Migration and Selection Enforced Multiple Phenotypic and Genotypic Changes in the Population of Phytophthora infestans in Israel During the Last 36-Year Period

Late blight caused by the oomycete pathogen Phytophthora infestans is a devastating disease of potato and tomato worldwide, including Israel. The population structure of this pathogen was monitored in potato and tomato fields in Israel during a 36-year period of 1983-2019. Isolates of the pathogen were tested for sensitivity to phenylamide fungicides, mating type, race structure, and genotype. The phenotypic and genotypic structure of the population from potato have changed greatly from one year to another, from one season to the next, within a season and within a single field. Major changes also occurred in the population collected from tomato crops. The mechanisms driving these multiple changes and the heterogeneous nature of the population in Israel are shown to derive from multiple migration events of the pathogen via seed tubers from Europe and from fitness-driven selection processes.


Introduction
Late blight incited by the oomycete Phytophthora infestans (Mont. De Bary) is one of the most devastating diseases of potato and tomato worldwide responsible for major crop losses (Anderson et al, 2004). It has a notorious history as the cause of the Irish Potato Famine in the 1800s (www.historyplace.com). The global economic burden of productivity loss and associated costs has been quantified at more than $6 billion per year (Haverkort et al., 2008).
P. infestans can easily overcome control strategies including host resistance and fungicides. Its lifestyle helps it to evade plant defenses, effectors that suppress host defenses and promote susceptibility, profuse sporulation with a short latent period that enables rapid dissemination, and a genome structure that promotes the adaptive evolution by fostering genetic diversity (Leesutthiphonchai et al., 2018).
This algal-like oomycete is a diploid heterothallic species, requiring mating of A1 and A2 mating types to form sexual oospores that might be an important source of primary inoculum (oospores) generating genetically diverse populations (Fry, 2016;Klarfeld et al., 2009;Martin et al., 2019). Some clonal lineages have been shown to be triploid, capable of producing offspring progenies when crossed with diploid or triploid isolates of opposite mating type (Hamed and Gisi, 2013). All progenies from sexual populations in nature are diploid, whereas nearly all dominant asexual lineages, including the most important pandemic lineages US-1 and 13_A2 are triploid. Such triploids possess significantly more allelic variation than diploids (Li et al 2017).
A series of epidemiologically successful clonal lineages that were distributed primarily by trade of seed tubers or young plants have dominated in many key production areas (Fry, 2016). The FAM-1 lineage was the primary wave of infection from Mexico to North America and Europe in the 19 th century. It was displaced by the US-1 lineage, which dominated global populations for decades and remains prevalent in some regions. The major US-1 migration consisted of the A1 mating type. Subsequent migrations into Europe and USA from Mexico included both mating types and genetic diversity has increased. Clonal lineages such as US-8 and, more recently, US-22 and US-23 in North America and 13_A2 in Europe have emerged to become locally dominant (Fry, 2016;Martin et al., 2019).
The movement of P. infestans between continents adds to the complexity of controlling late blight. This has resulted in the introduction of strains with novel virulence, increased fitness, fungicide resistance and of the opposite mating type to the one already present (Njoroge et al 2019).
Isolate characterization may be studied by means of several phenotypic and genotypic markers. Common phenotypic markers are mating type, virulence spectrum, and metalaxyl resistance (Cohen, 2000).
Genotypic markers include multilocus simple-sequence repeats (SSRs) that, due to their higher repeatability and resolution, have been selected as the standard genotyping strategy for P. infestans 3 (Cooke and Lees 2004;Lees et al., 2006). These SSR markers are embedded in the EuroBlight monitoring (www.euroblight.net) and USAblight (usablight.org) systems. SSRs and mitochondrial haplotypes are valuable tools for examining pathogen dispersal and the evolutionary history of P.
Diversity studies that consider both phenotypic and genotypic markers have enabled an understanding of population dynamics which has been useful for disease management. Knowing the population structure of P. infestans within a region helps to understand whether different strains have arisen and increased in frequency over time, hence providing a better insight into pathogen migrations. Disease management strategies can then be refined by knowledge of the phenotypic characteristics of the prevailing pathogen strains such as virulence profile or fungicide sensitivity. P. infestans populations can differ dramatically among countries and locations, and predictions concerning their phenotypic behavior need to be based on the correct population structure in a region (Njoroge et al 2019). For example, two migration cases from Europe to sub-Saharan Africa were recently reported: the EU_33_A2 lineage in Plateau State in Nigeria and the EU_13_A2 clone in potato crops in Senegal ).
According to Fry (2016), migration followed by selection has led to rapid changes in populations of P.
infestans over large regions. Because of such rapid population shifts, the characterization of a population in one year might not necessarily predict the population in the future. He stated that sexual recombination contributes little or nothing to the population diversity in a region even if both mating types are present (Fry, 2016).
Major changes in population structure of P. infestans were reported from Europe. The proportion of metalaxyl-resistant isolates fluctuates from year to year and within the season (Gisi and Cohen, 1996).
Almost concurrent with the appearance of resistant isolates was the discovery of the A2 mating type of P.
infestans in many European countries and in other parts of the world including Israel. No genetic correlation was found between resistance and mating type. Inheritance studies showed that resistance to metalaxyl is controlled by a single semi dominant gene. F1 isolates produced by mating of sensitive and resistant parental isolates of opposite mating types were partially resistant (Gisi and Cohen, 1996). Resistant isolates express equal or greater fitness than sensitive isolates. No correlation was detected between resistance and race structure (Gisi and Cohen, 1996 (Gore et al. 2019). In both reports, the authors speculated that the potato varieties imported in Turkey from other countries could be the reason for these changes.
According to Cooke et al (2019), the most obvious change in EU population in the years 2016-2018 was the decline in the combined frequency of the clones EU_13_A2, EU_6_A1 and EU_1_A1 from 60 to 40% of the population and the increase from 10 to 36% of the clones EU_36_A2, EU_37_A2 and EU_41_A2 in the sampled population.
Major changes in population structure of P. infestans were reported from U.S. Most clonal lineages were resistant to mefenoxam from the mid-1990s to 2009. However, recent lineages have been largely sensitive to mefenoxam. Savile et al. (2015) reported that the US-8 and US-11 clonal lineages were insensitive to mefenoxam while the US-20 to US-24 clonal lineages were sensitive to mefenoxam. These lineages are currently incompatible with tomatoes carrying the Ph2 and Ph3 genes for resistance (Fry 2016). A recent study revealed that sensitivity to mefenoxam in USA isolates is associated with two SNPs (single nucleotide polymorphism) whereas mating type is associated with one SNP (Ayala-Usma et al, 2019).
Three major genetic changes occurred in the Israeli population of P. infestans during the 18 years period of 1983-2000 (Cohen, 2002). The first in 1983, when resistance to metalaxyl and the A2 mating type were introduced. The second in 1993, when the A2 population was replaced by the A1 mating type while isolates with intermediate resistance to metalaxyl first appeared and the third in 1999, when A1-sensitive population with extreme virulence variation and high aggressiveness to tomato dominated. It was then concluded that these shifts in population structure might partially reflect the changes that occurred in Europe from which potato tubers were annually imported, but also implicate a rapid local evolution of new genotypes.
More than 20 cultivars of potato are grown in Israel in two seasons, autumn and spring. In the autumn, sowing takes place in September (with locally produced seed tubers) and harvest in the following December-January. In the spring, sowing (with seed tubers imported from Europe) takes place in January and harvest in the following May. Potato crops in both seasons are exposed to winter rain (November-March) or sprinkling overhead irrigation, which facilitates late blight epidemics and enables oospore production in the bottom leaves (Cohen et al 2000). Tomato are grown in greenhouses all year around, and in the summer, also in the open field.
During the season, landing of air borne sporangia from upwind fields of potato or tomato contribute to further enhancing the epidemics. Our data show that an infection event occurs when the following conditions prevail: RH of ≥ 90% at temperature of 10-21 0 C for a period of ≥6h .
In the current study we followed the annual and seasonal changes in the population structure of P, infestans in Israel since the last monitoring survey. Isolates that were collected during 2001-2019 were examined for phenotypic traits. Isolates that were collected during 2004-2019 were also tested for their genotypic traits. The results confirmed the high importance of imported tubers from Europe as a source of primary inoculum but also revealed the importance of local conditions in survival of some but not all immigrants. Some preliminary data on these findings were published (
Detached potato leaves (cvs. Nicola or Mondial or tomato leaves cv. ZH), laid on a wet filter paper in Petri dished at 20 0 C, were sprayed with 0, 1, 10 or 100 ppm ai of metalaxyl or mefenoxam and thereafter inoculated with sporangia of the test isolate. An isolate that sporulated on leaves treated with 0 and 1 ppm ai was considered sensitive; An isolate that sporulated on leaves treated with 0, 1 and 10 ppm ai was considered intermediate; An isolate that sporulated on leaves treated with 0, 1, 10 and 100 ppm ai was considered resistant.

Mating type
Data in Fig. 2 show the frequency of mating types among 1757 isolates collected during the 36 years period of 1983-2019. Not shown are 10 isolates that were homothallic and 42 isolates that were sterile. In the first nine years of the survey, 1983-1991, only A2 isolates occurred in the population (Fig. 2). As shown in Fig.1, these A2 isolates were R or S to metalaxyl. A1 isolates were first detected in 1993. They dominated the population for 17 years, from 1993 until 2009. Their frequency declined sharply in 2010 and remained low for additional 4 years. A1 has re-dominated the population during 2015-2019 (Fig.2). A unique mating type, A1A2, occurred at low frequency in potato fields. It was first detected in 2004 (Fig.2).
An A1A2 isolate produces no oospores when inoculated singly onto potato or tomato leaves but produces them when mix-inoculated with either A1 or A2 reference isolate. Due to the heavy rains that prevailed in 2010, late blight epidemics were unusually severe. During that year, 29% of the samples collected from potato fields carried both A1 and A2 mating types (Fig.2).  (2002). An A1 isolate produced oospore only when co-inoculated together with an A2 reference isolate onto detached leaves of potato or tomato. An A2 isolate produced oospores only when co-inoculated together with an A1 reference isolate onto detached leaves of potato or tomato. An A1A2 isolate failed to produce oospores when inoculated singly but did produce oospores when 7 co-inoculated together with either an A1 or an A2 reference isolate. An A1+A2 isolate is a mixture of A1 isolate (s) and A2 isolate (s) in a single lesion.

Genotype
SSR genotype analysis was employed to 651 isolates that were collected from potato crops during a 16- year period of 2004 to 2019. Results given in Fig.3A show that four genotypes occurred in the country, US-7 like, 23_A1, 13_A2 and 36_A2 at a proportion of 9.21, 79.26, 10.75 and 0.77%, respectively.
Genotype 23_A1 was major in the population every year but was temporarily over dominated by

Race structure
Race structure analysis was employed to 1165 isolates that were collected from potato fields in the autumn (Nov-Jan) and the spring (Feb-May) growing seasons during the 16-year period of 2004 to 2019.
Among the many isolates collected in Nov and Dec (n=219 and 280, respectively), 61 and 63 race combinations were identified. In Jan and Feb, the number of isolates and race combinations declined. In March, the number of isolates increased to 264; among them, a high number of 94 races were identified.
In Apr, both figures declined, reaching in May 63 isolates with 34 race combinations (Fig. 4A).

Latent infection in imported seed tubers
On January 11, 2017, 36 days after planting, five plants with foliar symptom of late blight were detected. No blight symptoms were observed on the surface of the mother seed tuber when the blighted plants were uprooted from the soil but necrotic symptoms typical to late blight were observed on the below-ground emerging stems. Typical sporulation of P. infestans developed on tuber slices which were cut away from the symptom-less mother seed tubers after surface disinfection with hypochlorite. The isolates recovered from these tubers were 13_A2, R to MFX and A2 mating type. They carried 10 virulence factors 1 2 3 4 5 6 7 9 10 11.

Population structure within a single field
Major genotypic and phenotypic changes occurred in a single organic plot within a month. In

Phenylamides sensitivity
During 1983-1991, the population comprised of R and S isolates (Fig.7A), as were the potato isolates during this period. The first I isolate appeared in 1993. During 1993-2016, S, I and R isolates comprised the population in various proportions (Fig.7A). The ratio between S: I: R during 2015-2016 was about 1:1:1.
They dominated the population for 17 years until 2009. A2 isolates re-dominated the population during the next 5 years from 2010 to 2014 (Fig.7B). A1 re-dominated the population during 2015-2016.
Interestingly, in 2005-2014, a few isolates were A1A2, namely producing oospores when mated with either A1 or A2 reference isolates. In 2010, a heavily rainy year, about one third of the isolates were a mixture of A1 and A2 mating types (Fig.7B).
Such complex races can infect and sporulate on all five differential lines of tomato, including NC 45, which carry both Ph2 and Ph3.

Genotype
The most prominent genotypes were US7-like and 23_A1. In 2010, genotype 13_A2 appeared for the first time. This genotype dominated the population in 2016 (Fig. 7D). In early March 2019, an unusual epidemic of late blight occurred in open-field tomato crops. Plants showed severe infection on basal stem and bottom leaves soon after planting, suspecting seed infection (Rubin and Cohen, 2004). All five isolates collected from these tomato plants belonged to genotype 23_A1.

Discussion
In a previous study (Cohen 2002), we reported that the Israeli population of P. infestans underwent three major genetic changes during 1983-2000. We then concluded that these shifts may partially reflect the changes that have occurred in Europe from which potato tubers are annually imported, but also implicate a rapid local evolution of new genotypes.
In the present study, we show that the population of P. infestans in Israel has continued to change during 2001-2019. Sensitivity to phenylamide (PA) fungicides (metalaxyl or mefenoxam) fluctuated, mating type replacement occurred frequently, new races displaced the old ones, and new genotypes showed up.
The frequency of S and R isolates showed opposite sinusoidal shapes. The frequency of resistant isolates had several peaks, of which the pick of 1985-1991 was highest (Fig. 4C). Isolates with intermediate resistance to PA fungicides appeared in 1993, parallel to the appearance of A1 isolates in the country, suggesting a possible role of oospores in their appearance. This corroborates with our earlier finding that hybrid isolates produced by mating of R and S isolates were intermediately resistant to PA fungicides ( Table 2 in Gisi and Cohen, 1996). This mating type is bisexual (not homothallic), capable of producing oospores when mated with either A1 or A2 reference isolates, but not when inoculated alone. In the rainy year of 2010, about 28% of the isolates were composed of both A1 and A2 isolates in the same lesion.
We identified four genotype of P. infestans in Israel: US7-like, 23_A1, 13_A2 and 36_A2 at a proportion In our previous survey of 1983-2000, we identified 56 race combinations. The most frequent races were 1 3 4 7 8 10 (18.8% of the isolates, 6 virulence factors) and 1 3 4 7 8 10 11 (34.6% of the isolates, 7 virulence factors). Only 1.9% of the races were more complex, carrying nine virulence factors. No race was found to carry 10 or 11 virulence factors. During the current monitoring period of 2004-2019 we detected 183 race combinations of which, race 1 3 4 7 9 was most frequent (31.2% of the isolates, 5 virulence factors). However, 2.5 and 1.8 % of the races were complex, carrying 10 and 11 virulence factors, respectively, indicating on expanded cultivar range of the population in the last decade. Such complex races were frequent in May, at the end of the spring season, suggesting adaptation to higher temperatures.
There are several reasons that might be responsible for the extreme diversity of the pathogen population in Israel and the rapid changes between years, between seasons and within a season or a field: (i) Annual migration of P. infestans from Europe is probably the major reason for the shift in population structure between years in Israel. We found that 1% of healthy-looking imported seed tubers carried latent infection with P. infestans. Genotypes, such as 13_A2 and 36_A2, showed up in Israel a year or two after their emergence in Europe. Nevertheless, many genotypes that are prevalent in Europe did not appeared in Israel, probably due their low fitness in potato tubers or low fitness under Israeli conditions. Tuber trade was also responsible for the rapid shift in the population structure in Turkey (Gore, et at 2019; Gunacti, et al., 2019). On the other hand, the lack of potato tuber trade between China and India was shown to prevent gene flow of P. infestans between the two countries (Wang et al 2019).
Similarly, in Peru the same four genotypes (EC-1, US-1, PE-3 and PE-7) persisted during the last two decades because import of seed potato from other continents is nonexistent and therefore, the risk of new lineages from overseas is low (Lindqvist-Kreuze et al, 2019) (ii) The shift in the population structure from one year to the next may also be related to the differential over-seasoning capacity of the isolates harbored in locally produced seed tubers, volunteer potato plants, or wild Solanaceae plants (Kadish and Cohen, 1992).
(iii) The rapid changes in the population structure that happen within a season, even in a single field within a month, might be attributed to fitness parameters that drive the outcome of the competitions between isolates (Kadish and Cohen 1988b;Kadish and Cohen 1988c;Kadish et al., 1990). Genotype 23_A1 that have emerged in Europe persisted well in Israel but not in Europe probably because it is better adapted to higher temperatures. As indicated by Ayala-Usma et al (2019), isolates may carry specific SNPs that are associated with their ability to grow at 15, 20 or 25 0 C.
(iv) Israeli isolates of P. infestans recovered from potato were all infective to tomato and vice versa, suggesting a continuous cross talk between the two populations. The rapid changes in the tomato population, including the appearance of highly complex races which can attack Ph2.Ph3 cultivars, seems to be driven by complex mechanisms. Tomatoes grown indoors in autumn and spring in the vicinity of potato crops may get infected by potato isolates growing outdoors. On the other hand, tomatoes which are prone to infection with oosporic variants of the pathogen arising from seed-borne oospores (Rubin et al., 2001;Rubin and Cohen, 2004;Cohen and Rubin, 2006) may transmit such variants to potato crops.
During the summer season tomato isolates are exposed to fungicide and high temperature selection. They may infect potato crops in the following autumn.

Bioassays
Sporangia from each sample (isolate) were used for three bioassays: sensitivity to phenylamide fungicides (PA, metalaxyl or mefenoxam, MFX), compatibility with differential cultivars (virulence phenotype) and mating type determination. Sensitivity to PA was conducted as described by Kadish and Cohen (1988a).
Pathogenicity assays to determine virulence factors in the pathogen were done with the 12 standard potato differentials, R0 to R11 (Black et al 1953) obtained from Syngenta Crop Protection, Stein, Switzerland.
Plants were clonally reproduced in vitro from stem segments, and then grown in 1-liter pots in the greenhouse. Leaflets were detached, placed in moistened petri dishes, lower side uppermost, and inoculated each with six 10-μl droplets (5,000 sporangia per ml), two leaflets per potato genotype. Plates were incubated in 18°C growth chambers with a 12h photoperiod for 10 days. Reaction to the pathogen was assessed as either compatible (expanded lesions with sporulation) or incompatible (hypersensitive response). Virulence factors of an isolate are given as a series of numbers representing the R-genes of the potato genotypes that produced compatible interaction with that isolate. In all tests, tomato leaflets (inbred ZH carrying no resistance genes) and potato Bintje (no resistance genes) were included. Mating type assays were conducted in vivo on detached leaflets of ZH tomato as described by Cohen et al. (1997). The occurrence of oospores in the test leaves was examined microscopically at 10 dpi.

Latent infection in imported potato seed tubers
Five hundred seed tubers cv Nicola which were imported from Holland in late 2016 were planted in a net house (no potato cultivation history) at BIU Farm on December 6, 2016. The developing plants were inspected daily (except Saturdays) for the appearances of late blight. On January 11, 2017 five plants with foliar symptoms of late blight were detected. They were uprooted, each placed in a sterile plastic bag and taken to the laboratory for further inspection.

Population structure within a single field
In spring 2017, we followed the changes in the population structure of P. infestans within a single potato field. Late blight infected leaves were collected from a 9-hactare potato field (cv. Nicola, organic management) at Kibbutz Nirim, Western Negev; 38 samples were collected on 3.4.2017 and 46 samples at four weeks later, on 1.5.2017.

Tomato isolates
Isolates from late blight affected tomato crops were collected throughout the country during 1983-2016.
They were tested for sensitivity to phenylamide fungicides (metalaxyl during 1983-2005

Genotype determination
For each sample of P. infestans 12 SSR loci (Pi02, Pi04, Pi4B, Pi63, Pi70, D13, G11, PinfSSR2, PinfSSR4, PinfSSR6, PinfSSR8 and PinfSSR11) were amplified according to a previously published multiplex protocol (Li et al., 2013). All 12 markers were run simultaneously in a 3730 capillary DNA analyzer according to the manufacturers default settings (Life Technologies, Grand Island, NY) and the alleles were edited and scored using GeneMapper 3.7 (Life Technologies) before being exported to a spreadsheet for comparison to other lineages (Martin et al., 2019).

Conclusions
The long-term data collected in Israel corroborate with studies from other parts of the world reaffirming the dynamic temporal behavior of P. infestans on potato and tomato crops (Fry, 2006;Martin et al., 2019;Njoroge et al., 2019). The unique practice in Israel of growing potato twice a year with seed tubers supplied locally in one season and imported from Europe in the other season, enhances the diversity of P. infestans via migration. Fitness and competition between variants of the pathogen further shape the structure of the population. It is indeed true that the characterization of a population in one year may not necessarily predict the population in the future (Fry, 2006).