Raspberry Ketone Analogs: Vapour Pressure Measurements and Attractiveness to Queensland Fruit Fly, Bactrocera tryoni (Froggatt) (Diptera: Tephritidae)

Abstract

The Queensland fruit fly, Bactrocera tryoni (Froggatt) (Q-fly), is a major horticultural pest in Eastern Australia. Effective monitoring, male annihilation technique (MAT) and mass trapping (MT) are all important for control and require strong lures to attract flies to traps or toxicants. Lure strength is thought to be related in part to volatility, but little vapour pressure data are available for most Q-fly lures. Raspberry ketone (4-(4-hydroxyphenyl)-2-butanone) and analogs that had esters (acetyl, difluoroacetyl, trifluoroacetyl, formyl, propionyl) and ethers (methyl ether, trimethylsilyl ether) in replacement of the phenolic group, and in one case also had modification of the 2-butanone side chain, were measured for their vapour pressures by differential scanning calorimetry (DSC), and their attractiveness to Q-fly was assessed in small cage environmentally controlled laboratory bioassays. Maximum response of one category of compounds, containing both 2-butanone side chain and ester group was found to be higher than that of the other group of compounds, of which either of 2-butanone or ester functionality was modified. However, linear relationship between vapour pressure and maximum response was not significant. The results of this study indicate that, while volatility may be a factor in lure effectiveness, molecular structure is the dominating factor for the series of molecules investigated.

About the Authors:

Soo J. Park

Contributed equally to this work with: Soo J. Park, Renata Morelli

* E-mail: [email protected]

Affiliation: Department of Chemistry and Biomolecular Sciences, Macquarie University, North Ryde, NSW 2109, Australia

ORCID http://orcid.org/0000-0002-5414-6772

Renata Morelli

Contributed equally to this work with: Soo J. Park, Renata Morelli

Affiliations Department of Biological Sciences, Macquarie University, North Ryde, NSW 2109, Australia, CAPES Foundation, Ministry of Education of Brazil, Brasilia/DF 70040-020, Brazil

Benjamin L. Hanssen

Affiliation: Department of Chemistry and Biomolecular Sciences, Macquarie University, North Ryde, NSW 2109, Australia

Joanne F. Jamie

Affiliation: Department of Chemistry and Biomolecular Sciences, Macquarie University, North Ryde, NSW 2109, Australia

Ian M. Jamie

Affiliation: Department of Chemistry and Biomolecular Sciences, Macquarie University, North Ryde, NSW 2109, Australia

Matthew S. Siderhurst

Affiliation: Eastern Mennonite University, Department of Chemistry, 1200 Park Road, Harrisonburg, VA, 22802, United States of America

Phillip W. Taylor

Affiliation: Department of Biological Sciences, Macquarie University, North Ryde, NSW 2109, Australia

Introduction

Many tephritid fruit flies are economically important horticultural pests in tropical and subtropical regions [1]. The Queensland fruit fly (Q-fly), Bactrocera tryoni (Froggatt), is amongst the most polyphagous and destructive fruit fly pests, having a major impact on the horticultural industries of Eastern Australia, and has also invaded several Pacific Islands [2]. Q-fly is a major quarantine pest, leading to strict regulatory requirements and post-harvest treatment for horticultural exports from affected areas.

For decades, most crops have been effectively protected from Q-fly infestation by cover sprays of organophosphate insecticides. Recently, however, these insecticides have come under review in Australia, and their use is now substantially restricted, leading to an urgent need for improved surveillance and alternative control measures [3]. Attract-and-kill techniques, including the male annihilation technique (MAT) and/or mass trapping (MT), are often used to control fruit flies as part of a larger system approach and/or area-wide management strategy. MAT involves the use of a male lure [4] combined with a toxicant, contained in a device or carrier in a way that prevents insecticide contamination of the crop or the environment. Effective MAT reduces the male population to such a low level that a large proportion of females are unable to find mates and so do not infest crops [5]. For Q-fly, cuelure (4-(4-acetoxyphenyl)-2-butanone, CL) has been the best available attractant since the 1960s [6]. Attract-and-kill approaches have been used successfully to eradicate Q-fly from Para Nui (Easter Island) in the South Pacific through the combination of cuelure-based MAT and spot spraying with protein-malathion bait [7].

Lure attractiveness is a crucial factor for successful surveillance, MAT and MT, and considerable research effort has been committed to developing more attractive lures [8–10]. Cuelure is a derivative of the naturally occurring phenylpropanoid, raspberry ketone (4-(4-hydroxyphenyl)-2-butanone, RK) [11]. RK also attracts Q-fly [12], though with less efficacy than CL. To find more effective lures, chemical analogs of RK that might be more volatile, more readily detected, or more intrinsically attractive have been explored, especially using melon fly, Bactrocera cucurbitae, as a model ‘RK responding’ Bactrocera [8–10, 12, 13]. One such compound is the formate ester of raspberry ketone (4-(4-formyloxy)phenyl)-2-butanone, melolure, ML) which, given its smaller ester group, is expected to be more volatile than CL. ML has been reported to be around twice as effective as CL at attracting melon fly [9, 12, 14, 15]. Even amongst the RK-responding Bactrocera, there are quite substantial differences in relative efficacy of RK analogs [16, 17]. While positive effects for one RK-responding species can provide some general guidance, lures ultimately need to be tested for each species. Guided by the promising results for melon fly, several studies have considered ML as a potential Q-fly attractant [16, 17]. One study, carried out in tropical conditions of North Queensland, found ML to be more effective than CL as a Q-fly lure [17], but another study, carried out in warm temperate areas of New South Wales, found ML to be less effective than CL [16].

Vapour pressure affects the amount of a compound present in the atmosphere and so is expected to influence lure efficacy. By correlating release rates and captures of melon fly for a set of RK analogs, Metcalf and Metcalf suggested there is a linear relationship between release rate and attractiveness [12]. The release rate was measured by comparing weight of lure-loaded cotton dental wicks before and after two day of use in traps. While release rate of a compound is related to its vapour pressure, Metcalf and Metcalf’s study did not directly measure actual vapour pressures of the lure. To date, there are no published vapour pressure data for RK analogs that are used as lures for Bactrocera fruit flies, although the boiling points of a few lure compounds under reduced pressure are available, having been recorded during the purification step of their syntheses [6, 18]. Measuring vapour pressures of lure compounds directly could allow a more comprehensive analysis of how physical properties of the compounds are related to efficacy in fied applications.

We report herein vapour pressure data for RK and eight structurally related compounds. The analogs had the same core RK structure, i.e., an aromatic ring with an alkyl chain and oxygen at the para position, but had functional group modifications that have been chosen to increase volatility. The chosen functional groups were esters (acetyl, difluoroacetyl, trifluoroacetyl, formyl, propionyl) and ethers (methyl ether, trimethylsilyl ether) in replacement of the phenolic group, and an ester in replacement of the ketone group. We also report results of the attractiveness of the eight compounds to Q-fly males under controlled conditions and the relationship between vapour pressure and attractiveness.

Materials and Methods

Sample information

Sample information and the structures of lures are summarised in Table 1 and Fig 1, respectively. Compound purity was determined by a combination of GC-FID and NMR analysis.

[Figure omitted. See PDF.]

Table 1. Sample information.

https://doi.org/10.1371/journal.pone.0155827.t001

[Figure omitted. See PDF.]

Fig 1. Chemical Structures of the Attractants.

https://doi.org/10.1371/journal.pone.0155827.g001

Differential Scanning Calorimetry (DSC)

Differential scanning calorimetry (DSC) has been used to determine the vapour pressure of diverse compounds, for instance, alkyl phosphonates [19], ethyl esters [20], 2-dialkyl aminoethanethiols [21], fatty acids [22] and precursors of chemical warfare agents [23]. In DSC, the isothermal boiling temperature of a pure compound is measured as a function of pressure. The thermal equilibrium of boiling is readily obtainable within a dynamically heated environment, and hence the measured pressure is equivalent to the vapour pressure of the compound undergoing isothermal boiling [24]. The experimentally obtained boiling point-pressure data of a compound are then fitted to the Antoine equation to derive Antoine parameters, which can then be used to calculate the vapour pressure of the compound at any given temperature. The American Society for Testing and Materials International Method ASTM E1782-14 [25] describes the use of DSC with pinhole-pans with a lower pressure limit of 0.2 kPa. This technique has been successfully used in vapour pressure measurements for a range of compounds at high temperatures [19–23].

The DSC experiments were conducted on a TA 2010 DSC instrument (TA Instruments, New Castle, DE) equipped with a standard DSC cell. The instrument preparation and operational methods were based on the ASTM E 1782–14 guidelines [25]. Reduced pressure was achieved using either a diaphragm vacuum pump (Vaccubrand GMBH CO KB, Wertheim), or oil sealed high vacuum pump (Edwards, Crawley) and the pressure of the system was maintained by a downstream bleed obtained by adjusting a needle valve. The absolute pressure inside the cell was measured using a calibrated barometer (A.L. Franklin Scientific Instruments, Sydney, NSW) with a precision of ±0.07 kPa.

The DSC system was calibrated in accordance with ASTM E967-08 [26]. The temperature calibration of the cell was performed using indium (m.p. = 156.60°C) and lead (m.p. = 327.502°C). The boiling point of water was determined at 101 kPa to validate instrument accuracy for the vapour pressure measurements, and the observed boiling point of water was within 0.3 K of the literature value [27]. Melting and boiling point calibrations were conducted several times during the experiments.

Samples of 7 to 20 mg were weighed on a micro-analytical balance (Mettler Toledo International Inc, AG 285, Columbus, OH) with precision of ±0.01 mg and placed in hermetic aluminium pans (TA instruments, New Castle, DE). At pressures of 5 kPa or greater, samples were sealed with hermetic lids that had a pinhole of diameter 75 μm (TA instruments, New Castle, DE) to allow pressure equilibration between the sealed pan and the instrument cell, without a substantial loss of the lure due to vaporisation before the boiling endotherm was reached [28]. For pressures lower than 5 kPa, hermetic lids with a larger pinhole size [29, 30] were prepared by manually punching the lids with a thin needle (Metler-Toledo International Inc, Columbus, OH). The pinhole sizes were examined by microscopy (Olympus, SZX 12 microscope, Shinjuku, Tokyo) to confirm that they were between 150 and 250 μm diameter.

The system was evacuated to a desired pressure and heating was initiated at a rate of 5°C min^-1 from room temperature. Heating continued through the boiling point temperature at the same rate and stopped when the endothermic curve returned to a stable baseline.

DSC Data Analysis

To obtain the Antoine parameters for each compound, DSC data were fitted to the following Antoine equation, where P is pressure, A, B and C are the Antoine equation parameters and T is temperature.(1)The Antoine equation is derived from the following Clausius-Clapeyron equation, where ΔH_vap is the enthalpy of vaporisation, and R is the universal gas constant.(2)The enthalpy of vaporisation can be calculated using the following equation, which was derived by combining the Clausius-Clapeyron Eq (2) with the derivative of the Antoine Eq (1) with respect to T.

(3)

The Antoine parameters were determined for each of the compounds by minimizing the sum of the squares of the differences between log(P) of the measured and calculated values.

Bioassay General Procedures

Q-fly pupae were obtained from a colony established at the Fruit Fly Production Facility at Elizabeth Macarthur Agricultural Institute, New South Wales, Australia, maintained under standardised conditions [31, 32]. Pupae (10 g, ~ 700 flies) were allowed to emerge in mesh cages (45 × 45 × 45 cm). Flies were provided with ad libitum access to a full diet (3: 1; sucrose to yeast hydrolysate enzymatic) and water. Cages were maintained at 25 ± 1°C and 65 ± 5% relative humidity on a 13L:11D photocycle, in which the first and last 30 min of the light phase were changed to simulate dawn and dusk by gradually ramping up the light level to full output, and down to darkness, respectively. Male flies were tested when they were 12–13 days old [33, 34] and sorted into experimental cages (30 × 30 × 90 cm) one day before experiments. The experimental cages were modified from a previous design [35–37] to provide better images of flies attracted to a compound. The ceiling of each cage was made of white mesh. The walls and floor were made of black mesh and the floor had two clear plastic windows (10 × 10 cm) located 50 cm apart from each other for camera views.

Attraction of Q-fly males to RK analogs

Attraction studies followed the methods of previous studies [35, 38]. One hundred male Q-flies were transferred to each cage and provided water and food as described in the Bioassay General Procedure until testing.

Experiments were carried out during the first two hours of photophase under controlled environment conditions (25 ± 1°C and 65 ± 5% relative humidity). Flies were allowed to acclimatise in the room for 4 minutes before testing. Two discs of 55 mm filter paper folded into a cone shape, one treatment and one control, were placed 50 cm apart each other on the top of each cage with the tip of the cone upward. The tip of one cone was treated with 30 mg of one of the lures as a pure compound (directly dispensed by volume, adjusting for density). A cylindrical plastic container (11 cm diameter, 5 cm high) was placed over each cone to direct the odour plume into the cage.

Images of the area directly below the treated and untreated paper cone were captured using two digital cameras (Cisco WVC 80N IP, Cisco Corp., San Jose) positioned 6 cm below and located 50 cm apart such that each camera faced either treated or untreated paper cone. The distance of 6 cm between the cameras and cages provided coverage of an approximately 95 cm² treatment area inside the cylindrical plastic container (referred to as ‘treated area’ and untreated area’, depending on the location of treated and untreated paper cones). Images were captured every 5 seconds for 3 mins. Each lure was tested in two different cages, with the treated and untreated locations reversed. Each cage of flies was tested only once. To avoid contamination, the different lures were tested in separate rooms that had 18 fresh air changes per hour on independent air handing systems. Each lure was tested five times with different batches of flies.

To quantify the number of flies that were attracted to the treated and untreated areas, digital still images were processed using custom macros for threshold-based image analysis in ImageJ, modifying the approach of Manoukis and Jang [38] as described in Siderhurst et al [35]. In brief, for each experimental run, one image was randomly chosen for each of the treatments, cropped to fit the treated or untreated area, threshold adjusted so that areas corresponding to the presence of flies were extracted from background, and analysed using the ImageJ measure command. These operations were recorded in a custom macro, which was then used to batch process all images for a treatment within an experimental run. This method yielded a total area of contrast, this being the total area covered by flies in the image. From a sample of images, the area representing each individual fly (number of pixels) was calculated. To calculate the total number of flies in each image, the total area extracted from background was divided by the area representing an individual fly. A concordance analysis was carried out to validate the use of image analysis to count the flies. Twenty images from the experiments were randomly selected and the number of flies counted both visually and using ImageJ. As the data were not normally distributed, Spearman rank correlation was used and showed a high level of concordance between numbers of flies counted by eye from images and by threshold-based image analysis in ImageJ (ρ = 0.99; S = 18.6; P<0.01).

The response probability of a lure was calculated as the ratio of the number of flies in the treated area to the total number of flies in the treated and untreated area in each replicate. The averaged response probability as a function of time was fitted in OriginPro 9 with an exponential growth function of the following form, where t is time, p_max is the maximum response probability, A is the preexponential factor, and k represents a growth constant.(4)The data were fitted from 0 to 170 seconds, where t = 0 seconds defines the time that the treatment commenced. The curves were constrained to pass through p = 0.5 when t = 0 seconds, to represent the random distribution of flies at the time of release of the treated and untreated filter papers. Convergence was determined by the step-changed in the reduced-χ² being less than 1×10⁻⁹.

An analysis of covariance (ANCOVA) was performed to assess the relationship between maximum response probability and vapour pressure, where the lures were clustered in two groups according to their molecular structure. One group contained both the ester functional group and the 2-butanone side chain, and the other group contained a modified functional group at either of the ester or butanone functionalities. The same procedure was applied for maximum response probability and growth constant.

DSC data and the response probability of compounds were analysed using OriginPro9.1 and Excel. Concordance analysis and ANCOVA was performed using R (R Core Team, 2014).

Results

DSC results

The experimental vapour pressure data for the nine lure compounds, the values calculated using Antoine parameters derived from those data and percentage deviation from the calculated values are shown in tables Tables A-I in S1 Table. Plots of log of vapour pressure of each compound versus reciprocal of temperature, including the regression fit to the Antoine vapour pressure equation and the reported boiling points if available, are displayed in Fig 2. The Antoine equation parameters for each compound are listed in Table 2.

[Figure omitted. See PDF.]

Fig 2. Vapour Pressure of Nine Compounds.

■, data generated by DSC methodology in the present study; previously reported data were presented in each plot; -----, regression fit to the Antoine equation.

https://doi.org/10.1371/journal.pone.0155827.g002

[Figure omitted. See PDF.]

Table 2. Antoine Equation Parameters Determined by Non-Linear Regression Analysis of the Vapour Pressure Data and Their Validity Range.

https://doi.org/10.1371/journal.pone.0155827.t002

Calculated values of vapour pressure and enthalpy of vaporisation of the lure compounds at the extrapolated temperatures, T = (298.15) K, based on the Antoine equation parameters (Table 2) are shown in Table 3.

[Figure omitted. See PDF.]

Table 3. Vapour Pressure (P) and Enthalpy of Vaporisation (ΔH_vap) of Raspberry Ketone and Eight Analogs at 298.15 K.

https://doi.org/10.1371/journal.pone.0155827.t003

Attraction of Q-fly males to RK analogs

All the compounds were attractive to Q-fly males, as the response probabilities were significantly higher than the random distribution value of 0.5 (Fig 3). Through the course of the bioassay (0 to 170 seconds), there was an exponential relationship between the number of flies associated with each lure and time, with maximum attractiveness represented by an asymptote (0.97 > reduced r² > 0.57 for all fits) (Fig 3). Parameters of the fitted exponential curves are given in Table 4. In comparison to cuelure (CL), the standard commercial product, raspberry ketone trifluoroacetate (RKTA) had a higher maximum response probability; raspberry ketone difluoroacetate (DF), melolure (ML) and propionate of raspberry ketone (PRK) were similar to cuelure (CL); the TMS derivative of RK (TMSRK), anisyl acetone (AA) and methyl 3-(4-acetoxyphenyl)propionate (MAPP) showed a lower response.

[Figure omitted. See PDF.]

Fig 3. Fly Response Probabilities over Time.

The response of flies to eight lures were measured over 170 seconds. An exponential function was fitted. Plots of the residuals are given.

https://doi.org/10.1371/journal.pone.0155827.g003

[Figure omitted. See PDF.]

Table 4. Exponential Growth Curve Parameters for Fitting Response Probability (p_max ± 95% confidence interval (CI)) of Q-fly Males in Bioassays to the Function.

https://doi.org/10.1371/journal.pone.0155827.t004

The obtained fit growth constant (k) was lower for ML, RKTA and CL than for DF, TMSRK and PRK. Reflecting the large confidence intervals resulting from comparatively poor model fits the growth constants for AA and MAPP were not significantly different to those of any other compounds (Table 4).

ANCOVA (overall model r² = 0.90) did not detect a significant relationship between vapour pressure and maximum response (F_1,4 = 4.91, P = 0.09). However, the two clusters into which tested compounds were categorized according to their molecular structure differed significantly in maximum response (F_1,4 = 64.02, P<0.01; Fig 4). There was no evidence of a difference between the clusters in the slope of the relationship between vapour pressure and maximum response (cluster x vapour pressure F_1,4 = 0.95, P = 0.39).

[Figure omitted. See PDF.]

Fig 4. Maximum Response Probability as a Function of Vapour Pressure.

CL: 4-(4-acetoxyphenyl)-2-butanone, DF: 4-(4-(2,2-difluoroacetoxy)phenyl)-2-butanone, RKTA: 4-(4-(2,2,2-trifluoroacetoxy)phenyl)-2-butanone, ML: 4-(4-formyloxyphenyl)-2-butanone, AA: 4-(4-methoxyphenyl)-2-butanone, TMSRK: 4-(4-((trimethylsilyl)oxy)phenyl)-2-butanone, PRK: 4-(4-propionyloxyphenyl)-2-butanone, MAPP: methyl 3-(4-acetoxyphenyl)propionate.

https://doi.org/10.1371/journal.pone.0155827.g004

Discussion

Vapour pressure measurement by DSC has proven useful in the determination of thermodynamic properties (vapour pressure and enthalpy of vaporisation) of the tested compounds, providing accurate data that are in good agreement with the available literature values [18, 39–41]. The average differences between observed and calculated vapour pressures (S1 Table) for the compounds in this study were 0.64% (CL), 0.72% (DF), 0.85% (RKTA), 0.74% (ML), 0.80% (AA), 0.94% (TMSRK), 0.66% (PRK), 0.69% (MAPP) and 0.95% (RK).

For CL, the boiling point reported by Winter [42] (107.5–109°C at ca. 0.008 kPa) is 23.5% below that predicted by the Antoine equation obtained in the present study. The manufacturer’s reported boiling point (Aldrich and TCI, 122–123°C at ca. 0.03 kPa) is 6.3% lower, a variation that would not be outside the manufacturer’s experimental uncertainty (Fig 2). For RK, a broad boiling point range (190–210°C at 2.4 kPa) reported by Berlin et al. [18] is 30.9% (calculated using an average of the boiling point range) greater than the Antoine equation prediction, suggesting that the applied pressure in that study was considerably higher than the reported value. Two other boiling point ranges (140–146 and 140–155°C) [40, 43] under reduced pressure (0.07 kPa) deviated from the Antoine equation by 16.4% and −13.8%, respectively (Fig 2). There are 11 reported boiling points of AA with an indication of applied pressure in the literature. These data were compared to the current Antoine equation predictions and showed a range of deviations, although six of the boiling points were in agreement (Fig 2) [18, 39, 41, 44–46]. The other data deviated significantly, indicating the applied pressures were different from those reported in those studies [40, 47–49]. A boiling point for TMSRK has been reported previously [50], but is different from the Antoine equation prediction (Fig 2), suggesting the applied pressure was considerably lower than the reported value. Most data in the literature were obtained during vacuum distillation purification in which applied pressure measurements can often be inaccurate. To date however, vapour pressure data of DF, RKTA, ML, MAPP and PRK have not been reported in the literature. It was expected that changing the hydroxyl group of RK to methyl, trimethylsilyl or ester groups would increase volatility, and that the trifluoroacetate of RK (RKTA), having a trifluoromethyl group, would have a higher vapour pressure than non-fluorinated esters. Indeed, the vapour pressure of RKTA was found to be the highest amongst the tested compounds. The methoxy moiety in AA and trimethylsilyl in TMSRK also significantly increased volatility, being the second and third most volatile compounds, respectively. Difluoroacetate of RK (DF) was found to be significantly less volatile than RKTA as anticipated. The vapour pressures of the propionate of RK (PRK) and methyl 3-(4-acetoxyphenyl)propionate (MAPP) were found to be slightly higher than RK as anticipated, but significantly lower than RKTA or DF. Overall, the order of vapour pressure data was consistent with predictions (Table 3).

The DSC method infers boiling temperatures at a variety of controlled pressures and hence the uncertainty in the vapour pressure data is dominated by the uncertainty in boiling onset (±1 K). The temperature accuracy of the instrumentation was determined by measuring the melting point of two standard metals in the range of interest for the compounds and the pressure accuracy was validated by measuring the boiling temperatures of a standard at various pressure values. The difference in vapour pressure that would result from a 1 K change in temperature can be readily calculated and gives the estimated uncertainty in the measured values of the vapour pressure within the range of the study. The uncertainties of the present data ranged from ±2.3 to ±3.3% at the highest temperature and from ±3.2 to ±4.2% at the lowest temperature of the compounds. Although the errors of the DSC data in the present study are acceptable, relatively larger variance was found at low temperatures. This could reflect a range of factors, such as curve broadening due to use of sub-optimal pinhole size, impurities and thermal decomposition, although these factors were minor contributors in the present study (addressed in Methods).

The bioassay results revealed that all the lures were attractive to Q-fly males. Furthermore, there was a distinct relationship between chemical structure and the maximum response. The p_max values of the compounds are clustered into two categories based on chemical structure. One category with the higher maximum response contains both the ester functional group and the 2-butanone side chain. The other with lower maximum response contains a modified functional group at either of the ester or butanone functionalities. This pattern indicates that, not unexpectedly, the lure efficacy is linked to appropriateness of structural moieties across the molecule, not only in one position. Within this constraint, increasing the release rate of a lure may improve its efficacy [12].

It has been suggested that the 2-butanone side chain is a primary constituent for attractiveness [12, 51]. Drew demonstrated that the 2-butanone side chain was a vital element for lure activity towards Q-fly males. In Drew’s study, twice as many flies were attracted to 2-butanone compared with CL, whereas fewer flies (0.7 times) were attracted to phenyl acetate compared with CL [51]. The response of dacine fruit flies to RK released by Bulbophyllum spp. orchids has been broadly reported [52, 53], and 2-butanone containing compounds have been attributed to the evolutionary adaptation of their receptor attuned to those phenylpropanoids [54]. Mature females of Q-fly are known to have three olfactory receptor cell types, one of which recognizes 2-butanone [55]. As males responded to all the compounds with 2-butanone chain that were tested by Drew [51], males should have an olfactory recognition system for the 2-butanone chain as well. However, any behavioural response follows olfactory receptor neuron response that is in turn, induced by sensitization of the odorant receptors in the neural membrane [56] that are activated by the conformation shift upon ligand binding to the odorant binding protein [57, 58]. The same binding protein can change the conformation depending on the ligand [59] so that stronger or weaker neural responses can be obtained from compounds with similar structure [60]. Apart from MAPP, both TMSRK and AA were less attractive to Q-fly than all the other compounds, which contain an ester group. This suggests that an ester group installed at the hydroxyl group of RK may play an important role in olfactory recognition, rather than the ester group simply increasing volatility, because the measured vapour pressures of TMSRK and AA were higher than most of the other compounds, except RKTA.

CL has been relied on as a fruit fly lure for decades, but there remains a need for more attractive lures. Illustrating the usefulness of a strong attractant, methyl eugenol is much more effective than CL for Bactrocera species that respond to this lure, and as a result the available tools for monitoring and control of methyl eugenol-responding species are more potent than those available for RK-responding species [61], i.e., a lure of potency comparable to methyl eugenol is currently not available for RK-responding flies, such as Q-fly. RKTA may be such a candidate, as it showed higher maximum response than CL in the present study and outcompetes both CL and ML in head-to-head competitive bioassays [35]. RKTA was similar to CL in speed of response and had by far the highest vapour pressure of all RK analogs tested in the present study. RKTA is an ester with a high volatility that makes the molecule more versatile as the ester functionality may play a role in fly olfactory recognition. However, RKTA is prone to hydrolysis [62], diminishing its superior performances, and hence an adequate formulation may be required to deliver a great potential for field applications. Under some conditions, ML may offer some advantages over CL [16]. In the present study, ML was found to have higher vapour pressure than CL but in bioassays it was very similar to CL in the maximum response and the response rate of flies. Because of the greater distance over which flies need to be attracted, differences in vapour pressure may yield stronger differences in capture efficacy when tested in field settings. A recent field study carried out in warm temperate conditions of New South Wales found no advantage in using ML rather than CL to monitor Q-fly [17]. However, another field study in tropical Queensland found that six Bactrocera species were significantly more attracted to ML than CL, including Q-fly and two other economically important species (B. frauenfeldi and B. bryoniae) [16]. Further research is needed to clarify the conditions under which ML might serve as a useful complement or alternative to CL in Q-fly management programs.

In summary, we have experimentally determined vapour pressures of nine RK-related lures and have found good agreement between vapour pressure data obtained and fit to the Antoine equations. The vapour pressures of compounds were consistent with predictions based on their modified functionalities. The attraction to Q-fly to eight RK analogs was tested in cages under controlled environment conditions to investigate the relationship between vapour pressure and attractiveness. All analogs were attractive to Q-fly. Amongst the tested compounds, molecular structure was the more important factor than vapour pressure in determining response. Within groupings of compounds with closely related structure, vapour pressure had only a small effect on lure efficacy in bioassays. The vapour pressure data determined in the present study will guide future investigation in this field as vapour pressure is a fundamental property of a chemical that can be used as a reference in a range of applications, such as quantitative assays in gas phase or formulation.

Supporting Information

[Figure omitted. See PDF.]

S1 Supporting Information. Synthetic Procedure and NMR Data.

https://doi.org/10.1371/journal.pone.0155827.s001

(DOCX)

S1 Table. Observed and Calculated Vapour Pressure Data for the Nine Lures.

https://doi.org/10.1371/journal.pone.0155827.s002

(DOCX)

Acknowledgments

We would like to thank Dr Christopher McRae for providing invaluable technical advice.

Author Contributions

Conceived and designed the experiments: SJP RM MSS PWT. Performed the experiments: SJP RM BH. Analyzed the data: SJP RM IMJ MSS PWT. Contributed reagents/materials/analysis tools: PWT JFJ. Wrote the paper: SJP RM IMJ MSS PWT.

Citation: Park SJ, Morelli R, Hanssen BL, Jamie JF, Jamie IM, Siderhurst MS, et al. (2016) Raspberry Ketone Analogs: Vapour Pressure Measurements and Attractiveness to Queensland Fruit Fly, Bactrocera tryoni (Froggatt) (Diptera: Tephritidae). PLoS ONE 11(5): e0155827. https://doi.org/10.1371/journal.pone.0155827

References

1. White IM, Elson-Harris MM. Fruit flies of economic significance: Their identification and bionomics. Oxon, UK.: CAB International; 1994.

2. Clarke AR, Powell KS, Weldon CW, Taylor PW. The ecology of Bactrocera tryoni (Diptera: Tephritidae): What do we know to assist pest management? Ann Appl Biol. 2011;158(1):26–54.

3. APVMA. Listing of chemical reviews: Australian Pesticides and Veterinary Medicines Authotiry; 2015 [cited 2015 September 24]. Available from: http://apvma.gov.au.

4. Howlett FM. VII. The effect of oil of citronella on two species of Dacus. Trans Entomol Soc Lond. 1912;60(2):412–8.

5. El-Sayed AM, Suckling DM, Byers JA, Jang EB, Wearing CH. Potential of “lure and kill” in long-term pest management and eradication of invasive species. J Econ Entomol. 2009;102(3):815–35. pmid:19610395

6. Beroza M, Alexander BH, Steiner LF, Mitchell WC, Miyashita DH. New synthetic lures for the male melon fly. Science. 1960;131(3406):1044–5. pmid:17808098

7. Batemam MA, Insungza V, Arreta P. The eradication of Queensland fruit fly from Easter Island. FAO Plant Prot Bull. 1973;21:114.

8. Metcalf RL, Metcalf ER, Mitchell WC. Benzyl acetates as attractants for the male oriental fruit fly, Dacus dorsalis, and the male melon fly, Dacus cucurbitae. Proc Natl Acad Sci. 1986;83(6):1549–53. pmid:16593663

9. Casaña-Giner V, Oliver JE, Jang EB, Carvalho L, Khirmian A, Demilo AB, et al. Raspberry ketone formate as an attractant for the melon fly (Diptera: Tephritidae). J Entomol Sci. 2003;38:120–6.

10. Casaña-Giner V, Oliver JE, Jang EB. Syntheses and behavioral evaluations of fluorinated and silylated analogs of raspberry ketone as attractants for the melon fly, Bactrocera cucurbitae (Coquilett). J Entomol Sci. 2003;38:111–9.

11. Bauer L, Birch A, Ryan A. Studies in relation to biosynthesis. VI. Rheosmin. Aust J Chem. 1955;8(4):534–8.

12. Metcalf RL, Metcalf ER. Plant kairomones in insect ecology and control. New York: Champmon and Hall; 1992.

13. Metcalf RL, Mitchell WC, Metcalf ER. Olfactory receptors in the melon fly Dacus cucurbitae and the oriental fruit fly Dacus dorsalis. Proc Natl Acad Sci. 1983;80(11):3143–7. pmid:16593321

14. Vargas RI, Piñero JC, Jang EB, Mau RFL, Stark JD, Gomez L, et al. Response of melon fly (Diptera: Tephritidae) to weathered SPLAT-spinosad-cue-lure. J Econ Entomol. 2010;103(5):1594–602. pmid:21061958

15. Shelly TE, Kurashima RS, Nishimoto JI. Field capture of male melon flies, Bactrocera cucurbitae (Coquillett), in Jackson traps baited with cue-lure versus raspberry ketone formate in Hawaii. P Hawaii Entomol Soc. 2012;44:63–70.

16. Royer JE. Responses of fruit flies (Tephritidae: Dacinae) to novel male attractants in north Queensland, Australia, and improved lures for some pest species. Aust J Entomol. 2015;54(4):411–26.

17. Dominiak BC, Campbell AJ, Jang EB, Ramsey A, Fanson BG. Field evaluation of melolure, a formate analogue of cuelure, and reassessment of fruit fly species trapped in Sydney, New South Wales, Australia. J Econ Entomol. 2015;108(3):1176–81. pmid:26470243

18. Berlin AY, Sherlin SM, Serebrennikova TA. Derivatives of zingerone. II. Zh Obshch Khim. 1949;19:569–76.

19. Butrow AB, Buchanan JH, Tevault DE. Vapor pressure of organophosphorus nerve agent simulant compounds. J Chem Eng Data. 2009;54(6):1876–83.

20. Akisawa Silva LY, Matricarde Falleiro RM, Meirelles AJA, Krähenbühl MA. Determination of the vapor pressure of ethyl esters by differential scanning calorimetry. J Chem Thermodyn. 2011;43(6):943–7.

21. Williams BR, Hulet MS, Brozena A, Miles RW, Tevault DE. Vapor pressure of 2-dialkyl aminoethanethiols. J Chem Eng Data. 2013;58(6):1679–84.

22. Matricarde Falleiro RM, Akisawa Silva LY, Meirelles AJA, Krähenbühl MA. Vapor pressure data for fatty acids obtained using an adaptation of the DSC technique. Thermochim Acta. 2012;547:6–12.

23. Brozena A, Tevault DE, Irwin K. Vapor pressure of thiodiglycol. J Chem Eng Data. 2014;59(2):307–11.

24. Seyler RJ. Parameters affecting the determination of vapor pressure by differential thermal methods. Thermochim Acta. 1976;17(2):129–36.

25. ASTM Standard E1782-14, Standard test method for determining vapour pressure by thermal analysis, ASTM International, West Conshohocken, PA, 2014.www.astm.org.

26. ASTM Standard E967-08(2014), Standard test method for temperature calibration of differential scanning calorimeters and differential thermal analyzers, ASTM International, West Conshohocken, PA, 2014.www.astm.org.

27. Wexler A. Vapor pressure formulation for water in range 0 to 100°C. A revision. J Res Natl Bur Stand. 1976;80A(3):775–85.

28. Barrall EM II. Precise determination of melting and boiling points by differential thermal analysis and differential scanning calorimetry. Thermochim Acta. 1973;5(4):377–89.

29. Brozena A. Vapor pressure of 1-octanol below 5 kPa using DSC. Thermochim Acta. 2013;561(0):72–6.

30. Butrow AB, Seyler RJ. Vapor pressure by DSC: Extending ASTM E 1782 below 5 kPa. Thermochim Acta. 2003;402(1–2):145–52.

31. Dominiak BC, Sundaralingam S, Jiang L, Jessup AJ, Barchia IM. Production Levels and Life History Traits of Mass Reared Queensland Fruit Fly Bactrocera tryoni (Froggatt) (Diptera: Tephritidae) during 1999/2002 in Australia. Plant Proct Q. 2008;23(3):131–5.

32. Dominiak BC, Sundaralingam S, Jiang L, Nicol HI. Longevity of Mass-Produced Bactrocera tryoni (Diptera: Tephritidae) Held without Food or Water. J Econ Entomol. 2014;107(6):2103–6. pmid:26470075

33. Perez-Staples D, Prabhu V, Taylor PW. Post-teneral protein feeding enhances sexual performance of Queensland fruit flies. Physiol Entomol. 2007;32(3):225–32.

34. Shelly T. Effects of methyl eugenol and raspberry ketone/cue lure on the sexual behavior of Bactrocera species (Diptera: Tephritidae). Appl Entomol Zool. 2010;45(3):349–61.

35. Siderhurst MS, Park SJ, Buller CN, Jamie IM, Manoukis NC, Jang EB, et al. Raspberry ketone trifluoroacetate, a new attractant for the Queensland fruit fly, Bactrocera tryoni (Froggatt). J Chem Ecol. 2016;42(2):156–62. pmid:26922349

36. Siderhurst MS, Jang EB. Attraction of female Oriental fruit fly, Bactrocera dorsalis, to Terminalia catappa fruit in wind tunnel and olfactometer tests. Form Entomol. 2006;26:45–55.

37. Khrimian A, Siderhurst MS, Mcquate GT, Liquido NJ, Nagata J, Carvalho L, et al. Ring-fluorinated analog of methyl eugenol: Attractiveness to and metabolism in the oriental fruit fly, Bactrocera dorsalis (Hendel). J Chem Ecol. 2009;35(2):209–18. pmid:19198948

38. Manoukis NC, Jang EB. The diurnal rhythmicity of Bactrocera cucurbitae (Diptera: Tephritidae) attraction to cuelure: Insights from an interruptable lure and computer vision. Ann Entomol Soc Am. 2013;106(1):136–42.

39. Drishaus I, inventor. Hydrogenation of α,β-unsaturated ketones. German patent DE 828244. 1952 Jan 17.

40. Tong ST, inventor. Method of producing 1-(4-methoxy-phenyl)-3-butanone and 1-(4-hydroxy-phenyl)-3-butanone. United Kingdom patent GB 1094417-A 1967 Dec 13.

41. Mori K. A new synthetic method of amino acids. II-III. II. Synthesis of tyrosine, 3,4-dihydroxyphenylalanine and phenylglycine. Nippon kagaku zasshi. 1960;81(3):464–7.

42. Winter M. Odeur et constitution XIX. Sur des homologues et analogues de la p-hydroxyphényl-1-butanone-3 («cétone de framboise»). Helv Chim Acta. 1961;44(7):2110–21.

43. Bunce RA, Reeves HD. Amberlyst-15 catalyzed addition of phenols to α,β-unsaturated ketones. Synthetic Commun. 1989;19(5–6):1109–17.

44. Goodall GD, Haworth RD. CCCXXVII.-The structure of isoanethole. J Chem Soc. 1930(0):2482–7.

45. Sawa Y, Masuda T, Hori M, Fujimura H. Studies on the syntheses of analgesics. VII. Syntheses of 1,2,3,4-tetrahydro-4,4-dimethylisoquinoline derivatives. Yakuga Zasshi. 1976;96(4):401–6.

46. Citterio A, Vismara E. 2-Arylalkyl ketones and 3-arylalkanals from arenediazonium salts and α,β-unsaturated carbonyl compounds. Synthesis. 1980;1980(04):291–2.

47. Barbero M, Cadamuro S, Dughera S. Arenediazonium o-benzenedisulfonimides in Heck-type arylation of allylic alcohols. Synthesis. 2006;2006(20):3443–52.

48. Yamashita M, Watanabe Y, Mitsudo T-a, Takegami Y. The reductive deacylation of the Knoevenagel condensates by the use of the tetracarbonylhydridoferrate(0) anion. A new synthetic route of ketones from aldehydes. B Chem Soc Jpn. 1978;51(3):835–8.

49. Krubsack AJ, Sehgal R, Loong W-A, Slack WE. Oxidations by thionyl chloride. Mechanism of 3-thietanone formation. J Org Chem. 1975;40(22):3179–82.

50. Saito T, Nishimoto Y, Yasuda M, Baba A. InCl3/I2-catalyzed cross-coupling of alkyl trimethylsilyl ethers and allylsilanes via an in situ derived combined Lewis acid of InCl3 and Me3SiI. J Org Chem. 2007;72(22):8588–90. pmid:17914847

51. Drew RAI. Behavioural strategies of fruit flies of the genus Dacus (Diptera: Tephritidae) significant in mating and host-plant relationships. B Entomol Res. 1987;77(01):73–81.

52. Tan K-H, Nishida R, Toong Y-C. Floral synomone of a wild orchid, Bulbophyllum cheiri, lures Bactrocera fruit flies for pollination. J Chem Ecol. 2002;28(6):1161–72. pmid:12184394

53. Tan K-H, Nishida R. Synomone or kairomone?–Bulbophyllum apertum flower releases raspberry ketone to attract Bactrocera fruit flies. J Chem Ecol. 2005;31(3):497–507. pmid:15898497

54. Raghu S. Functional significance of phytochemical lures to dacine fruit flies (Diptera: Tephritidae): An ecological and evolutionary synthesis. B Entomol Res. 2004;94(05):385–99.

55. Hull CD, Cribb BW. Olfaction in the Queensland fruit fly, Bactrocera tryoni. I: Identification of olfactory receptor neuron types responding to environmental odors. J Chem Ecol. 2001;27(5):871–87. pmid:11471941

56. Leal WS. Odorant reception in insects: Roles of receptors, binding proteins, and degrading enzymes. Annu Rev Entomol. 2013;58(1):373–91.

57. Stowers L, Logan DW. LUSH shapes up for a starring role in olfaction. Cell. 2008;133(7):1137–9. pmid:18585346

58. Laughlin JD, Ha TS, Jones DNM, Smith DP. Activation of pheromone-sensitive neurons is mediated by conformational activation of pheromone-binding protein. Cell. 2008;133(7):1255–65. pmid:18585358

59. Zhou J-J, Robertson G, He X, Dufour S, Hooper AM, Pickett JA, et al. Characterisation of Bombyx mori odorant-binding proteins reveals that a general odorant-binding protein discriminates between sex pheromone components. J Mol Biol. 2009;389(3):529–45. pmid:19371749

60. Nikonov AA, Peng G, Tsurupa G, Leal WS. Unisex pheromone detectors and pheromone-binding proteins in scarab beetles. Chem Senses. 2002;27(6):495–504. pmid:12142325

61. Steiner LF, Mitchell WC, Harris EJ, Kozuma TT, Fujimoto MS. Oriental fruit fly eradication by male annihilation. J Econ Entomol. 1965;58(5):961–4.

62. Park SJ, Siderhurst MS, Jamie IM, Taylor PW. Hydrolysis of Queensland fruit fly, Bactrocera tryoni (Froggatt), attractants: kinetics and implications for Biological activity. Aust J Chem. 2016:In Press.