Phytohormone crosstalk in the host-Verticillium interaction

ABSTRACT

Functional genomics can be applied to shed light on the Brassica napus – Verticillium interaction. RNAseq data indicated already that abscisic acid (ABA) is apparently involved in the early oilseed rape response to fungal infection with Verticillium longisporum isolate 43 (Vl43). A calreticulin (CRT1a) was identified as novel susceptibility factor for Vl43 infecting both Arabidopsis and oilseed rape. CRT1a is involved in calcium homeostasis and contributes in the endoplasmatic reticulum to the unfolded protein response. The latter function could either affect the correct folding of other susceptibility factors or of negative regulators in ethylene (ET) signaling. Which CRT1a function is affected in the mutants is currently unknown, but both hypotheses can explain that crt1a loss-of-function mutants display increased resistance to V. longisporum and enhanced expression of ethylene signaling related genes. This indicates that besides other phytohormones such as ABA or salicylic acid (SA) also ET plays a critical role in the plant-Verticillium interaction, which might be exploited to improve plant resistance.

KEYWORDS:

• Functional genomics
• susceptibility factors
• phytohormone crosstalk
• Brassica napus
• Arabidopsis
• Verticillium longisporum
• resistance breeding

A functional genomics approach to identify novel susceptibility factors involved in host-pathogen interactions

Oilseed rape is a rather young crop species with a complex allotetraploid genome, but low genetic variation which makes resistance breeding very challenging.¹ The hemibiotrophic vascular fungal pathogen Verticillium longisporum causes stem striping on oilseed rape and since there is no chemical control, resistance breeding is the method of choice.² However, the identification of new target genes for genome editing in pathogen resistance breeding is a major bottleneck. Functional genomics can be deployed to overcome this obstacle and identify candidate genes encoding novel susceptibility factors by differential expression analysis (Figure 1). Plant transcripts might be up-regulated during infection either due to their function in the host defense reaction or due to manipulation by the pathogen to gain an advantage (①). Bioinformatic analysis e.g. of RNAseq experiments, including meta-studies (②), helps to identify up-regulated genes, which are potentially interesting candidate genes for further analysis (③). Since oilseed rape and the model plant Arabidopsis are related Brassicaceous species, a pre-screen deploying Arabidopsis T-DNA mutants can be used to identify promising candidates for genome editing in oilseed rape (④). Mutation of the candidate genes in oilseed rape can be easiest accomplished by genome editing, e.g. via CRISPR/Cas. To achieve this, another bioinformatics approach has to be conducted to select a suitable region in the candidate gene to be targeted by a locus-specific sgRNA (⑤). After crop transformation and regeneration of plants (⑥) it is advisable to propagate the seeds first in order to obtain T2 plants with a more homogenous background concerning the target site mutations. In a final test these plants are challenged with the pathogen or other stress of interest and in best case some plants show resistance to fungal infection depending on the type of mutation (⑦). From 20 selected candidate genes analyzed by deploying corresponding Arabidopsis T-DNA lines in infection assays with the phytopathogenic fungus Verticillium longisporum (Vl43), CRT1a was identified as a novel susceptibility factor for this pathogen.³

Figure 1. Simplified pipeline to identify candidate genes encoding novel susceptibility factors by functional genomics, for details see text.

Phytohormone crosstalk in the host-Verticillium interaction

To get a clue about the possible mechanism(s) which might be influenced by the CRT1a mutation, expression of phytohormone related marker genes can be investigated and both Brassica napus and Arabidopsis mutants displayed enhanced expression of ethylene (ET) signaling related genes (such as EIN2) as well as higher resistance to Vl43. Comparing the stress-related phytohormone biosynthetic processes in both plant species by RNAseq from wild type plant roots harvested at 6 days post inoculation (dpi) with Vl43, it was found that the hormone responses appeared to differ between the crop and the model plant with respect to abscisic acid (ABA). Upon infection, ABA responses were suppressed in Brassica napus, but not altered in Arabidopsis. Expression of the JA/ET marker gene PDF1.2 in oilseed rape was much less pronounced (4-fold) than in Arabidopsis (200-fold), where this response was similarly strong as the SA response,⁴ probably because ABA is known to positively influence JA biosynthesis and signaling.⁵ However, on RNAseq level JA responses did not appear to be differentially regulated comparing both plant species. Expression of the SA marker gene PR1 was up-regulated in both oilseed rape (500-fold) and Arabidopsis (200-fold).⁴ This could be further explained by mutually antagonistic effects between SA and ABA^6-10 with SA levels being elevated when ABA is suppressed as observed in oilseed rape during early infection with Vl43 and confirmed by the finding that in the Arabidopsis ABA mutant nced3, PR1 expression was enhanced, while PDF1.2 and EIN2 expression were suppressed compared to Col-0 wild type.⁴ This was surprising, because also the interaction between ABA and ET signaling is reported to be mutually antagonistic^11,12 indicating a dominant negative effect of elevated SA signaling on ET and JA. In Arabidopsis, ABA responses were not suppressed, which is in line with observations that ABA levels tend to increase in Vl43 infected roots at 6 and 8 dpi¹³ and also in leaves early at 2 dpi¹⁴ or later at 15 dpi.¹⁵ Generally, ET and JA are believed to act in concert to orchestrate the plant defense response to necrotrophic pathogens and are antagonistically regulated by SA, being responsible for the defense responses to biotrophic pathogens.^16,17 Especially the SA inducible WRKY70 has been implicated in negative regulation of JA responses in Arabidopsis,¹⁸ wheat¹⁹ and cotton.²⁰ Compared to PDF1.2 and PR1 marker gene induction only a weak ET marker gene (EIN2) expression was observed in both Arabidopsis and oilseed rape,⁴ but over-expression of some ET responsive transcription factors (ERFs) increased Arabidopsis resistance to Vl43²¹ and other pathogens.²² In crt1a mutants of both plant species enhanced ET marker gene expression goes along with reduced susceptibility, indicating ET plays a more important role than previously anticipated and that there might be extensive phytohormone crosstalk determining the outcome of this compatible interaction. In Arabidopsis it was shown that infection with the Verticillium longisporum isolate VD11 gradually induced ET production and further reduced the fresh weight in the ET mutants ein2-1 (pos. regulator in ET signaling) and ein4-1 (an ET receptor, involved in negative regulation of ET signaling) as compared to Col-0 plants, while fresh weight increased in etr1-1 (another ET receptor, also negatively regulating ET signaling) mutants.²³ Though data concerning ET receptor mutants are somewhat ambiguous, all in all it appears that ET is required for the defense response to V. longisporum. In the Arabidopsis – V. longisporum interaction it has been demonstrated by mutant analysis that the fungus requires a COI1-dependent but JA-Ile-independent mechanism to efficiently colonize Arabidopsis.¹⁵ JA and MYC2 are required for full induction of monoterpenes production, which can stimulate germination and enhance V. longisporum invasion thus myc2-1 mutants display higher resistance.²⁴ In contrast, the coi1-16 mutant was more susceptible in the screen of various Arabidopsis genotypes and the eds8-1 or jar1-1 mutants were more resistant than Col-0.²³ However, the Arabidopsis npr1-1 mutant was more susceptible and showed decreased responses to the ethylene precursor ACC or MeJA pretreatments, indicating a major involvement of the SA pathway in this interaction.²³ Furthermore, these authors showed that SA-dependent PR1 was elevated highest at 7 dpi when compared to PDF1.2 and the ET-dependent PR4 expression.²³ It is important to note that the outcome of an interaction between phytohormones depends largely on their equilibrium, which could explain sometimes contradicting observations and makes prediction of phytohormone crosstalk difficult. For example, the relationship between ET and SA could be interpreted as being mutually antagonistic since a negative effect of SA on ET has been described in rice,²⁵ pear,²⁶ mustard,²⁷ melon,²⁸ oilseed rape²⁹ and Arabidopsis³⁰ or negative effects of ET on SA in Arabidopsis^31,32 and rice.³³ However, other reports indicate positive effects between both hormones, e.g. in Arabidopsis³⁴ and carrot³⁵ (SA on ET) or tomato³⁶ and Arabidopsis³⁴ (ET on SA), often depending on the concentration of applied phytohormones. Also during pattern triggered immunity (PTI) it has been shown in tobacco that low amounts of PAL-derived salicylic acid are required to induce ET synthesis, while high SA amounts could inhibit the elicitor-induced ET production.³⁷

Conclusion

The SA marker gene expression is higher in Brassica napus compared to Arabidopsis,⁴ which could be explained by reduced ABA levels¹⁰ and since transgenic Brassica napus NahG (SA-deficient) plants displayed increased susceptibility to V. longisporum,³⁸ basal resistance of oilseed rape to V. longisporum could be mainly attributed to this phytohormone. As discussed above, data concerning the JA response with respect to Arabidopsis infection with V. longisporum are not consistent. At least in Brassica napus the PDF1.2 expression was much lower than in Arabidopsis,⁴ which could be also explained by reduced ABA levels.⁵ Thus, all these differences between oilseed rape and Arabidopsis may be caused by the suppression of the ABA response which was observed in the oilseed rape – Verticillium interaction. Since ET responses appear to be activated in the crt1a mutants of Arabidopsis and Brassica napus, it is possible that ET presence at the time of infection inhibits disease development, while at later stages it might contribute to wilt development as discussed for tomato and V. dahliae.³⁹ A working model explaining the interplay of ABA, SA, JA and ET in the oilseed rape – Verticillium interaction is presented in Figure 2.

Figure 2. Working model of phytohormone crosstalk during the Brassica napus – V. longisporum interaction.① ABA is repressed in oilseed rape, but not in Arabidopsis. ② Therefore SA can accumulate higher compared to Arabidopsis. ③ Higher SA and lower ABA levels negatively influence JA, which is stronger activated in Arabidopsis. ④ ET responses could be activated in the crt1a mutants due to improper folding of negative regulators. ⑤ ET might directly benefit SA accumulation or indirectly via negatively affecting ABA ⑥. It has been proposed that the reduced ABA response might help the fungus to establish a long-lasting compatible interaction.⁴ while the mechanism underlying resistance in the crt1a background is currently unknown and could also involve correct folding of other – yet unknown – susceptibility factors ⑦.³

While manipulating ABA signaling could severely affect plant responses to abiotic stress, the crt1a mutant with enhanced ET signaling did not appear to negatively affect plant physiology under greenhouse conditions. Other hormone pathways are not strongly affected indicating ET signaling is only moderately activated,³ which could facilitate a quicker launch of SA responses (“priming”). Increasing genetic variation via CRISPR/Cas-mediated genome editing is therefore a promising approach to create new traits in crops. The deployment of functional genomics for the identification of novel target genes in resistance breeding is a pre-requisite to develop hypotheses which could be tested by genome editing via CRISPR/Cas. Such mutations are valuable resources to conduct further basic research in order to understand the molecular function of genes involved in the plant-pathogen interaction and resistance breeding. However, success in the greenhouse is not the final step in the breeding process and requires subsequent testing under field conditions, which is currently impossible in the EU.⁴⁰ In light of climate change it is urgently required to change the legal status of genome edited mutant plants to complete the breeding process and develop new crop varieties which meet all requirements, for the consumer, the environment and the farmers.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.