Differential plant response to toxins and elicitor proteins released by the potato and tomato pathogens Alternaria solani and Alternaria alternata

Early blight of potato and tomato, caused principally by Alternaria solani, results in extensive damage to foliar tissues. Symptoms are seen as expanding necrotic lesions. Lesions are sometimes surrounded by a halo of chlorotic tissue, however the basis for presence or absence of the distinctive halo have not been determined. To dissect the basis for lesion development a series of Alternaria-associated toxins and potential elicitor proteins were tested in potato and tobacco. Infiltration of leaves demonstrated that, while most of the pure toxins did not have a visible effect, two anthraquinones, bostrycin and altersolanon A, elicited a strong necrosis-mediated response. Extracts of solid and liquid cultures of A. solani yielded anthraquinones that also elicited a necrotic response. Two potential elicitor proteins were identified and cloned from the potato brown spot and early blight pathogens Alternaria alternata and A. solani, respectively. Agrobacterium-mediated infiltration and expression of the A. alternata xyloglucanase cel12-A and A. solani Hrip1 revealed that cel12-A had no apparent effect, whereas Hrip1 induced expanding tissue death and development of chlorotic halos. Development of chlorotic halos was dependent on leaf age, with older leaves exhibiting halos. It is proposed that specific toxin and elicitor production, in conjunction with tissue susceptibility mediate the outcomes of early blight symptoms.


Introduction
Early blight caused principally by Alternaria solani, is one of the primary diseases of potato and tomato. Additional foliar infections are caused by Alternaria alternata, which causes brown spot of potato. The primary economic impact of attack on mature tissue is that potato plants are in a critical tuber bulking stage, and tomatoes are laden with fruit that may fail to achieve full size and become sun-scorched (Shah et al. 2004). While there is a limited level of resistance in certain wild Solanum tuberosum subspecies, and the wild species Solanum raphinofolium (Herriott et al. 1986;Weber and Jansky 2012), there are currently no resistant cultivars available, thus growers must rely on fungicide applications.
While newer fungicides with greater target specificity reduce environmental impacts, resistance to these fungicides, by point mutations in the fungal gene encoding the target protein, have limited their use (Gudmestad et al. 2013).
Early blight disease is most prominent on older tissues. Infections by this necrotrophic pathogen results in expanding regions of dead tissue that may or may not be surrounded by a chlorotic halo. Cells at the site of infection undergo apoptosis or necrosis, limiting available plant defenses. It is not clear why older tissues seems to be more susceptible, nor is it clear why chlorotic halos are not consistently present around developing lesions. Toxins are often implicated in development of necrotic lesions and halos (Main 1971). Necrotophic pathogens are not limited by specific resistance genes, due to the disabling of the host cell metabolism before defenses can be fully activated. Previous research has uncovered a wide array of secondary metabolites produced by Alternaria spp. (Lou et al. 2013). In the case of Alternaria alternata there are many formae speciales (or pathovars) that infect a specific host with the aid of a specific toxin, that is not phytotoxic to a broad range of plants (Tsuge et al. 2013). Brown spot of potato, caused by Alternaria alternata, has not been associated with toxin production.
More recently, there is an increasing appreciation of the role of pathogen proteins that elicit a host hypersensitive response. One Alternaria protein, Hrip1, from Alternaria tenuissima, has thus far been identified (Kulye et al. 2012). To clarify the role of toxins and hypersensitive-eliciting proteins in the Alternaria pathosystem, we tested a series of toxins produced by Alternaria (Ichihara and Oikawa 1997), as well as two potential elicitor proteins, to determine their contribution to necrotic disease lesion development. Understanding the role of different pathogenicity factors deployed by Alternaria can aid in development of methods for overcoming or avoiding these factors. This information can contribute to selection of resistance traits or new bioengineering targets.

Materials and methods
Toxins Purified toxins were obtained from commercial sources as follows; Alternariol, bostrycin, tenuazonic acid (Cayman Chemical, Ann Arbor, MI) Altersolanol A (Santa Cruz Biotechnology, Dallas TX) tentoxin, AAL toxin (Sigma Chemical, St Louis, MO). Samples were dissolved in distilled water to a concentration of 1 ug/ul. Approximately 100 ul of sample or distilled water was infiltrated into tobacco and potato leaves using a needleless 1 ml syringe. Two small punctures 5 mm apart were made with a 20 gauge needle, and infiltration was performed over one puncture with a needle-free syringe (Suppl. Fig. 1).

Cultivation of Alternaria
Isolates of Alternaria solani from potato, and Alternaria alternata from tomato fruit, were cultured in potato dextrose broth and on potato dextrose agar (Difco, Becton Dickinson, Sparks, MD) in the dark at 23 °C for two weeks. Liquid cultures were decanted and then size fractionated using Nanosep centrifugal filters (Pall Life Sciences, Ann Arbor, MI). Filters were pre-rinsed with distilled water, then samples applied and spun for 5 min. Filter cutoffs were 30, 10 and 3 K. A portion of flow-through from each filter cutoff in microfuge tubes were placed in boiling water for 5 min to assess thermostability. Solid medium cultures were flooded with 20 ml of distilled water, allowed to soak for 2 h, followed by decanting of the fluid for processing in the same manner as fluid from liquid cultures.
Samples were also spotted onto thin layer chromatography plates and developed in methanol. Sections of the silica were scraped off the plates, soaked in water for 30 min. Water soluble material was tested for phytotoxicity by leaf infiltration.
Test plants Potato plants (Bintje and Kennebec) and tobacco (Nicotiana benthamiana) were grown in the greenhouse until 45 days old. Plants were maintained at ambient temperature (22 °C) and daylight exposure in the laboratory after treatments. A set of tobacco plants were maintained at 5 °C for two hours before and twelve hours after toxin treatment to determine if responses were necrotic or apoptotic.
Elicitor proteins A family 12 xyloglucanase/endoglucanase mRNA was identified from A. alternata using a BLASTp search and conserved domain sites as previously described (Costanzo et al. 2006). Specific members of this group of enzymes from other fungi have been shown to generate a hypersensitive response in tobacco (Jones 2015). The coding region was cloned from genomic DNA and ligated into the binary vector pBI 121. No family 12 ESTs were identified for A. solani.
A second set of elicitor protein encoding genes, encoding the Hrip1 protein (Kulye et al. 2012), were cloned from genomic DNA of A. solani and A. alternata. Clones were sequenced to confirm identity. The mRNA encoding regions were synthesized as gene blocks (IDT DNA, Iowa) and cloned into pBI 121 using Xba1 and Sac1 sites, along with cloning of the genomic copies. ClustalW alignments were made to identify variations in amino acid sequence.

Agrobacterium-mediated expression of elicitor proteins
Colonies of Agrobacterium tumefaciens LB 4404 harboring the Hrip1 and family 12 xyloglucanase constructs were grown in YM liquid medium overnight in shake culture, centrifuged and resuspended at an 0.5 (OD 600) in minimal medium plus acetosyringone. Cultures were incubated for three hours. Samples were infiltrated using the two-puncture leaf infiltration method used for toxin infiltration. Pictures were taken of leaf samples and DNA samples taken seven days later.
Analysis of potential defense response genes after toxin treatment Tissues were harvested three hours after infiltration, before hypersensitive responses and corresponding tissue collapse were evident. RNA samples preparation was performed using a Quick mRNA kit (ZymoResearch), and cDNA was generated using ProtoScript (New England Biolabs). Sample genes were PR 5 RING and RHOD (Ni et al. 2010). Semi-quantitative gene expression levels were determined using 2 μL of cDNA, gene-specific primers (Table 1), water and MyTaq Plant PCR 2X polymerase mix (Bioline) in a final volume of 25 μL. Cycle parameters were 95 °C for 3 min, followed by 30 cycles of primer annealing (primer-dependent temperature) for 15 s, extension at 72 °C for 1 min and denaturing at 95 °C for 30 s. A final extension was performed at 68 °C for 5 min. Samples were loaded (5 μL) and separated in a 1% agarose gel (100 V for 30 min) containing ethidium bromide. Bands were exposed using a BioRad ChemiDoc imager, at levels below pixel saturation, and quantified using the BioRad Quantity One software.
Genomic DNA fragmentation analysis Samples were harvested from toxin treated potato leaves 6 h after infiltration and total DNA isolated using protocols from a Total Plant DNA Kit (Qiagen). Samples of total DNA were also extracted from potato leaf tissue 7 days after Agroinfiltration. Samples were run on a 1% agarose gel for 25 min and visualized with ethidium bromide staining.

Results
Toxin activity Infiltration of Alternaria toxins that may affect solanaceous plants demonstrated that two anthraquinones, bostrycin (Charudattan and Rao 1982) and altersolanon A (Holenstein and Stoessl 1983;Suemitsu et al. 1984) caused rapid tissue collapse in both tobacco and potato. Tenuazonic acid (Lebruna et al. 1988) caused tissue collapse on tobacco but no obvious effect on potato. Water-soaking symptoms began at three to six hours after infiltration and dry lesions were fully developed at 24 h (Fig. 1). Culture filtrates from A. solani were yellowish-red and the color shifted based on pH, red in basic and yellow in acidic conditions suggesting the presence of anthraquinones (Richardson et al. 1988). Filtrates retained full toxicity in the flow-through from a 3 K cutoff filter. Boiling for 5 min did not reduce activity (Fig. 2). Toxic fractions were found to be soluble in methanol with an Rf of 0.95 on TLC plates using methanol as solvent. No toxic activity against tobacco or potato was identified in filtrates from A. alternata.
Infiltration of bostrycin, altersolanol A and the A. solani filtrates resulted in hypersensitive tissue collapse regardless of the age of leaf tissue. Toxin exposure resulted in a necrotrophic reaction based on tissue collapse under cold treatment. Analysis of plant DNA six hours after toxin infiltration showed no laddering usually seen in apoptotic cell death (Cheng et al. 2010), supporting a necrotic cell death (Fig. 3). Plant defense genes were not activated within three hours after infiltration of distilled water, altersolanol A or the A. solani culture filtrates, while there was a slight increase after bostrycin infiltration (Fig. 4).

Cloning of elicitor proteins and agroinfiltration mediated expression
Orthologs of the A. tenuissima Hrip1 (HQ713431.1) were identified using BLASTn. The orthologs were cloned from A. alternata and A. solani genomic DNA and sequenced. Both genomic clones had identical sites containing a 51 bp intron present in the A. tenuissima genomic copy (nt 350-401). While the A. alternata protein was nearly identical to A. tenuissima, A. gaisen (RYO46578) and A. arborescens (RY073447), the A. solani protein had numerous amino acid differences (Fig. 5). In some cases, the identical amino acid change was also present in the Hrip1 from two other necrotrophic pathogens Pyrenophora tritici-repentis (XP_001933412) and Bipolaris maydis (XP_014074103).
A single copy of Family 12 xyloglucanase was identified in A. alternata (XM_018526753) while no similar mRNA sequence was identified for A. solani.
Seven days after agroinfiltration, lesions with symptoms mimicking natural early blight infections were evident for each Hrip mRNA construct (Fig. 6). There was no response generated from expression of the A. alternata family 12 xyloglucanase. Only certain motifs at the amino terminus determine hypersensitive elicitation in family 12 enzymes, and A. alternata is apparently lacking this motif (Jones 2015). Leaf tissue age was a determinant in symptom expression. Young tissue showed small necrotic lesions and no halo, whereas mature tissue allowed for expanding necrotic regions surrounded by a distinct chlorotic halo. Samples of DNA from the lesions and surrounding tissue showed no obvious DNA laddering, suggesting the response was necrosis.

Discussion
Early blight has not been studied to the same degree as late blight, caused by Phytophthora infestans. Numerous effectors have been identified in the hemibiotrophic P. infestans pathosystem and numerous individual host resistance (R) genes have been identified (Haas et al. 2009). Conversely, there have thus far been no resistance genes identified for controlling Alternaria early blight. While some still unidentified genes are providing variable levels of reduced susceptibility in wild germplasm (Wolters et al. 2019), there may be a series of genes required to develop useful levels of resistance. A characteristic of early blight is the slow expansion of dry necrotic tissue that comprises the lesion. The infections occur primarily on mature tissue, and a chlorotic halo may or may not develop around the lesion. These features suggest that Alternaria solani, while damaging, is not as aggressive as P. infestans.
The variables in relation to Alternaria blight are production of necrotic factors, and host susceptibility to necrotizing factors. Various studies have used enzymes and toxins in screening for resistance to early blight (Lynch et al. 1991;Maiero et al. 1991;Shahbazi et al. 2011;Taheri 2019). Pathogen aggressiveness of Alternaria has been suggested to have a basis in enzyme production (Lawrence et al. 2000) although enzymatic capacity is not strong, as the pathogen is often limited by leaf veins, and lesions are slow to expand. Release of elicitor fragments from the host cell wall could play a role in cases of weak enzymatic activity. Host susceptibility in potato can be reduced by nitrogen applications (Abuley et al. 2019) indicating either increased susceptibility under reduced nitrogen and/or enhanced ability of Alternaria to produce necrotizing products under reduced nitrogen conditions. It is possible that shifts in carbon:nitrogen ratios in the plant induce production of toxins (Gressler et al. 2015). Anthraquinones such as Altersolanon A seem to be the most potent on tobacco and potato. The anthraquinones have been considered potent phytotoxins (Haraguchi et al. 1996). We  found that the anthraquinones produced in A. solani culture occurred on rich media, and there was little to no sporulation, even when exposed to UV and dessication, which normally favors sporulation. It is not known when the toxins might be produced during an infection, but they may be produced early in the infection when nutrient levels are higher. It would be most beneficial early in the infection when the necrotic activity would quickly yield a conducive infection site. Anthraquinones have mild antibiotic activity, thus a secondary benefit of production by a necrotrophic pathogen would be limiting competition by saprophytic microbes Okamura et al. 1993;Yanga et al. 2012).
In some cases toxins may have a differential effect based on leaf age (Barna and Gyorgyi 1992). The toxic activity of the anthraquinones did not vary depending on age of the leaf being infiltrated, indicating that the mature tissue was not more susceptible. This suggests that toxin production, not tissue susceptibility, could mediate development of infections.
Cultivation of Alternaria alternata did not lead to production of filtrates or extracted media that was phytotoxic to tobacco or potato. There are many well documented cases of A. alternata pathovars that infect specific hosts and produce a host-specific toxin that is critical to colonization of that host. Infection of tomato by A. alternata lycopersici is dependent on AAL toxin, and the corresponding host susceptible Asc1 gene (Brandwagt et al. 2002). There have not been any host-specific toxins identified from A. alternata isolates that cause brown spot of potato leaves, and we did not detect phytotoxin activity from cultures. The potato brown spot pathogen is a weak pathogen in comparison to A. solani. This would provide support for the toxin production capabilities of A. solani enhancing infections of solanaceous crops such as potato and tomato.
Altersolanon A causes a tissue collapse like fractionated filtrates from A. solani, while bostrycin is somewhat less active. The lower activity of bostrycin may allow time for a host response, resulting in the slight increase in defense transcript response. Testing of the toxins during cold exposure suggests that the toxic compounds induce a necrotic response as opposed to apoptosis. Apoptosis requires active metabolism and would have been limited by the cold conditions. Also, analysis of DNA from toxin infiltrated sites did not detect laddering which is typical of apoptosis. It is, however, appreciated that this study as with many others, involved sudden, broad exposure to toxins through infiltration, whereas there may spatial and temporal differences in toxin produced during an infection. Ultimately, identification of genes involved in Alternaria solani toxin production can be identified and mutants developed to fully determine their role.
Alternaria toxins have been the focus of the majority of host:pathogen studies, based on the concept that a necrotrophic fungus would kill cells in advance of colonization, and toxins would be the most likely candidates. In fact, there is good support for the role of toxins in infections caused by Alternaria alternata on specific hosts (Lou et al. 2013;Tsuge et al. 2013). An overlooked area in the realm of Alternaria infections is the role of proteins that elicit a hypersensitive response. Elicitor proteins have been identified in an increasing number of plant pathogens. Some examples include transglutaminase from Phytophthora (Brunner et al. 2002), xylanase from Botrytis (Brito et al. 2006;Frías et al. 2019), endoglucanase from Rhizoctonia (Ma et al. 2015) and specific family 12 xyloglucanase/endoglucanases (Jones 2015). In multiple cases the elicitor activity has been localized to a small portion of the protein; thirteen internal amino acids for the transglutaminase, twenty-five internal amino acids for the xylanase and 18 amino-terminal amino acids for the family 12 xyloglucanase. In each case, elicitor activity is independent of enzyme activity.
In Alternaria tenuissima, the Hrip1 protein was identified as an extracellular protein of unknown function that elicits a hypersensitive response in the tobacco Nicotiana tabacum cv. Samsun. Treatment of tobacco by infiltration of purified Hrip1 protein was also shown to induce subsequent systemic resistance to tobacco mosaic virus and tomato yellow leaf  (Kulye et al. 2012;Sokea et al. 2019). It is not currently known what portion of the protein is responsible for eliciting the hypersensitive response. In our study, we used a different method for delivery of Hrip1, using agroinfiltration-mediated expression in potato. When testing expression in different leaf tissue we found that the expression mimicked symptoms of natural infections with A. solani. Expression from mRNA constructs of genes from A. tenuissima, A. solani and A. alternata each mimicked the A. solani infected lesions, while a genomic construct from A. alternata gave a much smaller lesion. This may be due to poor processing of the fungal intron in plant cells. While Hrip1 caused necrosis and chlorosis in mature potato leaves, it did not have the same effect on younger tissue, where limited hypersensitive lesions were the only symptom. The A. alternata family 12 endoglucanase did not have an effect on any tissue tested.
In the original study of Hrip1, purified protein was infiltrated into tobacco tissue. This is like our infiltration of toxins where a rapid response is seen. Expression through Agrobacterium allows for gradual production of Hrip1, and plant cellular exposure. This may be what causes the chlorosis and would represent the cellular exposure to Hrip1 seen in a natural infection by A. solani. The fact that A. alternata Hrip1 also produces the chlorotic halo, while this is not a symptom of potato brown spot, suggests that Hrip1 may not be produced in significant quantities in brown spot lesions. The brown spot lesions of potato are generally much smaller than early blight, suggesting very limited colonization.
The Hrip1 chlorosis was generated on mature tissue, while younger tissue did not have chlorosis. Whatever the interaction is between Hrip1 and the plant cell, mature tissue seems more susceptible to its actions.