A simple and sensitive approach to quantify methyl farnesoate in whole arthropods by matrix-solid phase dispersion and gas chromatography–mass spectrometry

1 A simple and sensitive approach to quantify methyl farnesoate in whole arthropods by matrix-solid phase dispersion and gas chromatography-mass spectrometry Rosa Montes* a, Rosario Rodil a, Teresa Neuparth b, Miguel M. Santos b,c, Rafael Cela a, José Benito Quintana* a a Department of Analytical Chemistry, Nutrition and Food Sciences, IIAA Institute for Food Analysis and Research, University of Santiago de Compostela, Constantino Candeira S/N, Santiago de Compostela, Spain b CIMAR/CIIMAR – Interdisciplinary Centre of Marine and Environmental Research, Endocrine Disruptors and Emerging Contaminants Group, University of Porto, Avenida General Norton de Matos, S/N, Matosinhos, Portugal c FCUP - Department of Biology, Faculty of Sciences, University of Porto, Rua do Campo Alegre, 4169-007 Porto, Portugal *Corresponding authors: RM: [email protected]; Tel.: +34 881816035 JBQ: [email protected]; Tel.: +34 881816035 Abstract Methyl farnesoate (MF) is an arthropod hormone that plays a key role in the physiology of several arthropods’ classes being implicated in biological processes such as molting and reproduction. The development of an analytical technique to quantify the levels of this compound in biological tissues can be of major importance for the field of aquaculture/apiculture conservation and in endocrine disruption studies. Therefore, the aim of this study was to develop a simple and sensitive method to measure native levels of MF in the tissue of three representative species from different arthropods classes with environmental and/or economic importance. Thus, a new approach using whole organisms and the combination of matrix solid-phase dispersion with gas chromatography coupled to mass spectrometry was developed. This method allows quantifying endogenous MF at low levels (LOQs in the 1.2-3.1 ng/g range) in three arthropod species, and could be expanded to additional arthropod classes. The found levels ranged between 2 and 12 ng/g depending on the studied species and gender. The overall recovery of the method was evaluated and ranged between 69 and 96%. This is the postprint(accepted manuscript) version of the article published in Journal of Chromatography A https://doi.org/10.1016/j.chroma.2017.06.001 This manuscript version is made available under de CC-BY-NC-ND 4.0 license hppt://creativecommons.org/licenses/by-nc-cd/4.0 2 Keywords: sample preparation, gas chromatography, matrix solid-phase dispersion 3 1. Introduction Methyl Farnesoate (MF), one of the most important hormones in crustaceans and insects, is an isoprenoid structurally similar to insect juvenile hormone III (JHIII). This molecule is biosynthesized through the mevalonate pathway being the enzyme 3-hydroxy-3-methyl-glutaryl-CoA reductase (HMGR), the rate-limiting step in the production of MF [1]. MF plays central roles in the regulation of 5 crustacean development and reproduction [2-3]. It has been demonstrated that MF is implicated in crustaceans’ reproductive maturation by increasing the production of vitellogenin and neonates [4-5], or by stimulation of gonadal development and maturation [4]. These aforementioned studies showed strong and direct evidences that MF is a reproductive hormone in crustaceans. Therefore, by increasing the levels of MF in cultured crustaceans (injecting the MF directly into the animals, or 10 supplementing MF through the food) the productivity of crustacean aquaculture can be improved. Additionally, as MF is one of the major hormones of crustaceans, it is susceptible to be disrupted by environmental chemicals. A potential disturbance of the mevalonate signaling pathway by hypocholesterolaemic pharmaceuticals (statins) in arthropods has been associated with the reduction of reproductive performance in the amphipod Gammarus locusta and in the German cockroach 15 Blatella germanica [6-7] supporting the view that statins can lead to a reduction of MF levels by inhibiting HMGR. Therefore, the quantification of MF in arthropods can be used as a proxy to evaluate the effects of these endocrine disrupting chemicals. MF has been measured in different species of crustacean species [8-9] in several tissues such as hepatopancreas and mandibular organ and, specially, in hemolymph [8]. The presence of MF has 20 been also described in insects [8, 10], but in this case, the main role of MF is acting as intermediate in the JHIII synthesis that takes place in the corpus allatum. Although most of the descriptive works had located MF only in this tissue [8], recently, its presence in hemolymph in higher amount has been described [10]. Whereas there are some authors who have measured MF by high performance liquid 25 chromatography (HPLC) using ultraviolet (UV) detection [11-12], the reference technique, due to its higher selectivity is gas chromatography coupled to mass spectrometry (GC-MS) working in selected 4 ion monitoring (SIM) mode [13-15]. Moreover, chemical ionization (CI) has been the most used ionization technique for the analysis of this biomolecule. The use of isobutane, instead of the most frequently used methane, as reagent gas in GC-CI-MS has been described since it allows a softer 30 ionization giving the molecular ion [M+H]+ as the base peak of the spectrum (m/z 251). [10] In most studies describing the determination of MF in several organisms a previous extraction of hemolymph is mandatory [10-11, 16-17]. This step led to tedious and time-consuming protocols since a large amount of hemolymph, and consequently a large number of animals, is needed to proceed with the quantification of MF. Thus, by introducing the use of matrix solid-phase dispersion (MSPD) 35 this previous step can be avoided since the whole organism is used. MSPD is well-known in the environmental field, as extraction technique in the analysis of pollutants in a wide range of biota matrices [18]. This technique encompasses in the same step both extraction and purification simultaneously. This work describes the development and validation of a MSPD protocol for the direct extraction of 40 MF from three species of crustaceans and insects, including Bees that have a key ecological and economic role and have been facing a dramatic decline in several regions [19]. The final protocol can be applied in the quantification of MF in whole animals and can be very useful in conservation biology, ecotoxicological studies of endocrine disrupting chemicals, but also for the field of aquaculture and apiculture. 45 2. Material & Methods 2.1. Reagents and standards 50 Acetonitrile, ethyl acetate and n-hexane for analyses were purchased from Merck (Darmstadt, Germany). Bulk PSA (50 µm particle size) used as sample dispersant and Florisil (60–100 mesh) used as co-sorbent in the MSPD process were provided by Sigma-Aldrich (Milwaukee, WI, USA). Silica gel 60 (0.040-0.063 nm) tested also as co-sorbent was acquired from Merck. Empty solid-phase extraction polypropylene cartridges (15 mL volume) and polyethylene frits (20 μm) were purchased 55 5 from International Sorbent Technology (Mid Glamorgan, UK). Standards of MF and methyl heptadecanoate-d33 (as internal standard, I.S.) were acquired from Echelon Biosciences (Salt Lake City, UT, USA) and Sigma-Aldrich, respectively. 2.2. Samples and sample preparation 60 The individuals used for method development were provided by CIIMAR (Portugal). The selected species were Gammarus locusta (GL) (from CIIMAR culture, derived from field animals collected in Sado estuary, Portugal) and Artemia franciscana (AF) (GSL Artemia cysts purchased from Ocean Nutrition, CA, USA), two model crustaceans frequently used in ecotoxicological/aquaculture studies. Moreover, the ecological and economic relevant insect Apis mellifera (AM) (sampled in the north of 65 Portugal) was also studied. Each sample of GL and AM consisted of 3 individuals. This sample size corresponded to an average weight of 0.302 g in the case of male GL, 0.319 g for AM, and 0.147 g in the case of female GL. The AF samples contained 0.1 g of pooled 48h nauplii. All the samples were frozen and stored at -80 ºC until analysis. A scheme of the complete extraction protocol is shown in Figure 1. In the final protocol, the samples 70 were mixed with 0.5 g of PSA and dispersed in a glass mortar, with a pestle, until a visually homogeneous mixture was obtained. Twenty microliters of methyl heptadecanoate-d33 as I.S. (50 µg/mL in ethyl acetate) were added. Then, the blend was loaded into a cartridge containing, consecutively, a polyethylene frit, 0.5 g of sodium sulphate anhydrous and 1.5 g of florisil. A second frit was placed over the dispersed sample and compressed. MF was eluted by gravity with 1.5 mL of 75 ethyl acetate. The extracts were evaporated to dryness using a gentle nitrogen stream and reconstituted using 20 µL of ethyl acetate. The final extract was transferred to a 150 µL insert and directly injected in the GC-CI-MS system. 2.3 Instrumentation and determination conditions 80 MF was determined in a 7890A gas chromatograph combined with a 5975C quadrupole mass spectrometer from Agilent Technologies (Palo Alto, CA, USA), operated in positive CI mode with 6 isobutane as reagent gas [10]. The column, also supplied by Agilent was a HP-5MS (30 m × 0.25 mm i.d., d.f.: 0.25 μm). One microliter of extract or standard was injected in the splitless mode for 1 minute. Helium was used as carrier gas at a constant flow rate of 1 mL/min. The temperature set for injector 85 and transfer line was 280ºC, whereas MS source and quadrupole were maintained at 300ºC and 150ºC, respectively. The oven temperature was maintained at 90ºC during 1 minute and increased to 280ºC at 10ºC/min (hold for 10 min). An initial solvent delay of 11.5 min was used. MF was measured in SIM mode using m/z 251 protonated molecular ion for quantification and m/z 219 and 191 as qualifier ions. These main fragment ions represent the loss of CH3OH (m/z 219) 90 followed by the loss of CO (m/z 191), representatives of a methyl ester. The internal standard methyl heptadecanoate-d33 was measured using the m/z 317.5 as quantification ion and 267.5 and 282.5 as qualifier ions. 2.4 Recovery calculation and samples quantification 95 Calculations of the recovery were made for each type of matrix, male and female GL, AF and AM. Three MSPD replicates of pooled real individuals belonging each class were processed without (unspiked samples) and with addition of MF (spiked samples). A recovery factor was estimated using an internal standard calibration curve by subtracting the mean blank concentration. Then, when quantifying real samples, the concentration was also calculated using the internal standard calibration 100 curve and applying the recovery factor. 3. Results and discussion 3.1 Optimization of MSPD parameters 105 The optimization of the extraction procedure was carried out using adult females GL, since females were expected to have higher concentrations of natural MF, even presenting lower weight than male individuals. In each test, 3 females were used, thus, the mean sample weight was around 0.1 grams. The ratio between sample and dispersant was 1:5. The amount of dispersant was selected according to bibliography recommendations [20] whereas a high amount of clean-up sorbent (1.5 g) was used 110 7 due to the complexity of the extracts when the whole organism is used. The selection of PSA as dispersant sorbent was made due to its high selectivity when working with complex fatty samples [21]. This material consists of a silica gel base bonded with ethylenediamine-N-propyl groups and provides a mixed-mode retention mechanism. The studied clean-up sorbents were florisil and silica, two typical normal phase sorbents that interact 115 with sample components solely by adsorption. Figure 2A shows the recovery obtained for each sorbent using female GL samples with an addition amount of 3 ng, which corresponds to an addition level of 20.4 ng/g. When florisil was used as clean-up sorbent the overall recovery was 102 ± 4 %, whereas the use of silica led to a recovery value of 81 ± 9 %. The elution in both cases was made using 2 mL of ethyl acetate. In view of these results, florisil was selected as clean-up sorbent. 120 Then, the selection of elution solvent was made taking into account the low polarity of MF, thus, 2 mL of acetonitrile, hexane, ethyl acetate and hexane containing different percentages of ethyl acetate were used, in order to test the influence of a polarity increase in the selectivity of the elution process (see Fig. 2B). MF, with a structure similar to fatty acid methyl esters (FAME) was expected to elute in a less polar fraction than phospholipids or cholesterol in similar protocols for FAME fractionation 125 [22]. However, the best recoveries were achieved when using pure ethyl acetate or acetonitrile as elution solvents. The extracts obtained with both solvents were injected both in GC-PCI full scan and SIM modes and led to chromatograms with similar signals for MF and background. Thus, ethyl acetate was chosen due to its ease evaporation. Finally, the amount of solvent was optimized by collecting 3 consecutive fractions of 1 mL of solvent passed through an MSPD cartridge, with spiked female GL. 130 The percentage of MF in the second fraction was lower than 10%, thus 1.5 mL was selected as elution volume. Extracts were evaporated to dryness and reconstituted in a final volume of 20 µL as to increase the enrichment factor. The possible losses produced by evaporation/adsorption were not larger than 25% and were successfully corrected by the internal standard. 135 3.2 Method validation and analysis of real samples 8 The figures of merit of the proposed methodology for both, instrumental measures (i) and entire MSPD-GC-CI-MS method (m) are shown in Table 1. The limits of quantification (mLOQ) are reported for each studied species. These LOQs, defined for a signal to noise (S/N) of 10, were estimated from 140 the standard of lowest concentration (iLOQ) and the analyzed real samples without addition (mLOQ). Regarding male and female GL, the mLOQs were lower for female individuals (2 ng/g) than males (3.1 ng/g), since they led to cleaner extracts, due to the lower size of females. On the other hand, the mLOQ for AM and AF were established in 1.2 and 2.7 ng/g, respectively. The method recoveries were evaluated at two different addition levels, at low level, near the mLOQ 145 for each species and at high level, where the spiked amount was 5 times higher than the correspondent to low level. As shown in Table 2, recoveries ranged from 69 to 96% at the low spiked amount (1 or 2 ng depending on the species) and from 74 to 103% at the high spiked amount (5 or 10 ng). The recovery values obtained for the low level were applied in order to calculate concentrations in real samples. As an example, Figure 3 shows a chromatogram obtained for a real 150 sample of female GL with and without addition of MF (2 ng). As regards precision, the relative standard deviation (mRSD) was lower than 16 % for all the studies species, considered good since it included the inter-individual variability and the whole method performance. Real un-spiked samples were analyzed to calculate the average concentrations of MF in whole animals (see Table 2), by analyzing 3 replicates, which represented a total of 9 (3x3) individuals in 155 the case of GL and AM, and ca. 0.3 g (0.1g x3) in the case of AM. The native MF present in GL was 4 times higher in female individuals than males, 12 Vs 3 ng/g, respectively. Sex-specific differences in MF titer have been reported in several decapods species where males have higher levels of MF in the hemolymph than females [8]. The reason why MF levels are higher in GL females is not known and need future investigation. In the case of male GL, the levels found in real individuals were near 160 the mLOQs, aprox. 3 ng/g, thus, for individuals with lower native levels the method will be only suitable for detection issues, not for quantification. However, it could be also useful in exposure studies when the disruption mechanism led to an increase in MF production male GL. Regarding AF, to our knowledge there are no studies reporting the levels of MF. The concentration of MF found in this 9 study for AF, 3.5 ± 0.1 ng/g, was in the same order than that found for the males of the other studied 165 crustacean GL. The levels of MF in AM were around 2 ng/g. Other authors previously reported levels of MF in hemolymph of adult AM of 118 pg/µL [10]. In order to compare the obtained levels, we took into account that around 2 µL of hemolymph can be recovered from each adult AM [23] and that the average weight of each analyzed AM was 0.1 g. Thus, we obtained an extrapolated concentration 170 from Teal et al. [10] of 2.3 ng/g which is in agreement with our findings. 4. Conclusions A new simple and reliable method, employing MSPD, has been developed and validated for the analysis of MF in three ecologically and economically relevant species of arthropods avoiding the 175 extraction of hemolymph and simplifying the analytical protocol. This approach has proven to be useful to evaluate the concentration of this hormone in three arthropod species, and may be used in the future to foster studies in the field of conservation biology, mechanistic studies with endocrine disrupting chemicals and as a tool to increase the productivity of aquaculture and apiculture. However, further studies should be undertaken to compare the MF levels among different 180 physiological states of the animals since MF levels are expect to change according to animal age, molting and reproductive stage. Acknowledgements: This work has been financially supported by the Ministry of Economy and Competitiveness, Spain 185 (project no. CTM2014-56628-C3-2-R), the Galician Council of Culture, Education and Universities (ref. GRC2013-020) and FEDER/ERDF. T. Neuparth was supported by a Postdoctoral fellowship from the Portuguese Science and Technology Foundation (FCT), ref. SFRH/BPD/77912/2011. 190 0% 20% 40% 60% 80% 100% 120% Florisil Silica Recovery (%) Clean-up sorbent 0% 20% 40% 60% 80% 100% 120% Hex Hex/EtAc 70/30 Hex/EtAc 50/50 Hex/EtAc 30/70 EtAc ACN Recovery (%) Elution solvent A B 4 x10 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 + EIC(251.0000) SIM Female_WoutAD_1.D Counts vs. Acquisition Time (min) 11.7 11.8 11.9 12 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 12.9 13 13.1 13.2 13.3 + EIC(251.0000) SIM Female_WAD_1.D 3 x10 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 14 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8 14.9 15 15.1 15.2 15.3 15.4 15.5 MF I.S. + EIC(318.5000) SIM Female_WoutAD_1.D + EIC(318.5000) SIM Female_WAD_1.D