Brown planthopper honeydew-associated symbiotic microbes elicit momilactones in rice

ABSTRACT

Plants use many natural products to counter pests and diseases in nature. In rice, direct defense mechanisms include broad range of secondary metabolites, such as phenolamides (PA), diterpene phytoalexins, and flavonoid sakuranetin. Recently, accumulation of PAs in rice was shown to be under control of microbial symbionts in honeydew (HD), digestive waste from the rice brown planthopper (Nilaparvata lugens; BPH), but whether HD microbiota can also promote diterpene phytoalexins, momilactone A (MoA) and MoB, has not been reported. Here, we demonstrate that crude HD, but not a filtered one, induces MoA and MoB in rice, suggesting the involvement of BPH-HD endosymbionts. Consequently, microbial strains previously isolated from HD could promote MoA and MoB levels in wounded rice leaves, suggesting that rice indeed responds to BPH by cumulative chemical defense that involves both PA and diterpene phytoalexin pathways.

KEYWORDS:

• Honeydew-associated microorganisms
• momilactone A and B
• plant defense
• rice (Oryza sativa)
• rice brown planthopper (Nilaparvata lugens)

Symbiosis between insect pests and microbes has been known for a long time. For instance, Rickettsia sp. in Bemisia tabaci,¹ Sporothrix sp. in Bursaphelenchus xylophilus,² Leptographium procerum in Dendroctonus valens,³ Microsporidia in Harmonia axyridis,⁴ Streptomyces, γ-Proteobacteria and Amylostereum areolatum in Sirex noctilio,⁵ and Raffaelea lauricola in Xyleborus glabratus⁶ are just few examples to name with mutual relations. Close partnership between microbes and insects often negatively affects host plants, such as tomato yellow leaf curly virus (TYLCV) and tomato chlorosis virus (ToCV) in B. tabaci that simultaneously damage tomato plants.¹ On the other hand, plants evolved whole arsenal of natural products to counter pests and pathogens, and other competitive factors.⁷ For example, Arabidopsis thaliana accumulates camalexins and sesquiterpenes, and oat (Avena sativa) activates triterpenoid saponins as plant defense products against herbivores and/or pathogens.⁸ Rice (Oryza sativa) similarly quickly and universally accumulates phenolamides (PAs), p-coumaroylputrescine (CoP), and feruloylputrescine (FP) in response to attack by both chewing insects, i.e. feeding of the lawn armyworm (Spodoptera mauritia) and the rice skipper (Parnara guttata) larvae, and the attack of the sucking insect, the rice brown planthopper (Nilaparvata lugens, BPH).⁹ Momilactones (diterpenes) are another front defense chemicals in rice that are induced by fungal blast agent Magnaporthe oryzae,¹⁰ attack of sucking insect pests such as BPH,⁹ but not by mechanical wounding, suggesting their broad but not well-understood role in rice defense against multiple biotic stresses.

In particular, momilactone A (MoA) and MoB work as antimicrobial defense compounds,¹¹ and rice plants secrete MoA and MoB from their roots into the rhizosphere where these allelopathics inhibit the growth of competing weeds.¹² Variable levels of MoA and MoB were induced by different feeding guilds of insects,⁹ with BPH being the most effective inducer. Interestingly, BPH excretes sugar-rich digestive waste, honeydew (HD) on plants that supports microbial growth, and causes sooty appearances of rice leaves in paddy fields during heavy BPH seasons. It has been recently shown that HD contains symbiotic microbes^13,14 that are recognized by rice defense systems and trigger rice defense metabolism, as it was recently exemplified on PA class compounds in Wari et al.¹⁴ Here, we assess if HD-associated microbes could also affect momilactone levels that would explain their differential accumulation in response to BPH, while remaining unchanged in case of mechanical wounding.

Following protocols developed in our previous study,¹⁴ we first confirmed the accumulation of MoA and MoB in response to BPH feeding, using analytical methods for MoA and MoB described in Alamgir et al.⁹ As expected, BPHs (n = 10) attached to last fully expanded young leaf of 6-week-old rice seedlings, enclosed in ventilated cages for 4 d, significantly elevated MoA and MoB levels in the leaves (Figure 1A). Next, crude HD (2 μL) was gently rubbed on the leaf surface of the same-age leaves as before which clearly elicited MoA and MoB levels at 72 h, as shown in Figure 1B. We then used size-fractionated and heat-treated HD to characterize dominant elicitor that is required for momilactone induction. As elicitor activity was removable by 0.22 µm filter, and partially heat tolerant, dominant elicitor could be narrowed down to yet unknown component associated with microbial cells (Figure 1B). We then used our previously isolated symbionts from HD and test them for MoA and MoB induction with young rice leaves.¹⁴ Pattern wheel wounding on each side of leaf lamina was introduced prior to rubbing of microbes suspended in 15% sucrose (2 μL) to facilitate their easier penetration into rice cells. Previously shown as strongest inducer of PAs,¹⁴ symbiont 4–24 (Serratia marcescens subsp. marcescens) applied in 15% sugar also significantly induced MoA and MoB levels compared to sucrose-treated (mock) leaves. These levels were comparable to crude HD treatment on wounded leaves. As before,¹⁴ strain 2–08 was much less effective in inducing MoA and MoB levels (Figure 1C). Corroborating the importance of BPH HD-associated symbionts for elicitation of plant defense, our data indicate that, in addition to PAs and volatile organic compounds,¹⁴ rice cells also activate diterpene-based defense upon receiving HD information encoded by its inhabitants, facultative and/or obligate endosymbionts from BPH.

Figure 1. Accumulation of momilactones A and B in rice leaves. (A) Metabolite content measured in rice leaves by LC-MS/MS after 96 h infestation with BPH. Asterisks indicate statistically significant differences between treatment and control determined by Student’s t-test (*P < .05). n = 5–8; error bars = SEM; FM, fresh mass. (B) Momilactone A and B contents in rice leaves treated with raw or processed HD (filtered, 0.22 μM; or heat treated, 100°C, 20 min) determined by LC-MS/MS after 72 h treatment. Different letters show statistically significant differences between treatments by ANOVA (P < .05; Tukey HSD test). n = 3; error bars = SEM; FM, fresh mass; (C) Momilactone A and B contents in leaves after wounding with a serrated pattern wheel and applying microbial isolates from HD¹⁴ suspended in 15% (w/v) sucrose adjusted to OD₆₀₀ = 0.2. Metabolite levels were determined every 24 h after treatment by LC-MS/MS. Sucrose and HD were used as negative and positive controls, respectively. Different letters show statistically significant differences between treatments at the same time point determined by ANOVA (P < .05; Tukey HSD test). n = 4; error bars = SEM; W+ Suc, wounding with 15% sucrose; W+ Suc+2–08, wounding with isolate 2–08 suspended in 15% sucrose; W+ Suc+4–24, wounding with isolate 4–24 suspended in 15% sucrose; W+ HD, wounding with raw BPH honeydew; W, wounding; Suc, sucrose; HD, honeydew; FM, fresh mass.

In rice, multiple terpene synthase genes are induced by environmental factors that result in accumulation of terpenoid phytoalexins, momilactones, and phytocassanes.¹⁵ A number of transcription factors (TFs) have already been characterized as regulators of biosynthesis of rice terpenoids, and these factors are then likely to connect to the perception of BPH symbionts in rice.¹⁶ As sucking insect attack is often described as bearing pathogen signatures, it may suggest that due to a lack of mechanical damage as leading signal of attack, plants adopted less typical but reliable signals from the enemy to detect them. It is also possible to speculate that some of the antimicrobial signaling pathways became newly cross-linked to those for anti-herbivore defenses, or that plants have adopted broadly toxic chemicals as multifaceted unified weapons against variable biotic stress factors. It will be interesting to examine further plant defense pathways and crosstalk between pathogen (salicylic acid) and herbivore (jasmonate) signaling in plants that is usually considered negative in case of chewing insects but it can be perceived as more positive in case of sucking insect attack.

Author Contributions

D.W. designed and conducted experiments, analyzed data, wrote paper. K.M.A. designed and conducted experiments. K.M., Y.H., and H.N. conducted and assisted with experiments; T.S. conducted experiments, wrote paper. A.T. conducted experiments, wrote paper. I.G. designed, supervised and conducted experiments, and wrote paper.

Discloser of Potential Conflict of Interest

The authors report no conflict of interest.

Acknowledgments

We thank Dr. Masaya Matsumura (Kyushu Okinawa Agricultural Research Center, NARO) for providing field-collected BPHs; Ms. Yoshiko Fujitani for technical assistance during microbe isolations; Dr. Kazunori Okada (Tokyo University) for providing purified standards of momilactone A and B.

Additional information

Funding

This work was supported in part by Grants-in-Aid for Scientific Research [No. 24570026 and 16K08143 to I.G., No. 24780334 to T.S.], and by Japan Advanced Plant Science Research Network (LC-MS/MS instrumentation).