Molecular Evolution of Regulatory Genes in Spruces from Different Species and Continents: Heterogeneous Patterns of Linkage Disequilibrium and Selection but Correlated Recent Demographic Changes

Abstract

Genes involved in transcription regulation may represent valuable targets in association genetics studies because of their key roles in plant development and potential selection at the molecular level. Selection and demographic signatures at the sequence level were investigated for five regulatory genes belonging to the knox-I family (KN1, KN2, KN3, KN4) and the HD-Zip III family (HB-3) in three Picea species affected by post-glacial recolonization in North America and Europe. To disentangle neutral and selective forces and estimate linkage disequilibrium (LD) on a gene basis, complete or nearly complete gene sequences were analysed. Nucleotide variation within species, haplotype structure, LD, and neutrality tests, in addition to coalescent simulations based on Tajima’s D and Fay and Wu’s H, were estimated. Nucleotide diversity was generally low in all species (average π = 0.002–0.003) and much heterogeneity was seen in LD and selection signatures among genes and species. Most of the genes harboured an excess of both rare and frequent alleles in the three species. Simulations showed that this excess was significantly higher than that expected under neutrality and a bottleneck during the Last Glacial Maximum followed by population expansion at the Pleistocene/Holocene boundary or shortly after best explains the correlated sequence patterns. These results indicate that despite recent large demographic changes in the three boreal species from two continents, species-specific selection signatures could still be detected from the analysis of nearly complete regulatory gene sequences. Such different signatures indicate differential subfunctionalization of gene family members in the three congeneric species.Electronic supplementary materialThe online version of this article (doi:10.1007/s00239-010-9335-1) contains supplementary material, which is available to authorized users.

Accuracy of genomic selection models in a large population of open-pollinated families in white spruce

“…Weak population structure in white spruce from Québec has been reported in several studies using neutral genetic markers (Jaramillo-Correa et al, 2001;Namroud et al, 2008Namroud et al, , 2010 and was expected based on our previous analysis of a subset of the present discovery population . As a cautionary measure, we used the multidimensional scaling coefficients as population covariates for the GS analyses to control for any potential bias in prediction accuracy that could be brought about by such a population structure.…”

“…A population structure analysis was used to assess whether there were differences in allele frequencies among unobserved ancestral populations that could bias prediction accuracy estimates. Given the results obtained in previous studies conducted in the same area (JaramilloCorrea et al, 2001;Namroud et al, 2008Namroud et al, , 2010Beaulieu et al, 2011), we were expecting a weak or no population differentiation from SNPs. We used all available SNPs (m ¼ 6385) in multidimensional scaling and principal component analysis (Price et al, 2006) to estimate population covariates for each individual.…”

“…Although it is true that LD appears to be low in white spruce genes in essentially unrelated trees from natural populations, many exceptions were found where LD was sizeable (Namroud et al, 2010;Pavy et al, 2012a). Therefore, the true coverage would lie between the number of SNPs sampled and the number of gene loci sampled.…”

The challenge of separating signatures of local adaptation from those of isolation by distance and colonization history: The case of two white pines

“…This corresponds to a scenario where gene flow is reduced among ecologically distant populations at loci directly involved in local adaptation or at closely linked loci, while there are no selective constraints on gene flow among environments for the rest of the genome (Barton, 2000; Gavrilets & Vose, 2005; Wu, 2001). This result is not surprising considering the rapid decay of linkage disequilibrium in large outcrossing populations of conifers (Namroud, Guillet‐Claude, Mackay, Isabel, & Bousquet, 2010) and the high levels of gene flow across the range of P. strobus (Mehes, Nkongolo, & Michael, 2009; Nadeau et al., 2015), which should uniformize among‐population genetic variation at neutral loci. Provenance trial studies have previously found moderate among‐population genetic variation for adaptive traits in P. strobus (e.g., height growth, bud phenology, cold hardiness; Li, Beaulieu, Daoust, & Plourde, 1997; Joyce & Sinclair, 2002; Lu, Joyce, & Sinclair, 2003a,b).…”

“…In similar comparisons made between Picea mariana and P. glauca (divergence time >10 MYA), Prunier, Laroche, Beaulieu, and Bousquet (2011) found more adaptive similarities at the gene family level (paralogs) than at the gene level (orthologs). The redundancy of functions among recently duplicated genes in conifers could have offered the possibility for selection to act on paralogous genes in different species (Namroud et al., 2010). In distantly related P. glauca and Pinus contorta (divergence time ~140 MYA), an exome‐wide study detected 47 genes (~10–18% of top candidate genes) with convergent signatures of local adaptation to low temperatures (Yeaman et al., 2016).…”