Genetic Diversity among Forage Bermudagrass ... - PubAg - USDA

Genetic Diversity among Forage Bermudagrass (Cynodon spp.): Evidence from Chloroplast and Nuclear DNA Fingerprinting Mehmet Karaca, Sukumar Saha, Allan Zipf, Johnie N. Jenkins, and David J. Lang* ABSTRACT

for appropriate genetic analysis. The high value of bermudagrass as a forage crop clearly justifies new and innovative approaches toward evaluating and understanding the genetic relationships among different lines of the Cynodon species. DNA-based fingerprinting technologies have been applied in genetic studies in a wide range of plant species, including turf-type bermudagrass (Cato and Richardson, 1996; Caetano-Annolles et al., 1997; Arcade et al., 2000). An integrated study using different DNA fingerprinting techniques will be beneficial for the genetic improvement of forage bermudagrass. Four PCRbased DNA fingerprinting techniques including AFLP (Vos et al., 1995), CpSSRLP (Weising et al., 1998), RAPD (Costa et al., 2000), and DAMD (Bebeli et al., 1997) were used to evaluate genetic relationships among selected forage bermudagrass ecotypes and varieties.

Genetic analysis of forage bermudagrasses (Cynodon spp.) lags considerably behind other species, including the turf-type bermudagrasses. This research was undertaken to identify and characterize genetic relationships within and between forage bermudagrass ecotypes and varieties. Genetic relationships within 31 forage bermudagrass genotypes were determined by means of 15 amplified fragment length polymorphism (AFLP), 10 chloroplast-specific Simple sequence repeat length polymorphism (CpSSRLP), 10 random amplified polymorphic DNA (RAPD), and 10 directed amplification of minisatelliteregion DNA (DAMD) primers or primer pairs. The unweighted pair group method, using arithmetic averages (UPGMA) and the bootstrap analyses with 2000 replications, were used to calculate the relationships. Overall results indicated that forage bermudagrass genotypes have a narrow genetic base, with genetic similarity (GS) ranging from 0.608 to 0.977. The most genetically similar forage bermudagrass lines were ‘Tifton 78 WH’ and ‘Tifton 78’ (GS ⫽ 0.977), whereas ‘McDonald’ and ‘Alicia’ were the most genetically diverse bermudagrass lines (GS ⫽ 0.608). ‘Sumrall 007’, ‘Tanberg’, ‘Maddox’, McDonald, ‘Holly Springs’, ‘Murphy’, ‘Murphy II’, and ‘Stallings’ were distantly related to known varieties. The GS values of these genotypes were less than 0.85 compared with any other genotypes, indicating that these ecotypes were unique in their genome structure and provided genetic justification for their release as varieties. Close genetic relationships between some ecotypes and varieties were detected as follows: ‘Prairie I’, ‘Prairie II’, and ‘Prairie III’ to ‘Grazer’; ‘Lott I’, ‘Lott II’, and ‘Lancaster’ to ‘Callie’; and Tifton 78 WH to Tifton 78. Although DNA markers could differentiate these ecotypes, additional molecular, cytological, and phenological evidences are required to further confirm whether they could be considered as new varieties.

MATERIALS AND METHODS Plant Materials Plant materials included established varieties and locally selected ecotypes. An ecotype is a reselection based on a distinct phenotype from an existing variety. Forage bermudagrass ecotypes used in this study were first identified either by local farmers or extension specialists from an existing bermudagrass field (Edwards et al., 1993). A total of 31 forage bermudagrass genotypes, including varieties, were obtained from breeders, extension agents and farmers, and maintained in a greenhouse at Mississippi State University, Mississippi State, MS (Table 1).

B

ermudagrass varieties are planted as forage crops more than any other warm-season perennial species throughout the southeastern USA (Burton and Hanna, 1995). However, current investment in genetic improvement in this crop is not commensurate with its impact on the U.S. economy. Bermudagrass breeders frequently encounter three major challenges: (i) many of the so called “ecotypes” are selected from farmer fields with limited pedigree history and source of origin; (ii) limited knowledge is known for the genome of the forage-type bermudagrass, including DNA content, genome size, and ploidy level; and (iii) there is a lack of suitable methods

DNA Extraction Young and healthy plant leaf tissues (5–6 g) of each genotype were collected from at least two pots (vegetative clones) of greenhouse grown plants and mixed, washed in tap water, dried with clean paper towels, and placed in a ⫺80⬚C freezer until processed. An additional 10 plant samples (test or blind samples) were also collected from the greenhouse to verify reproducibility and accuracy of the techniques. Genomic DNAs were extracted by a modified CTAB (hexadecyltrimethylammonium bromide) procedure (Murray and Thompson, 1980; Karaca, 2001).

Mehmet Karaca, Dep. of Field Crops, Akdeniz Univ., Agricultural Faculty, Antalya, 07759 Turkey; David J. Lang, Dep. of Plant and Soil Sciences, Mississippi State Univ., Mississippi State, MS 39762; Sukumar Saha and Johnie N. Jenkins, USDA-ARS, P.O. Box 5367, Mississippi State, MS 39762; Allan Zipf, Dep. of Plant and Soil Science, Alabama A&M Univ., Normal, AL 35762. Mention of trademark or proprietary product does not constitute a guarantee or warranty of the product by the USDA, Mississippi State Univ. and Alabama A&M Univ. and does not imply its approval to the exclusion of other products that may also be suitable. Published as journal article no. J9901 of the Mississippi Agric. and Forestry Exp. Stn. Received 27 Aug. 2001. *Corresponding author ([email protected]).

Genomic DNA Amplification and Analysis Amplified Fragment Length Polymorphism (AFLP) The AFLP Small Plant Mapping Protocol for Genomes, P/N 4303051 (Perkin-Elmer, Norwalk, CT) was adapted with the modification noted. Sixty nanograms of genomic DNA, or negative control without template DNA, in a 3-␮L volume were added into a 0.2-mL PCR tube containing 7 ␮L reaction solution consisting of 1 ␮L MseI and EcoRI adapter pairs (Perkin-Elmer, Norwalk, CT), 1 ␮L 10⫻ T4 DNA ligase buffer [50 mM Tris-HCl (pH 7.8), 10 mM MgCl2, 10 mM dithiothreitol (DTT), 1 mM ATP, 25 ␮g/mL bovine serum albumin (BSA)],

Published in Crop Sci. 42:2118–2127 (2002).

2118

KARACA ET AL.: GENETIC DIVERSITY AMONG FORAGE BERMUDAGRASS

2119

Table 1. Bermudagrass lines studied in the present study. Entry Murphy Murphy II Gillihan Tifton 85 McDonald Prairie III Prairie I Prairie II Stallings Tanberg Dixie I Dixie II H. Springs Hardie Lancaster Maddox Tifton 78 WH Tifton 78 Tifton 44 Poplarville I Poplarville II W. Feeder Coastal Alicia Sumrall 007 Callie Lott I Lott II Grazer Russell‡ P. Rico Common

Type

Source†

Ecotype Ecotype Ecotype Variety Ecotype Ecotype Ecotype Ecotype Ecotype Ecotype Ecotype Ecotype Ecotype Variety Ecotype Ecotype Ecotype Variety Variety Ecotype Ecotype Variety Variety Variety Ecotype Variety Ecotype Ecotype Variety Variety Variety Variety

J. Murphy, Leake County, MS J. Murphy, Leake County, MS Shelby County, TN G. W. Burton, R. N. Gates and G. M. Hill, USDA, Coastal GA L. McDonald, Carthage, MS R. Ivy, Prairie, MS R. Ivy, Prairie, MS R. Ivy, Prairie, MS Weakley County, TN TX L. McDonald, Carthage, MS L. McDonald, Carthage, MS J. Johnson, Holly Springs, MS C. M Taliaferro and W. L. Richardson, Stillwater, OK M. W. Lancaster, Alcorn County, MS Simpson County, MS D. J. Lang, Mississippi State, MS G. W. Burton and W. G. Monson, Tifton, GA G. W. Burton and W. G. Monson, Tifton, GA N. Edwards, Poplarville, MS N. Edwards, Poplarville, MS L. Gordon, Oklahoma City, OK L. Sephens and W. W. Hanna, Tifton, GA C. Greer, Edna, TX G. Sumrall, Monticello, MS V. H. Watson, Mississippi State, MS H. Lott, Holcomb, MS H. Lott, Holcomb, MS G. W. Burton, W. G. Monson and R. Utley, Tifton, GA D. Bice, Russell County, AL Plant mixture of common and NK 37 Mississippi State, MS

† Crop registration dates are published by Karaca et al. (2000). ‡ Excluded from analysis because of plant contamination from other sources.

2 ␮L 0.5 M NaCl, 1 ␮L 1 mg/mL BSA, 1 unit MseI, 5 units EcoRI, 2 units T4 DNA ligase (Promega, Madison, WI), and mixed well. After a brief centrifugation, the samples were incubated for 2 h at 37⬚C in a thermal cycler, after which 190 ␮L of PCR-grade water were added and mixed well. Preselective amplification was performed in a 25-␮L volume containing 1⫻ GeneAmp PCR Gold Buffer (150 mM TrisHCl, pH 8.0, 500 mM KCl), 2.5 to 3.0 mM MgCl2, 0.25 ␮M MseI and EcoRI preselective amplification primers, 0.2 mM each dNTP, 0.5 unit AmpliTaq Gold DNA polymerase (Perkin-Elmer, Norwalk, CT), and 5 ␮L digestion-ligation reaction containing the template DNAs. After an initial hold at 95⬚C for 7 min, amplification was performed for 25 cycles of denaturation 94⬚C for 20 s, annealing at 56⬚C for 30 s, extension at 72⬚C for 2 min with a GeneAmp PCR System 9700 thermal cycler (Perkin-Elmer, Norwalk, CT). Ten microliters of the PCR reaction were added to 90 ␮L PCR grade water or buffer consisting of 10 mM Tris, 0.1 mM EDTA, pH 7.5, mixed well, and stored at ⫺20⬚C until needed. Selective amplification reactions were performed in a 25␮L reaction volume containing the same reactants as in the preamplification reaction, but with 0.25 ␮M MseI primer, 0.1 ␮M 5⬘ fluorescent dye-labeled EcoRI primer (EcoRI-FAM, -JOE, or -NED) and 5 ␮L preamplified product. A total of 15 AFLP primer combinations were used (Perkin-Elmer, Norwalk, CT, Table 2). PCR was carried out in a thermal cycler with the following profile: 7 min hold at 95⬚C, followed by a nine cycle pre-PCR consisting of 20 s at 94⬚C for denaturation, 30 s at 65⬚C for annealing, and 2 min at 72⬚C for extension. Annealing temperatures were reduced 1⬚C each cycle. PCR was continued for 30 more cycles at a 56⬚C annealing temperature with a final extension for 30 min at 72⬚C. Chloroplast-Specific Simple Sequence Repeat Length Polymorphism (CpSSRLP) The CpSSRLP PCR was performed with 80 ng DNA (negative control or known DNA sample as template), 0.15 ␮M

each of either HEX or FAM fluorescent-labeled (forward) and nonlabeled (reverse) chloroplast-specific SSR primers (Table 3, Research Genetics, Inc., Huntsville, AL), 0.2 mM each dNTP, 1⫻ GeneAmp PCR Gold Buffer, 2.5 to 3 mM MgCl2, and 0.5 units AmpliTaq Gold DNA polymerase in a 25-␮L reaction solution. PCR was carried out in a GeneAmp PCR System 9700 thermal cycler (Perkin-Elmer, Norwalk, CT) with the following profile: 7 min hold at 95⬚C, followed by a nine cycle pre-PCR consisting of 15 s at 94⬚C for denaturation, 30 s at 56⬚C for annealing and 2 min at 72⬚C for extension. Annealing temperature was reduced 1⬚C each cycle with the same PCR profiles. The PCR amplification was continued for 30 more cycles at a 47⬚C annealing temperature with a final extension for 30 min at 72⬚C. Table 2. AFLP primer combinations for PCR amplification of bermudagrass DNA samples. Preselective amplification used EcoRI adapter primer with no extension (⫺GAATTC), and MseI adapter primer ⫹ C with no extension (⫺TTAAC). Selective amplification used 5ⴕ EcoRI primer ⫹ 2N (⫺GAA TTCNN). 5ⴕ MseI primer ⫹ 3N (⫺TTAANNN). EcoRI primers E ⫹ 2N AA AA AG AG TG AA AC AG TC TC AA TT TT TA AC

MseI primers M ⫹ 3N CAG CTA CAA CTC CAT CAA CAT CTA CAA CTT CTT CTA CAA CTC CTA

2120

CROP SCIENCE, VOL. 42, NOVEMBER–DECEMBER 2002

Table 3. CpSSRLP primer combinations for PCR amplification of bermudagrass DNA samples. Primers† CCMP1R CCMP1F CCMP2R CCMP2F CCMP3R CCMP3F CCMP4R CCMP4F CCMP5R CCMP5F CCMP6R CCMP6F CCMP7R CCMP7F CCMP8R CCMP8F CCMP9R CCMP9F CCMP10R CCMP10F

Repeat in tobacco chloroplast DNA

Sequence CAGGTAAACTTCTCAACGGA CCGAAGTCAAAAGAGCGATT GATCCCGGACGTAATCCTG ATCGTACCGAGGGTTCGAAT CAGACCAAAAGCTGACATAG GTTTCATTCGGCTCCTTTAT AATGCTGAATCGATGACCTA CCAAAATATTTCGGAGGACTCT TGTTCCAATATCTTCTTGTCATTT AGGTTCCATCGGAACAATTAT CGATGCATATGTAGAAAGCC CATTACGTGCGACTATCTCC CAACATATACCACTGTCAAG ACATCATTATTGTATACTCTTTC TTGGCTACTCTAACCTTCCC TTCTTTCTTATTTCGCAGGAA GGATTTGTACATATAGGACA CTCAACTCTAAGAAATACTTG TTTTTTTTTAGTGAACGTGTCA TTCGTCGCCGTAGTAAATAG

Location

(TA)10

trnK intron

(A)11

5ⴕ to trnS‡

(T)11

trnG intron

(T)13

atpF intron‡

(C)7 (T)10

3ⴕ to rpse2‡

(T)5C(T)17

ORF 77-82

(A)13

atpB-rbcL

(T)6C(T)14

rpl20-rps12

(T)11

ORF 74b-psbB‡

(T)14

rpl2-rps19

† Sequences of chloroplast-specific primer pairs from Weising and Gardner (1998). ‡ Primers showed polymorphism in forage bermudagrass.

Capillary Electrophoresis (CE) Capillary electrophoresis was performed for separation of the amplified products of the AFLP and CpSSRLP techniques. The CE-AFLP technique used the FAM-JOE-NED module and CpSSRLP used the FAM-HEX-NED module. Fluorescently labeled undiluted AFLP (1–2 ␮L) and 1/30-diluted (in sterile water) CpSSRLP (2–3 ␮L) PCR products were mixed in 10 ␮L formamide (AMRESCO Inc., Solon, OH) with 0.2 ␮L of an internal size standard DNA labeled with a ROX dye, denatured at 95⬚C for 5 min, kept on ice for at least 4 min and loaded onto the automated ABI PRISM 310 Genetic Analyzer (PE Applied Biosystems, Foster City, CA). Variation in the amplified products was compared relative to the internal size standard DNA. Computer-assisted analysis of the data was performed by GeneScan software (PE Applied Biosystems, Foster City, CA). Additional running parameters in the ABI 310 system were: injection time 10 to 15 s, electrophoresis voltage 13 kV; injection voltage 15 kV; collection time 26 min; run temperature 60⬚C; and syringe pump time 180 s. Directed Amplification of Minisatellite-Region DNA (DAMD) The DAMD amplification reaction mixture (12.5 ␮L) contained 1⫻ GeneAmp PCR Gold Buffer (150 mM Tris-HCl, pH 8.0 500 mM KCl), 2.5 to 3.0 mM MgCl2, 0.2 to 0.4 ␮M primers (Table 4), 0.2 mM each dNTP, 0.5 units AmpliTaq Gold DNA polymerase (Perkin-Elmer, Norwalk, CT) and 80

to 100 ng of template DNA, negative control or blind DNA sample. The PCR profile was a 10 min hold at 95⬚C, followed by a nine cycle pre-PCR consisting of 15 s at 94⬚C for denaturation, 30 s at 56⬚C for annealing and 2 min at 72⬚C for synthesis. Annealing temperatures were reduced 1⬚C for each cycle. PCR was continued for 30 more cycles at a 47⬚C annealing temperature with a final extension for 30 min at 72⬚C. The samples were stored at ⫺20⬚C until needed. Random Amplified Polymorphic DNA (RAPD) The 25 ␮L of amplification reaction mixture contained 1⫻ GeneAmp PCR Gold Buffer (150 mM Tris-HCl, pH 8.0, 500 mM KCl), 2.5 to 3 mM MgCl2, 0.2 to 0.4 ␮M primer (Table 5, Operon Technologies, Alameda CA), 0.2 mM each dNTP, 0.5 units AmpliTaq Gold DNA polymerase (Perkin-Elmer, Norwalk, CT), and 80 to 100 ng of DNA template. Control samples containing the amplification mixture without template or inclusion of known DNA and blind DNA sample were also amplified to eliminate possible contamination or check the reproducibility. After an 8 min incubation at 94⬚C, amplification was carried out for 40 cycles consisting of 1 min denaturation at 94⬚C, 1 min annealing at 40⬚C, and 2 min extension at 72⬚C. Annealing temperature was reduced 1⬚C for every 2 cycles for the first 10 cycles and the remaining 30 cycles were carried out at 35⬚C. A 30 min extension at 72⬚C followed the last amplification cycle.

Table 4. Minisatellite core sequences used as primers for PCR amplification of bermudagrass DNA samples.

Table 5. Selected RAPD primers used for PCR amplification of bermudagrass DNA samples.

Primer†

Primers†

14C2 HBV-3 HBV-5 FVIIex8 FVIIe8-C TNZ-22 M13 33.6 6.2H(⫺) 6.2H(⫹)

Sequence GGCAGGATTGAAGC GGTGAAGCACAGGTG GTGGTAGAGAGGGGT ATGCACACACACAGG CCTGTGTGTGTGCAT CTCTGGGTGTGGTGC GAGGGTGGCGGCTCT AGGGCTGGAGG CCCTCCTCCTCCTTC AGGAGGAGGGGAAGG

† Primer sequences used in DAMD PCR analysis (Heath et al., 1993; Bebeli et al., 1997).

OPAA OPAE OPM-20 K-07 A-17 K-17 P-19 M-01 OPK-20 OPO-06???

Sequence GTGGGTGCCA CTGAAGCGCA AGGTCTTGGG AGCGAGCAAG GACCGCTTGT CCCAGCTGTG GGGAAGGACA GTTGGTGGCT GTGTCGCGAG CCACGGGAAG

† Primers were purchased from Operon Technologies, Inc. (Alameda, CA).

2121


Agarose Gel Electrophoresis Fifteen microliters of amplified products (RAPD or DAMD) were loaded on a 1.5 or 2% (w/v) agarose gel in 1⫻ DNA loading buffer [0.25% (w/v) bromophenol blue, 0.25% (w/v) xylene cyanol FF and 40% (w/v) sucrose in water]. Samples were then electrophoresed at 4 V/cm at constant voltage for 3 to 4 h in the presence of 1⫻ Tris borate EDTA buffer [89 mM Tris-borate, 2 mM EDTA (pH 苲8.3)] with 0.5 g/mL ethidium bromide (Sambrook et al., 1989). DNA fragments were visualized and photographed on a UV transilluminator for data analysis. Data Collection and Analyses Bands (agarose gel) or peaks (capillary electrophoresis) were scored as present (1) or absent (0), respectively. To determine best marker system(s) for forage-type bermudagrass genotypes, a total of 10 primers or primer pairs and 10 DNA samples were used. Binary scores from each of the four systems were analyzed separately and were combined for the final analysis. Dice genetic similarity indices (GSI) were calculated as SXY ⫽ 2nXY /(nX ⫹ nY), where nX and nY are the numbers of fragments in individuals X and Y, respectively, and nXY is the number of the fragments shared between individuals X and Y according to Nei and Li (1979). The dissimilarity (DXY ⫽ 1 ⫺ SXY ) matrices were analyzed by the Unweighted Pair Group Mean Average (UPGMA) or Neighbor Joining (NJ) method of Saitou and Nei (1987) using PAUP* software (Swofford, 1999). The matrices were generated by either distance or parsimony criteria in PAUP*. The bootstrap procedure, with 2000 random samplings, was also employed to determine genetic similarities calculated from the data sets obtained with the different marker systems using PAUP*. The linear correlation coefficients of the genetic similarities between the marker systems were calculated using Microsoft Excel 2000.

RESULTS AND DISCUSSION Forage bermudagrass genotypes, identified either by local farmers or extension specialists from an existing bermudagrass field, were called ‘ecotypes’ in this study to differentiate them from well-established cultigens that were designated as varieties (Table 1). To identify the most reproducible and accurate technique, this study used 10 additional plant lines, including varieties and

ecotypes as blind samples. Control samples containing the amplification mixture without template or inclusion of known DNA and unknown (blind) samples were also amplified to eliminate possible contamination as well as to check the reproducibility and accuracy of the techniques. DNA banding patterns from randomly chosen blind samples were compared lane by lane with known samples. Blind samples could be identified with using 10 primers or primer pairs (Table 2, 3, 4 and 5). A total of 1423 DNA fragments, including the 472 polymorphic markers were generated by means of 15 AFLP, 10 CpSSRLP, 10 DAMD, and 10 RAPD primers or primer pairs. Table 6 presents levels of polymorphism and comparison of informativeness of the four techniques used in this study. The AFLP technique produced the highest number of polymorphic bands per primer combination (25.9 bands per assay unit). The DAMD, RAPD, and CpSSRLP techniques produced 4.7, 3,7, and 0.5 polymorphic bands per assay unit, respectively (Table 6). Reproducibility and resolution of the DAMD and RAPD techniques were lower than the AFLP and CpSSRLP techniques. The CpSSRLP technique was equally reproducible to the AFLP technique (Fig. 1) and provided DNA markers that were useful in analyzing the maternal genome and phylogenetic relationships of Cynodon spp (Fig. 2). Tifton 78, Tifton 78 WH and ‘Poplarville I’ and ‘Poplarville II’ bermudagrass lines could not be differentiated by the CpSSRLP, DAMD (Fig. 3) and RAPD techniques (Fig. 4). However, the differentiation power of the CpSSRLP technique was lower than that of the AFLP technique because of its maternal nature and fewer number of polymorphic DNA markers. Our findings suggested that the AFLP technique was the most suitable technique (Table 6) and could differentiate all 31 forage-type bermudagrass genotypes. The Dice genetic similarity indices (GSI) for AFLP, CpSSRLP, DAMD, and RAPD data were calculated for each marker system using pair-wise distance or parsimony criteria. The GSI values were relatively lower when parsimony criteria were utilized. However, relationships among the genotypes were not affected. Similarly, the

Table 6. Level of polymorphism and comparison of informativeness obtained with AFLP, CpSSRLP, DAMD, and RAPD markers in 31 forage type bermudagrass genotypes. DNA technique Parameters Total number of assay units Total number of polymorphic bands Total number of bands Total number of bands per assay unit Total number of polymorphic bands per assay units Polymorphism (%) GSI Minimum Maximum Average Median Number of undifferentiated genotypes Reproducibility† Resolution‡

AFLP 15 389 1287 85.8 25.9 30.2 0.607 0.978 9.759 0.766 0 High High

CpSSRLP 10 5 22 2.2 0.5 22.7 0.615 1.000 0.863 0.846 16 High High

DAMD 10 47 74 7.4 4.7 63.5 0.603 1.000 0.743 0.731 2 High Low

RAPD 10 37 89 8.9 3.7 41.6 0.554 1.000 0.725 0.717 2 Medium Low

† Based on 100 bands/peaks that were consistent between PCR amplification when same samples and primers were reused. ‘High’ means that at least 95 bands were reproduced. ‘Medium’ means that at least 90 bands/peaks were reproduced. ‡ Resolution depended on the separation technique used. The AFLP and CpSSRLP techniques used capillary electrophoresis; therefore, they had a higher level of resolution.

2122


Fig. 1. AFLP Electropherograms. The x-axis represents the size (bp) of the DNA fragments and the y-axis indicates the amount of the amplified products in arbitrary numbers. Electropherogram A: Red and blue lines indicate AFLP DNA bands of Tifton 78 and Tifton 78 WH, respectively, using the AA-CAA primer pair. Electropherogram B: AFLPs between Callie (red) and Tifton 85 (blue) amplified with the AA-CAA primer pair. Note four polymorphic AFLPs indicated by arrows. Electropherogram C: A typical AFLP electropherogram showing amplified products ranging from 50 to 450 bp in length. The red line is the internal size standard.

UPGMA and NJ phylogenetic trees showed very similar dendograms (not shown). Results indicated that the forage bermudagrass genotypes had a narrow germplasm base, as the GSI ranged from 0.607 to 0.978 (AFLP), 0.615 to 1.000 (CpSSRLP), 0.603 to 1.000 (DAMD), and 0.554 to 1.000 (RAPD). The linear correlation coefficients of GSIs were calculated among the marker sys-

tems. High correlation coefficients between the AFLP, DAMD and RAPD GSI values were observed (Table 7). However, the linear correlation coefficients between CpSSRLP and other 3 techniques were lower. This low correlation was not surprising since the number of polymorphic bands in the CpSSRLP was lowest among all techniques tested. The UPGMA clusters with bootstrap frequency values are shown for each of the four techniques (Fig. 5) and a combination of the four techniques (Fig. 6). The AFLP and DAMD techniques resulted in identical clusters (Cluster A, B, C, D and E when compared with Fig. 2. CpSSRLP Electropherograms. An automated capillary electrophoretic system showing polymorphic fluorescently labeled fragments amplified with the CpSSRLP CCMP5R/F primer pair. PCR products are visualized as peaks on electropherograms. The x-axis represents the size (bp) of the DNA fragment and the y-axis shows the amount of the amplified products in arbitrary numbers. The red line is the internal size standard included in each electrophoretic run. Callie (1), paternal parent of Tifton 78, Tifton 78 (2), an F1 hybrid (Tifton 44 ⫻ Callie), Tifton 44 (3), a maternal parent of Tifton 78 and Coastal (4), a maternal parent of Tifton 44. Local ecotypes, Murphy (5) and Murphy II (6) showing differences in the chloroplast genome. These two ecotypes were included in the figure to emphasize the power of the CpSSRLP technique for differentiating two local ecotypes.


2123

Fig. 3. DAMD Agarose Gel Electrophoresis. Lanes 1 to 33 are Holly Springs, Tifton 44, Tifton 78, Tifton 78 WH, Callie, Coastal, Lott II, Murphy, Murphy II, Lott I, McDonald, World Feeder, Alicia, ‘Russell’,* Prairie III, Tifton 85, Prairie I, Prairie II, Pasto Rico, Gillihan, Sumrall 007, common type, Stallings, Maddox, Tanberg, Poplarville II, Lancaster, Hardie, Poplarville I, Dixie I, Dixie II, Grazer and negative control amplified using primer, MfVIIe8-C. *Excluded from the analysis due to plant contamination from other sources.

Fig. 4. RAPD Agarose Gel Electrophoresis. Lanes 1 to 33 are Callie (1), Tifton 44 (2), Tifton 78 (3), Tifton 78 WH (4), Coastal (5), Murphy (6), Murphy II (7); Lott I (8), Lott II (9), Prairie (10), Prairie II (11), Prairie III (12), Tifton 85 (13), Pasto Rico (14), World Feeder (15), Alicia (16), Russell* (17), Sumrall 007 (18), common (19), Stallings (20), Maddox (21), Tanberg (22), Poplarville I (23), Poplarville II (24), Lancaster (25), ‘Gillihan’ (26), Holly Springs (27), McDonald (28), Hardie (29), Dixie I (30), Dixie II (31), Grazer (32) bermudagrass lines and negative control (33), respectively, amplified with the RAPD primer OPM-20. *Excluded from the analysis due to plant contamination from other sources.

Fig. 6). Although the RAPD analysis generated A, D, and E clusters identical to that of the AFLP and DAMD techniques, some differences were observed in Clusters B and C. The bootstrap frequency values that support topology at each node in the UPGMA suggested that the combined data provided better overall bootstrap support than when DNA markers from a single technique were used (Fig. 6). Olson and McCauley (2000) also observed that combined DNA markers from mitochondria and chloroplast resulted in fully resolved clusters and had better overall bootstrap support than either of the phylogenies based on only one data set. Results indicated that the CpSSRLP technique could be used as a powerful DNA fingerprinting technique for parental genome identification. Compared with the other methods tested, the CpSSRLP technique was the easiest to analyze and equally reproducible to the AFLP technique, on the basis of this research. However, because of the conserved nature of the chloroplast genome, it produced the least number of polymorphic DNA markers among the four methods evaluated. To characterize locally selected forage-type bermuTable 7. Correlation of genetic similarity indices (GSI) between AFLP, CpSSRLP, DAMD, and RAPD techniques. AFLP CpSSRLP DAMD RAPD

AFLP

CpSSRLP

DAMD

RAPD

1.000

0.237 1.000

0.772 0.211 1.000

0.779 0.274 0.647 1.000

dagrass genotypes, we used GSI values of the varieties ‘Coastal’, ‘Tifton 44’, Tifton 78, and Callie (Table 8) from the combined data. We considered GSI value of 0.85 as a sufficient cutoff GS coefficient to identify a unique genotype. Since GSI values of the related varieties ranged from 0.861 (Callie and Tifton 78) to 0.962 (Tifton 44 and Coastal). Therefore, Sumrall 007, Tanberg, Maddox, McDonald, Holly Springs, Murphy, Murphy II, and Stallings were considered to be unique in their genome structure. However, GSI values of some other ecotypes were higher than 0.85 and they were considered to be related to some varieties: Prairie, Prairie II, and Prairie III to Grazer; Lott I, Lott II, and Lancaster to Callie; and Tifton 78 WH to Tifton 78. Although DNA markers could differentiate Poplarville I, Poplarville II, ‘Dixie I’, ‘Dixie II’, and ‘Gillihan’, additional molecular, cytological, and phenological evidences are required to confirm further whether they should be considered as related or identical genotypes. All 31 forage bermudagrass genotypes tested in the present study were divided into two major clusters (chlorotypes) based on the five polymorphic CpSSRLP markers (Fig. 5). The CpSSRLP technique suggested that Tifton 44, Tifton 78, and Tifton 78 WH have similar or identical chloroplast genomes (Fig. 2). Coastal, Tifton 44, Tifton 78, Tifton 78 WH, Lancaster, Lott I, and Lott II were related. The CpSSRLP technique also suggested that chloroplast genomes of Dixie I and Dixie II, Poplarville I and Poplarville II, Prairie I, Prairie II, and

2124


Fig. 5. Majority-rule consensus UPGMA Trees of 31 Forage Bermudagrass Genotypes. The trees were generated using the distance matrix based on Nei’s formula from 15 AFLP, 10 CpSSRLP, 10 RAPD, and 10 DAMD primers or primer pairs. Numbers on branches are bootstrap frequency values for 2000 bootstrap replicates.

Prairie III, Pasto Rico and common type were very similar. Murphy and Murphy II bermudagrass genotypes differed in their chloroplast genomes. On the basis of the AFLP, CpSSRLP, DAMD, and RAPD techniques, McDonald and Alicia were the most

genetically diverse bermudagrass genotypes (GSI ⫽ 0.608), whereas Tifton 78 and Tifton 78 WH were the most genetically similar genotypes (GSI ⫽ 0.977). The phylogenetic tree (Fig. 5) was generated on the basis of the similarity matrix (Table 8) with the combined data


2125

Fig. 6. Majority-rule consensus UPGMA tree combining AFLP, CpSSRLP, RAPD and DAMD data. The tree was generated using the distance matrix based on Nei’s formula from 15 AFLP, 10 CpSSRLP, 10 RAPD, and 10 DAMD primers or primer pairs. Numbers on branches are bootstrap frequency values for 2000 bootstrap replicates. Note combining information from all four markers systems showed better bootstrap support than any other phylogenies. A, B, C, D, and E are cluster groups discussed in the text.

from four techniques. The 31 bermudagrass genotypes were grouped into five major clusters: A, B, C, D, and E (Fig. 5). The only known pentaploid, ‘Tifton 85’ (Burton and Hanna, 1995), was in Cluster D, indicating its unique morphology and pedigree. Cluster A consisted of Tifton 78 WH, Lancaster, Lott I, and Lott II along with the varieties Coastal, Callie, Tifton 44, and Tifton 78. Ecotypes of the Cluster A were first identified in the Callie, Tifton 78, or Coastal bermudagrass fields. The results suggested that Lancas-

ter (found in a Coastal field) was probably a natural cross between Coastal and a heterozygous Callie bermudagrass. Lott I and Lott II (found in a Callie field) was a Callie line or a hybrid between Callie and Tifton 44. The close relationship between Tifton 78 and Tifton 78 WH was independently supported from the pedigree history since Tifton 78 WH is a surviving selection from Tifton 78 (D. Lang, 1997, personal communication). Cluster B consisted of Tanberg, Sumrall 007, Maddox, Gillihan, Dixie I, Dixie II, Poplarville I, Poplarville II,

1

1.000 0.858 0.861 0.857 0.791 0.778 0.808 0.750 0.805 0.796 0.704 0.751 0.751 0.895 0.801 0.792 0.770 0.772 0.767 0.887 0.911 0.723 0.739 0.723 0.750 0.717 0.690 0.682 0.625 0.677 0.774

1.000 0.926 0.920 0.822 0.763 0.806 0.770 0.790 0.780 0.726 0.769 0.767 0.867 0.797 0.799 0.754 0.784 0.778 0.842 0.869 0.735 0.762 0.748 0.756 0.736 0.688 0.678 0.614 0.680 0.772

2

1.000 0.977 0.812 0.757 0.806 0.763 0.798 0.782 0.719 0.774 0.755 0.873 0.787 0.792 0.753 0.768 0.778 0.844 0.868 0.724 0.749 0.727 0.754 0.722 0.687 0.672 0.609 0.665 0.766

3

1.000 0.817 0.761 0.804 0.764 0.799 0.782 0.714 0.773 0.756 0.868 0.788 0.791 0.757 0.767 0.777 0.846 0.863 0.728 0.752 0.734 0.758 0.729 0.682 0.664 0.609 0.666 0.766

4

1.000 0.756 0.787 0.783 0.786 0.785 0.712 0.774 0.772 0.807 0.794 0.775 0.773 0.757 0.764 0.817 0.818 0.742 0.765 0.758 0.772 0.758 0.689 0.679 0.619 0.691 0.745

5

1.000 0.815 0.751 0.806 0.792 0.692 0.744 0.744 0.793 0.801 0.813 0.766 0.775 0.761 0.802 0.789 0.758 0.776 0.760 0.738 0.751 0.708 0.690 0.658 0.694 0.780

6

1.000 0.801 0.859 0.846 0.731 0.802 0.787 0.826 0.852 0.850 0.792 0.826 0.837 0.830 0.825 0.778 0.802 0.778 0.797 0.779 0.723 0.735 0.636 0.704 0.826

7

1.000 0.802 0.811 0.706 0.773 0.773 0.758 0.789 0.787 0.801 0.777 0.789 0.774 0.762 0.761 0.782 0.762 0.767 0.760 0.698 0.694 0.608 0.684 0.770

8

1.000 0.917 0.731 0.789 0.799 0.830 0.856 0.862 0.805 0.840 0.819 0.835 0.828 0.757 0.787 0.760 0.790 0.802 0.737 0.729 0.641 0.721 0.886

9

1.000 0.726 0.768 0.790 0.810 0.855 0.863 0.796 0.844 0.824 0.817 0.805 0.748 0.783 0.770 0.790 0.786 0.718 0.712 0.630 0.713 0.839

10

1.000 0.705 0.709 0.713 0.714 0.704 0.706 0.718 0.715 0.709 0.714 0.695 0.710 0.693 0.703 0.709 0.666 0.665 0.651 0.677 0.721

11

1.000 0.845 0.766 0.794 0.790 0.794 0.776 0.762 0.783 0.771 0.815 0.834 0.787 0.845 0.802 0.712 0.694 0.635 0.703 0.757

12

1.000 0.763 0.777 0.771 0.792 0.770 0.761 0.774 0.761 0.814 0.852 0.804 0.787 0.803 0.724 0.705 0.646 0.712 0.769

13

1.000 0.823 0.821 0.776 0.784 0.790 0.909 0.939 0.738 0.754 0.741 0.775 0.742 0.696 0.681 0.621 0.682 0.801

14

1.000 0.934 0.807 0.829 0.817 0.836 0.821 0.763 0.774 0.746 0.777 0.779 0.707 0.713 0.636 0.715 0.811

15

1.000 0.801 0.836 0.817 0.822 0.809 0.777 0.780 0.763 0.795 0.789 0.733 0.728 0.635 0.719 0.846

16

1.000 0.774 0.769 0.791 0.774 0.781 0.781 0.771 0.797 0.779 0.707 0.695 0.643 0.720 0.785

17

1.000 0.801 0.800 0.785 0.759 0.788 0.749 0.781 0.770 0.705 0.709 0.634 0.704 0.805

18

1.000 0.798 0.791 0.758 0.781 0.769 0.774 0.750 0.703 0.691 0.612 0.688 0.782

19

1.000 0.941 0.746 0.770 0.744 0.780 0.744 0.687 0.682 0.624 0.678 0.798

20

1.000 0.729 0.749 0.726 0.766 0.727 0.684 0.680 0.623 0.675 0.792

21

1.000 0.858 0.832 0.774 0.800 0.712 0.701 0.643 0.728 0.756

22

1.000 0.853 0.801 0.828 0.717 0.705 0.649 0.725 0.771

23

1.000 0.818 0.866 0.723 0.693 0.637 0.727 0.779

24

1.000 0.831 0.735 0.706 0.642 0.720 0.784

25

1.000 0.743 0.715 0.645 0.736 0.781

26

1.000 0.854 0.683 0.769 0.746

27

29

30

31

1.000 0.707 1.000 0.781 0.784 1.000 0.742 0.638 0.717 1.000

28

† The distance matrix was analyzed by the UPGMA method of Saitou and Nei (1987) on the basis of the AFLP, CpSSRLP, RAPD, and MAFLP DNA markers. The distance between two lines was calculated as DXY ⫽ 1 ⫺ SXY, which can range from 0.0 to 1.0, when all bands in the two lines are the same (0.0), when there are no bands in common (1.0). Genetic similarity (SXY ), which can range from 1.0 to 0.0, when two lines are identical (1.0), when there are no bands in common between the two lines (0.0).

1 Callie 2 Tifton 44 3 Tifton 78 4 Tifton 78 WH 5 Coastal 6 Hardie 7 Maddox 8 Alicia 9 Dixie I 10 Dixie II 11 Tifton 85 12 Murphy 13 Murphy II 14 Lancaster 15 Poplarville I 16 Poplarville II 17 H. Springs 18 Sumrall 007 19 Tanberg 20 Lott I 21 Lott II 22 Prairie I 23 Prairie II 24 Prairie III 25 Stallings 26 Grazer 27 P. Rico 28 Common 29 McDonald 30 W. Feeder 31 Gillihan

Table 8. Similarity matrix for bermudagrass cultivars and ecotypes determined on the basis of AFLP, CpSSRLP, RAPD, and DAMD markers.†

2126 CROP SCIENCE, VOL. 42, NOVEMBER–DECEMBER 2002


and Holly Springs, along with the two varieties ‘Hardie’ and Alicia. Hardie is a sterile, vegetatively propagated bermudagrass (Taliaferro and Richardson, 1980). However, the Hardie genome contains portions of the genomes of Cynodon accessions; ’9945A’, ‘8153’, and ’9953’ (Taliaferro and Richardson, 1980). Therefore, ecotypes clustered with Hardie might be related to those Cynodon accessions. Cluster C consisted of ecotypes Stallings, Murphy, Murphy II, Prairie I, Prairie II, and Prairie III, along with the variety Grazer. As expected, ecotypes identified in the same location were genetically similar. For example, Prairie I, Prairie II, and Prairie III were related. These close relationships were also suggested from the DNA content analysis of the ecotypes (Karaca et al., 2000). The results suggested that these three ecotypes (Prairie I, Prairie II, and Prairie III) might have originated from Grazer. The diversity between the parents and among the F1s from which Grazer was selected indicates that it is highly heterozygous (Burton et al., 1986). Cluster E consisted of ‘Pasto Rico’, common type, McDonald, and World Feeder forage bermudagrass genotypes. Results suggested that the McDonald is a distant genotype and it is the only ecotype clustered in this group. The relationship between Pasto Rico and common type was not surprising since Pasto Rico is a mixture between a common and ‘NK 37’. Because of limited sample of NK 37, we could not include this variety in our analysis. However, observations of the banding patterns between Pasto Rico and common revealed that Pasto Rico almost always showed common type’s banding patterns. The extra bands from Pasto Rico, in comparison with that of common, were considered to be DNA markers representing NK 37. ACKNOWLEDGMENTS This research was in part supported by a research scholarship from the Turkish Ministry of Education and the Mississippi Agricultural and Forestry Experimental Station (MIS 162041). The authors thank Drs. C. E. Watson, J. V. Krans, D. P. Ma, and P. Williams for their helpful suggestions on the manuscript. The authors wish to thank Dr. A. G. Karaca for her help in sample collection and DNA extraction.

REFERENCES Arcade, A., F. Anselin, P.F. Rampant, M.C. Lesage, L.E. Paques, and D. Prat. 2000. Application of AFLP, RAPD and ISSR markers

2127

to genetic mapping of European and Japanese larch. Theor. Appl. Genet. 100:299–307. Bebeli, P.J., Z. Zhou, D.J. Somers, and J.P. Gustafson. 1997. PCR primed with minisatellite core sequences yields DNA fingerprinting probes in wheat. Theor. Appl. Genet. 95:276–283. Burton, G.W., and W.W. Hanna. 1995. Forage Breeding. In R.F Barnes et al. (ed.) Forages. Iowa State University Press, Ames, IA. Burton, G.W., and W.G. Monson, and P.R. Utley. 1986. Registration of ‘Tifton 72-84’ bermudagrass. Crop Sci. 26:389. Caetano-Annolles, G., L.M. Callahan, and P.M. Gresshoff. 1997. The origin of bermudagrass (Cynodon ) off-types inferred by DNA amplification fingerprinting. Crop Sci. 37:81–87. Cato, S.A., and T.E. Richardson. 1996. Inter- and intraspecific polymorphism at chloroplast SSR loci and the inheritance of plastids in Pinus radiata D. Don. Theor. Appl. Genet. 93:587–592. Costa, P., D. Pot, C. Dubos, J.M. Frigerio, C. Pionneau, C. Bodenes, E. Bertocchi, M.T. Cervera, D.L. Remington, and C. Plomion. 2000. A genetic map of Maritime pine based on AFLP, RAPD and protein markers. Theor. Appl. Genet. 100:39–48. Edwards, N.C., J. Askew, F. Boykin, W.B. Burdine, G. Cuomo, R. Elmore, C.H. Hovermale, D.M. Ingram, R. Ivy, B. Johnson, D. Lang, G.A. Pederson, R. Saunders, and J. Tomlinson. 1993. 1992–93 performance of forage crop varieties in Mississippi. MAFES Bull. No 247. Heath, D.D., G.K. Iwama, and R.H. Devlin. 1993. PCR primed with VNTR core sequence yields species specific patterns and hypervariable probes. Nucleic Acid Res. 21:5782–5785. Karaca, M., D.J. Lang, G.L. Yerk-Davis, S. Saha, and A.J. Ainsworth. 2000. Determination of DNA content and genome size in Cynodon species by flow cytometry. Crop Res. 20:1–12. Karaca, M. 2001. Characterization of Cynodon spp. and Gossypium spp. genomes using molecular and cytological techniques. Ph.D. Dissertation [DAI, 62, no. 05B (2001): p. 2119 ISBN: 0-493-261052]. Mississippi State University, Mississippi State, MS. Murray, M.G., and W.F. Thompson. 1980. Rapid isolation of high molecular weight plant DNA. Nucl. Acids Res. 8:4321–4325. Nei, M., and W.H. Li. 1979. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc. Natl. Acad. Sci. (USA) 76:529–5273. Olson, M.S., and D.E. McCauley. 2000. Linkage disequilibrium and phylogenetic congruence between chloroplast and mitochondrial haplotypes in Silene vulgaris. Proc. R. Soc. Lond. 267:1801–1808. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4: 406–423. Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Swofford, D.L. 1999. PAUP*. Phylogenetic analysis using parsimony (*and other Methods). Version 4. Sinauer Associates, Sunderland, MA. Taliaferro, C.M., and W.L. Richardson. 1980. Registration of ‘Hardie’ bermudagrass. Crop Sci. 20:413. Vos, P., R. Hogers, M. Bleeker, M. Reijans, T. Van de Lee, M. Hornes, A. Frijters, P. Peleman, M. Kuiper, and M. Zabeau. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23: 4407–4414. Weising, K., and R.C. Gardner. 1998. A set of conserved PCR primers for the analysis of simple sequence repeat polymorphism in chloroplast genomes of dicotyledonous angiosperms. Genome 42:9–19.