Viburnum L. (Adoxaceae) is a clade of approximately 165 species of shrubs and small trees that occur in northern temperate forests, the mountains of Central and South America, and subtropical montane forests of Southeast Asia. The phylogeny of Viburnum provides a clear understanding of relationships among major clades (Spriggs et al., 2015). However, evolutionary relationships within Viburnum clades that have experienced upward shifts in diversification rates, such as Oreinodentinus, are largely unresolved (Spriggs et al., 2015). Oreinodentinus is composed of Oreinotinus (ca. 32 species in Latin America; Killip and Smith, 1930; Morton, 1933) and Dentata (possibly three species native to eastern North America; Spriggs et al., 2015).

Phylogenetic analyses using plastid regions and the nuclear ribosomal internal transcribed spacer (ITS) region have supported the monophyly of Oreinotinus but have not fully resolved species relationships within the clade (Spriggs et al., 2015). Furthermore, relationships within the South American Oreinotinus clade are best described as a polytomy, and species boundaries are difficult to assess due to morphological similarity and ontogenetic variation. Although species in the South American Oreinotinus clade have been delimited based on morphological characters (Killip and Smith, 1930), our field studies suggest an evolutionary investigation will yield different species boundaries. More variable molecular markers are needed for such an analysis. Microsatellite loci (simple sequence repeats [SSRs]) have been developed to distinguish cultivated varieties of V. dilatatum Thunb. and closely related species (Dean et al., 2011) that belong to the distantly related Viburnum clade, Succotinus, of eastern Asia (Spriggs et al., 2015). Development of SSR loci specific to Oreinotinus will allow investigation of population dynamics and species boundaries within this group. We describe 16 novel microsatellite markers developed from V. hallii (Oerst.) Killip & A. C. Sm. (Oreinotinus) and V. trilobum Marshall (Opulus) and recovered from mining next-generation sequence (NGS) data for V. dentatum L. (Dentata) and V. triphyllum Benth. (Oreinotinus).

METHODS AND RESULTS

Construction of a microsatellite-enriched library and mining of NGS data were used to identify candidate loci. Viburnum hallii (collected from Ecuador) and V. trilobum (collected from Massachusetts, USA; Appendix 1) were used to construct microsatellite libraries (following V. Symonds, personal communication). Total genomic DNA was extracted from silica-dried leaves using a FastDNA kit (MP Biomedicals, Santa Ana, California, USA). DNA was digested using Sau3AI and was visualized using gel electrophoresis. Linkers constructed with SAU-LA and SAU-LB oligos were ligated to the DNA fragments for 16 h at 16°C. A nested PCR was used to verify linker ligation. PCR products were hybridized to a mix of (CA)_n and (GA)_n biotinylated probes. DNA fragments containing microsatellites were recovered from the PCR products using Streptavidin MagneSphere Paramagnetic Particles (Promega Corporation, Madison, Wisconsin, USA). SAU-LA primers were used to construct a second strand and repeat-enriched library. These products were cloned using a StrataClone PCR Cloning Kit (Agilent Technologies, Santa Clara, California, USA) and screened using T7 and M13R plasmid primers. Colonies with inserts containing repeat regions (144 selected from V. hallii and 144 from V. trilobum) were grown in liquid cultures and subjected to rolling circle amplification and sequencing in a single direction at the Interdisciplinary Center for Biotechnology Research at the University of Florida (Gainesville, Florida, USA). Using Primer3 (Rozen and Skaletsky, 1999), we designed primers for 22 fragments containing at least six dinucleotide repeats. PCR amplification of SSRs was as follows: 94°C for 3 min; 40 cycles of 94°C for 30 s, 52°C for 30 s, and 72°C for 45 s; and a 20-min extension at 72°C. Finally, we retained only those loci with sequences verified through cloning. Six to eight alleles from two different individuals per locus were isolated and sequenced using a StrataClone PCR Cloning Kit (Agilent Technologies) following the manufacturer's protocol with the exception of using half reactions. Sanger sequencing was performed at the DNA Analysis Facility on Science Hill at Yale University (New Haven, Connecticut, USA). Six loci were optimized.

We mined Illumina 100-bp paired-end NGS data from V. dentatum (collected from Connecticut, USA) and V. triphyllum (collected from Ecuador; SRP041815). First, data were assembled using reference-based assembly to a Lonicera L. plastid (M. Moore, personal communication) using the read mapping assembler in Geneious R8 (Biomatters, Auckland, New Zealand). Reads mapped to Lonicera were saved and thereafter considered plastid regions. Using the same approach, the unused reads were mapped to Helianthus L. mitochondria (NC023337). Mapped reads were saved and thereafter considered mitochondrial regions; unused reads were considered part of the nuclear genome. The plastid, mitochondrial, and nuclear sequence sets were subjected to de novo assembly using Velvet version 1.2.10 (Zerbino and Birney, 2008) within Geneious R8 (Biomatters). Putative SSRs that contained five or more perfect dinucleotide repeats were identified from the assemblies using MSATCOMMANDER (Faircloth, 2008) followed by primer design using Primer3 (Rozen and Skaletsky, 1999). Ten loci were optimized.

Table 1.

Characteristics of 16 microsatellite loci developed in Viburnum triphyllum and V. pichinchense.

All SSR loci were screened in 46 individuals from two populations of V. triphyllum (n = 16, n = 17) and one population of V. pichinchense Benth. (n = 13). Forward primers were complemented with a fluorescently labeled M13 primer (5′-CACGACGTTGTAAAAC-3′) for fragment detection (Schuelke, 2000). Loci were amplified in 10-µL reaction volumes containing 1 unit of GoTaq polymerase (Promega Corporation), 1× GoTaq Reaction Buffer, 25 mM MgCl₂, 0.5 mM dNTPs, 0.5 µM forward primer, 5 µM reverse primer, 8 µM labeled M13 primer, and 1.9–82.9 ng of DNA. Adding 1.5 µg of bovine serum albumin (BSA) improved amplification of some loci (Table 1). PCR amplification conditions were the same as described above. For locus DN16, a fluorescently labeled forward primer (5 µM per reaction) was used instead of the M13 system and amplified as follows: 94°C for 2 min; 30 cycles of 94°C for 20 s, 52°C for 20 s, and 72°C for 40 s; and final extension of 72°C for 5 min. Fragment analysis was performed at the DNA Analysis Facility on Science Hill at Yale University.

Polyploidy has been detected in two of the four Oreinotinus taxa that have been subject to chromosome counts (Egolf, 1962; Donoghue, 1982). Phylogenetic studies sampling a low-copy nuclear gene region (GBSSI) detected additional duplications in Viburnum clades where polyploidy had been confirmed, including Oreinotinus (Winkworth and Donoghue, 2004). Furthermore, we recovered three or more alleles per locus from 10% of the individuals sampled. For the purposes of these analyses, we first considered these viburnums to be allotetraploids. For nuclear loci, we calculated expected heterozygosity (H_e) and Shannon diversity index (H′) using ATETRA with 100,000 Monte Carlo simulations of possible allele combinations for partial heterozygotes (Van Puyvelde et al., 2010). We then calculated an additional diversity index that assumed autotetraploidy, Simpson diversity index (D), using the polysat package of R (Clark and Jasieniuk, 2011). For organellar loci, unbiased haploid diversity was calculated using GenAlEx (Peakall and Smouse, 2006, 2012). Rare alleles were grouped together as one haplotype.

Table 2.

Genetic properties of the 16 microsatellite loci for three populations of Viburnum triphyllum and V. pichinchense located in Ecuador.^a

Statistics per locus are in Table 2. Among nuclear loci, the number of alleles per locus per population varied from one to 13 alleles. H_e from 0 to 0.8975, H′ from 0 to 2.3670, and D from 0.0167 to 1.0000. For organellar loci, three to six alleles per locus per population were detected, and unbiased haploid diversity ranged from 0.756 to 0.853.

CONCLUSIONS

The 16 microsatellite loci developed for the South American V. triphyllum and V. pichinchense are variable and will be informative in studies of population dynamics and species boundaries among species of the Oreinotinus clade.