Concordance between molecular and morphological evidence of hybridization in the Dichanthelium acuminatum (Poaceae: Paniceae) subspecies complex.
According to Lelong (1986) D. acuminatum (and subspecies) is probably the most polymorphic and troublesome species in the genus. The difficult and confusing synonomy and circumscription is the result of extensive morphological variation among members of the complex (Shinners, 1944; Freckmann, 1981; Lelong, 1986). Lelong (1965) studied taxa in Dichanthelium, including the acuminatum complex, and concluded that hybridization probably played an important role in obscuring boundaries among taxa. Spellenberg (1975) studied populations of some subspecies in the western United States and proposed that autogamy and hybridization were common means that accounted for much of the morphological variation, and thus, the taxonomic difficulty encountered in the complex.
Our study attempts to document putative hybridization among members of the D. acuminatum complex and of the genus. A major impetus for this research was discovery of a ready-made molecular marker in D. a. lindheimeri (Lindheimer's rosettegrass) from DNA-sequence data (Hammer, 2010). Lindheimer's rosettegrass is a common taxon in the group and easily is found in the field, often growing sympatrically with other species of Dichanthelium, adding to utility of this taxon for studies of hybridization.
DNA sequencing of the nuclear, granule-bound-starch-synthase gene (GBSSI or waxy) in the D. acuminatum subspecies complex and other species of Dichanthelium (Hammer, 2010) provided preliminary data for our study. The GBSSI gene has utility for phylogenetic studies in plants (e.g., Mason-Gamer et al., 1998; Aliscioni et al., 2003) and introns in GBSSI are variable at lower taxonomic levels (Mason-Gamer et al., 1998; Small, 2004). The GBSSI gene was first characterized by van der Leij et al. (1991) with structure consisting of one untranslated and 13 translated exons; structure appeared to be conserved. Sequence data from the intron region between exons 10 and 11 (using F-for and K-bac primers; TGC GAG CTC GAC AAC ATC ATG CG and GCA GGG CTC GAA GCG GCT GG, respectively; Mason-Gamer et al., 1998) revealed a 15 base-pair deletion present thus far only in subspecies D. a. lindheimeri. The nucleotide sequence of the 15 base-pair stretch can be determined from a GBSSI alignment comprising sequence data from all members of the D. acuminatum complex and other species of Dichanthelium (Hammer, 2010) and is 5-TGCGGCGAGCAATGT-3'.
DNA sequencing of numerous D. a. lindheimeri revealed that the 15 base-pair deletion is homozygous. This provides a serendipitous molecular marker useful to indicate the haploid presence of D. a. lindheimeri in a particular diploid genome of Dichanthelium. Given the possibility that hybridization with other species in the genus has contributed to intraspecific and intrapopulation morphological variation in the complex (and in grasses in general), it seemed worthwhile to develop a molecular marker based on the deletion in D. a. lindheimeri and to use the marker to analyze DNA collected from populations of Dichanthelium in the wild. A putative hybrid individual would contain one copy of the GBSSI gene with the 15 base-pair deletion and one copy of the gene that does not contain the deletion at this locus.
For field surveys of individual plants, it would be too costly and laborious to sequence the GBSSI region from each specimen to determine presence or absence of the deletion. Thus, a less-laborious and less-expensive, non-sequence-based method was needed to take advantage of this marker. Analysis of DNA-sequence data in the fragment amplified with the F-for and K-bac primers revealed a DNA restriction site for the restriction endonuclease Fnu4HI (Fusobacterium nucleatum) within the GBSSI region containing the 15 base-pair deletion. This presented the opportunity to use a relatively fast and inexpensive technique of genetic analysis called polymerase-chain-reaction- restriction-fragment-length-polymorphism (PCR-RFLP) to assay for presence or absence of the marker to determine if a specimen is of possible hybrid origin.
MATERIALS AND METHODS--An alignment of GBSSI-sequence data representing the majority of taxa (subspecies) in the D. acuminatum complex was analyzed for single nucleotide polymorphisms (SNPs) using the software SNP2CAPS (Thiel et al., 2004). This analysis identified a restriction site for the restriction enzyme Fnu4HI within the 15 base-pair deletion identified in specimens of D. a. lindheimeri. The SNP2CAPS analysis indicated no other Fnu4HI restriction site within ca. 100 base pairs upstream or downstream from the position of the15 base-pair deletion. Based on this information, PCR primers (L1-f, GCC CTG CGT GTG TGC ATC C and L4-r, CGA CCT TGA TGG CGC GCT TC; Hammer, 2010) were designed using the software program OligoCalc (Kibbe, 2007), which flanked the deletion site to amplify an ca. 200 base-pair fragment from the GBSSI gene.
The SNP2CAPS software also was used to digest representative sequences of the D. acuminatum complex in silico with Fnu4HI to identify expected genotypes. From this analysis, three Fnu4HI genotypes were identified along with the predicted number of restriction fragments and their (in silico) lengths. Genotypes were: 1) homozygous D. a. lindheimeri, 1 fragment of 187 base pairs; 2) homozygous non-D. a. lindheimeri, 2 fragments, 106 and 96 base pairs; 3) heterozygous D. a. lindheimeri + non-D. a. lindheimeri, 4 fragments, 202 base-pair heteroduplex fragment, 187 base pair fragment from parent of D. a. lindheimeri; 106 and 96 base-pair fragments resulting from restriction of the 15 base-pair indel-sequence present in the non-D. a. parent of D. a. lindheimeri.
The third Fnu4HI genotype above (heterozygote) actually is deduced from knowledge of the restriction patterns of genotypes 1 and 2. The heteroduplex fragment in the heterozygote, post-restriction-reaction pool arises from annealing of some of the 187 base-pair fragments from the parent of D. a. lindheimeri to complementary strands (post-restriction) of the 106 and 96 base-pair fragments from the parent of non-D. a. lindheimeri. This heteroduplex fragment does not cut with Fnu4HI because of its hybrid nature. A single-stranded loop is formed on the 187 base-pair, D. a. lindheimeri fragment when it encounters the DNA sequence for the 15 base-pair deletion, which is present on parts of the 106 and 96 base-pair fragments. This partial single-stranded character of the heteroduplex molecule affects its electrophoretic mobility during gel electrophoresis.
We collected a small portion of leaf tissue from an individual of D. a. lindheimeri or other species of Dichanthelium in the same population as D. a. lindheimeri. Samples were placed in a 1.5-mL microcentrifuge tube containing desiccant material (t.h.e dessicant, EM Chemicals, Inc., Gibbstown, New Jersey). The entire plant was collected, preserved as a voucher specimen, and deposited at the S. M. Tracy Herbarium (TAES; Texas A&M university). we collected 91 plants from 14 populations in southeastern and eastern Texas during spring of 2005 and 2006. Specimens collected and those from which tissues were sampled for DNA analysis were identified to species or subspecies using the key in Freckmann and Lelong (2003).
DNA was extracted from leaves of individual specimens by a micropreparation method. Fresh tissue (ca. 1 [cm.sup.2]) was placed in a 1.5-mL microcentrifuge tube and macerated using a blue teflon pestle attached to a power-drill. After addition of extraction buffer, tissues were further hand-macerated for 10 s. After brief centrifugation to remove intact solids, nucleic acids were precipitated with addition of isopropyl alcohol. After centrifugation, the pellet was resuspended in 10 mM/L tris pH 7.5, 1 mM/L EDTA then briefly centrifuged to remove undissolved solids. Following addition of NaOAc, nucleic acids were again precipitated with isopropyl alcohol. After a final centrifugation, the nucleic acid pellet was resuspended in 10 mM/L tris pH 7.5, 1 mM/L EDTA.
Fragments of nuclear GBSSI were amplified using primers L1-f and L4-r. Polymerase chain reaction (PCR) was performed in 25 mL total-reaction volumes of 5 ng DNA template, 12.5 mL Go-Taq Green PCR Master Mix (Promega, Madison, Wisconsin) and 0.25 mM of each primer. PCR-cycling conditions were: 1 cycle of 2 min at 94[degrees]C of initial denaturation, followed by 35 cycles of denaturation at 94[degrees]C, primer annealing at 65[degrees]C for 30 s, primer extension at 72[degrees]C for 2 min. A final extension step consisted of one cycle of 10 min at 72[degrees]C. PCR products were electrophoresed on 1.5% agarose tris-borate-EDTA (TBE) gels, stained with ethidium bromide, and visualized under uV light to verify amplification products.
Successful PCR amplifications were digested for 1 h at 37[degrees]C with the restriction enzyme Fnu4HI (New England BioLabs, Ipswich, Massachusetts) in 10 [micro]L total reaction volumes of 0. 5 [micro]L Fnu4HI, 1.0 [micro]L 10X buffer, 5.0 [micro]L of waxy PCR product, and 3.5 [micro]L reaction grade water. Digestion products were electrophoresed on 4% agarose TBE gels, stained with ethidium bromide, and visualized under uV light to reveal banding patterns.
RESULTS--Results of restriction analysis revealed the following genotypes from the 91 specimens of plants that were screened: 55 specimens had a homozygous genotype of subspecies D. a. lindheimeri, 34 had a homozygous genotype of non-D. a. lindheimeri, and 2 (Hammer 303 and 328) had a putative heterozygous genotype of D. a. lindheimeri plus non-D. a. lindheimeri.
DISCUSSION--The PCR-RFLP results for specimens Hammer 303 (collected in Marysee Prairie Preserve, Liberty County, Texas) and Hammer 328 (collected in Fort Boggy State Park, Leon County, Texas) show the predicted GBSSI restriction-fragment pattern for a diploid genotype with one copy of the locus contributed by a D. a. lindheimeri parent and the other copy contributed by a Dichanthelium parent other than D. a. lindheimeri. To confirm the hybrid status of these specimens, samples of DNA from both specimens were amplified using the F-for and K-bac pairs of primers to produce a larger amplicon more suitable for direct sequencing, but that still contained the smaller amplicon amplified using the L1-f+L4-r pair of primers. The resulting PCR amplifications were sequenced both forward and reverse using the F-for and K-bac primers and sequence chromatograms were inspected for evidence of heterogeneity in the form of multiple peaks along stretches of nucleotides at the location of the deletion. The DNA-sequence chromatogram for specimen 328 is shown in Fig. 1, which displays the pattern of multiple peaks that would be expected from sequence data produced from a PCR pool that contained a heterogeneous mixture of fragments; i.e., some containing the 15 base-pair sequence and other fragments not containing this sequence. In Fig. 1, the secondary base peaks are labeled on both the forward and reverse strands and sequences of both are in agreement with nucleotide sequence for the 15 base-pair deletion (TGC GGC GAG CAA TGT). This pattern of double peaks is present in chromatograms for the other specimen identified as a putative hybrid, specimen 303, with the PCR-RFLP analysis. The GBSSI DNA-sequence data provide confirmation that both specimen 303 and 328 are the products of hybridization between a D. a. lindheimeri parent and another unknown species of Dichanthelium.
Morphological phenotypes for specimens 303 and 328 exhibit a mostly typical morphotype of D. a. lindheimeri. When collected in the field, these specimens were identified as D. a. lindheimeri. However, there was doubt as to the identity of specimen 328. In addition to lower culms being fuzzy, specimen 328 has one morphological character, a peduncle trichome 2.3 mm long, which is atypical in that specimens of D. a. lindheimeri typically have little if any pubescence on the peduncle. In contrast, the other hybrid specimen of D. a. lindheimeri (303) has no measurable pubescence on the peduncle. Specimens of D. a. fasciculatum typically have a peduncle trichome 1-2 mm long. Both specimens 303 and 328 share an additional character state that is atypical for D. a. lindheimeri in that both have an internode trichome ca. 2.0 mm long (1.9 mm for specimen 303 and 2.1 mm for 328). Specimens of D. a. lindheimeri typically are glabrous along the internodes of culms. Other than these exceptions to the typical morphotype of D. a. lindheimeri, specimens 303 and 328 exhibit a predominant phenotype of D. a. lindheimeri and the herbarium specimens were annotated accordingly (Hammer 303 = TAES 246125; Hammer 328 = TAES 246126). No specimen of D. a. fasciculatum was observed when specimens 303 and 328 were collected. However, an exhaustive inventory of species of Dichanthelium occurring in areas around collecting sites was not conducted. Regarding specimen 328 (small trailside population), one specimen of D. a. acuminatum was collected along with 328. Compared to a typical specimen of D. a. fasciculatum, a typical specimen of D. a. acuminatum has more pubescence on all parts of the plant. A hybrid cross between a D. a. lindheimeri and a D. a. acuminatum conceivably could produce the phenotype described for specimen 328.
[FIGURE 1 OMITTED]
Hammer (2010) conducted a principal-coordinates analysis (PCoA) using 15 morphological characters on 391 specimens representing the D. acuminatum subspecies complex. The two confirmed hybrid specimens of D. a. lindheimeri, 303 and 328, were included in that analysis. One of the cluster plots from the PCoA is presented in Fig. 2.
In Fig. 2, there is a divergence of the main cluster of specimens of D. a. lindheimeri from specimens 303 and 328. Overall degree of pubescence increases from left to right across Fig. 2. In the plot of axis 1 with axis 3 in Fig. 2, 303 and 328 cluster between the glabrous D. a. lindheimeri group to the left and the pubescent specimens of D. a. fasciculatum (none of which was collected in the same county as 303 and 328) to the right side. Results of the PCoA seem to support the hybrid status of the genotypes as confirmed by results of the PCR-RFLP. In addition, the morphometric analysis provides preliminary evidence that D. a. fasciculatum might be the second parent contributing to the hybrid genotypes of 303 and 328. However, the working hypothesis that 303 and 328 have a D. a. lindheimeri x D. a. fasciculatum genotype would need to be corroborated by additional research. This is especially true for specimen 328, given that a specimen of D. a. acuminatum and not D. a. fasciculatum was found nearby when 328 was collected. Specifically, PCR products should be cloned to deconstruct the GBSSI sequences into individual alleles. Sequences of individual alleles could then be compared to sequences of subspecies of the D. acuminatum complex and other species of Dichanthelium to look for homology and possible molecular confirmation that D. a. fasciculatum, or possibly another taxon, is the second parent for hybrid specimens 303 and 328.
[FIGURE 2 OMITTED]
This study has provided documentation of hybridization at the molecular level between D. a. lindheimeri and another taxon of the genus Dichanthelium. Multivariate analysis seemed to detect a statistically significant hybrid signal in the morphometric dataset, which has provided complementary evidence to support the molecular evidence of hybridization for specimens 303 and 328. The hybrid phenotypes of 303 and 328 provide tangible morphological evidence of the role that hybridization plays in obscuring morphological boundaries between subspecies and contributing to taxonomic difficulties in the D. acuminatum complex. More generally, molecular and morphological evidence for at least occasional outcrossing (sometimes resulting in hybridization) among D. a. lindheimeri and other members of the D. acuminatum complex has been demonstrated.
We thank D. Riskind of Texas Parks and Wildlife Department for permission to study and collect plant material in Nails Creek and Fort Boggy state parks. Thanks also are extended to M. Johnston and the Texas Land Conservancy for permission to study and collect plants in the Marysee Prairie Preserve. This study was supported partially by a grant from the Frank W. Gould Research Award in Plant Systematics (Texas A&M university) made to RLH.
ALISCIONI, S., L. M. GIUSSANI, F. O. ZULOAGA, AND E. A. KELLOGG. 2003. A molecular phylogeny of Panicum (Poaceae: Paniceae): tests of monophyly and phylogenetic placement within the Panicoideae. American Journal of Botany 90:796-821.
CRINS, W.J.1991. The genera of Paniceae (Gramineae: Panicoideae) in the southeastern United States. Journal of the Arnold Arboretum, Supplementary Series 1:171-312.
FRECKMANN, R. W. 1981. Realignments in the Dichanthelium acuminatum complex (Poaceae). Phytologia 48:99-110.
FRECKMANN, R. W., AND M. G. LELONG. 2003. Dichanthelium. Flora of North America north of Mexico (Flora of North America Editorial Committee, editors) 25:406-450.
GOULD, F. W. 1974. Nomenclatural changes in the Poaceae. Brittonia 26:59-60.
GOULD, F. W., AND C. A. CLARK. 1978. Dichanthelium (Poaceae) of the United States and Canada. Annals of the Missouri Botanical Garden 65:1088-1132.
HAMMER, R. L. 2010. Systematic and evolutionary studies in the Dichanthelium acuminatum (Poaceae: Paniceae) complex. Ph.D. dissertation, Texas A&M University, College Station.
HITCHCOCK, A. S. 1950. Manual of the grasses of the United States. Second edition revised by A. Chase. United States Department of Agriculture, Miscellaneous Publication 200:1-1040.
HITCHCOCK, A. S., AND A. CHASE. 1910. The North American species of Panicum. Contributions of the United States National Herbarium 15:1-396.
KIBBE, W. A. 2007. OligoCalc: an online oligonucleotide properties calculator. Nucleic Acids Research 35(supplement 2):W43-W46.
LELONG, M. G. 1965. Studies of reproduction and variation in some Panicum subgenus Dichanthelium. Ph.D. dissertation, Iowa State University, Ames.
LELONG, M. G. 1986. A taxonomic treatment of the genus Panicum (Poaceae) in Mississippi. Phytologia 61:251-269.
MASON-GAMER, R. J. 2004. Reticulate evolution, introgression, and intertribal gene capture in an allohexaploid grass. Systematic Biology 53:25-37.
MASON-GAMER, R. J., C. F. WEIL, AND E. A. KELLOGG. 1998. Granule-bound starch synthase: structure, function, and phylogenetic utility. Molecular Biology and Evolution 15:1658-1673.
SHINNERS, L. H. 1944. Notes on Wisconsin grasses--IV. Leptoloma and Panicum. American Midland Naturalist 32:164-180.
SMALL, R. L. 2004. Phylogeny of Hibiscus sect. Muenchhusia (Malvaceae) based on chloroplast rpL16 and ndhF, and nuclear ITS and GBSSI sequences. Systematic Botany 29:385-392.
SPELLENBERG, R. W. 1975. Autogamy and hybridization as evolutionary mechanisms in Panicum subgenus Dichanthelium (Gramineae). Brittonia 27:87-95.
THIEL, T., R. KOTA, I. GROSSE, N. STEIN, AND A. GRANER. 2004. SNP2CAPS: a SNP and INDEL analysis tool for CAPS marker development. Nucleic Acids Research 32(e5):1-5
VAN DER LEIJ, F. R., R. G. VISSER, A. S. PONSTEIN, E. JACOBSEN, AND W. J. FEENSTRA. 1991. Sequence of the structural gene for granule-bound starch synthase of potato (Solanum tuberosum L.) and evidence for a single point deletion in the amf allele. Molecular and General Genetics 228:240-248.
Submitted 11 August 2010.
Accepted 2 April 2012.
Associate Editor was Janis K. Bush.
RICK L. HAMMER, * STEPHAN L. HATCH, ALAN E. PEPPER, AND JAMES R. MANHART
Department of Biology, Texas A&M University, College Station, TX 77843 (RLH, AEP, JRM)
Department of Ecosystem Science and Management, Texas A&M University, College Station, TX 77843 (SLH)
Present address of RLH: Department of Biology, Hardin-Simmons University, Box 16170, Abilene, TX 79698
* Correspondent: email@example.com
|Printer friendly Cite/link Email Feedback|
|Author:||Hammer, Rick L.; Hatch, Stephan L.; Pepper, Alan E.; Manhart, James R.|
|Date:||Jun 1, 2012|
|Previous Article:||Ingestion and nutrient budget of Attwater's pocket gopher (Geomys attwateri).|
|Next Article:||Restoration of grasslands and nesting success of dickcissels (Spiza americana).|