Determining Ploidy Level and Nuclear DNA Content in Rubus by Flow Cytometry

Nuclear DNA flow cytometry was used to differentiate ploidy level and determine nuclear DNA content in Rubus. Nuclei suspensions were prepared from leaf discs of young leaves following published protocols with modifications. DNA was stained with propidium iodide. Measurement of fluorescence of 40 genotypes, whose published ploidy ranged from diploid to dodecaploid, indicated that fluorescence increased with an increase in chromosome number. Ploidy level accounted for 99% of the variation in fluorescence intensity (r2 = 0.99) and variation among ploidy levels was much higher than within ploidy levels. This protocol was used successfully for genotypes representing eight different Rubus subgenera. Rubus ursinus Cham. and Schldl., a native blackberry species in the Pacific Northwest, which has been reported to have 6x, 8x, 9x, 10x, 11x, and 12x forms, was extensively tested. Genotypes of R. ursinus were predominantly 12x, but 6x, 7x, 8x, 9x, 11x, and 13x forms were found as well. Attempts to confirm the 13x estimates with manual counts were unsuccessful. Ploidy level of 103 genotypes in the USDA-ARS breeding program was determined by flow cytometry. Flow cytometry confirmed that genotypes from crosses among 7x and 4x parents had chromosome numbers that must be the result of nonreduced gametes. This technique was effective in differentiating chromosome numbers differing by 1x, but was not able to differentiate aneuploids. Nuclear DNA contents of 21 diploid Rubus species from five subgenera were determined by flow cytometry. Idaeobatus, Chamaebatus, and Anaplobatus were significantly lower in DNA content than those of Rubus and Cylactis. In the Rubus subgenus, R. hispidus and R. canadensis had the lowest DNA content and R. sanctus had the highest DNA content, 0.59 and 0.75 pg, respectively. Idaeobatus had greater variation in DNA content among diploid species than the Rubus subgenus, with the highest being from R. ellipticus (0.69 pg) and lowest from R. illecebrosus (0.47 pg). makes the results of heteroploid crosses unpredictable (Thompson, 1997). Fertility is often poor if the hybrid is triploid, pentaploid, or an aneuploid with a chromosome number less than hexaploid. At higher ploidy levels, the odd euploid and aneuploid genotypes may be completely fertile (Lawrence, 1986; Waldo, 1950). In blackberry and raspberry breeding programs, knowing the ploidy level of Rubus genotypes is essential to predict crossing success and parental combinations that might produce problematic progenies. For germplasm enhancement, ploidy level can also serve as a valuable distinguishing taxonomic trait when collections of Rubus are evaluated (Thompson, 1995a, 1995b). In the Pacific Northwest, the native R. ursinus Cham. & Schldl. is widely distributed and has been a valuable genetic resource for developing trailing blackberry cultivars (Finn et al., 1997). Rubus ursinus has a known ploidy level range from hexaploid to dodecaploid except for septaploid (Brown, 1943). In 1993, R. ursinus was collected from throughout the Pacific Northwest and established in a common garden in Corvallis, Ore. (Anderson and Finn, 1996). These populations were evaluated for horticultural and taxonomic characteristics and superior individuals were identified. To incorporate this material most effectively into erect, semierect and trailing blackberry breeding germplasm, it would be useful to know the ploidy level of the selected genotypes. Successful chromosome counts have been made on at least 387 Rubus species, ≈40% of the known species in the genus, and 90 cultivars and selections (Thompson, 1995a, 1995b, 1997). Ploidy level in Rubus genotypes has usually been determined by chromosome counts in meristematic tissues, such as root or shoot tips, or pollen mother cells. However, these approaches can be Received for publication 2 Oct. 2001. Accepted for publication 14 May 2002. We thank Corwin Willard (Oregon State University) for his instruction in the use of the flow cytometer; Kathiravetpilla Arumuganathan (University of Nebraska) and Maxine Thompson (Corvallis, Ore.) for their helpful advice in the early stages of this work and the USDA–ARS National Clonal Germplasm Repository (Corvallis, Ore.) for providing plant material. Rubus is a large genus that includes an estimated 900 to 1000 species that are widely distributed in various ecosystems (Thompson, 1997). A number of commercial crops are members of this genus including red and black raspberries (R. idaeus L. and R. occidentalis L.) and blackberries (Rubus sp. L.). The naturally occurring range of chromosome numbers in Rubus species is from 2n = 2x = 14, the diploid state, to 2n = 14x = 98 or possibly 2n = 18x = 126, including odd-ploids and aneuploids (Moore, 1984; Thompson, 1995a). Rubus has 12 subgenera but the species and cultivars used in blackberry and raspberry breeding have largely been from the Rubus or Idaeobatus subgenera, respectively. The Rubus subgenera range from diploids to dodecaploids (Moore, 1984). Presumably, the present day species and cultivars have arisen primarily from the intercrossing of diploid species and occasionally polyploid species (Jennings, 1988). Genotypes representing many different Rubus ploidy levels have been used in blackberry breeding and released as cultivars. Heteroploid crosses are often used to introduce desirable genes from wild species or from elite material of different species or crop types, and can usually generate progenies with varying ploidy levels. Factors such as spontaneous doubling of chromosomes at an early stage of development or the union of a nonreduced gamete (produced by nonreduction of sporocytes during meiosis or reduction of polyploid premeiotic cells that appear nonreduced) with a reduced gamete, both of which are common in Rubus,