Late Blight of Tomato: A Brief History of the Disease and its Impact, and Current Knowledge of Resistance (with a focus in western Washington)

By P. Merscher, June 2017, Undergraduate, The Evergreen State College

(Figures unfortunately not included in this blog, but sources are still included in bibliography)

Originating in the New World, potatoes (Solanum tuberosum) and tomatoes (Solanum lycopersicum) were introduced to Europe around 1570 CE (Agrios, 2005). Despite initial hesitancy to adopt these crops as food items, the potato earned its place as a staple for many farmers and families in Ireland; this was largely due to potatoes producing more calories per unit of land than traditional grain crops (Agrios, 2005). By the 1840s, political, social, and economic factors fostered almost complete reliance on potatoes as sole food source (Agrios, 2005; Buchanan, 2012). The necessity and desperation of Irish farmers to produce as many calories on as little land as possible led to widespread adoption of a new cultivar called ‘Lumper,’ which was productive enough to feed a hungry family of four for a year on an acre of land; with the potential for such high yields, nearly every potato planted in Ireland was the Lumper genotype (Buchanan, 2012). Lumper’s downfall, in addition to poor taste, was its susceptibility to late blight, which would devastate the country in the mid-1840s (Agrios, 2005; Buchanan 2012). While late blight is believed to have first arrived in Europe in the early part of the decade, the disease destroyed nearly every potato tuber and plant in Ireland between 1845 and 1846, leading to the starvation of an estimated million and a half people and the emigration of a million and a half more to the United States (Agrios 2005; Buchanan 2012). In a zeitgeist still popularizing the theory of spontaneous generation, most people at the time believed the Irish potato famine was the result of the devil’s work or God’s punishment (Agrios, 2005). It wasn’t until 1861 that Anton deBary performed the experiments that revealed late blight was caused by a fungus (later reclassified as an oomycete), but nonetheless the disease has remained a global focus ever since (Agrios, 2005).

The Pathogen: Phytophthora infestans

Late blight is caused by the oomycete Phytophthora infestans, and is widely regarded as one of the most devastating plant diseases, specifically for potatoes and tomatoes (Agrios, 2005; Fry et al, 2015). Once infected with P. infestans, whole crops can be devastated within 7-10 days, and the disease can affect 41-100% of unprotected fields and 12-65% of fields sprayed with systemic fungicides (See Figure 1; Nowicki et al, 2012). For potato crops alone, the estimated annual cost of the disease is upwards of $6.8 billion due to decreased yields, tuber quality and storage ability as well as costs associated with fungicide applications (Nowicki et al, 2012; Fry et al, 2015).

The success of P. infestans as a pathogen can partly be attributed to its effective reproduction both sexually and asexually as shown in Figure 2 (Nowicki et al, 2012). As a hemibiotroph, P. infestans initially requires a living host for survival, so primary inoculum usually comes from infected potato tubers overwintered in storage, cull piles or left in the field (Agrios, 2005). As tubers germinate, P. infestans mycelium infects the seedlings and eventually forms sporangiophores through the stomata of stems and leaves; sporangiophores are specialized hyphae that produce sporangia (Link et al, 2012). Sporangia readily dislodge and spread via wind and water to cause new infections, which can grow to produce new sporangiophores in as few as four days (Agrios, 2005). When sporangia are released, indeterminate growth of sporangiophores can produce new, mature sporangia in just four hours in a moist environment (Fry et al, 2013; Agrios, 2005). Under ideal conditions, a single lesion can produce nearly 300,000 sporangia per day (Nowicki et al, 2013). In oomycetes, two methods of sporangium germination are known that are often species specific, but P. infestans exhibits both (Link et al, 2012). Sporangium can germinate directly on susceptible tissue by the formation of a germ tube, but can also germinate indirectly by the creation of zoospores – motile, asexual spores capable of chemotaxis (a movement response to chemical signals and gradients; Link et al, 2012).

Phytophthora infestans is a heterothallic species, meaning two strains of opposite mating types (A1 and A2) are required for fertilization and sexual reproduction (Agrios, 2005). A1 and A2 mating types do not differ morphologically, but produce different sex hormones required for sexual reproduction, a process that results in a thick-walled zygote called an oospore, which acts as a resting spore; they are produced in infected plant tissue and released as the tissue degrades (Link et al, 2012; Nowicki et al, 2013). When the oospore germinates, it produces sporangia, which then germinate directly by a germ tube or indirectly via zoospores causing new infections on host tissue (Link et al, 2012). Prior to 1980, all tested P. infestans isolates were mating type A1, but international dispersal of the A2 mating type out of Mexico has led to establishment of sexual populations in several locations, and sexual recombination events have increased virulence complexity, fungicide resistance, and overall pathogenicity of the oomycete leading to epidemics on regular occasions (Nowicki et al, 2013; Derie & Inglis, 2001). Additionally, oospores are much more resilient than asexually produced sporangia or zoospores; they are capable of overwintering on dead tomato and potato tissue as well as in the soil providing an additional source of primary inoculum the following growing season (Agrios, 2005).

Oomycetes like P. infestans are commonly referred to as water molds because moisture is essential to their proliferation. For example, a film of water is required for zoospore motility (Nowicki et al, 2013). Growth and reproduction of the pathogen and disease transmission are facilitated by humidity, particularly when coupled with cool temperatures, rain, heavy dew or fog (McGrath, 2015). Foliar symptoms appear 5-10 days after inoculation and begin as water-soaked lesions, usually at the edges of lower leaves, which rapidly enlarge to form brown, blighted areas (Agrios, 2005; Nowicki et al, 2013). White, downy growth is visible on the edges of lesions on the underside of leaves indicating sporangiophore production, which can be seen in Figure 3 (Agrios, 2005). Dry conditions and temperatures above 95°F cease disease progression, but abundant moisture and cool temperatures rapidly causes whole leaves, plants, and fruit to rot away in a matter of days or weeks as infection spreads (Agrios, 2005; Nowicki et al, 2013). Additionally, research has shown that continuous light inhibits sporulation, but no endogenous time-keeping mechanism has been identified in P. infestans suggesting sporulation is influenced by a plant signal from the host’s system for tracking periods of light and dark (Nowicki et al, 2013). As can be seen in Figure 2, spores can germinate directly through the leaf surface or through an opening like a stoma, initiating the biotrophic growth phase of P. infestans (Nowicki et al, 2013). During this phase, nutrients are obtained via a penetration peg from living plant cells and mycelium growth rapidly occurs in the extracellular space (Agrios, 2005). Additionally, hook-like haustoria differentiate from the branching hyphae and penetrate cell walls while remaining outside the cell plasma membrane (Hardham & Blackman, 2010). Extensive necrosis follows haustorial growth due to the secretion of effectors, including proteases, elicitins, and other degrative compounds, which facilitate nutrient uptake for the pathogen and initiate the necrotic growth phase (Bozkurt et al, 2011; Nowicki et al, 2013; Vleeshouwers et al, 2011). In general, effectors are pathogen-produced molecules that manipulate host cell structure and/or function, and facilitate pathogen fitness (Rouxel & Balesdent, 2010). In necrotic growth, the pathogen retrieves nutrients from plant cells that are dead or dying; this often coincides with the beginnings of reproductive growth cycles of P. infestans (Nowicki et al, 2013).  Haustorial and mycelial growth also affect the plant’s physiology by decreasing photosynthetic abilities, nutrient transport abilities, and by increasing transpiration, while also making tissues more susceptible to infection from secondary pathogens (Bozkurt et al, 2011; Agrios, 2005).

Discussing Plant Immunity

Since plants are without an adaptive immune system and sessile, they therefore are incapable of moving to escape attack from pests or pathogens. They have evolved sophisticated and complex chemical pathways in addition to preformed strategies as a means of defense and immunity (Jones & Dengl, 2006). The instructions for immunity and defense pathways are found in the plant’s genetic material, but pathogens play a role in initiating various defense responses even before infection because plants can perceive physical and chemical stimuli on their surface (Nowicki et al, 2013). Depending on receptors and defensive pathways present in the host, plants will respond differently, however responses to pathogen presence can occur rapidly (Nowicki et al, 2013). For tomato and P. infestans, the preformed line of defense involves physically resisting the penetration peg at the leaf surface, whose pressure acts as a signal for the formation of microfilaments beneath the pressure point within minutes (Hardham & Blackman, 2010). The pressure behind the penetration peg can reach up to 40 times the pressure in a typical car tire (Agrios, 2005). Additional preformed or passive defenses involve the synthesis of constitutive antimicrobial compounds and pathogen recognition molecules (Agrios, 2005; Shamrai, 2014).

In response to pathogen attack and inoculation, transmembrane pattern recognition receptors (PRRs) recognize microbe-associated molecular patterns (MAMPs), which are general elicitor molecules that are highly conserved and essential among large taxonomic groups of pathogens (Shamrai, 2014). Elicitor molecules are effectors that produce an immune response in the host, and therefore activate or change gene expression (Rouxel & Balesdent, 2010). In MAMP-triggered immunity, β-glucan acts as a signal in oomycete and fungal infections since it is crosslinked to cellulose or chitin, respectively, within the pathogen’s cell walls (Shamrai, 2014). MAMPs share a key feature in that these structures are not produced within the host; thus, their presence signifies an invader (Shamrai, 2014). MAMP-triggered immunity begins a cascade of events within cells triggering upregulation of constitutive defense compounds (those that degrade cell wall components and inhibit the pathogen’s effectors), activation of induced defense compounds, and changes in cellular organization and metabolism (Shamrai, 2014; Nowicki et al, 2013). For example, within 14 hours of inoculation with P. infestans, epidermal cells of a tomato host react with localized rearrangements of cytoplasmic contents by aggregating them together, producing reactive oxygen species (ROS), and accumulating callose (Hardham & Blackman, 2010). Additionally, a wide range of compounds is delivered to the site of infection including phytoalexins, phenolics, silicon, hydrogen peroxide, peroxidases, and other enzyme inhibitors (Hardham & Blackman, 2010). MAMP-triggered immunity evokes changes in host gene expression resulting in localized production of phytoalexins and other pathogenesis-related (PR) proteins (Hardham & Blackman, 2010). This more generalized recognition and response system does not stop disease development, but can severely slow its progression and inhibit colonization and reproduction of the pathogen (Shamrai, 2014; Newman et al, 2013). A successful pathogen is able to overcome MAMP-triggered immunity initiating a phase known as effector-triggered susceptibility (Jones & Dangl, 2006). The amount, type, and presence of receptors, pathway intermediates, and final gene products can vary immensely based on the host’s genotype, and pathway regulation is highly dependent on environmental conditions (Newman et al, 2013). Similarly, elicitors and MAMPs are all under both positive and negative selective pressure, so pathogen genotype plays a key role as well (Newman et al, 2013). It may be a reasonable hypothesis that an effective MAMP-triggered immune response contributes to horizontal or field resistance, which is broadly effective against many races of a pathogen, controlled by multiple genes or quantitative trait loci (QTLs), and considered more durable than vertical or race-specific resistance (Nowicki et al, 2013; Agrios, 2005). In regards to late blight of tomato, breeding for horizontal resistance is of questionable value due to the fast and prolific reproductive strategies of P. infestans (Nowicki et al, 2012).

Horizontal resistance can slow disease progression, but pathogen effectors continue to attack host cells during effector-triggered susceptibility; among these effectors are the pathogen’s avirulence (Avr) proteins, or virulence factors, which allow them to “manipulate host cellular pathways [to] gain entry into, multiply and move within, and eventually exit the host for new infection cycles” (Speth et al, 2008). The pathogen’s effectors, toxins, and avirulence proteins disrupt many host cell functions, immune responses and mimic hormones as can be seen in Figure 4, but the details around signaling and pathway disruption are only just now being slowly elucidated for different host-pathogen relationships, particularly for eukaryotic pathogens (Speth et al, 2008). In response, host plants produce resistance (R) proteins, which recognize avirulence proteins through their effects on targets or directly by bonding (Jones & Dangl, 2006). The complex evolutionary “arms race” often talked about between pathogens and their hosts generally refers to the relationship between avirulence genes of the pathogen and resistance genes of the plant host, and both sets of genes are highly polymorphic (Jones & Dangl, 2006; Agrios, 2005). Most R genes code for nucleotide binding (NB) and leucine-rich repeat (LRR) domains that can be found in the cytoplasm (Jones & Dangl, 2006; Zhang et al, 2014). When virulence factors are recognized by host R proteins, effector-triggered immunity is induced (Jones & Dangl, 2006). Effector-triggered immunity is an accelerated and amplified MAMP-triggered immune response, but generally confers disease resistance and results in the hypersensitive cell death response (Jones & Dangl, 2006). The relationship between virulence and plant immunity is referred to as the zigzag model as seen in Figure 5 (Jones and Dangl, 2006). A host with a corresponding R gene for a pathogen’s Avr gene is said to be vertically resistant, since it is under monogenic control and largely capable of ceasing disease and infection in a variety of environmental conditions (Agrios, 2005; Nowicki et al, 2012).

Both Avr genes and R genes are discontinuous or discrete traits, meaning they are distinct and do not exist on a continuum; individuals either have or do not have the avirulence or resistance in question (Speth et al, 2008). As previously stated, both sets of genes are also highly polymorphic with the existence of many different alleles within and between pathogen and host populations (Vleeshouwers et al, 2011). R genes are highly variable in wild populations, but much less so in domesticated crop genomes due to intense selection by humans throughout history (Gruber, 2017; Foolad et al, 2008; Jones & Dangl, 2006). Variability in Avr genes among pathogens is still extensive, especially for P. infestans (Nowicki et al, 2012). New Avr alleles and new combinations of alleles are results of genetic mutation (ie: addition, insertion, deletion), sexual recombination, gene and genotype flow, and genetic drift (Agrios, 2005). Populations of a pathogen species with similar virulence factors (Avr gene products) are grouped into races. Different crop cultivars are resistant to some races and not others, depending on the ability of the cultivar’s R proteins to recognize the pathogen race’s Avr proteins (Agrios, 2005). New pathogen races that overcome previous resistance in a crop or cultivar have resulted in devastating epidemics and challenges to breeders around the world. A mutation as simple as a single nucleotide change in an Avr sequence can be sufficient to alter the final gene product of a pathogen; assuming the variant genotype is successful, asexual reproductive cycles can establish new, more virulent races (Agrios, 2005). Ultimately, the durability of vertical resistance in a cultivar can be precarious. As a solution, pyramiding multiple and different R alleles at respective loci is a common practice in many breeding programs seeking stable, strong and durable disease resistance, including those researching late blight in tomatoes (Nowicki et al, 2013; Agrios, 2005). With regards to P. infestans, the previously mentioned migration of the A2 mating type in the 1980s led to potential for sexual reproduction, which is a major factor in the emergence of new, more virulent lineages of the pathogen (Saville et al, 2016). In a sample of 165 P. infestans isolates collected from wild Solanum spp. in Mexico’s Toluca Valley, 158 were found to be genotypically unique illustrating the differentiation capable of persisting within a sexually reproducing population (Flier et al, 2003). Since dispersal of the A2 mating type, novel races of P. infestans have emerged, including those that have changed host preference and established resistance to systemic fungicides such as metalaxyl/mefenoxam; metalaxyl resistance is conferred by a single incompletely dominant allele (Nowicki et al, 2013; Derie & Inglis, 2001). Figure 6 illustrates the genotype frequencies of P. infestans in submitted tomato samples to USABlight Project; Table 1 outlines the host preference and metalaxyl sensitivity of the genotypes in Figure 6 (USABlight, 2017).

Recent Impacts of Tomato Late Blight

Epidemics of late blight have increased all over the globe in recent decades possibly influenced by continued globalization of the food system, which seems to have exacerbated crop losses affecting both farmer livelihoods and the food supply. Returning to the Irish potato famine, analysis of historic herbarium samples indicate race HERB-1 was responsible for the outbreak; HERB-1 isolates fall into the H1a haplotype per PCR amplification of mtDNA genomes (Saville et al, 2016). Currently, the H1a haplotype is altogether rare or nonexistent in global populations, but persists in small populations of P. infestans in Mexico and South America (Saville et al, 2016). In the 1840s, dry rot was ruining Europe’s potato crops, and a burgeoning bat guano trade connected Peru to New England and Europe (Saville et al, 2016). Trade facilitated the migration of the pathogen to Europe as new potato tubers were shipped along with the guano to address Europe’s dry rot problem (Saville et al, 2016). Clonal and variant populations of HERB-1 were predominantly responsible for global late blight outbreaks until the mid-20^th century, when a similar migration of the sister lineage US-1 occurred, which belongs to haplotype H1b (Saville et al, 2016). US-1 lineages dominated the global population of P. infestans until the 1970s and ‘80s, and while more virulent than HERB-1 lineages (particularly on tomato), US-1 races were manageable thanks to resistant cultivars and the development of systemic fungicides (Nowicki et al, 2012). Saville et al (2016) used the largest globally sourced collection of historic and modern samples to identify 11 different haplotypes, seven of which are represented in modern US aggressive strains (see Figure 7); the widespread and diverse distribution represented by modern populations of P. infestans may be a consequence of the continued quest for new and expanding global markets, and the translocation of propagules preceding production.

Mitochondrial DNA analysis can reveal insights regarding geographic centers of origin, which for P. infestans has recently shown to be “more complicated than previously thought” (Martin et al, 2015). The earliest scientists commenting in the 1800’s proposed an Andean center of origin mostly based on indigenous people’s accounts of the disease (Saville et al, 2016). Modern molecular support for a South American Andean origin also includes increased allelic diversity and persistence of US-1 lineages in Ecuador; identification of some of the oldest mutations in nuclear and mtDNA present only in South American isolates; higher levels of nucleotide diversity at multiple loci; presence of more diverse haplotypes; and earlier divergence of clonal lineages of P. infestans and the intraspecies hybrid P. andina (Martin et al, 2015). However, equally strong arguments can be made for a central Mexican origin. Most notably, the oldest sexually reproducing population of P. infestans can be found in Mexico’s Toluca Valley (Flier et al, 2003; Fry et al, 2013). While both A1 and A2 mating types are represented in South America, the two are geographically isolated (Martin et al, 2015). Additionally, P. infestans’ closest relatives (P. ipomoeae and P. mirabilis) are endemic to Mexico (Martin et al, 2015). The first 16 R genes identified for P. infestans virulence factors were also identified in the wild Mexican potato relative, Solanum demissum, further suggesting an origin in Mexico; notably 11 of these R genes have been introgressed into the potato genome (Derie & Inglis, 2001). Using haplotype phasing, Martin et al (2015) most recently concluded that P. infestans likely originated in Ecuador or Peru, but diversified in central Mexico on potato after a host jump from a wild relative. To further complicate matters, 7% of isolates recently collected and analyzed from late blight lesions on Ethiopian potato and tomato revealed a previously unreported haplotype (Shimelash et al, 2016). Regardless of its geographic origin, P. infestans is currently distributed worldwide and continues to cause epidemics affecting food security and grower livelihood.

In the Karnataka state in southwest India, where 46,000 ha of tomatoes are grown for fresh market, growers witnessed crop losses up to 100% in 2009 and 2010 (Chowdappa et al, 2013). While late blight on potato has been recorded in the region since 1953, the disease was not considered important prior to 2007 (Fry et al, 2015). Chowdappa et al (2013) assayed isolates of P. infestans during the 2009 and 2010 epidemics, and concluded the increased disease severity was due to a newly arrived genotype of the pathogen, 13_A2 – a particularly aggressive strain on both potato and tomato that wreaked havoc in Great Britain from 2005 to 2008 (Fry et al, 2015; Chowdappa et al, 2013). It is believed that the migration of 13_A2 genotype was facilitated by the importation of tons of British seed potatoes to India prior to 2009 (Chowdappa et al, 2013). Similar recent patterns of P. infestans genotype migration have also been documented between Russia and China, and are also being considered in recent outbreaks in Oman and Nigeria, where late blight has historically been a nonissue; allozyme analysis for the Oman and Nigeria outbreaks has not yet been published (Fry et al, 2015).

2009 also saw a late blight epidemic in the eastern United States causing “many home gardeners and organic producers to lose most if not all their tomato crop,” and CSA farmers unable to supply tomatoes to subscribers (Fry et al, 2013). The outbreak resulted from infected plants grown in southern greenhouses for a chain of “big-box” retail stores that were shipped north (Fry et al, 2013; Buchanan, 2012). Unknowing gardeners bought and planted infected plants setting the stage for widespread inoculation by P. infestans spores (Fry et al, 2013; Buchanan, 2012). Rapid genotype assays revealed widespread distribution of P. infestans genotype US-22, but also found US-8, US-23, and US-24 genotypes (Fry et al, 2013). Unlike other late blight epidemics, the 2009 example is unique because it wasn’t caused by unseasonable weather nor the introduction of a new genotype (US-22 was already present, although US-8 was not), but rather the incredibly widespread distribution of inoculum from infected transplants at retail stores (Figure 8; Fry et al, 2013). Denial, surprise, panic, confusion, anger, and concern were among the public’s responses to the 2009 eastern US epidemic, which was the first encounter with the disease for many gardeners and non-commercial growers (Fry et al, 2013). According to Nowicki et al (2012), economic losses in the United States in 2009 from late blight of tomatoes totaled $46 million for fresh market types and $66 million for processing types. In terms of global gross production value, potatoes and tomatoes are the fifth and sixth most economically important crops behind rice, corn, wheat, and soy; the tomato market comprised $64 billion dollars in 2014, so crop losses can have a major effect on a grower’s socioeconomic status (FAOSTAT).  The 2009 US epidemic however, prompted increased interest and funding for more accurate analysis, tracking, and management practices for the disease, including genetically based resistance.

The overall global population of P. infestans is asexual, with novel and variant populations developing from mutation and subsequent dispersal, but several sexually reproducing populations have established outside of Mexico in recent decades facilitating genetic recombination and more aggressive races (Saville et al, 2016). Unexpected late blight severity since the turn of the century in Tunisia is being attributed to the establishment of a sexual population after introduction of the A2 mating type; sexually reproductive populations have also been documented in the Netherlands, Scandinavia, Poland, and Great Britain (Saville et al, 2016; Fry et al, 2015; Nowicki et al, 2012). Western Washington sees an annual epidemic of late blight, amplified by a temperate and coastal climate that favors disease development (Derie & Inglis, 2001). There is evidence that sexual reproduction has played a role in the area, and potential for residence of an active sexual population still exists (Derie & Inglis, 2001). Prior to 1994, US-6 was the dominant lineage present in Washington state, but US-11 would completely displace the former by 1996; both are A1 mating types, and US-11 is quite virulent on tomato (Derie & Inglis, 2001; D. Inglis, personal communication, April 12, 2017). Virulence complexity refers to multiple virulence factors characterizing an isolate or race, and P. infestans in western Washington has historically been quite complex (Derie & Inglis, 2001). Not all virulence factors are necessary though, and unnecessary virulence factors represent a fitness tradeoff (Derie & Inglis, 2001). US-11 isolates averaged 5.4 virulence factors, compared to 8.2 – 9.3 for the other strains tested; fewer unnecessary virulence factors may be one reason why US-11 replaced US-6 as the dominant genotype in the region. In 1997, US-7, US-8 and US-14 were all detected for the first time in western Washington (Derie & Inglis, 2001). As A2 mating types, the arrival of these new races also brought the potential for sexual recombination, although no field evidence nor lab experiments have produced viable oospores from western Washington isolates (Derie & Inglis, 2001). Representatives of both mating types have been found in the same field and on neighboring hosts, but never on the same host (Derie & Inglis, 2001). Interestingly, US-11, which remains as the dominant genotype west of the Cascades, is thought to be a sexually recombinant strain from US-6 with US-7 or US-8, suggesting sexual reproduction of P. infestans is possible and perhaps already occurred once in the region (Derie & Inglis, 2001; D. Inglis, personal communication, April 12, 2017).

Tomato Resistance to Late Blight

Phytophthora infestans is described as a pathogen with “high evolutionary potential,” given its remarkable ability to overcome plant resistance genes (Vleeshouwers et al, 2011). While sexual recombination is a source of novel P. infestans diversity, asexual lineages are highly adaptable, a feature underpinned by the genome’s peculiar architecture (see Figure 9); the genome itself is extremely large and unstable, while also largely repetitive and reportedly prone to mitotic recombination (Labate et al, 2005; Martin et al, 2015; Haas et al, 2009). The ability of P. infestans to overcome host immunity and fungicide resistance is exacerbated by its effective reproduction, producing 300,000 asexual spores in a single day, each of which can travel up to 30 miles via wind (Nowicki et al, 2013; McGrath, 2015).   Improved cultivar resistance based on pyramiding effective R genes is widely expected to be the greatest contribution to late blight control and management (Nowicki et al, 2013; Derie & Inglis, 2001; McGrath, 2015). This warrants discussion of the S. lycopersicum genome, which is described by Foolad et al (2008) as a “genetic bottleneck” based on low levels of isozymes and DNA marker polymorphisms. This is a common feature of domesticated species’ genomes, which represent a fraction of the genetic diversity present within their wild relatives (Jones & Dangl, 2006). Foolad et al (2008) compared sequence diversity in S. lycopersicum to wild relatives also in the Solanum section Lycopersicon, and concluded that domesticated tomatoes represent only 5% of the variation within the other species. The researchers added that many genes for desirable agricultural characteristics simply do not exist in the S. lycopersicum genome (Foolad et al, 2008). Similarly, Labate et al (2005) found the genetic variation in a single, wild population of S. pimpinellifolium to be about equivalent to the variation encompassed by 31 cultivars of modern tomato. Solanum pimpinellifolium has played a significant role in breeding for late blight resistance, but notably was identified as a “less variable” relative along with S. cheesmaniae; wild relatives described as “more variable” include S. chinense, S. harborchaites, S. peruvianum, and S. pennellii (Foolad et al, 2008). Whether used to form new crops or improve existing ones, wild crop relatives are set to be part of the answer to addressing crop losses that contribute to food insecurities worldwide (Gruber, 2017). Wild relatives represent an “immense library” of valuable traits with potential to improve the quality and resilience of modern crops, and “researchers are trying to endow domesticated crops with some of these traits through interbreeding” (Gruber, 2017). This is evidenced by more recent SNP analysis, which suggests higher levels of diversity in several regions of the S. lycopersicum genome than previously thought, an observation attributed to introgressions from wild relatives (Sim et al, 2012; Nowicki et al, 2013). Notably multiple major late blight resistance genes and QTLs associated with resistance have been identified and mapped in tomato, thanks in large part to a few of its wild relatives (Nowicki et al, 2013). A review of the currently identified late blight resistance genes and their role in breeding programs follows.

The first late blight resistance gene, Ph-1, was discovered in S. pimpinellifolium accessions West Virginia 19 and 731 during the 1950s (Nowicki et al, 2013). Ph-1 is a completely dominant allele conferring resistance to late blight caused by P. infestans race-0, a member of the US-1 lineage (Nowicki et al, 2013). In 1962, the cultivar ‘Rockingham’ was the first commercially released variety with the introgressed Ph-1 allele, and was used to map the Ph-1 gene to the distal end of chromosome 7 (Nowicki et al, 2013). New varieties such as ‘New Yorker’ (slicing type) and ‘Nova’ (processing type) were bred and subsequently released on the market (Nowicki et al, 2013). Since US-1 lineages have altogether been replaced by more virulent strains of P. infestans, Ph-1 resistance is no longer considered useful, and varieties expressing this allele currently show complete late blight susceptibility (Nowicki et al, 2013; McGrath, 2015).

Identification of Ph-2 occurred during the same screening of S. pimpinellifolium accessions in the 1950s that revealed Ph-1, but Ph-2 was observed in West Virginia 700 (Nowicki et al, 2013). Ph-2 is an incompletely dominant allele that confers resistance to multiple races of late blight, but can fail in the presence of more aggressive isolates, and only reduces the rate of disease (Nowicki et al, 2013). Moreau et al (1998) mapped Ph-2 to the long arm of chromosome 10, and identified a few PCR-based markers that are still used in marker assisted breeding (Nowicki et al, 2013). Characterization of the underlying biochemistry of Ph-2 has proved difficult as its expression is highly dependent on environment, plant age, plant organ and pathogen race (Moreau, 1998). Ph-2 has been incorporated into many tomato cultivars and remains an element of many current breeding programs, but virulence variability in P. infestans has overall reduced the efficacy of Ph-2-based resistance (Moreau, 2998; Nowicki et al, 2013). The breakdown of Ph-1 and Ph-2 by emergent new races prompted further screening of wild tomato relatives.

In the 1990s, strong late blight resistance to a wide range of isolates was identified in S. pimpinellifolium accession L3708, and promptly named Ph-3 (Zheng et al, 2014). Restriction and amplified fragment length polymorphisms and PCR-based markers have been identified to facilitate marker assisted selection and incorporate Ph-3 into tomato, which has successfully been done by many university breeding programs (Nowicki et al, 2013; TOMI Project, 2017). Interestingly, despite P. infestans overcoming Ph-2 conferred resistance, Ph-3 has shown to be most effective when combined with Ph-2 (TOMI Project, 2017). Further research has shown late blight resistance in L3708 was not controlled by Ph-3 alone, but rather Ph-3 is epistatically involved with other genes, so specific alleles of other genes are necessary for complete resistance (Nowicki et al, 2013; Zheng et al, 2014). Additionally, isolates have emerged that can overcome Ph-3 resistance (Zheng et al, 2014).

Several QTLs potentially conferring late blight field resistance were reported in accessions LA1033, LA2009, and LA1777 of S. habrochaites, and described as Ph-4 (Brouwer & St. Clair, 2004). Accession LA2009 contained QTLs conferring late blight resistance on all 12 chromosomes, but severe linkage drag has prevented these from being useful (Nowicki et al, 2013). Researchers developed near isogenic lines of these accessions, and subsequently identified QTLs for canopy density, fruit size and yield, plant type and maturity time in the same associated regions; late maturity and large plant size were both associated with late blight resistance in these lines (Brouwer & St. Clair, 2004). Because of the difficulty introgressing Ph-4, the significance of this resistance is still unknown (Nowicki et al, 2013).

The most recently described late blight resistance genes were identified in a Pennsylvania State University screening of S. pimpinellifolium, in which a single accession (PI270443) was selected for further characterization after showing resistance to seven isolates of P. infestans (Merk et al, 2012; Nowicki et al, 2013). PI270443 was crossed with a susceptible individual, and F2 and F3 generations were evaluated (Merk et al, 2014). Analysis revealed two genomic regions on chromosome 1 (Ph-5-1) and chromosome 10 (Ph-5-2) associated with late blight resistance, and a high rate of heritability for the genes (Merk et al, 2012). Ph-5-2 is colocalized with Ph-2, however it remains unknown if Ph-5-2 represents a separate, closely linked gene, or a different allele of Ph-2 (Merk et al, 2012). Both Ph-5-1 and Ph-5-2 are currently being utilized by university breeding programs and private plant breeders, but there is still much to discover about Ph-5-1 and Ph-5-2 (Nowicki et al, 2013; McGrath, 2015). Presumably, more resistance genes have yet to be identified in older and heirloom tomato varieties, and wild relatives (McGrath, 2015). Appendix A is an adapted table from McGrath (2015) summarizing resistant varieties from recent field evaluations and laboratory assays with detached leaflets.

Furthermore, resistance to late blight has been achieved through stimulating induced systemic response in tomato host foliage. In induced systemic response, plant defensive responses are preconditioned by a treatment resulting in resistance or tolerance to subsequent pathogens and disease pressure (Vallad & Goodman, 2003). The last several decades of research have uncovered much about biological elicitors that can be used to prime systemic defenses as a means of resistance (Vallad & Goodman, 2003). For example, Cohen et al (1994) observed that solutions of [beta]-aminobutyric acid sprayed over tomato can provide 92% protection for up to 11 days against P. infestans, including when treatment was applied one day after inoculation. Unlike other isomers tested, [beta]-aminobutyric acid induced the accumulation of three known PR proteins (P14a, [beta]-1,3 glucanase, and chitinase), without a significant increase in ethylene production (Cohen et al, 1994). Notably the observed resistance was only tested on foliage and not on fruit. Several products are available to growers to stimulate induced systemic response, and further research is being done to increase the understanding and availability of products on the commercial market.

Currently, there is great international concern about late blight disease of tomato and potato as economically, nutritionally, and calorically important crops are threatened by the quickly evolving pathogen. The Irish potato famine represents a historically true, yet extreme example of crop epidemics, but local late blight epidemics have increased in frequency, severity and in previously unaffected areas over the last several decades.  Breeding for genetically-based resistance is considered the greatest hope, with expectations of finding valuable resistance traits in wild relatives and older cultivars. Challenges remain however with the pathogen’s ability to reproduce extremely efficiently, and mutate or recombine rapidly to overcome major resistance genes, as has been evidenced by the decreasing efficiency of P. infestans R genes Ph-2 and Ph-3 (Nowicki et al, 2013). Additionally, standardized procedures for tracking, describing and analyzing P. infestans isolates in the lab have not yet been outlined (Nowicki et al, 2012). Difficult questions face breeders seeking late blight resistance in tomato; the value of horizontal versus vertical resistance or a combination of both, the role of bio-stimulants initiating systemic induced responses, and further identification and characterization of resistance genes are all foci of both public and private breeding programs. Given market constraints, breeders must also be sure to preserve certain agronomic qualities in resistant cultivars (ie: yield, shipping ability, flavor), which adds time and labor to the development and evaluation process. Despite the concern and challenges, there is great excitement about prospects of improved genetic resistance to late blight in tomato, and groups like the Tomato Organic Management Improvement Project (TOMI) and the Northern Organic Vegetable Improvement Collaboration (NOVIC) in the United States are eagerly screening new and old cultivars as well as new breeding lines. The significance of this research remains to be seen, but is eagerly awaited.