Plasmid engineering of aphid alarm pheromone in tobacco seedlings affects the preference of aphids

ABSTRACT

Plants producing sufficient amount of aphid alarm pheromone by expressing (E)-β-Farnesene (EβF) synthase gene may contribute to plant protection by reducing aphid populations. However, terpene biosynthesis varies among plant species and developmental stages. In the present study, volatile headspace analysis of tobacco seedlings with MaβFS1 (an EβF synthase from the Asian peppermint Mentha asiatica) failed to generate EβF. We further targeted MaβFS1 to the tobacco plastid, using a chloroplast targeting sequence, either with or without the AtFPS1 gene for the biosynthesis of the precursor farnesyl diphosphate. When both MaβFS1 and AtFPS1 genes were targeted to the chloroplast, low levels of EβF were detected in stably transformed tobacco seedlings; resulting in specific repellence of the green peach aphid, Myzus persicae. These data indicate that redirecting the EβF biosynthetic pathway from its natural cytosolic location to the chloroplast is a valid strategy. This redirecting strategy may be very useful for other crop plants that do not naturally produce EβF or other repellent volatiles.

Aphids are globally distributed and among the most economically important agricultural pests, due to their direct ingestion of phloem sap, production of honeydew, and transmission of various phytopathogenic viruses.^1-3 Insect pheromones, such as the aphid alarm pheromone, have been widely considered as potential alternatives to current pesticides, and are applicable for various integrated pest management (IPM) programs.^4-6 (E)-β-farnesene (EβF) is the main or sole component of the alarm pheromone for most species of aphids. EβF may cause dispersal and repellence of aphids, as well as increase foraging by natural enemies including parasitic wasps and predators (e.g. ladybird beetles, syrphids, and lacewings).^4,7 Interestingly, EβF also occurs in the volatiles of several plant species that are either constitutively emitting EβF and/or do so in response to herbivore. For caterpillar-damaged maize and trichomes of the wild potato Solanum berthaultii under natural conditions, the released EβF could act as an aphid deterrent.^8,9

EβF synthase enzyme gene that catalyzes the conversion of the cytosolic farnesyl diphosphate (FPP) to EβF has been identified in several plant species, including Yuzu Citrus junos,¹⁰ Douglas fir Pseudotsuga menziesii,¹¹ sweet wormwood Artemisia annua,^12,13 chamomile Matricaria recutita,¹⁴ and peppermint (Mentha piperita and Mentha asiatica).^15-17 Following the identification of EβF synthase genes, efforts were made to reduce aphid infestation through genetic engineering. Expression of an EβF synthase from M. piperita, enabled Arabidopsis plants to produce EβF to keep the pest aphid Myzus persicae away and to increase the foraging behavior of Diaeretiella rapae.¹⁸ Similar results have been documented for transgenic tobacco,^13,17,19 rice,²⁰ wheat,⁴ Indian mustard Brassica juncea,²¹ and alfalfa Medicago sativa.²²

Mechanisms of terpene biosynthesis vary among plant species and developmental stages.^23-26 Previous studies documented that transgenic tobacco at the flowering stage produced small amounts of EβF.^13,17,19 Younger plants are more susceptible to aphids in nature, and during the young plant stage, aphid infestations significantly increase the economic losses of crop plants in the field.² In this context, we studied whether tobacco seedlings could generate sufficient EβF to repel aphids. Volatiles from both the control and MaβFS1 (an EβF synthase gene isolated from M. asiatica) seedlings were collected and analyzed via GC-MS. GC-MS analysis showed that the profile of MaβFS1 seedlings was similar to that of control seedlings, and no unique EβF peak was observed (Figure S1), indicating that MaβFS1 seedlings did not emit a detectable amount of EβF.

The precursor of sesquiterpenes FPP is naturally located in the cytosol; however, the plastid-isoform FPP synthase has been reported to occur in some species such as Arabidopsis, tomato, tobacco, rice, and wheat.^27-29 Plant plastids (e.g. mitochondria and chloroplast) are an ideal compartment for satisfactory biosynthesis of sesquiterpenes via the retargeting of one or multiple biosynthetic enzymes, because of the high flux of IPP and DMAPP from MEP pathway in this organelle.^7,30,31 We proposed targeting sesquiterpene synthases into plastids to increase the EβF biosynthesis in our previous reviews.^2,7 Here, two chloroplast-targeting constructs [CMaβFS1-pBI121 and CMaβFS1:CAtFPS1-pBI121 (Figure 1(a,b))] were transferred into the tobacco. A total of 12 transgenic events with CMaβFS1-pBI121 and eight events with CMaβFS1:CAtFPS1-pBI121 were obtained. Positive lines from the T₀–T₂ generations were confirmed by PCR. Expression of the genes in T₂ generations was confirmed via qRT-PCR with the primer pairs listed in table S1. Out of these, we chose three lead T₂ lines each of CMaβFS1 (CMa2, CMa6, and CMa11) and CMaβFS1:CAtFPS1 lines (CMaCAt3, CMaCAt4, and CMaCAt7; Figure 1(c–f)) to conduct GC-MS analyses. Tobacco seedlings expressing chloroplast-targeted forms of MaβFS1 plus AtFPS1 (CMaβFS1:CAtFPS1 lines) exhibited a unique EβF peak that was not present in the control (Figure 1(g)). The emission levels of EβF from CMaCAt3, CMaCAt4, and CMaCAt7 lines were up to 4.43 ng/d per g fresh tissues. In contrast, we did not detect EβF in the headspace of tobacco seedlings expressing the chloroplast form of MaβFS1 alone that was unexpected. These results indicate that redirecting the EβF biosynthesis pathway from cytosolic to plastid by overexpressing MaβFS1 plus AtFPS1 in the chloroplast of tobacco seedlings increases the amount of emitted EβF.

Figure 1.

Generation of tobacco lines with plastid-form of MaβFS1 alone or MaβFS1 plus AtFPS1. a) Schematic of CMaβFS1 expression cassette. b) Schematic of CMaβFS1:CAtFPS1 expression cassette. Nos P, Nos promoter; Nos T, Nos terminator; CaMV 35S P, CaMV 35S promoter. c) PCR identification of CMaβFS1 in T₂ lines. d) PCR identification of CAtFPS1 in T₂ lines. e) Gene expression analysis of CMaβFS1 in T₂ lines. f) Gene expression analysis of CAtFPS1 in T₂ lines. M, DL2000 DNA marker; P, CMaβFS1:CAtFPS1-pBI121 plasmid; 1, control with the blank vector pBI121; 2–4, CMaβFS1 lines CMa2, CMa6, and CMa11, respectively; 5–7, CMaβFS1:CAtFPS1 lines CMaCAt3, CMaCAt4, and CMaCAt7, respectively. g) Volatile analysis of tobacco seedlings expressing plastid-form of MaβFS1 alone or MaβFS1 plus AtFPS1.

The continuous release of EβF in transgenic plant contributed to repelling aphids and reducing aphid colonization and plant damage.^13,17-22 In the present research, one MaβFS1 transgenic line Ma4, one CMaβFS1 transgenic line CMa2, and three CMaβFS1:CAtFPS1 transgenic lines (CMaCAt3, CMaCAt4, and CMaCAt7) were chosen for the Y-tube olfactometer assay. Y-tube results demonstrated that two CMaβFS1:CAtFPS1 transgenic seedlings (CMaCAt3 and CMaCAt7) were significantly repellent to alate aphids (Figure 2(a); P < 0.05 or P < 0.01, χ²-test). Compared with untransformed and control tobacco seedlings, no different behavioral activity to entrainment volatiles of MaβFS1 and CMaβFS1 lines was detected (Figure 2(a); P > 0.05, χ²-test). Further bioassays were performed with each of the transgenic and control seedlings placed in a 30 × 30 × 30 cm insect cage as indicated in Figure 2(b), with introduction of 50 alate aphids that were starved for 2 h. The number of aphids on each plant was counted after 30 and 60 min, respectively, with the aphids staying on the net cover were recorded as no choice. Compared with the control at the time-point of 30 min, aphids on CMaCAt3 and CMaCAt7 were decreased by about 34.9% (P < 0.05, t-test, n = 3) and 44.2% (P < 0.01, t-test, n = 3), respectively (Figure 2(d)). At the 60 min, the number of aphids was reduced by approximately 31.0% in CMaCAt3 (P < 0.05, t-test, n = 3), and 42.7% in CMaCAt7 (P < 0.01, t-test, n = 3), respectively (Figure 2(e)). The CMaβFS1:CAtFPS1 transgenic seedlings present here exhibited a repellent effect to minimize aphid infestation.

Figure 2.

Preference of green peach aphid Myzus persicae on transgenic and control tobacco seedlings. WT, wild type tobacco seedling; C1, control seedling with the blank vector pBI121; Ma4, representative line of MaβFS1 seedling; CMa2, representative line of CMaβFS1 seedling; CMaCAt3, CMaCAt4, and CMaCAt7 are three representatives of CMaβFS1:CAtFPS1 lines. a) Behavioral response of M. persicae tested in a Y-tube olfactometer. The behavioral responses of 50 alate aphids to the volatiles collected from different transgenic lines were recorded. Choices between odor sources were analyzed with χ2-test. Asterisks indicate significant differences between treatments (*p < 0.05, **p < 0.01). ns indicates no significant difference at the 5% level (χ2-test). b) Schematic of the setup used for aphid repellence test in the greenhouse. Fifty alate aphids were placed in the midpoint circle at the start of each assay. The number of aphids was counted after 30 min and 60 min, respectively. c) The repellence test between wild type and control seedlings. Values shown in the figure are means ± SEM from three biological replicates. ns indicates no significant difference at the 5% level (t-test). d) The repellence test between control and CMaCAt3 seedlings. Values shown in the figure are means ± SEM. Single asterisk indicates significant differences at the 5% level between treatments (t-test). e) The repellence test between control and CMaCAt7 seedlings. Asterisks indicate significant differences between treatments (t-test; **P < 0.01).

The data reported in this paper strongly support our hypothesis that plastid-engineered tobacco seedlings generate sufficient EβF to repel aphids. Several points are germane. First, MaβFS1 seedlings did not emit a detectable amount of EβF. Second, targeting sesquiterpene synthases to plastids leads to demonstrable increases in EβF emissions. Third, aphid bioassays document aphid repellency in two engineered tobacco lines, CMaCAt3 and CMaCAt7. Taken together, these points amount to a potent argument that plastid-engineered crop plants may produce ecologically relevant emissions of pest-repellent volatiles. We recognize, however, that EβF and other plant volatile emissions highly volatile. Broadly considered, the volatile nature of the emissions is a necessary pre-condition for chemical communication between insects and their potential herbivores. It follows that the volatility leads to rapid dilution of the chemicals into the environment, and effective communication requires substantial biosynthesis and continued release of the volatiles. We note, too, that many plant volatiles are susceptible to oxidation,³² which also drives selection for high biosynthesis and release. For deployment in agriculture, GM plants must release sufficient levels of EβF to cause the natural aphid alarm repellency response. In the context of field-level applications, our work has developed a model system for targeting enzymes for biosynthesis of plant volatiles to plastids. Future research should focus on improving the model to achieve high volatile production in crop plants for pest management. For a specific example, other multipoint metabolic engineering strategies to increase carbon flux through the mevalonate (MVA) or MEP pathway leading to FPP and EβF synthesis remain to be evaluated.

Funding Statement

This work is funded by the National Natural Science Foundation of China (Grant no. 31601379), the Education Department of Henan Province (grant no. 14A210004), and the National Key Research and Development Program of China (No. 2016YFD0300400).

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.