Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Alternative method’s results for the non targeted determination of xenobiotics in food by means of high resolution and accuracy mass spectrometry

Alternative method’s results for the non targeted determination of xenobiotics in food by means... The application of a high resolution and accurate mass spectrometry (HRAMS) approach to detect xenobiotics in dif- ferent food matrices by means of non targeted determination by UHPLC-Orbitrap followed by data processing analy- sis was described. Three case studies were reported to demonstrate the possibility to identify unexpected substances in different food commodities overcomes targeted method. This innovative approach could lay the foundation for its applicability to routine analysis in the near future giving the possibility to open new horizons to the research of a wide range of xenobiotics. Keywords: High resolution accurate mass spectrometry, Non-target screening, Food, Xenobiotics, Differential analysis Introduction expensive, but enable to lower the analytes detection lim- Nowadays, food safety has become a major issue of pub- its (up to sub ppt-levels) (Kaufmann et al. 2015). But, as a lic concern and it is a key concern for governments, food result, new or unexpected substances can go undetected. industry and manufacturers. Frauds, mistakes and acci- Therefore, the development of new strategies is needed dents can pollute food with a great deal of chemicals and in order to ensure a more efficient and rigorous food compromise its safety (Hollender et al. 2019). quality control. The application of a non targeted screen - For food control, targeted methods are the most ing method able to identify a wide range of xenobiotics applied. The current framework focuses only on sub - is an innovative approach involving lower costs and time stances expected to be found in specific food matrices analyses (Cavanna et al. 2018, Herrera-Lopez et al. 2014). according to compounds established by authorities in the In this approach, target inclusion lists are not used, since field. For this reason, these target analyses focus on the the molecules to be detected are not known a priori. detection of one or few classes of compounds. In many The recent advances in mass spectrometry, mainly high cases the extraction procedures are complex and also resolution mass spectrometry (HRMS), such as Orbit- rap a new type of mass analyzer invented by Makarov (Hu et  al. 2005; Makarov and Scigelova 2010; Zubarev and Makarov 2013; Bletsou et  al. 2015; Guo et  al. 2020; *Correspondence: r.scarpone@izs.it Department of Bromatology and Residues in Foods for Human Mol et al. 2012) and Time of Flight (TOF) (García-Reyes and Animal Consumption, Istituto Zooprofilattico Sperimentale et  al. 2007; Ki et  al. 2019; Peters et  al. 2010), together dell’Abruzzo e del Molise ‘G.Caporale’, Via Campo Boario snc, with the use of appropriated software (Milman 2015), 64100 Teramo, TE, Italy © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Scarpone et al. FoodContamination (2021) 8:5 Page 2 of 8 ™ ™ have enabled the development of untargeted approaches Scientific Accucore aQ C18 column (100 × 2.1 mm (Kaufmann 2014). with particle diameter of 2.6 μm). The oven and autosa - These untargeted approaches are a powerful tool for mpler temperature were set, respectively, at 40 °C and counteracting the continuous development of food con- 15 °C. The injection volume was 10 μL. The mobile phases tamination with the identification of novel compounds consisted of water (A) and methanol (B) both containing also by using retrospective data analysis. However, at the 5 mm ammonium formate and 0.1% of FAc. The analysis − 1 same time, some critical aspects have also to be taken was done at a flow rate of 0.4 mL min using the follow- into account. If, on the one hand, the data collected ing gradient elution: at the beginning, 20% phase B was with this approach need to be evaluated with the use of constant for 0.5 min, and it was increased up to 98% in multivariate statistical models (Riedl et  al. 2015), on the 10 min. The latter was maintained for 4 min, and then other hand a harmonized workflow is required, including switched back to the initial 20% in 0.5 min and kept standardized protocols and quality requirements. This is constant for 4 min giving a total runtime of 19 min. The to guarantee an efficient framework for data evaluation UHPLC system was connected to the single stage Orbit- and communication (Hollender et al. 2019). rap mass spectrometer Q Exactive from ThermoFisher Nevertheless, a preliminary attempt of non targeted Scientific (Bremen, Germany) through a heated elec - analysis harmonization was published by the US Phar- trospray interface (HESI-II) operating in positive/nega- macopeia (2016). In this document, the criteria to build a tive ionization (Makarov and Scigelova 2010). The HESI set of reference and test samples able to provide a reliable parameters in positive polarity were the following: elec- predictive model are reported. This preliminary work is trospray voltage of 3.2 kV; sheath gas of 40 arbitrary units; a good starting point but it is too generic, thus unable to and auxiliary gas of 25 arbitrary units; capillary tempera- provide suggestions for each analytical technique. ture 250 °C and auxiliary temperature 300 °C. Instead, The aim of this work is to apply a high resolution and in negative polarity were: electrospray voltage of 3.8 kV; accurate mass spectrometry (HRAMS) approach to sheath gas of 40 arbitrary units; and auxiliary gas of 15 detect xenobiotics in different food commodities by arbitrary units; capillary temperature 250 °C and auxiliary means of non targeted determination by UHPLC-Orbit- temperature 250 °C. rap followed by data processing analysis with dedicated The analysis was performed with 4 consecutive compound libraries. UHPLC-HRAMS runs in ESI+: 3 with a FullScan-All Ion Fragmentation (FullScan-AIF); 1 in Data Dependent Scan Materials and methods Mode (ddMS2). Chemicals and reagents The FullScan-AIF runs were acquired with resolv - Acetonitrile (ACN), methanol (MeOH) and water, all ing power of 140,000 FWHM for parental ions and LC-MS grade, were purchased from Sigma Aldrich mass range of 110–1200 m/z, AGC target of 3e6, max (Steinheim, Germany). Acetic acid (AAc) and formic IT 500 ms; the FullScan-AIF acquisition of all fragments acid (FAc) for mass spectrometry were also obtained was set with a resolving power of 35,000 FWHM and from Sigma Aldrich (Steinheim, Germany). QuEChERS 63.3–700 m/z as mass range, AGC target of 3e6, max IT Extraction Packets 5982-7650 (4 g of magnesium sul- 150 ms. The data dependent scan mode run was set with phate, 1 g of sodium citrate, 0.5 g sodium hydrogen cit- a “Homemade Exclusion list” of about 70 compounds in rate sesquihydrate, 1 g of sodium chloride) were obtained a mass range from 110 to 950 m/z and carried out with from Agilent Technologies (Santa Clara, USA). Finally, resolving power of 70,000 FWHM for parental ions and Pierce LTQ Velos ESI Negative and Positive Ion Calibra- 17,500 FWHM for all fragmentation products, using a tion Solution were provided by Thermo Fisher Scientific mass accuracy ≤2 ppm. All the ions that are present in (Waltham, MA, USA). the “environmental-laboratory system” were included in In the acquisition, lock masses were used: in positive an exclusion list that was generated by performing the ionization (ESI+) was disooctyl phthalate [M + H] with blank reagent chromatographic run. All the chromato- m/z = 391.28429 and in negative ionization (ESI-) was graphic runs were carried out using a stepped energy col- formic acid dimer [M  + Na-2H] with m/z = 112.98563. lision of 20, 35, and 60 eV. Other 4 consecutive UHPLC-HRAMS runs in ESI- UHPLC‑HRAMS analysis were carried out: 3 with a FullScan-All Ion Fragmenta- The analysis was based on a previous work from the same tion (FullScan-AIF); 1 in Data Dependent Scan Mode laboratory as described in 2020 by Scarpone et al. Briefly, (ddMS2). The resolving power, mass range and stepped the UHPLC chromatographic analysis was performed energy collision were the same set for the positive acqui- ™ ™ using Dionex Ultimate 3000 (Thermo Scientific , San sition mode. Instead, the “Homemade Exclusion list” was Jose, USA) equipped with an analytical column Thermo not used. S carpone et al. FoodContamination (2021) 8:5 Page 3 of 8 Case histories and analytical procedure Pesticide Residues in Fruits and Vegetables and regarded The technological improvements in mass spectrometry the untargeted screening of pesticides using multi-resi- give new possibilities for greatly increasing the scope of due methods analysis. The samples were spiked with 17 multi-residue methods (MRM) analysis helping the eve- pesticides, but the participants did not received informa- ryday work carried out by laboratories. Whereas full-scan tion about the list. measurements are theoretically the best approach for MS The blank matrix was characterized by the same beans screening, developments in targeted measurements also grown in the same field. offer the potential for a substantially increased scope of Two hundred grams portions of homogenate samples analysis. were weighed out into screw-capped polyethylene plas- The reason to conducing proficiency tests on screening tic bottles, sealed and stored at − 20 °C prior the distri- methods was to gather information from laboratories as bution to the participants. The extraction was carried to the type of software they use for processing data and out as reported by Scarpone et al. (2020) with the use of to evaluate the analytical method applicability in routine QuEChERS (EN 15662:2008 2008) without clean-up. The analysis. sample extracts were diluted 1:10 in water and injected in The collected data from the participants, used by EU LC-MS for the analysis. Reference Laboratories for Residues of Pesticides, could It was asked to participant laboratories to screen the help the improvement of data processing software. test items using a wide-scope screening methods nor- In these analyses, for the qualitative screening meth- mally applied for official monitoring purposes. The evalu - ods, the acceptable false-negative rate was 5% according ation test was based on qualitative information, but it to SANTE/12682/2019 Document and to the 2002/657/ was also requested to estimate the concentration, only for CE Commission Decision (Commission Decision n. 657/ informative purposes. EC 2002, European Commission Directorate General for Health and Food Safety, 2019). Case study 3 The third case study consisted in the EURL-FV-12 Profi - ciency Test. It was based on the pesticide-residue analy- sis of onion. The lab code of our laboratory was 053. This Case study 1 Proficiency Test, like the EURL-FV-10, was also organ - Our laboratory was involved in a proficiency test (RIKILT ized by the EURL for Pesticide Residues in Fruits and test 2018-11), organized by Wageningen University & Vegetables and regarded the untargeted screening of pes- Research, for unknown anti-microbiologically active ticides using multi-residue methods analysis. compounds in water in order to test and validate proce- Approximately 200 g of onion test item treated with dures for identification of ‘unknown’ substances having pesticides were stored at − 20 °C prior to shipment to antimicrobial activity. Our laboratory was identified by participants. The blank matrix sample was not the same lab code number 9837. of the delivered one. The selected pesticides used to spike Two delivered samples (sample material 1 and 2) con- the sample were 17 and no information was available to sisted approximately of 20 mL of rain water. The partici - the laboratories. pating laboratories were asked to identify all microbially Also in this case, the extraction was carried out active compounds in the samples and to report only the as reported by Scarpone et  al. (2020) with the use of identity of these. Quantification was outside of the scope QuEChers (EN 15662:2008 2008) without clean-up. The of this test. sample extracts were diluted 1:10 in water for the LC-MS The water samples were analyzed without any extrac - analysis. tion, concentration and/or clean-up. The analysis has The aim of this test was to evaluate laboratory capabil - been performed in positive and negative ionization mode, ity using large-scope quantitative and screening meth- in presence and absence of salt in the mobile phases of ods during routine analysis, for detecting and identifying UHPLC analysis. The blank matrix used to mapping the unexpected pesticides at levels at, or above 0.01 mg/kg. matrix components was the Milli-Q water. Database searching For this work different software and databases were used Case study 2 to identify the unknown compounds, such as online The second case study reported the European Union Ref - libraries (ACToR: Aggregated Computational Toxicology erence Laboratory-Fruit and Vegetable-10 Proficiency Resource, Drug Bank, FDA, Nature Chemical Biology, Test (EURL-FV10) in which the sample matrix was green Nature Communications, PubMed, Nature Chemis- beans with pods. The lab code of our laboratory was 067. try, Royal Society of Chemistry, Science Base, Springer This Proficiency Test was organized by the EURL for Scarpone et al. FoodContamination (2021) 8:5 Page 4 of 8 Nature), Compound Discoverer 2.0, and also the library good speed and resolution thanks to their solid-core from Thermo Fisher (EFS HRAM Compound Database). particles. The first tests were carried out with Thermo ™ ™ Scientific Accucore aQ C18 column in positive and Results and discussion negative modes, in presence and absence of salts in the Method applicability and advantages mobile phase of the UHPLC analysis. Then, Acquity The proposal method allows the identification in a sin - C18 column was also tested in the same conditions gle sample of a wide range of xenobiotics, not previously of the first one. Analyzing the chromatograms, it was preselected, that can be detected by means of untargeted observed that several signals presented a better separa- screening methods accomplished by UHPLC-Orbitrap tion, peak shape and efficiency using the Accucore aQ followed by data processing analysis. This method could C18 column in presence of salts in the mobile phases. be applied to a variety of food matrices and raw data can In all trials in absence of salts, no significant signals be processed many times for performing chemometric were detected, and any compounds were identified. and differential analyses. After these trials, a differential analysis between the The analyses of different matrices could present some blank matrix and the samples was carried out. The problems due to the different matrix effect, that consists blank matrix was not the same of the samples, but as in an ion suppression or enhancement of analyte caused reported in a previous study (Scarpone et al. 2020), it is by co-eluting matrix components (Uclés et al. 2017). This not significant to use the same blank matrix. problem can be reduced with the dilution of the extract To identify the compounds, different libraries were in most of the cases as also reported by Ferrer et al. (Fer- consulted (see paragraph 2.4). rer et al. 2011). The dilution has the advantage to reduce For the sample material 1, the results of the different the matrix effect but at the same time to lower the LODs trials gave a double putative identification matching to of the analytes that risk not being detected. The occur - the library: linezolid (C H FN O ) and vepandavir 16 20 3 4 ring phenomena of signal suppression or enhancement (C H N O P ). Easily, it was possible to identify puta- 13 32 6 3 2 in MS detection cannot be attributed to only one cause, tively in positive mode the first compound as correct but it depends on a synergic effect of all the conditions based on accurate mass measurements and informative involved. The interfering compounds, that co-elute in the fragmentation spectra matching with the library data- chromatographic separation, can be components of the base (Supplementary Information, Additional file 1 a). sample, or released during the extraction process or rea- For the sample material 2, analysing the results in gents added to the mobile phase (Furey et al. 2013; Gos- FullScan-AIF range [110,0000 – 1200,0000 m/z], it can etti et al. 2010). be claimed that the differential analysis performed by Compound Discoverer 2.0 allowed the putative identifi - cation of boromycin, boromycin deprotonated and also Case study 1 N-formyl-boromycin (boromycin-CHO) both in posi- Any information was available on the chemical and tive and negative mode and in presence of salts. physical characteristic of the unknown compounds and In Additional file  1b (Supplementary Information), to prevent their stability during the analytical procedure the putative identification of boromycin (C H BNO ), 45 74 15 and to avoid their loss, no treatment was applied to the has been shown ([M + H]   = 880.52014 m/z) at samples. The focus was to improve the sensitivity of the 10.98 min. screening method and also improve the data evaluation Then, the analysis was performed in Data Dependent procedures. Scan (ddMS2) to confirm the previous identification. The information collected from HRAMS allows the Operating in ESI- mode, the boromycin was detected identification of unknown compounds with relatively at 878.52069  m/z (Fig.  1a) with a retention time of high degree of confidence without reference analytical 11.05 min with a specific spectra. standards. The identification was carried out only using The boromycin profile pattern was characterized the mass spectra database. The test analyzes were car - by 879.52356 m/z (C H BN0 ); 878.52069 m/z 45 74 15 + + ried out directly on the two samples of water delivered (C H BN0 ); 877.52393 m/z (C H BN0 ) and 45 73 15 45 72 15 for the proficiency test using two analytical columns. 880.52612 m/z (C H BN0- H ). 45 74 15 These selected columns have similar characteristic; The characteristic fragments that contributed they have the same dimensions (100 × 2,1 mm) and to the putative identification were 752.42841 m/z, both were packed with spherical solid core particles. 666.58533 m/z, 155.10989 m/z and 116.07271 m/z The only difference was the particle size dimension (1.8 (Fig. 1b). and 2.6 μm). They were tested considering their opti - In ESI+ mode, the same compound was detected mal retention of a wide range of polar analytes with as [M + H] with a m/z= 880.51776 at RT 10.96 min S carpone et al. FoodContamination (2021) 8:5 Page 5 of 8 Fig. 1 Putative identification of Boromycin (C H BNO ): a Full-Scan in ESI-: experimental and theorical spectra; b ddMS2 acquisition in ESI-: 45 74 15 parental ion and fragmentation products c ddMS2 acquisition in ESI+: parental ion and fragmentation products Scarpone et al. FoodContamination (2021) 8:5 Page 6 of 8 Table 1 Estimated concentrations of the detected pesticides and the fragmentation pattern was characterized by two fragments  751.54974 and 93.07027 m/z (Fig.  1c). Robust Mean CV (%) Estimated It was detected also the boromycin adduct with Na at (mg/kg) concentration (mg/kg) 902.50049 m/z (Supplementary Information, Additional file 1b). Benalaxyl 0.041 22 0.034 Furthermore, the theoretical fragmentation pattern of Clomazone 0.010 21 0.0085 boromycin was also studied to confirm the putative iden - Emamectin 0.0090 34 0.0090 tification and to avoid to bump into isobaric compounds. Etoxazole 0.013 26 0.0098 Fenpyrazamine 0.0080 27 0.0085 Isopryrazam 0.0090 29 0.0012 Case study 2 Metrafenone 0.013 21 0.011 The results of this Proficiency Test were reported in the Penflufen 0.010 15 0.011 EUPT-FV-SM10 Final Report published by EURL. The Penthiopyrad 0.067 23 0.067 pesticides used to spike the samples, at different lev - Prosulfocarb 0.011 22 0.011 els between 0.01 mg/kg and 0.1 mg/kg, were decided Spinetoram 0.035 26 0.029 upon the Quality Control Group. The pesticide treat - Spirotetramat 0.072 38 0.012 ments were carried out just post-harvest used pesti- Sulfoxaflor 0.0090 24 0.011 cide analytical standards. The selected pesticides were: Tetramethrin 0.011 16 0.010 benalaxyl, clomazone, cyfluthrin, emamectin, etoxa - zole, fenpyrazamine, isopyrazam, metrafenone, pen- flufen, pentachloroaniline, penthiopyrad, proquinazid, prosulfocarb, spinetoram, spirotetramat, spirotetramat by ChemSpider and EFS HRAM Compound database as metabolite BYI08330-enol, sulfoxaflor and tetramethrin. proquinazid. Seventy-three laboratories participated to this Profi - The standards of these analytes were available in our ciency Test and only 69 of them submitted the results. laboratory and the identified compounds were also quan - These laboratories analysed the test item using method tified and the data processed. In Table  1, it is reported the based on gas or liquid chromatography, or both, com- estimated concentration of the identified pesticides. The bined with mass spectrometry detection. detection limit ranged between 0.0025 to 0.0050 mg/kg. Fourteen pesticides, corresponding to 78% of reported Spirotetramat-enol was also present in the test item at pesticides by laboratory, were detected in our laboratory the same range of concentration of others, but it was not by liquid chromatography analysis combined with high detected; spirotetramat-ketohydroxy was present below resolution mass spectrometry. 0.01 mg/kg and it was detected and reported and not all It was possible in our laboratory to detect and identify the laboratories reported the concentration. 14 pesticides: benalaxyl, clomazone, emamectin, etoxazole, fen- Case study 3 pyrazamine, isopyrazam, metrafenone, penflufen, pen - For this test, it is available at this moment only a pre- thiopyrad, prosulfocarb, spinetoram, spirotetramat, liminary report by EURL, where it was reported which sulfoxaflor and tetramethrin. Three pesticides were not analytes were present in the test item, the list of labora- detected: cyfluthrin, pentachloroaniline and proquinazid. tories that participated and the table with the compounds In particular, the detection of cyfluthrin and pentachlo - detected and non-detected by each laboratories. Only roaniline was failure because they are analytes suitable 17% of laboratories reported the entire results from all for the GC-MS determination. analytes, while 83% reported fewer analytes. Instead, proquinazid was not detected with our screen- In this case study, no matrix reference was available as ing method because its signal was low, and it was not in a real routine case analysis. depart from the blank matrix with the set statistical The pesticide treatments were carried out post-harvest parameters (p-value < 0.05 and log fold charge =1). The using standard analytical solutions. The spiked pesticides differential analysis showed about 600 compounds to be were 17: alachlor, cyanofenphos, diuron, dodemorph, identified, but the proquinazid signal was not included endrin, fluacryprim, fonofos, isoprocarb, metamitron, in this group, but present among the matrix signals. In metazachlor, metobromuron, monolinuron, prometryne, retrospective analysis, it was present a low signal of the propazine, propoxur, simazine, and tetrachlorvinphos. parent ion at [M + H]   = 373.04000 m/z and also its ion In this Proficiency Test, 64 laboratories participated fragments at 330.99480 and 288.94750 m/z, recognized and only three of them have not reported results. Thir - teen pesticides (76% of the reported pesticides by S carpone et al. FoodContamination (2021) 8:5 Page 7 of 8 laboratory) were detected by our screening method using In the case of extremely complex matrices, like onion, it Compound Discoverer 2.0: diuron, dodemorph, fonofos, is possible, that the differential analysis cannot highlight isoprocarb, metamitron, metazachlor, metobromuron, signals of some analytes at low concentrations that could monolinuron, prometryne, propazine, propoxur, sima- be covered by the matrix signals. zine, and tetrachlorvinphos (Supplementary Information, The databases contemplate only compounds that Additional  file  2). Alachlor, cyanofenphos, endrin and present parent ion and this is a limitation for the com- fluacryprim were not reported. pounds instable that present a high fragmentation. The Alachlor was not identified by Compound Dis - libraries could be implemented with the fragmentation coverer 2.0 because it was not present its molecular pattern of the compound to help the identification. ion. Our system identified its characteristic fragment ([M + H]   = 238.09875 m/z) as ketamine (C H ClNO) 13 16 at 8.072 min as retention time. Therefore, the putative Conclusion identification of this analyte, that is not a stable com - The results of these studies are most encouraging and pound, was compromised by the absence of the molec- this method could be used more and more as screens/ ular ion and the fragment did not match to the library filters, to make routine laboratory work easier and database. The database should be implemented with frag - faster. mentation pattern of the compound to better recognize a These three interlaboratory tests on wide-scope wide range of analytes. screening methods showed that such an approach can The detection of cyanofenphos and fluacryprim was substantially expand the scope of xenobiotics analysis. affected by an error of the laboratory operator. Onion is This is especially useful for unknown compounds, like a very complex matrix, that presented a high number of some pesticides, not frequently found in food and feed, compounds to be identified (about 395 compounds) and or not monitored by the laboratories because they are in this case it is necessary great attention in the elabora- not part of the EU-Coordinated Programme. The use of tion and interpretation of the results and at last but not screening methods can greatly increase the chance of least, the adequately trained operator. detecting less commonly found xenobiotics. However, In a retrospective analysis, the cyanofen- the tests also revealed that verification of the screening phos (C H N OPS) showed a low signal methods performance (i.e. validation) are necessary to 9 15 5 2 [M + H]  = 304.05490 m/z at retention time of 8.558 min increase the reliability of such methods and guidelines with the presence also of its characteristic fragments for such validation have been prepared and included in (156.98710 and 258.01360 m/z). This signal was recog - the Document SANTE/12682/2019 (European Com- nized by Chemspider database. In the case of fluacryprim mission Directorate General for Health and Food (C H F N O ), it was only detected by 20 laboratories. Safety, 2019). 20 21 3 2 5 Its pseudomolecular ion ([M + H]   = 427.14670 m/z) Furthermore, this screening method did not give false was present at retention time of 9.045 min and also the positive results. characteristic profile fragments in ddMS2 acquisition In the near future, the perspectives should be focused (205.08582 and 145.06476 m/z). This molecule was cor - on the implementation and development of database rectly identified by Compound Discoverer 2.0. and software to help the identification of xenobiotics. Finally, endrin is not an analyte suitable for the LC-MS determination and for this reason it was not detected. Supplementary Information These identifications were carried out without the ana - The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s40550- 021- 00086-x. lytical standards and the concentrations, for this reason, were not reported. Additional file 1. Differential analysis and result table for sample material 1 (a) and sample material 2 (b) by means of Compound Discoverer 2.0. Limitations Additional file 2. Identification results for the onion sample by means of This non targeted method presented in this paper pre - Compound Discoverer 2.0. sents some limitations. For these kinds of analyses, it is really important the Acknowledgements attention in the processing of data and the expertise of Not applicable. the operator. The data collected with this approach are Authors’ contributions many and need to be evaluated with the use of specific All authors that contributed to the research paper are listed at the beginning. software to process data. The author(s) read and approved the final manuscript. Scarpone et al. FoodContamination (2021) 8:5 Page 8 of 8 Funding non-targeted antitussive adulterants in herbal medicines by Q-Orbitrap Not applicable. HRAMS and screening database. Int J Mass Spectrom 447:116250. https:// doi. org/ 10. 1016/j. ijms. 2019. 116250 Availability of data and materials Herrera-Lopez S, Hernando MD, García-Calvo E, Fernández-Alba AR, Ulasze- The RIKILT data are not accessible without a PT registration. It is possible to wska MM (2014) Simultaneous screening of targeted and nontargeted access the data with a specific request. Our lab code was 9837. contaminants using an LC-QTOF-MS system and automated MS/MS For the EURL-FV10 (lab code 067), the data are available in https:// www. eurl- library searching. J Mass Spectrom 49:878–893. https:// doi. org/ 10. 1002/ pesti cides. eu/ docs/ public/ tmplt_ artic le. asp? CntID= 1082& LabID= 500& Lang= jms. 3428 EN. Hollender J, Van Bavel B, Dulio V, Farmen E, Furtmann K, Koschorreck J, Kunkel Instead for EURL-FV12 (lab code 053), the data are available in https:// www. U, Krauss M, Munthe J, Schlabach M, Slobodnik J, Stroomberg G, Ternes T, eurl- pesti cides. eu/ user files/ file// Preli minary_ Report_ SM12. pdf . Thomaidis NS, Togola A, Tornero V (2019) High resolution mass spectrom- etry-based non-target screening can support regulatory environmental monitoring and chemicals management. Environ Sci Eur 31(1):42–53. Declarations https:// doi. org/ 10. 1186/ s12302- 019- 0225-x Hu Q, Noll RJ, Li H, Makarov A, Hardmanc M, Cooks RG (2005) The Orbitrap: a Competing interests new mass spectrometer. J Mass Spectrom 40:430–443. https:// doi. org/ 10. Rossana Scarpone, Rachele Rocchi, Federico Bacà, Roberta Rosato, Francesco 1002/ jms. 856 Chiumiento and Giacomo Migliorati declare that they have no conflict of Kaufmann A (2014) Combining UHPLC and high-resolution MS: a viable interest. approach for the analysis of complex samples? TrAC Trend Anal Chem 63:113–128. https:// doi. org/ 10. 1016/j. trac. 2014. 06. 025 Received: 12 February 2021 Accepted: 18 October 2021 Kaufmann A, Butcher P, Maden K, Walker S, Widmer M (2015) Reliability of veterinary drug residue confirmation: high resolution mass spectrometry versus tandem mass spectrometry. Anal Chim Acta 856:54–67. https:// doi. org/ 10. 1016/j. aca. 2014. 11. 034 Ki NY, Jisu H, Beom KH, Kyung KH, Bong JM, Han BO, Jongki H (2019) Rapid References screening of sulfonamides in dietary supplements based on extracted Bletsou AA, Junho J, Hollender J, Archontaki E, Thomaidis NS (2015) Targeted common ion chromatogram and neutral loss scan by LC-Q/TOF-mass and non-targeted liquid chromatography-mass spectrometric workflows spectrometry. J Food Drug Anal 27:164–174. https:// doi. org/ 10. 1016/j. for identification of transformation products of emerging pollutants in jfda. 2018. 08. 006 the aquatic environment. TrAC Trend Anal Chem 66:32–44. https:// doi. Makarov A, Scigelova M (2010) Coupling liquid chromatography to Orbitrap org/ 10. 1016/j. trac. 2014. 11. 009 mass spectrometry. J Chromatogr A:3938–3945. https:// doi. org/ 10. 1016/j. Cavanna D, Righetti L, Elliott C, Suman M (2018) The scientific challenges in chroma. 2010. 02. 022 moving from targeted to non-targeted mass spectrometric methods Milman BL (2015) General principles of identification by mass spectrometry. for food fraud analysis: a proposed validation workflow to bring about a TrAC Trend Anal Chem 69:24–33. https:// doi. org/ 10. 1016/j. trac. 2014. 12. harmonized approach. Trends Food Sci Technol 80:223–241. https:// doi. org/ 10. 1016/j. tifs. 2018. 08. 007 Mol HGJ, Zomer P, De Koning M (2012) Qualitative aspects and validation of a Commission Decision n.657/EC (2002) Implementing Council Directive 96/23/ screening method for pesticides in vegetables and fruits based on liquid EC concerning the performance of analytical methods and the interpre- chromatography coupled to full scan high resolution (Orbitrap) mass tation of results spectrometry. Anal Bioanal Chem 403:2891–2908. https:// doi. org/ 10. EN 15662:2008 (2008) European Committee for Standardization (2008) CEN EN 1007/ s00216- 012- 6100-x 15662: Foods of plant origin-determination of pesticide residues using Peters RJB, Stolker AAM, Mol JGJ, Lommen A, Lyris E, Angelis Y, Vonaparti A, Sta- GC-MS and/or LC-MS/MS following acetonitrile extraction/partition and mou M, Georgakopoulos C, Nielen MWF (2010) Screening in veterinary clean-up by dispersive SPE-QuEChERS method. https:// www. en- stand drug analysis and sports doping control based on full-scan, accurate- ard. eu/ csn- en- 15662- foods- of- plant- origin- multi method- for- the- deter mass spectrometry. TrAC Trend Anal Chem 29:1250–1268. https:// doi. org/ minat ion- of- pesti cide- resid ues- using- gc- and- lc- based- analy sis- follo 10. 1016/j. trac. 2010. 07. 012 wing- aceto nitri le- extra ction- parti tioni ng- and- clean- up- by- dispe rsive- Riedl J, Esslinger S, Fauhl-Hassek C (2015) Review of validation and reporting spe- modul ar- quech ers- method/ of non-targeted fingerprinting approaches for food authentication. Anal European Commission Directorate General for Health and Food Safety (2019) Chim Acta 885:17–32. https:// doi. org/ 10. 1016/j. aca. 2015. 06. 003 Analytical quality control and method validation for pesticide residues Scarpone R, Rosato R, Chiumiento F, Cipolletti C, Sergi M, Compagnone D analysis in food and feed (SANTE/12682/2019), pp 1–48 (2020) Preliminary study to develop an alternative method for the non- Ferrer C, Lozano A, Agüera AA, Jiménez G, Fernández-Alba AR (2011) targeted determination of xenobiotics in food by means of ultra high Overcoming matrix effects using the dilution approach in multiresidue performance liquid chromatography coupled to high resolution and methods for fruits and vegetables. J Chromatogr A 1218(42):7634–7639. accuracy mass spectrometry. Food Anal Methods 13:1099–1110. https:// https:// doi. org/ 10. 1016/j. chroma. 2011. 07. 033 doi. org/ 10. 1007/ s12161- 020- 01727-1 Furey A, Moriarty M, Bane V, Kinsella B, Lehane M (2013) Ion suppression; a Uclés S, Lozano A, Sosa A, Parrilla Vázquez P, Valverde A, Fernández-Alba AR critical review on causes, evaluation, prevention and applications. Talanta (2017) Matrix interference evaluation employing GC and LC coupled to 115:104–122. https:// doi. org/ 10. 1016/j. talan ta. 2013. 03. 048 triple quadrupole tandem mass spectrometry. Talanta 174:72–81. https:// García-Reyes JF, Hernando MD, Molina-Díaz A, Fernández-Alba AR (2007) doi. org/ 10. 1016/j. talan ta. 2017. 05. 068 Comprehensive screening of target, non-target and unknown pesticides US Pharmacopeia (2016) Guidance on developing and validating non-tar- in food by LC-TOF-MS. TrAC Trend Anal Chem 26:828–841. https:// doi. geted methods for adulteration detection. Appendix XVIII:2053–2067 org/ 10. 1016/j. trac. 2007. 06. 006 Zubarev RA, Makarov A (2013) Orbitrap mass spectrometry. Anal Chem Gosetti F, Mazzucco E, Zampieri D, Gennaro MC (2010) Signal suppression/ 85:5288–5296. https:// doi. org/ 10. 1021/ ac400 1223 enhancement in high-performance liquid chromatography tandem mass spectrometry. J Chromatogr A 1217:3929–3937. https:// doi. org/ 10. 1016/j. Publisher’s Note chroma. 2009. 11. 060 Springer Nature remains neutral with regard to jurisdictional claims in pub- Guo C, Liping G, Weijian W, Jiawei L, Liming Z, Sheng X, Yanxia Z, Ruiqing X, lished maps and institutional affiliations. Xunjie Z, Feng S (2020) Rapid screening and identification of targeted or http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Food Contamination Springer Journals

Alternative method’s results for the non targeted determination of xenobiotics in food by means of high resolution and accuracy mass spectrometry

Loading next page...
 
/lp/springer-journals/alternative-method-s-results-for-the-non-targeted-determination-of-ZWE9JTqQBt

References (25)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
eISSN
2196-2804
DOI
10.1186/s40550-021-00086-x
Publisher site
See Article on Publisher Site

Abstract

The application of a high resolution and accurate mass spectrometry (HRAMS) approach to detect xenobiotics in dif- ferent food matrices by means of non targeted determination by UHPLC-Orbitrap followed by data processing analy- sis was described. Three case studies were reported to demonstrate the possibility to identify unexpected substances in different food commodities overcomes targeted method. This innovative approach could lay the foundation for its applicability to routine analysis in the near future giving the possibility to open new horizons to the research of a wide range of xenobiotics. Keywords: High resolution accurate mass spectrometry, Non-target screening, Food, Xenobiotics, Differential analysis Introduction expensive, but enable to lower the analytes detection lim- Nowadays, food safety has become a major issue of pub- its (up to sub ppt-levels) (Kaufmann et al. 2015). But, as a lic concern and it is a key concern for governments, food result, new or unexpected substances can go undetected. industry and manufacturers. Frauds, mistakes and acci- Therefore, the development of new strategies is needed dents can pollute food with a great deal of chemicals and in order to ensure a more efficient and rigorous food compromise its safety (Hollender et al. 2019). quality control. The application of a non targeted screen - For food control, targeted methods are the most ing method able to identify a wide range of xenobiotics applied. The current framework focuses only on sub - is an innovative approach involving lower costs and time stances expected to be found in specific food matrices analyses (Cavanna et al. 2018, Herrera-Lopez et al. 2014). according to compounds established by authorities in the In this approach, target inclusion lists are not used, since field. For this reason, these target analyses focus on the the molecules to be detected are not known a priori. detection of one or few classes of compounds. In many The recent advances in mass spectrometry, mainly high cases the extraction procedures are complex and also resolution mass spectrometry (HRMS), such as Orbit- rap a new type of mass analyzer invented by Makarov (Hu et  al. 2005; Makarov and Scigelova 2010; Zubarev and Makarov 2013; Bletsou et  al. 2015; Guo et  al. 2020; *Correspondence: r.scarpone@izs.it Department of Bromatology and Residues in Foods for Human Mol et al. 2012) and Time of Flight (TOF) (García-Reyes and Animal Consumption, Istituto Zooprofilattico Sperimentale et  al. 2007; Ki et  al. 2019; Peters et  al. 2010), together dell’Abruzzo e del Molise ‘G.Caporale’, Via Campo Boario snc, with the use of appropriated software (Milman 2015), 64100 Teramo, TE, Italy © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Scarpone et al. FoodContamination (2021) 8:5 Page 2 of 8 ™ ™ have enabled the development of untargeted approaches Scientific Accucore aQ C18 column (100 × 2.1 mm (Kaufmann 2014). with particle diameter of 2.6 μm). The oven and autosa - These untargeted approaches are a powerful tool for mpler temperature were set, respectively, at 40 °C and counteracting the continuous development of food con- 15 °C. The injection volume was 10 μL. The mobile phases tamination with the identification of novel compounds consisted of water (A) and methanol (B) both containing also by using retrospective data analysis. However, at the 5 mm ammonium formate and 0.1% of FAc. The analysis − 1 same time, some critical aspects have also to be taken was done at a flow rate of 0.4 mL min using the follow- into account. If, on the one hand, the data collected ing gradient elution: at the beginning, 20% phase B was with this approach need to be evaluated with the use of constant for 0.5 min, and it was increased up to 98% in multivariate statistical models (Riedl et  al. 2015), on the 10 min. The latter was maintained for 4 min, and then other hand a harmonized workflow is required, including switched back to the initial 20% in 0.5 min and kept standardized protocols and quality requirements. This is constant for 4 min giving a total runtime of 19 min. The to guarantee an efficient framework for data evaluation UHPLC system was connected to the single stage Orbit- and communication (Hollender et al. 2019). rap mass spectrometer Q Exactive from ThermoFisher Nevertheless, a preliminary attempt of non targeted Scientific (Bremen, Germany) through a heated elec - analysis harmonization was published by the US Phar- trospray interface (HESI-II) operating in positive/nega- macopeia (2016). In this document, the criteria to build a tive ionization (Makarov and Scigelova 2010). The HESI set of reference and test samples able to provide a reliable parameters in positive polarity were the following: elec- predictive model are reported. This preliminary work is trospray voltage of 3.2 kV; sheath gas of 40 arbitrary units; a good starting point but it is too generic, thus unable to and auxiliary gas of 25 arbitrary units; capillary tempera- provide suggestions for each analytical technique. ture 250 °C and auxiliary temperature 300 °C. Instead, The aim of this work is to apply a high resolution and in negative polarity were: electrospray voltage of 3.8 kV; accurate mass spectrometry (HRAMS) approach to sheath gas of 40 arbitrary units; and auxiliary gas of 15 detect xenobiotics in different food commodities by arbitrary units; capillary temperature 250 °C and auxiliary means of non targeted determination by UHPLC-Orbit- temperature 250 °C. rap followed by data processing analysis with dedicated The analysis was performed with 4 consecutive compound libraries. UHPLC-HRAMS runs in ESI+: 3 with a FullScan-All Ion Fragmentation (FullScan-AIF); 1 in Data Dependent Scan Materials and methods Mode (ddMS2). Chemicals and reagents The FullScan-AIF runs were acquired with resolv - Acetonitrile (ACN), methanol (MeOH) and water, all ing power of 140,000 FWHM for parental ions and LC-MS grade, were purchased from Sigma Aldrich mass range of 110–1200 m/z, AGC target of 3e6, max (Steinheim, Germany). Acetic acid (AAc) and formic IT 500 ms; the FullScan-AIF acquisition of all fragments acid (FAc) for mass spectrometry were also obtained was set with a resolving power of 35,000 FWHM and from Sigma Aldrich (Steinheim, Germany). QuEChERS 63.3–700 m/z as mass range, AGC target of 3e6, max IT Extraction Packets 5982-7650 (4 g of magnesium sul- 150 ms. The data dependent scan mode run was set with phate, 1 g of sodium citrate, 0.5 g sodium hydrogen cit- a “Homemade Exclusion list” of about 70 compounds in rate sesquihydrate, 1 g of sodium chloride) were obtained a mass range from 110 to 950 m/z and carried out with from Agilent Technologies (Santa Clara, USA). Finally, resolving power of 70,000 FWHM for parental ions and Pierce LTQ Velos ESI Negative and Positive Ion Calibra- 17,500 FWHM for all fragmentation products, using a tion Solution were provided by Thermo Fisher Scientific mass accuracy ≤2 ppm. All the ions that are present in (Waltham, MA, USA). the “environmental-laboratory system” were included in In the acquisition, lock masses were used: in positive an exclusion list that was generated by performing the ionization (ESI+) was disooctyl phthalate [M + H] with blank reagent chromatographic run. All the chromato- m/z = 391.28429 and in negative ionization (ESI-) was graphic runs were carried out using a stepped energy col- formic acid dimer [M  + Na-2H] with m/z = 112.98563. lision of 20, 35, and 60 eV. Other 4 consecutive UHPLC-HRAMS runs in ESI- UHPLC‑HRAMS analysis were carried out: 3 with a FullScan-All Ion Fragmenta- The analysis was based on a previous work from the same tion (FullScan-AIF); 1 in Data Dependent Scan Mode laboratory as described in 2020 by Scarpone et al. Briefly, (ddMS2). The resolving power, mass range and stepped the UHPLC chromatographic analysis was performed energy collision were the same set for the positive acqui- ™ ™ using Dionex Ultimate 3000 (Thermo Scientific , San sition mode. Instead, the “Homemade Exclusion list” was Jose, USA) equipped with an analytical column Thermo not used. S carpone et al. FoodContamination (2021) 8:5 Page 3 of 8 Case histories and analytical procedure Pesticide Residues in Fruits and Vegetables and regarded The technological improvements in mass spectrometry the untargeted screening of pesticides using multi-resi- give new possibilities for greatly increasing the scope of due methods analysis. The samples were spiked with 17 multi-residue methods (MRM) analysis helping the eve- pesticides, but the participants did not received informa- ryday work carried out by laboratories. Whereas full-scan tion about the list. measurements are theoretically the best approach for MS The blank matrix was characterized by the same beans screening, developments in targeted measurements also grown in the same field. offer the potential for a substantially increased scope of Two hundred grams portions of homogenate samples analysis. were weighed out into screw-capped polyethylene plas- The reason to conducing proficiency tests on screening tic bottles, sealed and stored at − 20 °C prior the distri- methods was to gather information from laboratories as bution to the participants. The extraction was carried to the type of software they use for processing data and out as reported by Scarpone et al. (2020) with the use of to evaluate the analytical method applicability in routine QuEChERS (EN 15662:2008 2008) without clean-up. The analysis. sample extracts were diluted 1:10 in water and injected in The collected data from the participants, used by EU LC-MS for the analysis. Reference Laboratories for Residues of Pesticides, could It was asked to participant laboratories to screen the help the improvement of data processing software. test items using a wide-scope screening methods nor- In these analyses, for the qualitative screening meth- mally applied for official monitoring purposes. The evalu - ods, the acceptable false-negative rate was 5% according ation test was based on qualitative information, but it to SANTE/12682/2019 Document and to the 2002/657/ was also requested to estimate the concentration, only for CE Commission Decision (Commission Decision n. 657/ informative purposes. EC 2002, European Commission Directorate General for Health and Food Safety, 2019). Case study 3 The third case study consisted in the EURL-FV-12 Profi - ciency Test. It was based on the pesticide-residue analy- sis of onion. The lab code of our laboratory was 053. This Case study 1 Proficiency Test, like the EURL-FV-10, was also organ - Our laboratory was involved in a proficiency test (RIKILT ized by the EURL for Pesticide Residues in Fruits and test 2018-11), organized by Wageningen University & Vegetables and regarded the untargeted screening of pes- Research, for unknown anti-microbiologically active ticides using multi-residue methods analysis. compounds in water in order to test and validate proce- Approximately 200 g of onion test item treated with dures for identification of ‘unknown’ substances having pesticides were stored at − 20 °C prior to shipment to antimicrobial activity. Our laboratory was identified by participants. The blank matrix sample was not the same lab code number 9837. of the delivered one. The selected pesticides used to spike Two delivered samples (sample material 1 and 2) con- the sample were 17 and no information was available to sisted approximately of 20 mL of rain water. The partici - the laboratories. pating laboratories were asked to identify all microbially Also in this case, the extraction was carried out active compounds in the samples and to report only the as reported by Scarpone et  al. (2020) with the use of identity of these. Quantification was outside of the scope QuEChers (EN 15662:2008 2008) without clean-up. The of this test. sample extracts were diluted 1:10 in water for the LC-MS The water samples were analyzed without any extrac - analysis. tion, concentration and/or clean-up. The analysis has The aim of this test was to evaluate laboratory capabil - been performed in positive and negative ionization mode, ity using large-scope quantitative and screening meth- in presence and absence of salt in the mobile phases of ods during routine analysis, for detecting and identifying UHPLC analysis. The blank matrix used to mapping the unexpected pesticides at levels at, or above 0.01 mg/kg. matrix components was the Milli-Q water. Database searching For this work different software and databases were used Case study 2 to identify the unknown compounds, such as online The second case study reported the European Union Ref - libraries (ACToR: Aggregated Computational Toxicology erence Laboratory-Fruit and Vegetable-10 Proficiency Resource, Drug Bank, FDA, Nature Chemical Biology, Test (EURL-FV10) in which the sample matrix was green Nature Communications, PubMed, Nature Chemis- beans with pods. The lab code of our laboratory was 067. try, Royal Society of Chemistry, Science Base, Springer This Proficiency Test was organized by the EURL for Scarpone et al. FoodContamination (2021) 8:5 Page 4 of 8 Nature), Compound Discoverer 2.0, and also the library good speed and resolution thanks to their solid-core from Thermo Fisher (EFS HRAM Compound Database). particles. The first tests were carried out with Thermo ™ ™ Scientific Accucore aQ C18 column in positive and Results and discussion negative modes, in presence and absence of salts in the Method applicability and advantages mobile phase of the UHPLC analysis. Then, Acquity The proposal method allows the identification in a sin - C18 column was also tested in the same conditions gle sample of a wide range of xenobiotics, not previously of the first one. Analyzing the chromatograms, it was preselected, that can be detected by means of untargeted observed that several signals presented a better separa- screening methods accomplished by UHPLC-Orbitrap tion, peak shape and efficiency using the Accucore aQ followed by data processing analysis. This method could C18 column in presence of salts in the mobile phases. be applied to a variety of food matrices and raw data can In all trials in absence of salts, no significant signals be processed many times for performing chemometric were detected, and any compounds were identified. and differential analyses. After these trials, a differential analysis between the The analyses of different matrices could present some blank matrix and the samples was carried out. The problems due to the different matrix effect, that consists blank matrix was not the same of the samples, but as in an ion suppression or enhancement of analyte caused reported in a previous study (Scarpone et al. 2020), it is by co-eluting matrix components (Uclés et al. 2017). This not significant to use the same blank matrix. problem can be reduced with the dilution of the extract To identify the compounds, different libraries were in most of the cases as also reported by Ferrer et al. (Fer- consulted (see paragraph 2.4). rer et al. 2011). The dilution has the advantage to reduce For the sample material 1, the results of the different the matrix effect but at the same time to lower the LODs trials gave a double putative identification matching to of the analytes that risk not being detected. The occur - the library: linezolid (C H FN O ) and vepandavir 16 20 3 4 ring phenomena of signal suppression or enhancement (C H N O P ). Easily, it was possible to identify puta- 13 32 6 3 2 in MS detection cannot be attributed to only one cause, tively in positive mode the first compound as correct but it depends on a synergic effect of all the conditions based on accurate mass measurements and informative involved. The interfering compounds, that co-elute in the fragmentation spectra matching with the library data- chromatographic separation, can be components of the base (Supplementary Information, Additional file 1 a). sample, or released during the extraction process or rea- For the sample material 2, analysing the results in gents added to the mobile phase (Furey et al. 2013; Gos- FullScan-AIF range [110,0000 – 1200,0000 m/z], it can etti et al. 2010). be claimed that the differential analysis performed by Compound Discoverer 2.0 allowed the putative identifi - cation of boromycin, boromycin deprotonated and also Case study 1 N-formyl-boromycin (boromycin-CHO) both in posi- Any information was available on the chemical and tive and negative mode and in presence of salts. physical characteristic of the unknown compounds and In Additional file  1b (Supplementary Information), to prevent their stability during the analytical procedure the putative identification of boromycin (C H BNO ), 45 74 15 and to avoid their loss, no treatment was applied to the has been shown ([M + H]   = 880.52014 m/z) at samples. The focus was to improve the sensitivity of the 10.98 min. screening method and also improve the data evaluation Then, the analysis was performed in Data Dependent procedures. Scan (ddMS2) to confirm the previous identification. The information collected from HRAMS allows the Operating in ESI- mode, the boromycin was detected identification of unknown compounds with relatively at 878.52069  m/z (Fig.  1a) with a retention time of high degree of confidence without reference analytical 11.05 min with a specific spectra. standards. The identification was carried out only using The boromycin profile pattern was characterized the mass spectra database. The test analyzes were car - by 879.52356 m/z (C H BN0 ); 878.52069 m/z 45 74 15 + + ried out directly on the two samples of water delivered (C H BN0 ); 877.52393 m/z (C H BN0 ) and 45 73 15 45 72 15 for the proficiency test using two analytical columns. 880.52612 m/z (C H BN0- H ). 45 74 15 These selected columns have similar characteristic; The characteristic fragments that contributed they have the same dimensions (100 × 2,1 mm) and to the putative identification were 752.42841 m/z, both were packed with spherical solid core particles. 666.58533 m/z, 155.10989 m/z and 116.07271 m/z The only difference was the particle size dimension (1.8 (Fig. 1b). and 2.6 μm). They were tested considering their opti - In ESI+ mode, the same compound was detected mal retention of a wide range of polar analytes with as [M + H] with a m/z= 880.51776 at RT 10.96 min S carpone et al. FoodContamination (2021) 8:5 Page 5 of 8 Fig. 1 Putative identification of Boromycin (C H BNO ): a Full-Scan in ESI-: experimental and theorical spectra; b ddMS2 acquisition in ESI-: 45 74 15 parental ion and fragmentation products c ddMS2 acquisition in ESI+: parental ion and fragmentation products Scarpone et al. FoodContamination (2021) 8:5 Page 6 of 8 Table 1 Estimated concentrations of the detected pesticides and the fragmentation pattern was characterized by two fragments  751.54974 and 93.07027 m/z (Fig.  1c). Robust Mean CV (%) Estimated It was detected also the boromycin adduct with Na at (mg/kg) concentration (mg/kg) 902.50049 m/z (Supplementary Information, Additional file 1b). Benalaxyl 0.041 22 0.034 Furthermore, the theoretical fragmentation pattern of Clomazone 0.010 21 0.0085 boromycin was also studied to confirm the putative iden - Emamectin 0.0090 34 0.0090 tification and to avoid to bump into isobaric compounds. Etoxazole 0.013 26 0.0098 Fenpyrazamine 0.0080 27 0.0085 Isopryrazam 0.0090 29 0.0012 Case study 2 Metrafenone 0.013 21 0.011 The results of this Proficiency Test were reported in the Penflufen 0.010 15 0.011 EUPT-FV-SM10 Final Report published by EURL. The Penthiopyrad 0.067 23 0.067 pesticides used to spike the samples, at different lev - Prosulfocarb 0.011 22 0.011 els between 0.01 mg/kg and 0.1 mg/kg, were decided Spinetoram 0.035 26 0.029 upon the Quality Control Group. The pesticide treat - Spirotetramat 0.072 38 0.012 ments were carried out just post-harvest used pesti- Sulfoxaflor 0.0090 24 0.011 cide analytical standards. The selected pesticides were: Tetramethrin 0.011 16 0.010 benalaxyl, clomazone, cyfluthrin, emamectin, etoxa - zole, fenpyrazamine, isopyrazam, metrafenone, pen- flufen, pentachloroaniline, penthiopyrad, proquinazid, prosulfocarb, spinetoram, spirotetramat, spirotetramat by ChemSpider and EFS HRAM Compound database as metabolite BYI08330-enol, sulfoxaflor and tetramethrin. proquinazid. Seventy-three laboratories participated to this Profi - The standards of these analytes were available in our ciency Test and only 69 of them submitted the results. laboratory and the identified compounds were also quan - These laboratories analysed the test item using method tified and the data processed. In Table  1, it is reported the based on gas or liquid chromatography, or both, com- estimated concentration of the identified pesticides. The bined with mass spectrometry detection. detection limit ranged between 0.0025 to 0.0050 mg/kg. Fourteen pesticides, corresponding to 78% of reported Spirotetramat-enol was also present in the test item at pesticides by laboratory, were detected in our laboratory the same range of concentration of others, but it was not by liquid chromatography analysis combined with high detected; spirotetramat-ketohydroxy was present below resolution mass spectrometry. 0.01 mg/kg and it was detected and reported and not all It was possible in our laboratory to detect and identify the laboratories reported the concentration. 14 pesticides: benalaxyl, clomazone, emamectin, etoxazole, fen- Case study 3 pyrazamine, isopyrazam, metrafenone, penflufen, pen - For this test, it is available at this moment only a pre- thiopyrad, prosulfocarb, spinetoram, spirotetramat, liminary report by EURL, where it was reported which sulfoxaflor and tetramethrin. Three pesticides were not analytes were present in the test item, the list of labora- detected: cyfluthrin, pentachloroaniline and proquinazid. tories that participated and the table with the compounds In particular, the detection of cyfluthrin and pentachlo - detected and non-detected by each laboratories. Only roaniline was failure because they are analytes suitable 17% of laboratories reported the entire results from all for the GC-MS determination. analytes, while 83% reported fewer analytes. Instead, proquinazid was not detected with our screen- In this case study, no matrix reference was available as ing method because its signal was low, and it was not in a real routine case analysis. depart from the blank matrix with the set statistical The pesticide treatments were carried out post-harvest parameters (p-value < 0.05 and log fold charge =1). The using standard analytical solutions. The spiked pesticides differential analysis showed about 600 compounds to be were 17: alachlor, cyanofenphos, diuron, dodemorph, identified, but the proquinazid signal was not included endrin, fluacryprim, fonofos, isoprocarb, metamitron, in this group, but present among the matrix signals. In metazachlor, metobromuron, monolinuron, prometryne, retrospective analysis, it was present a low signal of the propazine, propoxur, simazine, and tetrachlorvinphos. parent ion at [M + H]   = 373.04000 m/z and also its ion In this Proficiency Test, 64 laboratories participated fragments at 330.99480 and 288.94750 m/z, recognized and only three of them have not reported results. Thir - teen pesticides (76% of the reported pesticides by S carpone et al. FoodContamination (2021) 8:5 Page 7 of 8 laboratory) were detected by our screening method using In the case of extremely complex matrices, like onion, it Compound Discoverer 2.0: diuron, dodemorph, fonofos, is possible, that the differential analysis cannot highlight isoprocarb, metamitron, metazachlor, metobromuron, signals of some analytes at low concentrations that could monolinuron, prometryne, propazine, propoxur, sima- be covered by the matrix signals. zine, and tetrachlorvinphos (Supplementary Information, The databases contemplate only compounds that Additional  file  2). Alachlor, cyanofenphos, endrin and present parent ion and this is a limitation for the com- fluacryprim were not reported. pounds instable that present a high fragmentation. The Alachlor was not identified by Compound Dis - libraries could be implemented with the fragmentation coverer 2.0 because it was not present its molecular pattern of the compound to help the identification. ion. Our system identified its characteristic fragment ([M + H]   = 238.09875 m/z) as ketamine (C H ClNO) 13 16 at 8.072 min as retention time. Therefore, the putative Conclusion identification of this analyte, that is not a stable com - The results of these studies are most encouraging and pound, was compromised by the absence of the molec- this method could be used more and more as screens/ ular ion and the fragment did not match to the library filters, to make routine laboratory work easier and database. The database should be implemented with frag - faster. mentation pattern of the compound to better recognize a These three interlaboratory tests on wide-scope wide range of analytes. screening methods showed that such an approach can The detection of cyanofenphos and fluacryprim was substantially expand the scope of xenobiotics analysis. affected by an error of the laboratory operator. Onion is This is especially useful for unknown compounds, like a very complex matrix, that presented a high number of some pesticides, not frequently found in food and feed, compounds to be identified (about 395 compounds) and or not monitored by the laboratories because they are in this case it is necessary great attention in the elabora- not part of the EU-Coordinated Programme. The use of tion and interpretation of the results and at last but not screening methods can greatly increase the chance of least, the adequately trained operator. detecting less commonly found xenobiotics. However, In a retrospective analysis, the cyanofen- the tests also revealed that verification of the screening phos (C H N OPS) showed a low signal methods performance (i.e. validation) are necessary to 9 15 5 2 [M + H]  = 304.05490 m/z at retention time of 8.558 min increase the reliability of such methods and guidelines with the presence also of its characteristic fragments for such validation have been prepared and included in (156.98710 and 258.01360 m/z). This signal was recog - the Document SANTE/12682/2019 (European Com- nized by Chemspider database. In the case of fluacryprim mission Directorate General for Health and Food (C H F N O ), it was only detected by 20 laboratories. Safety, 2019). 20 21 3 2 5 Its pseudomolecular ion ([M + H]   = 427.14670 m/z) Furthermore, this screening method did not give false was present at retention time of 9.045 min and also the positive results. characteristic profile fragments in ddMS2 acquisition In the near future, the perspectives should be focused (205.08582 and 145.06476 m/z). This molecule was cor - on the implementation and development of database rectly identified by Compound Discoverer 2.0. and software to help the identification of xenobiotics. Finally, endrin is not an analyte suitable for the LC-MS determination and for this reason it was not detected. Supplementary Information These identifications were carried out without the ana - The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s40550- 021- 00086-x. lytical standards and the concentrations, for this reason, were not reported. Additional file 1. Differential analysis and result table for sample material 1 (a) and sample material 2 (b) by means of Compound Discoverer 2.0. Limitations Additional file 2. Identification results for the onion sample by means of This non targeted method presented in this paper pre - Compound Discoverer 2.0. sents some limitations. For these kinds of analyses, it is really important the Acknowledgements attention in the processing of data and the expertise of Not applicable. the operator. The data collected with this approach are Authors’ contributions many and need to be evaluated with the use of specific All authors that contributed to the research paper are listed at the beginning. software to process data. The author(s) read and approved the final manuscript. Scarpone et al. FoodContamination (2021) 8:5 Page 8 of 8 Funding non-targeted antitussive adulterants in herbal medicines by Q-Orbitrap Not applicable. HRAMS and screening database. Int J Mass Spectrom 447:116250. https:// doi. org/ 10. 1016/j. ijms. 2019. 116250 Availability of data and materials Herrera-Lopez S, Hernando MD, García-Calvo E, Fernández-Alba AR, Ulasze- The RIKILT data are not accessible without a PT registration. It is possible to wska MM (2014) Simultaneous screening of targeted and nontargeted access the data with a specific request. Our lab code was 9837. contaminants using an LC-QTOF-MS system and automated MS/MS For the EURL-FV10 (lab code 067), the data are available in https:// www. eurl- library searching. J Mass Spectrom 49:878–893. https:// doi. org/ 10. 1002/ pesti cides. eu/ docs/ public/ tmplt_ artic le. asp? CntID= 1082& LabID= 500& Lang= jms. 3428 EN. Hollender J, Van Bavel B, Dulio V, Farmen E, Furtmann K, Koschorreck J, Kunkel Instead for EURL-FV12 (lab code 053), the data are available in https:// www. U, Krauss M, Munthe J, Schlabach M, Slobodnik J, Stroomberg G, Ternes T, eurl- pesti cides. eu/ user files/ file// Preli minary_ Report_ SM12. pdf . Thomaidis NS, Togola A, Tornero V (2019) High resolution mass spectrom- etry-based non-target screening can support regulatory environmental monitoring and chemicals management. Environ Sci Eur 31(1):42–53. Declarations https:// doi. org/ 10. 1186/ s12302- 019- 0225-x Hu Q, Noll RJ, Li H, Makarov A, Hardmanc M, Cooks RG (2005) The Orbitrap: a Competing interests new mass spectrometer. J Mass Spectrom 40:430–443. https:// doi. org/ 10. Rossana Scarpone, Rachele Rocchi, Federico Bacà, Roberta Rosato, Francesco 1002/ jms. 856 Chiumiento and Giacomo Migliorati declare that they have no conflict of Kaufmann A (2014) Combining UHPLC and high-resolution MS: a viable interest. approach for the analysis of complex samples? TrAC Trend Anal Chem 63:113–128. https:// doi. org/ 10. 1016/j. trac. 2014. 06. 025 Received: 12 February 2021 Accepted: 18 October 2021 Kaufmann A, Butcher P, Maden K, Walker S, Widmer M (2015) Reliability of veterinary drug residue confirmation: high resolution mass spectrometry versus tandem mass spectrometry. Anal Chim Acta 856:54–67. https:// doi. org/ 10. 1016/j. aca. 2014. 11. 034 Ki NY, Jisu H, Beom KH, Kyung KH, Bong JM, Han BO, Jongki H (2019) Rapid References screening of sulfonamides in dietary supplements based on extracted Bletsou AA, Junho J, Hollender J, Archontaki E, Thomaidis NS (2015) Targeted common ion chromatogram and neutral loss scan by LC-Q/TOF-mass and non-targeted liquid chromatography-mass spectrometric workflows spectrometry. J Food Drug Anal 27:164–174. https:// doi. org/ 10. 1016/j. for identification of transformation products of emerging pollutants in jfda. 2018. 08. 006 the aquatic environment. TrAC Trend Anal Chem 66:32–44. https:// doi. Makarov A, Scigelova M (2010) Coupling liquid chromatography to Orbitrap org/ 10. 1016/j. trac. 2014. 11. 009 mass spectrometry. J Chromatogr A:3938–3945. https:// doi. org/ 10. 1016/j. Cavanna D, Righetti L, Elliott C, Suman M (2018) The scientific challenges in chroma. 2010. 02. 022 moving from targeted to non-targeted mass spectrometric methods Milman BL (2015) General principles of identification by mass spectrometry. for food fraud analysis: a proposed validation workflow to bring about a TrAC Trend Anal Chem 69:24–33. https:// doi. org/ 10. 1016/j. trac. 2014. 12. harmonized approach. Trends Food Sci Technol 80:223–241. https:// doi. org/ 10. 1016/j. tifs. 2018. 08. 007 Mol HGJ, Zomer P, De Koning M (2012) Qualitative aspects and validation of a Commission Decision n.657/EC (2002) Implementing Council Directive 96/23/ screening method for pesticides in vegetables and fruits based on liquid EC concerning the performance of analytical methods and the interpre- chromatography coupled to full scan high resolution (Orbitrap) mass tation of results spectrometry. Anal Bioanal Chem 403:2891–2908. https:// doi. org/ 10. EN 15662:2008 (2008) European Committee for Standardization (2008) CEN EN 1007/ s00216- 012- 6100-x 15662: Foods of plant origin-determination of pesticide residues using Peters RJB, Stolker AAM, Mol JGJ, Lommen A, Lyris E, Angelis Y, Vonaparti A, Sta- GC-MS and/or LC-MS/MS following acetonitrile extraction/partition and mou M, Georgakopoulos C, Nielen MWF (2010) Screening in veterinary clean-up by dispersive SPE-QuEChERS method. https:// www. en- stand drug analysis and sports doping control based on full-scan, accurate- ard. eu/ csn- en- 15662- foods- of- plant- origin- multi method- for- the- deter mass spectrometry. TrAC Trend Anal Chem 29:1250–1268. https:// doi. org/ minat ion- of- pesti cide- resid ues- using- gc- and- lc- based- analy sis- follo 10. 1016/j. trac. 2010. 07. 012 wing- aceto nitri le- extra ction- parti tioni ng- and- clean- up- by- dispe rsive- Riedl J, Esslinger S, Fauhl-Hassek C (2015) Review of validation and reporting spe- modul ar- quech ers- method/ of non-targeted fingerprinting approaches for food authentication. Anal European Commission Directorate General for Health and Food Safety (2019) Chim Acta 885:17–32. https:// doi. org/ 10. 1016/j. aca. 2015. 06. 003 Analytical quality control and method validation for pesticide residues Scarpone R, Rosato R, Chiumiento F, Cipolletti C, Sergi M, Compagnone D analysis in food and feed (SANTE/12682/2019), pp 1–48 (2020) Preliminary study to develop an alternative method for the non- Ferrer C, Lozano A, Agüera AA, Jiménez G, Fernández-Alba AR (2011) targeted determination of xenobiotics in food by means of ultra high Overcoming matrix effects using the dilution approach in multiresidue performance liquid chromatography coupled to high resolution and methods for fruits and vegetables. J Chromatogr A 1218(42):7634–7639. accuracy mass spectrometry. Food Anal Methods 13:1099–1110. https:// https:// doi. org/ 10. 1016/j. chroma. 2011. 07. 033 doi. org/ 10. 1007/ s12161- 020- 01727-1 Furey A, Moriarty M, Bane V, Kinsella B, Lehane M (2013) Ion suppression; a Uclés S, Lozano A, Sosa A, Parrilla Vázquez P, Valverde A, Fernández-Alba AR critical review on causes, evaluation, prevention and applications. Talanta (2017) Matrix interference evaluation employing GC and LC coupled to 115:104–122. https:// doi. org/ 10. 1016/j. talan ta. 2013. 03. 048 triple quadrupole tandem mass spectrometry. Talanta 174:72–81. https:// García-Reyes JF, Hernando MD, Molina-Díaz A, Fernández-Alba AR (2007) doi. org/ 10. 1016/j. talan ta. 2017. 05. 068 Comprehensive screening of target, non-target and unknown pesticides US Pharmacopeia (2016) Guidance on developing and validating non-tar- in food by LC-TOF-MS. TrAC Trend Anal Chem 26:828–841. https:// doi. geted methods for adulteration detection. Appendix XVIII:2053–2067 org/ 10. 1016/j. trac. 2007. 06. 006 Zubarev RA, Makarov A (2013) Orbitrap mass spectrometry. Anal Chem Gosetti F, Mazzucco E, Zampieri D, Gennaro MC (2010) Signal suppression/ 85:5288–5296. https:// doi. org/ 10. 1021/ ac400 1223 enhancement in high-performance liquid chromatography tandem mass spectrometry. J Chromatogr A 1217:3929–3937. https:// doi. org/ 10. 1016/j. Publisher’s Note chroma. 2009. 11. 060 Springer Nature remains neutral with regard to jurisdictional claims in pub- Guo C, Liping G, Weijian W, Jiawei L, Liming Z, Sheng X, Yanxia Z, Ruiqing X, lished maps and institutional affiliations. Xunjie Z, Feng S (2020) Rapid screening and identification of targeted or

Journal

International Journal of Food ContaminationSpringer Journals

Published: Nov 8, 2021

Keywords: High resolution accurate mass spectrometry; Non-target screening; Food; Xenobiotics; Differential analysis

There are no references for this article.