Genome variability of some Gentiana L . species in nature and in culture in vitro : RAPD-analysis

Aim. 2nvestigation of intraspecies and somaclonal variability of G. acaulis, G. cruciata and G. punctata. Methods. Random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR), gel-electrophoresis. Results. It was established a species specificity for indicators of gentians genetic heterogeneity – the proportion of polymorphic amplicons and genetic distances. A level of intraspecies variation decreased in the direction of G. acaulis > G. punctata > G. cruciata. Changes in gentians tissue culture by the proportion of polymorphic amplicons were shown to vary within the range of 10–15 % and failed to extend beyond the intraspecies variation. Conclusions. Genome variation of G. acaulis, G. cruciata and G. punctata was studied through the use of RAPD-PCR. The gentians were found to be characterized by different variability in both nature and culture in vitro. In the tissue culture there were discovered considerably smaller number changes as compared to intraspecies variation.

Introduction.Gentiana L. species are valuable medicinal and decorative plants; growing in Ukraine mainly in the Carpathian Mountains.The diminishing of natural habitats of these species, decrease in their amount and violation of structure of their natural populations due to anthropogenic impact are the reasons of including many of them into the Red Data Book of Ukraine [1].In its turn, it limits their usage as raw material for pharmaceutics.The provision of efficient protection of gentians requires not only thorough study of these species (including their gene pool), but also the application of novel biotechnologies, in particular, obtaining their cell, tissue and organ cultures in vitro [2][3][4].
Besides, we evaluated somaclonal variation of tissue cultures of different Gentiana species by cytogenetic analysis and blot-hybridization [9][10][11][12].The results of molecular and genetic research revealed the changes in the Gentian genome in vitro which are in the range of intraspecies variability [9].In many cases the variability in vitro is not revealed at the DNA level, but is detected while studying karyotype [6,[10][11][12].
The aim of current work is further study on intraspecies and somaclonal variability of G. acaulis, G. cruciata and G. punctata by RAPD-PCR that will extend our knowledge of the peculiarities of Gentian genome variability in both nature and culture in vitro.
Materials and Methods.The plants from natural populations were used as material for investigation: G. acaulis (mountains Turkul, Rebra, Brebeneskul, all in Chornogora Ridge, Ukrainian Carpathians), G. cruciata (Krenychi village, Kyiv region and reservation "Medobory", Ternopil' region), G. punctata (mountains Pozhyzhevska and Breskul, Chornogora Ridge, mountain Troyaska, Svydovets Ridge, Ukrainian Carpathians), we also used plants of these species grown in sterile conditions (aseptic plants).Genetic variability of gentians in nature was estimated analysing of eight samples of G. acaulis from three populations (four plants -Rebra populations, two -Turkul population, and two -Brebeneskul population), six samples of G. cruciata (three plants from each of two populations -Medobory and Krenychi) and six samples of G. punctata from three populations (three plants from Breskul, two -from Troyaska and one -from Pozhyzhevska).Besides the plants, we used tissue culture of root origin: G. acaulis (Turkul mountain) at 7 th , 30 th and 72 nd passages, G. cruciata (Krenychi village) at 9 th passage, G. cruciata (reservation "Medobory") at 8 th , G. punctata (Troyaska mountain) at 10 th and G. punctata (Pozhyzhevska mountain) at 15 th passages of growth, obtained from aseptic plants.One passage of all calluses lasted for four weeks.The conditions of obtaining and growing calluses of the gentians investigated are described in the works [2,3].The variability of tissue culture of these species was compared with a donor plant, which was also used for the analysis of intraspecies polymorphism.
DNA isolation and gel-electrophoresis of amplification products were performed by the methods previously chosen [10].The conditions of polymerase chain reaction with the random amplified polymorphic DNA and nucleotide sequences of the primers are described in [8].
The results of processing RAPD-products electrophoregrams are presented as a binary matrix, where the presence or absence of similar in size amplicons is indicated as "1" or "0", respectively.Genetic distances were calculated according to Nei and Li's model on the basis of composed matrices using POPGENE 1.31 programme [13,14].The unweighted pair group method with arithmetic mean (UPGMA) was applied to build a dendrogram of relations between investigated objects, using MEGA 3.1 programme [15].The ratio of polymorphic amplicons (P) was determined by the formula Ð = Amount of polymorphic amplicons Total amount of amplicons

%. ×100
Results and Discussion.According to the RAPD-analysis of Gentiana samples with 21 arbitrary primers, 957 fragments of 250-3000 b.p. were obtained.Not all the primers produced clear amplification products, therefore calculations for G. cruciata, G. punctata, and G. acaulis were performed using 21, 19, and 17 primers, respectively.The total amount of clear restored amplicons for G. acaulis was 313 (18.4 per primer on average), G. cruciata -315 (15 per primer), G. punctata -329 (17.3 per primer).Some typical RAPD-spectra for G. acaulis plants are presented in Fig. 1, a.The percent of polymorphic amplicons (P) in investigated samples out of the total selection of primers was 63%.RAPD-spectra of the donor plant (Turkul mountain) and its tissue culture were similar, but not identical (Fig. 1, b).The level of genetic polymorphism between these samples is not high consisting only 12%.The highest similarity was shown for the RAPD-spectra of same callus samples, isolated at 30 th and 72 nd passages (Fig. 1, b).
The results of RAPD-analysis were used to build a dendrogram of genetic similarity of G. acaulis samples (Fig. 2).The dendrogram shows that the investigated objects form two clusters.The first cluster consists of the G. acaulis samples originated from the population localised on Rebra Mountain, the second -of the objects from two other populations, namely Turkul and Brebeneskul.The first cluster contains a subcluster, formed by wildling plants from Rebra population, and two separate branches, presented by aseptic plants from the same population.The second cluster divides clearly into two subclusters: one -the samples from Turkul Mountain, the other -plants from Brebeneskul Mountain.The values of genetic distances between specific genotypes of these populations are very close.In their turn, the samples of Turkul population show a separate branch, presented by a wildling plant, and two subclusters, the first of which is formed by a donor plant and a tissue culture of 7 th passage, isolated from it, and the second one -by the callus of 30 th and 72 nd passages from another plant, which was not studied in the current work.The dendrogram shows (Fig. 2) -it is samples of callus, cultivated for a long time that have the highest level of genetic proximity.The results presented show a higher genetic proximity of G. acaulis samples from Turkul and Brebeneskul populations, which form one cluster of the dendrogram, and remoteness of Rebra samples as they form a separate cluster.
According to Nei and Li the genetic distance between G. acaulis plants is in the range of 0.170-0.438(Table 1).These indices are in the range of values of interspecies variability, previously obtained by us (the average value of genetic distances between seven species is 0.54), and genetic distance between G. acaulis and G. verna (0.532), the closest in the dendrogram [8].The level of genetic differences between the callus of G. acaulis (Turkul Mountain) of 7 th passage and the donor plant is 0.122 which is in the range of genetic distances between the plants of these species.
For the investigated G. cruciata plants P equals 30%.The comparison of RAPD-spectra of donor plants and their tissue cultures, (Fig. 3), revealed that the levels of genetic polymorphism between them are 10% and 12% for Medobory and Krenychi populations, respectively.
The results of RAPD-analysis were used to build the dendrogram of genetic similarity of G. cruciata samples (Fig. 4), which were divided into two clusters according to the population.The first cluster, presented by the samples from Krenychi population, has two separate subclusters, formed by wildling and aseptic plants respectively, and one more plant from nature and tissue culture from an aseptic plant.The second cluster has samples from Medobory population, distributed in a different way: two plants belong to one subcluster, while aseptic plant and its tissue culture form separate branches.The dendrogram demonstrates that in both cases the donor plants and their calluses do not belong to any cluster, which may testify to some changes in the genome in G. cruciata tissue culture.The highest level of genetic proximity was demonstrated by two plants from Medobory population.
According to Nei and Li, the values of genetic distances between G. cruciata plants (Table 2) are in the range of 0.046 to 0.215, which is in good agreement with the previously determined index of interspecies variability of gentians (0.54), as well as genetic distance to the species, which is the closest in the dendrogram -G.pneumonanthe (0.498) [8].The comparison of the tissue culture and primary plant from Krenychi and Medobory populations revealed that genetic distance between them is 0.132 and 0.1, respectively.Some typical RAPD-spectra for G. punctata plants are presented in Fig. 5.The percentage of polymorphic amplicons in investigated G. punctata plants is 39%.According to this index, the level of genetic polymorphism between the primary plant of Pozhyzhevska and Troyaska populations and their tissue culture is relatively low, amounting to 11% and 15% respectively.
The results of RAPD-analysis were used to build a dendrogram of genetic similarity between investigated G. punctata samples (Fig. 6), forming two clusters.The first one comprises G. punctata objects (Troyaska Mountain), while the second one has representatives of two other populations of this species -Breskul and Pozhyzhevska populations.The second cluster may be divided into two subclusters: one is formed by three plants from Breskul Mountain, and the second one consists of the samples from Pozhyzhevska Mountain.It means that in the dendrogram of genetic similarity the division of G. punctata objects corresponds to their belonging to a specific population.Two plants of Breskul population demonstrated the highest genetic proximity.
According to Nei and Li, the genetic distance between G. punctata plants is in the range of 0.063 to 0.255 (Table 3).These values do not exceed the average index of interspecies variability (0.54), previously obtained by us, and the distance between Genetic distances by Nei and Li [14] between G. acaulis samples following RAPD-analysis G. punctata and the genetically closest G. lutea (0.351) [8].The genetic distance between a donor plant and tissue culture of G. punctata (Troyaska Mountain) is 0.158.This index is somewhat lower (0.113) for the samples of Pozhyzhevska population.The comparison of results of RAPD-analysis of G. acaulis, G. cruciata, and G. punctata shows that they are characterized by a different level of variability in both nature and culture in vitro.The amount of polymorphic amplicons for three species was relatively high: G. acaulis -63%, G. punctata -39% and G. cruciata -30%.The level of genetic polymorphism of a species is believed to on a number of factors, the most significant of them being biological peculiarities of a species (for instance, ecologic flexibility, a way of reproduction, etc.), sizes of populations, geographic location, etc. [16][17][18].RAPD-PCR study on genetic variability of different in size populations of a rare species, Gentianellla germanica L., a representative of Gentianella Moench  Nei and Li [14] between G. cruciata samples following RAPD-analysis  genus, close to Gentiana, revealed positive correlation between genome variability and sizes of populations [19].There is an assumption that growth in extreme conditions may cause increase in genetic heterogeneity which ensures survival of a species [16].In particular, the authors explain high level of genome polymorphism in Ungernia victoris Vved.Ex Artjushenko by growth in severe conditions and narrow areal of this endemic species [20].
Literature data on the impact of geographic location of a species on the level of genetic variability are ambiguous.For instance, the level of genome variability of populations of Orobanche gracilis Sm. (var.gracilis) depends on their location and is higher in the northern regions of Spain comparing to the southern ones [18].However, RAPD-analysis of polymorphism of representatives of Lemnaceae family did not reveal any connection between intraspecies differentiation and geographic location of samples.The authors explain this fact by the absence of territorial isolation of specific populations [21].
Generally, the grouping of investigated objects on dendrograms of genetic similarity corresponds to their  3 Genetic distances by Nei and Li [14] between G. punctata samples following RAPD-analysis Breskul mountains (two neighbouring peaks of Chornogora Ridge) are distributed on the dendrogram of genetic similarity in the boundaries of one cluster, whereas the samples from Troyaska population, which is much farther from two former ones (Svydovets Ridge), form a separate cluster.However, it is not possible to explain the data on G. acaulis, by geographic location only.
The results demonstrate high level of genetic similarity between the species of Turkul and Brebeneskul populations, which form one cluster on the dendrogram, while the Rebra samples, forming a separate cluster, are much farther from them, though Turkul Mountain is geographically closer to Rebra Mountain than to Brebeneskul Mountain.It is likely conditioned by genetic isolation of the mentioned populations, maybe due to the biological specificities of G. acaulis , which has the highest capability of vegetative reproduction among the investigated species [22].The similar regularity was revealed while studying these plants from other mountain ridges of Europe , in different parts of the Alps and the Pyrenees [23].In particular, G. acaulis samples from the same mountain ridge, the Maritime Alps, are located in different clades of the dendrogram according to the results of RAPD-analysis.The authors also revealed considerable genetic distance between the populations of G. clusii, one more representative of Ciminalis section, a species, which is very close by its biological properties to G. acaulis .
Taking into account our own and literature data we assume that the biological specificities of G. acaulis are a genuine reason of partial genetic distinguishing of the investigated populations and certain genotypes in the boundaries of one locality.
The level of somaclonal variation by P for all the investigated species is 2-5 times lower than intraspecies polymorphism and fluctuates in the range of 10-15%.The same tendency was revealed while studying the genome variability of U. victoris tissue culture: the authors found out that it is much lower in tissue culture than intraspecies variability [24,25].The level of somaclonal variability was in the range of intraspecies variability in the investigation of tissue culture and intact plants of some Iris L. genus plants [26].However, it is known that in vitro cultivation may cause considerable genome rearrangements, the scope of which in Rauwolfia serpentina is comparable with the interspecies rearrangements [27] and in Panax ginseng -with intraspecies rearrangements [28].
The selection of G. acaulis objects demonstrated the highest level of genetic proximity for the callus samples, isolated at 30 th and 72 nd passages of cultivation.It may testify to relative genetic stability of the mentioned callus during the third-sixth years of in vitro cultivation.This assumption seems to be proven by the results of cytogenetic investigations which demonstrated that with age the ratio of polyploid cells in this tissue culture increases, but the level of aneuploid cells in six-year-old callus does not change significantly in comparison with one-year-old one [29].Preservation of relative genetic stability during long-term cultivation was also revealed in the study on 45S rDNA in the tissue culture of 23 rd -50 th passages (2-5 years of in vitro cultivation) of another Gentian species -G.lutea [9].
Conclusions.RAPD-PCR method was applied to study genome variability of G. acaulis, G. punctata and G. cruciata.It was shown that gentians are notable for different level of genetic heterogeneity in both nature and culture in vitro.The content of polymorphic amplicons, characterizing the level of intraspecies variability, is as follows: G. acaulis -63%, G. punctata -39% and G. cruciata -30%.Clusterization of the objects on dendrograms corresponds to their belonging to a specific population which testifies to relative genetic isolation of populations (localities).The level of somaclonal variation by P for all the investigated species is 2-5 times lower than that of intraspecies polymorphism, and is in the range of 10-15%.

Table 2
Genetic distances by