A Comparison of the Effects of FATTY ACID DESATURASE 7 and HYDROPEROXIDE LYASE on Plant–Aphid Interactions

The spr2 mutation in tomato (Solanum lycopersicum), which disrupts function of FATTY ACID DESATURASE 7 (FAD7), confers resistance to the potato aphid (Macrosiphum euphorbiae) and modifies the plant’s C6 volatile profiles. To investigate whether C6 volatiles play a role in resistance, HYDROPEROXIDE LYASE (HPL), which encodes a critical enzyme in C6 volatile synthesis, was silenced in wild-type tomato plants and spr2 mutants. Silencing HPL in wild-type tomato increased potato aphid host preference and reproduction on 5-week old plants but had no influence on 3-week old plants. The spr2 mutation, in contrast, conferred strong aphid resistance at both 3 and 5 weeks, and silencing HPL in spr2 did not compromise this aphid resistance. Moreover, a mutation in the FAD7 gene in Arabidopsis thaliana also conferred resistance to the green peach aphid (Myzus persicae) in a genetic background that carries a null mutation in HPL. These results indicate that HPL contributes to certain forms of aphid resistance in tomato, but that the effects of FAD7 on aphids in tomato and Arabidopsis are distinct from and independent of HPL.


Introduction
FATTY ACID DESATURASE 7 (FAD7) is an ω-3 fatty acid desaturase (FAD) that is widely similar in sequence throughout the plant kingdom, and that desaturates 16-and 18-carbon fatty acids with two double bonds (C16:2 and C18:2) to generate fatty acids with three double bonds (C16:3 and C18:3) [1,2]. In diverse plant species, expression levels of genes encoding FAD7 and other FADs increase in response to some stresses and decrease in response to others, suggesting that modulation of desaturase activity plays a role in stress adaptation [3][4][5][6]. Moreover, artificial manipulation of desaturase activity through silencing or other genetic modifications alters plants' susceptibility to a variety of abiotic and biotic stresses, enhancing resistance to some stresses and compromising resistance to others [7][8][9][10]. In short, FADs appear to influence stress resistance.
One form of stress resistance that is negatively correlated with FAD7 activity is resistance to aphids, a group of phloem-feeding insects that include many serious agricultural pests. The suppressor of prosystemin-mediated responses2 (spr2) mutant in tomato, which has a null mutation in FAD7, has enhanced resistance to potato aphids (Macrosiphum euphorbiae) compared to wild-type plants [10]. Aphid resistance in spr2 comprises both antixenosis (i.e., decreased host preference) and antibiosis (increased mortality and decreased fecundity). Moreover, population growth of the green peach aphid (Myzus persicae) is significantly lower on Arabidopsis thaliana mutants with null mutations in FAD7 than on wild-type controls [10]. These results indicate that the FAD7 enzyme or its products negatively regulate aphid resistance in more than one plant family.
One way that FAD7 could possibly influence plant defenses against aphids is by affecting the profile of substrates available to the hydrogen peroxide lyase (HPL) pathway. HPL generates six-carbon aldehydes and alcohols (C6 volatiles) from fatty acid hydroperoxides that are produced from C18:2 and C18:3 by 13-lipoxygenase (13-LOX) (Figure 1) [11]. Loss of function of FAD7 results in decreased C18:3 and increased C18:2, and in tomato, this shift in precursors has been shown to result in dramatically altered C6 volatile composition [12,13]. In particular, decreased levels of C18:3 result in significantly lower (Z)-3-hexenal and (Z)-3-hexenol than observed in wild-type plants, and enhanced C18:2 levels result in elevated levels of hexanal and hexanol [12,13]. Conversely, overexpression of the FAD7 gene in tomato increases the production of C6 volatiles derived from 18:3 and decreases the production of 18:2-derived compounds [14]. The impact of altered FAD7 activity on volatile profiles generated by the HPL pathway could potentially alter aphid host selection and/or survival and fecundity on foliage.  Several lines of evidence indicate that the HPL pathway can influence direct defenses against insects. In in vitro tests, C6 volatiles including hexanal, (E)-2-hexenal, hexanol, (E)-2-hexenol, and (Z)-3-hexenol have been shown to reduce aphid fecundity [15]. Moreover, artificial manipulation of HPL gene expression can influence insect resistance. In potato, antisense suppression of a gene encoding a 13-HPL resulted in increased aphid fecundity [16]. Similarly, a null mutation in an HPL homolog rendered rice more susceptible to another piercing-sucking insect, the rice brown plant hopper (Nilaparvata lugens) [17]. In Arabidopsis, however, overexpression of HPL did not influence the host preference, fecundity, or growth of the green peach aphid, even though it resulted in a fifty-fold increase in C6 volatile production [18]. Furthermore, silencing HPL in coyote tobacco (Nicotiana attenuata) decreased the feeding behavior and growth of a tobacco hornworm (Manduca sexta) [19], and overexpression of an HPL gene from tea (Camellia sinensis) in tomato decreased resistance to another chewing insect, Prodenia litura [20]. These studies indicate that the HPL pathway influences insect resistance, but that its effects vary in different species combinations, and that further work is needed to understand the role of HPL in specific plant-insect interactions.
The goals of this study were to determine whether the HPL pathway contributes to direct plant defenses against aphids in wild-type tomato, and whether it is required for the enhanced aphid resistance observed in spr2 plants, which have impaired FAD7 function. We examined the effects of silencing HPL in wild-type plants on aphid infestations at two different stages of plant development (3 and 5 weeks after planting), in comparison with the effects of the spr2 mutation. Silencing HPL increased aphid host preference, offspring production, and offspring survival on 5-week old plants, but did not affect aphid infestations on 3-week old plants, and did not influence the survival of adult aphids at either stage of plant development. In contrast, the spr2 mutation had a strong suppressive effect on adult survival, fecundity, and offspring survival at both 3 and 5 weeks after planting. Thus, the timing and effects of aphid resistance associated with HPL differ from those of spr2-dependent resistance. We also developed a tomato line (spr2HPL-RNAi) that is deficient in both FAD7 function and HPL expression in order to determine if loss of function of HPL would compromise aphid resistance associated with spr2. The spr2HPL-RNAi line showed similar levels of aphid resistance as the spr2 parent, despite dramatically reduced levels of HPL expression. Similarly, bioassays in Arabidopsis indicated that loss of function of FAD7 could confer aphid resistance even in a genotype that carries a null mutation in HPL (the fad7-1 mutant, which we confirmed to be homozygous for the hpl mutation). In summary, our results indicate that both HPL and FAD7 influence antixenosis and antibiosis against aphids in tomato, but that the effects of FAD7 on aphids are distinct from and independent of HPL.

Confirmation of Silencing of HPL in Tomato
A transgenic line in which expression of HPL was targeted for silencing by RNA interference (RNAi) [21] was utilized for this study. PCR detection of the kanamycin resistance gene NPTII confirmed the presence of the transgene in individuals of the HPL-RNAi line, and semi-quantitative PCR confirmed that HPL transcript abundance was lower in HPL-RNAi plants than in untransformed controls (cv. Flora-Dade) ( Figure 2). Analysis of hexanal, (Z)-3-hexenal, (E)-2-hexenal, and (Z)-3-hexen-1-ol levels by gas chromatography (Supplementary Figure S1A-D) also revealed that average C6 volatile production was lower in 5-week-old HPL-RNAi plants than in wild-type Flora-Dade.

Influence of HPL and spr2 on Aphid Survival and Fecundity on Tomato
Survival of adult aphids and offspring production were measured in no-choice aphid bioassays to assess the effects of silencing HPL on aphid antibiosis in a wild-type tomato cultivar (cv. Flora-Dade). For comparison, antibiotic aphid resistance was also quantified in spr2 compared to its wild-type control (cv. Castlemart). Assays were performed with more than one age of plant (3-and 5-week old plants) just in case the effects of HPL on antibiotic defenses varied with plant age. Based on previous reports, HPL activity and volatile production both can vary over the course of development [22,23].
At both stages of development tested here, adult survival was significantly lower on spr2 than on the wild-type control Castlemart six days after inoculation (3-week old plants: p = 0.042; Figure 3A. 5-week old plants: p < 0.0001; Figure 3B). The average number of live offspring per cage, which is a measure of adult fecundity, was also more than 50% lower on spr2 than on Castlemart at either developmental stage (p < 0.0001; Figure 3C,D).
In contrast to spr2, which influenced aphid infestations at both 3 and 5 weeks, the effects of silencing HPL on aphid populations varied with plant age. The live offspring on HPL-RNAi were not significantly different from its control Flora-Dade at 3 weeks (p = 0.977; Figure 3C), but were approximately 29% higher than on wild-type at five weeks (p = 0.005; Figure 3D). Adult survival did not differ between HPL-RNAi and Flora-Dade at either age (p > 0.05; Figure 3A,B). These data indicate that HPL contributes to antibiotic defenses against juvenile aphids in five-week old plants, whereas the spr2 mutation promotes antibiosis against adults and juveniles at both 3 and 5 weeks after planting.

Influence of HPL and spr2 on Aphid Host Preference on Tomato
To measure the effects of HPL and spr2 on aphid settling behavior, pair-wise choice tests were performed to compare the HPL-RNAi line or spr2 to their respective wild-type controls. Choice tests were performed with 5-week-old plants, since both HPL and spr2 influence antibiotic defenses at this life stage ( Figure 3). In comparisons between spr2 and wild-type (cv. Castlemart) plants, aphids initially dispersed onto the two genotypes equally, with no significant difference in adult numbers at 1 h after introduction into the choice arena (p = 0.19; Figure 4A). Over time, the numbers of adults on spr2 decreased and the numbers on wild-type controls increased; compared to wild type plants, spr2 had significantly fewer aphids at 6 h (p = 0.005), 24 h (p < 0.0001), and 48 h (p < 0.0001) after inoculation ( Figure 4A). Offspring were first observed at 6 h, and their abundance was significantly lower on spr2 compared to wild-type control at 24 h (p < 0.0001), and 48 h (p < 0.0001) ( Figure 4B).
When HPL-RNAi was compared to its wild-type control (cv. Flora-Dade), the numbers of adult aphids were significantly higher on HPL-RNAi at 6 h (p = 0.038), 24 h (p = 0.009) and 48 h (p = 0.038) ( Figure 4C); and the numbers of offspring were significantly higher on HPL-RNAi at 24 h (p = 0.044) and 48 h (p = 0.030) ( Figure 4D). These results indicate that aphid host preference is enhanced on the HPL-RNAi line compared to wild-type plants, whereas the spr2 mutation decreases aphid host preference. The effects of spr2 on aphid host preference appeared to be much greater than the effects of HPL-RNAi; for example, at 48 h, the number of juveniles on spr2 was 97% lower than the number on wild-type plants, whereas silencing HPL caused only a 37% change in offspring numbers. Aphid settling behavior was assessed by recording on which plant the adults were located, and how many offspring they produced at 1 h, 6 h, 24 h and 48 h after inoculation (HAI). Asterisks (*) indicate statistically significant differences at α = 0.05 according to Matched pairs one-sided t-tests, and error bars represent SEM (n ≥ 10).

Silencing HPL in the spr2 Mutant
To determine whether the HPL pathway might contribute to enhanced aphid resistance in spr2, the HPL-RNAi line was crossed with spr2, and PCR genotyping was used to identify F 2 progeny that were homozygous for the spr2 mutation and positive for the HPL-RNAi construct. RT-qPCR confirmed that HPL transcript abundance was significantly reduced in these plants (spr2HPL-RNAi), as well as in the HPL-RNAi parent line ( Figure 5). Compared to the parental lines, the spr2HPL-RNAi line had intermediate levels of hexanal, a C6 volatile that accumulates to high levels in spr2 (Supplementary Figure S1A). No-choice assays were performed to assess aphid performance on spr2HPL-RNAi and the parental lines. Bioassays were performed five weeks after planting to focus on a time point when both spr2 and HPL-RNAi impact aphid resistance. As in previous assays, adult survival and live offspring production were significantly lower on spr2 than on the wild-type control Castlemart (p < 0.05; Figure 6A,B), and live offspring production was higher on HPL-RNAi than on Flora-Dade (p < 0.05; Figure 6B). For both measures of aphid performance, numbers on spr2HPL-RNAi were similar to numbers observed on spr2 (p > 0.05, Figure 6A,B), and were significantly lower than numbers observed on the wild-type controls Castlemart and Flora-Dade or on the HPL-RNAi parental line (p < 0.0001; Figure 6A,B). These data indicate that levels of aphid resistance in spr2HPL-RNAi are comparable to levels of resistance in spr2, and that silencing HPL in an spr2 background does not compromise aphid resistance mediated by spr2. Figure 5. HPL Expression in spr2HPL-RNAi. RT-qPCR was used to compare HPL expression in the wild-type cultivar Flora-Dade (FD), a transgenic line in which the HPL gene was silenced (HPL-RNAi), and progeny of the spr2 X HPL-RNAi cross that were homozygous for the spr2 mutation in the FAD7 gene and positive for the HPL-RNAi transgene (spr2HPL-RNAi). Expression values were normalized using the housekeeping gene Rpl2 and calculated relative to the wild-type control. Relative expression data were analyzed by one-way ANOVA, and mean separations were performed using Tukey-Kramer HSD. Bars having the same letter are not significantly different at α = 0.05, and error bars represent SEM (n ≥ 3). Figure 6. The single and combined effects of impairments in FAD7 and HPL on aphid survival and reproduction. A no-choice assay on 5-week-old plants was used to measure aphid performance on F 2 progeny of the spr2 X HPL-RNAi cross compared to aphid performance on the parental lines (spr2 and HPL-RNAi) and their respective wild-type controls, Castlemart (CM) and Flora-Dade (FD). All progeny used for this assay were confirmed by PCR to be homozygous for the spr2 mutation and positive for the HPL-RNAi transgene. Data was analyzed by one-way ANOVA and Tukey-Kramer HSD. Bars having the same letter are not significantly different at α = 0.05, and error bars represent SEM (n ≥ 10).

FAD7 and HPL in Arabidopsis
In parallel, we also explored whether the homologous FAD7 gene in Arabidopsis required a functional copy of the HPL gene to influence aphid performance. The Columbia ecotype (Col-0) was previously reported to carry a ten-nucleotide deletion in HPL that eliminates the gene's function [24]. PCR genotyping revealed that the fad7-1 mutant, which originated from Col-0, is also homozygous for the hpl mutation ( Figure 7A). This fad7-1 hpl mutant supported significantly fewer aphids than the wild-type genotypes Columbia and Nossen (p < 0.0001; Figure 7B). Thus, aphid resistance conferred by loss of function of the FAD7 protein is independent of HPL in Arabidopsis as well as in tomato.
Unlike Columbia, the Nossen (Nos) ecotype carries a functional allele of the HPL gene ( Figure 7A) [24]. Aphid population growth did not differ significantly between these two ecotypes (p = 0.2293; Figure 7B). These results suggest that HPL does not have a major impact on aphid infestations on Arabidopsis, or that its effects are outweighed by other differences between these two genotypes. Figure 7. Aphid performance on Arabidopsis genotypes with and without mutations in FAD7 and HPL. PCR was used to determine the presence of wild-type and mutant alleles of the HPL gene in Columbia (Col-0), Nossen, and fad7-1 ((A), NTC = no template control). A no-choice test was used to assess performance of the green peach aphid on these genotypes ((B), n = 20). Aphid numbers were analyzed by one-way ANOVA, and mean separations were performed using Tukey's HSD. Bars having the same letter are not significantly different at α = 0.05, and error bars represent SEM.

Discussion
One objective of this study was to determine whether HPL contributes to plant defenses against aphids in wild-type tomato plants. Silencing HPL expression in wild-type plants (cv. Flora-Dade) had no effect on aphid infestations three weeks after planting (Figure 3), but resulted in enhanced aphid host preference, reproduction, and survival when infestations occurred about five weeks after planting (Figures 3 and 4). These data suggest that HPL contributes to both antixenotic and antibiotic defenses against aphids on tomato, and that these defenses vary between 3-and 5-week old plants. Potentially, the activity of the HPL pathway may vary with plant age; for example, in rice, HPL enzyme activity is low in seedlings and peaks twelve weeks after sowing [22]. Therefore, the fact that silencing HPL at three weeks had little effect on aphids could be due to relatively low levels of HPL activity in three-week old plants.
A second objective was to determine if the HPL pathway contributes to aphid resistance in the spr2 mutant, which is defective in a chloroplast-localized fatty acid desaturase FAD7. Like HPL, the spr2 mutation also impacts C6 volatile synthesis because it alters the relative abundance the fatty acid substrates for volatile synthesis. Because the spr2 mutation enhances C18:2 accumulation and decreases C18:3 synthesis, it promotes production of C18:2-derived volatiles such as hexanal and inhibits accumulation C18:3 derivatives such as (Z)-3-hexenol [12,13]. However, the results of our aphid bioassays suggest that products of the HPL pathway do not have a causal role in the enhanced levels of aphid resistance observed in spr2. Whereas silencing HPL did not influence aphid infestations on three-week old plants, the spr2 mutation has just as strong an impact on aphids in three-week old seedlings as in five-week old plants (Figure 3). Moreover, when HPL is silenced in the spr2 mutant line, it has no effect on aphid performance on this line ( Figure 6). These results indicate that HPL expression is not essential to aphid resistance in spr2.
Consistent with our observations in tomato, the fad7-1 mutation in Arabidopsis also enhances aphid resistance even in a genetic background that carries the hpl mutation (Figure 7), even though this mutation has previously been shown to suppress C6 volatile synthesis [22]. Thus, FAD7 activity in tomato and Arabidopsis modulates direct defenses against aphids independent of HPL activity. Our results also suggest that HPL may not have a strong, direct impact on aphid infestations on Arabidopsis, since aphid population growth was similar on ecotypes with (Col-0) and without (Nossen) the hpl mutation ( Figure 7). This is consistent with a previous report that overexpression of HPL had no influence green peach aphid population increase or host preference on Arabidopsis even though overexpression increased C6 volatile levels by over 40-fold [18].
In conclusion, HPL contributes to basal aphid resistance in wild-type tomato plants, but enhanced aphid resistance in mutants with impaired FAD7 function is independent of HPL gene expression. In Arabidopsis, although HPL aids in indirect defenses against aphids by recruiting parasitoids wasps [18], HPL does not appear to contribute significantly to direct defenses against aphids in the fad7 mutant. These results indicate that fatty acid metabolism in plants can influence plant-aphid interactions through routes independent of C6 volatile synthesis. It is unlikely that loss of function of FAD7 would impact the nutritional quality of plants for aphids, because polyunsaturated fatty acids are naturally rare in the phloem sap on which aphids feed [25]. Our prior work also indicates that aphid resistance in spr2 is not influenced by jasmonate signaling, although it requires salicylic acid accumulation and Non-expressor of Pathogenesis Related Proteins (NPR1) [10]. These findings emphasize the need for further work to understand the mechanisms through which components of primary metabolism including fatty acid desaturation influence plant defense signaling and immunity.

Tomato Culture
Five tomato (Solanum lycopersicum) genotypes were used in this study: a mutant line with impaired FAD7 activity called suppressor of prosystemin-mediated responses2 (spr2) [26], a transgenic line silenced for HPL (HPLi-1653-3 [21], referred to here as HPL-RNAi), a line deficient in both FAD7 function and HPL expression (spr2HPL-RNAi), and two wild-type cultivars, Castlemart and Flora-Dade. The spr2 mutant carries a point mutation that results in loss of function of FAD7 [2], and Castlemart is the genetic background that was originally used to develop spr2. The creation of the HPL-RNAi line in the tomato cultivar Flora-Dade was previously described [21]. In brief, a 330-bp fragment comprising bases 562-881 of the HPL open reading frame in the sense orientation and a 595-bp fragment comprising bases 562-1154 of the HPL open reading frame in the antisense orientation were expressed in cv. Flora-Dade under the control of the Figwort mosaic virus 35S in order to induce silencing of HPL. The authors previously demonstrated that the HPL RNAi line used in this study has significantly reduced HPL mRNA accumulation and C6 volatile production in the fruits as well as foliage [21]. Since fruits are not typically produced until at least 8 weeks after planting, this data indicated that silencing was persistent in mature plants. The spr2HPL-RNAi line was produced by crossing spr2 and HPL-RNAi (described below). Tomato plants (Solanum lycopersicum) were grown in LC1 Sunshine potting mix (Sungro Horticulture, Bellevue, WA, USA) with 15-9-12 Osmocote slow-release fertilizer (Scotts-MiracleGro Company, Marysville, OH, USA) at 23 • C and L16:D8 photoperiod in an environmental growth chamber (Conviron, Winnipeg, MB, Canada), and watered daily with a dilute nutrient solution containing 1000 ppm CaNO 3 (Hydro Agri North America, Tampa, FL, USA), 500 ppm MgSO 4 (Giles Chemical Corp, Waynesville, NC, USA), and 500 ppm 4-18-38 Gromore fertilizer (Gromore, Gardena, CA, USA).

Development of the spr2HPL-RNAi Line
A tomato line with impairments in both FAD7 and HPL was developed by manually transferring pollen from HPL-RNAi to spr2 and screening the (spr2 × HPL-RNAi) F 2 generation for individuals that were positive for the HPL-RNAi transgene and homozygous for the spr2 mutation. Screening for the spr2 mutation was performed by PCR using two allele-specific primer sets that target a single nucleotide polymorphism as described by Avila et al. [10]. Presence of the HPL-RNAi transgene in F 2 plants was determined by amplifying the selectable marker NPTII (Neomycin phosphotransferase II) using forward (5 -GCAATATCACGGGTAGCCAA-3 ) and reverse (5 -GCCGTGTTCCGGCTGTCA-3 ) primers. NPTII PCR was performed at 95 • C for 5 min; 95 • C for 45 s, 50 • C for 45 s, and 72 • C for 45 s (30 cycles); and final extension at 72 • C for 5 min.

Arabidopsis Culture and Materials Development
Arabidopsis plants (Arabidopsis thaliana) were grown in a peat, vermiculite, perlite (4:3:2 ratio, Sungro Horticulture, Bellevue, WA, USA) soil mixture supplemented with 15-9-12 Osmocote Plus fertilizer (Scotts-MiracleGro Company, Marysville, OH, USA) at 23 • C and L13:D11 photoperiod in a growth chamber (Conviron, Winnipeg, MB, Canada). The plants were fertilized weekly with Miracle Gro ® all-purpose plant food (Scotts-MiracleGro Company, Marysville, OH, USA). Arabidopsis ecotype Columbia (Col-0, CS70000) was obtained from the Arabidopsis Biological Resource Center (Ohio State University, Columbus, OH, USA), and the Nossen ecotype and the fad7-1gl1 mutant (developed in a Col-0 genetic background) were obtained from Dr. Jyoti Shah (University of North Texas). Because the fad7-1gl1 mutant carries a mutation (gl1) in a gene required for trichome production (GLABRA1) in addition to a mutation in FAD7, this mutant was crossed with Col-0 to develop another line (fad7-1) with impaired FAD7 function but normal trichome development. In the F 2 generation, plants with trichomes were screened by PCR with primer sets specific to the mutant and wild-type alleles of FAD7 to select for plants homozygous for the mutant fad7-1 allele [27]. Plants that lacked the gl1 mutant phenotype and that were homozygous for the fad7-1 mutation were then propagated to generate seeds for subsequent assays. All plants were observed to confirm the presence of trichomes before they were used for experiments.

Identification of the HPL Mutation in the Arabidopsis fad7-1 Mutant
The fad7-1 mutant in Arabidopsis was screened by PCR for the presence of a 10-bp deletion (from 161 bp to 170 bp) in the HYDROPEROXIDE LYASE (HPL) gene that occurs naturally in the ecotype Columbia (Col-0), and that results in a non-functional HPL protein [24]. Col-0 was included as a positive control for the mutant allele, and the Nossen ecotype was included as a positive control for the wild-type HPL allele. Genomic DNA was extracted using an extraction buffer that was made by diluting Edwards solution (200 mM Tris-HCl (pH 7.5), 250 mM NaCl, 25 mM EDTA, and 0.5% SDS) by 10-fold with TE buffer (10 mM Tris-HCl (pH 8) and 1 mM EDTA) [28,29]. Each sample was used for two separate PCR reactions: one with a primer set that amplifies only the wild-type HPL allele (At4g15440.1) (forward 5 -GGACCGTTTAGATTACTTCTGGTT-3 , reverse 5 -CGGAAGTCTCCGATGAGAAC-3 ), and another reaction with a primer set that specifically targets the mutant hpl allele with the 10 nts deletion (5 -GACCGTTTAGATTCCAAGGAC-3 , reverse 5 -CGGAAGTCTCCGATGAGAAC-3 ). PCR amplification was performed at 95 • C for 5 min, followed by 30 cycles of 95 • C for 45 s; 55 • C for 45 s, and 72 • C for 45 s, and a final extension at 72 • C for 5 min. PCR products were separated by electrophoresis on 1% agarose gels.

RNA Isolation and Gene Expression Analysis
For analysis of gene expression, total RNA was extracted from approximately 100 mg of flash-frozen leaf tissue using TRIzol reagent and chloroform (Invitrogen Corp., Carlsbad, CA, USA) using the manufacturer's instructions. cDNA was synthesized from 1 µg of total RNA per sample using Superscript III reverse transcriptase and oligo dT (18) primers in a 20 µL reaction volume (Invitrogen Corp., Carlsbad, CA, USA). Transcript abundance in the cDNA was then quantified by semi-quantitative PCR or real-time PCR. For semi-quantitative PCR, 50 ng of cDNA was used as a template and the final concentration for each primer was 0.4 µM. The PCR program was: 95 • C for 5 min; 95 • C for 30 s, 52 • C for 30 s, and 72 • C for 30 s (22 cycles); and final extension at 72 • C for 5 min. PCR products were detected on a 1% agarose gel. Real-time quantitative PCR was performed on 2 µL of cDNA in a 20 µL reaction volume using the QuantiTect SYBR Green PCR Kit (Qiagen, Inc., Valencia, CA, USA) on a StepOnePlus Real-time PCR system (Applied Biosystems, Foster City, CA, USA). The RT-qPCR program was: 95 • C activation for 10 min, followed by 40 cycles of amplification and quantification (denaturation at 95 • C for 15 s, annealing at 52 • C for 30 s, extension 72 • C for 30 s with a single fluorescence detection). Melting curves were generated at 60-95 • C with a heating rate of 0.3 • C per second. Three biological replicates for each genotype and two technical replicates for each biological replicate were used. Transcript abundance of tomato HPL (Solyc07g049690; GenBank accession AF230372.1) was measured using primers previously described by Shen and coworkers [21] (5 -AGCTACGGATTGCCGTTAGT-3'/ 5 -TTTCCATTCTCTTGGTGAAGAA-3 ). Data were normalized to the expression levels of the endogenous control Ribosomal Protein L2 (RPL2) using primers previously described by Avila and coworkers [10] (5 -GAGGGCGTACTGAGAAACCA-3 /5 -CTTTTGTCCAGGAGGTGCAT-3 ). Gene expression was calculated relative to the wild-type control for each genotype comparison using the methodology described by Pfaffl [30].

Aphid Survival and Fecundity on Tomato
No-choice assays were performed to evaluate potato aphid survival and fecundity. Wingless adult potato aphids within 24 h of emergence to adulthood were confined to single leaflets of intact plants using clip cages (4 adults per cage, 2 cages per plant, 10-15 replicate plants per genotype), and the numbers of living and dead adults and offspring were recorded at six days after infestation (6-DAI). Both 3-and 5-week-old plants were inoculated to determine if aphid resistance varied with age. The positions of the cages were standardized for all plants in each assay; cages were placed on the terminal leaflet of the 2-3 fully-expanded leaves below the apical meristem. Plants were maintained in a growth chamber at 23 • C and 16L: 8D photoperiod during the bioassay.

Aphid Host Preference on Tomato
Settling behavior of the potato aphid was measured on intact tomato plants by placing adult aphids on choice arenas that allowed them to move back and forth between paired tomato genotypes. Each arena consisted of a Styrofoam platform (15 cm diameter) that was placed underneath two paired leaflets: the terminal leaflet of the third fully expanded leaf below the apical meristem of each of the paired plants. Wingless adult potato aphids within 24 h of emergence to adulthood (14 adults per arena) were placed between the leaflets and confined to the arena using a vented petri dish lid with a soft gasket that prevented damaging the petioles (Supplementary Figure S2). The number of adult aphids on each plant were recorded at 1 h, 6 h, 24 h and 48 h after release. Offspring production was also monitored because it is a well-established marker of host plant acceptance [31]. In each experiment, ten to fifteen replicate pairs per combination of genotypes were tested using five-week-old tomato plants, and each experiment was performed at least twice.

Aphid Survival and Fecundity on Arabidopsis
To measure aphid performance on Arabidopsis, plants were inoculated with the green peach aphid (3 wingless newly-emerged adults/plant; 15 plants/genotype) when first flower buds were visible (developmental stage 5.1 according to [32]). After infestation, plants were covered with sleeve cages and maintained for 7 days in a growth chamber (23 • C; 65% relative humidity; L13:D11 photoperiod). The numbers of live and dead adults and offspring aphids on each plant were scored 7 days after infestation (DAI) in this no-choice assay.

Statistical Analysis
All statistical analyses were done with JMP ® v.11 (SAS Institute Inc., Cary, NC, USA). Host preference assays were analyzed by matched pairs one sided t-tests within each time point, and no-choice assays were analyzed by one-way ANOVA, with α = 0.05.

Supplementary Materials:
The following are available online at http://www.mdpi.com/1422-0067/19/4/1077/ s1. Figure S1: Comparison of C6 volatile levels in five-week-old plants with modifications in fatty acid desaturation and/or HYDROPEROXIDE LYASE expression. Figure S2: Design of choice assays to study aphid host preference on tomato.