Castilleja sessiliflora Pursh is a generalist, hemiparasitic perennial in the Orobanchaceae Vent. with a wide host range (Crosswhite and Crosswhite, 1970). It is native to the shortgrass prairies of the Great Plains from southern Canada to northern Mexico, and extends east into Wisconsin and Illinois; it is classified as endangered in Illinois. The majority of the hemiparasites in the Orobanchaceae were previously placed within the Scrophulariaceae Juss.; however, molecular systematic studies demonstrated that the traditional circumscription of Scrophulariaceae was largely artificial (e.g., Olmstead et al., 2001).
Because the seeds of C. sessiliflora are gravity dispersed, it is likely that pollen movement plays a more important role in gene flow than seed dispersal. Several characteristics of the flowers of C. sessiliflora led Pennell (1935) to speculate that it is pollinated by lepidopterans. However, Crosswhite and Crosswhite (1970) observed that flowers in Wisconsin, USA, were only visited by Bombus fervidus Fabr. queens, although their observations were restricted to daytime hours when bee activity is typically high and when crepuscular insects are generally inactive. More recent observations in Illinois and Colorado reveal that C. sessiliflora is visited by at least one hawkmoth species, Hyles lineata Fabr. (J. Fant and K. Skogen, unpublished data). Interestingly, C. sessiliflora is the only known member of the genus for which hawkmoth visitation has been documented, and the effects of moth pollination on gene flow remain largely unexplored in this important pollinator group.
Despite being a relatively speciose genus, population genetic studies within Castilleja Mutis ex L.f. are surprisingly limited (two allozyme and one amplified fragment length polymorphism [AFLP] study) and have not used taxon-specific markers. Here, we characterize 12 microsatellite loci in C. sessiliflora for use in studies of gene flow, genetic structure, and diversity, and report cross-amplification in 24 additional Castilleja species.
METHODS and RESULTS
Microsatellite-enriched genomic libraries were developed by Genetic Identification Services (Chatsworth, California, USA; Jones et al., 2002). Libraries were enriched for four repeat motifs—(CA)n, (AAC)n, (AAG)n, and (ATG)n—and from a total of 144 sequenced clones, microsatellites were found in 22 out of 24 sequences for CA, 31 of 40 sequences for AAC, 28 of 40 sequences for AAG, and 26 of 40 sequences for ATG. Of the 107 sequences identified as containing microsatellites, PCR primers were designed for 33 regions ([CA]6, [AAC]9, [AAG]11, and [ATG]7) in DesignerPCR version 1.03 (Research Genetics, Huntsville, Alabama, USA) using the default parameters. These primer pairs were tested on a subset of C. sessiliflora individuals.
Genomic DNA was extracted from silica-dried leaf material using QIAGEN DNeasy kits (QIAGEN, Valencia, California, USA) for C. sessiliflora samples and the modified 2× cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle, 1987) for remaining species. DNA quantity was determined using a Nanodrop 2000 (Thermo Fisher Scientific, Wilmington, Delaware, USA), and samples were diluted to a final concentration of 5 µg/mL. To visualize samples, each forward primer was modified with the addition of an M13 sequence to the 5′ end (5′-CACGACGTTGTAAAACGAC-3′; Schuelke, 2000). An initial 10-µL PCR was conducted using 5 ng of template DNA, 25 µM of modified forward and reverse primer, and proprietary PCR MasterMix 2× (50 units/mL Taq DNA polymerase and buffer plus 400 µM of each dNTP; Promega Corporation, Madison, Wisconsin, USA). This PCR was run at 94°C for 3 min, followed by 15 cycles of 94°C for 40 s, 57°C for 40 s, and 72°C for 90 s, with a final extension of 72°C for 10 min. To this PCR product, an additional 5 µL of PCR mixture was added with the forward and reverse primer substituted with 25 µM of M13 primer labeled with either WellRED Black (D2), Green (D3), or Blue (D4) fluorescent dye (Sigma-Aldrich, St. Louis, Missouri, USA). With the additional label, the PCR was rerun at 94°C for 3 min, 27 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min, with a final extension of 72°C for 10 min. Products were analyzed and scored using a CEQ 8000 Genetic Analysis System version 9.0 (Beckman Coulter, Brea, California, USA).
Characteristics of 15 microsatellite primers tested on three Castilleja sessiliflora populations, two located in Colorado and one in Illinois, USA.
Results of initial primer screening in three populations of Castilleja sessiliflora. a
We tested a total of 33 primer pairs on a subset of C. sessiliflora individuals to identify primers that were polymorphic and amplified reliably. Of these, seven did not amplify (GenBank accession no.: JX983112, JX983116–JX983119, JX983121, JX983129), seven were weak or did not amplify consistently (GenBank accession no.: JX983114, JX983122–JX983127), one was monomorphic (GenBank accession no.: JX983120), three produced multiple peaks (GenBank accession no.: JX983113, JX983115, JX983128), and 15 were polymorphic (Table 1). The 15 polymorphic primer pairs were further tested on 32 individuals from each of three populations of C. sessiliflora (Colorado City [CC], Colorado, USA, Hilpman & Todd s.n., Chicago Botanic Garden Herbarium [CHIC] 15799; David's Canyon [DC], Colorado, USA, Hilpman & Todd s.n., CHIC 16794; and Illinois Beach State Park [IBSP], Illinois, USA, for which no herbarium specimens were collected and specific GPS coordinates are withheld, due to its conservation status in Illinois; Table 2). To evaluate the utility of these primers beyond the target species, the primer pairs were also tested on two individuals from a diverse sampling of Castilleja species representing the morphological and geographic diversity of the genus, with special attention given to the North American species from the rapidly radiating perennial clade of Castilleja (Tank and Olmstead, 2008). This sampling included 20 species from western North America (two annual and 18 perennial) and four species from Central and South America (one annual and three perennial; Table 3).
Results of cross-amplification of primers on 24 Castilleja species (two individuals screened per species). a,b
Primers were tested for the potential of null alleles, by population and globally, using exact tests in MICRO-CHECKER (van Oosterhout et al., 2004). Potential null alleles were identified in three of the 15 loci tested (CaSe_A01, CaSe_A101, and CaSe_A102; Table 1). Linkage disequilibrium was tested for each pair of loci across all populations using Fisher's method in GENEPOP (Raymond and Rousset, 1995). Of the 107 possible loci pairs, significant linkage disequilibrium (P < 0.05) was identified between CaSe_B103 and CaSe_D119 and between CaSe_B116 and CaSe_C104, although this was nonsignificant when sequential Bonferroni corrections were applied. For all loci, we report the following descriptive parameters: sample size, mean number of alleles, number of private alleles, observed heterozygosity, expected heterozygosity, and departure from Hardy—Weinberg equilibrium (HWE) (Tables 1 and 2; calculated in GenAlEx; Peakall and Smouse, 2006). The 12 loci that showed reliable amplification and allelic polymorphisms varied from three to 20 alleles per locus (Table 1). One locus, CaSe_B116, produced two separate peak regions separated by 36 bp; this may represent a duplicate annealing site within close vicinity. Significant departure from expected proportions under HWE was observed in five of the 12 loci for at least one population, although no loci showed significant deviation in all populations (Table 2). Finally, 15 loci (including CaSe_A01, CaSe_A101, and CaSe_A102) were tested on two individuals from different populations of 24 additional Castilleja species (Table 3). Nine loci produced bands in all species tested, while of the remaining loci, one worked in 23 species, four worked in 22 species, and one worked in 16 species. Monomorphism by loci varied from one to 10 (Table 3). Some loci produced more than two bands; this may suggest evidence of differences in ploidy, which is common among perennial Castilleja species, or, alternatively, these extra bands may be a result of stutter or spurious peaks that might disappear with more stringent and optimized PCR conditions.
Twelve microsatellite loci developed in C. sessiliflora were polymorphic and amplified reliably in the samples analyzed. In addition, all loci cross-amplified in 24 additional Castilleja species, with most loci revealing polymorphisms in more than half of the species tested. These loci will be useful for assessing patterns of gene flow, genetic diversity, and structure within and among populations of C. sessiliflora and other Castilleja species, and will contribute to investigations of species delimitation in this diverse and complex genus.