Introduction

Agrostis comprises more than 200 species of perennial cool-season turfgrasses belonging to the Poaceae family (Hitchcock 1971). Creeping bentgrass (Agrostis stolonifera L.) can be propagated sexually by seeds and is used for golf greens and fairways in temperate climates (Warnke 2003), whereas colonial (A. capillaris L.) and velvet (A. canina L.) bentgrasses are rarely used in golf courses (Brilman 2003; Rummele 2003). Two other Agrostis species are reported as being used as turfgrass in the United States, dryland (A. castellana Boiss. and Reut.) and redtop (A. gigantea Roth) (Chakraborty et al. 2005). The growth habit of creeping and velvet bentgrasses is stoloniferous and adapted to cool and humid regions, whereas colonial bentgrass has a bunch-type growth habit (Brilman 2003). Like other grass species, Agrostis species have different ploidy levels with different genome composition. Velvet bentgrass is a diploid, with x = 7 and 2n = 2x = 14, having only bivalents in the first metaphase of meiosis with an A1A1 genome composition (Jones 1956; Nelson 1985). Creeping bentgrass is reported to be allotetraploid, with 2n = 4x = 28 chromosomes with an A2A2A3A3 genome (Warnke 2003). Colonial bentgrass also has 2n = 4x = 28 chromosomes but has a different genome composition (A1A1A2A2 or A2A2A2A2) (Rummele 2003).

Sequence related amplified polymorphism (SRAP) technology has been recognized as one of the most variable types of DNA sequences found in plants. This system has been employed for mapping and gene tagging in Brassica (Li and Quiros 2001). SRAP has also been employed to evaluate genetic diversity and phenetic relationships among a broad range of cool- and warm-season turfgrass species (Budak et al. 2004a), tomato (Lycopersicon esculentum L. Mill.) (Ruiz et al. 2005), and squash (Cucurbita moschata) (Ferriol et al. 2003), and genetic relationships among subpopulations of single species (Budak et al. 2004b, 2005a). It has proven to be more informative than AFLP (amplified fragment length polymorphism), RAPD (rapid amplified polymorphic DNA), ISSR (inter-simple sequence repeat), and SSR (simple sequence repeat) markers (Ferriol et al. 2003; Budak et al. 2005a, 2004c). SRAP is based on two-primer amplifications where the primers are 17 or 18 nucleotides long. Primers consist of a core sequence of 13 or 14 bases, where the 5′-most 10 or 11 bases are nonspecific, followed by the sequence CCGG in the forward primer and AATT in the reverse primer. The core sequence is followed by three selective nucleotides at the 3′ end of each primer (Li and Quiros 2001).

The objective of this study was to investigate the presence and nature of SRAP at the cDNA level (not genomic DNA) in three bentgrass species: creeping bentgrass (Agrostis stolonifera), velvet bentgrass (A. canina), and colonial bentgrass (A. capillaris).

The study also includes characterization of a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene identified from velvet bentgrass, because of its critical role in many cellular processes, from both an enzymological and a molecular point of view. For instance, up-regulation of this GAPDH gene has been shown to increase tolerance under various stress conditions, such as in response to environmental stress in maize (Russell and Sachs 1989), Arabidopsis thaliana (Yang et al. 1993), and Craterostigma plantagineum (Velasco et al. 1994), and biotic stress in potatoes (Laxalt et al. 1996).

Materials and Methods

Plant Materials

Approximately 1,000 pure live seeds from each of ’Southshore’ creeping bentgrass (Agrostis stolonifera), ‘SR7200’ velvet bentgrass (A. canina), and ‘Heriot’ colonial bentgrass (A. capillaris) species were planted in the greenhouse to insure that the resulting plant populations represented the diversity of each species. The seeds were planted in pots 15 cm in diameter and 15 cm deep, with a soil mixture of 350 g kg−1 peat, 320 g kg−1 vermiculite, 90 g kg−1 soil, and 240 g kg−1 sand. The greenhouse was maintained at 25±1°C with supplemental light supplied by metal halide lamps on a photoperiod of 15 h day/9 h night. The pots were saturated biweekly with a soluble fertilizer (21 N, 3.5 P, 15 K) at 200 mg l−1 nitrogen (Budak et al. 2005b). Individual seedlings of each Agrostis species were used for cDNA amplification using 56 SRAP primer pair combinations.

cDNA Library Construction

Total RNA was isolated from 0.2 g fresh leaf and root tissues using an RNeasy Plant Mini Kit (Qiagen), and poly(A)+RNA was isolated from total RNA using the Oligotex mRNA Mini Kit. Results were visualized by electrophoresis in formaldehyde (2.2 M)-agarose (1.2%) gel in MOPS buffer (0.02 M MOPS, 2 mM CH3COONa, 1 mM EDTA) at pH 7.0. Isolated RNA was then treated with DNaseI and precipitated before the determination of its concentration and purity spectrophotometrically. A λ-ZAP cDNA library was constructed with 5 μg of mRNA from leaf and root tissue following the manufacturer’s instructions (Stratagene, La Jolla, CA).

Amplification, Cloning, and Sequencing of cDNA Using SRAP

Total RNA (2.5 μg) was converted to cDNA using an Omniscript reverse transcription kit (Qiagen) according to the manufacturer’s instructions. First-strand cDNA was synthesized from the poly(A)+ RNA with Moloney Murine Leukemia Virus (MMLV) reverse transcriptase. In total, 56 SRAP oligonucleotide combinations (Table 1) were used for PCR amplification of cDNA sequences from three bentgrass species. Amplification of cDNA fragments was performed in 20 μl PCR reactions (Budak et al. 2005a, 2006). Each reaction mixture contained 2 μl first-strand cDNA, 2 μl 10 × PCR buffer without MgCl2, 2.5 mM MgCl2, 200 μM dNTP mix, 1 μM forward primer, 1 μM reverse primer, 1.25 U Taq DNA polymerase (Promega). Amplifications were carried out using an MJ Research PTC-100 thermocycler programmed for 32 cycles of 1 min at 94°C, 1 min at 50°C, 1 min at 72°C, followed by a final extension at 72°C for 5 min before cooling to 24°C. The PCR products obtained from first-strand cDNA synthesis were size-fractionated in 6% polyacrylamide (PAGE)/8 M urea gel. The gel containing 10 μl of amplification product was separated at 180 V for 4 h.

Table 1 SRAP primers used in this study
Full size table

Amplified fragments were excised from gels and extraction was performed using the QiaexII Gel Extraction Kit (Qiagen). DNA was inserted into a PGEM-T Easy Vector System I (Promega, Madison, WI) and propagated in E. coli DH5 (Life Technology) grown cells in Luria Bertani medium containing 100 μg/ml ampicillin. Three clones were selected for sequence analysis.

Sequence Analysis

The sequences were first entered in VecScreen http://www.ncbi.nlm.nih.gov/VecScreen/VecScreen.html, which aligns the entered sequences with common vector sequences. Using this tool, the vector sequences were identified and removed. The Blast algorithm (Altschul et al. 1997) was used to analyze the DNA sequences (http://www.ncbi.nlm.nih.gov). The discontinuous MegaBlast tool (http://www.ncbi.nlm.nih.gov) was used because the regular MegaBlast tool gave only exact or highly similar matches, causing some similar sequences to be ignored. The “nr” database was used (all GenBank + RefSeq Nucleotides + EMBL + DDBJ + PDB sequences). The sequences obtained were then submitted to dbEST, GenBank, under the accession numbers EB812786–EB812823. Sequence analysis of the cDNA fragment EB812822 was further translated to protein sequence using the Translate tool at the Expasy Web site (http://www.expasy.org/tools/dna.html). Sequence alignment of the corresponding protein sequence of the EB812822 EST fragment was done by ClustalW (Vector NTI version 9.0) against various plant and mammalian (wheat, maize, barley, corn, banana, lobster, halophyte, carnation) cytosolic phosphorylating GAPDH sequences.

Expressed Sequence Tag (EST) Analysis

Minimum evolutionary analysis was examined with a Dayhoff model tree constructed using the PAM matrix (Mega program, version 3.1). The PAM distance matrix primarily corrects for multiple substitutions based on the model of amino acid substitution outlined as substitution-rate matrices and uses the PAM 001 matrix (Schwarz and Dayhoff 1979). Using a substitution-rate matrix (Q), the matrix F, which consists of the observed proportions of amino acid pairs between a pair of sequences with their divergence time t, is given by the following equation:

$$ {F}({t})={A\rm e}^{2tQ} $$

where A is the diagonal matrix of the equilibrium amino acid frequencies for Q. From the equation, the evolutionary distance d = 2tQ was computed by a maximum-likelihood method. The eigen values for the PAM matrix were obtained from the program source code of Phylip version 3.6 (Felsenstein 1993).

Results and Discussion

Identification and Characterization of ESTs Using SRAP

A total of 56 SRAP oligonucleotide combinations (Table 1) were used for cDNA amplifications of three bentgrass species: creeping bentgrass (Agrostis stolonifera), velvet bentgrass (A. canina), and colonial bentgrass (A. capillaris). The PCR products generated 80 fragment bands in the size range 250–950 bp. Fragment bands smaller than 100 bp were not visible on the gel. A total of 38 cDNA fragment bands that appear to be unique within each species (19 from creeping, 10 from colonial, and 9 from velvet bentgrass species) were excised from the gel. Three cDNA clones per fragment were sequenced and analyzed further. Some SRAP markers worked in the creeping bentgrass genome but did not amplify cDNA in colonial and velvet bentgrasses. This response might be because velvet bentgrass does not contain the amplification site when compared to creeping bentgrass. This degree of site loss likely reflects genomic divergence in Agrostis species.

The SRAP primer combinations used targeted and amplified ORFs. The possible ORFs were determined by the ORF finder program at http://www.ncbi.nlm.nih.gov/gorf/gorf.html (Table 2). A GenBank search with the BlastX algorithm (Altschul et al. 1997) was performed to compare the sequences of ESTs obtained to the nr databases. A homology search revealed similarity in 68% (26/38) of ESTs to functionally diverse proteins cloned from other plants, bacteria, virus species, and zebrafish species. Identified ESTs revealed significant similarity to proteins involved in DNA synthesis, transcription, and transposon elements such as ribonucleotide reductase, putative reverse transcriptase, and retrotransposon protein, and proteins involved in structures like Matrilin 2 (MATN2, cell wall-associated kinases); extensin-like proteins; and metabolism-related enzymes such as phosphoglucomutase, GAPDH, and catalase. Particularly, the corresponding protein of the EST sequence, EB812822, reveals clear sequence homology (93%) to the ubiquitous and highly conserved cytosolic phosphorylating GAPDH (GapC) enzyme of Triticum aestivum (EF592180). Hence GAPDH and its deduced amino acid sequence were further analyzed. Sequence alignments of representative homologs from plants and eukaryotes are shown in Fig. 1.

Table 2 ESTs and homologous sequences obtained by BlastX search at the NCBI network server
Full size table
Fig. 1
figure 1

Sequence alignment of velvet bentgrass GAPDH and other phosphorylating plant and eukaryotic GAPDH amino acid sequences obtained using ClustalW program. Asterisks (*) indicate fully conserved residues. Bold letters in positions marked by a colon (:) indicate strong conservation among the species. Gray letters in positions marked by a dot (.) indicate weak conservation among the species. There is no consensus among the species for other amino acids

Full size image

Sequence Analysis of Cytosolic GAPDH

The partial peptide sequence obtained is 112 amino acids long (complementary amino acids span from 227 to 330). One monomeric unit of the GAPDH comprises two conserved functional domains, an N-terminal NAD-binding domain (residues 1–147 and 312–330) and a C-terminal catalytic domain (residues 148–311) (Yun et al. 2000; Branlant and Branlant 1985) (in the text all the residue numberings are given according to the lobster GAPDH sequence unless otherwise noted) (Fig. 2). Some of the known residues that are highly conserved in the GAPDH sequences are present in velvet bentgrass cDNA. The two critical amino acids for the catalytic mechanism of GAPDH, Cys 149 and His 176, were not sequenced in the cDNA fragment of velvet bentgrass. Arg 231 is conserved in the obtained EST, which implies the presence of the putative second anion recognition site. Also, Tyr 311 and Ser 238, which are identified as conserved residues in the cDNA fragment obtained, are the two hydrogen-bonded residues to His 176 in the two dimeric subunits of the homotetramer (Moras et al. 1975). One of the interfacial hydrogen-bond forming residues, Glu 276, in the catalytic domain (Skarzynski et al. 1987) is also conserved in the EST sequence obtained (Fig. 2). The active site loop, 220s loop, whose sequence is highly conserved among GAPDH, is partially obtained in the velvet bentgrass (residues Gly 227 to Phe 230). As observed for the available GAPDH structures, this loop acts like a latch and closes toward the catalytic Cys 149 and His 176 residues to constrain the active site by moving the inorganic phosphate-binding site away from the active region (Kim and Hol 1998).

Fig. 2
figure 2

Crystal structure of the lobster GAPDH monomer (Protein Data Bank code 1GPD). Residues represented as spheres, Ser 148 and Asp 312, are the linkage residues between the two domains of GAPDH, the N-terminal NAD-binding domain (green) and the C-terminal catalytic domain (blue). The yellow region of the ribbon diagram corresponds to the homologous parts of the GAPDH EST sequence obtained for velvet bentgrass. NAD+ is shown as a stick model in orange. The S-loop is shown with a black arrow. The figure was produced using the program Pymol (DeLano 2002)

Full size image

EST Analysis

A minimum evolution dendrogram was constructed among the sequences (Fig. 3) in order to further analyze the EST obtained. The cophenetic correlation between ultrametric similarities of the tree matrix was r = 0.93, suggesting the cluster analysis performed for this data strongly represents the similarity matrix. Eight clustering groups were found and are depicted in Fig. 3. The first group consists of 9 ESTs, including colonial, creeping, and velvet bentgrasses. There are 10 ESTs in the third group. Groups II and IV contain creeping and velvet bentgrass ESTs, but group V has only creeping bentgrass ESTs. Colonial bentgrass is known to be more related to creeping bentgrass (Macbryde 2005); however, SRAP analysis revealed more common ESTs among creeping and velvet bentgrass species. Group VII consisted of five ESTs from creeping and colonial bentgrasses, which might be an example of a gene from a common ancestor that later diverged during speciation. The two ESTs, EB812811 from colonial bentgrass and EB812801 from creeping bentgrass, were found to span the extremes of the dendrogram, indicating that these ESTs have the highest divergence. These results indicated that the SRAP primer combinations, targeted at the transcribed genes of three bentgrass species, revealed the existence of common ESTs. Different clustering groups might indicate that these cDNA sequences isolated from three C3 Agrostis species appear to be part of a multigene family.

Fig. 3
figure 3

Minimum evolution tree comparing translated EST sequences obtained from Agrostis species. Numbers at the nodes are the bootstrap percentages based on 500 replications. GenBank accession numbers EB812786–EB812804 are for creeping bentgrass (black type), EB812805–EB812814 for colonial bentgrass (gray screen), and EB812815–EB812823 for velvet bentgrass (gray type)

Full size image

Conclusion

Cloning and sequencing of a set of cDNA to visualize transcript polymorphism are reported using SRAP technology in three bentgrass species. There are no reports on the use of SRAP for revealing transcribed bentgrass species genes or any other turfgrass species. The ESTs identified in this study could potentially be used in turfgrass breeding and genetics programs as functional markers. Integration of these ESTs to the existing linkage map of turfgrass species provides high-density coverage in selected genomic regions. Minimum evolutionary tree clustering indicated that ESTs obtained using SRAP could be used for comparative genomics analysis of transcribed genes among the grass species.

Because GAPDH plays a critical role in many cellular processes, from both an enzymological and a molecular point of view, this enzyme appears to be an attractive candidate for use as the target enzyme for Agrostis species. GAPDH can be used as a molecular biomarker because of its ubiquitous distribution and slow rate of evolution (Fothergill-Gilmore and Michels 1993). Also, up-regulation of plant GAPDH under stress conditions implies that enhanced rates of glycolysis are one of the immediate responses to stress (Russell and Sachs 1989; Yang et al. 1993; Velasco et al. 1994; Laxalt et al. 1996). The identification, molecular characterization, and understanding of the processes by which plant GAPDH activities limit cellular damage may represent a critical strategy for surviving stress in plants; therefore, they could have a potential to improve stress tolerance in these species.