The Best of Santalum Album: Essential Oil Composition, Biosynthesis and Genetic Diversity in the Australian Tropical Sandalwood Collection

The best of Santalum album: Essential oil composition, biosynthesis and genetic diversity in the Australian tropical sandalwood collection

Christopher G. Jones

Apologies to Kerstin Bach and Carlos Santana…

This thesis is presented for the degree of Doctor of Philosophy

The University of Western Australia

School of Plant Biology

Faculty of Natural and Agricultural Sciences June 2008

DECLARATION FOR THESIS CONTAINING PUBLISHED WORK AND / OR WORK PREPARED FOR PUBLICATION

This thesis contains published work and / or work prepared for publication, some of which has been co-authored. The bibliographic details of the works and where they appear in the thesis are set out below. (The candidate must attach to this declaration a statement detailing the percentage contribution of each author to the work. This must be signed by all authors. Where this is not possible, the statement detailing the percentage contribution of authors should be signed by the candidate‟s co-ordinating supervisor).

The following manuscripts have either been published or have been submitted to scientific journals.

Chapter 2: Jones CG, Plummer JA (2008) “Sandalwood”. In 'A Compendium of Transgenic Crop Plants: Forest Tree Species'. (Eds C Kole and TC Hall) pp. in press. (Blackwell Publishing: Oxford UK)

Chapter 3: Jones CG, Barbour EL, Plummer JA and Byrne M. “Genetic diversity of the Australian tropical sandalwood collection – biogeographic origins and tree improvement potential”. Submitted to Tree Genetics and Genomes (March 2008)

Chapter 4: Jones CG, Ghisalberti EL, Barbour EL, Plummer JA. (2006) “Quantitative co-occurrence of sesquiterpenes; a tool for elucidating their biosynthesis in Indian Sandalwood, Santalum album L”. Phytochemistry, 67, 2463-2468

Chapter 5: Jones CG, Keeling CI, Ghisalberti EL, Barbour EL, Plummer JA and Bohlmann J. “Isolation of cDNAs and functional characterisation of two multi-product TPS enzymes from sandalwood, Santalum album L”. Accepted in Archives of Biochemistry and Biophysics (May 2008).

The contribution of work for each manuscript is 90% by the candidate of this thesis, Christopher G Jones and 10% for all other authors, except for Chapter Five where the candidate contributed 70%, C I Keeling 20% and 10% for all other authors.

Candidate, Christopher G Jones

Signature:……………………………………………………………………

Co-ordinating supervisors, Julie A Plummer and Emilio L Ghisalberti

Signatures:……………………………………………………………………

ii Statement of Original Contribution

The research presented in this thesis constitutes an original contribution to high-value tropical forestry, genetics and secondary metabolite biosynthesis. The ideas and hypotheses are my own, with help and critique from my supervisors to test these. In the experimental sections, other people who have helped in any significant capacity are dutifully acknowledged as co-authors or in the acknowledgements sections of each manuscript.

This thesis is presented as a summary of my PhD course at the University of Western

Australia, and has not been used elsewhere for credit towards any other endeavours at any other institution.

Christopher Graham Jones

June 2008

iii Abstract

An investigation into the causes of heartwood and essential oil content of Australian plantation sandalwood, Santalum album was undertaken. Genetic diversity of 233 S. album, five S. austrocaledonicum and fifteen S. macgregorii trees growing in the Forest Products Commission arboretum, Kununurra WA, was assessed using nuclear and chloroplast RFLPs. Santalum spicatum was chosen as an out-group. Nuclear genetic diversity of the S. album collection was very low, with observed and expected heterozygosity levels of 0.047. This was lower than the results previously reported in the literature for trees in India, however a different technique was used. Based on allelic patterns, the collection was able to be categorised into 19 genotypes; each representing some shared genetic origin. Some groups were highly redundant with 56 trees being represented, while others were populated by just one tree. The essential oil yield and heartwood contents of trees from these genetic groups were compared. Yields were highly variable both within and between groups of trees which share a common genetic history, suggesting a significant environmental component was contributing to the observed phenotype, despite identical soil and climatic conditions. Ancestral lineages were tested using chloroplast RFLPs, although a lack of shared mutations between species made this difficult. Only one S. album tree originating from Timor was resolved using nuclear RFLPs, with the other trees being grouped with material sourced from India. There was no resolution of Indian S. album from Timorese using chloroplast RFLPs, however one S. album tree grown from Indian seed possessed a single unique mutation. The low genetic diversity of the Australian S. album collection is likely to be a combination of incomplete seed sourcing and highly restricted gene flow during the evolution of the species. Combined with information gathered on the phylogeny of the genus by other researchers, S. album is postulated to have originated from an over-sea dispersal out of northern Australia or Papua New Guinea 3 to 5 million years ago. Essential oil yield and composition was assessed for 100 S. album trees growing in the collection, ranging in age from 8 to 17 years. Oil content of heartwood ranged from 30 mg g-1 to 60 mg g-1, and the transition zone 36 mg g-1 to 90 mg g-1. Sapwood contained almost no sesquiterpene oils. Despite the highly variable total oil yields, the chemical profile of the oil did not vary, suggesting there was limited genetic diversity within this region of the genome. Strong, positive correlations existed between iv sesquiterpenoids in the essential oil of S. album. This was particularly evident in the santalenes and -bergamotene, although trends were also seen in the curcumenes, bisabolene and bisabolol, and an unidentified group of sesquiterpenes. Co-occurrence patterns indicate shared chemical intermediates from which compounds are partitioned, resulting in multiple product formation. To further test the multiple product hypothesis, two terpene synthase (TPS) genes were isolated from S. album leaf and wood cDNA. These were cloned and the enzymes were heterologously expressed in Escherichia coli. One enzyme, SasesquiTPS1 converted the universal sesquiterpene precursor farnesyl diphosphate into germacrene D-4-ol, helminthogermacrene, -bulnesene, -muurolene - and -selinene. The other enzyme, SamonoTPS1 converted geranyl diphosphate into (+) -terpineol and ()-limonene, with small amounts of sabinene, linalool, - terpinolene, myrcene and geraniol. These represent the first TPS genes to be isolated from sandalwood and will enable further elucidation of oil biosynthesis genes. This thesis compiles a three-pronged approach to understanding the underlying causes of oil yield variation in S. album. As a species for which so little is known, the research presented here provides a major leap forward for tree improvement, breeding and silviculture. Hence the best of Santalum album research is presented.

v Acknowledgements

I am indebted to my supervisors, Julie Plummer and Emilio Ghisalberti for their support, advice and guidance throughout my entire PhD. I‟m yet to meet a student with a sour thing to say about either person. What I have appreciated most is the freedom to explore widely, but also gave sound advice at keeping it relevant and answer the question at hand. This project would not have been possible without the generous support of the Forest Products Commission of Western Australia and the dedication of my industry supervisor, Liz Barbour. Without Liz, the sandalwood project would have surely lost momentum. I am most grateful to Margaret Byrne and the Science Division at the West Australian Department of Environment and Conservation for her solid knowledge of population genetics and molecular biology. Often, there were few people in the entire state of WA who could help me with a specific problem, and Margaret‟s advice was always worth the wait. I owe more than morning tea to Bronwyn and

Maggie at the lab, for if it weren‟t for them, the radiation work would have been on the

6 o‟clock news for sure. Also, I thank Lin Wong for her help with DNA extractions and digest preparations. I am most grateful to Jörg Bohlmann, who kindly let me into his lab at the University of British Columbia, Vancouver. I learnt molecular biology for the first time, as well as the finer details of terpenoid biosynthesis in his lab. Without his generous support, this project would not be nearly as exciting. I look forward to many future collaborations with Jörg and his research group.

I am appreciative for the funding provided to this Linkage Project by the

Australian Research Council and the Forest Products Commission. In addition to this, the generous awards, scholarships and travelling fellowships I have received throughout my PhD have made life much easier. In particular, The Commonwealth Government

Department of Agriculture, Fisheries and Forestry Science and Innovation Award

vi (2005) enabled me to conduct the gene discovery and biochemical expression studies detailed in Chapter 5. This has opened several doors for future research, including a collaborative project with the Michael Smith Laboratories at UBC. Travel costs were minimised courtesy of the Mike Carroll travelling fellowship and the UWA

Postgraduate Students Association Travel Award, as well as the Mary Janet Lindesay of

Yanchep fellowship. For this I am most thankful.

To the many friends and mentors who have enriched my life and learning, I offer my gratitude. Special regards are due to my good friend Vince Bowry, who helped me through my undergraduate years and sustained my interest in chemistry.

Particular thanks go to my closest friends and colleagues Nic George and Cam Beeck,

Leida Williams, Stuart Pearse, Ben Biddulph, Eleftheria Dalmaris, Pat Mitchell,

Lindsay Bell and Al Grigg. My new friends from UBC; Carsten, Björn, Britta, Eric,

Chris, Betty, Wing-Ka and Karen, thanks and I hope to see you all soon!

I must also mention Robert and Lisa of Artisans café in Brisbane. If it were not for them employing me as a kitchen-hand, I would never have saved enough money to move to Perth and start a new life. See, I told you you‟d get a mention in my thesis!

Finally I wish to acknowledge the unwavering love and support from my parents. They have been keen to support my every endeavour, academic or otherwise.

There have been some pretty lean times in recent years, and without their support I‟d definitely be hungry and broke. More so.

vii …As the afternoon wore on, the conversation turned to botany and in particular a new book that made a fuss about house-plants responding to music and human voices. For Tim, the very idea was ludicrous.

„Why would a plant give a shit about Mozart?‟ I remember him saying. „And even if it did, why should that impress me? I mean, they can eat light. Isn‟t that enough?‟

He went on to speak of photosynthesis the way an artist might describe colour. He said that at dusk, the process is reversed and that plants actually release small amounts of light. He referred to sap as the green blood of plants, explaining that chlorophyll is structurally almost the same as the pigment of our blood, only the iron in haemoglobin is replaced by magnesium in plants. He spoke of the way plants grow, a seed of grass producing sixty miles of root hairs in a day, six thousand miles over the course of a season; a field of hay exhaling five hundred tons of water into the air each day; a flower pushing it‟s blossom through three inches of pavement; a single catkin of a birch tree producing five million grains of pollen; a tree living for four thousand years. Unlike every botanist I had known, he was not obsessed with classification. For him Latin names were like koans or lines of verse. He remembered them effortlessly, taking particular delight in their origins.

„When you say the names of the plants,‟ he said at one point, „you say the names of the gods.‟…

-Wade Davis, remembering his friend and contemporary, the late Timothy Plowman.

Davis, W. (1996) “One River” Simon and Schuster Paperbacks, New York.

viii Table of contents

Title page …i Publications arising from this thesis …ii Statement of original contribution …iii Abstract …iv Acknowledgements …vi Table of contents …ix Definition of Terms Used …xi List of Tables and Figures …xiii

Chapter One: General Introduction …1

Chapter Two: Review of the Literature …5 References …22

Chapter Three: Genetic diversity of an Australian tropical sandalwood (Santalum album) collection – biogeographic origins and tree improvement potential …29 Abstract …31 Introduction …32 Materials and Methods …34 Results …36 Discussion …43 Acknowledgements …50 References …51

Chapter Four: Quantitative co-occurrence of sesquiterpenes; a tool for elucidating their biosynthesis in Indian Sandalwood, Santalum album L. …55 Abstract …57 Introduction …58 Results and discussion …60 Experimental …67 Acknowledgements …70 References …71

Chapter Five: Isolation of cDNAs and functional characterisation of two multi-product terpene synthase enzymes from sandalwood, Santalum album L. …73 Abstract …75 Introduction …76 Results …77 Discussion …86 Experimental …92 Acknowledgements …100 References …100

Chapter Six: General Discussion …104 Genetic diversity of the FPC Kununurra collection …104 Chemical diversity …109 Terpenoid biosynthesis in S. album …112 Conclusion …115 Complete list of references …116

Appendix I. Photographs …130

x Definition of terms used in this thesis Australo-Papua: The geographical region incorporating eastern Indonesia, Papua New Guinea (PNG) and northern Australia; sometimes referred to as Wallacea. bp: Base-pair CaMV35S: Cauliflower mosaic virus promoter Core (wood): Sample of stem wood taken through the diameter of standing trees Core collection: Collection of germplasm representing the genetic diversity of a species NPTII: Neomycin phosphotransferase II (gene) NOS: Nopaline synthase promoter

DXP 5-P: Deoxy-D-xylulose 5-phosphate Carbocation: positively charged molecular ion cDNA: Complimentary DNA (derived from mRNA) DEC: Department of Environment and Conservation, Western Australia DMAPP: Dimethylallyl diphosphate E. coli: Escherichia coli (bacteria) Essential oil: Oil consisting of terpenoids which can be extracted from plant material EST: Expressed sequence tag. Usually from the 3‟- end of cDNA FPC: Forest Products Commission of Western Australia FPP: Farnesyl pyrophosphate Genetic diversity: A measure of how variable the genome of a species is Genome: The sum of all genes within an organism Genotype: The genetic constitution, latent or expressed, of an organism. Unique to an individual and its clones. Group of trees bearing identical genetic markers GUS: -glucuronidase enzyme (gene) GPP: Geranyl diphosphate Growth characteristics: Measurable physical traits of a tree such as height, girth, bole height, diameter at breast height, health and form Haplotype: Unique combination of mutations which are transmitted together

He, Ho: Expected and observed heterozygosity, respectively Heartwood: Discoloured wood located in the centre of a mature tree stem. Typically not water conducting; extractives are often deposited here Heterozygosity: Presence of more than one allele at a given loci

xi Isozyme (allozyme): Technique used to infer allele variation using enzyme electrophoresis IPP: Isopentyl diphosphate Locus (pl. loci): Discrete location on a chromosome

Monoterpene: Volatile 10 carbon chemical based on the isoprene (C5) unit Ma: Million years before present mRNA: Messenger RNA MVA: Mevalonic acid NOS: Nopaline synthase promoter NT: Northern Territory OPP: Diphosphate anion Phenotype: The observed physical appearance of a tree. The combined outcome of genotype and the growing environment Phytoalexin: Chemical produced by an organism to ward off pests and/or pathogens Polytomy: A section of phylogeny where several taxa cannot be resolved to a dichotomy Qld: Queensland QTL: Quantitative trait loci RAPD: Randomly amplified polymorphic DNA Ray parenchyma: Live cells which radiate from the centre of a stem to the cambium RFLP: Restriction fragment length polymorphism Sapwood: Water conducting secondary xylem located to the outside of heartwood yet inside the cambium of a tree

Sesquiterpene: 15 carbon chemical based on the isoprene (C5) unit. Less volatile than monoterpenes and often fragrant SSR: Short sequence repeat; microsatellite T-DNA: Transferred DNA of the Agrobacterium tumefaciens tumour-inducing plasmid TPS: Terpene synthase enzyme (gene); terpene cyclase enzyme (gene) Transition zone: Region of xylem at the interface of sapwood and heartwood uidA: gene encoding GUS WA: Western Australia

xii List of tables and figures

Tables:

Chapter 3:

Table 3.1. Mean heterozygosity statistics of Santalum spp. from the Kununurra arboretum compared with those of natural populations. …37

Table 3.2 Mutations detected in the chloroplast genomes of 4 Santalum species. …40

Chapter 5:

Table 5.2. GC-MS profile of SasesquiTPS9 when incubated with FPP. Composition determined by cold on-column GC-MS analysis as well as GC-FID (splitless injector). 1MassFinder terpenoid RI library and Adams (1995). † Overlapping peaks. …84

Figures:

Chapter 1: Fig 1.1 Map showing the distribution of Santalum in the Australo-Papuan region …2

Chapter 3: Figure 3.1. UPGMA phenogram of the 19 genotypes of S. album and five genotypes of S. austrocaledonicum identified in trees from the FPC Kununurra collection. Number of trees within each group shown in parenthesis. Groups containing individuals sourced from Timor and the Northern Territory (NT) are also indicated. All other seed was

xiii sourced from India. Santalum spicatum serves as a distant out group, with five individuals from northern and southern populations of Western Australia. …38

Figure 3.2. Principal component analysis of the 19 S. album genotypes and related Santalum species. S. album genotypes 18 and 19 are discussed in the text. Genotype 2 is highlighted for reference. …39

Figure 3.3. Phylogenetic parsimony tree of Santalum haplotypes, generated from PAUP v4.0. Numbers above branches represent individual mutations. Question mark indicates S. album trees with the chloroplast genome of S. austrocaledonicum. …41

Figure 3.4. A: Extractable oil yields (mg g-1 dry wt.) of wood cores from 30 cm above ground level and B: Heartwood contents (%) of these core samples from selected S. album genotypes (±S.E.). n = 4, 7, 11, 5, 3, 4 and 3 for genotypes 1 to 13 respectively …42

Chapter 4: Fig 4.1 Major sesquiterpene alcohols of Santalum album. …58

Fig 4.2 Proposed biosynthesis of the santalenes and bergamotene (6-9) …62

Fig 4.3 Proposed biosynthetic pathway of farnesyl diphosphate to the curcumenes (10 and 11) and cis-lanceol. …63

Fig 4.4 Alternative folding of acyclic precursor farnesyl diphosphate in Santalum album. …64

Fig 4.5 Co-occurrence of α- and β- santalol (1 and 2 respectively) in S. album oil …65

Fig 4.6 A: Co-occurrence of compounds 6-9 in S. album oil: a; 7 vs. 8 (y=7.54 x + 0.13, r=0.98), b; compounds 8 vs. 9 (y=5.08 x + 0.74, r=0.98), c; compounds 6 vs. 8 (y=1.28 x + 0.31, r=0.88), d; compounds 9 vs. 7 (y=1.21 x + 0.16, r=0.90), e; compounds 6 vs. 9 (y=0.92 x + 0.18, r=0.85), f; compounds 6 vs. 7 (y=0.18 x + 0.002,

xiv r=0.98). B: Co-occurrence of compounds 6 and 9 with the age of wood cores highlighted …66

Chapter 5: Fig 5.1. Transverse section of S. album heartwood (lower) – sapwood (upper) transition zone (40x magnification). Essential oil and other extractives are deposited in the ray parenchyma (vertical features) during heartwood formation. …78

Fig 5.2. Alignment of SamonoTPS1 amino acid sequence with other closely related monoterpene synthases; VVTPS: Vitis vinifera α-terpineol synthase (Martin and Bohlmann 2004), CLLS: Citrus limon (-)-limonene synthase (Lucker et al. 2002), CTGS: Cinnamomum tenuipilum geraniol synthase (Yang et al. 2005). Consensus indicated on bottom row. FEPQY motif, common to both SamonoTPS1 and SasesquiTPS9 is indicated by an overhead line (─). Motifs common to all TPS indicated by a solid overhead line (▀). …80

Fig 5.3. Alignment of SasesquiTPS1 with other closely related angiosperm sesquiterpene synthase sequences; VVGDS: Vitis vinifera (-)-germacrene D synthase and VVVS: Vitis vinifera (+)-valencene synthase (Lucker et al. 2004), PBGDS: Populus trichocarpa x deltoides (-)-germacrene D synthase (Arimura et al. 2004). Consensus indicated on bottom row. FEPQY motif highlighted with an overhead line (─). Double dagger indicates Y538 of SaSesquiTPS1. Motifs common to all TPS indicated by a solid overhead line (▀). …81

Figure 5.4. Mass spectra of monoterpenes identified in the product mixture of SamonoTPS1, incubated with GPP in the presence of Mn2+. A: -pinene, B: sabinene, C: myrcene, D: limonene, E: -terpinolene, F: linalool, G: -terpineol, H: geraniol. I: GC-FID chromatogram (DB-WAX) with peak numbering for reference. …83

Figure 5.5. Mass spectra of sesquiterpenoids identified in the product mixture of SasesquiTPS9 when incubated with FPP in the presence of Mg2+. A: -humulene, B: - muurolene, C: -selinene, D: -selinene, E: -bulnesene, F: helminthogermacrene, G:

xv trans-nerolidol, H: germacrene D-4-ol, I: GC-MS chromatogram (cold on-column injection, DB-WAX) with peak numbering for reference. …85

Figure 5.6. Scheme showing hypothesised transformation of GPP by SamonoTPS1…88

Figure 5.7. Scheme showing hypothesised biosynthetic transformations of farnesyl and nerolidyl diphosphate (boxed) into compounds found in the product mixture of SasesquiTPS9. Broken arrows indicate possible heat induced dehydration and rearrangement reactions …90

Chapter 6: Figure 6.1. Possible evolutionary origins of S. album (Sa), S. lanceolatum (Sla), S. macgregorii (Sm), S. yasi (Sy) and S. spicatum (Ss). Gene flow is represented by arrows, while bottlenecks are indicated by dashed lines. Ti = Timor, Ba = Babar, F = Flores, S = Sumba, W = Wetar. Dark blue indicates seas deeper than 120m. Geological reconstructions inferred from Hall (2001). Distribution of Santalum inferred from Applegate and McKinnell (1993) Harbaugh (2007) and Paul (1990) …106

xvi Chapter 1. General introduction

Tropical, or Indian sandalwood, Santalum album, is a small hemi-parasitic tree found growing in India, Sri Lanka, Indonesia and northern Australia (Fig 1.1). Its fragrant heartwood is highly sought after in Asian and European markets for carving, incense manufacture and perfumes. The immense value and scarcity of the resource in its natural environment has led to over-exploitation, while an opportunity for plantations to satisfy demand has opened. Western Australia has led the way in S. album plantation establishment, however early harvest trials have delivered highly variable produce.

Sandalwood heartwood contains an essential oil which is composed primarily of sesquiterpenes; highly reduced 15 carbon molecules. The main components of the oil are

-, - and epi--santalol, and -trans-bergamotol. Also present are the olefinic santalenes and bergamotene. These chemicals are deposited in the heartwood in mature trees; usually after around ten to fifteen years. The rate and timing of oil deposition is poorly understood, and the causes of within-plantation differences are equally unclear. The problem is felt at an economic level at harvest since many trees may not yield fragrant heartwood. As the plantation sandalwood industry continues to grow, farmers and investors will desire reliable seed, known to produce oil-rich heartwood within a profitable timeframe.

Selection of superior phenotypes can be better guided by a knowledge of the genetic diversity which exists within the breeding population.

Phenotype may be defined as the sum of all observable traits expressed by a plant, and is the combined result of both genetic predisposition and the environment in which the organism exists. Santalum album growing in the Kununurra collection shows highly variable phenotypes, particularly in terms of heartwood content, essential oil yield and other growth characteristics like diameter at breast height, total height, health and vigour.

Fig 1.1. Distribution of Santalum species in the Australo-Papuan region. Distribution of Santalum inferred from Applegate and McKinnell (1993), Harbaugh (2007) and Paul (1990).

It remains to be determined if this variation is due primarily to genes or the growing environment.

Chapter Two of this thesis serves as a review of the literature concerning all aspects of sandalwood research to date. It covers the taxonomy, distribution and biology of the genus, the cultural significance and value of sandalwood, and how some aspects of sandalwood management and agronomy may be improved through the application of biotechnology. The body of the review looks closely at how modern molecular and biochemical approaches can further aid sandalwood tree improvement. This review is to appear as a book chapter in Jones CG, Plummer JA (2008) Sandalwood. In 'A

Compendium of Transgenic Crop Plants: Forest Tree Species'. (Eds C Kole and TC Hall) in press. (Blackwell Publishing: Oxford UK).

The problems of heartwood oil variability, production and genetic predisposition, highlighted in this review are addressed through testing three main hypotheses:

2 1. That there is substantial genetic diversity in the Australian Santalum album collection. Different genotypes, or genetic groups are likely to be present, given the geographically diverse seed collections. Trends in essential oil yield and composition may be reflected by these groupings within the collection. Also, it is hypothesised that other species found in northern Australia - coral sea region; S. austrocaledonicum, S. macgregorii and S. spicatum from the south, are sufficiently similar in genomic and chloroplast DNA sequence to construct a broad phylogenetic history of the genus.

Combined with other phylogenetic and biogeographic data, a plausible evolutionary path may be devised.

2. That the essential oil yield of the Australian Santalum album collection is variable in terms of oil yield and heartwood content, as well as varying in chemical composition. It has become apparent to the plantation sector that heartwood oil yields can be quite variable, however a thorough assessment of the level and nature of variation is needed. A mean oil yield of typical woodlots at varying stages of maturity would enable estimates of potential returns. Moreover, any tree improvement efforts must be compared to a baseline mean.

This has not been determined. Compositional differences in the essential oil profile would also be expected given the wide collections of seed and environmental conditions, including host species, soil types and light availability. Unique trends in the proportions of specific chemicals may be indicative of specific chemotypes, or imply linked biosynthetic origins.

3. That heartwood oil is synthesised by specific terpene synthase enzymes, within living tissues of the stem. Heartwood is largely non-living material, although ray parenchyma is metabolically active and acts as a conduit from the cambium to the interior of stems and roots. Biosynthesis may occur here. Terpenoids are synthesised from activated isoprenyl chains such as geranyl diphosphate or farnesyl diphosphate. These

3 chemicals are cyclised and further rearranged by terpene synthases (TPS). It is proposed that several TPS are active in sandalwood tissues and that their regulation is integral to understanding overall oil production.

The first hypothesis is tested in Chapter Three. Genetic diversity of S. album trees sampled from the Kununurra collection was analysed by nuclear and chloroplast RFLPs.

Several other closely related species were included for comparison. Heterozygosity levels, phylogenetic relationships and genotypic groupings were determined. Appropriate avenues for tree improvement of the core collection are discussed, and the evolutionary origins of the species are postulated.

The second hypothesis is tested in Chapter Four, where the essential oil yield and heartwood content of one hundred trees growing in the FPC Kununurra collection was determined. This enabled a baseline understanding of yield variation and compositional differences. Trends in the relative proportions of the chemicals within the essential oil are examined and their significance discussed.

The third hypothesis is tested in Chapter Five. cDNA was isolated from S. album wood and leaf tissue and probed for candidate TPS genes. These genes are well characterised throughout the plant kingdom and were prime candidates for oil biosynthesis in sandalwood. Functional characterisation of the genes confirmed their roles in terpenoid biosynthesis, and provides many avenues for further experimentation.

Sandalwood growing in the FPC collection located in Kununurra, Western Australia was tested for genetic and essential oil yield variation. In addition, the genes responsible for terpenoid biosynthesis were isolated and characterised. These findings are highly valuable for the plantation industry and adds to our knowledge of the species as a whole.

Chapter Six is a general discussion of the major findings and highlights gaps in the

4 knowledge where further work is needed. It discusses the overall significance of the results within the context of sandalwood tree improvement.

5 Chapter 2. Review of the Literature

The following chapter is due to appear in a book series titled “A Compendium of Transgenic Crop Plants” edited by Chittaranjan Kole and Timothy C Hall in early 2008. Chapter 11 is dedicated to Sandalwood.

SANDALWOOD Christopher G. Jones* and Julie A. Plummer School of Plant Biology (M084) The University of Western Australia, 35 Stirling Highway, Crawley 6009, Western Australia, Australia Corresponding author: [email protected]

INTRODUCTION 1.1 Taxonomy, Distribution and Ecology Sandalwood is the general name given to a small group of parasitic shrubs or trees in the genus Santalum, family Santalaceae. The taxonomic and molecular phylogeny of these species has recently been reviewed, with a comprehensive summary on 16 species and their variants (Harbaugh and Baldwin 2007). Members of the genus Santalum are widespread, growing throughout India, Indonesia, Australia, New Guinea, Hawaii and many South Pacific islands. The term sandalwood refers to those species which bear fragrant heartwood and have been utilized commercially for this purpose. Santalum album, often referred to as East Indian or tropical sandalwood has long been harvested for its valuable, odorous wood (Rai and Sarma 1990). The genetic origins of S. album appear to be East and West Timor, as well as a small population in northern Australia (Harbaugh and Baldwin 2007). A more detailed study of the Indo-Papuan species S. macgregorii and S. album in particular has found few mutations in the chloroplast genome, suggesting recent speciation (C. Jones et. al., Chapter 3, this thesis). In India, it occurs throughout the Indian subcontinent, particularly in the southern states of Karnataka, Kerala, Tamil Nadu and Andhra Pradesh (Rai 1990). Santalum spicatum is one of Western Australia‟s oldest export industries and it is still harvested commercially. To a lesser extent, S. lanceolatum, commonly known as „plumbush‟ or northern sandalwood, has been harvested in Queensland and South Australia (Applegate et al. 1990b; Applegate and McKinnell 1993; Statham 1990). In the Pacific, S. yasi, S. austrocaledonicum and S. insulare were extensively utilized (Ehrhart 1998; Jiko 1993). Sandalwoods are slow growing root hemi-parasites, meaning they derive some water and nutrients from the roots of neighboring trees and shrubs (Loneragan 1990; Radomiljac and

7 McComb 1998; Radomiljac et al. 1999; Tennakoon and Cameron 2006). Host trees vary, depending on the Santalum species and their native ecosystem. In India, S. album utilizes a wide range of hosts, and occasionally self-parasitizes (Rai 1990). Santalum album occurring in Timor similarly parasitizes numerous species, including Eugenia guajava, Casurina junghuhniana, Cassia siamea, Schleichera oleosa and Pterocarpus indicus (Suriamihardja and Suriamihardja 1993). In Australia, S. spicatum associates with Acacia accuminata and Eremophlia spp. while S. lanceolatum is often found in association with Acacia cambagei and Melaleuca spp. (Applegate and McKinnell 1993). Growth rates for each species will vary according to rainfall, soil type and climate. Tropical S. album may take 40-60 years to reach maturity (Rai 1990). In contrast, S. spicatum is the slowest growing species taking up to 100 years to reach a commercial size (Applegate et al. 1990a). Inflorescences of Santalum spp. are usually auxiliary or terminal cymose panicles, petals are 4-or 5-merous, fused to the perianth and purple-brown in color (Hewson and George 1984). The flowers have inferior ovaries yielding one single-seeded drupe. Flesh of the fruit is typically purple, however S. spicatum fruits are dry and fibrous, enclosing a hard nut (Hewson and George 1984). Although only limited work has been done on the cytological features of sandalwood, root- tip squashes have indicated that S. album is diploid and has 20 chromosomes (Kulkarni et al. 1998). Other members of the genus are believed to be tetraploid, particularly the non- oil-yielding S. acuminatum (Byrne et al. 2003b) and Papua New Guinean S. macgregorii (C. Jones et al., Chapter 3). Several other species within the genus may also be polyploid, and this is currently being investigated (Harbaugh, personal communication). Santalum album and S. spicatum are outcrossing insect-pollinated trees, however self-pollination is possible (Rugkhla et al. 1997).

1.2 History and Economic Importance All of the commercially utilized sandalwoods are pulled from the ground at harvest, the upper limbs are cut and bark removed. Logs are then cut into billets which are either processed or exported directly (Loneragan 1990). Usual down-stream processing involves grinding the wood into a fine powder and processed into incense or joss sticks. These are used in many Asian ceremonies and religious events (Loneragan 1990; Rai and Sarma 1990). The ground wood is also distilled by either hydro- or steam-distillation to produce the neat oil, often used as a fixative for many perfumes. Supercritical CO2 has been

8 utilized as an alternative to distillation with similar results (Moretta et al. 1998; Piggot et al. 1997). Santalum album yields more heartwood oil than other members of the genus, with typical values around 6.2% (w/w) (Shankaranarayana and Parthasarathi 1987; Shankaranarayana et al. 1998; Verghese et al. 1990). Extractable oil yield can vary from zero to nearly 9% (w/w) from within one tree (Jones et al. 2006). Santalum yasi from Fiji is reported to yield about 5% oil, S. austrocaledonicum around 3-5%, and S. spicatum around 2% (Applegate et al. 1990a). Sandalwood oil production from India is variable. However, the Food and Agriculture Organization of the United Nations reports an estimated 40 tons per annum over the period of 1987 to 1993 (Anonymous 1995). Indonesia, the world‟s second largest producer of S. album products, produced an average of 15 tons of oil per annum for the same period. In the 10 years from 1995 to 2005, Indonesia provided the bulk of S. album products to the world markets, while India‟s production levels fell significantly due to scarcity and legislated resource protection. In March 2006, the highest auction price in India for cleaned S. album logs was Aus$105,400 per ton (Source: Tropical Forestry Services Pty. Ltd., Australia). Sandalwood, despite being a renewable plant resource, is suffering from significant decline in India, Indonesia, the South Pacific and Australia. This is primarily due to over- harvesting and illegal poaching of native stands, or by biological hindrance from grazing animals or sandal spike disease (Loneragan 1990; Rai and Sarma 1990; Statham 1990). Harvesting of S. spicatum from native stands is likely to continue for some time although considerable effort is going into developing the species into an extensive plantation crop (Brand 2002; Radomiljac 1998; Woodall and Robinson 2002a;b). In Western Australia, S. spicatum and S. album are being planted in the south west and far north of the state, respectively. Plantations of S. album, as well as small plots of S. austrocaledonicum and S. macgregorii are being trialed at the Frank Wise Agricultural Research Station, Kununurra, Western Australia. The oldest S. album trees from these trials are approaching 20 years, and preliminary experiments have shown yields of valuable heartwood oil after 10 years. Santalum lanceolatum is being trialed in Queensland (Bristow et al. 2000), while a concerted effort to re-establish S. austrocaledonicum as a tree crop are also underway in Vanuatu (Tate et al. 2006). Santalum insulare from eastern Polynesia is critically endangered due to over-harvesting, and as such a conservation program is being implemented (Butaud et al. 2005).

9 1.3 Traditional Breeding and Improvement Programs Despite being utilized for centuries, sandalwood has received little attention from a tree improvement perspective. Efforts to understand the complex ecology of the species has largely hampered field trials, and as such most tree improvement work has been devoted to selection of superior phenotypes form naturally occurring populations. Heritability studies for desirable traits such as tree form, health, early heartwood onset, high heartwood content and essential oil yield have not been conducted extensively. Although once complete, these investigations would certainly provide a solid rationale for further tree improvement. Probably the most significant steps towards genetic improvement of sandalwood species has been through genetic diversity studies; identifying restriction fragment length polymorphisms (RFLPs), microsatellite or simple sequence repeat (SSR) polymorphisms and random amplified polymorphic DNAs (RAPDs). By gaining some measure of the level of variation within and between sandalwood populations, researchers have been able to assess potential for improvement. Santalum spicatum, S. austrocaledonicum, S. album and S. insulare have been studied intensively in this regard. Santalum spicatum was assessed for genetic diversity using both nuclear and chloroplast RFLP probes (Byrne et al. 2003a;b) revealing modest levels of genetic variation but few rare alleles and considerable inbreeding coefficients compared to other widespread forest species such as Eucalyptus spp. (Hines and Byrne 2001). Although this particular study used anonymous alleles, these polymorphisms may represent genotypic variation in gene structure and function, which could in turn affect important phenotypic traits like form, growth rate and heartwood oil production. Santalum austrocaledonicum, a sandalwood tree confined to the South Pacific islands of Vanuatu and New Caledonia, was recently examined for population structure and genetic diversity using microsatellite markers (Bottin et al. 2005). In contrast to S. spicatum, S. austrocaledonicum has a very limited distribution, bound by ocean in all directions, significantly restricting gene flow and germplasm dispersal. This probably explains the high mean fixation index, sometimes referred to as the inbreeding coefficient

Fis, compared to S. spicatum. Based on this comparison, one may predict that there is considerable scope for improvement in S. spicatum compared to S. austrocaledonicum, however caution must be exercised when comparing different techniques and unknown alleles. Whilst microsatellites and RFLPs represent co-dominant genetic markers, meaning data can be pooled for most forms of analyses, the analytical methods are fundamentally

10 different. Often the detected polymorphisms may not be directly related to functional genes. Santalum insulare, a sandalwood growing throughout the French Polynesia was studied intensively by chloroplast microsatellite polymorphisms (Butaud et al. 2005). As populations of S. insulare are restricted to islands, gene flow between populations was particularly low. About one-third (36%) of the total genetic diversity was between archipelagos, 31% between islands within a given archipelago, and 33% within islands. Further studies on the same populations found that clonality was common, and that up to 58% of the trees examined were likely to be clones of another individual (Lhuiller et al. 2006). Clonal reproduction in nature is common for many species of sandalwood. Santalum lanceolatum in Victoria, south-eastern Australia, for example, is represented by only one small population of clones (Warburton et al. 2000). The level of genetic variation in S. lanceolatum in its northern distributions (Queensland, Northern Territory and northern Western Australia) is not well understood. Clonal reproduction of S. lanceolatum has been used to advantage, as regrowth from pulled trees appears to result in healthier root suckers (Bristow et al. 2000). Genetic diversity in Santalum album has been studied using isozyme variation (Suma and Balasundaran 2003) and RAPDs (Shashidhara et al. 2003; Suma and Balasundaran 2004). Both these techniques have helped researchers to understand the extent of diversity remaining within the natural distribution of sandalwood in India. Unfortunately, no comprehensive study has included Indonesian material, although a study by RFLPs of S. album growing in northern Western Australia has included some Timorese material for comparison (C. Jones et al., Chapter 3 this thesis). Of all of the work compiled on the genetic diversity of S. album, the consensus seems to be that while little diversity exists within populations, there appears to be significant diversity between natural populations. Suma and Balasundaran (2003) found that 78% of all variation was described by differences between populations, while the remainder was within individual populations. The populations studied were geographically isolated (several hundred kilometers apart) with different climatic zones. Gene flow is hence quite low between populations. Shashidhara and co-workers (2003) found that RAPDs were effective in distinguishing genotypes and estimating relatedness. Levels of genetic dissimilarity were low (15%) between nearby populations, and high (91%) when compared to S. spicatum (the designated out-group). This would be expected as S. spicatum is unique to arid south-western Australia and resides some 6,000 km away. However, when the out-

11 group was excluded from the same analysis, 75% was the highest level of genetic dissimilarity. From these findings one may conclude that the level of genetic diversity for S. album in India is sufficient for a tree improvement program. Traditional breeding methods for woody perennials, most notably open-pollinated seed orchards, are well established for many forest species particularly Eucalyptus (Chaix et al. 2003). Seed orchards have been established for S. album in India and Australia using this principle. The phenotypes of the progeny seed collected from such orchards are difficult to assess due to natural variability in the plantation environment. While RAPD is an accepted means for estimating relatedness, the technique has been shown to overestimate within population variability in the absence of a suitable out-group (Powell et al. 1996). In the Forest Products Commission collection of S. album in the Ord River Irrigation Area, Kununurra, Western Australia, genetic variation estimated by RFLPs was low, but phenotypic variation in oil yield was quite substantial. This may be due to subtle nucleotide sequences undetected by this technique (C. Jones et al., Chapter 3 this thesis). In addition to genetic improvement, improved plantation management has taken the sandalwoods to a new level of productivity. For many years, the seemingly simple process of growing seedlings, planting them in the field and maintaining trees until harvest has been very difficult and often inefficient. It was not until the early 1990s that research into suitable silvicultural practices was conducted, and eventually a fairly reliable production system for S. album was published (Radomiljac 1998). Likewise, S. spicatum recently underwent a similar investigation (Brand 2002; Woodall and Robinson 2002a;b). In both instances, it was essential for a short-term pot host to be grown alongside Santalum sp. seedlings. At the time of planting into the field, a medium-term host must be well- established nearby and a suitable long-term (longer than 5-7 years) host must eventually be present. Sufficient haustorial connections must be made in the early stages of development. Since all of the sandalwoods are root hemi-parasites, early host-parasite relationship is key to their survival. In a recent comprehensive study of the histology and anatomy of S. album haustoria, Tennakoon and Cameron (2006) found that proto-haustoria will develop in young seedlings, which remain unattached to host roots. But the formation of a penetration peg, which taps into a hosts root xylem is not present in these immature organs. Once attached to the host, a peg is formed. Furthermore, the presence of parenchymous tissue in the host-parasite border supports the hypothesis that cross- membrane solute transport is occurring, and hence some solute selectivity could be involved. Most significantly for the plantation industry, if this process is selective in terms

12 of the nutrients and chemical signals that are moved from host to sandalwood tree, the influence of host species and root connectivity is bound significantly to affect tree growth, vigor and essential oil productivity. Not surprisingly, the complex ecology of the sandalwoods makes phenotypic selection difficult. Plant nutrient and water demands are complicated further by host contributions, as well as planting density, pruning regime and light penetration/requirements. All of these factors contribute to the overall health and vigor of the sandalwood tree, making genetic selection fraught with error. Simple oil yield and heartwood content analysis of standing trees have been used as a selection tool (Jones et al. 2007), but the genetic contribution to any variation identified is substantially masked by agronomic practices and variations in environment niches across the plantation. In the last 20 years, tissue culture has provided some useful avenues into large-scale clonal propagation of sandalwood. Somatic embryogenesis (Rai and McComb 2002; Rao and Bapat 1995; Rugkhla and Jones 1998) along with molecular studies on intra- and interspecific hybrids (McComb and Jones 1998; Rugkhla et al. 1997) have been major advances, and trees cloned in this fashion have shown growth characteristics comparable to naturally regenerated trees. The ability to clonally reproduce large numbers of specific genotypes is essential for any future transgenic work. Currently tissue culture technique is proving useful for propagation but the process is expensive and time consuming, and application to genetic transformation is lacking. Unlike the usual applications of tissue culture for plantlet regeneration, Valluri and co- workers (1991) devised a process whereby S. album cell cultures were grown in a bioreactor where phenolic compounds were measured as an indicator of secondary metabolic activity. This rationale ensured the optimum growing abiotic conditions were discovered before further exploitation of the system. Bioreactors offer the benefit of having all necessary genetic machinery operating in a hygienic, controlled environment.

1.4 Scope for Transgenics in Sandalwood Clearly, traditional tree improvement methods can, and should be used to deliver more desirable trees. However, the application of transgenic technology to sandalwood may offer a faster route to the same end. One cannot disregard the value of knock-out or knock- down sandalwood genotypes to improving our understanding of gene structure, transcription and regulation. Further to experimental applications, highly active promoters

13 may be introduced to economically important genes such as terpenoid synthases, prenyltransferases and transcription factors. Unfortunately almost none of the Santalum genome or transcriptome has been sequenced, let alone functionally characterized. A thorough evaluation of all relevant genes is essential before embarking upon the introduction of foreign genes to sandalwood. The long rotation time for sandalwood trees is probably the main driving force to develop transgenic lines. Traditional crossing and selection trials would take at least 40 to 50 years to evaluate, provided a suitable phenotype screen is established (that is, one which accounts for environmental variability). Given the long period between harvests for sandalwood, this may not be of concern. However if genetically superior seedlings could be identified at the nursery stage, or even propagated in vitro, growers could plant sandalwood with certainty of better long term returns.

2 TRANSGENICS DEPLOYED SO FAR AND POTENTIAL OPTIONS FOR THE FUTURE 2.1 Current Status Transgenic sandalwood has only ever been produced in two experiments, both a decade apart. In 1998, a proof of principle experiment by Veena and Rao (1998) inserted the bacterial derived marker genes β-glucuronidase and the selective gene neomycin phosphotransferase II (NPTII) bound by a binary vector into the S. album genome using Agrobacterium tumefaciens. A suite of binary vectors containing the marker genes were trialed; pKIWI105, pBI121, pIG121-Hm. Of these pKIWI105 proved to be the most efficient. Standard assays for GUS and NPTII were used to select transformed S. album embryos and polymerase chain reaction (PCR) with insert-specific primers confirmed the new genes were present in the regenerated seedlings. The vectors, pKIWI 105 and 110 are binary vectors which have been used previously in other plant gene transformations (Janssen and Gardner 1989; Yao et al. 1995). The T- DNA component of the vector inserted into S. album used a nopaline synthase promoter (NOS) to drive expression of neomycin phosphotransferase II gene, enabling kanamycin- based antibiotic selection. The construct also contained a P35S cauliflower mosaic virus promoter to drive expression of the uidA gene (encoding the β-glucuronidase enzyme).

14 GUS expression was restricted only to true positives by utilizing defections in both pKIWI105 and pIG121-Hm. The uidA gene in pKIWI105 lacked a ribosomal binding site while the vector pIG121-Hm contained an intron in the same gene construct. This ensured that GUS expression would only occur under plant transcriptional machinery and not bacterial. Veena and Rao (1998) used PCR to confirm the presence of GUS genes and hence cells were shown to be transgenic, however the shoots emerging from infected embryos were chimeric. Chimeras are a common problem, although repeated sub-cloning may have eventually rectified this. Until recently, this was the only example in the literature of the introduction of foreign genes into a sandalwood species. Just this year, Shekhawat and co-workers (2008) repeated the experiment, only instead of using cotyledon-stage plantlets, embryogenic cell suspension cultures were infected with A. tumefaciens containing the pCAMBIA 1301 vector. Stable, high level GUS protein expression was noted. The procedure relies heavily on tissue culture, and probably for this reason it seems the application of large-scale transgenics to Santalum has not yet taken off. Although not technically a transgenic experiment, there is a growing body of evidence that gene transfer between parasitic plants and their hosts has occurred in the past, resulting in polyphyletic distributions (Davis et al. 2005; Nickrent et al. 2004). While still poorly understood, there may be some merit to the possibility of genetic information being transferred between morphologically and genetically distant angiosperms. If this is the case, then there is considerable potential for transfer of specific genes via the parasitic connections sandalwoods have with their hosts. Further work is clearly needed to investigate this phenomenon.

2.2 Potential Application of Resistance Gene Transfer There are many potential applications for the insertion of foreign genes into sandalwood, not least resistance to sandal spike disease. Spike disease is a major concern for growers in India, where the majority of natural stands remain. Symptoms include yellowing of leaves and in severe cases defoliation, poor form and low fruit set. Efforts to combat the phytoplasma responsible for the disease have been made but follow-up research is lacking (Khan et al. 2004; Lakshmi Sita 1996). If resistance genes were to prove useful against spike disease, there is certainly potential to insert it into the genome of sandalwood. To this end, Lakshmi Sita and Bhattacharya (1998) isolated and cloned a proline-rich protein

15 (PRP) cDNA from S. album leaves. Proline-rich proteins serve as biopolymers involved in maintenance of cell wall integrity, and may be involved in systemic resistance to pathogens. At present, this appears to be the only functional genetic analysis of potential resistance mechanisms in sandalwood. Further work in this area is certainly needed.

2.3 Potential Improvement by Metabolic Engineering Sandalwood is highly prized for essential oils contained within the heartwood. The oils consist of the sesquiterpene alcohols α-santalol, β-santalol, α-trans-bergamotol and epi-β- santalol. These compounds make up nearly 90% of the distilled oil (Verghese et al. 1990). These appear to be the 12-hydroxy products of the olefin precursors, α- and β-santalene, α- bergamotene and epi-β-santalene. Other minor compounds make up the remainder of the extracted oil. The biosynthesis of these oils appears to occur in the transition zone between sapwood and heartwood (Jones et al. 2006). Terpene biosynthesis begins with the 5 carbon unit of isopentyl diphosphate (IPP). This compound is synthesized from either the mevalonic acid (MVA) pathway, localized in the cytoplasm of cells (Dewick 2002; Gershenzon and Croteau 1990), or it can be derived from plastids via the deoxy-D-xyluose 5-phosphate (DXP) pathway (Arigoni et al. 1997; Rohmer 1999). Condensation with the isomer of IPP, dimethylallyl diphosphate (DMAPP) leads to the universal 10 carbon monoterpene precursor geranyl diphosphate (GPP). Further condensation with DMAPP leads to the 15 carbon farnesyl diphosphate (FPP) while the diterpene precursor geranyl geranyl diphosphate (GGPP) is derived from a third condensation with DMAPP (Gershenzon and Croteau 1990). Terpene synthases (TPS) are responsible for converting the linear precursors GPP, FPP and GGPP into cyclic compounds. Scores of TPS have been isolated from a vast array of plant species and most have been functionally characterized by either heterologous expression in Escherichia coli or from cell free extractions. The TPS gene family is large, with high homology among taxonomic groups (Bohlmann et al. 1998). Many of these genes encode proteins which catalyze the formation of multiple products (Munck and Croteau 1990). Site directed mutagenesis studies by Hyatt and Croteau (2005) found that only very subtle amino acid changes can have a large impact on the terpenoids produced. Sandalwood oil biosynthesis has only recently received detailed attention. Our research group has found strong correlations between the santalenes and bergamotene, suggesting multiple product formation from a single synthase enzyme in S. album (Jones et al. 2006).

16 This work has been followed up with gene isolation, functional characterization and expression studies (C. Jones et al., Chapter 5 this thesis). The gene structure and expression properties of TPS in sandalwood may be exploited through modification or selection of specific genotypes conducive to high essential oil production. However, before this kind of work can proceed, one must understand the normal pattern of oil formation in wild type sandalwood. Normally, in the case of S. album, fragrant heartwood does not form until around 10 years (Rai 1990) and it continues to expand outwards as the tree matures. Ideally, a tree worthy of selection would yield heartwood earlier, and the ongoing expansion of fragrant heartwood would continue throughout the trees life at a faster rate. At some point in the first few years of tree development a signal induces oil biosynthesis. Interestingly, almost all members of the Santalum genus yield sesquiterpenes; and even the related African sandalwood Osyris tenuifolia (Santalaceae) produces sesquiterpenes in its heartwood. All of these species are root parasites, so it is curious to note the link between root parasitism and secondary metabolite formation in vascular tissues. As detailed by Tennakoon and Cameron (2006) the haustorial connection involves exposure to soil-borne pathogens which may in turn elicit oil production. Oil biosynthesis appears to be localized to ray parenchyma (C. Jones, Chapter 5 this thesis), which seems logical as only ray parenchyma cells are metabolically active. Studies linking host species and haustorial connection to oil yield have not been compiled. Such investigations would surely present a valuable insight to this phenomenon. With respect to transgenic sandalwood, expression of the sesquiterpene metabolic pathway is probably controlled very tightly until a certain age, suggesting over-expression may cost the plant significantly. Terpenoids are highly reduced chemicals which bear a considerable metabolic cost (Gershenzon 1994), so production of sesquiterpenes must be limited until sufficient photosynthetic capacity is attained. A comprehensive knowledge of secondary metabolism pathways and their regulation in sandalwood is required before it can be exploited. Another potential application of transgenics in sandalwood is the knock-out of undesirable genes. For example, S. spicatum attracts a lower value than Indian sandalwood for various reasons, one being the complex mixture of sesquiterpenes in the extracted oil. One of the many sesquiterpene compounds present in distilled oil is farnesol (Howes et al. 2004; Piggot et al. 1997). Farnesol has been implicated as a potential allergen (Hostynek and Magee 1997; Schnuch et al. 2004). If the farnesol synthase gene could be silenced or excised from the genome, the allergenic properties of the distilled oil may be reduced. At

17 present the distilled oil may be rectified by fractional distillation and preparative chromatography, such that farnesol can be isolated from the oil. While the application of farnesol-free transgenic sandalwood may seem rather superfluous, it would at least provide an immediate commercial use of knock out sandalwood lines.

2.4 Potential for Transforming Sandalwood Genes into Other Organisms

Further to transforming sandalwood with foreign genes, sandalwood genes may be inserted into other species. There is growing interest in this technology as the santalene chemical structure is particularly difficult to synthesize (Christenson and Willis 1980; Krotz and Helmchen 1994; Zhong and Schlosser 1994). Transformation of faster growing annual plants with synthase genes derived from Santalum may provide a quicker and easier route to the complex tricyclic santalenes. There is a growing body of research confirming multi- gene transformations, including the recent transformation of caffeine biosynthesis in tobacco (Uefuji et al. 2005). Escherichia coli or yeast can also be engineered to produce medicinally important plant metabolites, as has been proven with the anti-malarial compound artemisin (Chang and Keasling 2006). Sandalwood fragrances may also be produced in the same way, while careful regulation of the cytochrome P 450 responsible for hydroxylation of the santalenes completes the process. Oil yields may be low compared to that of a mature tree however the process can be automated for a continuous supply. New compounds are still being discovered in various sandalwood species (Brunke et al. 1995; Kim et al. 2005). Some chemicals found in sandalwood tissues have shown potential for medicinal benefit (Kim et al. 2006). It may be possible to engineer these compounds into host plants, fungi or bacteria as their natural abundance in sandalwood is too low to be economically feasible.

2.5 Testing of Transgenic Sandalwood If sandalwood were to be transformed with genes for pathogen resistance or increased essential oil production, the product would require rigorous testing. Again to do this, uniform field trials must be established along with non-transformed controls. At present this is particularly difficult given the aforementioned issues with growing parasitic trees. Glasshouse trials would be useful for initial pathogen resistance tests, however the long-

18 term expression patterns of the gene(s) would need to be monitored and confirmed in the field. Heartwood oil yields can be monitored over time using established non-destructive sampling techniques (Jones et al. 2007). As no mapping projects have been undertaken on sandalwood, nor has the genome been extensively sequenced, this will be difficult and time consuming. Nonetheless, if transformation of new genes into sandalwood is successful and the desired outcomes are expressed in the phenotype, growers will want the ability to continue growing these trees from seed. Studies on sandalwood mating systems are somewhat deficient in the literature. Rugkhla and co-workers (1997) found that both S. spicatum and S. album were preferentially outcrossing, however both species were able to self-pollinate. Fruit development and seed set also depended heavily on genotype. In circumstances where clonal propagation has occurred and genotypes are appropriate, self-pollination can occur and produce viable seed. This process must be taken into consideration if transgenic sandalwood species are released. Seed collected from these plants will also require testing to confirm hereditary patterns. Further crosses must be done in order to obtain a relatively pure line, however this will take considerable time given the long rotation cycle of sandalwood plantations. Flowering and fruit production are observed after only a few years, but the transgenic traits may not be revealed for many more years. It may also be necessary to plant the transgenic sandalwood at some distance from natural populations as cross-pollination is likely to occur. This should apply to all genetically improved plantations whether they are transgenic or bred using more traditional methods.

2.6 Regulatory Measures Adopted In Australia, all genetic engineering and biotechnology regulation is covered by the Commonwealth Government Office of Gene Technology Regulator. Legislation is covered extensively in the Gene Technology Act 2001, and all research, development and extension is bound by this act. Release of products created through transgenic technology is covered by other offices; in the case of foodstuffs the Australian and New Zealand Food Safety Authority and for other products, including potential transgenic sandalwood products, the Therapeutic Goods Administrator. Indian transgenic research is also bound by legislation covered by the Environmental Protection Act 1986, and the rules concerning genetic engineering were added in 1989.

19 Several governing authorities are responsible for monitoring the research and release of genetically engineered goods, particularly the Ministry of Environment and Forests, and the Department of Biotechnology. Any transgenic work on sandalwood which is done in either of these sandalwood producing countries is bound by the regulations set out by these authorities.

3 FUTURE ROAD MAP 3.1 Product Expectations While there is certainly potential for transgenic sandalwood or other plants with sandalwood genes to be developed, there is often a large disparity between promised qualities and actual results. Production of transgenic woody plants is not a trivial task as it often relies heavily on tissue culture. Strong selective markers have improved the odds of acquiring positive transformants, but the random location within the genome of an insert could make their regulation more difficult to follow. Therefore, the following approach for transgenic research in all economically and non- economically important sandalwoods should be adopted. A significant and sustained genetic mapping project should be undertaken, along with sequencing as much of the genome as possible. Genetic diversity studies should be followed up with gene flow and mating system studies, such that breeders will have known parameters in mind when performing crosses and developing new seed lines. Expression studies should be initiated concurrently, investigating transcriptional and metabolic responses to the many environmental factors which affect sandalwood growth and development. Functional characterization of as many secondary metabolic pathway genes should also be carried out and correlated with metabolite profiles in vivo. Extensive field trials must be established with all variables controlled, such that systematic processes can be fully explored. Promoter regions and transcription factors for many important pathways should also be considered; the sequences of these regions must be deduced. Once these significant barriers to understanding normal gene expression in the sandalwoods are brought down, transgenic lines can be considered. Trials of different techniques of transformation should be considered. For example gene-gun technology has not yet been trialed in sandalwood, and may prove more efficient than bacterial transformation. If young sandalwood flowers are inoculated with Agrobacterium, and the

20 embryos transformed the need for tissue culture could be avoided (Clough and Bent 1998). Selectable markers like antibiotic resistance could be optimized, and the use of herbicide resistance constructs may also improve selection. Retention of herbicide resistance qualities in regenerated sandalwood plants may be useful, as weed management in plantations often risks killing sandalwood via weed-parasite haustorial connections.

3.2 Addressing Risks and Concerns

Transgenic sandalwood faces the same challenges and uncertainties as every other transgenic organism produced to date. The public have grown wary of transgenic technology and the outrageous promises often claimed. Sandalwood is perceived as a natural product, far superior to any man-made chemical perfume. The same market which desires natural sandalwood will also question any transgenic lines of sandalwood, or even metabolically engineered bacteria as „un-natural‟. Europe is a major producer of perfumes, and one of the largest customers of sandalwood oil. Current bans on genetically engineered products may hamper any efforts to test this technology. No doubt organizations which intend on capitalizing on transgenic sandalwood will have a strong marketing challenge ahead of them. But in addition to marketing, the product must be safe and unlikely to harm consumers, or natural populations of sandalwood. Ultimately, the product must actually work and deliver the desired properties for which it was created. The sandalwoods have a very low weed risk, given the demise of their natural distribution. Hence it is unlikely that transgenic lines will „escape‟. Based on the current status of the technologies, market and industry, transgenic sandalwood is unlikely to attract much commercial attention in the near future. Growing the plant under normal conditions and traditional breeding methods are already challenging enough. Furthermore, as the product is well-established and has been shown to yield fragrant heartwood under plantation conditions, the need for transgenics is still low.

3.3 Expected Technologies Transgenic sandalwood as a final product represents only part of the technology. As researchers seek to understand the genetic control of essential oil production, further

21 applications in other species may become apparent or indeed using ex planta systems. Little work has been done since 1998, suggesting modern techniques could be used to optimize transformation and selection. New T-DNA vectors and genotypes of Agrobacterium have since been developed, and are likely to work far more efficiently than those described by Veena and Rao (1998). Tissue culture-free techniques of transformation should also be considered, particularly flower-dipping inoculation. Santalum inflorescences vary widely, but generally have large numbers of individual flowers per stem. This would save considerable time but may in turn result in fewer transformants. Recognizing the natural promoters and terminators of the sandalwood genome will also help determine the make up of T-DNA constructs. Comparison of TPS genes and their regulatory elements among the Santalum genus will substantially aid our understanding of divergence in oil yield and composition.

References Anonymous (1995) 'Flavours and fragrances of plant origin.' (Food and Agriculture Organization of the United Nations: Rome)

Applegate GB, Chamberlain J, et al. (1990a) A state of knowledge synthesis and summary from the April 1990 symposium. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990'. (Eds L Hamilton and CE Conrad) pp. 1-11. (US Forest Service: Honolulu, Hawaii)

Applegate GB, Davis AGW, Annable PA (1990b) Managing sandalwood for conservation in North Queensland, Australia. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990, Honolulu, Hawaii'. (Eds L Hamilton and CE Conrad) pp. 12-18. (US Forest Service: Honolulu, Hawaii)

Applegate GB, McKinnell FH (1993) The management and conservation status of Santalum species occurring in Australia. In 'Sandalwood in the Pacific Region'. (Ed. FH McKinnell) pp. 5-11. (ACIAR: Canberra)

Arigoni D, Sagner S, Latzel C, Eisenreich W, Adelbert B, Zenk M (1997) Terpenoid biosynthesis from 1-deoxy-D-xyluose in higher plants by intramolecular skeletal rearrangement. Proc. Natl. Acad. Sci. USA 94, 10600-10605.

Bohlmann J, Meyer-Gauen G, Croteau R (1998) Plant terpenoid synthases: Molecular biology and phylogenetic analysis. Proc. Natl. Acad. Sci. USA 95, 4126-4133.

Bottin L, Verhaegen D, Tassin J, Olivieri I, Vaillant A, Bouvet JM (2005) Genetic diversity and population structure of an insular tree, Santalum austrocaledonicum in New Caledonian archipelago. Mol. Ecol. 14, 1979-1989.

22 Brand JE (2002) Review of the influence of Acacia species on establishment of sandalwood (Santalum spicatum) in Western Australia. Conservation Science W. Aust. 4, 125-129.

Bristow M, Taylor D, Robson K (2000) Queensland sandalwood (Santalum lanceolatum): Regeneration following harvesting. Sandalwood Research Newsletter 11, 4-8.

Brunke EJ, Vollhardt J, Schumaus G (1995) Cyclosantalal and epicyclosantalal - new sesquiterpene aldehydes from east Indian sandalwood. Flavour and Fragrance Journal 10, 211-219.

Butaud J-F, Rives F, Verhaegen D, Bouvet J-M (2005) Phylogeography of eastern Polynesian sandalwood (Santalum insulare), an endangered tree species from the Pacific: a study based on chloroplast microsatellites. J. Biogeogr. 32, 1763–1774.

Byrne M, MacDonald B, Brand JE (2003a) Phylogeography and divergence in the chloroplast genome of Western Australian sandalwood (Santalum spicatum). Heredity 91, 389-395.

Byrne M, MacDonald B, Brand JE (2003b) Regional genetic differentiation in Western Australian sandalwood (Santalum spicatum) as revealed by nuclear RFLP analysis. Theor. Appl. Genet. 107, 1208-1214.

Chaix G, Gerber S, Razafimaharo V, Vigneron P, Verhaegen D, Hamon S (2003) Gene flow estimation with microsatellites in a Malagasy seed orchard of Eucalyptus grandis. Theor. Appl. Genet. 107, 705-712.

Chang MCY, Keasling JD (2006) Production of isoprenoid pharmaceuticals by engineered microbes. Nat. Chem. Biol. 2, 674-681.

Christenson PA, Willis BJ (1980) East Indian sandalwood oil. 2. Stereoselective synthesis of (+/-) - epi--santalene and (+/-) - epi--santalol. J. Org. Chem. 45, 3068-3072.

Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. The Plant Journal 16, 735-743.

Davis CC, Anderson WR, Wurdack KJ (2005) Gene transfer from a parasitic flowering plant to a fern. Proceedings of the Royal Society B 272, 2237–2242.

Dewick PM (2002) The biosynthesis of C5-C25 terpenoid compounds. Natural Product Reports 19, 181-222.

Ehrhart Y (1998) Descriptions of some sandal tree populations in the South West Pacific: Consequences for the silviculture of these species and provenances. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 105-112. (ACIAR: Canberra)

Gershenzon J (1994) Metabolic costs of terpenoid accumulation in higher plants. J. Chem. Ecol. 20, 1281-1328.

23 Gershenzon J, Croteau R (1990) Regulation of monoterpene biosynthesis in higher plants. In 'Recent advances in phytochemistry, Vol. 24: Biochemistry of the mevalonic acid pathway to terpenoids.'. (Eds GHN Towers and HA Stafford) pp. 99-160. (Plenium Press: New York)

Harbaugh DT, Baldwin BG (2007) Phylogeny and biogeography of the sandalwoods (Santalum, Santalaceae): Repeated dispersals throughout the Pacific. Am. J. Bot. 64, 1028- 1040.

Hewson HJ, George AS (1984) Santalum. In 'Flora of Australia'. (Australian Government Publishing Service: Canberra)

Hines B, Byrne M (2001) Genetic differentiation between mallee and tree forms in the Eucalyptus loxophleba complex. Heredity 87, 566-572.

Hostynek JJ, Magee PS (1997) Fragrance allergens: Classification and ranking by QSAR. Toxicology In Vitro 11, 377-384.

Howes M-JR, Simmonds MSJ, Kite GC (2004) Evaluation of the quality of sandalwood essential oils by gas chromatography-mass spectrometry. Journal of Chromatography A 1028, 307-312.

Hyatt DC, Croteau R (2005) Mutational analysis of a monoterpene synthase reaction: Altered catalysis through directed mutagenesis of (-)-pinene synthase from Abies grandis. Arch. Biochem. Biophys. 439, 222-233.

Janssen B-J, Gardner RC (1989) Localized transient expression of GUS in leaf discs following cocultivation with Agrobacterium. Plant Molecular Biology 14, 61-72.

Jiko LR (1993) Status and current interest in sandalwood in Fiji. In 'Sandalwood in the Pacific Region'. (Ed. FH McKinnell) pp. 13-18. (ACIAR: Canberra)

Jones CG, Ghisalberti EL, Plummer JA, Barbour EL (2006) Quantitative co-occurrence of sesquiterpenes; a tool for elucidating their biosynthesis in Indian sandalwood, Santalum album. Phytochemistry 67, 2463–2468.

Jones CG, Plummer JA, Barbour EL (2007) Non-destructive sampling of Indian sandalwood (Santalum album L.) for oil content and composition. J. Essent. Oil. Res. 19, 157-164.

Khan JA, Srivastava P, Singh SK (2004) Efficacy of nested-PCR for the detection of phytoplasma causing spike disease of sandal. Current Science 86, 1530-1533.

Kim TH, Ito H, Hatano T, Takayasu J, Tokuda H, Nishino H, Machiguchi T, Yoshida T (2006) New antitumor sesquiterpenoids from Santalum album of Indian origin. Tetrahedron 62, 6981–6989.

Kim TH, Ito H, Hayashi K, Hasegawa T, Machiguchi T, Yoshida T (2005) Aromatic constituents from the heartwood of Santalum album L. Chemical & Pharmaceutical Bulletin 53, 641-644.

Krotz A, Helmchen G (1994) Total syntheses, optical rotations and organoleptic properties of sandalwood fragrances, (Z)- and (E)--santalol and their enantiomers, ent--santalene. Liebigs Ann. Chem. 6, 601-609.

Kulkarni RS, Fakrudin B, Shashidhar KS (1998) Tree improvement efforts in sandal: the need to employ novel strategies. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 151-153. (ACIAR: Canberra)

Lakshmi Sita G (1996) Molecular basis of disease resistance in sandlewood. Denrome 3, 7- 8.

Lakshmi Sita G, Bhattacharya A (1998) cDNA cloning and characterisation of a proline- (or hydroxyproline-) rich protein from Santalum album L. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 29-34. (ACIAR: Canberra)

Lhuiller E, Butaud J-F, Bouvet J-M (2006) Extensive clonality and strong differentiation in the insular Pacific tree Santalum insulare: Implications for its conservation. Annals of Botany 98, 1061–1072.

Loneragan OW (1990) 'Historical Review of Sandalwood (Santalum spicatum) Research in Western Australia.' (Department of Conservation and Land Management: Perth)

McComb JA, Jones MGK (1998) Interspecific hybridisation between Santalum album and S. spicatum. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 36-41. (ACIAR: Canberra)

Moretta P, Ghisalberti EL, Piggot MJ, Trengove RD (1998) Extraction of oil from Santalum spicatum by supercritical fluid extraction. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 83-85. (ACIAR: Canberra)

Munck SL, Croteau R (1990) Purification and characterization of the sesquiterpene cyclase patchoulol synthase from Pogostemon cablin. Arch. Biochem. Biophys. 282, 58-64.

Nickrent DL, Blarer A, Qiu Y-L, Vidal-Russell R, Anderson FE (2004) Phylogenetic inference in Rafflesiales: the influence of rate heterogeneity and horizontal gene transfer. BMC Evolutionary Biology 4, 40.

Piggot MJ, Ghisalberti EL, Trengove RD (1997) Western Australian sandalwood oil: extraction by different techniques and variations of the major components in different sections of a single tree. Flavour and Fragrance Journal 12, 43-46.

Powell W, Morgante M, Andre C, Hanafey M, Vogel J, Tingey S, Rafalski A (1996) The comparison of RFLP, RAPD, AFLP and SSR (microsatellite) markers for germplasm analysis. Mol. Breeding 2, 225-238.

Radomiljac A (1998) The influence of pot host species, seedling age and supplementary nursery nutrition on Santalum album L. (Indian sandalwood) plantation establishment

25 within the Ord River Irrigation Area, Western Australia. Forest Ecology and Management 102, 193-201.

Radomiljac A, McComb JA (1998) Nitrogen-fixing and non-nitrogen-fixing woody host influences on the growth of the root hemi-parasite Santalum album L. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 54-57. (ACIAR: Canberra)

Radomiljac A, McComb JA, Pate JS (1999) Gas exchange and water relations of the root hemi-parasite Santalum album L. in association with legume and non-legume hosts. Annals of Botany (London) 83, 215-224.

Rai SN (1990) Status and cultivation of sandalwood in India. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990,'. (Eds L Hamilton and CE Conrad) pp. 66-71. (US Forest Service: Honolulu, Hawaii)

Rai SN, Sarma CR (1990) Depleting sandalwood production and rising prices. The Indian Forester 116, 348-355.

Rai VR, McComb JA (2002) Direct somatic embryogenesis from mature embryos of sandalwood. Plant Cell, Tissue and Organ Culture 69, 65-70.

Rao PS, Bapat VA (1995) Somatic embryogenesis in sandalwood Santalum album L. In 'Somatic embryogenesis in woody plants'. (Eds S Jain, P Gupta and R Newton) pp. 153- 170. (Klewer Academic Publishers: Dordrecht)

Rohmer M (1999) The discovery of a mevalonate-independent pathway for isoprenoid biosynthesis in bacteria, algae and higher plants. Natural Product Reports 16, 565-574.

Rugkhla A, Jones MGK (1998) Somatic embryogenesis and plantlet formation in Santalum album and S. spicatum. J. Exp. Bot. 49, 563–571.

Rugkhla A, McComb JA, Jones MGK (1997) Intra- and interspecific pollination of Santalum spicatum and S. album. Aust. J. Bot. 45, 1083-1095.

Schnuch A, Uter W, Geier J, Lessmann H, Frosch PJ (2004) Contact allergy to farnesol in 2021 consecutively patch tested patients. Results of the IVDK. Contact Dermatitis 50, 117- 121.

Shankaranarayana KH, Parthasarathi K (1987) On the content and composition of oil from heartwood at different levels in sandal. Indian perfumer 31, 211-214.

Shankaranarayana KH, Ravikumar G, Rajeevalochan AN, Theagarajan KS, Rangswami CR (1998) Content and composition of oil from the central and transition zones of the sandalwood disc. In 'Sandal and its Products'. (Eds A Radomiljac, HS Ananthapadmanabho, RM Welbourn and KS Rao) pp. 86-88. (ACIAR: Canberra)

Shashidhara G, Hema MV, Koshy B, Farooqi AA (2003) Assessment of genetic diversity and identification of core collection in sandalwood germplasm using RAPDs. Journal of Horticultural Science & Biotechnology 78, 528-536.

Shekhawat UKS, Ganapathi TR, Srinivas L, Bapat VA, Rathore TS (2008) Agrobacterium- mediated genetic transformation of embryogenic cell suspension cultures of Santalum album L. Plant Cell Tiss Organ Cult 92, 261-271.

Statham P (1990) The sandalwood industry in Australia. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990'. (Eds L Hamilton and CE Conrad) pp. 26- 38. (US Forest Service: Honolulu, Hawaii)

Suma TB, Balasundaran M (2003) Isozyme variation in five provenances of Santalum album in India. Aust. J. Bot. 51, 243-249.

Suma TB, Balasundaran M (2004) Genetic diversity of eight Santalum album L. provenances of India based on RAPD analysis. Indian Journal of Genetics 64, 167-168.

Suriamihardja H, Suriamihardja S (1993) Sandalwood in Nusa Tenggara Timur. In 'Sandalwood in the Pacific Region'. (Ed. FH McKinnell) pp. 39-43. (ACIAR: Honolulu, Hawaii)

Tate H, Sethy M, Tungon J (2006) Grafting of sandalwood in Vanuatu. Sandalwood Research Newsletter 21, 7.

Tennakoon KU, Cameron DD (2006) The anatomy of Santalum album L. (sandalwood) haustorium. Canadian Journal of Botany 84, 1608-1616.

Uefuji H, Tatsumi Y, Morimoto M, Kaothien-Nakayama P, Ogita S, Sano H (2005) Caffeine production in tobacco plants by simultaneous expression of three coffee N- methyltrasferases and its potential as a pest repellant. Plant Molecular Biology 59, 221- 227.

Valluri JV, Treat WJ, Soltes EJ (1991) Bioreactor culture of heterotrophic sandalwood (Santalum album L.) cell suspensions utilizing a cell-lift impeller. Plant Cell Reports 10, 366-370.

Veena S, Rao KS (1998) Introduction and expression of marker genes in sandalwood (Santalum album L.) following Agrobacterium-mediated transformation. Plant Science 131, 53-63.

Verghese J, Sunny TP, Balakrishnan KV (1990) (Z)-(+)--santalol and (Z)-(-)--santalol concentration, a new quality determinant of east Indian sandalwood oil. Flavour and Fragrance Journal 5, 223-226.

Warburton CL, James EA, Fripp YJ, Trueman SJ, Wallace HM (2000) Clonality and sexual reproductive failure in remnant populations of Santalum lanceolatum (Santalaceae). Biological Conservation 96, 45-54.

Woodall GS, Robinson CJ (2002a) Direct seeding of Acacias of different form and function as hosts for sandalwood (Santalum spicatum). Conservation Science W. Aust. 4, 130-134.

27 Woodall GS, Robinson CJ (2002b) Same day plantation establishment of the root hemiparasite sandalwood (Santalum spicatum (R.Br.) A D.C.:Santalaceae) and hosts. Journal of the Royal Society of Western Australia 85, 37-42.

Yao J-L, Cohen D, Atkinson R, Richardson K, Morris B (1995) Regeneration of transgenic plants from the commercial apple cultivar Royal Gala. Plant Cell Reports 14, 407-412.

Zhong GF, Schlosser M (1994) A short and simple access to both enantiomers of epi-- santalene and (Z)-epi--santalol. Synlett 3, 173-174.

28 Chapter 3: Genetic diversity of an Australian tropical sandalwood (Santalum album) collection – biogeographic origins and tree improvement potential

This chapter determines the level of genetic diversity in the FPC Kununurra S. album collection. Once genotypes were allocated to groups of trees, comparisons of heartwood and essential oil yield could be made. Factors which influence genetic diversity and the implications for tree improvement are considered.

This manuscript has been submitted to Tree Genetics and Genomes (March 2008).

Genetic diversity of an Australian tropical sandalwood

(Santalum album) collection – biogeographic considerations tree

improvement potential

Christopher G Jones1*, Julie A Plummer1, Elizabeth L Barbour1,2, Margaret Byrne1,3

1. School of Plant Biology, University of Western Australia, 35 Stirling Hwy Crawley WA 6009 Australia. Phone: +61 8 6488 1992 Email: [email protected] 2. Forest Products Commission of Western Australia, 117 Great Eastern Highway Rivervale WA, 6103, Australia 3. Western Australian Department of Environment and Conservation, Science Division. Locked Bag 104, Bentley Delivery Centre, WA 6983, Australia

* To whom all correspondence should be addressed.

30 Abstract

Nuclear RFLP analysis of the Australian Santalum album core germplasm collection revealed surprisingly low levels of genetic diversity for a tree species.

Heterozygosity levels in the collection were very low (Ho and He= 0.047). Nineteen groups were identified within the 233 S. album genotypes sampled. Only one tree originating from

Timor was differentiated from Indian material, with the remaining Timorese germplasm sharing the same genotypes as seed collected from India. Chloroplast RFLP analysis revealed no genetic distinction between S. album from Timor and India, which suggests that S. album has recently been dispersed to India from Timor. Santalum album collected from a small population in the Northern Territory was also undifferentiated by chloroplast

DNA analysis. A hypothesis concerning the evolution of Santalum album is proposed.

31 Introduction

Sandalwood, Santalum album (Santalaceae) is a small root-parasitic tree renowned for its fragrant heartwood. The ground wood can contain as much as 9% essential oil

(Jones et al. 2006) and is used in incense and joss stick manufacture, while the oil is used as a fixative in many high-end perfumes. Santalum album has been widely cultivated throughout much of south-east Asia for centuries. It is reported to be native to the Banda

Sea islands of Flores, Sumba and Timor (McKinnell 1990) and occurs widely throughout

India (Rai 1990). A small population of S. album also grows on the north coast of the

Northern Territory (NT) of Australia (McKinnell 1990). Because of a long history of human intervention, the precise evolutionary origin of S. album is somewhat unsure (Rai

1990). The biogeographic origins of the Pacific and eastern Australian Santalum species have been discussed in a recent phylogenetic study (Harbaugh and Baldwin 2007). The authors conclude that Polynesian species of Santalum were the result of long-range dispersal events out of eastern Australia. Unfortunately no Indonesian S. album was included in their study, so sound conclusions on the origins of the species could not be made. Uncertainty over the origin of S. album remains, however it seems unlikely that S. album evolved on the Indian continent in isolation from the other closely related members of the genus, which clearly show an Australian origin.

The high value and relative scarcity of sandalwood has lead to sustained pressure on natural stands in India and Timor, threatening their survival. In an effort to reduce wild- harvest demand and develop a local industry, commercial plantations have been established in Australia in far northern Western Australia, the Northern Territory and Queensland. The

Forest Products Commission of Western Australia (FPC) manages ongoing field trials of sandalwood in the Ord River Irrigation Area, Kununurra, WA. The collection consists mostly of S. album and a wide variety of host species, but specimens of other commercially

32 significant sandalwoods, S. austrocaledonicum (New Caledonia and Vanuatu) and S. macgregorii (Papua New Guinea) are also present in some trials. In addition to optimising silvicultural practice for this relatively new crop, the plantation serves as an arboretum; capturing germplasm from many seed collections. Subsequent plantations have shown high growth rates and improved seedling survival but have variable heartwood content and essential oil yield (Jones et al. 2006). Addressing heartwood oil content is a priority, since the economic value of sandalwood is governed mostly by heartwood content, with the best quality logs exceeding A$100,000 per ton (WesCorp Pty. Ltd, Pers. Comm).

Santalum album growing in the FPC Kununurra collection originates from families and populations across India, with a few accessions from Timor and the Northern Territory of Australia. It is likely that some component of the observed phenotypic variation in heartwood content and essential oil yield may stem from genetic variability. The collection was compiled by various groups over many years, and the precise origins of some material is uncertain. Effective breeding and selection for commercially important traits requires a good knowledge of the genetic diversity of the core collection. This study sought to address these issues.

DNA based molecular markers are an effective means of determining genetic relatedness and provide estimates of genetic diversity (Parker et al. 1998). DNA mutations in the nucleus occur relatively frequently compared to chloroplast or mitochondrial DNA, and are subject to recombination during meiosis. This makes nuclear DNA markers, particularly RFLPs and microsatellites or short sequence repeats (SSRs) valuable tools for studying mating systems and population structure, whereas chloroplast DNA is best suited to studying ancient lineages (Newton et al. 1999). At the time of commencing this study,

RFLPs were the only nuclear DNA markers available, although SSR primers have since been developed for S. austrocaledonicum (Bottin et al. 2005a). As the collection contains

33 S. album accessions from several geographical regions, chloroplast RFLPs were employed to identify ancestral lineages and determine whether Indonesian S. album is genetically distinct from Indian material. These results are discussed in light of what is known on the biogeography of the region throughout late Neogene period.

Mature trees (>15 yr) were also sampled for heartwood content and essential oil yield using methods described previously (Jones et al. 2006). Genotypic clades and heartwood variability were compared to determine whether there was any association between heartwood oil yield and genotypic groups, which may represent common genetic or geographic origins. Trees from a particular genetic group which show promising heartwood traits may prove useful as initial breeding populations or clonal lines.

Materials and Methods

Collection of plant material

Santalum album was sampled in spring 2004. Fully expanded leaves were collected, placed in sealed plastic bags and kept refrigerated at 4°C until returned to the laboratory where they were stored at -80°C. Collections were made from 233 S. album trees within the collection. The vast majority of the sampled trees were derived from an Indian core collection, while seven of the sampled trees were grown from seed which originated from

Timor. One of the sampled S. album trees was derived from a Northern Territory population. Santalum austrocaledonicum (New Caledonia and / or Vanuatu) and S. macgregorii (Papua New Guinea) were also growing in the arboretum. Five trees from each species were included in the study for comparison. Santalum spicatum (south-western

Western Australia) was also included as a more distant out-group. DNA was extracted from leaf material as described in Byrne et al. (2003b).

34 RFLP analysis

Variation in DNA was assayed using RFLP markers developed for S. spicatum

(Byrne et al. 2003b). DNA was digested with BglII and EcoRV, subjected to agarose gel electrophoresis and transferred to membranes by Southern blotting (Southern 1975).

Hybridisation was carried out with 21 RFLP probes (s3B, s9B, s38B, s41E, s43E, s44B, s47E, s51E, s57B, s59B, s62E, s64B, s66E, s75B, s77E, s80E, s83E s84B, s90B, s93E and s95E) which were derived from a genomic library of S. spicatum (Byrne et al. 2003b). The probes were hybridised to each filter and visualised using 32P dCTP and autoradiography.

The B or E annotation indicates which enzyme (BglII or EcoRV) was used in combination with each probe.

Chloroplast DNA analysis of Santalum spp.

Genetic differences between Indian, Indonesian and Northern Territory S. album, as well as phylogenetic relationships between S. macgregorii, S. austrocaledonicum and S. spicatum were assessed using chloroplast probes from tobacco (Shinozaki et al. 1986;

Sugiura et al. 1986) and Petunia hybrida (Systma and Gottlieb 1986). Genomic DNA was digested with three more restriction enzymes (BclI, Dra and Xba) and Southern blotted using the method described above. These filters were combined with the filters previously made using BglII and EcoRV, and hybridised with radio-labelled chloroplast probes; p1, p4, p3, p6, p8, p10 from Petunia and PtBa1 from tobacco as per Byrne et al. (2003a), however probe PtBa1 was only used in combination with filters using EcoRV and BglII.

Filters were autoradiographed for up to a week at -80°C depending on the strength of the radio-labelled probe.

35 Statistical analyses

Nuclear RFLP banding patterns were interpreted based on the Mendelian multi- allelic model. As the collection did not contain populations of trees, rather a large collection of individuals, Nei‟s genetic distance and F statistics could not be reliably determined. Each species in the study was treated as a single population. Genetic diversity parameters He and Ho, and genetic distance matrices were calculated using GenAlEx version 6 (Peakall and Smouse 2001). Santalum macgregorii consistently showed four bands for most nuclear probes trialled, indicating it was tetraploid. Owing to the difficulty in analysing a combination of diploids and tetraploids, S. macgregorii was not included in the nuclear RFLP analysis. Phylogenetic relationships were inferred using the Phylip package (Felsenstien 1993). A phylogenetic tree was constructed using NEIGHBOUR and unweighted paired genetic mean analysis (UPGMA). Principal component analysis was performed using GenAlEx, treating each species as a population. Chloroplast RFLP data were scored for absence or presence of specific mutations. Haplotypes were determined based on these mutations and a phylogenetic tree was inferred from the haplotype data using PAUP v4.0. Branch support strength was determined by bootstrapping 10000 times.

Heartwood oil content and heartwood proportion

Wood cores from 30 cm above ground level were sampled from trees aged 15-17 years using methods described previously (Jones et al. 2006; Jones et al. 2007).

Results

Nuclear RFLP analysis

Of the 21 nuclear probes trialled, 20 were able to be scored as co-dominant markers using classic Mendelian principles. Probe s83E was not scored as it revealed multiple

36 bands per lane, indicative of multiple loci. Only three loci were polymorphic in the entire

S. album collection; these yielded two (s59B and s38B) or three (s95E) alleles per locus.

Santalum austrocaledonicum was considerably more polymorphic, with 7 loci being variable across the 5 individuals studied. Of the polymorphic loci, s47E and s66E had three alleles, with the remainder having only two. Santalum spicatum was the most variable of the species, with variation detected across all probes. The level of polymorphic loci for the three species were 25 %, 45 % and 65 % for S. album, S. austrocaledonicum and S. spicatum respectively. Santalum macgregorii was not included in the nuclear study due to tetraploid banding patterns, however considerable diversity of alleles was seen across the 15 specimens.

Table 3.1. Mean heterozygosity statistics of Santalum spp. from the Kununurra arboretum compared with those of natural populations.

Species n Method Ho He Reference S. album (arboretum) 233 RFLP 0.047 0.047 This study S. austrocaledonicum (arboretum) 5 RFLP 0.140 0.137 This study S. spicatum (selected individuals) 5 RFLP 0.30 0.21 This study S. album (natural) 100 Isozyme 0.126a 0.08a (Suma and Balasundaran 2003) S. austrocaledonicum (natural) 431 SSR 0.45 0.66 (Bottin et al. 2005b) S. spicatum (23 natural pops.) 100 RFLP 0.201 0.210 (Byrne et al. 2003b) a = arithmetic mean of 5 population values

Variation within S. album was very low, with only 19 different genotypes present in the 233 individuals sampled. In comparison, the five individuals sampled from S. austrocaledonicum and S. spicatum were all represented by different genotypes. Analysis of genetic relationships by UPGMA dendogram showed genotypes clustering into groups representing the three species (Fig. 3.1). This is also evident in principal component analysis (Fig. 2). Some of the sampled trees grown from seed sourced form Timor were

37 not resolved from the rest of the Indian material. Of the seven Timorese trees, only one

(genotype 19) was clearly distinguished from the rest of the group.

Figure 3.1. UPGMA phenogram of the 19 genotypes of S. album and five genotypes of S. austrocaledonicum identified in trees from the FPC Kununurra collection. Number of trees within each group shown in parenthesis. Groups containing individuals sourced from Timor and the Northern Territory (NT) are also indicated. All other seed was sourced from India. Santalum spicatum serves as a distant out group, with five individuals from northern and southern populations of Western Australia.

Figure 3.2. Principal component analysis of the 19 S. album genotypes and related Santalum species. S. album genotypes 18 and 19 are discussed in the text. Genotype 2 is highlighted for reference.

Chloroplast RFLP analysis

Analysis of the banding patterns revealed thirty-eight site and length mutations in the chloroplast genome of the four species used in this study (Table 3.2). These mutations were associated with three chloroplast haplotypes in the 233 S. album individuals The majority of S. album showed a single haplotype. The other two haplotypes were represented by S. album individual 9, and individuals 16 and 158 respectively. Similarly, there was little haplotype diversity in S. macgregorii with all 15 individuals bearing the one haplotype. In contrast, both S. austrocaledonicum and S. spicatum showed greater variation, with three haplotypes present in the five individuals of each species.

Relationships between haplotypes are shown in Figure 3.3, using S. spicatum as an out- group. The haplotypes are clustered according to species, except for one S. album haplotype which clustered with the three S. austrocaledonicum haplotypes. This haplotype was represented by two individuals.

39 Table 3.2. Mutations detected in the chloroplast genomes of 4 Santalum species.

Probe / enzyme Mutation Size (kb) Species / Individuals 1. p1+p4 Bcl Length 11→10 S. spicatum (all) 2. p1+p4 Bcl Length 9.4→9 S. spicatum (Wiluna) 3. p3 BglII Site gain 7 → 5+2 S. album (all) 4. p3 BglII Length 7 → 9 S. spicatum (all) 5. p3 BglII Site gain 3.5 → 2 + 1.5 S. spicatum (all) 6. p3 EcoRV Length 14→13 S. macgregorii 7. p3 Bcl Length 8 → 5.5 S. spicatum (all) 8. p3 Bcl Length 1.5 → 1.6 S. spicatum (all) 9. p3 Dra Length 2 → 1.8 S. album 16, 158 10. p3 Dra Site gain 8 → 5.6 + 2.4 S. macgregorii 11. p3 Dra Length 1.7 → 1.6 S. macgregorii 12. p3 Dra Site gain 7.4→2.9 + 4.5 S. spicatum (all) 13. p3 Dra Length 2 + 0.2 → 2.2 S. spicatum (all) 14. p3 Xba Length 5 → 4 S. austrocaledonicum 42 and 46 15. p6 EcoRV Length 1.1 → 1.2 S. album (all) 16. p6 Bcl Site loss 5 + 4 → 9 S. album 16, 158 17. p6 Bcl Site gain 7 → 4 + 3 S. spicatum (all) 18. p6 Dra Length 4.3 → 5 S. macgregorii 19. p6 Dra Site loss 4.5 + 1 → 5.5 S. spicatum (all) 20. p8 BglII Site loss 5 + 1 → 6 S. macgregorii 21. p8 EcoRV Length 9.5 → 9.3 S. macgregorii 22. p8 EcoRV Length 9.5 → 8 S. spicatum (Nyabing) 23. p8 Bcl Site gain 8 → 5 + 3 S. spicatum (Kokerbin, Dumbelyung, Hyden) 24. p8 Dra Length 1.4 → 1.2 S. austrocaledonicum (all) 25. p8 Dra Length 1.4 → 1.35 S. album 16, 158 26. p10 Xba Length 16 → 12 S. album 27. p8 Dra Site gain 10 → 3 + 7 S. spicatum (all) 28. p8 Xba Site loss 6 + 5 → 11 S. album 29. p8 Xba Length 5 → 4.8 S. spicatum 30. p8 Xba Site loss 1 + 0.8 → 1.8 S. album 16, 158 31. p8 Dra Length 1.4 → 1 S. album 9 32. pTBa1 BglII Length 6 → 6.4 S. album 33. pTBa1 BglII Site loss 4 + 2 → 6 S. austrocaledonicum 45 34. pTBa1 BglII Length 6.4 → 6.3 S. spicatum 35. pTBa1 EcoRV Length 3.5 → 6 S. macgregorii 36. pTBa1 EcoRV Site loss 8 + 4 → 12 S. album 37. pTBa1 EcoRV Length 3 → 4 S. austrocaledonicum 38. pTBa1 EcoRV Site loss 3 + 8 → 11 S. macgregorii

Bootstrap values indicate strong support for the separation of S. album haplotypes

(98%), and the S. austrocaledonicum clade (96%). Strong support is also found for S. spicatum (100%). The relationship between S. album, S. austrocaledonicum and S. macgregorii shows a polytomy, and is unresolved in this analysis.

Evaluation of heartwood characteristics

Trees aged between 15 and 17 years which had been analysed previously for heartwood content and oil yield (Jones et al. 2006) were grouped and numbered according to their genotype as determined by nuclear RFLPs (Figs 3.1 and 3.2). Due to the limited number of trees per group, errors were large. These trees represent collections from India as no Timorese trees were older than 8 years. Thus no useful comparison of heartwood and oil characteristics between Indian and Timorese trees could be made. There was no significant difference in heartwood and oil yield between genetic groups, although group 6 showed the greatest consistency (Fig 3.4).

42 Discussion

The level of genetic diversity of sampled S. album in the FPC Kununurra collection was remarkably low for a tree species, and considerably lower than previously reported figures for natural stands of S. album in India (Suma and Balasundaran 2003). Nuclear

RFLP analysis showed genetic differentiation of one Timorese tree, with all other Timorese germplasm undifferentiated from material collected from India. Differentiation was not seen using chloroplast RFLP analysis either, with Timorese, Indian and Northern Territory

S. album trees bearing identical haplotypes. A selection of five S. austrocaledonicum and five S. spicatum individuals had higher levels of nuclear and chloroplast DNA variation than the entire S. album collection. In comparison, S. macgregorii had no chloroplast variation whatsoever, yet possessed a highly variable tetraploid genome. These findings have important implications for the conservation, selection and deployment of germplasm from the FPC Kununurra collection. They also provide some additional evidence in support of the theory that S. album was probably the result of dispersal to Timor from

Australia, and has since been carried to India in more recent times.

Genetic diversity of S. album collection

The S. album germplasm captured in the Kununurra collection is highly homozygous (Ho = 0.047, He = 0.047, SE ± 0.026) compared to other members of the genus

(Table 3.1). While Suma and Balasundaran (2003) found that isozyme variation in S. album growing naturally in India was more diverse than that seen in the FPC Kununurra collection, the He and Ho values quoted are still rather low for a tree species (Austerlitz et al. 2000). The isozyme technique used by Suma and Balasundaran is generally less likely to reveal polymorphisms than RFLPs (Aldrich and Doebley 1992). Indeed, comparisons in eucalypts show that RFLPs generally detect twice the level of variation compared to

43 isozymes (Byrne 2008). The low variation detected in the S. album collection would suggest that either the RFLPs analysed were not particularly polymorphic, or more likely, the FPC collection is not representative of the species diversity seen in India and possibly

Indonesia. Few S. album trees possessed rare alleles. Genotypes 18 and 19 were differentiated from the other S. album trees by two rare alleles at loci s90B and s64B.

Genotype 19 is a tree grown from seed collected in Timor, while genotype 18 is a younger tree grown from seed collected on site at Kununurra. No record of its parent exists. It is likely that genotype 18 is progeny of 19, however several more polymorphic loci would be needed for reliable parental testing.

A molecular marker which detects higher levels of polymorphism may prove more appropriate, given the limited number of polymorphic loci revealed by RFLP. The probes used in this study brought up sufficient variation in the species from which the library was constructed (S. spicatum) (Byrne et al. 2003b) with over 60 % of loci being polymorphic in

S. spicatum. This was not the case for S. album. Microsatellite loci have been shown to give estimates of heterozygosity nearly twice that of RFLPs in soybean (Powell et al. 1996) and in eucalypts (Byrne 2008) and may prove to be considerably more variable than RFLPs in S. album.

Low genetic diversity in the S. album collection is also reflected in the highly consistent essential oil profiles (Jones et al. 2006). One hundred trees were analysed for essential oil composition, but no significant variation in the proportions of each chemical was detected. Chemical variation is common within plant genera, for example Juniperus

(Adams 1994), Eucalyptus (Li et al. 1995) and Toona sp. (Maia et al. 2000). Chemical variation within a single species also occurs but this may be more common in divergent populations, such as has been found in Chamelaucium uncinatum (Egerton-Warburton et al. 1998), Artemisia herba-alba (Segal et al. 1987) and to a lesser extent Cannabis sativa

44 (Hillig 2004). Bottin and co-workers (2007) found low chemical diversity within S. austrocaledonicum essential oil, however a major component of the oil, Z-lanceol was quite variable. Some island populations were quite abundant in Z-lanceol, while others were not. Terpenoids are effectively metabolic dead-ends, with limited evidence of catabolism (Wildung and Croteau 2003). This means their composition would vary little under seasonal fluctuations, and they may be used reliably as operational taxonomic units

(Sneath and Sokal 1973).

Low heterozygosity in the FPC Kununurra collection may not be entirely due to incomplete seed sourcing. If a plant population is subjected to significant gene flow restriction (i.e. reduction of the effective breeding population) and nearest-neighbour pollination is common, homozygosity increases rapidly over subsequent generations

(Turner et al. 1982). Bottlenecks can occur in nature through climatic or geographical changes, or through founder events. Aridity or glacial encroachment, or oceanic inundation, can result in vicariance leading to small refuge populations with limited gene flow. Limited dispersal events would result in similar circumstances, with very small populations being established (Michaux 2001). Several tree species in natural systems exhibit low genetic diversity, which may be attributed to either of these circumstances.

Acacia mangium, which is native to Papua New Guinea, the islands of Aru, Ceram and

Sula, Sidei and far north Queensland has low genetic diversity both within populations and as a species (Butcher et al. 1998). Larger populations which were closer to surrounding populations had higher heterozygosity levels than smaller, more geographically isolated populations. Likewise, refugial populations may be created through climate changes, resulting in a significantly narrowed gene-pool with which to mate. Species such as western red cedar, Thuja plicata are thought to have exceptionally low heterozygosity

(Ht=0.07) due to recent glacial encroachment and retreat, combined with limited seed

45 dispersal characteristics (Glaubitz et al. 2000). In the most extreme cases, founder populations from very recent dispersal events, or severe population decline and bottlenecks can result in no detectable genetic variation (Peakall et al. 2003; Waters and Schaal 1991).

Given what is now known about the chloroplast and ribosomal genetic diversity of the Santalum genus (Harbaugh and Baldwin 2007), and the results presented here, it is hypothesised that S. album in Timor was founded by a very narrow genetic base and has experienced limited gene-flow since establishment. It is possible that S. album evolved in northern Australia from a Santalum progenitor and was subsequently dispersed to the island of Timor soon after its formation 3 to 5 million years ago (Anonymous 2003).

Climatic oscillations during the Pleistocene ice ages have lead to significant fluctuations in sea levels in this region (Outlaw and Voelker 2007). This would lead to variable extents of land connectivity, and hence gene flow between populations in northern Australia and

Timor. Onset of arid conditions may have led to the extirpation of the ancestral form in northern Australia. Santalum album in Northern Territory may represent a relict population but is more likely to be the result of reintroduction from Indian seed. More extensive sampling from the northern coastline of Australia would clarify this. The lack of differentiation between Timorese and Indian germplasm strongly supports the hypothesis that S. album was introduced to India in very recent geological time, possibly through human mediation. Santalum album is a preferential out-crossing tree which is capable of self-pollination (Rugkhla et al. 1997). It is also capable of clonal reproduction through root suckering, especially after disturbance (e.g. upon harvest or after a fire event). There is a high probability of self-pollination and hence increased homozygosity if all neighbouring trees are clones. This may well have been accentuated through excessive harvest by humans.

46 Santalum album trees 16 and 158 had chloroplast haplotypes which clustered with

S. austrocaledonicum (Fig 3.3). As the same RFLP membranes were used with chloroplast probes, sample mix-up is unlikely, indicating a chloroplast capture event has occurred, which can result after crosses between related species. The phenomena is commonly detected in broad phylogenetic studies where the ancestral origins are the subject of interest

(Rieseberg et al. 1990; Smith and Sytsma 1990; Soltis et al. 1991; Wolfe and Elisens

1995). Some seed growing in the FPC Kununurra collection is derived from arboreta in

India, and it is likely that a cross between S. austrocaledonicum and S. album has occurred within these arboreta. No information on the origins or layout of these arboreta is available. Trees 16 and 158 are unlikely to be F1 hybrids derived from crosses within the

FPC trials since they clustered with other S. album genotypes and were not intermediate with S. austrocaledonicum in the nuclear RFLP study. However, at some point the S. austrocaledonicum genome has been diluted enough to not be detected by RFLP.

Selection and breeding

Plant breeding is the process of building an improved population from a selection of superior phenotypes while removing undesired genes from the population. In doing this, genetic diversity is narrowed and homozygosity increases. Breeders aim to minimise this trend by starting with a diverse breeding population and maintaining a modest selection intensity (i.e. selecting a large group from which to breed the next generation). However, superior trees may still be selected even if overall genetic diversity is low, as is the case with S. album. Improvements have been achieved in Eucalyptus cladocalyx, where in provenance trials some populations expressed good growth characteristics despite low levels of heterozygosity at both a population level and a species level (McDonald et al.

47 2003). The broad genetic groups of S. album identified in this study may serve as initial breeding populations.

Variation present in phenotypic traits such as heartwood content and essential oil yield indicates selection potential in the FPC Kununurra collection. In comparing the genotypes as determined by RFLPs, there was little differentiation in heartwood content and essential oil yield. That is, groups of trees which share genotypes (suggesting a common genetic history, either geographical or familial) did not express differing heartwood characteristics. These comparisons are ambiguous given the wide variety of ages represented in the collection and the high phenotypic variation within single groups, suggesting the growing environment is playing a significant role in phenotypic variability.

The extent to which the growing environment contributes to phenotype must therefore be accurately determined. All species in the Santalum genus and almost all members of the family Santalaceae are root hemi-parasites (Hewson and George 1984).

Even if identical soil and climatic conditions are chosen, the varying levels of haustorial connectivity will influence access to water and nutrients. Sandalwood trees in the FPC

Kununurra collection are growing on a wide variety of host species at different planting densities, which simply compounds this environmental effect. Statistical analysis of host species, planting density and sandalwood growth characteristics in the FPC Kununurra collection is currently being undertaken. While the ages of trees vary considerably, some estimate of environmental variability may me attained. In order to fully confirm the role environment plays on heartwood production and essential oil yield, large scale trials of clonal material growing under varying circumstances would be the only way to accurately quantify these effects.

Genetic control of essential oil production in S. album may also be occurring at a nucleotide level, undetectable by anonymous markers such as RFLPs. This could be tested

48 by genotyping individual trees with a focus on characterised genes and their regulatory elements. For example, Cseke et al. (1998) reported differences in the promoter region of linalool synthase of Clarkia brewerii, compared to linalool synthase from its closest living ancestor, C. concinna. Genes from both species share high similarity, indicative of a shared ancestry. However C. brewerii produces large amounts of linalool in flower tissues while C. concinna does not. Using northern blot analysis Ceske et al. (1998) concluded that regulation at a transcription level was responsible for the observed chemical differences, and that promoter sequences may play a significant role in differential gene expression. If certain genotypes of S. album have highly active promoters which result in high yielding phenotypes, PCR primers could be developed based on these sequences or other related regulatory elements, and used for selection of superior genotypes.

Deployment approaches

Since clonal propagation of S. album is difficult at a production scale, seed will continue to be the most convenient mode of deployment. Therefore, in order to maintain genetic diversity, orchard design is critical and panmixic pollination is highly desired as it helps to limit homozygosity in the progeny (Olsson et al. 2001). This issue is highlighted by a recent study of a family trial for the Western Australian native sandalwood, S. spicatum, where the trial established was representative of the genetic diversity in the southern populations of the species but the progeny of the trial were less diverse due to apparent lack of synchronous flowing among all families (Muir et al. 2007). Similarly, S. album growing in the FPC Kununurra collection is at high risk of further homogenisation if random mating is not promoted. Seed orchards of both Picea glauca (Stoehr and El-

Kassaby 1997) and Psuedotsuga menziesii (El-Kassaby and Ritland 1995) were found to have similar genetic diversity and allelic representation as the original wild sources, and

49 second generation plants from these orchards are equally diverse. Phenotypic traits are also uncompromised by the high selection pressure in Picea glauca (Stoehr et al. 2005) courtesy of a diverse breeding population. Also of concern to breeders is the risk of pollen introduction from nearby plantations. Well over 1000 Ha of S. album has been planted in the Ord River Irrigation Area (Done et al. 2004), much of which is within 5 km of the FPC collection. It will be essential then to limit the flow of pollen into the orchard, particularly if the surrounding plantations arise from a narrow genetic base through selection. There is no risk of affecting natural populations, as S. album is not endemic to the Kimberly region of Western Australia.

From the results of this study, it is clear that more collections of S. album from different areas of its distribution are required. Given the inconsistency between genetic diversity in natural populations in India and that in the FPC Kununurra collection, more work must be done to clarify the cause of this difference. Alternative molecular markers such as SSRs may reveal higher genetic diversity than the techniques used here. More importantly, the extent to which the growing environment affects the observed differences in heartwood content and essential oil yield must be determined. Long-term trials of clonal material should be tested under a variety of host species and planting densities currently employed in the plantation sector. As phenotypic variation in oil yield is very high, the underlying cause of this variation must be determined.

Acknowledgements:

The authors are grateful to the Australian Research Council, Forest Products

Commission of Western Australia, and the Western Australian Department of Environment and Conservation for generous financial support. This work is part of Australian Research

50 Council Linkage Project LP0454919. Many thanks to Bronwyn MacDonald and Lin Wong for excellent technical assistance.

References

Adams RP (1994) Geographic variation and systematics of monospermous Juniperus (Cupressaceae) from the Chihuahua desert based on RAPDs and terpenes. Biochem. Syst. Ecol. 22, 699-710.

Aldrich PR, Doebley J (1992) Restriction fragment variation in the nuclear and chloroplast genomes of cultivated and wild Sorghum bicolor. Theor. Appl. Genet. 85, 293-302.

Anonymous (2003) Geology of Timor Leste. In 'Atlas of mineral resources of the ESCAP region ST/ESCAP/2267'. (United Nations Economic and Social Commission for Asia and the Pacific (UNESCAP) www.unescap.org/esd/water/publications/mineral/amrs/vol17: Bangkok)

Austerlitz F, Mariette S, Machon N, Gouyon P-H, Godelle B (2000) Effects of colonization processes on genetic diversity: Differences between annual plants and tree species. Genetics 154, 1309–1321.

Bottin L, Isnard C, Lagrange A, Bouvet J-M (2007) Comparative molecular and phytochemical study of the tree species Santalum austrocaledonicum (Santalaceae) distributed in the New-Caledonian archipelago. Chem. Biodivers. 4, 1541-1556.

Bottin L, Vaillant A, Sire P, Cardi C, Bouvet JM (2005a) Isolation and characterization of microsatellite loci in Santalum austrocaledonicum, Santalaceae. Mol. Ecol. Notes 5, 800- 802.

Bottin L, Verhaegen D, Tassin J, Olivieri I, Vaillant A, Bouvet JM (2005b) Genetic diversity and population structure of an insular tree, Santalum austrocaledonicum in New Caledonian archipelago. Mol. Ecol. 14, 1979-1989.

Butcher PA, Moran GF, Perkins HD (1998) RFLP diversity in the nuclear genome of Acacia mangium. Heredity 81, 205-213.

Byrne M (2008) Phylogeny, diversity and evolution of eucalypts. In 'Plant Genome: Biodiversity and Evolution. Vol. 1 Part E'. (Eds A Sharma and A Sharma) pp. 303-346. (Oxford & IBH Publishing Co: New Delhi)

Byrne M, MacDonald B, Brand JE (2003a) Phylogeography and divergence in the chloroplast genome of Western Australian sandalwood (Santalum spicatum). Heredity 91, 389-395.

Byrne M, MacDonald B, Brand JE (2003b) Regional genetic differentiation in Western Australian sandalwood (Santalum spicatum) as revealed by nuclear RFLP analysis. Theor. Appl. Genet. 107, 1208-1214.

51 Cseke L, Dudareva N, Pichersky E (1998) Structure and evolution of linalool synthase. Mol. Biol. Evol. 15, 1491–1498.

Done C, Kimber P, Underwood R (2004) Development of the Indian sandalwood industry on the Ord River irrigation area. In 'Prospects for high-value hardwood timber plantations in the 'dry' tropics of northern Australia'. Mareeba pp. 1-8. (http://www.plantations2020.com.au/reports/pfnq/acrobat/1-7_done_et_al.pdf)

Egerton-Warburton LM, Ghisalberti EL, Considine JA (1998) Infraspecific variability in the volatile leaf oils of Chamelaucium uncinatum (Myrtaceae). Biochem. Syst. Ecol. 26, 873-888.

El-Kassaby Y, Ritland K (1995) Impact of selection and breeding on the genetic diversity in Douglas-fir. Biodivers. Conserv. 5, 795-813.

Felsenstien J (1993) PHYLIP (Phylogeny Inference Package). V 3.5c (Department of Genetics, University of Washington, Seattle)

Glaubitz JC, El-Kassaby Y, Carlson J (2000) Nuclear restriction fragment length polymorphism analysis of genetic diversity in western red cedar. Can. J. Forest Res. 30, 379-389.

Harbaugh DT, Baldwin BG (2007) Phylogeny and biogeography of the sandalwoods (Santalum, Santalaceae): Repeated dispersals throughout the Pacific. Am. J. Bot. 64, 1028- 1040.

Hewson HJ, George AS (1984) Santalum. In 'Flora of Australia'. (Australian Government Publishing Service: Canberra)

Hillig KW (2004) A chemotaxonomic analysis of terpenoid variation in Cannabis. Biochem. Syst. Ecol. 32, 875-891.

Jones CG, Plummer JA, Barbour EL (2007) Non-destructive sampling of Indian sandalwood (Santalum album L.) for oil content and composition. J. Essent. Oil. Res. 19, 157-164.

Li H, Madden JL, Potts BM (1995) Variation in volatile leaf oils of the Tasmanian Eucalyptus species-1. Subgenus Monocalyptus. Biochem. Syst. Ecol. 23, 299-318.

Maia BHLNS, de Paula JR, Sant'Ana J, Da Silva MF, Fernandes JB, Vieira PC, Costa M, Ohashi OS, Silva JNM (2000) Essential oils of Toona and Cedrela species (Meliaceae): taxonomic and ecological implications. J. Brazil. Chem. Soc. 11, 629-639.

McDonald WM, Rawlings M, Butcher PA, Bell JC (2003) Regional divergence and inbreeding in Eucalyptus cladocalyx (Myrtaceae). Aust. J. Bot. 51, 393-403.

52 McKinnell FH (1990) Status of management and silviculture research of sandalwood in Western Australia and Indonesia. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990'. (Eds L Hamilton and CE Conrad) pp. 19-25. (US Forest Service: Honolulu, Hawaii)

Michaux B (2001) Dispersal versus vicariance, artifice rather than contest. In 'Faunal and floral migrations and evolution in SE Asia-Australasia'. (Eds I Metcalfe, JMB Smith, M Morwood and I Davidson) pp. 311-318. (A A Balkema Publishers: Lisse)

Muir K, Byrne M, Barbour EL, Cox MC, Fox JED (2007) High levels of outcrossing in a family trial of Western Australian sandalwood (Santalum spicatum). Silvae Genet. 56, 222- 230.

Newton AC, Allnutt TR, Gillies ACM, Lowe AJ, Ennos RA (1999) Molecular phylogeography, intraspecific variation and the conservation of tree species. Trends Ecol. Evol. 14, 140-145.

Olsson T, Lindgren D, Li B (2001) Balancing genetic gain and relatedness in seed orchards. Silvae Genet. 50, 222-227.

Outlaw DC, Voelker G (2007) Pliocene climatic change in insular Southeast Asia as an engine of diversification in Ficedula flycatchers. J. Biogeogr. doi:10.1111/j.1365- 2699.2007.01821.x.

Parker PG, Snow AA, Schug MD, Booton GC, Fuerst PA (1998) What molecules can tell us about populations: Choosing and using a molecular marker. Ecology 79, 361-382.

Peakall R, Ebert D, Scott LJ, Meagher PF, Offord CA (2003) Comparative genetic study confirms exceptionally low genetic variation in the ancient and endangered relictual conifer, Wollemia nobilis (Araucariaceae). Mol. Ecol. 12, 2331-2343.

Peakall R, Smouse P (2001) GenAlEx, Genetic analysis in Excel. Population genetic software for teaching and research. V 5 (Australian National University, Canberra)

Powell W, Morgante M, Andre C, Hanafey M, Vogel J, Tingey S, Rafalski A (1996) The comparison of RFLP, RAPD, AFLP and SSR (microsatellite) markers for germplasm analysis. Mol. Breeding 2, 225-238.

Rieseberg L, Beckstrom-Sternberg S, Doan K (1990) Helianthus annuus ssp. Texanus has chloroplast DNA and nuclear ribosomal RNA genes of Helianthus debilis ssp. cucumerifolius. Proc. Natl. Acad. Sci. USA 87, 593-597.

Rugkhla A, McComb JA, Jones MGK (1997) Intra- and interspecific pollination of Santalum spicatum and S. album. Aust. J. Bot. 45, 1083-1095.

53 Segal R, Feuerstein I, Danin A (1987) Chemotypes of Artemisia hera-alba in Israel based on their sesquiterpene lactone and essential oil composition. Biochem. Syst. Ecol. 15, 411- 416.

Shinozaki K, Ohme M, et al. (1986) The complete nucleotide sequence of the tobacco chloroplast genome: Its gene organization and expression. EMBOJ 5, 2043-2049.

Smith RL, Sytsma KJ (1990) Evolution of Populus nigra (Sect. Aigeiros): Introgressive hybridization and the chloroplast contribution of Populus alba (Sect. Populus). Am. J. Bot. 77, 1176-1187.

Sneath PHA, Sokal RR (1973) 'Numerical Taxonomy.' (W. H. Freeman and Co.: San Fransisco)

Soltis DE, Soltis PS, Collier TG, Edgerton ML (1991) Chloroplast DNA variation within and among genera of the Heuchera group (Saxifragaceae): Evidence for chloroplast transfer and paraphyly,. Am. J. Bot. 78, 1091-1112.

Southern EM (1975) Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503-517.

Stoehr M, El-Kassaby Y (1997) Levels of genetic diversity at different stages of the domestication cycle of interior spruce in British Columbia. Theor. Appl. Genet. 94, 83-90.

Stoehr M, O'Neill G, Hollefreund C, Yanchuk A (2005) Within and among family variation of orchard and wild-stand progeny of interior spruce in British Columbia. Tree Genet. Genomes 1, 64-68.

Sugiura M, Shinozaki K, Zaita N, Kusuda M, Kumano M (1986) Clone bank of the tobacco (Nicotiana tabacum) chloroplast genome as a set of overlapping restriction endonuclease fragments: mapping of eleven ribosomal protein genes. Plant Science 44, 211-217.

Suma TB, Balasundaran M (2003) Isozyme variation in five provenances of Santalum album in India. Aust. J. Bot. 51, 243-249.

Systma KJ, Gottlieb LD (1986) Chloroplast DNA evolution and phylogenetic relationships in Clarkia Sect. Peripetasma (Onagraceae). Evolution 40, 1248-1261.

Turner ME, Stephens JC, Anderson WW (1982) Homozygosity and patch structure in plant populations as a result of nearest-neighbor pollination. Proc. Natl. Acad. Sci. USA 79, 203- 207.

Waters ER, Schaal BA (1991) No variation is detected in the chloroplast genome of Pinus torreyana. Can. J. Forest Res. 21, 1832–1835.

Wildung MR, Croteau R (2003) Genetic engineering of peppermint for improved essential oil composition and yield. Transgenic Res. 14, 365–372.

Wolfe AD, Elisens WJ (1995) Evidence of chloroplast capture and pollen-mediated gene flow in Penstemon sect. Peltanthera (Scrophulariaceae). Syst. Biol. 20, 395-412.

54 Chapter 4. Quantitative co-occurrence of sesquiterpenes; a tool for elucidating their biosynthesis in Indian Sandalwood, Santalum album L.

This experimental chapter examines the essential oil profiles and yields of trees growing in the FPC Kununurra collection. Co-occurrence patterns of specific chemicals contained in the oils are described and discussed with respect to their biosynthesis. This has been published in Phytochemistry, 67, 2463-2468.

55 Quantitative co-occurrence of sesquiterpenes; a tool for elucidating

their biosynthesis in Indian Sandalwood, Santalum album

Christopher G. Jones a,b*, Emilio L. Ghisalberti b, Julie A. Plummer a,

Elizabeth L. Barbour c

a School of Plant Biology (M084), Faculty of Natural and Agricultural Sciences, University of Western

Australia, 35 Stirling Hwy, Crawley, WA 6009, Australia.

b School of Biomedical, Biomolecular and Chemical Sciences (M313), Faculty of Life and Physical

Sciences, University of Western Australia, 35 Stirling Hwy, Crawley, WA 6009, Australia.

c Forest Products Commission of WA, 117 Great Eastern Highway, Rivervale, WA 6103, Australia.

______* Corresponding author, Tel.: +61 08 64881992; fax: +61 08 64881108

E-mail address: [email protected]

56 Abstract

A chemotaxonomic approach was used to investigate biosynthetic relationships between heartwood sesquiterpenes in Indian sandalwood, Santalum album L. Strong, linear relationships exist between four structural classes of sesquiterpenes; α- and β- santalenes and bergamotene; γ- and β-curcumene; β-bisabolene and α-bisabolol and four unidentified sesquiterpenes. All samples within the heartwood yielded the same co-occurrence patterns, however wood from young trees tended to be more variable. It is proposed that the biosynthesis of each structural class of sesquiterpene in sandalwood oil is linked through common carbocation intermediates. Lack of co-occurrence between each structural class suggests that four separate cyclase enzymes may be operative. The biosynthesis of sandalwood oil sesquiterpenes is discussed with respect to these co-occurrence patterns.

Extractable oil yield was correlated to heartwood content of each wood core and the oil composition did not vary significantly throughout the tree.

Keywords: essential oil; Santalum album, α-santalene, β-santalene; α-trans-bergamotene, epi-β-santalene, santalols, heartwood

57 Introduction

Indian sandalwood, Santalum album L., is a hemi-parasitic tree native to India and

Indonesia that is highly valued for its fragrant heartwood. The heartwood of mature trees

(>10 years old) contains essential oils, chiefly the sesquiterpene alcohols cis-α-santalol (1), cis-β-santalol (2) (Verghese et al. 1990), α-trans-bergamotol (3), epi-cis-β-santalol (4) (Fig

4.1) along with small quantities of trans-β-santalol and cis-lanceol (5) (Howes et al. 2004).

The parent hydrocarbons, α-santalene (6), β-santalene (7), α-bergamotene (8) and epi-β- santalene (9) are also present in the oil, as well as α-curcumene (10), γ-curcumene (11), β- curcumene (12), β-bisabolene (13) and α-bisabolol (14) (Figs. 4.2 and 4.3) (Adams et al.

1975; Braun et al. 2003; Howes et al. 2004). Compositional differences have been noted throughout the roots and stem of S. spicatum (Piggot et al. 1997) and S. album

(Shankaranarayana et al. 1998). Although a route to 1 and 2 from farnesyl diphosphate was proposed almost 40 years ago (Parker et al. 1967), it has not been studied thoroughly due to slow onset of oil formation and the location of key enzymes within woody tissue. The former renders in vivo isotopic feeding experiments impractical, while the latter makes extraction of functional enzyme systems difficult due to the low number of living cells in wood.

Figure 4.1. Main sesquiterpene alcohols of Santalum album.

58 Terpenoid variation can be used to explore hypothetical biogenic pathways. A quantitative, mathematical relationship following the general linear form y = a b + C between individual compounds in an oil profile would suggest that the biosynthesis of such chemicals is not only closely related, but may involve the same enzyme system or at least a common intermediate (Zavarin 1970). The resulting proportionality of these compounds would be determined by flexibility at the active site of an enzyme (allowing more than one isomer to be produced), and the mechanistic probability of hydride shifts or Wagner-

Meerwein type rearrangement of the intermediate. Multiple product formation has been noted in various sesquiterpene synthases extracted from plants (Colby et al. 1998; Munck and Croteau 1990), and by site-directed mutation of active site residues in functional enzymes (Hyatt and Croteau 2005; Little and Croteau 2002). In vivo essential oil profiling, with the objective of identifying co-occurrences in these species, has not been widely performed on these species.

Alternatively, the co-occurrence of structurally similar compounds may be due to several enzymes working in unison. In this instance, the correlation would most likely be weaker than the common intermediate hypothesis, as several complicating factors would be involved. Quantitative co-occurrence would best explain structural isomers that share a common intermediate. Correlation between product and functionalised precursor may be apparent, however the strength of correlation is likely to be considerably lower than between structural isomers. Such relationships may yield negative linear slopes, where substitution of a precursor compound results in an increase in the proportion of product

(Zavarin 1970; Zavarin et al. 1970).

In this study, plantation grown S. album of various ages was analysed for heartwood oil content and composition. Variation in composition was considered both within individual

59 trees, age groups and amongst the entire population. The biosynthesis of sandalwood sesquiterpenes is described with respect to these co-occurrences.

Results and Discussion

Both qualitative and quantitative co-occurrence patterns were discovered in the essential oil of S. album heartwood. Percent composition data was the most informative for investigating quantitative co-occurrence patterns, as peak area data revealed only qualitative relationships. Product-precursor relationships were apparent, but correlation was weak. Samples in which the concentration of oil was very low (particularly 100 cm cores from young trees) gave more variable co-occurrence ratios. When a single age group of trees were considered, slight differences could be seen in the essential oil composition between younger and older wood. A higher proportion of sesquiterpene alcohols was observed for 30 cm cores, while 100 cm cores had a slightly lower proportion. The amount of heartwood in each core sample was correlated to the overall oil yield. The total oil content of central and transition zone heartwood was fairly consistent at 3.1 to 3.6 % dry wt.

Zavarin (1970) noted that both quantitative and qualitative co-occurrence patterns may exist in the essential oil composition of conifers and many other plants. Here, an example of qualitative co-occurrence in sandalwood oil was α-santalol (1) and β-santalol (2). Both shared a strong, positive linear correlation (r = 0.99) when peak area data was used.

However, when percent composition data was used (Fig. 4.5), a cluster of points arises around 53% and 23% for α-santalol and β-santalol respectively, with no significant correlation. This highlights the importance of using percent composition data when examining co-occurrences. Percent composition data is essentially adjusted by the presence

60 of everything else in the oil, whereas peak area data is not. For this reason, only co- occurrences which arose from percent composition data were investigated.

Striking co-occurrence patterns were present amongst the olefinic sesquiterpenes (6-9)

(Fig. 4.6A). Generally, older wood cores tended to follow the trendline very closely, but cores form younger trees were more variable. An example of this variation is illustrated in

Figure 4.6B.

Oil deposition most probably initiates in the heartwood-sapwood transition zone and after a certain concentration of oil is reached, biosynthesis ceases. The oxidation of sesquiterpenes to sesquiterpenols must also occur in this period and would affect the overall composition of the oil for this sample. Younger heartwood, or wood which has only recently been synthesised may contain lower amounts of alcohols 1-4.

A number of mono- and sesquiterpene synthases are remarkable in the fact that they can transform a single substrate into multiple products. Examples of synthases that can utilise both enantiomers of an acyclic terpene 3-diphosphate R-15 and S-15 (Fig 4.2) have also been described (Colby et al. 1998; Munck and Croteau 1990; Wise and Croteau 1999).

Moreover, some cyclases can generate enantiomers from the corresponding l-diphosphate

(16, 17) (Fig 4.2) due to the inherent chirality arising from the folding of an acyclic substrate into a right- and left-handed helical conformation (Colby et al. 1998; Munck and

Croteau 1990).

The proposed biosynthetic pathway to the santalenes and α-bergamotene (Fig 4.2) has been adapted from the model pathways proposed for mono- (Wise and Croteau 1999) and sesquiterpene examples (Kollner et al. 2004) with the added constraint imposed by the known absolute configuration of these sesquiterpenes (Chapuis et al. 1998; Krotz and

Helmchen 1994). Interestingly, epi-β-santalene (9) appears to be generated from a sequence that leads to products antipodal to those that include 6, 7 and 8.

61 Another possibility is that some amount of 3S-nerolidyl diphosphate (S-15) may cyclise in the anti-exo conformation as observed for the monoterpene analogues (Fig 4.4) (Wise and Croteau 1999). In either case, one would expect to see epi-- santalene and epi-α- bergamotene in the oil profile, unless these isomers were not resolved under the chosen chromatographic conditions.

Figure 4.2. Proposed biosynthesis of the santalenes and bergamotene (6-9)

Figure 4.3. Proposed biosynthetic pathway of farnesyl diphosphate to the curcumenes (10 and 11), bisabolene and bisabolol (12 and 13) and cis-lanceol (5)

Correlation was also found between γ-curcumene (10) and β-curcumene (11) (y =

0.588 x + 0.28, r = 0.88). It is presumed that these two monocyclic sesquiterpenes are synthesised from the bisabolyl cation (18) via a 1, 2 hydride shift followed by loss of a proton (Fig 4.3). As neither of these compounds shared a linear relationship with either the santalene series or β-bisabolene, the biosynthesis of the curcumenes may be a separate process unrelated to the other sesquiterpenes which utilises a different cyclase.

Figure 4.4. Alternative folding of acyclic sesquiterpene precursor farnesyl diphosphate in Santalum album.

Although only minor components, the percent composition of β-bisabolene (12) was correlated (r = 0.76) to that of α-bisabolol (13). The synthesis of 12 probably occurs via the initial cyclisation product from farnesyl or nerolidyl diphosphate, followed by loss of a proton from C-14 (Fig 4.3). It would appear that the formation of curcumenes 11 and 12 could be closely related to that of 13 and 14, but no correlation between these compounds was observed. It is possible that different cyclases are involved and generate the enantiomeric bisabolyl intermediate, whereby the 6R- compound leads to 12 and 13 and the

6S- to 10 and 11 following a 6,7-hydride shift.

Figure 4.5. Co-occurrence of α- and β-santalol (1 and 2 respectively) in S. album oil

The fourth series of sesquiterpenols in S. album heartwood eluted between α-curcumene and α-bisabolol and had a molecular ion of 220 m/z (data not shown). Strong linear correlations were present between these three compounds, while having no correlation with the santalenes, curcumenes or bisabolene. The full identity of these compounds is not known, but based on their correlations being independent from the other sesquiterpenes it seems probable that a separate cyclase/synthase is responsible for their synthesis.

Cores extracted from wood deemed younger (5 to 8 year-old trees at 100 cm above ground) tended to be further from the main trend lines than those derived from mature wood (14-17 year-old trees at 30 cm above ground). Sapwood contained almost no extractable oil with only the main component of sandalwood oil, α-santalol, occasionally present in very small concentrations. Transition zone wood contained an average of 36 ±

6.0 mg g-1 dry wt. of extractable oil, while the central heartwood yielded 30 ± 5.1 mg g-1 dry wt. of oil. The maximum oil yield for both transition zone and heartwood was 90.6 mg g-1 dry wt. and 60.5 mg g-1 dry wt. respectively. No differences were observed in either the

65 yield or composition of the extracted oil from both zones. The current understanding of heartwood formation suggests that as the secondary xylem ages, water conducting tissues stop functioning or become blocked (Zimmermann 1983). This occurs as the tree grows, presumably in seasonal bursts. Seasonal growth was difficult to identify in wood discs as these trees had been flood irrigated regularly throughout the dry season. Due to the slow- growing habit of sandalwood, seasonal changes in composition or yield may not have been obvious when using a 9 mm wide sample of transition wood.

Figure 4.6. A: Co-occurrence of compounds 6-9 in S. album oil: a; 7 vs. 8 (y=7.54 x + 0.13, r=0.98), b; compounds 8 vs. 9 (y=5.08 x + 0.74, r=0.98), c; compounds 6 vs. 8 (y=1.28 x + 0.31, r=0.88), d; compounds 9 vs. 7 (y=1.21 x + 0.16, r=0.90), e; compounds 6 vs. 9 (y=0.92 x + 0.18, r=0.85), f; compounds 6 vs. 7 (y=0.18 x + 0.002, r=0.98). B: Co-occurrence of compounds 6 and 9 with the age of wood cores highlighted.

66 Total essential oil yield from S. album trees varied considerably amongst the sampled population. Presence or absence of essential oil depended mostly on tree age, and generally trees under 8 years yielded little fragrant heartwood. Extractable oil yield was compared to the proportion of heartwood in cores from both heights, resulting in the equation: oil yield = 42.803 x proportion of heartwood - 0.4501, r = 0.87, P<0.01). Despite this general trend, many samples from 15 year old trees yielded very little fragrant heartwood. Several factors involving the structure of storage tissues and the location of essential oils within the heartwood undoubtedly contribute to the variation seen in this relationship, and these in turn may be related to growing conditions, host tree species and genotype. Increased rates of sesquiterpene biosynthesis, either genetically or environmentally induced, may also contribute to this disparity. Further work is required in order to fully understand the process of heartwood formation and secondary metabolite biosynthesis in wood.

Experimental

Site description

Santalum album trees were growing at the Frank Wise Agricultural Research Station,

Kununurra, Western Australia. Soil was heavy clay, very plastic when wet and very hard when dry. The Northcote key classifies the soil type as Ug 5.29, coloured dark brown (2.5

Y 4/2) with neutral to slightly alkaline pH (Radomiljac 1998). Kununurra has a tropical climate with a very distinct monsoon wet season from December through to April, and an average annual rainfall of 790 mm. Mean temperature ranges from 15-30C in winter and

25-38C in summer (Australian Bureau of Meteorology, Commonwealth Government

Perth). The plot was within the Ord River Irrigation Area where flood irrigation was carried out monthly during dry seasons. Sandalwood trees were growing alongside various hosts,

67 including Acacia trachycarpa, Azadirachta indica, Dalbergia sp. Khaya sp. Melaleuca sp.

Pongamia sp. Sesbania formosa, and S. grandifolia. Both host and sandalwood trees were form pruned.

Within-tree essential oil variation (destructive sampling)

Discs (5 cm thick) from ground level were cut from four 10 year-old sandalwood trees in May 2003. These were dried and stored. The upper surface of each disc was sanded using a belt sander so that central heartwood, sapwood and the transitional zone between these two zones could be easily distinguished. Air-dried wood samples were taken from these zones by drilling parallel to the grain of the wood and collecting the shavings. Three replicates of shavings were taken from the sapwood, transition zone (to the inside of the transition line) and central heartwood using a drill equipped with a 9 mm drill bit. This yielded enough wood shavings for subsequent analyses. These were stored at room temperature in plastic bags until analysis.

Core sampling of plantation trees (non-destructive sampling)

Core sampling was performed in spring 2004. Cores (diameter 12 mm) were extracted from the diameter of 98 trees of ages ranging from 5 years to 17 years. Cores from two heights, 30 cm and 100 cm above ground level were taken using a petrol-powered drill with a specialised coring bit. Trees were re-sealed with caulking polymer to prevent infection. In total, 196 cores were extracted at two heights (30 cm and 100 cm) from 98 trees, ranging in age from 9 to 18 years. Heartwood from trees aged 5-8 years, and at 100 cm above ground was deemed „younger‟. Heartwood from trees aged 14-17 years and 30 cm above ground was deemed „older‟. Once removed these were placed into paper bags and transported to the laboratory where they were air-dried and stored at ambient temperature (20 ± 5°C).

68 Oven drying of all wood samples was not performed as this risked volatilisation of the heartwood oils. Core length and heartwood proportion was measured. In some instances the heartwood was difficult to distinguish from sapwood. Wood cores were ground in a

Rhetz-Mhule grinder to 500 μm and stored in re-sealable plastic bags until analysis.

Heartwood was not separated from sapwood prior to grinding.

Solvent extraction of wood

Sandalwood shavings (4-5 g) were placed into 50 ml volumetric flasks and a 1 ml aliquot of internal standard, (±)-camphor (4.00 g l-1) was added. The flasks were then filled just short of the mark with spectroscopic grade hexane. Volumetric flasks were stored at

21°C for at least 2 days and periodically shaken before being topped up to the mark and analysed by GC-FID.

Gas chromatographic analysis of sandalwood extracts

GC analysis was performed using a Shimadzu GC-17A instrument equipped with an

AT-WAX column (Alltech, 30 m, 0.25 mm inside diameter, 0.25 μm film thickness) and flame ionisation detection. Injection volume was 1.5 µl, the injection port temperature was

200°C and detector 250°C. Helium (2.4 ml min-1) was the carrier gas an and a split ratio of

20:1 was used. Oven temperature was ramped from 80o at 4°C min-1 to 180°C and held for

20 min (total run time 45 min). Peak identification was facilitated by application of GC-MS using a Shimadzu QP2010 instrument under identical column conditions. Integration was performed using Shimadzu GC-Solutions software. Areas were recorded for all detectable peaks, and percent composition was calculated by taking area of peak divided by total chromatogram area x 100. Samples which contained small amounts of total oil tended to overestimate the proportion of the major components (1-4).

Acknowledgements

The authors are grateful for the financial assistance of the Australian Research Council and the Forest Products Commission of Western Australia through linkage project LP 0454919.

Thanks also go to Mr Greg Cawthray for technical assistance with GC analysis.

Preliminary results of this investigation were presented at the Regional Workshop on

Sandalwood Research Development and Extension in the Pacific Islands and Asia, Nadi,

Fiji. Held 28th November – 1st December 2005

70 References

Adams DR, Bhatnagar SP, Cookson RC (1975) Sesquiterpenes of Santalum album and Santalum spicatum. Phytochemistry 14, 1459-1460.

Braun NA, Meier M, Pickenhagen W (2003) Isolation and chiral GC analysis of - bisabolols - trace constituents from the essential oil of Santalum album L. (Santalaceae). Journal of Essential Oil Research 15, 63-65.

Chapuis C, Barthe M, Muller BL, Schulte-Eke KH (1998) Preparation and absolute configuration of (-)-(E)--trans-bergamotenone. Helvetica Chimica Acta 81, 153-162.

Colby SM, Crock J, Dowdle-Rizzo B, Lemaux PG, Croteau R (1998) Germacrene C synthase from Lycopersicon esculentum cv. VFNT Cherry tomato: cDNA isolation, characterisation, and bacterial expression of the multiple product sesquiterpene cyclase. Proc. Natl. Acad. Sci. USA 95, 2216-2221.

Howes M-JR, Simmonds MSJ, Kite GC (2004) Evaluation of the quality of sandalwood essential oils by gas chromatography-mass spectrometry. Journal of Chromatography A 1028, 307-312.

Kollner TG, Schnee C, Jonathan G, Degenhardt J (2004) The variability of sesquiterpenes emitted from two Zea mays cultivars is controlled by allelic variation of two terpene synthase genes encoding stereoselective multiple product enzymes. The Plant Cell 16, 1115-1131.

Little DB, Croteau R (2002) Alteration of product formation by directed mutagenesis and truncation of the multiple-product sesquiterpene synthases -selinene synthase and - humulene synthase. Archives of Biochemistry and Biophysics 402, 120-135.

Munck SL, Croteau R (1990) Purification and characterization of the sesquiterpene cyclase patchoulol synthase from Pogostemon cablin. Archives of Biochemistry and Biophysics 282, 58-64.

Parker W, Roberts JS, Ramage R (1967) Sesquiterpene biogenesis. Quarterly Reviews 21, 331-363.

71 Radomiljac A (1998) The influence of pot host species, seedling age and supplementary nursery nutrition on Santalum album L. (Indian sandalwood) plantation establishment within the Ord River Irrigation Area, Western Australia. Forest Ecology and Management 102, 193-201.

Wise ML, Croteau R (1999) Biosynthesis of monoterpenes. In 'Comprehensive Natural Products Chemistry'. (Ed. DE Cane) pp. 97-153. (Elseiver: Oxford)

Zavarin E (1970) Qualitative and quantitative co-occurrence of terpenoids as a tool for elucidation of their biosynthesis. Phytochemistry 9, 1049-1063.

Zavarin E, Snajberk K, Reichert T, Tsien E (1970) On the geographic variability of the monoterpenes from the cortical blister oleoresin of Abies lasiocarpa. Phytochemistry 9, 377-395.

Zimmermann MH (1983) 'Xylem structure and the ascent of sap.' (Springer-Verlag: Berlin)

72 Chapter 5. Isolation of cDNAs and functional characterisation of two multi-product TPS enzymes from sandalwood, Santalum album L.

This final experimental chapter considers the biosynthesis of terpenoids by TPS genes in S. album. Isolation and functional characterisation of two TPS genes was carried out, along with subsequent GC-MS analysis. This chapter was accepted in Archives of Biochemistry and Biophysics in May 2008, and due to appear in print soon.

73 Isolation of cDNAs and functional characterisation of two multi-product

terpene synthase enzymes from sandalwood, Santalum album L.

Christopher G Jones a,c*, Christopher I Keeling b, Emilio L Ghisalberti c,

Elizabeth L Barbour d, Julie A Plummer a and Jörg Bohlmann b

a School of Plant Biology (M084), Faculty of Natural and Agricultural Sciences, University of Western

Australia, 35 Stirling Hwy, Crawley, WA 6009, Australia. b Michael Smith Laboratories, University of British Columbia 301-2185 East Mall, Vancouver, BC, Canada

V6T 1Z4

c School of Biomedical, Biomolecular and Chemical Sciences (M313), Faculty of Life and Physical Sciences,

University of Western Australia, 35 Stirling Hwy, Crawley, WA 6009, Australia.

d Forest Products Commission of Western Australia, 117 Great Eastern Highway, Rivervale, WA 6103,

Australia.

* Corresponding author. Tel.: +61 8 6488 1992; fax: +61 8 6488 1108

Email address: [email protected] (C.G. Jones)

74 Abstract

Sandalwood, Santalum album (Santalaceae) is a small hemi-parasitic tropical tree of great economic value. Sandalwood timber contains resins and essential oils, particularly the santalols, santalenes and dozens of other minor sesquiterpenoids. These sesquiterpenoids provide the unique sandalwood fragrance. The research described in this paper set out to identify genes involved in essential oil biosynthesis, particularly terpene synthases (TPS) in

Santalum album, with the long-term aim of better understanding heartwood oil production.

Degenerate TPS primers amplified two genomic TPS fragments from S. album, one of which enabled the isolation of two TPS cDNAs, SamonoTPS1 (1731 bp) and

SasesquiTPS1 (1680 bp). Both translated protein sequences shared highest similarity with known TPS from grapevine (Vitis vinifera). Heterologous expression in E. coli produced catalytically active proteins. SamonoTPS1 was identified as a monoterpene synthase which produced a mixture of (+)--terpineol and ()-limonene, along with small quantities of linalool, myrcene, ()--pinene, (+)-sabinene and geraniol when assayed with geranyl diphosphate. Sesquiterpene synthase SasesquiTPS1 produced the monocyclic sesquiterpene alcohol germacrene D-4-ol and helminthogermacrene, when incubated with farnesyl diphosphate. Also present were -bulnesene, -muurolene, - and -selinenes, as well as several other minor bicyclic compounds. Although these sesquiterpenes are present in only minute quantities in the distilled sandalwood oil, the genes and their encoded enzymes described here represent the first TPS isolated and characterised from a member of the Santalaceae plant family and they may enable the future discovery of additional TPS genes in sandalwood.

75 Introduction

Santalum album L. is a small hemi-parasitic tree found growing in southern India,

Sri Lanka, eastern Indonesia and northern Australia. The timber is highly sought after for its fine grain, high density and excellent carving properties. The fragrant wood is usually ground and steam distilled, with the essential oil serving as a fixative for many high-end perfumes. Centuries of over-exploitation has led to the demise of sandalwood in natural stands. Large plantations are being established throughout northern Australia to satisfy demand and conserve remaining reserves. Santalum album heartwood contains up to 6 % dry wt. sesquiterpene oils (Jones et al. 2006) predominantly - and -santalol, -trans- bergamotol and epi--santalol, along with the sesquiterpene olefins - and -santalene, - bergamotene and epi--santalene, -bisabolene, -, - and -curcumene (Howes et al.

2004). A series of aromatic and phenolic compounds have also been identified in the heartwood of S. album (Kim et al. 2005). The amount of heartwood oil produced in a tree varies considerably, even under near-identical growing conditions (Jones et al. 2007).

Causes of this yield variation are not well understood, but it is likely to be the result of both genetic and environmental factors.

Almost nothing is known about the biosynthesis of sesquiterpenoids in S. album or how essential oil production is regulated. Biosynthetic schemes were proposed decades ago (Parker et al. 1967) but these schemes have not been studied biochemically. Recent investigations into co-occurrence patterns of several sesquiterpenes in wood extracts suggest multiple products are formed from common carbocation intermediates (Jones et al.

2006). Multiple product formation by single terpene synthase (TPS) enzymes is well documented in the literature (Deguerry et al. 2006; Kollner et al. 2006; Martin et al. 2004;

Steele et al. 1998; Tholl et al. 2005). Terpenoid biosynthesis is widely understood to initiate through either the mevalonic acid pathway or the deoxy-D-xylulose pathway to

76 isoprenoid diphosphate precursors like geranyl diphosphate (GPP) and farnesyl diphosphate (FPP), and these are in turn converted by TPS to the terpenoid products typical of many plant essential oils (Davis and Croteau 2000). Plant terpenoids may be involved in defence, and the production of multiple compounds from a single TPS enzyme may provide an evolutionary advantage through increased resistance against potential herbivores or pathogens (Gershenzon and Dudareva 2007; Huber et al. 2004; Keeling and

Bohlmann 2006). We proposed that TPS are active in the heartwood and leaves of mature sandalwood trees, and that the diverse mono- and sesquiterpenoid profile of S. album likely stems from several multi-product TPS enzymes collectively involved in total oil biosynthesis. Essential oil is located in the heartwood of stems and roots of sandalwood, with biosynthesis probably occurring in parenchyma. Microscopic sections of S. album wood show the accumulation of oil and other extractives in ray parenchyma (Fig. 5.1). The sapwood-heartwood transition zone is therefore a logical place for the isolation of TPS mRNAs and cDNA cloning. We describe here the functional characterization of the first two TPS cDNAs from S. album.

Results

Identification of TPS gene fragments amplified from genomic DNA of S. album.

Degenerate primers (primers p5F and p8R) were based on conserved regions of several published angiosperm sesqui-TPS genes; valencene synthase from Vitis vinifera

(grapevine; AAS66358), germacrene D synthase from Populus trichocarpa x deltoides

(poplar; AAR99061), -cadinene synthase from Gossypium hirsutum (cotton; AAD51718) and -caryophyllene synthase from Artemisia annua (AAL79181). These primers amplified an intron-free 180 bp fragment from S. album genomic DNA. Upon cloning and sequencing of the PCR product, two independent TPS gene sequences were apparent.

Fig 5.1. Transverse section of S. album heartwood (lower) – sapwood (upper) transition zone (40x magnification). Essential oil and other extractives are deposited in the ray parenchyma (vertical features) during heartwood formation.

Based on BLAST searches against the NCBI GenBank nr database

(www.ncbi.nih.gov) the deduced amino acid sequences from these fragments both showed strong similarity to known sesquiterpene synthases from grapevine and cotton. The predicted peptide sequences of the two gDNA fragments cloned from S. album differed by

13 amino acids (87% identity, 100% similarity).

Full length cDNA cloning of two TPS genes from S. album.

The partial genomic TPS sequences enabled the design of PCR primers to be used for iterative rapid amplification of cDNA ends (RACE) which yielded the 3‟- and 5‟- cDNA ends and eventually the complete cDNA clones for two distinct TPS genes,

SamonoTPS1 and SasesquiTPS1 (Figs. 5.2 and 5.3). SasesquiTPS1 showed highest similarity to other angiosperm sesqui-TPS. SasesquiTPS1 was isolated from wood cDNA.

Although all initial primer design was targeted at sesquiterpene synthases, the other cDNA

78 SamonoTPS1, isolated from leaf cDNA showed highest similarity to previously discovered monoterpene synthases, predominantly grapevine -terpineol synthase. Alignments of the two S. album TPS deduced amino acid sequences with other angiosperm mono-TPS and sesqui-TPS are shown in Figures 5.2 and 5.3 respectively.

SamonoTPS1 features motifs typical of the monoterpene synthase gene TPS-b subfamily (Bohlmann et al. 1998) including the aspartate rich DDXXD metal ion binding site (positions 323-327) and the RRX8W motif (positions 33-44) which is implicated in diphosphate group migration (Williams et al. 1998). SasesquiTPS1 also shares the

R(R/P)X8W motif (positions 19-29) and the DDXXD motif. Y538 of SasesquiTPS1 resides in approximately the same position as Y520 of tobacco epi-aristolochene synthase

(TEAS) (Facchini and Chappell 1992; Starks et al. 1997) (Fig 5.3.). This residue has been shown to be responsible for protonation of germacrene A in TEAS (Rising et al. 2000).

Functional characterisation of SamonoTPS1 and SasesquiTPS1.

The proteins encoded by cDNAs SamonoTPS1 and SasesquiTPS1 were expressed as full-length His-tagged fusion proteins (N-terminal His6) in E. coli. Recombinant proteins were purified using Ni2+ affinity chromatography and confirmed to be of the expected size of ~66 kDa by SDS-PAGE and Western blot analysis using His6-specific antibody. Enzyme assays using SamonoTPS1 recombinant enzyme with GPP as substrate in the presence of Mn2+ produced a mixture of mainly (+)--terpineol and ()-limonene as well as a selection of minor monoterpene products as determined by GC-MS (Figure 5.4;

Table 5.1).

79 ▄▄▄▄▄▄ VVTPS ----MALSMLSSIPNLITHTRLPIIIKS--SSCKASPRGIKVKIGNSNCEEIIVRRTANY 54 CLLS MSSCINPSTLVTSANGFKCLPLATNKAA--IRIMAKNKPVQCLVSAKYDNLIVDRRSANY 58 CTGS ----MALQMIAPFLSSFLPNPRHSLAAHGLTHQKCVSKHISCSTTTPTYSTTVPRRSGNY 56 SamonoTPS1 ------MDAFATSPTSALIKA------VNCIAHVTPMAGEDSSENRRASNY 39 . : . : . **:.** ▄▄▄▄▄ VVTPS HPTIWDYDYVQSLRSDY--VGETYTRRLDKLKRDVKPMLGKVKKPLDQLELIDVLQRLGI 112 CLLS QPSIWDHDFLQSLNSNY--TDETYRRRAEELKGKVKIAIKDVTEPLDQLELIDNLQRLGL 116 CTGS KPSIWDYDFVQSLGSGY--KVEAHGTRVKKLKEVVKHLLKETDSSLAQIELIDKLRRLGL 114 SamonoTPS1 KPSTWDYEFLQSLATSHNTVQEKHMKMAEKLKEEVKSMIKGQMEPVAKLELINILQRLGL 99 :*: **::::*** :.: * : .:** ** : ..: ::***: *:***:

VVTPS YYHFKDEIKRILNSIYNQYNRHEEWQKDDLYATALEFRLLRQHGYDVPQDVFSRFKDDTG 172 CLLS AYRFETEIRNILHNIYNN-NKDYVWRKENLYATSLEFRLLRQHGYPVSQEVFNGFKDDQG 175 CTGS RWLFKNEIKQVLYTISSD-NTSIEMRKD-LHAVSTRFRLLRQHGYKVSTDVFNDFKDEKG 172 SamonoTPS1 KYRFESEIKEELFSLYKD--GTDAWWVDNLHATALRFRLLRENGIFVPQDVFETLKDKSG 157 : *: **:. * .: .: : *:*.: .*****::* *. :**. :**. *

VVTPS SFKACLCEDMKGMLCLYEASYLCVQGESTMEQARDFAHRHLGKGLEQN---IDQNLAIEV 229 CLLS GF---IFDDFKGILSLHEASYYSLEGESIMEEAWQFTSKHLKEVMISKSMEEDVFVAEQA 232 CTGS CFKPSLSMDIKGMLSLYEASHLAFQGETVLDEARAFVSTHLMD-IKEN---IDPILHKKV 228 SamonoTPS1 KFKSQLCKDVRGLLSLYEASYLGWEGEDLLDEAKKFSTTNLNNVKESIS---SNTLGRLV 214 * : *.:*:*.*:***: :** :::* * :* . . . : .

VVTPS KHALELPLHWRMPRLEARWFIDVYEKRQDMNPILLEFAKLDFNMVQATHQEDLRHMSSWW 289 CLLS KRALELPLHWKVPMLEARWFIHVYEKREDKNHLLLELAKMEFNTLQAIYQEELKEISGWW 292 CTGS EHALDMPLHWRLEKLEARWYMDIYMREEGMNSSLLELAMLHFNIVQTTFQTNLKSLSRWW 288 SamonoTPS1 KHALNLPLHWSAARYEARWFIDEYEKEENVNPNLLKYAKLDFNIVQSIHQQELGNLARWW 274 ::**::**** ****::. * :.:. * **: * :.** :*: .* :* :: ** _____ ▄▄▄▄▄ VVTPS SSTRLGEKLNFARDRLMENFLWTVGVIFEPQYGYCRRMSTKVNTLITIIDDVYDVYGTMD 349 CLLS KDTGLGEKLSFARNRLVASFLWSMGIAFEPQFAYCRRVLTISIALITVIDDIYDVYGTLD 352 CTGS KDLGLGEQLSFTRDRLVECFFWAAAMTPEPQFGRCQEVVAKVAQLIIIIDDIYDVYGTVD 348 SamonoTPS1 VETGL-DKLSFVRNTLMQNFMWGCAMVFEPQYGKVRDAAVKQASLIAMVDDVYDVYGSLE 333 . * ::*.*.*: *: *:* .: ***:. : . ** ::**:*****:::

VVTPS ELELFTDVVDRWDIN-AMDPLPEYMKLCFLALYNSTNEMAYDALKEHGLHIVSYLRKAWS 408 CLLS ELEIFTDAVARWDINYALKHLPGYMKMCFLALYNFVNEFAYYVLKQQDFDMLLSIKNAWL 412 CTGS ELELFTNAIDRWDLE-AMEQLPEYMKTCFLALYNSINEIGYDILKEEGRNVIPYLRNTWT 407 SamonoTPS1 ELEIFTDIVDRWDIT-GIDKLPRNISMILLTMFNTANQIGYDLLRDRGFNGIPHIAQAWA 392 ***:**: : ***: .:. ** :. :*:::* *::.* *::.. . : : ::*

VVTPS DLCKSYLLEAKWYYSRYTPSLQEYISNSWISISGPVILVHAYFL--VANPITKEALQSLE 466 CLLS GLIQAYLVEAKWYHSKYTPKLEEYLENGLVSITGPLIIAISYLS--GTNPIIKKELEFLE 470 CTGS ELCKAFLVEAKWYSSGYTPTLEEYLQTSWISIGSLPMQTYVFAL--LGKNLAPESSDFAE 465 SamonoTPS1 TLCKKYLKEAKWYHSGYKPTLEEYLENGLVSISFVLSLVTAYLQTEILENLTYESAAYVN 452 * : :* ***** * *.*.*:**:... :** . : : : : :

VVTPS RYHNIIRWSSMILRLSDDLGTSLDELKRGDVPKSIQCYMYETGASEEDARKHTSYLIGET 526 CLLS SNPDIVHWSSKIFRLQDDLGTSSDEIQRGDVPKSIQCYMHETGASEEVAREHIKDMMRQM 530 CTGS KISDILRLGGMMIRLPDDLGTSTDELKRGDVPKSIQCYMHEAGVTEDVARDHIMGLFQET 525 SamonoTPS1 SVPPLVRYSGLLNRLYNDLGTSSAEIARGDTLKSIQCYMTQTGATEEAAREHIKGLVHEA 512 ::: .. : ** :***** *: ***. ******* ::*.:*: **.* :. :

VVTPS WKKLNE-DGAVESPFPETFIGIAMNLARMAQCMYQHGD----GHGIEYGETEDRVLSLLV 581 CLLS WKKVNAYTADKDSPLTRTTTEFLLNLVRMSHFMYLHGD----GHGVQNQETIDVGFTLLF 586 CTGS WKKLNE--YLVESSLPHAFIDHAMNLGRVSYCTYKHGDGFSDGFGDPGSQEKKMFMSLFA 583 SamonoTPS1 WKGMNK-CLFEQTPFAEPFVGFNVNTVRGSQFFYQHGD----GYAVTESWTKDLSLSVLI 567 ** :* ::.:... :* * : * *** *.. . ::::

VVTPS EPIPS-----LSFE------590 CLLS QPIPLEDKD-MAFTASPGTKG 606 CTGS EPLQVDEAKGISFYVDGGSA- 603 SamonoTPS1 HPIPLNEED------576 .*:

Fig 5.2. Alignment of SamonoTPS1 amino acid sequence with other closely related monoterpene synthases; VVTPS: Vitis vinifera -terpineol synthase (Martin and Bohlmann 2004), CLLS: Citrus limon ()-limonene synthase (Lucker et al. 2002), CTGS: Cinnamomum tenuipilum geraniol synthase (Yang et al. 2005). Consensus indicated on bottom row. FEPQY motif, common to both SamonoTPS1 and SasesquiTPS1 is indicated by an overhead line (─). Motifs common to all TPS indicated by a solid overhead line (▀).

80 ▄▄▄▄▄▄▄▄▄ VVGDS MSVQSSGVLLAPSKNLSPEVGRRCANFHPSIWGDHFLSYASEFTNTDDHLKQHVQQLKEE 60 PBGDS MSVEGSAIFSTAT--VEPNVSRRSAGYSPSIWGDHFLSYATDSMETSD--KAEHKKLKEE 56 SasesquiTPS1 MENQKVPISSVPN---LKELSRPIANFPPSIWGDRFINYACEDENEQAQKERQVEELKEQ 57 VVVS MSTQVSASSLAQI---PQPKNRPVANFHPNIWGDQFITYTPEDKVTRACKEEQIEDLKKE 57 *. : . .* *.: *.****:*:.*: : : . :.**::

VVGDS VRKML-MAADDDSAQKLLLIDAIQRLGVAYHFESEIDEVLKHMFDGSVVS------AE 111 PBGDS VKREL-MANINKPSQTLDFIDAIQRLGISYHFEIEIDEILREMYKSHCDFDNGDDDDHHH 115 SasesquiTPS1 VRREL-AATVDKPLQQLNIIDATQRLGIAYHFENEIEESLEHIYLHTYVENN---CFQGS 113 VVVS VKRKLTAAAVANPSQLLNFIDAVQRLGVAYHFEQEIEEALQHICN-SFHDCN---DMDG- 112 *:: * * .. * * :*** ****::**** **:* *..:

VVGDS EDVYTASLRFRLLRQQGYHVSCDLFNNFKDNEGNFKESLSSDVRGMLSLYEATHFRVHGE 171 PBGDS NDLYAISLKFRLLRQQGYKISCDVFGKFKNSQGTFNDSLANDTRGILSLYEATHLRVHGD 175 SasesquiTPS1 HDLYSVALWFRLLRQDGYKVSCDVFDKFRDYEDNFKNSLMEDAKGLLELYEATHLSVHGE 173 VVVS -DLYNIALGFRLLRQQGYTISCDIFNKFTDERGRFKEALISDVRGMLGLYEAAHLRVHGE 171 *:* :* ******:** :***:*.:* : .. *:::* .*.:*:* ****:*: ***:

VVGDS DILDEALAFTTTHLQSATKHSSNPLAEQVVHALKQPIRKGLPRLEARHYFSVYQADDSHN 231 PBGDS EVLEEALVFTTSHLEFLATHSSSPLRAKINHALKQPVRKNIPRLEARHYFSIYQEDPSCS 235 SasesquiTPS1 EMLDDALEFAKTRLESIVNHLNYPLAEQVRHALYRPLRKGLPRLEAVYFFRIYEAYDSHN 233 VVVS DILAKALAFTTTHLKAMVESLGYHLAEQVAHALNRPIRKGLERLEARWYISVYQDEAFHD 231 ::* .** *:.::*: . . * :: *** :*:**.: **** :: :*: . ____ VVGDS KALLKLAKLDFNLLQKLHQKELSDISAWWKDLDFAHKLPFARDRVVECYFWILGVYFEPQ 291 PBGDS EVLLNFAKLDFNILQKQHQKELSEIANWWKELDFAKKLPFARDRVIECYFWILGVYLEPE 295 SasesquiTPS1 KALLKLAKLDFNLLQSLHKKELSDMARWWKSLDFAAKFPFARDRLVEGYFWVLGVYFEPQ 293 VVVS KTLLELAKLDFNLVQSLHKEELSNLARWWKELDFATKLPFARDRLVEGYFWMHGVYFEPQ 291 :.**::******::*. *::***::: ***.**** *:******::* ***: ***:**: _ ▄▄▄▄▄ VVGDS FFFARRILTKVIAMTSIIDDIYDVYGTLEELELFTEAVERWDISAIDQLPEYMRVCYQAL 351 PBGDS YFLARRILTKVIAMTSVIDDIYDVYGTPEELELFTDAIERWEITAVDQLPEYMKVTYKAL 355 SasesquiTPS1 YSLARKIIIKVFTMISTIDDIYDAYGTLDELELFTKAMQRWDVGSLDQLPEYMKPCYKSI 353 VVVS YLRGRRILTKVIAMTSILDDIHDAYGTPEELKLFIEAIERWDINSINQLPEYMKLCYVAL 351 : .*:*: **::* * :***:*.*** :**:** .*::**:: :::******: * ::

VVGDS LYVYSEIEEEMAKEGRSYRLYYAKEAMKNQVRAYYEEAKWLQVQQIPTMEEYMPVALVTS 411 PBGDS LDVYTEIEENMVTEERSYRVYYAKEAMKNLVRAYYLESKWFHQKHTPTMEEYMAVALVTS 415 SasesquiTPS1 LDVYNEIEEEMDNQGSLFRMHYAKEVMKKLVEGYMDEAKWCHEKYVPTFEEYMPVALVTS 413 VVVS LDVYKEIEEEMEKEGNQYRVHYAKEVMKNQVRAYFAEAKWLHEEHVPAFEEYMRVALASS 411 * **.****:* .: :*::****.**: *..* *:** : : *::**** ***.:*

VVGDS AYSMLATTSFVGMGDAVTKESFDWIFSKPKIVRASAIVCRLMDDMVFHKFEQKRGHVASA 471 PBGDS GYAMLAATSFVGMGHVVTKDSFDWLFRGPKILKASEIICRLMDDIVSHKFEQKRGHVASS 475 SasesquiTPS1 GYTFLTTISYLGMGEIASKEAFDWLFSHPPVIEASESVCRLMDDMRSHKFEQERGHVASG 473 VVVS GYCLLATTSFVGMGEIATKEAFDWVTSDPKIMSSSNFITRLMDDIKSHKFEQKRGHVTSA 471 .* :*:: *::***. .:*::***: * :: :* : *****: *****:****:*.

VVGDS VECYMKQHGASEQETPNEFPQPVREAWKDINEECLIPTAVPMPILMRVLNLARVIDVIYK 531 PBGDS IECYMKQHGTTEQETVHEFRKQVTDAWKDLNEEFLHPTAVPMPLLTRMLNLARVIDVVYK 535 SasesquiTPS1 IECYMKQYGVTEEEAHDEFRKQLVKAWKDINEECLRPYRVPKPLLMRILNLTRVIDVIYK 533 VVVS VECYMKQYGVSEEQVYSEFQKQIENAWLDINQECLKPTAVSMPLLARLLNFTRTMDVIYK 531 :******:*.:*::. ** : : .** *:*:* * * *. *:* *:**::*.:**:** ‡ VVGDS NEDGYTHFGAVLKDFVTSMLIDPVPI 557 PBGDS DEDGYTNAGTVLKDLVSALLIDPVPM 561 SasesquiTPS1 NEDGYTHVKKAMKDNIASLLIDPMIV 559 VVVS EQDSYTHVGKVMRDNIASVFINAVI- 556 ::*.**: .::* :::::*:.:

Fig 5.3. Alignment of SasesquiTPS1 with other closely related angiosperm sesquiterpene synthase sequences; VVGDS: Vitis vinifera ()-germacrene D synthase and VVVS: Vitis vinifera (+)-valencene synthase (Lucker et al. 2004), PBGDS: Populus trichocarpa x deltoides ()-germacrene D synthase (Arimura et al. 2004). Consensus indicated on bottom row. FEPQY motif highlighted with an overhead line (─). Double dagger indicates Y538 of SasesquiTPS1. Motifs common to all TPS indicated by a solid overhead line (▀).

81 Chemical ionisation assessment of -terpineol in negative mode revealed an [M-1]- peak of m/z = 153, indicative of the alcohol. When (E,E)-FPP was used as substrate, only small amounts of one compound, -bisabolene, was produced (data not shown). Under the chosen assay conditions, the catalytic activity and product profile of SamonoTPS1 on either substrate was not substantially affected by exchanging the Mn2+ metal ion with Mg2+

(Table 5.1).

Table 5.1. Product profile of SamonoTPS1 when incubated with GPP in the presence of either Mn2+ or Mg2+ ions. Composition data based on GC-FID analysis using HP5-MS column. 1Taken from MassFinder terpenoid RI library and Adams [39]. RIs were also compared to an in-house monoterpene standard collection. Composition (%) RI Lit. RI1 RI Mn2+ Mg2+ Compound (HP5-MS) (DB-5) (DB-WAX) 1 (-)--pinene 935 936 1013 3.3 3.3 2 (+)-sabinene 974 973 1112 3.0 2.6 3 myrcene 992 987 1158 3.4 5.5 4 (-)-limonene 1030 1025 1188 33.6 35.3 5 -terpinolene 1098 1084 1236 1.6 1.0 6 linalool 1100 1086 1551 5.0 3.1 7 (+)--terpineol 1193 1176 1694 44.0 38.7 8 (E)-geraniol 1224 1235 1849 5.9 10.5

83 Enzyme assays of SasesquiTPS1 with FPP as substrate and Mg2+ ion containing buffer produced a complex mixture of mono- and bicyclic sesquiterpenes. The main products as analysed by GC-MS and GC-FID were the cyclic alcohol germacrene D-4-ol, and helminthogermacrene (Fig. 5 and Table 2). On-column injection with chemical ionization in negative mode using methane as the ionization gas confirmed the assignment of alcohols based upon the presence of [M-1]- of m/z = 221 for germacrene D-4-ol.

High GC injector temperatures resulted in a large proportion of -elemene, and a suite of other bicyclic sesquiterpenes including - and -cadinene. Cold on-column GC-

MS analysis revealed no -elemene, and considerably less of the cadinenes. Several minor components with mass spectra and retention indices matching - and -selinene, - humulene and nerolidol were detected in proportions below 5 % (Fig. 5.5 and Table 5.2).

In the presence of Mn2+, a similar product profile was detected however the overall yield was almost 10-fold lower. When GPP was used as substrate, SasesquiTPS1 only produced very small amounts of linalool, regardless of which metal ion was used (data not shown).

Table 5.2. GC-MS profile of SasesquiTPS1 when incubated with FPP. Composition determined by cold on- column GC-MS analysis as well as GC-FID (splitless injector). 1 MassFinder terpenoid RI library and Adams [39]. † Overlapping peaks. Composition (%) RI Lit. RI1 RI GC-FID On-column Compound (HP-5) (DB-5) (DB-WAX) (HP5MS) (DB-WAX) 1 -elemene 1400 1389 1594 37.6 - 2 -guaiene 1447 1440 - 1.6 - 3 -humulene 1465 1455 1680 2.2 2.2 4 patchoulene? 1485 1473 1695 0.9 - 5 -muurolene 1492 1474 1718 5.9 11.4 6 -selinene 1497 1486 1729 2.7 1.8 7 -selinene 1506 1494 1733 4.1 3.4 8 -bulnesene 1515 1503 1723 4.1† 10.0 9 helminthogermacrene 1516 1503 1772 4.1† 29.4 10 -cadinene 1527 1507 1764 1.4 - 11 -cadinene 1532 1520 1768 3.8 - 12 E-nerolidol 1570 1550 2037 0.7 2.5 13 germacrene D-4-ol 1590 1574 2057 19.1 38.9

Figure 5.5. Mass spectra of sesquiterpenoids identified in the product mixture of SasesquiTPS1 when incubated with FPP in the presence of Mg2+. A: -humulene, B: -muurolene, C: -selinene, D: -selinene, E: -bulnesene, F: helminthogermacrene, G: trans-nerolidol, H: germacrene D-4-ol, I: GC-MS chromatogram (cold on-column injection, DB-WAX) with peak numbering for reference.

85 Discussion

With the plantation sandalwood industry growing rapidly in Asia and Australia, there is a substantial need for well-characterized germplasm. Knowledge and utilization of the genes involved in essential oil production may advance efforts towards sandalwood tree improvement, further development of sustainable of sandalwood plantations, and thereby conservation efforts to protect sandalwood trees in natural forests. In sandalwood trees, some environmental stimuli may induce terpenoid production, while specific genotypes of sandalwood may have a predisposition to high rates of heartwood oil deposition. Cloning of TPS involved in sandalwood oil production may enable a better molecular characterization of these genotypes and a better understanding of the molecular processes of phenotypic variability. In addition, the cloning of genes in sandalwood terpenoid biosynthesis may also enable the microbial production of sandalwood oil compounds.

We describe the cDNA cloning and functional characterization of two TPS genes,

SamonoTPS1 and SasesquiTPS1, isolated from leaf and wood tissues of the essential oil- producing sandalwood tree S. album. Functional characterisation of SamonoTPS1 and

SasesquiTPS1 identified both enzymes as multi-product mono- and sesquiterpene synthases, respectively. The terpenoids produced by SamonoTPS1 and SasesquiTPS1 are not found in substantial quantities in the distilled oil of sandalwood, suggesting that additional TPS genes are expressed in S. album which may contribute to the majority of heartwood essential oil. Since TPS make up large gene families (Bohlmann et al. 1998;

Martin et al. 2004) and since cloned TPS can be employed for the characterization of terpenoid modifying cytochrome P450 enzymes (Ro et al. 2005), cloning of the first TPS genes from S. album will enable the discovery of additional TPS as well as other metabolic pathway enzymes for sandalwood essential oil biosynthesis. With the long-term goal of sandalwood tree improvement and enhanced production of sandalwood oil, our ongoing

86 cloning efforts are focusing on the TPS genes expressed in the heartwood-sapwood transition zone.

Based on sequence inference, SamonoTPS1 is a member of the TPS-b subfamily and SasesquiTPS1 is a member of the TPS-a subfamily of the plant TPS gene family

(Bohlmann et al. 1998; Dudareva et al. 2003; Martin and Bohlmann 2004; Martin et al.

2004). Both SamonoTPS1 and SasesquiTPS1 share highest similarity with known mono-

TPS and sesqui-TPS, respectively, from grapevine. SamonoTPS1 shares 45% identity and

66% similarity with grape ()--terpineol synthase (Martin and Bohlmann 2004), while sesquiterpene synthase SasesquiTPS1 shares 59 % identity and 76 % similarity with grapevine (+)-valencene synthase (Lucker et al. 2004). The next most closely related TPS gene sequences to SasesquiTPS1 are from poplar and cotton. These similarities are reasonable given the relatively close phylogenetic relationship between the Santalaceae family and the Rosid subclass of angiosperms, which includes the Vitales, Malpighiales and Malvales orders (Bremer et al. 2003).

Figure 5.6 outlines the likely mechanisms for transformation of GPP into monoterpenes found in the product profile of SamonoTPS1. These transformations follow similar mechanisms to those described for previously characterised plant monoterpene synthases (Wise and Croteau 1999). Geraniol may be produced from mild acid hydrolysis of GPP or through water quenching of the initial carbocation intermediate (Yang et al.

2005). Intermediate (9) is responsible for the formation of a myriad of monoterpene compounds (Bohlmann et al. 1997; Hyatt and Croteau 2007). erpineol may result from nucleophilic attack of (9) by water. The circumstances in which water is available to the active site are not known, and future efforts of enzyme structure analysis are needed to resolve this question. Some clues may come from a comparison with 1,8-cineole synthase from Salvia fruticosa, where water is presumed to access the active site through

87 deformation of the 14 helix. Water may be H-bonded to a neighbouring Asn residue

(Kampranis et al. 2007). Although an Asn residue is not present in the corresponding region in SamonoTPS1, a Gln residue is located slightly downstream at position 346 which may play a similar role.

Figure 5.6. Scheme showing hypothesised transformation of GPP by SamonoTPS1.

88 Biosynthesis of the sesquiterpenes catalysed by SasesquiTPS1 can be explained by several rearrangements of a germacrenyl cation (Fig. 7). Cyclisation of FPP between C1 and C10 affords the (E,E)-germacradienyl cation intermediate (1). However, FPP may be converted to (6E)-nerolidyl diphosphate (NPP) prior to cyclisation between C1 and C10, leading to the (Z,E)-germacradienyl cation (2) (Cane 1990). As sesquiterpenes derived from both (E,E)-FPP and (6E)-NPP are found in the product mixture of SasesquiTPS1, the active site must accommodate both configurations. Germacrene D-4-ol may be formed through several possible mechanisms, requiring further investigation, all involving water attacking the carbocation. To our knowledge, germacrene D-4-ol as the product of a TPS enzyme has previously only been found with an engineered TPS enzyme (Yoshikuni et al.

2006).

Intermediate (1) may lose a proton from C12, resulting in the formation of germacrene A (6). No free germacrene A was found in the product mixture, even under cold on-column conditions. Under hot injector conditions, -elemene was a major product, indicating operation of a thermocyclic Cope rearrangement of (6) to -elemene. -

Elemene is rarely found in nature (Herout 1971) and is generally accepted to be an artefact of high injector temperatures during GC-analysis (de Kraker et al. 1998). Furthermore, under hot injector conditions there was a distinct increase in both  and -cadinene, and a decrease in -muurolene compared to cold on-column injection. There is the possibility of a dehydration route from germacrene D-4-ol to -and -cadinene via intermediates (3) and

(4) under hot GC conditions, given their absence under milder GC conditions (broken arrows in Fig. 5.7).

Figure 5.7. Scheme showing hypothesised biosynthetic transformations of farnesyl and nerolidyl diphosphate (boxed) into compounds found in the product mixture of SasesquiTPS1. Broken arrows indicate possible heat induced dehydration and rearrangement reactions.

Formation of another component, -bulnesene, can only occur through the re- ionisation of (6), which requires the addition of a proton to C3. The majority of germacrene A may not leave the active site, as its hydrophobic nature would prevent re- entry. Nonetheless, the role of germacrene A as an intermediate in sesquiterpene

90 biosynthesis is well established (Cane and Tsantrizos 1996). Concerted ring closure between C2 and C6 generates a tertiary carbocation at C7 (7), a transformation which has been studied in tobacco epi-aristolochene A synthase (TEAS). Addition of a proton from

Y520 of TEAS to germacrene A whilst located at the active site enables further transformations to occur via the eudesmane carbocation (Rising et al. 2000). This may be occurring in SasesquiTPS1, however the exact location of the proton donor remains unknown. Based on the protein alignment in Figure 3, Y538 is likely the same residue in

SasesquiTPS1 and would be a potential candidate for site-directed mutational analysis to test this concept. Also present in the product mixture of SasesquiTPS1 are - and - selinene. Both of these compounds may be formed from free germacrene A under mildly acidic conditions (de Kraker et al. 1998) however enzyme catalysed conversion of (6) to the selinenes is possible.

In conclusion, we have described the first cloning and functional characterization of two multi-product TPS genes from sandalwood. This work is relevant for research in an important tropical tree species, as it enables the future cloning and characterization of other members of the TPS gene family and potentially members of the cytochrome P450 family involved in essential oil biosynthesis in S. album and related species. The results of this work may have important long-term implications for sandalwood plantations, silviculture, and tree breeding as well as for the fragrance industry.

Experimental

Chemicals and reagents

All reagents, solvents, antibiotics and precursor chemicals were purchased from commercial sources. Farnesyl diphosphate and geranyl diphosphate were from Sigma (St.

Louis, MO, USA). Restriction endonucleases and T4 DNA ligase were from New England

91 Biolabs (Ipswich, MA). Chiral standards (+) and ()-pinenes, (+) and ()-limonenes were purchased from Aldrich, ()-linalool from Fluka ()--terpineol from SAFC (via

Sigma/Aldrich) and ()-sabinene was purchased from ChromaDex (Irvine, CA).

Plant material collection, DNA and RNA extraction

Five increment cores were taken from three mature Santalum album trees growing in Perth, Western Australia. Cores went through the diameter of the trunk at approximately

30 cm above ground level. Once taken from the tree, the core was snap frozen in liquid nitrogen and stored in a cryogenic shipper. Mature and emerging leaves were also collected, frozen and stored at -150°C. The samples were transported to UBC Vancouver,

Canada for nucleic acid extraction.

Genomic DNA was extracted from leaves using the technique outlined by Byrne et al. (2003b). Leaves (1 g) were ground in a chilled mortar and pestle with liquid nitrogen before being added to DNA extraction buffer (0.18 M sorbitol, 5% polyethylene glycol, 25 mM Tris pH 7.5, 2.5 mM EDTA pH 8, 0.05% BSA, 0.025% spermidine, 0.1% NaSO3 and

0.5 mg l-1 -mercaptoethanol). Total RNA was extracted from leaves and wood using a modified version of Kolosova et al. (2004). Leaves (3 g) were ground in liquid nitrogen and added to RNA extraction buffer (200 mM Tris-HCl, pH 8.5, 1.5% lithium dodecyl sulphate, 300 mM LiCl, 10 mM EDTA, 1% w/v sodium deoxycholate, 1% w/v Tergitol

Nonidet® P-40). 5 mM Thiourea, 1 mM aurintricarboxylic acid, 10 mM dithiothreitol, and

2% (w/v) polyvinylpolypyrrolidone (PVPP) were added just prior to use. Wood cores (15 g) were ground into a fine powder in a coffee grinder with liquid nitrogen and solid CO2 and added to extraction buffer. Several tubes were combined at the TE / NaCl resuspension step to concentrate the RNA sample. All solutions were prepared from DEPC treated and / or autoclaved water.

Primer design and genomic DNA amplification

Alignment of several previously characterised angiosperm sesquiterpene synthases enabled degenerate primer design based on several highly conserved regions, as per Chang et al. (2000). Forward primer p5F (5‟-TGG TGG AAR GAY TTI GAY TTY-3‟) and reverse primer p8R (5‟-DGT WCC RTA CGC RTC RTA CGT RTC RTC-3‟) were based on the WWK(E/D/S)LDF and DD(T/I)YDAYGT motifs, respectively. All primers were purchased from IDT Pty. Ltd (Coralville, IA). Annealing temperatures for all primers were calculated using OligoAnalyser (IDT Pty. Ltd.) and Primer3 (Rozen and Skaletsky 2000).

Degenerate primers p5F and p8R (2 M), along with a 2 x mixture of thermostable Taq

Jumpstart™ DNA polymerase (Invitrogen, Carlsbad CA) were added to 20 ng of genomic

DNA in a final volume of 20 l. The mixture was cycled using the following conditions; initial denaturing step of 4 min at 94°C, then 35 cycles of 2 min at 94°C, 30 s at 50°C, and

1.5 min at 72°C. A final extension time of 7 min completed the cycle. The PCR product was run on a 1% agarose gel in TAE buffer and visualised with SybrSafe® stain

(Invitrogen). A band of 180 bp was gel purified, cloned into Topo T/A pCR 2.1 vector

(Invitrogen) and sequenced using the BigDye™ terminator method. BLAST searches of the 180 bp fragment revealed 2 unique clones (from 4 colonies) with high similarity to - cadinene synthase from cotton.

RACE strategies

Single strand cDNA for 3‟ RACE was generated from S. album leaf and wood total

RNA using the First Choice® RLM-RACE kit (Ambion, Austin TX) as per the manufacturers instructions. 3‟ RACE nested PCR was performed using universal 3‟ end anchor primers and gene-specific primers (outer primer 5‟-AGA CTA GTT GAG GGT

93 TAT TTT TGG GTA C-3‟, inner primer 5‟-TTG AGC CAC AGT ACT CCC TTG CAC

G-3‟). The 1.1 kb PCR product was gel purified, cloned into pCR 2.1 vectors and sequenced. The deduced amino acid sequence from this gene fragment closely resembled other previously published sesquiterpene synthases and was designated SesquiTPS-3. PCR program was as follows: initial denature step of 2 min 94°C, then 35 cycles at 94°C for 30 s, annealing temperature of 57°C for 30 s (increased to 60°C in round 2) followed by 2 min extension at 72°C.

5‟ RACE was performed using SMART™ RACE (Clontech, Mountain View CA).

Leaf and wood total RNA was reverse transcribed with PowerScript™ M-MLV reverse transcriptase and SMART II™ oligonucleotide as per the manufacturers instructions. The resulting single strand cDNA was used as a template for nested PCR with universal 5‟ end anchor primers and gene specific primers (outer primer 5‟-AGC TCA TCT ACT GTA

CCG TAT GCA TC-3‟, inner primer 5‟-CGT GCA AGG GAG TAC TGT GGC TCA

AAG-3‟). A BLAST search of the cloned and sequenced 1.1 kb 5‟ gene fragment closely resembled published monoterpene synthases and was designated MonoTPS-5‟. This fragment was probably from another gene which had serendipitously been amplified instead of the original putative sesquiterpene synthase gene. The nucleotide sequence translates to the FEPQY motif present in both S. album TPS genes (Figs 2 and 3). A second pair of nested gene specific primers (outer primer 5‟-CCA AGC TCC CAA CAT

CCC ACC-3‟, inner primer 5‟-CTT TAG CGT AGT GCA TCC GAA AC-3‟) were used to amplify a 1.2 kb product. The cloned and sequenced gene fragment was homologous to previously published sesquiterpene synthases, and designated SesquiTPS-5‟. Concurrently,

3‟ RACE using monoterpene synthase gene specific primer 5‟-CAA TCA ATT CAT CAG

CAG GAG CTT GGC AAC CTA GC-3‟ and 3‟ RACE universal anchor primers revealed a

94 1.2 kb product which was homologous to the 3‟ end of other published angiosperm monoterpene synthase genes. This 3‟ end fragment was designated MonoTPS-3‟.

Full length clones of both the putative monoterpene and sesquiterpene synthases were prepared from SMART™ RACE cDNA template. For the putative sesquiterpene synthase gene, a pair of nested start primers (outer forward primer 5‟-GCT CCC GTC TAT

GCT TTG AAG CAT TCC-3‟, inner forward primer 5‟-C ATC ATA CTG AAC TTC TCA

TTT GCC CTC ATG G-3‟) and nested stop primers (outer reverse primer 5‟-CTG AGG

AAA AGC TAT AAA TCC AAC GGA TGC-3‟, inner reverse primer 5‟-TCA AAC AAT

CAT GGG ATC TAT AAG TAG GGA TGC-3‟) were used to amplify an open reading frame of 1680 bp incorporating start and stop codons (underlined). Nested PCR conditions were as follows: 94°C for 2 min, then 35 rounds of 94°C for 30 s, annealing for 30 sec at

58°C and extension at 68°C. A final extension period of 7 min at 68°C completed the cycle. Accuprime™ proofreading DNA polymerase was used in all reactions. Likewise, forward (5‟-GCT TGA TCC GCC ATG GAT GCC TTT GCC ACT-3‟) and reverse (5‟-C

TCA ATC CTC CTC GTT CAG TGG AAT AGG GTG GAT G-3‟) primers were used to amplify the putative monoterpene synthase gene from leaf and wood cDNA. One round of

40 cycles of PCR (using identical conditions as the sesquiterpene synthase gene, except annealing temperature was set at 62°C) produced the full length gene with an open reading frame of 1731 bp. Both genes were cloned into Topo pCR II Zero Blunt™ vectors

(Invitrogen).

Bacterial expression and protein isolation

Both putative terpene synthase genes were cloned into a pET28b (+) expression vector (Novagen, San Diego CA) using sticky end PCR (Zeng 1998). For the putative sesquiterpene synthase, sesquiTPS1 primers F1 (5‟-TATG GAA AAT CAA AAA GTG

95 CCT ATT TCT TCT GTC C-3‟) and R1 (5‟-T TCA AAC AAT CAT GGG ATC TAT

AAG TAG GGA TGC AAT G-3‟) were cycled 35 times with Accuprime™ proofreading

DNA polymerase using the cloned full length gene as template. PCR conditions were as follows, denature for 2 min at 94°C, then 35 cycles of 30 sec 94°C, annealing for 30 sec at

58°C and extension for 2 min at 68°C. For the second round, identical conditions were used with primers F2 (5‟-TGG AAA ATC AAA AAG TGC CTA TTT CTT CTG TCC) and R2 (5‟-AG CTT TCA AAC AAT CAT GGG ATC TAT AAG TAG GGA TGC-3‟).

Two micrograms of DNA from each PCR product were combined and denatured at 95°C for 5 min, then re-annealed at 20°C for 20 min. This mixture now contained hybridised

PCR product bearing NdeI and HindIII overhangs suitable for ligation into similarly digested pET28b (+) vector. Ligation of the sticky end PCR product was done at 16°C overnight with T4 DNA ligase. The circular plasmid, which contained the sesquiterpene synthase gene in frame with a Ni-affinity tag (His6) was first transformed into Top 10 chemically competent cells and grown on LB plates with kanamycin (50 g ml-1).

Directional cloning of the putative monoterpene synthase gene, monoTPS1 was carried out using the same method, except primers F1 (5‟-T ATG GAT GCC TTT GCC

ACT TCT CCG ACC TC-3‟), R1 (5‟-TTC AAT CCT CCT CGT TCA GTG GAA TAG

GGT GGA TG-3‟) and F2 (5‟-TGG ATG CCT TTG CCA CTT CTC CGA CCT C-3‟), R2

(5‟-AGC TTT CAA TCC TCC TCG TTC AGT GGA ATA GGG TG-3‟) were used to create sticky ends suitable for ligation into NdeI and HindIII digested pET28b (+) vector.

PCR conditions were the same as for the sesquiterpene synthase gene, except the annealing temperature was raised to 64°C.

For both genes, transformants were grown on selective LB plates (kanamycin 50 g ml-1), and the DNA extracted via plasmid preparation (Invitrogen). pET28b (+) vectors containing the insert were sequence verified and transformed into chemically competent

96 C41 E. coli cells (Avidis, Saint-Beauzire, France) containing the pRARE 2 plasmid isolated from Rosetta 2 competent cells (Novagen). Colonies were grown on LB plates containing kanamycin and chloramphenicol (50 g ml-1). Three independent colonies were picked and grown in a shaker overnight at 37°C in 5 ml of LB with the same antibiotics. 2 ml of this culture was added to 400 ml of TB containing kanamycin and chloramphenicol. The cell suspension was grown for 4 hours with shaking at 37°C until the OD600 = 0.8. Isopropyl--

D-thiogalacto-pyranoside (IPTG) was added to a final concentration of 0.2 mM and the mixture was shaken overnight at 16°C. Cell suspension was harvested by centrifugation at

4°C and the cell pellet frozen at -80°C for future use.

Cell pellets (~7g) were resuspended in 15 ml of cold binding buffer (20 mM

HEPES pH 7.5, 150 mM NaCl and 20 mM imidazole), 1 ml of 10 mg ml-1 lysozyme in binding buffer and 500 l of 200 mM PMSF. While on ice, the mixture was stirred thoroughly with a glass rod for 30 min. The lysate was then sonicated on ice using a

Bronson ultrasonic probe until translucent. The lysate was centrifuged at 20 000 x g at 2°C for 30 min. Supernatant was collected and filtered through a 0.22 m syringe filter on ice.

Protein purification and enzyme assays

The cleared lysate (~20 ml) was loaded onto an ÄKTApurifier 10 FPLC with a 1 ml

HisTrap HP column (GE Healthcare, Piscataway NJ, USA). After washing with 60 ml binding buffer, the proton was eluted with elution buffer (20 mM HEPES pH 7.5, 150 mM

NaCl and 350 mM imidazole) and then desalted on a BioRad Econo-Pac 10 DG desalting column (BioRad, Hercules CA, USA) into elution buffer which lacked imidazole. The remaining eluent (4 ml) was spun in a 30 kDa Amicon Ultra-4 Centrifugal Filter Unit

(Millipore, Billerica MA, USA) to concentrate the protein. Aliquots of the purified proteins were frozen in liquid N2 and stored at -80°C until use. Proteins were analysed

97 using SDS-PAGE, western blotting and immunolabelling with His6 antibody (Sigma).

Antibody was visualised by the NBT/BCIP assay (Roche, USA).

Enzyme assays for both recombinant proteins were done in triplicate using GC vials as per O‟Maille et al. (2004). For each assay, 10 l of protein (25 mg ml-1) was added to a reaction buffer containing 25 mM HEPES, 10 % glycerol 5 mM DTT and 10 mM of either

Mg2+ or Mn2+. 25 l of substrates FPP and GPP (~1 mg ml-1) were added and the mixture gently shaken. Final enzyme concentration was 0.5 mg ml-1. Vials were overlaid with 500

l of pentane to trap volatile products and incubated at 30°C for 1 h.

GC-MS analysis and product identification

Product mixtures of both enzymes were analysed on an Agilent 6890 GC with a

5973 MSD using helium as carrier gas. Two columns were used for peak identification;

DB-WAX capillary column (30 m, 0.25 mm ID, 0.25 m film thickness) and an HP5-MS capillary column (30 m, 0.25 mm ID, 0.25 mm film thickness). For the DB-WAX column, injector temperature was 240°C, detector temperature was set at 250°C, injection volume 1

l and column flow rate was 1 ml min-1. Oven temperature program was as follows; held for 3 min at 40°C, then 3°C min-1 to 110°C, 10°C min-1 to 180°C, then a final ramp of 15

°C min-1 to 240°C where this temperature was held for 5 min. For the HP5-MS column, conditions were the same, except the oven program was held for 1 min at 40°C, then ramped at 7.5°C min-1 to 250°C and held. All mass spectra were detected at 70 eV in multiple ion scan mode.

Cold on-column GC-MS analysis of SasesquiTPS1 product mixture was performed on a DB-WAX column using the same instrument with the following conditions: Injector initial temperature was 20°C (cooled below ambient with liquid CO2) and tracked with the

98 oven temperature program; initial 20°C for 3 min, then ramped at 8°C min-1 to 240°C and held for 10 min. Injection volume 1 l and column flow rate was 1 ml min-1.

Percent composition of the enzyme product mixtures was confirmed by GD-FID using splitless injection of 1 l onto a 30 m HP5-MS capillary column. Conditions were as follows: Injector 250°C, detector 250°C, oven initial temperature was 40°C for 3 mins, then ramped at 8°C min-1 to 250°C and held for 10 min. Column flow was 1 ml min-1. Chiral

GC analysis of SaMonoTPS1 was by GC-MS using a 30 m Cyclodex B capillary column and splitless injection (1 μl). Injector: 230°C, detector 230°C, oven 55°C for 1 min, then ramped at 10°C min-1 to 230°C, and held for 10 min. Column flow was 1 ml min-1.

The identities of products from SamonoTPS1 incubations were confirmed by comparison of retention times to an in-house collection of authentic chiral standards using

HP5, DB-WAX and Cyclodex B stationary phases. Mass spectra were compared to the

NIST 2005 library and (Adams 1995). Sesquiterpene product identification was more difficult as several compounds share identical fragmentation patterns. Since authentic standards were unavailable the product mixtures were run on both stationary phases and the retention indices recorded. These were compared to published retention indices e.g.

(Adams 1995; Pala-Paul et al. 2005) and MassFinder (www.massfinder.com). All spectra were compared to the NIST 2005 spectral library. Likely fragmentation patterns of germacrene D-4-ol were interpreted, and were in agreement with the major peaks found in the spectra (Fig. 5 H) as well as those found by Yoshikuni et al. (Yoshikuni et al. 2006).

99 Acknowledgements

This research was funded by the Australian Research Council and the Forest

Products Commission of Western Australia as part of linkage project LP0454919. Work in the laboratory of J.B. was supported in part with funds from the Natural Sciences and

Engineering Research Council of Canada (NSERC). CGJ is grateful for scholarships received for research in the UBC Michael Smith Laboratories in Vancouver Canada, and to complete this work, particularly the DAFF Science and Innovation Award for 2005, the

Mike Carroll travelling fellowship and the UWA Post Graduate Students Association travel award. We thank Britta Hamberger, Harpreet K. Sandhu and Lina Madilao, all are in the

UBC Michael Smith Laboratories, for excellent technical assistance.

References

Adams RP (1995) 'Identification of essential oil components by gas chromatography / mass spectrometry.' (Allured Publishing Corporation: Carol Stream, Illinois)

Arimura G-I, Huber DPW, Bohlmann J (2004) Forest tent caterpillars (Malacosoma disstria) induce local and systemic diurnal emissions of terpenoid volatiles in hybrid poplar (Populus trichocarpa x deltoides): cDNA cloning, functional characterization and patterns of gene expression of (-)-germacrene D synthase PtdTPS1. The Plant Journal 37, 603-616.

Bohlmann J, Meyer-Gauen G, Croteau R (1998) Plant terpenoid synthases: Molecular biology and phylogenetic analysis. Proc. Natl. Acad. Sci. USA 95, 4126-4133.

Bohlmann J, Steele CL, Croteau R (1997) Monoterpene synthases from grand fir (Abies grandis). J. Biol. Chem. 272, 21784-21792.

Bremer B, Bremer K, Chase MW, Reveal JL, Soltis DE, Soltis PS, Stevens PF (2003) An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141, 399–436.

Byrne M, MacDonald B, Brand JE (2003) Regional genetic differentiation in Western Australian sandalwood (Santalum spicatum) as revealed by nuclear RFLP analysis. Theor. Appl. Genet. 107, 1208-1214.

Cane DE (1990) Enzymatic formation of sesquiterpenes. Chemical Reviews 90, 1089-1103.

100 Cane DE, Tsantrizos YS (1996) Aristolochene synthase. Elucidation of the cryptic germacrene A synthase activity using the anomalous substrate dihydrofarnesyl diphosphate. Journal of the American Chemical Society 118, 10037-10040.

Chang Y-J, Song S-H, Park S-H, Kim S-U (2000) Amorpha-4,11-diene synthase of Artemisia annua: cDNA isolation and bacterial expression of a terpene synthase involved in artemisinin biosynthesis. Arch. Biochem. Biophys. 383, 178-184.

Davis EM, Croteau R (2000) Cyclization enzymes in the biosynthesis of monoterpenes, sesquiterpenes, and diterpenes. Topics in Current Chemistry 209, 53-95. de Kraker J-W, Franssen MCR, de Groot A, Konig WA, Bouwmeester HJ (1998) Germacrene A biosynthesis; the committed step in the biosynthesis of bitter sesquiterpene lactones in chicory. Plant Physiology 117, 1381-1392.

Deguerry F, Pastore L, Wu S, Clark A, Chappell J, Schalk M (2006) The diverse sesquiterpene profile of patchouli, Pogostemon cablin, is correlated with a limited number of sesquiterpene synthases. Arch. Biochem. Biophys. 454, 123-136.

Dudareva N, Martin D, Kish CM, Kolosova N, Gorenstein N, Faldt J, Miller B, Bohlmann J (2003) (E)--ocimene and myrcene synthase genes of floral scent biosynthesis in snapdragon: function and expression of three terpene synthase genes of a new terpene synthase subfamily. The Plant Cell 15, 1227–1241.

Facchini P, Chappell J (1992) Gene family for an elicitor-induced sesquiterpene cyclase in tobacco. Proc. Natl. Acad. Sci. USA. 89, 11088-11092.

Gershenzon J, Dudareva N (2007) The function of terpene natural products in the natural world. Nat. Chem. Biol. 3, 408-414.

Herout V (1971) Biochemistry of sesquiterpenoids. In 'Aspects of terpenoid chemistry and biochemistry'. (Ed. TW Goodwin) pp. 53-94. (Academic Press: London)

Howes M-JR, Simmonds MSJ, Kite GC (2004) Evaluation of the quality of sandalwood essential oils by gas chromatography-mass spectrometry. Journal of Chromatography A 1028, 307-312.

Huber DPW, Ralph S, Bohlmann J (2004) Genomic hardwiring and phenotypic plasticity of terpenoid-based defenses in conifers. J. Chem. Ecol. 30, 2399-2418.

Hyatt DC, Croteau R (2007) Structure of limonene synthase, a simple model for terpenoid cyclase catalysis. Proc. Natl. Acad. Sci. USA 104, 5360–5365.

Jones CG, Plummer JA, Barbour EL (2007) Non-destructive sampling of Indian sandalwood (Santalum album L.) for oil content and composition. J. Essent. Oil. Res. 19, 157-164.

101

Kampranis SC, Ioannidis D, et al. (2007) Rational conversion of substrate and product specificity in a Salvia monoterpene synthase: Structural insights into the evolution of terpene synthase function. The Plant Cell 19, 1994-2005.

Keeling CI, Bohlmann J (2006) Genes, enzymes and chemicals of terpenoid diversity in the constitutive and induced defence of conifers against insects and pathogens. New Phytologist 170, 657-675.

Kim TH, Ito H, Hayashi K, Hasegawa T, Machiguchi T, Yoshida T (2005) Aromatic constituents from the heartwood of Santalum album L. Chemical & Pharmaceutical Bulletin 53, 641-644.

Kollner TG, O‟Maille PE, Gatto N, Boland W, Gershenzon J, Degenhardt J (2006) Two pockets in the active site of maize sesquiterpene synthase TPS4 carry out sequential parts of the reaction scheme resulting in multiple products. Arch. Biochem. Biophys. 448, 83-92.

Kolosova N, Miller B, Ralph S, Ellis B, Douglas C, Ritland K, Bohlmann J (2004) Isolation of high-quality RNA from gymnosperm and angiosperm trees. BioTechniques 36, 821-824.

Lucker J, Bowen P, Bohlmann J (2004) Vitis vinifera terpenoid cyclases: functional identification of two sesquiterpene synthase cDNAs encoding (+)-valencene synthase and (-)-germacrene D synthase and expression of mono- and sesquiterpene synthases in grapevine flowers and berries. Phytochemistry 65, 2649-2659.

Lucker J, El Tamer MK, Schwab W, Verstappen FWA, van der Plas LHW, Bouwmeester HJ, Verhoeven HA (2002) Monoterpene biosynthesis in lemon (Citrus limon) cDNA isolation and functional analysis of four monoterpene synthases. European Journal of Biochemistry 269, 3160–3171.

Martin D, Bohlmann J (2004) Identification of Vitis vinefera (-)--terpineol synthase by in silico screening of full-length cDNA ESTs and functional characterisation of recombinant terpene synthase. Phytochemistry 65, 1223-1229.

Martin D, Faldt J, Bohlmann J (2004) Functional characterization of nine Norway spruce TPS genes and evolution of gymnosperm terpene synthases of the TPS-d subfamily. Plant Physiology 135, 1908–1927.

O‟Maille PE, Chappell J, Noela JP (2004) A single-vial analytical and quantitative gas chromatography–mass spectrometry assay for terpene synthases. Anal. Biochem. 335, 210- 217.

Pala-Paul J, Perez-Alonso MJ, Velasco-Negueruela A, Vadare J, Villa AM, Sanz J, Brophy JJ (2005) Essential oil composition of the different parts of Eryngium bourgatii Gouan from Spain. Journal of Chromatography A 1074, 235-239.

Parker W, Roberts JS, Ramage R (1967) Sesquiterpene biogenesis. Quarterly Reviews 21, 331-363.

102 Rising KA, Starks CM, Noel J, Chappell J (2000) Demonstration of germacrene A as an intermediate in 5-epi-aristolochene synthase catalysis. Journal of the American Chemical Society 122, 1861-1866.

Ro D-K, Arimura G-I, Lau SYW, Piers E, Bohlmann J (2005) Loblolly pine abietadienol / abietadienal oxidase PtAO (CYP720B1) is a multifunctional, multisubstrate cytochrome P450 monooxygenase. Proc. Natl. Acad. Sci. USA 102, 8060-8065.

Rozen S, Skaletsky HJ (2000) Primer3 on the WWW for general users and for biologist programmers. In 'Bioinformatics Methods and Protocols: Methods in Molecular Biology'. (Eds S Krawetz and S Misener) pp. 365-386. (Humana Press: Totowa, NJ)

Starks CM, Back K, Chappell J, Noel JP (1997) Structural basis for cyclic terpene biosynthesis by tobacco 5-epi-aristolochene synthase. Science 277, 1815-1820.

Steele CL, Crock J, Bohlmann J, Croteau R (1998) Sesquiterpene synthases from Grand Fir (Abies grandis). J. Biol. Chem. 273, 2078-2089.

Tholl D, Chen F, Petri J, Gershenzon J, Pichersky E (2005) Two sesquiterpene synthases are responsible for the complex mixture of sesquiterpenes emitted from Arabidopsis flowers. The Plant Journal 42, 757–771.

Williams DC, McGarvey DJ, Katahira EJ, Croteau R (1998) Truncation of limonene synthase preprotein provides a fully active „pseudomature‟ form of this monoterpene cyclase and reveals the function of the amino-terminal arginine pair. Biochemistry 37, 12213-12220.

Wise ML, Croteau R (1999) Biosynthesis of monoterpenes. In 'Comprehensive Natural Products Chemistry'. (Ed. DE Cane) pp. 97-153. (Elseiver: Oxford)

Yang T, Li J, Wang H-X, Zeng Y (2005) A geraniol-synthase gene from Cinnamomum tenuipilum. Phytochemistry 66, 285-293.

Yoshikuni Y, Martin VJJ, Ferrin TE, Keasling JD (2006) Engineering cotton (+)-- cadinene synthase to an altered function: Germacrene D-4-ol synthase. Chemistry & Biology 13, 91-98.

Zeng G (1998) Sticky end PCR: New method for sub-cloning. BioTechniques 25, 206-208.

103 Chapter 6. General discussion

There are several issues affecting the productivity of sandalwood plantations in far northern Western Australia, but overwhelmingly the economic value of S. album is very much dependant on its ability to yield fragrant heartwood. Currently, Australian plantation grown sandalwood aged 10-17 years is highly variable in terms of heartwood content and oil yield, as well as important growth characteristics. This research set out to identify the genetic diversity of the FPC Kununurra S. album collection, and whether the broad genetic groups identified shared similar heartwood characteristics. Essential oil composition was studied intensively, identifying trends in chemical profiles within and between trees. The biosynthesis of sesquiterpenes from sandalwood was also explored using functional gene expression. In this chapter, the significance of the findings to scientists and the plantation industry are discussed, and the remaining gaps in our knowledge indicated. Future research efforts should be aimed at addressing these shortcomings, with possible experiments suggested.

Genetic diversity of the Kununurra collection

This study found that overall genetic diversity of the Kununurra S. album collection was low for a tree species, and the genetic groupings determined by nuclear RFLPs did not possess differing heartwood characteristics. As noted in Chapter 3, low genetic diversity is probably the result of incomplete seed sourcing, propensity toward self pollination and root-suckering upon disturbance, and the evolutionary history of the species. Despite the low genetic variation, highly variable heartwood characteristics are apparent. These results have significant implications for tree improvement and germplasm deployment. Firstly, it is possible that the S. album collection is incomplete, and does not represent the diversity of the species. An earlier study using isozymes identified considerably more polymorphic

104 loci in natural populations of S. album in India (Suma and Balasundaran 2003). Typically isozymes reveal fewer polymorphisms than RFLPs (Newton et al. 1999). It would therefore be prudent to source more germplasm from the Northern Territory, Indonesia,

India, and possibly any other natural or naturalised populations in South East Asia.

Increasing the diversity of an initial breeding population will decrease the incidence of homozygosity in progeny, and create a more robust gene-pool from which selections can be made (El-Kassaby and Ritland 1995). However, genetic diversity alone should not be interpreted as an indication of production forestry potential, as many tree species with low population and overall genetic diversity have been exploited and improved (McDonald et al. 2003).

Selection of superior phenotypes and clonal propagation of this germplasm may prove to be the most effective means of deployment for S. album, however for the foreseeable future seed will continue to be the easiest means of propagating S. album.

Therefore it is important that seed orchards be well planned, maximising the probability of wide crosses and panmixic pollination. This will ensure self-pollination and subsequent homozygosity in the progeny is minimised (Turner et al. 1982). Trees which exhibit superior heartwood and essential oil production may be selected from the breeding pool and the progeny tested.

As discussed in Chapter 3, genetic diversity can be reduced throughout the course of evolution. The Santalum genus appears to have had an Austral origin (Harbaugh and

Baldwin 2007) and it is likely that S. album was derived from an ancestral form in northern

Australia, and achieved its current distribution from a very narrow genetic base through dispersal in the last 3 Ma. Sea level fluctuations and aridity throughout the Plio-

Pleistocene may have resulted in extinction of the ancestral type and extirpation of the species in Australia, limiting gene flow in the island-bound species (Fig 6.1). Seed derived

105 from Timor was not substantially different to Indian material, so it is likely that seed was carried from Timor to the Indian sub-continent by humans in recent millennia.

Figure 6.1. Possible evolutionary origins of S. album (Sa), S. lanceolatum (Sla), S. macgregorii (Sm), S. yasi (Sy) and S. spicatum (Ss). Gene flow is represented by arrows, while bottlenecks are indicated by dashed lines. Ti = Timor, Ba = Babar, F = Flores, S = Sumba, W = Wetar. Dark blue indicates seas deeper than 120m. Geological reconstructions inferred from Hall (2001). Distribution of Santalum inferred from Applegate and McKinnell (1993) Harbaugh (2007) and Paul (1990).

106 The island of Timor is believed to have formed through upward thrust as the

Australian and Asian plates collided, approximately 3 to 5 million years ago (Anonymous

2003; Hall 2002). Seed dispersal across the Timor trough probably occurred during a glacial maximum of the late Pliocene. Harbaugh and Baldwin (2007) show phylogenetic evidence of a shared ancestry between the Fijian sandalwood S. yasi and S. album. This is supported by the ease of hybridisation between the two species (Doran et al. 2005) compared to distant relatives like S. spicatum and S. album (Rugkhla et al. 1997). It seems reasonable, then, to suggest that S. yasi became established on Fiji around the same time as

S. album was established in Timor. A detailed genetic diversity study of naturally occurring S. yasi would also be expected to reveal similarly low genetic diversity, as the circumstances leading to its establishment on Fiji would be equivalent to those of S. album.

In this study, chloroplast RFLPs found polytomy between S. austrocaledonicum, S. macgregorii and S. album samples, meaning lineages could not be inferred. However

Harbaugh and Baldwin (2007) found S. macgregorii to bear a more ancestral haplotype than the northern Australian S. lanceolatum using combined ribosomal and chloroplast markers. Santalum album was found to be more ancestral than all of the abovementioned species, with strong bootstrap support. This may suggest that the common northern ancestor experienced further fragmentation, with the south-east Australian species S. leptocladum (Harbaugh 2007) and S. macgregorii / S. lanceolatum groups being derived from different progenitor to S. album. Range expansion as far north as PNG, followed by retreat may have occurred, leaving relictual populations. The close ancestry shared by S. lanceolatum and S. macgregorii (Harbaugh 2007) indicates gene flow across the Torres

Strait land bridge may have continued for some time, with vicariance and climate changes eventually contributing to speciation.

107 Another important finding in this study was that heartwood content and essential oil yield were not commensurate with the genetic groups identified by nuclear RFLP analysis.

These groups of trees share genotypes, suggesting a common genetic history, either familial or geographical. Despite the sampling limitations owing the variety of tree ages and the number of sampled trees, no group had unique heartwood contents or oil yields.

Substantial variation in these traits was seen across the collection, as well as within genotypes, indicating the growing environment probably plays a major role in determining phenotypes. The genetic component of phenotypic variation would normally be assessed through provenance trials similar to the ones currently under investigation, however in these circumstances host species, host genotype, nitrogen-fixing soil biota and planting density all vary. The extent to which host species and stand density affects growth characteristics, heartwood content and essential oil yield in sandalwood growing in the

Kununurra collection is currently being assessed by the Forest Products Commission. This study, combined with future field trials of clonal lines or seed families under varying host species and planting densities will assist in identifying the environmental circumstances conducive to heartwood formation.

It is also possible that some trees might have a genetic predisposition toward high or low heartwood oil yield, but at a level which is not synonymous with the broad groups identified by anonymous RFLP loci. A more polymorphic marker such as SSRs, could aid in the location of quantitative trait loci (QTL) through association mapping, however the long breeding cycle of sandalwood would problematic. Microsatellite polymorphisms based on expressed sequence tag libraries generally tend to offer a much higher probability of linking economic traits to genotypes (Gupta et al. 2003). The application of functional genomics may also result in “perfect markers” – polymorphic loci linked directly to gene function (Varshney et al. 2005).

108 Chemical diversity in S. album

Essential oil content and composition studies on the FPC Kununurra collection revealed highly variable heartwood contents, ranging from almost none to nearly 70% in cross-sectional area. Extractable oil yield was somewhat commensurate with heartwood content (Chapter 4) and followed a general linear trend. Heartwood is a poorly understood aspect of tree physiology. Its onset and development may be influenced by several complex physiological processes (Hillis 1987; Plomion et al. 2001). In the case of sandalwood, essential oils were only found in heartwood and although the concentration of oil per gram of wood varied, the amount of heartwood contributed most to overall yield.

Therefore any effort to increase the oil yield of trees, should probably be aimed at increasing heartwood content. Heritability of heartwood formation appears to vary with species, ranging from modest (0.3 to 0.4) in black walnut, Juglans nigra (Woeste 2002) to high (0.6-0.99) for Larch (Larix sp.) (Paques 2001). Heritability studies into heartwood content need to be done for S. album. Careful planning of the field trial is essential as the growing environment is certain to play a significant role in heartwood development, and this is further complicated by the parasitic nature of sandalwood.

Essential oil profiles of wood cores extracted from S. album trees growing in the

FPC Kununurra collection were remarkably similar. Over 50 compounds were detected by

GC-FID, but only ten compounds made up the bulk of the oil. Proportions of each compound varied little across the 100 trees indicating the enzymes responsible for oil biosynthesis were highly conserved. Chemical variation has been used as a taxonomic indicator in several genera of plants (Adams 1994; Li et al. 1995; Pureswaran et al. 2004).

Less commonly, differences are found within a species, but the literature seems to indicate this is more common in divergent populations of short-lived perennials or annuals

(Egerton-Warburton et al. 1998; Hillig 2004; Segal et al. 1987). Nucleotide

109 polymorphisms resulting in changes in amino acid residues at the active sites of terpenoid biosynthesis enzymes ultimately change the catalytic properties of the enzyme (Keeling et al. 2008). While this would certainly result in chemical differences in the phenotype, such mutations have a low probability of occurring. Moreover, throughout the evolution of the

Santalum genus, such mutations could have been deleterious as a specific suite of chemicals may serve as phytoalexin response to a wide variety of pests and diseases.

Changes in these proportions may alter their effectiveness, rendering the plant unfit.

Most intriguing was the co-occurrence of structurally similar sesquiterpenes. When the proportions various sesquiterpenes were plotted against each other, a straight-line correlation through the origin was revealed. One compound would only be present unless the other was also present in a commensurate proportion. The slope of the line towards either axis was an indication of their relative abundance. The santalenes and bergamotene, the curcumenes, bisabolene and bisabolol, and several unidentified sesquiterpenes all showed this trend. This was evidence of shared intermediates, with the chemical mechanism resulting in partitioning of products. Terpene synthases may yield multiple products, in quite specific proportions (Deguerry et al. 2006). Changes in residues near or at the active site would change these proportions, or afford new compounds. Conveniently, this allowed the development of a hypothesis that multiple product forming TPS enzymes were active in the heartwood of S. album. This was tested in Chapter 5 and found to be the case.

It remains uncertain what role sesquiterpenes play in the sandalwoods. It is likely that they serve as a chemical defence against pathogens or wood-specific predators like termites. The fact that there are so many different sesquiterpenes may be the result of broad-spectrum insurance, with the possibility of synergy between compounds

(Gershenzon and Dudareva 2007). Santalum has evolved in tropical savannah and semi-

110 arid environments which are commensurate with termites, and this may be why several species have evolved to produce sesquiterpene-rich heartwood. Several chemical phenotypes may exist in Western Australian sandalwood, S. spicatum. As this species has higher genetic diversity, a wide natural distribution and a long evolutionary history, it is likely that specific chemicals within the heartwood oil will vary from population to population. It may also be that populations from the northern extent of its distribution have evolved to contain more insecticidal chemicals than southern populations given the higher incidence of termites in warmer climates. The chemical diversity of other species of sandalwood should be tested thoroughly, as evolutionary forces may have favoured specific chemotypes. Work on the chemical diversity of Polynesian sandalwoods, S. austrocaledonicum (Bottin et al. 2007) and S. insulare (Butaud et al. 2006) has been initiated, with some interesting results. Other chemicals found in sandalwood leaf and bark tissues are only just beginning to be explored (Kim et al. 2005) and certainly warrant further investigation.

Volatile compounds may also be involved in plant-host interactions. An experiment using the common parasitic weed dodder, Cuscuta pentagona found that volatile emissions are used in host selection (Runyon et al. 2006). Certain chemicals, either emitted by host plants (tomato and wheat), total extractives or individual chemicals, acted as cues for the exploring dodder seedlings. This revealed compelling evidence for interaction between plants, and of great interest to sandalwood host selection mechanisms.

This should be explored further with sandalwood seedlings, and may improve our understanding of early growth performance with respect to host species.

111 Terpenoid biosynthesis in S. album

Heartwood extractives in S. album appear to be localised in ray parenchyma (Fig 1,

Chapter 5), indicating sesquiterpene biosynthesis begins there. Ray parenchyma is the only living tissue in wood, and it serves as a conduit between the cambium and the inner stem

(Zimmermann 1983). Signal molecules may be synthesised there and transported throughout the stem so as to initiate other defence responses (Hudgins and Franceschi

2004). Hence, the heartwood seemed a logical place to attempt extracting transcripts of

TPS genes. Wood cores through the diameter of living S. album trees were used to obtain heartwood cDNA. Using the methods outlined in Chapter 5, two TPS genes were functionally characterised.

The two heterologously expressed enzymes, a monoterpene synthase and a sesquiterpene synthase were responsible for the formation of (+)--terpineol and germacrene D-4-ol respectively, as major products. While these compounds are not found in quantities detectable in the distilled essential oil of S. album, they do however confirm that terpenoid biosynthesis occurs in heartwood. The monoterpene synthase was isolated from transcripts found in both leaf and wood, but seemed to be more abundant in leaf tissue, based on the fact that fewer rounds of PCR were required to amplify the gene. As predicted in Chapter 4, these TPS enzymes were capable of producing several compounds, supporting the hypothesis.

Terpene synthases form a large gene family (Bohlmann et al. 1998), and are represented in both plants and fungi (Cane 1990). Forty TPS genes have been identified in the genome of Arabidopsis thaliana, and most of these have been functionally characterised (Aubourg et al. 2002). Two of these genes are involved in primary metabolism, particularly plant hormone biosynthesis (gibberellins) but the majority appear to be dedicated to secondary metabolism. One can only ponder the number of additional

112 TPS genes in the sandalwood genome, not least those dedicated to essential oil biosynthesis. Based on the work done here, several more TPS genes may be isolated from cDNA and the genome of several species of Santalum. Plans are underway to isolate the santalene synthases from sandalwood, with the long-term aim of better understanding their regulation.

Also important in the biosynthesis of sesquiterpenes in sandalwood is the prenyltransferases, responsible for condensation of isopentyl diphosphate and dimethylallyl diphosphate into GPP, FPP and so on. Regulation of oil production may be regulated at this critical branch point (Gershenzon and Croteau 1990). The main components of sandalwood oil, - and -santalol are probably synthesised through the action of a hydroxylase enzyme. This is clearly a key step in the pathway, and it may be that all 12-

OH compounds found in sandalwood oil are made by this class of enzyme. Cytochrome

P450s have been characterised in several plant species and are often highly substrate specific (Karp et al. 1990; Ro et al. 2005; Ro et al. 2006).

Of great importance to the improvement of sandalwood is understanding sesquiterpene oil biosynthesis regulation. As mentioned earlier, heartwood formation and oil biosynthesis are closely linked, so increasing production of one may result in an increase of the other. Genes and enzymes involved in the formation of phenolics and other extractives in sandalwood heartwood should also be explored, as expression of these transcripts are likely to be commensurate with terpenoid biosynthesis. Several points in the metabolic pathway may be critical to oil production in sandalwood, particularly prenyltransferases, TPS and cytochrome P450s. Regulation may be at a transcriptional level, with promoter sequences differing in high-yielding trees. Evidence of nucleotide variation on transcript levels is seen in the herbaceous annual, Clarkia brewerii (Cseke et al. 1998). The vastly different heartwood oil yields seen in S. album may also be due to

113 genetic predispositions at a nucleotide level. Promoters and transcription factors have been identified as key players in terpenoid biosynthesis regulation (Dudareva et al. 1996; Xu et al. 2004).

Molecular markers based on terpenoid biosynthesis may prove useful in selecting superior genotypes. (Kirst et al. 2004) identified microsatellites based on expressed sequence tags that were closely linked to economically important traits in eucalypts. A functional genomics approach, which examines all genes being expressed in specific tissues over time will vastly improve our understanding of factors controlling oil production in sandalwood. This can be done through construction of an EST library, which enables key genes to be explored in detail. Their expression levels may be monitored in seedling and mature tree experiments using gene microarrays (Kirst et al. 2004). Influence of host species, elicitors such as salicylic acid or methyl jasmonate and abiotic stresses could be tested, and may aid in the optimisation of sandalwood silviculture.

Finally, an important discovery in this research was the application of biotechnology to sandalwood oil biosynthesis. Isolation of TPS genes and their heterologous expression in micro-organisms paves the way for bioreactor technology.

Considerable progress has been made in the development of similar systems for antimalarial Artemisia drugs which are difficult to synthesis chemically (Ro et al. 2006).

Bacteria or yeast systems with appropriate expression vectors may produce all of the necessary enzymes involved in secondary metabolite biosynthesis. Application of this technology to sandalwood has obvious benefits – instead of waiting 25 years for a variable yield, a bioreactor may produce essential oils in just days. Enzymes may be altered using site-directed mutagenesis to tailor the final products. New fragrances could be developed in this manner at a comparatively low cost. Alternatively, key enzymes may be

114 transformed into annual plants, enabling new culinary herbs. The list of possibilities is endless.

Conclusion

This study of the genetic diversity, essential oil chemistry and terpene biosynthesis in S. album has shed new light on the origins of heartwood and essential oil yield variability. More work must be done in several fields of sandalwood production, as this chapter has highlighted. The implications of this work have already aided in seed orchard design and future biotechnological avenues. The areas covered by this thesis dealt with important sandalwood production issues, and are equally relevant in the development of this relatively new crop. Hopefully, it has explored the best of Santalum album.

115 Complete list of references

Adams DR, Bhatnagar SP, Cookson RC (1975) Sesquiterpenes of Santalum album and Santalum spicatum. Phytochemistry 14, 1459-1460.

Adams RP (1994) Geographic variation and systematics of monospermous Juniperus (Cupressaceae) from the Chihuahua desert based on RAPDs and terpenes. Biochem. Syst. Ecol. 22, 699-710.

Adams RP (1995) 'Identification of essential oil components by gas chromatography / mass spectrometry.' (Allured Publishing Corporation: Carol Stream, Illinois)

Aldrich PR, Doebley J (1992) Restriction fragment variation in the nuclear and chloroplast genomes of cultivated and wild Sorghum bicolor. Theor. Appl. Genet. 85, 293-302.

Anonymous (1995) 'Flavours and fragrances of plant origin.' (Food and Agriculture Organization of the United Nations: Rome)

Aubourg S, Lecharny A, Bohlmann J (2002) Genomic analysis of the terpenoid synthase (AtTPS) gene family of Arabidopsis thaliana. Mol. Genet. Genomics 267, 730-745.

116 Austerlitz F, Mariette S, Machon N, Gouyon P-H, Godelle B (2000) Effects of colonization processes on genetic diversity: Differences between annual plants and tree species. Genetics 154, 1309–1321.

Bohlmann J, Meyer-Gauen G, Croteau R (1998) Plant terpenoid synthases: Molecular biology and phylogenetic analysis. Proc. Natl. Acad. Sci. USA 95, 4126-4133.

Bohlmann J, Steele CL, Croteau R (1997) Monoterpene synthases from grand fir (Abies grandis). J. Biol. Chem. 272, 21784-21792.

Bottin L, Vaillant A, Sire P, Cardi C, Bouvet JM (2005a) Isolation and characterization of microsatellite loci in Santalum austrocaledonicum, Santalaceae. Mol. Ecol. Notes 5, 800- 802.

Brand JE (2002) Review of the influence of Acacia species on establishment of sandalwood (Santalum spicatum) in Western Australia. Conservation Science W. Aust. 4, 125-129.

Braun NA, Meier M, Pickenhagen W (2003) Isolation and chiral GC analysis of - bisabolols - trace constituents from the essential oil of Santalum album L. (Santalaceae). J. Essent. Oil. Res. 15, 63-65.

Bristow M, Taylor D, Robson K (2000) Queensland sandalwood (Santalum lanceolatum): Regeneration following harvesting. Sandalwood Research Newsletter 11, 4-8.

Brunke EJ, Vollhardt J, Schumaus G (1995) Cyclosantalal and epicyclosantalal - new sesquiterpene aldehydes from east Indian sandalwood. Flavour and Fragrance Journal 10, 211-219.

Butaud J-F, Raharivelomanana P, Bianchini J-P, Faure R, Gaydou EM (2006) Leaf C- glycosylflavones from Santalum insulare (Santalaceae). Biochem. Syst. Ecol. 34, 433-435.

Butcher PA, Moran GF, Perkins HD (1998) RFLP diversity in the nuclear genome of Acacia mangium. Heredity 81, 205-213.

117

Byrne M, MacDonald B, Brand JE (2003a) Phylogeography and divergence in the chloroplast genome of Western Australian sandalwood (Santalum spicatum). Heredity 91, 389-395.

Byrne M, MacDonald B, Brand JE (2003b) Regional genetic differentiation in Western Australian sandalwood (Santalum spicatum) as revealed by nuclear RFLP analysis. Theor. Appl. Genet. 107, 1208-1214.

Cane DE (1990) Enzymatic formation of sesquiterpenes. Chemical Reviews 90, 1089-1103.

Cane DE, Tsantrizos YS (1996) Aristolochene synthase. Elucidation of the cryptic germacrene A synthase activity using the anomalous substrate dihydrofarnesyl diphosphate. Journal of the American Chemical Society 118, 10037-10040.

Chang MCY, Keasling JD (2006) Production of isoprenoid pharmaceuticals by engineered microbes. Nat. Chem. Biol. 2, 674-681.

Chapuis C, Barthe M, Muller BL, Schulte-Eke KH (1998) Preparation and absolute configuration of (-)-(E)--trans-bergamotenone. Helvetica Chimica Acta 81, 153-162.

Christenson PA, Willis BJ (1980) East Indian sandalwood oil. 2. Stereoselective synthesis of (+/-) - epi--santalene and (+/-) - epi--santalol. J. Org. Chem. 45, 3068-3072.

Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. The Plant Journal 16, 735-743.

Cseke L, Dudareva N, Pichersky E (1998) Structure and evolution of linalool synthase. Mol. Biol. Evol. 15, 1491–1498.

Davis CC, Anderson WR, Wurdack KJ (2005) Gene transfer from a parasitic flowering plant to a fern. Proceedings of the Royal Society B 272, 2237–2242.

118 Davis EM, Croteau R (2000) Cyclization enzymes in the biosynthesis of monoterpenes, sesquiterpenes, and diterpenes. Topics in Current Chemistry 209, 53-95. de Kraker J-W, Franssen MCR, de Groot A, Konig WA, Bouwmeester HJ (1998) Germacrene A biosynthesis; the committed step in the biosynthesis of bitter sesquiterpene lactones in chicory. Plant Physiology 117, 1381-1392.

Dewick PM (2002) The biosynthesis of C5-C25 terpenoid compounds. Natural Product Reports 19, 181-222.

Doran JC, Thomson L, Brophy JJ, Goldsack RJ, Bulai P, Faka'osi T, Mokoia T (2005) Variation in heartwood oil composition of young sandalwood trees in the South Pacific (Santalum yasi, S. album and F1 hybrids in Fiji, and S. yasi in Tonga and Niue). Sandalwood Research Newsletter, 3-7.

Dudareva N, Cseke L, Blanc VM, Pichersky E (1996) Evolution of floral scent in clarkia: Novel patterns of S-linalool synthase gene expression in the C. breweri flower. The Plant Cell 8, 1137-1148.

Egerton-Warburton LM, Ghisalberti EL, Considine JA (1998) Infraspecific variability in the volatile leaf oils of Chamelaucium uncinatum (Myrtaceae). Biochem. Syst. Ecol. 26, 873-888.

El-Kassaby Y, Ritland K (1995) Impact of selection and breeding on the genetic diversity in Douglas-fir. Biodivers. Conserv. 5, 795-813.

Facchini P, Chappell J (1992) Gene family for an elicitor-induced sesquiterpene cyclase in tobacco. Proc. Natl. Acad. Sci. USA. 89, 11088-11092.

Felsenstien J (1993) PHYLIP (Phylogeny Inference Package). In. (Department of Genetics, University of Washington: Seattle)

119

Gershenzon J (1994) Metabolic costs of terpenoid accumulation in higher plants. J. Chem. Ecol. 20, 1281-1328.

Gershenzon J, Croteau R (1990) Regulation of monoterpene biosynthesis in higher plants. In 'Recent advances in phytochemistry, Vol. 24: Biochemistry of the mevalonic acid pathway to terpenoids.'. (Eds GHN Towers and HA Stafford) pp. 99-160. (Plenium Press: New York)

Gershenzon J, Dudareva N (2007) The function of terpene natural products in the natural world. Nat. Chem. Biol. 3, 408-414.

Glaubitz JC, El-Kassaby Y, Carlson J (2000) Nuclear restriction fragment length polymorphism analysis of genetic diversity in western red cedar. Can. J. Forest Res. 30, 379-389.

Gupta PK, Rustgi S, Sharma S, Singh R, Kumar N, Balyan HS (2003) Transferable EST- SSR markers for the study of polymorphism and genetic diversity in bread wheat. Mol. Genet. Genomics 270, 315-323.

Hall R (2001) Cenozoic reconstructions of SE Asia and the SW Pacific: changing patterns of land and sea. In 'Faunal and floral migrations and evolution in SE Asia-Australasia'. (Eds I Metcalfe, JMB Smith, M Morwood and I Davidson) pp. 35-56. (A.A. Balkema Publishers: Lisse)

Hall R (2002) Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations. J. Asian Earth Sci. 20, 353-431.

Harbaugh DT (2007) A taxanomic revision of Australian northern sandalwood (Santalum lanceolatum, Santalaceae). Aust. Syst. Bot. 20, 409-416.

Harbaugh DT, Baldwin BG (2007) Phylogeny and biogeography of the sandalwoods (Santalum, Santalaceae): Repeated dispersals throughout the Pacific. Am. J. Bot. 64, 1028- 1040.

Herout V (1971) Biochemistry of sesquiterpenoids. In 'Aspects of terpenoid chemistry and biochemistry'. (Ed. TW Goodwin) pp. 53-94. (Academic Press: London)

Hewson HJ, George AS (1984) Santalum. In 'Flora of Australia'. (Australian Government Publishing Service: Canberra)

Hillig KW (2004) A chemotaxonomic analysis of terpenoid variation in Cannabis. Biochem. Syst. Ecol. 32, 875-891.

Hillis WE (1987) 'Heartwood and tree exudates.' (Springer-Verlag: New York)

Hines B, Byrne M (2001) Genetic differentiation between mallee and tree forms in the Eucalyptus loxophleba complex. Heredity 87, 566-572.

120 Hostynek JJ, Magee PS (1997) Fragrance allergens: Classification and ranking by QSAR. Toxicology In Vitro 11, 377-384.

Howes M-JR, Simmonds MSJ, Kite GC (2004) Evaluation of the quality of sandalwood essential oils by gas chromatography-mass spectrometry. Journal of Chromatography A 1028, 307-312.

Huber DPW, Ralph S, Bohlmann J (2004) Genomic hardwiring and phenotypic plasticity of terpenoid-based defenses in conifers. J. Chem. Ecol. 30, 2399-2418.

Hudgins JW, Franceschi VR (2004) Methyl jasmonate-induced ethylene production is responsible for conifer phloem defense responses and reprogramming of stem cambial zone for traumatic resin duct formation. Plant Physiology 135, 2134–2149.

Hyatt DC, Croteau R (2007) Structure of limonene synthase, a simple model for terpenoid cyclase catalysis. Proc. Natl. Acad. Sci. USA 104, 5360–5365.

Janssen B-J, Gardner RC (1989) Localized transient expression of GUS in leaf discs following cocultivation with Agrobacterium. Plant Molecular Biology 14, 61-72.

Jiko LR (1993) Status and current interest in sandalwood in Fiji. In 'Sandalwood in the Pacific Region'. (Ed. FH McKinnell) pp. 13-18. (ACIAR: Canberra)

Jones CG, Plummer JA, Barbour EL (2007) Non-destructive sampling of Indian sandalwood (Santalum album L.) for oil content and composition. J. Essent. Oil. Res. 19, 157-164.

Karp F, Mihaliak CA, Harris JL, Croteau R (1990) Monoterpene biosynthesis: specificity of the hydroxylations of (-)-limonene by enzyme preparations from peppermint (Mentha piperita), spearmint (Mentha spicata) and perilla (Perillia frutescens) leaves. Arch. Biochem. Biophys. 276, 219-226.

Keeling CI, Bohlmann J (2006) Genes, enzymes and chemicals of terpenoid diversity in the constitutive and induced defence of conifers against insects and pathogens. New Phytologist 170, 657-675.

121 Keeling CI, Weisshaar S, Lin RPC, Bohlmann J (2008) Functional plasticity of paralogous diterpene synthases involved in conifer defense. Proc. Natl. Acad. Sci. USA 105, 1085- 1090.

Khan JA, Srivastava P, Singh SK (2004) Efficacy of nested-PCR for the detection of phytoplasma causing spike disease of sandal. Current Science 86, 1530-1533.

Kim TH, Ito H, Hatano T, Takayasu J, Tokuda H, Nishino H, Machiguchi T, Yoshida T (2006) New antitumor sesquiterpenoids from Santalum album of Indian origin. Tetrahedron 62, 6981–6989.

Kim TH, Ito H, Hayashi K, Hasegawa T, Machiguchi T, Yoshida T (2005) Aromatic constituents from the heartwood of Santalum album L. Chemical & Pharmaceutical Bulletin 53, 641-644.

Kirst M, Myburg AA, De Leon JPG, Kirst ME, Scott J, Sederoff R (2004) Coordinated genetic regulation of growth and lignin revealed by quantitative trait locus analysis of cDNA microarray data in an interspecific backcross of Eucalyptus. Plant Physiology 135, 2368-2378.

Kolosova N, Miller B, Ralph S, Ellis B, Douglas C, Ritland K, Bohlmann J (2004) Isolation of high-quality RNA from gymnosperm and angiosperm trees. BioTechniques 36, 821-824.

Lakshmi Sita G (1996) Molecular basis of disease resistance in sandlewood. Denrome 3, 7- 8.

122 Lhuiller E, Butaud J-F, Bouvet J-M (2006) Extensive clonality and strong differentiation in the insular Pacific tree Santalum insulare: Implications for its conservation. Annals of Botany 98, 1061–1072.

Li H, Madden JL, Potts BM (1995) Variation in volatile leaf oils of the Tasmanian Eucalyptus species-1. Subgenus Monocalyptus. Biochem. Syst. Ecol. 23, 299-318.

Loneragan OW (1990) 'Historical Review of Sandalwood (Santalum spicatum) Research in Western Australia.' (Department of Conservation and Land Management: Perth)

McDonald WM, Rawlings M, Butcher PA, Bell JC (2003) Regional divergence and inbreeding in Eucalyptus cladocalyx (Myrtaceae). Aust. J. Bot. 51, 393-403.

McKinnell FH (1990) Status of management and silviculture research of sandalwood in Western Australia and Indonesia. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990'. (Eds L Hamilton and CE Conrad) pp. 19-25. (US Forest Service: Honolulu, Hawaii)

123 Michaux B (2001) Dispersal versus vicariance, artifice rather than contest. In 'Faunal and floral migrations and evolution in SE Asia-Australasia'. (Eds I Metcalfe, JMB Smith, M Morwood and I Davidson) pp. 311-318. (A A Balkema Publishers: Lisse)

Muir K, Byrne M, Barbour EL, Cox MC, Fox JED (2007) High levels of outcrossing in a family trial of Western Australian sandalwood (Santalum spicatum). Silvae Genet. 56, 222- 230.

Munck SL, Croteau R (1990) Purification and characterization of the sesquiterpene cyclase patchoulol synthase from Pogostemon cablin. Arch. Biochem. Biophys. 282, 58-64.

Newton AC, Allnutt TR, Gillies ACM, Lowe AJ, Ennos RA (1999) Molecular phylogeography, intraspecific variation and the conservation of tree species. Trends Ecol. Evol. 14, 140-145.

O‟Maille PE, Chappell J, Noela JP (2004) A single-vial analytical and quantitative gas chromatography–mass spectrometry assay for terpene synthases. Anal. Biochem. 335, 210- 217.

Olsson T, Lindgren D, Li B (2001) Balancing genetic gain and relatedness in seed orchards. Silvae Genet. 50, 222-227.

Outlaw DC, Voelker G (2007) Pliocene climatic change in insular Southeast Asia as an engine of diversification in Ficedula flycatchers. J. Biogeogr. doi:10.1111/j.1365- 2699.2007.01821.x.

Paques LE (2001) Genetic control of heartwood content in larch. Silvae Genet. 50, 69-75.

Parker PG, Snow AA, Schug MD, Booton GC, Fuerst PA (1998) What molecules can tell us about populations: Choosing and using a molecular marker. Ecology 79, 361-382.

Parker W, Roberts JS, Ramage R (1967) Sesquiterpene biogenesis. Quarterly Reviews 21, 331-363.

Paul J (1990) The status of Sandalwood (S. macgregorii) in Papua New Guinea. In 'Proceedings of the Symposium on Sandalwood in the Pacific April 9-11, 1990'. (Eds L Hamilton and CE Conrad) pp. 76-78. (Forest Service, USDA: Honolulu, Hawaii)

124

Peakall R, Smouse P (2001) GenAlEx, Genetic analysis in Excel. Population genetic software for teaching and research. In. (Australian National University: Canberra)

Plomion C, Leprovost G, Stokes A (2001) Wood formation in trees. Plant Physiology 127, 1513-1523.

Powell W, Morgante M, Andre C, Hanafey M, Vogel J, Tingey S, Rafalski A (1996) The comparison of RFLP, RAPD, AFLP and SSR (microsatellite) markers for germplasm analysis. Mol. Breeding 2, 225-238.

Pureswaran DS, Gries R, Borden JH (2004) Quantitative variation in monoterpenes in four species of conifer. Biochem. Syst. Ecol. 32, 1109-1136.

Radomiljac A (1998) The influence of pot host species, seedling age and supplementary nursery nutrition on Santalum album L. (Indian sandalwood) plantation establishment within the Ord River Irrigation Area, Western Australia. Forest Ecology and Management 102, 193-201.

Rai SN, Sarma CR (1990) Depleting sandalwood production and rising prices. The Indian Forester 116, 348-355.

Rai VR, McComb JA (2002) Direct somatic embryogenesis from mature embryos of sandalwood. Plant Cell, Tissue and Organ Culture 69, 65-70.

125 Rieseberg L, Beckstrom-Sternberg S, Doan K (1990) Helianthus annuus ssp. Texanus has chloroplast DNA and nuclear ribosomal RNA genes of Helianthus debilis ssp. cucumerifolius. Proc. Natl. Acad. Sci. USA 87, 593-597.

Rising KA, Starks CM, Noel J, Chappell J (2000) Demonstration of germacrene A as an intermediate in 5-epi-aristolochene synthase catalysis. Journal of the American Chemical Society 122, 1861-1866.

Ro D-K, Paradise EM, et al. (2006) Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature 440, 940-943.

Rohmer M (1999) The discovery of a mevalonate-independent pathway for isoprenoid biosynthesis in bacteria, algae and higher plants. Natural Product Reports 16, 565-574.

Rugkhla A, Jones MGK (1998) Somatic embryogenesis and plantlet formation in Santalum album and S. spicatum. J. Exp. Bot. 49, 563–571.

Rugkhla A, McComb JA, Jones MGK (1997) Intra- and interspecific pollination of Santalum spicatum and S. album. Aust. J. Bot. 45, 1083-1095.

Runyon JB, Mescher MC, De Moraes CM (2006) Volatile chemical cues guide host location and host selection by parasitic plants. Science 313, 1964-1967.

Schnuch A, Uter W, Geier J, Lessmann H, Frosch PJ (2004) Contact allergy to farnesol in 2021 consecutively patch tested patients. Results of the IVDK. Contact Dermatitis 50, 117- 121.

Segal R, Feuerstein I, Danin A (1987) Chemotypes of Artemisia hera-alba in Israel based on their sesquiterpene lactone and essential oil composition. Biochem. Syst. Ecol. 15, 411- 416.

Shankaranarayana KH, Parthasarathi K (1987) On the content and composition of oil from heartwood at different levels in sandal. Indian perfumer 31, 211-214.

126