Buchloë dactyloides (Nutt.) Engelm. (buffalograss; Poaceae) is a low-growing, perennial C4 grass that is a dominant component of shortgrass prairies of the North American Great Plains (Shearman et al., 2004). Beyond this significant ecosystem role, buffalograss has been widely adopted as a drought-tolerant turfgrass alternative, particularly notable as a native-species option in North America. Like many dominant Great Plains grasses, B. dactyloides comprises an autopolypoid series, including diploids (2n = 20), tetraploids, pentaploids, and hexaploids (Johnson et al., 2001). Preserving the full range of buffalograss phenotypic and genotypic diversity and utilizing this diversity for crop improvement will require an understanding of the distribution of genetic variation among cytotypes and across its large geographic range.
Beyond numerous methodological advantages (Guichoux et al., 2011), microsatellites, or simple sequence repeat (SSR) markers, are an attractive genetic tool for studies of wide-ranging polyploid series given their codominant nature and applicability to museum-derived DNAs. Because SSR data are routinely obtainable from DNA extracted from museum tissue (Wandeler et al., 2007), these samples can be used to quickly and economically obtain comparative genotypic data from all portions of a large geographic range. Currently no buffalograss-specific SSR loci are available, as previous studies have relied on a mixture of dominant and codominant loci that were designed for other taxa (Budak et al., 2004). In this study, a set of SSR loci are designed from B. dactyloides genomic sequence data. The variability of these loci are then evaluated in six populations from numerous portions of the buffalograss range.
METHODS AND RESULTS
Silica gel–dried tissue was preserved from a B. dactyloides individual collected in Kiowa Co., Colorado, USA. A voucher specimen (Hadle 2228) has been deposited at the Arthur L. Youngman Herbarium at Wichita State University (WICH). Extraction with a DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany) yielded 2.4 µg of DNA, and an Illumina paired-end shotgun library was prepared from 1 µg of sheared DNA following the Illumina TruSeq DNA Library Kit protocol (Illumina, San Diego, California, USA) using a multiplex identifier adapter index. The library was sequenced (100-bp paired-end reads) on an Illumina HiSeq 2000. Five million of the resulting reads were screened with PAL_FINDER_v0.02.03 (Castoe et al., 2012) to extract those containing di-, tri-, tetra-, penta-, and hexanucleotide repeats. Such reads were batched to a local installation of Primer3 version 2.0.0 (Rozen and Skaletsky, 1999) for primer design. Single-copy loci were targeted by selecting those for which either primer sequence occurred 1–7 times among the 5 million reads.
Characteristics of 15 Buchloë dactyloides loci.a
Of the 5012 loci that met this criterion, 48 were selected and screened following O'Bryhim et al. (2013). Fifteen polymorphic loci were then evaluated in six B. dactyloides populations collected throughout the Great Plains (Appendix 1). Within each population, material from eight individuals separated by >30 m was preserved in silica gel. Voucher specimens for all individuals are archived at WICH. No cross-species amplification was attempted given the monotypic nature of the genus. Floral buds from one individual in five of the six populations (Appendix 1) were fixed in 3:1 100% ethanol : glacial acetic acid for 24 h and then transferred to 70% ethanol. Fixed anthers were macerated, stained with 1% acetocarmine, squashed following standard methods, and examined with brightfield microscopy at 1000× magnification. DNA was extracted with the high-throughput protocol outlined in Beck et al. (2012). Amplifications comprised 2 µL (20–25 ng) of genomic DNA, 2.5 µL of QIAGEN Multiplex PCR Master Mix, 0.8 µL of primer mix (see Table 1), and 2.7 µL of H2O. Loci were either tri- or tetraplexed, with individual primer concentrations optimized to reduce unequal locus amplification (Table 1). Cycling conditions included initial DNA denaturation at 95°C (15 min); 30 cycles of 94°C (30 s), 57°C (90 s), 72°C (60 s); followed by a final extension at 60°C (30 min). All samples were genotyped at the University of Chicago Comprehensive Cancer Center DNA Sequencing and Genotyping Facility (Chicago, Illinois, USA). Alleles were called with GeneMarker 1.91 (SoftGenetics, State College, Pennsylvania, USA), and diversity measures were calculated with GenoDive 2.0 (Meirmans and Van Tienderen, 2004). All measures were calculated using observed data only; missing data due to uncertain dosage were not inferred. Note that observed heterozygosity in these polyploid genotypes was calculated as “gametic heterozygosity,” the chance that two alleles drawn from an individual are different (Moody et al., 1993).
Genetic diversity in six Buchloë dactyloides populations at 14 newly developed microsatellite loci.a
All but one of the 15 loci were both highly variable and interpretable. One locus (Buda_4) was often difficult to interpret due to excessive stutter and was excluded from further analyses. A total of 271 alleles were identified, with 8–38 (mean 19.4) alleles per locus (Table 1). Mean expected heterozygosity (0.86) (Table 2) was notably higher than the average (0.65) for 71 SSR-based studies of outcrossing plants reviewed by Nybom (2004). This high level of within-population genetic variation is consistent with B. dactyloides' predominantly (but not exclusively) dioecious life history (Huff and Wu, 1992). Although variable and interpretable, one locus (Buda_14) consistently exhibited more alleles than expected given the known cytotypes in a set of 79 chromosome-counted specimens analyzed as part of a broader study (Hadle et al., unpublished). All five chromosome-counted specimens were tetraploids (2n = 20II). Consistent with their 4x cytotype (Appendix 1), each of these five chromosome-counted specimens exhibited a maximum of three or four alleles at the remaining 13 loci (excluding Buda_4 and_14).
The 14 optimized loci exhibited substantial variability within populations from numerous portions of the buffalograss range, and allelic variation at a set of 13 loci was consistent with the known cytotype in a set of chromosome-counted specimens. Ongoing studies in a set of >550 samples have established that these loci are readily amplifiable in herbarium-extracted DNAs of a wide age range (Hadle et al., unpublished), which will allow for the rapid determination of multilocus genotypes in a large set of samples representing all portions of the buffalograss range. These data have the potential to inform many aspects of buffalograss germplasm conservation and breeding programs, enhancing the conservation, crop, and ecosystem value of this dominant Great Plains grass.
The authors thank the Kiowa, Pawnee, Rita Blanca, and Thunder Basin National Grasslands for permission to sample. This work was supported by the Wichita State University Department of Biological Sciences and by the National Science Foundation (EPS-0903806) with matching support from the Kansas Board of Regents. Manuscript preparation was partially supported by the Department of Energy (DE-FC09-07SR22506) to the University of Georgia Research Foundation. Bioinformatics support came from Biostatistics and Bioinformatics Shared Resource of the University of Colorado Cancer Center (5P30CA046934).