UBC Faculty Research and Publications

The Eucalyptus terpene synthase gene family Külheim, Carsten; Padovan, Amanda; Hefer, Charles; Krause, Sandra T; Köllner, Tobias G; Myburg, Alexander A; Degenhardt, Jörg; Foley, William J Jun 11, 2015

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

Item Metadata


52383-12864_2015_Article_1598.pdf [ 4.02MB ]
JSON: 52383-1.0223299.json
JSON-LD: 52383-1.0223299-ld.json
RDF/XML (Pretty): 52383-1.0223299-rdf.xml
RDF/JSON: 52383-1.0223299-rdf.json
Turtle: 52383-1.0223299-turtle.txt
N-Triples: 52383-1.0223299-rdf-ntriples.txt
Original Record: 52383-1.0223299-source.json
Full Text

Full Text

RESEARCH ARTICLEynducl its.de rCogmixture with antiseptic effects which is utilised in phar- that this variation is built on the largest family of terpeneKülheim et al. BMC Genomics  (2015) 16:450 DOI 10.1186/s12864-015-1598-x(GGPP - C20) are formed in the chloroplast, and farnesylpyrophosphate (FPP - C15) in the cytosol. These prenylEnvironment, Australian National University, Canberra 0200, AustraliaFull list of author information is available at the end of the articlemaceuticals and as a scent and flavour. Eucalyptus ter-penes also act to mediate ecological interactions includingdeterrents to insect herbivory [1,2], attractants [3] andrepellents to vertebrate herbivores [4], cues to other toxicsynthase genes of any plant yet sequenced.Terpenes are formed from C5 precursors, which areproduced either through the methylerithritol phosphatepathway (MEP) in the chloroplast or the mevalonatepathway (MVA) in the cytosol [15]. Geranyl pyrophos-phate (GPP - C10) or geranylgeranyl pyrophosphate* Correspondence: Carsten.kulheim@anu.edu.au1Research School of Biology, College of Medicine, Biology and theduce Eucalyptus oil, a 1,8-cineole-dominated terpenoid106 putative functional TPS genes in E. grandis and E. globulus, respectively. All but one TPS from E. grandis wereexpressed in at least one of seven plant tissues examined. Genomic clusters of up to 20 genes were identified.Many TPS are expressed in tissues other than leaves which invites a re-evaluation of the function of terpenes inEucalyptus.Conclusions: Our data indicate that terpenes in Eucalyptus may play a wider role in biotic and abiotic interactionsthan previously thought. Tissue specific expression is common and the possibility of stress induction needs furtherinvestigation. Phylogenetic comparison of the two investigated Eucalyptus species gives insight about recentevolution of different clades within the TPS gene family. While the majority of TPS genes occur in orthologouspairs some clades show evidence of recent gene duplication, as well as loss of function.Keywords: Eucalyptus, Myrtaceae, Terpene synthase, Essential oil, Monoterpenes, Sesquiterpenes, Evolution,Biodiversity, HerbivoryBackgroundEucalyptus dominates Australian forests and woodlandsand is truly the “essence” of Australia as well as beingthe dominant hardwood plantation tree in the world.High growth rates make eucalypts a desirable hardwoodplantation tree for pulp, sawmills and biofuels. Foliar ter-penes give eucalypts their characteristic odour, they areindustrially important and mediate many ecological in-teractions. Several Eucalyptus species are used to pro-constituents [5], mediators of resistance to fungal infec-tion [6], allelopathic agents [7], attractants for parasit-oids and pollinators [8], determinants of leaf litterdecomposition rates [9,10], mitigators to heat stress [11],as well as significant contributors to biogenic hydrocar-bons in cities [12].Although no terpenes are found exclusively in euca-lypts, striking variation can be observed in the foliar ter-pene profile within a single species [13] or even withinindividual branches of a single tree [14]. Here we showThe Eucalyptus terpene sCarsten Külheim1*, Amanda Padovan1, Charles Hefer2, SaJörg Degenhardt3 and William J Foley1AbstractBackground: Terpenoids are abundant in the foliage of Eas being valuable economically and influencing ecologicaintra- specific variation of terpenes is common in eucalypResults: The genome sequences of Eucalyptus grandis anand compared to other plant species. We investigated thfunctionally characterized five TPS genes from E. grandis.grandis has the largest number of putative functional TPS© 2015 Külheim et al.; licensee BioMed CentraCommons Attribution License (http://creativecreproduction in any medium, provided the orDedication waiver (http://creativecommons.orunless otherwise stated.Open Accessnthase gene familyra T Krause3, Tobias G Köllner4, Alexander A Myburg5,alyptus, providing the characteristic smell as wellnteractions. Quantitative and qualitative inter- andE. globulus were mined for terpene synthase genes (TPS)elative expression of TPS in seven plant tissues andmpared to other sequenced plant genomes, Eucalyptusenes of any sequenced plant. We discovered 113 andl. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/4.0), which permits unrestricted use, distribution, andiginal work is properly credited. The Creative Commons Public Domaing/publicdomain/zero/1.0/) applies to the data made available in this article,Külheim et al. BMC Genomics  (2015) 16:450 Page 2 of 18pyrophosphates are the substrate for terpene synthases,which catalyse the production of mono-(C10) from GPPor di-(C20) terpenes from GGPP as well as hemiterpenes(C5) directly from isopentyl pyrophosphate in thechloroplast, or sesquiterpenes (C15) from FPP in the cyto-sol. The products of these multi product enzymes can befurther modified by oxygenation by cytochrome P450monooxygenases, or methylation by methyl transferasesto form additional compounds.The terpene synthase gene family (TPS) has been di-vided into three classes and seven sub-families: Class Iconsists of TPS-c (copalyl diphospate and ent-kaurene),TPS-e/f (ent-kaurene and other diterpenes as well assome mono- and sesquiterpenes) and TPS-h (Selaginellaspecific); class II consists of TPS-d (gymnosperm spe-cific) and class III of TPS-a (sesquiterpenes), TPS-b (cyc-lic monoterpenes and hemiterpenes) and TPS-g (acyclicmonoterpenes) [16]. These clades have been identifiedthrough sequencing and functional studies of a widerange of plants (Arabidopsis thaliana: [17],Vitis vinifera:[18], Solanum lycopersicum: [19], Selaginella moellendorfii:[20], Populus trichocarpa: [21]).Often, plants that emit or store few terpenes, haveonly a small number of TPS genes, such as Arabidopsisthaliana (32 putative functional and 8 pseudo TPSgenes: [17]) and poplar (38 putative functional genes:[21]). Other plants that have a more complex blend ofterpenes, which are often stored in trichomes or otherglandular structures, tend to contain a larger number ofTPS genes (e.g. grape with 69 putative functional and 63pseudo TPS genes [18]). Eucalypts contain a high diver-sity of terpenes, which are stored mainly in schizogenoussecretory cavities in the leaf and flower buds [22]. Giventhe striking variation in terpenes that can occur within asingle Eucalyptus species, we hypothesise that eucalyptswill contain a highly diverse and abundant family of ter-pene synthase genes which can be differentially regulatedto generate unique terpene profiles in different parts orat different times of the plant as required. We describethe Eucalyptus grandis TPS gene family from the re-cently sequenced reference genome of E. grandis (V1.1annotation, www.phytozome.net) [23] and also makecomparisons with a second species of Eucalyptus thathas been sequenced (E. globulus) to understand how thisgene family is arranged and how it has evolved over therelatively short time (ca. 12 million years) that these twospecies separated (Thornhill, Külheim and Crisp unpub-lished) and how it has contributed to the success of thegenus Eucalyptus across its native and introduced range.ResultsDiscovery of putative TPS genes from the E. grandis genomeWe identified 172 loci in the Eucalyptus grandis genomewith a high sequence similarity to known terpene synthasegenes from other species. Loci that spanned fewer thanthree exons (out of seven for most TPS sub families) werenot further considered which resulted in the exclusion of20 loci. Loci were considered putatively functional if theywere full length (with the exception of genomic areaswhere sequence data was missing), which had fewer thantwo frame shifts and stop codons (combined) or showedevidence of gene expression either through the presenceof expressed sequence tag (EST) models or RNAseq datathat matched the loci (Additional file 1: Table S1). Publi-cally available RNAseq data (www.phytozome.net), aswell as our own RNAseq dataset, were used to validate thegenomic sequences. The resulting list of genes was thenclassified as (i) full-length, expressed with no prematurestop codons or frame shifts (89), (ii) full-length, expressedwith up to two stop codons or frame shifts (23), (iii)full-length, with no expression and no frame shift orstop codon (1), (iv) pseudogenes with more than twoframe shifts or stop codons (39) and (v) partial genes(20) (Additional file 1: Table S2). A total of 113 loci ingroups (i) (ii) and (iii) were considered to be putativelyfunctional TPS genes (Additional file 1: Table S1).Six of the eight currently recognized TPS subfamiliesare present in E. grandis, the exceptions being theGymnosperm specific TPS-d subfamily and the Selaginellamoellendorffii specific TPS-h subfamily (Figure 1). Genecopy numbers in diterpene synthase genes (TPS-c and –e)were similar to those found in other plant species(Table 1). However, the other TPS subfamilies are muchmore abundant in E. grandis compared to other plantspecies (Table 1).Putative terpene synthases in E. globulusWe discovered 106 terpene synthase genes in E. globulususing a strategy similar to that applied to E. grandis, butwithout any gene expression data due to the lack of suchdatasets (Table 1). We further discovered 37 putativepseudogenes which contained multiple stop codons orframe shifts (Additional file 1: Table S3). Overall, the dif-ferences in gene copy numbers between E. globulus andE. grandis were small, with slightly fewer sesquiterpenesynthases (TPS-a), two more mono- and hemi-terpenesynthases (TPS-b) as well as TPS-f copies, similar copynumbers in TPS-c and –e and three fewer copies in theTPS-g subfamily (Table 1). Most pseudogenes were foundamong the TPS-a subfamily (16), with slightly fewer in theTPS-b (13) and four each in the TPS-f and –g subfamilies(Additional file 1: Table S3).Genomic organization of terpene synthase genes fromEucalyptus grandisMost E. grandis TPS genes are organized in moderate tolarge groups (tandem gene arrays) in the genome. Eachgroup contains only genes of the same TPS subfamilyFigure 1 (See legend on next page.)Külheim et al. BMC Genomics  (2015) 16:450 Page 3 of 18PSgenbyKülheim et al. BMC Genomics  (2015) 16:450 Page 4 of 18and genes are found in dense clusters. In some cases,the distance between two EgranTPS genes is less than1 kb, e.g. between EgranTPS026 and EgranTPS027, thedistance is only 508 bp between stop and start codon.On scaffold 6, there are 17 TPS genes and pseudogenesfound within 317 kb, one TPS gene for every 18.6 kb(Additional file 1: Table S4 and Figure 2). The clusteringof putative isoprene/ocimene synthases is even greaterand on scaffold 11, eight isoprene synthase genes arefound within 107 kb – one gene per 13.4 kb (Additionalfile 1: Table S4).Genes from different steps in the terpene biosyntheticpathway are found nearby the clusters of E. grandisTPS genes, such as a prenyl transferase on scaffold 11(Additional file 1: Table S4). The TPS-b cluster on scaffold4 is also co-incident with eight quantitative trait loci(QTL) for foliar monoterpenes from Eucalyptus nitensand one for foliar sideroxylonal discovered by Henery andco-workers [24]. Other genes of plant secondary metabol-ism are also found close to clusters of terpene synthases(e.g. phenylalanine ammonia lyase, anthocyanidin reduc-tase, flavonol reductase) as are many transcription factors(Additional file 1: Table S4).Most (82 out of 92) EgranTPS genes of subfamilies –a,−b and –g, contain seven exons, with the exception ofthree TPS-a genes (EgranTPS016, EgranTPS017 andEgranTPS019) which appear to have recently obtained aseventh intron near the 3′ end of exon seven and sevengenes for which sequence data is incomplete (Figure 3).Genes from the remaining subfamilies, −c, −e and -f,contain between 9 and 14 exons, with the exception ofEgranTPS92 which has just seven exons. The length ofthe introns is largely conserved with a few exceptions,notably intron 1, which is highly variable across subfam-ilies -a, −b, and -g (Figure 3).Global expression profiling of terpene synthase genesfrom Eucalyptus grandisA heat map showing relative transcript abundance of113 TPS genes in three biological replicates of seven(See figure on previous page.)Figure 1 Phylogeny of the 113 putatively functional terpene synthase (TE. grandis TPS gene family rooted at the branching of type I and III TPSwith main products of isoprene or ocimene. Bootstrap values supporteddesignated O. Subfamilies are described on the branches.tissues is shown in Figure 4. The hierarchical clusteranalysis of the tissues shows that the six wood tissuesamples (immature xylem and phloem) form one cluster,three root tissues a separate cluster and the ‘green tissue’comprised of young leaf, mature leaf, shoot tips andflowers a third cluster. The only two tissues where allthree biological replicates cluster together are root andflower. The hierarchical clustering of genes shows acluster of highly expressed genes in the ‘green tissue’,dominated by TPS-b1 genes (I). The next major clusteris also expressed mostly in ‘green tissue’, but to a lesserextent (II). Gene expression cluster II contains mostlygenes from the TPS-a group with several TPS-b1 and -b2genes. Cluster (III) is moderately expressed in ‘greentissue’ with one sub-cluster expressed in root as well.Genes that cluster by gene expression patterns are amixture of most subfamilies, excluding TPS-c and -e.Cluster (IV) shows highest levels of expression inroot tissue, with minor expression in ‘green tissue’ andwood tissue. All subfamilies bar isoprene/ocimene syn-tases (TPS-b2) are represented in this cluster. Cluster (V)has the overall lowest level of gene expression scatteredover all tissues, containing genes from TPS-a, −b1and -g. Finally, cluster (VI) is a single gene, which isexpressed constitutively in all tissues at moderate tohigh levels. This gene encodes a putative ent-kaurenesynthase, which produces the precursor for gibberellicacid.Phylogenetic analysis of Eucalyptus TPS genes incomparison to other plant speciesThe phylogenetic analysis shown here includes allknown TPS genes from Arabidopsis thaliana, Populustrichocarpa, Vitis vinifera (cultivar Pinot Noir), Solanumlycopersicum, Oryza sativa, Sorghum bicolor, Eucalyptusgrandis and E. globulus. The phylogenies are divided intogroupings of (i) TPS-a (Figure 5 and Additional file 2:Figure S1), (ii) TPS-b and -g (Figure 6 and Additionalfile 2: Figure S2) and (iii) TPS-c, −e and -f (Figure 7 andAdditional file 2: Figure S3). The phylogenies containedtoo many terminal nodes to be easily viewable (210 for(i), 161 for (ii)) and the full phylogenies are thereforepresented in the supplemental files, while the figuresshow only TPS genes from Arabidopsis thaliana, Vitisvinifera and Populus trichocarpa in comparison to botheucalypt species studied here.For the TPS-a subfamily, the most closely relatedgenes to the eucalypt TPS clade are from Vitis vinifera) genes in Eucalyptus grandis. Maximum likelihood analysis of thees. Clade TPS-b2 contains acyclic terpene synthases from the rosids≥ 80% are designated * while those with bootstrap values≥ 95% areand Populus trichocarpa (Figure 5). The Eucalyptus TPSgenes form a paralogous cluster, which is differentfrom some of the Populus/Vitis TPS-a clades (Figure 2,Additional file 2: Figure S1). A comparison of the twoeucalypt species shows that 31 out of 52 genes arefound in orthologous pairs for E. grandis (Additionalfile 1: Table S1) and 31 out of 45 for E. globulus. Forseveral genes in E. grandis, the E. globulus ortholog isTable 1 Gene copy numbers of TPS genes from plant species with a fully annotated genome separated by subfamiliesTerpene type E. grandis E globulus V. vinifera P. trichocarpa A. thaliana S. lycopersicum S. bicolor O. sativa S. moellendorffii P. patentsTPS-a sesqui 52 45 29 13 23 12 15 19 0 0TPS-b1 mono 27 28 8 10 6 8 2 0 0 0TPS-b2 isoprene/ocimene 9 10 2 2 0 0 0 0 0 0TPS-c di 2 2 2 2 1 2 1 3 3 2TPS-e mono, sesqui, di 3 2 1 2 1 5 3 9 3 0TPS-f mono, sesqui, di 7 9 0 1 1 0 0 0 0 0TPS-g mono, sesqui, di 13 10 15 2 1 2 3 1 0 0TPS-h di 0 0 0 0 0 0 0 0 8 0Total 113 106 57 32 33 29 24 32 14 2Gene copy numbers differ between studies; numbers shown here are derived from the downloaded sequences from each species.Külheimetal.BMCGenomics (2015) 16:450 Page5of18Külheim et al. BMC Genomics  (2015) 16:450 Page 6 of 18missing and vice versa (Figure 5). There are also severalgene duplication events that post-date the separation ofthe two species. It is notable that the Arabidopsis TPSgenes have longer branches than both grape and eucalyptssuggesting a longer period of gene differentiation withoutnovel gene duplication events.Genes from TPS subfamilies –b and – g are shown inFigure 6 and Additional file 2: Figure S2. TPS-g forms aclade embedded within TPS-b. Eucalypts have twoclades of TPS-g genes, one closely related to the onlyArabidopsis TPS-g gene (At1g61680) gene and the otherclosest to two poplar genes. Six out of 13 and 10 TPS-ggenes are in orthologous pairs for E. grandis and E.globulus, respectively (Additional file 1: Table S1), withseveral gene duplication events in the clade close toAt1g61680 since the separation of these two eucalyptspecies.Figure 2 Genomic organization of a 400 kb multi-gene cluster fromEucalyptus grandis. A cluster of 10 putative functional (green arrows)and seven TPS pseudogenes (light green arrows) as well as othergenes (white arrows) in the area are shown. Close phylogeneticrelationships are indicated by brackets.The TPS-b subfamily falls into three separate clades,all of which contain multiple species (Additional file 2:Figure S2). The middle clade contains putative isoprene/ocimene synthases, and is described as TPS-b2, whilethe other two clades are described as TPS-b1 and con-tain putative cyclic monoterpene synthases (Figure 6).For the TPS-b1 19 out of 27 and 28 for E. grandis andE. globulus, respectively, occur in orthologous pairs(Additional file 1: Table S1). Roughly half of the iso-prene/ocimene synthases occur in orthologous pairs (5/9and 5/10 for E. grandis and E. globulus, respectively) withseveral gene duplications and loss of function events post-speciation (Additional file 1: Table S1).Subfamilies TPS–c, −e and –f are involved in the for-mation of di-terpenes as well as mono- and sesquiter-penes. In the TPS-c subfamily, two out of two genesoccur in orthologous pairs, while in TPS-e, two ortholo-gous pairs are found, but E. grandis contains one extragene (Figure 7, Additional file 1: Table S1). Gene copynumbers are very similar in these two families comparedto other species (Table 1, Figure 7). However, there hasbeen a significant expansion of the TPS-f family in euca-lypts compared to other plant species. Only three ortho-logous pairs occur between the two eucalypt species outof seven and nine genes (Additional file 1: Table S1). Re-cent duplications (e.g. EglobTPS119 and EglobTPS123)as well as loss of function (e.g. the E. globulus ortholo-gue of EgranTPS098) of genes have occurred.Functional characterization of full-length Eucalyptusgrandis TPS genesOf the nine genes synthesised for characterization stud-ies, five have predicted transit-peptide sequences at theN-terminus, which suggests they are monoterpene syn-thases. Based on sequence characteristics, the remainingfour genes are likely to be sesquiterpene synthases. Wewere able to express all nine genes in the vector, however,only three were able to produce significant amounts ofterpenes and two produced traces of terpenes (Additionalfile 1: Table S5). We characterised a bicyclogerma-crene synthase (EgranTPS041), an isoledene synthase(EgranTPS013) and a γ-terpinene synthase (EgranTPS059),all of which are multi-product enzymes which producedbetween 5 and 15 terpenes (Figure 8).DiscussionEucalyptus is widely known for its foliar terpenes andthe diversity of eucalypts in Australia is matched by alarge diversity of terpenes in different species. Our ana-lysis of two Eucalyptus genomes shows that this diversityarises from the largest gene family of terpene synthasesyet described. More than 2000 papers have been pub-lished on Eucalyptus terpenes over the past century andthey have been implicated in many processes within theFigure 3 (See legend on next page.)Külheim et al. BMC Genomics  (2015) 16:450 Page 7 of 18l teExof taarEgranKülheim et al. BMC Genomics  (2015) 16:450 Page 8 of 18plant, within the ecosystem and as large-scale influenceson the environment via their impact on forest fires andin the atmosphere. The discovery of the diversity of TPSgenes in the genome of Eucalyptus as well as recentdiscoveries of the control of quantitative variation andnew evolutionary analyses of Myrtaceae terpenes [15]provides the first opportunity to understand how vari-ation in Eucalyptus terpenes arises and how they can bemanipulated in the world’s most widely planted hard-wood tree.Both Eucalyptus grandis and E. globulus have approxi-mately four times as many putative functional TPS genesas Arabidopsis thaliana, three times as many as Populustrichocarpa and twice as many TPS genes as Vitis vinif-era. Conifers may also have a large and diverse TPS genefamily given the diversity of TPS genes that have beencharacterized [25], but despite recent assemblies of coni-fer genome sequences [26,27], accurate estimates of genecopy numbers are not available. Amongst the speciesthat have been sequenced, the larger gene families areassociated with species that have specialized storageorgans for terpenes including Eucalyptus and grape [18].No studies have reported on the sub-cellular localizationof terpene synthases in eucalypts but they are presumedto occur in cells of the secretory cavities where the ter-penes are stored along with other non-volatile constitu-ents such as oleuropeyl glucose esters [28,29].When comparing the branch lengths on the TPS phy-logenies between each of the species in this study, it isclear that species with overall low numbers of TPS geneshave longer branch lengths and those with large TPSgene families have shorter branch lengths. Species with asmall TPS gene family such as Arabidopsis thaliana havefew genes with long branches. This is especially apparent(See figure on previous page.)Figure 3 Gene structure and motif representation of putative functionamanually determined by comparison to previously characterized genes.show missing genome sequences and the red line shows the location oRR(x)8W and (ii) NSE/DTE are indicated after the gene name where “+” isacids changed and “–” means that the motif is not present. Gene namesis red, TPS-b (EgranTPS053 – EgranTPS088) is blue, TPS-c (EgranTPS089 –TPS-f (EgranTPS094 – EgranTPS100) is pink and TPS-g (EgranTPS101 – Egin the TPS-a subfamily where Arabidopsis has a large num-ber of genes (Figure 5 and Additional file 2: Figure S1).Eucalyptus and Vitis have comparably small branchlengths, indicating rapid ongoing evolution as comparedto Arabidopsis.All but one study of the terpenes of E. grandis havefocused on the foliar oils of the mature leaf. The profilesare complex with one study [30] detecting 99 componentsby capillary gas chromatography and the remainder be-tween 38 and 67 [31-35]. The monoterpene fraction isdominated by either α-pinene or 1,8-cineole, which maybe indicative of the occurrence of two terpene chemotypesin E. grandis. Since these two monoterpenes arise fromdifferent carbocations, it is likely that they are prod-ucts of different TPS. However, none of the TPS genesthat we characterized produced large amounts of eithercompound.All published studies show a complex mixture of up to30 sesquiterpenes with bicyclogermacrene and spathule-nol usually dominant, but with a large diversity of othersesquiterpenes that were minor components of the oil.We characterized a bicyclogermacrene synthase that alsoproduced a further four compounds and a second ses-quiterpene synthase that produced 15 sesquiterpenes.Thus it may not require a large number of sesquiterpenesynthases to be expressed in mature leaf to explain thecomplex oil profiles described previously.Given the abundance of secretory cavities and highlevels of terpene (monoterpene and isoprene) emissionsfrom leaves of Eucalyptus it was not surprising that thelargest proportion of E. grandis TPS genes are highlyexpressed in ‘green tissues’ (mature and young leaf butalso floral buds and shoot tips). All of the TPS subfam-ilies found highly expressed in ‘green tissues’ belong tothe TPS-a, −b1 and -b2 family, producing mono-,sesqui- and hemiterpenes as well as three acyclic terpenesynthases from subfamily -g. However, we also found asignificant group of TPS that were expressed in roots,xylem and phloem. Early studies identified secretory cav-ities in bark (including the root bark) the stem pith, andphloem, depending on the species [36] but the few stud-ies of the essential oils in these tissues suggest that theyare largely non-terpenoid [37]. Two of three samplescluster closely together in all tissues, which was not sur-prising as these trees derive from clonal material. Therpene synthases from Eucalyptus grandis. Exon-intron structures werens are shown by full boxes, while introns are depicted as lines. Arrowshe DDxxD motif. The presence and quality of conserved motifs (i)perfect motif match, “•” has one amino acid changed, “>” two aminoe coloured according to subfamily; TPS-a (EgranTPS001 – EgranTPS052)ranTPS090) is purple, TPS-e (EgranTPS091 – EgranTPS093) is green,TPS113) is yellow.third sample, however, was a half-sib of the other twoand showed some differences in gene expression.The root-specific cluster of TPS genes that we identi-fied may be involved in allelopathic interactions [7] and/or, as in other plants, in mediating interactions with herbi-vores and other soil organisms [38]. A further possibilityis that root-derived terpenes may play a role in mycor-rhizal colonization which is essential for many species ofEucalyptus [39,40]. Clearly there is a need for a broad in-vestigation of terpenes in roots of eucalypts from both thechemical and molecular perspectives.Figure 4 Gene expression of 113 terpene synthase (TPS) genes from Eucalyptus grandis in seven plant tissues. The log2 expression valuesdetermined from FPKM values of RNAseq data are shown. Cluster analysis of tissues and genes was performed and the E. grandis TPS subfamiliesare indicated. The tissues investigated are: YL: young leaf, ST: shoot tips, ML: mature leaf, Fl: flower, Rt: root, Ph: phloem and Xy: xylem.Külheim et al. BMC Genomics  (2015) 16:450 Page 9 of 18Külheim et al. BMC Genomics  (2015) 16:450 Page 10 of 18One possible explanation for the large number of TPSgenes in Eucalyptus is that many are only involved in in-ducible responses. For example, Danner and co-workersisolated three TPS from Populus that were only expressedafter gypsy moths had fed on the plants [41]. Other stud-ies have identified TPS genes that are only expressed ontreatment by methyl jasmonate [42]. However, the evi-dence for induced production of terpenes in Eucalyptus ismixed. The only comprehensive study involving methyljasmonate found no evidence of an increasedFigure 5 Phylogeny of the TPS-a subfamily. Maximum likelihood analysis of TPand Arabidopsis thaliana. Functionally characterized enzymes are indicated withbicyclogermacrene. Bootstrap values supported by≥ 80% are disignated * whilefrom E. grandis and P. trichocarpa were selected as outgroups.concentration of foliar terpenes or terpene phloroglucinoladducts in clones of a hybrid E. grandis × E. camaldulen-sis. Henerey and co-authors [24] argued that the inducedfoliar responses seen in deciduous trees and pines shouldnot be expected in evergreen broadleaved trees like euca-lypts because herbivory is much harder to predict in timeand space. Furthermore, the terpene storage organs in eu-calypts are primarily formed during leaf development [22],so induction may be difficult because the size and num-bers of oil glands may not be very plastic.S-a subfamily from E. grandis and E. globulus in comparison to Vitis viniferaan * with main products: *1 isoledene / cadinene, *2 germacrene D, *3those with bootstrap values≥ 95% are designated O. Two TPS-b genesKülheim et al. BMC Genomics  (2015) 16:450 Page 11 of 18There is however some evidence of induction of ter-penes in woody tissues. Eyles et al., [6] found significantconcentrations of terpenes in new phloem of wound tis-sue of E. globulus and observed traumatic secretory cav-ities in the wood of E. globulus after pruning and fungalattack [43]. They argued that these inducible terpenesplayed a broad role in defence against infection after anydamage to the phloem, either mechanical or biological.Our observations of a diverse range of TPS expressed inthe phloem and wood tissues supports Eyles et al.’s [43]call for further investigation of terpenes in these tissues.Further studies in eucalypts should focus on expres-sion of terpene pathway genes including TPS after insectfeeding, fungal infection and jasmonate treatments in arange of tissues and such studies may reveal a morecomplex pattern of induction. Such studies would helpidentify which (if any) of the TPS genes in the genomeare inducible. This may help to explain the large size ofthe gene family.A group of terpenoid compounds have been presentsince the evolution of the genus Eucalyptus ca. 50 millionyears ago and Keszei et al. [44] argued based on a phyl-ogeny of TPS fragments from more than 20 species thatearly divergence and maintenance of function has sug-gested a closer correlation between sequence homologyand function than seen in other plant groups. The com-parison of TPS genes between E. grandis and E. globulusenables us to estimate both recent evolutionary activity ofthe terpenes synthase family (during the last 12 millionyears) and distant evolutionary activity. The results showthat most TPS genes in E. grandis have evolved more than12 million years ago, but also that there is still ongoingevolution as indicated by novel gene duplication and lossof function events. There are broad differences in theamount of conservation vs. evolution between TPS sub-families and even within clades of the larger subfamilies.TPS-g and -f have the lowest proportion of orthologouspairs with 46 and 43%, respectively, while TPS-b and -chave the highest proportion with 82 and 100%, respect-ively. This shows that genes in subfamilies -b and -c arehighly conserved. Overall, the similarities between genefamilies in the two species are very high and it would beinteresting to compare further species pairs or evengroups of Eucalyptus species to learn more about geneevolution in this gene family. Recent advances in DNA se-quencing have enabled a large number of whole-genomesequencing projects and if a closely related species has awell-assembled and annotated genome, gene families caneasily be characterized from such species. Species fromgenera like Arabidopsis and Populus are good candidates.Isoprene is the smallest terpenoid compound, but hasa significant impact on the planet’s atmosphere [45]through enhancement of aerosol and ozone formation.The physiological functions of isoprene in plants includeheat protection [11,46], ozone tolerance [47] and toler-ance of reactive oxygen species [48]. The biosyntheticcost for each isoprene molecule is very high (includingat least 20 ATP and 14 NADPH), so the benefits forplants that emit large amounts such as Eucalyptus mustbe significant. Isoprene is the major component of euca-lypt hydrocarbon emissions (64-100%) with the remainderbeing monoterpenes. Eucalyptus grandis and E. globulusare among the highest emitters of isoprene and hydrocar-bons of any plant investigated to date [12] (although thishas been challenged by Winters et al., [49]). The ability toemit isoprene has evolved many times in plants and is ar-gued to be a response to local environmental conditions[45]. Therefore, the high number of putative isoprene syn-thase genes found in E. grandis and E. globulus could bereflective of the environment eucalypts evolved in, thoughfunctional characterizations of each gene are needed toensure they do not encode functional ocimene synthases.Australia is the hottest and driest continent and isopreneemissions have been implicated in mitigating heat stress[46]. Clade TPS-b2 contains putative isoprene/oci-mene synthases, there are two recent gene duplications inE. globulus (EglobTPS099 and EglobTPS103; EglobTPS098and EglobTPS100) and three putative losses of func-tion (orthologous gene is missing in other species inEgranTPS079, EglobTPS102 and EgranTPS078) indicativeof rapid ongoing evolution. Eucalyptus grandis has a subtropical to tropical distribution with high rainfall, whileE. globulus is found in temperate regions of southernVictoria and lowland Tasmania, but at sites that can ex-perience high temperatures and dry summers. Charac-terizing the Eucalyptus isoprene synthases should clarifytheir role in mediating local heat stress and climateadaptation.Genomic clusters of genes that encode multiple stepsof biosynthetic pathways for plant secondary metabo-lites are an emerging theme in plant biology [50,51].These clusters which typically span 100 s of kb, can con-tain parts of, or entire biosynthetic pathways and theirevolutionary origin and biological function is of wideinterest. In this study we have focused on smaller genomicareas to investigate clusters of duplicated TPS genes andtheir evolutionary fate. Nevertheless, we found multiplegenes that are or may be involved in terpene biosyn-thesis, such as isopentenyl pyrophosphate isomerase ona cluster with three putative functional TPS-b loci, fiveTPS-b pseudogenes and five methyltransferases thatmay be involved in further modifications of terpenestructures. Future work could investigate the largerscale genomic surroundings of the TPS gene clusters.Clusters of homologous genes evolve through gene du-plication events and are widespread for enzymes that di-versify the plant’s defense. There are examples for bothTPS genes [18] and other gene families such as KunitzKülheim et al. BMC Genomics  (2015) 16:450 Page 12 of 18protease inhibitors involved in insect herbivory defense[52], nucleotide binding site-leucine-rich repeat genesinvolved in fungal disease resistance [53] or polyphenoloxidases involved in bacterial and fungal disease resist-ance [49]. Due to the evolutionary origin of tandemgene duplications, genes within clusters have highersimilarity to each other than to genes that are not in theFigure 6 Phylogeny of the TPS-b and -g subfamilies. Maximum likelihoodcomparison to Vitis vinifera and Arabidopsis thaliana. Functionally characteri*2 isoprene (characterized by [56]), *3 γ-terpinene. Bootstrap values supportare designated O. A single TPS-a gene from E. grandis was selected as outgrcluster. We observed that the amino acid similarity ofgenes within clusters were twice as high as those of thewhole gene sub-family (Additional file 1: Table S6), con-firming their evolutionary origin from unequal crossing-over as suggested by Ober [54].It has been difficult to functionally characterize ter-pene synthases from the Myrtaceae for several reasonsanalysis of TPS-b and -g subfamilies from E. grandis and E. globulus inzed enzymes are indicated with an * with main products: *1 β-pinene,ed by≥ 80% are designated * while those with bootstrap values ≥ 95%oup.Külheim et al. BMC Genomics  (2015) 16:450 Page 13 of 18including high levels of plant secondary metabolitespresent in leaves. Few studies have been successful[44,55,56]. Here, we used an alternative approach to cap-ture functional terpene synthases via de-novo gene syn-thesis rather than gene amplification from cDNA. Oursuccess rate for genes without target peptide (all sesquiter-pene synthases) was high (75%), with the one unsuccessfulgene (EgTPS032) also showing comparably low levelsof gene expression across all tissues (Additional file 1:Figure 7 Phylogeny of the TPS-c, −e and -f subfamilies. Maximum likelihoodcomparison to Vitis vinifera and Arabidopsis thaliana. Bootstrap values supportare designated O. A single TPS-b gene from E. grandis was selected as outgroTable S5). Our success rate with the characterizationof genes that putatively contain transit peptides (acyc-lic and cyclic monoterpenes from subfamilies TPS-band –g) was much lower at 40% and this is possibly due toerrors in assigning the location/end of the transit peptideas well as exon-intron boundaries. All genes that we at-tempted to functionally characterize were expressed, mostlyin green tissue, but also in xylem and roots (Additionalfile 1: Table S5).analysis of TPS-c, −e and –f subfamilies from E. grandis and E. globulus ined by≥ 80% are designated * while those with bootstrap values≥ 95%up.Külheim et al. BMC Genomics  (2015) 16:450 Page 14 of 18Most of the investigations of terpenes in Eucalyptushave been driven by the prospects of extracting industri-ally useful products and the mature leaf is the mostabundant source of terpenes in the plants. Nonethelessthe focus on terpenes in mature leaf tissue of Eucalyptushas meant that the functional and ecological interpreta-tions of their role have been skewed towards questionssuch as defence against insect and vertebrate herbivory[57-59]. In contrast, reports of terpenes in other tissuessuch as roots and wound tissue [6] are sparse. OurFigure 8 Gas chromatographic profile of the products formed in vitro by twere incubated with GPP and FPP and products were analyzed as describeresults call for a re-interpretation of the role of terpenesin Eucalyptus since almost half the gene family isexpressed in woody tissues and of those found ingreen tissues many are restricted to juvenile leaf. Thissuggests that there are many important functions ofterpenes in eucalypts that have been underappreciatedbut which may be critical in explaining the success ofeucalypts continent wide in Australia and in manyother regions of the world where they have subsequentlybeen cultivated.he enzyme activities of several Eucalyptus grandis TPS genes. Enzymesd in the Methods.Külheim et al. BMC Genomics  (2015) 16:450 Page 15 of 18ConclusionsIn this study, we have analysed the terpene synthasegene families from two Eucalyptus species, E. grandisand E. globulus. We investigated the genomic structure,gene diversity, evolutionary pathways and gene expres-sion across seven tissues. High levels of intraspecific ter-pene diversity is reflected by the largest number of TPSgenes found in a single species to date. Most of the TPSgenes, as well as large numbers of TPS pseudogenes, arefound in large genomic clusters which may originatefrom many gene duplication events leading to the ter-pene diversity found in the genus. Nearly half of the TPSgenes were expressed in non-green tissues (root, phloemand xylem) and the role of terpenes in these tissuesremains unknown. A phylogenetic analysis of the genefamily revealed that most TPS clades have evolved littlesince the separation of the two Eucalyptus species, whilea few clades show signs of recent evolution as shown bygene duplications and loss of function in one or botheucalypt species.MethodsDiscovery and annotation of putative terpene synthasegenes in Eucalyptus grandisThe Eucalyptus grandis 8x BRASUZ1 genome (V1.1)sequence assembly (http://www.phytozome.net/eucalyptus.php) was used in a BLAST search based on conservedregions (Additional file 3) of all terpene synthase subfam-ilies with standard BLAST parameters. A preliminary list ofgenomic regions containing TPS genes was created andredundancies removed. In order to find full-length genesthe up- and down-stream areas of the conserved regionswere screened and a reverse BLAST search was used toconfirm the identity of the putative TPS gene. Of the 172loci that showed significant similarities to known TPSgenes, 59 were excluded due to more than two frame shiftsor stop codons in the translated gene. These pseudo-genes are however of evolutionary interest and weretherefore retained separately for later analysis of geneclusters (Additional file 1: Table S2). Loci with fewer thanthree exons (TPS genes generally contain between 7 and13 exons depending on subfamily) were not included inthe list, but those loci that did not cover the full genelength due to lack of assembled DNA sequence (fallingoff the scaffold) were included. Genes were manuallyaligned to known TPS genes from multiple species andsub-families (Arabidopsis: http://arabidopsis.org/, grape,tomato and Backhousia citriodora from Genebank (http://www.ncbi.nlm.nih.gov)) to determine exon-intron borders.After preliminary phylogenetic analysis of amino acid se-quences (using Phyml [60]) putative functional genes weresorted into TPS subfamilies TPS-a, −b, −c, −e -f and -g bysequence similarity, then sorted by linkage group (LG)and then by position within LG. TPS-b was furtherdivided into TPS-b1 for cyclic monoterpene synthases andTPS-b2 for putative isoprene/ocimene synthases. Onecharacterised isoprene synthase from Eucalyptus globulus(AB266390) falls within this subgroup and is the synonymof EglobTPS106 [55]. Gene annotations were given fromthe first TPS-a gene (EgranTPS001) to the last TPS-g gene(EgranTPS113) in the assembled genome scaffolds (V1.1).Discovery of putative terpene synthase genes inEucalyptus globulusIllumina paired end shotgun short reads (75 bp) gener-ated for a clonal genotype (X46) of Eucalyptus globulusgenome (Department of Energy-Joint Genome Institute)were aligned against the Eucalyptus grandis genome as areference using CLC Genomics Workbench (CLC, Aarhus,Denmark). Of the 224 million single short reads 215million were matched (95.8%) against the reference se-quence using default parameters for reference assemblyin CLC Genomics Workbench. The remaining readswere assembled de-novo using standard parameters inCLC Genomics workbench into 5,739 contigs with anaverage length of 771 bp. Consensus sequences fromboth reference and de-novo assembly were importedinto Geneious Pro (version 5.6.4, [61]) and BLASTsearches with representatives from each subfamily of E.grandis TPS genes were performed. All hits with Evalues < e−10 were considered and redundancies re-moved. The genomic regions surrounding the BLASThits (±2,500 bp) were investigated and reverse BLASTedagainst the non-redundant database at Genebank (http://www.ncbi.nlm.nih.gov). Coding regions for each genewere extracted and genes were examined for full length,stop codons and frame shifts. In total, 143 loci withhigh similarity to terpene synthase genes were discov-ered and 106 putative functional (full length geneswith up to 2 frame shifts or stop codons) genes wereconsidered for downstream analysis (Additional file 1:Table S3). Gene annotations were given in a similarway as in E. grandis, but were numbered through to includepseudogenes, thereby having a range from EglobTPS001 toEglobTPS143.Phylogenetic analysis of terpene synthase genes fromEucalyptus and other plantsTerpene synthase genes from Arabidopsis thaliana(http://arabidopsis.org/), Populus trichocarpa (http://www.plantgdb.org/PtGDB), Vitis vinifera (http://www.plantgdb.org/VvGDB/), Solanum lycopersicum [19], Sorghum bi-color (http://www.plantgdb.org/SbGDB/) and Oryza sativa(http://www.plantgdb.org/OsGDB/) were retrieved anddivided by TPS subfamily. Amino acid alignments weremade for subfamilies/groups of subfamilies, using Clus-talW [62] in Genious Pro (version 5.6.4, [61]) using stand-ard parameters. The alignments were then manuallyKülheim et al. BMC Genomics  (2015) 16:450 Page 16 of 18adjusted with a focus on diagnostic conserved regionssuch as the RLLR, DDXXD and NSE/DTE motifs. Thealignment was then truncated to ensure that sites werehomologous. All of exon 1 and parts of exon 2 were ex-cluded due to high levels of variation, resulting from thepresence of chloroplast target peptides in some genes andgenerally high levels of diversity. To create a phylogeny,we first tested which amino acid substitution model pro-vided the maximum likelihood tree with the best AICcvalue (Akaike’s information criterion value, correctedfor samples size) and further tested whether gammadistribution estimation and/or proportion of invariablesites estimation improved the AICc value using Phyml[60]. The amino acid substitution models that wetested were: EHO, EX2, EX3, JTT, LG, UL2, UL3 andWAG. The tree with the highest AICc value wasusing the JTT model with estimation of invariable sitesand estimation of gamma distribution. The phylogeny ofthe EgranTPS gene family was determined using 100bootstrap replicas. The phylogenies were visualized usingFigTree v1.3.1 [63].RNA isolation and identification of terpene synthasetranscripts in seven tissue types of Eucalyptus grandisRNA was isolated from shoot tips, young leaves, matureleaves, floral buds, immature xylem and phloem tissuescollected from three actively growing E. grandis trees asdescribed previously [64]. Root tissues were sampledfrom rooted cuttings of the same E. grandis genotypes.Sampled tissues were immediately frozen in liquid nitro-gen and stored at −80°C until RNA extraction. TotalRNA was isolated as described previously [65] and usedfor Illumina RNAseq analysis at the Center for GenomeResearch & Biocomputing at Oregon State University.The RNA read pairs (50 bp paired end) were mapped topredicted gene models in the draft E. grandis genome se-quence (Department of Energy-Joint Genome InstituteV1.0, http://www.phytozome.net) using TopHat version1.3 [66] and regions which were identified as terpenesynthase genes. Gene expression values (fragments perkilobase of coding region per million mapped fragments,FPKM) were calculated for each predicted locus usingCufflinks version 1.1.0 [67].Analysis of transcript abundance and cluster analysis ofgenes and tissuesFor each of the 21 RNAseq datasets (seven different tis-sues in triplicate, deposited at the Eucalyptus GenomeIntegrative Exploreer: http://eucgenie.org/), the log2 ofthe FPKM value was used to create a heat map inExpander (version 6.0) (EXPression ANalysis andDisplayER; Ron Shamir’s Computational genomics group,University of Tel Aviv, Israel). Hierarchical cluster analysisfor both tissues and genes was performed in GenStat(version 14.1) with Euclidian matrix formation and thecomplete linkage method.Functional characterization of recombinant proteinsFrom the list of putative functional terpene synthasesfrom E. grandis, we took only those that were full lengthand had no frame shift mutations or premature stop co-dons. We focussed on the terpene synthases highlyexpressed in leaves, since these are likely to be the mostimportant in defence against folivores. We chose be-tween one and two sequences from each group withinthe TPS-a, −b and -g subfamilies; a total of 9 sequences(Additional file 1: Table S5). The coding sequences minusputative chloroplast target peptide region were synthesised(quality: research grade) and inserted into the pUC57 vec-tor (Genescript, NJ).We cloned the coding sequences into the pASK-IBA37+expression vector (IBA GmbH, Germany), using methodsdescribed by Köllner et al. [68,69]. Briefly, the codingsequences were amplified from the pUC57 vector usingprimers with overhangs complimentary to the vectoroverhangs. The amplicon was ligated into the pASK-IBA37+ vector; these constructs were introduced intoTOP10 E. coli cells (Invitrogen, CA) and fully sequencedto confirm the absence of PCR-induced errors. Liquidcultures were grown at 37°C to OD600 between 0.5 and0.6. The expression of pASK-IBA37+ constructs in TOP10cells was induced with anhydrotetracycline (final concen-tration 200 μg.l). After 20 h incubation at 18°C, the cellswere collected by centrifugation and disrupted by a 4 ×30 s treatment at 50% with a sonicator (Branson Sonifier250, Germany) in chilled extraction buffer (50 mM Tris–HCl pH 7.5, 10% (v/v) glycerol, 5 mM MgCl2, 5 mMdithiothreitol, 5 mM Na-ascorbate pH 7.0, 0.5 mM phe-nylmethylsulfonyl fluoride). The cell fragments were re-moved by centrifugation at 14,000 × g, and the supernatantwas desalted into the assay buffer (10 mM Tris–HClpH 7.5, 10% glycerol (v/v), 1 mM dithiothreitol) by pas-sage through a Econopac 10DG column (Bio-Rad, CA).Standard assays containing 30 μL of the bacterial ex-tract in assay buffer with 13.2 ng/μl GPP or (E,E)-FPP,10 mM MgCl2 and 58 μl of assay buffer (10 mM Tris–HCl, 10% glycerol, 1 mM DTT; pH 7.5), in a glass gaschromatograph assay tube (Macherey-Nagel, Germany).A solid phase microextraction fibre consisting of 100 μmpolydimethylsiloxane (SUPELCO, PA) was placed into theheadspace of the tube during a 45 min incubation at 35°C.The solid phase microextraction fibre was then directlyinserted into the injector of a gas chromatograph for pro-duct analysis.A Shimadzu model GC-2010 gas chromatograph wasemployed with the carrier gas hydrogen at 1 ml·min−1,splitless injection (injector temperature: 220°C) with an EC-5 column (5% phenyl-methylpolysiloxane, 30 m × 0.25 mmCK drafted the manuscript and all co-authors edited and approved theKülheim et al. BMC Genomics  (2015) 16:450 Page 17 of 18final manuscript.AcknowledgementsWe thank Dr Erich Lassak and Dr Joe Brophy for discussions about the natureof terpenes in Eucalyptus. The work of CK, AP and WJF is supported by theAustralian Research Council (LP110100184 and DP14101755) and the RuralIndustries Research and Development Corporation. SK, TK and JD are supportedby European Commission (QLRT-2001-01930 and MRTN-CT-2003-504720), theGerman Science Foundation (DE8372-2), and the Max Planck Society. CH andAM acknowledge support from the South African Department of Science andTechnology (DST), Sappi and Mondi, through the Wood and Fibre MolecularGenetics (WFMG) Programme, the Technology and Human Resources forIndustry Programme (THRIP) and the National Research Foundation (NRF) ofSouth Africa.Author details1Research School of Biology, College of Medicine, Biology and theEnvironment, Australian National University, Canberra 0200, Australia.2Department of Botany, University of British Columbia, Vancouver, BCV6T1Z4, Canada. 3Institut für Pharmazie, Martin-Luther UniversitätHalle-Wittenberg, 06120 Halle, (Saale), Germany. 4Department ofi.d. × 0.25 μm film thickness, (Grace, Deerfield, IL)). Thetemperature program for monoterpenes was from 50°C (3-min hold) to 150°C at 7°C·min−1, then to 300°C (2 minhold) at 100°C·min−1; and for sesquiterpenes from 80°C(3 min hold) to 200°C at 7°C·min−1, then to 300°C (2 minhold) at 100°C. min−1. Terpenes were detected with a massspectrometer (GCMS-QP 2010 Plus, Shimadzu), which wascoupled to the gas chromatograph. Terpenes were identi-fied with the Shimadzu software “GCMS Postrun Analysis”with the mass spectral library “Wiley8” (Hewlett Packard,Palo Alto).Availability of supporting dataThe data sets supporting the results of this article areincluded within the article and its additional files.Additional filesAdditional file 1: This file contains 6 supplemental tables.Additional file 2: This file contains 3 supplemental figures.Additional file 3: This file contains 2 supplemental text documents.AbbreviationsTPS: Terpene synthase; MEP: Methylerithritol phosphate pathway;MVA: Mevalonate pathway; GPP: Geranyl pyrophosphate; GGPP: Geranylgeranylpyrophosphate; FPP: Farnesyl pyrophosphate; LG: Linkage group; AICc: Akaike’sinformation criterion value, corrected for samples size; FPKM: Fragments per kbof coding region per million mapped fragments; EST: Expressed sequence tag;QTL: Quantitative trait loci.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsCK and WJF conceived the study, CK and AP carried out the genomicanalysis and prepared all figures. CH and AAM conducted the RNAseqexperiment and primary analysis (mapping and expression count) andSTK, TGK and JD conducted the functional characterization of TPS genes.Biochemistry, Max Planck Institute for Chemical Ecology, 07745 Jena,Germany. 5Department of Genetics, Forestry and Agricultural BiotechnologyInstitute, Private Bag X20, Pretoria 0028, South Africa.Received: 17 December 2014 Accepted: 28 April 2015References1. Edwards PB, Wanjura WJ, Brown WV, Dearn JM. Mosaic resistance in plants.Nature. 1990;347(6292):434–4.2. Stone C, Bacon PE. Relationships among moisture stress, insect herbivory,foliar cineole content and the growth of river red gum (Eucalyptuscamaldulensis). J Appl Ecol. 1994;31(4):604–12.3. Hume ID, Esson C. Nutrients, antinutrients and leaf selection by captivekoalas (Phascolarctos cinereus). Aust J Zool. 1993;41(4):379–92.4. Southwell IA. Essential Oil Content of Koala Food Trees. In: Bergin TJ, editor.The Koala: Proceedings of the Taronga Symposium on Koala Biology,Management and Medicine. Sydney: Zoological Parks Board of New SouthWales; 1978. p. 62–78.5. Lawler IR, Stapley J, Foley WJ, Eschler BM. Ecological example ofconditioned flavor aversion in plant-herbivore interactions: effect ofterpenes of Eucalyptus leaves on feeding by common ringtail and brushtailpossums. J Chem Ecol. 1999;25(2):401–15.6. Eyles A, Davies NW, Yuan ZQ, Mohammed C. Host responses to naturalinfection by Cytonaema sp. in the aerial bark of Eucalyptus globulus.Forest Pathology. 2003;33:317-3317. Del Moral R, Muller CH. The allelopathic effects of Eucalyptus camaldulensis.Am Midl Nat. 1970;83:254–82.8. Giamakis A, Kretsi O, Chinou I, Spyropoulos CG. E. camaldulensis: volatilesfrom immature flowers and high production of 1,8-cineole and β-pinene byin vitro cultures. Phytochemistry. 2001;58:351–5.9. Molina A, Reigosa MJ, Carbelleria A. Release of allelochemical agents fromlitter throughfall and topsoil of plantations of Eucalyptus globulus (Labill) inSpain. J Chem Ecol. 1991;17:147–60.10. Trenbath BR, Fox LR. Insect frass and leaves from Eucalyptus bicostata asgermination inhibitors. Australian Seed Science News. 1976;2:34–9.11. Sharkey TD, Yeh SS. Isoprene emission from plants. Annu Rev Plant PhysiolPlant Mol Biol. 2001;52:407–36.12. The Arabidopsis Genome I. Analysis of the genome sequence of theflowering plant Arabidopsis thaliana. Nature. 2000;408(6814):796–815.13. Keszei A, Brubaker CL, Foley WJ. A molecular perspective on terpenevariation in Australian Myrtaceae. Aust J Bot. 2008;56(3):197–213.14. Padovan A, Keszei A, Wallis IR, Foley WJ. Mosaic eucalypt trees suggestgenetic control at a point that influences several metabolic pathways.J Chem Ecol. 2012;38:914–23.15. Külheim C, Yeoh SH, Wallis IR, Laffan S, Moran GF, Foley WJ. The molecularbasis of quantitative variation in foliar secondary metabolites in Eucalyptusglobulus. New Phytol. 2011;191:1041–53.16. Chen F, Tholl D, Bohlmann J, Pichersky E. The family of terpene synthases inplants: a mid-size family of genes for specialized metabolism that is highlydiversified throughout the kingdom. Plant J. 2011;66(1):212–29.17. Aubourg S, Lecharny A, Bohlmann J. Genomic analysis of the terpenoidsynthase (AtTPS) gene family of Arabidopsis thaliana. Mol Genet Genomics.2002;267(6):730–45.18. Martin DM, Aubourg S, Schouwey MB, Daviet L, Schalk M, Toub O, et al.Functional annotation, genome organization and phylogeny of thegrapevine (Vitis vinifera) terpene synthase gene family based on genomeassembly, FLcDNA cloning, and enzyme assays. BMC Plant Biol. 2011;10–226.19. Falara V, Akhtar TA, Nguyen TTH, Spyropoulou EA, Bleeker PM, SchauvinholdI, et al. The tomato terpene synthase gene family. Plant Physiol.2011;157(2):770–89.20. Li G, Kollner GT, Yin Y, Jiang Y, Chen H, Xu Y, et al. Non-seed plantSelaginella moellendorfii has both seed plant and microbial types of terpenesynthases. Proc Natl Acad Sci. 2012;109:14711–5.21. Irmisch S, Jiang Y, Chen F, Gershenzon J, Koellner T. Terpene synthases andtheir contribution to herbivore-induced volatile emission in western balsampoplar (Populus trichocarpa). BMC Plant Biol. 2014;14:270.22. Carr DJ, Carr SGM. Oil glands and ducts in Eucalyptus L’Herit. II.Development and structure of oil glands in the embryo. Aust J Bot.1970;18:191–212.23. Myburg AA, Grattapaglia D, Tuskan GA, Hellsten U, Hayes RD, Grimwood J,et al. The genome of Eucalyptus grandis. Nature. 2014;510(7505):356–62.24. Henery ML, Moran GF, Wallis IR, Foley WJ. Identification of quantitative trait lociinfluencing foliar concentrations of terpenes and formylated phloroglucinolcompounds in Eucalyptus nitens. New Phytol. 2007;176(1):82–95.compounds from Eucalyptus spp. in southern Australia. Atmos Environ.2009;43:3035–43.50. Kliebenstein DJ, Osbourn A. Making new molecules - evolution of pathwaysfor novel metabolites in plants. Curr Opin Plant Biol. 2012;15(4):415–23.51. Takos AM, Rook F. Why biosynthetic genes for chemical defensecompounds cluster. Trends Plant Sci. 2012;17:383–8.52. Philippe RN, Ralph SG, Külheim C, Jancsik SI, Bohlmann J. Poplar defenseagainst insects: genome analysis, full-length cDNA cloning, and transcriptomeand protein analysis of the poplar Kunitz-type protease inhibitor family.Külheim et al. BMC Genomics  (2015) 16:450 Page 18 of 1825. Padovan A, Keszei A, Külheim C, Foley W. The evolution of foliar terpenediversity in Myrtaceae. Phytochem Rev. 2014;13(3):695–716.26. Keeling CI, Bohlmann J. Genes, enzymes and chemicals of terpenoiddiversity in the constitutive and induced defence of conifers against insectsand pathogens. New Phytol. 2006;170(4):657–75.27. Birol I, Raymond A, Jackman SD, Pleasance S, Coope R, Taylor GA, et al.Assembling the 20 Gb white spruce (Picea glauca) genome fromwhole-genome shotgun sequencing data. Bioinformatics. 2013.28. Nystedt B, Street NR, Wetterbom A, Zuccolo A, Lin Y-C, Scofield DG, et al.The Norway spruce genome sequence and conifer genome evolution.Nature. 2013;497(7451):579–84.29. Heskes AM, Goodger JQD, Tsegay S, Quach T, Williams SJ, Woodrow IE.Localization of oleuropeyl glucose esters and a flavanone to secretorycavities of Myrtaceae. PLoS ONE. 2012;7, e40856.30. Heskes AM, Lincoln CN, Goodger JQD, Woodrow IE, Smith TA.Multiphoton fluorescence lifetime imaging shows spatial segregation ofsecondary metabolites in Eucalyptus secretory cavities. J Microsc.2012;247:33–42.31. Elaissi A, Medini H, Simmonds M, Lynen F, Farhat F, Chemli R, et al. Variationin volatile leaf oils of seven Eucalyptus species harvested from Zernizaarboreta (Tunisia). Chem Biodivers. 2011;8:362–72.32. Boland DJ, Brophy JJ, House APN. Eucalyptus Leaf Oils, Use, Chemistry,Distillation and Marketing. Inkata Press: Melbourne; 1991.33. Coffi K, Soleymane K, Harisolo R, Balo T, Claude C, Pierre C, et al.Monoterpene hydrocarbons, major components of the dried leavesessential oils of five species of the genus Eucalyptus from Côte d’Ivoire.Nat Sci. 2012;4:106–11.34. De Olivera Flavia NM, Pedro FH, Paula JR, Seraphin JC, Estefano FP. Seasonalinfluence on the essential oil compositions of Eucalyptus urophylla S. T.Blake and E. grandis W. Hill ex Maiden from Brazilian Cerrado. J Essent OilRes. 2008;20:555–60.35. Ogunwande AI, Olawore NO, Adeleke AK, Konig AW. Chemical compositionof the essential oils the leaves of three Eucalyptus species growing inNigeria. J Essent Oil Res. 2003;15:297–301.36. Carr SGM, Carr DJ. Oil glands and ducts in Eucalyptus L’Hérit. I. The phloemand the pith. Aust J Bot. 1969;17:471–513.37. Lassak EV, Southwell IA. The bark oil of Eucalyptus crenulata. Phytochemistry.1969;69:667–8.38. Wenke K, Ka M, Piechulla B. Belowground volatiles facilitate interactionsbetween plant roots and soil organisms. Planta. 2010;231:499–506.39. Akiyama K, Matsuzaki K, Hayashi H. Plant sesquiterpenes inducehyphal branching in arbuscular mycorrhizal fungi. Nature.2005;435(7043):824–7.40. Steffen RB, Antoniolli ZI, Steffen GPK, da Silva RF. Essential oil of Eucalyptusgrandis Hill ex Maiden in stimulating mycorrhizal Sibipiruna seedlings(Caesalpinia peltophoroides Benth.). Ciencia Florestal. 2012;22(1):69–78.41. Danner H, Boeckler GA, Irmisch S, Yuan JS, Chen F, Gershenzon J, et al.Four terpene synthases produce major compounds of the gypsy mothfeeding-induced volatile blend of Populus trichocarpa. Phytochemistry.2011;72:897-90842. Martin DM, Gershenzon J, Bohlmann J. Induction of volatile terpenebiosynthesis and diurnal emission by methyl jasmonate in foliage of norwayspruce. Plant Physiol. 2003;72:897-908.43. Eyles A, Davies NW, Mohammed CM. Traumatic oil glands induced bypruning in the wound-associated phloem of Eucalyptus globulus: Chemistryand histology. Trees-Struct Funct. 2004;18(2):204–10.44. Keszei A, Brubaker CL, Carter R, Kollner T, Degenhardt J, Foley WJ.Functional and evolutionary relationships between terpene synthases fromAustralian Myrtaceae. Phytochemistry. 2010;71(8–9):844–52.45. Sharkey TD, Wiberley AE, Donohue AR. Isoprene emission from plants: whyand how. Ann Bot. 2008;101(1):5–18.46. Behnke K, Ehlting B, Teuber M, Bauerfeind M, Louis S, Hasch R, et al.Transgenic, non-isoprene emitting poplars don’t like it hot. Plant J.2007;51(3):485–99.47. Loreto F, Mannozzi M, Maris C, Nascetti P, Ferranti F, Pasqualini S. Ozonequenching properties of isoprene and its antioxidant role in leaves.Plant Physiol. 2001;126(3):993–1000.48. Affek HP, Yakir D. Protection by isoprene against singlet oxygen in leaves.Plant Physiol. 2002;129(1):269–77.49. Winters AJ, Adams MA, Bleby TM, Rennenberg H, Steigner D, Steinbrecher R,et al. Emissions of isoprene, monoterpene and short-chained carbonylNew Phytol. 2009;184(4):865–84.53. Nagy ED, Bennetzen JL. Pathogen corruption and site-directedrecombination at a plant disease resistance gene cluster. Genome Res.2008;18:1918–23.54. Ober D. Gene duplications and the time thereafter - examples from plantsecondary metabolism. Plant Biol. 2010;12(4):570–7.55. Sharkey TD, Gray DW, Pell HK, Breneman SR, Topper L. Isoprene synthasegenes form a monophyletic clade of acyclic terpene synthases in the TPS-bterpene synthase family. Evolution. 2013;67(4):1026–40.56. Padovan A, Keszei A, Kollner TG, Degenhardt J, Foley WJ. The molecularbasis of host plant selection in Melaleuca quinquenervia by a successfulbiological control agent. Phytochemistry. 2010;71(11–12):1237–44.57. Andrew RL, Wallis IR, Harwood CE, Henson M, Foley WJ. Heritable variationin the foliar secondary metabolite sideroxylonal in Eucalyptus conferscross-resistance to herbivores. Oecologia. 2007;153(4):891–901.58. Lawler IR, Foley WJ, Eschler BM. Foliar concentration of a single toxincreates habitat patchiness for a marsupial folivore. Ecology.2000;81(5):1327–38.59. Matsuki M, Foley WJ, Floyd RB. Role of volatile and non-volatile plantsecondary metabolites in host tree selection by Christmas beetles.J Chem Ecol. 2011;37:286–300.60. Guindon S. Bayesian estimation of divergence times from large sequencealignments. Mol Biol Evol. 2010;27:1768–81.61. Drummond AJ, Ashton B, Buxton S, Cheung M, Cooper A, Duran C, et al.Geneious v5.4. In.: Available from http://www.geneious.com/; 2011.62. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA,McWilliam H, et al. Clustal W and Clustal X version 2.0. Bioinformatics.2007;23(21):2947–8.63. Morariu VI, Srinivasan BV, Raykar VC, Duraiswami R, Davis LS. Automaticonline tuning for fast Gaussian summation. Adv Neural Inf Process Syst.2008;2008:1113–20.64. Mizrachi E, Hefer CA, Ranik M, Joubert F, Myburg AA. De novo assembledexpressed gene catalog of a fast-growing Eucalyptus tree produced byIllumina mRNA-Seq. BMC Genomics. 2010;11:681.65. Chen W, Böcker W, Brosius J, Tiedge H. Expression of neural BC200 RNA inhuman tumours. J Pathol. 1997;183:345–51.66. Trapnell C, Pachter L, Salzberg SL. TopHat: discovering splice junctions withRNA-Seq. Bioinformatics. 2009;25:1105.67. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, Van Baren MJ, et al.Transcript assembly and quantification by RNA-Seq reveals unannotatedtranscripts and isoform switching during cell differentiation. Nat Biotechnol.2010;28:511–5.68. Köllner TG, Schnee C, Gershenzon J, Degenhardt J. The variability ofsesquiterpenes cultivars is controlled by allelic emitted from two Zea maysvariation of two terpene synthase genes encoding stereoselective multipleproduct enzymes. Plant Cell. 2004;16(5):1115–31.69. Köllner TG, Gershenzon J, Degenhardt J. Molecular and biochemicalevolution of maize terpene synthase 10, an enzyme of indirect defense.Phytochemistry. 2009;70:1139–45.


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