In Vitro Germination and Dormancy Responses of Hydrangea macrophylla and Hydrangea paniculata Seeds to Ethyl Methane Sulfonate and Cold Treatment

To determine the optimal conditions for mutagenesis of Hydrangea macrophylla (Thunb.) Ser. and Hydrangea paniculata Siebold, cold-treated and untreated seed from representative cultivars were exposed to 0.5%, 2.5%, and 5.0% ethyl methane sulfonate (EMS). Most untreated H. macrophylla seed exposed to 2.5% and 5.0% EMS had substantially lower germination percentages. H. paniculata seed germination percentages did not differ from controls until EMS concentrations reached 5.0%. Cold treatment of H. macrophylla and H. paniculata seed made germination more tolerant to all concentrations of EMS tested and increased germination of most H. macrophylla cultivars when compared with lower dosages. Germination of cold-treated seed from H. paniculata cultivars rose substantially above control levels even at the highest dosages. For the most part, we observed more nondormant seed, or seed available for germination, but less viability with increasing EMS dosage in untreated Hydrangea L. seed versus controls. In contrast, cold-treated seed displayed higher levels of mutagen tolerance and nondormancy versus controls when exposed to these same doses of EMS. Ethyl methane sulfonate (EMS) is a mutagen that chemically alters DNA through random reaction and modification of oxygen present in nucleotide bases (such as the O6 of guanine and O4 of thymine) as well as in DNA phosphate groups (Sega, 1984). This, in turn, causes mismatched base pairing to occur. Although EMS is capable of causing double strand breakage in higher organisms (Hoffmann, 1980; Sega, 1984), almost all erroneous DNA repair of EMS damage in plant genomes yields transition mutations with the original G/C pair being replaced by an A/T pair (Greene et al., 2003). In spite of the production of similar mutation types during EMS exposure, the rates of chemical mutagenesis are known to differ significantly among plant species (Filkowski et al., 2004; van Harten, 1998; Waugh et al., 2006) and even within species (Henikoff and Comai, 2003). As a result, proper application of chemical mutagens like EMS often requires protocols to be independently established for targeted species, cultivars, and ecotypes. Alkylating agents like EMS have been used to increase allelic diversity in plants and generate breeding lines (see http://wwwmvd.iaea.org for a current list of reported mutant varieties) for more than 4 decades (see van Harten, 1998 for historical review and application). Recent reverse genetic advances such as TILLING (Targeting Induced Local Lesions IN Genomes, Henikoff et al., 2004; McCallum et al., 2000; Till et al., 2003) have broadened the application of EMS to plant science, empowering scientists to screen for specific mutant alleles in the absence of a mutant phenotype, even when very little genome sequence information exists (Henikoff and Comai, 2003). In addition to providing functional genomic information, genetic variations created by chemical mutagenesis are exempt from regulatory approval requirements for transgenic crops. Hydrangea cultivars are among the topselling deciduous flowering shrubs in the United States; traditional breeding efforts for this genus are focused on the development of desirable and divergent ornamental traits (Reed, 2000a, 2000b; Rinehart et al., 2006). To our knowledge, there are no published parameters for EMS application to H. macrophylla or H. paniculata and there are no published reports as to the relative effectiveness of this mutagen in these species. In addition, nothing is known about the effects of EMS on the viability, dormancy, or germination of Hydrangea seeds. We performed a series of experiments using increasing EMS dosage in an effort to estimate the range of damage tolerated by Hydrangea seeds and to better understand variation within individual cultivars regarding viability, dormancy, and germination. We defined viability in terms of whether a seed was alive or dead, dormancy as a physiological state in which the germination processes of viable seed are inhibited, and germination as the appearance of a radicle from the seed of any size. Materials and Methods Seed collection. Commercial cultivars of H. macrophylla subsp. serrata Makino ‘Blue Bird’, ‘Beni Gaku’, ‘Intermedia’, ‘Omacha’, and H. macrophylla subsp. macrophylla ‘Coerulea’, ‘Lady in Red’, ‘Nikko Blue’, ‘Seafoam’, ‘Tokyo Delight’, and ‘Veitchii’ were obtained from Amethyst Hill Nursery (Aurora, OR) and were grown outside under 50% shade at the USDA-ARS Southern Horticultural Laboratory in Poplarville, MS, or at the Mississippi State Truck Crops Branch Experiment Station in Crystal Springs, MS. Whole infructescences from these open-pollinated H. macrophylla plants were collected in October and November of 2007 and seed were harvested by crushing seed pods manually between fingers or mechanically grinding them with the aid of an electric food chopper (Black and Decker Corp., Towson, MD). After harvesting, seed were initially separated from larger particulates by repeated passage through a tea strainer. Seeds and smaller particulates were then rolled around on loose paper to separate seed from the finer particulates, which clung more fastidiously to the paper. Open-pollinated seed from H. paniculata ‘Big Ben’ and ‘Pink Diamond’ were obtained from Dr. Sandy Reed at the Tennessee State University Nursery Crop Research Station in McMinnville, TN. Dried whole infructescences were removed and shaken to release seed from seed pods. Seeds obtained from these plants were large and relatively clean so no further processing was needed. All collected seed were stored at 24 C in ambient relative humidity, ranging from 55% to 70%, until their use in experiments. Cold treatment, chemical treatment, and sterilization of seed. Seed used in experimental studies were first aliquoted into polypropylene microtubes and cold-treated at 4 C for 6 weeks under dry conditions. Untreated seed used in this study were kept under identical dry conditions at 24 C. One day before plating, all seed were imbibed with agitation in solution containing 300 ppm of Trichlor (Trichloro-s-triazinetrione; Pool Time Products, Buford, GA), 1300 ppm gibberellic acid (GA3; PhytoTechnology Laboratories, Shawnee Mission, KS), and 1000 ppm KNO3 (Acros Organics, Morris Plains, NJ) for 24 h. For seed used as controls, the resultant supernatant (often containing a mixture of dissolved dirt and washed off debris) was decanted and a second wash of 2500 ppm Trichlor was applied to seed with agitation for 10 to 15 min and subsequently decanted. For seed that were chemically mutagenized, the initial supernatant was decanted and a solution of 0.1 M potassium phosphate buffer (Sigma-Aldrich, St. Louis, Received for publication 6 Oct. 2008. Accepted for publication 12 Feb. 2009. To whom reprint requests should be addressed; e-mail [email protected]. 764 HORTSCIENCE VOL. 44(3) JUNE 2009 MO), pH 7.0, containing 0.5%, 2.5%, or 5.0% EMS (Sigma-Aldrich) was added to respective tubes. These tubes were then agitated at 24 C for 3 h after which the EMS solution was decanted into 50% sodium thiosulfate. Seed in each tube were then briefly suspended in 500 ppm Trichlor and the resultant wash decanted again into 50% sodium thiosulfate. A final wash of 2500 ppm Trichlor was applied to seed with agitation for 10 to 15 min and then decanted into 50% sodium thiosulfate. After final washes with 2500 ppm Trichlor, all respective seed, controls as well as seed exposed to EMS, were suspended in a solution containing appropriate volumes of 0.2% agar and 1000 ppm Trichlor. One hundred to 500 sterilized seeds were sown on plates and three randomized replicate plates were used per treatment. Incubations of seed were done on solid media within 100 · 15-mm petri dishes (VWR International, West Chester, PA) that contained 0.5% Phytoblend agar (Caisson Laboratories, North Logan, UT), 0.2% Plant Preservative Mixture (PPM; Plant Cell Technology, Inc., Washington, DC), 0.5 · Gamborgs B5 + vitamins (PhytoTechnology Laboratories; Gamborg et al., 1968), and 130 ppm GA3 (final concentration). Light conditions for experimental treatments. Plated seed were placed under 40-W Sylvania Gro-Lux lights (Danvers, MA) for 21 d at 24 C with red cellophane completely covering the lights; approximate visible wavelength emission ranged from 610 to 750 nm as measured by a Perkin Elmer (Waltham, MA) Lambda3B spectrophotometer. Photosynthetic photon flux was measured with an Apogee QMSW-5S quantum meter (Apogee Instruments Inc., Logan, UT) and was 15 mmol m s levels for lights covered with red cellophane. Illuminance, as measured by a Control Company Model #4332 light meter (Control Company, Friendswood, TX), was 650 lx for red (cellophane-covered) light reaching the plates. For all experiments, seed were subject to a 16:8 light:dark cycle. Seed staining and image acquisition. After 21 d of incubation, total numbers of germinated seed on plate were manually quantified with the aid of an Olympus CO11 Stereo microscope (Olympus America Inc., Center Valley, PA). Seed were counted as germinated if the seedcoat had been clearly broken and a radicle of any size had emerged. After germination counts were recorded for each plate, 10 mL of freshly made 0.2% 2,3,5-triphenyl-2H-tetrazolium chloride (TTC; Alfa Aesar, Ward Hill, MA) was added to each plate (enough to completely cover and submerge all the remaining seed) and each plate was immediately put in complete darkness for 48 h. Each plate was then photographed at 7.1 megapixel resolution (Fig. 1) using optimal ambient light conditions with a Canon Powershot A560 digital camera (Canon USA Inc., Lake Success, NY). Image processing and seed quantification. Digital image of plates were processed with the aid of Picasa software (Google Inc., Mountain View, CA) and ImageJ (Abramoff et al., 2004) to count the total number of seeds on a plate and the number of seeds stained red. In short, images were first tuned in Picasa by turning highlights up to 100%, shadows up to 50%, and adjusting fill light until all the contaminating artifacts of the petri dish in the picture were minimized. This resultant image, which shows primarily only seed and some contaminating particulate matter, was then split into three separate but different image reproductions with the aid of Picasa’s filtered black and white effect feature. To generate the first reproduction, green color was filtered from the image. In this image, all seed (dormant and nondormant) as well as particulate matter were darkened and visually detectable. A second image reproduction was generated by filtering all the green and yellow color out of the initial colored image. In this image, germinated, dormant, or dead seed were no longer visually detectable, and only nondormant (red) seed and contaminating particles were visible. A third image reproduction was generated by filtering all the green, yellow, and red color out of the initial image. The only visible constituents of this resulting image were contaminating particles present within the seed mixture, most often dirt or pieces of seed pods. ImageJ’s ‘‘Analyze Particles’’ feature was used to quantify the number of seed-sized particles in each image. We calculated that the average pixelized area of H. macrophylla seeds measured 150 to 200 pixels ( 0.25 to 0.40 mm) and the pixelized average area of H. paniculata seeds ranged from 350 to 900 pixels ( 0.65 to 1.0 mm). Using identical binary threshold sensitivities for each image, we quantified H. macrophylla or H. paniculata seed by slightly expanding pixel area cutoff parameters so that particles ranging from 50 to 300 pixels were tabulated on H. macrophylla plate images and a range of 300 to 1200 pixels was used to tabulate H. paniculata. The number of particles identified by ImageJ from the first reproduction of each plate image included all seed (dormant and nondormant) as well as the contaminating particulates, which were visible. The number of particles counted in the second reproduction of a plate image (showing particulates and TTC-stained seed) was subtracted from the first yielding a total number of dead or dormant (yellow) seed that were present initially on a sample plate. Subtracting the number of particles in the third image, which contained only contaminant particles, from the particles counted in the second image yielded the total number of red (viable and nondormant) seed that were present on each sample plate. Thus, the total number of yellow and unstained (dormant or dead) seeds for each sample was calculated subtracting the number of particles in image 2 from image 1, and the total number of TTCstained, nondormant seeds for each sample was calculated subtracting the number of particles in image 3 from the particles counted in image 2. Calculations and statistics. Percentage of nondormant seed for each plate was calculated by adding the numbers of germinated and red seed and dividing the sum by the total of red (R), yellow (Y), and germinated (G) seed: (G + R)/(G + R + Y). Percent germination of nondormant seed on a plate was calculated by dividing the number of germinated seed by the sum of the germinated seed and the red seed on a plate: G/(G + R). The germination index, or total germination expected per amount of total seed, was calculated by multiplying percent nondormant seed by percent germinated, nondormant seed. Univariate analysis of variance (ANOVA, a = 0.05) and Ryan-Einot-Gabriel-Welsch Range post hoc tests (a = 0.05) were conducted with SPSS software (SPSS Inc., Chicago, IL) to analyze significance. When transformation was required to meet assumptions of normality or equal variances, data sets were Tukey normal scored using the formula (r – 1/3)/(w + 1/3), where r is the rank and w is the sum of the case weights (Tukey, 1962). Eta-square (h) and Type II error statistics (b) statistics were also calculated for each ANOVA to verify variance contribution and the power of our statistical tests (Sokal and Rohlf, 1995).