Assessment of Genetic Stability of Propagated Plantlets of Four Sea Buckthorn (Hippophae) Cultivars and the Establishment of Genetic Relationships Between them by ISSR Markers

Abstract

The common sea buckthorn or seaberry (Hippophae rhamnoides) is important environmentally, commercially, and as a new berry crop. Some morphological differences among mother plants and lines obtained by cuttings often lead a purchaser to question if the propagated plantlets are true-to-name cultivars. Intersimple sequence repeat (ISSR) markers were employed to assess the genetic stability of lines obtained by cuttings of four H. rhamnoides cultivars and the genetic relationships among them. Fifteen ISSR primers generated 174 bands among the 35 samples of four cultivars, 158 of which (90.8%) were polymorphic. Percentages of polymorphic and unique bands between mother plant and propagated plantlets ranged from 53.21% to 77.17% and from 7.09% to 13.76%, respectively. Cluster analysis based on ISSR data indicated that at a Jaccard coefficient of 0.78, mother plant and lines obtained by cuttings for each cultivar were grouped into different subclusters, respectively, which could be further clustered into different sub-subclusters. Our data first indicated that the genotype was affected by the cutting process to a certain degree, because there were some morphological and low molecular variations between the mother plant and lines obtained by cuttings. Information generated from this study can be used to select parents for hybrid development to maximize desirable agronomic traits in a breeding program aimed at developing segregating populations to map genes controlling special traits in H. rhamnoides.

KEYWORDS:

• breeding
• genetic relationship
• genetic stability
• Hippophae rhamnoides
• molecular marker
• multiplication by cutting

INTRODUCTION

The genus Hippophae L. is a deciduous, perennial shrub or tree of the Elaeagnaceae. It has six species and 12 subspecies worldwide (Bartish et al., 2002; Sun et al., 2002; Swenson and Bartish, 2002) that are restricted to the Qinghai-Xizang Plateau and adjacent areas except for Hippophae rhamnoides L., which is naturally distributed from Asia to Europe and has been introduced into Africa, South America, and North America. H. rhamnoides adapts well to extreme conditions, including drought, salinity, and alkalinity. Commercially, it is a multipurpose plant with orange, red, or yellow berries and has substantial agricultural, ecological, nutritional, medical, and ornamental values (Cheng et al., 2003; Zadernowski et al., 2002).

Hippophae rhamnoides is being planted on a large scale in the world for commercial aims and environmental improvements. This requires a great number of seedlings of specific cultivars. In its natural condition, reproduction of H. rhamnoides occurs through seeds with a low germination rate as well as root suckers. Some previously published results indicated that the application of micropropagation is a valid tool for propagation of this species (Liu et al., 2007; Montpetit and Lalonde, 1988; Mou, 1995), but there are low regeneration rates and survival rates after transplantation (Ruan et al., 2007). In contrast, multiplication by cuttings that have over an 80% survival rate is still the main protocol used for propagation of H. rhamnoides cultivars (Ruan et al., 2007).

Micropropagation protocol is severely hindered due to incidences of somaclonal variations (Joshi and Dhawan, 2007), as in case of oil palm, where aberrant flowering patterns were observed among the regenerated plants (Matthes et al., 2001). Somaclonal variation mostly occurs as a response to the stress imposed on the plant in culture conditions and is manifested in the form of DNA methylations, chromosome rearrangements, and point mutations (Kaeppler and Phillips, 1993; Peschke et al., 1987; Phillips et al., 1994; Xu et al., 2004). There are many woody species commercially produced by micropropagation (Matthes et al., 2001), and variability induced by cutting multiplication has also been reported in Sansevieria trifasciata (http://www.yhagri.gov.cn/documents/docdetail.asp? documenti =19397&sub_menuid=76) and Vigna radiata (Zhang et al., 2002). Hence a stringent quality check in terms of genetic similarity of the lines obtained by cuttings becomes important not only for seedling production but also for lawsuits requiring careful technical investigation (unauthorized commercialization of patented cultivars).

Success of any crop-breeding program is based on the knowledge of and availability of genetic variability for efficient selection (Ali et al., 2008). Genetic similarity estimates among genotypes are helpful in selecting parental combinations for creating segregating populations so as to maintain genetic diversity in a breeding program (Becelaere et al., 2005) and the classification of germplasm into heterotic groups for hybrid-crop breeding (Menz et al., 2004). The search for and establishment of heterotic groups can be based on geographical origin, agronomical traits, pedigree data, or molecular marker data (Melchinger, 1999). Neither pedigree- nor morphologically based estimates may reflect the actual genetic difference of the studied populations (Almanza-Pinzon et al., 2003; Fufa et al., 2005; van Beuningen and Busch, 1997). On the other hand, molecular markers are not influenced by the environment and reflect genetic similarity (and differences) and do not require previous pedigree information (Bohn et al., 1999), which is valuable for plants for which pedigree information is lacking.

Various types of molecular markers are available for genome analysis in trees and shrubs (Teixeira da Silva et al., 2005) and have been used to assess the genetic uniformity of the micropropagated plantlets (Martins et al., 2004). This is apparent in studies conducted to screen somaclonal variations produced in tissue-cultured plants such as in neem (Singh et al., 2002), tea (Singh et al., 2004), and soybean (Hofmann et al., 2004). Intersimple sequence repeat (ISSR) markers in particular have been reported to be very useful to unravel genetic diversity among closely related genotypes, as there are high polymorphic bands (Agostini et al., 2008; Bornet et al., 2002; Prevost and Wilkinson, 1999; Sreedhar et al., 2007; Vijayan et al., 2006) and to detect variations among tissue culture-produced plants (Chowdari et al., 1998; Leroy et al., 2000; Leroy et al., 2001; Rahman and Rajora, 2001). It has been extensively used in genetic-diversity studies in many plants, including Hagenia abyssinica (Feyissa et al., 2007), fig (Ikegami et al., 2009), mulberry (Kar et al., 2008) and papayas (Carrasco et al., 2009). Joshi and Dhawan's (2007) results, based on ISSR markers, confirmed the clonal fidelity of the tissue culture-raised Swertia chirayita plantlets and corroborated the fact that axillary multiplication is the safest mode for multiplication of true-to-type plants.

Genetic diversities, variations, originations, and phylogenetic relationships of H. rhamnoides have been studied at a molecular level (isozyme, random amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP), ISSR, chloroplast DNA (cpDNA) and internal transcribed spacer (ITS); Bartish et al., 1999; Chen et al., 2008; Ruan et al., 2007; Tian et al., 2004). Among the six subspecies of H. rhamnoides, ssp. sinensis and ssp. mongolica are mainly and widely distributed in Asia and Europe. Selection and breeding programs for these two subspecies have been running since the 1960s to improve their adaptability, yield, and quality. Some specific cultivars have been successfully selected and bred in many countries (e.g., China, Russia, and Germany) and distributed using plants from multiplication by cuttings. However, some morphological differences among mother plants and lines obtained by cuttings (Table 1) not only make a purchaser question if the plantlets obtained by cuttings are true-to-name cultivars but has also lead to commercial disputes (unauthorized commercialization of patented cultivars).

TABLE 1 Plant Cultivars Used in this Study that are Planted Widely in China and Other Countries and Potential Materials for Breeding Fine and Resistant Lines

Cultivars	Code (identify) (ages)	Origin	Description	Morphological differences among mother plant and lines obtained by cuttings
Yalishandashierhao (YLSD)	YM (mother plant, accession no. FXYLSD9901) (8)YL1–8 (lines obtained by cuttings) (5)	Russia	height about 2.5–4.0m, obvious trunk, no-thorn, orange columnar fruit (Fig. 1A), mean fruit weight 0.462g, resistance to dried-shrink disease	Leaf type: thick and long elliptic for mother plant, but oblate for lines obtained by cuttings;Columnar fruit for mother plant, but near spherical for some lines obtained by cutting
Hongmaxiongzhu (HMXZ)	HM (mother plant, accession no. FXHMXZ9923) (8)HL1–13 (lines obtained by cuttings) (4)	Russia	Male parent, height about 2.0–3.0m, no-thorn, more pollen, susceptible to dried-shrink disease	Dark brown bark for mother plant, but light brown for lines obtained by cuttings;No thorn for mother plant, but few or no thorn for lines obtained by cuttings
Zhongguowucixiongzhu (ZGWCXZ)	ZM (mother plant, accession no. FXZGWC9758) (10)ZL1–2 (lines obtained by cuttings) (6)	China	Male parent, height about 2.0–4.0m, no-thorn, more pollen, strong adaptability, fast growth, resistance to dried-shrink disease	No-thorn for mother plant, but few thorns for lines obtained by cuttings
Hongmaxiongzhu (WHSJ)	WM (mother plant, accession no. FXWH9810) (9)WL1–8 (lines obtained by cuttings) (5)	Introduced from Russia, but selection in China	Height about 2.0–3.0m, no-thorn, fruit without stone (Fig. 1B), long fruit stalk, resistance to dried-shrink disease	No-thorn and long fruit stalk for mother plant, but few or no-thorn and short fruit stalk for lines obtained by cuttings;Without stone for fruits of mother plant, but few fruits with stone

FIGURE 1 Fruits of ‘Yalishandashierhao’ (A) and profile view of ‘Wuheshaji’ fruit without stone (B).

In this study, we first report the use of ISSR markers to assess genetic stability of propagated plantlets of four H. rhamnoides cultivars; then, we analyze genetic relationships among the four cultivars with different characteristics. These data provide a scientific basis for the identification of genetic stability of propagated Hippophae plantlets and the development of crossing strategies in a breeding program.

MATERIALS AND METHODS

Plant Materials and DNA Extraction

Four H. rhamnoides cultivars with different characteristics (Table 1), which are being widely planted on a large scale in China and other countries, were used in this study, including ‘Yalishandashierhao’ (‘YLSD’; ssp. mongolica), ‘Hongmaxiongzhu’ (‘HMXZ’; ssp. mongolica), ‘Zhongguowucixiongzhu’ (‘ZGWCXZ’; ssp. sinensis), and ‘Wuheshaji’ (‘WHSJ’; ssp. mongolica; Fig. 1B). In our cultivar-cutting orchard there are just four original genotypes (only one original genotype within a cultivar and the rest of the plants should be clones of these). There are some morphological differences between mother plants and lines obtained by cuttings (Table 1). The mother plant and lines obtained by cuttings for each cultivar were used as donor materials for multiplication by cuttings and have some differences in morphological characteristics. They were selected as the samples in this study (Table 1) to investigate ISSR analyses for assessing the genetic stability of lines obtained by cuttings. In addition, the four cultivars used in this study are potential materials for breeding specific lines with different characters. Genetic relationship estimates among genotypes based on molecular markers is helpful in selecting parental combinations for creating segregating populations so as to maintain genetic diversity in a breeding program (Becelaere et al., 2005) and clone the desired gene. To develop optimal strategies for breeding, we estimated the genetic relationships among the four cultivars using ISSR markers.

Young leaves were collected from each sample of each cultivar in 2007. Total genomic DNA was extracted from leaf tissue using the protocol of Ruan et al. (2004).

ISSR Analysis

Amplification reactions of ISSR analyses were carried out in a total volume of 20 µL in a 0.2 mL microfuge tube: 2µl of the 10X reaction buffer, 6 mM MgCl₂, 600 µM dNTPs, 0.5 µM primer, 4% formamide, 20 ng DNA template, and 2U Taq DNA polymerase (TakaRa, China). The thermocycler program for PCR was set at 1.5 minutes at 94°C; followed by 45 cycles of 45 seconds at 94°C, 45 seconds annealing at 53°C, 1.5 minutes extension at 72°C; and a single 7 minutes at 72°C extension cycle. The samples were then placed at 4°C until gel analysis was performed. Primer selection was performed on the samples of four cultivars using 40 primers purchased from the University of British Columbia. Based on the number and quality of polymorphic fragments at 52°C, 52.5°C, 53°C, 53.5°C, and 54°C, fifteen ISSR primers (see Table 2) were selected and amplification was carried out on all samples in a GeneAmp PCR system 9700 (Applied Biosystems). Amplification products were separated on 1.5% agarose gels, and visualized under AutoChemi System ^TM UVP Bioimaging Systems after ethidium bromide staining.

TABLE 2 Assessment of Genetic Stability of Propagated Plantlets Based on ISSR Markers

UBC primer	Primer sequence	Range of band size (bp)	Number of bands (total bands, and in parcentheses, polymorphic bandsa and unique bandsb)
			YLSD	HMXZ	ZGWCXZ	WHSJ
807	(AG)8T	300–1,300	8 (5, 0)	9 (7, 0)	11 (10, 1)	11 (6, 1)
808	(AG)8C	300–1,000	7 (0, 0)	6 (5, 0)	9 (6, 4)	5 (4, 0)
809	(AG)8G	300–1,500	11 (11, 1)	13 (13, 0)	11 (5, 3)	10 (10, 1)
810	(GA)8T	300–1,100	6 (1, 0)	5 (1, 0)	4 (0, 0)	6 (1, 0)
811	(GA)8C	300–1,000	4 (1, 0)	7 (4, 0)	8 (8, 2)	7 (5, 2)
815	(CT)8G	150–1,500	9 (0, 0)	13 (4, 3)	8 (1, 0)	9 (1, 0)
816	(CA)8T	150–400	7 (5, 1)	8 (5, 0)	7 (3, 0)	6 (0, 0)
817	(CA)8A	200–800	5 (4, 1)	10 (10, 1)	9 (2, 2)	10 (8, 0)
818	(CA)8G	150–450	6 (4, 1)	7 (6, 1)	2 (0, 0)	3 (2, 0)
825	(AC)8T	200–700	7 (1, 1)	7 (5, 0)	5 (2, 0)	5 (1, 0)
827	(AC)8G	200–500	5 (3, 1)	5 (3, 0)	5 (0, 0)	5 (0, 0)
835	(AG)8YC	150–1,500	9 (9, 1)	9 (9, 1)	9 (9, 3)	9 (4, 2)
840	(GA)8YT	100–1,300	15 (14, 5)	12 (12, 2)	5 (4, 0)	8 (7, 5)
881	GGGT(GGGG)2G	300–1,300	5 (4, 0)	6 (5, 1)	10 (7, 4)	8 (5, 0)
887	(AGT)(ACG)(AGT)(TC)7	250–1,500	11 (8, 0)	10 (9, 0)	6 (1, 1)	10 (3, 0)
Total			115 (70, 12)	127 (98, 9)	109 (58, 15)	102 (57, 11)
^apolymorphic band means the band did not appear in all samples of one cultivar.
^bunique band means the band only appeared in mother plant or any one line obtained by cutting of one cultivar.

Data Analysis

The amplified fragments of each ISSR marker were scored as “1” and “0”, where “1” indicated the present of a band and “0” indicated its absence. The Jaccard (1908) coefficient was calculated by using NTSYS2.02 (Rohlf, 2000). A similarity tree was produced based on the Jaccard coefficient with the unweighted-combination group method using arithmetic averages (UPGMA) and the sequential agglomerative hierarical nested (SAHN)-clustering program. The goodness of fit of the clustering was tested using the matrix comparison (MXCOMP) program, which directly compares the original similarity matrix and the cophenetic value matrix. Cluster analysis and MXCOMP were conducted by using NTSYS2.02 (Rohlf, 2000).

RESULTS

ISSR Analysis

Fifteen ISSR primers generated 174 ISSR bands for all 35 samples, 158 of which (90.8%) were polymorphic. The number of bands over all 35 samples detected by individual primers ranged from six for the UBC primer 827 to 19 for the UBC primer 840, with a mean of 11.6 bands per primer. Among the four cultivars, the number of unique bands detected by an individual primer ranged from one for the UBC primers 809, 810, 825, and 827 to six for the UBC primer 840, with a mean of 3.7. Fifteen primers generated 38 unique bands among four cultivars, two of which, generated from the UBC primer 811, appeared in all lines of ‘HMXZ’ obtained by cuttings, and one generated from the UBC primer 816 appeared in all lines of ‘WHSJ’ obtained by cuttings. In our study all the primers amplified bands between 100 bp to 1,500 bp molecular size range. Amplifications produced by the UBC primer 881 were shown in Figure 2.

FIGURE 2 Amplifications produced by the UBC primer 881. Lane M represents the 100 bp DNA Ladder Marker and other lanes represent the mother plants or lines obtained by cuttings of four Hippophae cultivars, means of code in Table 1.

Table 1 shows total bands, polymorphic bands, and unique bands generated from each cultivar using 15 ISSR primers. Among mother plants and lines obtained by cuttings, the percentages of polymorphic and unique bands ranged from 53.21% to 77.17% and ranged from 7.09% to 13.76%, respectively. Additionally, 17.1% polymorphic bands for ‘YLSD’ were unique bands, as well as 9.2% for ‘HMXZ’, 25.86% for ‘ZGWCXZ’, and 19.3% for ‘WHSJ’.

Cluster Analysis

Based on the Jaccard coefficient, a UPGMA dendrogram was constructed (Fig. 3). The measure of goodness of fit of the cluster analysis was given by the “cophenetic correlation” (r) calculated with the MXCOMP program. The dendrogram had a matrix correlation r  =  0.893, which was interpreted as a good fit.

FIGURE 3 Dendrogram of 35 samples of four Hippophae cultivars by UPGMA cluster analysis based Jaccard coefficient, using 158 ISSR polymorphic bands obtained by 15 primers, means of code in Table 1.

At a Jaccard coefficient of 0.67, the following clusters were formed: cluster I (group IV) included one ssp. sinensis cultivar (‘ZGWCXZ’) from China; cluster II consisted of three ssp. mongolica cultivars from Russia, including ‘YLSD’ (group I), ‘HMXZ’ (group II), and ‘WHSJ’ (group III). At a level of 0.78, mother plant and all lines obtained by cuttings for each cultivar were grouped into different single subcluster (Fig. 3), respectively. Group I could be further clustered three sub-subclusters. Group Ia included YM and YL8, five plantlets (YL2–YL6) were clustered in group Ib, and group Ic consisted of YL1 and YL7. In group II, 13 plantlets of ‘HMXZ’ were clustered in three subgroups, group IIa included five plantlets (HL1–HL5), five plantlets (HL7, HL10–HL13) were clustered in group IIb, and the remaining three (HL6, HL8, and HL9) were consisted of group IIc. For ‘WHSJ’, group IIIa included WM and WL6, three plantlets (WL2–WL4) in group IIIb were not discriminated clearly by the dendrogram, and the remaining four plantlets (WL1, WL5, WL7, and WL8) in group IIIc were separate each other. For ‘ZGWCXZ’, two plantlets in group IVb were discriminated from ZM at a Jaccard coefficient of 0.79.

DISCUSSION

Our results of ISSR bands and cluster analysis indicate that multiplication by cuttings affect the genotype of Hippophae to a certain degree, because there were some morphological and low molecular variations among mother plants and lines obtained by cuttings.

ISSR markers detected the highest levels of polymorphism (Joshi and Dhawan, 2007; Sreedhar et al., 2007), that is, they provide the highest level of discrimination between any pair of genotypes (Powell et al., 1996). Devarumath et al. (2002) found that ISSR fingerprinting detected more polymorphic loci (12.8%) than RAPD fingerprinting (4.28%) in micropropagated tea clones. The number of marker bands amplified per cultivar (ranged from 102 to 127) and percentages of polymorphic bands (ranged from 53.21% to 77.17%; Table 2) in our study is sufficient to reveal variations. This is evidenced by the comparable number of bands scored in various plant taxa by employing ISSR-based marker assays. Kumar et al. (2009) used 64 bands with 98.14% polymorphism to reveal genetic diversity among 12 Jatropha species, and Carrasco et al. (2009) used 114 bands with 55.3% polymorphism to reveal genetic diversity among 333 samples of Vasconcellea pubescens.

Between mother plant and lines obtained by cuttings of the four Hippophae cultivars, low molecular variations were also detected by ISSR analysis (Table 2, Fig. 3) and there are some differences in morphological characteristics (Table 1). These indicate that multiplication by cuttings of H. rhamnoides affect genotypes to some degree. Cluster analysis (Fig. 3) provides some evidence for cutting induced variation. In fact when we look at the genetic distance between individuals within a cultivar we see that it is high relative to the total genetic variation (up to half the total Jaccard coefficient value). However, variation induced by cuttings occurs within cultivars. This is because all mean genetic similarities (Jaccard coefficient) among mother plant and lines obtained by cuttings for each cultivar in this study were more than 0.78, which is the upper limit of genetic similarities (ranged from 0.29 to 0.78) among 15 different Hippophae cultivars. At a Jaccard coefficient of 0.78, fifteen Hippophae cultivars were independent of each other and not clustered into any group (Ruan and Li, 2005), and in this study, mother plant and lines obtained by cuttings for ‘YLSD’ were group into one cluster, as well as ‘WHSJ’, ‘HMXZ’, and ‘ZGWCXZ’.

It is generally recognized that somaclonal variation mainly occurs in tissue culture (Huang and Liu, 2006; Matthes et al., 2001; Palombi and Damiano, 2002) and that there are typically no problems concerning genetic stability when multiplying cultivars by cuttings. Many woody species are commercially vegetatively reproduced, but some variation is occasionally found in lines obtained by cuttings, such as plantlets of Sansevieria trifasciata.

Transposition in response to environmental stress is a common phenomenon by which plants acquire genetic diversity (Flavell et al., 1994; Mcclintock, 1984). Genetic variation in the scion plants of Vigna radiata seems not to be caused simply by DNA transformation, but most likely the nutrient and hormone stress induced mutation (Zhang et al., 2002). Some morphological and molecular variations between the mother plant and lines obtained by cuttings for Hippophae used in this study may be related to the cutting process, because the bases of cuttings were often dipped for 1 hour in a mixed 5,000 ppm solution containing indole butyric acid, indoleacetic acid, naphthylacetic acid, and rooting reagent before planting, which may be induce some variation (Phillips et al., 1994).

Information generated from this study can be used to select parents for hybrid development to maximize desirable agronomic traits in a breeding program aimed at developing segregating populations to map genes controlling special traits in sea buckthorn. The use of hybrids between some species in exotic conditions frequently results in significant genetic gains possibly due to either true heterotic effects and/or a combination of traits. The resulting information on genetic relationships can be used as a guideline for the development of crossing strategies in a breeding program (Le Thierry d' Ennequin et al., 2000). Genotype ‘ZGWCXZ’ is adaptable to harsh environments and is fast growing, and ‘YLSD’ has promising characteristics (e.g., large fruits, long fruit-stalks, and high vitamin-C contents in seeds and leaves). Crossing between them (Jaccard coefficient between ‘ZM’ and ‘YM’  =  0.44) may be useful in breeding specific lines with each parent's desirable traits. Similarly, crossing between ‘WHSJ’ and ‘ZGWCXZ’ (Jaccard coefficient between them  =  0.50) may be useful in breeding of advanced lines with promising characteristics, such as fruit without stones, fast growth, and strong adaptability to harsh environments. Dried-shrink disease is the main problem restricting the sustainable development of Hippophae, and molecular biotechnology may be able to provide a solution to overcome this problem (Ruan et al., 2007). ‘YLSD’ and ‘HMXZ’ can be used in breeding programs aimed at developing cultivars with resistance to dried-shrink disease. This is because ‘YLSD’ is resistant to dried-shrink disease, but ‘HMXZ’ is susceptible to this disease.

ACKNOWLEDGEMENTS

This research was funded by the Keystone Project of the Agriculture Science and Technology Research Item of the Liaoning Scientific and Technological Committee (No. 2007207005) and the preparative fund of Dalian Nationalities University (No. 0212-112012).