Curators of plant genetic resources collections must preserve germplasm possessing known useful characteristics as well as material displaying general genetic diversity. In order to ensure that both types of germplasm are included in a collection, germplasm curators require three fundamental types of information about each accession: taxonomic identity, genetic identity, and genetic relationship. Because simple sequence repeat DNA fragments (SSRs) have been successfully used to determine the genetic identity of grape clones, we conducted a study to determine if SSRs would supply all three types of information for the accessions in the cold-hardy Vitis (grape) germplasm collection. SSR fragments were amplified at six different loci for 23 accessions of cold-hardy grape spanning the range of species diversity in the collection. The minimum number of different alleles found at a locus was 9; the maximum was 26. Heterozygosity values ranged between 0.565 and 0.783, while gene diversity values were in the range 0.785 to 0.944. Two hundred fifty-two pairs of plants out of a possible 253 could be distinguished by their SSR profiles. Nei's genetic identities were computed between all pairs of plants and used in a UPGMA cluster analysis. The relationships obtained did not correspond well to expected relationships based on geography and taxonomy. Four species of grapes were represented by two or more accessions in this study. No DNA fragments found at these six loci served to unambiguously distinguish one species from another. Thus, SSR fragments from the six loci studied were useful in determining genetic identity of accessions, but were not helpful in determining genetic relationships or taxonomic identities. We are searching for additional loci that are informative for these types of information. Meanwhile we highly recommend SSRs for determining genetic identity in germplasm resources collections.
Simple sequence repeats (SSRs) are highly polymorphic regions of DNA that can be used for the molecular characterization of apple (Malus) germplasm. SSR markers are sufficiently variable to distinguish between individual plants in wild Malus species. In this study, accessions of Malus hupehensis were screened for fragment length variation in PCR amplified simple sequence repeat regions of DNA. The fragment length phenotype produced by five SSR primer pairs showed no variation between two lineages of M. hupehensis collected in the Changjiang (Yangtse) River valley. One lineage was collected by E.H. Wilson in 1908 near the city of Ichang, Hubei Province. The second lineage was collected by cooperators at China's Southwest Agricultural University (SWAU) in 1997 near the city of Chongqing (Chungking). M. hupehensis Plant Introduction No. 588760 from the National Plant Germplasm System lacks provenance, but displays a fragment length phenotype identical to both the Wilson and SWAU lineages. The spread of a clone may be aided by asexual reproduction through seed, which is not uncommon in polyploid apples. Two seedlings each of 15 maternal trees from the SWAU lineage were assayed for ploidy level by flow cytometry. The DNA content per nucleus for all SWAU progeny fell within the range for triploids, 2.19 to 2.68 pg DNA/nucleus. It appears that plant explorers in China separated by almost 90 years have succeeded in sampling a single clonal lineage of M. hupehensis.
Molecular DNA markers based on the RAPD (random amplified polymorphic DNA) assay are gaining use in germplasm assessment. RAPD markers are simple, relatively inexpensive, and highly informative. We used five primers to assess 26 Brassica oleracea breeding lines from the IVF and nine accessions from the PGRU. The test array included eight subspecies of B. oleracea. We generated 90 RAPD markers and were able to unambiguously discriminate among all 35 test entries, but could not separate subspecies within B. oleracea. Genetic similarity between subspecies ranged from 0.629 to 0.738. Average similarity within accessions was 0.96, confirming the suspected homogeneity of breeding lines. Nevertheless, significant genetic diversity was found among kohlrabi, broccoli, and cabbage accessions. Similarity analysis of breeding lines and hybrids confirmed their pedigree relationships. Interestingly, B. o. subsp. costata `Couve Nabica' showed closer similarity to B. napus subsp. oleifera `Jet Neuf' than to other B. o. materials and B. o. subsp. italica `Packman' showed higher similarity to some cabbages than to other broccolis. Results provide further evidence that diversity assessment using RAPDs is broadly applicable and useful in germplasm conservation and utilization.
The U.S. National Plant Germplasm System (NPGS) currently holds 36 separate accessions of the `Yichang' clone of Malus hupehensis (Pamp.) Rehd. The `Yichang' clone originally entered the United States in 1908 as seed collected for the Arnold Arboretum by E.H. Wilson near Yichang, Hubei Province, China. The original description of M. hupehensis omits fruit characters, and botanists frequently augment these omissions with descriptions of the `Yichang' clone. Apomixis occurs in Malus, including M. hupehensis, and is strongly associated with elevated ploidy levels. Simple sequence repeats (SSRs) were used to characterize 65 accessions of M. hupehensis. To check for polyploidy, a set of M. hupehensis accessions was evaluated with flow cytometry. The simple sequence repeat phenotypes and ploidy information revealed the `Yichang' clone under various accession names in arboreta. It was neither known nor suspected that the U.S. National Plant Germplasm System held many duplicate accessions of the `Yichang' clone prior to their molecular characterization. Germplasm conservation decisions for Malus species can benefit from an increased knowledge of the genetic variation or lack thereof in naturalized populations and ex situ collections.
We discuss a series of studies in our units employing molecular genetic markers in collection management, primarily for identity and diversity assessment and partitioning of genetic variation. Isozymes, random amplified polymorphic DNAs (RAPDs), and simple sequence repeat DNAs (SSRs) have been used for these purposes. We analyzed a range of Brassica oleracea accessions at six isozyme loci. Unique isozyme profiles (or fingerprints) were found for 40% of the individual genotypes within accessions. While isozymes were extremely valuable for partitioning genetic variability between and among subspecies, they failed to identify accessions and subspecies. Furthermore, relationships found did not correspond to those predicted by taxonomy. In a study of three species of Chinese vegetable brassicas using 112 RAPD markers, we were able to unambiguously distinguish all 52 accessions studied, despite some intra-accession variability. In addition, cluster analysis correctly grouped all individuals of the same species, but below that rank, taxonomic groupings occasionally broke down. RAPD profiles were found that unambiguously distinguished the three Brassica species from one another, but, for subspecies, no such profiles were found. In another RAPD study of B. oleracea subsp. capitata (cabbage), a closely related set of cultivars were not distinguishable, although more distantly related cultivars were. We had disappointing results with a RAPD study of Vitis accessions. DNA was extracted from the leaves of 23 greenhouse-grown and 52 field-grown vines. Twelve of the 23 greenhouse vines were rooted cuttings collected from 12 of the field-grown vines. Unfortunately, the RAPD profiles of all vines grown in the same location (whether greenhouse or vineyard) were more similar to one another than were profiles from the same clone grown in the two different locations. We are studying whether this result is due to physiological differences in plants growing under different conditions, to differences between PCR reagent lots, to pathogen infestation, or to DNA sample contamination. In a study of 23 accessions representing 15 Vitis species and three species hybrids, we used six different SSR markers to identify individual genotypes. We were able to unambiguously distinguish all genotypes, except two that were identical at all six loci. Review of planting records revealed that the two genotypes were probably the same grape clone. SSR results were not congruent with known taxonomic relationships or geographic origin of genotypes. The SSR polymorphisms found in even this small subset of the Vitis collection in principle make possible the identification of more than 130 trillion different genotypes. This high level of polymorphism, however, makes our particular SSR loci of limited use for identification of species and for the determination of genetic relationships. Molecular genetic markers offer a powerful, efficient approach to assessing questions of identity, relationship, and diversity in germplasm collections, but markers need to be selected based on their suitability for the particular task.
To comprehend genetic identity and relatedness in Malus germplasm held in situ and ex situ, we are employing simple sequence repeat (SSR) DNA fragment information in combination with passport and horticultural data. SSRs offer certain advantages for characterizing large arrays of germplasm efficiently. They are abundantly dispersed throughout plant genomes and are exceedingly polymorphic. In addition, they can be PCR-amplified and detected by automated fluorescence-based technology. A size-fractionated DNA library of M. ×domestica cv Golden Delicious was screened to identify SSR loci. Eight loci were found to be reliably informative and were used to prepare locus-specific primer pairs. Characterization of the 75 M. ×domestica accessions included in the core subset of the USDA-ARS Malus germplasm collection revealed six of the eight loci were polymorphic within the array. The number of alleles per locus ranged from two to 21. Throughput was enhanced by multiplexing, allowing simultaneous use of two or three primer pairs. With improved genetic characterization of Malus germplasm, we intend to better develop and relate the core subset to the rest of the collection and to in situ Malus genetic resources. SSR markers appear to be an efficient and reliable tool to expedite this process.
The USDA–ARS active collection of Malus includes over 2500 accessions maintained as field-grown trees at the Plant Genetic Resources Unit (PGRU), Geneva, N.Y. Nearly 30% of this collection is presently cryopreserved as dormant buds at the National Seed Storage Laboratory, Fort Collins, Colo., as a backup security collection. Successful bud-grafting recovery rates (≥40%) after one to four years of cryogenic storage have been documented for over 675 of 750 accessions tested. However, current protocols dictate budwood collection at PGRU from late December through early March, when buds are thought to be optimally acclimated for desiccation and slow freezing to –30°C, our pretreatment for cryopreservation. This causes a processing bottleneck. Our observations suggest temporary storage of budwood at –4°C after field harvest is possible, but we had not tested this directly. Therefore, we collected budwood from four accessions representing different levels of cold tolerance on six dates from January to March, 1995. Dormant buds were processed for cryopreservation monthly after storage in sealed bags at –4°C for 1 to 6 months. Recovery rates ranged from 55% to 100%. Neither collection date nor length of storage at –4°C affected rate of recovery. These results suggest we can significantly increase the throughput and efficiency of our cryopreservation efforts, thereby enhancing management and security of the Malus collection.
One of the primary progenitors of the cultivated apple (Malus ×domestica) is M. sieversii, a species native to the forested regions of Central Asia. Despite the horticultural importance of M. sieversii, little is known about its genetic variation. In this study, isozyme diversity at 18 loci was determined for 259 open-pollinated offspring belonging to 31 different maternal half-sib families collected from 14 different populations in 4 regions of central Asia. Genetic diversity statistics were computed from the resulting allele and phenotype frequencies. Cluster analysis of half-sib families showed that there was some grouping based on geographic region, but 16 of the half-sib families were most closely related to half-sib families from other regions. AMOVA, the analysis of molecular variance, indicated that most of the enzyme variability (85%) was attributable to differences among half-sib families within populations, none could be assigned to populations within regions, and 15% was due to differences among regions. In addition, no alleles were found that were both fixed in a region and unique to that region. These results suggest that plants belonging to M. sieversii effectively form a single panmictic population. Thus, a thorough sampling of a few large populations will efficiently capture most of the genetic diversity present in wild M. sieversii.
Simple-sequence repeats (SSRs) are efficient and informative DNA markers with great potential for germplasm characterization. When used to characterize large arrays of accessions, such as the core subset of the USDA/ARS Malus collection, SSRs may be more effective than other approaches, such as restriction fragment length polymorphism (RFLP) and random amplified polymorphic DNA (RAPD). For example, SSRs can be PCR-amplified and fluorescence-based detected; they also appear to be abundantly disbursed throughout plant genomes and yield abundant polymorphisms in most taxa studied. We are conducting an extensive screening of a size-fractionated library of Malus ×domestica cv. Golden Delicious to identify and characterize selected SSR loci. We are applying genetic information revealed by SSR loci in combination with passport and horticultural data to better comprehend genetic identity and relatedness in Malus germplasm collections and help develop the Malus core subset. Ultimately, application of molecular marker data will permit improved conservation and use of genetic resources.