Pea endocarp tissue was obtained from immature pea pods harvested directly from greenhouse-grown (Samish) peas. The pod halves were separated, and the elicitor treatments were applied to the exposed endocarp tissue.

Immature, 2-cm-long pea pods were cut in half. For each sample, one piece (∼1 cm in diameter) was immersed in deionized water in a single well of a white 24-well microplate (PerkinElmer). After an overnight incubation, the solution in the well was exchanged with assay solution containing 100 μM of L-012 (luminol analog; Wako) and 20 μg/ml of horseradish peroxidase (Sigma–Aldrich), with or without SA. The luminescence from each well was measured using an EnSpire multimode plate reader (PerkinElmer).

The bean pathogen F. solani f.sp. phaseoli, Snyder and Hansen (Fsph) (ATCC no. 38135) was donated from the Doug Burke lab, and the pea pathogen F. solani f.sp. pisi (Fspi) was obtained from Lindon Porter, IAREC, Prosser, WA.

Plant tissue was extensively ground in a mortar with liquid N₂, glass beads, and the nucleic acids were extracted in buffer no. 1 [5 M sodium perchlorate, 0.5 M Tris base, 2.5% (w/v) SDS, 0.05% (w/v) NaCl, 0.05 M EDTA]. DNA/RNA were precipitated with 95% (v/v) ethanol, and the pellet was redissolved in water, subsequently extracted with chloroform/phenol, and redissolved in water. The RNA was precipitated from the extract by treating the solution with 2 M lithium chloride. The RNA pellet and the ethanol-precipitated DNA from the supernatant were quantitated in a spectrophotometer at 260 nm. Aliquots of the total DNA were electrophoretically separated on standard 1% (w/v) agarose gels. In addition, 30 μg of the total of each treatment was incorporated into 1 ml of 1% (w/v) CHEF gel (in a 1.5 diameter well) under alkaline conditions to cause DNA strand separation. The solidified gel disk was overlaid with 1 ml of alkaline buffer (30 ml 1 N NaOH and 8 ml 0.5 M EDTA/L) and rotated for 48 h. The DNA fragments eluted into the overlay were precipitated and separated on standard agarose gels. All of the treatments were repeated with similar results.

Changes in the nuclear structure and nuclei diameter were imaged with a fluorescent microscope following staining with the DNA-specific dye, DAPI. Subsequently, the diameters of the nuclei from the digital images were uniformly amplified by photocopying, and 45 nuclei from each treatment were manually measured.

The procedures for the total RNA isolation and purification were performed as described above. The total RNA was subjected to qRT-PCR using a CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories, Inc.) The primers used were described in our previous research (Hadwiger and Tanaka, 2015).

An “oxidative burst” is the rapid release of ROS from stressed plant cells that develops when they come into contact with different pathogens (Grant and Loake, 2000). To detect this early ROS response in pea, a luminol-based oxidative burst assay was performed. As shown in Figure 1, SA treatment induced an oxidative burst with a peak at ∼20 min, whereas water treatment (mock) did not induce an oxidative burst. This method is quite robust and sensitively captures the dynamic changes in ROS production at an early time point in the pea endocarp.

Salicylic acid applied 20 min prior to the inoculum at certain concentrations significantly reduced the linear growth of the true pea pathogen F. solani f.sp. pisi (Fspi) on the endocarp surface. The gradation of action relative to the SA concentration was reproducible over two extensive trials. One of the trials is presented in Table 1, while the other is not shown. The growth of the pea pathogen Fspi on the pea endocarp surface is less than that on water-treated tissue.

Cytological readings (Table 1) of the fungal growth began to demonstrate measurable inhibition after 24 h (Figure 2). An SA dilution series treatment down to the 0.03 μM showed suppressive effects. The characteristic changes in the background hypersensitivity discoloration of the adjacent pea cells suggest that there was a plant-based change in the suppressive effect. Nearly complete and optimal resistance occurred close to the 0.15 μM SA concentration.

Salicylic acid had no significant direct effect on in vitro Fspi growth in liquid media. The microscopic examination of growth after 24 h in Vogel’s media indicated that the Fspi spores germinated and grew uniformly at the concentrations used in Figure 2 (data not shown).

Following the report that SA has an affinity for DNA (Neaualt et al., 1996), it was of interest to examine changes in pea DNA damage in the nucleus and elsewhere within the pea cells. SA applied to the cuticle-free surface of the pea endocarp tissue rapidly caused cytologically detectable changes in the plant nuclei (Figure 3 and Table 2). These changes were related to the SA concentration and the duration of the SA exposure.

DNA fragmentation appeared rapidly (50 min post treatment) and variably with the range of SA treatment concentrations and was generally consistent throughout multiple experiments (Figure 4). Fragmentation was more intense for the treatment with 100 to 6.75 μM SA and for tissues treated with Fsph spores. The specific mechanistic impact of SA on the pea DNA responsible for initiating chromatin transcription is not known for either pea or animal tissues. Maximal transcription of PR genes may depend on a “perfect storm” of conditions and the fragility of chromosomal regions adjacent to the promoter and open reading frame of the gene. Regions of dispersed pea chromatin that are also regions of intense transcription have been detected by electron microscopy (Hadwiger, 2015a) as resistance is developing. Genes within eukaryotic tissues can possess the requisite transcription complex with the proper transcription factors in place and still be silent or stalled (Li et al., 2007). We suggest that there may be stalled PR genes that are activated following major DNA or chromatin structural changes caused by the non-specific SA insults within the adjacent regions.

The reported interaction of SA and DNA did not cause major changes to plasmid DNA (Figure 5). There were detectable, faster migrating DNA molecules generated at the highest SA concentrations. How these minor changes would reflect on the structure of DNA incorporated into pea chromatin is not known. This result may indicate that the DNA fragmentation caused by SA in living tissue could involve additional components.

Induction of PR genes is correlated with the activation of plant defense. We measured the transcriptional induction of the pea PR genes, DRR206, Defensin, PR10, and PR1b in the presence of SA. The results indicate that the expression of the PR genes induced by SA took place mostly at concentrations between 1.5 and 50 μM (Figure 6). The induction levels were comparable to those caused by Fsph.

The elicitation of pisatin, a phytoalexin, serves to indicate the activity of a series of secondary metabolism enzymes from phenylalanine through phenylpropanoid structures to isoflavonoid and other phenolics, many of which have fungal-suppressive properties (Bailey and Mansfield, 1982). Pisatin accumulation is often associated with the induction of immunity in peas (Hadwiger, 2008). The data in Table 3, with a high SA concentration range (15–1000 μM), and in Table 4, with a lower range (0.7–100 μM), recorded at 24 h indicate detectible levels of SA-induced pisatin. The response with both ranges indicates a much lower pisatin accumulation than that induced by the intact microconidia of Fsph during the authentic non-host resistance response. The 1.5 μM SA treatment optimally induced pisatin. However, this value is much lower than the level induced by spores. This result suggests that SA is not a major elicitor of this secondary metabolism route of defense responses in pea at this or higher concentrations of the SA elicitor (Table 4).

The low-level effect of SA on phytoalexin synthesis indicates a departure from the mechanisms of other signals for non-host resistance in pea. However, SA is capable of inducing a response that suppresses the true pathogen of pea and approaches total resistance. The following assay of pisatin production indicates that its effect can be additive to that induced by Fsph, a bean pathogen.

The pisatin levels (Table 5) indicate a marginal increase in synthesis enhanced in the presence of both Fsph and specific SA concentrations. Because of the low strength of the SA-induced pisatin increase, it is likely that the modeling effect of SA on chromatin differs in approach or substance from the DNA single strand cleavage generated by Fsph DNase (Klosterman, et al., 2001).

The enzymatic action of DNase has also been implicated in initiating the transcription of plant defense genes by directly altering nuclear chromatin via single DNA strand nicking. The resultant DNA damage has to be subtle enough to alter chromatin structure in a manner that benefits the pathogen and yet does not initiate processes that could cause immediate cell death (Choi et al., 2001). DNA damage by microbial enzymes that cause double stranded breaks has also been reported (Song and Bent, 2014), and it is likely that this higher level damage is more of a challenge to the plant than the single strand nicking caused by Fsph DNase. Interestingly, the abundance of double strand breaks is reduced by plant defense responses, suggesting that the mechanisms for activating DNA repair processes may share some similarity with the induction of PR genes (Gasser et al., 2005; Yan et al., 2013).

Some current possibilities for the origin, presence, and availability of the SA signal are described in Figure 7. SA is synthesized by bacteria and some fungi (Harper and Hamilton, 1988). SA and methyl-SA can be found in the plant tissue prior to infection and be stored as a byproduct (Maeda and Dudareva, 2012). Hydrogen peroxide is generated in inoculated plant tissue (Coquoz et al., 1988) as tissue damage occurs. In tomatoes, the wound hormone systemin is also produced (Orozco-Cardenas and Ryan, 1999). Hydrogen peroxide can also generate increases in SA. Plants biosynthesize SA using the phenylalanine/cinnamic acid pathway or alternately via benzoic acid (Chen et al., 2009). Both hydrogen peroxide and SA are capable of damaging host DNA. Fungal DNase can directly cleave a single DNA strand. The gene for this potent elicitor has been identified in all fungi whose DNA has been sequenced (Hadwiger and Polashock, 2013). All the DNase proteins are translated with a “signalP peptide” that enables proteins to pass through membranes. Many other eliciting components may be released from fungi, such as the chitosan heptamer that is released from the fungal cell wall (Kendra et al., 1989).

Since SA has recently been reported (Bau et al., 2013) to interact with DNA and has the potential to indirectly influence the state of nuclear DNA by its catalytic inhibition of topoisomerase II, it has the potential to also influence nuclear DNA in plant cells. Single-strand nicks may have previously escaped observance due to their low abundance within the total, large genomic DNA yields from plants. The alkaline processing and agarose trapping of the total DNA enabled the detection of released DNA fragments (see Materials and Methods) that occur in the very early hours of the inductive treatments (Figure 5). This fragmentation was temporally associated with the initiation of defense responses. The sheer presence of the SA association with host DNA is not likely to result in DNA alterations without the presence of a contributing factor such as the direct in vitro association of various SA concentrations.

Pea PR genes map to multiple chromosomes and often reside in regions that also map as QTLs (Pilet-Nayel et al., 2002). PR genes are ubiquitously present in plant genomes and possess antifungal properties, and therefore, they are potentially the major contributors (Chiang and Hadwiger, 1991; Almeida et al., 2000) to disease resistance. It appears that it is the additive effect of multiple PR genes that develops the complete non-host resistance. Additionally, the pea PR genes analyzed share some homology with the PR genes induced by SA in Arabidopsis (Sels et al., 2008).

Gene expression is initiated within chromatin, the site of transcription. The DNA transcription within the region of defense genes can be up regulated or down regulated depending on the associated chromatin structure. Chromatin is a complex of proteins and DNA packed into nucleosomes (Li et al., 2007). The DNA of a particular gene can be accessed by transcription complexes following the alteration of DNA supercoiling or modifications of the relevant proteins. In pea endocarp tissue, PR gene activation is influenced by DNA alterations, by ubiquitination/histone modifications and by a reduction in the architectural transcription factor HMG A (Klosterman et al., 2003; Isaac et al., 2009).

DNA damage can result in the stalling of elongating RNA polymerase II (Lagerwerf et al., 2011) and the attraction of chromatin remodelers to damaged sites. Chromatin modifications function by either disrupting chromatin contacts or affecting the recruitment of non-histone proteins to chromatin (Kouzarides, 2007). Histone modifications can dictate the higher-order chromatin structure in which DNA is packaged, affecting many biological processes. Chromatin structure itself imposes obstacles on all aspects of transcription that are mediated by RNA polymerase II (Li et al., 2007). The resultant chromatin regulation affects the binding of transcription factors and the initiation and elongation steps of transcription. SA reportedly has an affinity for DNA, and similar to other previously described DNA-specific agents, it can cause DNA damage (Neaualt et al., 1996; Yan et al., 2013). The somewhat preferential transcription of PR genes gives credence to the observed selective expression of plant defense resulting from general challenges to sensitive chromatin structures. For transcription to occur, the DNA helix must be opened as the polymerase threads the separated strands through the enzyme (Ma et al., 2013). The upstream torque disrupts the DNA double strand structure and stalls the polymerase, while release of this torsional stress allows the polymerase to resume transcription.

Chitosan, a fungal-derived elicitor of PR genes, can compete with histones for sites on DNA and can insert itself into the minor groove of DNA. Fungal DNase (Fsph DNase), a second major elicitor of PR gene expression, causes single strand cleavage in double stranded DNA, enabling the release of tension within the DNA helical structures (Gerhold et al., 1993). Multiple regulatory substances are released from fungal spores following inoculation on their respective host tissues. Of current interest are the proteins that exit the fungal cell via their SignalP sequences (Ramachandran et al., 2016). The functional properties of defense proteins (described above) can range from those that specialize in digesting the cell wall barriers to those that are metabolic enzymes or proteins with unknown function. The regulatory functions affected within the host tissue may also occur either through receptors and subsequent signal cascades that target transcription factors or as direct insults to the organization of sites within chromatin that result in increases in transcription of host genes (Hartney et al., 2007). Although the signaling of compounds of fungal origin can be shown to specifically complex with membrane proteins and modulate the plant’s response, the resulting signaling cascade to the site of defense gene transcription is less well understood than in many other host/parasite interactions (Presti et al., 2015). High throughput genomic analyses may be able to detect a multiplicity of potential effectors with the potential to traverse the host-parasite barrier. If so, future research should focus on determining which effectors display potential to play a major role in the processes that result in the development of resistance or enable a susceptible response.

The biotic and abiotic elicitors of PR genes, such as the single strand cleaving DNase elicitor from Fsph, require a SignalP sequence (Hadwiger and Polashock, 2013). Since homologous fungal genes for DNase production are present in all of the fungal genomes sequenced to date, this chromatin modeling may be implicated in similar signaling in many other plant/fungal interactions (Hadwiger and Polashock, 2013). DNase activity has also been shown to be released from spores of rust (Puccinia striiformis), Verticillium dahliae, Colletotrichum coccodes and yeast cells (Hadwiger and Polashock, 2013). The universality of this enzyme suggests that it could be a general elicitor of the non-host resistance response, protecting plants from pathogens known to be out of their host range. DNase enzymatic action has also been implicated in the initiation of plant defense gene transcription by directly altering nuclear chromatin via single DNA strand nicking. The resultant DNA damage must be subtle enough to alter the chromatin structure for the benefit of the pathogen but not initiate processes that could cause immediate cell death.