Discussion
We have assessed the capacity of chitosan to induce resistance againstB. cinerea in different plant species and have linked its effect with priming of defence mechanisms. We have identified a formulation of chitosan that unlike some other formulations, can be easily dissolved in water and does not require infiltration. This opens possibilities to identify early-acting priming mechanisms in elicitor-induced resistance. Moreover, it enables opportunities for upscaling the use of chitosan as an elicitor of resistance in large-scale experiments due to the high-throughput nature of spraying the elicitor onto plants.
Treatments with chitosan resulted in induced resistance in S. lycopersicum (Fig 1a), S. melongena (Fig S1), Arabidopsis (Fig 1b) and N. benthamiana (Fig 4a) at a range of concentrations, which indicates that there are similar defence mechanisms acting in the response to fungal PAMPs. Moreover, treatments with chitosan resulted in the activation of basal resistance processes such as the deposition of callose at the cell wall (Fig 1c,d), which is considered an important factor for penetration resistance against invading pathogens (Oide et al., 2013). Expression of resistance was dependent on the concentration of chitosan used in Arabidopsis. In contrast, in tomato and aubergine the levels of resistance did not depend on the chitosan concentration. Moreover, chitosan-induced callose deposition in tomato and Arabidopsis did not follow a classical dose-response curve and the most effective treatments that activated callose were the lower concentrations of the elicitor (Fig 1c,d). This is likely to be dependent on the antimicrobial effects of chitosan (Fig S2) at higher concentrations. Other elicitors have been shown to trigger induced resistance phenomena at lower concentrations. For example, meJA treatment results in more effective protection against the pathogen Fusarium oxysporum f.sp.lycopersici when applied at lower concentrations (Król et al., 2015). In contrast, high doses of MeJA had detrimental effects on physiological processes and overall decreased protection efficiency. This, together with the observation that low concentrations of chitosan do not directly impact pathogen growth (Fig S2) suggests that there is a concentration threshold in the effect of chitosan-induced resistance.
Foliar applications of chitosan have been widely used to control disease development caused by numerous pests and pathogens (El Hadrami et al. 2010). However, few studies have investigated the role of chitosan as a priming agent and most have focused on its use as a seed priming elicitor mainly to improve germination and yield (Guan et al., 2009, Hameed et al., 2013). Here, we show that chitosan-induced resistance is based on priming of defence mechanisms. Our experiments confirmed that chitosan-induced resistance is not associated with growth reduction (Fig S3a), was durable and maintained for at least two weeks after treatment (Fig 2a), and that is based on a stronger accumulation of callose at the site of attack and accumulation of JA (Fig 2d) and JA-ile (Fig S3b). These results demonstrate that fungal growth arrest after chitosan treatment is not directly mediated by the toxicity effect of the chemical, as the infected leaves were formed after treatment and therefore were not sprayed with the elicitor. Moreover, these results demonstrate similar priming mechanisms after chitosan treatment to other elicitors, including Hx, which has been linked with priming of callose and JA against B. cinerea (Fernández-Crespo et al., 2017, Wang et al., 2014). Interestingly, however, despite many reported antagonistic and other crosstalk interactions between plant hormones (Robert-Seilaniantz et al., 2011), the concentrations of other plant hormones, SA and ABA, were not affected. This suggests that priming by chitosan does not result in the downregulation of other hormone-dependent signalling pathways, thereby maintaining an effective resistance status against other stresses.
In order to further explore priming of defence and to unravel the transcriptional mechanisms behind chitosan-induced resistance, we performed transcriptome analysis. In our experiment, using a concentration of chitosan that is associated with priming but with no direct antimicrobial effect, we identified early-acting differential transcriptomic changes. Results demonstrate that chitosan treatments do not result in major transcriptional changes (Fig 3a). In contrast, comparison of treatment against Water + Mock revealed and Chitosan +B. cinerea shows a higher number of DEGs (Fig 3b,c), thus responding to the priming nature of the elicitor in the first instance.
Panther enrichment analysis showed that at 6hpi, the number of down-regulated DEGS was more than three times up-regulated ones for Chitosan + B. cinerea (203 down-regulated and 57 DEGs up-regulated). This suggests that tomato plants might repress susceptible factors in order to reduce B. cinerea manipulation of host defences (El Oirdi et al., 2011, Temme and Tudzynski, 2009). Interestingly, some of the down-regulated transcripts have cysteine-type peptidase activity (Table S1). These proteins have been reported to have a role in immunity against pathogens including B.cinerea (Pogány et al., 2015). Other down-regulated genes are related to plant hormone activity; including ethylene AP2/ERF transcription factors and ABA PYL receptors (SlABAPYL4), reported to be involved in defence responses, which act as positive or negative regulators of JA/ET-dependent defences against B. cinerea (Cantu et al., 2009, Moffat et al., 2012). Up-regulated genes included transcripts with peroxidase and transcription regulatory activity, such as peroxidase 5, SlMYB20, SLWRKY51 and SlWRKY72, CONSTANS-like protein with zinc finger binding domain and NAC domain protein and a RING-type E3 ubiquitin transferase involved in protein degradation. These genes have been linked with defence responses (Serrano et al., 2018), which could be result in priming of the tomato immune system against B. cinerea infection.
Transcriptomic (Table S2, Fig S4a) and qRT-PCR (Fig S4b) analyses showed that chitosan can prime ACRE75 for a faster and stronger expression after infection with B. cinerea . ACRE genes have been linked to plant defence responses. Similar genes were previously identified in tobacco cells to exhibit rapid Cf-9–dependent change in expression through gene-for-gene interaction between the biotroph pathogen Cladosporium fulvum avirulence gene (Avr9) and tomato resistance Cf-9 gene (Durrant et al., 2000). To determine the role of ACRE genes in priming by chitosan, we searched for other ACRE genes showing similar expression profiles to ACRE75 and this revealed that ACRE180 displays a similar priming profile. This was more evident at 9 hpi (Fig S4b) than at 6 hpi, suggesting that the role of ACRE180 is later time than ACRE75. Subcellular localisation may indicate why priming of these genes does not occur at the same time; whereas ACRE75 accumulates exclusively in the nucleus and nucleolus (Fig S6e,f), ACRE180 accrues in the ER and peroxisomes (Fig S6g,h). This suggests different molecular functions of these proteins as they tag different cell organelles. Moreover, it could be plausible that ACRE75 and ACRE180 are part of the same signalling pathway, one working upstream of the other, therefore justifying the delayed transcription and activity of ACRE180.
The roles of ACRE75 and ACRE180 in chitosan-induced priming were investigated by overexpressing these genes in transient and stable systems, in N. benthamiana and Arabidopsis, respectively. Moreover, we aimed to identify any N. benthamiana and Arabidopsis ACRE75 and ACRE180 analogues. BLAST analysis of tomato ACRE75 identified a very low amino acid identity sequence (39%) and ACRE180 failed to identify any Arabidopsis homologue. In contrast, N. benthamianaACRE75 and ACRE180 homologues were putatively identified. ACRE75 and ACRE180 lack signal peptides, which suggests they might encode small proteins involved in signalling or antimicrobial activity within the infected cell. Similar to the exclusive production of glucosinolates compounds in Brassica plants (Matthaus & luftmann 2000) it is likely that ACRE75 and ACRE180 are involved in the production of unique compounds to Solanaceae plants. Overexpression of SlACRE75 andSlACRE180, and their N. benthamiana orthologues results in induced resistance against B. cinerea (Fig 4b,c). Therefore, our results confirm involvement of ACRE genes in plant immunity and suggest an involvement in chitosan-induced priming due to their expression profiles. Interestingly, the induced resistance effect was greater in Arabidopsis plants overexpressing ACRE75 in comparison to ACRE180 (Fig 4c), which could corroborate our evidence of earlier activity of ACRE75, therefore being more effective during early resistance response. More work in needed to unravel the molecular function of ACRE75 and ACRE180 in the expression of priming mechanisms. Nevertheless, fine-tuning of priming-based mechanisms under the control of SlACRE75 ,SlACRE180 , NbACRE75 and NbACRE180 could facilitate its incorporation into other crop species for the enhancement of cross tolerance to old and emergent pest and pathogens, and other challenges. The results unveiled potential molecular pathways involved in chitosan-induced priming of resistance in tomato against B. cinerea , potentially applicable to other crops.