High-Throughput Genome-Wide Genotyping, Targeted Sequencing and Association Mapping of Adaptive and Breeding Traits in Loblolly Pine (Pinus taeda L.) Populations

Loblolly pine (Pinus taeda L.) is the most important forest tree for timber and paper production in the southeastern US and worldwide. We used genotypic data based on a genome-wide set of 4,264 SNPs representing expressed loci to study association between SNPs and adaptive and breeding value related traits in more than 1,000 trees in the first-generation selection (FGS) and second generation selection (SGS) breeding populations of loblolly pine from the Western Gulf Forest Tree Improvement Program. The SNP genotypes of ~450 trees from 27 counties across East Texas were also used in the landscape genomics study to find significant associations with environmental variables, such as monthly and annual temperature and precipitation, and two complex variables - Growing Degree Days representing annual growing season, and Aridity Index representing ratio of precipitation to evapo-transpiration. Correlation of county specific allele frequencies with various environmental variables was tested using Bayesian methods. Clinal variation in allele frequencies was tested with logistic regression. An F_SToutlier approach was used to identify signatures of selection.

Numerous associations were found between various phenotypic traits, environmental variables and SNPs. Signatures of diversifying and balancing selection were found in markers representing important biological functions. These results present the first step in the application of marker assisted selection (MAS) to the Western Gulf Forest Tree Improvement Program (WGFTIP) for loblolly pine and will contribute to the knowledge base necessary for genomic selection technology. Implications of these results for breeding, genetic conservation and towards understanding local adaptation will be discussed.

In addition we tested the genome target sequencing technique for re-sequencing particular genome regions and for high-throughput SNP discovery and genotyping by sequencing (GBS). The Agilent SureSelect Target Enrichment method was applied to capture unigene-based targeted genomic sequences in loblolly pine. We used 35,386 out of 35,550 unigenes (http://bioinfolab.muohio.edu/txid3352v1) to design 647,634 oligonucleotide hybridization probes (baits). To make this approach more affordable for population genomic studies, we tested different levels of multiplexing. The individual DNA samples were barcoded and multiplexed in the pools composed of from two (2X) up to eight (8X) indexed individual DNA samples for hybridization and following genotyping by sequencing. After capturing the targeted sequences, all samples were sequenced using Illumina Hiseq2000 and paired-end sequencing (2×100 bp). The obtained several hundred million high quality reads per pool were mapped to the original unigenes and to the draft loblolly pine reference genome assembly provided by the PineRefSeq project (http://pinegenome.org/pinerefseq) using CLCbio, BWA and SAMtools software. On average, more than 90% of the reads in multiplexed pools were mapped to the draft loblolly pine reference genome assembly. SNP detection was done with each library using SAMtools. On average, 6.5M SNPs were detected in the multiplexed pools. The results indicate that multiplexing strategies used in this target sequencing study were efficient and cost-effective. We plan to use this approach to genotype nearly the entire exome in more than 400 individual trees in an association mapping population to link allelic variation with phenotypic variation in important economic, adaptive and breeding traits.