Chitosan increases Pinus pinaster tolerance to the pinewood nematode (Bursaphelenchus xylophilus) by promoting plant antioxidative metabolism

The pine wilt disease (PWD), for which no effective treatment is available at the moment, is a constant threat to Pinus spp. plantations worldwide, being responsible for significant economic and environmental losses every year. It has been demonstrated that elicitation with chitosan increases plant tolerance to the pinewood nematode (PWN) Bursaphelenchus xylophilus, the causal agent of the PWD, but the biochemical and genetic aspects underlying this response have not been explored. To understand the influence of chitosan in Pinus pinaster tolerance against PWN, a low-molecular-weight (327 kDa) chitosan was applied to mock- and PWN-inoculated plants. Nematode population, malondialdehyde (MDA), catalase, carotenoids, anthocyanins, phenolic compounds, lignin and gene expression related to oxidative stress (thioredoxin 1, TRX) and plant defence (defensin, DEF, and a-farnesene synthase, AFS), were analysed at 1, 7, 14, 21 and 28 days post-inoculation (dpi). At 28 dpi, PWN-infected plants elicited with chitosan showed a sixfold lower nematode population when compared to non-elicited plants. Higher levels of MDA, catalase, carotenoids, anthocyanins, phenolic compounds, and lignin were detected in chitosan-elicited plants following infection. The expression levels of DEF gene were higher in elicited plants, while TRX and AFS expression was lower, possibly due to the disease containment-effect of chitosan. Combined, we conclude that chitosan induces pine defences against PWD via modulation of metabolic and transcriptomic mechanisms related with plant antioxidant system.


Results
Nematode population in infected plants. Total nematode count inside stem tissues of inoculated plants without chitosan elicitation significantly increased by 44-fold from 1 to 28 dpi, indicating a successful infection and multiplication of PWN inside plant tissues over time (Fig. 1). In contrast, with chitosan elicitation the number of nematodes was low and statistically identical over the entire experimental period. In fact, only about 0.14-fold of the number of nematodes found in non-elicited plants were recovered from plants inoculated in the presence of chitosan, being the effect of the chitosan treatment significant (P < 0.0001) ( Table 1).
Oxidative stress-related mechanisms. Plants inoculated with PWN displayed a progressive and significant increase in MDA concentration throughout the experimental period, regardless of chitosan elicitation ( Fig. 2A), with a significant effect of the chitosan treatment and the DPI (Table 1). In fact, at the end of the experimental period, PWN-inoculated plants elicited with chitosan had significantly higher MDA concentration than non-elicited plants (by 1. 8-fold). Similarly to MDA, during the first 14 days of infection, catalase activity was not significantly affected in PWN-infected plants, regardless of chitosan elicitation, which remained similar to catalase levels in water-inoculated plants (Fig. 2B). Non-elicited inoculated plants presented a progressive increase in catalase levels and, at 28 dpi, these were 4.5-fold significantly higher than at 1 dpi. In chitosan-elicited plants, a more drastic and significant increase was registered after 14 dpi, with catalase activity increasing by 25-fold at 21 dpi, resulting in a significant interaction between chitosan treatment and DPI (Table 1).
Plants behaved similarly regarding the accumulation of carotenoids (Fig. 3A) and anthocyanins (Fig. 3B). From 1 to 28 dpi PWN-infected plants without chitosan elicitation showed a progressive and significant increase of 15-and sevenfold in carotenoid and anthocyanin concentrations, respectively. In chitosan-elicited plants, carotenoid and anthocyanin concentrations did not vary significantly between control and PWN-infected plants. www.nature.com/scientificreports/ Hence, no significant effect was found for the chitosan treatment for both analytes (Table 1). However, their concentration abruptly increased at 28 dpi by 21-fold in carotenoids concentration and 13-fold in anthocyanins concentration (when compared to the 21 dpi time-point), with a significant effect of the DPI and its interaction with the chitosan treatment (Table 1). For all time-points, total phenolic concentration was about two times higher in PWN-infected plants with chitosan elicitation (P < 0.0001), when compared to non-elicited plants (Fig. 4A, Table 1). At 14 dpi, plants treated with chitosan presented the highest levels of total phenolic compounds. Moreover, in both elicited and non-elicited plants, polyphenols concentration was drastically reduced in infected plants at the end of the assay (28 dpi), returning to concentration levels similar to non-infected plants. In general, lignin concentration was     www.nature.com/scientificreports/ similar between PWN-infected plants regardless of chitosan elicitation (Fig. 4B). However, at 28 dpi, inoculated plants showed increased lignin concentration (as compared with water-inoculated control plants), particularly with chitosan elicitation (by ca. twofold). No significant interaction between chitosan treatment and DPI was found for both total phenolic compounds and lignin accumulation (Table 1).

Relative expression of stress-related genes.
Regarding the expression of TRX1 gene (Fig. 5A), nonelicited PWN-infected plants showed an increase by more than twofold at 7 dpi, which was followed by an eightfold increase at 14 dpi, when compared to 1 dpi. The expression decreased to less than half of the values seen at 14 dpi, and remained at this level until the end of the experiment. For elicited PWN-infected plants, TRX1 relative expression was also significantly increased at 7 dpi, however, it immediately decreased in the following time-points and was generally lower than the expression values in non-elicited plants. Following PWN infection, DEF expression ( Fig. 5B) was differently modulated throughout the experiment depending on the chitosan treatment (P < 0.05). In non-elicited plants, from 1 to 7 dpi it was significantly reduced to almost null values and maintained at this level until 21 dpi, until its expression increased again, by about twofold, at 28 dpi. In elicited plants, there was a significant decrease in DEF expression at 7 dpi, but at 14 dpi its expression increased again, remaining unaltered until the end of the experimental period. Chitosan elicitation significantly impacted the expression pattern of AFS (Fig. 5C), with non-elicited PWN-infected plants having a progressive increase in its expression over time, reaching the highest value at 28 dpi (ninefold higher than at 1 dpi). Elicited plants showed constant and relatively low AFS expression throughout the experiment, which was twofold higher in PWN-infected plants than in controls. Chitosan treatment and DPI had a significant effect in the expression of all genes, also having a significant interaction among them (Table 1).

Discussion
When the PWN infects susceptible plant species, there is a primary phase of the disease (early phase), where the nematode population remains low, without producing any visual symptoms. When the second phase develops (advanced stage), PWN population increases exponentially, beginning a more destructive phase for the plant due to nematode feeding and reproduction on the epithelial cells 6,7,36 . Previous histological and biochemical works on the PWN population dynamics demonstrated that the PWD progresses strongly between 7 and 14 days after infection 37,38 , which is in accordance to what was observed in the present work in the absence of chitosan elicitation. It has been previously shown that chitosan application prevents nematode multiplication in plant tissues over time 24 , and reduces the number of juveniles in soil 39 . The present work seems to attest for the promising role of chitosan elicitation of P. pinaster plants against the PWD, as the nematode population remained at low levels throughout the entire experimental period, indicating that this compound prevented, or, at least, slowed disease progression.
In susceptible Pinus spp. host plants, PWN damages xylem and phloem parenchyma and cortex cells, which can compromise water transport, lead to the peroxidation of the unsaturated lipids of cellular membranes and to ROS formation, culminating in leaf necrosis and often in plant death 7,40 . MDA, a secondary compound of lipid peroxidation reactions, is frequently used as an indicator of cell damage induced by free radicals 41 . Interestingly, despite preventing nematode reproduction in plants tissues, chitosan did not prevent cellular damage, leading to increased MDA accumulation. This could be a result of the abrupt induction of plant defence mechanisms upon PWN infection, with consequent formation of ROS 42 . To counteract the negative effects of intracellular ROS levels, increased or de novo synthesis of several antioxidant enzymes occur. One of those enzymes, catalase, can capture H 2 O 2 and convert it in water and oxygen molecules, which do not have a damaging impact to plant cells 43,44 . Here, following PWN inoculation, catalase activity was significantly induced throughout the progression of the disease, although more drastically in chitosan-elicited plants, putatively representing the activation of plants' antioxidant defences. Hydrogen peroxide-mediated oxidative burst has been observed in several plant species supplemented with chitosan, and it is generally acknowledged that it is required for chitosan-induced defence responses 25,45 . Increased H 2 O 2 content in chitosan-elicited plants in generally accompanied by the accumulation of several antioxidant enzymes, among which catalase and peroxidase seem to be the most important [46][47][48] . The application of chitosan may, therefore, lead to a stronger, and possibly more efficient, antioxidant response against PWN infection, which may contribute to an increased tolerance against the pathogen.
Carotenoids and anthocyanins are metabolites involved in plant protection by preventing photosystem damage, participating in signalling, ROS scavenging and inhibiting the growth of noxious microorganisms [49][50][51] . It appears that their concentration increased in a similar manner to the PWN population, MDA concentration and catalase activity, as they gradually increased in inoculated plants (as compared to the controls), in both elicited and non-elicited plants. Previous studies have reported that chitosan-elicitation promotes plant defences through increased activity of several defence-related enzymes, e.g. phenylalanine ammonia-lyase, involved in the biosynthesis of phenolic compounds and terpenoids (which include, among several others, carotenoids and anthocyanins) 52,53 .We hypothesise that increased concentrations of carotenoids and anthocyanins in PWNinfected tissues, especially by the end of the experimental period, may be a coping mechanism to prevent further cellular damage caused by the pathogen 54 . As observed for catalase, chitosan elicitation seems to promote the accumulation of these metabolites, thus corroborating their potential in improving plants' antioxidant defences. Along with carotenoids and anthocyanins biosynthesis, phenolics and lignin are also involved in plant resistance to biotic stress 15,55,56 . In general, total phenolic compounds concentration increased at 14 dpi (with statistical significance in PWN-infected elicited plants), which seems to correspond to the beginning of the advanced stage of the PWD. This is concordant with a previously reported positive relationship found between nematode migration and phenolics concentration 57 . In the case of lignin, chitosan-elicited plants had higher accumulation than non-elicited ones, which may indicate an intensified tissue lignification after pathogen infection, possibly  58 . Chitosan also induces a generalized H 2 O 2 -mediated hypersensitive response, involved in cell wall lignification processes that confer protection against pathogen-induced mechanical damage 30,59 . The higher accumulation of carotenoids, anthocyanins, phenolics and lignin observed in the present study, particularly during the later stages of PWN infection and with chitosan elicitation, are in agreement in these previous works, and attest the potential of chitosan elicitation in promoting Pinus spp. defences against the pathogen. www.nature.com/scientificreports/ In plants, thioredoxins act as signalling molecules through their reduction-regulatory properties in response to pathogen attacks 60,61 . In the present work, significant overexpression of TRX1 was observed in non-elicited plants at 14 dpi, corroborating the hypothesis that the advanced stage of the PWD was triggered around this time-point. In PWN-infected elicited plants, TRX1 expression levels were not induced at this time-point, which is coherent with the lower nematode population and increased antioxidative activity previously discussed. Plant defensins, on the other hand, are usually associated with antimicrobial and antifungal activity, constituting the first defensive line against pathogens 62,63 . It is interesting to note that, here, DEF expression pattern was very distinct between chitosan-elicited and non-elicited plants following PWN infection. It seems that, without chitosan elicitation, plants were not able to activate the expression of this gene as the PWN population progressed, whereas in elicited plants DEF expression was significantly increased along time, possibly contributing to the lower extent of disease progression observed. Also involved in plant defence is α-farnesene (AFS), an herbivoreinduced plant volatile, usually increased in plants suffering pathogen attacks, being key in disease resistance 18,64 .
Here, non-elicited plants which showed increased nematode population over time, also displayed higher levels of AFS expression, particularly at 28 dpi, when the highest number of nematodes were registered. Contrastingly, chitosan-elicited plants had lower levels of AFS expression, possibly due to the low nematode progression and multiplication within their tissues.
In general, chitosan increased P. pinaster tolerance to the PWN, resulting in decreased nematode density, increased accumulation of metabolites involved in antioxidant activity, and differential expression of stressrelated gene expression. This further supports the potential of chitosan in the prevention and/or treatment of the PWD in Pinus spp., which should be further explored and considered for the development of sustainable disease control strategies. Nematode culture and plant inoculation. Fourteen days following plant elicitation, a virulent strain of PWN (65 GO, isolated from Góis, Portugal) was inoculated into plant stems. Nematodes were maintained in agar plates with Botrytis cinerea mycelia for 14 days, after which they were extracted from the growing media using the Baermann funnel technic 65 . Plant inoculation was performed with 2 000 nematodes at ca. 3 cm from the top of each plant as described by Nunes da Silva et al. 24 .

Methods
Experimental design and plant sampling. A total of 150 P. pinaster plants were used in this experiment. Chitosan was applied to a group of 75 plants, in which 25 were mock-inoculated with water (non-inoculated controls) and 50 were inoculated with PWN. An additional group of 75 plants served as non-elicited control, where a 0.5% solution of acetic acid (without chitosan) was added to the substrate: 25 plants were mockinoculated with water (non-inoculated controls) and 50 were inoculated with PWN. At 1, 7, 14, 21 and 28 days post-inoculation (dpi) five biological replicates of each treatment were sampled. At each time-point, leaves were recovered for the quantification of lipid peroxidation (MDA), catalase, carotenoids and anthocyanins, and stems were used for total soluble phenolics and lignin quantification and gene expression analysis. Five additional PWN-inoculated plants were sampled at the same time-points for whole-stem PWN quantification.

Nematode quantification in plant tissues.
In each time-point, five inoculated plants from each treatment were randomly selected for PWN quantification. Leaves were removed, and the entire stem was cut into small portions (ca. 0.5 cm) and used for nematode quantification using the Baermann funnel technique for 24 h at 25 °C. Nematode density in stem tissues was estimated taking into consideration whole stem fresh weight.
Lipid peroxidation and catalase. Quantification of MDA, a sub-product of lipid peroxidation, was performed by mixing vigorously 100 mg of leaf tissue previously homogenized in liquid nitrogen with 10 mL of 0.5% thiobarbituric acid in 20% of trichloroacetic acid and incubated at 100 °C for 30 min. After the incubation period, the reaction was terminated by transferring the tubes into ice. Samples were centrifuged during 10 min at 5 000 g and the supernatant was filtered, after which absorbances were measured at 450, 532 and 600 nm and MDA estimated as described by Yang et al. 66 .
Quantification of catalase activity was performed following the protocol by Ruley et al. 67 , adding 1.5 mL of phosphate buffer 1 M to 100 mg of leaf tissue previously homogenized with liquid nitrogen. Samples were vigorously mixed for 2 min, centrifuged at 5 000 g for 10 min at 4 °C and the supernatant was recovered and diluted 3 times with phosphate buffer. Three hundred and forty-four microlitres of a 73 mM H 2 O 2 solution (in 0.5 M Tris/ HCl pH 7.0) were then added to 666 μL of sample extract. Enzyme activity was measured for 3 min at 240 nm. One unit of enzyme corresponded to a decrease of 0.001 in absorbance.  69 . Lyophilized leaf tissues previously homogenized with liquid nitrogen (100 mg) were extracted with 5 mL of methanol for 24 h in the dark at 4 °C. One hundred μL of sample extract were mixed with 5 mL of ultrapure water and 0.5 mL of Folin-Denis reagent, stirred vigorously and incubated at room temperature for 5 min. After incubation, 1.5 mL of sodium carbonate at 35% (w/v) was mixed with each sample and let to react in the dark for 2 h, after which 2.9 mL of ultrapure water were added and absorbances were determined at 760 nm. Soluble phenolic compounds were determined using a quercetin calibration curve.
Lignin was determined by the acetyl bromide method 70 using the leaf biomass used for phenolics extraction, which was subjected to sequential 24 h extractions with water, acetone and hexane. Samples were dried at 60 °C for 48 h, and 10 mg of biomass were mixed with 500 µL of glacial acetic acid and 500 µL of acetyl bromide 25%. The mixture was digested at 50 °C for 2 h with vigorous stirring, after which samples were centrifuged for 10 min at 15 000 g. One hundred microlitre of sample were recovered to a new microcentrifuge tube and homogenized with 200 µL of glacial acetic acid, 150 µL of NaOH 0.3 M, 50 µL of hydroxylamine hydrochloride 0.5 M and 500 µL of glacial acetic. Absorbances were recorded at 280 nm and lignin concentration in each sample was determined through a standard calibration curve.
Gene expression analysis. From the five biological replicates from each treatment, three were randomly selected for gene expression analysis. After stem tissue homogenization with liquid nitrogen, RNA extraction was performed using the RNeasy Mini Kit (QIAGEN GmbH, Hilden, Germany) following the manufacturer's protocol. RNA quality and quantity were determined spectrophotometrically and single-strand cDNA was synthesised using iScrip cDNA Synthesis Kit (Bio-Rad Laboratories, Inc, California, USA) following the manufacturer's instructions. The transcript levels of genes encoding proteins related to oxidative stress (thioredoxin 1, TRX) and plant defence (defensin, DEF; α-farnesene synthase, AFS) were analysed. Primer sequences were retrieved from available published literature 18 . Reactions of RT-qPCR were carried out in a thermal cycler CFX96 Touch Deep Well Real-Time PCR Detection System using iQ SYBR Green Supermix (Bio-Rad Laboratories, Inc, California, USA) and data visualized with the software Bio-Rad CFX Manager 3.1. The amplification protocol cycle was 95 °C for 3 min and 40 cycles at 95 °C for 15 s, 30 s at 54 °C and 71 °C for 30 s. The comparative CT method (ΔΔCT) was used for the relative quantification of gene expression values using the geometric mean of the expression of the control transcripts ubiquitin and 18S ribosomal RNA genes, and the plants inoculated with water (non-inoculated controls) as reference sample 71 . For each sample and target gene, two technical replicates were analysed. Statistical analysis. Data were analysed using GraphPad Prism v6.0 software (GraphPad Software, California, USA) and significant differences between chitosan treatments were determined using two-way analysis of variance (ANOVA) followed by Tukey's post hoc test with P = 0.05. Due to the large number of datasets, and to facilitate result viewing and discussion, results concerning MDA, catalase, carotenoids, anthocyanins, phenolic compounds and lignin are presented as the difference between the averages of PWN-and water-inoculated plants. Bars showing positive values represent an increase in PWN-inoculated plants relatively to water-inoculated plants, whereas negative results represent a decrease. Propagation of uncertainty was used to calculate the standard error of each value.