Date Published: October 18, 2018
Publisher: Public Library of Science
Author(s): Marian V. Lea, John Syring, Tara Jennings, Richard Cronn, Leo P. Bruederle, Jennifer Ramp Neale, Diana F. Tomback, Dusan Gomory.
Pinus albicaulis (whitebark pine) is a widely-distributed but rapidly declining high elevation western North American tree and a candidate for listing under the U.S. Endangered Species Act. Our objectives were to develop reliable nuclear microsatellite markers that can be used to assess within-population genetic diversity as well as seed and pollen migration dynamics, and to validate markers using two geographically proximal P. albicaulis populations. We identified 1,667 microsatellite-containing sequences from shotgun DNA libraries of P. albicaulis. Primer pairs were designed for 308 unique microsatellite-containing loci, and these were evaluated for PCR amplification success and segregation in a panel of diploid needle tissue. DNA was extracted with an SDS protocol, and primers were screened through gel electrophoresis. Microsatellites were genotyped through fluorescent primer fragment analysis. Ten novel and 13 transferred loci were found to be reproducible in analyses based on 20 foliage samples from each of two locations: Henderson Mountain, Custer Gallatin National Forest, Montana, and Mt. Washburn, Yellowstone National Park, Wyoming (USA). Transferred loci had higher numbers of alleles and expected heterozygosities than novel loci, but also revealed evidence for a higher frequency of null alleles. Eight of the 13 transferred loci deviated significantly from Hardy-Weinberg Equilibrium, and showed large positive FIS values that were likely inflated by null alleles. Mantel’s tests of transferred and novel markers showed no correlation between genetic and geographic distances within or among the two sampled populations. AMOVA suggests that 91% of genetic variability occurs within populations and 9% between the two populations. Studies assessing genetic diversity using these microsatellite loci can help guide future management and restoration activities for P. albicaulis.
North American forest trees have been exposed to a number of unprecedented health challenges that are likely to be exacerbated by warming temperatures . These challenges include outbreaks of native insects and especially bark beetles , mortality of older age classes from drought , and the accidental introduction and spread of destructive exotic pests and pathogens (e.g., ). In fire-prone communities in the western United States, intervals between fires are projected to decline over the next century, as climate warms [5, 6], and this may lead to metapopulation extirpation and the extinction of some woody plant species [7, 8]. As populations regenerate following disturbance, the recovery of community structure and genetic diversity is extremely important [9, 10].
To develop novel microsatellite loci for P. albicaulis, we identified di- and tri-nucleotide repeats using the procedure outlined in Jennings et al. , a strategy that was designed to identify very short (< 80bp) paired-end sequence data from early-generation Illumina sequencers. Genomic DNA from diploid needle tissue from a single tree (Custer Gallatin National Forest, Montana; 45.443°N, -110.005°W, 2,470 m a.s.l.) was used to make a standard Illumina library containing internal ‘barcoding’ adapters (barcode 5’-CACT; ). This library was pooled with 15 additional barcoded libraries from other conifer species, and five pmol were sequenced on an Illumina Genome Analyzer II using 80 bp paired end reads (University of Oregon Genomics Core Facility, Eugene, OR, USA). Libraries were sorted by barcode using a custom Perl script (https://github.com/listonlab/utility-scripts/blob/master/bcsort_v5.pl). For this strategy, paired microreads are concatenated and separated by 50 N’s (e.g., forward read + 50 N’s + reverse complement reverse read), and identical to nearly-identical sequences (> 95% identity) are filtered to a single representative sequence using cd-hit-454 (Niu et al. 2010). This allowed us to search for dinucleotide (>5 repeats) and trinucleotide (>4 repeats) motifs in the Read 1 or Read 2 sequences using SSR_pipeline , and to design flanking amplification primers in the Read 1 and Read 2 sequences using BatchPrimer3  and synthesized at Integrated DNA Technologies (IDT, Coralville, IA, USA). We used default settings for BatchPrimer3, with the exception of the following parameters: product length (min = 100 bp, max = 200 bp), primer length (min = 17 bp, max = 25 bp, optimum = 19 bp), melting temperature (min = 48°C, max = 63°C, optimum = 54°C); number of primers per sequence = 1. The original sequencing reads are deposited at the NCBI Short Read Archive under accession SRR7944190. Primers for 49 nSSR loci were screened for transferability to P. albicaulis, including 14 loci from P. cembra L. (Swiss stone pine) [64, 65], three loci from P. parviflora (Japanese white pine) , 13 loci from P. koraiensis Siebold & Zucc. (Korean pine) , and 19 loci from P. strobus L. (eastern white pine)  (S1 Table).
Of the sequences screened, we identified 1,667 (1,341 di- and 326 tri-nucleotide) microsatellite-containing sequences from shotgun DNA libraries of P. albicaulis. Primer pairs were designed for 308 unique microsatellite-containing loci, and these were evaluated for PCR amplification success and segregation in a panel of diploid needle tissue. Of these markers, ten novel P. albicaulis-derived loci amplified consistently and cleanly and were polymorphic in the two screened populations (Tables 1 and 2). The total number of alleles (NA) for these novel loci ranged from 2–9 in the combined populations, with an average of 3.3 alleles per locus (Table 2). Observed heterozygosity (HO) ranged from 0.02–0.50 (x¯ = 0.18) and expected heterozygosity (HE) ranged from 0.02–0.40 (x¯ = 0.18). All loci conformed with Hardy–Weinberg Equilibrium (HWE) (Table 2).
Our primary objective—the development of reliable nSSR markers useful for assessing genetic diversity in populations of P. albicaulis—was accomplished with the identification of ten novel and 13 transferred microsatellite loci. Of the 308 novel microsatellites identified from shotgun sequencing of P. albicaulis, only ten loci were sufficiently polymorphic and reliable to be identified for marker development, corresponding to a success rate of ~3%. This is approximately one-fifth the success rate of SSRs developed in Chamaecyparis lawsoniana (A. Murray) Parl. using these same methods . This conversion rate is extremely low for species of genus Pinus, which have already been characterized by low conversion rates, attributed to the complexity of their genomes (> 20Gbp) and the presence of large numbers of paralogous SSR families , , . The low yield is likely due to the selection of shorter motifs, which are known to show lower mutation rates than long microsatellite repeats . The success rate of SSR marker transfer from other white pines to P. albicaulis was 13 out of 49—almost 27%. The average number of alleles per locus of the transferred microsatellites (7.2) is more than twice that of the novel microsatellites (3.2), suggesting that the transferred loci may be useful in studies requiring high individual identification power, or detecting differences in spatial differentiation. For studies sensitive to the presence of null alleles and error in heterozygote identification, the markers derived from P. albicaulis are likely to perform better, albeit at the cost of variability. Notably, transferability of loci from the two east Asian species (P. koraiensis and P. parviflora) was substantially higher than for the European (P. cembra) and North American (P. strobus) species tested. This study highlights the opportunity for screening newly published novel microsatellites from closely related species in P. albicaulis, such as recently developed for P. sibirica . While these were not included in the present study, they also have a high likelihood of transferability, due to the close genetic affinity between P. albicaulis and P. sibirica.