2352 Biotechnol. & Biotechnol. eq. 25/2011/2

Article DOi: 10.5504/bbeq.2011.0040 b&e

Biotechnol. & Biotechnol. eq. 2011, 25(2), 2352-2356Keywords: Tribulus terrestris, South Bulgaria, steroidal saponins, hPlc

IntroductionTribulus terrestris (Zygophyllaceae) (Fig. 1) is a small annual herb that grows worldwide, especially in the Mediterranean region and the subtropical area. it is used in the folk medicine in Bulgaria, china, india and other countries against sexual impotency, oedemas, abdominal distension and cardiovascular diseases. it is also included in many dietary supplements claimed to have a biostimulating activity (2). Several studies have shown that steroidal saponins are among the compounds responsible for the biological activities of T. terrestris (1, 3, 4, 5).

Fig. 1. Tribulus terrestris in flowers. Struma Valley

this study aimed to estimate the variability of the main biologically active compounds protodioscin, prototribestin,

dioscin and rutin in Tribulus terrestris populations in four floristic regions in South Bulgaria. Cluster analysis was performed to find out the relationship between the chemical profile and the ecological conditions.

Materials and MethodsPlant materialthe literature information and our practice in the previous projects allowed to select an easy and effective approach related with the transect method, in this case oriented along the ecological gradient “type of soil- degree of humidity”. this approach gave the chance to eliminate the nonproductive areas and to cover a larger territory with a high degree of approximation. According to us the statistics “gaps” can be accepted about 5%, including the statistic error. the following methodological steps were made:

• collection of the information from the SoM, SoA, So herbariums;

• Allocation and indication of the localities with their GPS coordinates;

• collection of the samples in the phase of mass blossoming and first fruits;

• calculation of the resource in three control measurements per 1 m2;

• extrapolation of the collected data to the territory of the given locality;

• Measurements of the rate fresh:dry weight (3.5:1);• the projective coverage was taken as a corrective value;• the materials collected were stored in a cool place with

good aeration;• the materials were stored in paper bags after aerial

drying;

INTRASPECIFIC VARIABILITY OF BIOLOGICALLY ACTIVE COMPOUNDS OF DIFFERENT POPULATIONS OF TRIBULUS TERRESTRIS L. (ZYGOPHYLLACEAE) IN SOUTH BULGARIA

irina lazarova1, Antoaneta ivanova1, Pepa Mechkarova1, Dimitar Peev2, Natalia Valyovska2

1Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences, Sofia, Bulgaria2Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences, Sofia, Bulgariacorrespondence to: irina lazarovae-mail: airi_airi@abv.bg

ABSTRACTThe biologically active compounds protodioscin, prototribescin, dioscin and rutin were investigated in the samples originating from 25 natural populations distributed in South Bulgaria, including four floristic regions. The samples were collected along the ecological cline - “type of soil – degree of humidity” and analyzed by an HPLC method. The variability of the investigated compounds was evaluated and cluster analysis results were interpreted in order to discover intraspecific variability and eventual correlation with some ecological conditions.

2353Biotechnol. & Biotechnol. eq. 25/2011/2

• the localities were indicated on the map of Bulgaria (scale 1:500 000) with UtM grid, 10 km2);

• the voucher specimens were deposited in SoM (6).

Analytical methodAn hPlc system la chrom elite consisting of l-2130 pump equipped with a gradient controller and UV detector L-2400 was used. the separation was performed on 250 x 4.6 mm i. d., 5 µm, inertsil oDS-2 column (tokyo, Japan) with MetaGuard Pursuit direct connect guard column from Varian was used for all separations. the mobile phase which consisted of phosphoric acid buffer with ph-3 (A) and acetonitrile (B) was used for gradient elution. The flow rate was adjusted to 1.0 ml/min, the detection wavelength was at 203 nm. All separations were performed at ambient temperature.

Sample preparationThe finely powdered plant material (1g, leaves and fruits, 1:1) was extracted three times with 5.0 ml of 50% aqueous acetonitrile by sonication for 15 min. the extracts were combined, after filtration in 20.00 ml volumetric flask and the volume was adjusted to 20 ml with the solvent used for extraction. Prior to injection, all samples were filtered through a 0.45-mm Chromafil 0-45/25 Machery-Nagel. every sample solution was injected in triplicate with an injection volume of 20ml.

Cluster analysesAn unsupervised technique was applied, cluster analysis, to discover similarities within the obtained concentration data of the main compounds – protodioscin, prototribestin, dioscin and rutin. Squared euclidean distance was always used as the interval measure for clustering using distinct linkage methods: between-groups linkage, within-groups linkage and Ward’s method. All statistical analyses and graphical display were done with STATISTICA 7 for Windows (StatSoft, Inc., Tulsa, OK, USA).

Results and DiscussionQuantitative specificsthe comparative investigation of the main biologically active components of Tribulus terrestris from four floristic regions in South Bulgaria was carried out and the results were presented in Fig. 2, Table 1 and Table 2.

the content of some other 13 samples was established to vary from 1.3 mg/g to 1.96 mg/g. (Table 2, Fig. 2).

the analysis revealed no clear geographical and ecological dependence of the obtained results. For example: the comparison between the samples from the mountain (Rila Monastery, the town of Krumovgrad) or from the Black Sea coast (the town of Sozopol, the town of Pomorie) were not informative.

the highest content of protodioscin (more than 13.3 mg/g) was established in the Porominovo (13.42 mg/g) and Maglizh (13.30 mg/g) samples.

Fig. 2. location of the studied samples from T. terrestris

TABLE 1Floristic regions in Bulgaria

№ Floristic region1* Black Sea coast2 north-eastern Bulgaria3 Danube Plain4 Balkan foothill region5 Balkan Mountains (eastern, central, Western)6 Sofia Region7 Znepole Region8 Vitosha Region9 West frontier mountains10* Struma Valley11 Belasitsa Mountain12 Slavyanka Mountain13 Mesta Valley14 Pirin Mountain (Southern, northern)15 Rila Mountain16 Sredna Gora Mountain (eastern, Western)17* Rhodopa Mountains (Western, central, eastern)18 thracian Plain19* tundzha hilly region20 Strandzha Mountain

*collection of the investigated samples

A lower content (less than 7.00 mg/g) was established in the following samples from: ovchi Kladenets (3.79 mg/g); Rupite (3.49 mg/g); Zadgorsko (5.51 mg/g); Acheloy (6.29 mg/g). the content in the other 17 samples varied from 7.15 mg/g to 10.18 mg/g dry weight.

the highest values were estimated in two different geographical and ecological regions (the Black Sea coast and Struma valley).

2354 Biotechnol. & Biotechnol. eq. 25/2011/2

Fig. 3. Chemical profile of biologically active compounds in Tribulus terrestris from the Black Sea coast floristic region

Fig. 4. Chemical profile of biologically active compounds in Tribulus terrestris from the Struma Valley floristic region

TABLE 2comparative investigation of the main components of Tribulus terrestris in four floristic regions in South Bulgaria- origins and data

№ Locality Rutinmg/g DW Protodioscin Prototribestin Dioscin

Black Sea coast (01)a

09.1.1 tsarevo 1.40±0.00 15.28±0.09 8.65±0.03 < loDb

09.1.2 Kavatsi 0.70±0.01 13.42±0.30 8.10±0.13 < loD09.1.3 Kaya 1.81±0.08 11.96±0.16 4.65±0.06 < loD09.1.4 Sozopol 0.95±0.01 7.18±0.22 5.63±0.08 4.35±0.1709.1.5 Aheloy 1.60±0.15 6.29±0.11 4.55±0.38 < loD09.1.6 Pomorie 2.35±0.02 8.26±0.15 7.96±0.03 < loD

Struma Valley (10)09.10.1 Petrich 2.28±0.05 10.47±0.33 11.07±0.24 < loD09.10.2 Sandanski 2.22±0.02 8.47±0.22 5.06±0.16 < loD09.10.3 Rupite 1.94±0.06 4.39±0.27 2.31±0.11 < loD09.10.4 Drangovo 1.87±0.04 9.88±0.08 6.02±0.01 2.33±0.0109.10.5 Rila Monastery bifurcation 2.51±0.00 14.90±0.02 10.59±0.00 < loD09.10.6 Porominovo 0.62±0.00 13.42±0.14 11.51±0.30 < loD09.10.7 Bumar 2.21±0.06 12.47±0.25 8.44±0.16 1.19±0.0109.10.8 Kopilovtsi 2.24±0.04 6.30±0.18 4.11±0.07 1.51±0.01

East Rhodopa Mountains (17)09.17.1 Dzhebel 0.92±0.00 8.72±0.02 11.37±0.08 1.71±0.0209.17.2 Momchilgrad 0.71±0.00 8.16±0.04 6.99±0.12 1.34±0.0009.17.3 Zagorsko 1.96±0.02 5.51±0.00 6.77±0.03 < loD09.17.4 Krumovgrad 0.30±0.00 7.15±0.01 6.17±0.01 1.41±0.0209.17.5 Kardzhali 0.87±0.00 10.78±0.05 11.63±0.03 1.14±0.00

Tundzha hilly region (19)09.19.1 Pirne 0.92±0.00 4.75±0.03 3.01±0.00 < loD09.19.2 Maglizh 0.42±0.00 13.30±0.14 9.25±0.11 < loD09.19.3 Vetren 1.66±0.02 6.97±0.04 4.51±0.07 < loD09.19.4 ovchi Kladenets 1.31±0.00 3.79±0.07 2.52±0.01 1.52±0.0009.19.5 Bolyarovo 2.41±0.03 8.39±0.17 7.62±0.22 1.98±0.0009.19.6 Balgarovo 1.96±0.03 9.74±0.06 5.81±0.13 < loD

2355Biotechnol. & Biotechnol. eq. 25/2011/2

the only similarity of the regions was related to the highest annual temperature.

there were not any geographical and/or ecological correlations in the rest of the samples (Table 2, Fig. 2, Fig. 7 and Fig. 8).

the highest content (more than 9 mg/g) of prototribescin was observed in the samples from the following locations: Kardzhali (11.63 mg/g); Dzhebel (11.37 mg/g); Petrich (11.07 mg/g); the bifurcation for the Rila Monastery (10.59 mg/g); Maglizh (9.25 mg/g).

A lower content (less than 5.25 mg/g) was estimated in the samples from the following locations: Rupite 1 (2.31 mg/g); ovchi Kladenets (2.52 mg/g); Pirne (3.01 mg/g); Kopilovtsi (4.11 mg/g); Acheloy 1 (4.55 mg/g), cape Kaya (4.65 mg/g). the content in the other 12 samples varied from 5.81 mg/g (Balgarovo) to 8.65 mg/g dry weight (tsarevo). the chemical investigation of prototribescin showed that there were not any geographical and/or ecological correlations (Table 2, Fig. 2, Fig. 7 and Fig. 8).

Fig. 5. Chemical profile of biologically active compounds in Tribulus terrestris from East Rhodopa Mountains floristic regiona - floristic regionb - limit of detection

only 10 samples contained dioscin. the highest content was measured in the sample from Sozopol (4.35 mg/g). A lower presence of the compound was observed in the samples from Kopilovtsi (1.51 mg/g), ovchi Kladenets (1.52 mg/g); Bolyarovo (1.98 mg/g); and Drangovo (2.33 mg/g).

the quantity of dioscin was less than the limit of detection in the other 15 samples under investigation. there were not geographical or ecological correlations in the samples containing dioscin.

there was a correlation between the small amounts of dioscin and comparatively high content of protodioscin, which depends on their biotransformation into one another (Table 2, Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 8).

the overall synthetic activity relevant to the four biologically active compounds could be presented as follows: the highest productivity was established in the samples from the Black Sea coast, the samples from the Struma Valley showed about 5% lower values and the eastern Rhodopes samples had 20% less active contents.

Fig. 6. Chemical profile of biologically active compounds in Tribulus terrestris from the Tundzha hilly floristic region

Fig. 7. Chemical profile of protodioscin and prototribestin in Tribulus terrestris in four floristic regions in South Bulgaria

Fig. 8. Chemical profile of protodioscin and dioscin in Tribulus terrestris in four floristic regions in South Bulgaria

the lowest content (about 31%) was established in the samples from the tundzha hilly region (Table 2, Fig. 7, Fig. 8).

the highest productivity of particular population fragments was found in the samples from the habitats of tsarevo, Porominovo, Kavatsi, Kardzhali, Bumar and Vetren.

the tsarevo and Kavatsi samples belong to the Black Sea coast region. the rest of the samples were gathered from different floristic regions in Bulgaria.

Although the analysis of compound distribution did not show a clearly expressed geographical or other ecological dependence, it should be noted that the general conditions in the South-eastern region of Bulgaria along the line Acheloy-Pomorie-

2356 Biotechnol. & Biotechnol. eq. 25/2011/2

Sozopol-Kavatsi-tsarevo are most suitable for biosynthesing of protodioscin, prototribestin, dioscin and rutin.

the graphical expression of the numerical data illustrates that the differences between the samples from different floristic regions dominated over the random distribution of the results from a single region. it was clearly shown that the relative high content of protodioscin was accompanied by a very low biosynthesis of dioscin (Fig. 8).

Cluster analysisthe highest degree of similarity was established between

the samples from the localities of Rupite-Pirne-ovchi Kladenets, Acheloy-Vetren-Kopilovtsi, Momchilgrad-Krumovgrad, Kavatsi-Maglizh.

these samples were situated about 2 units from the line of full similarity, while the criteria for real similarity requires this distance to be less than 0.5 units. in this respect they showed only a low degree of similarity.

The relative deficiency of similarity was confirmed by the position of the other cluster elements- from 12 to 49 unit distance (Fig. 9).

Tree Diagram for 25 CasesWard`s method

Euclidean distances

0 10 20 30 40 50Linkage Distance

Ovchi kladenetsPirne

RupiteZagorsko

KopilovtsiVetrenAheloy

KrumovgradMomchilgrad

BolyarovoPomorieSozopol

DrangovoBalgarovo

SandanskiKaya

DzhebelKardzhali

PetrichBumar

MaglizhKavatsi

PorominovoRilski monastery

Tsarevo

Fig. 9. Dendrogram obtained with the Ward’s method of linkage. it shows the distribution of biologically active compounds in Tribulus terrestris from 25 localities of four floristic regions in South Bulgaria

in spite of the relatively low degree of similarity the following clusters could be described:

I. tsarevo-Rila Monastery-Porominovo-Kavatsi-Maglizh-Bumar-Petrich-Kardzhali-Dzhebel with 25 units of general distance. The samples in this cluster belonged to 4 floristic regions in Bulgaria.

II. cape Kaya-Sandansky-Balgarovo-Drangovo-Sozopol-Pomorie-Bolyarovo-Momchilgrad-Krumovgrad with 15 units of general distance.

the samples in this cluster were closer in respect of the geographical distribution, forming two sub-clusters for the Black Sea coast and eastern Rhodopes scattered with samples of other origins.

III. Acheloy-Vetren-Kopilovtsi-Zadgorsko-Rupite-Pirne-ovchi Kladenets with about 16-unit distance.

The samples in this cluster belonged to different floristic regions and ecological conditions.

IV. the second and third clusters were rated at a similarity value of about 17 units and there was no similarity with cluster (I) because of the distance of 49 units from the line of full similarity (Fig. 9).

ConclusionsThe investigated samples expressed significant quantitative and qualitative amplitudes varying from full absence of some compounds (dioscin) to differences in concentration four times more than the others (prototribestin).

these differences could not be clearly correlated with geographical or local ecological conditions as it is illustrated graphically in Fig. 7 and Fig. 8 and the cluster diagram in Fig. 9.

According to the theoretical data the species Tribulus terrestris forms panmict populations due to its entomophily. it was shown that there was a high similarity between particular populations or fragments and there were some elements of geographically or ecologically oriented cline. the very low degree of similarity (8 to 49 units) of the samples taken along the Black Sea coast indicated the existence of geographically oriented variability in the direction north-South (Pomorie to tsarevo).

no clear tendency could be seen in the cluster of the eastern Rhodopes as samples taken at a distance of about 10-20 km from each other (Momchilgrad-Krumovgrad, Dzhebel-Zagorsko) had a low similarity value (about 7 units) in contrast to the real possibility for entomophilic crosspollination.

in the qualitative and quantitative differences between separate samples, the lower group similarity in the cluster analysis gave arguments to suppose very significant population fragmentation- and broken gene flow.

in the cases described above it could be suggested that the modification variability dominated as a response to the combination of local ecological conditions.

Acknowledgmentsthis work was supported by the national Science Fund of Bulgaria (BnSF), grant Do 02-246.

REFERENCES1. Bedir E. and Khan I.A. (2000) J. nat. Prod., 63, 1699-1701.2. De Combarieu E., Fuzzati N., Lovati M., Mercalli E.

(2003) Fitoterapia, 74, 583-591.3. Fang S., Hao C., Liu Z., Song F., Liu S. (1999) Planta

Med., 65, 68-73.4. Kostova I. and Dinchev D. (2005) Phytochemistry Rev., 4,

111-137.5. Kostova I., Dinchev D., Rentsch G.H., Dimitrov V.,

Ivanova A. (2002) Z. naturforsch. c, 57(1-2), 33-38.6. Peev D. and Valyovska N. (2010) Biotechnol. & Biotechnol.

eq., 24(special edition 2/on-line), 66-70.