Genetic divergence among elephantgrass cultivars GENETIC DIVERGENCE AMONG ELEPHANTGRASS CULTIVARS ASSESSED BY RAPD MARKERS IN COMPOSIT SAMPLES

ABSTRACT: Elephantgrass Pennisetum purpureum( Schum.) is native to regions of tropical Africa and wasintroduced in Brazil around 1920 through plantings imported from Cuba. It is currently one of the mostwidespread forage plants throughout the country. At first, there were two cultivars, Napier and Mercker, withwell defined characteristics. New genotypes arose and it is believed that the large number of cultivars existingtoday in germplasm bank is due to duplicates. DNA markers for cultivar characterization are a very valuabletool, especially in situations where morphological and isoenzymatic markers have already been used as inthe case of elephantgrass. Thus RAPD markers were used to estimate the genetic divergence among theNapier group elephantgrass cultivars from the elephantgrass Active Germplasm Bank at EMBRAPA DairyCattle. The polymerase chain reaction with 37 arbitrary primers from the OPERON Technologies series supplied94 polymorphic and 73 monomorphic bands. From the matrix of complement of the Nei index, cluster analysisby the Tocher optimization method formed three clusters. Pearson correlation among genetic distance estimatesobtained from the DNA markers and the isoenzymatic markers showed the consistency of both the methodsin assessing genetic divergence among elephantgrass cultivars. No duplicates were found in the treatmentsassessed.Key words:


INTRODUCTION
Elephantgrass (Pennisetum purpureum Schum.) a tetraploid (2n = 4x = 28) and perennial forage grass (Manara, 1973;Brunken, 1977), is native to regions of Tropical Africa and was introduced in Brazil around 1920 by vegetative propagation of plantings imported from Cuba.It is currently one of the most widespread forage species in the country (Otero, 1961).At first there were two cultivars with well defined characteristics, 'Napier' and 'Mercker'.New genotypes arose with time, either by introduction or by breeding of existing varieties, using the variation resulting from the sexual process.It is believed that part of the accessions currently in the species Germplasm Bank (Embrapa Dairy Cattle) are duplicates because the same cultivar was introduced from different origins but the original identifications were not kept.
The genetic variability among elephantgrass cultivars, a primordial condition to carry out an efficient breeding program, is still not completely known.Several Scientia Agricola, v.59, n.4, p.623-627, out./dez.2002 studies have assessed and reported elephantgrass for its yield potential (Mozzer et al. 1970;Pedreira & Campos, 1975).Carvalho et al. (1972) gathered 30 elephantgrass introductions in three big groups, including variables such as stand, tillering, inflorescence and flowering period.Alcântara et al. (1980) characterized 25 elephantgrass introductions using 15 descriptors and concluded that it is extremely difficult to separate the introductions by morphological characterization.The assessment of morphological traits and multivariate analysis allowed for a better interpretation of the data, the formation of heterotic groups (Daher et al. 1997a;Botrel et al. 1995;Tcacenco & Botrel, 1994) and the discarding of variables of little interest for genetic divergence assessment (Tcacenco, 1988;Daher et al. 1997b).Biochemical and isoenzymatic makers have been used to minimize the effects of environment-genotype interaction, (Augustin & Tcacenco, 1993;Daher et al. 1997a;Bach et al. 1998).
DNA markers are valuable for cultivar characterization (Ferreira & Grattapaglia, 1996) especially when morphological and isoenzymatic markers have already been used, as in the case of elephantgrass.This study was carried out to estimate the genetic divergence among elephantgrass cultivars of the 'Napier´ group from the Active Germplasm Bank at Embrapa Dairy Cattle by RAPD markers in composit samples.

Genetic material
Nine elephantgrass accessions of the 'Napierǵ roup from the Germplasm Bank at Embrapa Dairy Cattle (Table 1) located in Coronel Pacheco, MG, were used in this study.Nine young leaf blades were taken from seven plants per accession, squashed in liquid nitrogen and placed in an ultrafreezer (-86°C) in closed 15 mL 'Falcont ubes.The study was carried out in Campos dos Goytacazes, RJ, Brazil.

DNA extraction
Approximately 200 mg of squashed material from the leaf blades was transferred to 1.5 mL eppendorf tubes and immersed in liquid nitrogen.DNA was extracted according to the protocol by Doyle and Doyle (1987) with modifications; 800 mL of pre-heated extraction buffer containing 1% CTAB, 1.4 mol L -1 NaCl were added to the tube containing 1% CTAB, 1.4 mol L -1 NaCl, 20 mmol L -1 EDTA, 100 mmol L -1 Tris-HCl (pH 8.0), 1% PVP and 0.1% 2-mercaptoetanol.It was incubated at 65°C for 30 to 40 minutes and gently shaken every 10 minutes, and then centrifuged at 13400 g for five minutes.The supernatant (600 µL) was transferred to new tubes and an equal volume of chloroform:isoamilic alcohol (24:1) was added and continually inverted until an emulsion was obtained.This step was repeated, and after new centrifuging, the supernatant was transferred to new tubes and chilled isopropanol was added, with gentle inversion and placed overnight in the refrigerator.Samples were then centrifuged at 13400 g for 10 minutes, and a pellet obtained (precipitate) that was washed twice in 300 µL 70% ethanol and once in 300 µL 95% ethanol, dried at room tempetature, re-suspended in 200 µL Tris-EDTA solution (10 mmol L -1 Tris CHCl, 1 mmol L=-1 EDTA, pH 8.0) and incubated with RNAse at a final concentration of 40 µg mL -1 at 37°C for 30 minutes.After adding 20 µL NACl 5 mol L -1 and 140 µL chilled isopropanol the mixture was incubated overnight at 4 o C. It was again centrifuged at 13400 g for 10 minutes, dried and finally the pellet was re-suspended in 200 µL of TE solution.The DNA concentrations in the samples were estimated using fluorometric methods (DyNA Quant 200 Fluorometer, Hoefer Scientific, San Francisco, USA) and then standardized at 10 ng µL -1 concentration.

Polymerase chain reaction
The amplification reactions were carried out according to Williams et al. (1990) modified in a final volume of 25 µL containing the reagents at the following concentrations 10 mmol L -1 Tris-HCl, pH 8.3; 50 mmol L -1 KCl; 2.4 mmol L -1 MgCl 2 ; 100 µmol L -1 dATP, dCTP, dGTP and dTTP; 0.3 µmol L -1 primer (Table 2); 20 ng genomic DNA and a unit of Taq DNA polymerase (Pharmacia Biotech, EUA).A thermal cycler was used (Perkin Elmer GeneAmp PCR System 9600) programmed for 95°C for 1 minute followed by 45 cycles of one minute at 94°C, 1 minute at 36°C and 2 minutes at 72°C and a final step for extension of 7 minutes at 72°C, using the fastest temperature transition mode available (1°C/1seg.).The amplification products (bands) were analyzed by electrophoresis and visualized after staining by ethydium bromide.

Band reading and data analysis
The RAPD profiles of each accession were obtained by the presence (1) or absence ( 0

Statistical analysis
Hierarchical clustering (closest neighbor) methods were used and optimization proposed by Tocher, quoted by Rao (1952)

RAPD Polymorphism
Thirty-seven primers were analyzed and a total of 167 amplification products (Table 2) was obtained with a mean of 4.51 bands per primer.This value is similar to the results reported by Virk et al. (1995) in rice, Sweeney & Danneberger (1994) in Poa annua L., Haley et al. (1994) in common bean and Harvey & Botha (1996) in Saccharum, in genetic divergence assessment by the RAPD technique.

Data Analysis
Analysis of the clustering by the Tocher optimization method based on the matrix of the complement of the Nei & Li index indicated three distinct groups, possibly forming heterotic groups (Table 3).The clustering by the closest neighbor hierarchical method (Figure 2) was very similar to that obtained by the Tocher method and only differed in the final disposition of the groups and the greater discrimination among the accessions.
Generally reduced separation was observed among the Mercker and the Napier type cultivars (Figure 2), contrary to information in the literature about the start of its introduction in Brazil, when two cultivars with well defined characteristics were known, Mercker and Napier (Otero, 1961).The Mercker accession (number 2 of the elephantgrass BAG) used in this study may not be an accession from the Mercker group, but rather from the Napier group as its original identification was probably lost during introduction to the Active Germplasm Bank.
The Mineiro cultivar clustered with the Napier cultivar by Tocher clustering (Group 1, Table 3), and were also placed together on the same branch of the dendrogram obtained by the hierarchical closest neighbor method (Figure 2).The Napier and Mineiro accessions were also observed within the same patterns of isoenzymatic bands by the isoenzymatic standard obtained by Daher et al. (1997a), and this confirms that the Mineiro cultivar is derived from selection among Napier cultivar progenies, currently forming one of the cultivars that occupies greatest cultivated area (Pereira, 1993).
The Napier accessions Volta Grande (4) and Napier no. 2 (8) were also very similar, based on the results of this study (Figure 2) and on the results obtained with the isoenzymatic markers in Daher et al. (1997a).Similarly, the distribution of the Pusa Napier no. 2 (6), Pusa Gigante Napier (15) and Pusa Napier no. 1 (25) accessions in distinct and more distant groups compared to the other accessions (Groups 1, 2 and 3, respectively) in Table 3 and on independent branches, as shown in the dendrogram (Figure 2), was in line with the clustering obtained using the isoenzymatic markers presented in Daher et al. (1997a) where these were also placed in distinct groups.Furthermore, this illustrates the agreement between the genetic composition, expressed by the molecular markers (isoenzymes and RAPD) and the geographic diversity, related to the Indian origin of these materials confirming the more distant positioning of the Pusa Napier no. 2 (6), Pusa Gigante Napier (15) and Pusa Napier no. 1 (25) accessions compared to the others.
When the seven highest dissimilarity values (complement of the Nei & Li index) detected by RAPD were compared with the seven highest dissimilarity values (complement of the Jaccard index) detected by the isoenzymatic systems in Daher et al. (1997a), only two combinations did not coincide (2 x 15 and 8 x 15) while the accession pairs 8 x 25, 4 x 25, 25 x 27, 15 x 25 remained the most divergent in the two assessments (Table 4).On the other hand, only two pairs of accessions (4x8 and 27x30) coincided among the seven lowest values by least dissimilarity values.
The Pearson correlation coefficient among the dissimilarity estimates (Divergence) obtained by the RAPD technique and isoenzymes was 0.5176, and was significant by the t test (P < 0.01).This indicates the consistency of both methods in obtaining genetic divergence estimates among elephantgrass accessions.
The use of compound samples was effective to conclude on the interpopulation genetic diversity.However, individual samples will have to be used to understand the intrapopulation genetic variability.

CONCLUSIONS
The markers enabled each one of the cultivars to be discriminated indicating that they are genetically different.The use of compound samples showed the existence of genetic diversity among the cultivars called Napier.
The genetic distance estimates obtained from RAPD DNA markers and the isoenzymatic markers were highly correlated, indicating the consistency of the methods for assessing the genetic divergence among elephantgrass accessions.
) of high intensity bands and the complement of the Nei & Li index (1 -IAB) (Nei & Li, 1979) was calculated as IAB = 2a/ (2a+b+c) where: a = number of bands present in both accessions, simultaneously; b = number of bands present only in accession A; c = number of bands present only in accession B.
based on the matrix obtained from the arithmetical complement of the Nei & Li similarity index (1979).The statistical analyses were performed by the programs GENES, 0.1.0version and STATISTICA version 95 (Statsoft Inc, Tulsa, Oklahoma, USA).

Figure 1 -
Figure 1 -λ phage marker digested with EcoRI and Hind III and bandprofiles generated by RAPD with the OPAE-04 primer in 1.2% agarose gel involving the nine elephant grass accessions described in Table1.

Table 1 .
Figure 2 -Clustering obtained by the hierarchical closest neighbor method based on the arithmetical complement of the Nei & Li similarity index of nine elephant grass cultivars (Pennisetum purpureum Schum.)assessed by LMGV/ CCTA/UENF, in Campos dos Goytacazes, RJ.Scientia Agricola, v.59, n.4, p.623-627, out./dez.2002 a The number sequence indicates the entry of the clone in the group.b Table 1 lists the clones.

Table 2 -
Primers from the Series OPERON Technologies with the respective numbers of polymorphic and monomorphic bands.

Table 4 -
Relationship of the seven first pairs of accessions that presented the greatest and least dissimiliarity estimates by the RAPD technique (arithmetical complement of the Nei & Li similarity index) and isoenzymes (arithmetical complement of the Jacard).