Evaluation of phytochemical composition of fresh and dried raw material of introduced Chamerion angustifolium L. using chromatographic, spectrophotometric and chemometric techniques

Phytochemistry xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Phytochemistry

journal homepage: www.elsevier .com/locate /phytochem

Evaluation of phytochemical composition of fresh and dried rawmaterial of introduced Chamerion angustifolium L. usingchromatographic, spectrophotometric and chemometric techniques

http://dx.doi.org/10.1016/j.phytochem.2015.02.0050031-9422/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: [email protected] (A. Maruška).

Please cite this article in press as: Kaškoniene, V., et al. Evaluation of phytochemical composition of fresh and dried raw material of introduced Chaangustifolium L. using chromatographic, spectrophotometric and chemometric techniques. Phytochemistry (2015), http://dx.doi.org/1j.phytochem.2015.02.005

Vilma Kaškoniene a, Mantas Stankevicius a, Tomas Drevinskas a, Ieva Akuneca a, Paulius Kaškonas b,Kristina Bimbiraite-Surviliene a, Audrius Maruška a,⇑, Ona Ragazinskiene c, Olga Kornyšova a,Vitalis Briedis d, Rasa Ugenskiene e

a Faculty of Natural Sciences, Vytautas Magnus University, Kaunas, Lithuaniab Institute of Metrology, Kaunas University of Technology, Lithuaniac Kaunas Botanical Garden of Vytautas Magnus University, Sector of Medicinal Plants, Lithuaniad Faculty of Pharmacy, Lithuanian University of Health Sciences, Kaunas, Lithuaniae Institute of Oncology, Lithuanian University of Health Sciences, Lithuania

a r t i c l e i n f o

Article history:Received 9 December 2014Received in revised form 30 January 2015Available online xxxx

Keywords:IntroductionChamerion angustifolium L.Volatile compoundsPhenolic compoundsFlavonoidsChemometric analysis

a b s t r a c t

Due to the wide spectrum of biological activities, Chamerion angustifolium L. as medicinal plant is used forthe production of food supplements. However, it should be kept in mind that quality (biological activity)of the herb depends on its geographic origin, the way of raw material preparation or extraction andchemotype. The purpose of this study was to evaluate and compare the compositions of volatile, non-volatile compounds and antioxidant activities of C. angustifolium grown in Kaunas Botanical Garden afterthe introduction from different locations in Lithuania. The compositions of fresh and air-dried sampleswere compared.

The profile of volatile compounds was analyzed using headspace solid phase microextraction coupledwith GC/MS. trans-2-Hexenal (16.0–55.9% of all volatiles) and trans-anethole (2.6–46.2%) were deter-mined only in the dried samples, while cis-3-hexenol (17.5–68.6%) only in fresh samples. Caryophyllenes(a- and b-) were found in all analyzed samples, contributing together from 2.4% to 52.3% of all volatilesaccording to the origin and preparation (fresh or dried) of a sample. Total amount of phenolic compounds,total content of flavonoids and radical scavenging activity (using 2,2-diphenyl-1-picrylhydrazyl (DPPH))were determined using spectrophotometric assays. The variation of total phenolic compounds contentwas dependent on the sample origin, moreover, drying reduced amount of phenolics 1.5–3.5 times.The DPPH free radical scavenging activity was in the range of 238.6–557.1 mg/g (expressed in rutinequivalents) in the fresh samples and drastically reduced to 119.9–124.8 mg/g after drying. The qualita-tive analysis of phenolic compounds in the aqueous methanolic extracts of C. angustifolium was per-formed by means of HPLC with UV detection. Oenothein B and rutin were predominant in thesamples; also caffeic and chlorogenic acids, and quercetin were determined. Chemometric methods,namely principal component analysis, hierarchical cluster analysis and K-means clustering analysis, wereapplied for evaluation of the results. Chemometric analysis showed existence of different chemotypes ofC. angustifolium L. and their relation to the geographic origin.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

As phytotherapy and herbal medicine are very popular nowa-days, it is necessary to fulfill a demand of qualitative plant materialfor the consumers and manufacturers. Cultivation of medicinal

plants can ensure the production of raw material and productsbased on the most popular plants. However it is not revealedhow the cultivation (introduction) does affect the phytochemicalcomposition, which is the main factor affecting therapeutic (biolo-gical) activity of the herb. It is already known that biological activ-ity and accumulation of bioactive compounds in the plant dependon its phenotype and more precisely on geographic origin, climaticconditions (ecotype) or genotype. Phenotype differences are

merion0.1016/

http://dx.doi.org/10.1016/j.phytochem.2015.02.005

mailto:[email protected]


http://www.sciencedirect.com/science/journal/00319422

http://www.elsevier.com/locate/phytochem



2 V. Kaškoniene et al. / Phytochemistry xxx (2015) xxx–xxx

reflected in metabolism and chemotype of the plants (Aprotosoaieet al., 2010; Rota et al., 2008; Kaškoniene et al., 2013).

Chamerion angustifolium L. Holub (also called Epilobium angusti-folium L.) is a perennial medicinal plant of the Onagraceae family. Itis native plant of many countries of Northern Hemisphere, includ-ing Lithuania. In the beginning of this century, the anticancer prop-erties of C. angustifolium were discovered (Vitalone et al., 2001,2003; Kiss et al., 2006a,b; Schepetkin et al., 2009). The mainresponsibility is attributed to the oenothein B, which belongs tothe group of oligomeric ellagitannins (Moilanen et al., 2013). Alsoantioxidant (Hevesi Tóth et al., 2009; Kiss et al., 2011), antibacteri-al (Battinelli et al., 2001) anti-inflammatory (Kiss et al., 2011), andanti-aging (Ruszova et al., 2013) properties of Epilobium specieswere established.

The aim of this study was to evaluate the phytochemical com-position (volatile and non-volatile compounds) and DPPH freeradical scavenging activity of C. angustifolium L. introduced to Kau-nas Botanical Garden from the different locations of Lithuania. Theidentic growing conditions allow to eliminate the differences insunlight exposure, amount of precipitations, effect of the tem-perature and humidity, and soil quality in different locations tothe phytochemical composition of C. angustifolium, in order to eval-uate the presence or absence of different chemotypes of C. angusti-folium growing in Lithuania.

To our knowledge, this is the first comparative study of the phy-tochemical composition of C. angustifolium plants introduced fromthe different locations. Such study is needed to reveal how the cul-tivation (introduction) of the herbs affects the volatile compoundscomposition, qualitative and quantitative composition of phenoliccompounds and radical scavenging activity, depending on thepreparation of raw material (fresh or dried). It is needed for deter-mination of the impact of introduction on the phytochemical com-position of C. angustifolium and presence of different chemotypes ofthis medicinal plant.

2. Results and discussion

Eleven samples of C. angustifolium were collected from differentlocations in Lithuania and introduced to the experimental field inKaunas Botanical Garden of Vytautas Magnus University. One sam-ple already growing in the medicinal plant collection at KaunasBotanical Garden was also analyzed. All samples were cultivatedin controlled experimental field at the same climatic and soil con-ditions. Spectrophotometric assays of total phenolic compounds,

Fig. 1. Total phenolic content (TPC), total flavonoids content (TFC) and radic

Please cite this article in press as: Kaškoniene, V., et al. Evaluation of phytochemangustifolium L. using chromatographic, spectrophotometric and chemoj.phytochem.2015.02.005

total flavonoids and DPPH free radical scavenging as well as HPLCand GC/MS analysis were performed for the fresh and dried sam-ples of C. angustifolium. Data obtained were classified using chemo-metric methods.

2.1. Spectrophotometric assays

The variations in total phenolic content, total flavonoids contentand radical scavenging activity (RSA) were observed between thetested samples and between fresh and dried material (Fig. 1). Thetotal phenolic content in fresh samples varied from 180.2 to324.3 mg/g (expressed as rutin equivalents); the total phenoliccompounds content reduced significantly in all samples after dry-ing, i.e., to 70.6–126.8 mg/g. The drying had the smallest impact onthe variation of total flavonoids content, which was in the range of15.0–38.8 and 14.3–41.0 mg/g (expressed as rutin equivalents) infresh and dried samples, respectively. The total flavonoids contentremain the same in two samples (KBG and MOL), increased by 1.2–2.2 times in four samples (ALE, VAR, TRA and ALY), and reduced by1.2–1.6 times in the rest of the samples (Fig. 1).

Drying had the highest impact on the RSA of tested samples:RSA reduced 2.0–4.5 times, i.e., RSA was 238.6–557.1 mg/g (ex-pressed as rutin equivalents) in the fresh samples, while this valuevaried from 119.9 to 124.8 mg/g in the dried samples. The correla-tion coefficient between radical scavenging activity and total phe-nolic content was 0.55 and 0.68 in fresh and dried samples,respectively; while total content of flavonoids did not correlatewith RSA of the tested samples (correlation coefficient in both cas-es was close to zero).

It is evident, that air drying had negative impact to the RSA of C.angustifolium. Data coincide with the findings of Chan et al. (2012),who compared the total phenolic compounds content, total flavo-noids content and free DPPH radical scavenging of fresh and driedLabiatae herbs. Total phenolic compounds content, total flavonoidscontent and radical scavenging activity reduced by 18–89%, 19–85%, and 23–95%, respectively, almost in all tested samples, exceptoregano and marjoram, where increase of some values wasobserved. In our case, the loss of total phenolic compounds was34.7–71.4%, the loss of 11.6–39.0% of total flavonoids was observedin 7 out of 12 samples, and RSA reduced from 49.0% to 77.7% in allthe samples. Chan et al. (2012) compared three different dryingtechniques, i.e., oven drying at 50 and 80 �C, and microwave pre-treatment before oven drying at 50 �C, while in our study the airdrying was used, thus the drying temperature was ca. 25 �C. Some

al scavenging activity (RSA) in the tested samples of C. angustifolium L.

ical composition of fresh and dried raw material of introduced Chamerionmetric techniques. Phytochemistry (2015), http://dx.doi.org/10.1016/



https://www.researchgate.net/publication/257321239_Characterization_of_bioactive_plant_ellagitannins_by_chromatographic_spectroscopic_and_mass_spectrometric_methods?el=1_x_8&enrichId=rgreq-5e755e22-4a14-43ab-808d-31b3757fc095&enrichSource=Y292ZXJQYWdlOzI3Mjc5MzI2ODtBUzoyMDEwMTQ3MzgxOTg1MzFAMTQyNDkzNzA1Njk1Ng==

https://www.researchgate.net/publication/257747393_Antioxidant_and_antibacterial_properties_of_some_fresh_and_dried_Labiatae_herbs?el=1_x_8&enrichId=rgreq-5e755e22-4a14-43ab-808d-31b3757fc095&enrichSource=Y292ZXJQYWdlOzI3Mjc5MzI2ODtBUzoyMDEwMTQ3MzgxOTg1MzFAMTQyNDkzNzA1Njk1Ng==

https://www.researchgate.net/publication/257747393_Antioxidant_and_antibacterial_properties_of_some_fresh_and_dried_Labiatae_herbs?el=1_x_8&enrichId=rgreq-5e755e22-4a14-43ab-808d-31b3757fc095&enrichSource=Y292ZXJQYWdlOzI3Mjc5MzI2ODtBUzoyMDEwMTQ3MzgxOTg1MzFAMTQyNDkzNzA1Njk1Ng==

Fig. 2. The HPLC/UV profile of some fresh and dried C. angustifolium L. samples (1 –gallic acid; 2 – oenothein B; 3 – rutin; 4 – quercetin). Note: extracts from fresh anddried herb were prepared using the same amount of material, i.e., 500 mg.

V. Kaškoniene et al. / Phytochemistry xxx (2015) xxx–xxx 3

studies, however, show an increase of the total phenolic contentand antioxidant activity in sun-dried or dried at 40 �C herbs com-paring with fresh samples (Yi and Wetzstein, 2011).

A loss of phenolic content during the drying process may beexplained by their degradation due to the enzymatic and non-enzymatic reactions (López-Nicolás and García-Carmona, 2009;Al-Rawahi et al., 2013). Mrad et al. (2012) suggest that a decreasein total phenolic content during drying may also be attributed tothe binding of polyphenols with other compounds (like proteins)or to transformations in the chemical structure of polyphenolswhich cannot be extracted or determined using appropriatemethods.

Capecka et al. (2005) found a great decrease of carotenoids andascorbic acid in the herbs from Lamiaceae species: the loss of car-otenoids was up to 65%, while the loss of ascorbic acid was morethan 90%. As it is known that these compounds also possess radicalscavenging activity (Stahl and Sies, 2003, 2005; Zhang andHamauzu, 2004), it is possible that these compounds can have sig-nificant impact on the increased RSA of fresh C. angustifolium sam-ples. However, the reduction of RSA can be explained not only bythe obvious quantitative losses of compounds with antioxidantactivity, but also the synergy occurring between antioxidants andanother plant constituents as suggested by some authors (Miloset al., 2000; Zheng and Wang, 2001).

2.2. HPLC/UV analysis

HPLC/UV analysis revealed that the changes of phenolic com-pounds in C. angustifolium extracts were mostly quantitative ratherthan qualitative after drying (Fig. 2). Nevertheless, the profile of allextracts was similar. Qualitative analysis showed that oenothein Btogether with rutin were predominant in all samples tested.Remarkably lower amounts of gallic and chlorogenic acids andquercetin were determined in the samples (chlorogenic acid wasnot detected in some of the tested samples). The variation of themain compounds is presented in Fig. 3. It is evident, that oenotheinB was more sensitive to the drying than rutin. The decrease ofoenothein B after drying was up to 5.0 times, while rutin – up to2.2 times.

It is evident, that C. angustifolium extracts have a few character-istic compounds, eluting at 12.6 min, 28.5 min, and 29.3 min,between 33.0 and 35.0 min, and 36.6 min. Those compounds werenot identified due to the lack of reference compounds. Chlorogenicacid (low content), caffeic acid and oenothein B (medium content)and quercetin-7-O-glucoronide (high content) were found inC. angustifolium collected in Hungary (Hevesi Tóth et al., 2009).Interestingly, in the study of Hevesi Tóth et al. (2009), C. angustifoliumdistinguished from other Epilobium species (i.e., Epilobium parviflorumSchreb., Epilobium roseum L., Epilobium tetragonum L. and Epilobiummontanum L.) by medium amount of oenothein B and the presenceof high amount of quercetin-7-O-glucoronide, while this compoundwas not detected and oenothein B found in remarkably higheramounts in other species. According to the published data, quercetinglycosides (particularly quercetin glucuronide) dominate inC. angustifolium extracts in contrast to other Epilobium species,where myricetin glucosides are the main compounds (Ducreyet al., 1995; Hevesi Tóth et al., 2009). Caffeic and chlorogenic acidswere also determined in C. angustifolium by Ruszova et al. (2013)from Czech Republic. Glycosides of other flavonoids, like myricetin,quercetin and kaempferol, were determined in the latter study.

As it was mentioned above, oenothein B was predominant in alltested samples, although its content varied considerably (Fig. 3).The highest amounts were determined in the samples introducedfrom Moletai (MOL) and Šiauliai (SIA) districts; while the lowest– form Švencionys (SVE) district. It was reported, that oenotheinB, a dimeric macrocyclic ellagitannin, is one of the main


biologically-active components in Epilobium taxa, and Epilobiumspecies accumulate this compound in high concentrations(Ducrey et al., 1997). The main responsibility for the wide rangeof C. angustifolium biological activities (antioxidant, antitumor,antibacterial, antiviral, imunomodulatory, etc.) is attributed tooenothein B (Kiss et al., 2006a,b, 2011; Schepetkin et al., 2009).However, oenothein B was not the only compound responsiblefor DPPH free radical scavenging activity, because correlationbetween the amount of oenothein B and RSA obtained by spec-trophotometric assay was weak, i.e., correlation coefficient was




Fig. 3. The variation of oenothein B and rutin in the tested C. angustifolium L. samples. Note: HPLC/UV peak area of fresh samples was recalculated taking into account themoisture content.


0.40 and 0.36 in fresh and dried samples, respectively. Therefore,RSA of C. angustifolium extracts is complex.

2.3. Volatile compounds composition

Totally 42 different compounds were detected in the headspaceof C. angustifolium by means of solid phase microextraction cou-pled with GC/MS. As it was expected, the higher diversity of com-pounds was found in the headspace of fresh samples rather thandried, i.e., 30 and 25 different compounds were identified in freshand dried samples, respectively. The significant qualitative andquantitative differences were observed not only between freshand dried, but also between different samples (Table 1). The iden-tified components involve different classes of chemical compoundsincluding alcohols, aldehydes, esters, terpenes, saturated andunsaturated, linear and branched hydrocarbons. The drying ofC. angustifolium had two effects to the composition of the herb,i.e., the quantitative changes of compounds and disappearance/appearance of several compounds. Some compounds, like trans-2-hexenal, phenylacetaldehyde, nonanal, menthol, cis-anethole,decanal, trans-anethole, c-terpinen-7-al, b-ionone, and tetrade-cane, were found only in dried samples of C. angustifolium.cis-3-Hexenol, cis-3-hexenyl isobutanoate, cis-3-hexenylbutyrate,c-elemene, e-amorphene, and dihydroapofarnesal were presentonly in the fresh samples. Only 5 out of 42 compounds weredetected in all samples, i.e., 2-methyl-cis-3-hexenylbutanoate,b-bourbonene, b-caryophyllene, a-caryophyllene, b-cubebene,and c-cadinene. The quantitative composition varied dependingon the sample origin. For example, percentage content of trans-2-hexenal in dried samples varied from 16.0% to 55.9%, cis-3-hexenolin fresh samples was in the range of 17.5% and 68.6%, amount ofb-caryophyllene varied from 1.5% to 30.5% (Table 1). Possibly theslow drying (room temperature at the dark) causes oxidation ofalcohols to aldehydes, therefore several aldehydes (like trans-2-hexenal, phenylacetaldehyde, nonanal, decanal, c-terpinen-7-al)were detected in dried samples only. No compound or group ofthe compounds predominant in all the samples investigated weredetermined, therefore the existence of different chemotypes ofC. angustifolium growing in Lithuania is proved. The literature dataabout the volatile compounds composition of C. angustifolium isscare.

As results show, the drying had negative effect due to theremarkable loss of volatiles – the total GC peak area reduced by2.0–8.4 times. The loss and changes of volatile compounds compo-sition was also demonstrated by other authors (Okoh et al., 2008;


Shunying et al., 2005). However the appearance of several newcompounds, like anethole, can have a positive effect on the qualityof C. angustifolium. It was found that anethole is effective againstyeast, bacterial and fungal strains (De et al., 2002; Huang et al.,2010). High amounts of anethole (89.5%) were found in Illiciumverum fruit essential oils by Huang et al. (2010) and in Backhousiaanisata (71.2–93.7%) by Blewitt and Southwell (2000). After dryingthe percentages of b-caryophyllene and a-caryophyllene increasedin 9 out of 12 samples. It was found, that Zingiber nimmonii rhi-zome oil rich in caryophyllene (60.9%) showed significant inhibito-ry activity against some fungi, like Candida glabrata, Candidaalbicans and Aspergillus niger and the bacteria Bacillus subtilis andPseudomonas aeruginosa, however no activity was observed againstthe fungus Fusarium oxysporum (Sabulal et al., 2006). The eval-uation of antimicrobial activities of the volatile compounds of C.angustifolium was beyond the scope of this study, however accord-ing to the above-mentioned published data the volatile compoundsof this herb also can be of therapeutic value.

It is evident, that drying has negative effect on the quality ofplant material (reduction of total amount and several phenoliccompounds, flavonoids, DPPH free radical activity, volatile com-pounds). However air-drying is the most popular method usedfor the preparation, conservation and storage of plant raw material(like herbs or spices) for a longer period of time. Some studies rec-ommend freeze-drying instead of drying at ambient and elevatedtemperatures (especially higher), since freeze-drying causes lowerdegradation of biologically active compounds (Materska, 2014;Krokida and Philippopoulos, 2006; Chan et al., 2012), but on theother hand, freeze-drying requires a special equipment and higherenergy resources, so drying at low temperatures is recommended.

2.4. Chemometric analysis

Statistical analysis was started with representative data setsbuilding. As the number of the measurements was relatively small,the well-known Monte Carlo algorithm was employed to randomlygenerate data, assuming Normal (known as Gaussian) distributionfunction, having two parameters – l and r, where l is the meanand r is the standard deviation of the variable.

To build data sets, corresponding to the volatile compoundscomposition of the tested fresh and dried C. angustifolium L. sam-ples, the Monte Carlo algorithm with parameters (lvi, rv, N) wasapplied, where lvi is the mean estimate of the i-th volatile com-pound (Table 1), i is the compound index from 1 to 42 (includingfresh and dried samples), and rv is the largest observed standarddeviation estimate of all identified volatile compounds (including




Table 1Identified compounds in the samples of C. angustifolium L.

Identified compounda RT KBGb KRF ALE MOL VAR SVE TRA ALY UKM PRI SIA ZAR

Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried Fresh Dried

1 trans-2-Hexenal 3.307 25.5 40.3 55.9 47.1 30.6 41.1 16.0 23.1 20.3 40.2 39.3 43.02 cis-3-Hexenol 3.369 31.4 68.6 55.0 64.8 38.1 38.0 66.9 46.6 37.2 17.5 49.7 53.13 Benzyl alcohol 6.511 0.5 0.8 0.5 0.6 0.6 0.6 0.3 3.1 0.6 0.54 Phenylacetaldehyde 6.709 1.7 3.3 2.0 0.7 2.4 1.2 1.5 1.9 2.35 cis-3-Hexenol propanoate 7.967 2.7 0.4 1.8 3.2 0.8 2.8 1.0 2.3 0.6 2.2 1.7 1.9 2.1 1.46 3-Methylbutyl-2-

methylbutyrate8.062 1.3 2.5

7 Nonanal 8.075 2.2 11.4 5.3 4.5 4.0 4.6 2.1 2.0 6.7 9.68 Phenethyl alcohol 8.379 4.1 2.6 1.8 3.6 1.5 3.7 1.8 3.3 1.8 0.7 3.3 2.2 1.7 1.7 2.3 4.7 4.6 2.9 0.9 3.0 0.7 2.49 cis-3-Hexenyl isobutanoate 9.085 0.6 1.2 1.3 1.2 0.7 1.5 0.2 0.4 1.2 0.310 Menthol 9.860 0.3 1.2 0.5 1.5 0.4 0.7 0.4 1.1 0.7 1.311 cis-3-Hexenyl butyrate 10.181 2.9 10.8 8.5 4.9 4.9 10.4 2.4 2.4 3.5 5.1 5.0 3.012 cis-Anethole 10.545 1.6 0.8 1.7 1.5 1.6 0.8 0.8 2.6 2.0 1.2 1.613 Decanal 10.665 0.9 1.9 2.1 1.9 2.2 0.8 1.2 0.7 1.7 2.5 2.514 2-Methyl-cis-3-

hexenylbutanoate11.351 7.1 0.9 11.0 1.7 13.3 1.4 12.2 0.9 11.3 1.1 20.0 1.4 3.5 0.6 1.9 1.5 3.2 0.7 11.3 0.9 2.2 1.1 1.2 0.9

15 trans-Anethole 12.814 9.0 6.2 2.6 10.0 10.3 11.5 10.9 3.4 46.2 16.4 12.9 7.916 cis-3-Hexen-1-yl-2-methyl

crotonate13.792 2.2

17 c-Terpinen-7-al 14.371 0.8 1.6 1.0 0.4 1.4 1.5 1.718 Eugenol 14.695 0.3 0.5 0.3 3.2 0.519 b-Bourbonene 15.389 2.3 7.0 0.4 2.5 0.6 2.0 0.5 2.0 3.5 5.9 0.5 2.6 0.5 7.0 1.9 7.8 4.8 0.8 2.8 1.7 2.3 1.7 1.3 3.120 c-Elemene 15.549 0.9 0.2 0.9 0.4 1.4 1.021 Tridecane 15.622 1.5 0.5 1.3 0.6 1.2 0.2 0.3 0.6 0.7 1.522 b-Caryophyllene 16.273 21.3 24.1 1.5 11.0 5.7 9.9 4.3 9.4 17.5 19.8 6.0 12.7 5.2 30.5 18.7 26.3 18.0 7.7 17.3 10.0 13.6 9.4 12.8 13.023 b-Cubebene 16.494 1.7 1.7 0.8 0.6 0.6 0.4 0.7 1.6 1.4 0.8 1.0 1.4 1.7 1.7 1.7 2.8 0.7 2.2 0.7 1.8 0.8 2.2 0.924 a-Caryophyllene 17.108 15.5 17.9 1.0 8.1 4.2 7.3 3.2 7.1 11.9 13.0 4.8 9.3 2.8 21.8 14.4 22.1 10.7 3.2 12.7 7.2 9.6 7.2 7.9 9.425 c-Cadinene 17.777 5.8 2.7 0.3 1.4 2.6 1.1 1.3 1.3 3.9 2.2 3.8 1.8 8.4 3.1 5.8 2.4 11.9 3.1 9.5 1.9 8.6 2.0 12.5 1.926 b-Ionone 17.878 2.6 1.5 2.6 1.1 2.6 0.8 1.3 1.3 2.3 2.8 2.827 e-Amorphene 18.227 0.5 0.4 0.3 0.6 1.0 0.7 0.7 1.028 d-Cadinene 18.776 0.5 1.2 0.6 0.3 0.5 0.3 0.6 0.6 0.5 0.6 0.9 0.7 0.4 0.6 0.7 0.7 0.6 0.829 b-Bisabolene 19.655 0.5 2.430 Dihydroapofarnesal 19.928 1.2 0.9 1.0 0.6 2.7 0.8 0.8 0.6 3.6 1.1 0.531 Tetradecane 20.430 0.6 1.3 2.4 2.6 0.5 0.9 1.1 2.0Total GC peak area (relative units) 196.1 72.1 222.6 26.6 279.1 65.4 204.1 34.6 235.8 69.1 187.9 38.2 123.6 136.0 155.2 78.4 176.03 125.5 363.8 56.0 220.0 43.7 225.8 31.3

Relative standard deviation did not exceed 6.7%.a Compounds, which amount was less than 1.0% according to peak area relative to the total peak area, i.e., isopentylbutyrate, 2,6-dimethyl-7-octen-4-one, n-decane, cis-3-hexenyl-pentanoate, n-dodecane, a-copaene, cis-3-

hexenyl hexanoate, isogermacrene D, trans-a-farnesene, n-hexadecanol, tetrahydrofurfuryl propionate, are not included in the Table.b Sample codes as in Table 3.

V.K

aškonieneet

al./Phytochemistry

xxx(2015)

xxx–xxx

5

Pleasecite

thisarticle

inpress

as:K

aškoniene,V.,

etal.Evalu

ationof

phytochemicalcom

positionof

freshand

driedraw

materialof

introducedCham

erionangustifolium

L.using

chromatographic,

spectrophotometric

andchem

ometric

techniques.Phytochem

istry(2015),

http://dx.doi.org/10.1016/j.phytochem

.2015.02.005




fresh and dried samples), which, expressed in relative form,equaled to rv = 6.8%, and N = 1000 is a chosen number of datapoints to be generated. After employing the algorithm, consideringthe defined data distribution, formed initial data matrices were fil-tered rejecting volatile compounds contributing less than 1% andstandardized subtracting the calculated mean estimate and divid-ing to standard deviation estimate. These processing steps resultedin building volatile compounds data matrices of sizes of[12,000 � 12] and [12,000 � 18], representing 12 fresh and 12dried samples, described by 12 and 18 volatile compounds,respectively.

As the formed variable space was composed of 12 and 18dimensions, respectively, which obviously were correlated (notorthogonal), principal components analysis was performed. PCAis a linear transformation that transforms the data set to a newcoordinate system such that the new set of variables, called princi-pal components, are linear functions of the original variables andare uncorrelated. The greatest variance by any projection of thedata set comes to lie on the first coordinate, the second greatestvariance on the second coordinate, etc. Although the space of theprincipal components after the analysis results in the same sizeas the data matrix, decorrelation allows retaining only those prin-cipal components that have the highest variances, while rejectingthe rest. This rejection shrinks considerably the newly composedorthogonal principal components space. The number of PCA com-ponents was selected according to applied several decision criteria(Cattell’s scree plot test, Kaiser’s test, variance explained test, etc.),which led to keeping only first 3 and 4 principal components,explaining 84.43% and 86.00% of the total variance, correspondingto fresh and dried samples, respectively. Therefore two matrices ofsizes of [12,000 � 3] and [12,000 � 4] were formed for successivesamples classification techniques.

Table 2Classification of C. angustifolium L. samples using hierarchical cluster analysis.

Sample Volatile composition Phenolic composition

Fresh Dried Fresh Dried

PRI 1 1 1 1SVE 1 1 1 1MOL 1 1 2 1ALE 1 1 1 1KRF 1 1 1 1ALY 2 2 2 2KBG 2 2 2 2VAR 2 2 2 2TRA 3 2 1 2UKM 3 1 2 1ZAR 3 1 1 1SIA 3 1 1 1

Sample codes as in Table 3.

Table 3Tested samples of Chamerion angustifolium L.

Sample code Introduction year Growing and

KBG Native Kaunas BotanKRF 2012 Kazluz R�uda f

ALE 2012 Aleksotas, KaMOL 2012 Moletai distrVAR 2012 Varena distr.SVE 2012 Švencionys dTRA 2012 Trakai distr.,ALY 2013 Alytus cityUKM 2013 Ukmerge disPRI 2013 Prienai distr.SIA 2013 Šiauliai distrZAR 2013 Zarasai distr.


Data matrices corresponding to the composition of the phenoliccompounds (i.e., oenothein B, rutin and other three unidentifiedcompounds) of fresh and dried samples was prepared for classifica-tion in a similar manner. Again, the Monte Carlo algorithm, assum-ing Normal distribution function, with the following parameters(lpj, rpj, N), where lpj and rpj are the mean and standard deviationestimates of the jth phenolic compound, j is the compound indexfrom 1 to 5 and N = 1000 is the chosen number of data points beinggenerated, was applied. After the data standardization procedure,PCA was performed. Principal components retaining decision crite-ria allowed to retain 2 and 2 principal components, having the lar-gest variances and explaining 80.56% and 87.25% of the totalvariance, corresponding to fresh and dried samples, respectively.Thus two phenolic composition data matrices of size of[12,000 � 2] were prepared for further statistical mining.

The same procedure of data matrices formation was run onspectrophotometric data, corresponding to antioxidant propertiesand total amounts of the phenolic compounds and flavonoids.The Monte Carlo algorithm, with assumed Normal distributionfunction, with parameters (lsl, rsl, N), where lsl and rsl are themean and standard deviation estimates of the lth variable of thespectrophotometric assay, l is the variable index from 1 to 3,N = 1000 is the chosen number of data points to be generated,was performed. After the data standardization procedure, PCAwas used to eliminate correlation of the data. PCA componentsrejection criteria suggested to keep first 2 and 2 principal compo-nents, explaining 83.83% and 90.85% of the total variance, corre-sponding to fresh and dried samples, respectively. These formeddata matrices of spectrophotometric assay of size of [12,000 � 2]were used for the samples clustering.

There were two techniques, hierarchical clustering analysis andK-means clustering analysis, adopted for classification of the ana-lyzed fresh and dried C. angustifolium L. samples to groups. HCA

Spectrophotometric data Classification

Fresh Dried

2 1 Southern-middle Lithuania2 1 Southern-middle Lithuania1 1 Northern-middle Lithuania1 1 Northern-middle Lithuania1 1 Northern-middle Lithuania2 2 Southern Lithuania2 2 Southern Lithuania2 2 Southern Lithuania1 2 –2 2 –1 1 Northern Lithuania1 1 Northern Lithuania

collection place Geographic coordinates

ical Garden, Kaunas city 54�52017.7500 , 23�54039.3900

orest, Kaunas distr. 54�43059.0200 , 23�29027.7600

unas city 54�53026.2700 , 23�53057.1400

., Stirniai 55�1603.800 , 25�36041.9400

, Panara 54�6018.8100 , 24�6041.400

istr., Svirkos 55�10013.7700 , 26�26026.8200

Uzutrakis forest 54�4009.9100 , 24�56051.1500

54�23053.4700 , 24�2048.9200

tr., Vidiškiai 55�18030.8500 , 24�51055.900

, Strielciai 54�4004.8200 , 23�57052.1100

., Kuziai 55�58056.5500 , 23�8022.2900

, Deguciai 55�39038.4500 , 26�3027.6600




Fig. 4. K-means clustering analysis scatter plots of fresh samples: (a) volatilecompound composition data, (b) phenolic compounds composition data, (c)spectrophotometric data.


grouped the samples according to the evaluated similarity (recip-rocal to distance) measure and calculated linkage. The distancesand linkages were calculated for the previously formed data matri-ces using Spearman’s distance function and weighted average dis-tance linkage rule. The above described distance calculationfunction and linkage rule for HCA were chosen in accordance withcalculated cophenetic correlation coefficient, which is a measure ofhow faithfully a HCA dendrogram preserves the pairwise distancesbetween the original data points. The cophenetic correlation coef-ficients were 0.83, 1.00 and 1.00 for volatile composition data, phe-nolic composition data and spectrophotometric data of the freshsamples, respectively and 0.93, 1.00 and 1.00 for volatile composi-tion data, phenolic composition data and spectrophotometric dataof the dried samples, respectively. The calculated HCA dendro-grams, were cut at the level of 70% of the maximum calculated dis-tance to build clusters and samples are considered too dissimilarbeyond that point. The results of HCA are summarized in Table 2.for clearer view and better comparability.

Analyzing the HCA classification results it was clearly seen 3, 2,2, 2, 2 and 2 groups suggested for the volatile composition of freshand dried samples, phenolic composition of fresh and dried sam-ples and spectrophotometric data of fresh and dried samples. Thecomparison of these groups revealed that their composition didnot fully coincide comparing either the results achieved using dif-ferent type of data (volatile composition, phenolic composition andspectrophotometric data) or data representing different state ofsamples (fresh or dried). As the samples were tended to migrate,i.e., some of them, attributed to one group because of similarityfrom one point of view, were too dissimilar and therefore labeledas members of other group from other point of view, it was obviousthat there were more classes of samples than the HCA resultsrevealed. Indeed, after closer examination of the dendrograms pre-sented in table form (Table 2), 6 different groups of samples couldbe distinguished: (1) ZAR and SIA; (2) MOL, ALE and KRF; (3) PRIand SVE; (4) ALY, KBG and VAR; (5) TRA; (6) UKM. The samplesin each presented group always fell into the same class, i.e.,retained similarity, despite the analyzed data or state of the sam-ple, except one case of sample MOL – fresh sample MOL fell intodifferent group analyzing the phenolic compounds compositiondata). Looking at the map of Lithuania, these groups can be corre-lated to geographic C. angustifolium L. collection locations: 1stgroup corresponds to northern Lithuania, 2nd and 3rd groups rep-resent the northern-middle Lithuania and southern-middle Lithua-nia and 4th group correspond to southern Lithuania. 2nd and 3rdgroups could be aggregated as one group to represent the middleLithuania, but there will be an exception: fresh samples PRI andSVE (noted as southern-middle Lithuania) clustered differentlyfrom MOL, ALE and KRF samples (noted as northern-middle Lithua-nia) in respect of the spectrophotometric data. The characteristicsof samples TRA and UKM were too different to be connected toany of previously presented groups, thus they left as separateclasses. The origin of the sample KBG, which has been growing inKaunas Botanical Garden, is unknown as the plant was broughtthere from its original location and planted artificially. This samplewas classified as southern Lithuanian despite the fact that KaunasBotanical Garden lies in the middle Lithuania.

Since HCA exposes similarities between samples and links themtogether according to calculated distance using mean values of theinput data and do not take into account variances, another classifi-cation method, K-means clustering, based on iterative computa-tion, was applied to do the classification. This tool, however,requires specifying a number of the initial clusters, which can bereduced by the algorithm if empty cluster is created during compu-tation. This number was calculated using rule of thumb – sqrt(K),where K – is the number of variables. To find the global classifica-tion solution, i.e., to minimize the sum, over all clusters, of the


within-cluster sums of point-to-cluster-centroid distances, KMCAwas replicated 200 times (for each classification task) reportingthe best achieved result. The results of the accomplished KMCA




Fig. 5. K-means clustering analysis scatter plots of dried samples: (a) volatilecompounds composition data, (b) phenolic compounds composition data, (c)spectrophotometric data.


are presented in Fig. 4 (for fresh samples) and Fig. 5 (for dried sam-ples). The scatter plots are presented in the 3D space of the princi-pal components, PCv 1 – PCv 2 – PCv 3, in the case of the volatile


compounds data (Figs. 4a and 5a), in the 2D space of the principalcomponents, PCp 1 – PCp 2, in the case of the phenolic compositiondata, i.e., oenothein B and other four characteristic compounds,(Figs. 4b and 5b) and in the 2D space of principal components,PCs 1 – PCs 2, in the case of the spectrophotometric data, represent-ing antioxidant properties and total amounts of the phenolic com-pounds and flavonoids (Figs. 4c and 5c).

KMCA results showed 3, 3, 2, 2, 2 and 2 groups doing classifica-tion in respect of the volatile composition of fresh and dried sam-ples, phenolic composition of fresh and dried samples andspectrophotometric data of fresh and dried samples. After closerlook at presented data scatter plots, slightly different formedgroups can be noticed comparing to previous classification tech-nique. Nevertheless, the earlier presented groups of samples, rep-resenting different geographical locations of Lithuania, wereconfirmed. However, two exceptions should be noted: fresh sam-ple PRI and dried sample MOL were classified into different groupsanalyzing the volatile compounds composition data. Besides, K-means clustering results exposed bigger differences betweennorthern-middle Lithuania and southern-middle Lithuania samplegroups.

3. Conclusions

Phytochemical composition of C. angustifolium herb collected indifferent locations of Lithuania was analyzed. The results revealedthat not only the composition was dependent on the geographicalorigin, but also on sample preparation procedure (i.e., fresh or dry).Comparing dried and fresh samples, it was found that the first con-tained from 1.5 to 3.5 times less amount of total phenolic com-pounds and exhibited from 2.0 to 4.5 times lower radicalscavenging activity, while reduction of the content of total flavo-noids from 1.2 to 1.6 times was observed only in several samplesas the other samples showed increase from 1.2 to 2.2 times. Corre-lation of the radical scavenging activity with the content of pheno-lic compounds was higher than with the content of flavonoids.

HPLC analysis allowed to detect oenothein B, rutin, quercetin,gallic acid in all fresh and dried tested samples. Chlorogenic acidwas present in the most of the samples.

Solid phase microextraction coupled with GC/MS analysisrevealed the predominance of trans-2-hexenal (16.0–55.9%) andtrans-anethole (2.6–46.2%) in the dried samples only, while cis-3-hexenol (17.5–68.6%) dominated in the fresh samples. Caryophyl-lenes (a- and b-), found in all analyzed samples, were in the rangefrom 2.4% to 52.3% of all volatiles, varying on the plant geo-graphical origin and sample preparation.

The classification of the fresh and dried C. angustifolium herb inrespect of volatile compounds composition, phenolic compoundscomposition and spectrophotometric data, representing antioxi-dant properties and total amounts of the phenolic compoundsand flavonoids using hierarchical clustering analysis and K-meansclustering analysis was applied. Four groups were determined,which were attributed to northern, northern-middle, southern-middle and southern Lithuania according to the origin of thesamples.

4. Experimental

4.1. Plant material

Eleven samples of C. angustifolium L. were collected from elevendifferent locations in Lithuania at 2012 and 2013, and introducedto ex situ collections of medicinal and aromatic plants sector ofKaunas Botanical Garden of Vytautas Magnus University. A com-parative study of the phytochemical composition of the plants





grown after introduction was carried out. The tested samples andtheir codes are listed in Table 1. A sample of C. angustifolium nativegrown in the medicinal plant collection of Kaunas Botanical gardenwas used for the comparative purposes.

Samples were identified by prof. O. Ragazinskiene. The aerialpart of C. angustifolium herb was cut during the massive bloomingvegetation period. Part of the plant was analyzed immediately aftercollection, while another part was dried at room temperature in awell-ventilated shadow place. 500 mg of fresh or dried materialwas extracted with 20 ml 75% aqueous methanol for 24 h in Titer-Tek orbital shaker (Germany). Extracts were filtered through apaper filter and afterwards through 0.22 lm pore-size disposablemembrane filter. The extracts were stored at +4 �C and used forspectrophotometric and HPLC analysis.

The moisture content in fresh and air dried samples was deter-mined by PMB-53 Moisture balance (Adam Equipment, Kingston,UK) according to the recommendations of manufacturer. The mois-ture content in the fresh samples was in the range of 65.3–85.5%,while air dried samples contained 9.0–9.5% of the moisture.

4.2. Spectrophotometric assays

Total phenolic compounds content and total flavonoids contentwere determined using Folin–Ciocalteu method and colorimetricaluminum trichloride assay, respectively, as described byKaškoniene et al. (2015). The determination of radical scavengingactivity was performed using 2,2-diphenyl-1-picrylhydrazyl(DPPH) free radical according to Kaškoniene et al. (2015). The mea-surements were compared to a calibration curve of rutin solutionsand expressed in mg of rutin equivalents per gram of herb sample.All experiments were performed in triplicate.

In order to compare the spectrophotometric data of fresh anddried samples, the results of total phenolic content, total flavonoidscontent and radical scavenging activity of fresh samples wererecalculated per one gram of dry mass considering the moisturecontent in the tested samples.

4.3. HPLC analysis

The qualitative comparison of phenolic compounds in C. angus-tifolium extracts was performed using HPLC system with UV detec-tor. HPLC system consisted of Varian 9012 Solvent Delivery System(USA), Linear 206 PHD (Linear Instruments, USA) UV detector,Perkin Elmer Series 200 LC AutoSampler (Norwalk, CT, USA), andClarity™ chromatographic workstation (DataApex, Prague, CzechRepublic).

Separation was carried out using the reversed phase analyticalcolumn LiChrospher 100 RP-18e, 125 mm � 4.6 mm i.d. (Merck,Germany), packed with C18 stationary phase with the particle sizeof 5 lm. Detection was carried out at 254 nm wavelength. The fol-lowing eluents were used: A – water with 0.05% trifluoroaceticacid; B – HPLC grade methanol with 0.05% trifluoroacetic acid.The gradient program was as follows: from 0 to 5 min an increasefrom 1% to 25%, from 5 min to 10 min held at 25%, from 10 min to38 min from 25% to 60%, from 38 min to 40 min from 60% to 99%,from 40 min to 45 min 99% of component B, and then 3 min toreturn to the initial conditions. Equilibration time of 10 min wasallowed before the injections. The flow rate was 0.75 ml/min, andthe injection volume was 10 lL.

The identification of compounds was performed comparing theretention time of analyte and reference compounds.

4.4. Solid phase microextraction – GC/MS analysis

Volatile compounds composition was determined using head-space solid phase microextraction (SPME) coupled with GS/MS.


Extraction of C. angustifolium volatiles was performed on PDMS/DVB Stable Flex fiber (polydimethylsiloxane/divinylbenzene coat-ing thickness 65 lm, Supelco, Bellefonte, PA).

50 mg of dried or 25 mg of fresh samples were added into 10 mlvials and thermostated for 15 min at 50 �C. The analysis was car-ried out using a GC/MS system (GCMS-QP2010, Shimadzu, Tokyo,Japan). A Restec (Bellefonte, USA) RTX-5MS (30 m � 0.25 mmi.d. � 0.25 lm film thickness) GC column was used. The oven tem-perature gradient started at 30 �C and raised to 200 �C at 5 �C/min,then raised to 280 �C at 20 �C/min and was held for 2 min. Helium(99.999%) was used as carrier gas with a constant flow rate of1.2 ml/min (AGA, Lithuania). The injector temperature was keptat 230 �C, injection was performed in the split mode (1:10). Themass spectrometric detector was operated in electron impactmode (70 eV). The ion source and interface temperatures wereset at 220 �C and 260 �C correspondingly. Identification of com-pounds was performed according to their mass spectra library(NIST v1.7). Positive identification was assumed when good match-es (80% and more) of the mass spectra were achieved.

4.5. Chemometric analysis

Statistical data mining was performed using MATLAB v8.2(R2013b) software on Windows 8.1 x64 platform. Representativedata sets for the successive analysis from small number of observa-tions were built applying the well-known Monte Carlo data gen-eration algorithm. The Normal (Gaussian) distribution function ofthe measurement data was assumed. Generated data sets werestandardized by subtracting the mean and dividing to the standarddeviation. Data representing volatile compounds composition wereprocessed additionally, including filtering, in order to exclude com-pounds, having magnitudes lower than 1%, from further statisticalanalysis.

The chemometric data exploration included principal compo-nents analysis (PCA), hierarchical clustering analysis (HCA) andK-means clustering analysis (KMCA). PCA was used as a tool fordata compression by transforming correlated initial set of variablesto non-correlated set of principal components. HCA and KMCAhelped to classify the analyzed C. angustifolium L. samples havingsimilar characteristics to classes and visualize the classificationresults in the space of the principal components. Both clusteringtechniques were applied to link the samples to clusters accordingto the composition of the volatiles, the composition of the phenoliccompounds and the spectrophotometric data representing antiox-idant properties and total amounts of the phenolic compounds andflavonoids.

Acknowledgment

The study was financially supported by the Research Council ofLithuania (Grant No. MIP 084/2012).

References

Al-Rawahi, A.S., Rahman, M.S., Guizan, I.N., Essa, M.M., 2013. Chemical composition,water sorption isotherm, and phenolic contents in fresh and dried pomegranatepeels. Drying Technol. 31, 257–263.

Aprotosoaie, A.C., Spac, A., Hancianu, M., Miron, A., Tanasescu, V.F., Dorneanu, V.,Stanescu, U., 2010. The chemical profile of essential oils obtained from fennelfruits (Foeniculum vulgare Mill.). Farmacia 58, 46–53.

Battinelli, L., Tita, B., Evandri, M.G., Mazzanti, G., 2001. Antimicrobial activity ofEpilobium spp. extracts. Il Farmaco 56, 345–348.

Blewitt, M., Southwell, I.A., 2000. Backhousia anisata Vickery, an alternative sourceof (E)-anethole. J. Essent. Oil Res. 12, 445–454.

Capecka, E., Mareczek, A., Leja, M., 2005. Antioxidant activity of fresh and dry herbsof some Lamiaceae species. Food Chem. 93, 223–226.

Chan, E.W.C., Kong, L.Q., Yee, K.Y., Chua, W.Y., Loo, T.Y., 2012. Antioxidant andantibacterial properties of some fresh and dried Labiatae herbs. Free Radic.Antioxid. 2, 20–27.


http://refhub.elsevier.com/S0031-9422(15)00052-7/h0005


















De, M., De, A.K., Sen, P., Banerjee, A.B., 2002. Antimicrobial properties of star anise(Illicium verum Hook f). Phytother. Res. 16, 94–95.

Ducrey, B., Wolfender, J.L., Marston, A., Hostettmann, K., 1995. Analysis of flavonolglycosides of thirteen Epilobium species (Onagraceae) by LC–UV andthermospray LC–MS. Phytochemisry 38, 129–137.

Ducrey, B., Marston, A., Gohring, S., Hartmann, R.W., Hostettmann, K., 1997.Inhibition of 5 a-reductase and aromatase by the ellagitannins oenothein A andoenothein B from Epilobium species. Planta Med. 63, 111–114.

Hevesi Tóth, B., Blazics, B., Kéry, Á., 2009. Polyphenol composition and antioxidantcapacity of Epilobium species. J. Pharm. Biomed. Anal. 49, 26–31.

Huang, Y., Zhao, J., Zhou, L., Wang, J., Gong, Y., Chen, X., Guo, Z., Wang, Q., Jiang, W.,2010. Antifungal activity of the essential oil of Illicium verum fruit and its maincomponent trans-anethole. Molecules 15, 7558–7569.

Kaškoniene, V., Kaškonas, P., Maruška, A., Ragazinskiene, O., 2013. Essential oils ofBidens tripartita L. collected during period of 3 years composition variationanalysis. Acta Physiol. Plantarum 35, 1171–1178.

Kaškoniene, V., Kaškonas, P., Maruška, A., 2015. Volatile compounds compositionand antioxidant activity of bee pollen collected in Lithuania. Chem. Pap. 69,291–299.

Kiss, A., Kowalski, J., Melzig, M.F., 2006a. Induction of neutral endopeptidase activityin PC-3 cells by an aqueous extract of Epilobium angustifolium L. and oenotheinB. Phytomedicine 13, 284–289.

Kiss, A.K., Bazylko, A., Filipek, A., Granica, S., Jaszewska, E., Kiarszys, U., Kosmider, A.,Piwowarski, J., 2011. Oenothein B’s contribution to the anti-inflammatory andantioxidant activity of Epilobium sp. Phytomedicine 18, 557–560.

Kiss, A., Kowalski, J., Melzig, M.F., 2006b. Effect of Epilobium angustifolium L. extractsand polyphenols on cell proliferation and neutral endopeptidase activity inselected cell lines. Pharmazie 61, 66–69.

Krokida, M.K., Philippopoulos, C., 2006. Volatility of apples during air and freezedrying. J. Food Eng. 73, 135–141.

López-Nicolás, J., García-Carmona, F., 2009. Enzymatic and non-enzymaticdegradation of polyphenols. In: Fruit and Vegetable Phytochemicals:Chemistry, Nutritional Value, and Stability. Wiley-Blackwell Publishing, Ames,Iowa, USA, pp. 101–129.

Materska, M., 2014. Bioactive phenolics of fresh and freeze-dried sweet and semi-spicy pepper fruits (Capsicum annuum L.). J. Funct. Food 7, 269–277.

Milos, M., Mastelic, J., Jerkovic, I., 2000. Chemical composition and antioxidanteffect of glycosidically bound volatile compounds from oregano (Origanumvulgare L. ssp. hirtum). Food Chem. 71, 79–83.

Moilanen, J., Sinkkonen, J., Salminen, J.P., 2013. Characterization of bioactive plantellagitannins by chromatographic, spectroscopic and mass spectrometricmethods. Chemoecology 23, 165–179.


Mrad, N.D., Boudhrioua, N., Kechaou, N., Courtois, F., Bonazzi, C., 2012. Influence ofair drying temperature on kinetics, physicochemical properties, total phenoliccontent and ascorbic acid of pears. Food Bioprod. Process. 90, 433–441.

Okoh, O.O., Sadimenko, A.P., Asekun, O.T., Afolayan, A.J., 2008. The effects of dryingon the chemical components of essential oils of Calendula officinalis L. Afr. J.Biotechnol. 7, 1500–1502.

Rota, M.C., Herrera, A., Martinez, R.M., Sotomayor, J.A., Jordan, M.J., 2008.Antimicrobial activity and chemical composition of Thymus vulgaris, Thymuszygis and Thymus hyemalis essential oils. Food Control 19, 681–687.

Ruszova, E., Cheel, J., Pávek, S., Moravcova, M., Hermannová, M., Matejková, I.,Spilkova, J., Velebny, V., Kubala, L., 2013. Epilobium angustifolium extractdemonstrates multiple effects on dermal fibroblasts in vitro and skin photo-protection in vivo. Gen. Physiol. Biophys. 32, 347–359.

Sabulal, B., Dan, M., Kurup, R., Pradeep, N.S., Valsamma, R.K., George, V., 2006.Caryophyllene-rich rhizome oil of Zingiber nimmonii from South India: chemicalcharacterization and antimicrobial activity. Phytochemistry 67, 2469–2473.

Schepetkin, I.A., Kirpotina, L.N., Jakiw, L., Khlebnikov, A.I., Blaskovich, C.L., Jutila,M.A., Quinn, M.T., 2009. Immunomodulatory activity of oenothein B isolatedfrom Epilobium angustifolium. J. Immunol. 183, 6754–6766.

Shunying, Z., Yang, Y., Huaidong, Y., Yue, Y., Guolin, Z., 2005. Chemical compositionand antimicrobial activity of the essential oils of Chrysanthemum indicum. J.Ethnopharmacol. 96, 151–158.

Stahl, W., Sies, H., 2003. Antioxidant activity of carotenoids. JMAM 24, 345–351.Stahl, W., Sies, H., 2005. Bioactivity and protective effects of natural carotenoids.

BBA Mol. Basis Dis. 1740, 101–107.Vitalone, A., Bordi, F., Baldazzi, C., Mazzanti, G., Saso, L., Tita, B., 2001. Anti-

proliferative effect on a prostatic epithelial cell line (PZ-HPV-7) by Epilobiumangustifolium L. Il Farmaco 56, 483–489.

Vitalone, A., Guizzetti, M., Costa, L.G., Tita, B., 2003. Extracts of various species ofEpilobium inhibit proliferation of human prostate cells. J. Pharm. Pharmacol. 55,683–690.

Yi, W., Wetzstein, H.Y., 2011. Effects of drying and extraction conditions on thebiochemical activity of selected herbs. HortScience 46, 70–73.

Zhang, D., Hamauzu, Y., 2004. Phenolics, ascorbic acid, carotenoids and antioxidantactivity of broccoli and their changes during conventional and microwavecooking. Food Chem. 88, 503–509.

Zheng, W., Wang, S.Y., 2001. Antioxidant activity and phenolic compounds inselected herbs. J. Agric. Food Chem. 49, 5165–5170.





















































































Documents

Evaluation of phytochemical composition of fresh and dried raw material of introduced Chamerion angustifolium L. using chromatographic, spectrophotometric and chemometric techniques