Agricultura 10: No 1-2: 37-43 (2013)
Copyright 2013 by University of Maribor
Genetic diversity of Christmas rose (Helleborus niger L.) natural local populations as revealed by AFLP markers
Andrej ŠUŠEK1*
1 University of Maribor, Faculty of Agriculture and Life Sciences, Pivola 10, 2311 Hoče, Slovenia
ABSTRACT
Three naturally-growing local populations of Christmas rose (Helleborusniger L.) from two locations in Slovenia and one from neighbouring Croatia, were selected for molecular analysis based on AFLPs. 27 primer pairs were evaluated during the selective amplification step of the AFLP, and ten primer combinations were generated on the basis of the number of bands. The results from molecular analysis showed that the variation within populations was larger than that between populations. The molecular analysis showed that the genetic differences amongst populations were not high enough to state that the individuals of H. niger were grouped in clearly distinctive clusters. Small differences between the two Alpine populations (Bohinjska Bela and Peca) suggest that these are not isolated at the level of sexual reproduction. The migration of genes through pollen dissemination is probably always present, although pollen cannot migrate far.
Key words: Christmas rose, Helleborus niger, population diversity, molecular markers, AFLP
INTRODUCTION
The Christmas rose (Helleborusniger L.) that belongs to the family Ranunculaceae (Tamura 1993) is becoming more and more popular on the market as an ornamental plant (Armstrong 2002). However, it is a relatively new species regarding market-oriented production. It is important for its promotion to organise efficient and systematic genetic breeding, and to improve production technology. The efficiency of genetic breeding depends on several factors, such as available genetic resources, breeding methodology, breeders' experiences, and financial support.
Slovenia, one of the smaller European countries, appears to be very rich in terms of its biotic diversity (Martincic, 1999). The main reasons for such diversity are its specific geographical position as well as its diverse climate, relief, and bedrock. H.niger grows within all phytogeographical regions (Wraber 1969) (Fig. 1) and Slovenia is considered as being at the centre of diversity regarding this species. The variation within the genus Helleborus appears to be extremely high and the majority of the more important traits are at least partly influenced by the environment (Susek et al. 2005). Reliable data about existing diversity and variation within germplasm collections will be extremely important for breeders in the future.
The traditional evaluation of genetic variability, based on a description of morphological traits, cannot always ensure accurate estimates of the genetic variation. Many more reliable estimates of genetic variation can be obtained by
Figure 1: Localities of Helleborusniger in Slovenia (Jogan 2001). The map was adapted with the permission of the CKFF (the Slovenian Centre for the Cartography of Fauna and Flora)
using biochemical (e.g. isozymes) and molecular markers.
The AFLP technique has been successfully used for evaluating genetic relationships within the genus Helleborus. Meiners et al. (2011) analysed the taxonomic subdivision within the genus Helleborus. The results supported the previously-suggested division of the genus and thereby approved AFLP data as being applicable for phenetic
Correspondence to: E-mail: Andrej Šušek@um.si
analyses. AFLP markers have found the widest application regarding the analyses of genetic diversity below the species level within plants, particularly during investigations into population structure and when differentiating between both cultivated and natural/rare populations (Gupta et al. 1999, Nissim et al. 2004). Such analyses are crucial from the applied breeding point of view and for genetic conservation, and the rapidity with which AFLP markers enable the delivering of crucial information within intense time-constraints.
Patzak (2001) compared its potential for assessing the genetic diversity of ten hop (Humulus lupulus L.) varieties by using RAPD, STS, ISSR, and the AFLP molecular approaches. All the molecular methods accurately distinguish between the tested varieties, with the exception of three clones, which can only be distinguished by AFLPs. The similarities between varieties were revealed by using cluster analysis, and similar results were found for all the molecular methods.
When AFLPs reveal low-levels ofgenotypical differentiation between sub-populations' and relatively high variations within populations, they cannot be used for determining the populations' geographical origins. The reason could be the migration (of seeds or pollen) amongst sub-populations. This migration is probably present frequently. If seed migrations were excluded, the most important factor could be the migration of pollen. Examples can be the studies of wild North American populations of the weedy-plant Arabidopsis thaliana (L.) Heynh, as conducted by Jorgensen and Mauricio (2004), and the analysis of population structure regarding Arctophila fulva var. pendulina (Laest) Holmb. (Kreivi et al. 2005). Similar results were also obtained by Nissim et al. (2004) when studying the genetic variations of wild populations and cultivated genotypes of Anemone coronaria (L.). Most of the genetic variations (about 80 %) were documented within populations and only about 20 % between populations.
Data about the level of genetic diversity within and between populations of H. niger are crucial for genetic breeders. They are also very important for the establishment and management of germplasm collections. Information about the genetic structuring of populations is crucial during evalution studies. However, at this moment in time, no such information is available.
AFLPs can be very helpful for studying the genetic relationships amongst populations, and for determining or identifying ecotypes, hybrid lines, and cultivars. The AFLP technique can also represent a useful tool during the breeding process of the Christmas rose, thus enabling early determinations of traits that can only normally be determined when the plants are fully mature, and traits that can be determined only within specific environments (e.g., in the presence of certain diseases or pests). We studied the genetic structures of three naturally-grown populations that were chosen on the basis of results from morphological analysis. The aim of this study was to estimate the usefulness of the AFLP molecular technique when examining the genetic diversities amongst and within populations.
MATERIALS AND METHODS Plant material
The molecular analysis included randomly-chosen Christmas rose genotypes (individual plants) from two locations within Slovenia (Bohinjska Bela and Peca) and one in neighbouring Croatia (Zumberak). These locations were selected according to the results of morphological analysis that involved 8 populations (Susek et al. 2005). From each selected population, we took (for the molecular analyses) 16 plants (a total of 48 plants). Each individual within the population was collected at a distance of 50-500 m from the others, in order to avoid clones.
DNA isolation and AFLP analysis
Genomic DNA was extracted from the leaves according to the established CTAB method using NucleoSpin Plant kits (Macherey-Nagel, GmbH & Co.KG, Düren, Germany) by following the manufacturer's protocol. The DNA quantities and qualities were estimated by measuring the absorbance at 260 nm and 280 nm, using an Eppendorf BioPhotometer (Eppendorf AG, Hamburg, Germany), and by gel electrophoresis by means of 0.8 % agarose gel electrophoresis.
AFLP analysis was performed according to Vos et al. (1995), with slight modifications. The AFLP core reagent and starter primer kits were purchased from ScienceTec (LICOR Inc., Biotechnology division, Lincoln, Nebraska, USA). Approximately 100 ng of genomic DNA was double-digested using the restrictive enzymes mix EcoRI/MseI. The genomic DNA was incubated within a 5 x reactive buffer (50 mM Tris-HCl (pH 7.5), 50 mM Mg-acetate, and 250 mM K-acetate). Digestion was carried out within a final volume of 12.5 ^l at 37 oC for 2 h, and then heated to 70 oC for 15 minutes in order to deactivate the enzymes.
MseI and EcoRI adapters (Table 1) were subsequently ligated to the digested DNA fragments by adding 12 ^l of adapter ligation solution (EcoRI/MseI adapters, 0.4 mM ATP, 10 mM Tris-HCl (pH 7.5), 10 mM Mg-acetate, 50 mM K-acetate), and 0.5 ^l of T4 DNA ligase (1 unit/^l in 10 mM Tris-HCl (pH 7.5), 1 mM DTT, 50 mM KCl, 50 % glycerol (v/v)). The ligation was incubated for 2 h at 20 oC. After the reaction, the digested-ligated DNA fragments were diluted (1:10) and used as templates for the first amplification reaction (the pre-amplification step prior to the selective amplification).
Pre-amplification reactions were performed within a 25.5 ^l volume containing a 2.5 ^l of 10 x PCR buffer for AFLP (100 mM Tris-HCl, pH 8.3, 15 mM MgCl2, 500 mM KCl), 20 ^l of pre-amplification primer mix (Table 2), 2.5 ^l of diluted ligation and 0.5 ^l of Taq DNA polymerase (Roche Molecular biochemicals, Penzberg, Germany). The PCR amplification was carried out in a PTC-100 Programmable Thermal Controller using the following cyclic profile:
Table 1: Adapter and +3 primer sequences (5'-3') used for AFLP analysis
Name
Enzyme
Type
Sequence (5'-3')
E-0
M-0
EcoRI
MseI
Adapter
Adapter
5'-CTCGTAGACTGCGTACC
CATCTGACGCATGGTTAA-5'
5'-GACGATGAGTCCTGAG
TACTCAGGACTCAT-5'
M-CAG M-CTG M-CAC M-CTA
Msel Msel Msel Msel
Unlabelled primer +3 Unlabelled primer +3 Unlabelled primer +3 Unlabelled primer +3
GAT GAG TCC TGA GTA ACA G GAT GAG TCC TGA GTA ACT G GAT GAG TCC TGA GTA ACA C GAT GAG TCC TGA GTA ACT A
E-ACA	EcoRI	IRDye 700 labelled primer +3
E-ACC	EcoRI	IRDye 700 labelled primer +3
E-AAG	EcoRI	IRDye 700 labelled primer +3
GAC TGC GTA CCA ATT CAC A GAC TGC GTA CCA ATT CAC C GAC TGC GTA CCA ATT CAA G
E-ACT	EcoRI	IRDye 800 labelled primer +3
E-AGC	EcoRI	IRDye 800 labelled primer +3
GAC TGC GTA CCA ATT CAC T GAC TGC GTA CCA ATT CAG C
Step
Temperature (oC)
Time
1.	denaturation
2.	annealing
3.	extension
4.
94 56 72 4
30 seconds 60 seconds 60 seconds hold
20 cycles total
for long-term usage
Gel electrophoresis was conducted using a Li-Cor Long ReadIR DNA sequencer (LI-COR Inc., Biotechnology division, Lincoln, Neb., USA).
Data analysis
After the reaction, pre-amplified DNA fragments were diluted (1:40) and used as templates for the selective amplification reaction. A volume of 11 ^l PCR reaction combined with a 2 ^l diluted (1:40) pre-amplification product, 0.5 ^l of IRDye 700 labelled EcoRI primerA, 0.5 ^l of IRDye 800 labelled EcoRI primerB, 2 ^l of MseI primer, and 6 ^l of TaqMix. The TaqMix for 33 reactions contained 40 ^l of 10xPCR amplification buffer (100 mM Tris-HCl, pH 9, 15 mM MgCl2, 500 mM KCl), 158 ^l of deionised water and 2 ^l Taq DNA polymerase (5units/^l) (Roche Molecular Biochemicals, Penzberg, Germany). The second amplification was carried out in a PTC-100 Programmable Thermal Controller by programming a touch-down cyclical profile (Don et al., 1991). Touchdown PCR was performed by decreasing the annealing temperature from 66 °C to 56 °C by 0.7/30 s increments per cycle with each of the initial 12 cycles, followed by a touchdown annealing temperature for the remaining 25 cycles at 56 °C for 30 s with a final cycle of 72 °C for 1 min and a hold at 4 °C.
Following the amplification reaction, each product was mixed with 5.0 ^l of blue-stop solution (98 % formamide,10 mM EDTA pH 8.0, and bromo-phenol-blue and xylene cyanol as tracking dyes). The resulting mixtures were heated at 94 oC for 3 min., and then quickly cooled by ice and covered to reduce exposure to light. The samples were stored at -20 oC
Image analysis was performed using Saga 2 software (LICOR Inc., Biotechnology division, Lincoln, Neb., USA). After the automated analysis, the placements of lanes, calibrations, desmile lines, and the scoring of bands was reviewed by the visual inspections of two independent readers. Only the bands' scores by both readers were accepted as reproducible. When considering of bands, the diffused or faint bands were not taken into account and neither were those bands that occurred at the extremes of the amplified size-range. Infrequent bands were eliminated. Only the bands with a frequency above 20 % were taken into account. Each AFLP marker was treated as a unit character and scored as a binary code (1/0).
Only polymorphic bands were used during the construction of a binary matrix, reflecting the presence or absence of those fragments obtained by AFLP, in different genotypes. The estimates of genetic similarities (GS) amongst all genotypes were calculated according to the Jaccard similarity coefficient (1908). The genetic distance was calculated as GD = 1 - J, using the data from the Jaccard's similarity coefficient matrix. The resulting distance matrices were subject to clustering methods by the Neighbour Joining Tree (NJTree) method.
The reliabilities and robustness of the phenograms were tested by bootstrap/jackknife analyses by drawing 1,000 samples.
Data analyses were done by Darwin, the non-commercial
software of CIRAD (Perrier et al. 2003). A bootstrap value was given to each edge and indicated the occurrence frequencies of this edge in the bootstrapped trees.
Principal component analysis (PCA) was used in an
attempt to detect geographical patterns amongst the H.niger
populations.
RESULTS
During the selective amplification step, the duplex of one MseI with two IRDye labelled EcoRI primers was used for exploiting the capability of the LI-COR system when detecting two different dyes. The 27 primer pair combinations were evaluated with regard to their capacities for generating polymorphic bands using 4 genotypes, randomly chosen from the 3 populations being investigated (Table 2). According to their capacities to generate polymorphic bands, it was decided to use a five duplex of one MseI with two IRDye labelled EcoRI primers: (1) M-CAG, E-ACA (IRDye 700 labelled), E-AGC (IRDye 800 labelled); (2) M-CTG, E-ACC (IRDye 700 labelled), E-AGC (IRDye 800 labelled); (3) M-CAC, E-AAG (IRDye 700 labelled), E-ACT (IRDye 800 labelled); (4) M-CTA, E-ACC (IRDye 700 labelled), E-AGC (IRDye 800 labelled); (5) M-CTG, E-ACA (IRDye 700 labelled), and E-ACT (IRDye 800 labelled). Each selected primer combination generated a greater number of bands. The number of polymorphic fragments for each primer pair varied from 5 (for M-CTG/E-ACC) to 23 (for M-CAG/E-ACA), with an average of 8.9 per primer pair (Table 3). The sizes of the AFLP fragments were determined by comparing an AFLP standard marker to the AFLP patterns. The AFLP fragments' sizes ranged from approximately 45 to 400 base pairs (bp). The polymorphic fragments were distributed across the entire size-range, with the major proportion being between 75 and 300 bp.
In order to assess the usefulness of AFLPs as phenetic markers, a similarity-matrix based on Jaccard's coefficient was constructed for estimating the level of relatedness amongst individuals from the three natural populations used during the study. Calculation of the Jaccard coefficient was based on the presence or absence of discrete characters (AFLP markers). The similarity matrix was then used to cluster the data using the NJTree method and to conduct PCA analysis. The 48 individuals appeared as being scattered along the resulting dendrogram constructed by the Jaccard coefficient, and by the NJTree clustering method (Fig. 2). Individuals
Table 3: Number of polymorphic fragments and their size-range
Primer combination	Scored polymorphic bands	Fragment size-range (bp)
M-CAG/E-ACA	23	50-370
M-CAG/E-AGC	10	50-370
M-CTG/E-ACC	5	40-340
M-CTG/E-AGC	6	50-360
M-CAC/E-AAG	13	45-340
M-CAC/E-ACT	5	50-400
M-CTA/E-ACC	6	50-370
M-CTA/E-AGC	8	45-370
M-CTG/E-ACA	8	50-255
M-CTG/E-ACT	5	50-370
Totals	89	-
Mean	8.9	-
Number		
of primer	10
combination
Maximum	23	-
Minimum	5	-
Standard	5 59
deviation_'_
from the Zumberak population were clustered into two sub-clusters of 5 and 10 individuals, and a sub-cluster of 6 individuals was identified for the Peca population. All the other individuals were clustered in groups of two to four individuals, irrespective of their geographic origins. The resulting dendrogram showed low bootstrap values for the main edges with a certain number of null values.
The similarity matrix was also used as input data for principal component analysis (PCA). The scatter-plot representation of the PCA showed a clear separation of 9 individuals from the Zumberak population in relation to the first two principal axes of variation (Fig. 3). When the plan was formed by axes 1 and 2 only, 18.11 % (9.76 % + 8.35 %) of the total variation was explained by the markers. The second
Table 2: Primer pair combinations tested for conducting selective amplification
_M-CAA	M-CAC M-CAG M-CAT M-CTA M-CTC M-CTG M-CTT
E-AAC	o »o	•
E-AAG o	•
E-ACA	o • •	•
E-ACC	• •	o	•	o
E-ACG •	o o
E-ACT o	• o0
E-AGC	• •
E-AGG	o • o
plane (axes 3 and 4) accounted for 13.29 % (6.82 % + 6.47 %)	individuals. Similarly, in the case of the third plane (axes 4
of the total variation. This meant that the markers could not and 5), only 12.17 % (6.47 % + 5.70 %) of the total variation explain the highest portion of variation amongst the studied could be explained.
«44 • 40
>39 • 48 -•28 —•12
Sub-cluster Zumberak
> Sub-cluster Peca
Sub-cluster Zumberak
■ 45
Figure 2: Unrooted dendrogram of individual plants from wild populations based on individual scoring for each AFLP band. The numbers represent individuals from various populations: Bohinjska bela (from 1 to 16); Peca (from 17-32); Zumberak (from 33 to 48). Neighbour-joining clustering was performed, using the genetic similarity coefficient of Jaccard calculated between every pair of the genotype. Bootstrap values (%) for each branch point are indicated
3
57
2
2
24
• 26
27
25
19
22
5
10
42
0
11
0
7
15
23
0
11
• 29
20
• 30
0
22
41
37
21
• 17
13
16
47
34
33
36
35
38
41
37
14
0- L7
43
46
34
4
8
• 32
72
31
0
0.2
31 • 12*
< u
-.3	-.25	-.2	-.15	-.1	-.05
32	9
27
DARwin 4.0 ©CIRAD-Flhor - 17.7.2006 -
PCA 1 (9.76 %)
Figure 3: Principal component analysis (PCA) of AFLP data for H.
niger naturally-growing populations. The numbers represent individuals from various populations: Bohinjska Bela (from 1 to 16); Peca (from 17-32); Zumberak (from 33 to 48). The percentage of total variance explained by the first two principal components is shown on the abscissa and ordinate axes, respectively.
35
23
22
41
30
21
25
42
15
29
44
20
48
16
05
19
40	28
10
17
18
05
25
35
24
26 2
25
15
45
-.2	4
14
DISCUSSION
The AFLP technique was used for evaluating the genetic diversities within and between three naturally-grown populations of Christmas rose. As this was the first work on the natural diversity of H.niger, sixteen individuals were selected from each population in order to gain some knowledge of their genetic diversities. For technical reasons, we decided to use different individuals as repetitions, being aware that this could not eliminate all the possible errors associated with the representativeness of the data.
For the AFLP assay, we used a relatively wide-range of primer combinations. From amongst the 27 AFLP primer combinations tested, 17 were discarded owing to the reduced number of obtained fragments, polymorphic or not. The genetic diversities were finally evaluated using 10 primer combinations that generated 89 polymorphic fragments and made the fingerprinting of each individual genotype possible.
Evaluation of the genetic diversity of the Christmas rose within naturally-occurring Slovenian populations did not show strong differentiations. Although several individuals from the same location appeared to be very tightly clustered,
many of them were distributed all over the dendrogram (Fig.
2).	The heterogeneity within clusters was higher than between clusters. PCA analysis showed that only a low percentage of the total variations amongst these individuals could be explained by the markers (Fig. 3). However, geographical distance tended to be a factor associated with genetic similarities. The Zumberak population, which was geographically the most distant when compared to the Bohinjska Bela and Peca populations, also tended to be genetically different to the two latter populations, to a certain extent (Figs. 2 and
3).	A similar situation was observed by Susek et al. (2005) on the basis of morphological traits where, on the basis of the first principal component, the Zumberak population was separated from the Peca and Bohinska Bela, which could not be distinguished one from the other.
One interpretation of the structure in the neighbour joining-tree and in the scatter plot from the PCA analysis could be that these Alpine populations were not genetically isolated. The only substance that enables genetic communication amongst populations is probably pollen, which is carried by various insects (e.g., bees).
ACKNOWLEDGMENTS
derung Sloweniens. Vegetatio 1969;17:176-199.
I would like to express my gratitude to Dr. Diederik van Tuinen of INRA-Dijon (France), for his technical advice and support during molecular analyses.
Received: October 31, 2012 Accepted in final form: April 17, 2013
RFERENCES
1.	Armstrong H. Patience with Helleborus pays off. Flow-erTECH 2002;5(3):8-11.
2.	Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS. "Touchdown" PCR to circumvent spurious priming during gene amplification. Nucl. Acids Res.1991;19:4008.
3.	Gupta PK, Varshney RKV, Sharma PC, Ramesh B. Molecular markers and their applications in wheat breeding. Plant Breed. 1999;118:369-390.
4.	Jaccard P. Nouvelles recherches sur la distribution florale. Bull. Soc. Vaud. Sci. Nat. 1908;44:223-270.
5.	Jogan N. Gradivo za Atlas flore Slovenije. Materials for the Atlas of Flora of Slovenia. Center za kartografijo favne in flore, Miklavž na dravskem polju, 2001.
6.	Jorgensen S, Mauricio R. Neutral genetic variation among wild North American populations of the weedy plant Arabidopsis thaliana is not geographically structured. Mol. Ecol. 2004;13:3403-3413.
7.	Kreivi M, Rautiainen P, Aspi J, Hyvarinen M. Genetic structure and gene flow in an endangered perennial grass, Arctophila fluva var. pendulina. Conserv. Genetics 2005;6:683-696.
8.	Meiners, J, Debener T, Schweizer G, Winkelmann T. Analysis of the taxonomic subdivision within the genus Helleborus by nuclear DNA content and genome-wide DNA markers. Sci. Hortic. 2011;128:38-47
9.	Nissim Y, Jinggui F, Arik S, Neta P, Uri L, Avner C. Phe-notypic and genotypic analysis of a commercial cultivar and wild populations of Anemone coronaria. Euphytica 2004;136:51-62.
10.	Patzak J. Comparison of RAPD, STS, ISSR and ADLP molecular methods used for assessment of genetic diversity in hop (Humulus lupulus L.). Euphytica 2001;121:9-18.
11.	Perrier X, Flori A, Bonnot F. Methods of data analysis. In: Hamon P, Seguin M, Perrier X, Glaszmann JC (eds.), Genetic diversity of cultivated tropical plants. Cirad, Moellier, 2003:31-63.
12.	Šušek A, Ivančič A, Lemoine MC, Guillemin JP, Ca-neill J, Šiško M, Janžekovič F, Praprotnik L. Variability of Christmas rose (Helleborus niger L.) populations and its potential use in genetic breeding. Acta. Biol. Cracov. Bot. 2005;47(2):129-135.
13.	Tamura M. Ranunculaceae. In: Kubizki K, Rohwer JG, Bittrich V (eds.), The families and genera of vascular plants: flowering plants-dicotyledons, 2. Berlin 1993; 563-583.
14.	Vos P, Hogers R, Bleeker M, Rijans M, Van de Lee T, Hornes M, Frijters A, Pot J, Kuiper M, Zabeau M. AFLP: A new technique for DNA fingerprinting. Nucl. Acids Res. 1995;23:4407-4414.
15.	Wraber M. Pflanzengeographische Stellung und Glie-