UBC Faculty Research and Publications

Slow but not low: genomic comparisons reveal slower evolutionary rate and higher dN/dS in conifers compared… Buschiazzo, Emmanuel; Ritland, Carol; Bohlmann, Jörg; Ritland, Kermit Jan 20, 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12862_2011_Article_2013.pdf [ 724.44kB ]
JSON: 52383-1.0221287.json
JSON-LD: 52383-1.0221287-ld.json
RDF/XML (Pretty): 52383-1.0221287-rdf.xml
RDF/JSON: 52383-1.0221287-rdf.json
Turtle: 52383-1.0221287-turtle.txt
N-Triples: 52383-1.0221287-rdf-ntriples.txt
Original Record: 52383-1.0221287-source.json
Full Text

Full Text

Genome evolution and evolutionary systemsbiologyBuschiazzo et al.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8 (20 January 2012)RESEARCH ARTICLE Open AccessSlow but not low: genomic comparisons revealslower evolutionary rate and higher dN/dS inconifers compared to angiospermsEmmanuel Buschiazzo1,2*, Carol Ritland1, Jörg Bohlmann1,3 and Kermit Ritland1Background: Comparative genomics can inform us about the processes of mutation and selection across diversetaxa. Among seed plants, gymnosperms have been lacking in genomic comparisons. Recent EST and full-lengthcDNA collections for two conifers, Sitka spruce (Picea sitchensis) and loblolly pine (Pinus taeda), together with fullgenome sequences for two angiosperms, Arabidopsis thaliana and poplar (Populus trichocarpa), offer anopportunity to infer the evolutionary processes underlying thousands of orthologous protein-coding genes ingymnosperms compared with an angiosperm orthologue set.Results: Based upon pairwise comparisons of 3,723 spruce and pine orthologues, we found an averagesynonymous genetic distance (dS) of 0.191, and an average dN/dS ratio of 0.314. Using a fossil-establisheddivergence time of 140 million years between spruce and pine, we extrapolated a nucleotide substitution rate of0.68 × 10-9 synonymous substitutions per site per year. When compared to angiosperms, this indicates adramatically slower rate of nucleotide substitution rates in conifers: on average 15-fold. Coincidentally, we found athree-fold higher dN/dS for the spruce-pine lineage compared to the poplar-Arabidopsis lineage. This jointoccurrence of a slower evolutionary rate in conifers with higher dN/dS, and possibly positive selection, showcasesthe uniqueness of conifer genome evolution.Conclusions: Our results are in line with documented reduced nucleotide diversity, conservative genomeevolution and low rates of diversification in conifers on the one hand and numerous examples of local adaptationin conifers on the other hand. We propose that reduced levels of nucleotide mutation in large and long-livedconifer trees, coupled with large effective population size, were the main factors leading to slow substitution ratesbut retention of beneficial mutations.BackgroundDetermining the mutational and the selective forcesresponsible for evolution has overarching implications inbiology, e.g. in understanding what makes species uniqueand how organisms respond to biotic and abiotic chal-lenges. Identifying the rate of evolution and the patternsof nucleotide substitution underlying DNA evolution hasthus become a fundamental goal of molecular genomics[1,2]. Key to the central dogma of molecular biology, pro-tein-coding sequences (hereafter referred to as genes)have classically been regarded as a major unit ofevolution. Substitutions at synonymous (silent) and non-synonymous (replacement) sites are commonly distin-guished to differentiate between neutral (or at least weak)and active selective forces acting on genes, respectively.In pairwise comparisons of orthologous genes, the ratioof non-synonymous distance (i.e. number of substitutionsper non-synonymous site; dN) over synonymous distance(dS) gives a general but conservative indication of themode and strength of selection [1,2]. An excess of non-synonymous substitutions (dN/dS > 1) suggests adaptiveor diversifying selection, while an excess of synonymousmutations (dN/dS < 1) indicates purifying selection, andno difference between synonymous and non-synonymousmutation rates (dN/dS = 1) is taken as evidence forneutrality [3].* Correspondence: elbuzzo@gmail.com1Department of Forest Sciences, University of British Columbia, 2424 MainMall, Vancouver, BC V6T 1Z4, CanadaFull list of author information is available at the end of the articleBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8© 2012 Buschiazzo et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Large-scale sequence datasets now exist, allowing com-parisons to be made for thousands of genes in alldomains of life. Synonymous and non-synonymous sub-stitution rates have been found to vary widely within andbetween taxa [4-7]. From early studies based on a limitednumber of species and genes to the era of genomics andsystems biology [8,9], a complex blend of non-mutuallyexclusive biological, biochemical and demographicmechanisms emerged to explain these variations. Whileintraspecies differences are believed to be influenced byselection on protein structure and function (reviewed in[10-14]), interspecies differences are influenced by (i) theefficacy of the DNA repair machinery, (ii) life historytraits (e.g. generation time), (iii) metabolic rate, (iv) effec-tive population size (random genetic drift), (v) purifying(background) selection and (vi) reproductive strategy.Some factors (i - iii) influence the way mutations appear,while others (iv - vi) influence their fixation over genera-tions (reviewed in [9,13,14]).Among plants, most of the attention in comparativeevolutionary studies has been focused on flowering plants[4,5,14,15], and interest is now growing for other planttaxa as more sequence data is produced. Gymnospermsare separated from angiosperms by ~300 million years ofevolution [16]. Expectedly, many biological features ofgymnosperms and angiosperms differ greatly, includingseed morphology, life span, diversification rate, pollina-tion processes, environmental requirements and responseto environmental stresses. With ~600 extant species,conifers make up about two thirds of all gymnospermspecies, and are the dominant plants in most temperateand boreal ecosystems. Conifers have an immense ecolo-gical and economical value such as practical forestry eco-nomics, immediate ecological value of forest ecosystemsand in the long term, large capacity for carbon sequestra-tion. Biological differences between angiosperms andconifers and the need for long-lived conifer species tocope with challenges such as insect pests and environ-mental changes, underscore the importance of under-standing the molecular and functional evolution ofconifer genomes.The genetic architecture of conifers has been addressedby a wide variety of studies, mainly in pine (Pinus [17])and spruce (Picea [18]). Approaches include quantitativetrait locus mapping [19-21], candidate gene approaches[22,23], association mapping [24,25], BAC sequencing[26,27], transcriptome analysis [28,29], characterization ofgene families [30] and proteome analyses [31], and combi-nations thereof [32]. Missing from past endeavors, how-ever, are large-scale comparative comparisons thatinvestigate both evolutionary rates and the selective forcesacting on conifer genes.In this study, we take advantage of the existing largeand high-quality sequence data in two conifer species,Sitka spruce (Picea sitchensis) and loblolly pine (Pinustaeda), consisting of a collection of bona fide full-lengthcDNA sequences (FL-cDNAs) [33,34] and UniGenesconstructed from several EST libraries, respectively.Together with whole-genome gene sets available for twoangiosperms, Arabidopsis thaliana and Populus tricho-carpa; a rich data set exists to identify rates and patternsof evolution between conifer species and between coni-fer and angiosperm species. We find evidence for signifi-cantly slower evolutionary rates in conifers. In starkcontrast, we find a significantly higher dN/dS ratio inconifers as compared to angiosperms, indicating perhapshigher adaptation. We also investigate these patternsacross functional categories of genes.MethodsProtein-coding sequences for conifers and angiospermsConifer sequencesClustered ESTs from loblolly pine were downloaded fromNCBI UniGene (build 10, which had 18,921 clusters).Sitka spruce FLcDNAs came from the Treenomix II pro-ject [35]; as of Nov. 10 2009, this collection comprised10,665 FLcDNAs, of which 3,218 clustered in contigs. Weused all individual FLcDNAs because our approach ulti-mately removes any redundant or duplicated sequences.Open reading frame (ORF) search in conifer genesAll possible ORFs (from start to stop codons) found inspruce FLcDNAs were queried against the plant Uni-ProtKB SwissProt and trEMBL datasets [36], with pre-dicted proteins from Sitka spruce [33] removed from thetrEMBL dataset. Only ORFs from the 5,680 spruceFLcDNAs that had no hit against the SwissProt datasetwere queried against the trEMBL dataset. ORFs from3,296 spruce FLcDNAs had no homology with either ofthe plant UniProtKB datasets; for those, the longest ORFwas arbitrarily selected for further analysis. A singleFLcDNA with no ORF structure in its sequence wasdiscarded.We did not use the same strategy for loblolly pinebecause the pine UniGene set may contain only a trun-cated portion of the actual coding sequence. For conifers,we looked in each member of the UniGene set for theORF among all possible ORFs with the same frame as thelongest overlapping sequence with the best-scoringBLAST query against the spruce ORFs. Of 18,921 pineUniGenes, we found 7,627 ORFs in the same frame asspruce ORFs.Orthology of conifer genesWe used the reciprocal best hit (RBH) approach [37,38]to infer putative 1:1 orthologues between spruceFLcDNAs and pine UniGene sequences, using BLASTwith -e threshold = 10-20. We found a total of 4,774RBHs, of which 4,250 contained a complete ORF inpine.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 2 of 14Angiosperm orthologuesA. thaliana was chosen because it represents the bestcharacterized plant genome. Poplar was included in theanalyses as the first completely sequenced tree genome.A. thaliana coding sequences were downloaded fromthe TAIR9 annotation release [39]. Poplar codingsequences (annotation 1.1) were downloaded from theJGI Genome Portal [40]. We used Ensembl Comparapredictions through the BioMart server [41] to select alist of orthologous genes from Arabidopsis and poplar.Only 1:1 and apparent 1:1 orthologous coding sequenceswere retained for analysis, finalizing a set of 5,108orthologues.AlignmentGymnosperm (spruce-pine) and angiosperm (Arabidopsis-poplar) orthologous coding sequences were aligned usingDIALIGN-TX [42] with highest sensitivity (-L option).Gaps in the alignments and gap-free regions > 7 bp, inter-preted as non-homologous by DIALIGN-TX, wereexcluded from the analysis. Finally, alignments shorterthan 30 amino acids were discarded. The RBH coniferorthologue set contained 3,883 alignments and the angios-perm gene set totaled 5,073 successfully aligned 1:1orthologues.Data analysisSubstitution ratesPairwise distances at non-synonymous (dN), synonymous(dS) and 4-fold degenerate (4D) sites (d4) were estimatedfor individual genes in both gymnosperm and angiospermalignment sets using codeml (PAML 4.0) [43,44], with set-tings seqtype = 1, CodonFreq = 2, Runmode = -2, andtransition-transversion ratio () estimated from the data.Genes showing signs of saturated divergence wereexcluded because codeml results are reliable for moderateranges of sequence divergence. For conifers, we discarded42 orthologues with dN/dS = 98.99 and 118 with dS > 0.5,and for angiosperms, we discarded two genes with dN > 5and 996 genes with dS > 4. Threshold dS values weredetermined by plotting dN as a function of dS and exclud-ing outliers from the main distribution. Final RBH ortho-logue sets (see Additional file 1) contained 3,723 conifergenes (average gap-free length = 510 bp) and 4,080 1:1angiosperm genes (average gap-free length = 387 bp). 95%confidence intervals for evolutionary estimates were calcu-lated based on 1,000 bootstrap replicates using R [45].Absolute rates of substitution at coding sites (μ) in pair-wise comparisons were inferred using the formula:μ =d2Twith d the distance at synonymous (dS), non-synon-ymous (dN) or 4D (d4) sites; T divergence time betweenspruce and pine, or between Arabidopsis and poplar.Divergence times are documented from fossil records,between ~120 and ~160 MYA for conifers [46-51], andbetween ~105 and ~115 MYA for Arabidopsis andpoplar [52]. Unless mentioned otherwise, we used 140MYA and 110 MYA, respectively, as working divergencetimes.Analyses of functional categoriesFunctions of conifer orthologues were inferred usinganalogy with Arabidopsis proteins for GO annotations,and with plant proteins for descriptive annotation. Indetail, spruce ORFs were queried against the TAIR9protein-coding genes and the plant UniprotKB databaseusing BLASTX (-e threshold = 10-5). Of the 3,983 besthits against Arabidopsis, 1,230 contained an ORF thatsuccessfully aligned to loblolly pine ORFs and wereassigned the GO annotation corresponding to that ofthe best Arabidopsis hits, when available.For statistical comparisons among conifer genes, weused gene set enrichment analysis tools in the Babelo-mics platform [53], a web application that implementsthreshold-independent statistics (FatiScan and logisticregression) to investigate asymmetrical distributions ofGO terms, KEGG pathways and InterPro domains withinour list of annotated genes ranked by dN/dS. Fatiscanuses a Fisher exact test over a collection of partitions ofthe ranked list of genes, while the logistic model is usedto find association of each functional block with the highor low values of the ranked list; under- and over-repre-sented functional terms are then extracted. Prior to theseanalyses, we removed 43 genes that showed no non-synonymous substitution. For other functional analyses,we used the ‘GO Slim’ classification system provided byTAIR database [54].ResultsSubstitution rates in conifer protein-coding genesWe aligned the sequences of 3,723 spruce-pine ortholo-gous genes and inferred the number of pairwise synon-ymous (dS) and non-synonymous (dN) substitutions persite (see Table 1, Additional file 1). Mean dS was 0.191(95% confidence interval [CI] = 0.188, 0.193), meaningthat on average, one mutation occurred about every fivesites along both lineages since the common ancestor.Mean dN was lower than dS (0.049; CI = 0.048, 0.050),reflecting the expected elevated mutational constraint onnon-synonymous sites.Based on fossil records, the Pinus-Picea divergenceoccurred between 120 and 160 MYA [46-51]. Assumingan average divergence time of 140 MYA and that rateswere equivalent along both lineages, we inferred an aver-age rate of 0.68 × 10-9 (95% CI = 0.67 × 10-9, 0.69 × 10-9)substitutions per site per year at synonymous sites (μS, seeTable 1). However, to fully account for the uncertainty ofBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 3 of 14divergence time between pine and spruce, we also con-sider that this time is between 120 and 160 MYA, givingthe actual estimate of μS as lying between 0.60 × 10-9 and0.80 × 10-9.The neutral theory of molecular evolution predicts thatthe evolutionary rate at neutral sites corresponds to theactual mutation rate in an organism [55]. Because neutral-ity at synonymous sites is disputed [56], distance in a sub-set of synonymous sites known as 4-fold degenerate (μ4D)sites (i.e. sites where a change to any of the four nucleo-tides will not alter the amino acid during translation)stands as a better proxy to estimate the mutation rate.From our comparison in conifers, we inferred distance atμ4D sites (d4) at 0.177 (95% CI = 0.174, 0.179), whichtranslates into a substitution rate of 0.64 × 10-9 per 4D siteper year (μ4, see Table 1), and a range of 0.55 × 10-9 and0.74 × 10-9 using the extreme estimates of divergence timebetween spruce and pine.dN/dS in conifer protein-coding genesIdeally, dN/dS should be estimated at every site to findevidence of selection (which is only possible when com-paring more than two species in a phylogenetic context)and not averaged over the entire gene. However, an over-representation of non-synonymous substitutions can beused as a crude indication of either adaptive evolution orat least relaxed constraint in protein-coding genes. MeandN/dS in conifer genes was 0.314 (95% CI = 0.299,0.329). Of the 3,723 pairwise comparisons, 100 (2.68%)had a dN/dS > 1 (Additional file 2). We note the presenceof genes that are involved in abiotic and biotic stressresponse; some examples are protein kinases, proteinphosphatases, heat shock proteins, leucine-rich repeatproteins, histone modification proteins, glycosyltrans-ferases, and transcription factors (see Table 2).Genes with dN/dS lower than 1 can in fact be underpositive selection at specific sites [3] and dN/dSmeasured over the whole gene length is thus consideredtoo conservative to identify genes or groups of genesputatively under positive selection. Hence, we alsoapplied a segmentation test and a logistic regression testto look for functional groups of genes that are signifi-cantly and coordinately associated to high and/or lowvalues of dN/dS. Based on 1,230 GO-annotated conifergenes, we found that heat shock proteins, genes involvedin signal transduction and regulation of transcriptionand nucleic acids seem more likely to evolve underreduced constraint; whereas genes involved in transla-tion, protein assembly, chlorophyll biosynthesis and cel-lular organization are under strong selective constraint(Additional File 3).Comparison between gymnosperms and angiospermsWe compared evolutionary distances between tworepresentative conifer taxa, Sitka spruce and loblollypine, and two representative angiosperm taxa, Arabidop-sis and poplar (see Table 1). Mean dN in 4,080 Arabi-dopsis-poplar orthologous genes was 0.202 (95% CI =0.199, 0.205), mean dS was 2.184 (95% CI = 2.164,2.206), and mean d4 was 2.006 (95% CI = 1.985, 2.026).Based on a relatively confident divergence time of ~110million years [52], we inferred an average synonymousmutation rate μS of 9.93 × 10-9 substitutions per yearalong the lineages separating Arabidopsis and poplar(CI = 9.84 × 10-9, 10.03 × 10-9). This is 15-fold higherthan the average mutation rate found in conifer ortholo-gues (see Table 1). Even using the lowest estimate ofdivergence time between spruce and pine, μS is morethan 10-fold higher in angiosperms. Absolute rates ofsubstitution are calculated assuming equal rates on thepoplar and the Arabidospis lineages, but it has been sug-gested that the evolutionary rate in the poplar branch isone-sixth that of the Arabidopsis branch since divergence[57,58]. Using this factor, we obtained μS estimates ofTable 1 Substitution rates in conifer protein-coding genes compared to angiosperm genesPairwise comparison Gene number dS d4 dN μS (×10-9) μ4D (×10-9) μN (×10-9) dN/dSGymnosperms: 3,723 0.1908 0.1769 0.0492 0.68 0.64 0.18 0.3137Sitka spruceLoblolly pineAngiosperms: 4,080 2.1846 2.0057 0.2019 9.93 9.12 0.92 0.0924Arabidopsis 17.02 15.63 1.57Poplar 2.84 2.61 0.26Fold-changeAngiosperm:conifers 11.4:1 11.4:1 4.1:1 14.6:1 14.4:1 5.2:1 1:3.4Arabidopsis:conifers 25.0:1 24.7:1 9.0:1Poplar:conifers 4.2:1 4.1:1 1.5:1Mean genetic distances at synonymous (dS), 4-fold degenerate (d4) and non-synonymous (dN) sites are expressed as a number of substitutions per site. Absolutesubstitution rates are expressed in substitutions per synonymous (μS), four-fold degenerate (μ4D) and non-synonymous (μN ) site per year. Species-specific ratesfor angiosperms were estimated based on the 1:6 difference in evolutionary rate between poplar and Arabidopsis [57,58].Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 4 of 142.84 × 10-9 in the poplar lineage and 1.70 × 10-8 in theArabidopsis lineage (Additional file 1), which compareswell with 1.50 × 10-8, a previously known rate in Arabi-deae [59]. However, this rate has since been revised to7.5 × 10-9 with the recent finding that the divergencetime between A. thaliana and A. lyrata is about twice thepreviously known time, i.e. ~10 MYA instead of ~5 MYA[60].We also found a difference in μN between gymnos-perms and angiosperms (0.18 × 10-9 and 0.92 × 10-9mutations per year, respectively), representing a five-folddifference. If we account for the differential ratebetween the two angiosperm species, the difference forμN is 9-fold and 1.5-fold with Arabidopsis and poplar,respectively (see Table 1). Figure 1.A illustrates the dif-ference in dS and dN distributions between conifers andangiosperms, in particular the strikingly low dS esti-mates for conifers.Overall, our results indicate a relative over-representa-tion of non-synonymous mutations versus synonymousmutations in conifer species compared to angiospermspecies. Consequently, mean dN/dS is higher in conifersthan in angiosperms, i.e. 0.3137 and 0.0924, respectively,on average, and the distribution of dN/dS values for coni-fers extends towards and over unity (Figure 1.B). Whilewe found 100 conifer genes with dN/dS > 1 out of 3,723orthologues, there was a single Arabidopsis-poplarorthologue out of 4,080 orthologues that showed signs ofpositive selection over the entire alignment (dN/dS =1.8565). This gene (ORF25; TAIR ID: ATMG00640; Uni-Prot ID: Q04613) encodes a plant b subunit of mitochon-drial ATP synthase.We compared dN/dS between functional categories inconifers and gymnosperms, and consistently found higherdN/dS in conifers in most functional GO Slim categories(Figure 2; Mann-Whitney test, P < 0.05). However, ‘DNA/RNA metabolism’ (biological processes; P = 0.37), and‘chloroplast’ and ‘ribosome’ (cellular component; P = 0.46and P = 0.62, respectively) showed no significant difference.If synonymous mutations, and even more so mutationsat 4D sites, follow a neutral mode of evolution, we wouldexpect no significant difference in average μS betweenfunctional categories (Additional file 4). However, therewere significant disparities among some of the functionalcategories, even when considering the ‘more neutral’mutations at 4D sites (Kruskal-Wallis test; H = 52.831,P < 0.001), a surprising finding because it goes againstthe neutral expectancy. Interestingly, a recent study inbirds has found evidence for selective constraints at 4Dsites in the avian genome [61], and completes previousevidence accumulated in mammals [56]. Taken together,these results should call for careful attention when usingdS as an estimate of neutral mutation rate, especiallywhen inferring positive selection from dN/dS estimatesTable 2 Conifer genes involved in defense, resistance and response against insects with dN/dS > 1Spruce cloneIDPine UniGeneIDdN/dSUniProtIDSpecies Putative functionWS02821_B21 DT625383 7.3061 A7P5L0 Vitis vinifera Protein phosphatase/Serine/threonine phosphatasesWS0297_D22 CX645632 7.0185 A7QNM9 Vitis vinifera leucine-rich repeat family protein/binding proteinWS02725_C02 DR097823 6.5839 A7P656 Vitis vinifera Protein phosphatase 2C/hydrolase/metal-bindingWS02757_H19 DR165429 4.4902 Q9SE11 Funaria hygrometrica Chloroplast-localized small heat shock protein (HSP20)familyWS02758_N18 DR160912 4.4589 Q0DTD2 Oryza sativa subsp.japonicaHeat shock protein DnaJWS02741_E07 DT634060 3.3785 Q588B8 Cryptomeria japonica Glycoside Hydrolase Family 17WS02761 N01 CO365391 3.0817 A7PWA7 Vitis vinifera Heat shock protein DnaJWS02817_M06 DR093347 2.9656 A7NWZ2 Vitis vinifera serine/threonine-specific protein kinaseWS0272_J12 DR015390 2.3311 A7QFY4 Vitis vinifera Heat shock protein DnaJWS0454_E20 DR049906 2.2728 A0MMD5 Litchi chinensis Xyloglucan endotransglycosylase (Glycoside hydrolasefamily)WS02774_M01 DR060506 2.1005 Q6VAA9 Stevia rebaudiana UDP-glycosyltransferaseWS02749_F04 CO164226 1.7421 Q9MA24 Arabidopsis thaliana GlycosyltransferaseWS0288_C08 DR022129 1.5822 A7QTB5 Vitis vinifera Glycoside hydrolaseWS0292_O15 DR681862 1.294 A7P0R3 Vitis vinifera heat shock protein (hsp70)WS02729_N15 DR689530 1.1356 Q8LHS7 Oryza sativa subsp.japonicaHistone deacetylaseWS02716_E18 AI784893 1.1314 A5AWM3 Vitis vinifera Pathogenesis-related transcriptional activator PTI6WS0298_F15 DT638459 1.0692 A7QTU5 Vitis vinifera GlycosyltransferaseWS02725_E03 U39301 1.0481 A0ERF9 Cathaya argyrophylla Caffeic acid ortho-methyltransferaseBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 5 of 14or when applying molecular clocks. The present studydoes not claim positive selection but merely reports evo-lutionary trends; our results are therefore not signifi-cantly affected by the assumed neutrality of dS.DiscussionOur findings, based upon large-scale sampling ratherthan a small set of genes, are of significance for under-standing the differences in patterns of evolution betweenFrequency of genesdN/dS0.0 0.5 1.0 1.5 2.00100200300400ConifersAngiosperms0 1 2 3 4024681012Number of substitutions per sitedN conifersdS conifersdN angiospermsdS angiospermsDensity of genesABFigure 1 Distribution of evolutionary estimates for conifer and angiosperm protein-coding genes. A. Smoothed density plots of dS anddN estimates. B. Histogram plots of dN/dS estimates. Conifer genes with dN/dS > 2 (n = 38) are not shown. Breaks = 200.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 6 of 14transcription (65, 167)unknown biological processes (197, 1390)signal transduction (49, 134)other biological processes (153, 205)other metabolic processes (650, 1391)other cellular processes (690, 1524)protein metabolism (264, 492)developmental processes (123, 344)response to abiotic or biotic stimulus (144, 245)response to stress (148, 223)DNA or RNA metabolism (14, 128)electron transport or energy pathways (45, 65)transport (127, 298)cell organization and biogenesis (107, 257) binding or activity (6, 22)other molecular functions (49, 165)transcription factor activity (56, 144)hydrolase activity (150, 479)other enzyme activity (291, 637)other binding (193, 583)unknown molecular functions (153, 1080)protein binding (143, 392)transferase activity (156, 452)kinase activity (65, 202)DNA or RNA binding (94, 338)transporter activity (85, 197)nucleotide binding (133, 429)nucleic acid binding (56, 188)structural molecule activity (71, 54) (42, 70)other cellular components (74, 363)unknown cellular components (181, 1025)cell wall (299, 63)plasma membrane (214, 284)mitochondria (115, 253)plastid (159, 354)other membranes (265, 550)nucleus (184, 381)extracellular (18, 45)other cytoplasmic components (371, 653)other intracellular components (444, 886)cytosol (44, 82)Golgi apparatus (63, 46)chloroplast (103, 867)ribosome (70, 47) 2 dN/dS estimates in conifer and angiosperm genes across Arabidopsis’ GO slim functional categories. Mean dN/dS values forconifer (full circle) and angiosperm (open circle) protein-coding genes. Conifer genes were BLASTed against Arabidopsis gene transcripts, whoseGO Slim annotations were used for homologous conifer genes. Brackets represent the standard error of the mean. A: Biological processes; B:Molecular functions; C: Cellular component.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 7 of 14conifers and angiosperms. First, we found that evolu-tionary rates are dramatically lower in conifers than inangiosperms. Second, we find that such differences varyacross functional categories of genes.Classically, interspecific studies of protein-codinggenes in conifers have involved very few loci. Kusumi etal. [62] studied evolutionary rates of 11 genes in theCupressacea. Bouillé and Bousquet [63] compared poly-morphisms of three nuclear genes in Picea. Morerecently, Palmé et al. [64] scrutinized patterns of selec-tion in 21 nuclear genes in a pine phylogeny while Chenet al. [65] carried out similar analyses for 10 genes infour spruce species. Large-scale comparative approachesare needed to grasp global evolutionary trends represen-tative of conifer genomes.Genome-scale sequencing of conifer genomes is com-ing of age [26,27], in particular for two economically andenvironmentally important species of the Pinaceae: Sitkaspruce and loblolly pine. EST datasets for these specieshave previously been used in a comparative framework tofind conifer-specific genes [66] and studying the evolu-tion of gene families [67] and of xylem-specific genes[68] in vascular plants. Here, we carried out the firstcomparative study of substitution rates and mutationalpatterns in a sizable fraction of the conifer gene set - orthat of any gymnosperm.Lower rates of evolution in conifers as compared toangiospermsAre evolutionary rates slower in conifers and gymnospermsthan in angiosperms?We estimated evolutionary measures at 3,723 coniferorthologues and 4,080 angiosperm orthologues. As in anypartial list of ESTs (i.e. not genome-wide), there mighthave been an unintentional selection of particular func-tional categories of genes, but we believe that our gene setis large enough to be representative of the genome as awhole. We found a much smaller dS in conifers than inangiosperms (0.1908 and 2.1846, respectively; see Table 1).A practical consequence of this difference is that we dis-carded almost 10 times as many angiosperm genes beforefinal analysis; these genes showed a significant level ofgenetic saturation compared to conifer genes. Geneticsaturation artificially reduces sequence divergence becausemultiple mutations at any given site of a particularly fast-evolving gene cannot be ruled out. All considered, not dis-carding these genes would only increase the difference indS between conifers and angiosperms. Estimates of dNwere also lower in conifers than in angiosperms (0.0492and 0.2019, respectively), but the difference was not asdramatic as for dS (see Table 1, Figure 1.A), suggestingthat substitutions at synonymous sites are particularly con-strained - or that those at non-synonymous sites are lessconstrained, at equal mutation rate, in conifers ascompared to angiosperms. Although the causes for thispattern of substitutions in conifer genes are unclear, theanswer resides in what seems a unique picture of muta-tional processes and/or selective influences that affect con-ifer genes (see below).Using published divergence times, we inferred an aver-age synonymous mutation rate of 0.68 × 10-9 substitutionsper site per year in conifer genes (see Table 1); this is 15times less than the average rate in 4,080 Arabidopsis-poplar orthologues (μS = 9.93 × 10-9). If we account forthe lower (1:6) rate in the poplar lineage [57], the differ-ence is 25 times less in conifers than in Arabidopsis (μS =17.02 × 10-9), and four times less than poplar (μS = 2.84 ×10-9). We compiled a list of substitution rates that havebeen published for gymnosperms and angiosperms (Addi-tional File 5), and our findings fall well into the range ofrates reported for the two seed plant groups. For example,two phytochrome genes were shown to evolve at a synon-ymous rate of 0.48 × 10-9 per year in Pinus sylvestris andPicea abies [69]. For angiosperms, a rate of 1.5 × 10-8 peryear was commonly accepted for Arabidopsis [59] and theresulting 1:6 rate in poplar (2.5 × 10-9 per year) is also verysimilar to our results (Table 1). However, with a diver-gence time between A. thaliana and A. lyrata recentlyrevised at ~10 MY [60], the current estimate of the muta-tion rate in Arabidopsis has doubled. Although it isunclear how this relates to our results, it is important toacknowledge the uncertainty that exists in our results, inthe 1:6 poplar:Arabidopsis ratio and in timing divergence,even when relaxed molecular clocks are used.Interestingly, at the population level, conifers also exhi-bit lower nucleotide diversity despite high gene flow andlow population structure [65,70,71]. In addition, low sub-stitution rate and low nucleotide diversity in conifers areparalleled with reports of relatively low evolutionary ratesabove the nucleotide level. For example, angiosperms arehighly diversified while gymnosperms have experienced avery low speciation rate [72]. At least in birds, diversifica-tion has been shown to be positively correlated withmutation rate [73]. At the chromosome level, not only isthere little variation in the number of haploid coniferchromosomes (n = 11-13) with only scarce evidence ofwhole genome duplication and polyploidy [74] but com-parative genome maps also suggest that macrosynteny isconserved; making it possible to easily navigate acrossgenomes [75] and suggesting that conifer chromosomesare ‘fossilized’. There is on the contrary, a high rate ofchromosome evolution in angiosperms [72], as well asfrequent polyploidy and genome duplication events.Finally, Jaramillo-Correa et al. [76] found that recombi-nation, which has been correlated with levels of geneticdiversity, is lower in conifers compared to angiosperms.There are only a few known exceptions to this generaltrend of lower evolutionary rates in gymnosperms.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 8 of 14Conifers have larger genomes than angiosperms [74],partly due to larger gene families and abundance ofpseudogenes and partly due to a very high content inrepetitive DNA such as transposable elements [27,74].Possible elevated rates of gene duplication and transpo-sition could have occurred along the gymnosperm line-age to cause this genome expansion, with evidence todate suggesting that these events were ancient [77].Despite these exceptions, conifers exhibit dramaticallyslower evolutionary rates compared to angiosperms, inparticular substitution rates in protein-coding genes,suggesting the existence of conifer-specific evolutionarymechanisms.What are the causes for the slow substitution rates inconifer genomes?Substitution rates vary depending on rates at whichmutations appear in individuals and are fixed in thepopulation [9,13].First, the rate at which mutations appear is affected bythe efficacy of the DNA repair machinery, generationtime, and metabolic rate. In animals, mitochondrial genesevolve ten times faster than nuclear genes, but theinverse situation is found in plants [4]. This differencemay at least in part originate from the presence of theDNA repair gene recA in plant mitochondrial genomes,and its absence in those of animals [78]. To our knowl-edge, there is no information on the efficiency of the con-ifer DNA repair system compared to that of angiospermspecies. Life history traits such as generation time ortotal life span are factors that are commonly called forthto explain differences in evolutionary rates detectedbetween species, e.g. in mammals [79], in invertebrates[80] and in plants [81]. In angiosperms, rates of evolutionare higher in annuals than in perennials [15]. Our datasupports this finding as Arabidopsis (an annual) hashigher rates than trees. This accords with the germlinetheory of mutations [82]. However, generation timeeffects will be unknown until we can reconcile the differ-ence between cell lineage division time and generationtime in plants [14]. Conifers exhibit lower values ofnucleotide diversity at the population level despite highgene flow and low population structure [65,70,71] sug-gesting that trees accumulate fewer mutations per unit oftime than other plants and thus generation time is notsufficient to explain the annual-perennial difference inmutation rates. Finally, the low metabolite rate of conifertrees, with their large body size and temperate to borealhabitats [83], as well as reduced recombination rates [76],could generate fewer nucleotide substitutions in theirgenomes.Second, the fixation rate of new mutations depends onthe interplay between random genetic drift (i.e. effectivepopulation size and population structure), purifying(background) selection and reproductive strategy. Largepopulation sizes and extensive gene flow are often sug-gested as the causes of low synonymous polymorphismfound in conifer populations [58]. Both empirically andtheoretically, grey areas remain about the effect of effec-tive population size (Ne), population subdivision andselection on the pattern of nucleotide divergence betweenspecies [84-86]. Our results however support the inverserelationship between Ne and neutral substitution ratethat is expected by the “nearly neutral theory of molecu-lar evolution” [87]. In addition, with low diversificationrate in conifers [72], there have been fewer speciation-associated bottleneck events than in angiosperms, thuscontinuous low diversity between populations. That coni-fers are mainly outcrossing (selfing is generally avoidedthrough high early inbreeding depression) is only addingto the homogenization of populations. Indeed, studieshave shown that there is weak population structure inSitka spruce [88] and loblolly pine [89]. Finally, the influ-ence of background selection and other selective forcessuch as hitchhiking on the genomic reduction of substi-tution rate in conifers is mostly unknown, although selec-tive sweeps following bottlenecks have been reported forseveral loci [22,23,90].Teasing out the evolutionary mechanisms controllingthe rate of evolution in any organism is a daunting task.When comprehensive data are available across severalconifer and other gymnosperm species, comparative ana-lyses will help elucidate if, in what manner and to whatextent typical conifer features such as low metaboliterate, long generation time, large effective population andlow genetic structure affect substitution rates [91,92].Is the evolutionary slow-down similar between conifer andangiosperm trees?Conifers have high levels of genetic diversity within popu-lation but experience low nucleotide substitution rates andlow speciation rates. Strikingly, the same trend can beseen in angiosperm trees and all trees (angiosperm andgymnosperm) share common attributes that may explainthis similarity such as perenniality, outcrossed mating sys-tem and large population sizes [58,82]. However, vast evi-dences point at a more pronounced slow-down in coniferscompared to angiosperm trees, for example: recombina-tion rate [76], nucleotide diversity [58] and substitutionrates. In this study, we found that conifers have a lowersubstitution rate at both synonymous and non-synon-ymous sites than poplar (see Table 1). The existence ofconifer-specific factors that explain this difference is there-fore likely; gymnosperms have evolved separately fromangiosperms for about 300 MY. However, the exact natureand influence of these factors are still to be determined.High adaptability of conifers to their environmentWe found that mean dN/dS was about three timeshigher in conifers than in angiosperms (0.3137 vs.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 9 of 140.0924, respectively; see Table 1) despite much lowersubstitution rates in conifer protein-coding genes, andthat this trend was found throughout almost all func-tional categories. Higher dN/dS in conifers could be dueto a general low mutation rate and a high selective con-straint on synonymous mutations, which seems at oddswith the neutral expectancy but cannot be completelyruled out, or a general very low mutation rate but a pro-portionally lower constraint (relative to angiospermgenes) at non-synonymous sites. Assuming a relativelyhigh rate of amino acid change in conifer proteins, highaverage estimates of dN/dS in conifers have importantevolutionary implications, especially in light of the dis-tinctive biology of conifer trees.Characteristics of fast-evolving genes and functional genecategoriesAmong 100 conifer genes with dN/dS > 1, we found alarge fraction of genes involved in abiotic and biotic stressresponse. For example, we found two protein phospha-tases with dN/dS > 6, and one protein kinase with dN/dS~3 (see Table 2). Protein phosphatases and kinases actin tandem to regulate signaling pathways for plant stresstolerance or avoidance [93]. Four heat shock proteins, oneleucine-rich repeat protein, one histone modification pro-tein, two glycosyltransferases, four glycoside hydrolases,and seven transcription factors are also gene productsinvolved in defense, resistance and/or stress response.Other genes with dN/dS > 1 were involved in cell signal-ing, development and growth, vesicle trafficking andDNA/RNA binding. These single-gene results were paral-leled by a gene set analysis on 1,230 annotated genesranked by dN/dS, where functional categories involvedwith heat shock proteins, signal transduction and in theregulation of transcription and nucleic acids were morelikely to contain genes with high dN/dS (Additional File3). Conifers, like other long-lived sessile plants, requireresponsiveness and plasticity to defend themselves againstvarious herbivores and pathogens, as well as abiotic stres-ses (e.g. temperature and drought). This plasticity can forexample be obtained by regulating transcription andDNA/RNA binding proteins, which could explain whythese groups of genes seem to have experienced adaptiveselection in conifer lineages. In contrast, categories ofgenes involved in translation, protein assembly, cellularorganization and chlorophyll biosynthesis are under strongselective constraint (low dN/dS) because these processesare highly conserved across either the tree of Life, oracross photosynthetic organisms (i.e. chlorophyllbiosynthesis).Adaptability of conifersThe conifer divergence was dramatically slower atsynonymous sites than at non-synonymous sites (11-foldvs. 4-fold), suggesting that more adaptive mutations(and deleterious mutations, but see below) are fixed inconifers than in angiosperms. Indeed, there was a singleArabidopsis-poplar orthologue gene with a dN/dS > 1while values for other orthologues were below 0.6. Con-versely, we found a distribution of conifer dN/dS ratiossignificantly deviated near unity (Figure 1.B), with 100genes showing values suggesting positive selection (dN/dS > 1). In addition, all GO Slim functional categoriesshowed a significantly higher dN/dS in conifers than inangiosperms, with the exception of DNA/RNA metabo-lism and translation, which are evolutionary stable pro-cesses (Figure 2).A threshold of unity is usually applied to determine if agene shows signs of adaptive evolution, but this thresholdis overly conservative in the case of pairwise comparisonsover the whole length of the alignment. Algorithms existto identify adaptive mutations at specific sites and/or onspecific branches of a species tree, even when dN/dS < 1over the entire gene, but there is an implicit requirementfor comparisons of at least three species [3]. At the timeof this study, loblolly pine and Sitka spruce had signifi-cantly more publicly available sequences than any otherconifer, and we chose to restrain our study to two speciesand several thousands of genes, rather than opting foradditional species but a few hundreds of genes. Withmore sequences becoming available for conifer species[94], it will be possible to test for positive selection usingmodels of evolution across a tree composed of three ormore species.An overarching goal of modern biology is to uncover thegenetic architecture of biological adaptations. Our studysuggests that there is a substantial amount of adaptive sub-stitutions in two conifer species and we expect that thisfinding will be generalized to other conifer taxa, especiallyin environments where conifers compete in extreme eco-logical niches. For example, the Vietnamese pine hasevolved broad leaves, i.e. flattened needles, to compete forlight with evergreen angiosperm trees in tropical forests[95]. In Western North America, lodgepole pine hasevolved large and thick-scaled cones where squirrels areabsent but crossbills are present, while crossbills evolvelarger beaks [96]. An arms race between conifers and her-bivorous insects, such as bark beetles, results in the diver-sification of constitutive defense and stress-induced genesin conifers [97]. Sitka spruce and loblolly pine, like mostconifers in their natural environment, have been con-fronted by various endemic herbivorous pests, which wespeculate could be reflected by high dN/dS estimates atgenes involved in defense and stress response.Why do conifers show more signs of adaptive evolutionthan most plant lineages?Our results show that the low mutational rate seen inconifer genes is congruent with higher dN/dS, i.e. higheradaptability at the amino acid level, compared to angios-perm genes. At first, this relationship might seemBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 10 of 14contradictory and counter-intuitive; it is accepted thatmutations are the foundation for adaptation. In conifers,a combination of factors seems to have promoted astaggering high rate of fixation for non-synonymousmutations, despite a generalized low mutation rate.Little evidence has been found for adaptive evolution inangiosperm genes. In Arabidopsis thaliana and A. lyrata,purifying selection is the determinant force acting onamino acid substitutions [98]. In addition, Gossmann etal. [99] found little or no signal of adaptation in ninepairs of angiosperm species, except in sunflowers. Otherexceptions to this rule are European aspen [100] and thecrucifer Capsella grandiflora [101], where 30% and 40%of amino acid substitutions have been fixed by naturalselection, respectively. What differentiates sunflowersand C. grandiflora from the other studied angiospermsare low population genetic structure and especially largeeffective population size (Ne > 500,000). European aspenhas a lower reported Ne (118,000) but it has been arguedthat 500,000 individuals may not be unrealistic [100].Strasburg et al. [102] compared different species of sun-flowers, and found a positive correlation between Ne andlevels of adaptive divergence. Sunflowers, Europeanaspen and C. grandiflora are also outcrossing species butan excess of non-synonymous mutations was found inthe outcrossing A. lyrata [98], so mating system mayonly have limited effect on selective pressure comparedto demographic factors. Lastly, selfing A. thalianaappears to have rare adaptive substitutions, likely due toconsequent population subdivision and reduced Nethrough different bottleneck episodes [98,103,104].In conifers, investigations of sequence divergence atthe genome level have not been performed yet. Rese-quencing and comparative data have already provided alarge body of evidence that several individual genes inconifers species have evolved under positive selection[58,64,89]. In addition, there are various examples oflocal adaptation in conifer species, whereby a specificpopulation within the range of the species has expresseda phenotype adapted to an environmental constraint[105-107]. Concurrent with our results, the overall pic-ture from the study of molecular evolution of conifergenes is that ecology, demography, life history and gen-ome stability of conifers are favorable for the fixation ofnon-synonymous mutations. While fixation of deleter-ious mutations is reduced by outcrossing and largeeffective population size, most non-synonymous muta-tions are likely beneficial mutations in the conifer phyla.In addition, although deleterious mutations could befixed through bottlenecks and selective sweeps, it hasbeen shown that the time to establishment of complexadaptations is minimized in species with a large effectivepopulation size, even in the advent of deleterious inter-mediate steps [108].ConclusionsLarge-scale and genomewide comparative approaches gobeyond comparisons of small groups of candidate genesand provide global evolutionary trends. In this study, wefound that there was a dramatic slow-down in the overallmutation rate of conifer orthologues compared to angios-perm orthologues. This finding is compatible with anincrease in the fixation of non-synonymous mutations,which can be beneficial for adaptation. Large effectivepopulation size is likely the main factor that contributesto this trend, along with low population structure, lowrecombination and outcrossing mating system.Several genome sequencing projects in conifer speciesare now funded including for loblolly pine, Douglas fir,sugar pine, white spruce and Norway spruce. These datawill allow phylogenetic comparisons of much greaterpower then we currently employ. Not only should the pre-sent approach be expanded to a phylogenetic context, butfuture studies may also apply comparative methods totease out the evolutionary processes under various demo-graphic and ecological scenarios [91,92]. Finally, resequen-cing large numbers of candidate genes, once a referencegenome sequence is established, will further identify themode and strength of selection in conifer genomes.Additional materialAdditional file 1: Evolutionary measures for angiosperm andgymnosperm orthologues. Includes gene/transcript/EST IDs, ORFlength, aligned and analyzed length, and dN, dS and dN/dS estimates.Additional file 2: Annotation and dN/dS values for coniferorthologous genes. A more detailed description based on UniProt,PFAM and Interpro searches is provided for the 100 genes that showeddN/dS > 1, as well putative function where relevant.Additional file 3: Gene set analyses of conifer annotated genes(Fatiscan and logistic regression methods). Includes references toBabelomics and statistical methods, and results of over-representedcategories of genes with high and low dN/dS (adjuste p < 0.05, flasediscovery rate correction) in InterPro, KEGG pathways, and GO functionalcetegories.Additional file 4: dS estimates in conifer and angiosperm genesacross Arabidopsis’ GO Slim functional categories. Mean dS values forconifer (full circle) and angiosperm (open circle) protein-coding genes.Conifer genes were BLASTed against Arabidopsis gene transcripts, whoseGO Slim annotations were used for homologous conifer genes. Bracketsrepresent the standard error of the mean. A: Biological processes; B:Molecular functions; C: Cellular component.Additional file 5: Literature survey for plant mutation rates.List of abbreviationsBAC: Bacterial Artificial Chromosome; cDNA: complementary DNA; EST:Expressed Sequence Tag; FLcDNA: Full-length cDNA; GO: Gene Ontology;MYA: Million Years Ago; ORF: Open Reading Frame; RBH: Reciprocal Best Hit;4D: 4 fold degenerate.Acknowledgements and FundingThis work was supported by Genome British Columbia, Genome Canada,and the Province of British Columbia (Treenomix II/Conifer Forest HealthBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 11 of 14grant to KR and JB). We thank Stephen Ralph for the production of the Sitkaspruce FL-cDNA, Nancy Liao at the Michael Smith Genome Sciences Centrefor bioinformatics work, and Elizabeth Flavall for editing the manuscript.Author details1Department of Forest Sciences, University of British Columbia, 2424 MainMall, Vancouver, BC V6T 1Z4, Canada. 2School of Natural Sciences, Universityof California, Merced, 5200 North Lake Road, Merced, CA 95343 USA.3Michael Smith Laboratories, University of British Columbia, 2185 East Mall,BC V6T 1Z4, Canada.Authors’ contributionsEB participated in the design of the study, performed the analyses, anddrafted the manuscript. KR conceived of the study, and participated in itsdesign, analysis, and final write-up. JB was involved with the initial grantproposal, with the identification of genes important for secondarymetabolites, and grant leadership. CR was involved in project management.All authors read, revised and approved the final manuscript.Received: 28 July 2011 Accepted: 20 January 2012Published: 20 January 2012References1. Nielsen R: Molecular signatures of natural selection. Annu Rev Genet 2005,39:197-218.2. Hurst LD: Genetics and the understanding of selection. Nat Rev Genet2009, 10:83-93.3. Yang Z, Nielsen R: Codon-substitution models for detecting molecularadaptation at individual sites along specific lineages. Mol Biol Evol 2002,19:908-917.4. Wolfe KH, Li WH, Sharp PM: Rates of nucleotide substitution vary greatlyamong plant mitochondrial, chloroplast, and nuclear DNAs. Proc NatlAcad Sci USA 1987, 84:9054-9058.5. Drouin G, Daoud H, Xia J: Relative rates of synonymous substitutions inthe mitochondrial, chloroplast and nuclear genomes of seed plants. MolPhylogenet Evol 2008, 49:827-831.6. Britten RJ: Rates of DNA sequence evolution differ between taxonomicgroups. Science 1986, 231:1393-1398.7. Kumar S, Subramanian S: Mutation rates in mammalian genomes. ProcNatl Acad Sci USA 2002, 99:803-808.8. Nishant KT, Singh ND, Alani E: Genomic mutation rates: what high-throughput methods can tell us. Bioessays 2009, 31:912-920.9. Lanfear R, Welch JJ, Bromham L: Watching the clock: Studying variation inrates of molecular evolution between species. Trends Ecol Evol 2010,25:495-503.10. Rocha EPC: The quest for the universals of protein evolution. TrendsGenet 2006, 22:412-416.11. Pál C, Papp B, Lercher MJ: An integrated view of protein evolution. NatRev Genet 2006, 7:337-348.12. Warnecke T, Weber CC, Hurst LD: Why there is more to protein evolutionthan protein function: splicing, nucleosomes and dual-coding sequence.Biochem Soc Trans 2009, 37:756-761.13. Baer CF, Miyamoto MM, Denver DR: Mutation rate variation inmulticellular eukaryotes: causes and consequences. Nat Rev Genet 2007,8:619-631.14. Gaut B, Yang L, Takuno S, Eguiarte LE: The patterns and causes ofvariation in plant nucleotide substitution rates. Annual Review of Ecology,Evolution, and Systematics 2011, 42:245-266.15. Yue JX, Li J, Wang D, Araki H, Tian D, Yang S: Genome-wide investigationreveals high evolutionary rates in annual model plants. BMC Plant Biol2010, 10:242.16. Hedges SB, Dudley J, Kumar S: TimeTree: a public knowledge-base ofdivergence times among organisms. Bioinformatics 2006, 22:2971-2972.17. Plomion C, Chagné D, Pot D, Kumar S, Wilcox P, Burdon R, Prat D,Peterson D, Paiva J, Chaumeil P, et al: Pines. In Forest Trees. Edited by: KoleC. Berlin Heidelberg: Springer-Verlag; 2007:29-92.18. Bousquet J, Isabel N, Pelgas B, Cottrell J, Rungis D, Ritland K: Spruce. InForest Trees. Edited by: Kole C. Berlin Heidelberg: Springer-Verlag;2007:93-114.19. Hurme P, Sillanpaa MJ, Arjas E, Repo T, Savolainen O: Genetic basis ofclimatic adaptation in Scots pine by Bayesian quantitative trait locusanalysis. Genetics 2000, 156:1309-1322.20. Ukrainetz N, Ritland K, Mansfield S: Identification of quantitative trait locifor wood quality and growth across eight full-sib coastal Douglas-firfamilies. Tree Genet Genom 2008, 4:159-170.21. Pelgas B, Bousquet J, Meirmans P, Ritland K, Isabel N: QTL mapping inwhite spruce: gene maps and genomic regions underlying adaptivetraits across pedigrees, years and environments. BMC Genomics 2011,12:145.22. Eckert AJ, Wegrzyn JL, Pande B, Jermstad KD, Lee JM, Liechty JD, Tearse BR,Krutovsky KV, Neale DB: Multilocus patterns of nucleotide diversity anddivergence reveal positive selection at candidate genes related to cold-hardiness in coastal Douglas-fir (Pseudotsuga menziesii var. menziesii).Genetics 2009, 183:289-298.23. Ersoz ES, Wright MH, González-Martínez SC, Langley CH, Neale DB:Evolution of disease response genes in loblolly pine: insights fromcandidate genes. PLoS ONE 2010, 5:e14234.24. Quesada T, Gopal V, Cumbie WP, Eckert AJ, Wegrzyn JL, Neale DB,Goldfarb B, Huber DA, Casella G, Davis JM: Association mapping ofquantitative disease resistance in a natural population of loblolly pine(Pinus taeda L.). Genetics 2010, 186:677-686.25. Holliday JA, Ritland K, Aitken SN: Widespread, ecologically relevantgenetic markers developed from association mapping of climate-relatedtraits in Sitka spruce (Picea sitchensis). New Phytol 2010, 188:501-514.26. Hamberger B, Hall D, Yuen M, Oddy C, Hamberger B, Keeling CI, Ritland C,Ritland K, Bohlmann J: Targeted isolation, sequence assembly andcharacterization of two white spruce (Picea glauca) BAC clones forterpenoid synthase and cytochrome P450 genes involved in coniferdefence reveal insights into a conifer genome. BMC Plant Biol 2009, 9:106.27. Kovach A, Wegrzyn J, Parra G, Holt C, Bruening G, Loopstra C, Hartigan J,Yandell M, Langley C, Korf I, Neale D: The Pinus taeda genome ischaracterized by diverse and highly diverged repetitive sequences. BMCGenomics 2010, 11:420.28. Pavy N, Boyle B, Nelson C, Paule C, Giguère I, Caron S, Parsons LS,Dallaire N, Bedon F, Bérubé H, et al: Identification of conserved corexylem gene sets: conifer cDNA microarray development, transcriptprofiling and computational analyses. New Phytol 2008, 180:766-786.29. Verne S, Jaquish B, White R, Ritland C, Ritland K: Global transcriptomeanalysis of constitutive resistance to the white pine weevil in spruce.Genome Biology and Evolution 2011.30. Keeling C, Weisshaar S, Ralph S, Jancsik S, Hamberger B, Dullat H,Bohlmann J: Transcriptome mining, functional characterization, andphylogeny of a large terpene synthase gene family in spruce (Piceaspp.). BMC Plant Biol 2011, 11:43.31. Lippert DN, Ralph SG, Phillips M, White R, Smith D, Hardie D, Gershenzon J,Ritland K, Borchers CH, Bohlmann J: Quantitative iTRAQ proteome andcomparative transcriptome analysis of elicitor-induced Norway spruce(Picea abies) cells reveals elements of calcium signaling in the earlyconifer defense response. Proteomics 2009, 9.32. Hall DE, Robert JA, Keeling CI, Domanski D, Quesada AL, Jancsik S,Kuzyk MA, Hamberger B, Borchers CH, Bohlmann J: An integratedgenomic, proteomic and biochemical analysis of (+)-3-carenebiosynthesis in Sitka spruce (Picea sitchensis) genotypes that areresistant or susceptible to white pine weevil. The Plant Journal 2011,65:936-948.33. Ralph SG, Chun HJ, Kolosova N, Cooper D, Oddy C, Ritland CE, Kirkpatrick R,Moore R, Barber S, Holt RA, et al: A conifer genomics resource of 200,000spruce (Picea spp.) ESTs and 6,464 high-quality, sequence-finished full-length cDNAs for Sitka spruce (Picea sitchensis). BMC Genomics 2008,9:484.34. Lippert D, Yuen M, Bohlmann J: Spruce proteome DB: a resource forconifer proteomics research. Tree Genet Genom 2009, 5:723-727.35. Treenomix - Conifer Forest Health. [http://www.treenomix.ca/].36. Schneider M, Lane L, Boutet E, Lieberherr D, Tognolli M, Bougueleret L,Bairoch A: The UniProtKB/Swiss-Prot knowledgebase and its PlantProteome Annotation Program. J Proteomics 2009, 72:567-573.37. Hirsh AE, Fraser HB: Protein dispensability and rate of evolution. Nature2001, 411:1046-1049.Buschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 12 of 1438. Jordan IK, Rogozin IB, Wolf YI, Koonin EV: Essential genes are moreevolutionarily conserved than are nonessential genes in bacteria.Genome Res 2002, 12:962-968.39. The Arabidopsis Information Resource. [http://www.arabidopsis.org/].40. JGI Genome Portal. [http://genomeportal.jgi-psf.org/].41. Smedley D, Haider S, Ballester B, Holland R, London D, Thorisson G,Kasprzyk A: BioMart - biological queries made easy. BMC Genomics 2009,10:22.42. Subramanian A, Kaufmann M, Morgenstern B: DIALIGN-TX: greedy andprogressive approaches for segment-based multiple sequencealignment. Algorithms Mol Biol 2008, 3:6.43. Yang Z: PAML: a program package for phylogenetic analysis bymaximum likelihood. Comput Appl Biosci 1997, 13:555-556.44. Yang Z: PAML 4: Phylogenetic Analysis by Maximum Likelihood. Mol BiolEvol 2007, 24:1586-1591.45. R Development Core Team: R: A language and environment for statisticalcomputing. R Foundation for Statistical Computing, Vienna, Austria. ISBN3-900051-07-0, URL. 2011 [http://www.R-project.org/].46. Savard L, Li P, Strauss SH, Chase MW, Michaud M, Bousquet J: Chloroplastand nuclear gene sequences indicate late Pennsylvanian time for thelast common ancestor of extant seed plants. Proc Natl Acad Sci USA 1994,91:5163-5167.47. Wang XQ, Tank DC, Sang T: Phylogeny and divergence times in Pinaceae:Evidence from three genomes. Mol Biol Evol 2000, 17:773-781.48. Miller C: Mesozoic conifers. Bot Rev 1977, 43:217-280.49. Lin C-P, Huang J-P, Wu C-S, Hsu C-Y, Chaw S-M: Comparative chloroplastgenomics reveals the evolution of Pinaceae genera and subfamilies.Genome Biol Evol 2010, 2:504-517.50. Alvin KL: Further conifers of the Pinaceae from the Wealden formation ofBelgium Bruxelles: Institut Royal des Sciences Naturelles; 1960.51. Gernandt DS, Magallón S, Geada López G, Zerón Flores O, Willyard A,Liston A: Use of simultaneous analyses to guide fossil-based calibrationsof Pinaceae phylogeny. Int J Plant Sci 2008, 169:1086-1099.52. Bell CD, Soltis DE, Soltis PS: The age and diversification of theangiosperms re-revisited. Am J Bot 2010, 97:1296-1303.53. Medina I, Carbonell J, Pulido L, Madeira SC, Goetz S, Conesa A, Tarraga J,Pascual-Montano A, Nogales-Cadenas R, Santoyo J, et al: Babelomics: anintegrative platform for the analysis of transcriptomics, proteomics andgenomic data with advanced functional profiling. Nucl Acids Res 2010, 38:W210-213.54. Swarbreck D, Wilks C, Lamesch P, Berardini TZ, Garcia-Hernandez M,Foerster H, Li D, Meyer T, Muller R, Ploetz L, et al: The ArabidopsisInformation Resource (TAIR): gene structure and function annotation.Nucleic Acids Res 2008, 36:D1009-1014.55. Kimura M: The Neutral Theory of Molecular Evolution Cambridge: CambridgeUniversity Press; 1983.56. Chamary JV, Parmley JL, Hurst LD: Hearing silence: non-neutral evolutionat synonymous sites in mammals. Nat Rev Genet 2006, 7:98-108.57. Tuskan GA, Difazio S, Jansson S, Bohlmann J, Grigoriev I, Hellsten U,Putnam N, Ralph S, Rombauts S, Salamov A, et al: The genome of blackcottonwood, Populus trichocarpa (Torr. & Gray). Science 2006,313:1596-1604.58. Savolainen O, Pyhäjärvi T: Genomic diversity in forest trees. Curr OpinPlant Biol 2007, 10:162-167.59. Koch MA, Haubold B, Mitchell-Olds T: Comparative evolutionary analysisof chalcone synthase and alcohol dehydrogenase loci in Arabidopsis,Arabis, and related genera (Brassicaceae). Mol Biol Evol 2000,17:1483-1498.60. Beilstein MA, Nagalingum NS, Clements MD, Manchester SR, Mathews S:Dated molecular phylogenies indicate a Miocene origin for Arabidopsisthaliana. Proc Natl Acad Sci USA 2010, 107:18724-18728.61. Künstner A, Nabholz B, Ellegren H: Significant selective constraint at 4-folddegenerate sites in the avian genome and its consequence fordetection of positive selection. Gen Biol Evol 2011.62. Kusumi J, Tsumura Y, Yoshimaru H, Tachida H: Molecular evolution ofnuclear genes in Cupressacea, a group of conifer trees. Mol Biol Evol2002, 19:736-747.63. Bouillé M, Bousquet J: Trans-species shared polymorphisms atorthologous nuclear gene loci among distant species in the coniferPicea (Pinaceae): implications for the long-term maintenance of geneticdiversity in trees. Am J Bot 2005, 92:63-73.64. Palmé A, Pyhäjärvi T, Wachowiak W, Savolainen O: Selection on nucleargenes in a Pinus phylogeny. Mol Biol Evol 2009, 26:893-905.65. Chen J, Kallman T, Gyllenstrand N, Lascoux M: New insights on thespeciation history and nucleotide diversity of three boreal sprucespecies and a Tertiary relict. Heredity 2010, 104:3-14.66. Ujino-Ihara T, Tsumura Y: Screening for genes specific to coniferousspecies. Tree Physiology 2008, 28:1325-1330.67. Volokita M, Rosilio-Brami T, Rivkin N, Zik M: Combining comparativesequence and genomic data to ascertain phylogenetic relationships andexplore the evolution of the large GDSL-lipase family in land-plants. MolBiol Evol 2010.68. Li X, Wu H, Southerton S: Comparative genomics reveals conservativeevolution of the xylem transcriptome in vascular plants. BMC Evol Biol2010, 10:190.69. García-Gil MR, Mikkonen M, Savolainen O: Nucleotide diversity at twophytochrome loci along a latitudinal cline in Pinus sylvestris. Mol Ecol2003, 12:1195-1206.70. Heuertz M, De Paoli E, Kallman T, Larsson H, Jurman I, Morgante M,Lascoux M, Gyllenstrand N: Multilocus patterns of nucleotide diversity,linkage disequilibrium and demographic history of Norway spruce [Piceaabies (L.) Karst]. Genetics 2006, 174:2095-2105.71. Pyhäjärvi T, Garcia-Gil MR, Knurr T, Mikkonen M, Wachowiak W,Savolainen O: Demographic history has influenced nucleotide diversity inEuropean Pinus sylvestris populations. Genetics 2007, 177:1713-1724.72. Levin DA, Wilson AC: Rates of evolution in seed plants: Net increase indiversity of chromosome numbers and species numbers through time.Proc Natl Acad Sci USA 1976, 73:2086-2090.73. Lanfear R, Ho SYW, Love D, Bromham L: Mutation rate is linked todiversification in birds. Proc Natl Acad Sci USA 2010, 107:20423-20428.74. Ahuja MR, Neale DB: Evolution of genome size in conifers. Silvae Genet2005, 54:126-137.75. Pelgas B, Beauseigle S, Achere V, Jeandroz S, Bousquet J, Isabel N:Comparative genome mapping among Picea glauca, P. mariana × P.rubens and P. abies, and correspondence with other Pinaceae. Theor ApplGenet 2006, 113:1371-1393.76. Jaramillo-Correa J, Verdu M, Gonzalez-Martinez S: The contribution ofrecombination to heterozygosity differs among plant evolutionarylineages and life-forms. BMC Evol Biol 2010, 10:22.77. Friesen N, Brandes A, Heslop-Harrison JS: Diversity, origin, and distributionof retrotransposons (gypsy and copia) in conifers. Mol Biol Evol 2001,18:1176-1188.78. Lin Z, Kong H, Nei M, Ma H: Origins and evolution of the recA/RAD51gene family: Evidence for ancient gene duplication and endosymbioticgene transfer. Proc Natl Acad Sci USA 2006, 103:10328-10333.79. Welch J, Bininda-Emonds O, Bromham L: Correlates of substitution ratevariation in mammalian protein-coding sequences. BMC Evol Biol 2008,8:53.80. Thomas JA, Welch JJ, Lanfear R, Bromham L: A generation time effect onthe rate of molecular evolution in invertebrates. Mol Biol Evol 2010,27:1173-1180.81. Smith SA, Donoghue MJ: Rates of molecular evolution are linked to lifehistory in flowering plants. Science 2008, 322:86-89.82. Petit RJ, Hampe A: Some evolutionary consequences of being a tree.Annu Rev Ecol Evol Syst 2006, 37:187-214.83. Gillooly JF, Allen AP, West GB, Brown JH: The rate of DNA evolution:Effects of body size and temperature on the molecular clock. Proc NatlAcad Sci USA 2005, 102:140-145.84. Whitlock MC: Fixation probability and time in subdivided populations.Genetics 2003, 164:767-779.85. Woolfit M: Effective population size and the rate and pattern ofnucleotide substitutions. Biol Lett 2009, 5:417-420.86. Li J, Li H, Jakobsson M, Li SEN, SjÖDin PER, Lascoux M: Joint analysis ofdemography and selection in population genetics: where do we standand where could we go? Mol Ecol 2011.87. Ohta T: The nearly neutral theory of molecular evolution. Annu Rev EcolSyst 1992, 23:263-286.88. Gapare WJ, Aitken SN: Strong spatial genetic structure in peripheral butnot core populations of Sitka spruce [Picea sitchensis (Bong.) Carr.]. MolEcol 2005, 14:2659-2667.89. Eckert AJ, van Heervaarden J, Wegrzyn JL, Nelson CD, Ross-Ibarra J,Gonzalez-Martinez SC, Neale DB: Patterns of population structure andBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 13 of 14environmental associations to aridity across the range of Loblolly pine(Pinus taeda L., Pinaceae). Genetics 2010, 185:969-982.90. Palmé AE, Wright M, Savolainen O: Patterns of divergence among coniferESTs and polymorphism in Pinus sylvestris identify putative selectivesweeps. Mol Biol Evol 2008, 25:2567-2577.91. Mayrose I, Otto SP: A likelihood method for detecting trait-dependentshifts in the rate of molecular evolution. Mol Biol Evol 2011, 28:759-770.92. Lanfear R: The local-clock permutation test: a simple test to comparerates of molecular evolution on phylogenetic trees. Evolution 2011,65:606-611.93. Chae L, Pandey GK, Luan S, Cheong YH, Kim KN: Protein kinases andphosphatases for stress signal transduction in plants. In Abiotic StressAdaptation in Plants. Edited by: Pareek A, Sopory SK. Bohnert HJ: SpringerNetherlands; 2010:123-163.94. Rigault P, Boyle B, Lepage P, Cooke JEK, Bousquet J, MacKay JJ: A whitespruce gene catalog for conifer genome analyses. Plant Physiol 2011,157:14-28.95. Brodribb TJ, Feild TS: Evolutionary significance of a flat-leaved Pinus inVietnamese rainforest. New Phytol 2008, 178:201-209.96. Benkman C: Diversifying coevolution between crossbills and conifers. EvoEdu Outreach 2010, 3:47-53.97. Raffa KF, Berryman AA: Interacting selective pressures in conifer-barkbeetle systems: A basis for reciprocal adaptations? Amer Nat 1987,129:234-262.98. Foxe JP, Dar VU, Zheng H, Nordborg M, Gaut BS, Wright SI: Selection onamino acid substitutions in Arabidopsis. Mol Biol Evol 2008, 25:1375-1383.99. Gossmann TI, Song B-H, Windsor AJ, Mitchell-Olds T, Dixon CJ, Kapralov MV,Filatov DA, Eyre-Walker A: Genome wide analyses reveal little evidencefor adaptive evolution in many plant species. Mol Biol Evol 2010,27:1822-1832.100. Ingvarsson PK: Natural selection on synonymous and nonsynonymousmutations shapes patterns of polymorphism in Populus tremula. Mol BiolEvol 2010, 27:650-660.101. Slotte T, Foxe JP, Hazzouri KM, Wright SI: Genome-wide evidence forefficient positive and purifying selection in Capsella grandiflora, a plantspecies with a large effective population size. Mol Biol Evol 2010,27:1813-1821.102. Strasburg JL, Kane NC, Raduski AR, Bonin A, Michelmore R, Rieseberg LH:Effective population size is positively correlated with levels of adaptivedivergence among annual sunflowers. Mol Biol Evol 2011, 28:1569-1580.103. Nordborg M, Hu TT, Ishino Y, Jhaveri J, Toomajian C, Zheng H, Bakker E,Calabrese P, Gladstone J, Goyal R, et al: The pattern of polymorphism inArabidopsis thaliana. PLoS Biol 2005, 3:e196.104. Cao J, Schneeberger K, Ossowski S, Gunther T, Bender S, Fitz J, Koenig D,Lanz C, Stegle O, Lippert C, et al: Whole-genome sequencing of multipleArabidopsis thaliana populations. Nat Genet 2011, 43:956-963.105. King JN, Alfaro RI, Cartwright C: Genetic resistance of Sitka spruce (Piceasitchensis) populations to the white pine weevil (Pissodes strobi):distribution of resistance. Forestry 2004, 77:269-278.106. Mimura M, Aitken SN: Local adaptation at the range peripheries of Sitkaspruce. J Evol Biol 2010, 23:249-258.107. Grivet D, Sebastiani F, Alia R, Bataillon T, Torre S, Zabal-Aguirre M,Vendramin GG, Gonzalez-Martinez SC: Molecular footprints of localadaptation in two Mediterranean conifers. Mol Biol Evol 2010, 28:101-116.108. Lynch M, Abegg A: The rate of establishment of complex adaptations.Mol Biol Evol 2010, 27:1404-1414.doi:10.1186/1471-2148-12-8Cite this article as: Buschiazzo et al.: Slow but not low: genomiccomparisons reveal slower evolutionary rate and higher dN/dS inconifers compared to angiosperms. BMC Evolutionary Biology 2012 12:8.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitBuschiazzo et al. BMC Evolutionary Biology 2012, 12:8http://www.biomedcentral.com/1471-2148/12/8Page 14 of 14


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items