Resistance to Phytophthora infestans in tomato wild relatives

1 Plant Pathology Research Institute, Agricultural Research Center, Giza, 12619, Egypt. 2 Department of Plant Pathology, Faculty of Agriculture, Cairo University, Giza, 12613, Egypt. 3 Kazusa DNA Research Institute, 2-6-7 Kazusa-Kamatari, Kisarazu, Chiba 292-0818, Japan. 4 Horticulture Department, Faculty of Agriculture, University of Kafrelsheikh, Kafr El-Sheikh 33516, Egypt. 5 World Vegetable Center, P. O. Box 42, Shanhua, Tainan 74199, Taiwan.


INTRODUCTION
Economically, tomato (Solanum lycopersicum L.) is the fourth most important crop in the world after rice, wheat, and soybean (Nowicki et al., 2013), with global production of 162 million tons and a net value of more than $62 billion in 2014 (FAO, 2017).In Egypt, the tomato crop production value was estimated at $1.7 billion in 2014 (FAO, 2017).Diseases caused by different organisms including fungi, bacteria, virus, and nematodes can limit tomato production.Late blight, caused by the hemibiotrophic oomycete Phytophthora infestans (Mont.)de Bary, is considered a major threat to tomato production in tropical and subtropical regions (Lima et al., 2009;Elsayed et al., 2012).The pathogen attacks all above-ground parts of the plant including leaves, petioles, stems and fruit at any growth stage, causing blights, necrosis, blotches and rots that reduce yield and fruit quality (Lievens et al., 2004).The disease can spread and kill plants rapidly when favorable environmental conditions of high humidity and low temperature (18°C) prevail (Haq et al., 2008;Stroud et al., 2016).
P. infestans is a highly diverse pathogen that can reproduce sexually and asexually.During asexual reproduction P. infestans produces lemon-shaped sporangia, while sexual reproduction results in the production of large, thick-walled oospores capable of surviving for several years in soil or plant debris (Smart and Fry, 2001).The widespread potential of sexual reproduction may increase the risk of host resistance breakdown, due to the development of new aggressive races and genotypes (McDonald and Linde, 2002).Recently, the genetic diversity in P. infestans is most often characterized using simple sequence repeat (SSRs) markers (Cooke and Lees, 2004;Lees et al., 2006;Li et al., 2010Li et al., , 2013)).The data is used to assign genotypes to isolates, for example in Europe Cooke et al. (2012) uses a system where 2 to 25 alleles per locus were detected in many European isolates using 11 SSR markers.These genotypes, however, do not reflect information regarding the specific race of the pathogen, with the same SSR genotype containing different races (Li et al., 2012) Late blight management strategies include the use of several fungicides.Apart from the economic impact of their use, fungicides have been shown to reduce efficacy and resistance problems, particularly formulations containing metalaxyl (mefenoxam) (Randall et al., 2014;Saville et al., 2015;Montes et al., 2016).It can be difficult to detect P. infestans in the field during the initial stages of infection, from where it can rapidly develop into severe epidemics due to the pathogen's short life cycle.Most fungicides are ineffective once the pathogen had been established in the field.Host resistance has the potential for being a key component in managing late blight of tomato; it would reduce fungicide use and provide costeffective, environmentally safe management strategies against the pathogen.
A few resistant varieties of tomato have been developed through the introgression of resistance genes obtained from wild tomato species (Panthee and Gardner, 2010).To date, five late blight resistance genes have been identified.Ph-1, a single dominant allele located on chromosome 7, confers resistance to P. infestans race 0 but it has been overcome by new races of the pathogen.
The second gene, Ph-2, a partially dominant allele located on chromosome 10, confers partial resistance to some P. infestans isolates and often fails in the presence of aggressive isolates.Ph-3, a single dominant allele located on chromosome 9, provides increased resistance against some aggressive isolates like Pi-16 from Taiwan that can overcome Ph-1 and Ph-2 (Irzhansky and Cohen, 2006).However, Ph-3 can also be overcome by some isolates (Kim and Mutschler, 2006;Chen et al., 2008;de Miranda et al., 2010).A new resistant gene, Ph-5, has been identified recently on chromosome 1 and 10, and confers strong resistance to several P. infestans isolates including those overcoming the aforementioned four resistance genes (Foolad et al., 2008).Currently, the Ph-2 and Ph-3 genes are available in tomato breeding lines (e.g.'NC1 CELBR', 'NC2 CELBR') and hybrid cultivars (Gardner and Panthee, 2010;Zhang et al., 2014).The continuous cycle of resistance being overcome by new P. infestans races warrants continuous efforts to identify additional sources of resistance from commercial cultivars, or other wild relatives of tomato in order to improve future breeding programs.The objectives of this study were to (i) investigate the level of resistance to P. infestans in wild relatives S. habrochaites, S. pennellii, S. pimpinellifolium and S. peruvianum and (ii) determine whether tomato accessions containing late blight resistance genes (Ph-1, Ph-2 and Ph-3) could provide acceptable resistance to P. infestans isolates present in Egypt.

Plant material and growth condition
Forty-eight wild tomato accessions were evaluated, including 24 accessions of S. pimpinellifolium, 12 accessions of S. habrochaites, eight accessions of S. pennellii, and four accessions of S. peruvianum, along with four accessions of S. lycopersicum containing Ph resistance genes.Seed was obtained from the C. M. Rick Tomato Genetics Resource Center (TGRC), University of California, Davis (LA numbers).Tomato variety 'Super Strain B', received from the Horticulture Research Institute, Agricultural Research Center (ARC), Egypt, which is known to be susceptible to late blight, was included as a control.In the greenhouse (25±2 °C, 16/8 h day/night), seeds of all accessions and the susceptible control were sown in 209-cell seedling trays containing 40 ml of peat moss-vermiculite mixture (1:1 volume) per cell plug.Plants were watered daily and fertilized weekly with N: P: K 15-15-15.Five weeks after sowing, seedlings of accession and control plants were transplanted into 20 cm pots containing potting soil, which were used for whole plant assays in 2013 and 2014.Seven-week-old plants were moved from the greenhouse to a growth room (20 ± 2°C, 90% relative humidity (RH), 16/8 h day/night) at the Plant Pathology Research Institute at ARC for inoculation assays.

Isolate selection and maintenance
Fourteen P. infestans isolates were collected from naturally infected tomato plants from 2013 to 2014 epidemics occurring in Beheira, Means disease severity rating followed by ± standard error.Means followed by different letters are significantly (P = 0.05) different, whereas means followed by the same letter are not significantly different.
The susceptible 'Super strain B' control was artificially inoculated with the 10 P. infestans isolates collected in 2013, as described for the germplasm testing, in order to select the most virulent isolate for late blight screening.Among the 10 evaluated isolates, EG_7 was the most virulent (Table 1).Isolate EG_7 was identified as genotype 23_A1_12 and mating type A1 based on identification carried out by the laboratory of Dr. David E. L. Cooke at the James Hutton Institute, Dundee, UK.This isolate was chosen as the inoculum source in the first round of germplasm tests conducted in 2013 and 2014.In 2014, another four P. infestans isolates were collected from naturally infected tomato plants (Table 1).These isolates were used to re-test germplasm from the first round of testing, due to the severe epidemics they caused in tomato fields in 2014.

Inoculum production
Sporangia and zoospores were produced as described by Chen et al. (2008).Tomato leaves collected from 6-week-old plants of the susceptible 'Super strain B' genotype were placed on moist filter paper in 140 mm sterilized Petri plates.The abaxial surface of the leaves were injured at the center using a sterile 10 µl micropipette tip and inoculated with 30 µl of a sporangial suspension obtained from 20-day-old rye agar plates.Leaflets were incubated for 48 h at 18°C in darkness, followed by incubation at 18°C for 10 days with a 12-h photoperiod.Subsequently, tomato leaflets were placed in a glass beaker containing 500 mL sterilized distilled water and gently shaken using a vortex to dislodge sporangia from the leaflets.The suspension was filtered through four layers of sterile muslin cloth.The concentration of the sporangia was determined using a haemocytometer, and was adjusted to 15×10 4 sporangia/ml.The suspension was chilled at 4°C for 2 to 4 h prior to inoculation to encourage zoospore release from the sporangia.

Disease assessment using P. infestans isolate EG_7
Eight-week-old plants of 48 wild accessions along with four tomato genotypes containing Ph genes and a susceptible control were inoculated with P. infestans isolate EG_7 using whole plant assays.
Plants were sprayed with a suspension of P. infestans zoospores using a hand sprayer until complete leaf coverage and excess runoff was observed.The inoculated plants were covered with a plastic tunnel to increase humidity and kept at 16 to 18°C in the dark in a growth chamber for 24 h under 100% RH.The inoculated plants were then grown at 18 to 20°C and 90% RH with a 12 h photoperiod for 7 to 12 days.Plants were arranged following a completely randomized design with five replicate plants per accession.
The experiment was conducted twice from November to January in 2013 and 2014.Plants were evaluated individually at 10 days post-inoculation (dpi) by visually scoring the disease severity using a scale of 0 to 6 as described by Chen et al. (2014), where 0 indicates no symptoms (immune); 1 indicates <5% leaf area affected and small lesions (highly resistant); 2 indicates 6 to 15% leaf area affected and restricted lesions (resistant); 3 indicates 16 to 30% leaf area affected and/or water-soaked flecks on stems (moderate resistant); 4 indicates 31 to 60% leaf area affected and/or a few stem lesions (moderate susceptible); 5 indicates 61 to 90% leaf area affected and expanding stem lesions (susceptible); 6 indicates 91 to 100% of leaf area affected, extensive stem damage, or a dead plant (highly susceptible).

Reassessment of disease resistance using P. infestans isolates EG_9, EG_10, EG_11 and EG_12
The four P. infestans isolates collected in 2014 were used to confirm the resistance identified in S. habrochaites accession LA1777, and to re-access the response of the five tomato accessions containing Ph resistance genes against different P. infestans isolates.The susceptible control 'Super Strain B' was also included in the re-testing.Preparation of isolates and disease assessments were carried out as described previously.The experiment was repeated twice only with LA1777 along with the susceptible control 'Super Strain B' inoculated with isolate EG_11.

Statistical analysis
Statistical procedures were performed using the statistical software SAS (version 9.1; SAS Institute, Cary, NC).The percentage of late blight severity was transformed using Arcsine square root transformation.Back-transformed data are presented in tables.All data were subjected to one-way analysis of variance (ANOVA).Mean separations were determined using the Tukey-Kramer honestly significant difference (HSD) test (P = 0.05).

Isolate selection
To identify the most virulent isolate, 10 P. infestans isolates collected from tomato fields in 2013, were artificially inoculated on the susceptible control 'Super Strain B' under controlled greenhouse conditions.All P. infestans isolates infected the susceptible control, with disease severity ranging from 99% for the most aggressive isolate EG_7 to 22.7% for the least aggressive isolate EG_2 (Table 1).In addition to isolate EG_7, isolates EG_5 and EG_8 also exhibited high virulence causing disease severities of 75 and 80%, respectively.Based on these results, P. infestans isolate EG_7 was selected and used to further screen late blight resistance in 48 tomato wild tomato relatives along with tomato genotypes containing the Ph genes and the susceptible control.

Disease assessment using P. infestans isolate EG_7
Late blight severity on the 48 wild accessions, four tomato genotypes containing Ph genes, and the susceptible control 'Super Strain B', was evaluated using a whole plant assay under controlled greenhouse conditions (Table 2).The ANOVA revealed highly significant differences among treatments (P < 0.0001).Results from our two experiments conducted in 2013 and 2014 were similar.No genotype was immune to P. infestans EG_7 and all susceptible control plants had a 100% blight severity (dead).Mean disease ratings of all tested accessions eight weeks after sowing ranged from 4.5 to 100% (Table 2).Lower disease severities were identified for S. habrochaites accessions LA1777, LA1352, LA2855, LA1347, LA1718 and LA1295, with disease severities ranging from 4.5 to 13.5%.S. lycopersicum accessions containing Ph-2 and Ph-3 genes had significantly lower disease severity values, whereas LA3152 (Ph-2), LA3151 (Ph-2) and LA4286 (Ph-3) were moderately resistant.Conversely, all S. peruvianum and S. pennellii accessions were susceptible or highly susceptible to isolate EG_7.The majority of the evaluated S. pimpinellifolium accessions were either susceptible or highly susceptible to EG_7, whereas LA 1269 and LA1578 were moderately resistant with disease severity 22 and 23%, respectively.No correlation was seen between the geographic origin of tomato genotypes and their resistance to P. infestans.For example, all resistant accessions in this study originated from Peru and Ecuador, but many susceptible accessions also came from these regions.

Reassessment of disease resistance using P. infestans isolates EG_9, EG_10, EG_11 and EG_12
The late blight resistance identified in the 2013 and 2014 experiments in S. habrochaites accession LA1777 and the S. lycopresicum genotypes containing Ph genes was verified using a new set of P. infestans isolates including EG_9, EG_10, EG_11 and EG_12, which were collected from tomato epidemcis in 2014 (Table 3).Responses to individual isolates varied.Genotypes inoculated with the aggressive isolate EG_12 had 13.3 to 100% disease ratings, while genotypes inoculated with EG_10, the least aggressive isolate, had disease severities of 0 to 78.3% (Table 3).LA1777 was the most resistant genotype, exhibiting the lowest disease severity values against all selected isolates of P. infestans compared to other genotypes containing Ph genes and the susceptible control, with disease severity means ranging from 0 to 13.3%.The susceptible control had a significantly higher disease severity compared to LA1777 and tomato genotypes containing Ph genes when inoculated with isolates EG_10 and EG_11.Interestingly, all S. lycopersicum accessions containing Ph-2 and Ph-3 genes were found to be moderately resistant or resistant to isolates EG_9 and EG_10, but were susceptible to isolates EG_11 and EG_12 with mean severity ratings > 68.3%.

DISCUSSION
Late blight is one of the most devastating diseases of tomato, especially in cool and moist environments.In Egypt, it is almost impossible for grower's to produce tomatoes from November to February, when cool and moist weather predominate and favour P. infestans epidemics.Knowledge on the effectiveness of resistance genes in existing cultivars can help growers in Egypt to manage late blight.Efforts must thus continue to assess the stability of resistance and whether such cultivars are adaptable to all tomato-growing regions.
commonly grown or introduced in Egypt have not been assessed for their resistance to local populations of P. infestans.Therefore, a large number of accessions and genotypes were screened for their resistance to late blight using a highly virulent isolate EG_7 under controlled greenhouse conditions.Our results showed that S. habrochaites accessions LA 1777, LA1352, LA2855, LA 1347, LA1718 and LA1295 exhibited high levels of resistance to late blight, with LA1777 being the most resistant.Similar results have been reported in S. habrochaites accessions by Abreu et al. (2008) in BGH6902, Brouwer et al. (2004) in LA2099, and Li et al. (2011) in LA1777.
Although the resistance mechanism is unknown, these accessions have high densities of glandular trichomes type IV and VI (Muigai et al., 2003;Momotaz et al., 2010;Bergau et al., 2015), which are often able to secrete exudates with antifungal activities as has been shown in a wild potato species (S. berthaulltii) that are resistant to P. infestans (Lai et al., 2000).Additional resistance mechanisms in LA1777 could be the secretion of a wide range of proteins that have been shown in other plants to play a role in the degradation of microbial cell walls, and in blocking pathogen-released elicitors (Ferreira et al., 2007).The resistance present in LA 1777 will, however, be difficult to confer to tomato since S. habrochaites is a distant relative of cultivated tomato and challenges such as self-incompatibility, segregation distortion, and linkage drag complicate the transfer of useful genes into tomato (Covey et al., 2010;Elizondo and Oyanedel, 2010;Labate and Robertson, 2012;Haggard et al., 2013).The sequencing of the tomato genome will be helpful in this regard since it has resulted in the availability of thousands of molecular markers that can facilitate the mapping and introgression of beneficial genes from wild species into cultivated tomato.
Tomato genotypes containing Ph-2 and Ph-3 genes, LA3152, LA3151 and LA4286, demonstrated acceptable levels of resistance when inoculated with P. infestans EG_7.These genes have thus far been used in commercial cultivars and may directly benefit farmers by reducing their dependence on fungicides to control P. infestans, thus lowering their production cost.These results support previous reports indicating that Ph-2 and Ph-3 confers resistance to P. infestans.Ph-2 is a completely dominant gene that confers partial resistance to P. infestans and in this case isolate EG_7, and Ph-3 is a partial dominant allele that confers stronger resistance to P. infestans when it is homozygous.Notably, resistance to P. infestans EG_9, and EG_10 isolates exhibited by accession LA3152 (containing Ph-2) was significantly higher than the resistant accessions LA3151, LA1269 and LA4286 containing Ph-2 or Ph-3, suggesting the presence of unknown resistance gene(s) potentially involved in resistance to late blight in accession LA3152.Further, P. infestans is a highly diverse pathogen that can develop virulence toward the Ph resistance genes (Goodwin et al., 1995;Bradshaw et al., 2006;Chen et al., 2014).The fluctuation of P. infestans populations may be due to sexual recombination and transposable elements (Vetukuri et al., 2012).Based on the results of this study, tomato accessions possessing Ph-2 and Ph-3 resistance genes were susceptible to P. infestans isolates EG_12 and EG_11, whereas LA1777 had a high level of resistance against all five evaluated P. infestans isolates.In addition, the tomato accession possessing the resistance gene Ph-1 was also susceptible to all five P. infestans isolates EG_7, EG_9, EG_10, EG_11, and EG_12.
The differences in the effect of Ph genes against different P. infestans isolates could be due to isolates used in this study probably having different race compositions.For example, isolates EG_2 and EG_7 were collected from the same county, but differed in their virulence to the host plant.Differences between these isolates have significant implications for local tomato breeding programs against P. infestans in Egypt, where the diversity of pathogen populations, the existence of different physiological races within P. infestans, and polygenic inheritance can hamper breeding efforts (Singh and Singh, 2006;Zhang and Kim, 2007;Harbaoui et al., 2011;Pule et al., 2013;Tian et al., 2016).Further studies are required to characterize P. infestans populations, define races that infect tomato, and determine their spatial structure in order to deploy tomato breeding programs effectively in Egypt.
Only two accessions of S. pimpinellifolium, LA1269 and LA1578, were found to be moderately resistant.These red-fruited accessions are closely related to tomato and therefore fewer backcrosses may be required to introgress late blight resistance into tomato compared to the green-fruited species S. pennellii and S. habrochaites (Rick, 1971;Peralta et al., 2008).LA1269, also known in the literature as L3708, was described as being resistant to multiple strains of P. infestans (Chunwongse et al., 2002).Such resistance has been linked to the Ph-3 gene and qPh2.1 QTL, which confers resistance to isolate Pi733 in Taiwan (Black et al., 1996;Chen et al., 2014).Although breeders might be interested in attaining high levels of resistance, the use of partially resistant cultivars can be useful and can help to reduce the number of fungicide applications (Stevenson et al., 2007).Interestingly, some accessions found to be resistant in other studies were moderate or susceptible in this study.For example, S. pennellii LA716 was reported as highly resistant by Eshed and Zamir (1994) and Smart et al. (2007), but was found to be susceptible in this study with a mean rating of 5.0.Conversely, the most resistant S. habrochaites accession LA1352 identified in this study was highly susceptible to aggressive P. infestans isolate T1, 2, 3, 4 in Taiwan, with a mean rating 6. Differences in resistance in different studies could be due to the use of different experimental designs or more likely differences in P. infestans races.sequence repeat was recently conducted (Li et al., 2013) genotyping 40 isolates of P. infestans collected from different regions in Egypt from 2012 to 2014, which revealed genetic variability among P. infestans isolates (Arafa et al., unpublished data).This could have implications for resistance screening.However, since SSR genotypes cannot be linked to race composition, race typing will also have to be conducted to determine the extent of variation in races in Egypt and their relevance in resistance screenings.
The results of the study show that resistance in S. habrochaites accession LA1777 was higher than the resistance conferred by Ph-2 and Ph-3 genes, indicating that the LA1777 genome contains a of late blight resistance.This is in agreement with the study of Li et al. (2011), who also found that LA1777 has a good level of resistance to several isolates of P. infestans.Li et al. (2011) has identified five loci derived from LA1777 (Rlbq4a, Rlbq4b, Rlbq7, Rlbq8 and Rlbq12) associated with lesion size, which were co-localized with previously described QTLs from S. habrochaites LA2099 except QTL Rlbq4b (Brouwer et al., 2004).Pyramiding resistance alleles from LA1777 and available late blight resistance genes such as Ph-2 and Ph-3 in cultivated tomato could enhance the durability of resistance to late blight.
visually disease severity ratings throughout 2013 and 2014 experiments followed by ± standard error (± SE; n=5).Means followed by different letters are significantly (P = 0.05) different.y Disease response: HS = highly susceptible, S = susceptible, MS = moderately susceptible, MR = moderately resistant, R = resistant, and HR = highly resistant.

Table 1 .
Geographic locations where Phytophthora infestans isolates were collected in Egypt, which were evaluated for virulence. P

Table 2 .
Mean late blight disease severity ratings for 48 wild tomato accessions and controls inoculated with P. infestans isolate EG_7.