Plant, Cell and Environment (2017) 40,2663–2677 doi: 10.1111/pce.13015 Original Article Solanum dulcamara’s response to eggs of an insect herbivore comprises ovicidal hydrogen peroxide production Daniel Geuss , Sandra Stelzer, Tobias Lortzing & Anke Steppuhn Molecular Ecology, Dahlem Centre of Plant Sciences, Institute of Biology, Freie Universität Berlin, Haderslebener Strasse 9, 12163 Berlin, Germany ABSTRACT (Beyaert et al. 2012; Pashalidou et al. 2012; Bandoly et al. 2015; Bandoly et al. 2016; Austel et al.2016).Overall,plants Plants can respond to insect oviposition, but little is known likely evolved to perceive and respond to the oviposition of her- about which responses directly target the insect eggs and how. bivorous insects to prevent feeding damage by the larvae hatch- Here, we reveal a mechanism by which the bittersweet night- ing from these eggs (Hilker & Meiners 2006). However, our shade Solanum dulcamara kills the eggs of a generalist noctuid knowledge about the nature of most plant responses to insect herbivore. The plant responded at the site of oviposition by oviposition that directly affects the insect eggs as well as the Spodoptera exigua with formation of neoplasms and chlorotic mechanisms by which plants can kill insect eggs is still restricted. tissue, accumulation of reactive oxygen species and induction One of the best characterized plant responses that directly of defence genes and proteins. Transcriptome analysis revealed reduce egg survival is the release of ovicidal benzyl benzoate that these responses were reflected in the transcriptional into watery lesions at the oviposition sites of the planthopper reprogramming of the egg-laden leaf. The plant-mediated egg Sogatella furcifera on rice plants (Seino et al. 1996; Suzuki mortality on S. dulcamara was not present on a genotype lack- et al. 1996). One major and several minor quantitative trait loci ing chlorotic leaf tissue at the oviposition sites on which the associated with watery lesions and egg mortality are mapped eggs are exposed to less hydrogen peroxide. As exposure to hy- (Yamasaki et al.2003;Yanget al. 2014). Rice genotypes that drogen peroxide increased egg mortality, while catalase supple- do or do not exhibit this ovicidal response show global devia- mentation prevented the plants from killing the eggs, our tions in gene expression in response to S. furcifera infestation, results suggest that reactive oxygen species formation directly but the biochemical pathways underlying this response and acts as an ovicidal plant response of S. dulcamara. their regulation remain to be determined. Other plants respond to insect oviposition with growth re- Key-words: egg-killing; herbivory; hypersensitive response; in- sponses that physically affect the eggs. Egg deposition by the – duced plant defence; microarray; phytohormones; plant insect leaf beetle Pyrrhalta viburni on stems of Viburnum species interactions; ROS. elicits tissue production at the oviposition site that displaces the egg cap, partially crushes the eggs and encases egg masses, thereby reducing egg survival (Desurmont & Weston 2011). Pea and physalis plants also produce plant tissues in form of neoplasms underneath the eggs of pea weevils or of the lepi- INTRODUCTION dopteran herbivore Heliothis subflexa, respectively (Doss Plants deploy various defences against insect herbivory, and et al. 2000; Petzold-Maxwell et al. 2011). Neoplasm formation many are inducible by herbivore attack. Inducible plant de- is characterized by limited non-meristematic growth and is as- fences are elicited by signals associated with the damage that sociated with reduced pea weevil infestations of pea plants fl the feeding herbivores inflict and by signals of the herbivore it- and reduced egg hatching rates of H. sub exa on physalis self such as components of their oral secretions (Bonaventure plants, but the mechanisms for these effects remain unknown. 2012). Moreover, several plant species respond already to the It has been suggested that neoplasm formation facilitates egg oviposition of herbivorous insects (Hilker & Fatouros 2015). removal from the plant by physical ablation or predation (Doss Whereas some of the plant responses to oviposition may result et al. 2000; Petzold-Maxwell et al.2011). in dropping, crushing, desiccation or intoxication of the insect At the sites of insect oviposition, several plant species of the eggs, others can repel herbivorous insects or attract egg Brassicaceae and the Solanaceae exhibit chlorotic or necrotic predators and parasitoids that kill the eggs (Hilker & Fatouros responses, which are paralleling a hypersensitive response 2015). In addition to plant responses that directly or indirectly (HR) that is commonly described for pathogen infections of reduce egg survival, previous insect oviposition can affect de- plant tissue (Shapiro & DeVay 1987; Balbyshev & Lorenzen velopment of the feeding larvae that hatch from these eggs 1997; Petzold-Maxwell et al. 2011; Fatouros et al. 2014; Bittner et al. 2017). In most of these plant species, this response is asso- Correspondence: Anke Steppuhn. e-mail: [email protected] ciated with reduced egg survival, although the mechanism © 2017 The Authors Plant, Cell & Environment Published by John Wiley & Sons Ltd 2663 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 2664 D. Geuss et al. remains unclear. Shapiro and DeVay (1987), who initially de- Although the response of A. thaliana to oviposition is not af- scribed HR-like necrosis at oviposition sites on Brassica nigra, fected on JA-deficient mutants, the transcriptional analyses suggested a lethal desiccation of the eggs as cause of humidity- of its responses also revealed the regulation of JA- dependent egg mortality on necrotic tissues. In accordance responsive genes (Little et al. 2007). with this hypothesis, the mortality of sawfly eggs was suggested In this study, we investigated the responses of Solanum to increase on pine foliage that was either desiccated (Codella dulcamara, a wild relative of potato and tomato, to oviposition. & Raffa 2002) or showed HR-like responses (Bittner et al. We discovered that the plant kills the eggs of the noctuid moth, 2017). However, egg desiccation does not occur on potato Spodoptera exigua, and asked for the mechanism underlying leaves exhibiting HR-like necrosis at the oviposition sites of the negative effect on this generalist herbivore. Therefore, we the Colorado potato beetle, which is generally not affected in (1) characterized the plant’s response to the moth’s oviposition its egg hatching rate (Balbyshev & Lorenzen 1997). Instead, on physiological and transcriptional levels and (2) we exam- the HR-like necrosis reduces egg attachment to the leaf surface ined which of the responsive plant traits is functionally linked and the reduced larval infestation of plants exhibiting the HR- to the egg killing. like response in a field trial is attributed to predation of dropped eggs by ground predators. The response of physalis plants to H. subflexa eggs also involves necrosis, and even chlo- MATERIALS AND METHODS rotic neoplasms are reported, but the contribution of each of Plant and insects these responses to the reduction of egg hatching rates on responding plants and to the effect that more eggs vanish on We used S. dulcamara L. (Solanaceae) plants originating from these plants under field conditions remains elusive (Doss different populations in the vicinity of Berlin (Erkner: 0 ″ 0 ″ 0 ″ et al. 2000; Petzold-Maxwell et al.2011). 52°41 88.8 N; 13°77 34.1 E, Grunewald: 52°27 44.4 N; 0 ″ 0 ″ 0 ″ The HR-like response to eggs of pierid butterflies on 13°11 24.6 E, Mehrow: 52°34 06.4 N; 13°38 04.0 Eand 0 ″ 0 ″ Arabidopsis thaliana leaves is further associated with the ac- Siethen 52°16 53.7 N; 13°11 18.7 E) and from the 0 ″ 0 ″ cumulation of reactive oxygen species (ROS) such as hydro- Netherlands (Friesland: 52°58 36.2 N5°3059.4 E). Except for one experiment, plants were grown from stem cuttings of gen peroxide (H2O2), increased levels of the phytohormone salicylic acid (SA) and the induction of SA-responsive genes 6 to 7-week-old plants. Stem segments that included two nodes beneath the eggs (Little et al. 2007; Bruessow et al. 2010; were planted into 0.75 L pots with one node within and one Gouhier-Darimont et al. 2013). Therefore, B. nigra plants above the soil. The microarray experiment was performed with fl that exhibit the HR-like response to oviposition by pierid plants grown from seeds and thus re ected the transcriptomic butterflies show enhanced expression of the SA-responsive response of several genotypes from three S. dulcamara marker gene PR1 (pathogenesis-related protein 1; Fatouros populations. The seeds were incubated in darkness on steril- – et al. 2014). Comprehensive analyses of transcriptome regu- ized wet sand (2 4 mm grain size) in plastic containers lation reveal large overlaps in A. thaliana’s response to (20 × 20 × 6.5 cm; Gerda, Schwelm, Germany) sealed with fi Pieris brassicae eggs and infection by the bacterial pathogen cling lm at 4 °C. After 12 d, the containers were transferred Pseudomonas syringae, particularly in defence-related and to the greenhouse, and 10 d later, individual seedlings were ® stress-related genes (Little et al. 2007; Gouhier-Darimont transferred to 0.75 L pots with soil. The soil (Einheits Erde , fi et al. 2013), suggesting similarities in the HRs elicited by in- type: Pro Substrat Classic, Sinntal-Jossa, Germany) of plants sect oviposition and biotrophic pathogens. Yet, Pieris eggs for all experiments was covered with about 1 cm of sand – are not affected by this response of A. thaliana (Gouhier- (2 4 mm grain size) to prevent fungus gnats infestation. The Darimont et al. 2013), which has been mainly evaluated for plants were grown in the greenhouse with a 16/8 h its effects on the feeding larvae hatching from the eggs light/dark cycle and a photon irradiance between 190 and μ À2 À1 (Bruessow et al.