Biological processes involved in plant responses to insect eggs
We first explored whether the investigated plant species show conformable transcriptional reprogramming in response to insect eggs.
Taking into account all GS included in the analysis for all plant–insect combinations tested and significantly regulated in at least one of them, we found down-regulation of 649 and up-regulation of 969 GS in response to egg deposition (Supplementary Table S1). Of these, we identified an overlap of 52 down-regulated and 310 up-regulated GS with conformable regulation across the different plant–insect combinations (Fig. 2a, Supplementary Table S2). This indicates that plant species with very different lifestyles share up to 32% of transcriptional regulation in biological processes after egg deposition from different insect species.
Comparison of GAGE analyses of the transcriptional up-regulation in four different plant species in response to (a) insect eggs (E vs. C) and (b) larval feeding (F vs. C). The species are given below†. The heatmaps depict false discovery rate-adjusted p-values (FDR) according to the colour key of up-regulated gene sets (GS): I. GS significantly enriched in at least one plant species, II. Heatmap sections of the conformably enriched GS in at least three out of four (a) and four out of five (b) plant–insect interactions. GS in II were re-ordered according to the biological function. For detailed descriptions of the GS see Supplementary Table S2 (E vs. C up and F vs C up). AA: amino acid; ABA: abscisic acid; C & CD: cytokinesis & cell differentiation; CH: carbohydrates; Develop., …: development, morphogenesis & reproduction; ET: ethylene; JA: jasmonic acid; Local.: localisation; Ox. burst: oxidative burst; SA: salicylic acid; SAR: systemic acquired resistance; Signal. & transduct.: signalling and transduction; UPR: unfolded protein response. †At: Arabidopsis thaliana-Pieris brassicae; Na-M: Nicotiana attenuata-Manduca sexta; Na-S: N. attenuata-Spodoptera exigua; Sd: Solanum dulcamara-S. exigua; Um-1 h/-6 h/-24 h: Ulmus minor-Xanthogaleruca luteola after 1/6/24 h of egg deposition or larval feeding. For detailed experimental setup description see Fig. 1.
The 52 down-regulated GS represent 8% of conformable regulation across the plant–insect combinations and were mostly related to regulation of gene expression and some developmental, morphological and cell cycle processes (Supplementary Table S2).
Of the 310 conformably up-regulated GS, 34% belong to generic stress responses, roughly equally distributed between abiotic and biotic stress responses (Fig. 2a, Supplementary Table S2). The majority of GS associated with “biotic stress” were related to plant immune responses, which comprise hypersensitive response (HR)-like responses and cell death, accumulation of reactive oxygen species (ROS), systemic acquired resistance (SAR), endoplasmic reticulum stress and unfolded protein responses (UPR). Insect egg depositions can cause obvious phenotypic leaf tissue modifications, such as necrosis/chlorosis and neoplasm formation at the site of egg deposition. These egg-induced leaf modifications occur in several plant species, including A. thaliana, Brassica nigra, S. dulcamara and P. sylvestris, thus resembling a HR-like symptom, which is linked to the accumulation of ROS11,12,13,25,26,27. This egg-induced change in leaf traits might result in egg desiccation or detachment of eggs from leaves. Some plant species may rely on ROS signalling to initiate formation of chlorotic or necrotic leaf tissue at the site of egg deposition10; others use extensive ROS accumulation to directly kill the eggs12. An oxidative burst is an essential signalling component for the formation of necrotic lesions, which are typical for HR-like responses. To our surprise, plant species like tobacco and elm, which do not show obvious HR-like symptoms in response to eggs, display transcriptional activation of innate immune responses similar to that of A. thaliana and S. dulcamara (Fig. 2a).
Further conformably up-regulated GS after egg deposition included, among others, GS involved in small and macromolecule metabolism, metabolism of organic acids, amines, cyclic carbohydrates and phenylpropanoids, and GS related to phytohormones (Fig. 2a).
In response to pathogens, ROS synergistically amplify the SA signal to induce HR-like symptoms and the expression of PATHOGENESIS-RELATED (PR) defence genes such as PR1e.g.28,29,30. PR genes are also more strongly expressed in response to eggs in several plant species12,15,31,32. We found strong conformable up-regulation of SA-related GS in response to eggs (Fig. 3a, Hormones (H); for abbreviations see Supplementary Table S3). This effect is quite weak only in S. dulcamara, although this plant species accumulates SA in response to insect eggs, as has been shown in phytohormone measurements by Geuss et al.12. Hence, the ROS- and SA-mediated induction of immune responses and PR gene expression in response to eggs is conserved amongst different plant species and both might contribute to direct plant defence against insect eggs. Interestingly, JA, abscisic acid (ABA) and ethylene (ET) signalling are also part of the conformable response to insect eggs (Fig. 2a and Fig. 3a).
Comparison of GAGE analyses of phytohormones from five different plant–insect combinations† depicting plant responses to (a) insect eggs (E vs. C) and (b) larval feeding (F vs. C), eggs with subsequent feeding (EF vs. C) and the alterations in plant responses to feeding by prior egg deposition (EF vs. F). The heatmap depicts false discovery rate-adjusted p-values (FDR) according to the colour key for up- or down-regulated gene sets (GS). Black boxes indicate GS which could not be assigned to the plant species or for which enrichment scores were not calculated due to a lack of data (E vs. C; Na-M). For a detailed description of the phytohormone-related GS 1H-71H (H) see Supplementary Table S3. ABA: abscisic acid; AUX: auxin; CK: cytokinin; ET: ethylene; GA: gibberellic acid; JA: jasmonic acid; SA: salicylic acid. †At: Arabidopsis thaliana-Pieris brassicae; Na-M: Nicotiana attenuata-Manduca sexta Na-S: N. attenuata-Spodoptera exigua; Sd: Solanum dulcamara-S. exigua; Um-1 h/-6 h/-24 h: Ulmus minor-Xanthogaleruca luteola after 1/6/24 h of egg deposition and larval feeding, NA: not annotated. For a detailed experimental setup description see Fig. 1.
Taken together, the different plant–insect combinations showed a considerable overlap in their transcriptomic responses to insect eggs, including up-regulation of GS related to generic stress responses and down-regulation of GS related to development and cell cycle processes. This would suggest a conserved plant response to insect egg depositions, regardless of whether the egg deposition is associated with leaf damage (as is the case for U. minor).
Biological processes involved in plant responses to insect feeding and their similarities to plant responses to insect eggs
Using the same methodology as for the E vs. C comparison, we searched for conformable plant transcriptional responses to feeding herbivores (F vs. C) across the plant–insect combinations we investigated. Then, we compared the conformable responses to feeding and to insect eggs with each other to identify a subset of GS that responds analogously in both treatments.
Larval feeding led to down-regulation of 972, and to up-regulation of 911 GS in at least one of the plant–insect combinations (Supplementary Table S1).
Of the down-regulated GS, 16% were conformably down-regulated (Supplementary Table S2). Similar to the response to eggs, the feeding-responsive down-regulated GS included especially those associated with regulation of gene expression by epigenetic and post-transcriptional modifications, developmental and morphological processes and cell cycle processes (Fig. 4a, green intersection, Supplementary Tables S2 and S4).
Venn diagrams with the number of gene sets (GS) that showed conformable (a) down-regulation and (b) up-regulation across the different plant–insect combinations when comparing the plant response to eggs (E vs. C, Fig. 2a II), to feeding (F vs. C, Fig. 2b II) and to eggs followed by feeding (EF vs. F, Fig. 5b). For detailed descriptions of uniquely or commonly enriched GS see Supplementary Table S4.
A considerable fraction of GS (28%) was conformably up-regulated in response to feeding (Fig. 2b, Supplementary Table S2). The vast majority (78%) of these GS overlapped with the conformable response to eggs (Fig. 4b, green intersection). The GS in this analogous response to eggs and to feeding included most stress- and plant immune response-related GS like ROS production, phytohormonal regulation and large parts of the metabolism-related GS, e.g. biosynthesis of aromatic compounds and phenylpropanoid metabolism (Supplementary Table S4), but lacked the GS related to nucleoside/-tide metabolism, which responded only to egg deposition.
As expected33, the conformable plant response to feeding includes many JA-related processes, accompanied by ABA and ET signalling. However, we also found a surprisingly consistent enrichment of GS related to immune responses, SA and ROS signalling in feeding-induced leaves (Figs. 2b and 3b).
In all of the plant species investigated here, JA-related responses dominated the plant response to feedinge.g.12,34. However, some studies found SA levels to be slightly enhanced after herbivory by P. brassicae in A. thaliana17 and by M. sexta and S. exigua in N. attenuata35,36, but not in S. dulcamara37. Elevated SA levels frequently antagonise JA-mediated plant defences against herbivory38,39; they are therefore usually considered to be beneficial for chewing herbivorese.g.40. However, activation of SA signalling is not always advantageous for the herbivore16,17,41. JA and SA are embedded in a complex phytohormonal signalling network which determines, as a whole, the metabolic outcome affecting biotic stressors like insects33. Subtle changes in SA levels may therefore fine-tune a JA-dominated response within this phytohormonal network and vice versa42.
Overall, the conformable feeding-induced transcriptional response observed in the different plant–insect combinations was remarkably similar to the conformable response to insect egg deposition. Developmental, morphogenesis and growth processes were down-regulated in response to eggs and feeding, indicating that metabolic resources might be shifted towards defence and stress reaction (Fig. 4a, Supplementary Table S4). The up-regulation of immune-related stress responses, phytohormonal regulation and secondary metabolism-related GS were almost identical in the conformable responses to egg deposition and to feeding (Fig. 2). The particularly large overlap in ROS-related stress responses and the involvement of multiple phytohormonal signalling pathways might indicate a more fundamental role of ROS signalling in plant responses to insect eggs beyond the formation of defensive HR-like symptoms. ROS are not only important as a second messenger during establishment of HR, but are closely connected with the hormonal signalling network and metabolic reprogramming after herbivore attack43,44.
Modification of plant transcriptional responses to larval feeding by prior egg deposition
A comparison of the transcriptomes of feeding-damaged plants with and without prior egg deposition (EF vs. F) revealed 84 down-regulated and 630 up-regulated GS in at least one of the plant–insect combinations (Supplementary Table S1).
We did not detect any conformably down-regulated GS (Supplementary Table S2), whereas 39 GS were conformably up-regulated across the plant–insect combinations (Fig. 5a, Supplementary Table S2). Almost all (36) of the latter GS were also found in the analogous responses to feeding and eggs (Fig. 4b, grey intersection, Supplementary Table S4). They account for a core set of 18% of the GS analogously regulated by eggs and by feeding in most of the plant species we tested. These GS indicate additive or synergistic effects when egg deposition precedes larval feeding. It includes mostly biotic stress and immune responses with regulation of cell death, but also hormonal responses, particularly the response to JA and phenylpropanoid biosynthesis (Fig. 5b, Supplementary Table S2).
Comparison of GAGE analyses of the transcriptional up-regulation in four different plant species comparing (a) the response to larval feeding with and without prior egg deposition (EF vs. F; species are given below†). The heatmap depicts false discovery rate-adjusted p-values (FDR) according to the colour key of all up-regulated gene sets (GS) significantly enriched in at least one plant species; (b) Heatmap-section of a) with conformably enriched GS in at least four out of five plant–insect combinations (EF vs. F) and FDR values of the same GS in comparisons between untreated controls and egg-deposited (E vs. C), feeding damaged (F vs. C) or egg deposited and feeding damaged (EF vs. C) plants. GS in b) were re-ordered according to their biological function. For detailed descriptions of the GS see Supplementary Table S2 (EF vs. F up). JA: jasmonic acid; Phenylprop.: phenylpropanoids; Second. metabol.: secondary metabolites; Signal. & transduct.: signalling and transduction. †At: Arabidopsis thaliana-Pieris brassicae; Na-M: Nicotiana attenuata-Manduca sexta Na-S: N. attenuata-Spodoptera exigua; Sd: Solanum dulcamara-S. exigua; Um-1 h/-6 h/-24 h: Ulmus minor-Xanthogaleruca luteola after 1/6/24 h of egg deposition and larval feeding, respectively, NA: not available. For detailed experimental setup description, see Fig. 1.
In summary, a considerable percentage of the activated GS involved in the analogous egg and feeding responses is further enhanced when plants experience both stimuli in succession. This suggests a conserved herbivore alarm response that is initiated by insect egg deposition and affects the transcriptional response induced by feeding in an additive or synergistic manner.
The plant’s transcriptional response to eggs, larval feeding and to the combination of eggs followed by larval feeding involves several phytohormone pathways
Our analysis, and the earlier original publications12,17,21,22,24 to which our analysis refers, found prominent regulation of GS related to phytohormone signalling. Therefore, we compared the enrichment of all GS associated with phytohormone signalling and metabolism that were up- or down-regulated in at least one of the species combinations (Fig. 3, Supplementary Table S3).
In response to eggs (Fig. 3a, E vs. C), many GS related to generic hormone responses were up-regulated in most plant–insect species combinations. GS related to JA, SA, ABA and ET follow this pattern. GS related to other phytohormones showed a more differentiated response pattern. A few auxin (AUX)-related GS were up-regulated, but their number differed between the plant–insect combinations. Gibberellic acid (GA)- and steroid-related GS were up-regulated in N. attenuata, but the latter were down-regulated in A. thaliana. Solanum dulcamara’s response to eggs involved a less clear enrichment of the ET- and SA-related GS than the responses to eggs by the other plant species.
Elevated transcription of SA-related GS in response to insect eggs has frequently been described but the conserved induction of JA-, ET-, and ABA-related GS is surprising. Enhanced activation of JA signalling is plausible for U. minor because the leaves in this dataset were mechanically wounded to mimic the leaf damage inflicted by the beetles during egg deposition24. The wound stimulus alone might have elicited the induction of JA-mediated pathways in the egg treatment45,46. Egg deposition in U. minor also elicits the emission of terpene volatiles, which is frequently linked to JA-dominated signalling events47. However, all of the other plant species we analysed also showed activation of JA-related GS in response to lepidopteran egg deposition, which does not damage any plant tissue (Figs. 2a and 3a). Some studies indicate that egg deposition without tissue damage might indeed elicit JA-related responses in plants. Solanum lycopersicum enhances the expression of a proteinase inhibitor (PI) gene in response to H. zea eggs, which correlates with increased JA levels18, and lepidopteran egg deposition on S. dulcamara results in JA-dependent enhanced leaf PI activity12, which is also linked to ABA and ET signalling37,48.
The distinct egg-induced changes in the expression of GS related to phytohormones suggest that plant responses to egg deposition do not only rely on SA- and ROS-related responses. Egg deposition rather causes a complex reorganisation of the dynamic phytohormonal signalling network, which is remarkably similar across the different plant–insect combinations.
The plant responses to larval feeding (Fig. 3b, F vs. C, up) involved similar phytohormonal GS as the responses to the eggs. They included strong up-regulation of JA-, SA-, ABA- and ET-related GS in most plant–insect combinations. The only exception was S. dulcamara, which showed strong up-regulation of JA-related GS, but clear down-regulation of ABA-, SA- and ET-related GS (Fig. 3b, F vs. C, down), which could indicate a weaker inducible response to larval feeding in this species, which maintains a quite effective constitutive defence due to its high content of steroidal alkaloids. Alternatively it might be a side effect of weaker feeding damage in the experimental setup used (see “Methods” section, Additional evaluation of Solanum dulcamara transcriptome data). In general, more AUX-related GS were up-regulated in response to feeding than in response to eggs.
Insect egg deposition enhanced the hormonal plant response to larval feeding (Fig. 3b, EF vs. F, up). The feeding-induced up-regulation of JA- and SA-related GS was further enhanced by prior egg deposition in all plant–insect combinations, although egg-treated elms showed this enrichment only at the onset of larval feeding (after 1 h), but not later on. ABA- and, to a lesser degree, ET-related GS were also commonly up-regulated when egg deposition preceded larval feeding. In U. minor, egg deposition also caused down-regulation of a fraction of genes in ET-related GS and further enrichment of down-regulated ET-related GS after 24 h of feeding (Fig. 3b, EF vs. F, down). Although in S. dulcamara phytohormonal regulation was only moderately affected by egg deposition and feeding alone, the combination of the two treatments led, as in the other plant–insect combinations, to increased expression of genes in ABA-, ET-, JA- and SA-related GS. In N. attenuata, feeding-induced expression of ET-related GS was only enriched after M. sexta egg deposition.
Taken together, JA-, ABA- and SA-related GS were more strongly enriched in all plant–insect combinations when egg deposition preceded larval feeding (Fig. 3). ET-related GS were strongly affected by both stimuli alone, but showed only faint additive or synergistic responses when eggs preceded larval feeding. Previous studies had already suggested that either JA- or SA-mediated pathways are further intensified in response to feeding when egg deposition occurs prior to larval feeding (Table 1). Our analysis corroborates these findings but furthermore suggests an interplay of several phytohormones mediating the improved anti-herbivore defence in egg-deposited and subsequently feeding-damaged plants. Although egg deposition and larval feeding are very different stimuli, it becomes quite clear that both affect the phytohormonal network at the transcriptional level in a similar way. Egg depositions and larval feeding may therefore trigger similar changes in metabolism-related GS across the plant–insect combinations (Fig. 2), which might contribute to the enhanced anti-herbivore defence we observed following egg deposition.
Phenylpropanoid metabolism and its involvement in the egg-mediated plant defence response to larvae
The results of the plant–insect combinations studied here suggest that regulation of the phenylpropanoid pathway is linked to the impaired performance of herbivores on previously egg-deposited plants (Table 1). Our analysis also shows that induction of the phenylpropanoid pathway by feeding damage was enhanced by prior egg deposition. Phenylpropanoids are well known for their diverse roles in anti-herbivore defence49,50,51. The phenylpropanoid pathway is widely branched52, and each branch leads to end products which may impair feeding herbivorese.g.53,54. We applied GAGE to evaluate the transcriptional regulation of those pathway branches that were regulated in at least one of the plant–insect combinations. In this way we could determine whether certain branches of the phenylpropanoid pathway showed analogous regulation patterns that could explain the egg-mediated enhancement of the plant’s defence against feeding herbivores across the different plant–insect combinations (Fig. 6, Phenylpropanoids (P); for abbreviations see Supplementary Table S3).
Comparison of GAGE analyses associated with the phenylpropanoid pathway between five different plant–insect combinations†, depicting plant responses to (a) insect eggs (E vs. C) and (b) larval feeding (F vs. C), eggs with subsequent feeding (EF vs. C) and the alterations in plant responses to feeding by prior egg deposition (EF vs. F). The heatmap depicts false discovery rate-adjusted p-values (FDR) according to the colour key for up- or down-regulated gene sets (GS). Black colour indicates GS which could not be assigned to the plant species or for which enrichment scores were not calculated due to a lack of data (E vs. C; Na-M). For a detailed description of the 20 phenylpropanoid pathway-associated GS (1P-20P) see Supplementary Table S3. Phen.pro: phenylpropanoids; Coum: coumarins; Fla-oids: flavonoids; Fla-ol: flavonols; Anthocy: anthocyanins. †At: Arabidopsis thaliana-Pieris brassicae; Na-M: Nicotiana attenuata-Manduca sexta Na-S: N. attenuata-Spodoptera exigua; Sd: Solanum dulcamara-S. exigua; Um-1 h/-6 h/-24 h: Ulmus minor-Xanthogaleruca luteola after 1/6/24 h of egg deposition and larval feeding, respectively, NA: not annotated. For detailed experimental setup description, see Fig. 1.
In response to eggs (Fig. 6a, E vs. C, up), we found conformable up-regulation of generic phenylpropanoid-related GS across all of our plant–insect combinations. These GS are mainly coumarin- and flavonoid-related. Ulmus minor showed up-regulation of flavonoid-related GS only after 6 h of egg deposition, although some of them were down-regulated after 1 h and after 24 h. In addition, lignin-related GS were clearly up-regulated after egg deposition in A. thaliana and S. dulcamara.
In response to feeding (Fig. 6b, F vs. C), GS in all branches of the phenylpropanoid pathway were up-regulated in almost all of the plant–insect combinations tested, with S. dulcamara being the only exception that did not show a response in its flavonol and anthocyanin-related GS. The response of N. attenuata to the herbivore species studied (M. sexta, S. exigua) differed with respect to the regulation of anthocyanin-related GS. Feeding by M. sexta led to up-regulation, but the response to feeding by S. exigua resulted in a more diffuse response pattern with less clear up-regulation, and even some down-regulation, of those GS. The late feeding-induced up-regulation of the phenylpropanoid-related GS in U. minor illustrates that there is a lag between the onset of feeding and the induction of this pathway, and that it is apparent even at the transcriptomic level.
When egg deposition preceded larval feeding (Fig. 6b, EF vs. F, up), pronounced up-regulation of several of the phenylpropanoid-related GS was found in all plant–insect combinations in response to feeding. This egg-enhanced response to feeding primarily affected GS related to flavonoids and anthocyanins in S. dulcamara and A. thaliana. Previously egg-deposited A. thaliana and U. minor both showed enhanced transcription in lignin-related processes in response to feeding damage. In N. attenuata, we found enhanced up-regulation of coumarin- and anthocyanin-related GS after egg deposition and feeding by S. exigua, but not by M. sexta. Interestingly, S. exigua feeding alone (without prior egg deposition) hardly induced any anthocyanin-related responses. Insect egg deposition on U. minor resulted in a stronger feeding-induced up-regulation of GS linked to lignin-related processes after a 6 h feeding period.
Our analysis shows that egg-enhanced activation of phenylpropanoid-related gene expression after feeding is indeed a conserved response across the plant species investigated. Metabolite analyses in A. thaliana showed increased flavonol levels in egg-deposited and feeding-damaged plants, while in N. attenuata caffeoylputrescine, a phenylpropanoid-polyamine conjugate, was found to be responsible for the reduced performance of S. exigua on egg-deposited plants17,19,20. Larvae of the elm leaf beetle suffered higher mortality on previously egg-deposited elm, and this was accompanied by an increased uptake of a flavonoid (kaempferol 3-O-robinoside-7-O-rhamnoside)55.
The huge diversity in plant secondary metabolites, including phenylpropanoids, likely facilitates plant defence as it hampers the counter-adaptations of herbivores feeding on those plants56. It is almost certain that each of the distantly related host plants investigated here holds a different phenylpropanoid profile. Almost all branches of the phenylpropanoid pathway in all plant–insect combinations were feeding-induced, but the modification of this induction profile by eggs was specific to the plant–insect combination analysed.
It appears that the egg-mediated modification of feeding-induced gene expression in the phenylpropanoid pathway in general is a conserved response, but the specific branches of this pathway seem to be affected in a plant-, and perhaps even herbivore-, specific way. Accordingly, plants might use the egg stimulus not only to prepare against impending herbivory in general, but to fine-tune the feeding-induced phenylpropanoid defences according to the specific herbivore they are likely to encounter. This idea is further supported by the finding that N. attenuata exhibits altered transcriptomic responses to feeding by S. exigua and M. sexta when the plant has received the eggs of the respective other herbivore prior to feeding21.
Source: Ecology - nature.com
