Despite a very intensive development of herbal science, there are hardly any publications concerning weed amino acids (AA) profile. Thus, our objective was to apply a method to detect free amino acids (FAA) using liquid chromatography-electrospray ionization coupled to a triple quadrupole mass spectrometer (LC-ESI-MS/MS) and to establish the free amino acid profile of a very common weed - heartsease (Viola tricolor) found in different types of habitat in the Lower Silesia region in Poland. This plant species was selected based on its frequency of occurrence, worldwide habitat, edibility, implications with alternative and modern medicine. Heartsease (Viola tricolor L.) from Violaceae family is a traditional medicinal plant and has been documented in the Pharmacopoeia of Europe [1,2,3]. Moreover, farmers recognize this plant as a weed as it can spread rapidly, the seeds have excellent adaptation to germination which can lead to a crop yield loss when the weeds compete with crops , Over the past few years a number of herbicides have been discovered which inhibit amino acids metabolism; therefore, studies relating to weed amino acids profile are of paramount importance . Rop et al.  observed that flowers of Viola x wittrockiana possess the highest level of mineral elements among 12 edible flowers that were investigated. Viola tricolor could turn out to be a suitable and cheap source of amino acids for consumption, the cosmetic industry, pharmaceutical applications. or medicine. Data exists that show its favorable effect on human health - to cure skin diseases, rheumatic pains, eczema, asthma and respiratory problems. Amino acids participate in the biosynthesis of polyphenols and alkaloids which contribute to antioxidant properties in plants [7,8,9]. It is also worth mentioning, that there are promising projects underway that study cyclotide isolation from Viola tricolor and other species from the Violaceae family which contain anti-tumur properties [10,11]. Tang et al.  isolated 14 cyclotides from Viola tricolor, including seven novel cyclotides, using tandem mass spectrometry and NMR spectroscopy, some of which show cytotoxic activities against five cancer cell lines.
Hellinger et al. [13,14] shed light on the unexplored variety of plant-derived cyclotides of the Violaceae family and discovered that an aqueous Viola tricolor extract contained bioactive cyclotides, with immunosuppressive activity.
Metabolite profiling aimed at analyzing a small number of known metabolites belonging to the same compound classes is one of the three general approaches to analyze small molecules, among metabolic fingerprinting and metabolomics. Tandem mass spectrometry (MS/ MS) allows interfering signals to be filtered out after molecular ions are broken into fragments between steps of mass analysis. Mass spectrometry coupled with ultra high-performance liquid chromatography (UHPLC) is a powerful tool enabling accurate and reliable metabolite analysis in a short time . Also, the sample preparation is very important, and the choice of the drying method could be a goal itself when analyzing plants used for herbal medicine. Zhu et al.  used UHPLC-TQ-MS coupled with multivariate statistical analysis to characterise amino acids, nucleosides and nucleobases in Angelicae Sinensis Radix was obtained under different drying methods.
Sami et al.  found 17 amino acids in the okra plant that were collected from four different regions in Egypt, 11 of which were essential. Authors found that the major amino acid was aspartic acid, which dry weight concentrations were 2.91–4.92 g/100 g. However, the authors did not provide information about the locations; therefore, no conclusions could be drawn between the relation between amino acid profile and habitat conditions. Sun et al.  selected 13 wild edible mushrooms species found in China and analyzed their free amino acid composition using reversed phase liquid chromatography The total free amino acid (TAA) content ranged from 1462.6 mg/100 g to 13,106.2 mg/100 g on a dry weight basis. Furthermore, the authors did a principal component analysis and cluster analysis to show that essential amino acids composition and content might be an important parameter in separating the mushroom species.
This work identifies (both qualify and quantify) the composition of 19 FAA in a Viola tricolor matrice for the first time, due to important biological functions and a possible value for industry; it also shows free amino acid variability related to different environmental conditions using the discriminant and cluster analysis method.
2 Materials and Methods
Samples of heartsease (Viola tricolor) used in this study were collected from a field in Zagródki (A) N50o 59’ 4.9943” E17o6’ 47.3708” and Wrocław (B) N51o4’ 40.2254” E17o2’ 35.0415 at the end of June 2015. The aboveground biomass was obtained at stage five (flower development), with both stems and leaves. Three specimens per habitat were collected, immediately transported to the laboratory; the samples were frozen and ground using pestle and liquid nitrogen.
2.2 Determination of Free Amino Acids Composition
2.2.1 Sample preparation
The samples (0.5 g) were extracted using LC-MS grade water, followed by a 15 min sonification in an ultrasonic bath. Homogenates were centrifuged at 11000 rpm for 15 min at 4oC to obtain supernatants. EZ:faast(TM) Free Amino Acid kit (Phenomenex, Torrance, CA, USA) was used for AA analysis. The procedure was performed according to the optimized and validated method presented by Dziągwa-Becker et al. . In short, the procedure consisted of solid phase extraction, where the plant extract was passed through the sorbent tip that bound the amino acids, derivatization using propyl chloroformate and liquid-liquid extraction .
2.2.2 Determination of free amino acids by high performance liquid chromatography with electrospray tandem mass spectrometry (LC-MS/MS)
The analysis was conducted using a high-performance liquid chromatograph Shimadzu 8030 (Shimadzu, Kyoto, Japan) with a binary solvent manager, autosampler and column oven. An EZ:faast(TM)4u AAA-MS column, 3 μm, 250 × 2.0 mm (Phenomenex, Torrance, CA, USA) at a flow rate of 0.25 mL min-1 was used for the chromatographic separation. The column temperature was maintained at 35°C. The mobile phase consisted of water/methanol (A/B) gradient both 10 mM ammonium formate where the methanol percentage was changed linearly as follows: 0 min, 68%; 13 min, 83%; 13.01 min, 68%; 18 min, 68%. The abovementioned chromatographic conditions were applied to all analyzed samples. All AAs were analyzed in a positive ionization mode, showing an abundant [M + H]+ ion for each derivatised amino acid. The sample volume injected into the HPLC system was 1 μL. The tandem mass spectrometer LCMS-8030 (Shimadzu, Kyoto, Japan) with ultra fast polarity switching and ultra fast Multiple Reaction Monitoring (MRM) transitions was used for analysis. Nitrogen was drying as well as nebulising gas, obtained from pressurized air in a N2 LC-MS pump, working at a flow rate 15 L min-1 and 3 L min-1, respectively. The desolvation line temperature was 250°C and the heat block temperature was 400°C. The collision-induced dissociation gas (CID) was argon 99.999% (Linde, Wrocław, Poland) at a pressure of 230 kPa. A dwell time of 10 ms was selected. LabSolution Ver. 5.6 (Shimadzu, Kyoto, Japan) software was used to process the quantitative data. LC-MS grade solvents were used, obtained from Fluka Analytical (St. Louis, MO, USA) .
2.3 Soil properties determination
The physico-chemical proprieties of the soils were determined as follows: soil pH was tested potentiometrically in 1 M KCl, potassium and phosphorus content was tested according to the Egner-Riehm method, the granulomertic composition was conducted using the sieve method by Casagrande in Prószyński modification. Granulometric group was specified according to USDA. The organic carbon content was determined according to Tiurin’s method, soil organic matter (SOM) was performed on the basis of Corg content [21,22] (Table 2).
2.4 Meteorological data
Meteorological data came from two IUNG weather stations. Temperature measurements were made at a standard shelter height (2m) in accordance with established on-site meteorological guidelines (Table 3).
2.5 Data Analysis
Assessing the variability of the free amino acid content in Viola tricolor populations in Wrocław and Zagródki, multivariate analysis of variance (MANOVA) and discriminant analysis method was performed using the statistical data analysis program STATISTICA 6 (StatSoft Polska, Poland), laid down by Morrison , Caliński and Chudzik , Krzysko . The analysis enabled the assessment of the amino acid content in a dimension created by two variables - different habitats. Cluster analysis was conducted using Ward’s method to compare the achieved results.
Ethical approval: The conducted research is not related to either human or animals use.
3.1 Viola tricolor amino acids profile
All 19 free amino acids were found in Viola tricolor samples collected from two locations in Poland - Zagródki and Wrocław. Their concentrations are exhibited in Table 1. The total free amino acids (TAA) content in the analyzed samples ranged from 9938.0 to 11393.8 mg/kg of fresh weight. To the best of our knowledge, this is the first work that shows the presence of 19 FAA belonging to the group of 23 proteinogenic amino acids. Plants can synthesize all proteinogenic amino acids. Among them. 9 amino acids are known to be essential, meaning humans can them from their diet. All essential AA are present in Viola tricolor and their content amounts of 2703.9 and 1202.3 mg/kg in Zagródki and Wrocław respectively. The ratio between EAA and TAA was 0.272 and 0.106 for Zagródki and Wrocław, respectively. As shown in Table 1, Figure 1 and 2, it was possible to determine all t19 free amino acids: alanine, arginine, aspartic acid, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, valine, asparagine, andglutamine.
3.2 Habitat conditions
Both soils originated from farmland. The soil from Wrocław was slightly acidic with light texture and classified as sandy loam (USDA Soil Texture Calculator https://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167). This soil that was very high in phosphorus, potassium and Soil Organic Matter (SOM) resulted from intensive cultivation and fertilization. Also, the pH level was controlled by intensive limning. Sandy loam soils are susceptible to over-drying.
The soil from Zagródki had a high content of clay fraction and was classified as sandy clay loam . This type of soil is not intensively cultivated and has lower phosphorus and potassium content than Wrocław, but a slightly higher SOM content and similar pH level. Soils with sandy clay loam texture are susceptible to poor drainage and gley properties (Table 2).
The second quarter of 2015 was warm, with high precipitation levels. Especially in June, there was heavy rain (Table 3).
4.1 Amino acids profile
Ala is the most abundant AA found in both habitats. Its ratio to the TAA content was 0.535 and 0.700. The divergence of FAA content in Viola tricolor from two locations can be a result of many conditions interfering, co-affecting and interchanging the AA profile. Among them are the use of pesticides, weather conditions, soil properties, microorganisms, harvesting time and growth conditions. No one single factor can change the amino acids profile, but more importantly, the group of interrelating factors can lead to enormous alterations. Furthermore, Viola tricolor has a tendency to crossbreed with other species from the Violaceae family and form hybrids . Therefore, the different AA profile from two habitats can be a proof of genetic variability.
4.2 Statistical Analysis of Variability
The variability of the investigated amino acids with respect to different habitat conditions was statistically assessed using the methods of discriminant and cluster analysis. Statistical analysis showed significant differences among the free amino acid content in the investigated habitats (Table 1). Only Met showed similar concentrations in both Viola tricolor populations. The majority of amino acids showed a higher concentration in Zagródki. Only the content of Ala, Gln, Glu and Asp was higher in Wrocław habitat in comparison to Zagródki. A multidimensional analysis of variance (MANOVA) showed a significant differentiation of free amino acids in Viola tricolor in relation to different habitats (Table 4).
Wilk’s lambda distribution for the total discriminant, calculated as the ratio of determinant matrix variance and intra-group covariance to determinant matrix variance and total covariance shows that the hypothesis of centroid equality of the two habitats should be rejected at p<0,0001 significance level. Partial Wilk’s lambda and F test results indicate that the Zagródki habitat affected the amino acids variability more than the Wrocław habitat; however, in Wrocław the differences between AA were also significant. Table 5 shows squared Mahalanobis distances between two amino acids formed by two habitats. The Mahalanobis distance is similar to the standard Euclidean distance, however it additionally shows the correlation between the two variables. The larger the distances shown in the table, the farther the amino acids are located and also the bigger discriminant power has the proposed model in order to differentiate the examined amino acids.
Mean concentration values of the investigated amino acids from two habitats varied significantly in most of the cases. Insignificant Mahalanobis distances were found between Asn and Phe, Phe and Arg. No significant differences were calculated between Arg and Asn, Phe, His and Tyr. Considerable similarities were found between the concentration of His and Tyr in the investigated locations. Cluster analysis was performed to compare the concentration of 19 AA in the dimension formed by two habitats. The closer the location of AA, the higher similarity between groups of the investigated plants. While analyzing the aggregation of the control subject, four clusters could be distinguished (Figure 3).
The first cluster was formed by 10 amino acids - Gly, Thr, Ser, Lys, Asn, Phe, Met, Trp, Arg and His. They were characterized by a significant Euclidean distance in comparison to the second cluster formed by four amino acids - Pro, Ile, Leu and Glu. The two clusters were in a considerable distance from the third cluster, formed by Val, Gln, Asp. Interestingly, Ala formed the fourth one-element cluster. Ala was present in a far bigger concentration compared to the other analyzed compounds. The tree diagram shows the Euclidean distances between the AA in two dimensional spaces. However, it does not include the correlation between the two habitats (for example similar temperature and precipitation range during Viola tricolor’s vegetation). Therefore, some of the Mahalanobis distances do not illustrate the distance of each AA in the Euclidean space. However, there is some tendency that can be observed. In Table 5 as well as in the tree diagram Ala is far distant from the other AA. The insignificant differences in the content of Asp and Phe as well as in Tyr and His are confirmed by Mahalanobis distances and by the tree diagram.
Plants grown in both habitats contained all of the 19 analyzed amino acids, some of which had significant concentrations, like Ala, Val, Asp and Gln. It cannot be unequivocally determined whether the results of the study on the influence of habitat conditions on the amino acids composition of Viola tricolor are permanent. Therefore, it is necessary to continue the research.
Discriminant analysis demonstrated that the investigated habitats significantly differentiated the free amino acids content of Viola tricolor. Only Met showed similar concentration in both Viola tricolor populations.
Zagródki habitat is far more diverse with regard to free amino acid content in comparison to Wrocław.
No significant difference with regard to free amino acid content in Viola tricolor was found between Asn and Phe, Tyr and His and also between Tyr and Arg.
Ala is by far the most abundant amino acid in Viola tricolor.
Viola tricolor is prone to genetic variability and FAA content can be one of the evidence.
National Science Centre financially supported this project holding the decision number: UMO-2013/09/N/NZ9/01960.
Rimkiene S., Ragazinskiene O., Savickiene N., The cumulation of Wild pansy (Viola tricolor L.) accessions: the possibility of species preservation and usage in medicine, Medicina, 2003, 39, 411-416. PubMedGoogle Scholar
European Pharmacopoeia, European Directorate for the Quality of Medicines Healthcare, 7th ed., Volumes 1 and 2, Supplements 1 and 2. Council of Europe, Strasbourg, 2011. Google Scholar
Gonçalves A.F.K., Friedrich R.B., Boligon A.A., Piana M., Beck R.C.R., Athayde M.L., Anti-oxidant capacity, total phenolic contents and HPLC determination of rutin in Viola tricolor (L) flowers, Free Rad. Antiox., 2012, 2, 32-37. CrossrefGoogle Scholar
Konovalova I.S., Babich N.A., Reproductive features of weed plants of forest nurseries, J. Int. Sci. Publ.: Ecol. Saf., 2014, https://www.scientific-publications.net/en/article/1000131/.
Rop O., Mlcek J., Jurikova T., Neugebauerova J., Vabkova J., Edible Flowers—A New Promising Source of Mineral Elements in Human Nutrition, Molecules, 2012, 17, 6672-6683. CrossrefWeb of ScienceGoogle Scholar
Tiong S.H., Looi C.Y., Hazani H., Arya A., Paydar M., Wong W.F., et al., Antidiabetic and Antioxidant Properties of Alkaloids from Catharanthus roseus (L.) G. Don., Molecules, 2013, 18, 9770-9784. CrossrefWeb of ScienceGoogle Scholar
Bi W., He C., Ma Y., Shen J., Zhang L.J., Peng Y., et.al., Investigation of free amino acid, total phenolics, antioxidant activity and purine alkaloids to assess the health properties of non-Camellia tea, Acta Pharm. Sinica B, 2016, 6, 171-181. Web of ScienceGoogle Scholar
Tang J., Wang C.K., Pan X., Yan H., Zeng G., Xu W., et al., Isolation and characterization of cytotoxic cyclotides from Viola tricolor, Peptides, 2010, 31, 1434–1440. Web of ScienceCrossrefPubMedGoogle Scholar
Hellinger R., Koehbach J., Fedchuk H., Sauer B., Huber R., Gruber C.W., et al., Immunosuppressive activity of an aqueous Viola tricolor herbal extract, J. Ethnopharmacol., 2014, 151, 299–306. CrossrefWeb of SciencePubMedGoogle Scholar
Hellinger R., Koechbach J., Soltis D.E., Carpenter E.J., Wong G.K., Gruber C.W., Peptidomics of Circular Cysteine-Rich Plant Peptides: Analysis of the Diversity of Cyclotides from Viola tricolor by Transcriptome and Proteome Mining, J. Proteome. Res., 2015, 14, 4851–4862. CrossrefPubMedWeb of ScienceGoogle Scholar
Zhu S., Guo S., Duan J., Qian D., Yan H., Sha X., et al., UHPLC-TQ-MS Coupled with Multivariate Statistical Analysis to Characterize Nucleosides, Nucleobases and Amino Acids in Angelicae Sinensis Radix Obtained by Different Drying Methods, Molecules, 2017, http://www.mdpi.com/1420-3049/22/6/918. PubMedWeb of Science
Sami R., Lianzhou J., Yang L., Ma Y., Jing J., Evaluation of Fatty Acid and Amino Acid Compositions in Okra (Abelmoschus esculentus) Grown in Different Geographical Locations, Biomed. Res. Int., 2013, https://www.hindawi.com/journals/bmri/2013/574283/. PubMedWeb of Science
Sun L., Liu Q., Bao C., Fan J., Comparison of Free Total Amino Acid Compositions and Their Functional Classifications in 13 Wild Edible Mushrooms, Molecules, 2017, http://www.mdpi.com/1420-3049/22/3/350. PubMedWeb of Science
Dziągwa-Becker M.M., Marin Ramos J.M., Topolski J.K., Oleszek W.A., Determination of free amino acids in plants by liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS), Anal. Methods., 2015, 7, 7574-7581. CrossrefWeb of ScienceGoogle Scholar
EZ:faast(TM), User’s Manual, Phenomenex, Torrance, 2013. Google Scholar
Drozd J., Licznar M., Licznar S.E., Weber J., Gleboznawstwo z elementami mineralogii i petrografii, 2nd ed., Wydawnictwo Akademii Rolniczej we Wrocławiu, Wrocław, 1998. Google Scholar
Kabała C., Karczewska A., Metodyka analiz laboratoryjnych gleb i roślin, 4th ed., 2008. Google Scholar
Morrison D.F., Multivariate Statistical Methods, Mc Graw – Hill, 2nd ed., New York, 1976. Google Scholar
Caliński T., Chudzik H., Grupowanie populacji na podstawie wyników wielozmiennej analizy wariancji, Algorytmy biometryczne i statystyczne, 1980, 9, 139-167, (in Polish). Google Scholar
Krzyśko M., Analiza Dyskryminacyjna, Wydawnictwa Naukowo-Techniczne, Warszawa, 1990. Google Scholar
Ditzler C., Scheffe K., Monger H.C. (Eds.), Soil Survey Manual, USDA Handbook 18, Soil Science Division Staff, Government Printing Office, Washington DC, 2017. Google Scholar
Słomka A., Sutkowska A., Szczepaniak M., Malec P., Mitka J., Kuta E., Increased genetic diversity of Viola tricolor L. (Violaceae) in metal-polluted environments, Chemosphere, 2011, 83, 435–442. Web of ScienceCrossrefPubMedGoogle Scholar
About the article
Published Online: 2018-08-20
Conflict of interest: Authors state no conflict of interest.
Citation Information: Open Chemistry, Volume 16, Issue 1, Pages 833–841, ISSN (Online) 2391-5420, DOI: https://doi.org/10.1515/chem-2018-0098.
© 2018 Magdalena Dziągwa-Becker et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. BY-NC-ND 4.0