Effects of Bactericera cockerelli Herbivory on Volatile Emissions of Three Varieties of Solanum lycopersicum

Domesticated tomato (Solanum lycopersicum L.) crops have presented an increased susceptibility to pests under field and greenhouse conditions. Among these pests is tomato/potato psyllid, Bactericera cockerelli Sulc (Hemiptera: Triozidae), a major pest in solanaceous crops. In this study, we evaluated volatile organic compound (VOC) emissions from the headspace in three healthy varieties of tomato plants (Floradade, Micro-Tom and wild) under greenhouse conditions using solid-phase microextraction and gas chromatography–mass spectrometry (SPME/GC-MS). Later, independent bioassays were performed to evaluate VOC emissions with three varieties infested with nymphs of B. cockerelli. The results in healthy plants showed markedly different VOC profiles in each variety (14 compounds for wild, 17 for Floradade and 4 for Micro-Tom). Plants infested with nymphs showed changes in VOC emissions distinctly in Floradade and wild varieties. We suggest that these qualitative differences in VOC profiles by the degree of domestication could explain the preferences of B. cockerelli.


Introduction
Plant crops can interact with many organisms in the environment, including phytopathogens and insects. Plant defenses involve different mechanisms to avoid this type of attack, and among the direct defenses, the salicylic acid (SA) pathway is activated during plant-bacteria hemibiotrophic and phloem-feeding insect interactions [1,2]. Meanwhile, the jasmonic acid/ethylene (JA/ET) pathway is activated by attacks from necrotrophic pathogens and chewing insects [3,4]. In addition, plant defenses also involve indirect mechanisms to counteract the damage caused by pathogens and herbivores [5], and these are mediated by volatile organic compounds (VOCs) with different functions in the plant.
Those plants that were infested with psyllids showed a differential profile for the eight VOCs identified, in which the compounds (+)-4-carene, β-phellandrene and (+)-sabinene (peak area > 10 8 ) were the most abundant ( Figure 3). Nevertheless, only the compound α-pinene was statistically different between wild and Micro-Tom plants (P = 0.000154).  The three varieties of tomato plants emitted qualitative and quantitative changes in VOC emissions. For example, Floradade plants presented eight compounds and the majority of these compounds were reduced in infested plants (Table 1, Figure 4B); further, seven VOCs were exclusive to healthy plants (  The three varieties of tomato plants emitted qualitative and quantitative changes in VOC emissions. For example, Floradade plants presented eight compounds and the majority of these compounds were reduced in infested plants (Table 1, Figure 4B); further, seven VOCs were exclusive to healthy plants (Table 1). In wild plants, five compounds were exclusive to healthy plants and two compounds were found as exclusive to infested plants (Table 1). Micro-Tom plants presented very poor emissions in all assays performed here (Table 1).

Discussion
In this study, we evaluated VOC emissions of three varieties of tomato in healthy and psyllidinfested plants. Previous studies in our research group found that B. cockerelli adults presented a strong oviposition preference toward the domesticated S. lycopersicum cv. Floradade and a lower preference for S. lycopersicum cv. Micro-Tom and S. lycopersicum cv. wild [25]. These differences in the

Discussion
In this study, we evaluated VOC emissions of three varieties of tomato in healthy and psyllid-infested plants. Previous studies in our research group found that B. cockerelli adults presented a strong oviposition preference toward the domesticated S. lycopersicum cv. Floradade and a lower preference for S. lycopersicum cv. Micro-Tom and S. lycopersicum cv. wild [25]. These differences in the preferences of psyllids could be attributed to VOC blends in healthy plants. For this purpose, we performed some bioassays and the VOC profiles were obtained for healthy and infested plants (Table 1). Considering the qualitative differences among varieties of tomato VOC, we speculated that some volatiles could act as attractants or repellents for B. cockerelli adults, even though in this study we did not perform an electroantennography analysis to validate the results. In similar studies, healthy tomato plants (S. lycopersicum cv. Castlemart) emitted 11 VOCs in the constitutive blend. α-pinene, o-cymene, (+)-4-carene, and β-phellandrene were found among these compounds [24]. The shared compounds with VOC patterns from the headspace of healthy tomato plants were obtained from three varieties, although our results triggered 14 VOCs for wild, 17 VOCs for Floradade and 4 VOCs for Micro-Tom (Table 1). The oviposition behavior by B. cockerelli in solanaceous crops is mediated through the olfactory and visual stimuli to locate their hosts [26]. Some factors, such as natal host plant and plant phenology, also improve the settling and oviposition [23]. In other studies, wild varieties of tomato (Solanum peruvianum) that bore the Mi-1.2 gene, presented a reduced oviposition of B. cockerelli, compared with susceptible varieties that lacked the gene [27]. In field conditions, B. cockerelli showed a strong preference for potato and tomato compared with eggplant and pepper [26].
To determine whether nymph feeding promoted differences in VOC emissions in the three varieties of tomato, two herbivory bioassays were performed on tomato leaves. In Floradade-nymph interactions, α-thujene was detected as the most abundant compound after 2 h of herbivory (Table 1). This compound was detected in basil plants emissions (Ocimum gratissimum L. and O. basilicum L.) and may be involved in the oviposition deterrence against Tuta absoluta [28]. Regarding VOCs induced in wild plants, β-phellandrene and (+)-4-carene were detected as the most abundant ones (Table 1). This could correlate with the reduced oviposition of B. cockerelli [25] and also with the role of monoterpenes involved with repellence in solanaceous plants against whiteflies, such as terpinenes and phellandrenes [9]. Interestingly, Micro-Tom plants infested with nymphs presented a reduced number of VOCs emitted, and α-pinene was the most abundant one (Table 1). Besides, we did not perform leaves surface analysis to evaluate the differences in trichome numbers. The synthesis of diverse trichome-borne metabolites influences host plant selection by herbivores under natural conditions [29]. Prager et al. [30] showed that total alkaloid and phenolics levels were greater in three breeding lines of potato compared with commercial cultivar. These differences might explain the reduced psyllid oviposition and aphid reproduction observed. Although our experiments were performed with B. cockerelli CaLso-positive, 2 h of herbivory by nymphs (4th and 5th stages) was sufficient to induce VOC emissions for leaf damage. In this sense, to transmit CaLso in solanaceous, it is necessary to have at least 24 h of feeding by B. cockerelli [31]. CaLso manipulates host tomato defenses within 1 week, and when the pathogen population is increased in the phloem, it can evoke changes in the expression of defense genes [21]; this modification of gene expression may induce changes in VOC emissions to attract the psyllid vector [32]. Therefore, we speculate that the changes in VOC emissions in each tomato variety were promoted by B. cockerelli herbivory and not by CaLso. Mustafa et al. [33] showed that bittersweet nightshade (Solanum dulcamara) improved the development and body size of three haplotypes of B. cockerelli (Hemiptera: Triozidae) compared to potato plants. Lately, the study of wild relatives has been expanded to understand those mechanisms involved for indirect defense against phytophagous insects, with an aim to improve the commercial crops.

Plant Growth
Tomato seeds (Solanum lycopersicum) from three varieties were germinated in peat-moss until two leaves were presented. Seedlings from domesticated (S. lycopersicum cv. Floradade, Crown Seeds, Homestead, FL, USA), Micro-Tom (S. lycopersicum cv. Micro-Tom), and wild type (S. lycopersicum cv. wild) were transplanted in 1.5 L pots containing peat-moss:perlite (70:30 v/v). The plants were watered every 3 days and fertilized once a week with Steiner's nutritive solution (25%). Tomato plants were maintained under greenhouse conditions at 28 ± 2 • C, RH = 45%, and were allowed to grow until they reached 20 cm height or had four to five fully extended leaves for the bioassays. The seeds of Micro-Tom employed were kindly provided by Dr. John Délano-Frier (CINVESTAV-Unidad Irapuato), whereas the wild type seeds were collected in summer 2015 in Irapuato (20 • 34 34.3" N 101 • 23 24.3" W), Guanajuato, Mexico. Wild type tomato was taxonomically characterized as a member of Solanum lycopersicum and was deposited with catalogue number XAL0148356 in Herbario XAL by Carlos Durán Espinoza from Instituto de Ecología A.C. in Veracruz, Mexico.

Insect Rearing
The potato/tomato psyllids (Bactericera cockerelli Sulc) were reared on tomato plants under greenhouse conditions. B. cockerelli (CaLso positive) was maintained in cages of 50 × 50 × 50 cm with the same conditions described above. Psyllids were donated by Dr. Alberto Flores from Departamento de Parasitología of Universidad Autónoma Agraria Antonio Narro (UAAAN), in Saltillo, Mexico.

Headspace Collection and Analysis of Plant Volatiles
Three varieties of tomato plants (2 weeks after transplanting) were selected to collect the volatile emissions (healthy plants). Firstly, to obtain volatile profiles, we employed the headspace-solid-phase microextraction and gas chromatography-mass spectrometry (HS-SPME-GCMS) system. Individual plants were placed into a polypropylene desiccator previously adapted with a charcoal-activated trap, and flow-air (6 L/min) was passed on the plant for 5 min (Floradade); polyethylene bags (Sunbag, MERCK KGaA, Darmstadt, Germany) were utilized for Micro-Tom and wild. To obtain the VOC profiles, three varieties with three biological replicates (n = 3) were performed for a single bioassay in healthy tomato plants (psyllid-free).
To evaluate the VOCs released in the three varieties during B. cockerelli nymph feeding, two bioassays were performed independently with three biological replicates for all varieties. With a small brush, 25 nymphs of B. cockerelli (fourth and fifth stages) were gently placed on each plant and allowed 2 h feeding on the tomato leaves (nymphs were left in plants). After feeding, individual plants were collected, using either bags or a desiccator, to evaluate the headspace of each treatment by SPME. For SPME, gray fiber (50/30 µm, DVB/CAR/PDMS, MERCK KGaA, Darmstadt, Germany) was used in all evaluations. The fiber was exposed during 2.5 h. Healthy plants (nymphs-free) were exposed at the same time for SMPE. The bioassays were performed at 28 ± 2 • C in greenhouse conditions. All fibers were injected and desorbed in a gas chromatograph (GC 7890A, Agilent Technologies, Wilmington, DE, USA) coupled to mass selective detector (Agilent Technologies MSD 5975 C). All analyses were performed under running conditions: initial oven temperature of 60 • C for 1 min, increasing 8 • C/min until 280 • C for 1 min. Injector temperature was 230 • C. The VOCs were analyzed employing a capillary column Agilent Technologies HP-5MS (30 m × 0.25 mm × 0.25 µm). After 15 minutes, the fiber was removed from the injector and stored until use. All the compounds were identified by a NIST92 and Wiley library, according the percentage of quality presented (> 90%); retentions times were compared with library, and Kovats indices also were determined for individual compounds.

Data Analysis
All VOCs collected from headspace of the three varieties of tomato (healthy and infested) were graphed using GraphPad Prism version 5 for Windows, GraphPad Software, La Jolla, California USA. Prior to analysis, the raw data of peak areas were log transformed and were tested using univariate ANOVA, the non-parametric Kruskal-Wallis test. Statistical analyses were performed using R Statistical Software (R Core Team 2004).