- Email: [email protected]
Chemical profiling of herbarium samples of solanum (Solanaceae) using mass spectrometry
Phytochemistry Letters 36 (2020) 99–105
Contents lists available at ScienceDirect
Phytochemistry Letters journal homepage: www.elsevier.com/locate/phytol
Chemical profiling of herbarium samples of solanum (Solanaceae) using mass spectrometry
T
João Victor Mendes Resendea, Najla M.D. de Sáb, Marcelo Trovó Lopes de Oliveirac, Rosana Conrado Lopesc, Rafael Garrettb, Ricardo Moreira Borgesa,* a
Natural Product Research Institute Walter Mors (IPPN), Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil Chemistry Institute, Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil c Institute of Biology, Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil b
ARTICLE INFO
ABSTRACT
Keywords: Herbarium LC–MS Molecular network Solanaceae Solanum Chemical space
Herbarium collections are broadly available for scientific evaluations but surprisingly few studies explored their rich chemical diversity. Considering the systematic organization and the storage conditions Herbarium collections are kept, we wonder if there are still secondary metabolites of interest after years of storage and how this data could be used to discriminate different species within the same genus. Thus, using a set of 25 Solanum (Solanaceae) samples selected randomly from the RFA Herbarium we designed a fast method to extract and analyze them using LC-HRMS/MS. This pilot study shows the broad chemical space of samples stored in Herbarium collections. Also, we performed multivariate analysis (PCA and PLS-DA) using data from two species, S. argenteum and S. pseudoquina, to evaluate if we could discriminate them based on their chemical profiles and we successfully showed sample grouping despite even 10 years of difference between their collection and their different collection sites. Thus, herbarium exsiccates was proven to be a reliable source of samples for NP chemistry studies. By this means, we make a plea in favor of the use of chemical profiling as a tool for taxonomists in collaboration with chemists for classification studies and to consider keeping an extract collection along with the exsiccates.
1. Introduction Chemical profiling of raw extracts using Mass Spectrometry (MS) has being successfully implemented as routine in natural products (NP) chemistry laboratories. When such profiling studies are made to identify known compounds the term dereplication is used (Beutler, 1990; Yang et al., 2013). Advantages, such as high sensitivity and selectivity, point to MS as the technique of choice for NP dereplication because it enables the identification of dozens of compounds among hundreds of detected features (m/z values). Using state-of-art methods to aid compound identification from a MS dataset, researchers are reaching an impressive increase of annotation rates (Beauxis and Genta-Jouve, 2019; Borges et al., 2018; da Silva et al., 2018; Ernst et al., 2019; Rogers et al., 2019; Silva et al., 2018; Soares et al., 2019; Wandy et al., 2017). The advent of the visualization of fragmentograms through vector similarities into networks was a milestone that enabled an intuitive propagation and compound annotation of even unknown compounds (unknowns); the term molecular networking (MN) was coined and GNPS is becoming a valuable tool for compound annotation of NP.
⁎
One major bottleneck for compound identification is the inexistence of a comprehensive tandem mass spectrometry database (DB). Thus, dereplication might be best defined as not the identification of all known compounds, but the identification of well-cataloged known compounds in a certain sample. Again, GNPS enables users to contribute to a massive DB (Wang et al., 2016) directly with their mass spectra even when they are putatively identified. In any case, it’s suggested the use of a species (genera or family)-filtered DB of the studied organism for obvious chemotaxonomic reasons. On the other hand, our group has been developing a DB independent method for annotation of MS data using molecular formula restrictions for both precursors and their fragments (Borges et al., 2018; Soares et al., 2019). This method relies mainly on the expected fragmentation patterns of classes of NP and enables users to search even for unknown compounds that fit certain molecular formula restrictions widening the range of identification over DB dependent procedures. Here we show the use of liquid chromatography-high resolution tandem mass spectrometry (LC-HRMS/MS) data of raw extracts of Solanum L. (Solanaceae) samples housed in the RFA Herbarium (Federal
Corresponding author. E-mail address: [email protected] (R. Moreira Borges).
https://doi.org/10.1016/j.phytol.2020.01.021 Received 17 November 2019; Received in revised form 24 January 2020; Accepted 31 January 2020 1874-3900/ © 2020 Published by Elsevier Ltd on behalf of Phytochemical Society of Europe.
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
Fig. 1. The pie chart shows the various classes of metabolites putatively identified by tandem mass spectrometry in the study.
Fig. 2. PCA (A) and PLS-DA (B) score plots for the discrimination of S. pseudoquina and S. argenteum species; the PLS-DA performance plots (C) and VIP scores (D). Data analysis was performed at Metaboanalyst.
University of Rio de Janeiro, Brazil) in order to test the efficiency of previously stored botanical collections for chemotaxonomic studies. The hypothesis revolves around the idea that many secondary metabolites should be present after the common 40−60 °C oven drying and
often long storage times. The characterization of degradation products that might have been produced within this period is beyond the scope of this paper. It also relies on the fact that surprisingly few studies have used herbarium samples from a natural products point of view (Afzan 100
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
Table 1 VIP list of candidates with scores > 2 from the PLS-DA analysis identified at GNPS. m/z
Compound ID
Cluster ID (GNPS)
727.205
5-hydroxy-3-[(2S,3R,4R,5S)-3-hydroxy-5-(hydroxymethyl)-4-[(2S,3R,4S,5S,6R)-3,4,5-trihydroxy-6-(hydroxymethyl)oxan-2-yl]oxyoxolan-2-yl]oxy-2(4-hydroxyphenyl)-7-[(2S,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxychromen-4-one 3-[(2S,3R,4S,5R,6R)-6-[[(2R,3R,4R,5S,6S)-3,5-dihydroxy-6-methyl-4-[(2S,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxyoxan-2-yl] oxymethyl]-3,4,5-trihydroxyoxan-2-yl]oxy-5,7-dihydroxy-2-(4-hydroxyphenyl)chromen-4-one no MSMS spectra (not in any network) spirosol-5-en-3-yl O-6-deoxy-alpha-L-mannopyranosyl-(1- > 2)-O-[6-deoxy-alpha-L-mannopyranosyl-(1- > 4)]- beta-D-Glucopyranoside (NCGC00179912-02) Not identified - epiminocholesta acetate derivative 3-[4,5-dihydroxy-6-(hydroxymethyl)-3-[3,4,5-trihydroxy-6-(hydroxymethyl)oxan-2-yl]oxyoxan-2-yl]oxy-5,7-dihydroxy-2-(4-hydroxyphenyl) chromen-4-one Not identified - triglycoside spirosolane derivative Not identified - triglycoside epiminocholestane derivative 2-(3,4-dihydroxyphenyl)-3-[(2S,3R,4S,5S,6R)-4,5-dihydroxy-3-[(2R,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxy-6-[[(2R,3R,4R,5R,6S)3,4,5-trihydroxy-6-methyloxan-2-yl]oxymethyl]oxan-2-yl]oxy-5,7-dihydroxychromen-4-one Not identified - triglycoside epiminocholestane derivative Not identified Not identified - triglycoside epiminocholestane derivative no MSMS spectra (not in any network) Not identified - triglycoside epiminocholestane derivative kaempferol-3-O-rutinoside | Nicotiflorin (ReSpect:PT104230)
87
741.219 505.347 912.499 486.316 611.158 926.509 882.483 757.211 868.501 347.244 884.497 926.51 866.488 595.161
6 – 198 277 181 54 32 202 1 88 2 – 222 12
Fig. 3. MN evidencing glycoside flavones and glycoalkaloids from S. pseudoquina and S. argenteum species. VIP scores from PLS-DA analysis are shown together with molecular structure.
et al., 2019; Jafari et al., 2018; Katavi, 2005; Tasca et al., 2018) and none of those targeting insights into the chemical space of the collected samples. With more than 2700 species, Solanaceae is a highly diverse plant family. At least 1330 of those species belong to Solanum, by far the richest genus of the family (Yadav et al., 2016). Solanum species are widely distributed in the Neotropics (Kaunda and Zhang, 2019). Our ongoing efforts to profile the chemical space of Solanum, which is a model for taxonomic and evolutionary studies (Soltis and Soltis, 2000), led us to assay the use of herbarium samples for compound identification as a model that can also be used for other diverse plant groups. Considering the richness of the genus, many diverse compounds are expected to be found. Among those, phenolic compounds and their glycosides, carotenoids, steroids, glycoalkaloids and tropane alkaloids are examples of bioactive and economic relevant compounds.
yielded around 2−5 mg of dry residue per sample. Such a yield points to the possibility of increasing the amount of the plant material up to 200 mg (practical limit for the 1.5 mL microtube for extraction/ homogenization) to obtain around 10 mg of dry residue to enable deeper profiling analysis using NMR and different Bioassays for prospection. It was shown that no difference in the antibacterial activity was found when comparing long storage with fresh samples (Eloff, 1999) proving promising perspectives for the use of herbarium samples in drug discovery. Sample availability in each specimen needs to be evaluated for collection in reverence to invaluable herbarium specimens, but then to collect only one part of a leave might underrepresent that sample. This method conceives the destruction of the collected sample and this must be kept in mind. Though, only a minute fraction of the extracts is used for the proposed analytical platform and the rest might be kept under safe storage conditions to compose a bank of extracts. Due to the sheer size of the acquired LC–MS/MS dataset, positiveand negative-ESI sub-experiments were separated using MSConvert and MN were directly used to visualize the chemical profile. In this section,
2. Results and discussion The extraction procedure used in this study was straightforward and 101
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
Fig. 4. The biggest MN with the epiminocholestane glycoalkaloids from S. pseudoquina and S. argenteum species. VIP scores from PLS-DA analysis are shown together with molecular structure.
we aimed for a broad visualization of the chemical profile of the whole sample set. Thus we submitted the pre-processed dataset from all the 25 samples for molecular networking (Figure S1). 185 nodes (63 in negative and 133 in positive ion mode) among the calculated networks was annotated using by GNPS’ routing through DB matching and, among those, 83 compounds were putatively identified to the Metabolomics Standards Initiative level 2 (Sumner et al., 2007) with mass errors below 7 ppm and visual validation of each fragmentogram. Phenolics compounds, aminoacids, fatty acids and other metabolites were identified (Fig. 1). As expected, many phenylpropanoid derivatives were annotated despite saponins (including glycoalkaloids) being markers for Solanum species. Probably it is because saponins are less documented compared to phenylpropanoids, which are commonly identified in other plants and cataloged in DB. The most populated MN found under ESI positive mode contained features identified as saponins: (3beta,22R,25R)-26-(beta-DGlucopyranosyloxy)-22-hydroxyfurost-5-en-3-yl 6-deoxy-alpha-L-mannopyranosyl-(1- > 2)-[beta-D-glucopyranosyl-(1- > 4)]-beta-D-glucopyranoside (C51H84O23); (3beta,22beta,25R)-26-(beta-D-glucopyranosyloxy)22-hydroxyfurost-5-en-3-yl O-6-deoxy-alpha-L-mannopyranosyl-(1- > 2)O-[6-deoxy-alpha-L-mannopyranosyl-(1- > 4)]-beta-D-Glucopyranoside (C51H84O22); spirosol-5-en-3-yl O-6-deoxy-alpha-L-mannopyranosyl-(1> 2)-O-[6-deoxy-alpha-L-mannopyranosyl-(1- > 4)]-beta-DGlucopyranoside (C45H73NO15). Spread among other MNs in both positive and negative ion modes various phenolics were detected, including different flavonoids and their mono-, di- and triglycoside derivatives, many phenylpropanoids, some benzoic acid derivatives and coumarins. Flavones, such as kaempferol, luteolin and quercetin (flavonol) were found to be the main aglycone among the glycosides with fragments at m/z 287.05, 303.05, respectively under positive ion mode. Among the identified
phenylpropanoids, caffeic acid, p-coumaric acid, ferulic acid and 3,5Dimethoxy-4-hydroxycinnamic acid and others were identified together with theirs quinic acid conjugates; the fragments at m/z 191.06 and 173.04 are characteristic under negative ion mode. 2.1. Species-based grouping through mass spectrometry-multivariate analysis Once the detection of secondary metabolites from Herbarium samples was proven to be possible, we wondered if it would be possible to discriminate the different plant species based on their chemical profile using multivariate analysis. LC-HRMS data from S. pseudoquina and S. argenteum species (n = 3) were processed and aligned in MZMine and the resulting. csv table was exported for the multivariate analysis calculations at Metaboanalyst web platform. Despite only three replicates from the two plant species were available for the study, the authors strongly recommend that a bigger number of replicates should be pursuit to ensure a more significant statistical meaning in the multivariate analysis, especially for chemotaxonomic studies. PCA and PLS-DA score plots indicated a clear trend for the distinction between S. pseudoquina and S. argenteum species even though 10 years of difference between their collection and their different collection sites (Fig. 2A and B, respectively). Components 1 and 2 explained more than 60 % of variance. By using these components, the PLS-DA analysis showed r2 = 0.999 and q2 = 0.555 (Fig. 2C), indicating that the model was reliable. To aid compound annotation, the same processed dataset submitted to MetaboAnalyst (with S. pseudoquina and S. argenteum) was then submitted to MN in GNPS. Even though it is impossible to assure the compounds identity only from MS data, because substitutions might occur at different positions in a molecule, key fragments are used to validate diagnostic features of certain 102
103
360 1370 1552 7276 3903 3824 359 5847 7185 473 2037 4687 6261 8939 6208 2991 9249
18115 18999 5462 7855 784 448 4
P. Occhioni P. Occhioni E. Pereira P. Occhioni E. Pereira P. Occhioni P. Occhioni P. Occhioni P. Occhioni E.M. Occhioni P. Occhioni P. Occhioni P. Occhioni P. Occhioni P. Occhioni L.T.H. Dombrowski P. Occhioni
G. Hatschbach G. Hatschbach & C. Keczicki P. Occhioni P. Occhioni O.M. Barth A.M. de Barros N.G. Antas
4575 6327 4435 16362 4423 9237 4574 14647 16378 21474 6769 11344 15467 20006 15390 13664 20479 40631 11835 11832 13453 17250 29167 29108 39313
Voucher Number S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S.
argenteum Dunal* argenteum Dunal* argenteum Dunal* asperum Rich. asperum Rich. caavurana Vell. caavurana Vell. campaniforme Roem. & Schult. campaniforme Roem. & Schult. capsicoides All. cinnamomeum Sendtn. cladotrichum Dunal cladotrichum Dunal didymium Dunal didymium Dunal pseudoquina A. St.-Hil.* pseudoquina A. St.-Hil.* pseudoquina A. St.-Hil.* sanctae-catharinae Dunal schwackeanum L.B.Sm. & Downs swartzianum Roem. & Schult. swartzianum Roem. & Schult. vaillanti Dunal variabile Mart. sisymbriifolium Lam.
Specie
12/14/1967 04/10/1968 02/28/1973 12/12/1975 01/12/1960 09/22/1990 06/11/2009
03/22/1945 07/21/1956 "—" 01/05/1975 06/24/1958 x/01/1970 04/25/1945 01/25/1974 04/28/1975 01/14/1986 x/11/1960 11/19/1971 09/16/1974 05/25/1979 09/14/1974 x/10/1969 x/11/1980
Filing Date
Rio de Janeiro, Estr. Vista Chinesa (Rio de Janeiro state) Parque Nacional da Serra dos Órgãos (Rio de Janeiro state) Rio de Janeiro, Avenida Niemeyer (Rio de Janeiro state) Linhares, Estr. para Regência margem Rio Doce (Espírito Santo state) Jacarepaguá, Três Rios (Rio de Janeiro state) Guanabara, Restinga da Tijuca (Rio de Janeiro state) Estrada do Redentor (Rio de Janeiro state) Serra dos Órgãos; Estrada Velha Teresópolis X Petrópolis (Rio de Janeiro state) BR 262. km.195 (Espírito Santo state) Ilha da Marambaia, Praia Grande (Rio de Janeiro state) Serra dos órgãos (Rio de Janeiro state) Estr. BR-101, KM. 48 alto da serra (Paraná state) Parque Nacional de Itatiaia Estr. Registro Planalto Km 7 e 8 a 1500 metros de altitude (Rio de Janeiro state) Parque Nacional de Itatiaia, Lago Azul (Rio de Janeiro state) Parque Nacional de Itatiaia, caminho do maromba (Rio de Janeiro state) Curitiba, Capão da Imbuia (Paraná state) Parque Nacional de Itatiaia, Estr. para as macieiras Km. 14 (Rio de Janeiro state) Simão Pereira (Minas Gerais state) Guaraqueçaba, Rio de Cedro (Paraná state) Guaraqueçaba (Paraná state) Serra de Friburgo, 1000 m altitude (Rio de Janeiro state) Parque Nacional de Itatiaia, Estr. Registro (Rio de Janeiro state) Casa de J. Spanner, PNI (Rio de Janeiro state) Estr. de São Paulo, Jardim São Francisco (São Paulo state) Seropédica, Pedreira ENFOL (Rio de Janeiro state)
Provenance
* S. pseudoquina and S. argenteum occurred three times in this first pilot batch. These samples were used to access speciation within this proposed approach.
Collector Number
Collector ID
Table 2 List of samples collected from the RFA Herbarium.
J.V. Mendes Resende, et al.
Phytochemistry Letters 36 (2020) 99–105
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
structure. Thus, candidate compounds likely to be taxonomic markers from the VIP plot list from the PLS-DA analysis (Fig. 2D) were identified (Table 1). Three main MN were found to contain all the fifteen VIP candidates and they are separated into clusters of flavonoid glycosides and many glycoalkaloids (Figs. 3 and 4). The biggest MN shows 89 nodes of glycoalkaloids classified as epiminiocholestane derivatives according literature (Fig. 4) (Soares et al., 2019). Key fragments at m/z 253.19 and 271.20 (or at m/z 255.09 and 273.10 in positive ion mode) can be highlighted as characteristic of the hydroxy-tetracyclic structure of steroids (Man et al., 2009). The PLS-DA results suggest that a more suitable prediction model could be reached using a higher number of replicates. On the other hand, power analysis using this current study as a pilot indicated that to reach a power of 0.8, close to 50 replicates would be required (Figure S2). Such a high sample size is hard to achieve for most of NP studies, but it is important to be able to understand such parameter to establish the significance of the proposed study. Herbarium exsiccates was proven to be a reliable source of samples for NP chemistry studies. Even assuming some degradation due to heat drying and long-term storage (over 50 years in some cases), many meaningful metabolites were kept stable. Little can be speculated about the stability in situ of compounds in dried plants and different investigation shall be designed accordingly. Interestingly, samples from the same species (e.g species of S. argenteum and S. pseudoquina) grouped together even with 10 years of difference between their collection and their collection site (Fig. 2A and B; Figure S3). The same is true when all the 25 samples are analyzed together, they cluster in one group showing a trend to the occurrence of similar compounds even after decades of storage in situ. The extraction of up to 50 mg with 3 ml of solvent yielded a satisfactory amount of dry residue for LCeMS/MS analysis. Depending on the complexity of the raw extract, milligram and sub-milligram scale miniaturized procedure should be considered for a bioguided fractionation approach. Still, the results shown here indicates a promising approach for taxonomists in collaboration with NP researchers to include the LCeMS/MS based chemical profile within classification studies and to consider keeping an extract collection along with the exsiccates. Higher number of samples is strongly suggested for statistical significance for species-driven classification and grouping.
formic acid in methanol in gradient elution mode (0 − 0.5 min 5 % B; 0.5 − 3 min 30 % B; 3 − 12 min, 70 % B; 12 − 15 min 95 % B; 15 − 17 min 95 % B; and 17.1 − 20 min 5 % B). The UHPLC system was coupled to a QExactive Plus high resolution and accurate mass spectrometer (Thermo Scientific) equipped with an electrospray ion source (ESI) operating in both positive and negative ionization modes. Source ionization parameters were: spray voltage 3.8 kV; capillary temperature 300 °C; S-Lens level 70, sheath and auxiliary gas 40 and 25 (arbitrary units), respectively. Samples were analyzed in the scan range of m/z 150–1500 at resolution of 70,000 (positive and negative full scan) followed by data-dependent MS2 (ddMS2 Top3 experiments) using a resolution of 17,500 and normalized collision energy (NCE) stepped 35 − 50 %. The acquired data were converted to mzXML files using the software MSConvert (ProteoWizard; available at proteowizard.sourceforge.net/ tools.shtml); polarity filter was used to separate positive and negative sub-experiments to simplify the MN interpretation. MZMine was used for feature-based processing before submitting the exported files into GNPS for MN calculations. MZMine parameters are given as SI (Table S1). The networks were plotted using the Cytoscape (http://www. cytoscape.org). The MN for both positive and negative ionization modes is accessible at the GNPS Web site with the following links: (positive) https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task= b0a773793d7e4774938445d5f9d84f67; (negative) https://gnps.ucsd. edu/ProteoSAFe/status.jsp?task= c7e62fbb6b774b0599a082aa83738b16. GNPS parameters are given as SI (Table S2). For further evaluation, the two species with tree samples each were selected as a pilot study to probe species-driven grouping capability of this method; Solanum argenteum and S. pseudoquina. MZMine was used to align and export the matrix to be analyzed at the Metaboanalyst (web server at https://www.metaboanalyst.ca/) for multivariate analysis under the protocol Statistical Analysis and Power Analysis; for multivariate analysis analysis the raw data was used without polarity separation. Metaboanalyst parameters are given as SI (Figure S4). The performance of the PLS-DA method was evaluated using the parameters r2 and q2 for fitting and predictability. Additionally, this dataset used for multivariate analysis was submitted to GNPS for a simpler MN; this is accessible in the following: https://gnps.ucsd.edu/ProteoSAFe/ status.jsp?task=6a1c707e017345cabd6f3c892fce2c73.
3. Experimental
Declaration of Competing Interest
3.1. Sample collection and sample preparation
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Two or three random loose leaves from 15 Solanum specimens (25 samples) were collected from the RFA Herbarium (Institute of BiologyFederal University of Rio de Janeiro, Brazil; Table 2). Note that the collection of this sample amount was not detrimental to the vouchers. 50 mg of each sample was homogenized with 80 % methanol (1.50 mL) and sonicated for 8 min at room temperature. The extraction mixtures were centrifuged at 10,000 g for 10 min at 4 °C and the supernatants were filtered and concentrated to dryness under vacuum. This procedure was repeated twice for maximum recovery. After resuspension in methanol (1.00 mL), aliquots of 100.00 μl were diluted in 50 % methanol (0.50 mL) for MS data acquisition.
Acknowledgments This research was supported in part by the PIBIC/CNPQ, CAPES and FAPERJ funding agencies. We would like to acknowledge the RFA Herbarium of UFRJ for kindly sharing their samples for this study and we hope this study serves as a model for many others using exsiccate collections. Appendix A. Supplementary data
3.2. Liquid chromatography-high resolution tandem mass spectrometry analysis
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.phytol.2020.01.021.
Liquid chromatography analysis was performed on a Dionex UltiMate 3000 UHPLC system (Thermo Scientific) consisting of a quaternary solvent delivery pump and a column oven compartment. Samples (5 μL) were injected using a TriPlus RSH Autosampler (Thermo Scientific) and separated on a Thermo Hypersil GOLD RP C18 column (2.1 × 100 mm, 3.0 μm particle size) at 350 μL min−1 maintained at 40 °C. The mobile phase consisted of (A) 0.1 % formic acid and (B) 0.1 %
References Afzan, A., Bréant, L., Bellstedt, D.U., Grant, J.R., Queiroz, E.F., Wolfender, J.L., Kissling, J., 2019. Can biochemical phenotype, obtained from herbarium samples, help taxonomic decisions? – a case study using Gentianaceae. Taxon. Beauxis, Y., Genta-Jouve, G., 2019. MetWork: a web server for natural products
104
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al. anticipation. Bioinformatics 35, 1795–1796. Beutler, J.A., 1990. Dereplication of phorbol bioactives: lyngbya majuscula and Croton cuneatus. J. Nat. Prod. 53, 867–874. Borges, R.M., Taujale, R., de Souza, J.S., de Andrade Bezerra, T., Silva, E.L.E., Herzog, R., Ponce, F.V., Wolfender, J.L., Edison, A.S., 2018. Dereplication of plant phenolics using a mass-spectrometry database independent method. Phytochem. Anal. 29, 601–612. da Silva, R.R., Wang, M., Nothias, L.F., van der Hooft, J.J.J., Caraballo-Rodriguez, A.M., Fox, E., Balunas, M.J., Klassen, J.L., Lopes, N.P., Dorrestein, P.C., 2018. Propagating annotations of molecular networks using in silico fragmentation. PLoS Comput. Biol. 14, e1006089. Eloff, J.N., 1999. It is possible to use herbarium specimens to screen for antibacterial components in some plants. J. Ethnopharmacol. 67, 355–360. Ernst, M., Kang, K.B., Caraballo-Rodríguez, A.M., Nothias, L.F., Wandy, J., Wang, M., Rogers, S., Medema, M.H., Dorrestein, P.C., van der Hooft, J.J.J., 2019. MolNetEnhancer: enhanced molecular networks by integrating metabolome mining and annotation tools. Metabolites 9 xx-xx. Jafari, T., Alanne, A.L., Issakainen, J., Pihlaja, K., Sinkkonen, J., 2018. Suitability of dried herbarium specimens for NMR metabolomics of mushrooms. A comparison of four species of the genera Kuehneromyces and Hypholoma (Strophariaceae). Fungal Biol. 122, 138–146. Katavi, P.L., 2005. Chemical Investigations of the Alkaloids from the Plants of the Family Elaeocarpaceae. School of Science/Natural Product Discovery (NPD) Griffith University, Faculty of Science p. 319. Kaunda, J.S., Zhang, Y.J., 2019. The genus Solanum: an ethnopharmacological, phytochemical and biological properties review. Nat. Prod. Bioprospect. 9, 77–137. Man, S., Gao, W., Zhang, Y., Jin, X., Ma, C., Huang, X., Li, Q., 2009. Characterization of steroidal saponins in saponin extract from Paris polyphylla by liquid chromatography tandem multi-stage mass spectrometry. Anal. Bioanal. Chem. 395, 495–505. Rogers, S., Ong, C.W., Wandy, J., Ernst, M., Ridder, L., van der Hooft, J.J.J., 2019. Deciphering complex metabolite mixtures by unsupervised and supervised substructure discovery and semi-automated annotation from MS/MS spectra. Faraday Discuss. Silva, E.L., Almeida-Lafeta, R.C., Borges, R.M., Staerk, D., 2018. Athenolide a, a new steroidal lactone from the leaves of Athenaea martiana (Solanaceae) determined by means of HPLC-HR-MS-SPE-NMR analysis. Chem. Biodivers. 15. Soares, V., Taujale, R., Garrett, R., da Silva, A.J.R., Borges, R.M., 2019. Extending compound identification for molecular network using the LipidXplorer database independent method: a proof of concept using glycoalkaloids from Solanum pseudoquina A. St.-Hil. Phytochem. Anal. 30, 132–138. Soltis, E.D., Soltis, P.S., 2000. Contributions of plant molecular systematics to studies of molecular evolution. Plant Mol. Biol. 42, 45–75.
Sumner, L.W., Amberg, A., Barrett, D., Beale, M.H., Beger, R., Daykin, C.A., Fan, T.W., Fiehn, O., Goodacre, R., Griffin, J.L., Hankemeier, T., Hardy, N., Harnly, J., Higashi, R., Kopka, J., Lane, A.N., Lindon, J.C., Marriott, P., Nicholls, A.W., Reily, M.D., Thaden, J.J., Viant, M.R., 2007. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) metabolomics standards initiative (MSI). Metabolomics 3, 211–221. Tasca, J.A., Smith, C.R., Burzynski, E.A., Sundberg, B.N., Lagalante, A.F., Livshultz, T., Minbiole, K.P.C., 2018. HPLC-MS detection of pyrrolizidine alkaloids and their Noxides in herbarium specimens dating back to the 1850s. Appl. Plant Sci. 6, e1143. Wandy, J., Zhu, Y., van der Hooft, J.J.J., Daly, R., Barrett, M.P., Rogers, S., 2017. Ms2lda.org: web-based topic modelling for substructure discovery in mass spectrometry. Bioinformatics. Wang, M., Carver, J.J., Phelan, V.V., Sanchez, L.M., Garg, N., Peng, Y., Nguyen, D.D., Watrous, J., Kapono, C.A., Luzzatto-Knaan, T., Porto, C., Bouslimani, A., Melnik, A.V., Meehan, M.J., Liu, W.T., Crusemann, M., Boudreau, P.D., Esquenazi, E., Sandoval-Calderon, M., Kersten, R.D., Pace, L.A., Quinn, R.A., Duncan, K.R., Hsu, C.C., Floros, D.J., Gavilan, R.G., Kleigrewe, K., Northen, T., Dutton, R.J., Parrot, D., Carlson, E.E., Aigle, B., Michelsen, C.F., Jelsbak, L., Sohlenkamp, C., Pevzner, P., Edlund, A., McLean, J., Piel, J., Murphy, B.T., Gerwick, L., Liaw, C.C., Yang, Y.L., Humpf, H.U., Maansson, M., Keyzers, R.A., Sims, A.C., Johnson, A.R., Sidebottom, A.M., Sedio, B.E., Klitgaard, A., Larson, C.B., P, C.A.B, Torres-Mendoza, D., Gonzalez, D.J., Silva, D.B., Marques, L.M., Demarque, D.P., Pociute, E., O’Neill, E.C., Briand, E., Helfrich, E.J.N., Granatosky, E.A., Glukhov, E., Ryffel, F., Houson, H., Mohimani, H., Kharbush, J.J., Zeng, Y., Vorholt, J.A., Kurita, K.L., Charusanti, P., McPhail, K.L., Nielsen, K.F., Vuong, L., Elfeki, M., Traxler, M.F., Engene, N., Koyama, N., Vining, O.B., Baric, R., Silva, R.R., Mascuch, S.J., Tomasi, S., Jenkins, S., Macherla, V., Hoffman, T., Agarwal, V., Williams, P.G., Dai, J., Neupane, R., Gurr, J., Rodriguez, A.M.C., Lamsa, A., Zhang, C., Dorrestein, K., Duggan, B.M., Almaliti, J., Allard, P.M., Phapale, P., Nothias, L.F., Alexandrov, T., Litaudon, M., Wolfender, J.L., Kyle, J.E., Metz, T.O., Peryea, T., Nguyen, D.T., VanLeer, D., Shinn, P., Jadhav, A., Muller, R., Waters, K.M., Shi, W., Liu, X., Zhang, L., Knight, R., Jensen, P.R., Palsson, B.O., Pogliano, K., Linington, R.G., Gutierrez, M., Lopes, N.P., Gerwick, W.H., Moore, B.S., Dorrestein, P.C., Bandeira, N., 2016. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nat. Biotechnol. 34, 828–837. Yadav, R., Rathi, M., Pednekar, A., Rewachandani, Y., 2016. A detailed review on Solanaceae family. Eur. J. Pharm. Med. Res. 3, 369–378. Yang, J.Y., Sanchez, L.M., Rath, C.M., Liu, X., Boudreau, P.D., Bruns, N., Glukhov, E., Wodtke, A., de Felicio, R., Fenner, A., Wong, W.R., Linington, R.G., Zhang, L., Debonsi, H.M., Gerwick, W.H., Dorrestein, P.C., 2013. Molecular networking as a dereplication strategy. J. Nat. Prod. 76, 1686–1699.
105
Contents lists available at ScienceDirect
Phytochemistry Letters journal homepage: www.elsevier.com/locate/phytol
Chemical profiling of herbarium samples of solanum (Solanaceae) using mass spectrometry
T
João Victor Mendes Resendea, Najla M.D. de Sáb, Marcelo Trovó Lopes de Oliveirac, Rosana Conrado Lopesc, Rafael Garrettb, Ricardo Moreira Borgesa,* a
Natural Product Research Institute Walter Mors (IPPN), Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil Chemistry Institute, Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil c Institute of Biology, Federal University of Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil b
ARTICLE INFO
ABSTRACT
Keywords: Herbarium LC–MS Molecular network Solanaceae Solanum Chemical space
Herbarium collections are broadly available for scientific evaluations but surprisingly few studies explored their rich chemical diversity. Considering the systematic organization and the storage conditions Herbarium collections are kept, we wonder if there are still secondary metabolites of interest after years of storage and how this data could be used to discriminate different species within the same genus. Thus, using a set of 25 Solanum (Solanaceae) samples selected randomly from the RFA Herbarium we designed a fast method to extract and analyze them using LC-HRMS/MS. This pilot study shows the broad chemical space of samples stored in Herbarium collections. Also, we performed multivariate analysis (PCA and PLS-DA) using data from two species, S. argenteum and S. pseudoquina, to evaluate if we could discriminate them based on their chemical profiles and we successfully showed sample grouping despite even 10 years of difference between their collection and their different collection sites. Thus, herbarium exsiccates was proven to be a reliable source of samples for NP chemistry studies. By this means, we make a plea in favor of the use of chemical profiling as a tool for taxonomists in collaboration with chemists for classification studies and to consider keeping an extract collection along with the exsiccates.
1. Introduction Chemical profiling of raw extracts using Mass Spectrometry (MS) has being successfully implemented as routine in natural products (NP) chemistry laboratories. When such profiling studies are made to identify known compounds the term dereplication is used (Beutler, 1990; Yang et al., 2013). Advantages, such as high sensitivity and selectivity, point to MS as the technique of choice for NP dereplication because it enables the identification of dozens of compounds among hundreds of detected features (m/z values). Using state-of-art methods to aid compound identification from a MS dataset, researchers are reaching an impressive increase of annotation rates (Beauxis and Genta-Jouve, 2019; Borges et al., 2018; da Silva et al., 2018; Ernst et al., 2019; Rogers et al., 2019; Silva et al., 2018; Soares et al., 2019; Wandy et al., 2017). The advent of the visualization of fragmentograms through vector similarities into networks was a milestone that enabled an intuitive propagation and compound annotation of even unknown compounds (unknowns); the term molecular networking (MN) was coined and GNPS is becoming a valuable tool for compound annotation of NP.
⁎
One major bottleneck for compound identification is the inexistence of a comprehensive tandem mass spectrometry database (DB). Thus, dereplication might be best defined as not the identification of all known compounds, but the identification of well-cataloged known compounds in a certain sample. Again, GNPS enables users to contribute to a massive DB (Wang et al., 2016) directly with their mass spectra even when they are putatively identified. In any case, it’s suggested the use of a species (genera or family)-filtered DB of the studied organism for obvious chemotaxonomic reasons. On the other hand, our group has been developing a DB independent method for annotation of MS data using molecular formula restrictions for both precursors and their fragments (Borges et al., 2018; Soares et al., 2019). This method relies mainly on the expected fragmentation patterns of classes of NP and enables users to search even for unknown compounds that fit certain molecular formula restrictions widening the range of identification over DB dependent procedures. Here we show the use of liquid chromatography-high resolution tandem mass spectrometry (LC-HRMS/MS) data of raw extracts of Solanum L. (Solanaceae) samples housed in the RFA Herbarium (Federal
Corresponding author. E-mail address: [email protected] (R. Moreira Borges).
https://doi.org/10.1016/j.phytol.2020.01.021 Received 17 November 2019; Received in revised form 24 January 2020; Accepted 31 January 2020 1874-3900/ © 2020 Published by Elsevier Ltd on behalf of Phytochemical Society of Europe.
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
Fig. 1. The pie chart shows the various classes of metabolites putatively identified by tandem mass spectrometry in the study.
Fig. 2. PCA (A) and PLS-DA (B) score plots for the discrimination of S. pseudoquina and S. argenteum species; the PLS-DA performance plots (C) and VIP scores (D). Data analysis was performed at Metaboanalyst.
University of Rio de Janeiro, Brazil) in order to test the efficiency of previously stored botanical collections for chemotaxonomic studies. The hypothesis revolves around the idea that many secondary metabolites should be present after the common 40−60 °C oven drying and
often long storage times. The characterization of degradation products that might have been produced within this period is beyond the scope of this paper. It also relies on the fact that surprisingly few studies have used herbarium samples from a natural products point of view (Afzan 100
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
Table 1 VIP list of candidates with scores > 2 from the PLS-DA analysis identified at GNPS. m/z
Compound ID
Cluster ID (GNPS)
727.205
5-hydroxy-3-[(2S,3R,4R,5S)-3-hydroxy-5-(hydroxymethyl)-4-[(2S,3R,4S,5S,6R)-3,4,5-trihydroxy-6-(hydroxymethyl)oxan-2-yl]oxyoxolan-2-yl]oxy-2(4-hydroxyphenyl)-7-[(2S,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxychromen-4-one 3-[(2S,3R,4S,5R,6R)-6-[[(2R,3R,4R,5S,6S)-3,5-dihydroxy-6-methyl-4-[(2S,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxyoxan-2-yl] oxymethyl]-3,4,5-trihydroxyoxan-2-yl]oxy-5,7-dihydroxy-2-(4-hydroxyphenyl)chromen-4-one no MSMS spectra (not in any network) spirosol-5-en-3-yl O-6-deoxy-alpha-L-mannopyranosyl-(1- > 2)-O-[6-deoxy-alpha-L-mannopyranosyl-(1- > 4)]- beta-D-Glucopyranoside (NCGC00179912-02) Not identified - epiminocholesta acetate derivative 3-[4,5-dihydroxy-6-(hydroxymethyl)-3-[3,4,5-trihydroxy-6-(hydroxymethyl)oxan-2-yl]oxyoxan-2-yl]oxy-5,7-dihydroxy-2-(4-hydroxyphenyl) chromen-4-one Not identified - triglycoside spirosolane derivative Not identified - triglycoside epiminocholestane derivative 2-(3,4-dihydroxyphenyl)-3-[(2S,3R,4S,5S,6R)-4,5-dihydroxy-3-[(2R,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxy-6-[[(2R,3R,4R,5R,6S)3,4,5-trihydroxy-6-methyloxan-2-yl]oxymethyl]oxan-2-yl]oxy-5,7-dihydroxychromen-4-one Not identified - triglycoside epiminocholestane derivative Not identified Not identified - triglycoside epiminocholestane derivative no MSMS spectra (not in any network) Not identified - triglycoside epiminocholestane derivative kaempferol-3-O-rutinoside | Nicotiflorin (ReSpect:PT104230)
87
741.219 505.347 912.499 486.316 611.158 926.509 882.483 757.211 868.501 347.244 884.497 926.51 866.488 595.161
6 – 198 277 181 54 32 202 1 88 2 – 222 12
Fig. 3. MN evidencing glycoside flavones and glycoalkaloids from S. pseudoquina and S. argenteum species. VIP scores from PLS-DA analysis are shown together with molecular structure.
et al., 2019; Jafari et al., 2018; Katavi, 2005; Tasca et al., 2018) and none of those targeting insights into the chemical space of the collected samples. With more than 2700 species, Solanaceae is a highly diverse plant family. At least 1330 of those species belong to Solanum, by far the richest genus of the family (Yadav et al., 2016). Solanum species are widely distributed in the Neotropics (Kaunda and Zhang, 2019). Our ongoing efforts to profile the chemical space of Solanum, which is a model for taxonomic and evolutionary studies (Soltis and Soltis, 2000), led us to assay the use of herbarium samples for compound identification as a model that can also be used for other diverse plant groups. Considering the richness of the genus, many diverse compounds are expected to be found. Among those, phenolic compounds and their glycosides, carotenoids, steroids, glycoalkaloids and tropane alkaloids are examples of bioactive and economic relevant compounds.
yielded around 2−5 mg of dry residue per sample. Such a yield points to the possibility of increasing the amount of the plant material up to 200 mg (practical limit for the 1.5 mL microtube for extraction/ homogenization) to obtain around 10 mg of dry residue to enable deeper profiling analysis using NMR and different Bioassays for prospection. It was shown that no difference in the antibacterial activity was found when comparing long storage with fresh samples (Eloff, 1999) proving promising perspectives for the use of herbarium samples in drug discovery. Sample availability in each specimen needs to be evaluated for collection in reverence to invaluable herbarium specimens, but then to collect only one part of a leave might underrepresent that sample. This method conceives the destruction of the collected sample and this must be kept in mind. Though, only a minute fraction of the extracts is used for the proposed analytical platform and the rest might be kept under safe storage conditions to compose a bank of extracts. Due to the sheer size of the acquired LC–MS/MS dataset, positiveand negative-ESI sub-experiments were separated using MSConvert and MN were directly used to visualize the chemical profile. In this section,
2. Results and discussion The extraction procedure used in this study was straightforward and 101
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
Fig. 4. The biggest MN with the epiminocholestane glycoalkaloids from S. pseudoquina and S. argenteum species. VIP scores from PLS-DA analysis are shown together with molecular structure.
we aimed for a broad visualization of the chemical profile of the whole sample set. Thus we submitted the pre-processed dataset from all the 25 samples for molecular networking (Figure S1). 185 nodes (63 in negative and 133 in positive ion mode) among the calculated networks was annotated using by GNPS’ routing through DB matching and, among those, 83 compounds were putatively identified to the Metabolomics Standards Initiative level 2 (Sumner et al., 2007) with mass errors below 7 ppm and visual validation of each fragmentogram. Phenolics compounds, aminoacids, fatty acids and other metabolites were identified (Fig. 1). As expected, many phenylpropanoid derivatives were annotated despite saponins (including glycoalkaloids) being markers for Solanum species. Probably it is because saponins are less documented compared to phenylpropanoids, which are commonly identified in other plants and cataloged in DB. The most populated MN found under ESI positive mode contained features identified as saponins: (3beta,22R,25R)-26-(beta-DGlucopyranosyloxy)-22-hydroxyfurost-5-en-3-yl 6-deoxy-alpha-L-mannopyranosyl-(1- > 2)-[beta-D-glucopyranosyl-(1- > 4)]-beta-D-glucopyranoside (C51H84O23); (3beta,22beta,25R)-26-(beta-D-glucopyranosyloxy)22-hydroxyfurost-5-en-3-yl O-6-deoxy-alpha-L-mannopyranosyl-(1- > 2)O-[6-deoxy-alpha-L-mannopyranosyl-(1- > 4)]-beta-D-Glucopyranoside (C51H84O22); spirosol-5-en-3-yl O-6-deoxy-alpha-L-mannopyranosyl-(1> 2)-O-[6-deoxy-alpha-L-mannopyranosyl-(1- > 4)]-beta-DGlucopyranoside (C45H73NO15). Spread among other MNs in both positive and negative ion modes various phenolics were detected, including different flavonoids and their mono-, di- and triglycoside derivatives, many phenylpropanoids, some benzoic acid derivatives and coumarins. Flavones, such as kaempferol, luteolin and quercetin (flavonol) were found to be the main aglycone among the glycosides with fragments at m/z 287.05, 303.05, respectively under positive ion mode. Among the identified
phenylpropanoids, caffeic acid, p-coumaric acid, ferulic acid and 3,5Dimethoxy-4-hydroxycinnamic acid and others were identified together with theirs quinic acid conjugates; the fragments at m/z 191.06 and 173.04 are characteristic under negative ion mode. 2.1. Species-based grouping through mass spectrometry-multivariate analysis Once the detection of secondary metabolites from Herbarium samples was proven to be possible, we wondered if it would be possible to discriminate the different plant species based on their chemical profile using multivariate analysis. LC-HRMS data from S. pseudoquina and S. argenteum species (n = 3) were processed and aligned in MZMine and the resulting. csv table was exported for the multivariate analysis calculations at Metaboanalyst web platform. Despite only three replicates from the two plant species were available for the study, the authors strongly recommend that a bigger number of replicates should be pursuit to ensure a more significant statistical meaning in the multivariate analysis, especially for chemotaxonomic studies. PCA and PLS-DA score plots indicated a clear trend for the distinction between S. pseudoquina and S. argenteum species even though 10 years of difference between their collection and their different collection sites (Fig. 2A and B, respectively). Components 1 and 2 explained more than 60 % of variance. By using these components, the PLS-DA analysis showed r2 = 0.999 and q2 = 0.555 (Fig. 2C), indicating that the model was reliable. To aid compound annotation, the same processed dataset submitted to MetaboAnalyst (with S. pseudoquina and S. argenteum) was then submitted to MN in GNPS. Even though it is impossible to assure the compounds identity only from MS data, because substitutions might occur at different positions in a molecule, key fragments are used to validate diagnostic features of certain 102
103
360 1370 1552 7276 3903 3824 359 5847 7185 473 2037 4687 6261 8939 6208 2991 9249
18115 18999 5462 7855 784 448 4
P. Occhioni P. Occhioni E. Pereira P. Occhioni E. Pereira P. Occhioni P. Occhioni P. Occhioni P. Occhioni E.M. Occhioni P. Occhioni P. Occhioni P. Occhioni P. Occhioni P. Occhioni L.T.H. Dombrowski P. Occhioni
G. Hatschbach G. Hatschbach & C. Keczicki P. Occhioni P. Occhioni O.M. Barth A.M. de Barros N.G. Antas
4575 6327 4435 16362 4423 9237 4574 14647 16378 21474 6769 11344 15467 20006 15390 13664 20479 40631 11835 11832 13453 17250 29167 29108 39313
Voucher Number S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S. S.
argenteum Dunal* argenteum Dunal* argenteum Dunal* asperum Rich. asperum Rich. caavurana Vell. caavurana Vell. campaniforme Roem. & Schult. campaniforme Roem. & Schult. capsicoides All. cinnamomeum Sendtn. cladotrichum Dunal cladotrichum Dunal didymium Dunal didymium Dunal pseudoquina A. St.-Hil.* pseudoquina A. St.-Hil.* pseudoquina A. St.-Hil.* sanctae-catharinae Dunal schwackeanum L.B.Sm. & Downs swartzianum Roem. & Schult. swartzianum Roem. & Schult. vaillanti Dunal variabile Mart. sisymbriifolium Lam.
Specie
12/14/1967 04/10/1968 02/28/1973 12/12/1975 01/12/1960 09/22/1990 06/11/2009
03/22/1945 07/21/1956 "—" 01/05/1975 06/24/1958 x/01/1970 04/25/1945 01/25/1974 04/28/1975 01/14/1986 x/11/1960 11/19/1971 09/16/1974 05/25/1979 09/14/1974 x/10/1969 x/11/1980
Filing Date
Rio de Janeiro, Estr. Vista Chinesa (Rio de Janeiro state) Parque Nacional da Serra dos Órgãos (Rio de Janeiro state) Rio de Janeiro, Avenida Niemeyer (Rio de Janeiro state) Linhares, Estr. para Regência margem Rio Doce (Espírito Santo state) Jacarepaguá, Três Rios (Rio de Janeiro state) Guanabara, Restinga da Tijuca (Rio de Janeiro state) Estrada do Redentor (Rio de Janeiro state) Serra dos Órgãos; Estrada Velha Teresópolis X Petrópolis (Rio de Janeiro state) BR 262. km.195 (Espírito Santo state) Ilha da Marambaia, Praia Grande (Rio de Janeiro state) Serra dos órgãos (Rio de Janeiro state) Estr. BR-101, KM. 48 alto da serra (Paraná state) Parque Nacional de Itatiaia Estr. Registro Planalto Km 7 e 8 a 1500 metros de altitude (Rio de Janeiro state) Parque Nacional de Itatiaia, Lago Azul (Rio de Janeiro state) Parque Nacional de Itatiaia, caminho do maromba (Rio de Janeiro state) Curitiba, Capão da Imbuia (Paraná state) Parque Nacional de Itatiaia, Estr. para as macieiras Km. 14 (Rio de Janeiro state) Simão Pereira (Minas Gerais state) Guaraqueçaba, Rio de Cedro (Paraná state) Guaraqueçaba (Paraná state) Serra de Friburgo, 1000 m altitude (Rio de Janeiro state) Parque Nacional de Itatiaia, Estr. Registro (Rio de Janeiro state) Casa de J. Spanner, PNI (Rio de Janeiro state) Estr. de São Paulo, Jardim São Francisco (São Paulo state) Seropédica, Pedreira ENFOL (Rio de Janeiro state)
Provenance
* S. pseudoquina and S. argenteum occurred three times in this first pilot batch. These samples were used to access speciation within this proposed approach.
Collector Number
Collector ID
Table 2 List of samples collected from the RFA Herbarium.
J.V. Mendes Resende, et al.
Phytochemistry Letters 36 (2020) 99–105
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al.
structure. Thus, candidate compounds likely to be taxonomic markers from the VIP plot list from the PLS-DA analysis (Fig. 2D) were identified (Table 1). Three main MN were found to contain all the fifteen VIP candidates and they are separated into clusters of flavonoid glycosides and many glycoalkaloids (Figs. 3 and 4). The biggest MN shows 89 nodes of glycoalkaloids classified as epiminiocholestane derivatives according literature (Fig. 4) (Soares et al., 2019). Key fragments at m/z 253.19 and 271.20 (or at m/z 255.09 and 273.10 in positive ion mode) can be highlighted as characteristic of the hydroxy-tetracyclic structure of steroids (Man et al., 2009). The PLS-DA results suggest that a more suitable prediction model could be reached using a higher number of replicates. On the other hand, power analysis using this current study as a pilot indicated that to reach a power of 0.8, close to 50 replicates would be required (Figure S2). Such a high sample size is hard to achieve for most of NP studies, but it is important to be able to understand such parameter to establish the significance of the proposed study. Herbarium exsiccates was proven to be a reliable source of samples for NP chemistry studies. Even assuming some degradation due to heat drying and long-term storage (over 50 years in some cases), many meaningful metabolites were kept stable. Little can be speculated about the stability in situ of compounds in dried plants and different investigation shall be designed accordingly. Interestingly, samples from the same species (e.g species of S. argenteum and S. pseudoquina) grouped together even with 10 years of difference between their collection and their collection site (Fig. 2A and B; Figure S3). The same is true when all the 25 samples are analyzed together, they cluster in one group showing a trend to the occurrence of similar compounds even after decades of storage in situ. The extraction of up to 50 mg with 3 ml of solvent yielded a satisfactory amount of dry residue for LCeMS/MS analysis. Depending on the complexity of the raw extract, milligram and sub-milligram scale miniaturized procedure should be considered for a bioguided fractionation approach. Still, the results shown here indicates a promising approach for taxonomists in collaboration with NP researchers to include the LCeMS/MS based chemical profile within classification studies and to consider keeping an extract collection along with the exsiccates. Higher number of samples is strongly suggested for statistical significance for species-driven classification and grouping.
formic acid in methanol in gradient elution mode (0 − 0.5 min 5 % B; 0.5 − 3 min 30 % B; 3 − 12 min, 70 % B; 12 − 15 min 95 % B; 15 − 17 min 95 % B; and 17.1 − 20 min 5 % B). The UHPLC system was coupled to a QExactive Plus high resolution and accurate mass spectrometer (Thermo Scientific) equipped with an electrospray ion source (ESI) operating in both positive and negative ionization modes. Source ionization parameters were: spray voltage 3.8 kV; capillary temperature 300 °C; S-Lens level 70, sheath and auxiliary gas 40 and 25 (arbitrary units), respectively. Samples were analyzed in the scan range of m/z 150–1500 at resolution of 70,000 (positive and negative full scan) followed by data-dependent MS2 (ddMS2 Top3 experiments) using a resolution of 17,500 and normalized collision energy (NCE) stepped 35 − 50 %. The acquired data were converted to mzXML files using the software MSConvert (ProteoWizard; available at proteowizard.sourceforge.net/ tools.shtml); polarity filter was used to separate positive and negative sub-experiments to simplify the MN interpretation. MZMine was used for feature-based processing before submitting the exported files into GNPS for MN calculations. MZMine parameters are given as SI (Table S1). The networks were plotted using the Cytoscape (http://www. cytoscape.org). The MN for both positive and negative ionization modes is accessible at the GNPS Web site with the following links: (positive) https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task= b0a773793d7e4774938445d5f9d84f67; (negative) https://gnps.ucsd. edu/ProteoSAFe/status.jsp?task= c7e62fbb6b774b0599a082aa83738b16. GNPS parameters are given as SI (Table S2). For further evaluation, the two species with tree samples each were selected as a pilot study to probe species-driven grouping capability of this method; Solanum argenteum and S. pseudoquina. MZMine was used to align and export the matrix to be analyzed at the Metaboanalyst (web server at https://www.metaboanalyst.ca/) for multivariate analysis under the protocol Statistical Analysis and Power Analysis; for multivariate analysis analysis the raw data was used without polarity separation. Metaboanalyst parameters are given as SI (Figure S4). The performance of the PLS-DA method was evaluated using the parameters r2 and q2 for fitting and predictability. Additionally, this dataset used for multivariate analysis was submitted to GNPS for a simpler MN; this is accessible in the following: https://gnps.ucsd.edu/ProteoSAFe/ status.jsp?task=6a1c707e017345cabd6f3c892fce2c73.
3. Experimental
Declaration of Competing Interest
3.1. Sample collection and sample preparation
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Two or three random loose leaves from 15 Solanum specimens (25 samples) were collected from the RFA Herbarium (Institute of BiologyFederal University of Rio de Janeiro, Brazil; Table 2). Note that the collection of this sample amount was not detrimental to the vouchers. 50 mg of each sample was homogenized with 80 % methanol (1.50 mL) and sonicated for 8 min at room temperature. The extraction mixtures were centrifuged at 10,000 g for 10 min at 4 °C and the supernatants were filtered and concentrated to dryness under vacuum. This procedure was repeated twice for maximum recovery. After resuspension in methanol (1.00 mL), aliquots of 100.00 μl were diluted in 50 % methanol (0.50 mL) for MS data acquisition.
Acknowledgments This research was supported in part by the PIBIC/CNPQ, CAPES and FAPERJ funding agencies. We would like to acknowledge the RFA Herbarium of UFRJ for kindly sharing their samples for this study and we hope this study serves as a model for many others using exsiccate collections. Appendix A. Supplementary data
3.2. Liquid chromatography-high resolution tandem mass spectrometry analysis
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.phytol.2020.01.021.
Liquid chromatography analysis was performed on a Dionex UltiMate 3000 UHPLC system (Thermo Scientific) consisting of a quaternary solvent delivery pump and a column oven compartment. Samples (5 μL) were injected using a TriPlus RSH Autosampler (Thermo Scientific) and separated on a Thermo Hypersil GOLD RP C18 column (2.1 × 100 mm, 3.0 μm particle size) at 350 μL min−1 maintained at 40 °C. The mobile phase consisted of (A) 0.1 % formic acid and (B) 0.1 %
References Afzan, A., Bréant, L., Bellstedt, D.U., Grant, J.R., Queiroz, E.F., Wolfender, J.L., Kissling, J., 2019. Can biochemical phenotype, obtained from herbarium samples, help taxonomic decisions? – a case study using Gentianaceae. Taxon. Beauxis, Y., Genta-Jouve, G., 2019. MetWork: a web server for natural products
104
Phytochemistry Letters 36 (2020) 99–105
J.V. Mendes Resende, et al. anticipation. Bioinformatics 35, 1795–1796. Beutler, J.A., 1990. Dereplication of phorbol bioactives: lyngbya majuscula and Croton cuneatus. J. Nat. Prod. 53, 867–874. Borges, R.M., Taujale, R., de Souza, J.S., de Andrade Bezerra, T., Silva, E.L.E., Herzog, R., Ponce, F.V., Wolfender, J.L., Edison, A.S., 2018. Dereplication of plant phenolics using a mass-spectrometry database independent method. Phytochem. Anal. 29, 601–612. da Silva, R.R., Wang, M., Nothias, L.F., van der Hooft, J.J.J., Caraballo-Rodriguez, A.M., Fox, E., Balunas, M.J., Klassen, J.L., Lopes, N.P., Dorrestein, P.C., 2018. Propagating annotations of molecular networks using in silico fragmentation. PLoS Comput. Biol. 14, e1006089. Eloff, J.N., 1999. It is possible to use herbarium specimens to screen for antibacterial components in some plants. J. Ethnopharmacol. 67, 355–360. Ernst, M., Kang, K.B., Caraballo-Rodríguez, A.M., Nothias, L.F., Wandy, J., Wang, M., Rogers, S., Medema, M.H., Dorrestein, P.C., van der Hooft, J.J.J., 2019. MolNetEnhancer: enhanced molecular networks by integrating metabolome mining and annotation tools. Metabolites 9 xx-xx. Jafari, T., Alanne, A.L., Issakainen, J., Pihlaja, K., Sinkkonen, J., 2018. Suitability of dried herbarium specimens for NMR metabolomics of mushrooms. A comparison of four species of the genera Kuehneromyces and Hypholoma (Strophariaceae). Fungal Biol. 122, 138–146. Katavi, P.L., 2005. Chemical Investigations of the Alkaloids from the Plants of the Family Elaeocarpaceae. School of Science/Natural Product Discovery (NPD) Griffith University, Faculty of Science p. 319. Kaunda, J.S., Zhang, Y.J., 2019. The genus Solanum: an ethnopharmacological, phytochemical and biological properties review. Nat. Prod. Bioprospect. 9, 77–137. Man, S., Gao, W., Zhang, Y., Jin, X., Ma, C., Huang, X., Li, Q., 2009. Characterization of steroidal saponins in saponin extract from Paris polyphylla by liquid chromatography tandem multi-stage mass spectrometry. Anal. Bioanal. Chem. 395, 495–505. Rogers, S., Ong, C.W., Wandy, J., Ernst, M., Ridder, L., van der Hooft, J.J.J., 2019. Deciphering complex metabolite mixtures by unsupervised and supervised substructure discovery and semi-automated annotation from MS/MS spectra. Faraday Discuss. Silva, E.L., Almeida-Lafeta, R.C., Borges, R.M., Staerk, D., 2018. Athenolide a, a new steroidal lactone from the leaves of Athenaea martiana (Solanaceae) determined by means of HPLC-HR-MS-SPE-NMR analysis. Chem. Biodivers. 15. Soares, V., Taujale, R., Garrett, R., da Silva, A.J.R., Borges, R.M., 2019. Extending compound identification for molecular network using the LipidXplorer database independent method: a proof of concept using glycoalkaloids from Solanum pseudoquina A. St.-Hil. Phytochem. Anal. 30, 132–138. Soltis, E.D., Soltis, P.S., 2000. Contributions of plant molecular systematics to studies of molecular evolution. Plant Mol. Biol. 42, 45–75.
Sumner, L.W., Amberg, A., Barrett, D., Beale, M.H., Beger, R., Daykin, C.A., Fan, T.W., Fiehn, O., Goodacre, R., Griffin, J.L., Hankemeier, T., Hardy, N., Harnly, J., Higashi, R., Kopka, J., Lane, A.N., Lindon, J.C., Marriott, P., Nicholls, A.W., Reily, M.D., Thaden, J.J., Viant, M.R., 2007. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) metabolomics standards initiative (MSI). Metabolomics 3, 211–221. Tasca, J.A., Smith, C.R., Burzynski, E.A., Sundberg, B.N., Lagalante, A.F., Livshultz, T., Minbiole, K.P.C., 2018. HPLC-MS detection of pyrrolizidine alkaloids and their Noxides in herbarium specimens dating back to the 1850s. Appl. Plant Sci. 6, e1143. Wandy, J., Zhu, Y., van der Hooft, J.J.J., Daly, R., Barrett, M.P., Rogers, S., 2017. Ms2lda.org: web-based topic modelling for substructure discovery in mass spectrometry. Bioinformatics. Wang, M., Carver, J.J., Phelan, V.V., Sanchez, L.M., Garg, N., Peng, Y., Nguyen, D.D., Watrous, J., Kapono, C.A., Luzzatto-Knaan, T., Porto, C., Bouslimani, A., Melnik, A.V., Meehan, M.J., Liu, W.T., Crusemann, M., Boudreau, P.D., Esquenazi, E., Sandoval-Calderon, M., Kersten, R.D., Pace, L.A., Quinn, R.A., Duncan, K.R., Hsu, C.C., Floros, D.J., Gavilan, R.G., Kleigrewe, K., Northen, T., Dutton, R.J., Parrot, D., Carlson, E.E., Aigle, B., Michelsen, C.F., Jelsbak, L., Sohlenkamp, C., Pevzner, P., Edlund, A., McLean, J., Piel, J., Murphy, B.T., Gerwick, L., Liaw, C.C., Yang, Y.L., Humpf, H.U., Maansson, M., Keyzers, R.A., Sims, A.C., Johnson, A.R., Sidebottom, A.M., Sedio, B.E., Klitgaard, A., Larson, C.B., P, C.A.B, Torres-Mendoza, D., Gonzalez, D.J., Silva, D.B., Marques, L.M., Demarque, D.P., Pociute, E., O’Neill, E.C., Briand, E., Helfrich, E.J.N., Granatosky, E.A., Glukhov, E., Ryffel, F., Houson, H., Mohimani, H., Kharbush, J.J., Zeng, Y., Vorholt, J.A., Kurita, K.L., Charusanti, P., McPhail, K.L., Nielsen, K.F., Vuong, L., Elfeki, M., Traxler, M.F., Engene, N., Koyama, N., Vining, O.B., Baric, R., Silva, R.R., Mascuch, S.J., Tomasi, S., Jenkins, S., Macherla, V., Hoffman, T., Agarwal, V., Williams, P.G., Dai, J., Neupane, R., Gurr, J., Rodriguez, A.M.C., Lamsa, A., Zhang, C., Dorrestein, K., Duggan, B.M., Almaliti, J., Allard, P.M., Phapale, P., Nothias, L.F., Alexandrov, T., Litaudon, M., Wolfender, J.L., Kyle, J.E., Metz, T.O., Peryea, T., Nguyen, D.T., VanLeer, D., Shinn, P., Jadhav, A., Muller, R., Waters, K.M., Shi, W., Liu, X., Zhang, L., Knight, R., Jensen, P.R., Palsson, B.O., Pogliano, K., Linington, R.G., Gutierrez, M., Lopes, N.P., Gerwick, W.H., Moore, B.S., Dorrestein, P.C., Bandeira, N., 2016. Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nat. Biotechnol. 34, 828–837. Yadav, R., Rathi, M., Pednekar, A., Rewachandani, Y., 2016. A detailed review on Solanaceae family. Eur. J. Pharm. Med. Res. 3, 369–378. Yang, J.Y., Sanchez, L.M., Rath, C.M., Liu, X., Boudreau, P.D., Bruns, N., Glukhov, E., Wodtke, A., de Felicio, R., Fenner, A., Wong, W.R., Linington, R.G., Zhang, L., Debonsi, H.M., Gerwick, W.H., Dorrestein, P.C., 2013. Molecular networking as a dereplication strategy. J. Nat. Prod. 76, 1686–1699.
105