Sampling from Breeding and Association Populations
White spruce pedigrees: A major resource for our project has been a partial diallel cross planted made in 1992 by MoFR. This diallel involves trees in the breeding program evaluated for weevil resistance. From the 44 crosses with 75 progeny per cross that they planted, we selected a diallel involving 3 putative resistant females and 2 susceptible males (6 x 75 = 450 total sample size). These trees were scored for number of weevil attacks, kills and failed attacks, each year dating from 1997.
For construction of the genetic map of white spruce for both COS markers and candidate genes for resistance and adaptation, we work with two 90-member full sib families, both of which share "PG29" as the female parent (the two male parents are different). PG29 is a tree for which we have an extensive EST collection of ca. 45,000 sequences. Many of our candidate genes that we work with have SNPs that can be identified in silico from this collection, for high throughput SNP assay via the Illumina system.
- Sitka Spruce pedigrees: Relying upon the work of John King, the Sitka spruce breeder for the BC MoFR, we have identified the best pedigree for QTL mapping in Sitka spruce: a complete 3x3 diallel (excluding self crosses). 90 progeny are growing in a replicated design; the trees were planted in 2004 and there are as yet no weevil attacks.
- Loblolly pine pedigrees: For a comparative map with loblolly pine, our collaborators at UC Davis have supplied us with DNA from two mapping populations, both of size 90, which they use in their own mapping project.
- White spruce association populations: Large-scale progeny trials conducted by the BC MoFR starting in 1972 involving 173 families from the Prince George breeding zone and 116 families from the Prince Rupert breeding zone, are being used for a study of the association between candidate SNPs and weevil resistance. Details of sampling strategy are provided below.
- Sitka spruce association populations: Branch cuttings were rooted from 500 individuals collected from across the species range. The rooted cuttings will also provide genetic material for a growth chamber experiment designed to separate candidate gene expression and cold acclimation phenotype responses to low temperature and critical night length. We have also sown the offspring of 215 open-pollinated mother trees from southern Vancouver Island to test associations between candidate SNPs and cold acclimation and phenology phenotypes within a single large population.
Provenance Expression Profiling
To detect genes that evolve by natural selection among spruce species, which can serve to identify candidate genes for SNP association studies, we have designed and conducted an expression profiling study involving seven spruce species. We collected mRNA from 7 species of spruce, using bud tissue standardized for phenology. A microarray experiment involving these species was performed (3 individuals per species, 21 hybridizations total). We identified genes that evolve via expression more rapidly than expected (diversifying selection), or less rapidly than expected (normalizing selection), by comparing among vs. within-species variation of gene expression levels. Among the 21,334 clones in the spruce microarray, 42% were inferred to be under diversifying selection, 50% were neutral, and 8% were under stabilizing selection. Estimates of genetic distance of AFLPs between species can be used to remove the confounding effect of genetic distance upon these comparisons (manuscript in progress, contact Kermit Ritland for more information).
We will also collect material to examine patterns of evolution within white spruce (among provenances within BC). Heritability of gene expression, and interactions with the environment, will be estimated using controlled crosses conducted by the BC MoFR.
eQTL and mQTL Mapping
In May 2006, tissue was collected from the 6 crosses described above, totaling 417 trees. The top three-quarters of each leader was harvested and the bark-phloem and xylem tissue separated and collected for mRNA and DNA isolation. At the time of sampling, weevil attack and severity was recorded as i) no attack, ii) very low, iii) low, iv) high or v) very high. As well, the BC MoFR, and the Canadian Forest Service, had extensive data on past weevil infestation. After scoring and collection, and removal of all extraneous samples (i.e. dead leader, other insect infestation other than weevil, aphids, physical damage, etc.), DNA extraction was performed. Also, upper laterals that appear to be leader-like were taken from 263 individuals of the four crosses, fixed in FAA and stored in 70% ethanol. Those samples will be investigated for 4-5 attributes that can be rapidly detected through light-microscopy.
As of October 2007, RNA extraction is completed and the microarrray assays are largely completed (expected completion mid-November 2007).
For genetic mapping, we constructed a framework genetic linkage map with 253 informative gene markers, using the Illumina genotyping system. We did a preliminary QTL study based on available oviposition data using the marker regression method; in addition we identified more loci which showed significant genotype differences for extreme phenotypes (including accumulated weevil attacks). We used those marker loci to identify the best four crosses in a diallel mating scheme and to design our microarray experiment.
For eQTL mapping, we are using the "distant pair method", with the goal of increasing the power of detecting eQTLs related with weevil resistance by choosing (after genotyping) sib-pairs that are most mutually unrelated to each other. Currently we have extracted RNA from 188 individuals that have been identified as members of distant pairs, to be paired up on 94 microarray slides. In order to be able to relate expression results to constitutive resistance traits directly measured on the tree we collected material for histology purposes.