Multiplex LAMP Detection of the Genus Phytophthora and Four Phytophthora Species P. ramorum, P. lateralis, P. kernoviae, and P. nicotianae, with a Plant Internal Control

Phytophthora species cause destructive plant diseases worldwide. All Phytophthora species, except for one, are listed as plant quarantine organisms in Japan. The exception, Phytophthora nicotianae is considered to be a domestic species. The injurious pests Phytophthora ramorum, Phytophthora lateralis, and Phytophthora kernoviae are invasive pathogens that cause tree mortality worldwide, mainly in the United States and the United Kingdom. To effectively control Phytophthora diseases, we established detection methods that utilize the loop-mediated isothermal amplification (LAMP) of the genus Phytophthora and the four species P. ramorum, P. lateralis, P. kernoviae, and P. nicotianae. LAMP primers for P. ramorum, P. lateralis, and P. kernoviae were newly designed in the present study. Our multiplex assay includes the detection of plant DNA as an internal control. When the optimum ratio between plant and pathogen primers was used in multiplex LAMP assays, 1 pg to 100 fg of pathogen DNA was detected with similar sensitivity to that in simplex LAMP assays. The detection of plant DNA in the absence of pathogens enables us to check for and avoid undesirable negative results caused by enzyme inactivation or the contamination of amplification inhibitors from plant tissues. The total time from sample collection to results is approximately 120 min, and, thus, our multiplex LAMP assay may be used as an accurate and time-saving detection method for Phytophthora pathogens.

Several Phytophthora species cause destructive diseases in forest trees and nursery plants (Lévesque, 2011;Roy et al., 2014;Jung et al., 2018). Three invasive species, Phytophthora ramorum, Phytophthora lateralis, and Phytophthora kernoviae cause widespread tree mortality and are potential threats to forests globally (Brasier et al., 2005;Rizzo et al., 2005;Webber, 2009;Robin et al., 2011). Phytophthora ramorum Werres, de Cock & Man in 't Veld is a causal agent of sudden oak death. It causes bleeding cankers and foliar lesions, which lead to the decline of the tree (Werres et al., 2001). This disease was initially recognized in 1994-1995 and has attacked forest ecosystems in North America (Rizzo and Garbelotto, 2003) and Europe (Brasier et al., 2004). P. ramorum has been reported from at least 109 plant species in 51 genera and 25 families (as of 2019; USDA), and has been detected in 30 countries (Invasive Species Compendium, last updated on 28 Jan 2021). The various host plants of P. ramorum are divided into two groups based on the risk of secondary infection; they are either transmissive (sporulating) or dead-end (non-sporulating) hosts. In forests in California and Oregon in the United States (US), P. ramorum causes trunk cankers on susceptible oak species (Quercus spp.) without sporulation, but does not cause leaf or twig symptoms . On the other hand, bay laurel (Umbellularia californica) infected with P. ramorum shows leaf blight followed by the appearance of abundant sporangia on the leaves (Davidson et al., , 2005. Furthermore, when P. ramorum infected tanoak (Lithocarpus densiflorus) in North America  and also Japanese larch (Larix kaempferi) in the United Kingdom (UK)  sporangia appeared on the foliage and were the source of inocula dispersed by rain. In global terms, the pathogen may spread rapidly and widely due to the growing demand for ornamental trees (Brasier et al., 2004;Brasier and Webber, 2010). Tree species of the genera Camellia, Pieris, Rhododendron, Syringa, and Viburnum are considered to be the key hosts involved in moving this pathogen to new geographical areas (Grünwald et al., 2008).
P. ramorum was initially detected in the UK in 2002 (Lane et al., 2003), and affected locations were surveyed to assess the extent of pathogen occurrence (Brasier et al., 2005). During these surveys, another invasive pathogen, P. kernoviae Brasier, Beales & S. A. Kirk, was isolated in 2003(Brasier et al., 2005. P. kernoviae forms caduceus sporangia on plant foliage, suggesting its aerial or splash dispersal (Brasier et al., 2005). Some Rhododendron plants in the UK were previously found to be infected by both P. ramorum and P. kernoviae (Webber, 2009). Abundant sporangia are produced on the leaves of infected rhododendrons, and this host has played a key role in the spread of P. ramorum and P. kernoviae in the natural environment (Webber, 2009). P. kernoviae has a wide host range, infecting at least 43 plant species in 21 genera and 14 families (USDA Agricultural Research Service). In addition to the UK (Brasier et al., 2005), P. kernoviae has been found in New Zealand (Scott and Williams, 2014) and recently in Chile (Sanfuentes et al., 2016).
P. lateralis Tucker & Milbrath is another tree pathogen that mainly infects Cupressaceae plants. In North America, it has killed Port Orford cedar (Chamaecyparis lawsoniana) and occasionally Pacific yew (Taxus brevifolia) (Hansen et al., 2000). P. lateralis was initially detected in North America in the 1920s (Hansen et al., 2000), and by the 1950s, had spread throughout the native range of Port Orford cedar in southern Oregon and northern California . In 2009, a major outbreak of Port Orford cedar mortality caused by P. lateralis was reported in France (Robin et al., 2011). P. lateralis produces abundant chlamydospores that enable it to survive in dry and hot environments (Jung et al., 2018). This species also produces caduceus sporangia, suggesting its aerial dispersal (Robin et al., 2011).
In the quarantine control of Phytophthora species, such as P. kernoviae, symptomless infections are assumed even if a phytosanitary certificate is attached (Fichtner et al., 2012). Therefore, imported plants must be carefully inspected for the presence of these pathogens, particularly in the case of transmissive host plants. Various effective detection methods have been developed, such as recombinase polymerase amplification-based methods (Miles et al., 2015) and TaqMan-based real-time PCR methods (Schena et al., 2006;Feau et al., 2019). However, high throughput sampling is technically difficult because DNA extraction may be timeconsuming. Therefore, we developed a loop-mediated isothermal amplification (LAMP)-based method (Notomi et al., 2000;Mori et al., 2001;Nagamine et al., 2002;Mori and Notomi, 2009). LAMP has the advantages of high tolerance to amplification inhibitors (Kaneko et al., 2007) and, thus, simple and easy DNA extraction methods are applicable for LAMP-based detection.
LAMP detection multiplexed with an internal control is useful for the accurate detection of plant pathogens. The internal control shows whether the amplification was successful, thereby allowing us to check for and avoid undesirable negative results caused by enzyme inactivation or amplification inhibitors from plant tissues. We used a previously published plant primer set for this internal control (Tomlinson et al., 2010), and made modifications to expand the number of detectable plant families. Regarding pathogen detection, we used primer sets for the detection of all species in the genus Phytophthora, and four species-specific primer sets for the detection of P. ramorum, P. lateralis, P. kernoviae, and P. nicotianae. We designed new primer sets for the three pathogens P. ramorum, P. lateralis, and P. kernoviae. Primers for P. ramorum (Tomlinson et al., 2007) and P. kernoviae (Tomlinson et al., 2010) have previously been reported; however, we designed new primers that were a better fit for our experimental conditions (different enzyme and reaction buffer), reduced undesirable amplifica-tion, and increased the specificity of detection. This was necessary due to the high levels of sequence similarity among the target and non-target species in the primer-tagged DNA region. The non-target species Phytophthora morindae Z. G. Abad & S. C. Nelson (Nelson and Abad, 2010) is phylogenetically closely related to the target species P. kernoviae. In Japan, all species in the genus Phytophthora, except for one, are listed as plant quarantine organisms. The exception is P. nicotianae, which is considered to be a domestic species. In our LAMP-based detection method, we included the genus-specific primer set for Phytophthora (Hieno et al., 2020); therefore, we may identify plant materials infected with any Phytophthora species, and the species-specific primer set for P. nicotianae (Hieno et al., 2019) to identify plants infected only with P. nicotianae. Our multiplex LAMP-based detection method with the internal plant control may be used for the effective quarantine control of Phytophthora pathogens in plants.

Isolates and mycelial DNA extraction
The isolates used in the present study are listed in Table S1. They were grown on V8 agar plates at 25°C (V8 agar [Miller, 1955] was used with the following modifications: 162 mL V8 juice [Campbell Japan] was mixed with 2.5 g CaCO 3 for 30 min, then transferred to a 50-mL tube and centrifuged at 4,000 rpm for 10 min. The supernatant was collected and diluted with deionized water to a total volume of 1 L. Agar [2% {w/v}] was added, and the mixture was autoclaved. The resulting V8 agar medium was plated on 6-cm plastic Petri dishes [5 mL per plate]). DNA were extracted from mycelial colonies using the PrepMan TM Ultra Sample Preparation Reagent (Thermo Fisher Scientific). A small amount of the mycelial mat was collected by scraping with an inoculating needle, transferred to a 1.5-mL Eppendorf tube containing 100 μL of 50% PrepMan Ultra Reagent, and incubated at 100°C for 10 min. After a 3-min incubation at room temperature, the sample was centrifuged at 21,880×g for 3 min. The supernatant (approximately 80 μL) was transferred to a new 1.5-mL tube and the total DNA concentration was measured using the Quanti-Fluor ® dsDNA System (Promega) with a Qubit ® 2.0 Fluorometer (Invitrogen). The concentration was adjusted to 100 pg μL -1 with sterilized deionized water (SDW) and stored at 4°C until used.

LAMP primer design
The sequences of the LAMP primer sets are shown in Table 1. The accession numbers of sequences used for primer design are shown in Fig. S1, S2, S3, and S4. The primer sets for P. ramorum, P. lateralis, and P. kernoviae were designed using Primer Explorer V5 software (Eiken Chemical, https://primerexplorer.jp). The primers for P. ramorum and P. lateralis were based on the cytochrome c oxidase subunit I (cox1) gene of the isolates CBS 101553 (from the CBS-KNAW collection, Westerdijk Fungal Biodiversity Institute, the Netherlands) and CPHST BL 42 (from the USDA-APHIS Center for Plant Health Science and Technology, Beltsville Laboratory, the United States), respectively ( Fig. S1 and S2). The primers for P. kernoviae were based on the rDNA-ITS region of the isolate P1571 (Fig. S3). Each LAMP primer set included the F3, B3, FIP, and BIP primers, but not the F-or B-Loop primer (Table 1 and Fig.  S1, S2, and S3).
The plant LAMP primer set previously reported (Tomlinson et al., 2010) was modified based on multiple alignments in the cox1 gene ( Fig. S4 and Table S2) and additional primers were designed to expand the number of detectable plant families. The plant LAMP primer set consisted of 8 primers: F3, B3G, FIPA (F1cA to  S4).

Extraction of plant DNA
The plant samples used in the present study are listed in Table 2  and Table S3. DNA samples were extracted from inoculated and non-inoculated plants using Kaneka Easy DNA Extraction Kit version 2 (Kaneka). Regarding thin plant tissue samples, such as leaves, buds, and flower buds, a 5-mm square of tissue was collected in a 2-mL screw-cap tube (BIO-BIK, INA OPTIKA) containing a stainless steel bead (0.25 inches, SUS304; AS ONE). Regarding stem and rhizome samples, the tissue was shaved off with a knife and two pieces (lengths of 5 mm) were collected in a tube as described above. Solution A of the Kaneka kit (200 μL) was added and the tube was vortexed using Vortex-Genie2 with a Turbomix attachment (Scientific Industries) for 5 min. The tube was briefly centrifuged to collect all drops at the bottom and then heated at 98°C for 8 min in a heat block. After cooling to room temperature, 28 μL of Solution B was added, mixed well, and then centrifuged at 2,000×g for 3 min in a micro-centrifuge (XX42CF0RT, CHIBITAN-R; Merck Millipore). The supernatant was transferred to a 1.5-mL Eppendorf tube and centrifuged under the same conditions. The supernatant was transferred to a new 1.5-mL tube and diluted 20 times with SDW. Samples were stored at 4°C until used as templates.
When weak or no detection results were obtained, extracted DNA was subjected to additional purification using the OneStep TM PCR Inhibitor Removal Kit (Zymo Research). The non-diluted DNA extract (50 or 100 μL) was purified according to the manufacturer's protocol. Samples were stored at 4°C until used as templates.

Pathogen inoculation
The inoculated host plants and Phytophthora isolates used in the present study are listed in Table 2. Tomato and eggplant fruits inoculated with P. nicotianae and P. capsici were prepared as previously described (Hieno et al., 2020). Since P. nicotianae and P. capsici infect the same host, P. capsici was used in this experiment to confirm the species-specific detection of P. nicotianae.
The attached and detached tree leaves of rhododendron (Rhododendron sp.), Japanese andromeda (Pieris japonica ssp. japonica), and common camellia (Camellia japonica) were also used in this experiment. Isolates of P. ramorum, P. lateralis, P. kernoviae, Phytophthora foliorum, and Phytophthora hibernalis were used to inoculate the leaves. P. foliorum and P. hibernalis are phylogenetically closely related to P. ramorum and P. lateralis; all four species belong to subclade 8c (Abad et al., 2019). Therefore, P. foliorum and P. hibernalis were used to confirm the speciesspecific detection of P. ramorum and P. lateralis. These isolates were grown on V8 agar plates for 4 d, and 7-mm mycelial disks were taken from actively growing colonies. Detached leaves with their petioles wrapped in wet paper were placed on lattices in plastic tube boxes with a wet paper towel at the bottom, and each was inoculated with one or two of the mycelial disks. The boxes were lidded to maintain high humidity. Leaves on trees were inoculated with mycelial disks and wrapped in parafilm, and each entire leaf was then covered with a plastic Ziploc bag. The whole plant was placed in a large plastic case to maintain high humidity. After the inoculation, leaves and plants were incubated in a growth chamber with a cycle of 12 h, 20°C, light and 12 h, 15°C, dark for 3-5 d. After the incubation, 5-mm squares of infected tissue were examined for DNA extraction as described above.

LAMP assays
The LAMP reaction was conducted in a total volume of 25 μL containing 1× Isothermal Master Mix (fluorescent dye; Optigene), 4 mg mL -1 bovine serum albumin (BSA, fraction V; Sigma), primer mixture(s), and a DNA template: 1 μL (default) or 2 μL (for DNA extracted from plants). The 1× primer mixture for each pathogen contained 0.2 μM each of the F3 and B3 primers (0.05 μM for P. nicotianae) and 1.6 μM each of the FIP and BIP primers. The 1× mixture for the previously designed plant LAMP primers (Tomlinson et al., 2010) contained 0.2 μM each of the F3 and B3 primers, 1.6 μM each of the FIP and BIP primers, and 0.8 μM each of the F-Loop and B-loop primers. The 1× mixture for the plant LAMP primer set modified in the present study contained 0.2 μM each of the F3 and B3G primers, 0.8 μM each of the FIPA and FIP2 primers, 1.6 μM of the BIPA primer, 0.4 μM each of the F-Loop and F-Loop2 primers, and 0.8 μM of the B-loop primer. In simplex LAMP, the primer mixtures for the pathogens were used at 1× concentration and the previously designed or modified plant primer mixture were used at 0.08× concentration. In multiplex LAMP, the primer mixtures for the pathogens were used at 1× concentration. The modified plant primer mixture was tested at 0.2×, 0.15×, 0.1×, 0.08×, and 0.05× concentrations. A reaction mix containing SDW instead of DNA was used as a negative control. The LAMP assay was performed on the portable real-time fluorometer Genie ® II (Optigene) with the following conditions: pre-heat at 65 or 68°C for 5 min; amplification at 65 or 68°C for 60 min; and an annealing curve analysis at 98 to 80°C, ramping at 0.05°C s -1 . The temperatures of the pre-heat and amplification steps were optimized at 65°C for all primer sets, except Phytophthora genusspecific primers, which were optimized at 68°C. Therefore, multiplex assays were performed at 65°C for all assays, except those that included the Phytophthora genus-specific primers, which were performed at 68°C. The format of the raw data (.gen) was changed to a text file using Genie Explorer version 2.0.6.3 software (Optigene), and amplification curves and annealing curves were then created using Microsoft Excel. The sensitivity of each LAMP assay was assessed with at least three experiments. The specificity and detectability of each assay was evaluated at least twice for each DNA sample.

Primer design and specificity tests of species-specific LAMP primer sets
Regarding the species-specific detection of P. ramorum, P. lateralis, and P. kernoviae, we designed LAMP primer sets (Table 1) based on multiple alignments of the cox1 gene for P. ramorum and P. lateralis ( Fig. S1 and S2), and the rDNA-ITS sequence for P. kernoviae (Fig. S3). The specificity of each primer set was tested using mycelial DNA, which was extracted from Phytophthora and closely related genera. There were 61 taxa including subspecies (101 isolates) of Phytophthora, 12 species (12 isolates) of Pythium, one species of Phytopythium, and one isolate each of the following soil-borne pathogens: Aphanomyces cochlioides, Fusarium oxysporum f. sp. fragariae, Plasmodiophora brassicae, Rhizoctonia solani, Saprolegnia parasitica, Sclerotinia sclerotiorum, and Verticillium albo-atrum (Table  S1). The results shown in Table S1 indicated that the primer sets were specific for each of the three species.

Primer modification and evaluation of the detection range of the plant LAMP primer set
We modified a previously designed plant LAMP primer set (Tomlinson et al., 2010) to expand the number of detectable plant families. Based on multiple alignments of the cox1 gene ( Fig. S4 and Table S2), primer sequences were modified and additional primers were designed (Table 1). The modified plant LAMP primer set was tested with DNA extracted from 176 plant species representing 155 genera and 110 families (Table S3). The reaction was performed using a 0.08× primer mixture at 65°C for 60 min; this concentration was selected as the optimum for species-specific multiplex LAMP reactions at 65°C. Under these conditions, we detected 140 plant species of 124 genera belonging to 87 families, including 10 species of 8 genera belonging to 5 families that were not amplified using the previously designed primers (Tomlinson et al., 2010) (Table S3). Among the weakly and non-detected species, additional DNA purification improved detectability with the modified primer set for 12 plant species of 12 genera belonging to 12 families (Table S3).

Sensitivity of species-specific LAMP primer sets
To test the detection limits of the LAMP primers for P. ramorum, P. lateralis, and P. kernoviae, we used mycelial DNA extracted from each species, with amounts ranging between 1 fg and 100 pg. P. ramorum PR-06-021, Pr-1, and CBS 101553; P. lateralis WPC P3361 (from the World Phytophthora Genetic Resource Collection, the United States), 450, and ATCC 201856 (from the American Type Culture Collection, the United States); and P. kernoviae Pk-1, 2654, and P1571 were tested. The specificity of each amplification was confirmed with a single peak in the annealing curve analysis at approximately 82°C for P. ramorum (Fig. S5D), 83°C for P. lateralis (Fig. S5E), and 86°C for P. kernoviae (Fig. S5F). The detection limits for P. ramorum (isolates PR-06-021 and Pr-1), P. lateralis (isolates WPC P3361 and ATCC 201856), and P. kernoviae (isolates Pk-1 and P1571) were 10, 100, and 10 fg, respectively. Representative data are shown in Fig. S5. On the other hand, we found 10-fold lower sensitivities for P. ramorum isolate CBS 101553, P. lateralis isolate 450, and P. kernoviae isolate 2654 (data not shown).

Selection of optimum primer ratios for multiplex LAMP assays
The optimum primer ratios for multiplex LAMP assays designed to detect pathogens in plant samples were selected. "Optimum" indicates the concentration of plant primers that may be used without disturbing pathogen detection, while functioning as an internal control for non-infected plants. We used the 1× concentration of the pathogen primer mixes combined with different concentrations of the plant primer mix, and tested these mixtures using reaction temperatures of 65 or 68°C in the pre-heat and amplification steps.
To select the optimum ratio in the species-specific multiplex LAMP detection of P. ramorum, P. lateralis, P. kernoviae, and P. nicotianae, we used the P. ramorum primers in reactions performed at 65°C. The P. ramorum species-specific primer set (at the 1× concentration) was mixed with different concentrations of the modified plant primer set (0.15×, 0.1×, 0.08×, and 0.05×), and then tested in assays with mycelial DNA extracted from P. ramorum Pr-1 and/or plant DNA extracted from the leaves of Rhododendron sp. (Fig. 1). Four samples were prepared and used for these multiplex LAMP assays: 1) SDW as a negative control, 2) plant DNA only (71.5 ng), 3) pathogen (P. ramorum) DNA only (1 pg), and 4) a mixture of plant DNA (71.5 ng) and pathogen DNA (1 pg). In reactions containing the 0.15× or 0.1× plant primers with both types of DNA, pathogen DNA (peak at approximately 82°C) and plant DNA (peak at approximately 84°C) were both simultaneously detected (Fig. 1A and B). In these cases, the peaks from pathogen DNA were lower than those when only pathogen DNA was included in the reaction. This suggested that pathogen detection was disturbed by competitive amplification with plant DNA. In reactions with the 0.08× or 0.05× plant primer and both types of DNA, only pathogen DNA was detected (Fig. 1C and D). In the reaction with the 0.08× plant primer and both DNAs, the pathogen peak was the reactions (Fig. 1D). Based on these results, the optimum higher than that when only pathogen DNA was included primer ratio for the reaction temperature of 65°C was 1× P. (Fig. 1C), implying that detection of the pathogen was not ramorum species-specific primers with 0.08× plant primers. disturbed. The 0.05× plant primer mix was not able to detect With this ratio, there was no significant interference with plant DNA (missing peak at approximately 84°C) in any of pathogen detection, and plant DNA was still detected in the absence of pathogen DNA. This ratio was also optimum with the other pathogen primers (data not shown).
In multiplex LAMP assays designed to specifically detect all Phytophthora species, the amplification reaction was performed at 68°C. We used Phytophthora genus-specific primers (1×) combined with different concentrations of the modified plant primers (0.2×, 0.15×, 0.1×, and 0.05×) and assayed with the DNA samples described above. In reactions containing the 0.2× plant primer and both types of DNA, pathogen DNA (peak at approximately 86°C) and plant DNA (peak at approximately 84°C) were simultaneously detected (Fig. 1E). The pathogen peak was lower than that when plant DNA was absent. This suggested that the amplification of pathogen DNA was under competition. When the reaction contained the 0.15×, 0.1×, or 0.05× plant primer and both types of DNA, only pathogen DNA was detected (Fig. 1F, G, and H). Furthermore, the 0.05× plant primer gave the very weak detection of plant DNA in the reaction containing only plant DNA (Fig. 1H). Based on these results, the optimum primer ratio in reactions at 68°C was 1× Phytophthora genus-specific primers with 0.15× plant primers because this ratio showed no evidence of interference in the amplification of pathogen DNA, but also allowed for the amplification of plant DNA when pathogen DNA was absent.

Sensitivity of multiplex LAMP assays for the detection of pathogens
The pathogen detection limits of multiplex LAMP assays were tested using specific primer sets for the genus Phytophthora (Hieno et al., 2020), P. ramorum, P. lateralis, and P. kernoviae (the present study), and P. nicotianae (Hieno et al., 2019) multiplexed with modified plant primers (Tomlinson et al., 2010). These primer sets were mixed at the optimized ratios described above. Mycelial DNA (ranging between 1 fg and 100 pg) and/or plant DNA (71.5 ng) extracted from Rhododendron sp. leaves were used for the assay. Fig. 2 shows representative data for assays with 100 fg, 1 pg, and 100 pg mycelial DNA. Pathogen DNAs were detected as peaks at approximately 82°C for P. ramorum Pr-1 ( Fig. 2A), 83°C for P. lateralis WPC P3361 (Fig. 2B), 86°C for P. kernoviae P1571 (Fig. 2C), and 87°C for P. nicotianae CBS 305.29 (Fig. 2D). When Phytophthora genus-specific primers were used to detect P. ramorum Pr-1 DNA, the peak occurred at approximately 85°C (Fig. 2E). Plant DNA was detected as a peak at approximately 84°C ( Fig. 2A, B, C, D, and E). The detection limits were 100 fg in multiplex LAMP assays using the primer sets for P. ramorum, P. kernoviae, and the genus Phytophthora ( Fig. 2A, C, and E), and 1 pg for assays with the primer sets for P. lateralis and P. nicotianae ( Fig. 2B and  D).

Detection of pathogens in inoculated plants
Multiplex LAMP assays for the detection of pathogens in inoculated plants were tested ( Fig. 3 and Table 2). The attached and detached leaves of rhododendron (Rhododendron sp.) and Japanese andromeda (Pieris japonica ssp. japonica) were inoculated with P. ramorum, P. lateralis, P. kernoviae, P. foliorum, and P. hibernalis. The detached leaves of common camellia (Camellia japonica), which were inoculated with P. ramorum, P. lateralis, and P. kernoviae, were examined. Tomato (Solanum lycopersicum) and eggplant (Solanum melongena) fruits were inoculated with P. capsici and P. nicotianae. On days three to five postinoculation, extracted DNA from the infected tissues were tested for the detection of the pathogen using multiplex LAMP assays. Target pathogen DNA with its specific primer set was detected as a single peak at the same temperatures as those described above for the sensitivity assays (see Fig. 2). Representative data are shown in Fig. 3. In control assays of samples from non-inoculated plants, plant DNA was amplified with a peak at approximately 84°C (Fig. 3). Therefore, multiplex LAMP assays detected infection by the target pathogen, and the modified plant primers functioned as an internal control.

Discussion
To detect important pathogenic species of the genus Phytophthora, we designed species-specific LAMP primer sets for P. ramorum, P. lateralis, and P. kernoviae and tested their specificity with numerous Phytophthora species. Intraspecies variation was present in the genome regions used for primer design. Therefore, we retrieved sequence data for P. ramorum (28 isolates), P. lateralis (42 isolates), and P. kernoviae (22 isolates) from the NCBI nucleotide database (https://www.ncbi.nlm.nih.gov/nucleotide/) to confirm the conservation of the species-specific sequences in the designed primers. In addition, the primer sets were tested with mycelial DNA from the isolates of P. ramorum (30 isolates), P. lateralis (4 isolates), and P. kernoviae (9 isolates) that had originated from different countries and/or host plants. All the isolates tested were detected by each speciesspecific primer. Therefore, intra-species sequence variations among the isolates examined did not affect our speciesspecific detection.
We modified a previously reported plant LAMP primer set (Tomlinson et al., 2010) based on a multiple alignment analysis of plant DNA (Fig. S4). Additional FIP2 and F-loop2 primers were designed that annealed to plant DNA not amplified by the original primers. Additionally, the original primers contained mixed bases at some sites, and we selected the bases at these sites that were least likely to result in primer dimerization. The modified primer set (Table 1) was able to detect 140 plant species of 124 genera belonging to 89 families, and these included 10 plant species of 8 genera belonging to 5 families that were not detected with the unmodified primers (Table S3). The newly detectable species were mainly in the Pinales (Cupressaceae, Pinaceae, and Taxaceae). To the best of our knowledge, our plant LAMP primer set has the broadest detectability available. Plants containing saponin, such as Pittosporum tobira, were more difficult to detect (Vincken et al., 2007;Itamiya and Kikkawa, 2016); however, additional DNA purification also improved their detection.
An internal control (the detection of plant DNA) is important when screening plant samples for infecting pathogens because it shows that negative results are not due to poor DNA extraction or other errors. The simultaneous amplification of the pathogen and plant DNA by multiplex LAMP is preferred, but is very difficult. In natural samples, the relative amounts of DNA originating from pathogens and plants markedly vary depending on the severity of the infection. In LAMP assays, the strong amplification of more abundant DNA may overwhelm the amplification of less abundant DNA (Kubota and Jenkins, 2015). Furthermore, the sensitivity of the assay may be reduced in multiplex reactions (Kubota and Jenkins, 2015;Yasuhara-Bell et al., 2018). Based on these findings, it is very important to optimize the primer ratio. We found optimum ratios of 1× species-specific primers with 0.08× plant primers for assays at 65°C, and 1× genus-specific primers with 0.15× plant primers for assays at 68°C (Fig. 1). The concentrations of the pathogen primers were similar to those in simplex LAMP to maintain sensitivity, and the concentration of the plant primer set was markedly lower in order to prevent pathogen detection being impeded by the presence of plant primers. Under these optimum primer ratios, we detected the target DNA of each pathogen in multiplex assays with similar sensitivity to simplex assays (between 1 pg and 100 fg) even though plant DNA was present at a markedly larger amount (71.5 ng) than pathogen DNA (Fig. 2). When we tested low concentrations of pathogen DNA (1 pg for P. lateralis and 100 fg for P. ramorum), the annealing peak of pathogen DNA was indistinct or formed dual peaks with plant DNA (Fig. 2). This result suggests that the amplification of pathogen DNA was slightly reduced under competition with plant DNA. Nevertheless, we selected the primer ratios indicated above because 0.08× was the lowest concentration of plant primers that clearly detected plant DNA (Fig. 1) . 3. Detection of Phytophthora pathogens in inoculated plants using multiplex LAMP assays. Plant DNA extracted from inoculated plants and mycelial DNA (100 pg per reaction) extracted from each Phytophthora species were subjected to multiplex LAMP assays, with SDW and noninoculated plants as controls. Plant primers were mixed with specific pathogen primers and tested with plant and mycelial DNA as follows: (A) P. ramorum primers, detached Japanese andromeda leaves inoculated with P. ramorum CBS 101330, and mycelial DNA from P. ramorum Pr-1; (B) P. lateralis primers, detached Japanese andromeda leaves inoculated with P. lateralis CBS 168.42, and mycelial DNA from P. lateralis WPC P3361; (C) P. kernoviae primers, detached Japanese andromeda leaves inoculated with P. kernoviae CBS 122051, and mycelial DNA from P. kernoviae P1571; (D) P. nicotianae primers, tomato fruits inoculated with P. nicotianae GK10Eg1, and mycelial DNA from P. nicotianae CBS 305.29; and (E) Phytophthora genus-specific primers, rhododendron leaves inoculated with P. ramorum CBS 101330, and mycelial DNA from P. ramorum Pr-1. Plant:pathogen primer ratios were 0.08:1 in (A, B, C, and D) and 0.15:1 in (E). After amplification at 65°C (A, B, C, and D) or 68°C (E) for 60 min, fluorescence derivative data during the annealing phase (98 to 80°C) were obtained. Open and black arrowheads indicate peaks derived from plant DNA and Phytophthora species DNA, respectively. SDW: sterilized deionized water. See Table 2 for more results.
In the present study, a multiplex LAMP detection method was developed for the genus Phytophthora and four species in the genus, P. ramorum, P. lateralis, and P. kernoviae, and P. nicotianae. Our multiplex assay includes the detection of plant DNA as an internal control. The detection of plant DNA in the absence of pathogens enables us to prevent undesirable negative results. The total time from sample collection to results is approximately 120 min. Therefore, our multiplex LAMP assay may be used as an accurate and time-saving detection method for Phytophthora pathogens.