Endophyte Infection and Methyl Jasmonate Treatment Increased the Resistance of Achnatherum sibiricum to Insect Herbivores Independently

Alkaloids are usually thought to be responsible for protecting endophyte-infected (EI) grasses from their herbivores. For EI grasses that produce few alkaloids, can endophyte infection enhance their resistance to herbivores? Related studies are limited. In the Inner Mongolian steppe, Achnatherum sibiricum is highly infected by Epichloë endophytes, but produces few alkaloids. Locusts are the common insect herbivores of grasses. In this study, A. sibiricum was used as plant material. Methyl jasmonate (MJ, when applied exogenously, can induce responses similar to herbivore damage) treatment was performed. The effects of endophyte infection and MJ treatment on the resistance of A. sibiricum to Locusta migratoria were studied. We found that locusts preferred EF (endophyte-free) plants to EI plants in both choice and no-choice feeding experiments. Endophyte infection enhanced the resistance of A. sibiricum to locusts. Endophyte infection decreased soluble sugar concentrations, while it increased the total phenolic content and phenylalanine ammonia lyase (PAL) activity, which may contribute to the resistance of A. sibiricum to locusts. There was an interaction effect between MJ treatment and endophyte infection on the growth of the host. MJ treatment was a negative regulator of the plant growth-promoting effects of endophyte infection. There was no interaction effect between MJ treatment and endophyte infection on the defense characteristics of the host. In groups not exposed to locusts, MJ treatment and endophyte infection had a similar effect in decreasing the soluble sugar content, while increasing the total phenolic content and the PAL activity. In groups exposed to locusts, the effect of MJ treatment on the above characteristics disappeared, while the effect of endophyte infection became more obvious. All of these results suggest that even for endophytes producing few alkaloids, they could still increase the resistance of native grasses to insect herbivores. Furthermore, endophyte infection might mediate the defense responses of the host, independent of jasmonic acid (JA) pathways.


Introduction
Epichloë endophyte species are characterized by their endophytic lifestyles in aerial parts of cool-season grasses [1]. Up to 30% of cool-season grasses have been reported to be associated with In the no-choice feeding experiment, the amount of mass consumed by L. migratoria was significantly affected by endophyte infection (F = 117.35, p < 0.01), MJ treatment (F = 8.688, p < 0.01), and the interaction between endophyte infection and MJ treatment (F = 8.348, p < 0.01). Endophyte infection significantly reduced the leaf consumption of the host plants, and this effect was not affected by MJ treatment. For EF plants, the amount of biomass consumed by the locusts decreased with MJ treatment (F = 9.560, p < 0.01), but a significant reduction occurred only in the MJH treatment, as shown in Figure 1B.

Growth and Biomass
Plant height was only significantly affected by MJ treatment, as shown in Table 1, and MJ treatment reduced plant height (CK = 31.16 cm; MJL = 27.98 cm; MJH = 25.64 cm). Leaf number, tiller number, and shoot biomass were significantly affected by the interaction between MJ treatment and endophyte infection, as shown in Table 1. MJ treatment inhibited the vegetative growth of both the EI and EF plants, but the inhibition degree was stronger for the EI than the EF plants. MJL significantly reduced the leaf number and shoot biomass of the EI plants, but a significant reduction in the leaf number and shoot biomass of the EF plants occurred only with the MJH treatment. MJL and MJH resulted in a 15% and 23% reduction in tiller number for the EF plants, respectively, and a 24% and 40% reduction in tiller number for the EI plants, respectively, as shown in Figure 2. In the no-choice feeding experiment, the amount of mass consumed by L. migratoria was significantly affected by endophyte infection (F = 117.35, p < 0.01), MJ treatment (F = 8.688, p < 0.01), and the interaction between endophyte infection and MJ treatment (F = 8.348, p < 0.01). Endophyte infection significantly reduced the leaf consumption of the host plants, and this effect was not affected by MJ treatment. For EF plants, the amount of biomass consumed by the locusts decreased with MJ treatment (F = 9.560, p < 0.01), but a significant reduction occurred only in the MJH treatment, as shown in Figure 1B.

Growth and Biomass
Plant height was only significantly affected by MJ treatment, as shown in Table 1, and MJ treatment reduced plant height (CK = 31.16 cm; MJL = 27.98 cm; MJH = 25.64 cm). Leaf number, tiller number, and shoot biomass were significantly affected by the interaction between MJ treatment and endophyte infection, as shown in Table 1. MJ treatment inhibited the vegetative growth of both the EI and EF plants, but the inhibition degree was stronger for the EI than the EF plants. MJL significantly reduced the leaf number and shoot biomass of the EI plants, but a significant reduction in the leaf number and shoot biomass of the EF plants occurred only with the MJH treatment. MJL and MJH resulted in a 15% and 23% reduction in tiller number for the EF plants, respectively, and a 24% and 40% reduction in tiller number for the EI plants, respectively, as shown in Figure 2.

Physiological Variables
There was no significant interaction effect between endophyte infection and MJ treatment on soluble sugar, total phenolics, phenylalanine ammonia lyase (PAL), and polyphenol oxidase (PPO), as shown in Table 2. In the no-feeding groups, soluble sugar content was significantly reduced by both endophyte infection and MJ treatment, while PAL was significantly enhanced by both endophyte infection and MJ treatment. Total phenolics and PPO were enhanced by endophyte infection and MJ treatment, separately, as shown in Figure 3. After locust feeding, the above physiological variables were significantly affected only by endophyte infection, but not by MJ treatment. Furthermore, endophyte infection reduced the plants' soluble sugar content, while it enhanced PPO and PAL activities, as shown in Figure 4.

Physiological Variables
There was no significant interaction effect between endophyte infection and MJ treatment on soluble sugar, total phenolics, phenylalanine ammonia lyase (PAL), and polyphenol oxidase (PPO), as shown in Table 2. In the no-feeding groups, soluble sugar content was significantly reduced by both endophyte infection and MJ treatment, while PAL was significantly enhanced by both endophyte infection and MJ treatment. Total phenolics and PPO were enhanced by endophyte infection and MJ treatment, separately, as shown in Figure 3. After locust feeding, the above physiological variables were significantly affected only by endophyte infection, but not by MJ treatment. Furthermore, endophyte infection reduced the plants' soluble sugar content, while it enhanced PPO and PAL activities, as shown in Figure 4.

Discussion
Epichloë endophytes can protect the host grass from insect herbivores, and endophyte-associated alkaloids are thought to be primarily responsible for feeding deterrence [14,[39][40][41][42][43][44]. In our experiments, only peramine was detected in one sample, and its concentration is not high enough to deter insects [17,18]. However, we found that infected A. sibiricum had significant resistance to locusts in both the choice and no-choice tests, as shown in Figure 1, which suggests that aside from alkaloid defense, additional mechanisms are likely to be involved in endophyte-associated insect deterrence.
Phytophagous insects depend on plants for their nutrients, so that their growth and development are affected directly by the makeup of the food source. Primary metabolites such as

Discussion
Epichloë endophytes can protect the host grass from insect herbivores, and endophyte-associated alkaloids are thought to be primarily responsible for feeding deterrence [14,[39][40][41][42][43][44]. In our experiments, only peramine was detected in one sample, and its concentration is not high enough to deter insects [17,18]. However, we found that infected A. sibiricum had significant resistance to locusts in both the choice and no-choice tests, as shown in Figure 1, which suggests that aside from alkaloid defense, additional mechanisms are likely to be involved in endophyte-associated insect deterrence.
Phytophagous insects depend on plants for their nutrients, so that their growth and development are affected directly by the makeup of the food source. Primary metabolites such as  (PAL, B), and polyphenol oxidase (PPO, C) of endophyte-infected (EI) or endophyte-free (EF) Achnatherum sibiricum with locust feeding. Different lower-case letters denote means that are significantly different (p < 0.05).

Discussion
Epichloë endophytes can protect the host grass from insect herbivores, and endophyte-associated alkaloids are thought to be primarily responsible for feeding deterrence [14,[39][40][41][42][43][44]. In our experiments, only peramine was detected in one sample, and its concentration is not high enough to deter insects [17,18]. However, we found that infected A. sibiricum had significant resistance to locusts in both the choice and no-choice tests, as shown in Figure 1, which suggests that aside from alkaloid defense, additional mechanisms are likely to be involved in endophyte-associated insect deterrence.
Phytophagous insects depend on plants for their nutrients, so that their growth and development are affected directly by the makeup of the food source. Primary metabolites such as carbohydrates are important nutrients for animal growth [45]. Secondary metabolites such as plant phenols play an important role in the resistance of the host plant to herbivores [46,47]. In the present study, we found that endophyte infection significantly increased the total phenolic content in the host plants, as shown in Table 2 and Figure 3. Similar results have been reported in perennial ryegrass, Lolium perenne [48,49] and tall fescue [50]. Moreover, PAL and PPO are enzymes that are involved in phenol oxidation, and they are correlated with plant defense mechanisms [51]. Higher levels of PAL [52,53] and PPO activity [54][55][56] have been reported to be associated with the resistance of plants to insects. Here, the higher phenol concentration, PAL and PPO activity, combined with the lower soluble sugar concentration of the EI leaves may be responsible for the higher locust deterrence of A. sibiricum.
JA is a ubiquitous wound hormone. Repeated wounding of mature leaves induces a rapid increase in endogenous JA concentrations, thus resulting in reduced leaf growth [57]. Exogenous applications of JA or MJ can mimic the temporal and quantitative characteristics of endogenous JA responses [58][59][60][61]. In the present study, MJ treatment inhibited the vegetative growth of both the EI and EF plants, as shown in Table 1 and Figure 2. This is in agreement with other studies that found a reduction in both root growth [62,63] and shoot growth [64][65][66] under exogenous JA treatments. On the contrary, jasmonates might act as negative regulators of plant primary metabolic products, such as sugars [23,[67][68][69][70][71] and as positive regulators of plant secondary metabolic products, such as phenolic compounds [72]. In the present study, we also found that MJ treatment decreased soluble sugar content, while it increased the PPO and PAL activities of the treated plants, as shown in Table 2 and Figure 3.
A JA-dependent pathway is mainly induced by chewing insects [73][74][75]. Additionally, these hormones also regulate the interactions of plants with beneficial organisms, such as the symbiosis with arbuscular mycorrhizal fungi (AMF) [76,77] and root endophytes [78]. For example, mycorrhization had a positive effect on floral traits, but its benefits were lost with JA application. JA signaling and JA biosynthesis mutants caused a significant reduction in Piriformospora indica colonization [78]. In the present study, MJ treatment reduced the growth-promoting effects of endophyte infection. Similar results have also been found regarding the effects of the root endophyte P. indica on Nicotiana attenuata [79]. Recently, P. indica has been found to promote the growth of JA-insensitive mutants more strongly than wild-type (WT) rice [80]. All these results suggest that JA signaling is a negative regulator of the plant growth-promoting effects induced by endophytes. It has been suggested that auxin and gibberellic acid (GA) released by endophytic fungi may explain growth enhancements in host plants [81][82][83]. The negative effect of JA on the growth-promoting effects of endophyte infection might be mediated via antagonistic interactions with the auxin/GA signaling pathway [84][85][86].
In the present study, MJ treatment and endophyte infection had a similar effect in decreasing plants' soluble sugar contents, while increasing the total phenolic content and PAL activity in the no-feeding groups, but there were no interactions between MJ treatment and endophyte infection, as shown in Table 2 and Figure 3. In a previous study, JA inhibitors were found to suppress both JA and volatile oil production, but fungal inoculation could still induce volatile oils [87]. All of these results suggest that endophyte infection might mediate the defense responses of the host via different pathways, such as H 2 O 2 -dependent pathways [87]. In the feeding groups, the effects of MJ treatment on the above characteristics disappeared, while the effects of endophyte infection became more obvious. The reason for this might be that endophyte pre-infection and late locust feeding led to a much stronger induction of the plants' defense responses, as similar results have also been found in AMF-plant systems [77]. Certainly, interactions among endophytes, endogenous hormone levels, and the host defense traits depend on the specific species, strain, or very likely, the plant-endophyte combination, and further research using more combinations of plants and endophytes is still needed.

Conclusions
In this study, we found that endophytes significantly improved the resistance of the host to locusts, although the host produced few alkaloids. Here, the lower soluble sugar concentration, higher total phenolics concentration, and higher PAL activity contributed to the higher resistance of the host plants to the locusts. MJ treatment inhibited the vegetative growth of both the EI and EF plants. MJ treatment is a negative regulator of the plant growth-promoting effects of endophyte infection. However, regarding plants' defense characteristics, the effects of endophyte infection may be independent of JA pathways.

Plant Material and Treatment
Seeds of A. sibiricum were collected from the National Hulunber Grassland Ecosystem Observation and Research Station of China. After examining the endophyte status of both seeds and seedlings following staining with aniline blue [88], we found that the endophyte infection rate was 100% (infection rate = the number of infected seedlings (or seeds)/total seedlings (or seeds) [88]). Endophyte-free (EF) seeds were obtained by high temperature treatment [89]. The seeds were planted in white plastic pots (20 cm in diameter and 20 cm in depth) filled with vermiculite. After 30 days' cultivation, the endophyte infection status of each plant was further inspected, and 10 plants of approximately equal size were maintained in each pot. Meantime, MJ treatments were processed. The experiment had a completely randomized 2 × 3 × 2 factorial design, with endophyte infection status (EI and EF), MJ treatment (no MJ control, CK; low MJ concentration, MJL; and high MJ concentration, MJH), and locust feeding (no-feeding and feeding) as the variables. There were six replicates per treatment group for the choice feeding experiment, and five replicates for the no-choice feeding experiment. The experimental plants were watered and fertilized with modified Hoagland nutrient solution as needed. The composition of the nutrient solution was (µM) [

Locusta Migratoria
Locust eggs were purchased from a local pet shop. Fourth instar nymphs were used in the experiments.

Methyl Jasmonate (MJ) Treatment
MJ (purity: 98% HPLC-grade) was bought from 3B Pharmachem (Wuhan, China) International Co., Ltd. MJ was firstly dissolved in absolute ethyl alcohol, and then dispersed in water to the desired concentration. MJ was applied with a backpack sprayer to plants until run-off; if necessary, the surrounding leaflets were shielded with a piece of paper. Control plants were sprayed with water that contained alcohol at the same concentration used for the MJ treatments. Neighboring plants were shielded from the spray with a large sheet of plastic. Plants were divided randomly into three treatment groups: control (no MJ), low MJ (25 mg/L, 0.025 mg of MJ per plant), and high MJ (400 mg/L, 0.4 mg of MJ per plant).

Choice Feeding Experiment
After starvation for 24 h, the fourth-instar locust nymphs were transferred to transparent plastic containers (19 cm height and 8 cm diameter). Locust nymphs were provided with 0.5 g leaf blades per pot per treatment as food. These leaf blades were cut 24 h after MJ treatment, and were transferred to the above containers simultaneously with the locusts. There were 12 plastic containers in total. Eight locusts were added to each of the six containers. Another six containers were used as a control to calculate the reduced leaf quality due to evaporation. After 1 h of exposure to herbivory, plants were harvested to record the fresh biomass.

No-Choice Feeding Experiment
Twenty-four hours after MJ treatment, we equipped every pot with a steel frame (45 cm height and 20 cm diameter), which was covered with a nylon stocking. Three fourth-instar locust nymphs, starved for 24 h, were introduced into half of these resulting cages. The feeding exposure lasted 3 h. Before and after the feeding, we immediately measured the length (L) and maximum width (W) of each leaf with a ruler. We estimated leaf area (S) according to the following equation: S = K (L × W), where K is 0.9, based on our previous study using A. sibiricum plants of similar sizes (unpublished data). At the same time, 10 fully expanded leaves per no-feeding pot were chosen to measure the area and were weighed separately for determination of specific leaf area (SLA). Leaf mass consumed was obtained from consumed leaf area divided by SLA.

Growth and Biomass
Growth variables were measured only in the no-feeding plants in the no-choice experiment. Twenty days after MJ treatment, regular measurement of tiller number, leaf number, and shoot height of the longest tiller were made on all ramets. Subsequently, the shoot and the root were harvested separately. The harvested material was oven-dried at 80 • C for biomass measurement [91].

Physiological Variables
Physiological variables were sampled before and 24 h after no-choice feeding, separately. The soluble sugar concentration was analyzed using the colorimetric method [92] [93]. The total phenolic concentration was determined according to Malinowski et al. [50]. Polyphenol oxidase (PPO) activities were determined spectrophotometrically according to the methods described in Thaler et al. [94], and phenylalanine ammonia lyase (PAL) was determined using modified procedures described by Assis et al. [95].

Statistical Analysis
All statistical analyses were performed with SPSS software (Version 21.0, SPSS, Chicago, IL, USA, 2012). Endophyte status and MJ treatment were analyzed using two-way analysis of variance (ANOVA). The differences between the means among different factors were compared using Duncan's multiple-range tests, and significance was set at p < 0.05.