Chemo-Enzymatic Synthesis of Optically Active γ- and δ-Decalactones and Their Effect on Aphid Probing, Feeding and Settling Behavior

The enantiomerically enriched γ- and δ-decalactones (4a and 4b) were prepared from corresponding racemic primary-secondary 1,4- and 1,5-diols (1a and 1b), as products of enzymatic oxidation catalyzed by different alcohol dehydrogenases. The results of biotransformations indicated that the oxidation processes catalyzed by alcohol dehydrogenase (HLADH), both isolated from horse liver and recombinant in Escherichia coli, were characterized by the highest degree of conversion with moderate enantioselectivity of the reaction. Useful, environmentally friendly extraction procedure of decalactones (4a and 4b) based on hydrodistillation using a Deryng apparatus was developed. Both racemic lactones (4a and 4b), as well as their enantiomerically enriched isomers, were tested for feeding deterrent activity against Myzus persicae. The effect of these compounds on probing, feeding and settling behavior of M. persicae was studied in vivo. The deterrent activity of decalactones (4a and 4b) against aphids depended on the size of the lactone ring and the enantiomeric purity of the compounds. δ-Decalactone (4b) appeared inactive against M. persicae while γ-decalactone (4a) restrained aphid probing at ingestional phase. Only (–)-(S)-γ-decalactone (4a) had strong and durable (i.e. lasting for at least 24 hours) limiting effect, expressed at phloem level.


Introduction
Compounds with the lactone moiety, regarding to their broad range of valuable properties, have been the subject of many studies. Among different biological activities, the most widely known is antimicrobial activity [1,2]. They also exhibit strong cytostatic activity. Compounds with the lactone structure on the one hand occur as antifeedants [3][4][5][6][7], on the other play pheromones role in the world of insects [8]. Lactones are described in the literature as important flavor compounds [9][10][11][12]. They exist as naturally occurring odorants in all major food groups and in many different beverages.
The odoriferous attributes of γand δ-decalactones (4a and 4b), which are widely spread in the nature, are well documented [10,13]. Privileged, occurring predominantly natural form of such lactones are the (R)-enantiomers, which are responsible for a pleasant, naturally fruity, aromatic note. Their odor is more intense compared to the (S)-isomers. (+)-(R)-γ-Decalactone (4a) exhibits intense, fruity with peach note odor, in opposite to (-)-(S)-γ-enantiomer (4a), which aroma is characterized as less intense, sweet, coconut with fruity note. γ-Decalactone (4a) is used in the formulation of fruit aromas such as strawberry and peach. Besides, it is also present in various fermented products such as bread and whisky [14]. (+)-(R)-δ-Decalactone (4b), which is sweet, creamy and has a nut-like odor, was identified in fat-containing food, such as cheese and butter. (+)-(R)-δ-Decalactone (4b), already used as food additive, is mixed with margarine to give a flavor similar to that of butter. Apart from the food industry it is used as a fragrance composition in the cosmetics industry and a flavor in the tobacco industry. Both lactones with their tempting aroma and low odour thresholds are currently one of the main products derived from aroma biotechnology.
γand δ-Decalactones (4a and 4b) have been widely investigated with regard to their flavor, but only few studies have been conducted on their other biological activities. Nevertheless, biological studies point at the very broad spectrum of acivities of γand δ-decalactones (4a and 4b). For example, it has been shown that both decalactones (4a and 4b) inhibit the growth of selected filamentous fungi (Aspergillus niger), yeast (Candida albicans) and bacteria (Staphylococcus aureus) [15]. δ-Decalactone (4b) has been detected in venom-producing mandibular glands of ponerine ant Pachycondyla apicalis [16] and (+)-(R)-δ-decalactone (4b) was reported as a component of the warning odor of North American porcupine Erethizon dorsatum [17]. γ-Decalactone (4a) as a component of the sex pheromone of scarab beetles Osmoderma eremita plays an important role in the communication process between insects [18][19][20]. It has been proven that only the (+)-(R)-γ-decalactone (4a) is active in relation to the female beetles being responsible for the characteristic fruity, peach and plum scent secreted by the male beetles. At the same time, (+)-(R)-γ-decalactone (4a) is used by the larval predator, the click beetle Elater ferrugineus, as a kairomone to locate O. eremita, which is its prey [20].
Biological properties of chiral compounds are often strongly related to the absolute configuration [21]. Therefore, looking for new stereoselective methods of synthesis of the optically pure molecules with various biological activities is one of the most developed areas in current chemistry. One of the method from green chemistry toolbox is biotransformation showing overwhelming advantages over synthetic chemistry approach. Synthesis involving enzymes allows stereochemical control of reactions and usually leads to optically active compounds, in the ideal case, enantiomerically pure. Biotransformations are characterized by high enantio-, regio-and chemoselectivity. Enzymes, through a detailed diagnosis of the three-dimensional structure of the compound by the enzymatic pocket, have high substrate specificity. It is also possible to provide reactions of low-reactive chemicals and reactions of not activated areas of the molecule. This often leads to receive products impossible to obtain by chemical synthesis. The ability to conduct biotransformation under mild conditions (temp, pH) reduces the number of possible side reactions. Moreover, biocatalysis decreases the consumption of chemicals, which is associated with a significantly smaller amount of toxic chemical waste.
Among different biocatalytic strategies of the synthesis of lactones [22][23][24][25], we have applied the enzymatic oxidation of aliphatic diols [26]. Encouraged by our latest valuable results [27], concerning biosynthesis of cis-and trans-whisky lactones, we have focused on enzymatic oxidation of long chain primary-secondary diols. In our current investigations, we have applied eight commercially available alcohol dehydrogenases in the stereoselective oxidation of 1,4and 1,5-decadiols (1a and 1b) to the corresponding γand δ-decalactones (4a and 4b).
Taking into consideration the aforementioned broad spectrum of lactone properties it was purposeful to check whether γand δ-decalactones (4a and 4b), besides their documented odoriferous, antimicrobial and attractant properties indicate also antifeedant activity. Our interests in this area is inspired by literature data describing the natural and synthetic compounds with γand δ-lactone function that are known as active feeding deterrents against many insects, including aphids [3][4][5][6]. In regard of this, both racemic γand δ-decalactones (4a and 4b), as well as their enantiomerically enriched isomers, were studied for feeding deterrent activity against the polyphagous peach potato aphid Myzus persicae (Sulz.).

Analysis
The purity of synthesized products was checked by thin layer chromatography (TLC), using aluminum foil plates coated with silica gel. Compounds were detected by spraying the plates with 1% Ce(SO 4 ) 2 and 2% H 3 [P(Mo 3 O 10 ) 4 ] in 10% H 2 SO 4 . The crude products were purified by preparative column chromatography using silica gel (Kieselgel 60, 230-400 mesh, Merck) with mixtures of hexane-acetone (in various ratios) as eluents.
The structures of the compounds were determined on the basis of 1 H NMR, 13 C NMR, Dept-135, HMQC, COSY and IR spectral data. 1 H NMR spectra were recorded in CDCl 3 solutions on Bruker Avance DRX 600 (600 MHz) spectrometer. IR spectra were determined using FTIR Thermo-Mattson IR 300 Spectrometer. Optical rotations were measured on an Autopol IV automatic polarimeter (Rudolph). The pH measurements were conducted on a HI 9321 Microprocesor pH Meter equipped with a glass electrode.

Enzymes and coenzymes
The following alcohol dehydrogenases were used: Horse Liver Alcohol Dehydrogenase (HLADH) from Sigma-Aldrich; Yeast Alcohol Dehydrogenase (YADH) from Sigma-Aldrich; Horse Liver Alcohol Dehydrogenase recombinant from Escherichia coli (HLADH-rec) from Evocatal GMBH; Lactobacillus Kefir Alcohol Dehydrogenase (LKADH) from Fluka; Primary Alcohol Dehydrogenase screening kit (PADH I, II and III) from Codexis. Bacillus Stearothermophilus Alcohol Dehydrogenase (BSADH) was a generous gift from Protein Biochemistry Institute, Naples, Italy.

Synthesis of substrates for enzymatic oxidation
Decane-1,4-diol (1a) and decane-1,5-diol (1b) were prepared by the reduction of commercially available corresponding racemic γand δ-decalactones (4a and 4b) with LiAlH 4 , according to the method described by us earlier [26] Enzymatic oxidation of racemic decane-1,4-diol (1a) and decane-1,5-diol (1b) Screening procedure. In the screening experiments diol (1a-b) (0.02 g) and coenzymes: NAD(P) + (0.007 g) and FMN (0.2 g) were dissolved in 0.1 M glycine-NaOH buffer (15 mL) at different value of pH (7.2, 8.5, 9.0) and temperature (25, 35, 45°C). The pH of the mixture was readjusted with 2 M NaOH. Then, the enzyme solution (0.002 g) dissolved in buffer (2 mL) was added dropwise. To determine the stability of the substrate the control sample, containing a buffer with the substrate without the addition of the enzyme, was conducted. Changes in pH during the biotransformation was monitored using a pH-meter. Samples (1 mL) of the reaction mixtures were taken after several time intervals (2,24,48 and 120 hours). The aqueous phase was acidified to pH = 3, washed with brine, then extracted with chloroform (1 mL). The extract was dried over anhydrous magnesium sulfate and analyzed by GC.
Preparative procedure. The preparative-scale transformations were performed fivefold in the same conditions as screening experiments using 0.1 g of diol (1a-b) (5 x 0.02 g), 0.035 g of NAD(P) + (5 x 0.007 g), 1 g FMN (5 x 0.2 g) and 0.01 g enzyme (5 x 0.002 g). When the reaction was completed the reaction mixture was extracted by hydrodistillation method using a Deryng apparatus. The yields and optical rotations of prepared products are given below. The precise spectral data for prepared diastereoisomeric mixture of hemiacetals, which are in agreement with literature data [28], were also presented.

Extraction procedure using a Deryng apparatus
After the biotransformation, the combined aqueous fractions were placed in a 500 mL round flask and 0.1M NaOH was added portionwise to pH = 12. The aqueous fraction was heated for 2 hours, after the boiling point was reached. The vapours were condensed by means of a cold refrigerant. The main purpose of medium alkalization was conversion of prepared lactones 4a and 4b, by opening of lactone ring, into hydroxyacid salts, which are water-soluble and nonvolatile compounds. After the extraction, the solvent (1 mL of cyclohexane) containing the volatile compounds-hemiacetals 2a-3a and 2b-3b was collected. Then, the reaction mixture was acidified by 0.1M HCl to pH = 3 and distilled again for 2 hours in Deryng apparatus. During the hydrodistillation of volatile compounds contained in the aqueous layer, γand δ-decalactones (4a and 4b) were extracted with 1 mL of cyclohexane. The unreacted diol 1a and 1b remained in the aqueous medium, was used to next enzymatic oxidation. In this way, purified and separated substrate as well as volatile products without time-consuming and toxic solventfree column chromatography were obtained.

Bioassays
Cultures of aphids and plants. Laboratory clone of peach-potato aphid Myzus persicae (Sulz.) was maintained on Chinese cabbage Brassica pekinensis in laboratory at 20°C, 65% of relative humidity, and L16:8D photoperiod. One to seven days old apterous females of M. persicae and 3 week old plants with 4-5 fully developed leaves were used for experiments. All experiments were carried out under the same conditions of temperature, relative humidity, and photoperiod. The bioassays were started at 10-11 a.m.
Behavioural Responses of Aphids. The effect of γand δ-decalactones (4a and 4b) was assessed by monitoring settling behaviour of free individuals and aphid stylet activities in plant tissues (aphid probing and feeding behaviour).
Aphid settling. Aphids settle on a plant only when they accept it as a food source [29]. Therefore, the number of aphids that settle and feed on a given substrate is a good indicator of its suitability. This bioassay allows to study aphid host preferences under semi-natural conditions. Aphids are given free choice between control and treated leaves. The studied γand δdecalactones (4a and 4b) were applied by immersing a leaf in 0.1% ethanolic solution of a given compound for 30 sec. Control leaves of similar size were immersed in 70% ethanol that was used as a solvent for the tested lactones. Treated and control leaves were placed in a Petri dish and allowed to dry for 1 hour to permit the evaporation of the solvent. Next, aphids were placed in the dish along the line that divided the arena into two halves so that aphids could choose between treated (on one half of a Petri dish) and control leaves (on the other half of the dish). Aphids that settled, i.e. they did not move and the position of their antennae indicated feeding [30] on each leaf were counted at 1h, 2h, and 24h intervals after access to the leaf (8 replicates, 20 viviparous apterous females/replicate). Aphids that were moving or out of any of the leaves were not counted.
The data were analyzed using one way analysis of variance ANOVA (STATISTICA 6.1. package). If aphids showed clear preference for the leaf treated with the tested compound (P<0.05), the compound was described as having attractant properties. If aphids settled mainly on the control half of the leaf (P<0.05), the compound tested in the respective choice-test was stated a deterrent. From the data thus obtained the relative index of deterrence (DI) was calculated on the basis of formula DI = C-T/C+T, where C represents the number of aphids that settled on the control half of the leaf and T represents the number of aphids that settled on the half of the leaf treated using tested compound. The values of DI range between 1 (ideal deterrent) and -1 (ideal attractant). The compounds that modified aphid settling behaviour (i.e., the settling deterrents) were selected for further detailed study on probing behaviour.
Aphid probing and feeding behavior. Aphid probing and, especially, the phloem sap uptake by M. persicae was monitored using the technique of electronic registration of aphid probing in plant tissues, known as electrical penetration graph (EPG), that is frequently employed in insect-plant relationship studies considering insects with sucking-piercing mouthparts. In this experimental set-up, aphid and plant are made parts of an electric circuit, which is completed when the aphid inserts its stylets into the plant. Weak voltage is supplied in the circuit, and all changing electric properties are recorded as EPG waveforms that can be correlated with aphid activities and stylet position in plant tissues [31]. The parameters describing aphid behaviour during probing and feeding, such as total time of probing, proportion of phloem patterns E1 and E2, number of probes, etc., are good indicators of plant suitability or interference of probing by chemical or physical factors in individual plant tissues [32]. Lactones 4a and 4b were applied to one leaf of a 3-week old plant by immersing in 0.1% ethanolic solution of a given compound for 30 sec. Control leaves of similar size were immersed in 70% ethanol that was used as a solvent for the tested halogenated lactones. Treated and control leaves were allowed to dry for 1 hour before the start of the experiment to permit the evaporation of the solvent. Aphids were attached to a golden wire electrode with conductive silver paint and starved for 1 h prior to the experiment. Probing behaviour of 12 apterous females per studied lactone/aphid combination was monitored for 8 h continuously with a four-channel DC EPG recording equipment. Each aphid was given access to a freshly prepared leaf. Signals were saved on the computer and analysed using the PROBE 3.1 software provided by W. F. Tjallingii (www.epgsystems.eu) The following aphid behaviours were distinguished: no penetration (waveform 'np'-aphid stylets outside the plant), pathway phase-penetration of non-phloem tissues (waveforms 'ABC'), salivation into sieve elements (waveform 'E1'), ingestion of phloem sap (waveform 'E2'), and ingestion of xylem sap (waveform 'G'). The parametres derived from EPGs were analysed according to their frequency and duration in configuration related to activities in peripheral and vascular tissues.
One-way analysis of variance (ANOVA) was carried out on the experimental results. All calculations were performed using the STATISTICA 6.1 package (StatSoft, Tulsa, OK, USA).

Chemo-enzymatic synthesis
Enzymatic oxidation of racemic decane-1,4-diol (1a). There are two possible approaches to obtain lactone from decane-1,4-diol (1a) (Fig 1). In the first one, the primary hydroxyl group of diol is chemoselectively oxidized to carboxylic group. Then, in acidic pH, hydroxycarboxylic acid formed is cyclized to lactone product (route I). The second pathway assumes two steps of enzymatic oxidation of diol. Primary-secondary diol is initially oxidized to appropriate hydroxyaldehyde and then spontaneously formed hemiacetals are further oxidized to corresponding lactone (route II). As detected by GC and confirmed by spectroscopic methods, the products of the first step of oxidation of decane-1,4-diol (1a), instead of the hydroxyaldehyde, were their corresponding hemiacetals 2a and 3a.
Due to the fact that HLADH isolated directly from horse liver is not commercially available now, we used also the enzyme HLADH recombinant in Escherichia coli. To extend our investigations, in looking for the most effective biocatalyst, we applied several other alcohol dehydrogenases in oxidation processes of decane-1,4-diol (1a) ( Table 1). It was interesting to evaluate the ability of different alcohol dehydrogenases and compare their stereoselectivity in transformation of such substrate.
Hemiacetals 2a and 3a, as a diastereoisomeric cis/trans mixture, were isolated by column chromatography from products mixture after 120 hours of oxidation catalyzed by PADH III. Identification of their structures were made on the basis of chemical shifts of the signals from proton H-5 in the 1 H NMR spectrum. Considering the space structure of individual diastereoisomers 2a and 3a, the position of hydroxyl group and hexyl group on the same side of the lactol ring will point to the cis isomer of hemiacetal (2a) (Fig 2). However, in the trans isomer (3a) both substituents are located in the opposite sides of the lactol ring. The ratio of cis/trans isomers (45% of 2a and 55% of 3a) was determined from the integration of signals of proton H-5. Doublet of triplets from the proton H-5 in the 1 H NMR spectrum of the trans isomer 3a is shifted downfield (δ = 4.19), in comparison with its location in the 1 H NMR spectrum of the cis isomer 2a (δ = 3.98). Such a difference in chemical shift is due to deshielding effect of the oxygen atom of the hydroxyl group located on the same side of hemiacetal ring as the proton H-5 in the trans-isomer 3a. The presence of one-proton multiplets from H-2 respectively at 5.45 (for 2a) and 5.54 ppm (for 3a) in the 1 H NMR spectrum indicates the presence of hemiacetal ring in the molecule.
Based on screening scale experiments important to notice is fact that, in contrary to formation of γ-lactone ring, all tested alcohol dehydrogenases, with exception of HLADH, were poor or inadequate biocatalysts to afford δ-lactone product (Table 3). It confirms that HLADH in oxidation process of decane-1,5-diol (1b) is irreplaceable enzyme. In preparative scale (+)-(R)-  Table 3. The composition (in % according to GC) of products mixture after 5 days of enzymatic oxidation of racemic decane-1,5-diol (1b).

Enzyme
OXIDATION CONDITIONS Substrate Products δ-decalactone (4b) with higher optical purity (ee = 56%) was achieved after five days of biotransformation catalyzed by HLADH recombinant in Escherichia coli (Table 4). Through spectroscopic analysis ( 1 H NMR, 13 C NMR, HMQC, COSY) the products of the first step of oxidation process were determined as diastereoisomeric cis/trans mixture of hemiacetals 2b and 3b. Identification of these products were made on the basis of chemical shifts of the signals from protons H-2 and H-6 in the 1 H NMR spectrum. The signals of these protons were assigned to the corresponding isomers on the basis of their chemical shifts. Assuming the chair conformation and equatorial position of alkyl group at C-6 (Fig 4) in the cis isomer hydroxyl group was also located in equatorial position, whereas in trans isomer it occupied axial position. In trans isomer the multiplet of H-6 is shifted downfield (δ = 3.92) by the axial hydroxyl group, in comparison with its location in the 1 H NMR spectrum of the cis isomer (δ = 3.40). On the other hand, the signal of equatorial proton H-2 in the trans isomer is shifted downfield (δ = 5.30) by deshielding effect of neighboring oxygen atom, in comparison with the chemical shift (δ = 4.69) of the axial proton H-2 found in the spectrum of the cis isomer. The ratio of cis/trans isomers (68% of 2b and 38% of 3b) was determined from the integration of multiplets of protons H-6 and H-2.
Chemical oxidation of diol 1b and hemiacetals 2b and 3b. As we plan to obtain both enantiomers of δ-decalactone (4b) we performed chemical oxidation of corresponding unreacted isomers of lactols 2b, 3b and diol 1b. The diastereoisomeric mixtures of 2b and 3b, isolated from both preparative transformations, were oxidized with pyridine dichromate to opposite isomer of δ-decalactone (4b). In this way enantiomerically enriched (-)-(S)-isomer (4b) with ee = 32% ([α] D 25 = -5.2°(c 3.3, CHCl 3 )) and ee = 58% ([α] D 25 = -13.7°(c 3.6, CHCl 3 )) was obtained as a product of two-step chemoenzymatic synthesis. Firstly, oxidation of diol 1b to lactol 2b and 3b catalyzed by HLADH isolated from horse liver as well as recombinant in Escherichia coli, followed by chemical oxidation took place. On the other hand the  Our goal of chemical oxidation of unreacted diol 1b and diastereoisomers of hemiacetals 2b, 3b formed during biotransformation was to evaluate the enantioselective step in enzymatic oxidation of primary-secondary 1,5-diols. Similarly to biotransformations of 1,4-diols, in oxidation of 1,5-diols the second stage of diol oxidation-oxidation of the corresponding hemiacetals to lactone-determines the enantioselectivity of the whole oxidation process. To our knowledge stereoselective enzymatic oxidation of diols to lactones is novel and effective method to obtain diastereomerically enriched hemiacetals, which can be further chemically oxidized to opposite lactone isomer.

Biological activity
Aphid settling. The behavior of Myzus persicae during settling on leaves varied depending on the decalactone applied (Fig 5). Racemic γ-decalactone (4a) showed a relatively high significant deterrent activity as soon as one hour after application (index of deterrence ID 1 = 0.6), but the deterrent effect ceased in the course of time (ID 24 = 0.2). However, considering the enantiomeric composition of γ-decalactone (4a), (-)-(S)-γ-decalactone discouraged aphids to settle on treated leaves significantly at least 24 hours that followed the application (ID 24 = 0.4), while (+)-(R)-γ-decalactone showed slight, but not statistically significant attractant properties (ID 24 = -0.1). Different response of aphids to pure enantiomers was probably the reason of the short-lived deterrent activity of the γ-decalactone (4a) racemate reported earlier. Neither the racemic δ-decalactone (4b) nor any of its enantiomers appeared to alter the host acceptance behavior by M. persicae. Aphid probing. The EPG recording revealed all kinds of aphid activities related to plant penetration: non-probing, pathway phase ('C') including the unidentified ('derailed') stylet movements 'F', phloem salivation and sap ingestion 'E1' and 'E2', respectively, and xylem sap uptake 'G'. 'F' and 'G' activities occurred sporadically irrespective of a treatment and were included in pathway phase for statistical analysis.
Despite individual variation within experimental groups (Fig 6), certain trends in aphid behavior could be observed. On control plants, aphid probing was rarely interrupted: there were twelve probes of 1.5 hour duration per aphid, on average during eight hours of continuous EPG recording. Consequently, plant penetration occupied 95% of experimental time and ca. 70% of that time was phloem phase that included 83% activities associated with sap ingestion. Sap ingestion periods were four hours long, on average, and more than 90% of aphids reached sieve elements within the 8-hour experiment, generally in 20 minutes within a successful probe. Phloem phase prevailed among probing activities from the second hour of the experiment onwards in the majority of aphids (Table 5, Figs 6-8). On racemic δ-decalactone (4b)treated plants, probing activities occupied ca. 93% of experimental time, 57% of probing was dedicated to phloem phase, and 41% to sap ingestion which was slightly, but not significantly, less than in aphids on control plants. There were ca. 16 probes per aphid, their mean duration was approximately less than one hour, and an average period of phloem phase was nearly two hours. Sap ingestion periods were 2.4 hour long on average. All aphids reached sieve elements  (Fig 7) and started ingestion usually in 20 minutes within a successful probe (Table 5). Moreover, phloem phase was the main aphid activity of nearly all experimental aphids on treated leaves (Fig 6), and it predominated over other activities from the second hour of the experiment onwards (Fig 8). In respect to the application of racemic γ-decalactone (4a), there was no significant decrease in the total duration of plant tissues penetration by aphids in comparison to control. However, aphid behaviour during probing in phloem vessels was altered: although the total duration of phloem phase was similar to that on control plants, the individual periods of sap ingestion were significantly shorter. The average phloem phase was 2.3 times shorter and mean duration of sustained ingestion was 3.5 times lower in aphids on treated plants than on control. Altogether, the total duration of sap ingestion during the 8-hour experiment was two times lower as compared to control although all aphids reached sieve elements (Table 5, Fig 6). Pathway activities (i.e., probing in non-phloem tissues) was a predominant activity during three successive hours after the onset of the experiment (Figs 6 and 8). Aphids on (-)-(S)γ-decalactone (4a)-treated leaves were not prevented from probing but their activities were limited mainly to penetration in non-phloem tissues. Total and average durations of phloem phase were 1.7 and 6.0 times shorter than on control leaves, respectively. Moreover, activities within phloem phase associated with sap ingestion were also constrained: the proportion of sap ingestion within phloem phase was 32%, while on control leaves 83% (Table 5). Pathway activities predominated during six hours of the 8-hour experiment in the majority of aphids and there were individuals (8% of the experimental group) that did not reach phloem phase during that time (Figs 6 and 7 and 8).
EPG-recorded signals provide an insight into the cryptic mode of aphid-plant interactions during probing. Due to the specific structure of sucking-piercing mouthparts, direct observation of aphid behaviour associated with feeding is impossible. Therefore, parameters derived from EPG experiments are good indicators of plant suitability for aphids. Additionally, the interpretation of the results shows the tissular localization of deterrent (or attractant) factors and, consequently, exposes the physiological effects in aphids. For example, long penetration time of non-phloem tissues as compared to total penetration time, high number of short vs. long probes before the first phloem phase, relatively long time to 1 st phloem phase within a probe, and a failure in finding sieve elements may be interpreted as preingestional effects of antifeedants that restrain aphid probing at the level of non-phloem tissues. Short (<10 min) probes are limited either to epidermis (<2 min.) or do not reach beyond mesophyll (2-10 min.). Similarly, the short total and mean durations of phloem sap ingestion and high proportion of salivation during penetration of phloem vessels may point to the ingestional level of feeding deterrence [32,33]. On suitable host plants, the sap ingestion periods may last for many hours with no interruption [34,35]. Supplementary to EPG experiments is the choicetest on aphid settling, which was also used in this work. In the case of compounds that are inactive at pre-and (or) ingestional levels, it can reveal postingestive activity if the compound deters aphid settling at least 24 hours after exposure. In our study, the application of δ-decalactone (4b), irrespective of its enantiomeric purity did not hinder aphid settling on leaves and it was confirmed in the EPG study. Aphid probes were relatively long and non-probing periods were relatively short and comparable to those on control plants. Likewise, the sap ingestion phases were relatively long and uninterrupted. In contrast, racemic γ-decalactone (4a) caused a negative response of aphids during settling. However, the effect was short-lived, which was further explained in the EPG study. It appeared that the enantiomeric purity was crucial in the expression and the durability of the deterrent properties. Only (-)-(S)-γ-decalactone (4a) had strong and durable (i.e. lasting for at least 24 hours) limiting effect, expressed at phloem level. Therefore, this compound (-)-4a can be considered an ingestive deterrent.
In summary, the deterrent activity of decalactones (4a and 4b) against aphids depended on the size of the lactone ring and the enantiomeric purity of the compounds. The decalactone (4b) containing 6-membered ring appeared inactive against M. persicae while the decalactone (4a) with 5-membered ring deterred aphid feeding at ingestional level. Only the levorotatory S isomer of 4a was active.