal the presence of Cff DNA on bean seeds and other plant materials, rapidly and with the naked eye, thus substantially reducing the risk of disease outbreaks by Cff due to its early detection.
Supplementary Materials: The following are available online at http://www.mdpi.com/2076-2607/8/11/1705/s1, Figure S1. Location of annealing sites of the four Cff -specific LAMP primers here designed on the target sequence. Table S1: Free Energy (∆G; kcal/mol) values for cross-dimers combination, and for self-dimers and harpins formation for the Cff LAMP primers designed in this study.