78 ACTA BOT. CROAT. 79 (1), 2020 Acta Bot. Croat. 79 (1), 78–86, 2020 CODEN: ABCRA 25 DOI: 10.37427/botcro-2020-010 ISSN 0365-0588 eISSN 1847-8476 Chemical and morphological diversity among wild populations of Hypericum aviculariifolium Jaub. et Spach subsp. depilatum (Freyn et Bornm.) N. Robson var. depilatum Cuneyt Cirak1*, Aysel Özcan2, Emine Yurteri2, Dursun Kurt1, Fatih Seyis2 1 Ondokuz Mayis University, Vocational High School of Bafra, Samsun, Turkey 2 Recep Tayyip Erdoğan University, Faculty of Agriculture and Natural Sciences, Department of Field Crops, Rize, Turkey Abstract – In this study, the chemical and morphological diversity among eleven wild populations of Hypericum aviculariifolium Jaub. et Spach subsp. depilatum (Freyn et Bornm.) N. Robson var. depilatum, an endemic Turkish species was studied. These populations were investigated for their contents of hypericin, pseudohypericin, hyper- forin, the chlorogenic, neochlorogenic, caffeic and 2,4-dihydroxybenzoic acids, hyperoside, quercitrin, isoquerci- trin, avicularin, 13,118 biapigenin, (+)-catechin and (–)-epicatechin as well as for their morphological traits, in- cluding density of leaf light and dark glands, leaf area, leaf length/width ratio and plant height. The top two-thirds of the plants representing thirty individuals was harvested at full flowering from eleven sites and analyzed for the content of bioactive compounds by high-performance liquid chromatography after being dried at room tempera- ture. Morphological characterization of the wild populations was performed on twenty randomly selected indi- viduals from each plant-growing locality. The content of the tested compounds, except for caffeic acid and avicu- larin, and some morphological traits, namely, the density of leaf translucent glands and black nodules and leaf area varied significantly with the investigated populations. It was observed that hypericin and pseudohypericin contents were connected positively with leaf black nodule density, but negatively with leaf area and the contents of hyperforin, quercitrin and 13,118-biapigenin were correlated positively with leaf translucent gland density. Data presented here could be useful in determining future targets for further wide-ranging studies on this en- demic species as well as in identifying superior germplasm in terms of high chemical content. Keywords: 13,118-biapigenin, chemical and morphological diversity, hyperforin, hypericin, Hypericum aviculariifolium, phenolic acids, pseudohypericin, quercitrin * Corresponding author e-mail: kalinor27@gmail.com Introduction Hypericum genus (Hypericaceae) includes over 400 spe- cies with world-wide distribution and is one of the 100 larg- est genera including twenty two percentage of angiosperm variety (Carine and Christenhusz 2010). Among its mem- bers, only Hypericum perforatum L. has been investigated widely so far. This species, herbal preparations of which have been utilized largely as a medicine in the treatment of mild to moderate depression, especially over the last three de- cades (Ng et al. 2017), is considered officinal. Hypericum spe- cies are well-known medicinal plants and have been used for centuries as traditional healing agents owing to their large number of pharmacological activities. All these species have traditionally been used for sedative, wound healing, disin- fectant and spasmolytic preparations in Turkish folk medi- cine with the local names of “sarıkantaron, askerotu, kılıç otu, kanlıot and kuzu kıran”. Turkey is a centre of great ex- tensity for the Hypericum genus and according to Güner et al. (2012) there are a total of 96 Hypericum taxa in the Turk- ish flora, of which 46 are endemic. Hypericum aviculariifo- lium Jaub. et Spach subsp. depilatum (Freyn et Bornm.) N. Robson var. depilatum [syn. Hypericum origanifolium var. depilatum (Freyn et Bornm.) N. Robson, sensu WFO 2019] is one of these endemic species (Davis 1988), growing wild in arid, stony and limy areas of Northern Turkey. The distri- mailto:kalinor27@gmail.com CHEMICAL DIVERSITY AMONG HYPERICUM AVICULARIIFOLIUM POPULATIONS ACTA BOT. CROAT. 79 (1), 2020 79 bution range of this endemic species is very localized by its exogenously dormant seeds (Cirak et al. 2007a). Its shoots are up to 30 cm in length with yellow inflorescence and typi- cal dark glands on all aerial parts (Fig. 1). Results of recent studies documenting the antibacterial (Gül et al. 2017) and antioxidant (Maltas et al. 2013) properties of H. aviculariifo- lium subsp. depilatum var. depilatum indicate that this en- demic species can be a substitute for widely known H. per- foratum L. Naphthodianthrones, principally represented by hyperi- cin and psudohypericin, the phloroglucinol derivatives ad- hyperforin and hyperforin, flavonoids such as rutin, hypero- side, quercetin and quercitrin, phenolic acids and essential oils with a wide range of bioactivities are considered to be the principal constituents of Hypericum plant taxa (Zhao et al. 2015). In the past, hypericins were indicated as the main chemicals responsible for the antidepressant activity of Hy- pericum extracts; however, recent studies have proved that antidepressant activity is revealed synergistically by both hy- pericins and hyperforin (Nabavi et al. 2018). Hyperforin and its derivatives were also reported to induce antitumor, anti- angiogenic and neuroprotective activities (Ma et al. 2018). Although hyperforin and hypericin have been indicated as providing essential support to the pharmacological activ- ities of Hypericum-derived products, some other ingredi- ents such as chlorogenic acid, quinic acid, hyperoside, rutin, quercitrin, quercetin and amentoflavone were also reported significantly to promote the antidepressant (Tusevski et al. 2018), neuroprotective (Silva et al. 2008), antioxidant and antimicrobial (Zorzetto et al. 2015) activities. A great number of Hypericum species have been subject- ed to studies, documenting their chemical content/compo- sition from Turkish flora as well as other growing localities of the world such as Brazil, Iran, Jordan, Serbia, Italy, Por- tugal, Tunisia, Peru and Lithuania (Cirak et al. 2016, and references therein). Results from the former works revealed significant differ- ences attributed to concentrations of the ingredients among the various Hypericum species from several taxa (Cirak et al. 2016); diversified populations of the same species from various geographic regions (Nogueira et al. 2008), various phenological stages of the same species (Abreu et al. 2004) and between various shoots as well, regenerated from the same in vitro culture (Cellarova et al. 1994). However, the precise pattern of bioactive compound accumulations inside and among members of Hypericum genus is not fully under- stood. It is not explained to what extent the chemical content and composition bear upon specific genotypes within spe- cies. It is also not clarified how far plant geographic origin affects the spectrum of phytochemicals. In our previous works, we reported H. aviculariifolium subsp. depilatum var. depilatum to include hypericins, hyper- forins, various flavonoids and phenolics as hyperoside, quercetine, chlorogenic acid, ru- tin, isoquercetine and quercitrine (Cirak et al. 2007b, 2013). However, population vari- ability of the chemical compounds as well as of morphologic traits has not yet been studied with respect to the endemic species. Hence, in the present work, our intention has been to specify for the first time the regional variability in the content of hypericin, pseu- dohypericin, hyperforin, the chlorogenic, neochlorogenic, caffeic and 2,4-dihydroxy- benzoic acids, hyperoside, quercitrin, iso- quercitrin, avicularin, (+)-catechin, (–)-epi- catechin and 13,118-biapigenin as well as five morphological traits including light and dark gland density on leaves, leaf area, leaf length/width ratio and plant height as well as the correlations between the chemical and morphological data among H. aviculariifo- lium subsp. depilatum var. depilatum popu- lations from eleven localities in the Middle Black Sea geographic region of Northern Turkey. In addition, neochlorogenic, caffeic and 2,4-dihydroxybenzoic acids, 13,118-bi- apigenin, isoquercitrin, avicularin, (+)-cate- chin and (–)-epicatechin were not detected previously in this endemic species. Hereby, we also report the first occurrence of the cor- responding compounds in H. aviculariifoli- um subsp. depilatum var. depilatum. Fig. 1. Hypericum aviculariifolium subsp. depilatum var. depilatum plant flowering in its native habitat (a), and its aerial parts with typical dark glands, namely leaves and stems (b), floral buds (c) and flowers (d, e). Dark and translucent glands on leaf under dissecting microscope (f ). Scale bars = 5 cm (a) and 1 cm (b–f ). CIRAK C, ÖZCAN A, YURTERI E, KURT D, SEYIS F 80 ACTA BOT. CROAT. 79 (1), 2020 Materials and methods Plant materials The plant materials were described in our previous stud- ies (see Cirak et al. 2013, Cirak and Bertoli 2013). The spe- cies were identified by Dr. Samim Kayikci, Mustafa Kemal University, Faculty of Arts and Sciences, Department of Bi- ology, Turkey. Voucher specimens were deposited in the herbarium of Ondokuz Mayis University Vocational High School of Bafra and the numbers of the voucher specimens are given in Tab. 1. Experimental procedures The aerial parts of H. aviculariifolium subsp. depilatum var. depilatum plants exemplify 30 shoots were harvested at flowering stage from eleven localities in Middle Black Sea geographic region of Northern Turkey (Tab. 1). The top two thirds of the plants was reaped between 14:00 pm and 15:00 pm. Conditions on the day of collection were clear and sun- ny at all sites and temperatures varied between 28 and 30 °C. The plant materials were dried at room temperature (20 ± 2 °C), and subsequently analyzed for chemical in gredients by HPLC. Morphological characterization of plants was made, as described previously in our previous study (Cirak et al. 2007b), on 20 randomly selected plants from each grow- ing locality according to plant height, leaf dark and trans- lucent gland density, leaf area, and leaf length/width ratio. Plant height was measured from the flowering crown of the primary stem to the base of the plant. Leaf area, leaf length/ width ratio and the number of dark and light spheroid nod- ules, were measured on 10 leaves of each selected plant from 11 different sites. The number of leaf dark and translucent glands was counted using a dissecting microscope (Fig. 1). For leaf area and leaf length/width ratio calculations, leaves were placed on aphotocopier, held flat and secure and cop- ied onto an A3 sheet (at 1:1 ratio). Placom Digital Planim- eter (Sokkisha Planimeter Inc., Model KP-90) was utilized to measure the actual leaf area of the copy. Leaf width (cm) was measured from tip-to-tip at the widest part of the lamina and leaf length (cm) was measured from lamina tip to the point of petiole intersection along the midrib. Preparation of plant extracts Air-dried plant material was mechanically ground us- ing a laboratory mill to obtain a homogeneous drug pow- der. Samples of about 0.1 g (weighed with 0.0001 g precision) were extracted in 10 mL of 100% methanol by ultrasonica- tion at 40 °C for 60 min in an ultrasonic bath. The prepared extracts were filtered through a membrane filter with a pore size of 0.22 µm (Carl Roth GmbH, Karlsruhe, Germany) and kept in a refrigerator at 4 °C until analysis. The extracts for naphthodianthrones analyses were exposed to light under xenon lamp (765 W/m2) for 8 min for the photoconversion of protohypericins into hypericins. HPLC Analyses and quantification Separation of the flavanoids and phenolic acids tested was carried out by using an RP-18 (5 µm, 250  4.0 mm) col- umn in a Shimadzu LC-2030C-3D HPLC device equipped with a DAD detector. The binary gradient elution method was used for detection of corresponding compounds. The mobile phase A consisted of water acidified with 0.3% phos- phoric acid as eluent A and acetonitrile containing 0.3% phosphoric acid as eluent B. The elution profile was used as following: 0-10 min 10% B, 10-30 min 25% B, 30-38 min 60% B, 38-45 60% B and 45-45.01 min 1% B. Flow rate was 0.6 mL min–1 at 25 °C column temperature. The extract in- jection volume was 10 µL. The calibration of components was obtained at 203 – 280 – 320 – 360 nm wavelengths using 5, 10, 20, 50, 100 and 200 ppm standard solutions. For hypericin, pseudohypericin and hyperforin, the same device, Shimadzu LC-2030C-3D HPLC equipped with a DAD detector, was used. Separation of these chemicals was carried out using an RP-18 (5 µm, 250  4.0 mm) column. The mobile phase of isocratic solution consisted of ethyl ac- etate, aqueous 0.1 M sodium dihydrogen phosphate solution was adjusted to pH 2.0 by using phosphoric acid and meth- anol (39:41:160 v/v). The flow rate was 1 mL min–1 at 40 °C column temperature. The volume of extract injected was 20 µL. The calibration of components was obtained at wave- lengths 207 and 589 nm using 1, 5, 10, 20, 50 and 100 ppm standard solutions. Analytical standards used for HPLC analysis and validation values of the method are shown in On-line Suppl. Tab. 1. The standards are also described in On-line Suppl. Tab. 2. Tab. 1. Geographical data and annual climatic conditions of Hypericum aviculariifolium subsp. depilatum var. depilatum-growing locali- ties from Northern Turkey. BMYO stands for “Bafra Meslek Yüksekokulu”, Vocational High School of Bafra, Turkey; Popul. - population; Latit - latitude; Long - longitude Elev - elevation, T - mean annual temperature; P - mean annual precipitation Popul. Collection date Voucher no. Latit (N) Long (E) Elev (m) T (°C) P (mm) Habitat 1 June 03, 2018 BMYO # 27/1 40° 54΄ 35° 25΄ 1053 08.78 765 Rocky and open slopes 2 June 03, 2018 BMYO # 27/2 40° 54΄ 35° 38΄ 1075 08.52 782 Rocky and open slopes 3 June 03, 2018 BMYO # 27/3 40° 55΄ 35° 25΄ 1293 08.07 821 Rocky and open slopes 4 June 03, 2018 BMYO # 27/4 40° 55΄ 35° 24΄ 1452 07.52 875 Rocky and open slopes 5 June 03, 2018 BMYO # 27/5 40° 50΄ 35° 09΄ 952 09.29 922 Igneous slopes and rock ledges 6 June 04, 2018 BMYO # 27/6 40° 50΄ 35° 10΄ 882 11.53 937 Pinus woodland 7 June 04, 2018 BMYO # 27/7 40° 49΄ 35° 09΄ 989 10.64 932 Arid pasturelands 8 June 04, 2018 BMYO # 27/8 40° 45΄ 35° 08΄ 1243 09.11 856 Stony riverside 9 June 04, 2018 BMYO # 27/9 40° 45΄ 35° 07΄ 1373 08.77 872 Stony riverside 10 June 04, 2018 BMYO # 27/10 40° 45΄ 35° 08΄ 1262 08.92 727 Stony riverside 11 June 04, 2018 BMYO # 27/11 41° 25΄ 36° 58΄ 441 12.64 982 Igneous slopes and rock ledges CHEMICAL DIVERSITY AMONG HYPERICUM AVICULARIIFOLIUM POPULATIONS ACTA BOT. CROAT. 79 (1), 2020 81 hydroxybenzoic acid and quercitrin were accumulated at sig- nificantly higher levels by plants of population-8 (0.58 and 7.10 mg g–1 DM, respectively). The highest accumulation level of hyperoside and isoquercitrin was reached in plants of pop- ulation 3 and population 2 (0.73 and 0.63 mg g–1 DM, respec- tively) (Tab. 2). The present results also indicate that H. avicu- lariifolium subsp. depilatum var. depilatum accumulates lower concentrations of hyperforin, hypericin, psedohypericin, neo- chlorogenic acid, hyperoside, isoquercitrin, (+)-catechin and (-)-epicatechin, comparable concentrations of avicularin and 13,118-biapigenin and higher concentrations of chlorogenic acid, caffeic acid, 2,4-dihydroxybenzoic acid and quercitrin when compared to H. perforatum, a well known commercial source of the compounds examined (Tab. 3). Significant variations (P < 0.01) were also observed in mean values of leaf dark and translucent gland density and leaf area among the investigated populations; however, leaf length/width ratio and plant height did not vary with plant growing localities (Tab. 4). Results of correlation analysis in- dicated an evident connection between leaf dark gland densi- ty/leaf area and hypericin/pseudohypericin contents of plants and leaf translucent gland density and hyperforin, quercitrin and 13,118-biapigenin contents of plants. No significant cor- relation was determined among the rest of the morphologi- cal traits and secondary metabolites tested (On-line Suppl. Tab. 3). The number of translucent glands and black nodules on leaf and leaf area varied considerably with the investigated populations. Leaf dark gland density was significantly high- er in plants of the populations 10, 9 and 8 whose hypericin and pseudoypericin contents were also found to be signifi- cantly higher. In a similar way, population 11, which accu- mulated the highest hyperforin, quercitrin and 13,118-biapi- genin contents, was found to be superior to the others with respect to leaf translucent gland density. Positive and signifi- cant relationships were determined between leaf dark gland density and hypericin (r2 = 0.86, P < 0.01) / pseudohypercin Data Analysis Data of secondary metabolites contents and morphologi- cal characters of plant material were subjected to one-way analysis of variance (ANOVA) and significant differences among mean values were tested with the Duncan Multiple Range Test (P < 0.01). Correlation analysis was performed to clarify the relationship between the chemical and mor- phological data, and principal component analysis (PCA) was carried out to elucidate the relationship of investigated populations regarding the chemical and morphological di- versity they exhibited by using the statistical software pack- age XLSTAT2010 Trial Version. PCA analysis is the two- dimensional visualization of the position of investigated accessions relative to each other. The principal components represent the axes which are the orthogonal projections for the values representing the highest possible variances in the case of PC1 and PC2. The obtained data were used to cre- ate scatter plot diagrams (Backhaus et al. 1989). Therefore, a factor analysis was performed, whereby each variable was used to calculate relationships between variable and inves- tigated factors. Based on the obtained data the cluster den- drogram was created. Results Results of the present study indicate that the contents of hypericin, pseudohypericin, hyperforin, the chlorogenic, ne- ochlorogenic and 2,4-dihydroxybenzoic acids, hyperoside, quercitrin, isoquercitrin, (+)-catechin, (-)-epicatechin and 13,118-biapigenin in plants differed greatly by populations (P < 0.01) whereas caffeic acid and avicularin were accumulated at similar levels in all growing localities. Plants from popula- tion-11 supplied the highest accumulation level of hyperfor- in, chlorogenic acid, neochlorogenic acid, 13,118-biapigenin, (+)-catechin and (–)-epicatechin (1.63, 21.71, 0.85, 2.09, 2.42 and 1.82 mg g–1 DM, respectively) whereas hypericin and pseudohypericin were yielded in the highest level by plants of population-10 (0.58 and 4.89 mg g–1 DM, respectively). 2,4-di- Tab. 2. Mean contents (mg g–1 DM) of different compounds: hypericin (a), pseudohypericin (b), hyperforin (c), chlorogenic acid (d), neochlorogenic acid (e), caffeic acid (f ), 2,4-dihydroxybenzoic acid (g), 13,118-biapigenin (h), hyperoside (i), isoquercitrin (j), querci- trin (k), avicularin (l), (+)-catechin (m), (–)-epicatechin (n) in Hypericum aviculariifolium subsp. depilatum var. depilatum populations (Popul.) located in Northern Turkey. Values are means of three replications and those, followed by different small letters in each column are significantly different (P < 0.01) according to Duncan’s Multiple Range test. Se = standard errors Popul. Compounds a b c d e f g h i j k l m n 1 0.14 d 1.17 d 0.25 c 12.13 b 0.47 c 0.25 0.26 b 1.30 b 0.11 c 0.51 a 2.53 e 0.65 1.38 b 1.05 b 2 0.28 c 1.93 c 0.21 c 9.91 c 0.69 b 0.55 0.42 a 1.46 b 0.24 b 0.63 a 2.97 e 0.65 1.26 b 0.79 c 3 0.39 b 2.10 c 0.07 de 7.64 c 0.45 c 0.26 0.09 c 1.29 b 0.73 a 0.38 b 3.65 d 0.65 1.00 bc 0.69 c 4 0.31 c 3.53 b 0.11 d 2.11 e 0.19 e 0.32 0.14 c 1.09 b 0.01 c 0.15 d 2.16 e 0.64 0.47 c 0.21 d 5 0.27 c 1.67 d 0.15 d 3.45 de 0.27 d 0.23 0.11 c 1.58 b 0.01 c 0.16 d 2.34 e 0.64 0.62 c 0.11 d 6 0.30 c 2.16 c 0.03 e 4.47 d 0.23 de 0.25 0.09 c 1.21 b 0.36 b 0.29 c 5.93 c 0.65 0.68 c 0.19 d 7 0.44 b 2.64 c 0.01 e 4.56 d 0.31 d 0.27 0.09 c 0.98 c 0.01 c 0.24 c 5.00 c 0.64 0.69 c 0.12 d 8 0.53 a 4.16 a 0.27 c 11.81 b 0.38 c 0.24 0.58 a 1.59 b 0.38 b 0.42 b 5.42 c 0.66 1.65 b 1.94 a 9 0.55 a 4.85 a 0.34 c 3.31 d 0.18 e 0.26 0.12 c 2.01 a 0.13 c 0.22 c 3.85 d 0.66 0.68 c 0.21 d 10 0.58 a 4.89 a 0.65 b 8.86 c 0.42 c 0.28 0.29 b 1.97 a 0.44 b 0.41 b 6.57 b 0.66 1.10 b 1.11 b 11 0.31 c 2.16 c 1.63 a 21.71 a 0.85 a 0.27 0.22 b 2.09 a 0.29 b 0.45 b 7.10 a 0.66 2.42 a 1.82 a Se 0.042 0.393 0.140 1.720 0.063 0.027 0.048 0.179 0.068 0.046 0.534 0.002 0.173 0.202 CIRAK C, ÖZCAN A, YURTERI E, KURT D, SEYIS F 82 ACTA BOT. CROAT. 79 (1), 2020 (r2 = 0.92, P < 0.01) contents and leaf translucent gland density and hyperforin (r2 = 0.75, P < 0.05), quercitrin (r2 = 0.71, P < 0.05) and 13,118-biapigenin (r2 = 0.77, P < 0.05) contents. As for the leaf area, the highest and lowest values were detected in population-1 and population-10 (11.38 and 4.82 cm2, re- spectively) yielding the highest and lowest levels of hyperi- cin and pseudohypericin accumulations. Likewise, the popu- lations producing higher amounts of hyperforin, quercitrin and 13,118-biapigenin had lower values of leaf area. Leaf area was found to be negatively correlated with the hypericin (r2 = 0.81, P < 0.01) and pseudohypericin (r2 = 0.86, P < 0.01) con- tents of plant material. Tab. 3. Comparison of the chemical content (mg g–1 DM) in Hypericum aviculariifolium subsp. depilatum var. depilatum (in the present study) and Hypericum perforatum, globally known commercial species of Hypericum genus (compiled from various relevant sources). Compound H. aviculariifolium subsp. depilatum var. depilatum* H. perforatum References Hyperforin 0.21–1.63 8.35–11.50 Greeson et al. 2001, Maggi et al. 2004, Couceiro et al. 2006 Hypericin 0.14–0.58 0.01–2.77 Sirvent et al. 2002, Southwell and Bourke 2001, Bagdonaite et al. 2010 Psedohypericin 1.17–4.89 0.05–6.75 Ayan and Cirak 2008, Bagdonaite et al. 2010, Büter and Büter 2002, Bagdonaite et al. 2012 Chlorogenic acid 2.11–21.71 1.11–2.19 Maggi et al. 2004, Cirak et al. 2007b, c Neochlorogenic acid 0.42–0.85 3.34–4.25 Jürgenliemk and Nahrstedt 2002 Caffeic acid 0.23–0.55 <0.01 Patocka 2003, Nahrstedt and Butterweck 1997 2,4–dihydroxybenzoic acid 0.09–0.58 trace Jürgenliemk and Nahrstedt 2002 Hyperoside 0.01–0.73 2.07–7.69 Maggi et al. 2004, Bagdonaite et al. 2012 Quercitrin 2.16–7.10 0.05–4.77 Martonfi and Repcak 1994, Radusiene et al. 2004 Isoquercitrin 0.22–0.63 3.19–6.99 Jürgenliemk and Nahrstedt 2002 Avicularin 0.64–0.66 0.32–0.96 Wu et al. 2002, Wei et al. 2009 13,118–biapigenin 1.21–2.09 1.78–2.65 Jürgenliemk and Nahrstedt 2002, Nahrstedt and Butterweck 1997 (+)–catechin 0.62–2.42 1.41–8.70 Ploss et al. 2001, Kalogeropoulos et al. 2010 (–)–epicatechin 0.11–1.94 20.6–118.9 Ploss et al. 2001, Kalogeropoulos et al. 2010 *The lowest and highest contents of the corresponding compounds, observed in the present study. Fig. 2. Principal component analysis biplot showing populations (1–11) and vectors of the chemicals and morphological traits based on 20 samples for each population. A two-dimensional (2D) visualization of the relative position of the phytochemicals tested was created by using the values of the principal components (the bioactive com- pounds examined here) relative to the investigated popu- lations. This was provided by utilizing the principal com- ponent analysis (PCA). A biplot was created to see the correlations between samples and the investigated traits. Re- sults of biplot analysis revealed that the investigated popula- tions could clearly be differentiated according to their chem- ical contents and morphological traits. Populations 3, 4, 5, 6, 7 and 9 were different in plant height and populations 1 and 2 were different with regards to caffeic acid content and leaf CHEMICAL DIVERSITY AMONG HYPERICUM AVICULARIIFOLIUM POPULATIONS ACTA BOT. CROAT. 79 (1), 2020 83 area. With the first two principal components, 68.19% of the present variation could be explained (Fig. 2). Further, the investigated populations were differentiated into two main groups namely, group A including popula- tions 1, 2, 3, 4, 5, 6, 7, 9 and group B consisting of popula- tions 8, 10, 11, with respect to their chemical contents and morphological traits on the dendogram, created by biplot analysis. As shown in Fig. 3, populations 1 and 2 and popula- tions 4 and 7 from group A were found to be similar chemi- cally and morphologically. heterogeneity of Hypericum plants from different origins is reported to influence the pharmacological activity of plant extracts significantly and to pose a great risk to the standard- ization of final Hypericum-derived products (Costa et al. 2016). Hence, there have been many investigations regard- ing population variability of bioactive compounds from Hy- pericum species. In H. perforatum L., the most common and commercially recognized species of the genus, wild popula- tions of Turkey (Cirak et al. 2007c), Canada, Australia, Ar- menia and Lithuania (Bagdonaite et al. 2010, and references therein) are shown to yield significantly different amounts of hyperforin, pseudohypericin and hyperforin. Essential oil composition is reported to differ significantly in accordance with the geographic origin of wild accessions of H. pulchrum L., H. humifusum L., H. perfoliatum L. and H. linarifolium Vahl. (Nogueira et al. 2008). Considerable differences were determined in concentra- tions of hypericins, hyperforin and various phenolics such as rutin, hyperoside, amentoflavone and quercetin in the four wild accessions of H. triquetrifolium Turra from Turkey. In a similar way, eleven populations of H. orientale L. and five wild populations of H. montbretii Spach and H. lydium Boiss. are reported to yield different quantities of hypericins, hyperfo- rins, phenolic acids and several flavonoids such as hypero- side, quercetin, amentoflavone, rutin, avicularin isoquerci- trin and quercitrin (Cirak et al. 2015, and references therein). Results of previous studies indicated the geographic ori- gin of plants as a distinct factor influencing the observed chemical variation among wild Hypericum populations. In a similar way, we observed significant differences in accumula- tion levels of 14 bioactive compounds among H. aviculariifo- lium subsp. depilatum var. depilatum from eleven geographic origins in the present work. Two populations of this endemic species are also reported to yield quantitatively and qualita- tively different amounts of essential oil (Cirak and Bertoli 2013). The investigated populations varied with the main environmental factors creating different growing conditions as they were located in different places of Northern Turkey as shown in Table 1. The wide variation observed in accu- mulation levels of the bioactive compounds tested among the populations could somewhat be attributed to adaptive Fig. 3. Dendrogram showing the differentiated groups of Hyperi- cum aviculariifolium subsp. depilatum var. depilatum populations (Group A represents populations 1, 2, 3, 4, 5, 6, 7 and 9; Group B represents populations 8, 10 and 11) regarding their chemical contents and morphological traits. Tab. 4. Mean values of the morphological characters evaluated in Hypericum aviculariifolium subsp. depilatum var. depilatum popula- tions located in Northern Turkey. Values are means of three replications and those, followed by different small letters in each column are significantly different (P < 0.01) according to Duncan’s Multiple Range test. Se = standard errors. Population Dark glands (per mm2) Light glands (per mm2) Leaf area (cm2) Leaf length/width Plant height (cm) 1 0.25 e 5.81 c 11.38 a 2.20 35 2 0.30 e 5.01 c 10.39 b 2.71 40 3 0.32 e 4.20 d 8.33 c 2.27 37 4 0.48 c 4.83 d 6.35 e 2.29 38 5 0.30 d 5.07 c 10.62 b 2.11 36 6 0.48 c 4.34 d 7.99 c 2.36 42 7 0.47 c 3.42 e 6.88 e 2.33 41 8 0.64 b 6.42 b 7.59 d 2.04 35 9 0.66 b 6.51 b 5.67 f 2.36 42 10 0.84 a 6.84 b 4.82 g 2.04 39 11 0.49 c 7.30 a 7.69 d 2.30 34 Se 0.055 0.371 0.631 0.056 0.878 Discussion Among the factors contributing to the variations in the phytochemical accumulation in the Hypericum genus, the geographic origin of plants is of considerable importance as the main environmental factors of a plant-growing habitat such as altitude, temperature, soil etc. influencing synthesis and accumulation of a given bioactive compound were di- verse mainly according to the growing sites. The chemical CIRAK C, ÖZCAN A, YURTERI E, KURT D, SEYIS F 84 ACTA BOT. CROAT. 79 (1), 2020 strategies of wild plants to changing environmental factors. It is also possible to evaluate the observed chemodiversity among populations as a result of genetic distinctness, but da- ta on the connection between the genetic and phytochemical structures in Hypericum spp. is scant and results are typically contradictory. For example, He and Wang (2013) reported only a partial correlation between chlorogenic acid, querce- tin, rutin and hyperoside concentrations and genetic data of 12 wild H. perforatum populations from China. Howev- er, Tonk et al. (2011) detect significant connections between hypericin content and random-amplified polymorphic DNA (RAPD) data for 19 field-grown H. perforatum clones indi- cating the necessity for further chemical and molecular re- searches on the genus Hypericum to differentiate exactly the genetic and environmental effects on the monitored chemi- cal variation among wild populations. The comparison of eleven wild populations of H. avicu- lariifolium subsp. depilatum var. depilatum revealed an intra- specific diversity in the distribution of light and dark glands corresponding with the accumulation of hypericins, hyper- forin, quercitrin and 13,118-biapigenin in the present study. Hypericum plants are categorized generally by three types of secretory structures namely, translucent glands, black or dark nodules and secretory canals (Kimáková et al. 2018). Among Hypericum chemicals, hypericins are reported to accumulate most extensively in the black nodules of aerial parts (Kornfeld et al. 2007) and previous results have proved the localization of hypericin and pseudohypericin in the dark glands of plant aerial parts in all species producing hy- pericins (Kusari et al. 2015, Kuchariková et al. 2016a, b). Be- sides, the absence of dark glands in aerial parts is described as an accurate indication of the absence of hypericins in sev- eral species of Hypericum such as H. brasiliense Choisy, H. caprifoliatum Cham. et Schltdl., H. carinatum Griseb. (Fer- raz et al. 2002), H. androsaemum L., H. kouytchense Levl., H. monogynum L., H. stellatum N. Robson, and H. canar- iense L. (Kuchariková et al. 2016a). In previous researches, a close relationship is observed between dark gland number of leaf and total hypericin content of plants in H. perfora- tum (Southwell and Campbell, 1991) and H. lydium (Cirak 2006). In accordance with the previous results, we observed a positive and significant relationship with high r2 values be- tween leaf dark gland density and the plant content of hy- pericin and pseudohypercin in the present work. We also detected that plant content of hypericins is negatively cor- related with leaf area, as reported by Cirak et al. (2007c) for H. perforatum. It may be speculated that in the enlargement of leaf area concludes in a decrease of dark gland density and that the inverse connection between leaf area and the content of hypericins might be attributed to this decrease. As for hyperforin, Soelberg et al. (2007) report that concen- trations of the bioactive compound in translucent glands of leaves surpassed that of original leaves by more than 100% in H. perforatum. Based on these results, the authors indicate translucent glands as the main site of hyperforin accumula- tion as confirmed latterly by Kusari et al. (2015). Hyperforin, besides, is reported to accumulate primarily in translucent glands of leaves in H. stellatum, H. annulatum Moris, H. an- drosaemum, H. kouytchense, H. monogynum, H. kalmianum L., H. balearicum L. and H. canariense (Kuchariková et al. 2016b). However, no attempt has been undertaken so far to investigate the correlation between number of translucent glands and content of hyperforin. We report here for the first time the significant and positive relationship between leaf translucent gland number and hyperforin accumulation lev- els which can be useful to explain sites of hyperforin synthe- sis and function of the bioactive compound within the genus Hypericum. As opposed to hypericins and hyperforin, it may not be feasible to ascertain a pattern of localization for flavo- noids as data on the localization of them on the aerial parts are discrepant and vary with species. Hyperoside, isoquer- citrin, quercetin and quercitrin were accumulated mainly in leaf dark glands of H. olympicum L., H. perforatum, and H. rumeliacum Boiss. and rutin was accumulated only in leaf black nodules of H. maculatum Crantz and H. erectum Thunb. However quercetin, the most prevalent flavonoid is reported to localize mainly in leaf translucent glands of H. rumeliacum (Kusari et al. 2015, Kuchariková et al. 2016a). Hyperoside and isoquercitrin, interestingly, are also report- ed to accumulate mainly in leaf translucent glands in H. kal- mianum (Kuchariková et al. 2016a). By contrast, hyperoside, isoquercitrin and quercitrin are accumulated primarily in both dark and translucent glands in H. humifusum leaves and quercetin and quercitin were accumulated principally in translucent glands and non-secretory structures in leaves of the species, namely, H. androsaemum, H. kouytchense, H. monogynum, H. stellatum (Kuchariková et al. 2016a). In the present study, we detected a positive and significant connec- tion between leaf translucent gland density and the content of quercitrin and 13,118-biapigenin indicating translucent glands as the main site for the accumulation of correspond- ing compounds in H. aviculariifolium subsp. depilatum var. depilatum. Principal component and cluster analyses are favored means for characterization of genotypes and their grouping on similarity. PCA is a beneficial statistical tool for the differ- entiation of plant materials, giving information on the varia- tion in chemical content/composition of several species. A combination of the two statistical tools provides broad in- formation of the traits making significant contributions to genetic diversity in crops. Biplot is another widely utilized procedure for graphical display of accession groups with the aim of searching for the relationships among agro-morpho- logical characters in several cultivars (Malik et al. 2014). In the present study, we used the above mentioned statistical tools to evaluate the chemical and morphological data of eleven H. aviculariifolium subsp. depilatum var. depilatum populations from Turkish flora. In conclusion, chemical and morphologic characteriza- tions of wild plant populations seem to be first step to de- fine superior germplasm and to provide improved chemi- cal profiles. Results of the present study indicate substantial correlations between hypericin, pseudohypericin, hyperfo- rin, quercitrin and 13,118-biapigenin contents and densi- CHEMICAL DIVERSITY AMONG HYPERICUM AVICULARIIFOLIUM POPULATIONS ACTA BOT. CROAT. 79 (1), 2020 85 ty of leaf dark and translucent glands and leaf area. Thus, these morphological characters could be utilized as selec- tion criteria in identifying germplasm that is superior with regard to high contents of the corresponding compounds. Data presented here could also be useful in determining the forthcoming goals for further wide-ranging studies on this endemic species as well as enriching data of current litera- ture on Hypericum chemistry. References Abreu, I.N., Porto, A.L.M., Marsaioli, A.J. Mazzafera. P., 2004: Distrubution of bioactive substances from Hypericum brasil- iense during plant growth. Plant Science 167, 949–954. Ayan, A.K., Cirak, C., 2008: Hypericin and pseudohypericin con- tents in some Hypericum species growing in Turkey. Pharma- ceutical Biology 46, 288–291. Backhaus, K., Erichson, B., Plinke, W., Weiber, R., 1989: Multivari- ate analysis methods. Springer Verlag, Heidelberg. Bagdonaite, E., Martonfi, P., Repcak, M., Labokas, J., 2010: Varia- tion in the contents of pseudohypericin and hypericin in Hy- pericum perforatum from Lithuania. Biochemical Systematics and Ecology 38, 634–640. Bagdonaite, E., Martonfi, P., Repcak, M., Labokas, J., 2012: Varia- tion in concentrations of major bioactive compounds in Hy- pericum perforatum L. from Lithuania. Industrial Crops and Products 35, 302–308. Büter, K.B., Büter, B., 2002: Ontogenetic variation regarding hy- pericin and hyperforin levels in four accessions of Hypericum perforatum L. Journal of Herbs Spices and Medicinal Plants 9, 95–100. Carine, M.A., Christenhusz, M.J.M., 2010: About this volume, the monograph of Hypericum by Norman Robson. Phyto- taxa 4, 1–4. Cellarova, E., Daxnerova, Z., Kimáková, K., Haluskova, J., 1994: The variability of hypericin content in the regenerants of Hy- pericum perforatum. Acta Biotechnologica 14, 267–274. Cirak, C., 2006: Hypericin in Hypericum lydium Boiss. growing in Turkey. Biochemical Systematics and Ecology 34, 897–899. Cirak, C., Bertoli, A., 2013: Aromatic profiling of wild and rare species growing in Turkey, Hypericum aviculariifolium Jaub. and Spach subsp. depilatum (Freyn and Bornm.: Robson var. depilatum and Hypericum pruinatum Boiss. and Bal. Natural Product Research 27, 100–107. Cirak, C., Kevseroğlu, K., Ayan, A.K., 2007a: Breaking of seed dormancy in a Turkish endemic Hypericum species: Hyperi- cum aviculariifolium subsp. depilatum var. depilatum by light and some pre-soaking treatments. Journal of Arid Environ- ments 68, 159–164. Cırak, C., Radusiene, J., Janulis, V., Ivanauskas, L., 2007b: Chemi- cal constituents of some Hypericum species growing in Tur- key. Journal of Plant Biology 50, 632–635. Cirak, C., Radusiene, J., Sağlam, B., Janulis, V., 2007c: Variation of bioactive substances and morphological traits in Hypericum perforatum populations from Northern Turkey. Biochemical Systematics and Ecology 35, 403–409. Cirak, C., Radusiene, J., Camas, N., Çalışkan, Ö., Odabas, M.S., 2013: Changes in the contents of main secondary metabolites in two Turkish Hypericum species during plant development. Pharmaceutical Biology 51, 391–399. Cirak, C., Radusiene, J., Ivanauskas, L., Jakstas, V., Camas, N., 2015: Changes in the content of bioactive substances among Hypericum montbretii Spach (Hypericaceae: populations from Turkey. Brazilian Journal of Pharmacognosy 24, 20–24. Cirak, C., Radusiene, J., Jakstas, V., Ivanauskas, L., Seyis, F., Yayla, F., 2016: Secondary metabolites of seven Hypericum species growing in Turkey. Pharmaceutical Biology 54, 2244–2253. Costa, J., Campos, B., Amaral, J.S., Nunes, M.E., Oliveira, B.P.P., Isabella, M., 2016: HRM analysis targeting ITS1 and matK loci as potential DNA mini-barcodes for the authentication of Hy- pericum perforatum and Hypericum androsaemum in herbal infusions. Food Control 61, 105–114. Couceiro, M.A., Afreen, F., Zobayed, S.M.A., Kozai, T., 2006: Variation in concentrations of major bioactive compounds of St. John's wort: Effects of harvesting time, temperature and germplasm. Plant Science 170, 128–134. Davis, P.H., 1988: Flora of Turkey and the East Aegean Islands. Edinburgh University Press, Edinburgh. Ferraz, A., Bordignon, S., Mans, D., Aline Schmitt, A., Ravazzolo, A.P., von Poser, G.L. 2002: Screening for the presence of hy- pericins in southern Brazilian species of Hypericum (Guttif- erae). Pharmaceutical Biology 40, 294–297. Greeson, JM., Sanford, B., Monti, D.A., 2001: St. John’s wort (Hy- pericum perforatum L.), a review of the current pharmaco- logical, toxicological and clinical literature. Psychopharma- cology 153, 402–414. Güner, A., Aslan, S., Ekim, T., Vural, M., Babaç, M.T., 2012: List of Turkish flora (Vascular plants). Publication of Nezahat Gökyiğit Botanical Garden and Flora Research Fondation, İstanbul. Retrieved December 23, 2019 from http://www.bi- zimbitkiler.org.tr/v2/hiyerarsi.php?c=Hypericum Gül, L.B., Özdemir, N., Gül, O., Cirak, C., Çon, A.H., 2017: Bio- active properties and antimicrobial activity of some Hyperi- cum species growing wild in Black Sea region of Turkey. In: Turker, S., Sahin, A.S., Ertekin, M., Soner, B.C., Unver, A., Dogu, S., Akyurek, H.E., Kozan, I. (eds.), Book of Abstract of the 1st International Medicinal and Aromatic Plants Con- gress, 1336–1336. Konya, Turkey: Necmettin Erbakan Uni- versity Publications. He, M., Wang, Z., 2013: Genetic diversity of Hypericum perfo- ratum collected from the Qinling Mountains of China. Bio- chemical Systematics and Ecology 50, 232–239. Jürgenliemk, G., Nahrstedt, A., 2002: Phenolic Compounds from Hypericum perforatum. Planta Medica 68, 88–91. Kalogeropoulos, N., Yannakopoulou, K., Gioxari, A., Chiou, A., Makris, D. P., 2010: Polyphenol characterization and encap- sulation in β-cyclodextrin of a flavonoid-rich Hypericum per- foratum (St. John's wort: extract, LWT). Food Science and Technology 43, 882–889. Kimáková, K, Petijová L, Bruňáková K, Čellárová E., 2018. Rela- tion between hypericin content and morphometric leaf pa- rameters in Hypericum spp.: A case of cubic degree polyno- mial function. Plant Science 271, 94–99. Kornfeld, A., Kaufman, P.B., Lu, C.R., Gibson, D.M., Bolling, S.F., Warber, S.L., Chang, S.C., Kirakosyan, A., 2007: The produc- tion of hypericins in two selected Hypericum perforatum shoot cultures is related to differences in black gland struc- ture. Plant Physiology and Biochemistry 45, 24–32. Kuchariková, A., Kusari, S., Sezgin, S., Spiteller, M., Cellárová, E., 2016a: Occurrence and distribution of phytochemicals in the leaves of 17 in vitro cultured Hypericum spp. adapted to outdoor conditions. Frontiers in Plant Science 7, 1616–15. Kuchariková, A., Kimáková, K., Janfelt, C., Cellárová, E., 2016b: Interspecific variation in localization of hypericins and phlo- roglucinols in the genus Hypericum as revealed by desorption electrospray ionization mass spectrometry imaging. Physio- logia Plantarum 157, 2–12. CIRAK C, ÖZCAN A, YURTERI E, KURT D, SEYIS F 86 ACTA BOT. CROAT. 79 (1), 2020 Kusari, S., Sezgin, S., Nigutova, K., Cellarova, E., Spiteller, M., 2015: Spatial chemo-profiling of hypericin and related phyto- chemicals in Hypericum species using MALDI-HRMS imag- ing. Analytical and Bioanalytical Chemistry 407, 4779–4791. Ma, L., Pan, X., Zhou, F., Liu, K., Wang, L., 2018: Hyperforin pro- tects against acute cerebral ischemic injury through inhibi- tion of interleukin-17A-mediated microglial activation. Brain Research 1678, 254–261. Maggi, F., Ferretti, G., Pocceschi, N., Menghini, L., Ricciutelli, M., 2004: Morphological, histochemical and phytochemical in- vestigation of the genus Hypericum of the Central Italy. Fito- terapia 75, 702–711. Malik, R., Sharma, H., Sharma, I., Kundu, S., Verma, A., Sheoran, S., Kumar, R., Chatrath, R., 2014: Genetic diversity of agro- morphological characters in Indian wheat varieties using GT biplot. Australian Journal of Crop Science 8, 1266–1271. Maltas, E., Uysal, A., Yildiztugay, E., Aladag, M.O., Yildiz, S., Kucukoduk, M., 2013: Investigation of antioxidant and anti- bacterial activities of some Hypericum species. Fresenius En- vironmental Bulletin 22, 862–869. Martonfi, P., Repcak, M., 1994: Secondary metabolites during flower ontogenesis of Hypericum perforatum L. Zahradnic- tvi 21, 37–44. Nabavi, S.M., Nabavi, S.F., Sureda, A., Caprioli, G., Lannarelli, R., Sokeng, A.J.T., Braidy, N., Khanjani, S., Moghaddam, A.H., Atanasov, A.G., Daglia, M., Maggi, F., 2018. The water ex- tract of tutsan (Hypericum androsaemum L.) red berries ex- erts antidepressive-like effects and in vivo antioxidant activity in a mouse model of post-stroke depression. Biomedicine & Pharmacotherapy 99, 290–298. Nahrstedt, A., Butterweck, V., 1997: Biologically active and other chemical constituents of the herb of Hypericum perforatum L. Pharmacopsychiatry 30, 129–134. Ng, Q.X., Venkatanarayanan, N., Ho, C.Y.X., 2017: Clinical use of Hypericum perforatum (St. John's wort: in depression: A meta-analysis. Journal of Affective Disorders 210, 211–221. Nogueira, T.M.J., Marcelo-Curto, A., Figueiredo, C., Barroso, J.G., Pedro, L.G., Rubiolo, P., Bicchi, C., 2008: Chemotaxonomy of Hypericum genus from Portugal, geographical distribution and essential oils composition of Hypericum perfoliatum, Hy- pericum humifusum, Hypericum linarifolium and Hypericum pulchrum. Biochemical Systematics and Ecology 36, 40–50. Patocka, J., 2003: The chemistry, pharmacology, and toxicology of the biologically active constituents of the herb Hypericum perforatum L. Journal of Applied Biomedicine 1, 61–70. Ploss, O., Petereit, F., Nahrstedt, A., 2001: Procyanidins from the herb of Hypericum perforatum. Pharmazie 56, 509–511. Radusiene, J., Bagdonaite, E., Kazlauskas, S., 2004: Morphological and chemical evaluation on Hypericum perforatum and H. mac- ulatum in Lithuania. Acta Horticulturae (ISHS) 629, 55–62. Silva, B., Oliveira, P.J., Dias, A., Malva, J.O., 2008: Quercetin, kaempferol and biapigenin from Hypericum perforatum are neuroprotective against excitotoxic insults. Neurotoxicity Re- search 13, 265–279. Sirvent, T., Walker, L., Vance, N., Donna, G., 2002: Variation in hy- pericins from wild populations of Hypericum perforatum L. in the Pasific Northwest of the U.S.A. Economic Botany 56, 41–49. Soelberg, J., Jorgensen, L.B., Jager, A.K., 2007: Hyperforin accu- mulates in the traslucent glands of Hypericum perforatum. Annals of Botany 99, 1097–1100. Southwell, I.A., Bourke, C.A., 2001: Seasonal variation in hyperi- cin content of Hypericum perforatum L. (St. John’s wort). Phy- tochemistry 56, 437–441. Southwell, I.A., Campbell, M.H., 1991: Hypericin content varia- tion in Hypericum perforatum L. in Australia. Phytochemis- try 30, 475–478. Tonk, F.A., Giachino, R.R.A., Sönmez, Ç., Yüce, S., Bayram, E., Telci, İ., Furan, M.A., 2011: Characterization of various Hy- pericum perforatum clones by hypericin and RAPD analy- ses. International Journal of Agricultural Biology 13, 31–37. Tusevski, O., Krstikj, M., Stanoeva, J.P., Stefova, M., Simic, S.G., 2018: Phenolic profile and biological activity of Hypericum perforatum L., Can roots be considered as a new source of natural com- pounds? South African Journal of Botany 117, 301–310. Wei, Y., Xie, Q., Donga, W., Ito, Y., 2009: Separation of epigallo- catechin and flavonoids from Hypericum perforatum L. by- high-speed counter-current chromatography and prepar- ativehigh-performance liquid chromatography. Journal of Chromatography A 1216, 4313–4318. WFO 2019: Hypericum origanifolium var. depilatum (Freyn et Bornm.) N. Robson. Retreived December 27, 2019 from http://www.worldfloraonline.org/taxon/wfo-0000746600. Wu, Y., Zhou, S.D., Li, P., 2002: Determination of flavonoids in Hypericum perforatum by HPLC analysis. Yao Xue Xue Bao 2000: 37, 280–282. Zhao, J., Liu, W., Wang, J., 2015: Recent advances regarding con- stituents and bioactivities of plants from the genus Hyperi- cum. Chemistry & Biodiversity 12, 309–349. Zorzetto, C., SAnchez-Mateo, C.C., Rabanal, R.M., Lupidi, G., Petrelli, D., Vitali, L.A., Bramucci, M., Quassinti, L., Caprioli, G., Papa, F., Ricciutelli, M., Sagratini, G., Vittori, S., Maggi, F., 2015: Phytochemical analysis and in vitro biological activity of three Hypericum species from the Canary Islands (Hyperi- cum reflexum, Hypericum canariense and Hypericum grandi- folium. Fitoterapia 100, 95–109. http://www.worldfloraonline.org/taxon/wfo-0000746600