Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Identification and quantification of surface wax compounds covering aerial organs of selected plant species Racovita, Radu C. 2016

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

Item Metadata

Download

Media
24-ubc_2016_september_racovita_radu.pdf [ 5.68MB ]
Metadata
JSON: 24-1.0308612.json
JSON-LD: 24-1.0308612-ld.json
RDF/XML (Pretty): 24-1.0308612-rdf.xml
RDF/JSON: 24-1.0308612-rdf.json
Turtle: 24-1.0308612-turtle.txt
N-Triples: 24-1.0308612-rdf-ntriples.txt
Original Record: 24-1.0308612-source.json
Full Text
24-1.0308612-fulltext.txt
Citation
24-1.0308612.ris

Full Text

IDENTIFICATION AND QUANTIFICATION OF SURFACE WAX COMPOUNDS COVERING AERIAL ORGANS OF SELECTED PLANT SPECIES by  Radu C. Racovita  Dipl.Eng., Politehnica University, 2008 M.Sc., The University of Western Ontario, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2016  © Radu C. Racovita, 2016   ii Abstract  The cuticle is an external barrier of aerial plant organs that prevents dessication. It is composed of hydrophobic waxes, i.e. complex mixtures of very-long-chain aliphatics, alicyclics and aromatics, lying on top of (epicuticular) and in between (intracuticular) a polyester matrix known as cutin. Wax compositions vary greatly between plant species, organs and tissues, both qualitatively and quantitatively. This thesis describes the identification and quantification of cuticular waxes of three plant species, including structure elucidation of novel compounds, chain length profiling, and wax compound distributions between intracuticular and epicuticular compartments for the first two species. Leaves of Aloe arborescens were found covered with 15 μg/cm2 wax on the adaxial side and 36 μg/cm2 on the abaxial side, with 3:2 and 1:1 ratios between epicuticular and intracuticular wax layers on each side, respectively. Along with ubiquitous wax compounds, three homologous series were identified as novel 3-hydroxy fatty acids (predominantly C28), their methyl esters (predominantly C28), and 2-alkanols (predominantly C31), and their biosynthetic pathways were hypothesized based on structural similarities and homolog distributions. The adaxial side of young and old Phyllostachys aurea leaves was found covered with 1.7 to 1.9 μg/cm2 each of epi- and intracuticular waxes. In addition to typical aliphatics and alicyclics, novel primary amides were identified, with their chain length profile peaking at C30, and found exclusively in the epicuticular waxes, hence near the true plant surface. Flag leaves and peduncles of Triticum aestivum cv. Bethlehem were found covered with 16 and 49 μg/cm2 wax, respectively, dominated by 1-alkanols in the case of the former and β-diketones and hydroxy-β-diketones for the latter. Along with previously reported wax classes, numerous   iii new classes were identified as homologous series: 2-alkanol esters, benzyl esters, phenethyl esters, p-hydroxyphenethyl esters, secondary alcohols, primary/secondary diols and their esters, hydroxy- and oxo-2-alkanol esters, 4-alkylbutan-4-olides, internally methyl-branched alkanes, and 2,4-ketols. Other new compounds were found as single homologs: C33 2,4-diketone, C31 mid-chain β-ketols, C30 mid-chain α-ketols and α-diketone, as well as C31 mid-chain ketones. Biosynthetic pathways are proposed in the thesis for the new compounds, based on common structural features and matching chain length patterns between related compound classes.                   iv Preface  A version of Chapter 2 has been published: Racovita, R.C., Peng, C., Awakawa, T., Abe, I., Jetter, R. Phytochemistry 2015, 113, 183-194. Reproduced with permission from Elsevier. I conducted all experiments, interpreted results and wrote most of the manuscript; co-authors helped with species selection (T.A.,I.A.), synthesized prior standards (C.P.), and contributed to data interpretation and writing (R.J.).  A version of Chapter 3 has been published: Racovita, R.C., Jetter, R. Phytochemistry 2016, Online publication http://dx.doi.org/10.1016/j.phytochem.2016.06.005. Reproduced with permission from Elsevier. I performed all experiments and data interpretation and wrote most of the manuscript; the co-author (R.J.) performed statistical analysis and contributed to data interpretation and writing.  A version of Chapter 4 has been published: Racovita, R.C., Hen-Avivi, S., Fernandez-Moreno, J.-P., Granell, A., Aharoni, A., Jetter, R. Phytochemistry 2016, Online publication http://dx.doi.org/10.1016/j.phytochem.2016.05.003. Reproduced with permission from Elsevier. I carried out most experiments and data interpretation, as well as wrote most of the manuscript; co-authors grew plants and supplied surface wax extracts (S.H.A., J.P.F.M., A.G., A.A.), and contributed to data interpretation and writing (R.J.).    v A version of Chapter 5 has been submitted for publication. I performed the experiments, interpreted results and wrote most of the manuscript; the co-author, Dr. Reinhard Jetter, contributed to figure editing, data interpretation and writing.  A version of Chapter 6 has been submitted for publication. I performed the experiments, interpreted results and wrote most of the manuscript; the co-author, Dr. Reinhard Jetter, contributed to figure editing, data interpretation and writing.  Certain paragraphs from Chapter 7 are concluding remarks from the manuscripts mentioned above. These paragraphs were written by me and edited by Dr. Reinhard Jetter.  The experiments, data analysis, interpretation and writing in the remaining chapters were my own contribution in close consultation with my supervisor, Dr. Reinhard Jetter.   vi Table of contents  Abstract...………………………………………………………………………………………...ii Preface…………………………………………………………………………………………....iv  Table of contents……………………………………………………………………...………....vi List of tables……………………………………………………………………………………. xii List of figures………………………………………………………………………………….. xiii List of abbreviations ................................................................................................................. xvii Acknowledgements .................................................................................................................... xix Dedication .....................................................................................................................................xx Chapter 1: Introduction to the plant cuticle................................................................................1 1.1. The plant cuticle: An adaptation to a dry environment…………………………..…..1 1.2. Chemical diversity of cuticle constituents…………………………………………..….1 1.2.1. Cutin/Cutan…………..……………………………………………………………1 1.2.2. Cuticular waxes……………………………………………………………………3 1.2.2.1. Aliphatics…………………..………………………………………………...3 1.2.2.2. Alicyclics……………………………..………………………………………5 1.2.2.3. Aromatics……………………..……………………………………………...7 1.3. Stratification of cuticle constituents………………………..…………………………..8 1.4. Functions of the plant cuticle…………………………..……………………………...10 1.5. Biosynthesis of VLC aliphatic cuticular waxes…………..…………………………..13 1.5.1. Biosynthesis of ubiquitous aliphatic waxes……………………………………..13 1.5.2. Biosynthesis of β-diketones in Poaceae………………………………………… 16   vii 1.6. Objectives………………..……………………………………………………………...17 Chapter 2: Very-long-chain 3-hydroxy fatty acids, 3-hydroxy fatty acid methyl esters and 2-alkanols from cuticular waxes of Aloe arborescens leaves ....................................................20 2.1. Introduction………………..…………………………………………………………...20 2.2. Experimental……………..…………………………………………………………….23 2.2.1. Preparation of total leaf wax extracts…………………………………………..23 2.2.2. Preparation of epicuticular and intracuticular wax extracts………………… 24 2.2.3. Qualitative and quantitative analyses of wax extracts………………………...24 2.2.4. Synthesis of reference compounds………………………………………………25 2.3. Results………………………………..…………………………………………………26 2.3.1. Structure elucidation of unknown compounds in A. arborescens leaf cuticular wax……………………………………………………………………………………….26 2.3.2 Quantitative compositional analysis of total cuticular wax of A. arborescens leaves.................................................................................................................................34 2.3.3. Distribution of A. arborescens leaf cuticular waxes between adaxial and abaxial surfaces, as well as epicuticular and intracuticular layers…………..……...40 2.4. Discussion…………………..………………………………………………………….. 44 2.4.1. Chain length distributions of ubiquitous wax constituents……………………44  2.4.2. Identification and chain length distributions of novel wax constituents…….. 47  2.4.3. Gradients between epi- and intracuticular wax layers………………………...52 Chapter 3: Composition of the cuticular waxes coating the adaxial side of Phyllostachys aurea leaves: Identification of very-long-chain primary amides .............................................53 3.1. Introduction…………………..………………………………………………………...53   viii 3.2. Experimental……………………..……………………………………………………. 57 3.2.1. Plant material and preparative thin layer chromatography (TLC)…………. 57 3.2.2. Preparation of epicuticular and intracuticular wax extracts………………… 58 3.2.3. Derivatization reactions………………………………………………………….59 3.2.4. Gas chromatography (GC) analysis of wax extracts…………………………..59 3.2.5. Synthesis of C30 amide standard………………………………………………...61 3.3. Results and discussion………………………………..………………………………..61 3.3.1. Identification of very-long-chain (VLC) fatty acid amides in the adaxial cuticular waxes of P. aurea leaves………..………………...……………………….....62 3.3.2. Compound class gradients between the epicuticular and intracuticular layers lining P. aurea leaves……..……………………………………………………...……..68 3.3.3. Individual compound gradients between the epicuticular and intracuticular layers lining P. aurea leaves…………..………………………...……………………...72 3.3.4. Ester isomer profiles in epicuticular and intracuticular layers on P. aurea leaves…………………………………………………………………………………….75 Chapter 4: Composition of cuticular waxes coating flag leaf blades and peduncles of Triticum aestivum cv. Bethlehem………..…………………..………………………………… 81 4.1. Introduction……………..……………………………………………………………...81 4.2. Experimental…………………………………………………………………………...85 4.2.1. Plant material and preparation of total wax extracts………………………….85 4.2.2. Derivatization reactions………………………………………………………….86 4.2.3. GC-MS and GC-FID analysis…………………………………………………...87       4.3. Results……………………………..……………………………………………………88    ix 4.3.1. Overall composition of wheat flag leaf blades and peduncles………………...88 4.3.2. Chain length distributions of common wheat compound classes……………..93 4.3.3. Chain length distributions of wheat wax esters……………………………….. 95  4.3.4. Chain length distributions of esterified alcohols and acids: ester isomer  compositions……………………………………………………………………….....…99 4.3.5. Composition of wheat wax polyketides and terpenoids……………………...107  4.4. Discussion……………..………………………………………………………………111 4.4.1. Wax composition differences between organs and new compound classes…111 4.4.2. Chain length distributions of ubiquitous wax compound classes……………113  4.4.3. Chain length and isomer distributions of aliphatic and aromatic esters……115  4.4.4. Chain length distributions of polyketide wax compound classes……………117 Chapter 5: Novel oxidized compounds and internally methyl-branched alkanes from cuticular waxes of Triticum aestivum cv. Bethlehem…………………...……………………119 5.1. Introduction……………………………………..…………………………………….119 5.2. Experimental……………………………..…………………………………………...122 5.2.1. Plant material…………………………………………………………………...122 5.2.2. Chemicals………………………………………………………………………..122 5.2.3. Preparation of wax extracts……………………………………………………123 5.2.4. Preparative thin layer chromatography………………………………………123 5.2.5. Derivatization reactions………………………………………………………...124 5.2.6. Gas chromatography…………………………………………………………...125 5.3. Results…………………………………………..……………………………………..126   x 5.3.1. TLC separation of cuticular waxes of T. aestivum flag leaf blades and peduncles……………………………………………………………………………….126 5.3.2. Structure elucidation of compound classes A - C……………………………. 128 5.3.3. Structure elucidation of compound classes D and E………………………… 138 5.3.4. Structure elucidation of compound classes F and G………………………… 144 5.3.5. Quantification of new compounds from cuticular waxes of T. aestivum cv. Bethlehem flag leaf blades and peduncles…………………………………………... 149 5.4. Discussion…………………………………………………………………………...... 153 5.4.1. Secondary alcohols, primary/secondary diols and diol esters in flag leaf wax……………………………………………………………………………………...153 5.4.2. Hydroxy-2-alkanol esters and oxo-2-alkanol esters in flag leaf and peduncle wax……………………………………………………………………………………...157 5.4.3. 4-Alkylbutan-4-olides and internally branched alkanes in flag leaf and peduncle waxes………………………………………………………………………...159 Chapter 6: Novel polyketides and polyketide-derived compounds from cuticular waxes of Triticum aestivum cv. Bethlehem …………………………………………….……………… 161 6.1. Introduction………………..………………………………………………………….161 6.2. Experimental………………..………………………………………………………...166 6.2.1. Plant material…………………………………………………………………...166 6.2.2. Wax extracts…………………………………………………………………….166 6.2.3. Preparative thin layer chromatography………………………………………167 6.2.4. Derivatization reactions………………………………………………………..167 6.2.5. Gas chromatography-mass spectrometry……………………………………..169   xi 6.3. Results…………………………..……………………………………………………..169 6.3.1. Structure elucidation of compound classes A – C…………………………….173 6.3.2. Structure elucidation of compound classes D and E………………………… 181 6.3.3. Structure elucidation of compound class F…………………………………...186 6.4. Discussion…………………………………..…………………………………………189 Chapter 7: Conclusions and future research directions…………………………………….198 7.1. Concluding remarks…………………………………………………………………. 198 7.2. Future research directions…………………………………………………………...204 7.2.1. Search for further novel wax compounds in other plant species…………….204 7.2.2. Investigations of some of the proposed biosynthetic pathways……………... 205 References……………………………………………………………………………………...208 Appendices……………………………………………………………………………………..227 Appendix A: Synthetic protocols for reference compounds from Chapter 2……….....227 A.1. General information……………………………………………………………..227 A.2. Synthesis scheme…………………………………………………...…………….227 A.3. Synthetic procedures and 1H NMR characterization data…………………….228           xii List of tables   Table 2.1. Composition of epi- and intracuticular waxes on the adaxial and abaxial sides of A. arborescens leaves……………………………………………………………………43 Table 5.1. Characteristic m/z fragments of trimethylsilyl ethers of secondary alcohols in leaf wax…………………………………………………………………………………………….131 Table 5.2. Characteristic m/z fragments of bis(trimethylsilyl) ethers of primary/secondary diols in leaf wax……………………………………………………………………………………134 Table 5.3. Characteristic m/z fragments of trimethylsilyl ethers of primary/secondary diol esters in leaf wax……………………………………………………………………………………137 Table 5.4. Characteristic m/z fragments of trimethylsilyl ethers of hydroxy-2-alkanol esters in leaf and peduncle wax………………………………………………………………………141 Table 5.5. Characteristic m/z fragments of oxo-2-alkanol esters in leaf and peduncle wax……144 Table 5.6. Characteristic m/z fragments of 4-alkylbutan-4-olides in leaf and peduncle wax…..147 Table 5.7. Characteristic m/z fragments of internally branched alkanes in leaf and peduncle wax…………………………………………………………………………………………….149 Table 6.1. Characteristic m/z fragments of trimethylsilyl ethers of 2,4-ketols in flag leaf and peduncle waxes of Triticum aestivum cv. Bethlehem…………………………………..186       xiii List of figures   Figure 1.1. Structures of most common monomers identified in the structure of cutin .................. 2 Figure 1.2. Structures of most common plant cuticular waxes………………………………………...4 Figure 1.3. 3β-Hydroxycholestane, the structural motif common to all sterols……………………..6 Figure 1.4. Typical oleanane (left) and ursane (right) triterpenoids from plant waxes…………..7 Figure 1.5. Cartoon of hypothetical cross-section through the cuticle……………………………….8 Figure 1.6. Cuticular crystallites formed by hydrocarbon chain alignment………………………..11 Figure 1.7. Biosynthesis of ubiquitous cuticular waxes in the ER of epidermal cells……………..15 Figure 1.8. Proposed biosynthesis of hentriacontan-14,16-dione…………………………………...17 Figure 2.1. Compounds identified in the total wax mixture of A. arborescens leaves…………….27 Figure 2.2. Structure elucidation of unknown compound series in fraction A of A. arborescens wax…………………………………………………………………………………...29 Figure 2.3. Structure elucidation of unknown compound series in fraction B of A. arborescens wax……………………………………………………………………………………………...31 Figure 2.4. Structure elucidation of unknown compound series in fraction C of A. arborescens wax……………………………………………………………………………………………...33 Figure 2.5. Compound class composition of total leaf wax mixture…………………………………35 Figure 2.6. Relative compositions of each compound class in the total leaf wax mixture………..37 Figure 2.7. Relative compositions of esterified acids and alcohols in total leaf wax mixture……39 Figure 2.8. Relative compound class compositions of wax layers on adaxial and abaxial leaf sides…………………………………………………………………………………………….42   xiv Figure 2.9. Proposed biosynthetic pathways to 3-hydroxy acids, 3-hydroxy FAMEs and 2-alkanols………………………………………………………………………………………...51 Figure 3.1. Compounds identified in the adaxial wax mixture of P. aurea leaves………………… 63 Figure 3.2. Structure elucidation of compound series A from the adaxial wax mixture of P. aurea leaves…………………………………………………………………………………………...65 Figure 3.3. Compound class distribution within the epicuticular and intracuticular wax mixtures coating the adaxial side of P. aurea leaves………………………………………………..70 Figure 3.4. Single constituent distribution within compound classes in the epicuticular and intracuticular wax layers on the adaxial side of P. aurea leaves……………………….73 Figure 3.5. Relative isomer compositions of ester homologs in epicuticular and intracuticular waxes on the adaxial side of young and old P. aurea leaves……………………………76 Figure 3.6. Profile of total esterified acids in the epicuticular and intracuticular waxes on the adaxial side of young and old P. aurea leaves……………………………………….78 Figure 3.7. Profile of total esterified alcohols in the epicuticular and intracuticular waxes on the adaxial side of young and old P. aurea leaves……………………………………….80 Figure 4.1. Compounds identified in the total wax mixtures of T. aestivum cv. Bethlehem flag leaf blades and peduncles…………………………………………………………………...89 Figure 4.2. Compound class compositions of total wax mixtures of T. aestivum cv. Bethlehem……………………………………………………………………………………..92 Figure 4.3. Chain length distributions within common compound classes of T. aestivum cv. Bethlehem waxes……………………………………………………………………………..94 Figure 4.4. Chain length distributions within wax ester classes in T. aestivum cv. Bethlehem waxes…………………………………………………………………………………………...96   xv Figure 4.5. Representative mass spectra of aromatic esters………………………………………….97 Figure 4.6. Fragmentation schemes of aromatic esters……………………………………………….98 Figure 4.7. Relative isomer compositions of 1-alkanol ester homologs in T. aestivum cv. Bethlehem waxes…………………………………………………………………... 100 Figure 4.8. Identification of C31 2-alkanol ester metamers………………………………………… 102 Figure 4.9. Relative isomer compositions of 2-alkanol ester homologs in T. aestivum cv. Bethlehem waxes…………………………………………………………………... 103 Figure 4.10. Relative compositions of esterified alcohols in T. aestivum cv. Bethlehem waxes………………………………………………………………………………………….105 Figure 4.11. Relative compositions of esterified acids in T. aestivum cv. Bethlehem waxes………………………………………………………………………………………….106 Figure 4.12. Mass spectra of various derivatives of 8- and 9-hydroxyhentriacontane-14,16-diones…………………………………………………………………………………………108 Figure 4.13. Fragmentation schemes of various derivatives of 8-hydroxyhentriacontane-14,16-dione…………………………………………………………………………………………..109 Figure 4.14. Relative compositions of polyketide and terpenoid  compound classes in T. aestivum cv. Bethlehem waxes………………………………………………………………………..110 Figure 5.1. Oxidized compounds and internally methyl-branched alkanes identified in the wax mixtures of T. aestivum cv. Bethlehem……………………………………………..127 Figure 5.2. Structure elucidation of secondary alcohols in wheat leaf wax………………………130 Figure 5.3. Structure elucidation of primary/secondary diols in wheat leaf wax………………..133 Figure 5.4. Structure elucidation of primary/secondary diol esters in wheat leaf wax………...136   xvi Figure 5.5. Structure elucidation of hydroxy-2-alkanol esters in wheat leaf and peduncle wax…………………………………………………………………………………………….140 Figure 5.6. Structure elucidation of oxo-2-alkanol esters in wheat leaf and peduncle wax…….143 Figure 5.7. Structure elucidation of 4-alkylbutan-4-olides in wheat leaf and peduncle wax…...146 Figure 5.8. Structure elucidation of internally branched alkanes in wheat leaf and peduncle wax…………………………………………………………………………………………….148 Figure 5.9. Total coverages of new compound classes in wheat leaf and peduncle waxes…….150 Figure 5.10. Chain length distributions of new compound classes in wheat leaf and peduncle waxes………………………………………………………………………………………….152 Figure 6.1. Polyketides and polyketide-derived compounds identified in the wax mixtures of T. aestivum cv. Bethlehem……………………………………………………………. 170 Figure 6.2. Identification of mid-chain β-diketone in wheat leaf and peduncle wax ……………172 Figure 6.3. Structure elucidation of mid-chain β-ketols in wheat leaf and peduncle wax………175 Figure 6.4. Structure elucidation of mid-chain α-diketone in wheat leaf and peduncle wax…...177 Figure 6.5. Structure elucidation of mid-chain acyloins in wheat leaf and peduncle wax……..180 Figure 6.6. Structure elucidation of 2,4-diketone in wheat leaf and peduncle wax……………...182 Figure 6.7. Structure elucidation of 2,4-ketols in wheat leaf and peduncle wax…………………185 Figure 6.8. Structure elucidation of ketones in wheat leaf and peduncle wax……………………188 Figure 6.9. Proposed biosynthetic pathway to subterminal -diketone (2,4-diketone) and sub-terminal -ketols (2,4-ketols)……………………………………………………………………………192 Figure 6.10. Proposed biosynthetic pathways to mid-chain ketones, α- and -diketones, and α- and -ketols………………………………………………………………………………………………..197 Figure A.1. Synthesis scheme…...………………………………………………………………228   xvii List of abbreviations  Ac: acetyl ACP: acyl carrier protein br: broad BSTFA: bis-N,O-(trimethylsilyl)trifluoroacetamide CER: Eceriferum CoA: coenzyme A d: doublet dd: doublet of doublets EAR: enoyl-ACP reductase ECR: enoyl-CoA reductase ER: endoplasmic reticulum FA: fatty acid FAE: fatty acid elongase FAME: fatty acid methyl ester FAR: fatty acyl-CoA reductase FAS: fatty acid synthase FATB: fatty acyl-ACP thioesterase B FDR: false discovery rate FID: flame ionization detection GC: gas chromatography HAD: β-hydroxyacyl-ACP dehydratase   xviii HCD: β-hydroxyacyl-CoA dehydratase KAR: β-ketoacyl-ACP reductase KAS: β-ketoacyl-ACP synthase KCR: β-ketoacyl-CoA reductase KCS: β-ketoacyl-CoA synthase LACS: long-chain acyl-CoA synthetase LC: long-chain m: multiplet MAH1: mid-chain alkane hydroxylase 1 MS: mass spectrum/mass spectrometry OS: (S)-2,3-oxidosqualene OSC: oxidosqualene cyclise PKS: polyketide synthase prim: primary Rf: retardation factor s: singlet sec: secondary t: triplet TLC: thin layer chromatography TMS: trimethylsilyl VLC: very-long-chain WS: wax (ester) synthase WSD1: wax ester synthase/acyl-coenzyme A : diacylglycerol acyltransferase 1   xix Acknowledgements  I would like to acknowledge the many people that made this thesis possible. There are no words to express my gratitude to my supervisor, Dr. Reinhard Jetter. I gratefully acknowledge his continuous guidance and support that have kept me going through to the end of my PhD program. I would like to also acknowledge the Jetter lab members, past and present, who have created the environment in which I carried out my research and studies at UBC. In addition, I would like to thank my colleague from the MacLachlan research group, Mr. Hessam Mehr, for his assistance with the acquisition of 1H-NMR spectra of synthetic standards. I must also acknowledge my collaborators on various research projects: Ms. Shelly Hen-Avivi and Dr. Asaph Aharoni (Weizmann Institute of Science, Israel) for kindly supplying wheat wax extracts and for numerous fruitful discussions, Ms. Josefina-Patricia Fernandez-Moreno and Dr. Antonio Granell (Polytechnic University of Valencia, Spain) for discussions regarding the wheat project, Mr. James M. Wigzell and Dr. Jas Pal S. Badyal (Durham University, UK) for fruitful discussions regarding the Phyllostachys aurea project, Dr. Takayoshi Awakawa and Dr. Ikuro Abe (University of Tokyo, Japan) for recommending Aloe arborescens as a species for study, Mr. Douglas Justice (UBC Botanical Garden) for kindly supplying Phyllostachys aurea plants, and Dr. John Kim (UBC Interface Analysis and Reactivity Laboratory) for assistance with ToF-SIMS instrumentation and numerous informative discussions. In addition, I thank Drs. Dan Bizzotto, Allan Bertram and Andrew MacFarlane for their advice and encouragement as members of my guidance committee throughout my PhD program. Lastly, I would like to express my eternal gratitude to my parents, whose love and support, coming from more than an ocean away, have kept me going for all these years.   xx Dedication  To my beloved father and mother   1 Chapter 1: Introduction to the plant cuticle   1.1. The plant cuticle: An adaptation to a dry environment When plants expanded their natural habitat beyond the planetary ocean about 500 million years ago, they had to undergo adaptations allowing them to retain their water reserves for longer times, essential to their survival (Schreiber, 2005). By far the most important among these is the development of a lipid coating over most aerial organs, known as the plant cuticle – a complex biocomposite, with many fine intricacies still undetermined to this day.   1.2. Chemical diversity of cuticle constituents There are two main cuticle components: a polymer matrix (cutin or cutan) and molecular waxes embedded in the cutin (intracuticular waxes) or atop of the cutin (epicuticular waxes).  1.2.1. Cutin/Cutan Unlike cuticular waxes, which are monomeric lipids soluble in a wide range of non-polar solvents, cutin is a polymer insoluble in all common solvents, which greatly limits the analytical tools available for its characterization. However, it can be depolymerized into the corresponding monomers by saponification, for example NaOCH3 (Graça et al., 2002) or BF3-catalyzed (Leide et al., 2011) methanolysis, and its chemical structure can be inferred thereof. The general consensus is that cutin is a three-dimensional polyester composed mainly of long-chain (LC, i.e. C16 and C18) saturated and unsaturated ω-hydroxy fatty acids, often with additional mid-chain epoxy, keto, and/or hydroxyl functionalities from which the 3D structure emerges. Other minor   2 constituents are long-chain mono- and di-carboxylic acids (with or without mid-chain hydroxy-functionalities) and phenolics of the flavonoid family (Laguna et al., 1999). The structures of the most important C16 and C18 monomers are shown in Fig. 1.1. Glycerol has also been identified in the monomer mixture as 1- and 2-monoacyl glycerides by partial saponification with CaO/CH3OH (Graça et al., 2002) and is believed to form cross-links between polyester chains (Pollard et al., 2008).   C16 Cutin MonomersOOHOHOOHOOHOOHOHOHOOHOHOHOOHOOHOHOOHOOHOOOHOOHOOHC18 Cutin MonomersOH OOHOH OOHOOH OOHOOOOHOOHOHOHOOHOHOHOOHOHOHOHOOHOHO Figure 1.1. Structures of the most common monomers identified in the structure of cutin.    3 Cutin is the dominant constituent of the plant cuticle, accounting for 40-80% of total cuticle mass (Heredia, 2003), but in most – although not all plant species (e.g. Ginkgo biloba) (Briggs, 1999) – a second lipophilic polymer has been identified in the residue left from the saponification of delipidated cuticles. This polymer is named cutan and even less is known about it, although it can sometimes account for even more mass than cutin (Pollard et al., 2008). One hypothesis is that it is a polyether, with C-O bonds emerging from ring-opening of highly reactive epoxy groups (found within some of the cutin monomers described above) in the presence of nucleophiles, such as OH groups from nearby sugars or cutin fragments (Jeffree, 2006). This biochemical maturation of cutin into cutan is supported in part by the fact that the mass of the non-saponifiable leaf polymer fraction increases six times as the leaf matures in the case of Clivia miniata, a species for which epoxy-ω-hydroxy-fatty acids represent a major cutin monomer (Schmidt, 1982). However, C-C bond cross-linkages and polyester monomers atypical for cutin, such as resorcinols and α-resorcilic acid, have also been described as part of cutan structure (Deshmukh et al., 2005).  1.2.2. Cuticular waxes Cuticular waxes exhibit an even greater variety and can be chemically classified in three main categories: aliphatics, alicyclics, and aromatics.  1.2.2.1. Aliphatics The aliphatics are represented by very-long-chain fatty acids (VLCFA, predominantly with even carbon-chain lengths ranging from C20 to C34) and other mono-, bi-, and even tri-functional compounds biosynthetically derived from them (in fact, from their thioesters with   4 coenzyme A) (Samuels et al., 2008). Some of these derivatives are also dominated by even chain lengths >C20: VLC aldehydes (Samuels et al., 2008), VLC alcohols (Samuels et al., 2008), their esters with either LC fatty acids (most commonly) or VLC fatty acids (less commonly) (Lai et al., 2007; Samuels et al., 2008), as well as acetates of such alcohols (Jetter et al., 2000). Other derivatives are dominated instead by odd chain lengths (typically C23-C35): mainly alkanes (dominant cuticular wax components in many species, including the model plant Arabidopsis thaliana) (Jetter and Kunst, 2008; Samuels et al., 2008) and various mid-chain oxygenated derivatives of them, such as secondary alcohols (Greer et al., 2007; Wen and Jetter, 2009; Wen et al., 2006a),  ketones (Greer et al., 2007; Wen and Jetter, 2009),  α- and β-diols (Wen and Jetter, 2009; Wen et al., 2006a),  α- and β-ketols (Holloway and Brown, 1977; Wen and Jetter, 2009), β-diketones (Evans et al., 1975a; von Wettstein-Knowles, 1976) and hydroxy-β-diketones (Tulloch and Weenink, 1969). The structures of most commonly encountered cuticular waxes are shown in Fig. 1.2. Very small amounts of compounds with carbon chain lengths of opposite parity than shown in the figure are also encountered in wax mixtures. CH3OHOnEven chain length Odd chain lengthCH3HOnCH3OHnCH3OOCH3nmCH3HnCH3OHCH3qpCH3OCH3qpCH3OCH3Oqp Figure 1.2. Structures of the most common plant cuticular wax compounds. Typical ranges are: n=11, …, 16; m+n=16, …, 25; p+q=12, …, 14.   5  Sometimes, VLC aldehydes are not found in their free monomeric form in cuticular wax, but rather as trimers (2,4,6-trialkyl-1,3,5-trioxanes), as is the case of three species of rose flowers (Rosa americana, R. imperial, R. virgo) (Mladenova, 1976). A series of other bifunctional VLC compounds, with even carbon-chain lengths, have also been described in the surface wax of various plant species:  1,2- and 1,3-diols, along with 1,2-diol monoacetates in Cosmos bipinnatus petal wax (Buschhaus et al., 2013a),  1,5-diols and 5-hydroxyaldehydes in Taxus buccata needles (Wen and Jetter, 2007), and 1,9 / 1,11 / 1,13 / 1,15-diols, ketols, keto-aldehydes, and ketol esters of LC fatty acids in Osmunda regalis fronds (Jetter and Riederer, 1999a). The analogous 5-hydroxyfatty acids have been found in the form of their δ-lactones as the major waxes in Cerithe minor (Jetter and Riederer, 1999b). Furthermore, there are some monofunctional compounds of odd chain length in cuticular waxes with functional groups other than mid-chain: 10-nonacosanol in the waxes of numerous gymnosperms (Franich et al., 1978; Osborne and Stevens, 1996; Alexander P Tulloch and Bergter, 1981; Tulloch, 1987) and some angiosperm species (Koch et al., 2006; Tulloch and Hoffman, 1982), 10-nonacosanone in the wax crystals on Osmunda regalis fronds (Jetter and Riederer, 2000), and a number of long-chain fatty acid esters of C13-C19 2-alkanols found in the wax of fruit capsules of Papaveraceae (Jetter and Riederer, 1996) or spikes of barley (von Wettstein-Knowles, 1976).  1.2.2.2. Alicyclics The alicyclics are represented by triterpenoids, which are typically minor components of cuticular wax, but in some species, for example Olea europaea (Bianchi et al., 1993), can make up the bulk of waxes. Triterpenoids are divided into 2 main classes: phytosterols and nonsterol   6 triterpenoids. Both of these are derived from the cytoplasmic mevalonate biosynthetic pathway, which generates their common precursor (3S)-2,3-oxidosqualene (OS) (Guhling et al., 2006; Stiti and Hartmann, 2012). This can then either be converted to sterols via cycloartenol (formed by cycloartenol cyclase), or to nonsterol triterpenoids by a wide range of other highly versatile oxidosqualene cyclases (OSCs) (Xu et al., 2004). For example, the genome of Arabidopsis thaliana encodes no less than 13 such enzymes (Phillips et al., 2006), and one of these (baruol synthase) can produce, on its own, 23 triterpenoids (Lodeiro et al., 2007). It follows naturally that the diversity of alicyclics in plant cuticular waxes is even greater than that of VLCFA derivatives and it would be nearly impossible to achieve a truly comprehensive review. Nonetheless, the most ubiquitous phytosterols are β-sitosterol, stigmasterol, campesterol, brassicasterol, and to a lesser extent cholesterol (Johnson et al., 1963). Their common structural motif is the tetracyclic cholestane ring with an OH group at position 3 (Fig. 1.3). CH3HHHCH3CH3 CH3CH3HOH  Figure 1.3. 3β-Hydroxycholestane, the structural motif common to all sterols.  Nonsterol triterpenoids are classified based on their parent carbon skeleton. Most common are pentacyclics belonging to the oleanane and ursane families (Stiti and Hartmann, 2012) such as β-amyrin, erythrodiol, and oleanolic acid of the former, and α-amyrin, uvaol, and ursolic acid of the latter family (Fig. 1.4).    7 CH3CH3RHCH3CH3HCH3HCH3 CH3OHRHCH3CH3HCH3HCH3 CH3OHCH3CH3R=CH3       -amyrinR=CH2OH erythrodiolR=COOH  oleanolic acidR=CH3       -amyrinR=CH2OH uvaolR=COOH  ursolic acid Figure 1.4. Typical oleanane (left) and ursane (right) triterpenoids from plant waxes.  While triterpenoids are found most often in free form in cuticular waxes, in some species alcohol triterpenoids (whether sterols or nonsterols) have been identified as either esters of LCFA (e.g. amyrin palmitates and stearates on the adaxial side of Rosa canina leaves) (Buschhaus et al., 2007a), as acetates (Meusel et al., 1994), or as acetal-bound aglycons of a special class of glycosides known as saponins (Vincken et al., 2007). Triterpenoid acids can also serve as sapogenins (i.e. aglycons in saponins), forming ester linkages with carbohydrate hydroxyl functions (Vincken et al., 2007).  1.2.2.3. Aromatics Aromatics are the least common of all cuticular wax classes, but they have been reported to occur in low amounts in selected species. Examples include VLC 5-alkylresorcinols in Secale cereale (Ji and Jetter, 2008) and Triticum aestivum (Adamski et al., 2013a), VLC 5-alkyl-m-guaiacols in Tamarix canariensis (Basas-Jaumandreu et al., 2014), VLC benzyl esters in Rosa canina (Buschhaus et al., 2007a) and Olea europaea (Bianchi et al., 1993), VLC 2-phenylethyl esters in Olea europaea (Bianchi et al., 1993) and Ligustrum vulgare (Buschhaus et al., 2007b), 4-hydroxyphenylpropyl, 3,4-dihydroxyphenylpropyl, and 3,4-dihydroxyphenylbutyl esters in Taxus baccata (Jetter et al., 2002; Wen and Jetter, 2007), as well as flavonoids in Pteridaceae   8 (Wollenweber, 1989), Myrtaceae (Wollenweber et al., 2000) and three species of Chrysothamnus (Stevens et al., 1999).  1.3. Stratification of cuticle constituents It is well acknowledged that some cuticular waxes are embedded within the amorphous cutin matrix (intracuticular waxes ICW) and can only be accessed by extraction with a nonpolar solvent (e.g. CHCl3), while other waxes lie atop of cutin (epicuticular waxes ECW) and can be selectively peeled off using adhesives (e.g. gum Arabic or frozen glycerol or water) (Buschhaus and Jetter, 2011). Thus, in cross-section, the cuticle appears layered into a cuticle proper made of an exterior ECW film, from which crystals may or may not protrude out of the surface (depending on species/organ/tissue) and an interior layer of cutin+ICW, as well as an additional cuticular layer containing cutin, ICW, and fibrillar extensions of epidermal cell wall sugars (Fig. 1.5). Stacking many such cross-sections into the third dimension would generate a 3D model of the cuticle.  Figure 1.5. Cartoon of hypothetical cross-section through the cuticle [modified from (Bargel et al., 2006)].   9  To date, there have been a number of GC analysis reports delineating quantitative and even qualitative compositional differences between ICW and ECW in various species (Buschhaus and Jetter, 2012; Buschhaus et al., 2007a, 2007b; Gniwotta et al., 2005; Guhling et al., 2005; Jetter et al., 2000; Ji and Jetter, 2008; van Maarseveen and Jetter, 2009; Wen et al., 2006b), as well as a recent review identifying certain common trends across these reports and speculating on some of the possible causes (Buschhaus and Jetter, 2011). The most important finding of the review was that alicyclic and aromatic wax compounds almost always accumulate preferentially (often exclusively) in the ICW. In addition, a similar preference, albeit less obvious, was noted for VLC primary alcohols and alkanediols, while the opposite seemed true for alkanes, fatty acids, and mid-chain secondary alcohols. Also noted were a few examples where homologs of same compound class having longer carbon chain predominated in the ECW, while shorter ones preferred the ICW. Overall, it was concluded that ICW incorporates preferentially more polar classes, presumably capable of stronger intermolecular bonding to cutin polar sites (e.g. hydrogen bonds), and in particular more compact, cyclic molecules and the shorter of VLC aliphatic homologs, since each of these will exclude from their own phases the other (longer) waxes due to shape/size mismatch. The seemingly contradictory epicuticular accumulation of fatty acids was easily explained by their solid state head-to-head dimerization (Bond, 2004; Leiserowitz, 1976; Moreno et al., 2006), causing their exclusion by shorter ICWs. A few exceptions where polar triterpenoids were abundant in ECW were noted (Markstadter et al., 2000; van Maarseveen and Jetter, 2009). A plausible explanation would be that there is a saturation limit of cutin with ICW beyond which even compact alicyclics are excluded into the ECW. Furthermore, specific wax interactions with the cell wall must be considered. Since   10 carbohydrates are capable of H-bonding even more than cutin, it seems reasonable to assume that the most polar of intracuticular waxes (compact triterpenoids with oxygenated functions like COOH or OH) will segregate in the intracuticular layer ICL, in part buffering the transition from hydrophilic cell wall to lipophilic cutin. In support of this view, oleanolic and ursolic acids, which do not form carboxy/carboxy dimers due to steric hindrance (Casado and Heredia, 1999), have only been reported as 100% intracuticular in all literature reports so far (Buschhaus et al., 2007a, 2007b; Jetter et al., 2000).  1.4. Functions of the plant cuticle As briefly mentioned before, the most important function of the cuticle is to prevent excessive astomatous transpiration, i.e. loss of water through the plant-atmosphere interface itself as opposed to the stomatal pores carefully controlled by guard cells (Buschhaus and Jetter, 2011). It has been described that water vapor diffuses through the cuticle via a hopping mechanism (Matas and Heredia, 1999), following two possible paths: the lipophilic path – which is the main path, passes straight through the numerous amorphous portions of the cuticle, and is also the trajectory followed by many other non-electrolytes (e.g. pesticides) (Buchholz, 2006), as well as the less important polar (“aqueous”) path – which is represented for the most part by the polysaccharide microfibrils cutting through the cuticular layer and also serves as the only access way for highly polar and charged compounds (e.g. inorganic ions) (Schreiber, 2005). These latter tracks appear to be located preferentially around stomata, along and at the base of trichomes, as well as over anticlinal epidermal cell walls (i.e. cell walls between cells and therefore perpendicular to the plant surface) (Schreiber, 2008). The path that is totally inaccessible to water and thus acts as the best cuticular barrier is represented by highly ordered crystalline areas   11 of the cuticle, most often resulting from alignment of the very long aliphatic chains of VLCFA and their derivatives (Fig. 1.6) (Riederer and Schneider, 1990). It follows from here that the permeability of cuticular membranes has little to do with the amount of wax, cutin, or the cuticle thickness, but is rather linked to the nature and relative amounts of individual wax compounds and their self-assembly (Kerstiens, 2006). For example, a dominance of VLC alkanes within a very narrow range of chain lengths will generate the most effective water-diffusion barrier via segregation into highly crystalline condensed phases (Vioque and Pastor, 1994), whereas having additional contributors with very different chain lengths (e.g. alkyl esters combining VLC alcohols with LC acids) (Leide et al., 2011) or with completely different molecular geometries (e.g. alicyclics and aromatics) (Buschhaus and Jetter, 2012; Leide et al., 2011) will decrease resistance to non-stomatal water loss.  Figure 1.6. Cuticular crystallites formed by hydrocarbon chain alignment [modified from (Riederer and Schneider, 1990)].    12 But the cuticle is more than just a mere barrier for water. It is also blocking damaging UV-B and UV-C radiation from reaching the epidermis when aromatic absorbents are present (Markstadter and Riederer, 1997). Due to wax crystals formed at the surface, it effects a change in reflectance and thus a shift from an appealing glossy green to a whitish glaucous appearance, which may be less attractive to herbivores (Eigenbrode and Espelie, 1995). A similar insect herbivore or pathogen deterrent effect has been associated with increased contents of triterpenoids (e.g. α- and β-amyrins) (Balsdon et al., 1995) when the insect probes into the plant surface, or, on the contrary, with reduced levels of a certain class in the cuticular wax mixture (e.g. VLC aldehydes) (Hansjakob et al., 2011), which would otherwise direct oviposition or fungal pre-penetration. In carnivorous plants of the genus Nepenthes, cuticular aldehydes actually assist with their feeding process, by forming slippery crystals on the inner side of their pitchers and thus directing insect prey into the digestive pool (Riedel et al., 2007, 2003). In myrmecophyte Macaranga plants, a more complicated plant defense system operates: mainly two triterpenoids, epi-taraxerol and taraxerone, generate wax surface crystals that are slippery to generalist ants, but not to symbiotic ants, which are thus provided shelter by the plant and in return protect it by deterring insects and other pests (Markstadter et al., 2000). Furthermore, in general, regardless of their chemical nature, wax crystals have the merit of creating a hydrophobic and sometimes superhydrophobic plant surface (especially when combined with anatomic surface protuberances, e.g. papillose cells on lotus leaves) (Barthlott et al., 1997) that triggers a self-cleaning effect by rolling off rain or dew droplets along with dust, spores and other contaminants. The cuticle is also the first shield against mechanical damage (e.g. wind-carried particulates) and, by developing early during ontogenesis, it prevents organ fusions (Weng et al., 2010).   13  1.5. Biosynthesis of VLC aliphatic cuticular waxes  1.5.1. Biosynthesis of ubiquitous aliphatic waxes The biosynthesis of aliphatic cuticular waxes has been studied extensively using the model plant species Arabidopsis thaliana and is known to occur in three stages: de novo synthesis of LC fatty acid, elongation to VLC fatty acyl-coenzyme A (acyl-CoA) intermediates and, finally, modification of the latter into final wax products (Kunst et al., 2006). The first step takes place in epidermal plastids, while the last two steps occur in the endoplasmic reticulum (ER), from where waxes are then exported through the plasma membrane and cell wall to the cuticle. The de novo synthesis begins with acetyl-CoA as a C2 starter unit which is elongated two carbons via condensation with malonyl-acyl carrier protein (malonyl-ACP) catalyzed by the β-ketoacyl-ACP synthase III (KASIII) enzyme (Clough et al., 1992). The resulting C4 β-ketoacyl-ACP is reduced to β-hydroxyacyl-ACP, then dehydrated to trans-Δ2-enoyl-ACP and, finally, reduced again to the C4 acyl-ACP (i.e. butanoyl-ACP), all steps being catalyzed by dedicated enzymes: two reductases and a dehydratase. Further elongation from C4 to C16 is effected by sequential condensations with malonyl-ACP catalyzed by KASI, while elongation from C16 to C18 acyl-ACP is afforded by KASII (Shimakata and Stumpf, 1982). The same two reductases and dehydratase are shared by all KAS enzymes in the repeated steps of reduction/dehydration/reduction that follow each condensation, such that all enzymes taken together are known as the enzymatic complex fatty acid synthase (FAS). This complex ultimately produces saturated C16 and C18 acyl-ACP intermediates, which are then hydrolyzed by   14 a fatty acyl-ACP thioesterase (FATB) to C16 and C18 free fatty acids, before export to the ER (Bonaventure et al., 2003). Upon export from plastid, free fatty acids undergo thioesterification with coenzyme A (CoASH), catalyzed by a long-chain acyl-CoA synthetase (LACS), of which nine have been detected in A. thaliana (Shockey et al., 2002). Similar to FAS-catalyzed elongation, elongation of C16 and C18 acyl-CoA to VLC acyl-CoA thioesters also requires cycles of four consecutive enzyme-catalyzed reactions (Fig. 1.7):  1) condensation with malonyl-CoA to give β-ketoacyl-CoA, catalyzed by a condensing enzyme known as β-ketoacyl-CoA synthase (KCS), of which 21 have been identified in Arabidopsis (Blacklock and Jaworski, 2006);  2) reduction to a β-hydroxyacyl-CoA by the β-ketoacyl-CoA reductase (KCR) (Beaudoin et al., 2002); 3) dehydration to trans-Δ2-enoyl-CoA by a putative β-hydroxyacyl-CoA dehydratase; 4) reduction to the two-carbon longer acyl-CoA by the enoyl-CoA reductase (ECR) (Kohlwein et al., 2001). Together, these enzymes form an enzymatic complex known as the fatty acyl elongase (FAE) and the only product released is the elongated acyl-CoA, whose final chain length seems to be dictated by the condensing enzyme (KCS) in the complex (Blacklock and Jaworski, 2006; Franke et al., 2009; Lee et al., 2009; Paul et al., 2006; Trenkamp et al., 2004).   15 de novo fatty acid biosynthesisCH3OCoASOnCH3OHCoASOnCH3CoASOnCH3CoASOnREPEAT KCS KCR HCDECRCH3CoASO nFARCH3OHn WS CH3OCH3OnmreductaseCH3OndecarbonylaseCH3CH3nhydroxylaseCH3CH3OHpCH3CH3OHqp pCH3CH3OppCH3CH3OqpFigure 1.7. Biosynthesis of ubiquitous cuticular waxes in the ER of epidermal cells.  The modification of acyl-CoAs into final wax products takes place according to one of two main pathways (Fig. 1.7): the acyl reduction pathway, producing primary alcohols and alkyl esters, and the decarbonylation pathway, yielding aldehydes, alkanes, secondary alcohols, ketones, and other minor hydroxylation products of alkanes (e.g. α-/β-diols and α-/β-ketols).    16 Primary alcohols are formed from VLC acyl-CoA substrates by a fatty acyl reductase (FAR), typically without release of the intermediate aldehyde (Fig. 1.7). In Arabidopsis, this FAR has been identified as ECERIFERUM4 (CER4) (Rowland et al., 2006). Primary alcohols are either exported to the cuticle, or are esterified with long-chain acyl-CoAs into alkyl esters by a wax (ester) synthase (WS), which in A. thaliana has been characterized and named wax ester synthase/acyl-coenzyme A : diacylglycerol acyltransferase (WSD1) (Li et al., 2008). An alternative reduction of VLC acyl-CoA by a different reductase yields VLC aldehydes, which are then decarbonylated by a decarbonylase to VLC alkanes (Fig. 1.7). In Arabidopsis, these two steps are catalyzed by the CER3 and CER1 enzymes, which form a heterodimer (Bernard et al., 2012). Presumably, the CER3 enzyme carries the reductase activity, while CER1 acts as the decarbonylase. Alkanes are either exported to the cuticle, or undergo single or multiple hydroxylations by a hydroxylase of the cytochrome P450 family of enzymes, to yield primarily secondary alcohols and ketones, but also diols and ketols to lesser extent (Wen and Jetter, 2009). In A. thaliana, this hydroxylase targets specifically carbons around the middle of the chain and has been named mid-chain alkane hydroxylase 1 (MAH1) (Greer et al., 2007).  1.5.2. Biosynthesis of β-diketones in Poaceae In the cuticular waxes of various species of Poaceae (Gramineae), the predominant compound class is often represented by β-diketones, whose biosynthesis is not well understood and is thus still the object of recent investigations (Hen-Avivi et al., 2016b; Schneider et al., 2016). However, a biosynthetic pathway has been proposed (von Wettstein-Knowles, 2012) and is depicted in Fig. 1.8 for the most commonly encountered homolog, hentriacontan-14,16-dione.    17 de novo fatty acid biosynthesisCH3CoASO 7   PKS CH3OCoASO 7   PKS CH3OOCoASO 76 x FAECH3OOCoASO 76 1. reductase2. decarbonylaseCH3OOCH376 Figure 1.8. Proposed biosynthesis of hentriacontan-14,16-dione.   The main difference versus the Arabidopsis pathway is that the C16 acyl-CoA, instead of undergoing elongation by the FAE enzymatic complex, undergoes two condensations with malonyl-CoA catalyzed by a type III polyketide synthase (PKS), an enzyme that no longer removes the β-oxygen functionality. This results in a triketide, which then undergoes six 2-carbon extensions by the FAE complex. Finally, the terminal carbon is removed by either the combination of reductase and decarbonylase enzymes, very similar to alkane formation, or alternatively by a combination of thioesterase and decarboxylase enzymes (the exact mechanism is not known).  1.6. Objectives As mentioned in section 1.2.2., past work exploring cuticular wax compositions of plant species other than A. thaliana has revealed novel cuticular wax constituents, some of which have   18 provided new insights into wax biosynthesis beyond the Arabidopsis model (Buschhaus et al., 2013a; Busta et al., 2016; Jetter and Riederer, 2000, 1999a, 1999b; Jetter, 2000; Vermeer et al., 2003; Wen and Jetter, 2009, 2007; Wen et al., 2006a). The prospect that previously unexplored plant species could harbour novel wax compounds, with divergent biosynthetic pathways from those well established in Arabidopsis, was the driving hypothesis of the work presented in this thesis.  The main objective of the thesis was to identify and, where possible, quantify novel cuticular wax constituents from the wax mixtures coating select vegetative organs of three plant species, and thus expand cuticular wax diversity even further.  An ancillary objective of the thesis was to provide insights into the potential biosynthetic pathways leading to these novel wax constituents. Several aspects of wax biosynthesis described for Arabidopsis may not apply to other plant species, so that new questions arise:  Can FAE elongation intermediates be intercepted by other enzymes (hydrolases, reductases, etc.) in species other than Arabidopsis?  Is there a pathway leading to fatty acid derivatives containing nitrogen operational in other species?  Are aldehydes released as intermediates en route to primary alcohols during wax biosynthesis in other plant species?  Do wax ester synthases exhibit substrate specificities for one or both of their acyl-CoA and alcohol substrates in species other than Arabidopsis?  Are there cytochrome P450 hydroxylases targeting positions other than mid-chain operational in other species?    19  What types of polyketide-like structures are yielded by polyketide synthases in species other than Arabidopsis?   Due to a current scarcity of cuticular wax composition studies on monocotyledons, the three plant species chosen were all monocots. Furthermore, the species were chosen to span a diverse range of habitats, as well as practical applications. The three chosen species were: candelabra aloe (Aloe arborescens), a medicinal plant species endemic to the dry regions of Southern Africa (Chapter 2); fishpole bamboo (Phyllostachys aurea), an ornamental species originating from the subtropical areas of South-East China (Chapter 3); and bread wheat (Triticum aestivum), a major crop species that is cultivated in temperate climates worldwide (Chapters 4-6).      20 Chapter 2: Very-long-chain 3-hydroxy fatty acids, 3-hydroxy fatty acid methyl esters and 2-alkanols from cuticular waxes of Aloe arborescens leaves    2.1. Introduction  Land plants have lipid coatings over all their non-woody aerial parts, enabling prolonged retention of water, essential to their survival. Known as the cuticle, this coating consists of a cutin polyester matrix and complex mixtures of solvent-soluble cuticular waxes (Schreiber, 2005). In most plant species, the predominant wax constituents are very-long-chain (VLC) saturated aliphatic compounds, either with a single functional group or no functionality at all. All these compounds occur as mixtures of homologs, with chain lengths commonly ranging from C24 to C34. The series of the constituent fatty acids, primary alcohols, esters and aldehydes are typically dominated by even-numbered chain lengths, while alkanes, secondary alcohols and ketones are dominated by odd-numbered homologs. In a few species, alicyclic compounds such as triterpenoids can be quite abundant, sometimes even more than VLC aliphatics (Manheim Jr. and Mulroy, 1978; van Maarseveen et al., 2009). The chemical structures and amounts of cuticular wax compounds can vary greatly between plant species, between organs of the same species, between tissues of the same organ, and sometimes even between different compartments of the cuticle covering the same tissue. Thus, it has been shown that, in certain species, wax constituents with somewhat higher polarity (e.g. alcohols) or with a more compact molecular geometry (e.g. triterpenoids) accumulate to greater extent within the more polar cutin matrix (in the intracuticular wax compartment), while   21 others with lower polarity (e.g. alkanes) accumulate instead outside of cutin (in the epicuticular wax layer deposited atop the matrix). In a few instances, subtle chain length partitioning effects have also been reported, where shorter homologs of fatty acids or alcohols accumulated more in the intracuticular wax and longer ones in the epicuticular wax (Buschhaus and Jetter, 2011).  The wax VLC homolog profiles are the consequence of divergent biosynthetic pathways leading to the different compound classes (Jetter et al., 2006; Kolattukudy, 1970). It is well established that wax biosynthesis proceeds by elongation of long-chain (C16 and C18) fatty acyl precursors in fatty acyl elongase (FAE) complexes to VLC fatty acyl-CoAs. This elongation involves a sequence of four reactions, each catalyzed by a dedicated enzyme: 1) a decarboxylative Claisen condensation with malonyl-CoA, catalyzed by the β-ketoacyl-CoA synthase (KCS), which extends the carbon chain by two carbons; 2) a reduction of the resulting β-ketoacyl-CoA to β-hydroxyacyl-CoA by the β-ketoacyl-CoA reductase (KCR); 3) a dehydration of β-hydroxyacyl-CoA to an α,β-unsaturated acyl-CoA by a β-hydroxyacyl-CoA dehydratase (HCD); and finally 4) a second reduction to saturated acyl-CoA effected by the enoyl-CoA reductase (ECR). Typically, none of the above-mentioned intermediates are released by the enzyme complex, such that the only products of elongation are saturated acyl-CoAs with even carbon chain lengths of C20 and higher. These acyl-CoA products are then either hydrolyzed to fatty acids and exported to the plant surface, or further processed via one of two parallel biosynthetic pathways: the acyl-reduction pathway, which converts them into even-numbered primary alcohols and esters, or the decarbonylation pathway, which, after a first reduction to still even-numbered aldehydes, then converts these into odd-numbered alkanes (which may, in some species, be transformed into mid-chain secondary alcohols and ketones) (Kunst et al., 2006; Samuels et al., 2008).   22   While much of current knowledge on wax formation stems from molecular genetic investigations using Arabidopsis thaliana, further insights can be expected from exploring the much greater chemical diversity of wax mixtures in other species as well. Accordingly, several recent publications described novel compounds in the waxes of diverse species, often with two functional groups in characteristic arrangements. Among these are VLC 1,3-diols and their monoacetates in petal wax of Cosmos bipinnatus (Buschhaus et al., 2013a), VLC 1,3-diols and 3-hydroxyaldehydes in leaf wax of Ricinus communis (Vermeer et al., 2003), VLC 1,5-diols and 5-hydroxyaldehydes in Taxus baccata needle wax (Wen and Jetter, 2007), VLC δ-lactones in Cerinthe minor leaf wax (Jetter and Riederer, 1999b), and a diverse range of 1,9-, 1,11-, 1,13- and 1,15-bifunctional VLC aliphatics identified in the wax of Osmunda regalis fronds, including ketols, ketoaldehydes, diols, and ketoalkyl esters (Jetter and Riederer, 2000, 1999a). In all these studies, the nature and the relative position of functional groups suggested biosynthetic relationships both between the compounds and with standard wax biosynthetic pathway intermediates. In line with these previous studies, the objective of the present work was to search for, identify and quantify further novel multifunctional compounds that would corroborate and expand the models for wax biosynthetic pathways. In this context, a further goal of this study was to localize potentially new wax compounds in the intracuticular or epicuticular layer, on both sides of the leaf, as a prerequisite for further studies into their biological functions. To complement previous information on dicotyledonous species, the present investigation aimed to analyze the leaf cuticular waxes of a monocot, Aloe arborescens. This species is of special interest, because its leaves are known to be rich in phenolic secondary metabolites with important therapeutical applications, such as 6-phenylpyrone 2’- and 4’-O-glucosides (Beppu et al., 2004; Gutterman and Chauser-Volfson, 2000; Okamura et al., 1996; Park et al., 1998), which   23 have anti-histamine and anti-inflammatory activity (Bastian et al., 2013) or have led to derivatives with promising cytotoxicity against human colorectal and hepatoma cancer cells (Jin et al., 2005), 2-alkylchromone 8-C-glucosides (Gutterman and Chauser-Volfson, 2000; Okamura et al., 1996; Park et al., 1998), some of which have been shown to be efficient inhibitors of UV-induced hyperpigmentation (Choi and Chung, 2003; Choi et al., 2002) or to have anti-inflammatory properties (Bastian et al., 2013; Speranza et al., 2005), and anthraquinones and anthrone 10-C-glucosides (Beppu et al., 2004; Gutterman and Chauser-Volfson, 2000; Okamura et al., 1996; Park et al., 1998), with a wide range of medical usages from laxative to antitumor agents (Choi and Chung, 2003; Srinivas et al., 2007).   2.2. Experimental  2.2.1. Preparation of total leaf wax extracts  Aloe arborescens plants were maintained in growth chambers at the University of British Columbia, under the following conditions: 16 h light / 8 h dark, 21ºC / 19ºC, ~150 μE m-2 s-1 photosynthetically active radiation. The type of soil used was Sunshine Mix No. 4 (JVK and Crofton Grower Service) and it was always allowed to dry before re-watering (typically, watering every 10-12 days). Top halves from mature leaves were excised with clean razor blades and submerged in 10 mL CHCl3 (Aldrich, ≥99%, 0.75% ethanol as stabilizer) at room temperature for 30 s. To ensure exhaustive extraction, this step was repeated with a fresh portion of 10 mL CHCl3 for another 30 s and the two extracts were combined. Care was taken to avoid contact between chloroform and the open cut or any aqueous gel or latex occasionally leaking through it. To this end, leaf halves were dipped into the solvent only up to a set mark and any   24 portions beyond the mark were cut off afterwards, such that only the extracted leaf areas remained to be photographed and then measured using the ImageJ program. One leaf half was used to prepare one total wax replicate for GC analysis. Ten leaf halves were used to prepare the extract for TLC analysis.   2.2.2. Preparation of epicuticular and intracuticular wax extracts   The epicuticular and intracuticular waxes were sampled separately using procedures described in detail elsewhere (van Maarseveen and Jetter, 2009).  In short, gum arabic was applied to one side of the leaf only, the resulting polymer films were peeled off, and the polymer and adhering epicuticular waxes were partitioned between water and chloroform, respectively. After the removal of epicuticular wax, the remaining intracuticular wax was extracted with chloroform using glass cylinders pressed gently onto the leaf surface. The same procedures were followed for both the adaxial and abaxial leaf surfaces.   2.2.3. Qualitative and quantitative analyses of wax extracts   For structure elucidation, total wax mixtures were fractionated by preparative TLC as described by Wen et al. (2006a), except that a mixture of CHCl3 and EtOH in a 98:2 ratio (v/v) was used as mobile phase. The resulting fractions were characterized by their retardation factor (Rf), i.e. the ratio between the distance travelled on the TLC plate by the corresponding fraction and the distance travelled by the eluent front.  Prior to GC analysis, all samples were transferred into autosampler vials where the solvent was removed at 50ºC under a stream of N2 (Praxair, ≥99.998%). Waxes were then   25 derivatized by refluxing in 10 μL N,O-bis(trimethylsilyl)trifluoroacetamide (Aldrich, GC grade) and 10 μL pyridine (Aldrich, ≥99.8%, anhydrous) at 70ºC for 30 min. Excess reagents were removed at 50ºC under a stream of N2 (Praxair, ≥99.998%), and a known amount (10.2 μg) of n-tetracosane (Alfa Aesar, ≥99%) was added as internal standard, to which the vast majority of wax compound classes had been shown to have relative response factors of 1.00 under almost identical GC-FID conditions (Riederer and Schneider, 1989). Finally, samples were dissolved in a known volume of CHCl3 for GC analysis.  Two different GC instruments were used for separation and detection of wax constituents, both equipped with the same type of capillary GC column (6890N, Agilent, Avondale PA, USA; 30 m long; type HP-1: 100% PDMS; 0.32 mm i.d.; df=0.1 µm), using on-column injection and following the same temperature program (2 min at 50ºC, ramp 40ºC/min to 200ºC, constant for 2 min, ramp 3ºC/min to 320ºC, constant for 30 min). The first employed He gas (Praxair, ≥99%) as mobile phase, at a flow rate of 1.4 mL/min, and was equipped with MS detector (5973N, Agilent, EI-70 eV), serving primarily the purpose of qualitative identification of separated wax compounds. The second used H2 carrier gas (Praxair, ≥99.95%) at 2.0 mL/min and an FID detector, for quantification of individual wax homologs based on normalization of peak areas against that of the internal standard. Quantitative ester isomer compositions were determined from GC-MS data as described by Lai et al. (2007).  2.2.4. Synthesis of reference compounds   The condensation of octacosanoic acid with Meldrum’s acid under Steglich esterification reaction conditions and subsequent methanolysis yielded methyl 3-oxotriacontanoate, which was   26 split into two portions used further in two separate reaction sequences. One portion was subjected to decarboxylative hydrolysis using aqueous base in the presence of a phase transfer catalyst, then the resulting methyl ketone was reduced with sodium borohydride to produce 2-nonacosanol reference standard. A second portion was directly reduced with sodium borohydride to give methyl 3-hydroxytriacontanoate, part of which was acid hydrolyzed to yield the 3-hydroxytriacontanoic acid standard. Detailed experimental protocols and 1H-NMR characterization data for all reaction products are described in Appendix A.  2.3. Results  This study aimed to provide a comprehensive chemical analysis of the lipid mixtures coating Aloe arborescens leaves. To this end, we first elucidated the structures of all major wax constituents, then quantified all compounds in the overall leaf wax mixture and, finally, sampled and analyzed the epicuticular and intracuticular wax layers selectively on both the adaxial and abaxial sides of the leaf.  2.3.1. Structure elucidation of unknown compounds in A. arborescens leaf cuticular wax  Preliminary analyses had shown that A. arborescens waxes contained many ubiquitously found compounds (Fig. 2.1) readily identified by comparing gas chromatography-mass spectrometry (GC-MS) data with previously published information. However, the wax mixtures also contained a number of unusual compounds whose identification required in-depth analysis based on partial purification by thin layer chromatography (TLC), MS signal assignments and structure confirmation using synthetic standards.   27 CH3OOCH3nCH3 OHnCH3 OnCH3Hn2.2a (n=11)2.2b (n=12)2.2f (n=16)...2.3a (n=12)2.3b (n=13)2.3e (n=16)...2.5a (n=13)2.5b (n=14)2.5d (n=16)...2.6a (n=13)2.6b (n=14)2.6e (n=17)...CH3OO CH3n m2.4a (n+m=19)2.4b (n+m=20)2.4f (n+m=24)...CH3OOHn2.1a (n=9)2.1b (n=10)2.1h (n=16)...OOHCH3CH3CH3CH3CH3 CH32       2.11a CH3CH3CH3HCH3CH3HCH3HCH3 CH3OHCH3HCH3CH3HCH3HCH3 CH3OHCH3CH3HCH3CH3HCH3HCH3 CH3OHCH3HCH3CH2HCH3CH3142.7aHCH3CH3142.7bOOCH3CH3OHn2.8a (n=10)2.8b (n=11)2.8e (n=14)...OOHCH3OHn2.9a (n=11)2.9b (n=12)2.9c (n=13)CH3CH3OHn2.10a (n=10)2.10b (n=11)2.10f (n=15)...       2.11b        2.11c        2.11d Figure 2.1. Compounds identified in the total wax mixture of A. arborescens leaves.      28 To enable structure elucidation of unknowns, the cuticular wax mixture extracted from A. arborescens leaves was separated by preparative TLC on silica gel (mobile phase CHCl3:EtOH 98:2, v/v). The five resulting fractions A – E were further analyzed by GC-MS and found to contain known VLC fatty acids 2.1 (A, Rf 0.31), VLC 1-alkanols 2.3 and terpenols 2.11 (C, Rf 0.40), VLC fatty acid methyl esters (FAMEs) 2.2 and aldehydes 2.5 (D, Rf 0.93), as well as VLC esters 2.4 and alkanes 2.6 (E, Rf 1.00). In addition, the three most polar fractions A, B (Rf 0.36) and C each contained a major homologous series of unknown compounds, based on GC-MS characteristics. Notably, all three compound classes appeared to share a common structural motif characterized by homologous fragments m/z 117+28n in their trimethylsilyl (TMS) derivative mass spectra (n being a homolog-dependent integer).  The three unknown compounds in fraction A were tentatively assigned to a homologous series of 3-hydroxy fatty acids 2.9. The TMS-derivatives of all three homologs had a prominent MS fragment m/z 147 [(CH3)2SiOSi(CH3)3]+ indicating the presence of two hydroxyl groups in the native compounds (Jetter et al., 1996; Richter and Burlingame, 1968; Rontani and Aubert, 2004). The formation of a fragment m/z 117 [(CH3)3SiOCO]+ together with the absence of a fragment m/z 103 [(CH3)3SiOCH2]+ indicated that one of the OH groups was located on a terminal carbon, not as a primary alcohol, but rather in a carboxyl group (Buschhaus et al., 2013; Jetter and Riederer, 1999b). The other functional group was recognized to be a non-terminal hydroxyl, based on one TMS-ether α-fragment m/z 233 common to all homologs and another α-fragment varying between homologs (m/z 481 in Fig. 2.2A). The fractionation of this unknown series with free fatty acids 2.1 was also suggestive of a functionalized carboxylic acid structure, however an alternative 2,4-diol structure could not be entirely ruled out thus far. We therefore synthesized C30 3-hydroxy fatty acid 2.9c (see Appendix A), and tested its GC-MS behaviour   29 against the unknown series in fraction A. The TMS derivative of the standard showed matching MS fragmentation patterns (Fig. 2.2A-C) and GC retention time (Fig. 2.2D) with the longest homolog in the unknown series, thus confirming that the A. arborescens wax compounds are indeed C26, C28 and C30 3-hydroxy fatty acids 2.9. 30hc m/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100Col 2 23373 481 147597 523117189 217        Figure 2.2. Structure elucidation of unknown compound series in fraction A of A. arborescens wax. Mass spectra of (A) C30 3-hydroxy fatty acid 2.9c isolated from A. arborescens leaf wax, and (B) synthetic C30 3-hydroxy fatty acid 2.9c. (C) Major fragmentations of C30 3-hydroxy fatty acid 2.9c; note that fragment m/z 597 can also originate from loss of any of the other five TMS-bound methyl groups in the molecular ion. (D) Overlay of selected ion m/z 233 chromatograms of the 3-hydroxy fatty acid series 2.9 in A. arborescens wax and of synthetic C30 3-hydroxy fatty acid 2.9c; this fragment is distinctive for the 3-hydroxy fatty acid series 2.9 in fraction A. 13CH3O(H3C)3SiOOSiCH3CH3CH3    233    481    597   523     117A  B D  C 30hc m/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100Col 2  233 73481  147597 523 1171892172D Graph 1Retention time [min]20 22 24 26 28 30 32 34 36 38 40Relative abundance of m/z 233 [%]01020304050607080901003-Hydroxy Acids in A. arborescens WaxC30 3-Hydroxy Acid Synthetic StandardC28C30C26  30  Similarly, the five compounds in fraction B were tentatively assigned to a homologous series of 3-hydroxy FAMEs 2.8. The absence of fragment m/z 147 and the presence of fragments m/z 73 [(CH3)3Si]+  and m/z 89 [(CH3)2SiOCH3]+ indicated that the compounds contained only one OTMS group (Vermeer et al., 2003; Wen and Jetter, 2007). A TMS-ether α-fragment varying between homologs (m/z 481 in Fig. 2.3A) suggested a secondary OH group in the same position as in the A series, while a second α-fragment m/z 175 common to all homologs indicated an additional terminal methoxycarbonyl group. All this information, together with the TLC behaviour of B indicating polarities between 3-hydroxy fatty acids 2.9 and primary alcohols 2.3, pointed to 3-hydroxy FAME 2.8 structures. To test this assignment, we synthesized C30 3-hydroxy FAME 2.8d (see Appendix A) and compared its GC-MS behaviour against compounds B (Fig. 2.3A-D). Based on the match between standard and wax characteristics, we conclude that the second unknown series is a series of 3-hydroxy FAMEs 2.8.         31 30hmm/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100Col 2 175577348115913353989 2D Graph 1Retention time [min]18 20 22 24 26 28 30 32 34 36 38Relative abundance of m/z 175 [%]01020304050607080901003-Hydroxy FAMEs in A. arborescens WaxC30 3-Hydroxy FAME Synthetic Standard  C28C30C26C24 C32 Figure 2.3. Structure elucidation of unknown compound series in fraction B of A. arborescens wax. Mass spectra of (A) C30 3-hydroxy fatty acid methyl ester (FAME) 2.8d isolated from A. arborescens leaf wax, and (B) of synthetic C30 3-hydroxy FAME 2.8d. (C) Major fragmentations of C30 3-hydroxy FAME 2.8d; note that fragment m/z 539 can also originate from loss of either of the other two TMS-bound methyl groups in the molecular ion. (D) Overlay of selected ion m/z 175 chromatograms of the 3-hydroxy FAME series 2.8 in A. arborescens wax and of synthetic C30 3-hydroxy FAME 2.8d; this fragment is distinctive for the 3-hydroxy FAME series 2.8 in fraction B.      13CH3OH3COOSiCH3CH3CH3    175    481    53930hm Stdm/z100 200 300 400 500 600Relative abundance [%]010203040506070809010017557 7348115913353989A B D C   32  Fraction C comprised six homologs tentatively identified as 2-alkanols 2.10 based on MS similarity to data previously reported for wax from one other plant species, Solanum tuberosum (Szafranek and Synak, 2006a). The diagnostic α-fragment m/z 117 common to all homologs in C, together with a chain length-variable α-fragment [M-15]+, indicated the presence of a 2-hydroxyl group in the native homologous series (Fig. 2.4A). This structure assignment was confirmed via preparation of C29 2-alkanol 2.10d (see Appendix A), whose characteristic MS fragments and GC retention time matched those of one compound in C (Fig. 2.4A-D). Thus, all odd-numbered 2-alkanol 2.10 homologs from C23 to C33 were identified in A. arborescens leaf wax.              33 Nonacosan-2-ol in Aloem/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100Col 2 1175773 4812D Graph 1Retention time [min]12 14 16 18 20 22 24 26 28 30 32Rel ative abundance of  m/z 117 [%]01020304050607080901002-Alkanols in A. arborescens WaxC29 2-Alkanol  Synthetic StandardC27 C29C25C23 C33C31 Figure 2.4. Structure elucidation of unknown compound series in fraction C of A. arborescens wax. Mass spectra of (A) C29 2-alkanol 2.10d isolated from A. arborescens leaf wax, and (B) synthetic C29 2-alkanol 2.10d. (C) Major fragmentations of C29 2-alkanol 2.10d; note that in (C), fragment m/z 481 can also originate from loss of any of the three TMS-bound methyl groups in the molecular ion. (D) Overlay of selected ion m/z 117 chromatograms of the 2-alkanol series 2.10 in A. arborescens wax and of synthetic C29 2-alkanol 2.10d; this fragment is distinctive for the 2-alkanol series 2.10 in fraction C.     Nonacosan-2-ol Standardm/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100Col 2 1175773481A B D C 13CH3OCH3(H3C)3Si    481    117  34  2.3.2 Quantitative compositional analysis of total cuticular wax of A. arborescens leaves    After establishing the various compound classes and specific homologs present in A. arborescens cuticular wax, we used GC with flame ionization detection (FID) to quantify all compounds in the overall wax mixture extracted from both sides of the leaf. The total leaf wax coverage was 25.6±0.9 μg/cm2, of which 95% was identified (1.1±0.2 μg/cm2 unidentified compounds).  n-Alkanes 2.6 were by far the most abundant compound class (8.8±0.4 μg/cm2) in the wax mixture, accompanied by lesser amounts of 1-alkanols 2.3, esters 2.4, aldehydes 2.5 and acids 2.1 (Fig. 2.4). Interestingly, the novel 3-hydroxy FAMEs 2.8 were relatively abundant (1.1±0.2 μg/cm2), whereas the 3-hydroxy acids 2.9, 2-alkanols 2.10 and non-hydroxylated FAMEs 2.2 were detected only in trace amounts (below 0.1 μg/cm2). The only branched-chain compounds were iso- and anteiso-alkanes 2.7 (below 0.1 μg/cm2). Cyclic compounds comprising both triterpenoids 2.11 (mainly lupeol 2.11d, along with β- 2.11b and α-amyrin 2.11c) and γ-tocopherol 2.11a constituted only 1% of the wax mixture (0.3±0.1 μg/cm2) (Fig. 2.5).       35  AcidsFAMEs1-AlkanolsEstersAldehydesn-AlkanesBranched Alkanes3-Hydroxy FAMEs3-Hydroxy Acids2-AlkanolsTerpenoidsNot IdentifiedWax cover age [g/cm2]012345678910.Figure 2.5. Compound class composition of total leaf wax mixture. Total coverages (μg/cm2) of compound classes within A. arborescens leaf wax. The inset shows an enlarged view of the area enclosed by dashed lines. Bars represent mean ± standard deviation (n = 5).      Total Wax Classes   AcidsFAMEs1-AlkanolsEstersAldehydesn-AlkanesBranched Alkanes3-Hydroxy FAMEsTerpenoidsNot IdentifiedWax cover age [g/cm2]0.000.010.020.030.040.05  36 Within all aliphatic compound classes, homologous series were present. The free fatty acid series 2.1 had the widest chain length range of all fractions, varying from C20 to C34 (including trace levels of odd-numbered homologs). The acid mixture 2.1 was further distinguished by a bimodal homolog distribution dominated by chain lengths C28 and C32 (Fig. 2.5). All other compound classes had narrower homolog distributions around only one maximum, with FAMEs 2.2, 1-alkanols 2.3, 3-hydroxy acids 2.9 and 3-hydroxy FAMEs 2.8 all dominated by respective C28 homologs and aldehydes 2.5 dominated by C32. The only compound classes with predominant odd-numbered homologs, n-alkanes 2.6 as well as 2-alkanols 2.10, had chain length profiles both peaking at C31. Similarly, the two branched alkanes 2.7 identified in A. arborescens leaf wax also had C31 chains (but 32 carbons overall). Finally, the fatty acid alkyl esters 2.4 had chain lengths ranging from C42 to C52, dominated by the C46 homolog 2.4c (Fig. 2.6).   37 2022242628303234242628303234262830323442444648505228303234272931333532 iso32 anteiso2426283032262830232527293133tocopherolamyrinamyrinlupeol  Relative composition [% of compound class]0102030405060708090Acids 1-AlkanolsFAMEs Terpenoids2-AlkanolsAcidsFAMEsAlkanesAldehydesEstersHydroxy  Hydroxy    Figure 2.6. Relative compositions of each compound class in the total leaf wax mixture. Relative abundances (%) of individual homologs or isomers (in the case of branched alkanes and terpenoids) from each compound class in the composition of A. arborescens leaf wax. Numbers on the x-axis indicate homolog chain length. Bars represent mean ± standard deviation (n = 5). Each group of bars adds up to 100%.      38 Each of the wax ester homologs 2.4 was found composed of several metamers, i.e. isomers with different combinations of acid 2.1 and 1-alkanol 2.3 chain lengths. Because our GC experimental conditions did not allow the separation and individual quantification of such isomers, their quantities were determined by MS instead. To this end, the same total wax extracts used for acquiring the GC-FID data described above were further analyzed by GC-MS, and the relative abundances of all Macid+1 product ions were used to measure relative quantities of esterified acids as previously described (Gülz et al., 1994; Jetter and Riederer, 1999a, 1996; Lai et al., 2007; Reiter et al., 1999; Shepherd et al., 1995). The esters contained mainly C16, C18 and C20 acids, with lesser amounts of C14 as well as C22-C26 acids (Fig. 2.7A). All esterified alcohols taken together had a chain length profile ranging from C20 to C34, with strong predominance of the C28 homolog 2.3b (Fig. 2.7B). The shorter esters (C42 2.4a and C44 2.4b) contained mainly C16 acid bonded to C26 2.3a and C28 2.3b alcohols, respectively. The predominant ester homolog (C46 2.4c) contained both C16 and C18 acid (esterified with C30 2.3c and C28 2.3b alcohols, respectively). In contrast, the longer ester homologs (C48-C52 2.4d-f) all had C20 acid 2.1a as the major constituent, esterified with C28-C32 alcohols 2.3b-d. Thus, the variation in overall ester chain length was due to both varying alcohol and acid chain lengths, with acid diversity driving the formation of central ester homologs C44-C48 2.4b-d and alcohol diversity driving the formation of all other ester homologs.   39 Ester chain length42 44 46 48 50 52Relative isomer composition [% of each est er  homol og]0102030405060708090100C14 acidC16 acidC18 acidC20 acidC22 acidC24 acidC26 acid Alcohol Chain Length20 22 24 26 28 30 32 34Relative compositi on of ester ified alcohol s [%]01020304050603.08.2 Total Esterified Alcohol Coverage:                  g/cm2 Figure 2.7. Relative compositions of esterified acids and alcohols in the total leaf wax mixture. (A) Relative abundances (%) of individual isomers (indicated by chain length of their acid moiety) of each ester homolog in A. arborescens leaf wax, and (B) resulting relative total abundances (%) of ester-bound alcohol homologs. Bars represent mean ± standard deviation (n = 5). Each group of bars in (A) adds up to 100%. A B   40 2.3.3. Distribution of A. arborescens leaf cuticular waxes between adaxial and abaxial surfaces, as well as epicuticular and intracuticular layers   To assess how the compounds identified and quantified in the overall wax mixture are distributed across various wax layers on both sides of the A. arborescens leaf, the epicuticular and intracuticular waxes from the adaxial and abaxial surfaces were sampled separately. Gum arabic was used as a glue for selective sampling of only the epicuticular wax layer from one side of the leaf, similar to previous studies (Jetter and Schäffer, 2001; Ji and Jetter, 2008; Wen et al., 2006b). To ensure complete removal of epicuticular waxes, four consecutive applications of the adhesive were performed on each side of each leaf. The remaining intracuticular waxes were extracted with chloroform applied in a glass cylinder of known diameter pressed against the leaf, a method also employed previously in the aforementioned studies. The epicuticular wax coverage on the adaxial side of the A. arborescens leaf was 9.0±0.9 μg/cm2, while the corresponding intracuticular wax amounted to 5.9±0.2 μg/cm2. In contrast, the abaxial epicuticular wax accumulated to 18.9±0.9 μg/cm2, and the abaxial intracuticular wax to 17.2±0.8 μg/cm2. Taken together, our results show an uneven wax distribution between the abaxial (70%) and the adaxial (30%) sides of the leaf (p<0.01). Based on the epi- and intracuticular coverages on both sides, an average total leaf coverage of 26 μg/cm2 was calculated, thus independently confirming the result of the total leaf wax extraction (see above). The total (epi- plus intracuticular) wax mixtures had very similar relative compositions (expressed in %) on both sides of the leaf, except for a slightly higher concentration of fatty acids in the adaxial wax (p<0.01).   41 In order to establish whether any wax constituents accumulate preferentially in a certain wax layer, the relative compound class compositions of the epicuticular and intracuticular wax mixtures were determined. Free fatty acids 2.1 and FAMEs 2.2 were found in higher concentrations in the epicuticular waxes than in the intracuticular layers on both leaf sides (adaxial and abaxial p<0.01), while 1-alkanols 2.3 (adaxial and abaxial p<0.01) and terpenoids 2.11 (adaxial and abaxial p<0.01) showed the opposite trends and preferentially accumulated in the intracuticular layer (Fig. 2.8A-B). The adaxial wax had relatively high concentration of n-alkanes 2.6 in the epicuticular (adaxial p<0.01) and 3-hydroxy acids 2.9 in the intracuticular wax (adaxial p<0.01), while the abaxial waxes had no corresponding gradients. Aldehyde 2.5 levels also differed between the wax layers, albeit with opposing gradients on the two leaf sides (adaxial and abaxial p<0.01). All other compound classes did not show preferential accumulation in either layer on either leaf side (Fig. 2.8A-B). The coverages of all individual compounds in the epicuticular and intracuticular compartments on the adaxial and abaxial leaf surfaces of A. arborescens are reported in Table 2.1. However, from this data, no apparent accumulation of wax homologs with either short or long chain lengths is noticeable for either compartment on either leaf surface.      42 AcidsFAMEs1-AlkanolsEstersAldehydesn-AlkanesBranched Alkanes3-Hydroxy FAMEs3-Hydroxy Acids2-AlkanolsTerpenoidsNot IdentifiedRe la tive  c omp osition of each layer       within the a baxial wax [%]010203040EpicuticularIntracuticularAcidsFAMEs1-AlkanolsEstersAldehydesn-AlkanesBranched Alkanes3-Hydroxy FAMEs3-Hydroxy Acids2-AlkanolsTerpenoidsNot IdentifiedRela tive c ompos iti on of each l ayer         within t he adaxial wa x [%]01020304050EpicuticularIntracuticularAB************ Figure 2.8. Relative compound class compositions of wax layers on adaxial and abaxial leaf sides. Relative abundances (%) of compound classes in epi- and intracuticular layers of (A) adaxial  and (B) abaxial wax of A. arborescens leaf. The insets show enlarged views of the areas enclosed by dashed lines. Bars represent mean ± standard deviation (n = 5). Asterisks mark significant differences between the arcsin-transformed epi- and intracuticular abundances of the respective compound class (two-tailed Student’s t test, p<0.01).   43 Table 2.1. Composition of epi- and intracuticular waxes on the adaxial and abaxial sides of A. arborescens leaves. Mean coverages ± standard deviations (n = 5) are given in µg/cm2. Compound Class Chain Length / Adaxial Wax (µg/cm2) Abaxial Wax (µg/cm2)  Isomer Epicuticular Intracuticular Epicuticular Intracuticular       Fatty acids 2.1 20 0.0042±0.0008 0.0012±0.0003 0.0099±0.0009 0.009±0.002  22 0.0028±0.0002 0.0003±0.0001 0.0052±0.0006 0.004±0.001  24 0.011±0.001 0.0025±0.0008 0.015±0.002 0.019±0.006  26 0.034±0.007 0.013±0.005 0.07±0.02 0.022±0.009  28 0.33±0.05 0.03±0.01 0.62±0.08 0.18±0.07  30 0.11±0.01 0.09±0.02 0.23±0.04 0.20±0.03  32 0.26±0.04 0.03±0.03 0.33±0.03 0.07±0.07  34 0.017±0.004 0.0017±0.0007 0.06±0.02 0.003±0.001       Fatty acid methyl esters 2.2 24 0.008±0.002 0.0032±0.0009 0.009±0.005 0.005±0.001 (FAMEs) 26 0.015±0.003 0.003±0.001 0.025±0.007 0.005±0.002  28 0.016±0.004 0.002±0.002 0.037±0.008 0.01±0.01  30 0.013±0.002 0.0009±0.0006 0.0096±0.0009 0.002±0.001  32 0.0004±0.0001 0.0002±0.0001 0.0020±0.0008 0.0004±0.0002       1-Alkanols 2.3 26 0.009±0.001 0.03±0.01 0.044±0.009 0.04±0.02  28 0.50±0.08 0.11±0.03 1.5±0.3 2.0±0.5  30 0.24±0.04 0.93±0.08 0.7±0.2 0.49±0.04  32 0.23±0.04 0.5±0.1 0.56±0.06 1.1±0.2  34 0.04±0.01 0.024±0.004 0.10±0.01 0.14±0.02       Esters 2.4 42 0.22±0.06 0.09±0.03 0.47±0.02 0.24±0.09  44 0.29±0.06 0.20±0.06 0.66±0.06 0.7±0.2  46 0.48±0.09 0.3±0.1 1.1±0.1 1.0±0.4  48 0.26±0.04 0.27±0.06 0.6±0.1 0.9±0.2  50 0.29±0.05 0.19±0.06 0.44±0.08 0.5±0.2  52 0.049±0.008 0.03±0.03 0.08±0.01 0.06±0.06       Aldehydes 2.5 28 0.11±0.02 0.25±0.08 0.3±0.1 0.25±0.09  30 0.11±0.02 0.019±0.005 0.4±0.1 0.4±0.1  32 0.33±0.06 0.6±0.2 1.2±0.1 0.31±0.08  34 0.26±0.06 0.05±0.02 0.8±0.2 0.2±0.1       n-Alkanes 2.6 27 0.014±0.002 0.003±0.001 0.047±0.006 0.004±0.002  29 0.025±0.006 0.05±0.02 0.09±0.01 0.03±0.01  31 3.8±0.3 0.9±0.1 6.0±0.6 5.0±0.6  33 0.20±0.01 0.05±0.01 0.39±0.06 0.5±0.1  35 0.010±0.001 0.0032±0.0005 0.023±0.003 0.0016±0.0002       Branched alkanes 2.7 32 iso 0.012±0.005 0.013±0.004 0.032±0.009 0.03±0.01  32 anteiso 0.04±0.02 0.015±0.007 0.09±0.04 0.04±0.02       3-Hydroxy fatty acid methyl esters 2.8 24 0.016±0.004 0.017±0.003 0.04±0.01 0.024±0.004 (3-Hydroxy FAMEs) 26 0.039±0.004 0.05±0.01 0.18±0.05 0.15±0.03  28 0.21±0.07 0.05±0.01 0.40±0.09 0.31±0.04  30 0.06±0.01 0.04±0.02 0.14±0.02 0.13±0.06  32 0.006±0.003 0.005±0.005 0.07±0.02 0.01±0.01       3-Hydroxy fatty acids 2.9 26 0.0007±0.0001 0.0011±0.0002 0.0026±0.0004 0.0036±0.0002  28 0.0037±0.0009 0.0046±0.0008 0.015±0.004 0.012±0.002  30 0.0005±0.0001 0.0011±0.0006 0.0023±0.0007 0.002±0.001       2-Alkanols 2.10 31 0.006±0.003 0.004±0.002 0.004±0.002 0.019±0.008  33 0.0034±0.0009 0.0024±0.0006 0.0008±0.0002 0.0017±0.0009       Terpenoids 2.11 γ-tocopherol not detected 0.008±0.002 not detected 0.009±0.002  β-amyrin not detected 0.017±0.008 not detected 0.11±0.03  α-amyrin not detected 0.011±0.002 not detected 0.04±0.01  lupeol not detected 0.09±0.01 not detected 0.3±0.1    44 2.4. Discussion  Our analyses of A. arborescens leaf waxes led to three major results: (1) the wax mixtures contained typical aliphatic constituents in characteristic chain length distributions; (2) we identified unusual constituents with 2- and 3-hydroxy functionalities; (3) the epicuticular and intracuticular layers on the adaxial and abaxial leaf sides differed both in their relative compound class compositions and their absolute wax amounts. All three findings have implications for the formation and accumulation of cuticular waxes in this species that will be discussed below.   2.4.1. Chain length distributions of ubiquitous wax constituents   Various Aloe species (including A. arborescens) were investigated before, but the analyses has been restricted to acids and alkanes so far. Our results are similar to the literature reports, however with two notable exceptions. First, fewer homologs have been detected in previous reports, with fatty acid chain lengths ranging only up to C31 (Herbin and Robins, 1969). Second, the C29 homolog has been reported to dominate the preeminent wax class of n-alkanes (Herbin and Robins 1968a, 1969), instead of C31 in our results (89%, see Fig. 2.6). It is noteworthy that both these discrepancies involve the longer homologs in respective compound classes, pointing to chain length bias originating from the very different methods used for sample preparation (filtration, distillation, decantation, solvent changes, fractionation) and analysis (desorption from celite, packed-column GC, isothermal elution). This is confirmed by Feakins and Sessions (2010), who, using protocols very similar to ours, found the homolog distribution of Echeveria runyeonii leaf wax alkanes to be shifted towards higher chain lengths than reported   45 before (Herbin and Robins, 1968b). We conclude that our findings of relatively high concentrations of longer homologs reflect the true composition of A. arborescens leaf wax more accurately than the earlier reports. Overall, the wax composition described here for A. arborescens is similar to those of other monocots and dicots. Previously reported wax coverages varied from ~1 μg/cm2 and ~15 μg/cm2 on Arabidopsis thaliana leaves and stems, respectively (Pascal et al., 2013), to ~600 μg/cm2 on Fuyu persimmon fruit (Tsubaki et al., 2013). A. arborescens leaves had an intermediate wax coverage (~26 μg/cm2), similar to the unrelated succulent Kalanchoe daigremontiana (~20 μg/cm2, van Maarseveen and Jetter, 2009). In particular, the aliphatic compound classes in A. arborescens leaf waxes had homolog distributions similar to those reported for many other species before, with predominantly odd-numbered chain lengths for alkanes (branched and unbranched), and mainly even-numbered acids, FAMEs, primary alcohols, aldehydes as well as esters. Such profiles are typical for plant cuticular waxes and are thought to originate from the different biosynthetic pathways leading to the various compound classes (Kolattukudy, 1970; Samuels et al., 2008). Wax biosynthesis is known to occur in two parallel branch pathways, leading from acyl CoA precursors either to primary alcohols and their alkyl esters or to aldehydes and on to alkanes one carbon shorter. These biosynthetic relationships between compound classes, mostly established in studies on model species including Arabidopsis thaliana, may help to interpret the homolog distribution of respective fractions of the A. arborescens leaf wax. The products of the two branch pathways have pairwise similarities in chain length profiles, with C32 and C31 predominating for aldehydes and (branched and unbranched) alkanes, respectively (see Fig. 2.6), and C28 for alcohols and ester alkyl moieties (compare Figs. 2.6 and 2.7). Thus, those compounds   46 on the same branch pathway have clearly matching chain length profiles, whereas the homolog patterns differed drastically between pathways. In contrast, the fatty acids have a bimodal distribution combining the predominant chain lengths of both branch pathways, likely reflecting the composition of the acyl CoA pool that serves as common precursors for both modification pathways. Therefore, our chemical data on acid, aldehyde, alkane, primary alcohol and ester profiles strongly suggest that the same biosynthetic pathways previously established for a few model species also operate in the distantly related A. arborescens. This species is characterized by a strong selectivity of the alcohol-forming branch pathway for C28 substrates, and an equally strong preference of the alkane-forming branch for C32 substrates instead. Further substrate specificity can be inferred for the enzyme(s) forming the alkyl esters, based on isomer profiles of the six ester homologs (Fig. 2.7). On the one hand, the matching chain length distributions of all esterified 1-alkanols (Fig. 2.7B) and free 1-alkanols (Fig. 2.6) suggests that esters are formed directly from the pool of free alcohols, without enzyme specificity towards the alcohol substrate. On the other hand, the predominance of C16, C18 and C20 acyl moieties in the overall ester mixture points to strong selectivity for acyl CoA substrates with these chain lengths. Therefore, our detailed homolog and isomer analyses suggest that A. arborescens wax esters are formed by an enzyme with access to the primary alcohols (formed on one wax biosynthetic branch pathway) and certain acyl CoA intermediates (of the fatty acid elongation machinery), albeit with drastically differing degrees of selectivity for both its substrates. It should be noted that FAMEs, as another set of esters, have a chain length distribution more similar to free acids (although not bimodal), and are therefore likely derived from them rather than from acyl CoA intermediates.    47 2.4.2. Identification and chain length distributions of novel wax constituents   Our second major result was the identification of three unusual compound classes with 2- and 3-hydroxyl groups. Similar medium- and long-chain 2-alkanols had been reported for waxes of various plant species, albeit in esterified form: odd-numbered homologs C9-C15 in cuticular waxes of Eucalyptus species (Horn et al., 1964), C11-C17 in waxes of Papaver species (Jetter and Riederer, 1996), and primarily C13 and C15 homologs in waxes from a wide range of Poaceae species (Tulloch, 1983; von Wettstein-Knowles and Netting, 1976a, 1976b; von Wettstein-Knowles, 1976), together with C9 (von Wettstein-Knowles et al., 1984). Odd-numbered C23-C33 2-alkanol homologs had also been found esterified in estolides from Juniperus scopulorum (A P Tulloch and Bergter, 1981), but in free form had only been reported once before, in Solanum tuberosum leaf wax, spanning all chain lengths from C25 to C30 (Szafranek and Synak, 2006a).  VLC 3-hydroxy fatty acids have not been reported before, in either free or esterified form. However, there are numerous accounts involving shorter 3-hydroxy acids and their derivatives. Acid hydrolysates of glycolipids from various bacterial strains had revealed normal, as well as anteiso and iso methyl-branched C9 to C16 3-hydroxy acids in Coxiella burnetti (Wollenweber et al., 1985), C12 to C18 in several Bacteroides species (Mayberry, 1980), and mostly n-C14 in Salmonella lipopolysaccharides (Rietschel et al., 1972). Even organic household dust had been found to contain primarily even-numbered C10-C18 3-hydroxy fatty acids of the same bacterial (Mielniczuk et al., 1993) or, alternatively, fungal origin (Saraf et al., 1997). Also, several species of Rhodotorula red yeast had been shown to produce extracellular glycolipids incorporating C16 and C18 3-hydroxy fatty acids (Tulloch and Spencer, 1964). In higher plants, certain long-chain 3-hydroxy fatty acids had been described in floral oils of various species,   48 either as mono- or di-glycerides, very often acetylated at the 3-OH group, and much less frequently in free form. Vogel was the first to identify C14, C16 and C18 3-acetoxy fatty acids as glycerides in Calceolaria floral oils (Vogel, 1971). Several reports followed shortly, describing C16, C18 and C20 3-acetoxy fatty acids in Krameria spp., both in free form and in glycerides (Seigler et al., 1978; Simpson et al., 1979, 1977). Unsaturated C16 and C18 3-acetoxy acids were also detected in Lysimachia ciliata floral oil, either free or as glycerides (Cane et al., 1983). Finally, 3-hydroxy fatty acids (typically C14, C16 and C18) have recently been reported in free form (as well as acetylated and bound in glycerides) in Diascia spp. (Kanchana et al., 2008) and Malpighia coccigera floral oils (Seipold et al., 2004). The strong structural resemblance between 3-hydroxy fatty acids, 3-hydroxy FAMEs and 2-alkanols (Figs. 2.2-2.4) co-occurring in the leaf wax mixture of A. arborescens suggests that these compounds are biosynthetically related. Furthermore, the very similar chain length profiles of 3-hydroxy FAMEs and 3-hydroxy acids (Fig. 2.6) makes it very plausible that these two compound classes are derived one from the other, probably in the same way that FAMEs are derived from fatty acids. Thus, a methyl transferase might convert 3-hydroxy acids into 3-hydroxy FAMEs (possibly an S-adenosylmethionine-dependent methyltransferase). The 3-hydroxy acids in turn may be formed in one step from corresponding 3-hydroxy acyl CoA intermediates of the fatty acid elongation cycle. To intercept these intermediates from the elongase complex, a thioesterase is required. We thus propose a simple two-step biosynthetic pathway (Fig. 2.9, center), leading from known wax biosynthesis intermediates to 3-hydroxy acids and 3-hydroxy FAMEs. Considering the predominant C28 chain lengths we observed for these novel compounds as well as for free acids and corresponding FAMEs in A. arborescens wax, it is even possible that the same thioesterase is involved in forming both hydroxy acids and   49 normal acids, and the same methyl transferase may be involved in converting them into hydroxy FAMEs and normal FAMEs, respectively. For the third class of unusual wax compounds, the 2-alkanols, three alternative pathways may be postulated based on their structure and homolog distribution: firstly, they could be formed by hydrolysis of 3-keto acyl CoA intermediates of elongation, decarboxylation to 2-ketones and reduction to the corresponding 2-alcohols (Fig. 2.9, top-right). A similar biosynthetic pathway has been formulated for (esterified) medium-chain 2-alkanols of Poaceae (Kunst et al., 2006; von Wettstein-Knowles, 2012, 1987). No additional enzymes have to be invoked for this pathway, since the decarboxylation step may occur spontaneously, and both reductase and thioesterase activities with fairly similar substrate requirements are present in the form of KCR and the esterase putatively involved in 3-hydroxy acid formation, respectively. 2-Ketones would be expected as intermediates of such a pathway, however they were not detected in A. arborescens wax. Both 2-alkanols and 2-ketones have to date only been found co-occurring in potato leaf wax, but with very different chain length profiles. Therefore, a biosynthetic relationship between them has been discussed very critically (Szafranek and Synak, 2006a).   Alternatively, the 2-alkanols might be formed by decarboxylation of 3-hydroxy acids, likely under enzymatic control (Fig. 2.9, center-right). This pathway would require a decarboxylase for which little precedence exists, and very strong chain length specificity would have to be invoked for the implicated enzyme to explain the different homolog distributions of its substrates and products (predominantly C28 hydroxy acids and C31 2-alkanols, respectively).  Finally, 2-alkanols might also be formed by reduction of 3-hydroxy acyl CoAs and decarbonylation of the 3-hydroxy aldehyde intermediates (Fig. 2.9, bottom-right). It should be noted that these intermediates were not detected in A. arborescens wax, but they have been   50 described in (not closely related) Ricinus communis (Vermeer et al., 2003). This pathway is analogous to formation of alkanes and might involve the same enzymes, likely homologs of the Arabidopsis CER3 and CER1 proteins (Bernard et al., 2012). The involvement of these enzymes could in fact explain the very similar chain length profiles of alkanes and 2-alkanols. Furthermore, this pathway would resemble the one proposed above leading to 3-hydroxy acids, in that both require intercepting the same elongation intermediates, thus together suggesting an unusual leak of the elongase enzyme complex. Similar interception of elongation intermediates by thioesterases has been reported, however only for medium-chain acyl-ACP intermediates from FAS complexes  in plant plastids (Yu et al., 2010) and the cytosol of sponge-associated bacteria (He et al., 2012). All arguments taken together, the third pathway seems to be most plausible based on our chemical evidence alone. However, the models discussed above (and possibly others) must be tested using genetic and biochemical tools, either in A. arborescens or in comparable models.     51 primary alcoholsalkyl estersde novo fatty acid biosynthesisCH3OCoASOnCH3OHCoASOnCH3 CoASOnCH3 CoASOnREPEAT KCS KCR HCDECRCH3 CoASOnCH3OOHO thioesterase reductase decarbonylase thioesterasen methyl-transferase reductaseCH3 On decarbonylaseCH3OHOCH 3OnCH3OHOHOn (decarboxylase)CO2CH3OCH3n reductaseCH3OHCH3COnCOCH3 CH3nCH3OHOnCO2 decarboxylase  Figure 2.9. Proposed biosynthetic pathways to 3-hydroxy acids, 3-hydroxy FAMEs and 2-alkanols. Full arrows describe the biosynthetic pathways most consistent with our A. arborescens leaf wax analysis results. Broken arrows describe two alternative pathways towards 2-alkanols that are neither supported, nor can be ruled out solely on the basis of our results herein. Intermediates shown in square brackets are expected to be unstable and thus undetectable in surface wax. Enzymes shown in round brackets denote situations where the respective reactions can also proceed non-enzymatically.      52 2.4.3. Gradients between epi- and intracuticular wax layers Substantial gradients were found in the chemical composition of the epicuticular and intracuticular wax layers on both sides of the A. arborescens leaf. The differences between adjacent layers were most pronounced for alicyclic compounds, both triterpenoids and tocopherols. This result is similar to results for many other species, where terpenoids had consistently been found accumulating predominantly (or exclusively) in the intracuticular wax layer (Buschhaus and Jetter, 2012; Buschhaus et al., 2007a, 2007b; Guhling et al., 2005; Jetter and Schäffer, 2001; van Maarseveen and Jetter, 2009).  Some of the aliphatic compound classes in A. arborescens also exhibited gradients between the epicuticular and the intracuticular wax mixtures, although only with relatively small differences in percentages between the layers. Acids were found at slightly higher concentrations in the epicuticular wax and primary alcohols in the intracuticular wax, similar to some other species investigated before (Buschhaus and Jetter, 2011). Our results on A. arborescens waxes thus help to further corroborate the previous notion that partitioning between both wax layers may be driven by differences in compound polarity (where acids must be considered fairly nonpolar due to formation of hydrogen-bonded head-to-head dimers). All other compounds showed either inconsistent effects between both sides of the A. arborescens leaf, or no gradients at all. Such conflicting behaviours for wax classes have been reported before (Buschhaus and Jetter, 2011). Interestingly, the unusual 3-hydroxy FAMEs and 2-alkanols did not partition differentially between wax layers on either leaf side, and 3-hydroxy acids showed a gradient between layers only on the adaxial side (see Fig. 2.8). Thus, these compounds with additional secondary functional groups show partitioning trends similar to those of relatively nonpolar mono-functional wax constituents.   53 Chapter 3: Composition of the cuticular waxes coating the adaxial side of Phyllostachys aurea leaves: Identification of very-long-chain primary amides   3.1. Introduction  In the evolutionary transition from aquatic to terrestrial environments, plants developed a lipid coating to protect their non-woody organs against excessive water loss (Schreiber, 2005). This skin, the cuticle, consists of the polymer cutin and a mixture of solvent-extractable waxes (Kolattukudy, 1970). Cutin is a cross-linked polyester incorporating mainly saturated and unsaturated long-chain (LC, C16-C18) hydroxy-fatty acids, epoxy-fatty acids, diacids, as well as glycerol (Pollard et al., 2008). Cuticular waxes are complex mixtures of very-long-chain (VLC, typically C20-C34) aliphatics, with no functionality (alkanes), a single terminal oxygen functional group (fatty acids, aldehydes, primary alcohols), or an in-chain group (secondary alcohols, ketones, alkyl esters). Most of the resulting compound classes occur as series of homologs with predominantly even-numbered hydrocarbon chains (acids, aldehydes, primary alcohols, esters), while some others have mainly odd-numbered homologs (alkanes, secondary alcohols, ketones) (Jetter et al., 2006). In many species, alicyclic compounds such as triterpenoids and tocopherols are present along with VLC aliphatics, in some species at fairly high concentrations (Bianchi et al., 1992; Markstadter et al., 2000; Nordby and McDonald, 1994).    54 Genetic and biochemical investigations have led to a detailed understanding of wax biosynthesis in Arabidopsis thaliana (Jetter et al., 2006; Li-Beisson et al., 2010; Samuels et al., 2008). In this model species, long-chain fatty acids are produced de novo in the plastids of epidermal cells and then transferred to the endoplasmic reticulum (ER) for further elongation to VLC acyls by a four-enzyme fatty acyl-CoA elongase (FAE) complex. The array of acyl-CoA homologs serves as precursors for modification into the wax products, along (i) an acyl reduction pathway yielding primary alcohols and alkyl esters, (ii) a decarbonylation pathway to aldehydes, alkanes, secondary alcohols and ketones, or (iii) hydrolysis to corresponding free fatty acids. Much evidence has been provided over several decades, showing that parts or all of the biosynthesis pathways thus defined for Arabidopsis may also be operational in diverse other species.  However, wax composition varies substantially between plant species, and thus further biosynthetic pathways, or variations of the Arabidopsis pathways, must lead to compounds not present in this model species. Therefore, further insights into wax composition, biosynthesis and function can be obtained from investigations into diverse plant lineages. For example, recent studies have revealed novel multi-functionalized wax compounds such as hydroxy alkyl esters in Funaria hygrometrica (Busta et al., 2016), 1,3-diol monoacetates in Cosmos bipinnatus wax (Buschhaus et al., 2013b). As another case point, 3-hydroxy fatty acids and their methyl esters were recently discovered in the leaf wax of the monocot Aloe arborescens (Racovita et al., 2015), in contrast to many earlier reports that had identified mostly standard wax compounds in Poaceae monocots (von Wettstein-Knowles, 2012). In all cases, comparisons between the chemical diversity within and between taxa led to new insights into wax biosynthesis. Thus,   55 detailed chemical investigations of diverse lineages have the potential to reveal further variation of wax compositions and to reveal the presence of new biosynthetic mechanisms.  Among the wax compound classes, the esters formed by combination of VLC alcohols and acids are of particular interest. Due to their extra-long hydrocarbon chains and molecular weights almost double that of other wax constituents, the esters may contribute substantially to the physiological function of the cuticle as transpiration barrier (Riederer and Schreiber, 1995). Their molecular properties also make them high-value industrial chemicals, for application as specialty lubricants, cosmetic and pharmaceutical products (Jetter and Kunst, 2008). Plant wax esters vary both in their overall chain length and in the chain lengths of their acyl and alkyl moieties, thus giving rise to homology and isomerism. The wax ester mixtures of a few plant species have been investigated in detail, and they were found to have characteristic patterns of overall ester homologs and acid/alcohol combinations. For example, Arabidopsis thaliana stem wax esters are formed mainly by C16 acid and C20-C32 alcohols (Lai et al., 2007). The esters in Camelina sativa leaf wax had C16 and C20-C22 acyl moieties combined with C22-C26 alkyls (Razeq et al., 2014), while those of Aloe arborescens had C16-C20 acyls and C20-C34 alkyls (Racovita et al., 2015), and those on barley cer-u69 spikes had C16 and C20 acyls connected to C22-C26 alkyls (von Wettstein-Knowles and Netting, 1976a). The compositions in some species resembled the homolog patterns of accompanying free alcohols, suggesting that the same alcohol pool may serve either for direct export to the plant surface or for formation of wax esters while still inside the epidermis cell. However, characteristic differences between chain length profiles of alcohols or acids before and after esterification further suggest substrate specificities of the ester synthase enzymes involved (Lai et al., 2007; Li et al., 2008).   56  Recent studies have revealed a stratification with substantial compositional gradients between the wax embedded within the cutin matrix (intracuticular wax) and that lying atop of it (epicuticular wax) (Buschhaus and Jetter, 2011). For many plant species, the somewhat more polar wax constituents (e.g., alcohols) and those with a more compact molecular geometry (e.g., terpenoids) were found in higher concentrations in the intracuticular wax. The least polar compound classes (e.g., alkanes) often accumulated preferentially in the epicuticular wax mixtures, where they affect the hydrophobicity of plant surfaces (Holloway, 1969) or may play a role in plant-pathogen (Gniwotta et al., 2005) and plant-insect interactions (Udayagiri and Mason, 1997). Depending on the species, both the intracuticular and the epicuticular wax layers may contribute to the transpiration barrier function of the cuticle (Jetter and Riederer, 2015). Overall, it is important to determine the chemical compositions of both the epi- and intracuticular wax mixtures towards a deeper understanding of their respective biological functions.  Leaves of the bamboo Phyllostachys aurea exhibit interesting surface properties (Badyal, pers. commun.), and this monocot species was hence chosen here for detailed wax analyses. No Phyllostachys wax compositions had been reported before, and studies of other bamboo genera (Bambusa, Dendrocalamus) had been restricted to the composition of one compound class, n-alkanes, for chemotaxonomy purposes (Li et al., 2012). Therefore, the present study aimed at a comprehensive analysis of the leaf cuticular waxes of Phyllostachys aurea, to identify all major compounds and to quantify respective compound classes, homolog distributions and isomer patterns in the epicuticular and intracuticular wax mixtures of young and old leaves.    57 3.2. Experimental  3.2.1. Plant material and preparative thin layer chromatography (TLC)  Leaves were harvested using clean metal tweezers and scissors from outdoor beds of old and young Phyllostachys aurea plants in June 2015 (Vancouver, British Columbia, Canada). Plant age was assessed based on the color of the culm: plants with a golden brown culm were designated as “old”, while those with a green culm as “young”.  For a first experiment, ten young leaves were submerged in 10 mL CHCl3 (Aldrich, ≥99%, 0.75% ethanol as stabilizer) and vigorously shaken for 30 seconds. The CHCl3 extract was transferred to another vial and the leaves were shaken with another portion of 10 mL CHCl3 for another 30 seconds. The second extract was added to the first one and the total volume reduced at 50°C under a gentle stream of N2 (Praxair, ≥99.998%). Compound classes in the resulting total wax extract were fractionated by preparative TLC, using the sandwich technique (Tantisewie et al., 1969), glass plates coated with silica gel 60 F254 as stationary phase (UniplateTM, Analtech Inc., layer thickness: 1 mm, size: 20x20 cm, without concentrating zone), and a mixture of hexane:methanol:ethyl acetate 85:10:5 (v/v/v) as mobile phase (Hexane: Aldrich, anh., ≥99%; Methanol: Aldrich, for HPLC, ≥99.9%; Ethyl acetate: Aldrich, anh., ≥99.8%). TLC bands were visualized under 365 nm UV light after spraying the plates with primuline (5 mg in 100 mL acetone/water 80/20, v/v), scratched off with clean spatulas into separate glass vials, and extracted twice with 10 mL each of fresh CHCl3 for 30 seconds, with agitation. The extracts were filtered through glass wool (Supelco), concentrated under N2 at 50°C, transferred to GC autosampler vial inserts, evaporated to dryness and stored until GC-MS analysis.         58 3.2.2. Preparation of epicuticular and intracuticular wax extracts  In a second experiment, four sets of leaves were sampled from each of young and old culms. To obtain the epicuticular wax extracts, each set of leaves was first painted with an aqueous solution of gum arabic (1.5 g/mL) over the full adaxial area of the leaves, using a paintbrush. After about 30 minutes, a solid film of gum arabic and adhering epicuticular waxes were lifted with clean tweezers and transferred to a glass vial containing 7 mL each of distilled water and CHCl3. The gum arabic application was repeated twice more and the resulting films added to the same vial and partitioned between water and CHCl3. The organic layer was carefully transferred to another glass vial using a Pasteur pipette; the aqueous layer was extracted once more with a fresh portion of 7 mL CHCl3, which was then combined with the first organic extract. The combined CHCl3 extracts were concentrated at 50°C under a stream of N2, quantitatively transferred to GC autosampler vial inserts, evaporated to dryness and stored until GC analysis. Typically, one biological replicate consisted in a set of four leaves, each with an adaxial area of ~8 cm2. Exact leaf areas were computed using the ImageJ software, after photographing the set of leaves with their adaxial side up on a white paper background in the presence of a ruler.   To obtain the intracuticular wax extracts, a glass cylinder (9.6 mm in diameter) was gently pressed onto the adaxial surface of leaves, following epicuticular wax removal, and 1.5 mL fresh CHCl3 were loaded inside it and agitated to improve extraction efficiency by bubbling air from a Pasteur pipette for 30 seconds. The CHCl3 extract of intracuticular waxes was then transferred to a clean glass vial, and the extraction was repeated twice more with another two   59 portions of 1.5 mL fresh CHCl3. Typically, the extracts from eight leaf locations (two per leaf) were pooled together and represented one replicate. The resulting extracts were concentrated at 50°C under N2, quantitatively transferred to GC autosampler vial inserts, evaporated to dryness and stored until GC analysis.  3.2.3. Derivatization reactions  In preparation for GC-MS/FID analysis, wax extracts were spiked with 10 μL of CHCl3 containing 1.02 mg/mL internal standard n-tetracosane (Alfa Aesar, ≥99%) and subjected to silylation with 10 μL N,O-bis(trimethylsilyl)trifluoroacetamide BSTFA (Aldrich, GC grade) in 10 μL pyridine (Aldrich, anh., ≥99.8%) for 30 min at 70ºC. Then, samples were taken to dryness at 50ºC under N2 and re-dissolved in 20 μL CHCl3. Under these conditions, amides did not undergo silylation, resulting in broad GC peaks and poor resolution for all homologs in this class. Therefore, after re-evaporation to dryness, samples were derivatized using 20 μL benzyl bromide (Aldrich, ≥98%) in the presence of 0.1 mg NaH (Aldrich, dry, ≥95%) at 70ºC for 1 hour. After aqueous work-up, products were partitioned into CHCl3, the organic layer was removed and the solvent evaporated to dryness. Lastly, the residue was subjected to silylation with BSTFA/pyridine again, re-dried, and then re-dissolved in 20 μL CHCl3.  3.2.4. Gas chromatography (GC) analysis of wax extracts  Two GC instruments were used for the qualitative and quantitative analysis of wax constituents, respectively. Shared characteristics of both instruments were the type of capillary   60 GC column (6890N, Agilent, Avondale PA, USA; 30 m long; type HP-1: 100% PDMS; 0.32 mm i.d.; df=0.1 µm) and the oven temperature program (on-column injection at 50ºC, constant for 2 min, ramp 40ºC min-1 to 200ºC, constant for 2 min, ramp 3ºC min-1 to 320ºC, constant for 30 min). The first instrument was operated with He gas (Praxair, ≥99%) as mobile phase at a flow rate of 1.4 mL/min, and was equipped with an MS detector (5973N, Agilent, EI-70 eV, m/z 50-750) for qualitative identification of the separated wax compounds. The second GC employed H2 as carrier gas (Praxair, ≥99.95%) at 2.0 mL/min and an FID detector for the quantification of individual wax constituents, based on normalization of their peak areas against that of the internal standard. It had been determined previously that most plant wax compounds have relative response factors of 1.00 with respect to n-tetracosane under almost identical GC-FID conditions (Riederer and Schneider, 1989). It should be noted that the exact FID response factors of the newly discovered acyl amide are currently unknown, and thus their quantification against the tetracosane standard had to be based on the assumption of response factors similar to other wax compounds. The current amide quantification may therefore slightly under- or overestimate the absolute amide amounts. The relative metamer compositions of alkyl ester were determined from GC-MS data as described elsewhere (Lai et al., 2007).   For statistical analysis, percentage values describing wax composition were arcsin-transformed, and then pair-wise comparisons were performed simultaneously on the entire dataset using Student’s t-tests (two-tailed, alpha = 0.05) and raw p values adjusted using a False Discovery Rate (FDR) equal to 5% with GraphPad Prism v6.0 software.     61 3.2.5. Synthesis of C30 amide standard  9.0 mg (0.02 mmol) triacontanoic acid (Aldrich, 98%) were dissolved in 1.00 mL freshly distilled dichloromethane (Fisher, 99.9%) in a 5 mL glass vial, with warming. 0.03 mL (0.41 mmol) of thionyl chloride (Aldrich, 97%) were added, the vial was sealed and the mixture was stirred for four hours with occasional heating up to 70°C to keep the triacontanoic acid dissolved. The resulting mixture was dried under a gentle stream of nitrogen at 50°C and re-dissolved in 1.00 mL dichloromethane. 0.50 mL (12.55 mmol) aqueous ammonia 28.0-30.0% (Aldrich) was added, and the two-phase system was vigorously stirred overnight. The organic layer was removed, and the aqueous phase extracted with 0.50 mL of fresh dichloromethane. Both organic solutions were combined, concentrated under nitrogen and loaded onto a preparative TLC glass plate coated with silica gel 60 F254 (layer thickness: 1 mm, size: 20x20 cm, with 4 cm concentrating zone), which was then developed with hexane:methanol:ethyl acetate 85:10:5 (v/v/v) as mobile phase. A white solid was obtained (1.0 mg, 11% yield) and found to contain >99% triacontanoic acid amide by GC-MS. 1H NMR (400 MHz, CDCl3): δ 5.31 (s, 2H, CONH2), 2.36 (t, 2H, J=7.6 Hz, CH2CONH2), 1.22-1.38 (br m, 54H, aliphatic CH2), 0.89 (t, 3H, J=6.7 Hz, CH3).     3.3. Results and discussion  The work herein aimed to compare the chemical composition of different layers within the cuticular waxes on the adaxial side of Phyllostachys aurea leaves on young and old culms. First, the structures of all major wax constituents were established by gas chromatography-mass   62 spectrometry (GC-MS; 3.3.1.), then the epicuticular and intracuticular wax layers were selectively sampled and their compound class composition (3.3.2.) and homolog patterns determined by GC with flame ionization detection (FID; 3.3.3.). Finally, the wax ester isomer distributions were quantified for each layer in further GC-MS analyses using signal ratios of specific acyl fragments (3.3.4.).  3.3.1. Identification of very-long-chain (VLC) fatty acid amides (3.1) in the adaxial cuticular waxes of P. aurea leaves  Qualitative GC-MS analysis of cuticular wax mixtures from the adaxial leaf side of both young and old bamboo plants identified homologous series (Fig. 3.1) of free fatty acids 3.2, primary alcohols 3.3, alkyl esters 3.4, aldehydes 3.5 and alkanes 3.6. They were accompanied by terpenoids 3.7, including several isomers of tocopherols, triterpenols and triterpenyl palmitates, all of which were identified by GC-MS comparison with authentic standards, or by MS characteristics matching those reported in the literature or mass spectral databases.  However, five small, evenly spaced GC peaks with MS features unprecedented in the wax literature warranted further investigation. All members of this compound class (series A) had identical, abundant fragment ions and molecular ions differing by multiples of 28 amu between them, indicating homologous compounds. For further structure elucidation, the P. aurea leaf wax mixture was fractionated by thin layer chromatography (TLC) using SiO2-coated glass plates and hexane:methanol:ethyl acetate 85:10:5 (v/v/v) as mobile phase (Nordby and McDonald, 1994). The fraction containing series A (Rf 0.06) was separated from free fatty acids   63 3.2 (Rf 0.23), primary alcohols 3.3 and most of the terpenols 3.7e-h (Rf 0.32), glutinol 3.7c and epifriedelanol 3.7d (Rf 0.38), aldehydes 3.5, alkyl esters 3.4 and terpenoid esters 3.7i-l (Rf 0.44), and alkanes 3.6 (Rf 1.00).   CH3OOHnCH3 OHnCH3 OnCH3Hn3.2a (n=7)3.2b (n=8)3.2j (n=16)...3.3a (n=10)3.3b (n=11)3.3g (n=16)...3.5a (n=11)3.5b (n=12)3.5f (n=16)...3.6a (n=12)3.6b (n=13)3.6d (n=15)...CH3OO CH3n m3.4a (n+m=16)3.4b (n+m=17)3.4i (n+m=24)...CH3ONH2n3.1a (n=12)3.1b (n=13)3.1e (n=16)...OOHRCH3CH3CH3CH3CH3 CH32       3.7a (R=H)       3.7b (R=CH3)CH3CH3CH3HHCH3CH3HCH3 CH3ROCH3       3.7c (R=H)       3.7i (R=C15H31CO)CH3CH3CH3HHCH3CH3HCH3OHCH3CH3       3.7dCH3CH3CH3HCH3CH3HCH3HCH3 CH3ROCH3HCH3CH3HCH3HCH3 CH3ROCH3CH3       3.7e (R=H)       3.7j (R=C15H31CO)        3.7f (R=H)       3.7k (R=C15H31CO)CH3CH3CH3HCH3CH3HCH3 CH3ROCH3       3.7g (R=H)       3.7l (R=C15H31CO)HCH3CH3HCH3HCH3 CH3OHCH3HCH3CH2       3.7h Figure 3.1. Compounds identified in the adaxial wax mixture of P. aurea leaves.    64 Compounds A had parent ions M with odd m/z, for example m/z 451 for the predominant homolog (Fig. 3.2A), indicating the presence of nitrogen. Taken together with the TLC results, this suggested fatty acid amide structures 3.1 for compounds A. This hypothesis was supported by M-43 fragments for all homologs, most likely resulting from loss of isocyanic acid (HNCO) from M. It was further confirmed by fragments m/z 59, 72 and 128 common to all compounds A, reminiscent of ions m/z 74, 87 and 143 characterizing fatty acid methyl esters, but 15 amu lighter due to the presence of –NH2 instead of –OMe. It seems likely that the signature fragments of compounds A thus originate via mechanisms similar to those described before for fatty acid methyl esters (Härtig, 2008). Overall, compounds A were thus far confirmed as a homologous series of VLC fatty acid amides 3.1 with even-numbered chain lengths ranging from C26 to C34.  To further test this structure assignment, the fraction enriched in series A was benzylated with excess benzyl bromide in the presence of excess NaH, resulting in a mixture of singly and doubly N-benzylated derivatives, both with MS features fully consistent with the primary amide parent structure (Fig. 3.2B and 3.2C). Finally, the C30 amide homolog 3.1c was synthesized from commercially available C30 fatty acid 3.2h via a Schotten-Baumann reaction. The synthetic standard showed identical MS features and GC retention time as one compound in series A, confirming its structure to be C30 amide 3.1c (Fig. 3.2D). It should be noted that the N-benzylated derivatives gave sharper GC peaks than the underivatized amides and were therefore used for GC-MS peak identification.    65 m/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100m/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100m/z100 200 300 400 500 600Relative abundance [%]0102030405060708090100CH3 NH2O  59(+H)  72  12811  239(+H)  9113PhCH2.  106(+H)CH3NO  148CH3 NHO    1391  106149(+H)  162A CB D1285972451408MM-431065771631540MM-91912391481065771541M91162149498M-43Retention time [min]40 42 44 46 48 50 52 54 56 58 60Relative abundance [%]0102030405060708090100P. aurea wax amidesC30 amide standard C30C32C28C26C34  Figure 3.2. Structure elucidation of compound series A from the adaxial wax mixture of P. aurea leaves. (A) Mass spectrum and fragmentation schemes of plant C30 amide 3.1c. (B) Mass spectrum and fragmentation schemes of N-benzylated plant C30 amide. (C) Mass spectrum and fragmentation schemes of N,N-dibenzylated plant C30 amide. (D) Overlay of extracted ion chromatograms (m/z 149) of the monobenzylated amide series from P. aurea adaxial leaf wax and the monobenzyl derivative of synthetic C30 amide; the fragment m/z 149 is distinctive for the monobenzyl derivatives of aliphatic amides.    66 Fatty acid amides had not been described in plant cuticular waxes before. More generally, nitrogen-containing compounds had been reported in only a few instances as constituents of plant wax mixtures, specifically in the form of monoterpenoid indole alkaloids, such as catharanthine, in the surface waxes of Catharanthus roseus leaves (Roepke et al., 2010), and of N-permethylated alkaloids in the waxes of several Papaveraceae species (Jetter and Riederer, 1996). In both cases, the N-containing compounds were found in surface extracts from species with fairly high concentrations of alkaloids in the underlying tissues, raising the possibility that these compounds were present due to contamination of wax samples (extracted from interior parts of respective organs). The discovery of fatty acid amides in bamboo wax may now raise similar questions. However, it should be noted that other internal lipids often found as contaminants of cuticular wax mixtures, such as acyl glycerides, were not detected in our bamboo wax samples, and that the acyl amides were found exclusively in the epicuticular waxes (see section 3.3.2.). Therefore, the fatty acid amides must be regarded as true components of the bamboo cuticle rather than contamination from internal tissues.  Interestingly, a compound with similar VLC acyl and amide feature, N-2-(p-hydroxyphenyl)ethyl C30 amide, had previously been isolated from macerated, dried leaves of Pseuderanthemum carruthersii (Nga et al., 2012). Unfortunately, the localization of this compound within the plant tissues was not studied. However, its acyl chain length and structural resemblance to 2-(p-hydroxyphenyl)ethyl esters of VLC fatty acids, which are fairly common cuticular wax constituents (Acevedo et al., 2000; Alfatafta et al., 1989; Oksuz and Topcu, 1992), both suggest that it may be formed in epidermal tissues (along with cuticular waxes). It is thus plausible that this amide too is accumulating in the cuticular wax of P. carruthersii.   67  Shorter-chain acyl amides are of pharmacological importance, for example oleamide present in human plasma and involved in the sleep/wake cycle (Mendelson and Basile, 2001) and vasodilation (Hiley and Hoi, 2007), linoleamide mediating Ca2+ flux (Lo et al., 2001), erucamide modulating water balance (Hamberger and Stenhagen, 2003), and elaidamide potentially functioning as an endogenous inhibitor of epoxide hydrolase (Morisseau et al., 2001). In this context, it is interesting to speculate whether the VLC primary amides occurring on the leaf surfaces of the P. aurea bamboo may also have biological functions, either affecting the physiological properties of the cuticle as a transpiration barrier or its ecological functions in interactions with microorganisms and herbivores.   It should also be noted that primary amides of unsaturated C20 and C22 fatty acids, such as erucamide, are produced on an industrial scale from seed oils of Limnanthes alba (Burg and Kleiman, 1991) and Crambe abyssinica (Nieschlag and Wolff, 1971), and serve as slip agents and antiblock agents for low-density polyethylene sheets (Shuler et al., 2004). Saturated amides, like stearamide and behenamide, have been used in the fabrication of water-repellent textiles (Nieschlag and Wolff, 1971).   While the biosynthetic pathways leading to fatty acid amides in plants are currently unknown, the biosynthesis of mammalian fatty acid amides has been investigated (Farrell and Merkler, 2008). There is experimental support for three proposed mammalian biosynthesis pathways: (1) direct amidation of fatty acyl-CoA intermediates by ammonia, catalyzed by cytochrome c (Driscoll et al., 2007), (2) two-step oxidation of N-acylethanolamines to N-  68 acylglycines (Chaturvedi et al., 2006; Prusakiewicz et al., 2002) and their cleavage by peptidylglycin α-amidating monooxygenase (PAM), and (3) a combined pathway involving cytochrome c-mediated production of N-acylglycines followed by PAM oxidation to the primary fatty acid amide (Mueller and Driscoll, 2007). It is tempting to speculate that the bamboo acid amides are formed on a pathway similar to (1) above, where ammonia directly amidates VLC acyl-CoA thioesters generated in the epidermal ER by the fatty acyl elongase (FAE) enzymatic complex. This scenario is assuming the presence of ammonia in or near the ER, possibly released during amino acid metabolism (Rosler et al., 1997). However, a plant pathway involving N-acylglycine intermediates must also be considered.          3.3.2. Compound class gradients between the epicuticular and intracuticular layers lining P. aurea leaves  In a second set of experiments, all compounds were quantified within the epicuticular and intracuticular wax layers of both young and old bamboo leaves. Based on previous experience with diverse other plant species (Buschhaus et al., 2007a, 2007b; Jetter and Schäffer, 2001), three consecutive treatments with gum Arabic were used to selectively remove the epicuticular wax layer, followed by extraction with CHCl3 to sample the intracuticular wax.  The epicuticular wax on the adaxial side of young leaves amounted to 1.81 ± 0.08 μg/cm2, while the adjacent intracuticular wax had a significantly lower coverage of 1.69 ± 0.06 μg/cm2 (p<0.05). In comparison, the epicuticular wax load on old leaves was 1.92 ± 0.06 μg/cm2,   69 again accompanied by significantly lower amounts of intracuticular wax at 1.71 ± 0.09 μg/cm2 (p<0.01).  The relative amounts of almost all compound classes differed between the epicuticular and intracuticular wax layers on young and old leaves (Fig. 3.3A/B). On the adaxial side of young leaves, free fatty acids 3.2, alkyl esters 3.4, aldehydes 3.5, alkanes 3.6 and amides 3.1 accumulated preferentially in the epicuticular layer, while primary alcohols 3.3 and terpenoids 3.7 were found at higher concentrations in the intracuticular wax (Fig. 3.3A). On old leaves, similar gradients were observed (Fig. 3.3B), except that alkanes 3.6 were found evenly distributed between layers. Interestingly, fatty acid amides 3.1 were found exclusively in the epicuticular wax of both young and old leaves.      Overall, our findings on compound class gradients between epi- and intracuticular wax layers on bamboo leaves are in accordance with reports from other plant species, with higher concentrations of primary alcohols 3.3 in the intracuticular compartment, and preferential accumulation of alkanes 3.6, aldehydes 3.5 and alkyl esters 3.4 in the epicuticular wax (Buschhaus and Jetter, 2012; Buschhaus et al., 2007a; Gniwotta et al., 2005; Racovita et al., 2015; Riedel et al., 2007; Wen et al., 2006). It has been speculated that such gradients may be due to the higher polarity of alcohols 3.3 and their hydrogen-bonding capability towards oxygen atoms of cutin (Buschhaus and Jetter, 2011), and our results thus support this hypothesis.       70 .AcidsAlcoholsAlkyl estersAldehydesAlkanesAmidesTerpenoidsNot identified   Relat ive composition [% of t otal wax]01020304050   Relative composition [ % of  t ot al wax]0102030405060EpicuticularIntracuticularAB * * * * * * * * * * *  * * Figure 3.3. Compound class distribution within the epicuticular and intracuticular wax mixtures coating the adaxial side of P. aurea leaves. Relative abundances (mass %) of compound classes in each wax layer on leaves of (A) young and (B) old plants. Bars represent mean ± standard deviation (n = 4). Asterisks mark discovery of significant differences between arcsin-transformed percentages based on Student’s t-test (p < 0.05, FDR 5%).   71  Fatty acids 3.2 accumulated in the epicuticular wax of bamboo leaves, again in accordance with previous reports on other species (Jetter and Schäffer, 2001; Racovita et al., 2015; Wen et al., 2006b). This finding is of special interest, since fatty acids 3.2 are known to form H-bonded head-to-head dimers in the solid state, resulting in relatively low overall polarities similar to alkyl esters 3.4 (Bond, 2004; Leiserowitz, 1976; Moreno et al., 2006). The accumulation of fatty acids 3.2 together with esters 3.4 in the epicuticular wax thus further underpins the hypothesis that epi-/ intracuticular partitioning is driven by polarity of wax compounds.   Interestingly, the bamboo fatty acid amides 3.1 accumulated exclusively in the epicuticular wax layer, an extreme partitioning unprecedented in the plant wax literature. This behaviour again suggests relatively low polarity, and might be explained by hydrogen-bonded molecular associations in the solid state, as described for crystals of tetradecanamide (Turner and Lingafelter, 1955) and decanamide (Brathovde and Lingafelter, 1958), as well as monolayers of dodecanamide (Bhinde et al., 2010a) and hexadecanamide (Bhinde et al., 2010b) deposited on graphite substrates. Of note, it has been shown that fatty acid amides 3.1, added as slip/antiblock agents to bulk polyolefin matrices, spontaneously migrate to the polymer surface (similar to those in the bamboo cuticle), where they serve to reduce the coefficient of friction and thus prevent the adherence of polymer sheets to one another (Dragnevski et al., 2009; Ramirez et al., 2002).     72  3.3.3. Individual compound gradients between the epicuticular and intracuticular layers lining P. aurea leaves  The chain length profiles of aliphatic compound classes were similar between the epicuticular and intracuticular wax mixtures, for both young and old plants (Fig. 3.4A/B). In particular, the primary alcohol 3.3 and aldehyde 3.5 fractions each had similar homolog distributions in the epicuticular and intracuticular waxes of both young and old leaves. In contrast, free fatty acids 3.2 showed bimodal chain length distribution with maxima at C16 and C28 in the epicuticular leaf wax of old leaves, whereas the adjacent intracuticular acids exhibited a trimodal profile peaking at C16, C22 and C30. For young bamboo leaves, only the C18 and C22 acids showed significantly higher relative amounts in the intracuticular compartment compared with the epicuticular wax, while on old leaves C16-C22 acids accumulated in the intracuticular layer, and C24 and C28 acids in the epicuticular wax. Similar patterns have been reported for fatty acids on Taxus baccata needles, where acid homologs up to C24 accumulated preferentially in the intracuticular layer, and those longer than C24 in the epicuticular layer (Wen et al., 2006b).  The homolog distributions of alkyl esters 3.4 and alkanes 3.6 were fairly similar between the epicuticular and intracuticular layers on both young and old leaves (Fig. 3.4A/B). The primary amides 3.1, found only in the epicuticular wax, were dominated by the C30 homolog 3.1c. The shorter homologs (C26 and C28) were more abundant than the longer ones (C32 and C34) on the adaxial surface of young leaves, while old leaves showed the opposite pattern.   73 1618202224262830323422242628303234363840424446485052242628303234252729312628303234TocopherolTocopherolGlutinolAmyrinAmyrinLupeolIsomultiflorenolEpifriedelanolGlutinyl palmitateAmyrenyl palmitateAmyrenyl palmitateIsomultiflorenyl palmitate    Rel at ive composi tion   [ % of  compound class]010203040500102030405060EpicuticularIntracuticularAcids AlcoholsAlkanesAldehydesEstersAmidesAB .******* ****** ******************** Figure 3.4. Single constituent distribution within compound classes in the epicuticular and intracuticular wax layers on the adaxial side of P. aurea leaves. Relative abundances (mass %) of individual VLC homologs or terpenoid isomers within each compound class on (A) young and (B) old leaves. Bars represent mean ± standard deviation (n = 4). Asterisks mark discovery of significant differences between arcsin-transformed percentages based on Student’s t-test (p < 0.05, FDR 5%). Minor odd-numbered acid, alcohol and aldehyde homologs and even-numbered alkane homologs were omitted for clarity (and thus respective groups of bars do not add up to 100%).    74  Most of the cyclic compounds also showed concentration gradients between the wax layers. In particular, γ-tocopherol 3.7a was found in greater proportion in the epicuticular layer of old leaves, while it was evenly distributed on young ones (Fig. 3.4A/B). α-Tocopherol 3.7b was also found in greater proportion in the epicuticular wax of old leaves. Glutinol 3.7c, β-amyrin 3.7e, and α-amyrin 3.7f accumulated preferentially in the intracuticular layer on both young and old leaves. In contrast, isomultiflorenol 3.7g and epifriedelanol 3.7d were found mainly in the epicuticular wax compartment on old leaves. Lupeol 3.7h was found evenly distributed between compartments regardless of plant age. Finally, glutinyl palmitate 3.7i and isomultiflorenyl palmitate 3.7l exhibited preferential accumulation in the epicuticular layer of young and old leaves.   Overall, terpenoids 3.7 thus accumulated to higher concentrations in the intracuticular layer, likely due to their more compact molecular geometry and polar functional groups, again consistent with many previous studies (Buschhaus and Jetter, 2012; Buschhaus et al., 2007a, 2007b; Racovita et al., 2015; Riedel et al., 2007, 2003). A notable exception to this pattern were triterpenoid esters 3.7i-l, with partitioning behaviour opposite to that of free triterpenols. This new finding suggests that the presence of an alkyl side chain, masking the polar functionality and substantially increasing the molecular volume, may drastically decrease their overall polarity and make the terpenoid esters 3.7i-l partition into the epicuticular layer (together with the alkyl esters 3.4).            75  3.3.4. Ester isomer profiles in the epicuticular and intracuticular layers on P. aurea leaves  In a separate experiment, the distribution of alkyl and acyl moieties within each of the alkyl ester homologs was determined. Since the ester metamers, comprising various combinations of acyls and alkyls all with the same total chain length, could not be resolved chromatographically, they had to be profiled based on mass spectral information. Average mass spectra were acquired from each ester homolog peak from epicuticular and intracuticular wax at both leaf ages, and relative abundances of product ions produced by McLafferty rearrangement with double hydrogen transfer (R-CO2H2+) were used to calculate isomer distributions, as reported before (Lai et al., 2007; Racovita et al., 2015; Razeq et al., 2014).   The ester homologs 3.4 within the epicuticular wax on young bamboo leaves had isomer compositions grouping them according to total ester chain length (Fig. 3.5A). The short ester homologs, with total carbon numbers C36-C40 3.4a-c, were mainly formed by acids with gradually increasing chain lengths (C16-C20) in combination with the same alcohol (C20). In contrast, the long ester homologs, C46-C52 3.4f-i, incorporated primarily one acid (C22) linked to various alcohols (C24-C30). The most abundant ester homolog, C48 3.4g, was thus formed mainly by combination of C22 acid and C26 alcohol. The mid-range ester homologs, C42-C44 3.4d-e, had isomer compositions transitioning between the patterns of the shorter and longer esters, on the one hand comprising relatively large percentages of C22 acid and on the other hand C20 alcohol, accompanied by various other combinations of acid and alcohols chain lengths. The epicuticular esters on old leaves had isomer compositions very similar to those on young leaves (Fig. 3.5B).    76 Ester chain length36 38 40 42 44 46 48 50 52102030405060708090100AIntracuticularEster chain length36 38 40 42 44 46 48 50 52102030405060708090100BIntracuticular        Relative isomeric composition                     [% of homolog]0102030405060708090100        Relative isomeric composition                    [% of  homolog]0102030405060708090100C14 acidC16 acidC18 acidC20 acidC22 acidC24 acidC26 acidC28 acidEpicuticular Epicuticular  Figure 3.5. Relative isomer compositions of ester homologs in epicuticular and intracuticular waxes on the adaxial side of young and old P. aurea leaves. Relative abundances (mass %) of individual metamers (indicated by their acyl moiety chain length) within each ester homolog in leaf wax from (A) young and (B) old plants. Bars represent mean ± standard deviation (n = 4, except for old leaf intracuticular n = 3).  The intracuticular esters 3.4 on both young and old leaves closely resembled those in the adjacent epicuticular layers (Fig. 3.5), with only a few shifts in the chain length profiles of esterified acid and alcohol moieties. In particular, the intracuticular C40 ester 3.4c contained higher proportions of C16, C22 and C24 acids than its epicuticular counterpart, irrespective of plant age. In contrast, the C42 3.4d and C44 3.4e esters had much lower percentages of C16 acid metamers than respective epicuticular ester homologs. Finally, the longest ester homologs, C46-C50 3.4f-h, comprised relatively large amounts of esterified C24 acid as well. Overall, the wax   77 esters 3.4 in the intracuticular wax were thus characterized by a slightly broader chain length distribution than those in the epicuticular layer. It should be noted that the isomer composition of the least abundant ester homologs (C36 3.4a, C38 3.4b and C52 3.4i) could not be determined, as their concentrations were below the MS detection limit.  Overall, our ester isomer analyses suggest that P. aurea leaves may harbor up to three wax ester synthases with different substrate chain length preferences to produce most of the ester isomers and homologs. Namely, one synthase may prefer C22 and (to lesser extent) C24 acyl-CoA for formation of C42-C52 esters 3.4d-i, a second enzyme may prefer C20 primary alcohol to form C36-C40 esters 3.4a-c, and a third enzyme C16 acyls for formation of C42 3.4d and C44 3.4e esters. However, the amounts and isomer composition of the latter two ester homologs may also be explained by availability of substrates (rather than enzyme specificity), as C26 and C28 alcohols had relatively high concentrations (compare Fig. 3.3A/B) and C16 acyl-CoA, as the first precursor common to many ER-localized lipid pathways (Lai et al., 2007; Ohlrogge and Browse, 1995), may also be expected to be abundantly available at the site of biosynthesis in the epidermal ER.  The cumulative amounts of esterified acids could be determined by pooling the percentages of all acyl moieties in individual ester homologs, taking coverages (g/cm2) of individual homologs into account and including triterpenoid esters. Independent of plant age, both the epicuticular and intracuticular esters were dominated by C16 acid, primarily due to the large amounts of triterpenoid palmitates (Fig. 3.6A/B). There were no significant differences between epicuticular and intracuticular amounts of esterified C16 acid at both ages. In contrast, most of the other acids accumulated in substantially lower amounts in the esters of the intracuticular layer of both old and young leaves.   78 Total ester ified acids [% of compound class]0.010.020.030.040.050.060.070.0EpicuticularIntracuticularA14 16 18 20 22 24 26 280.010.020.030.040.050.060.0BAcid chain lengthTotal  est erified acids [% of compound class] *****  Figure 3.6. Profile of total esterified acids in the epicuticular and intracuticular waxes on the adaxial side of young and old P. aurea leaves. Relative abundance (mass %) of each esterified acid homolog on the adaxial side of (A) young and (B) old P. aurea leaves. Bars represent mean ± standard deviation (n = 4). Asterisks mark discovery of significant differences between arcsin-transformed percentages based on Student’s t-test (p < 0.05, FDR 5%).    79 Similarly, the chain length profiles of total esterified primary alcohols and triterpenols could be calculated, and were also found to be very similar between young and old leaves (Fig. 3.7A/B). Both the epicuticular and intracuticular esterified alcohol profiles were bimodal, peaking at C26/C24 and, to a lesser degree, at C20. The majority of alcohol homologs accumulated preferentially in the epicuticular waxes. Among the triterpenols, only glutinol was found more in the epicuticular esters than the intracuticular.     80 Total esterified alcohols [ % of compound class]0.05.010.015.020.025.030.035.0A18 20 22 24 26 28 30 32 34 360.05.010.015.020.025.030.0EpicuticularIntracuticularBAmyrinAmyrinGlutinolIsomultiflorenolTotal esterified alcohols [ % of compound class]n-Alcohol chain length*********** Figure 3.7. Profile of total esterified alcohols in the epicuticular and intracuticular waxes on the adaxial side of young and old P. aurea leaves. Relative abundance (mass %) of each esterified VLC alcohol homolog and terpenol isomer on the adaxial side of (A) young and (B) old P. aurea leaves. Bars represent mean ± standard deviation (n = 4). Asterisks mark discovery of significant differences between arcsin-transformed percentages based on Student’s t-test (p < 0.05, FDR 5%).   81 Chapter 4: Composition of cuticular waxes coating flag leaf blades and peduncles of Triticum aestivum cv. Bethlehem   4.1. Introduction  The leaves, non-woody stems, flowers and fruit of terrestrial plants are coated by cuticles to prevent uncontrolled water loss. This lipid layer is composed of the polyester cutin and a mixture of cuticular wax, some of which is embedded in the cutin matrix (intracuticular wax) and some of which lies on top of it (epicuticular wax) (Buschhaus and Jetter, 2011). Cuticular wax composition varies both qualitatively and quantitatively between plant species, and in some cases between organs or even tissues of the same species.   Waxes are complex mixtures commonly composed of monofunctional derivatives of very-long-chain (VLC, i.e. >C20) fatty acids, including primary alcohols (1-alkanols), alkyl esters, aldehydes, alkanes, secondary alcohols, ketones, and free fatty acids (Jetter et al., 2006). Alicylic compounds, such as triterpenoids, are also often found mixed together with VLC aliphatics in cuticular waxes, sometimes in greater amount than the latter (Belge et al., 2014; Bianchi et al., 1993; Markstadter et al., 2000; Nordby and McDonald, 1994; van Maarseveen et al., 2009). Lastly, in several plant species, aromatic wax compounds have been reported as well, such as alkylresorcinols (Adamski et al., 2013; Ji and Jetter, 2008),  alkyl m-guaiacols (Basas-Jaumandreu et al., 2014), benzyl and phenethyl esters (Gülz and Marner, 1986; Jetter and Riederer, 1996; Rapley et al., 2004).   82   VLC aliphatic waxes occur as homologous series, typically with consecutive homologs separated by two methylene (CH2) units as a consequence of their biosynthesis. Extensive genetic and biochemical studies have led to a comprehensive understanding of cuticular wax biosynthesis in the model plant species Arabidopsis thaliana (Kolattukudy, 1970; Li-Beisson et al., 2010; Samuels et al., 2008). Accordingly, plant wax biosynthesis begins in the plastids of epidermal cells, where long-chain (LC, C16-C18) fatty acids are synthesized de novo. Next, these are transferred to the endoplasmic reticulum (ER), where a fatty acyl-CoA elongase (FAE) enzymatic complex extends them in increments of two carbons to VLC fatty acyl-CoA thioesters, which are then processed via two major pathways. On the acyl reduction pathway, fatty acyl-CoA reductase (FAR) enzyme(s) generate(s) fatty 1-alkanols, by reduction via fatty aldehyde intermediates that are not released, and wax ester synthase(s) link these fatty 1-alkanols with other (V)LC acyl-CoAs into alkyl esters. On the decarbonylation pathway, another reductase transforms fatty acyl-CoA substrates into aldehydes, which are then decarbonylated to n-alkanes. In some species (including Arabidopsis), alkanes can be further hydroxylated by cytochrome P450 enzymes to secondary alcohols, ketones, diols and ketols (Greer et al., 2007; Wen and Jetter, 2009).   However, the Arabidopsis model of wax biosynthesis does not account for the whole range of VLC aliphatic wax structures encountered in diverse species. Valuable insight can thus be obtained by exploring biosynthetic pathways operating in other plant species. In particular, Poaceae are thought to have a third wax biosynthetic pathway leading to β-diketones (Kunst et al., 2006; von Wettstein-Knowles, 2012, 1976). Based on genetic and biochemical evidence for   83 barley (Hordeum vulgare), it was inferred that this pathway comprises three enzyme activities (CER-Q, CER-C, CER-U) residing on a single multi-domain protein (von Wettstein-Knowles and Sogaard, 1981, 1980), including at least one polyketide 3-ketoacyl-CoA synthase (pkKCS) and a cytochrome P450 hydroxylase. However, a detailed understanding of the β-diketone pathway is currently missing. Substantial progress on wax biosynthesis can thus be expected in the near future based on the recently assembled genome information for the Poaceae crops barley and wheat. The molecular genetic investigations will have to build on reliable data on the chemical composition of the waxes of these model species.   Numerous reports have described wax compositions of wheat over more than four decades (Bianchi et al., 1980; Tulloch and Weenink, 1969; Wang et al., 2015a, 2015b). On the one hand, these previous studies all gave very similar qualitative results, describing very similar sets of typical wax constituents matching those on most other plant species, together with one to four β-diketones. However, the large majority of the investigations were carried out fairly long ago, with analytical equipment far inferior to modern instruments. Accordingly, many important wax components may have been missed, and specialty wax components that could inform speculations on wheat wax biosynthesis pathways and specific enzymes have not been described to date.   On the other hand, several previous reports suggested much quantitative variation in the relative amounts of ubiquitous wax compounds within respective wheat wax mixtures. Some of this variation has been ascribed to differences between wheat species or cultivars with different genetic backgrounds. Substantial variation was also linked to plant age, likely due to changes in   84 plant architecture and organ composition during plant ontogeny. However, many of the previous wheat wax studies focused either only on leaves or analyzed whole plants without distinguishing organs (Bianchi and Corbellini, 1977; Bianchi et al., 1980; Tulloch and Hoffman, 1973, 1971; Tulloch and Weenink, 1969; Tulloch, 1973). Only a few, mostly recent studies showed distinct wax compositions, in particular between leaves, stems and inflorescence parts (Adamski et al., 2013b; Wang et al., 2015a, 2015b). The findings thus far match those carried out in much more detail for barley (Lundqvist and von Wettstein, 1962; Lundqvist et al., 1968; von Wettstein-Knowles, 1969) pointing to particular differences between the lower and the upper portions of the wheat plants. Thus, the (lower) leaves seem to have wax mixtures dominated by 1-alkanols, while the inflorescence waxes contain relatively large percentages of -diketones. The organs in the transition zone are of particular interest, including the flag leaf (i.e. the top-most leaf directly below the inflorescence) and the peduncle (i.e. the stem segment between flag leaf and inflorescence).  In the light of the scattered information on wheat cuticle chemistry, the goal of the present investigation was to provide a comprehensive wax analysis of a wheat cultivar that is concurrently used in molecular genetic and biochemical investigations into the biosynthesis of wheat waxes (Hen-Avivi et al., 2016). To this end, the bread wheat (Triticum aestivum L.) cultivar Bethlehem was chosen, and wax analyses were performed using GC-MS to elucidate wax component structures, GC-FID for accurate quantification of individual compounds, and further quantification based on characteristic MS fragments to also assess isomer profiles within some of the compound classes. We concentrated our analyses on the blade of the uppermost leaf   85 (the flag leaf) and the top stem portion (the peduncle), as two organs expected to have 1-alkanol- and -diketone-dominated waxes, respectively.   4.2. Experimental  4.2.1. Plant material and preparation of total wax extracts  Triticum aestivum cv. Bethlehem plants were grown continuously in a greenhouse at the Weizmann Institute of Science (Rehovot, Israel) in a 12-14 h / 10-12 h light/dark cycle at 24-26ºC / 17-18ºC, respectively. A mix of 50% peat and 50% turf was used for plant growth, and plants were watered every 3-4 days with 300-500 mL water per 5 L pot. Flag leaves and peduncles were excised from mature plants using clean razor blades during the month of August 2013. Typically, one flag leaf blade with a total area of 40-50 cm2 (both sides) and, respectively, one peduncle with a cylindrical area of approx. 20 cm2 represented one biological replicate. Exact areas of leaf blades were determined by photographing them and using the ImageJ software to calculate the area of one side, then multiplying by 2 to account for both leaf sides. Exact areas of peduncles were calculated using the formula for lateral surface of a cylinder (π x D x L), after measuring the diameter D and length L of peduncles with a ruler.  Leaf specimens were rolled up and peduncles cut into smaller pieces such as to be fully submerged into 10 mL CHCl3 (Aldrich, ≥99%, 0.75% ethanol as stabilizer) containing also 5 μg of n-tetracosane (Alfa Aesar, ≥99%) as internal standard. After stirring for 30 s at room temperature, the CHCl3 was transferred to another vial, and a fresh portion of 10 mL CHCl3 was   86 added to the plant material and the extraction repeated for another 30 s. The two combined extracts were evaporated at 50ºC under a stream of N2 (Praxair, ≥99.998%), until the volume was low enough for transfer to 2 mL GC autosampler vials.  4.2.2. Derivatization reactions  In preparation for GC analysis, all samples were derivatized by silylation as follows: samples were taken to dryness under N2, then 10 μL N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA, Aldrich, GC grade) and 10 μL pyridine (Aldrich, ≥99.8%, anhydrous) were added, the vial was sealed and the mixture was refluxed at 70ºC for 20 min. Then, excess reagents were completely removed under N2 and the sample re-dissolved with 50 μL CHCl3. Under such conditions, acidic OH groups underwent silylation, but the enol tautomers of β-diketones did not to any significant extent. By increasing derivatization time to 30-120 min, some of the β-diketones formed trimethylsilyl ethers of their enol tautomers, but no complete conversion was achieved for any of the derivatization times tested. Thus, the 20 min derivatization time was chosen in all quantitative experiments to prevent silylation of β-diketones, as recommended by other authors (Stefan Schulz et al., 2000).    For further confirmation of hydroxyl group location in hydroxy-β-diketones, one specimen was acetylated prior to silylation and GC-MS analysis, as follows: to the dry wax mixture dissolved in 10 μL pyridine, 10 μL acetic anhydride (Aldrich, ≥98%) were added and the mixture was refluxed for 5 min at 70ºC, then was stored overnight at room temperature, before evaporating excess reagents and performing silylation as described above. For the same scope, a   87 second specimen was fully reduced, as follows: to the dry wax mixture dissolved in 50 μL diethyl ether (Aldrich, ≥99.7%, anhydrous, 1 ppm BHT as inhibitor),  0.1 mg LiAlH4 (Aldrich, ≥95%) were added, the vial was sealed and left to react at 70ºC overnight. Then, the reaction mixture was quenched with 10% aqueous H2SO4 and extracted three times with 60 μL each of fresh diethyl ether, the ether extracts were combined and evaporated under N2, and silylation was performed as described above.  4.2.3. GC-MS and GC-FID analysis         Two different GC instruments were used for separation, identification and quantification of wax constituents, both equipped with the same type of capillary GC column (6890N, Agilent, Avondale PA, USA; 30 m long; type HP-1: 100% PDMS; 0.32 mm i.d.; df=0.1 µm), both with on-column injection at 50ºC and programmed to follow the same temperature program (2 min at 50ºC, ramp 40ºC/min to 200ºC, constant for 2 min, ramp 3ºC/min to 320ºC, constant for 30 min). The first GC instrument employed He (Praxair, ≥99%) as carrier gas, at a flow rate of 1.4 mL/min, and was equipped with MS detector (5973N, Agilent, EI-70 eV, ionization source temperature 240ºC), serving primarily the purpose of identification of compounds in the plant wax mixtures. The second used H2 (Praxair, ≥99.95%) as mobile phase at 2.0 mL/min and was equipped with an FID detector, for quantification of individual wax homologs based on normalization of peak areas against that of the internal standard. The relative response factors of all cuticular wax classes with respect to the internal standard were approximated to 1.00, in agreement with past reports using the same GC operation conditions (Riederer and Schneider, 1989).   88 4.3. Results   The present work aimed to provide a comprehensive analysis of the wax mixtures coating select vegetative organs of mature wheat. Specifically, the blade of the flag leaf and the peduncle were to be investigated, as examples of surfaces characterized by the presence of platelet-shaped wax crystals and tubule-shaped wax crystals, respectively (Wang et al., 2015a). In the following, the results will be described in direct comparisons between both organs, first for the overall wax composition (2.1) and the proportions of acyl monomers contained in the mixtures (2.2), then the chain length distribution (2.3) and isomer composition of the ester-linked dimers (2.4), and finally the compositions of the polyketide and terpenoid wax constituents (2.5).  4.3.1. Overall composition of wheat flag leaf blades and peduncles  Preliminary screening by gas chromatography-mass spectrometry (GC-MS) revealed well over 100 different compounds in each of the two wax mixtures extracted from wheat flag leaf blades and peduncles. The large majority of these were present in both wheat wax mixtures, and they were identified as compounds falling into 14 different classes (Fig. 4.1). Four of these constituent classes were recognized as free fatty acids, aldehydes, n-alkanes, and 1-alkanols, hence compounds found ubiquitously in the wax of many species. Five other compound classes found on both wheat organs were identified as esters combining acyl moieties with various aliphatic or aromatic alcohols. Further wheat wax constituents were assigned polyketide structures, two of them purely aliphatic (ß-diketone and hydroxy-ß-diketones) and two combining aliphatic and aromatic moieties (alkylresorcinols and methyl alkylresorcinols). Finally, a few compounds found only in peduncle wax were identified as terpenoids.   89 CH3OOHnCH3 OHnCH3 On4.1a (n=9); ... ; 4.1g (n=15) 4.2a (n=9); ... ; 4.2h (n=16) 4.3a (n=11); ... ; 4.3f (n=16)CH3Hn4.4a (n=12); ... ; 4.4e (n=16)CH3OO CH3n m4.5a (n+m=17); ... ; 4.5i (n+m=25)CH3OO CH3CH3n m4.6a (n+m=12); ... ; 4.6e (n+m=16)CH3OOn4.7a (n=13); ... ; 4.7c (n=15)CH3OOn4.8a (n=10); ... ; 4.8e (n=14)CH3OOOHn4.9a (n=13); ... ; 4.9d (n=16)CH3OCH3O6 7      4.10aOCH3OCH3OH2 7      4.11a3CH3 OHOHnCH3 OHOHCH3n4.12a (n=9); ... ; 4.12f (n=14) 4.13a (n=9); ... ; 4.13f (n=14)OOHCH3CH3CH3CH3CH3CH3 CH32        4.14aCH3CH3CH3HCH3CH3HCH3HCH3 CH3RO4.14b (R=H)4.14e (R=C21H43CO)CH3HCH3CH3HCH3HCH3 CH3ROCH3CH34.14c (R=H)4.14f (R=C21H43CO)HCH3CH3HCH3HCH3 CH3OHCH3HCH3CH2               4.14dOCH3OCH3OH2 7      4.11b3 Figure 4.1. Compounds identified in the total wax mixtures of T. aestivum cv. Bethlehem flag leaf blades and peduncles.   90  While flag leaf blade and peduncle waxes were found to share most compound classes, they differed drastically in the relative proportions of these constituents (Fig. 4.2). The mixture extracted from flag leaves was dominated by 1-alkanols 4.2, representing 55% of the GC-detected compounds (8.82 ± 0.73 μg/cm2) (Fig. 4.2A). They were accompanied by substantial amounts of alkanes 4.4 (9%; 1.46 ± 0.15 μg/cm2), 1-alkanol esters 4.5 (9%; 1.51 ± 0.14 μg/cm2), ß-diketone 4.10 (6%; 0.92 ± 0.25 μg/cm2) and hydroxy-ß-diketones 4.11 (8%; 1.22 ± 0.40 μg/cm2). Relatively small portions of acids 4.1 (1%; 0.15 ± 0.03 μg/cm2), aldehydes 4.3 (3%; 0.40 ± 0.07 μg/cm2), benzyl esters 4.7 (0.2%; 0.03 ± 0.004 μg/cm2), phenethyl esters 4.8 (0.1%; 0.02 ± 0.004 μg/cm2), p-hydroxyphenethyl esters 4.9 (1%; 0.18 ± 0.06 μg/cm2), 2-alkanol esters 4.6 (1%; 0.16 ± 0.03 μg/cm2), alkylresorcinols 4.12 (2%; 0.25 ± 0.02 μg/cm2) and methyl alkylresorcinols 4.13 (2%; 0.33 ± 0.02 μg/cm2) were present. Taken together, 96% of the wax mass detected by GC could be identified. The wheat flag leaf had a total wax coverage of 16 ± 1 μg/cm2, including all identified and unidentified compounds.   In contrast to the flag leaf wax, the mixture extracted from wheat peduncles was dominated by ß-diketone 4.10 (37%; 17.91 ± 1.92 μg/cm2) and hydroxy-ß-diketones 4.11 (44%; 21.20 ± 2.38 μg/cm2) (Fig. 4.2B). Alkanes 4.4 (7%; 3.40 ± 0.77 μg/cm2), 1-alkanols 4.2 (2%; 1.13 ± 0.14 μg/cm2) and 1-alkanol esters 4.5 (2%; 1.13 ± 0.09 μg/cm2) were present at much lower concentrations than in the leaf wax mixture. Relatively small portions of acids 4.1 (1%; 0.44 ± 0.05 μg/cm2), aldehydes 4.3 (1%; 0.46 ± 0.06 μg/cm2), benzyl esters 4.7 (0.2%; 0.10 ± 0.02 μg/cm2), phenethyl esters 4.8 (0.1%; 0.05 ± 0.01 μg/cm2), p-hydroxyphenethyl esters 4.9 (2%; 0.84 ± 0.23 μg/cm2), 2-alkanol esters 4.6 (1%; 0.51 ± 0.06 μg/cm2), alkylresorcinols 4.12 (0.3%;   91 0.17 ± 0.03 μg/cm2) and methyl alkylresorcinols 4.13 (0.3%; 0.16 ± 0.01 μg/cm2) were detected in the peduncle wax, thus in relative amounts similar to those in flag leaf wax. Terpenoids 4.14 accumulated to 0.6% (0.27 ± 0.07 μg/cm2) in the peduncle wax mixture, leaving only 2% of the wax mixture unidentified. The wheat peduncle had a total wax coverage of 49 ± 5 μg/cm2, and thus more than triple the wax quantity covering the flag leaf blade.   92 Acids1-AlkanolsAldehydesAlkanes1-Alkanol esters2-Alkanol estersBenzyl estersPhenethyl estersp-Hydroxyphenethyl estersDiketonesHydroxy-DiketonesAlkylresorcinolsMethyl alkylresorcinolsTerpenoidsNot identifiedCover age [g/cm2]0123418 ± 2 g/cm221 ± 2 g/cm2Coverage [g/ cm2]0.00.51.01.52.08.8 ± 0.7 g/cm2.AB Figure 4.2. Compound class compositions of total wax mixtures of T. aestivum cv. Bethlehem. Coverages (μg/cm2) of compound classes within wax mixtures covering the (A) flag leaf blade and (B) peduncle. Bars represent mean ± standard deviation (n = 5).      93  4.3.2. Chain length distributions of common wheat compound classes   The compound classes typically found in plant cuticular waxes (fatty acids, 1-alkanols, aldehydes, alkanes) were all present in the wheat wax mixtures as extended series of homologs (Fig. 4.3). The acid fraction 4.1 of the flag leaf wax was characterized by the presence of even-numbered homologs, ranging from C20 to C32 and with a bimodal distribution, having a broad maximum between C26 and C30 and a second maximum at C22.  The peduncle wax contained the same acid homologs 4.1, largely dominated by C28 acid alone and with a relatively even distribution across all other chain lengths. The primary alcohol (1-alkanol) fraction 4.2 within the wheat flag leaf wax mixture comprised even-numbered homologs between C24 and C32, with a very strong predominance of C28. The peduncle wax contained the 1-alkanol homologs 4.2 from C20 to C34, albeit with a more even distribution peaking both at C24 and at C30/C32.   The flag leaf wax comprised even-numbered aldehydes 4.3 ranging from C24 to C34, with C28 aldehyde dominating and accompanied by significant amounts of C30 and C32 aldehydes. The peduncle wax again contained the same homologs 4.3, but with relatively little C28 aldehyde and predominantly C30 and C32 aldehydes instead. Finally, the alkane fractions 4.4 in the wax mixtures from both wheat organs were characterized by odd-numbered homologs from C25 to C33, with very similar chain length distributions in the flag leaf and peduncle waxes peaking at C31 alkane. Odd-numbered acids, 1-alkanols, aldehydes, and even-numbered alkanes were only found in traces in both the flag leaf and peduncle waxes.    94   Relative compositi on [% of compound class]0102030405060708090100Acids 1-Alkanols AlkanesAldehydes2022242628303220222426283032342426283032342527293133  Relative compositi on [% of compound class]0102030405060708090 Figure 4.3. Chain length distributions within common compound classes of T. aestivum cv. Bethlehem waxes. Relative abundances (%) of individual homologs from each of four common compound classes in the wax mixtures covering the (A) flag leaf blade and (B) peduncle. Numbers on the x-axis indicate homolog chain length. Bars represent mean ± standard deviation (n = 5). Not all bar groups add up to 100% because minor homologs of opposite parity have been omitted for clarity.   95  4.3.3. Chain length distributions of wheat wax esters   Five of the compound classes found equally in wheat flag leaf and peduncle wax mixtures were characterized by ester linkages between (very-) long-chain fatty acids and various alcohols in extended series of homologs (Fig. 4.4). The 1-alkanol esters 4.5, resulting from esterification of acyls with primary alcohols, had overall carbon numbers ranging from C38 to C54 in both wheat wax mixtures. In the flag leaf wax, 1-alkanol esters 4.5 were found to have a bimodal chain length distribution peaking at C44 and C50, whereas those in the peduncle wax were dominated by the C44 ester homolog alone. The esters 4.6 formed by acyls and 2-alkanols had total carbon numbers ranging from C29 to C37 in both wheat wax mixtures, with similar homolog compositions culminating at C35 in flag leaf and peduncle waxes alike.   The three remaining classes of esters were formed by (very-) long-chain fatty acids and aromatic alcohols. Representative mass spectra for one homolog from each class (4.7b, 4.8d and 4.9c) are shown in Fig. 4.5, and the corresponding fragmentation mechanisms in Fig. 4.6. One of the aromatic alcohols, benzyl alcohol, was esterified with C28, C30 and C32 fatty acids to form esters with overall carbon numbers of C35, C37 and C39, respectively (Fig. 4.4). Among these homologs, the one comprising C30 acid dominated in both flag leaf and peduncle waxes. Broader ranges of acyl chain lengths were found esterified with phenethyl and p-hydroxyphenethyl alcohols, with apparent preference for C24-C28 and C32 acyl chain lengths, respectively (Fig. 4.4).      96   Relative composition [ % of compound class]010203040506070803840424446485052542931333537(28+7) 35(30+7) 37(32+7) 39(22+8) 30(24+8) 32(26+8) 34(28+8) 36(30+8) 38(28+8) 36(30+8) 38(32+8) 40(34+8) 42  Relative compositi on [% of compound class]0102030405060701-Alkanolestersp-Hydroxy-phenethylestersPhenethylesters2-AlkanolestersBenzylestersAB Figure 4.4. Chain length distributions within wax ester classes in T. aestivum cv. Bethlehem waxes. Relative abundances (%) of individual homologs from each of five wax ester classes in the wax mixtures covering the (A) flag leaf blade and (B) peduncle. Numbers on the x-axis indicate total homolog chain length. For aromatic esters, the individual chain lengths of their acyl part plus alcohol part are specified in round brackets. Bars represent mean ± standard deviation (n = 5). All bar groups add up to 100%.   97 m/zm/zABCm/z50 100 150 200 250 300 350 400 450 500 550 600 650Rel. Abundance [%]020406080Rel. Abundance [%]020406080100Rel. Abundance [%]0204060807157108433834159145171104835740773 1771928357179 Figure 4.5. Representative mass spectra of aromatic esters.  Mass spectra of (A) benzyl triacontanoate 4.7b, (B) phenethyl octacosanoate 4.8d, and (C) trimethylsilyl derivative of p-hydroxyphenethyl dotriacontanoate 4.9c from the cuticular wax mixture of T. aestivum cv. Bethlehem. Insets show 10-fold enhancements of the 400-650 amu mass range.  m/zRel. Abundance [%]0Rel. Abundance [%]071571084338341591451711048357407542Mm/zm/zBRel. Abundance [%]0711048357407m/zm/zm/zRel. Abundance [%]0Rel. Abundance [%]8371104835740773 1771928357179429401 554M-90  98 OOCH314OOCH313OOCH3OSi(CH3)315  179  108 (+H)  451  91 -H2O433  104 (-H)  407  192 (-H) -CH3 177    73ABC Figure 4.6. Fragmentation schemes of aromatic esters. Major MS fragmentations of (A) benzyl triacontanoate 4.7b, (B) phenethyl octacosanoate 4.8d, and (C) trimethylsilyl derivative of p-hydroxyphenethyl dotriacontanoate 4.9c from the cuticular wax mixture of T. aestivum cv. Bethlehem.        99 4.3.4. Chain length distributions of esterified alcohols and acids: ester isomer compositions  The three ester classes comprising simple aromatic alcohols occurred as homologous series due to variation of acyl chain lengths (as described in 4.3.3.), but did not exhibit isomerism. In contrast, each homolog within the fractions of 1-alkanol 4.5 and 2-alkanol esters 4.6 occurred as a complex mixture of metamers, i.e. isomers fusing complementary chain lengths of the acyl and alkyl moieties (1-alkyl and 2-alkyl, respectively). Metamers of the same 1-alkanol ester homolog co-eluted under the current GC conditions, so their relative composition had to be determined based on mass spectral rather than chromatographic information. To this end, the relative intensities of all Macid+1 product ions detected within one GC-MS peak were quantified and used to calculate metamer percentages, as described previously (Lai et al., 2007; Racovita et al., 2015; Razeq et al., 2014; von Wettstein-Knowles and Netting, 1976a).   In the wheat flag leaf wax, the C44-C54 1-alkanol esters 4.5d-i comprised predominantly C28 alcohol, whereas C38-C42 1-alkanol esters 4.5a-c incorporated mainly C22 alcohol along with C24 alcohol (Fig. 4.7A). A greater diversity of alcohol chain lengths was found within the peduncle 1-alkanol esters 4.5, where the C38-C44 ester homologs 4.5a-d contained predominantly C22 alcohol, the C46-C52 homologs 4.5e-h predominantly C24 and C26 alcohols, and the C54 ester 4.5i predominantly C26 and C32 alcohols (Fig. 4.7B).   100 Relat ive isomer  composition   [% of  each est er homolog]0102030405060708090100C18 1-alkanolC20 1-alkanolC22 1-alkanolC24 1-alkanolC26 1-alkanolC28 1-alkanolC30 1-alkanolC32 1-alkanolEster chain length38 40 42 44 46 48 50 52 54Relative isomer  composi tion   [% of each est er homolog]0102030405060708090AB  Figure 4.7. Relative isomer compositions of 1-alkanol ester homologs in T. aestivum cv. Bethlehem waxes. Relative abundances (%) of individual metamers (indicated by chain length of their 1-alkanol moiety) within each ester homolog in the composition of wax mixtures covering the (A) flag leaf blade and (B) peduncle. Numbers on the x-axis indicate total homolog chain length. Bars represent mean ± standard deviation (n = 5). All bar groups add up to 100%.        101 The current GC conditions led to partial separation of metamer peaks for some 2-alkanol ester homologs, while others remained ill-separated. Therefore, the overall isomer composition of this ester class had to be determined from their mass spectra averaged over multiple GC peaks encompassing one homolog (Fig. 4.8). The relative quantification of 2-alkanol ester isomers 4.6 was based on summation of intensities of the three most intense product ions, Macid, Macid+1, and Malkyl-1, again as previously described (Aasen et al., 1971; von Wettstein-Knowles and Netting, 1976a). For both organs studied, the C29 2-alkanol ester 4.6a comprised mainly C7, C9 and C13 2-alkanols, the C31-C33 esters 4.6b-c mainly C7 and C13 2-alkanols, and the C35-C37 esters 4.6d-e primarily C7 and C15 2-alkanols (Fig. 4.9). Thus, all 2-alkanol ester homologs incorporated 2-alkanols with fairly similar bimodal distributions.   102 m/z50 100 150 200 250 300 350 400 450Relative abundance [%]0102030405060708090100m/z50 100 150 200 250 300 350 400 450Relative abundance [%]0102030405060708090100m/z50 100 150 200 250 300 350 400 450Relative abundance [%]0102030405060708090100A BE57FC DRetention time [min]21.6 21.8 22.0 22.2 22.4Intensity [a.u.]0100200300400500600700m/z 256 (C16 acid)m/z 284 (C18 acid)m/z 312 (C20 acid)m/z 340 (C22 acid)m/z 368 (C24 acid)m/z 210 (C15 2-alkanol)m/z 182 (C13 2-alkanol)m/z 154 (C11 2-alkanol)m/z 126 (C9   2-alkanol)m/z 98-97 (C7 2-alkanol)m/z50 100 150 200 250 300 350 400 450Relat ive abundance [%]0102030405060708090100m/z50 100 150 200 250 300 350 400 450Rel ative abundance [%]010203040506070809010071859783111210125213239256 257M466M466577185978311115412526929531231357718597831111821252412672842855771859783111126125577185978311198125323340341297351368369325  239 210CH3OO CH3CH37 6    257(+2H)    256  (+H)  (-H)CH3OO CH3CH38 5    267  285(+2H)    182  284  (+H)  (-H)CH3OO CH3CH39 4    295  313(+2H)    154  312  (+H)  (-H)CH3OO CH3CH310 3    323  341(+2H)    126  340  (+H)  (-H)CH3OO CH3CH311 2    351  369(+2H)    98  368  (+H)  (-H) Figure 4.8. Identification of C31 2-alkanol ester metamers. Mass spectra of the esters consisting of (A) C16 acid and C15 2-alkanol, (B) C18 acid and C13 2-alkanol, (C) C20 acid and C11 2-alkanol, (D) C22 acid and C9 2-alkanol, and (E) C24 acid and C7 2-alkanol. (F) Single ion chromatograms of fragments characteristic of fatty acid and 2-alkanol homologs constituting the C31 esters 4.6b, showing (partial) GC separation of the metamers.   103  Relat ive isomer composition   [% of each ester homol og]01020304050607080C7   2-alkanolC9   2-alkanolC11 2-alkanolC13 2-alkanolC15 2-alkanolC17 2-alkanolEster chain lengthRelative isomer composi tion   [ % of each ester homol og]01020304050607029 31 33 35 37AB Figure 4.9. Relative isomer compositions of 2-alkanol ester homologs in T. aestivum cv. Bethlehem waxes. Relative abundances (%) of individual metamers (indicated by chain length of their 2-alkanol moiety) within each ester homolog in the composition of wax mixtures covering the (A) flag leaf blade and (B) peduncle. Numbers on the x-axis indicate total homolog chain length. Bars represent mean ± standard deviation (n = 5). All bar groups add up to 100%.        104  Combining the quantitative information on the isomer compositions of all aliphatic esters (Figs. 4.4, 4.7 and 4.9), the aromatic esters (Fig. 4.4), and the composition of the two isomeric amyrenyl behenates (see below), the overall profile of ester-bound alcohols could be calculated for the flag leaf and peduncle wax mixtures (Fig. 4.10). The series of esterified 1-alkanols were dominated by the C28 homolog in the leaf wax mixture, similar to the chain length profile of the free (i.e. non-esterified) 1-alkanols in this wax mixture (compare Fig. 4.3A). In contrast, the esterified 1-alkanols of the peduncle wax had a broad distribution around C24, with only minor resemblance to the homolog profile of the accompanying free alcohols (compare Fig. 4.3B). The esterified 2-alkanols had overall chain length profiles peaking at C7 and at C13/C15, reflecting the bimodal distributions within all homologs of this ester class (compare Fig. 4.9). Among the esterified aromatic alcohols, p-hydroxyphenethyl alcohol was by far the most abundant in both wheat wax mixtures. It should be noted that, in the peduncle wax, this aromatic alcohol was found esterified in quantities comparable to those of individual homologs of primary alcohols. Finally, relative minor amounts of α- 4.14c and β-amyrin 4.14b were found esterified in the peduncle wax mixture (see below).    105   Relat ive composi ti on [%]012345678910Total esterified alcohol coverage: 0.84 ± 0.09 g/cm21820222426283032 7 911131517  Relative composition [%]051015201-AlkanolsBenzyl alcohol2-AlkanolsPhenethyl alcoholp-Hydroxyphenethyl            alcohol-Amyrin-AmyrinTotal esterified alcohol coverage: 0.9 ± 0.1 g/cm259 ± 5 %AB Figure 4.10. Relative compositions of esterified alcohols in T. aestivum cv. Bethlehem waxes. Relative total abundances (%) of all types of ester-bound alcohol homologs in the composition of wax mixtures covering the (A) flag leaf blade and (B) peduncle. Bars represent mean ± standard deviation (n = 5).   106  Similar to the profiles of esterified alcohols, the overall chain length profiles could be calculated for all ester-bound fatty acids taken together (Fig. 4.11). In both wheat wax mixtures, the total esterified acids had bimodal homolog distributions with maxima at C22 and C32. However, the absolute maxima were reversed, with predominance of C22 acid in the flag leaf esters and C32 in the peduncle esters. The latter finding is mostly due to the much higher p-hydroxyphenethyl ester 4.9 load on peduncles (compare Figs. 4.2 and 4.4).  Acid chain length14 16 18 20 22 24 26 28 30 32 34Rel ative composition [%]05101520253035Total esterified acid coverage: 1.6 ± 0.3 g/cm2Relat ive compositi on [%]05101520253035 Total esterified acid coverage: 0.9 ± 0.1 g/cm2AB Figure 4.11. Relative compositions of esterified acids in T. aestivum cv. Bethlehem waxes. Relative total abundances (%) of ester-bound fatty acid homologs in the composition of wax mixtures covering the (A) flag leaf blade and (B) peduncle. Bars represent mean ± standard deviation (n = 5).    107 4.3.5. Composition of wheat wax polyketides and terpenoids   Finally, the wheat wax mixtures contained several classes of polyketides and terpenoids, some of them comprising series of homologs and others isomers. One of the aliphatic polyketide fractions consisted of a single compound, hentriacontane-14,16-dione 4.10a, in both flag leaf and peduncle waxes. In contrast, two hydroxy-β-diketones 4.11a-b with differing hydroxyl group positions were identified based on characteristic MS fragments of various derivatives (Figs. 4.12 and 4.13). The relative amounts of the two isomers could be assessed from the mass spectra of their co-eluting TMS derivatives, by summing intensities of their characteristic α-fragment pairs m/z 201 and 435, or m/z 215 and 421, respectively (note that the larger ions 435 and 421 are due to α-fragmentation and loss of H2O). The results showed that 8- 4.11a and 9-hydroxy-β-hentriacontane-14,16-dione 4.11b were present in roughly equal amounts within the flag leaf wax mixture, and also in the peduncle wax.  The aromatic polyketides, alkylresorcinols 4.12 and methyl alkylresorcinols 4.13, occurred as series of homologs with mainly odd-numbered side-chains. Their chain length profiles were very similar, both between the two fractions and between the wheat organs, in all cases dominated by the C23 homolog (Fig. 4.14).   Terpenoids were detected only in the peduncle wax mixture. Among them, the most (and equally) abundant were two isomeric alcohols, α- 4.14c and β-amyrin 4.14b (Fig. 4.14). They were accompanied by lupeol 4.14d (a third isomer), and by esters of the amyrins with behenic (i.e. C22) acid 4.14e-f. α-Tocopherol 4.14a was present in amounts comparable to lupeol 4.14d.   108 m/zm/zABCm/z50 100 150 200 250 300 350 400 450 500 550 600Rel. Abundance [%]020406080Rel. Abundance [%]02040608010021520173715783435421537462239281 309 35371738357215201129147103 283313373421407 497339399309 520100Rel. Abundance [%]020406080Rel. Abundance [%]02040608071738357609353215 323519435201 421D5746230969 100111363522296125 1672098350444423925126628132634933538113861m/zM-15M-90M-18 MM-60M-18-60 Figure 4.12. Mass spectra of various derivatives of 8- and 9-hydroxyhentriacontane-14,16-diones. Mass spectra of the mixture of 8- 4.11a and 9-hydroxyhentriacontane-14,16-dione 4.11b from the cuticular wax of T. aestivum cv. Bethlehem that was (A) trimethylsilylated on the 8/9-hydroxyl group, (B) trimethylsilylated on both the 8/9-hydroxyl group and the β-diketone enol functionality, (C) acetylated on the 8/9-hydroxyl group, and (D) fully reduced with lithium aluminum hydride and then fully trimethylsilylated on all hydroxyl groups of the resulting triol. Insets show 10-fold enhancements of the 400-620 amu mass range.  m/zm/zRel. Abundance [%]071738357609353215 323 51943520142157 69 100251 281534M-15M-90M-15-90m/zm/zRel. Abundance [%]071738357215201129147103 283313373421407497339309520Rel. Abundance [%]02040608083 609215 323 5195746230969 100111363522296125 1672098350444423925126628132634933538113861m/zM-90-90  109 OCH3OCH3OTMS  7 3    239  (-TMSOH)  323 (-H2O)     309    281    209(-TMSOH)     73    201    435(-H2O)OCH3OCH3OTMSTMS  7 3  323  (-TMSOH)    353     73    201    435(-TMSOH)     73OCH3OCH3OTMSTMS  7 3  323  (-TMSOH)    353     73    201    435(-TMSOH)     73OCH3OAcOCH3  6 3    171      239        251  (-AcOH)   61(+2H)-AcOH  111  281-H2O  263  296(+H)   326(+H)   363 (-AcOH)  381(-CH2CO)-AcOH 266-H2O 248  209 (-AcOH)     309   279(-AcOH)-H2O  291OCH3OCH3OTMSTMS TMS  7 2 3    201    601    313  373     73    489    429     73    73  -TMSOH  283-TMSOH  193  -TMSOH  339-TMSOH-TMSOH  309-TMSOH  511-TMSOH  421ABC  399D Figure 4.13. Fragmentation schemes of various derivatives of 8-hydroxyhentriacontane-14,16-dione. Major MS fragmentations of the 8-hydroxyhentriacontane-14,16-dione isomer 4.11a from the cuticular wax of T. aestivum cv. Bethlehem that was (A) trimethylsilylated on the 8 hydroxyl group, (B) trimethylsilylated on both the 8-hydroxyl group and both tautomeric β-diketone enol functionalities, (C) acetylated on the 8-hydroxyl group, and (D) fully reduced with lithium aluminum hydride and then fully trimethylsilylated on all hydroxyl groups of the resulting triol.    110   Relative composition [ % of  compound cl ass]0102030405060708090100Alkyl-resorcinolsHydroxy-diketonesDiketonesMethylAlkyl-resorcinols3131(19+6) 25(21+6) 27(23+6) 29(25+6) 31(27+6) 33(29+6) 35(19+6+1) 26(21+6+1) 28(23+6+1) 30(25+6+1) 32(27+6+1) 34(29+6+1) 36  Relative composition [% of compound class]0102030405060708090-Tocopherol-Amyrin-Amyrin-Amyrenyl behenate-Amyrenyl behenateLupeolAB Figure 4.14. Relative compositions of polyketide and terpenoid compound classes in T. aestivum cv. Bethlehem waxes. Relative abundances (%) of individual homologs from each of four polyketide compound classes and of individual terpenoid isomers in the wax mixtures covering the (A) flag leaf blade and (B) peduncle. x-Axis labels indicate total homolog chain length or the name of the terpenoid isomer. For alkylresorcinols 4.12 and methyl alkylresorcinols 4.13, the individual chain lengths of their alkyl side chains plus the aromatic ring plus (if applicable) the methyl group are specified in round brackets. Bars represent mean ± standard deviation (n = 5). All bar groups add up to 100%.  111 4.4. Discussion   The chromatographic analyses of cuticular waxes from T. aestivum cv. Bethlehem revealed three major results: (i) the composition and relative amounts of the various wax compounds differed dramatically between flag leaf blade and peduncle, including several compound classes that had not been described for wheat cuticular wax mixtures before; (ii) most compound classes occurred as homologous series with highly characteristic chain length profiles in both wheat organs; (iii) 1-alkanol and 2-alkanol esters occurred as complex mixtures of metamers, with characteristic chain length patterns of their esterified acid and alkanol moieties, respectively. All of these aspects will be discussed in light of the underlying wax biosynthetic mechanisms.  4.4.1. Wax composition differences between organs and new compound classes   There are several literature reports on the composition of cuticular waxes extracted from either whole plants (Bianchi and Corbellini, 1977; Bianchi et al., 1980) or only leaf blades (Tulloch and Hoffman, 1973; Tulloch, 1973; Wang et al., 2015b) of various wheat cultivars. Fairly recently, the compositions of waxes covering flag leaf blades and peduncles of the wheat cvs. Shango and Shamrock (Adamski et al., 2013b) and Ming 988, respectively (Wang et al., 2015a) were reported. Only the latter report quantified wax loads per surface area, thus enabling direct comparisons of wax amounts with our results. It should be noted that cv. Ming 988 had very similar total wax loads between organs (8.2 ± 0.2 μg/cm2 on flag leaf blades leaves and 8.8 ± 0.2 μg/cm2 on peduncles), while our analyses showed three times larger wax loads on peduncles than flag leaves for cv. Bethlehem. The wax load on flag leaves of cv. Bethlehem was   112 in turn twice that of Ming 988, overall suggesting higher wax biosynthetic activity in cv. Bethlehem.    Some of the wax compound classes reported here have not been detected in other wheat cultivars, such as 2-alkanol esters, esters of aromatic alcohols and, in the case of peduncles, aldehydes and terpenoids. While aldehydes and terpenoids are ubiquitous wax compounds, 2-alkanol esters are infrequently found in cuticular wax mixtures. However, they have been described as side products of the β-diketone pathway in other Poaceae, for example in the waxes of barley (Mikkelsen, 1984; von Wettstein-Knowles and Netting, 1976a, 1976c; von Wettstein-Knowles, 2012, 1987, 1976) and sorghum (Penny von Wettstein-Knowles et al., 1984). Esters of aromatic alcohols had not been identified before in Poaceae waxes, but in many other unrelated species. For example, benzyl esters have been found in cuticular waxes of Eucalyptus globulus (Jones et al., 2002; Rapley et al., 2004; Steinbauer et al., 2004), Fagus sylvatica (Gülz et al., 1989), Humulus lupulus (Gülz et al., 1993), Acer pseudoplatanus (Prasad and Gülz, 1990), Simmondsia chinensis (Gülz and Marner, 1986), as well as Papaver rhoeas and P. somniferum (Jetter and Riederer, 1996). Phenethyl esters have been reported in cuticular waxes of Eucalyptus globulus (Jones et al., 2002; Rapley et al., 2004; Steinbauer et al., 2004), Simmondsia chinensis (Gülz and Marner, 1986), Solanum tuberosum (Szafranek and Synak, 2006b), as well as P. rhoeas and P. somniferum (Jetter and Riederer, 1996). Finally, p-hydroxyphenethyl esters have been found in Buddleja cordata (Acevedo et al., 2000), Inula graveolens (Oksuz and Topcu, 1992) and Bongardia chrysogonum (Alfatafta et al., 1989). It is noteworthy that alkylresorcinols and methyl alkylresorcinols have been detected on both flag leaves and peduncles of the wheat cvs. Shango and Shamrock (Adamski et al., 2013b).    113  The major differences between our results and previous wheat wax analyses are in the much higher loads of 1-alkanols and 1-alkanol esters on flag leaf blades, and of β-diketone and hydroxy-β-diketones on peduncles of cv. Bethlehem (see Fig. 4.2). However, these findings are similar to those for cv. Ming 988 and various barley lines (von Wettstein-Knowles, 1969; Wang et al., 2015a), confirming that the acyl reduction wax biosynthetic pathway is strongly dominating wax production in flag leaves, while the β-diketone pathway dictates wax production in peduncles. Both pathways are thus differentially regulated to a high degree, and cv. Bethlehem will be an ideal tool to study biosynthetic mechanisms in comparisons between both organs.   4.4.2. Chain length distributions of ubiquitous wax compound classes   All ubiquitous compound classes were found as series of homologs with distinctive chain length profiles in the waxes of wheat cv. Bethlehem. The bimodal profile of flag leaf fatty acids, peaking at C28 and C22, was described before for T. aestivum cvs. Selkirk and Manitou (Tulloch and Hoffman, 1973), whereas cvs. Ming 988 (Wang et al., 2015a) and Xinong 2718 (Wang et al., 2015b) showed similar predominance of C28 fatty acid but not C22. In contrast, the cv. Bethlehem peduncle wax showed a single maximum at C28 fatty acid (see Fig. 4.3), in this characteristic resembling cv. Ming 988 (Wang et al., 2015a). All taken together, the wheat cultivars and organs differentially accumulate two acid chain lengths, thus pointing to differences either in substrate pool compositions or enzyme preferences. This finding may spur further studies into the biosynthesis of VLC free fatty acids, the mechanisms of which are not understood at present.     114 The cv. Bethlehem flag leaf wax contained high amounts of 1-alkanols, in which the C28 homolog strongly dominated, similar to diverse other wheat cultivars and organs (Adamski et al., 2013b; Bianchi and Corbellini, 1977; Bianchi et al., 1980; Tulloch and Hoffman, 1973; Tulloch, 1973; Wang et al., 2015a, 2015b). The matching 1-alkanols and acid profiles on cv. Bethlehem flag leaves, both peaking at C28, suggest preferential accumulation of the common precursor C28 acyl-CoA. It seems likely that at least one fatty acyl-CoA reductase (FAR) uses this substrate to form wax alcohols in leaf epidermal cells. The TaFAR5 enzyme has recently been found to convert C28 upon heterologous expression in tomato (Wang et al., 2015b), and to be highly expressed in leaves. Hence, it may well be involved in the formation of this central wax constituent.  The peduncle wax 1-alkanols of various wheat cultivars are dominated by the C24 homolog (Adamski et al., 2013b; Wang et al., 2015a), thus in part matching the bimodal profile with maxima at C24 and C30/C32 1-alkanols of cv. Bethlehem. Overall, this may suggest the presence of two FARs with substrate specificities different from the corresponding leaf enzyme. Interestingly, the wheat TaFAR1 and TaFAR5 enzymes have recently been found to exhibit some preference for C22/24 and C20 (as well as C28) substrates, respectively, but they are expressed at relatively low levels in peduncles (Wang et al., 2015a, 2015b). Therefore, our chemical results suggest involvement of other FARs in 1-alkanol formation in wheat peduncles, three more of which have been annotated in the genome.    The wheat cv. Bethlehem organs had very different wax aldehyde profiles, the C28 homolog dominating in flag leaves and the C30 and C32 homologs in peduncles. The leaf results match   115 previous reports, where aldehydes of various wheat cultivars were also dominated by the C28 homolog (Bianchi and Corbellini, 1977; Bianchi et al., 1980; Wang et al., 2015a, 2015b). Thus, the aldehyde distribution matches that of the accompanying 1-alkanols, and not that of the alkanes peaking at C29 in some cultivars and at C31 in cv. Bethlehem. Taken together, these results suggest a biosynthetic relationship between aldehydes and 1-alkanols in wheat leaves, likely involving step-wise reduction of acyl-CoA precursors (by one or two reductases) and release of aldehyde intermediates. This hypothesis contrasts with evidence from Arabidopsis, where 1-alkanols are largely formed without release of aldehyde intermediates by the FAR CER4 (Kunst and Samuels, 2003), while the parallel decarbonylation pathway leads via aldehydes to alkanes. Interestingly, the peduncle aldehyde pattern did resemble that of the accompanying alkanes, suggesting that in this organ the aldehydes are intermediates en route to alkanes. Consequently, peduncles (but not flag leaves) likely harbour an alkane-forming pathway similar to Arabidopsis, possibly involving decarbonylases with different substrate preference for C32 and C30 aldehyde substrates in cv. Bethlehem and other cultivars, respectively.  4.4.3. Chain length and isomer distributions of aliphatic and aromatic esters   The cv. Bethlehem 1-alkanol esters had even-numbered chain lengths, with a bimodal distribution on leaves (peaking at C44 and C50) and a unimodal distribution on peduncles (peaking only at C44), very similar to other wheat cultivars (Tulloch and Hoffman, 1973; Tulloch, 1973; Wang et al., 2015a). In contrast, the 2-alkanol esters had odd-numbered chain lengths and unimodal distributions on both organs (peaking at C35), a pattern closely resembling   116 that reported for the wax mixtures of cer-u69 barley spikes without awns (von Wettstein-Knowles and Netting, 1976a).  The chain length profiles of esterified acids (see Fig. 4.11A) in flag leaves of cv. Bethlehem suggested the presence of two wax ester synthases with high substrate specificity for C22 and C16 acyl-CoA, respectively. In contrast, the 1-alkanol substrates appeared to be used proportional to their availability and without specificity (compare Figs. 4.10A and 4.3A). Thus, esters C50 (C22 acyl + C28 alkyl) and C44 (C16 acyl + C28 alkyl) were most abundant, leading to a characteristic bimodal distribution. For peduncles, less substrate specificity was apparent for acyl-CoAs, although C22 acyl-CoA was still the predominantly incorporated substrate (see Fig. 4.11B). Instead, additional specificity for C22-C24 1-alkanol substrates could be inferred, based on the ester metamer profile quite different from that of free 1-alkanols (compare Figs. 4.10B and 4.3B).  The 2-alkanol ester compositions were nearly identical between cv. Bethlehem organs (compare Figs. 4.9A and B), and so were the resulting profiles of esterified 2-alkanols (compare Figs. 4.10A and B). The C15 and C13 2-alkanols dominated, similar to reports on cuticular waxes from barley (von Wettstein-Knowles and Netting, 1976a). However, on both wheat flag leaves and peduncles also C7 2-alkanol was incorporated in substantial amounts, leading to bimodal distribution. Of note, the C7 2-alkanol moiety had not been reported before in wax of any plant species. Overall, it appears that two 2-alkanol ester synthases may be active in wheat: one with high substrate specificity for the C7 2-alkanol and no acyl-CoA specificity, and another with high   117 specificity for the C18-C22 acyl-CoAs (and especially for C20 acyl-CoA) using the available 2-alkanol pool to produce the full array of 2-alkanol esters (especially C35 and C33).   The characteristic chain length distributions of aromatic esters, which were very similar between cv. Bethlehem leaves and peduncles (see Fig. 4.4), suggest the presence of several additional ester-forming enzymes with distinct acyl substrate chain length preferences. Thus, benzyl esters were formed preferentially from C30 acyl-CoA, phenethyl esters from C24-C28 acyl-CoAs, and p-hydroxyphenethyl esters from C32 acyl-CoA substrate. Lastly, triterpenoid esters were likely formed by a further enzyme fusing C16 acyl-CoA with the two most available triterpenol substrates, β- and α-amyrin, albeit only in the peduncles (compare Figs. 4.10B and 4.14B).  4.4.4. Chain length distributions of polyketide wax compound classes  The most abundant polyketides, β-diketones and hydroxy-β-diketones, occurred as single homologs with 31 carbons in cv. Bethlehem waxes, as described previously for almost all other wheat cultivars (Adamski et al., 2013b; Bianchi and Corbellini, 1977; Bianchi et al., 1980; Tulloch and Hoffman, 1973; Tulloch, 1973; Wang et al., 2015a, 2015b). The only notable exception to this pattern is cv. Demar 4, where small amounts of the C29 homolog were detected as well (Bianchi and Corbellini, 1977). The 1:1 isomer ratio between 8-hydroxy- and 9-hydroxy-β-diketones found here had also been reported before for cvs. Selkirk and Manitou (Tulloch and Hoffman, 1973). The current findings thus are in accordance with the β-diketone and hydroxy-β-diketone biosynthesis pathways outlined before (von Wettstein-Knowles, 2012).     118  Finally, the phenolic polyketide classes of alkylresorcinols and methyl alkylresorcinols were found in wheat cv. Bethlehem waxes as series of odd-numbered homologs with alkyl side chains from C19 to C29, peaking at C23. The distributions were thus very similar to those of the cvs. Shamrock and Shango (Adamski et al., 2013b), with the addition of C29 homologs not reported before. The biosynthesis of plant alkylresorcinols is known to be effected by a type-III polyketide synthase,  an alkylresorcinol synthase, that first performs the extension of an acyl-CoA substrate with three malonyl-CoAs to a tetraketide intermediate, and then its intramolecular aldol cyclization and decarboxylation to the final alkylresorcinol (Baerson et al., 2010). How the additional methyl group in methyl alkylresorcinols is introduced is currently unknown, but the matching side-chain homolog profiles of the two types of phenolic lipids suggest an at least partially shared pathway, perhaps with methylmalonyl-CoA replacing malonyl-CoA in one of the extension steps or a subsequent C-methylation of alkylresorcinols by a dedicated enzyme. In both wheat organs investigated here, it would appear that the alkylresorcinol synthase has a chain length preference for C24 acyl-CoA substrate, resulting in a predominance of homologs with C23 side chains.          119 Chapter 5: Novel oxidized compounds and internally methyl-branched alkanes from cuticular waxes of Triticum aestivum cv. Bethlehem   5.1. Introduction  Most above-ground organs of land plants are covered by a hydrophobic coating, known as the cuticle, which is sealing them against uncontrolled loss of water. The cuticle is composed of the polyester cutin and cuticular wax that can easily be extracted with organic solvents (Jetter et al., 2006). Cutin consists of saturated and unsaturated long-chain (LC, C16 and C18) hydroxy or epoxy fatty acids, linked via ester bonds either directly between fatty acids or via glycerol (Nawrath, 2006; Pollard et al., 2008). Cuticular waxes are typically very-long-chain (VLC, >C20) saturated aliphatic compounds, bearing one functional group or no functionality. Most commonly encountered are homologous series of even-numbered fatty acids, primary alcohols, alkyl esters and aldehydes, as well as odd-numbered alkanes, secondary alcohols and ketones (Jetter et al., 2006; Kolattukudy, 1970). In the wax mixtures of some plant species, alicyclics (e.g., triterpenoids) and aromatics (e.g., alkylresorcinols) can also be quite abundant (Basas-Jaumandreu et al., 2014; Bianchi et al., 1992; Ji and Jetter, 2008; Manheim Jr. and Mulroy, 1978; Nordby and McDonald, 1994).  The characteristic mixtures of various aliphatic wax constituents result from biosynthetic pathways that are fairly well understood, mainly due to extensive studies in the model species Arabidopsis thaliana (Kunst et al., 2006; Samuels et al., 2008). Wax biosynthesis begins in the epidermal plastids, where the fatty acid synthase (FAS) elongates acetyl-CoA into LC fatty   120 acids. These are then transferred to the endoplasmic reticulum (ER), where fatty acid elongase (FAE) complexes extend their chain length two carbons at a time to VLC acyls. The resulting VLC fatty acyl-CoAs can then be further processed into final wax compounds, by (1) reduction to primary alcohols which may be esterified with (V)LC fatty acids; or (2) partial reduction to aldehydes, their subsequent decarbonylation to alkanes, and, in some species, hydroxylation of the alkanes to secondary alcohols and ketones.    While the Arabidopsis model has proved invaluable for our understanding of cuticular wax biosynthesis, much can be learned from studying the diversity of wax structures in other plant species as well. Thus, numerous novel wax compounds were discovered recently in various plant species, suggesting divergence from the traditional pathways in Arabidopsis. Among these are many compounds with multiple functional groups, including diols (Buschhaus et al., 2013b; Jetter, 2000; Wen and Jetter, 2009; Wen et al., 2006a), hydroxyaldehydes (Vermeer et al., 2003; Wen and Jetter, 2007), ketols (Wen and Jetter, 2009), ketoaldehydes (Jetter and Riederer, 1999a), hydroxyacids (Racovita et al., 2015), hydroxyesters (Busta et al., 2016; Racovita et al., 2015), and ketoesters (Jetter and Riederer, 1999a). In many of these studies, the nature of the functional groups, their relative positions and the chain length profiles within respective homologous series could be used to infer the biosynthetic origin of functional groups or entire molecules. The work herein is an extension of these studies, with the objective of seeking, identifying and quantifying novel wax compounds to further our understanding of wax biosynthesis beyond the well-known constituents of the Arabidopsis wax mixture.    121 Bread wheat (Triticum aestivum) is rapidly becoming a new model species for wax biosynthesis studies (Adamski et al., 2013b; Hen-Avivi et al., 2016b; Millet et al., 2013; Wang et al., 2015a, 2015b), due to its importance as a major staple crop world-wide and its susceptibility to drought, combined with the recognized role of cuticular waxes in conferring drought resistance in this species. In a previous analysis of flag leaf blade and peduncle waxes of T. aestivum cv. Bethlehem by gas chromatography-mass spectrometry (GC-MS), we quantified various VLC fatty acids, primary alcohols, aldehydes, alkanes and 1-alkanol esters common to most plant species (Racovita et al., 2016). Furthermore, benzyl esters, phenethyl esters, p-hydroxyphenethyl esters were identified for the first time in wheat wax, together with various terpenoids. Finally, several compounds characteristic of Poaceae waxes were reported, including β-diketones, hydroxy-β-diketones, alkylresorcinols, methyl alkylresorcinols and 2-alkanol esters. However, numerous compounds in the cuticular wax mixtures of flag leaf blades and peduncles remained unidentified. To elucidate the molecular structures of these additional wheat wax constituents, we have now performed in-depth mass spectrometric analyses, using various derivatives of each novel structure for comparison of fragmentation patterns. Finally, the homolog and isomer patterns of the novel compound classes were assessed, to enable conclusions on their biosynthetic origins.    122 5.2. Experimental  5.2.1. Plant material  Flag leaves and peduncles were harvested from mature Triticum aestivum cv. Bethlehem plants during the months of August 2013 for total wax specimens and August 2014 for specimens for preparative thin layer chromatography (TLC). Plants were grown in greenhouses at Weizmann Institute of Science (Rehovot, Israel) on 50% peat – 50% turf, with watering every 3-4 days (~400 mL per 5 L pot). Growth conditions were: 12-14 h / 10-12 h light / dark cycles (180 μmol m-2 s-1 light), with temperatures of 24-26°C / 17-18°C, respectively. For total wax samples, one leaf blade with total area of 40-50 cm2 (both sides) and one peduncle with a projected surface area of ~20 cm2 were used per biological replicate. For preparative TLC samples, ten leaf blades and ten peduncles of the same size were used. Exact areas for leaf blades were determined by capturing them in photographs and using the ImageJ software to measure the area of one side, then multiplying by a factor of 2. For peduncles, areas were determined by measuring the length L and diameter D of the peduncle and then calculating the area of specimen with the formula: π x D x L.   5.2.2. Chemicals  The following chemicals were acquired from Sigma-Aldrich (Oakville ON, Canada) and used without further purification: chloroform (≥99%, with 0.75% ethanol as stabilizer), ethanol (≥99.8%, HPLC grade), pyridine (≥99.8%, anhydrous), N,O-bis(trimethylsilyl)trifluoroacetamide   123 (BSTFA, GC grade), acetic anhydride (≥98%), diethyl ether (≥99.7%, anhydrous, 1 ppm BHT as inhibitor), lithium aluminum hydride (≥95%), sulphuric acid (95-98%), O-methylhydroxylamine hydrochloride (≥98%), boron trifluoride-methanol solution (14%), primuline (50% dye content), acetone (≥99.9%, HPLC grade). n-Tetracosane (≥99%) was from Alfa Aesar (Ward Hill MA, USA). Gases were acquired from Praxair Canada (Vancouver BC, Canada): nitrogen (≥99.998%), helium (≥99%), and hydrogen (≥99.95%).   5.2.3. Preparation of wax extracts  Rolled-up leaves and peduncle pieces were extracted for 30 s at ambient temperature with 10 mL chloroform, to which 5 μg n-tetracosane were added prior to extraction, as an internal standard. The chloroform was then transferred to another vial and the extraction repeated for another 30 s with a fresh portion of 10 mL chloroform. The combined chloroform extracts were then evaporated to dryness under a stream of N2 at 50°C, leaving behind the wax mixtures for either preparative TLC or GC analysis.   5.2.4. Preparative thin layer chromatography  Fractionation of compound classes in the total wax extracts was carried out by preparative TLC, using the sandwich technique (Tantisewie et al., 1969). Glass plates coated with silica gel 60 F254 (Uniplate, Analtech, layer thickness: 1 mm, size: 20x20 cm, with 4 cm   124 concentrating zone) served as stationary phase, and a mixture of chloroform:ethanol 98:2 (v/v) served as mobile phase. At the end of separation, TLC plates were sprayed with primuline (5 mg in 100 mL acetone/water 80/20, v/v) and bands were visualized under 365 nm UV light. All bands were removed from the plate with clean spatulas into several glass vials, and each extracted twice with 10 mL portions of fresh chloroform for 30 s, at ambient temperature. Then, the combined extracts were filtered through glass wool (Supelco), partially evaporated under N2 at 50°C, transferred to 2 mL GC vials, evaporated to dryness and stored until GC-Mass Spectrometry (MS) analysis.    5.2.5. Derivatization reactions  Prior to GC analysis, all samples were silylated by refluxing with 10 μL N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) and 10 μL pyridine at 70°C for 20 min. Excess reagents were then removed under a gentle stream of N2 and the silylated waxes re-dissolved in 50 μL CHCl3.  Acetylation was carried out by refluxing a mixture of dry wax, 10 μL pyridine, and 10 μL acetic anhydride at 70°C for 5 min, then allowing it to stir overnight at ambient temperature. After removal of excess reagents under N2, silylation was carried out as described above.  Complete reduction of carbonyl and ester groups was achieved by dissolving the wax sample in 50 μL diethyl ether and adding 0.1 mg LiAlH4, then allowing the mixture to react   125 overnight at 70ºC. After quenching with 10% H2SO4, followed by three sequential extractions with 60 μL diethyl ether each, the combined extracts were evaporated to dryness and silylated as described above.  Carbonyl-containing compounds were transformed into their methoximes by heating with 20 μL of a saturated solution of O-methylhydroxylamine hydrochloride in pyridine:chloroform 7:3 (v/v) for 30 min at 70°C. The resulting mixture was partitioned between 50 μL distilled water and 50 μL chloroform and the chloroform fraction retained. After extracting the aqueous phase one more time with 50 μL fresh chloroform, the chloroform extracts were combined, evaporated to dryness and silylated as described above.  Transesterification was achieved by heating the waxes in 100 μL of 14% BF3-methanol solution at 70°C for 2 hours. Then, the products were isolated by partitioning between 50 μL distilled water and 50 μL diethyl ether and repeating the extraction two more times with fresh portions of 50 μL ether. After evaporation to dryness, silylation was carried out as described above.  5.2.6. Gas chromatography  Two Gas Chromatography (GC) instruments were used for identification and quantification of wax constituents, respectively, both equipped with the same type of capillary GC column (6890N, Agilent, Avondale PA, USA; length: 30 m; type: HP-1 100% PDMS; i.d.: 0.32 mm; df: 0.1 µm), both equipped with on-column injector and programmed to follow the same   126 temperature program (2 min at 50ºC, ramp 40ºC min-1 to 200ºC, constant for 2 min, ramp 3ºC min-1 to 320ºC, constant for 30 min). One GC instrument employed helium as mobile phase, at a flow rate of 1.4 mL/min, and was equipped with MS detector (5973N, Agilent, EI-70 eV, m/z 50-750). The other used hydrogen as carrier gas at 2.0 mL/min and was equipped with a flame ionization detector (FID). Wax compounds were quantified by normalizing their GC-FID peak areas against that of the internal standard, added in known amount. The relative response factors of all wax compound classes with respect to the internal standard were approximated to 1.00, in agreement with literature reports using the same GC-FID operation conditions (Riederer and Schneider, 1989).       5.3. Results  The principal goals of the work herein were to identify novel compounds in the cuticular waxes of the bread wheat (Triticum aestivum) cultivar Bethlehem (sections 5.3.1. – 5.3.4.) and to determine their relative quantities within the wax mixtures covering flag leaf blades and peduncles (section 5.3.5.).   5.3.1. TLC separation of cuticular waxes of T. aestivum flag leaf blades and peduncles  Preliminary experiments showed that the unknown wheat wax constituents belonged to seven different compound classes A-G (Fig. 5.1). Each of these comprised a series of compounds separated into equally spaced GC peaks with shared characteristic MS fragmentation patterns, and thus each class was recognized as a homologous series of compounds. To enable   127 their structure elucidation, the compound classes were separated and concentrated by preparative thin layer chromatography (TLC) using silica gel as stationary phase and CHCl3:EtOH 98:2 (v/v) as mobile phase.  5.2a (n=7)5.2b (n=8)5.2c (n=9)5.3a (n=7)5.3b (n=8)5.3e (n=11)...CH3OHCH3n5.1a (n=6)5.1b (n=7)5.1f (n=11)...5OHOHCH3n 5OOHCH3CH3O8 5nOCH3O CH3CH3OH32n5.4a (n=7)5.4b (n=8)5.4e (n=11)...OCH3O CH3CH3O32n5.5a (n=7)5.5b (n=8)5.5e (n=11)...OOCH3n5.6a (n=10)5.6b (n=11)5.6d (n=13)...HCH3Hn55.7a (n=8)5.7c (n=9)5.7e (n=10)CH3CH3Hn55.7b (n=8)5.7d (n=9)5.7f (n=10) Figure 5.1. Oxidized compounds and internally methyl-branched alkanes identified in the wax mixtures of T. aestivum cv. Bethlehem. For compound classes where several co-eluting isomers were detected per homolog, only the major isomer is shown for each homolog in each series.  The flag leaf blade wax mixture was separated into eleven fractions, which were analyzed individually by GC-MS. Among them, seven fractions were found to contain previously identified compound classes, namely 5-alkylresorcinols and methyl 5-alkylresorcinols (Rf 0.09), 2-(p-hydroxyphenyl)ethyl esters of VLC fatty acids (Rf 0.27), free VLC fatty acids (Rf 0.30), hydroxy-β-diketones (Rf 0.33), 1-alkanols (Rf 0.38), β-diketones along with small amounts of aldehydes (Rf 0.86), and several very non-polar wax classes such as n-alkanes, iso- and anteiso-  128 alkanes, esters of 1- and 2-alkanols, benzyl esters and 2-phenylethyl esters (Rf 1.00). The unknown compound classes were found to have widely varying TLC behaviour, with series A (Rf 0.72), C (Rf 0.54), D (Rf 0.35), as well as E and F (Rf 0.44) in fractions of their own, and series B co-eluting with free VLC fatty acids (Rf 0.30) and series G with the non-polar wax classes (Rf 1.00). The peduncle wax mixture yielded nine fractions, with identical Rf values and compositions very similar to corresponding leaf wax fractions. However, series A, B and C were not found in peduncle wax.  5.3.2. Structure elucidation of compound classes A - C  Based on TLC behaviour, fraction A exhibited polarity intermediate between aldehydes and primary alcohols, and it was thus hypothesized to contain VLC secondary alcohols 5.1. All six compounds in A yielded TMS derivatives with common diagnostic MS fragment m/z 73 [(CH3)3Si]+, homolog-dependent M-15 and M-90 ions, due to loss of methyl radical and (CH3)3SiOH, respectively, and no m/z 147 [(CH3)2SiOSi(CH3)3]+ (Fig. 5.2A), together  confirming the secondary alcohol structure (Diekman et al., 1967; Racovita et al., 2015; Wen et al., 2006a). Each homolog, in its TMS derivative mass spectrum, also displayed a pair of α-fragments, with one ion m/z 257 common to all homologs indicative of a hydroxyl group on C-12, and a second fragment varying with chain length (m/z 397 for the C33 homolog in Fig. 5.2A). These fragments were accompanied by further pairwise combinations of α-fragments differing by 14 Da (e.g., C-10 m/z 229, C-11 m/z 243, C-13 m/z 271, C-14 m/z 285). A summary of all identified homologs and regiomers within them together with their diagnostic MS fragments is presented in Table 5.1. Extracted ion chromatograms (EICs) of the shorter α-fragments in these   129 pairs revealed small retention time differences between them, but all well within the overall GC peak of the respective homolog. Taken together, the GC and MS information thus revealed the presence of positional isomers (regiomers) of secondary alcohols that were only partially GC-separated. All six homologs were found dominated by the C-12 isomer, accompanied by others bearing hydroxyls on neighbouring carbon atoms (Fig. 5.2B).   For structure confirmation, acetyl (Ac) derivatives of compounds A were prepared and analyzed by GC-MS, revealing a characteristic fragment m/z 61 [CH3COOH2]+ common to all homologs indicative of a hydroxyl group (Buschhaus et al., 2013b; Wen et al., 2006a), as well as M-60 fragments varying between homologs due to elimination of acetic acid AcOH (Fig. 5.2C). α-fragments were much less prominent than for TMS derivatives, but further loss of AcOH resulted in distinct pairs of fragments confirming the presence of different regiomers, with m/z 167 / 307 for the C-12 alcohol isomer, m/z 153 / 321 for the C-11 isomer, m/z 181 / 293 for C-13, etc. Taken together, the TLC behaviour and GC-MS data for TMS and Ac derivatives demonstrated that A was a series of secondary alcohol homologs 5.1, each comprising several regiomers with hydroxyls at and around C-12.      130  Figure 5.2. Structure elucidation of secondary alcohols in wheat leaf wax. (A) Mass spectrum of co-eluting TMS derivatives of C33 sec alcohol isomers and major fragmentations of main isomer. (B) EICs showing cumulative intensity of m/z 73 and 75, as well as intensities of short α-fragments of main isomer and the four next most abundant isomers. (C) Mass spectrum of co-eluting Ac derivatives of C33 sec alcohol isomers and major fragmentations of main isomer.           131 Table 5.1. Characteristic m/z fragments of trimethylsilyl ethers of secondary alcohols in leaf wax  Compound Fragments characteristic of homolog Fragments characteristic of isomer    Pentacosan-10-ol 350                                           425  229                                           313 Pentacosan-11-ol   243                                           299 Pentacosan-12-ol   257                                           285 Pentacosan-13-ol   271                             Heptacosan-9-ol 378                                           453  215                                           355 Heptacosan-10-ol   229                                           341 Heptacosan-11-ol   243                                           327 Heptacosan-12-ol   257                                           313 Heptacosan-13-ol   271                                           299 Heptacosan-14-ol   285    Nonacosan-9-ol 406                                           481  215                                           383 Nonacosan-10-ol   229                                           369 Nonacosan-11-ol   243                                           355 Nonacosan-12-ol   257                                           341 Nonacosan-13-ol   271                                           327 Nonacosan-14-ol   285                                           313 Nonacosan-15-ol   299    Hentriacontan-8-ol 434                                           509  201                                           425 Hentriacontan-9-ol   215                                           411 Hentriacontan-10-ol   229                                           397 Hentriacontan-11-ol   243                                           383 Hentriacontan-12-ol   257                                           369 Hentriacontan-13-ol   271                                           355 Hentriacontan-14-ol   285                                           341 Hentriacontan-15-ol   299                                           327 Hentriacontan-16-ol   313                             Tritriacontan-8-ol 462                                           537  201                                           453 Tritriacontan-9-ol   215                                           439 Tritriacontan-10-ol   229                                           425 Tritriacontan-11-ol   243                                           411 Tritriacontan-12-ol   257                                           397 Tritriacontan-13-ol   271                                           383 Tritriacontan-14-ol   285                                           369 Tritriacontan-15-ol   299                                           355 Tritriacontan-16-ol   313                                           341 Tritriacontan-17-ol   327                             Pentratriacontan-8-ol 490                                           565  201                                           481 Pentatriacontan-9-ol   215                                           467 Pentatriacontan-10-ol   229                                           453 Pentatriacontan-11-ol   243                                           439 Pentatriacontan-12-ol   257                                           425 Pentatriacontan-13-ol   271                                           411 Pentatriacontan-14-ol   285                                           397 Pentatriacontan-15-ol   299                                           383 Pentatriacontan-16-ol   313                                           369 Pentatriacontan-17-ol   327                                           355 Pentatriacontan-18-ol   341   132 Series B consisted of three compounds found in the same fraction as free fatty acids in flag leaf wax. The TMS derivatives of all compounds in B shared diagnostic MS fragments m/z 73 [(CH3)3Si]+ and m/z 75 [(CH3)2SiOH]+, m/z 103 [(CH3)3SiOCH2]+, and m/z 147 [(CH3)2SiOSi(CH3)3]+ and m/z 149 [(CH3)2SiOSi(CH3)2OH]+, together indicating the presence of one primary and one secondary hydroxyl group (Fig. 5.3A) (Jetter and Riederer, 1999a; Jetter, 2000; McCloskey et al., 1968). They were accompanied by homolog-dependent M-15 and M-15-90 ions, due to loss of methyl radical and loss of both methyl and (CH3)3SiOH, respectively. Pairs of α-fragments, including one ion m/z 257 common to all homologs and a second fragment varying with chain length (m/z 415 for the C28 homolog in Fig. 5.3A), indicated a hydroxyl group on the -12 carbon. Further α-fragments differing by 14 Da were present, and EICs revealed small retention time differences between them (Fig. 5.3B). The diagnostic MS fragments for all identified homologs and regiomers are presented in Table 5.2.   The Ac derivatives of compounds in fraction B showed a characteristic fragment m/z 61 [CH3COOH2]+ confirming the presence of at least one hydroxyl group (Fig. 5.3C), as well as homolog-dependent parent ions M and daughter ions due to loss of up to two acetyl groups and acetic acid molecules (M-43, M-60, M-60-43 and M-60-60), supporting the presence of a second hydroxyl functionality (Wen et al., 2006a). Different regiomers were discerned based on loss of CH2CO or AcOH from α-fragments, confirming the presence of a hydroxyl function on -12 (m/z 167 / 313) and, for example, -11 (m/z 153 / 327),-13 (m/z 181 / 299). Taken together, the TLC behaviour and GC-MS data for TMS and Ac derivatives demonstrated that B was a homologous series of prim/sec diols 5.2, with sec hydroxyl groups predominantly in the -12 position or on adjacent carbons.   133  Figure 5.3. Structure elucidation of primary/secondary diols in wheat leaf wax. (A) Mass spectrum of co-eluting TMS derivatives of C28 prim/sec diol isomers and major fragmentations of main isomer. (B) EICs showing intensities of m/z 73 and of short α-fragments of main isomer and the four next most abundant isomers. (C) Mass spectrum of co-eluting Ac derivatives of C28 prim/sec diol isomers and major fragmentations of main isomer.           134 Table 5.2. Characteristic m/z fragments of bis(trimethylsilyl) ethers of primary/secondary diols in leaf wax Compound Fragments characteristic of homolog Fragments characteristic of isomer    Hexacosane-1,17-diol 437     452     527  229                  415                   325 Hexacosane-1,16-diol   243                  401                   311 Hexacosane-1,15-diol   257                  387                   297 Hexacosane-1,14-diol   271                  373                   283 Hexacosane-1,13-diol   285                  359                   269    Octacosane-1,20-diol 465     480     555  215                  457                   367 Octacosane -1,19-diol   229                  443                   353 Octacosane -1,18-diol   243                  429                   339 Octacosane -1,17-diol   257                  415                   325 Octacosane -1,16-diol   271                  401                   311 Octacosane -1,15-diol   285                  387                   297 Octacosane -1,14-diol   299                  373                   283    Triacontane-1,21-diol 493     508     583  229                  471                   381 Triacontane-1,20-diol   243                  457                   367 Triacontane-1,19-diol   257                  443                   353 Triacontane-1,18-diol   271                  429                   339 Triacontane-1,17-diol   285                  415                   325  Series C comprised five compounds with very long GC retention times (>50 min), suggesting relatively high molecular weights and long carbon chains, likely in the form of esters linking two VLC moieties. Based on TLC behaviour, these compounds had polarities between primary and secondary alcohols, rendering alkyl ester structures with an additional secondary hydroxyl function plausible. The TMS derivatives of compounds C had a common diagnostic MS fragment m/z 73 [(CH3)3Si]+, but no m/z 147 [(CH3)2SiOSi(CH3)3]+ (Fig. 5.4A),  indicating the presence of only one hydroxyl group in the native compounds. The five homologs also had fragments characteristic of the acid components of esters, such as acylium ions Macid-17 and fragments formed via McLafferty rearrangement with double hydrogen transfer Macid+1 (m/z 323 and 341 for the C50 homolog containing C22 acid shown in Fig. 5.4A) (Kingston et al., 1974). The hydroxy ester structure thus confirmed, further ions could be inferred to result from TMS   135 transfer from the sec hydroxyl to the ester group, i.e. m/z 397 and 413 for the C50 homolog in Fig. 5.4A (Busta et al., 2016; von Wettstein-Knowles and Madsen, 1984), and from sequential loss of (CH3)3SiOH and an acid molecule from M (C22 acid with molecular weight 340 amu in Fig. 5.4A).   The TMS derivatives of C also showed pairs of α-fragments, one common to all homologs (m/z 257) indicating an -12 hydroxyl group, and another one varying with chain length (m/z 665 for the C50 homolog in Fig. 5.4A). Other pairs of α-fragments differing by 14 amu units again suggested positional isomers, and small retention time differences in EICs confirmed the presence of, among others, -10 (m/z 229), -11 (m/z 243), -13 (m/z 271), -14 (m/z 285) hydroxyls. (Fig. 5.4B; Table 5.3). Ac derivatives served to confirm the tentatively assigned structures of C, with fragments m/z 61 [CH3COOH2]+ as well as M-43 and M-60 fragments due to loss of an acetate moiety corroborating the presence of one hydroxyl function in all homologs (Fig. 5.4C). Further fragments due to ester-linked acids (Macid-17 and Macid+1), to loss of both acetic acid and the VLC acid from M (m/z 390) further confirmed the hydroxy ester structures. Lastly, α-fragmentation and subsequent loss of CH2CO gave rise to ions confirming the presence of an -12 hydroxyl group (m/z 593), together with several co-eluting positional isomers (e.g., m/z 607 for -11 and m/z 579 for -13 hydroxyls).   Finally, reduction of a separate aliquot of series C with excess LiAlH4 yielded a single bifunctional compound, C28 prim/sec diol, together with C16 to C24 prim alcohols. The structure of the C28 prim/sec diol was assigned based on identical GC and MS behaviour (of the TMS derivative) to the corresponding homolog in series B. Thus, the LiAlH4 derivatization confirmed   136 that compounds C were esters of prim/sec C28 diols 5.3. Taken together, our TLC and GC-MS data demonstrate that fraction C was a homologous series of esters 5.3 containing C16 to C24 fatty acids linked to the terminal hydroxyl of prim/sec C28 diols. Each ester homolog comprised several isomers, with sec hydroxyl groups at and around the -12 carbon.    Figure 5.4. Structure elucidation of primary/secondary diol esters in wheat leaf wax. (A) Mass spectrum of co-eluting TMS derivatives of C50 prim/sec diol ester isomers and major fragmentations of main isomer. (B) EICs showing intensities of m/z 73 and of short α-fragments of main isomer and the four next most abundant isomers. (C) Mass spectrum of co-eluting Ac derivatives of C50 prim/sec diol ester isomers and major fragmentations of main isomer.   137 Table 5.3. Characteristic m/z fragments of trimethylsilyl ethers of primary/secondary diol esters in leaf wax Compound Fragments characteristic of homolog Fragments characteristic of isomer    Octacosane-1,20-diol palmitate 239 257 313 329  215                                           623 Octacosane-1,19-diol palmitate   229                                           609 Octacosane-1,18-diol palmitate   243                                           595 Octacosane-1,17-diol palmitate   257                                           581 Octacosane-1,16-diol palmitate   271                                           567 Octacosane-1,15-diol palmitate   285                                           553 Octacosane-1,14-diol palmitate   299                                           539    Octacosane-1,20-diol stearate 267 285 341 357  215                                           651 Octacosane-1,19-diol stearate   229                                           637 Octacosane-1,18-diol stearate   243                                           623 Octacosane-1,17-diol stearate   257                                           609 Octacosane-1,16-diol stearate   271                                           595 Octacosane-1,15-diol stearate   285                                           581 Octacosane-1,14-diol stearate   299                                           567    Octacosane-1,20-diol arachidate 295 313 369 385  215                                           679 Octacosane-1,19-diol arachidate   229                                           665 Octacosane-1,18-diol arachidate   243                                           651 Octacosane-1,17-diol arachidate   257                                           637 Octacosane-1,16-diol arachidate   271                                           623 Octacosane-1,15-diol arachidate   285                                           609 Octacosane-1,14-diol arachidate   299                                           595    Octacosane-1,20-diol behenate 323 341 397 413  215                                           707 Octacosane-1,19-diol behenate   229                                           693 Octacosane-1,18-diol behenate   243                                           679 Octacosane-1,17-diol behenate   257                                           665 Octacosane-1,16-diol behenate   271                                           651 Octacosane-1,15-diol behenate   285                                           637 Octacosane-1,14-diol behenate   299                                           623    Octacosane-1,20-diol lignocerate 351 369 425 441  215                                           735 Octacosane-1,19-diol lignocerate   229                                           721 Octacosane-1,18-diol lignocerate   243                                           707 Octacosane-1,17-diol lignocerate   257                                           693 Octacosane-1,16-diol lignocerate   271                                           679 Octacosane-1,15-diol lignocerate   285                                           665 Octacosane-1,14-diol lignocerate   299                                           651      138 5.3.3. Structure elucidation of compound classes D and E  The five homologs in series D were, according to relative TLC retention, more polar than primary alcohols and the hydroxy esters in fraction C. Compounds D were tentatively assigned as VLC esters of hydroxy-2-alkanols 5.4 linked through the 2-OH group, based on similarity of their TMS derivative mass spectra with that of 7-hydroxypentadecan-2-ol eicosanoate reported in the literature (von Wettstein-Knowles and Madsen, 1984). The combination of TMS derivative M-15 and ion m/z 73 [(CH3)3Si]+ together with the lack of m/z 147 [(CH3)2SiOSi(CH3)3]+ confirmed the presence of only one hydroxyl function. Two ester metamers with differing acid (and, consequently, also alcohol) chain lengths were immediately apparent for each homolog in D, based on the presence of two homologous pairs of Macid-17 / Macid fragments (m/z 295 / 312 and 323 / 340 for the C35 homolog in Fig. 5.5A). Similarly, homologous pairs of TMS transfer fragments were observed (m/z 369 / 385 and 397 / 413 in Fig. 5.5A), indicative of two co-eluting metamers for each ester homolog. Furthermore, four pairs of complementary α-fragments were observed for each ester chain length (m/z 215 / 497, 201 / 511, 187 / 525, and 173 / 539), suggesting the presence of two regiomers for each of the two metamers. Careful analysis of overlap between EIC traces of regiomer-specific α-fragments and of acid-specific fragments (Fig. 5.5C) allowed identification of four co-eluting isomers within the C35 ester homolog, listed in decreasing relative abundance: (i) C20 acid + C15 2,8-diol (m/z 385 and 201); (ii) C20 acid + C15 2,7-diol (m/z 385 and 215); (iii) C22 acid + C13 2,7-diol (m/z 413 and 187); (iv) C22 acid + C13 2,8-diol (m/z 413 and 173). A summary of all identified homologs and isomers within them along with their diagnostic MS fragments is presented in Table 5.4.     139 The structures of the diol esters in D were confirmed by MS analysis of their Ac derivatives. They showed the characteristic m/z 61 [CH3COOH2]+ indicative of hydroxyl group presence in the native structure, as well as homolog-dependent M-60, M-Macid, M-Macid-43, M-Macid-60 fragments due to single or combined losses of acetyl- and fatty acyl-derived moieties (Fig. 5.5D). Fragments Macid-17 / Macid+1 further confirmed the presence of two metamers per homolog (m/z 295 / 313 and 323 / 341 in Fig. 5.5D). The α-fragments had very low intensity, but their product ions resulting from loss of acetic acid confirmed the presence of the sec hydroxyl (m/z 111 and 421 in Fig. 5.5D). Finally, reduction of an aliquot of the fraction with excess LiAlH4 gave rise to two new compounds with relatively short GC retention times (as TMS derivatives). Their mass spectra were unambiguously assigned to mixtures of pentadecane-2,7-diol plus pentadecane-2,8-diol (Fig. 5.5F) and tridecane-2,7-diol plus tridecane-2,8-diol, respectively, thus confirming the presence of four isomers per ester homolog in D. Taken together, the TLC and GC-MS data identified D as a series of ester homologs formed by linking various fatty acids with 7- and 8-hydroxy-2-tridecanol as well as 7- and 8-hydroxy-2-pentadecanol.               140  Figure 5.5. Structure elucidation of hydroxy-2-alkanol esters in wheat leaf and peduncle wax. (A) Mass spectrum of co-eluting TMS derivatives of C35 hydroxy-2-alkanol ester isomers. (B) Major fragmentations of all isomers in (A). (C) EICs showing cumulative intensity of m/z 73 and 75, as well as intensities of: metamer-shared α-fragments m/z 185 and 199, regiomer-characteristic α-fragments m/z 173, 187, 201 and 215, and TMS-transfer acid fragments m/z 385 and 413. (D) Mass spectrum of co-eluting Ac derivatives of C35 hydroxy-2-alkanol ester isomers. (E) Major fragmentations of all isomers in (D). (F) Mass spectrum and major fragmentations of TMS derivatives of co-eluting C15 diol isomers obtained via LiAlH4 reduction of the C35 hydroxy-2-alkanol ester isomer mixture (corresponding information for C13 diols not shown).   141 Table 5.4. Characteristic m/z fragments of trimethylsilyl ethers of hydroxy-2-alkanol esters in leaf and peduncle wax Compound Fragments characteristic of homolog Fragments characteristic of isomer type: -regiomer        -metamer    Tridecane-2,8-diol stearate 539 173 483 199 109  267 285 341 357 Tridecane-2,7-diol stearate  187 469 185  95              Pentadecane-2,8-diol palmitate  201 455 199 109  239 257 313 329 Pentadecane-2,7-diol palmitate  215 441 185  95                 Tridecane-2,8-diol arachidate 567 173 511 199 109  295 312 369 385 Tridecane-2,7-diol arachidate  187 497 185  95              Pentadecane-2,8-diol stearate  201 483 199 109  267 285 341 357 Pentadecane-2,7-diol stearate  215 469 185  95                 Tridecane-2,8-diol behenate 595 173 539 199 109  323 340 397 413 Tridecane-2,7-diol behenate  187 525 185  95              Pentadecane-2,8-diol arachidate  201 511 199 109  295 312 369 385 Pentadecane-2,7-diol arachidate  215 497 185  95                 Tridecane-2,8-diol lignocerate 623 173 567 199 109  351 368 425 441 Tridecane-2,7-diol lignocerate  187 553 185  95              Pentadecane-2,8-diol behenate  201 539 199 109  323 340 397 413 Pentadecane-2,7-diol behenate  215 525 185  95                 Tridecane-2,8-diol cerotate 651 173 595 199 109  379 396 453 469 Tridecane-2,7-diol cerotate  187 581 185  95              Pentadecane-2,8-diol lignocerate  201 567 199 109  351 368 425 441 Pentadecane-2,7-diol lignocerate  215 553 185  95                The five homologs in series E, with polarity between primary alcohols and prim/sec diol esters, were tentatively assigned oxo-2-alkanol ester structures 5.5 based on similarity of their mass spectra with that of 7-oxopentadecan-2-ol eicosanoate reported before (von Wettstein-Knowles and Madsen, 1984). Treatment of E with BSTFA left the compounds in E lacking the TMS fragment m/z 73, suggesting that they did not bear hydroxyl groups (Fig. 5.6A). Instead, they had fragments indicative of the 2-alkanol ester structure, such as Macid-17 (m/z 295 and 323 for the two metamers in Fig. 5.6A), Macid+1 (m/z 313 and 341 in Fig. 5.6A), and Malcohol-1 (m/z 225 and 197 in Fig. 5.6A). They also showed prominent homolog-independent α-fragments indicating the carbonyl position on the 2-alkanol moiety, either 7- or 8-oxo groups on C15 2-  142 alkanol (m/z 141 and m/z 127, respectively), and 7- or 8-oxo groups on C13 2-alkanol (m/z 113 and m/z 99, respectively; Fig. 5.6A-C and Table 5.5). Complementary α-fragments were not observed, but some of the closely-related fragments resulting from McLafferty rearrangement on the same side of the carbonyl function were sizeable (m/z 452 and 438 for the C35 homolog in Fig. 5.6A). Molecular ions M could not be detected under the current conditions.   To directly probe the presence of a carbonyl functionality, an aliquot of fraction E was derivatized with O-methylhydroxylamine and the corresponding methoximes analyzed by MS. The resulting homologs all showed fragments m/z 87 and m/z 100 diagnostic for methoximes (Jetter and Riederer, 1999a) (Fig. 5.6D), accompanied by prominent MS ions characterizing the 2-alkanol ester structure, such as Macid-17 (m/z 295 and 323 for the two metamers in Fig. 5.6D), Macid+1 (m/z 313 and 341 in Fig. 5.6D), and Malcohol-1 (m/z 254 and 226 in Fig. 5.6D). α-Fragments and related McLafferty fragments indicative of methoxime position were 29 amu higher than corresponding signals of the underivatized carbonyls (e.g., m/z 156 and m/z 481 in Fig. 5.6D instead of m/z 127 and m/z 452 in Fig. 5.6A). Finally, fragments M-31 due to loss of methoxy group indicated homolog chain lengths, while M ions were not detected.   For further structure confirmation, fraction E was subjected to reduction with LiAlH4, resulting in the same two pairs of diol isomers also formed by reduction of series D. Taken together, the TLC behaviour as well as the GC-MS characteristics of diverse derivatives demonstrated that fraction E was a homologous series of esters containing various fatty acids linked to 7- and 8-oxo-2-tridecanol1 as well as 7- and 8-oxo-2-pentadecanol2..                                                   1 IUPAC names: 2-hydroxytridecan-7-one and 12-hydroxytridecan-6-one. 2 IUPAC names: 2-hydroxypentadecan-7-one and 2-hydroxypentadecan-8-one   143  Figure 5.6. Structure elucidation of oxo-2-alkanol esters in wheat leaf and peduncle wax. (A) Mass spectrum of co-eluting isomers of C35 oxo-2-alkanol ester. (B) Major fragmentations of all isomers from (A). (C) EICs showing intensity of m/z 57, as well as intensities of: metamer-characteristic alkyl fragments m/z 197 and 225, regiomer-characteristic α-fragments m/z 99, 113, 127 and 141, and Macid+1 fragments m/z 313 and 341. (D) Mass spectrum of co-eluting isomers of methoxime derivatives of C35 oxo-2-alkanol ester. (E) Major fragmentations of all isomers from (D).   144 Table 5.5. Characteristic m/z fragments of oxo-2-alkanol esters in leaf and peduncle wax Compound Fragments characteristic of homolog Fragments characteristic of isomer type: -regiomer        -metamer    8-oxotridecane-2-ol stearate -   99 114 424          197 267 285  7-oxotridecane-2-ol stearate  113 128 410              8-oxopentadecane-2-ol palmitate  127 142 396          225 239 257  7-oxopentadecane-2-ol palmitate  141 156 382                 8-oxotridecane-2-ol arachidate -   99 114 452          197 295 313  7-oxotridecane-2-ol arachidate  113 128 438              8-oxopentadecane-2-ol stearate  127 142 424          225 267 285 7-oxopentadecane-2-ol stearate  141 156 410                 8-oxotridecane-2-ol behenate -   99 114 480          197 323 341  7-oxotridecane-2-ol behenate  113 128 466               8-oxopentadecane-2-ol arachidate  127 142 452          225 295 313 7-oxopentadecane-2-ol arachidate  141 156 438                 8-oxotridecane-2-ol lignocerate -   99 114 508          197 351 369  7-oxotridecane-2-ol lignocerate  113 128 494              8-oxopentadecane-2-ol behenate  127 142 480          225 323 341  7-oxopentadecane-2-ol behenate  141 156 466                 8-oxotridecane-2-ol cerotate -   99 114 536          197 379 397  7-oxotridecane-2-ol cerotate  113 128 522              8-oxopentadecane-2-ol lignocerate  127 142 508          225 351 369 7-oxopentadecane-2-ol lignocerate  141 156 494               5.3.4. Structure elucidation of compound classes F and G  Series F comprised four compounds, found in the same fraction as series E, tentatively identified as VLC 4-alkylbutan-4-olides (4-alkyl-γ-lactones) 5.6 based on similarity of their mass spectra with those of shorter-chain 4-alkylbutan-4-olides reported previously (McFadden et al., 1965). Similar to series E, compounds F could not be silylated (no m/z 73), suggesting that they lacked a hydroxyl group (Fig. 5.7A). All homologs had a diagnostic base peak m/z 85, likely formed via cleavage of the alkyl side chain, as well as m/z 100 formed via McLafferty   145 rearrangement. Chain length-dependent molecular ions were accompanied by fragments M-18, M-18-18 and M-18-44, likely due to loss of water and acetaldehyde (Table 5.6).   Transesterification of F with excess CH3OH/BF3 resulted in an open-chain product that could not be silylated, similar to the behaviour of 5-alkyl-δ-lactones under the same conditions (Jetter and Riederer, 1999b), albeit independent of derivatization time. The transesterification products of F were identified as methyl 4-methoxyalkanoates based on their shared α-fragment m/z 131, and a second homolog-dependent α-fragment (m/z 381 for the C28 homolog in Fig. 5.7B). Molecular ions were found accompanied by M-15, M-32, M-15-32, and M-32-32 due to loss of methyl radical and/or methanol molecule(s). Lastly, reduction of F with excess LiAlH4 followed by reaction with BSTFA resulted in silylated 1,4-diols  with MS characteristics (Fig. 5.7C) indicative of two hydroxyl groups (m/z 73 [(CH3)3Si]+, m/z 75 [(CH3)2SiOH]+, m/z 103 [(CH3)3SiOCH2]+, m/z 147 [(CH3)2SiOSi(CH3)3]+ and m/z 149 [(CH3)2SiOSi(CH3)2OH]+). Other diagnostic signals were the α-fragments m/z 233 (independent of homolog) and m/z 439 (depending on the homolog), along with the base peak m/z 143 resulting from loss of (CH3)3SiOH from the shorter α-fragment. Chain length-dependent M-15 (loss of methyl) and M-90 (loss of (CH3)3SiOH) further confirmed the 1,4-diol structures of the reduction products. All data for F taken together unambiguously established this as a homologous series of 4-alkylbutan-4-olides 5.6.     146  Figure 5.7. Structure elucidation of 4-alkylbutan-4-olides in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of C28 4-alkylbutan-4-olide 5.6b. (B) Mass spectrum and major fragmentations of its product of transesterification with CH3OH/BF3. (C) Mass spectrum and major fragmentations of TMS derivative of LiAlH4 reduction product from C28 4-alkylbutan-4-olide.           147 Table 5.6. Characteristic m/z fragments of 4-alkylbutan-4-olides in leaf and peduncle wax Compound Fragments characteristic of homolog   4-Docosylbutan-4-olide 332                     358                      376                      394   4-Tetracosylbutan-4-olide 360                     386                      404                      422   4-Hexacosylbutan-4-olide 388                     414                      432                      450   4-Octacosylbutan-4-olide 416                     442                      460                      478    Series G comprised six compounds in the least polar fraction of wheat wax, tentatively assigned as internally methyl-branched alkanes 5.7 by analogy with spectra of 15-methyl-alkanes (von Wettstein-Knowles, 2007). Accordingly, compounds G did not exhibit MS fragments indicative of silylation, and other derivatization reactions (such as acetylation, methoximation, transesterification or LiAlH4 reduction) did not alter the compounds in any way, thus confirming the pure hydrocarbon structures.  Homolog-dependent molecular ions M and fragments M-15 (due to loss of the methyl branch) were observed (Fig. 5.8A/C). They were accompanied by α-fragments diagnostic for the methyl branch position, with a predominant m/z 168/169 common to all even-numbered homologs 5.7a/c/e and indicative of methyl branching at C-11 (Fig. 5.8A), and m/z 182/183 common to all odd-numbered homologs 5.7b/d/f and indicative of methyl branching at C-12 (Fig. 5.8C). Complementary homolog-dependent α-fragments were m/z 308/309 for the C32 homolog 5.7e (Fig. 5.8A) and m/z 280/281 for the C31 homolog 5.7d (Fig. 5.8C). Also noticeable were other α-fragments of additional, less abundant regiomers bearing methyl branches at C-9 (m/z 140/141), C-13 (m/z 196/197), or C-15 (m/z 224/225) of even-numbered homologs (Fig. 5.8B), and at C-10 (m/z 154/155) or C-14 (m/z 210/211) of odd-numbered homologs (Fig. 5.8D). A summary of all in-chain branched alkane homologs and isomers identified here is presented in Table 5.7 together with their diagnostic MS fragments.   148 Overall, our MS data identified compounds G as a homologous series of alkanes 5.7 bearing methyl branches primarily on C-11 and C-12, respectively.       Figure 5.8. Structure elucidation of internally branched alkanes in wheat leaf and peduncle wax. (A) Mass spectrum of co-eluting isomers of C32 internally branched alkane and major fragmentations of the main isomer 5.7e. (B) EICs showing intensity of m/z 85, as well as cumulative intensities of regiomer-characteristic α-fragments m/z 140+141, 168+169, 196+197 and 224+225. (C) Mass spectrum of co-eluting isomers of C31 internally branched alkane and major fragmentations of the main isomer 5.7d. (B) EICs showing intensity of m/z 85, as well as cumulative intensities of regiomer-characteristic α-fragments m/z 154+155, 182+183, and 210+211.   149  Table 5.7. Characteristic m/z fragments of internally branched alkanes in leaf and peduncle wax Compound Fragments characteristic of homolog Fragments characteristic of isomer    9-Methylheptacosane 379                394  140/141                             280/281 11-Methylheptacosane   168/169                             252/253 13-Methylheptacosane   196/197                             224/225    10-Methyloctacosane 393                408  154/155                             280/281 12-Methyloctacosane   182/183                             252/253 14-Methyloctacosane   210/211                             224/225    9-Methylnonacosane 407                422  140/141                             308/309 11-Methylnonacosane   168/169                             280/281 13-Methylnonacosane   196/197                             252/253 15-Methylnonacosane   224/225                             224/225    10-Methyltriacontane 421                436  154/155                             308/309 12-Methyltriacontane   182/183                             280/281 14-Methyltriacontane   210/211                             252/253    9-Methylhentriacontane 435                450  140/141                             336/337 11-Methylhentriacontane   168/169                             308/309 13-Methylhentriacontane   196/197                             280/281 15-Methylhentriacontane   224/225                             252/253    10-Methyldotriacontane 421                436  154/155                             336/337 12-Methyldotriacontane   182/183                             308/309 14-Methyldotriacontane   210/211                             280/281  5.3.5. Quantification of new compounds from cuticular waxes of T. aestivum cv. Bethlehem flag leaf blades and peduncles  Gas chromatography with flame ionization detection (GC-FID) was used in a second set of experiments to quantify the newly identified compounds in the total wax mixtures of wheat flag leaves and peduncles. They had fairly low wax coverages over both organs, ranging from 0.008 ± 0.001 μg/cm2 for 4-alkylbutan-4-olides 5.6 to 0.10 ± 0.02 μg/cm2 for secondary alcohols 5.1 in the flag leaf waxes, and from 0.12 ± 0.03 μg/cm2 for 4-alkylbutan-4-olides 5.6 to 0.32 ±   150 0.08 μg/cm2 for internally branched alkanes 5.7 in peduncle waxes (Fig. 5.9A/B). Neither of the oxo-2-alkanol ester homologs could be quantified reliably enough to calculate respective coverages.   Figure 5.9. Total coverages of new compound classes in wheat leaf and peduncle waxes. Coverages (μg/cm2) of new compound classes identified in the total wax mixtures covering the (A) flag leaf blade and (B) peduncle of T. aestivum cv. Bethlehem. Bars represent mean ± standard deviation (n = 5).   151 As all new compound classes comprised series of homologs, their characteristic chain length distributions could be further assessed. Secondary alcohols 5.1 (compounds A) were found as a homologous series with odd-numbered carbon chains from C25 to C35 and a bimodal distribution peaking at C27 5.1b and C33 5.1e (Fig. 5.10A). In contrast, the prim/sec diols 5.2 (compounds B) ranged from C26 to C30, with only even-numbered homologs present and a single homolog, C28 5.2b, accounting for more than 90% of this fraction. Similarly, the prim/sec diol esters 5.3 (compounds C) contained the same alkyl moiety together with various fatty acids, resulting in a homologous series of even-numbered total chain lengths from C44 to C52 that peaked at C50 5.3d. All three compound classes were found only in flag leaf wax. Hydroxy-2-alkanol esters 5.4 (compounds D) were present in both flag leaf and peduncle wax mixtures and had nearly identical chain length profiles in both organs, spanning the odd-numbered homologs from C31 to C39 with a single maximum at C35 5.4c (Fig. 5.10A/B). The fraction of oxo-2-alkanol esters 5.5 (compounds E) comprised odd-numbered C31 to C39 homologs, peaking around C35 5.5c.  The wheat waxes contained even-numbered C26 to C32 4-alkylbutan-4-olides 5.6 (compounds F) with unimodal chain length distributions peaking at C28 5.6b in both organs, albeit with a considerably higher proportion of the C30 homolog 5.6c in peduncle wax as compared to flag leaf wax (Fig. 5.10A/B). Finally, the internally branched alkanes 5.7 (compounds G) also had very similar chain length profiles ranging from C28 to C33 in the wax mixtures from both organs, with even-numbered homologs 5.7a/c/e considerably more abundant than odd-numbered ones 5.7b/d/f (the methyl branch is included in the carbon number). Among the even-numbered alkanes, the C32 homolog 5.7e was most abundant, while C31 5.7d predominated among the odd-numbered homologs.   152  Figure 5.10. Chain length distributions of new compound classes in wheat leaf and peduncle waxes. Relative adundances (%) of each homolog from each of the six new compound classes identified in the total wax mixtures covering the (A) flag leaf blade and (B) peduncle of T. aestivum cv. Bethlehem. Numbers on the x-axis indicate homolog chain length. Bars represent mean ± standard deviation (n = 5). Each bar group adds up to 100%.     153 5.4. Discussion   Our in-depth analysis of the wax mixtures on flag leaf blades and peduncles of T. aestivum cv. Bethlehem revealed the presence of (i) homologous series of secondary alcohols, primary/secondary diols and primary/secondary diol esters, all with secondary hydroxyls on and around identical methylene units 12 carbons away from one chain end; (ii) esterified C13 and C15 2-alkanols with hydroxy- or keto-functions on C-7 or C-8; (iii) a homologous series of -lactones, as such derived from fatty acids with hydroxyl functions on C-4; and (iv) alkanes with (total) carbon numbers ranging from C28 to C33 and methyl branches on C-11 or C-12. Both the homolog and the isomer distribution of all seven compound classes can now be used to infer potential biosynthetic pathways leading to them.  5.4.1. Secondary alcohols, primary/secondary diols and diol esters in flag leaf wax  The wax mixture on wheat flag leaves comprised homologous series of secondary alcohols, primary/secondary diols and corresponding diol esters, thus compounds characterized by secondary hydroxyls. All three compound classes were found to contain mainly regiomers with secondary hydroxyls on a methylene unit 12 carbons away from one end of the hydrocarbon chain, designated as C-12 in the secondary alcohols or as the -12 carbon in the diols. These major isomers in each of the three series were accompanied by further, minor regiomers, characterized by hydroxyls on methylene units in the vicinity of C-12. The finding that all three compound classes shared very similar isomer patterns around their secondary hydroxyl functions suggests that they are biosynthetically related.    154  The isomer distribution in all three compound classes, centred around one carbon position with minor but significant admixtures of isomers with hydroxyls on adjacent carbons is reminiscent of the isomer mixtures of wax secondary alcohols in several other species. For example, C25-C33 secondary alcohols with hydroxyls on C-6 to C-14 have been reported for leaf and fruit capsule waxes of several Papaveraceae species (Jetter and Riederer, 1996), C29-C33 secondary alcohols with hydroxyls on C-12 to C-17 for Pisum sativum leaf wax (Wen et al., 2006a), and secondary alcohols, ketones, vicinal secondary/secondary diols and ketols with functional groups between C-13 and C-15 for Arabidopsis thaliana stem wax (Wen and Jetter, 2009). The latter Arabidopsis compounds are known to be formed by a single enzyme, the mid-chain alkane hydroxylase MAH1 (Greer et al., 2007). This cytochrome P450-dependent monooxygenase exhibits characteristically limited regio-specificity, catalyzing the (repeat) hydroxylation of several methylene units near the centre of the hydrocarbon chain. Accordingly, it has been proposed that homologous cytochrome P450 enzymes with similarly limited regio-specificity form the broad mixtures of secondary alcohol isomers found in the waxes of other plant species (Jetter et al., 2006).   It is important to note that a second important mechanism for introducing hydroxyl functions in wax molecules exists. In many plant species, wax secondary alcohols were identified with hydroxyl groups exclusively on even-numbered carbon atoms, such as 10-nonacosanol on Malus domestica fruit (Dong et al., 2012), 10-heptacosanol, 10-nonacosanol and 12-nonacosanol on Osmunda regalis fronds (Jetter and Riederer, 2000), or C23-C33 2-alkanols from Aloe arborescens leaves (Racovita et al., 2015). It has been proposed that such secondary   155 alcohols, with functional groups on every other rather than adjacent carbons, may be derived from β-hydroxyacyl-CoA intermediates of fatty acid elongation instead of cytochrome P450 hydroxylation (Racovita et al., 2015; Wen and Jetter, 2007).  Our current findings of broad isomer distributions for the wheat secondary alcohols suggest that they may be synthesized by a cytochrome P450 enzyme and not as derivatives of fatty acid elongation intermediates. We conclude that wheat likely possesses a MAH1-like enzyme hydroxylating preferentially on C-12 of C25-C35 alkane substrates. Interestingly, the resulting secondary alcohols had a bimodal chain length distribution, peaking at C27 and C33 (compare Fig. 5.10A), very different from that of the corresponding alkane precursors, with a single maximum at C31 (Racovita et al., 2016). This suggests either an unusual chain length preference of this hydroxylase for C33 and C27 alkane substrates, or else the presence of two very similar enzymes with similar regio-specificity but different substrate chain length preference.   The broad isomer distributions of the primary/secondary diols in the wheat leaf wax suggest that they are also formed by a cytochrome P450 hydroxylase. It seems plausible that the enzyme(s) converting alkanes into secondary alcohols (see above) can also hydroxylate primary alcohols into corresponding diols. Considering the clear predominance of C28 diol (compare Fig. 5.10A), very similar to the profile of precursor primary alcohols (Racovita et al., 2016), it appears that the pool of alcohols is used non-discriminatively by the enzyme(s). Interestingly, both the secondary alcohols and diols had in-chain hydroxyls mainly on the methylene unit 12 carbons away from the methyl (or ) terminus, while the distance to the second methyl or alcohol terminus varied. We thus propose that the wheat MAH1-like enzyme may achieve its   156 (limited) regio-specificity by tight binding of the short alkyl moiety of the alkane or primary alcohol substrates, effectively counting carbons in from the methyl terminus. It should also be noted that the C28 alcohol thus serving as substrate has an overall molecule geometry, including the carbon chain and the oxygen atom, resembling the C29 alkane homolog against which the enzyme(s) seemed to discriminate (compare Fig. 5.10A).    Primary/secondary diols with in-chain hydroxyl groups on several adjacent carbons, similar to those in wheat wax, had been reported for several other plant species before. For example, Pisum sativum leaves contain C26-C28 diols with 1,12- through 1,17-functionalities (Wen et al., 2006a), while the eustigmatophyte Nannochloropsis gaditana has C28-C36  diols with 1,13- through 1,19-geometries (Mejanelle et al., 2003). Again it seems likely that the secondary hydroxyls of these diols may be introduced by cytochrome P450 enzymes. In contrast, other diols were detected with the secondary hydroxyls exclusively on odd-numbered carbons (when counting from the primary OH), and are thought to be formed as elongation by-products. Examples for such compounds include the C28-C38 1,5-diols in Taxus baccata needles (Wen and Jetter, 2007), C28-C32 1,11-diols in Osmunda regalis fronds (Jetter and Riederer, 1999a), C32 1,9-, 1,11- and 1,13-diols in Myricaria germanica leaves (Jetter, 2000), or C30 1,11-, C32 1,13-, C34 1,15- and C36 1,17-diols in Azolla filliculoides whole plants (Speelman et al., 2009).    Finally, esters of primary/secondary diols have scarcely been reported, including the C46-C52 esters of C30 1,11-, C32 1,13-, C34 1,15- and C36 1,17-diols of the fern Azolla filliculoides (Speelman et al., 2009) and C40-C52 esters of C30 1,5-, 1,7- and 1,9-diols of the moss Funaria hygrometrica (Busta et al., 2016). It seems plausible that such esters are acylation products of   157 corresponding free primary/secondary diols, formed by wax ester synthases. The diol esters found in wheat wax had nearly identical homolog and regiomer distributions as the accompanying free diols, suggesting that the responsible wax ester synthase shows no preference for the diol substrate. Instead, based on the chain length distribution of diol esters peaking at C50 (compare Fig. 5.10A), it appears that the wax ester synthase shows high substrate preference for C22 acyl-CoA as its second substrate. Since the same preference was noted for a wax ester synthase forming unsubstituted VLC esters in the leaves of the same wheat cultivar (Racovita et al., 2016), it is very likely that the same wax ester synthase produces esters of both primary alcohols and diols. This conclusion is in accordance with our finding that only the primary hydroxyl of the primary/secondary diols was esterified, but not the secondary group.    5.4.2. Hydroxy-2-alkanol esters and oxo-2-alkanol esters in flag leaf and peduncle waxes  2-Alkanol esters have been identified in several grass species, typically as minor components associated with the much more prominent wax -diketones (Racovita et al., 2016; von Wettstein-Knowles and Netting, 1976a; von Wettstein-Knowles et al., 1984). Most previous analyses revealed only esters of 2-alkanols bearing no other functional groups, except for one report identifying 7-oxo-pentadecan-2-ol as a minor constituent of barley spike wax (von Wettstein-Knowles and Madsen, 1984). The same compound was thus now also detected in wheat waxes, together with its C13 homolog and 8-oxo isomers. Furthermore, we identified the four corresponding hydroxy-2-alkanols, with chain lengths and in-chain functional group positions matching those of the keto-2-alkanols, all esterified with various fatty acids.     158 The common overall chain length profiles and isomer distributions of the hydroxy-2-alkanol esters and oxo-2-alkanol esters identified here suggest a biosynthetic relationship between both compound classes. Considering that regiomers with functional groups on adjacent carbons were detected for all the 2-alkanol ester derivatives, it is likely that the in-chain functionalities are introduced by a cytochrome P450 hydroxylase similar to MAH1. Accordingly, we propose that wheat has a cytochrome P450 enzyme catalyzing either a single hydroxylation leading to the hydroxy-2-alkanol esters or a double-hydroxylation to the corresponding oxo-2-alkanol esters.   The apparent regio-specificity of this enzyme, as a C-7/C-8 hydroxylase, clearly differs from that of the cytochrome P450 discussed above for the formation of secondary alcohols and diols. Interestingly, both enzymes also appear to have different expression patterns, since the secondary alcohols/diols were found only in flag leaves, whereas the oxidized 2-alkanol esters were detected in both peduncle and flag leaf waxes (even though potential precursors for both product groups were likely present in both organs). Taken together, we conclude that T. aestivum cv. Bethlehem has at least two distinct cytochrome P450 enzymes involved in wax biosynthesis, one being a C-12-specific hydroxylase forming secondary alcohols and diols, and the other one a C-7/8-specific hydroxylase involved in formation of oxidized 2-alkanol esters.  The matching chain length distributions of the esterified hydroxy/oxo-2-alkanols and 2-alkanols (Racovita et al., 2016), both peaking at C13/15, indicate the latter could be the substrates for hydroxylation by this second hydroxylase targeting carbons C-7 and C-8 of the 2-alkanol moiety. The C-7 and C-8 positions found preferentially hydroxylated in 2-alkanols are very   159 similar to those of hydroxyl groups in oxidized -diketones (e.g., 8- and 9-hydroxy-hentriacontane-14,16-dione), relative to both ends of either the 2-alkanols or the alkyl moiety within the -diketones (Racovita et al., 2016). It is thus plausible that the same hydroxylase is involved in the formation of hydroxy/oxo-2-alkanol esters and hydroxy-β-diketones. In fact, the geometry of the -diketone, bearing a C13 alkyl tail on a -CO-CH2-CO- functionality, is fairly similar to that of the (C15) 2-alkanol esters, having a C13 alkyl tail on a –CHCH3-O-CO- functionality. The same cytochrome P450 may thus accept either the -diketone or the 2-alkanol ester as substrate for hydroxylation on C-7 or C-8. However, it cannot be ruled out that hydroxylation may occur in earlier stages of the pathways leading to 2-alkanol esters and β-diketones, rather than on the latter two products.  5.4.3. 4-Alkylbutan-4-olides and internally branched alkanes in flag leaf and peduncle waxes  While δ-lactones have been identified as prominent components of the wax mixture on leaves of Cerinthe minor (Jetter and Riederer, 1999b), the corresponding 4-alkylbutan-4-olides (4-alkyl-γ-lactones) have not been reported before as plant cuticular wax constituents. However, γ-lactones have been identified in several plant species, albeit without localizing them to a specific organ or tissue. For example, a homologous series of C24 to C30 γ-lactones was detected in the ground aerial parts of Flourensia cernua (Mata et al., 2003), the C32 γ-lactone in the aerial parts of Pluchea lanceolata (Ali et al., 2001), and an unsaturated C21 homolog in the stem bark of Garcinia mannii (Hussain and Waterman, 1982). While the biosynthetic pathways leading to these structures remain unknown, it seems possible that they are formed via α-hydroxylation of   160 acyl-CoA substrates similar to the reactions thought to lead to 1,2-bifunctional wax compounds (Buschhaus et al., 2013a). The resulting α-hydroxyacyl-CoA intermediates might be elongated further by the FAE complex, leading to 4-hydroxyacyl-CoAs that, upon intramolecular esterification, would yield 4-alkylbutan-4-olides.   Finally, alkanes with an in-chain methyl branch had been described before as constituents of insect cuticular waxes (Nelson and Sukkestad, 1975), of wool wax (Mold et al., 1966), but also of plant cuticular waxes, namely leaf waxes of walnut tree (Stránsky et al., 1970) and spike waxes of barley (von Wettstein-Knowles, 2007). It is important to note that, different from previous reports, the in-chain-branched alkane regiomers identified here in wheat wax have methyl groups separated by two carbons, located on odd-numbered carbons in the chain of even-numbered homologs and on even-numbered carbons in odd-numbered homologs. Based on these isomer patterns, we conclude that the methyl groups are most likely introduced during FAE-catalyzed elongation of acyl-CoA precursors, possibly by incorporating a methylmalonyl-CoA extender unit in lieu of the normal malonyl-CoA. This assertion is supported by the finding that even-numbered internally branched alkanes (having odd chain lengths) were more abundant than their odd-numbered homologs (with even chain lengths), in the same way that odd-numbered n-alkanes are more abundant than those with even chain lengths.        161 Chapter 6: Novel polyketides and polyketide-derived compounds from cuticular waxes of Triticum aestivum cv. Bethlehem   6.1. Introduction  The above-ground surfaces of primary plant organs are coated with a lipidic cuticle to minimize uncontrolled water loss to the dry atmosphere. The plant cuticle is composed of a polyester matrix known as cutin and cuticular waxes lying on top of cutin (epicuticular waxes) or embedded within the cutin matrix (intracuticular waxes) (Buschhaus and Jetter, 2011).   Cutin is a biopolymer incorporating a diversity of long-chain (LC, i.e. C16 and C18) monomers, most commonly ω-hydroxy fatty acids or dicarboxylic acids, sometimes with additional in-chain hydroxyl, epoxy, or keto functionalities that participate in cross-linking of polyester chains (Nawrath, 2006). Cuticular waxes are complex mixtures consisting mostly of very-long-chain (VLC, i.e. >C20) aliphatics, such as fatty acids, primary alcohols, alkyl esters, aldehydes, alkanes, secondary alcohols and ketones (Jetter et al., 2006). In many species, the wax mixtures also comprise alicyclics, including a wide variety of triterpenoids (Belge et al., 2014; Bianchi et al., 1993; Markstadter et al., 2000; Nordby and McDonald, 1994; van Maarseveen et al., 2009), and in some species also aromatics such as 5-alkylresorcinols (Adamski et al., 2013; Ji and Jetter, 2008), benzyl and phenethyl esters (Buschhaus et al., 2007a, 2007b; Gülz and Marner, 1986; Jetter and Riederer, 1996; Rapley et al., 2004), and 4-hydroxyphenylpropyl, 3,4-dihydroxyphenylpropyl and 3,4-dihydroxyphenylbutyl esters (Jetter et al., 2002; Wen and Jetter, 2007).   162  Aliphatic cuticular waxes typically occur as series of either even- or odd-numbered homologs, according to the biosynthetic mechanisms leading to the various product structures. From molecular genetic studies of the model plant species Arabidopsis thaliana it is well established that wax biosynthesis utilizes C16 and C18 fatty acids formed de novo in epidermal plastids (Kunst et al., 2006; Samuels et al., 2008). In the course of transport to the endoplasmic reticulum, these acids are activated into thioesters by long-chain acyl-CoA synthetase (LACS) enzymes. Fatty acyl elongase (FAE) complexes then extend the acyl chains with C2 units, to yield mixtures of homologous acyl-CoAs with even carbon numbers. In the FAE reaction cycle, first a ketoacyl-CoA synthase (KCS) enzyme catalyzes a Claisen condensation of the acyl-CoA substrate with malonyl-CoA, and then three other enzymes perform the stepwise reduction of the -keto group into a methylene, to yield an acyl-CoA two carbons longer than its precursor. FAE complexes with different KCS enzymes have different product chain length specificities, and the interplay of the different FAEs thus determines the overall chain length profile of the acyl-CoA product pool generated.    The elongated acyl-CoAs are finally converted into diverse end products by modification of their head-groups. On the acyl reduction pathway, a fatty acyl-CoA reductase (FAR) reduces the carboxyl functionality into primary alcohols, part of which are exported to the cuticle, while others are fused with (very-) long-chain acyl-CoAs by a wax synthase (WS) to produce alkyl esters. As neither of the reactions on this pathway affect the carbon structure of the substrates, the alcohol and ester products all have even carbon numbers. On the second pathway, a different reductase partially reduces acyl-CoAs to even-numbered aldehydes, some of which are exported   163 to the cuticle, while others are decarbonylated to odd-numbered alkanes. The alkanes may undergo single or double hydroxylation by a mid-chain alkane hydroxylase (MAH) to produce odd-numbered secondary alcohols, ketones, diols and ketols.  However, cuticular waxes of many plant species also include VLC aliphatics with more than one functional group. Over the past two decades, numerous new wax structures have been identified with one functional group at the end of the aliphatic chain, similar to the ubiquitous wax compounds described above, and an additional functionality inside the chain. Such primary/secondary bifunctional compounds were identified for example as alkanediols, ketoaldehydes, ketoalcohols and ketoalkyl esters in the wax of Osmunda regalis fronds (Jetter and Riederer, 1999a), or δ-lactones in leaf waxes of Cerinthe minor (Jetter and Riederer, 1999b). 1,3-Alkanediols and 3-hydroxyaldehydes were reported for leaf waxes of Ricinus communis (Vermeer et al., 2003), further alkanediols in Pisum sativum leaves (Wen et al., 2006a), 1,5-alkanediols and 5-hydroxyaldehydes in Taxus baccata needles (Wen and Jetter, 2007), 1,3- and 1,2-alkanediol acetates in Cosmos bipinnatus petals (Buschhaus et al., 2013b), 3-hydroxyacid derivatives in Aloe arborescens (Racovita et al., 2015), and 3-hydroxyacid and alkanediol esters in Funaria hygrometrica (Busta et al., 2016).  In the waxes of many other species, bifunctional compounds with two in-chain functionalities were identified. For example, both prim./sec. and sec./sec. diols were found in cuticular waxes of Myricaria germanica leaves (Jetter, 2000) and in the gymnosperm needle waxes (Wen et al., 2006b). Most prominently, sec./sec. diketones accumulate to relatively high concentrations in wax mixtures of diverse plant species, in most cases as single compounds with   164 odd carbon numbers and -constellation of the two carbonyl groups (i.e., with one methylene unit between them). Specifically, -diketones such as nonacosane-6,8-dione, hentriacontane-8,10-dione and tritriacontane-10,12-dione are found for example in Buxus sempervirens (Dierickx, 1973), tritriacontane-16,18-dione is found in Eucalyptus globulus (Horn et al., 1964), nonacosane-10,12-dione and hentriacontane-10,12-dione in Hosta ‘Krossa Regal’ (Jenks et al., 2002), and nonacosane-8,10-dione, nonaconsane-12,14-dione, hentriacontane-10,12-dione and hentriacontane-14,16-dione in various species of Rhododendron (Evans et al., 1975b). The cuticular waxes of many Poaceae contain particularly high concentrations of -diketones, most frequently hentriacontane-14,16-dione. Accordingly, this C31 compound dominates the waxes of Agropyron dasystachyum, A. riparium and A. elongatum (Tulloch, 1983), Hordeum vulgare (von Wettstein-Knowles and Netting, 1976c), and several species of wheat (Tulloch and Hoffman, 1973, 1971; Tulloch and Weenink, 1969).   Among the Poaceae, wheat is of particular interest due to its world-wide role as primary staple crop. There are many reports on the composition of cuticular wax mixtures on various wheat species, cultivars and organs (Bianchi and Corbellini, 1977; Bianchi et al., 1980; Tulloch and Hoffman, 1973, 1971; Tulloch and Weenink, 1969; Wang et al., 2015a, 2015b), documenting the presence of ubiquitous compound classes such as fatty acids, aldehydes, alkanes, primary alcohols and alkyl esters, together with -diketones such as hentriacontane-14,16-dione and its hydroxylated derivatives. Recently, the Triticum aestivum cultivar Bethlehem was selected for genetic investigations into -diketone biosynthesis (Hen-Avivi et al., 2016), and comprehensive wax analyses of this cultivar were required to establish a chemical reference dataset. Accordingly, a comparative analysis of the wax mixtures on flag leaves and   165 peduncles of wheat cv. Bethlehem was performed, confirming the presence of the typical wheat wax compounds, most prominently the mid-chain -diketone hentriacontane-14,16-dione together with 8- and 9-hydroxyhentriacontane-14,16-dione (Racovita et al., 2016). Several classes of aliphatic and aromatic esters were identified that had not been described in wheat wax before, and detailed quantification of their homolog and isomer compositions enabled predictions regarding the number and specificities of ester synthase enzymes involved in wheat wax biosynthesis (Racovita et al., 2016).    As many compounds in the wheat wax mixtures remained unidentified, a more detailed analysis of the cv. Bethlehem flag leaf and peduncle waxes was carried out, relying on the pre-separation of the complex mixture by thin layer chromatography (TLC). Several TLC fractions were found to contain novel compounds, and six of them were identified as homologous series of sec. alcohols, prim./sec. diols, esters of prim./sec. diols, esters of sec./sec. diols (hydroxy-alkan-2-ols), esters of sec./sec. ketols (oxo-alkan-2-ols) and -lactones. Thus, all the new compound classes were recognized as derivatives of the ubiquitous wax constituents, carrying additional secondary hydroxyl or keto functions. It was therefore hypothesized that all the novel compounds were formed by oxidation of respective wax precursors, likely mediated by P450-dependent monooxygenases (Racovita and Jetter, 2016).   While oxygenated derivatives of many ubiquitous wax compounds were thus identified, no homologs, isomers or derivatives of the predominant mid-chain -diketones were found. However, several compounds in the wheat wax mixtures and their TLC fractions could not be identified to date, leading us to speculate that some of them might indeed represent the -  166 diketone derivatives missing so far. Therefore, the goal of the present work was to search the flag leaf and peduncle waxes of the wheat cv. Bethlehem for novel -diketone derivatives. To this end, the wax mixtures were separated by preparative TLC to obtain large enough quantities for detailed GC-MS analysis, and six TLC fractions were found to contain unknown compounds. They were transformed into various derivatives to distinguish functional groups in the novel structures and assign isomer compositions.  6.2. Experimental  6.2.1. Plant material  Triticum aestivum cv. Bethlehem plants were grown in greenhouses at the Weizmann Institute of Science (Rehovot, Israel). The type of soil used was a mix of 50% peat and 50% turf and it was watered every 3-4 days, using approx. 400 mL water per 5 L pot. Growth conditions included alternating light / dark cycles of 12-14 h / 10-12 h, with temperatures of 24-260C / 17-180C and a photon flux during light cycles of 180 μmol m-2 s-1. For every preparative thin layer chromatography experiment, ten flag leaf blades with an area of 40-50 cm2 each (measured using ImageJ) and ten peduncles with a diameter of 2 mm and a length of 15-25 cm were harvested in August 2014 from mature wheat plants using clean razor blades.  6.2.2. Wax extracts   Leaves and peduncles were submerged into 10 mL chloroform (Aldrich, ≥99%, with 0.75% ethanol as stabilizer) at ambient temperature for 30 s, with agitation. The chloroform with   167 extracted waxes was transferred to another vial and the plant material was extracted a second time with another 10 mL chloroform for another 30 s. The two chloroform extracts were combined and dried under a stream of N2 (Praxair, ≥99.998%) at 500C. The remaining waxes were stored until fractionation by TLC.  6.2.3. Preparative thin layer chromatography   Fractionation of cuticular wax classes by preparative TLC was done using the sandwich technique (Tantisewie et al., 1969). A mixture CHCl3:EtOH 98:2 (v/v) was employed as mobile phase, while the stationary phase was represented by SiO2-coated glass plates (Uniplate Analtech, silica gel 60 F254 layer thickness: 1 mm, size: 20x20 cm, with 4 cm concentrating zone). Following separation, TLC bands were visualized under 365 nm UV light after spraying the plates with a solution of 5 mg primuline (Aldrich, 50% dye content) in 100 mL (CH3)2CO:H2O 80:20 (v/v). All bands were scraped off the plates with spatulas and collected into 20 mL scintillation vials, where they were extracted twice with 10 mL portions of CHCl3 at ambient temperature, for 30 s each. After filtration through glass wool (Supelco), the combined extracts were partially evaporated under N2 at 500C, transferred to 2 mL GC autosampler vials, evaporated again to dryness, and stored until GC-MS analysis.  6.2.4. Derivatization reactions   Derivatization reactions in preparation for GC-MS analysis were carried out using various derivatization reagents: N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA, Aldrich, GC   168 grade) for silylation; acetic anhydride (Aldrich, ≥98%) for acetylation; lithium aluminum hydride (Aldrich, ≥95%) for reduction; and O-methylhydroxylamine hydrochloride (Aldrich, ≥98%) for conversion of carbonyls into methoximes. Specific derivatization protocols are described below.   All samples were subjected to silylation before injection into the GC-MS, by refluxing in a mixture of 10 μL BSTFA and 10 μL pyridine at 70°C, for 20 min. Excess reagents were then evaporated under N2 and the silylated waxes re-dissolved with 50 μL CHCl3.  Acetylation of hydroxyl-containing specimens was performed by refluxing the dry wax in a mixture of 10 μL acetic anhydride and 10 μL pyridine at 70°C, for 5 min, followed by overnight stirring at ambient temperature. Following evaporation of excess reagents under N2, silylation was carried out as described above.  Wax samples for reduction were dissolved in 50 μL (CH3CH2)2O to which 0.1 mg LiAlH4 were added. The mixture was left to react overnight at 70ºC, with the vial cap closed. Then it was quenched with 10% H2SO4 and extracted three times with 60 μL (CH3CH2)2O each. The combined extracts were evaporated to dryness and silylated as described above.  Conversion of carbonyl-containing specimens into the corresponding methoximes was accomplished by heating them with 20 μL of a saturated solution of O-methylhydroxylamine hydrochloride in pyridine:CHCl3 7:3 (v/v) at 70°C, for 30 min. After partitioning between 50 μL distilled H2O and 50 μL CHCl3, the organic fraction was collected and the aqueous phase   169 extracted once more with 50 μL fresh CHCl3. Finally, the combined chloroform extracts were evaporated to dryness and silylated as described above.  6.2.5. Gas chromatography-mass spectrometry   The GC instrument used for identification and relative quantification of wax constituents was equipped with capillary GC column (6890N, Agilent, Avondale PA, USA; length: 30 m; type: HP-1 100% PDMS; i.d.: 0.32 mm; df: 0.1 µm), on-column injector at 500C, and MS detector (5973N, Agilent, EI-70 eV, ionization source temp.: 2400C, m/z 50-750). It employed helium (Praxair, ≥99%) as carrier gas at a flow rate of 1.4 mL min-1 and was programmed to follow the following temperature program: 2 min at 50ºC, ramp 40ºC min-1 to 200ºC, constant for 2 min, ramp 3ºC min-1 to 320ºC, constant for 30 min.  6.3. Results  This study aimed at a detailed analysis of the compounds associated with -diketones in wax mixtures on wheat flag leaves and peduncles. To enable structure elucidation, the wax mixtures were extracted from the two organs and separated by TLC (silica, mobile phase CHCl3:EtOH 98:2), and diverse unknown constituents were located in specific fractions. Their chemical structures were studied by inspecting MS fragmentation patterns of a variety of derivatives, in each case chosen to provide multiple lines of evidence for the presence and relative position of functional groups along the hydrocarbon backbone.     170 Based on their TLC behaviour and common MS fragmentation patterns, the unknown wheat wax constituents could be grouped into six different compound classes (Fig. 6.1). All six fractions, designated as A-F, were detected equally in flag leaf and peduncle waxes. Preliminary evidence suggested that compounds A-C bore structural similarities, however they were found to have widely differing polarities since class A co-eluted with wax -lactones and keto-alkan-2-ol esters (Rf 0.44), B with -diketones (Rf 0.86), and C with prim./sec. alkanediol esters (Rf 0.54). Similarly, classes D and E were found structurally related to each other, again despite differing polarities, as D also co-eluted with -diketones (Rf 0.86) and E with hydroxy--diketones (Rf 0.33). Finally, class F was isolated as a fraction of its own, with polarity between those of -diketones and alkyl esters (Rf 0.91).  6.6a (n=10)6.6b (n=11)6.6g (n=16)...OCH3 CH3O  n 6 6.1a (n=6)6.1b (n=7)6.1c (n=8)OHCH3 CH3O7 6OCH3 CH3OH676.2a 6.2bOCH3CH3O  7 66.3aOHCH3CH3O  7 66.4a OCH3CH3OH  7 66.4bCH3OCH3O  146.5aCH3OHCH3O  nCH3 CH3O7 6OCH3 CH3676.7a 6.7b Figure 6.1. Polyketides and polyketide-derived compounds identified in the wax mixtures of T. aestivum cv. Bethlehem.   171 Initially, the -diketones 6.1 contained in the same fraction as classes B and D were investigated for reference. As previously described (Racovita et al., 2016), the -diketone fraction was largely dominated by one compound, hentriacontane-14,16-dione 6.1b (ca. 97% of the -diketone class). It was identified by the characteristic MS fragmentation pattern of its trimethylsilyl (TMS) enol ether derivative, exhibiting -bond cleavage products together with a series of M, M-15 and M-90 ions (Fig. 6.2A). Two more aliquots of the same TLC fraction were subjected to derivatization with O-methylhydroxylamine and lithium aluminum hydride (LiAlH4), and the resulting bis-methoxime and reduction products showed characteristic - and McLafferty fragments confirming the presence of two keto groups in -constellation (Fig. 6.2B/C). The same wax fraction contained further mid-chain -diketones that had not been described in wheat wax before, and close inspection of the TMS derivative and LiAlH4 reduction mixtures identified the C29 structure nonacosane-14,16-dione 6.1a (<0.5% of the compound class) and two isomers of the C33 homolog, tritriacontane-14,16-dione 6.1c and tritriacontane-16,18-dione (ca. 1% and 2% of the compound class, respectively).   172  Figure 6.2. Identification of mid-chain β-diketone in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of the TMS derivatives obtained from the two enol tautomers of hentriacontane-14,16-dione 6.1b. (B) Mass spectrum and major fragmentations of the O-methylhydroxylamine derivative of hentriacontane-14,16-dione 6.1b. (C) Mass spectrum and major fragmentations of the TMS derivative of the diol resulting from LiAlH4 reduction of hentriacontane-14,16-dione 6.1b.   173  6.3.1. Structure elucidation of compound classes A – C  Compound class A was detected as a single GC peak in a wheat wax fraction of intermediate polarity (Rf 0.44). Based on MS similarity with published data for C29 β-ketols in Brassicaceae waxes (Holloway and Brown, 1977; Wen and Jetter, 2009), A was hypothesized to have mid-chain β-ketol structure 6.2. In particular, the TMS derivative of A had a fragment m/z 73 diagnostic for a hydroxyl group and a fragment m/z 130 indicative of a β-ketol structure (Wen and Jetter, 2009), while lacking a fragment m/z 147 indicative of a second hydroxyl function (Fig. 6.3A). Two further fragments, M-15 (due to loss of methyl radical) and M-90 (due to loss of (CH3)3SiOH), further confirmed the ketol structure and indicated an overall chain length of C31. Finally, the α-fragments of the OTMS group (m/z 327 and m/z 313) and of the carbonyl group (m/z 211 and 355) identified one ketol isomer 6.2a with keto and hydroxyl functionalities on C-14 and C-16, while pairs of further ions (m/z 285 and 355; m/z 239 and 327) revealed the presence of a second isomer 6.2b with reversed configuration of keto and hydroxyl functions at C-16 and C-14, respectively.   For structure confirmation, the fraction containing A was treated with LiAlH4, a derivatizing reagent that can probe keto groups by reducing them to hydroxyls. In the resulting mixture, the original compound A was replaced with a single new one. The mass spectrum of its TMS derivative showed the characteristic fragments m/z 73 and m/z 147 for diols (Jetter et al., 1996), together with M-15 and M-90 ions indicative of chain length (Fig. 6.3B). Two pairs of α-fragments (m/z 285 and m/z 313; m/z 429 and m/z 401) together with one pair of product ions   174 formed by loss of (CH3)3SiOH (m/z 339 and m/z 311) corroborated the presence of two OTMS groups, and the LiAlH4 reduction product was thus identified as hentriacontane-14,16-diol. Taking this result together with the TLC behaviour and GC-MS data of the TMS derivative, we conclude that A is a mixture of two co-eluting C31 mid-chain β-ketol regiomers, 16-hydroxyhentriacontan-14-one 6.2a and 14-hydroxyhentriacontan-16-one 6.2b. Thus, the compounds in A are -ketols that are  structurally related to the major -diketone present in wheat wax, hentriacontane-14,16-dione 6.1b, with identical chain length of C31 and position of functional groups on C-14 and C-16, but with either one of the keto groups replaced by a hydroxyl.         175  Figure 6.3. Structure elucidation of mid-chain β-ketols in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of co-eluting TMS derivatives of 16-hydroxyhentriacontan-14-one 6.2a and 14-hydroxyhentriacontan-16-one 6.2b. (B) Mass spectrum and major fragmentations of TMS derivative of the diol obtained via LiAlH4 reduction of 16-hydroxyhentriacontan-14-one 6.2a and 14-hydroxyhentriacontan-16-one 6.2b.     176 Class B comprised a single GC peak in the TLC fraction also containing the -diketones. Treatment of this fraction with BSTFA left B unchanged, its mass spectrum lacking features (m/z 73, 75) characteristic of hydroxyls (Fig. 6.4A) and instead suggesting the presence of two alkyl termini without functional groups (prominent m/z 57, 71, 85, etc.). A pair of α-fragments (m/z 211 and m/z 239) in conjunction with a molecular ion m/z 450 further suggested a C30 chain bearing an -diketo functionality on C-14 and C-15. Alternative structures, such as C29 or C31 mono-ketone isomers, seemed unlikely due to the lack of other fragmentations around the carbonyl group, such as McLafferty rearrangement with double hydrogen transfer characteristic of VLC ketones leading to fragments 16 amu higher than corresponding α-fragments (Vajdi et al., 1981).   To confirm the presence of two carbonyl groups in B, two more derivatives were generated and characterized by MS. First, condensation with O-methylhydroxylamine yielded a new compound with prominent fragment M-31, due to loss of a methoxy unit (m/z 477), and α-fragments 29 Da heavier than those of the native compound (m/z 240 and m/z 268) (Fig. 6.4B), together indicating the presence of at least one carbonyl function. Finally, reduction of B with excess LiAlH4 followed by silylation resulted in a single compound with characteristic fragments m/z 73 and m/z 147, prominent α-fragments m/z 285 and 313, and an ion M-15 at m/z 583 (Fig. 6.4C), and thus identified as triacontane-14,15-diol. This finding, together with the combination of spectral features of the other derivatives, unambiguously identified compound B as the mid-chain-diketone, triacontane-14,15-dione 6.3a. The two diols resulting from reduction of the-diketone triacontane-14,15-dione (Fig. 6.4C) and of the -diketone hentriacontane-14,16-dione (see Fig. 6.2C) share many prominent features, however the two compounds are distinguished by   177 different GC retention times, and by molecular ions as well as fragments m/z 522, 311 and 339 characteristic of the -diketone-derived diol.  Figure 6.4. Structure elucidation of mid-chain α-diketone in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of triacontan-14,15-dione 6.3a. (B) Mass spectrum and major fragmentations of O-methylhydroxylamine derivative of triacontan-14,15-dione 6.3a. (C) Mass spectrum and major fragmentations of TMS derivative of the diol obtained via LiAlH4 reduction of triacontan-14,15-dione 6.3a.   178 Compound class C was found in a single GC peak, detected in a fraction (Rf 0.54) slightly less polar than that containing the β-ketols (class A). Again, MS similarity with previously reported Brassica and Arabidopsis wax ketols (Holloway and Brown, 1977; Wen and Jetter, 2009) suggested that C was a mixture of isomeric α-ketols (acyloins) 6.4a-b. The mass spectrum of TMS-derivatized C showed an ion m/z 73 not accompanied by m/z 147, suggesting a single hydroxyl group, and α-fragments m/z 285 and 313 suggesting OH-group location 14 and 16 carbons in from one alkyl chain end, respectively (Fig. 6.5A). While the spectrum thus far closely resembled that of a (TMS-derivatized) secondary alcohol, nonacosan-14-ol, further MS features clearly distinguished the two compounds. In particular, the TMS derivative of C exhibited a fragment M-15 characteristic of a C31 ketol structure, and two -fragments at m/z 211 and 239, interpreted as C14 and C16 acylium ions, showing the presence of a carbonyl group. C was thus recognized as an α-ketol mixture 6.4a-b rather than a secondary alcohol.   Further confirmation of the α-ketol structure was provided by the MS characterization of three more derivatives of C. One of them, generated by acetylation with acetic anhydride, showed very prominent acylium fragments m/z 211 and 239, together with a series of ions M, M-43 (loss of acetyl) and M-60 (loss of acetic acid) (Fig. 6.5B), thus confirming the presence of two isomeric α-ketol structures 6.4a-b. A third aliquot of the TLC fraction was reduced with LiAlH4, resulting in the same α-diol as from B (data not shown), and therefore further corroborating the C30 structure bearing two functional groups on C-14 and C-15. Finally, to also directly probe the presence of a carbonyl group, another aliquot of the fraction was derivatized with O-methylhydroxylamine and then silylated. While the resulting oxime exhibited the same hydroxyl -fragments m/z 285 and 313 as the simple silyl derivative, the acylium ions were replaced by   179 fragments 29 Da heavier (m/z 240 and 268) (Fig. 6.5C), confirming the presence of one carbonyl function. This interpretation was underpinned by ions M-15 (loss of methyl), M-31 (loss of methoxy) and M-90 (loss of (CH3)3SiOH), and the positions of the carbonyl and hydroxyl groups on either C-14 or C-15 were again confirmed by several other α-fragments in the spectrum of the silylated oxime along with some of their product ions. Overall, the TLC behaviour of the native compound and our GC-MS results for various derivatives identified C as a mixture of two C30 mid-chain α-ketols, 15-hydroxytriacontan-14-one 6.4a and 14-hydroxytriacontan-15-one 6.4b. Interestingly, these structures are closely related to those of compound classes A and B, the former identified as mid-chain -ketols 6.2a-b differing from the α-ketols 6.4a-b by the presence of one methylene unit between functional groups, and the latter identified as a mid-chain-diketone 6.3a differing from the α-ketols 6.4a-b only in functional group oxidation state.    180  Figure 6.5. Structure elucidation of mid-chain acyloins in wheat leaf and peduncle wax. (A) Mass spectrum of major fragmentations co-eluting TMS derivatives of 15-hydroxytriacontan-14-one 6.4a and 14-hydroxytriacontan-15-one 6.5a. (B) Mass spectrum and major fragmentations of co-eluting Ac derivatives of 15-hydroxytriacontan-14-one 6.4a and 14-hydroxytriacontan-15-one 6.5a. (C) Mass spectrum and major fragmentations of co-eluting O-methylhydroxylamine/TMS derivatives of 15-hydroxytriacontan-14-one 6.4a and 14-hydroxytriacontan-15-one 6.5a.    181  6.3.2. Structure elucidation of compound classes D and E  Class D was represented by a single compound in the same TLC fraction as β-diketones and -diketones (B), and was therefore suspected to have diketone structure as well. D was not affected by treatment with BSTFA, and its mass spectrum accordingly lacked all fragments characteristic of OH groups (m/z 73, 75) (Fig. 6.6A). Instead, it showed characteristic α-fragments m/z 85 and 435, together with a base peak m/z 100 due to a McLafferty rearrangement (without double H transfer) indicating the presence of at least one carbonyl group. The two ions M and M-18 (due to loss of water) indicated a molecular weight of 492 Da, pointing to either a C33 diketone or C34 monoketone structure.  To further investigate the number and relative positions of carbonyls in D, the fraction was derivatized with O-methylhydroxylamine. The product had an M ion 58 amu higher than the original compound, and product ions due to loss of a methoxy radical (M-31), dimethylether (M-46), or a methoxy radical and methanol (M-31-32), together indicating the presence of two carbonyl groups (Fig. 6.6B). The diketo structure was confirmed by a McLafferty-fragment (m/z 158) accompanied by a product ion resulting from loss of methoxy (m/z 127), thus firmly establishing that D was a C33 diketone. This finding, taken together with the size of various -fragments and McLafferty rearrangement products, identified D as a subterminal -diketone, tritriacontane-2,4-dione 6.5a. Finally, this structure was confirmed by reduction with LiAlH4, leading to a new compound identified as (TMS-derivatized) tritriacontane-2,4-diol (see below also for E).   182   Figure 6.6. Structure elucidation of 2,4-diketone in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of tritriacontan-2,4-dione 6.5a. (B) Mass spectrum and major fragmentations of O-methylhydroxylamine derivative of tritriacontan-2,4-dione 6.5a.   183  Compound class E consisted of seven compounds recognized as a homologous series based on their equidistant GC separation and common MS fragmentation patterns. All compounds E formed TMS derivatives with similar MS characteristics, exhibiting a fragment m/z 73 [(CH3)3Si]+ but no m/z 147 [(CH3)2SiOSi(CH3)3]+ (Fig. 6.7A),  thus indicating the presence of only one hydroxyl group in the native compounds. All homologs also had a noticeable fragment m/z 130 diagnostic for β-ketols (Wen and Jetter, 2009), and an -fragment with TMS transfer (m/z 115) suggesting a 2-keto function. An α-fragment base peak m/z 159, common to all homologs, accompanied by product ions due to loss of CH4 (m/z 143) and CH2CO (m/z 117), indicated the presence of a 4-hydroxy-2-keto structure. Conversely, longer α-fragments were found to vary with homolog chain length (m/z 509 for the homolog in Fig. 6.7A), in parallel with respective M-15 ions indicative of molecular weight and thus total chain length. Given all the evidence summarized thus far, class E was tentatively identified as a homologous series of C25 to C37 4-hydroxy-2-ketones 6.6 (i.e., subterminal -ketols). A summary of all identified homologs and their diagnostic MS fragments is presented in Table 6.1.  To further test the structure assignment, compounds E were transformed into acetates. The resulting derivatives showed pairs of α-fragments of relatively low intensity together with product ions resulting from loss of acetic acid, one set of them homolog-independent (m/z 129 and m/z 69) and the other one varying with homolog chain length (m/z 419 and m/z 461 for the C33 homolog 6.6e in Fig. 6.7B). Other fragments indicative of chain length were due to loss of water, an acetyl radical, and/or acetic acid from the molecular ion (M-18, M-43, M-60, and M-  184 60-18, respectively). The acetate spectra of compounds E thus confirmed the subterminal -ketol  structures (2,4-ketols) 6.6.  To specifically test the existence of a carbonyl functionality, a third aliquot of the fraction was derivatized with O-methylhydroxylamine and then silylated. The mass spectrum of the product (Fig. 6.7C) showed a homolog-independent α-fragment (m/z 188) and its product ions due to loss of CH4 and CH2O (m/z 172 and 158, respectively), with the first two ions shifted 29 amu higher than for the same compound without oximation, and thus confirming the presence of one carbonyl group. Other diagnostic fragments were due to loss of CH3, CH3O, and/or (CH3)3SiOH from the molecular ion (M-15, M-31, M-90 and M-31-90, respectively), and the base peak corresponded to the second α-fragment of the OTMS functionality (m/z 509 for the C33 homolog 6.6e in Fig. 6.7C).  Lastly, reduction of compounds E with LiAlH4 followed by silylation resulted in TMS-derivatized 2,4-diols, in  the case of the C33 homolog 6.6e identical to that produced upon reduction of compound D 6.5a. Their structure was confirmed by signature MS fragments m/z 73 [(CH3)3Si]+ indicative of at least one hydroxyl, an ion m/z 147 [(CH3)2SiOSi(CH3)3]+ suggesting a diol, and pairs of α-fragments indicating hydroxyl groups on C-2 (m/z 117 and, for the  C33 homolog 6.6e, m/z 625) and on C-4 (m/z 233 and, for the  C33 homolog 6.6e,  m/z 509) (Fig. 6.7D). Fragments due to loss of CH3 and/or (CH3)3SiOH from the molecular ion (M-15, M-90 and M-15-90) further indicated the molecular weight and therefore chain length of the diol homologs. Taken together, the TLC behaviour of the native compounds and the GC-MS data for TMS derivatives, acetate esters, methoxime/TMS derivatives, and (TMS-derivatized) LiAlH4   185 reduction products unambiguously established that E was a homologous series of odd-numbered 4-hydroxy-2-ketones 6.6, spanning chain lengths from C25 to C37. Relative GC-MS peak areas showed that this series of subterminal ketols was dominated by the C33 homolog 6.6e.      Figure 6.7. Structure elucidation of 2,4-ketols in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of TMS derivative of C33 2,4-ketol 6.6e. (B) Mass spectrum and major fragmentations of Ac derivative of C33 2,4-ketol 6.6e. (C) Mass spectrum and major fragmentations of O-methylhydroxylamine/TMS derivative of C33 2,4-ketol 6.6e. (D) Mass spectrum and major fragmentations of TMS derivative of the diol obtained via LiAlH4 reduction of C33 2,4-ketol 6.6e.   186  Table 6.1. Characteristic m/z fragments of trimethylsilyl ethers of 2,4-ketols in flag leaf and peduncle waxes of Triticum aestivum cv. Bethlehem Compound Fragments characteristic of homolog   4-Hydroxypentacosan-2-one 397                                                                             439   4-Hydroxyheptacosan-2-one 425                                                                             467   4-Hydroxynonacosan-2-one 453                                                                             495   4-Hydroxyhentriacontan-2-one 481                                                                             523   4-Hydroxytritriacontan-2-one 509                                                                             551   4-Hydroxypentatriacontan-2-one 537                                                                             579   4-Hydroxyheptatriacontan-2-one 565                                                                             607   6.3.3. Structure elucidation of compound class F  Compound class F was detected as a single peak in the GC trace from a TLC fraction running between β-diketones and alkyl esters, and thus of only moderate polarity. The mass spectrum of F did not show any of the fragments characteristic of OH groups after treatment with BSTFA (such as m/z 73, 75), but instead exhibited α-fragments characteristic of ketones (Fig. 6.8A). Specifically, the set of α-fragments m/z 211 and m/z 267 indicated one ketone isomer, while the corresponding ion m/z 239 suggested a co-eluting isomer with keto group position shifted by two carbons. This interpretation was corroborated by McLafferty+1 fragments 16 amu higher than the -fragments (m/z 227 and 283; m/z 255). The molecular ion (m/z 450) indicated a total ketone chain length of C31.     187 To further probe the presence and location of the carbonyl functionality, methoxime derivatives were prepared from F. The resulting compounds showed diagnostic methoxime fragments (m/z 87 and m/z 100) (Jetter and Riederer, 1999a), α-fragments 29 amu higher than the original ketone isomers (m/z 240 and m/z 296; m/z 268), and corresponding McLafferty rearrangement fragments as well as product ions resulting from loss of methanol (Fig. 6.8B). The oxime spectrum thus confirmed the presence of two isomeric ketones. The chain length, C31, was corroborated by the molecular ion and its product fragment M-31 (due to loss of a methoxy unit).   Finally, the ketone mixture was reduced with excess LiAlH4 and the resulting isomeric secondary alcohols subjected to silylation. Their mass spectra showed the alcohol-characteristic fragment m/z 73, as well as α-fragments indicating the location of the hydroxyl function on C-14 (m/z 285 and m/z 341) or on C-16 (m/z 313) (Fig. 6.8C). Fragments due to loss of CH3 or (CH3)3SiOH from the molecular ion (M-15 and M-90), confirmed the C31 chain length. All taken together, our TLC and MS data thus unambiguously identified compound F as a mixture of mid-chain ketones, hentriacontan-14-one 6.7a and hentriacontan-16-one 6.7b.   188  Figure 6.8. Structure elucidation of ketones in wheat leaf and peduncle wax. (A) Mass spectrum and major fragmentations of co-eluting ketone isomers hentriacontan-14-one 6.7a and hentriacontan-16-one 6.7b. (B) Mass spectrum and major fragmentations of co-eluting O-methylhydroxylamine derivatives of hentriacontan-14-one 6.7a and hentriacontan-16-one 6.7b. (C) Mass spectrum and major fragmentations of co-eluting TMS derivatives of the LiAlH4 reduction products of hentriacontan-14-one 6.7a and hentriacontan-16-one 6.7b.   189 6.4. Discussion  In this work we discovered several new classes of compounds in the cuticular waxes covering wheat flag leaf blades and peduncles, together with new homologs and isomers of the predominant class of -diketones. Among the newly identified wheat wax constituents, only the ketones were found to have a single in-chain functionality, whereas all other new wheat wax structures featured two functional groups varying 1) in their relative configuration, 2) in their positions within the hydrocarbon chain, and 3) in their oxidation states.  Firstly, regarding the relative configuration of functionalities, the mid-chain -ketols (A), subterminal -diketone (2,4-diketone) (D) and subterminal ketols (2,4-ketols) (E) featured two functional groups on carbons separated by one methylene group and thus shared the relative group constellation with the previously identified -diketones. In contrast, the -diketones (B) and -ketols (C) had two functional groups on directly adjacent carbons. Secondly, the mid-chain -ketols (C), -diketones (B), -ketols (A) and ketones (F) shared very similar functional group positions within the hydrocarbon chains with the mid-chain -diketones, having predominantly C13H27 and C15H31 alkyl chains on either side of the functionalities. Again, this characteristic contrasted with some other wheat wax constituents, where the subterminal 2,4-diketone (D) and 2,4-ketols (E) had functional groups near one chain terminus, with one alkyl residue constant as CH3 and the other one widely varying around C29H59 (from C21H43 to C33H67). Thirdly, all the bifunctional compounds could be divided into two distinct classes also according to functional group oxidation states, either with two carbonyls (B and D, hence designated as diketones) or with one hydroxyl and one carbonyl function (A, C and E, hence designated as ketols). In summary, each of the novel wheat wax compound classes shared some   190 but not all structural features with several others, suggesting that all the compound classes are biosynthetically related. Based on their structural commonalities and differences, potential biosynthetic pathways leading to them can now be outlined.  Several of the new wheat wax compound classes had polyketide-like structures, suggesting that they are biosynthesized through condensation reactions with malonyl units catalyzed by polyketide synthase (PKS) enzymes. In particular, the subterminal 2,4)-diketone (D) may be derived from C30 fatty acyl-CoA, occurring as an intermediate of elongation and modification towards normal wheat wax compounds such as C30 alcohol and C29 alkane (Fig. 6.9). It seems plausible that this acyl-CoA serves as substrate for a PKS catalyzing two consecutive condensation reactions with malonyl-CoA extenders, and that the resulting C34 triketide (3,5-diketoacyl-CoA) may be hydrolyzed and decarboxylated to the 2,4-diketone product. It should be noted that the latter two reactions, involving thioester hydrolysis and loss of CO2 from the carboxylate, may both occur spontaneously but could also be enzyme-catalyzed, similar to the formation of methylketones in tomato trichomes (Yu et al., 2010b). It should further be noted that the initial reaction towards the 2,4-diketone is identical with a condensation occurring as part of VLCFA elongation en route to other wax constituents, where a KCS enzyme in a FAE complex is thought to utilize the same C30 fatty acyl-CoA and malonyl-CoA substrates. It can therefore not be excluded that, instead of a PKS catalyzing two condensation rounds, the first of them may be carried out by a KCS and only the second one by a PKS. In this scenario, the ketoacyl-CoA intermediate of elongation would have to be intercepted (either directly or by a thioesterase releasing a free acid that is re-activated to a thioester) and transferred to the PKS for further condensation with a malonyl unit.     191  Based on the structural similarity and matching major homolog chain length, the subterminal 2,4)-ketols also identified in wheat waxes (E) are very likely biosynthetically related to the 2,4-diketone (Fig. 6.9). It seems plausible that either the diketone itself or an intermediate along the pathway leading to it can serve as precursor for ketol formation, implying that a reductase enzyme may specifically transform the 4-keto group into the corresponding hydroxyl. Alternatively, also the 3-hydroxyacyl-CoA intermediate of the FAE complex may be intercepted and used as starter for condensation with malonate, possibly catalyzed by the same PKS as in (then parallel) 2,4-diketone formation. Based on the chain length profile of the wheat 2,4-ketols and 2,4-diketones, both peaking at C33, we conclude that this PKS may be relatively specific for C30 acyl-CoA or C32 ketoacyl-CoA substrate. It is thus distinguished from other PKS enzymes thought to participate in wheat wax formation, such as the PKS producing mid-chain β-diketones from C14 and/or C16 acyl substrates (see below) or the one thought to be involved in alkylresorcinol formation from C24 acyl-CoA and similar precursors (Racovita et al., 2016).  Interestingly, a homologous series of 2,4-diketones with chain lengths ranging from C25 to C31 had been identified in the suberin of Ericaceae roots (van Smeerdijk and Boon, 1987), thus spanning a fairly wide range of homologs but not including the C33 diketone identified in wheat wax. In contrast, 2,4-ketols like those described here had not been reported before. However, closely related isomers had been identified as 4,6-ketols (4-hydroxyalkan-6-ones and 6-hydroxyalkan-4-ones) in lipid mixtures from sunflower pollen, albeit with shorter chain lengths ranging from C19 to C27 (S Schulz et al., 2000).     192 2,4-Ketols2,4-KetolsC27H55 COS-CoAC27H55 COCH2 COS-CoAC27H55 CHOHCH2 COS-CoAC27H55 CH CH COS-CoAC29H59 COS-CoAC29H59 COCH2 COS-CoAC29H59 CHOHCH2 COS-CoAC29H59 CH CH COS-CoAC31H63 COS-CoA2,4-KetolseeC31H63 COCH2 COS-CoAC31H63 CHOHCH2 COS-CoAeC31H63 CH CH COS-CoAeeeC27H55 COCH2 COCH2 COS-CoAC27H55 CHOHCH2 COCH2 COS-CoAC29H59 COCH2 COCH2 COS-CoAC29H59 CHOHCH2 COCH2 COS-CoAC31H63 COCH2 COCH2 COS-CoAC31H63 CHOHCH2 COCH2 COS-CoAC27H55 COCH2 COCH2 COO-C27H55 CHOHCH2 COCH2 COO-C29H59 COCH2 COCH2 COO-C29H59 CHOHCH2 COCH2 COO-C31H63 COCH2 COCH2 COO-C27H55 COCH2 COCH3C27H55 CHOHCH2 COCH3C29H59 COCH2 COCH3C29H59 CHOHCH2 COCH3C31H63 COCH2 COCH3C31H63 CHOHCH2 COCH3C31H63 CHOHCH2 COCH2 COO-a/eb/ha/eb/ha/eb/hhhhffffffgggggghhhhhhcd2,4-Diketonecdcd Figure 6.9. Proposed biosynthetic pathway to subterminal -diketone (2,4-diketone) and subterminal -ketols (2,4-ketols). Normal wax biosynthesis proceeds via elongation of C28 to C30 acyl-CoA, and in repeat reaction cycles to C32 acyl-CoA and beyond, catalyzed by four enzymes (a: KCS; b: KCR; c: HCD; d: ECR) in the FAE complex (left). The -keto intermediates may be intercepted and used as substrate(s) for a condensation reaction (center) catalyzed by a PKS enzyme (e). The resulting triketide may undergo (enzymatic or spontaneous) hydrolysis (f) and decarboxylation (g) to a 2,4-diketone three carbons longer than the original acyl substrate (right). Alternatively, the PKS may also catalyze the formation of the -ketoacyl-CoA intermediate. Although such reaction sequences leading to C31 or C35 2,4-diketones are feasible, only the C33 2,4-diketone (shaded box) was identified in wheat wax. Corresponding reaction sequences starting with -hydroxyacyl-CoA intermediates may lead to 2,4-ketols of various chain lengths (shaded boxes). Alternatively, either the 2,4-diketone(s) or their precursors may be enzymatically reduced (h) to the corresponding -hydroxy intermediates.   193 The wheat mid-chain β-ketols strongly resembled the β-diketone dominating the wax mixtures, hentriacontane-14,16-dione, both in their chain length and functional group positions. Therefore, it may be surmised that the mid-chain β-ketols are either intermediates on the β-diketone biosynthesis pathway or side products of it. It has long been recognized that β-diketones are polyketide in nature, and it was hypothesized that their formation involves two PKS-catalyzed condensation steps (von Wettstein-Knowles, 2012). However, very recently Hen Avivi et al. (2016) showed that the synthesis of wheat mid-chain β-diketone instead proceeds via hydrolytic intercept of a thioester intermediate of fatty acid synthesis, likely C16 -ketoacyl-ACP, and only one PKS-catalyzed condensation reaction. By analogy, it may now be speculated that mid-chain β-ketols are formed by intercept of C16 -hydroxyacyl-ACP, another intermediate of fatty acid elongation, followed by a PKS-mediated condensation (Fig. 6.10). Either one or both of the reactions involved, the thioester hydrolysis and the PKS reaction, may be catalyzed by the same enzyme(s) as the corresponding reactions in β-diketone synthesis. Alternatively, the wheat mid-chain β-ketols might also be formed by reduction of either the β-diketone or one of the intermediates along the pathway leading to it.  Mid-chain β-ketols with structures very similar to those observed in wheat wax had previously been reported. Namely, C29 β-ketols were reported together with -ketols, secondary alcohols and ketones in Arabidopsis thaliana stem wax (Wen and Jetter, 2009) and in the leaf wax from four Brassica species (Holloway and Brown, 1977). It was then also established that the Arabidopsis -ketols and the accompanying compounds with secondary functionalities are all products of a cytochrome P450-dependent enzyme, mid-chain alkane hydroxylase (MAH1), which hydroxylates alkane substrates on C-13, C-14 and C-15. The somewhat variable group   194 positions in the Brassicaceae ketols hence contrast with the wheat mid-chain β-ketols, where functional groups were exclusively located on C-14 and C-16. Thus, the differences in isomer profiles reflect the very different biosynthetic origins of ketols in both cases, in Arabidopsis involving cytochrome P450 oxidation of alkanes and in wheat intercept of fatty acid synthesis intermediates and PKS reactions.  The formation of wheat wax mid-chain ketones may also be assessed in light of the biosynthetic pathways likely leading to the similar diketone and ketol compounds accompanying them. Due to common features in the overall molecular structures, most prominently the predominance of C13H27 and C15H31 alkyl chains in various homologs and isomers, it seems likely that the wheat ketones are formed as side products of the major wax polyketides (Fig. 6.10). The ketones can thus be viewed as derivatives of the mid-chain diketones, with one carbonyl group instead of two, suggesting that one of the reactions generating the diketo functionalities may be skipped. Accordingly, the -ketoacyl intermediate of -diketone formation might, instead of the PKS-catalyzed condensation leading to the mid-chain -diketone, be elongated by FAE systems to give ketone products. Alternatively, the β-keto thioester hydrolysis step of -diketone formation could be omitted, and a PKS-mediated reaction utilizing a fatty acyl instead of a -ketoacyl would lead to the ketones. Again, the potential biosynthesis pathways described thus as variants of -diketone formation differ from those leading to Brassicaceae wax ketones via MAH1-mediated hydroxylation (Greer et al., 2007).   Interestingly, the Brassicaceae ketones have their carbonyl group on odd- and even-numbered carbons, for example nonacosan-15-one in Arabidopsis wax, while the wheat ketones   195 were characterized by a functionality exclusively on even-numbered carbons. The parity of the keto group position can hence serve as a distinguishing feature for the ketone formation pathways, involving either hydroxylation of alkanes mediated by cytochrome P450 enzymes similar to MAH1 or polyketide-forming enzymes. Based on this conclusion, it may be speculated that other ketones with carbonyl groups exclusively on even-numbered carbons may also be formed by polyketide pathways. Such ketones are prominent wax constituents in several species, for example nonacosan-10-one in the waxes of Osmunda regalis (Jetter and Riederer, 2000) and several apple cultivars (Dodova-Anghelova and Ivanov, 1973; Dong et al., 2012), and hentriacontan-16-one in waxes of sandal (Chibnall et al., 1937), leek (Rhee et al., 1998), Annona senegalensis (MacKie and Misra, 1956) and beech (Gülz et al., 1989).   Finally, mid-chain α-ketols (acyloins) had also been reported before, often together with β-ketols and ketones sharing the same acyl groups, for example in the cuticular waxes of Arabidopsis thaliana (Wen and Jetter, 2009), four Brassica species (Holloway and Brown, 1977), or the fern Osmunda regalis (Jetter and Riederer, 2000). However, in all previous instances predominantly odd-numbered homologs of α-ketols were identified (especially C29), similar to the alkane accompanying them, and thus suggesting a biosynthetic relationship. Accordingly, an Arabidopsis enzyme, MAH1, was found to catalyze the oxidation of alkane precursors, via secondary alcohols and ketones, to α- and -diols as well as α- and -ketols (Greer et al., 2007; Wen and Jetter, 2009). In stark contrast, the wheat α-ketols had even carbon numbers (C30), hence making formation from alkane precursors unlikely. Instead, the alkyl moieties involved, C13H27 and C15H31, may suggest a head-to-head condensation of C14 and C16 acyl precursors via an unknown mechanism involving rarely observed C-C bond formation   196 between two carbonyl carbons. Interestingly, the wheat wax α-ketols were found as mixtures of isomers with reversed positions of the hydroxyl and carbonyl groups (similar to the Brassica ketols). Based on this observation, it may be speculated that the mid-chain α-ketols are formed by fusion of C14 and C16 n-aldehydes as novel wax biosynthesis precursors. Alternatively, reductive coupling of C14 and C16 acyl-CoAs or a combination of acyl-CoA and aldehyde substrates also seem feasible. Finally, a similar reaction, albeit with different redox requirements, may afford the C30 α-diketone as an intermediate en route to the ketols (Fig. 6.10).              197 C13H27 COS-ACPC13H27 COCH2 COS-ACPC13H27 CHOHCH2 COS-ACPC13H27 CH CH COS-ACPC15H31 COS-ACPC13H27 COHC13H27 CHOHCH2 COO-C13H27 CHOHCHOHC15H31C15H31 COCH2 COS-ACPC15H31 CHOHCH2 COS-ACPC15H31 CH CH COS-ACPC17H35 COS-ACPC15H31 COHC15H31 CHOHCH2 COO-C13H27 COC17H35C15H31 COC15H31C13H27 CHOHCH2 CHOHC13H27C13H27 CHOHCH2 CHOHC17H35C15H31 CHOHCH2 CHOHC15H31C13H27 COO-C15H31 COO-C13H27 CHOHCH2 CHOHC15H31Ketones-Ketols and -Diketones-Ketols and -DiketonesopoprrqqmmnnC13H27 COC17H35C15H31 COC15H31Ketonesi/mjkli/mjklss Figure 6.10. Proposed biosynthetic pathways to mid-chain ketones, α- and -diketones, and α- and -ketols. Fatty acid de novo synthesis proceeds via elongation of C14 to C16 acyl-ACP, and in a further reaction cycle to C18 acyl-ACP, catalyzed by four enzymes (i: KAS; j: KAR; k: HAD; l: EAR) in the FAS complex (center). It was recently shown (Hen-Avivi et al., 2016) that mid-chain -diketones are formed by hydrolytic intercept of -ketoacyl-ACP intermediates of elongation (m) followed by PKS-mediated condensation (n). We propose that corresponding mid-chain -ketols may be formed either by analogous intercept of -hydroxyacyl-ACP intermediates of elongation, or else by reduction of the -diketone or its precursor(s). Similarly, hydrolytic intercept of acyl-ACPs (o) or their reduction to aldehydes (p) may provide the substrate for head-to-head condensation (q) leading to -diketones and -ketols, respectively. Finally, either PKS-mediated condensation (r) or elongation of the -keto intermediate (s) may lead to the ketones identified in wheat waxes.    198 Chapter 7: Conclusions and future research directions   7.1. Concluding remarks  The present thesis has the main merit of expanding the current knowledge about plant cuticular wax chemical diversity. Numerous new structures have been identified in the three chosen species and elucidated using primarily gas chromatography-mass spectrometry in combination with chemical derivatization, sometimes supplemented by synthesis of authentic standards. In addition, comprehensive wax compositional studies for three new plant species have been performed and are presented in the thesis. For two of the three species, the partitioning of wax compounds between the intracuticular and epicuticular compartments of the cuticle has also been studied, thus supplementing previous studies on wax partitioning using other plant species. Insights into the biosynthesis of novel waxes are also provided and often challenge previously well-established concepts, such as that FAE intermediates are not released by the FAE complex or intercepted by other enzymes, that aldehydes are not released as intermediates en route to primary alcohols, or that nitrogen-containing fatty acid derivatives are not encountered as cuticular waxes.  Summaries of the most important findings from each thesis chapter are presented below:  In chapter 2, the chemical composition of Aloe arborescens leaf cuticular wax was characterized both qualitatively and quantitatively. The total wax load on the abaxial side was   199 more than double that on the adaxial side, and the epicuticular waxes constituted approximately 50% of the wax load on the abaxial side and 60% on the adaxial side, respectively. The waxes were found to contain VLC compound classes very similar to other species, with distinct gradients between the intracuticular and epicuticular layers pointing to polarity-driven partitioning. The aliphatic wax constituents had very characteristic chain length distributions, highlighting the distinct specificities of the two modifying pathways involved in wax biosynthesis: the acyl reduction pathway and the alkane-forming pathway. Three homologous series of unique 1,3-bifunctional and 2-monofunctional compound classes were identified as very-long-chain 3-hydroxy fatty acids,  3-hydroxy FAMEs and 2-alkanols. The co-occurrence of these compound classes with very similar functional groups, together with characteristic chain length distributions, suggested that they are biosynthetically related. While the pathways and enzymes involved in their formation can only be speculated based on our chemical results, they point to an important new possibility: VLC fatty acyl elongation intermediates bearing a 3-hydroxyl group may be intercepted by other enzymes successfully competing for substrate with the β-hydroxyacyl-CoA dehydratase (HCD) component of the FAE complex. One such enzyme may be a thioesterase forming free acids, and another one the reductase yielding aldehydes.    In the work presented in chapter 3, the adaxial side of young and old P. aurea bamboo leaves was studied and established to be covered with about 3.5 μg/cm2 of total (epicuticular and intracuticular) wax, which is a relatively low wax load fairly close to that of A. thaliana leaves (~1 μg/cm2). Epicuticular waxes represented slightly more than 50% of total wax for both plant ages. Along with ubiquitous VLC aliphatic wax classes (fatty acids, alcohols, alkyl esters, aldehydes, alkanes), substantial amounts of terpenoids were detected, which accounted for 30-  200 40% of the epicuticular waxes and ~50% of the intracuticular waxes, depending on plant age. In general, pronounced gradients were observed between the relative abundances of all compound classes in the epicuticular and intracuticular compartments, commensurate with their relative polarities. Most notably, fatty acid amides were identified here as a novel wax compound class and found to partition exclusively into the epicuticular wax layer, where they might have a role in plant-insect and plant-pathogen interactions, or in preventing organ fusions during ontogenesis, similar to artificial fatty amide slip agents added to polyolefine plastics to prevent sheet adhesion. The discovery of fatty acid amides also raises questions regarding the biosynthetic mechanism(s) incorporating nitrogen into VLC acyl compounds. Underlying processes can only be hypothesized at this stage, but most likely involve integration of the wax biosynthetic and aminoacid metabolic pathways.  The work presented in chapter 4 is a comprehensive compositional comparison of cuticular wax mixtures from two organs (flag leaf blade and peduncle) of Triticum aestivum cv. Bethlehem. The total wax coverage of peduncles was three times larger than that of flag leaf blades, and peduncle wax was dominated by C31 β-diketone and 8- and 9-hydroxy-β-diketones, while flag leaf wax contained more than 50% 1-alkanols. Accordingly, the (currently unknown) biosynthetic pathway to the β-diketones and the (at least partially parallel) acyl reduction pathway to wax 1-alkanols are differentially regulated in peduncles and in flag leaves. Interestingly, the chain length profile of aldehydes closely matched that of 1-alkanols on flag leaf blades, suggesting that aldehydes may be released as intermediates of the reduction leading to alcohols, in contrast to our current understanding of wax biosynthesis in Arabidopsis. However, on peduncles, the aldehyde profile matched that of alkanes, in line with those of Arabidopsis.   201 Diverse classes of esters were identified in wheat cv. Bethlehem waxes as complex mixtures of metamers, necessitating detailed quantitative analysis and suggesting the presence of diverse ester synthase enzyme with distinct substrate specificities.  In the work presented in chapter 5, seven new compound classes were identified and quantified in the wax mixtures covering flag leaf blades and peduncles of T. aestivum cv. Bethlehem. Three of them were secondary alcohols, primary/secondary diols and their esters with (very-) long chain acids, all of which were found as homologous series only in flag leaf wax. These three classes appeared biosynthetically related based on their similar secondary hydroxyl groups (on and around C-12 from the non-functionalized chain end). The hydroxy-2-alkanol esters and oxo-2-alkanol esters found in both organs are biosynthetically related to hydroxy-β-diketones, as suggested also by a common location of hydroxyl/oxo groups. We hypothesize that both compound groups, the secondary alcohols/diols and the oxidized 2-alkanol esters/β-diketones, are formed by two distinct P450 enzymes with C-12 and C-7/8 regio-specificity, respectively. In contrast, the two other compound classes identified here, 4-alkylbutan-4-olides and internally methyl-branched may be formed via -oxidation and through incorporation of methylmalonyl-CoA into fatty acyl-CoA intermediates, respectively. Overall, we thus propose three specific variations from the normal wax biosynthesis pathways to occur in wheat, in the form of P450 oxidation, -oxidation and methylmalonate incorporation.  In chapter 6, six new classes of cuticular wax compounds were identified in TLC fractions of wax mixtures coating flag leaf blades and peduncles of Triticum aestivum cv. Bethlehem. They included a homologous series of C25 - C37 2,4-ketols and C33 2,4-diketone, whose common   202 structures suggested polyketide-type biosynthesis. Based on the predominance of the C33 homologs in both compound classes, we conclude that the PKS likely involved in the formation of these subterminal ketols and diketones must have a preference for C30 acyl-CoA or C32 ketoacyl-CoA substrate, thus distinguishing it from other PKSs thought to be involved in the biosynthesis of other wheat wax constituents. Two such PKSs, forming mid-chain β-diketones and alkylresorcinols, have been recognized before. We further identified regiomers of a C31 mid-chain ketone and a C31 mid-chain β-ketol, which are likely side products of the β-diketone biosynthetic pathway. It seems plausible that such minor products may be formed when either a step of the β-diketone biosynthetic pathway is skipped, leading to a ketone instead, or when a reduction step is added, leading to the corresponding ketol. Finally, C30 14,15-diketone and a pair of corresponding C30 mid-chain -ketol regiomers were identified in wheat wax, clearly differing from previous literature reports of similar acyloins with odd carbon numbers. We hypothesize that the wheat -ketols are formed by head-to-head condensation of C14 and C16 acyl precursors, in the form of thioesters or aldehydes, and not by hydroxylation of alkanes like the acyloins found in Brassicaceae waxes.  Overall, several of the wheat wax compounds identified in chapter 6 have very similar mid-chain-functionalized structures, characterized by C13H27 and C15H31 alkyl chains on a group of central carbons carrying functional groups. Accordingly, the potential biosynthetic mechanisms generating all these compounds also share a common characteristic, in that they branch off the normal wax biosynthesis pathway in a relatively early stage, through interception or diversion of C14 and C16 elongation intermediates and/or their utilization by PKS enzymes. It may be inferred that either one or more PKS enzymes in wheat must be able to handle such long-chain (C14 and C16) acyl substrates. In contrast, the subterminal -diketone (2,4-diketone) and   203 subterminal -ketols (2,4-ketols) found in wheat wax are likely formed by a different PKS with preference for very-long-chain (C30) substrates. Another PKS is expected to be involved in the formation of alkylresorcinols from acyl substrates such as C24, catalyzing three consecutive condensation reactions with malonate plus a final cyclization. This information, together with the evidence provided here, now leads us to distinguish at least three different PKSs participating in wheat wax biosynthesis, with very different biochemical properties including preference for C14/16, C24 and C30 substrates.  In conclusion, the main objective of the thesis was achieved in that approximately 20 new wax compound classes were identified across the three plant species studied. Furthermore, valuable insight into cuticular wax biosynthesis was gained from exploring the potential biosynthetic origins of these compounds. Thus:  It was suggested here that, unlike in A. thaliana, in other species (such as A. arborescens and T. aestivum cv. Bethlehem) FAE and/or FAS elongation intermediates (namely β-keto and β-hydroxyacyl thioesters) can be intercepted by other enzymes (hydrolases, reductases, PKSs) to produce interesting new wax structures with terminal (1,3-), subterminal (2,4-), or mid-chain bifunctionalities.   A pathway to primary amides was proposed to operate in P. aurea, primary amides being the first example of nitrogen-containing fatty acid derivatives in cuticular waxes.  In the flag leaves of T. aestivum cv. Bethlehem, the correlation between chain length profiles of primary alcohols and aldehydes suggested the latter to be intermediates en   204 route to the former released by the fatty acyl reductase, again unlike in the Arabidopsis model.  It was found that wax ester synthases exhibit widely variable substrate specificity (for either their acyl-CoA or their alcohol substrate) depending on the plant species and even on the organ of the species.  Products of cytochrome P450 enzymes different from the Arabidopsis mid-chain alkane hydroxylase were identified in wheat, with hydroxyls at positions C-8/9 or ω-12.  A great variety of polyketides and polyketide-like structures were identified also in wheat, including midchain and subterminal β-diketones and ketols, as well as simple ketones, suggesting the great versatility of the PKS enzymes responsible for their biosynthesis.       7.2. Future research directions  7.2.1. Search for further novel wax compounds in other plant species   The present thesis decribes the identification of numerous new wax compounds, however its findings are limited by the fact that only three plant species were explored. By expanding the repertoire of plant species studied, further novel wax structures are expected to be discovered. It would be of particular interest to continue the exploration of monocotyledons, since they have proven to be so rich in new compounds in the work herein. This exploration could be targeted at entirely new families and genera, but it could equally aim at studying new species from the same   205 families and genera as those in the present work. For example, other species in the genus Aloe may reveal further bifunctional compounds with 1,3-constellation of functionalities, such as diols, ketols, hydroxy-aldehydes, keto-aldehydes, hydroxy-esters or keto-esters. Other cereal species whose cuticular waxes have been less extensively studied (sorghum, millet, oats, triticale, fonio, teff) may be a source of further polyketides or oxygenated compounds. Finally, other species in the genus Phyllostachys may harbour a wider range of cuticular wax compounds containing nitrogen, for example amines resulting from reduction of amides or secondary amides formed via acylation of amines similar to wax ester formation.  7.2.2. Investigations of some of the proposed biosynthetic pathways        Numerous biosynthetic routes have been proposed towards the novel compounds described in this thesis, but they are all hypothetical at this stage. It would thus be of interest to probe at least some of these pathways, using molecular biological and/or biochemical tools. The plant species most suitable for this kind of work is Triticum aestivum, whose genome has been fully sequenced (International Wheat Genome Sequencing Consortium, 2014). In particular, the genes encoding the hydrolase and polyketide synthase (PKS) responsible for the production of β-diketones and the gene encoding the cytochrome P450 enzyme responsible for the hydroxylation of β-diketones were established (Hen-Avivi et al., 2016). It would be of interest to check if the minor polyketide metabolites structurally related to β-diketones (i.e. β-ketols, ketones, α-diketones, α-ketols) are still present or not in plants with mutations of either or both of the hydrolase and PKS, thus proving or disproving the participation of these enzymes in the production of the aforementioned minor wax constituents. Also, in a P450-knock-out mutant   206 line, it would be interesting to search for the presence of hydroxy- and oxo-2-alkanol esters, in order to establish whether the same cytochrome P450 enzyme forming hydroxy-β-diketones is forming these oxidized 2-alkanol esters as well. Biochemically, it would be interesting to purify this enzyme and feed it 2-alkanol substrates to directly observe the hydroxylation process of these substrates, as well as β-diketone substrates. This would eliminate the alternative possibility that the hydroxylation happens earlier in the pathway, on one of the common intermediates of β-diketones and 2-alkanol esters, such as the β-ketoacids produced by the hydrolase.  Of further interest would be to identify in the wheat genome the second cytochrome P450-encoding gene responsible for the production of secondary alcohols and diols in wheat flag leaves. This gene would have to be highly expressed in the flag leaf blades, but with negligible expression level in the peduncle, a feature that could be used to select it from among multiple gene candidates. If a mutation in this gene results in loss of secondary alcohols and/or diols, its role in their biosynthesis would thus be confirmed. For the diol monoesters, the homolog of Arabidopsis WSD1 would have to be identified in the wheat genome. A mutation in this gene would result in loss of esters of primary alcohols in addition to diol esters, making it unclear whether diol esters are absent due to knock-out of the gene directly responsible for their production or due to lack of ester substrates for hydroxylation by the cytochrome P450. It would be more useful to isolate and purify the wheat wax ester synthase and feed it in vitro separately primary alcohol and primary/secondary diol substrates (each in combination with acyl-CoAs) and thus test its versatility in esterifying both types of substrates with primary OH.   While the genome of Phyllostachys aurea has not been sequenced, it would nonetheless be of great interest to identify, clone and characterize the gene encoding the cytochrome c putatively responsible for the biosynthesis of VLC primary amides. A mutation in this gene   207 resulting in loss of amides would clarify the participation of cytochrome c in amide biosynthesis. If the cytochrome c enzyme is isolated and purified, a feeding experiment with ammonia and acyl-CoA will confirm whether or not this simplest pathway is indeed the pathway leading to primary amides in P. aurea.     208 References   Aasen, A.J., Hofstetter, H.H., Iyengar, B.T.R., Holman, R.T., 1971. Identification and analysis of wax esters by mass spectrometry. Lipids 6, 502–507. Acevedo, L., Martínez, E., Castaneda, P., Franzblau, S., Timmermann, B.N., Linares, E., Bye, R., Mata, R., 2000. New phenylethanoids from Buddleja cordata subsp. cordata. Planta Med. 66, 257–261. Adamski, N.M., Bush, M.S., Simmonds, J., Turner, A.S., Mugford, S.G., Jones, A., Findlay, K., Pedentchouk, N., von Wettstein-Knowles, P., Uauy, C., 2013. The inhibitor of wax 1 locus (Iw1) prevents formation of β- and OH-β-diketones in wheat cuticular waxes and maps to a sub-cM interval on chromosome arm 2BS. Plant J. 74, 989–1002. Alfatafta, A.A., Abu Zarga, M.H., Sabri, S.S., Freyer, A.J., Shamma, M., 1989. An investigation of Bongardia chrysogonum. J. Nat. Prod. 52, 818–821. Ali, M., Siddiqui, N.A., Ramachandram, R., 2001. Phytochemical investigation of aerial parts of Pluchea lanceolata C.B. Clarke. Indian J. Chem. 40B, 698–706. Baerson, S.R., Schröder, J., Cook, D., Rimando, A.M., Pan, Z., Dayan, F.E., Noonan, B.P., Duke, S.O., 2010. Alkylresorcinol biosynthesis in plants: new insights from an ancient enzyme family? Plant Signal. Behav. 5, 1286–1289. Balsdon, J.A., Espelie, K.E., Bramans, S.K., 1995. Epicuticular lipids from Azalea (Rhododendron spp.) and their potential role in host plant acceptance by Azalea lace bug, Stephanitis pyrioides (Heteroptera: Tingidae). Biochem. Syst. Ecol. 23, 477–485. Bargel, H., Koch, K., Cerman, Z., Neinhuis, C., 2006. Evans Review No. 3 : Structure–function relationships of the plant cuticle and cuticular waxes — a smart material? Funct. Plant Biol. 33, 893. Barthlott, W., Neinhuis, C., Schott, C.L., 1997. Purity of the sacred lotus, or escape from contamination in biological surfaces. Planta 202, 1–8. Basas-Jaumandreu, J., López, J., de las Heras, F.X.C., 2014. Resorcinol and m-guaiacol alkylated derivatives and asymmetrical secondary alcohols in the leaves from Tamarix canariensis. Phytochem. Lett. 10, 240–248.   209 Bastian, P., Fal, A.M., Jambor, J., Michalak, A., Noster, B., Sievers, H., Steuber, A., Walas-Marcinek, N., 2013. Candelabra aloe (Aloe arborescens) in the therapy and prophylaxis of upper respiratory tract infections: traditional use and recent research results. Wiener Medizinische Wochenschrift 163, 73–9. Beaudoin, F., Gable, K., Sayanova, O., Dunn, T., Napier, J. a, 2002. A Saccharomyces cerevisiae gene required for heterologous fatty acid elongase activity encodes a microsomal beta-keto-reductase. J. Biol. Chem. 277, 11481–11488. Belge, B., Llovera, M., Comabella, E., Gatius, F., Guillen, P., Graell, J., Lara, I., 2014. Characterization of cuticle composition after cold storage of “Celeste” and “Somerset” sweet cherry fruit. J. Agric. Food Chem. 62, 8722–8729. Beppu, H., Kawai, K., Shimpo, K., Chihara, T., Tamai, I., Ida, C., Ueda, M., Kuzuya, H., 2004. Studies on the components of Aloe arborescens from Japan—monthly variation and differences due to part and position of the leaf. Biochem. Syst. Ecol. 32, 783–795. Bernard, A., Domergue, F., Pascal, S., Jetter, R., Renne, C., Faure, J.-D., Haslam, R.P., Napier, J.A., Lessire, R., Joubes, J., 2012. Reconstitution of plant alkane biosynthesis in yeast demonstrates that Arabidopsis ECERIFERUM1 and ECERIFERUM3 are core components of a very-long-chain alkane synthesis complex. Plant Cell 24, 3106–3118. Bhinde, T., Arnold, T., Clarke, S.M., 2010a. The structure of dodecanamide monolayers adsorbed on graphite. Prog. Colloid Polym. Sci. 137, 5–8. Bhinde, T., Clarke, S.M., Phillips, T.K., Arnold, T., Parker, J.E., 2010b. Crystalline structures of alkylamide monolayers adsorbed on the surface of graphite. Langmuir 26, 8201–6. Bianchi, G., Corbellini, M., 1977. Epicuticular wax of Triticum aestivum Demar 4. Phytochemistry 16, 943–945. Bianchi, G., Lupotto, E., Borghi, B., Corbellini, M., 1980. Cuticular wax of wheat: The effects of chromosomal deficiencies on the biosynthesis of wax components. Planta 148, 328–331. Bianchi, G., Murelli, C., Vlahov, G., 1992. Surface waxes from olive fruits. Phytochemistry 31, 3503–3506. Bianchi, G., Vlahov, G., Anglani, C., Murelli, C., 1993. Epicuticular wax of olive leaves. Phytochemistry 32, 49–52. Blacklock, B.J., Jaworski, J.G., 2006. Substrate specificity of Arabidopsis 3-ketoacyl-CoA synthases. Biochem. Biophys. Res. Commun. 346, 583–590. Bonaventure, G., Salas, J.J., Pollard, M.R., Ohlrogge, J.B., 2003. Disruption of the FATB gene in Arabidopsis demonstrates an essential role of saturated fatty acids in plant growth. Plant Cell 15, 1020–1033.   210 Bond, A.D., 2004. On the crystal structures and melting point alternation of the n-alkyl carboxylic acids. New J. Chem. 28, 104–114. Brathovde, J.R., Lingafelter, E.C., 1958. The crystal structure of decanamide. Acta Crystallogr. 11, 729–732. Briggs, D.E.G., 1999. Molecular taphonomy of animal and plant cuticles: selective preservation and diagenesis. Philos. Trans. R. Soc. B Biol. Sci. 354, 7–17. Buchholz, A., 2006. Characterization of the diffusion of non-electrolytes across plant cuticles: properties of the lipophilic pathway. J. Exp. Bot. 57, 2501–13. Burg, D.A., Kleiman, R., 1991. Meadowfoam fatty amides: preparation, purification, and use in enrichment of 5,13-docosadienoic acid and 5-eicosenoic acid. J. Am. Oil Chem. Soc. 68, 190–192. Buschhaus, C., Herz, H., Jetter, R., 2007a. Chemical composition of the epicuticular and intracuticular wax layers on adaxial sides of Rosa canina leaves. Ann. Bot. 100, 1557–1564. Buschhaus, C., Herz, H., Jetter, R., 2007b. Chemical composition of the epicuticular and intracuticular wax layers on the adaxial side of Ligustrum vulgare leaves. New Phytol. 176, 311–316. Buschhaus, C., Jetter, R., 2011. Composition differences between epicuticular and intracuticular wax substructures: how do plants seal their epidermal surfaces? J. Exp. Bot. 62, 841–853. Buschhaus, C., Jetter, R., 2012. Composition and physiological function of the wax layers coating Arabidopsis leaves: β-amyrin negatively affects the intracuticular water barrier. Plant Physiol. 160, 1120–9. Buschhaus, C., Peng, C., Jetter, R., 2013. Very-long-chain 1,2- and 1,3-bifunctional compounds from the cuticular wax of Cosmos bipinnatus petals. Phytochemistry 91, 249–256. Busta, L., Budke, J.M., Jetter, R., 2016. Identification of β-hydroxy fatty acid esters and primary, secondary-alkanediol esters in cuticular waxes of the moss Funaria hygrometrica. Phytochemistry 121, 38–49. Cane, J.H., Eickwort, G.C., Wesley, F.R., Spielholz, J., 1983. Foraging, grooming and mate-seeking behaviors of Macropis nuda (Hymenoptera, Melittidae) and use of Lysimachia ciliata (Primulaceae) oils in larval provisions and cell linings. Am. Midl. Nat. 110, 257–264. Casado, C.G., Heredia, A., 1999. Structure and dynamics of reconstituted cuticular waxes of grape berry cuticle (Vitis vinifera L.). J. Exp. Bot. 50, 175–182.   211 Chaturvedi, S., Driscoll, W.J., Elliot, B.M., Faraday, M.M., Grunberg, N.E., Mueller, G.P., 2006. In vivo evidence that N-oleoylglycine acts independently of its conversion to oleamide. Prostaglandins Other Lipid Mediat. 81, 136–149. Chibnall, A.C., Piper, S.H., El Mangouri, H.A., Williams, E.F., Iyengar, A.V. V, 1937. CCXLIV. The wax from the leaves of sandal (Santalum album Linn.). Biochem. J. 31, 1981–1986. Choi, S., Chung, M.-H., 2003. A review on the relationship between Aloe vera components and their biologic effects. Semin. Integr. Med. 1, 53–62. Choi, S., Lee, S.-K., Kim, J.-E., Chung, M.-H., Park, Y.-I., 2002. Aloesin inhibits hyperpigmentation induced by UV radiation. Clin. Exp. Dermatol. 27, 513–5. Clough, R.C., Matthis, L., Barnums, R., Jaworski, J.G., 1992. Purification and characterization of 3-ketoacyl-acyl carrier protein synthase III from spinach. J. Biol. Chem. 267, 20992–20998. Deshmukh, A.P., Simpson, A.J., Hadad, C.M., Hatcher, P.G., 2005. Insights into the structure of cutin and cutan from Agave americana leaf cuticle using HRMAS NMR spectroscopy. Org. Geochem. 36, 1072–1085. Diekman, J., Thomson, J.B., Djerassi, C., 1967. Mass spectrometry in structural and stereochemical problems. CXLI. The electron impact induced fragmentations and rearrangements of trimethylsilyl ethers, amines, and sulfides. J. Org. Chem. 32, 3904–3919. Dierickx, P.J., 1973. New β-diketones from Buxus sempervirens. Phytochemistry 12, 1498–1499. Dodova-Anghelova, M.S., Ivanov, C.P., 1973. Composition and structure of secondary fatty alcohols and ketones, isolated from Pyrus malus skin wax. Phytochemistry 12, 2239–2242. Dong, X., Rao, J., Huber, D.J., Chang, X., Xin, F., 2012. Wax composition of “Red Fuji” apple fruit during development and during storage after 1-methylcyclopropene treatment. Hortic. Environ. Biotechnol. 53, 288–297. Dragnevski, K.I., Donald, A.M., Clarke, S.M., Maltby, A., 2009. Novel applications of ESEM and EDX for the study of molecularly thin amide monolayers on polymer films. Colloids Surfaces A Physicochem. Eng. Asp. 337, 47–51. Driscoll, W.J., Chaturvedi, S., Mueller, G.P., 2007. Oleamide synthesizing activity from rat kidney: identification as cytochrome c. J. Biol. Chem. 282, 22353–63. Eigenbrode, S.D., Espelie, K.E., 1995. Effects of plant epicuticular lipids on insect herbivores. Annu. Rev. Entomol. 40, 171–194.   212 Evans, D., Knights, B.A., Math, V.B., Ritchie, A.L., 1975. β-Diketones in Rhododendron waxes. Phytochemistry 14, 2447–2451. Farrell, E.K., Merkler, D.J., 2008. Biosynthesis, degradation and pharmacological importance of the fatty acid amides. Drug Discov. Today 13, 558–568. Feakins, S.J., Sessions, A.L., 2010. Crassulacean acid metabolism influences D/H ratio of leaf wax in succulent plants. Org. Geochem. 41, 1269–1276. Franich, R.A., Wells, L.G., Holland, P.T., 1978. Epicuticular wax of Pinus radiata needles. Phytochemistry 17, 1617–1623. Franke, R., Höfer, R., Briesen, I., Emsermann, M., Efremova, N., Yephremov, A., Schreiber, L., 2009. The DAISY gene from Arabidopsis encodes a fatty acid elongase condensing enzyme involved in the biosynthesis of aliphatic suberin in roots and the chalaza-micropyle region of seeds. Plant J. 57, 80–95. Gniwotta, F., Vogg, G., Gartmann, V., Carver, T.L., Riederer, M., Jetter, R., 2005. What do microbes encounter at the plant surface? Chemical composition of pea leaf cuticular waxes. Plant Physiol. 139, 519–530. Graça, J., Schreiber, L., Rodrigues, J., Pereira, H., 2002. Glycerol and glyceryl esters of ω-hydroxyacids in cutins. Phytochemistry 61, 205–215. Greer, S., Wen, M., Bird, D., Wu, X., Samuels, L., Kunst, L., Jetter, R., 2007. The cytochrome P450 enzyme CYP96A15 is the midchain alkane hydroxylase responsible for formation of secondary alcohols and ketones in stem cuticular wax of Arabidopsis. Plant Physiol. 145, 653–667. Guhling, O., Hobl, B., Yeats, T., Jetter, R., 2006. Cloning and characterization of a lupeol synthase involved in the synthesis of epicuticular wax crystals on stem and hypocotyl surfaces of Ricinus communis. Arch. Biochem. Biophys. 448, 60–72. Guhling, O., Kinzler, C., Dreyer, M., Bringmann, G., Jetter, R., 2005. Surface composition of myrmecophilic plants: Cuticular wax and glandular trichomes on leaves of Macaranga tanarius. J. Chem. Ecol. 31, 2323–2341. Gülz, P.-G., Markstädter, C., Riederer, M., 1994. Isomeric alkyl esters in Quercus robur leaf cuticular wax. Phytochemistry 35, 79–81. Gülz, P.G., Marner, F.J., 1986. Esters of benzyl alcohol and 2-phenyl-ethanol-1 in epicuticular waxes from Jojoba leaves. Zeitschrift für Naturforschung.C. 41, 673–6. Gülz, P.G., Müller, E., Herrmann, T., Lösel, P., 1993. Epicuticular leaf waxes of the hop (Humulus lupulus). Chemical composition and surface structures. Zeitschrift für Naturforschung.C. 48, 689–696.   213 Gülz, P.-G., Müller, E., Prasad, R.B.N., 1989. Organ-specific composition of epicuticular waxes of beech (Fagus sylvatica L.) leaves and seeds. Zeitschrift für Naturforschung.C. 44c, 731–734. Gutterman, Y., Chauser-Volfson, E., 2000. The distribution of the phenolic metabolites barbaloin, aloeresin and aloenin as a peripheral defense strategy in the succulent leaf parts of Aloe arborescens. Biochem. Syst. Ecol. 28, 825–838. Hamberger, A., Stenhagen, G., 2003. Erucamide as a modulator of water balance: New function of a fatty acid amide. Neurochem. Res. 28, 177–185. Hansjakob, A., Riederer, M., Hildebrandt, U., 2011. Wax matters: absence of very-long-chain aldehydes from the leaf cuticular wax of the glossy11 mutant of maize compromises the prepenetration processes of Blumeria graminis. Plant Pathol. 60, 1151–1161. Härtig, C., 2008. Rapid identification of fatty acid methyl esters using a multidimensional gas chromatography-mass spectrometry database. J. Chromatogr. A 1177, 159–169. He, R., Wakimoto, T., Egami, Y., Kenmoku, H., Ito, T., Asakawa, Y., Abe, I., 2012. Heterologously expressed β-hydroxyl fatty acids from a metagenomic library of a marine sponge. Bioorg. Med. Chem. Lett. 22, 7322–5. Hen-Avivi, S., Savin, O., Racovita, R.C., Lee, W.-S., Adamki, N., Malitsky, S., Almekias-Siegl, E., Levy, M., Vautrin, S., Bergès, H., Friedlander, G., Kartvelishvily, E., Ben-Zvi, G., Alkan, N., Uauy, C., Kanyuka, K., Jetter, R., Distelfeld, A., Aharoni, A., 2016. A metabolic gene cluster in the wheat W1 and the barley Cer-cqu loci determines β-diketone biosynthesis and glaucousness. Plant Cell, in press, http://dx.doi.org/10.1105/tpc.16.00197. Herbin, G.A., Robins, P.A., 1968a. Studies on plant cuticular waxes. I. The chemotaxonomy of alkanes and alkenes of the genus Aloe (Liliaceae). Phytochemistry 7, 239–255. Herbin, G.A., Robins, P.A., 1968b. Studies on plant cuticular waxes. II. Alkanes from members of the genus Agave (Agavaceae), the genera Kalanchoe, Echeveria, Crassula and Sedum (Crassulaceae) and the genus Eucalyptus (Myrtaceae) with an examination of Hutchinson’s sub-division of the angi. Phytochemistry 7, 257–268. Herbin, G.A., Robins, P.A., 1969. Patterns of variation and development in leaf wax alkanes. Phytochemistry 8, 1985–1998. Heredia, A., 2003. Biophysical and biochemical characteristics of cutin, a plant barrier biopolymer. Biochim. Biophys. Acta 1620, 1–7. Hiley, C.R., Hoi, P.M., 2007. Oleamide: a fatty acid amide signaling molecule in the cardiovascular system? Cardiovasc. Drug Rev. 25, 46–60.   214 Holloway, P.J., 1969. Chemistry of leaf waxes in relation to wetting. J. Sci. Food Agric. 20, 124–128. Holloway, P.J., Brown, G.A., 1977. The ketol constituents of Brassica epicuticular waxes. Chem. Phys. Lipids 19, 1–13. Horn, D.H.S., Kranz, Z.H., Lamberton, J.A., 1964. The composition of Eucalyptus and some other leaf waxes. Aust. J. Chem. 17, 464–476. Hussain, R.A., Waterman, P.G., 1982. Lactones, flavonoids and benzophenones from Garcinia conrauana and Garcinia mannii. Phytochemistry 21, 1393–1396. International Wheat Genome Sequencing Consortium, 2014. A chromosome-based draft sequence of the hexaploid bread wheat (Triticum aestivum) genome. Science 345, 1251788. Jeffree, C.E., 2006. The fine structure of the plant cuticle, in: Riederer, M., Muller, C. (Eds.), Biology of the Plant Cuticle. Blackwell, Oxford. pp. 11–125. Jenks, M.A., Gaston, C.H., Goodwin, M.S., Keith, J.A., Teusink, R.S., Wood, K. V, 2002. Seasonal variation in cuticular waxes on Hosta genotypes differing in leaf surface glaucousness. HortScience 37, 673–677. Jetter, R., 2000. Long-chain alkanediols from Myricaria germanica leaf cuticular waxes. Phytochemistry 55, 169–176. Jetter, R., Klinger, A., Schaffer, S., 2002. Very long-chain phenylpropyl and phenylbutyl esters from Taxus baccata needle cuticular waxes. Phytochemistry 61, 579–587. Jetter, R., Kunst, L., 2008. Plant surface lipid biosynthetic pathways and their utility for metabolic engineering of waxes and hydrocarbon biofuels. Plant J. 54, 670–683. Jetter, R., Kunst, L., Samuels, A.L., 2006. Composition of plant cuticular waxes, in: Riederer, M., Muller, C. (Eds.), Biology of the Plant Cuticle. Blackwell, Oxford. John Wiley & Sons, pp. 145–181. Jetter, R., Riederer, M., 1996. Cuticular waxes from the leaves and fruit capsules of eight Papaveraceae species. Can. J. Bot. 74, 419–430. Jetter, R., Riederer, M., 1999a. Long-chain alkanediols, ketoaldehydes, ketoalcohols and ketoalkyl esters in the cuticular waxes of Osmunda regalis fronds. Phytochemistry 52, 907–915. Jetter, R., Riederer, M., 1999b. Homologous long-chain δ-lactones in leaf cuticular waxes of Cerinthe minor. Phytochemistry 50, 1359–1364.   215 Jetter, R., Riederer, M., 2000. Composition of cuticular waxes on Osmunda regalis fronds. J. Chem. Ecol. 26, 399–412. Jetter, R., Riederer, M., 2016. Localization of the transpiration barrier in the epi- and intracuticular waxes of eight plant species: water transport resistances are associated with fatty acyl rather than alicyclic components. Plant Physiol. 170, 921-934. Jetter, R., Riederer, M., Seyer, A., Mioskowski, C., 1996. Homologous long-chain alkanediols from Papaver leaf cuticular waxes. Phytochemistry 42, 1617–1620. Jetter, R., Schäffer, S., 2001. Chemical composition of the Prunus laurocerasus leaf surface. Dynamic changes of the epicuticular wax film during leaf development. Plant Physiol. 126, 1725–1737. Jetter, R., Schäffer, S., Riederer, M., 2000. Leaf cuticular waxes are arranged in chemically and mechanically distinct layers: evidence from Prunus laurocerasus L. Plant Cell Environ. 23, 619–628. Ji, X., Jetter, R., 2008. Very long chain alkylresorcinols accumulate in the intracuticular wax of rye (Secale cereale L.) leaves near the tissue surface. Phytochemistry 69, 1197–1207. Jin, G.-Z., Quan, H.-J., Koyanagi, J., Takeuchi, K., Miura, Y., Komada, F., Saito, S., 2005. 4’-O-Alkyaloenin derivatives and their sulfates directed toward overcoming multidrug resistance in tumor cells. Cancer Lett. 218, 15–20. Johnson, D.F., Bennett, R.D., Heftmann, E., 1963. Cholesterol in higher plants. Science 140, 198–199. Jones, T.H., Potts, B.M., Vaillancourt, R.E., Davies, N.W., 2002. Genetic resistance of Eucalyptus globulus to autumn gum moth defoliation and the role of cuticular waxes. Can. J. For. Res. 32, 1961–1969. Kanchana, D., Seipold, L., Schmidt, J., Gerlach, G., Dotterl, S., Ellis, A.G., Wessjohann, L.A., 2008. Non-volatile floral oils of Diascia spp. (Scrophulariaceae). Phytochemistry 69, 1372–1383. Kerstiens, G., 2006. Water transport in plant cuticles: an update. J. Exp. Bot. 57, 2493–9. Kingston, D.G.I., Bursey, J.T., Bursey, M.M., 1974. Intramolecular hydrogen transfer in mass spectra. II. The McLafferty rearrangement and related reactions. Chem. Rev. 74, 215–242. Koch, K., Hartmann, K.D., Schreiber, L., Barthlott, W., Neinhuis, C., 2006. Influences of air humidity during the cultivation of plants on wax chemical composition, morphology and leaf surface wettability. Environ. Exp. Bot. 56, 1–9.   216 Kohlwein, S.D., Eder, S., Oh, C., Martin, C.E., Gable, K.E.N., Bacikova, D., Dunn, T., 2001. Tsc13p is required for fatty acid elongation and localizes to a novel structure at the nuclear-vacuolar interface in Saccharomyces cerevisiae. Mol. Cell. Biol. 21, 109–125. Kolattukudy, P.E., 1970. Plant waxes. Lipids 5, 259–275. Kunst, L., Jetter, R., Samuels, A.L., 2006. Biosynthesis and transport of plant cuticular waxes, in: Riederer, M., Muller, C. (Eds.), Biology of the Plant Cuticle. Blackwell, Oxford. pp. 182–215. Kunst, L., Samuels, A.L., 2003. Biosynthesis and secretion of plant cuticular wax. Prog. Lipid Res. 42, 51–80. Laguna, L., Casado, C.G., Heredia, A., 1999. Flavonoid biosynthesis in tomato fruit cuticles after in vivo incorporation of 3H-phenylalanine precursor. Physiol. Plant. 105, 491–498. Lai, C., Kunst, L., Jetter, R., 2007. Composition of alkyl esters in the cuticular wax on inflorescence stems of Arabidopsis thaliana cer mutants. Plant J. 50, 189–196. Lee, S.-B., Jung, S.-J., Go, Y.-S., Kim, H.-U., Kim, J.-K., Cho, H.-J., Park, O.K., Suh, M.-C., 2009. Two Arabidopsis 3-ketoacyl CoA synthase genes, KCS20 and KCS2/DAISY, are functionally redundant in cuticular wax and root suberin biosynthesis, but differentially controlled by osmotic stress. Plant J. 60, 462–475. Leide, J., Hildebrandt, U., Vogg, G., Riederer, M., 2011. The positional sterile (ps) mutation affects cuticular transpiration and wax biosynthesis of tomato fruits. J. Plant Physiol. 168, 871–7. Leiserowitz, L., 1976. Molecular packing modes. Carboxylic acids. Acta Crystallogr. Sect. B 32, 775–802. Li, F., Wu, X., Lam, P., Bird, D., Zheng, H., Samuels, L., Jetter, R., Kunst, L., 2008. Identification of the wax ester synthase/acyl-coenzyme A: Diacylglycerol acyltransferase WSD1 required for stem wax ester biosynthesis in Arabidopsis. Plant Physiol. 148, 97–107. Li, R., Luo, G., Meyers, P.A., Gu, Y., Wang, H., Xie, S., 2012. Leaf wax n-alkane chemotaxonomy of bamboo from a tropical rain forest in Southwest China. Plant Syst. Evol. 298, 731–738. Li-Beisson, Y., Shorrosh, B., Beisson, F., Andersson, M.X., Arondel, V., Bates, P.D., Baud, S., Bird, D., DeBono, A., Durrett, T.P., Franke, R.B., Graham, I.A., Katayama, K., Kelly, A.A., Larson, T., Markham, J.E., Miquel, M., Molina, I., Nishida, I., Rowland, O., Samuels, L., Schmid, K.M., Wada, H., Welti, R., Xu, C., Zallot, R., Ohlrogge, J., 2010. Acyl-lipid metabolism. Arab. B. 8, 1–65.   217 Lo, Y., Tang, K., Chang, W., Lu, C., Cheng, J., Lee, K., Chou, K., Liu, C., Chen, W., Su, W., Law, Y., Jan, C., 2001. Effect of oleamide on Ca2+ signaling in human bladder cancer cells. Biochem. Pharmacol. 62, 1363–1369. Lodeiro, S., Xiong, Q., Wilson, W.K., Kolesnikova, M.D., Onak, C.S., Matsuda, S.P.T., 2007. An oxidosqualene cyclase makes numerous products by diverse mechanisms: a challenge to prevailing concepts of triterpene biosynthesis. J. Am. Chem. Soc. 129, 11213–11222. Lundqvist, U., von Wettstein, D., 1962. Induction of Eceriferum mutants in barley by ionizing radiations and chemical mutagens. Hereditas 48, 342–362. Lundqvist, U., von Wettstein-Knowles, P., von Wettstein, D., 1968. Induction of Eceriferum mutants in barley by ionising radiations and chemical mutagens, II. Hereditas 59, 473–504. MacKie, A., Misra, A.L., 1956. Chemical investigation of the leaves of Annona senegalensis. I. Constituents of the leaf wax. J. Sci. Food Agric. 7, 203–209. Manheim Jr., B.S., Mulroy, T.W., 1978. Triterpenoids in epicuticular waxes of Dudleya species. Phytochemistry 17, 1799–1800. Markstadter, C., Federle, W., Jetter, R., Riederer, M., Holldobler, B., 2000. Chemical composition of the slippery epicuticular wax blooms on Macaranga (Euphorbiaceae) ant-plants. Chemoecology 10, 33–40. Markstadter, C., Riederer, M., 1997. Attenuation of UV radiation by plant cuticles from woody species. Plant, Cell Environ. 20, 1079–1085. Mata, R., Bye, R., Linares, E., Macias, M., Rivero-Cruz, I., Perez, O., Timmermann, B.N., 2003. Phytotoxic compounds from Flourensia cernua. Phytochemistry 64, 285–291. Matas, A., Heredia, A., 1999. Molecular dynamics modellization and simulation of water diffusion through plant cutin. Zeitschrift für Naturforschung.C. 54, 896–902. Mayberry, W.R., 1980. Hydroxy fatty acids in Bacteroides species : D-(-)-3- hydroxy-15-methylhexadecanoate and its homologs. J. Bacteriol. 143, 582–587. McCloskey, J.A., Stillwell, R.N., Lawson, A.M., 1968. Use of deuterium-labeled trimethylsilyl derivatives in mass spectrometry. Anal. Chem. 40, 233–236. McFadden, W.H., Day, E.A., Diamond, M.J., 1965. Correlations and anomalies in mass spectra. Lactones. Anal. Chem. 37, 89–92. Mejanelle, L., Sanchez-Gargallo, A., Bentaleb, I., Grimalt, J.O., 2003. Long chain n-alkyl diols, hydroxy ketones and sterols in a marine eustigmatophyte, Nannochloropsis gaditana, and in Brachionus plicatilis feeding on the algae. Org. Geochem. 34, 527–538.   218 Mendelson, W.B., Basile, A.S., 2001. The hypnotic actions of the fatty acid amide, oleamide. Neuropsychopharmacology 25, S36–S39. Meusel, I., Leistner, E., Barthlott, W., 1994. Chemistry and micromorphology of compound epicuticular wax crystalloids (Strelitzia type). Plant Syst. Evol. 193, 115–123. Mielniczuk, Z., Mielniczuk, E., Larsson, L., 1993. Gas chromatography-mass spectrometry methods for analysis of 2- and 3-hydroxylated fatty acids: Application for endotoxin measurement. J. Microbiol. Methods 17, 91–102. Mikkelsen, J.D., 1984. Biosynthesis of esterified alkan-2-ols and β-diketones in barley spike epicuticular wax: Synthesis of radioactive intermediates. Carlsberg Res. Commun. 49, 391–416. Millet, E., Rong, J.-K., Qualset, C.O., McGuire, P.E., Bernard, M., Sourdille, P., Feldman, M., 2013. Production of chromosome-arm substitution lines of wild emmer in common wheat. Euphytica 190, 1–17. Mladenova, K., 1976. Trialkyltrioxanes in flower wax of some decorative roses. Phytochemistry 15, 419–420. Mold, J.D., Means, R.E., Stevens, R.K., Ruth, J.M., 1966. The paraffin hydrocarbons of wool wax. Homologous series of methyl alkanes. Biochemistry 5, 455–461. Moreno, E., Cordobilla, R., Calvet, T., Lahoz, F.J., Balana, A.I., 2006. The C form of n-hexadecanoic acid. Acta Crystallogr. Sect. C 62, o129–o131. Morisseau, C., Newman, J.W., Dowdy, D.L., Goodrow, M.H., Hammock, B.D., 2001. Inhibition of microsomal epoxide hydrolases by ureas, amides, and amines. Chem. Res. Toxicol. 14, 409–415. Mueller, G.P., Driscoll, W.J., 2007. In vitro synthesis of oleoylglycine by cytochrome c points to a novel pathway for the production of lipid signaling molecules. J. Biol. Chem. 282, 22364–22369. Nawrath, C., 2006. Unraveling the complex network of cuticular structure and function. Curr. Opin. Plant Biol. 9, 281–287. Nelson, D.R., Sukkestad, D.R., 1975. Normal and branched alkanes from cast skins of the grasshopper Schistocerca vaga (Scudder). J. Lipid Res. 16, 12–18. Nga, V.T., Nhan, T.T.T., Phung, N.K.P., Suong, N.N., 2012. Some fatty compounds from leaves of Pseuderanthemum carruthersii var. Atropurpureum. J. Eng. Technol. Educ. 34–39. Nieschlag, H.J., Wolff, I.A., 1971. Industrial uses of high erucic oils. J. Am. Oil Chem. Soc. 48, 723–727.   219 Nordby, H.E., McDonald, R.E., 1994. Friedelin, the major component of grapefruit epicuticular wax. J. Agric. Food Chem. 42, 708–713. Ohlrogge, J., Browse, J., 1995. Lipid biosynthesis. Plant Cell 7, 957–970. Okamura, N., Asai, M., Hine, N., Yagi, A., 1996. High-performance liquid chromatographic determination of phenolic compounds in Aloe species. J. Chromatogr. A 746, 225–231. Oksuz, S., Topcu, G., 1992. A eudesmanolide and other constituents from Inula graveolens. Phytochemistry 31, 195–197. Osborne, R., Stevens, J.F., 1996. Epicuticular waxes and glaucousness of Encephalartos leaves. Phytochemistry 42, 1335–1339. Park, M.K., Park, J.H., Kim, N.Y., Shin, Y.G., Choi, Y.S., Gyun, J., Kim, K.H., Lee, S.K., 1998. Analysis of 13 phenolic compounds in Aloe species by high performance liquid chromatography. Phytochem. Anal. 9, 186–191. Pascal, S., Bernard, A., Sorel, M., Pervent, M., Vile, D., Haslam, R.P., Napier, J. a, Lessire, R., Domergue, F., Joubès, J., 2013. The Arabidopsis cer26 mutant, like the cer2 mutant, is specifically affected in the very long chain fatty acid elongation process. Plant J. 73, 733–746. Paul, S., Gable, K., Beaudoin, F., Cahoon, E., Jaworski, J., Napier, J.A., Dunn, T.M., 2006. Members of the Arabidopsis FAE1-like 3-ketoacyl-CoA synthase gene family substitute for the Elop proteins of Saccharomyces cerevisiae. J. Biol. Chem. 281, 9018–9029. Phillips, D.R., Rasbery, J.M., Bartel, B., Matsuda, S.P., 2006. Biosynthetic diversity in plant triterpene cyclization. Curr. Opin. Plant Biol. 9, 305–14. Pollard, M., Beisson, F., Li, Y., Ohlrogge, J.B., 2008. Building lipid barriers: biosynthesis of cutin and suberin. Trends Plant Sci. 13, 236–246. Prasad, R.B.N., Gülz, P., 1990. Epicuticular waxes from leaves of maple (Acer pseudoplatanus L.). Zeitschrift für Naturforschung C 45c, 599–601. Prusakiewicz, J.J., Kingsley, P.J., Kozak, K.R., Marnett, L.J., 2002. Selective oxygenation of N-arachidonylglycine by cyclooxygenase-2. Biochem. Biophys. Res. Commun. 296, 612–617. Racovita, R.C., Hen-Avivi, S., Fernandez-Moreno, J.-P., Granell, A., Aharoni, A., Jetter, R., 2016. Composition of cuticular waxes coating flag leaf blades and peduncles of Triticum aestivum cv. Bethlehem. Phytochemistry, in press, http://dx.doi.org/10.1016/j.phytochem.2016.05.003. Racovita, R.C., Jetter, R., 2016. Novel oxidized compounds and internally methyl-branched alkanes from cuticular waxes of Triticum aestivum cv. Bethlehem. PLoS One, submitted.   220 Racovita, R.C., Peng, C., Awakawa, T., Abe, I., Jetter, R., 2015. Very-long-chain 3-hydroxy fatty acids, 3-hydroxy fatty acid methyl esters and 2-alkanols from cuticular waxes of Aloe arborescens leaves. Phytochemistry 113, 183–194. Ramirez, M.X., Hirt, D.E., Wright, L.L., 2002. AFM characterization of surface segregated erucamide and behenamide in linear low density polyethylene film. Nano Lett. 2, 9–12. Rapley, L.P., Allen, G.R., Potts, B.M., 2004. Susceptibility of Eucalyptus globulus to Mnesampela privata defoliation in relation to a specific foliar wax compound. Chemoecology 14, 157–163. Razeq, F.M., Kosma, D.K., Rowland, O., Molina, I., 2014. Extracellular lipids of Camelina sativa: Characterization of chloroform-extractable waxes from aerial and subterranean surfaces. Phytochemistry 106, 188–196. Reiter, B., Lechner, M., Lorbeer, E., Aichholz, R., 1999. Isolation and characterization of wax esters in fennel and caraway seed oils by SPE-GC. J. High Resolut. Chromatogr. 22, 514–520. Rhee, Y., Hlousek-Radojcic, A., Ponsamuel, J., Liu, D., Post-Beittenmiller, D., 1998. Epicuticular wax accumulation and fatty acid elongation activities are induced during leaf development of leeks. Plant Physiol. 116, 901–911. Riedel, M., Eichner, A., Jetter, R., 2003. Slippery surfaces of carnivorous plants: composition of epicuticular wax crystals in Nepenthes alata Blanco pitchers. Planta 218, 87–97. Riedel, M., Eichner, A., Meimberg, H., Jetter, R., 2007. Chemical composition of epicuticular wax crystals on the slippery zone in pitchers of five Nepenthes species and hybrids. Planta 225, 1517–1534. Riederer, M., Schneider, G., 1989. Comparative study of the composition of waxes extracted from isolated leaf cuticles and from whole leaves of Citrus: Evidence for selective extraction. Physiol. Plant. 77, 373–384. Riederer, M., Schneider, G., 1990. The effect of the environment on the permeability and composition of Citrus leaf cuticles II. Composition of soluble cuticular lipids and correlation with transport properties. Planta 180, 154–165. Riederer, M., Schreiber, L., 1995. Waxes - The transport barriers of plant cuticles, in: Hamilton, R.J. (Ed.), Waxes: Chemistry, Molecular Biology and Function. The Oily Press, West Ferry. pp. 131–156. Rietschel, E.T., Gottert, H., Lüderitz, O., Westphal, O., 1972. Nature and linkages of the fatty acids present in the lipid-A component of Salmonella lipopolysaccharides. Eur. J. Biochem. 28, 166–73.   221 Roepke, J., Salim, V., Wu, M., Thamm, A.M.K., Murata, J., Ploss, K., Boland, W., De Luca, V., 2010. Vinca drug components accumulate exclusively in leaf exudates of Madagascar periwinkle. Proc. Natl. Acad. Sci. 107, 15287–15292. Rosler, J., Krekel, F., Amrhein, N., Schmid, J., 1997. Maize phenylalanine ammonia-lyase has tyrosine ammonia-lyase activity. Plant Physiol. 113, 175–179. Rowland, O., Zheng, H., Hepworth, S.R., Lam, P., Jetter, R., Kunst, L., 2006. CER4 encodes an alcohol-forming fatty acyl-coenzyme A reductase involved in cuticular wax production in Arabidopsis. Plant Physiol. 142, 866–877. Samuels, L., Kunst, L., Jetter, R., 2008. Sealing plant surfaces: Cuticular wax formation by epidermal cells. Annu. Rev. Plant Biol. 59, 683–707. Saraf, A., Larsson, L., Burge, H., Milton, D., 1997. Quantification of ergosterol and 3-hydroxy fatty acids in settled house dust by gas comparison with fungal culture and determination of endotoxin by a Limulus amebocyte lysate assay . Quantification of Ergosterol and 3-Hydroxy Fatty Acids in Settled House. Appl. Environ. Microbiol. 63, 2554–2559. Schmidt, H.W.S.J., 1982. Development of plant cuticles: occurence and role of non-ester bonds in cutin of Clivia miniata Reg. leaves. Planta 156, 380–384. Schneider, L.M., Adamski, N.M., Christensen, C.E., Stuart, D.B., Vautrin, S., Hansson, M., Uauy, C., von Wettstein-Knowles, P., 2016. The Cer-cqu gene cluster determines three key players in a β-diketone synthase polyketide pathway synthesizing aliphatics in epicuticular waxes. J. Exp. Bot. 67, 2715-2730. Schreiber, L., 2005. Polar paths of diffusion across plant cuticles: new evidence for an old hypothesis. Ann. Bot. 95, 1069–1073. Schreiber, L., 2008. Characterisation of polar paths of transport in plant cuticles. Annu. Plant Rev. Biol. Plant Cuticle 23, 280. Schulz, S., Arsene, C., Tauber, M., McNeil, J.N., 2000. Composition of lipids from sunflower pollen (Helianthus annuus). Phytochemistry 54, 325–336. Seigler, D., Simpson, B.B., Martin, C., Neff, J.L., 1978. Free 3-acetoxyfatty acids in floral glands of Krameria species. Phytochemistry 17, 995–996. Seipold, L., Gerlach, G., Wessjohann, L., 2004. A new type of floral oil from Malpighia coccigera (Malpighiaceae) and chemical considerations on the evolution of oil flowers. Chem. Biodivers. 1, 1519–1528. Shepherd, T., Robertson, G.W., Griffiths, D.W., 1995. Compositional analysis of intact alkyl esters in leaf epicuticular wax of swede by capillary gas chromatography and electron-impact mass spectrometry. Phytochem. Anal. 6, 65–73.   222 Shimakata, T., Stumpf, P.K., 1982. Isolation and function of spinach leaf beta-ketoacyl-[acyl-carrier- protein] synthases. Proc. Natl. Acad. Sci. 79, 5808–5812. Shockey, J.M., Fulda, M.S., Browse, J.A., 2002. Arabidopsis contains nine long-chain acyl-coenzyme A synthetase genes that participate in fatty acid and glycerolipid metabolism. Plant Physiol. 129, 1710–1722. Shuler, C.A., Janorkar, A. V., Hirt, D.E., 2004. Fate of erucamide in polyolefin films at elevated temperature. Polym. Eng. Sci. 44, 2247–2253. Simpson, B.B., Neff, J.L., Seigler, D., 1977. Krameria, free fatty acids and oil-collecting bees. Nature 267, 150–151. Simpson, B.B., Seigler, D.S., Neff, J.L., 1979. Lipids from the floral glands of Krameria. Biochem. Syst. Ecol. 7, 193–194. Speelman, E.N., Reichart, G.-J., de Leeuw, J.W., Rijpstra, W.I.C., Sinninghe Damsté, J.S., 2009. Biomarker lipids of the freshwater fern Azolla and its fossil counterpart from the Eocene Arctic Ocean. Org. Geochem. 40, 628–637. Speranza, G., Morelli, C.F., Tubaro, A., Altinier, G., Durì, L., Manitto, P., 2005. Aloeresin I, an anti-inflammatory 5-methylchromone from cape aloe. Planta Med. 71, 79–81. Srinivas, G., Babykutty, S., Sathiadevan, P.P., Srinivas, P., 2007. Molecular mechanism of emodin action: transition from laxative ingredient to an antitumor agent. Med. Res. Rev. 27, 591–608. Steinbauer, M.J., Schiestl, F.P., Davies, N.W., 2004. Monoterpenes and epicuticular waxes help female autumn gum moth differentiate between waxy and glossy Eucalyptus and leaves of different ages. J. Chem. Ecol. 30, 1117–1142. Stevens, J.F., Wollenweber, E., Ivancic, M., Hsu, V.L., Sundberg, S., Deinzer, M.L., 1999. Leaf surface flavonoids of Chrysothamnus. Phytochemistry 51, 771–780. Stiti, N., Hartmann, M.-A., 2012. Nonsterol triterpenoids as major constituents of Olea europaea. J. Lipids 2012, 476595. Stránsky, K., Streibl, M., Kubelka, K., 1970. On natural waxes. XV. Hydrocarbon constituents of the leaf wax from the walnut tree (Juglans regia L.). Collect. Czechoslov. Chem. Commun. 35, 882–891. Szafranek, B.M., Synak, E.E., 2006. Cuticular waxes from potato (Solanum tuberosum) leaves. Phytochemistry 67, 80–90.   223 Tantisewie, B., Ruijgrok, H.W.L., Hegnauer, R., 1969. Die Verbreitung der Blausaure bei den Cormophyten 5: Uber cyanogene Verbindungen bei den Parietales und bei einigen weiteren Sippen. Pharm. Weekbl. 104, 1341–1355. Trenkamp, S., Martin, W., Tietjen, K., 2004. Specific and differential inhibition of very-long-chain fatty acid elongases from Arabidopsis thaliana by different herbicides. Proc. Natl. Acad. Sci. 101, 11903–11908. Tsubaki, S., Sugimura, K., Teramoto, Y., Yonemori, K., Azuma, J.-I., 2013. Cuticular membrane of Fuyu persimmon fruit is strengthened by triterpenoid nano-fillers. PLoS One 8, e75275. Tulloch, A.P., 1973. Composition of leaf surface waxes of Triticum species: variation with age and tissue. Phytochemistry 12, 2225–2232. Tulloch, A.P., 1983. Epicuticular waxes from Agropyron dasystachyum, Agropyron riparium and Agropyron elongatum. Phytochemistry 22, 1605–1613. Tulloch, A.P., 1987. Epicuticular waxes of Abies balsamea and Picea glauca: Occurrence of long-chain methyl esters. Phytochemistry 26, 1041–1043. Tulloch, A.P., Bergter, L., 1981. Epicuticular wax of Juniperus scopulorum. Phytochemistry 20, 2711–2716. Tulloch, A.P., Hoffman, L.L., 1971. Leaf wax of durum wheat. Phytochemistry 10, 871–876. Tulloch, A.P., Hoffman, L.L., 1973. Leaf wax of Triticum aestivum. Phytochemistry 12, 2217–2223. Tulloch, A.P., Hoffman, L.L., 1982. Epicuticular wax of Cirsium arvense. Phytochemistry 21, 1639–1642. Tulloch, A.P., Spencer, J.F.T., 1964. Extracellular glycolipids of Rhodotorula species. The isolation and synthesis of 3-D-hydroxypalmitic and 3-D-hydroxystearic acids. Can. J. Chem. 42, 830–835. Tulloch, A.P., Weenink, R.O., 1969. Composition of the leaf wax of Little Club wheat. Can. J. Chem. 47, 3119–3126. Turner, J.D., Lingafelter, E.C., 1955. The crystal structure of tetradecanamide. Acta Crystallogr. 8, 551–557. Udayagiri, S., Mason, C.E., 1997. Epicuticular wax chemicals in Zea mays influence oviposition in Ostrinia nubilalis. J. Chem. Ecol. 23, 1675–1687. Vajdi, M., Nawar, W.W., Merritt Jr., C., 1981. GC/MS analysis of some long chain esters, ketones and propanediol diesters. J. Am. Oil Chem. Soc. 58, 106–110.   224 Van Maarseveen, C., Han, H., Jetter, R., 2009. Development of the cuticular wax during growth of Kalanchoe daigremontiana (Hamet et Perr. de la Bathie) leaves. Plant, Cell Environ. 32, 73–81. Van Maarseveen, C., Jetter, R., 2009. Composition of the epicuticular and intracuticular wax layers on Kalanchoe daigremontiana (Hamet et Perr. de la Bathie) leaves. Phytochemistry 70, 899–906. Van Smeerdijk, D.G., Boon, J.J., 1987. Characterisation of subfossil sphagnum leaves, rootlets of ericaceae and their peat by pyrolysis-high-resolution gas chromatography-mass spectrometry. J. Anal. Appl. Pyrolysis 11, 377–402. Vermeer, C.P., Nastold, P., Jetter, R., 2003. Homologous very-long-chain 1,3-alkanediols and 3-hydroxyaldehydes in leaf cuticular waxes of Ricinus communis L. Phytochemistry 62, 433–438. Vincken, J.-P., Heng, L., de Groot, A., Gruppen, H., 2007. Saponins, classification and occurrence in the plant kingdom. Phytochemistry 68, 275–97. Vioque, J., Pastor, J., 1994. Leaf wax alkanes in the genus Coincya. Phytochemistry 36, 349–352. Vogel, S., 1971. Olproduzierende Blumen, die durch olsammelnde Bienen bestaubt werden. Naturwissenschaften 58, 58–59. Von Wettstein-Knowles, P., 1969. 5. Gene action. Barley Genet. Proc. 2, 146. Von Wettstein-Knowles, P., 1976. Biosynthetic relationships between β-diketones and esterified alkan-2-ols deduced from epicuticular wax of barley mutants. Mol. Gen. Genet. 144, 43–48. Von Wettstein-Knowles, P., 1987. Genes, elongases and associated enzyme systems in epicuticular wax synthesis, in: Stumpf, P.K., Mudd, J.B., Nes, W.D. (Eds.), The Metabolism, Structure, and Function of Plant Lipids. Springer, New York. pp. 489–498. Von Wettstein-Knowles, P., 2007. Analyses of barley spike mutant waxes identify alkenes, cyclopropanes and internally branched alkanes with dominating isomers at carbon 9. Plant J. 49, 250–264. Von Wettstein-Knowles, P., 2012. Plant waxes, in: eLS. John Wiley & Sons Ltd., Chichester. pp. 1–11. Von Wettstein-Knowles, P., Madsen, J.O., 1984. 7-Oxopentadecan-2-ol esters - a new epicuticular wax lipid class. Carlsberg Res. Commun. 49, 57–67.   225 Von Wettstein-Knowles, P., Mikkelsen, J.D., Madsen, J.O., 1984. Nonan-2-ol esters in sorghum leaf epicuticular wax and their collection by preparative gas chromatography. Carlsberg Res. Commun. 49, 611–618. Von Wettstein-Knowles, P., Netting, A.G., 1976a. Esterified alkan-1-ols and alkan-2-ols in barley epicuticular wax. Lipids 11, 478–484. Von Wettstein-Knowles, P., Netting, A.G., 1976b. Composition of epicuticular waxes on barley spikes. Carlsberg Res. Commun. 41, 225–235. Von Wettstein-Knowles, P., Sogaard, B., 1980. The cer-cqu region in barley: gene cluster or multifunctional gene. Carlsberg Res. Commun. 45, 125–141. Von Wettstein-Knowles, P., Sogaard, B., 1981. Genetic evidence that CER-CQU is a cluster-gene, in: Barley Genetics IV. Edinburgh University Press, Edinburgh. pp. 625–630. Wang, Y., Wang, M., Sun, Y., Hegebarth, D., Li, T., Jetter, R., Wang, Z., 2015a. Molecular characterization of TaFAR1 involved in primary alcohol biosynthesis of cuticular wax in hexaploid wheat. Plant Cell Physiol. 56, 1944–1961. Wang, Y., Wang, M., Sun, Y., Wang, Y., Li, T., Chai, G., Jiang, W., Shan, L., Li, C., Xiao, E., Wang, Z., 2015b. FAR5, a fatty acyl-coenzyme A reductase, is involved in primary alcohol biosynthesis of the leaf blade cuticular wax in wheat (Triticum aestivum L.). J. Exp. Bot. 66, 1165–1178. Wen, M., Au, J., Gniwotta, F., Jetter, R., 2006a. Very-long-chain secondary alcohols and alkanediols in cuticular waxes of Pisum sativum leaves. Phytochemistry 67, 2494–2502. Wen, M., Buschhaus, C., Jetter, R., 2006b. Nanotubules on plant surfaces: chemical composition of epicuticular wax crystals on needles of Taxus baccata L. Phytochemistry 67, 1808–1817. Wen, M., Jetter, R., 2007. Very-long-chain hydroxyaldehydes from the cuticular wax of Taxus baccata needles. Phytochemistry 68, 2563–2569. Wen, M., Jetter, R., 2009. Composition of secondary alcohols, ketones, alkanediols, and ketols in Arabidopsis thaliana cuticular waxes. J. Exp. Bot. 60, 1811–1821. Weng, H., Molina, I., Shockey, J., Browse, J., 2010. Organ fusion and defective cuticle function in a lacs1 lacs2 double mutant of Arabidopsis. Planta 231, 1089–100. Wollenweber, E., 1989. Exudate flavonoids in ferns and their chemosystematic implication. Biochem. Syst. Ecol. 17, 141–144. Wollenweber, E., Wehde, R., Dorr, M., Lang, G., Stevens, J.F., 2000. C-Methyl-flavonoids from the leaf waxes of some Myrtaceae. Phytochemistry 55, 965–970.   226 Wollenweber, H.-W., Schramek, S., Moll, H., Rietschel, E.T., 1985. Nature and linkage type of fatty acids present in lipopolysaccharides of phase I and II Coxiella burnetii. Arch. Microbiol. 142, 6–11. Xu, R., Fazio, G.C., Matsuda, S.P.T., 2004. On the origins of triterpenoid skeletal diversity. Phytochemistry 65, 261–291. Yu, G., Nguyen, T.T.H., Guo, Y., Schauvinhold, I., Auldridge, M.E., Bhuiyan, N., Ben-Israel, I., Iijima, Y., Fridman, E., Noel, J.P., Pichersky, E., 2010. Enzymatic functions of wild tomato methylketone synthases 1 and 2. Plant Physiol. 154, 67–77.                    227 Appendices Appendix A  : Synthetic protocols for reference compounds from Chapter 2  A.1 General information All materials, reagents and solvents were used as received from the vendor, except for tetrahydrofuran (THF) and CH2Cl2, which were distilled from Na/benzophenone and CaH2, respectively. All reactions were performed in oven-dried flasks (150°C), without inert atmosphere protection. Analytical TLC was performed on aluminium sheets coated with silica gel 60 F254 (layer thickness 0.2 mm, pore size 60 Å). Preparative TLC was performed on glass plates coated with silica gel 60 F254 (layer thickness 0.5 mm or 1 mm, 20x20 cm, with 4 cm concentration zone). TLC plates were sprayed with a dilute primuline solution (5 mg in 100 mL acetone/water 80/20, v/v) and visualized under 365 nm UV light. Column chromatography was carried out using silica gel (pore size 60 Å, 230−400 mesh). Compound purity was assessed via GC-MS, with derivatization and GC-MS conditions as described in the experimental section. 1H-NMR spectra were recorded from CDCl3 solutions at 25°C on a Bruker Avance 300 MHz spectrometer.   A.2 Synthesis scheme Synthetic standards of C30 3-hydroxy fatty acid, C30 3-hydroxy fatty acid methyl ester (FAME) and C29 2-alkanol were needed for structure confirmation. All three compounds were synthesized from the same starting material, C28 fatty acid, first converted in two steps to C30 3-keto FAME, and then either in two steps into the secondary alcohol or in one step each into the 3-hydroxy FAME and the 3-hydroxy fatty acid (Figure A1).   228  CH3 OHO131. EDC / DMAP2.OOOOCH3 CH3CH3OOOOHOCH3CH313CH3OHreflux CH3 OCH3OO13CH3 OCH3OO13KOH / Bu4NBrDMF / H2OCO2CH3 CH3O13       NaBH4CHCl3 / THFCH3 CH3OH13       NaBH4CHCl3 / THF CH3 OCH3OOH13CH3COOHClHCH3 OHOOH13 Figure A.1. Synthesis scheme.                   Reaction sequences employed for synthesis of authentic standards of C29 2-alkanol, C30 3-hydroxy FAME, and C30 3-hydroxy fatty acid.  A.3 Synthetic procedures and 1H NMR characterization data  5-Octacosanoyl Meldrum’s acid: 200 mg (0.471 mmol) octacosanoic acid (Aldrich, ≥98.5%), 451 mg (2.35 mmol) N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide hydrochloride (EDC, Aldrich, ≥98%), and 230 mg (1.88 mmol) 4-dimethylaminopyridine (DMAP, TCI, ≥99%) were dissolved together in 20 mL freshly distilled dichloromethane (Fisher, ≥99.9%). After stirring for 30 min at room temperature, 271 mg (1.88 mmol) Meldrum’s acid (TCI, ≥98%) were added, and the mixture was stirred overnight. The reaction mixture was vacuum filtered to remove precipitated urea. The organic phase was washed sequentially with diluted HCl (Fisher, 32-35%   229 original conc.) and NaHCO3 (Merck, ≥99.7%), then dried over anhydrous Na2SO4 (Merck, ≥99%), and the solvent removed under vacuum. The residue was taken up in chloroform and purified by column chromatography (using hexane:ethyl acetate 5:1 as mobile phase) to give the final product as a white solid (235 mg, 91% yield). 1H-NMR (300 MHz, CDCl3): δ 3.06 (2H, t, J=7.2 Hz, CH2CO), 1.73 (6H, s, C(CH3)2), 1.10-1.70 (50H, br m, aliphatic CH), 0.88 (3H, t, J=6.9 Hz, CH3).  Methyl 3-oxotriacontanoate: 235 mg (0.427 mmol) 5-octacosanoyl Meldrum’s acid were dissolved in 25 mL of methanol (Fisher, ≥99.9%) and refluxed for eight hours. Then the solution was allowed to cool to room temperature, the precipitate collected by vacuum filtration and washed with cold methanol to afford the practically pure product as a white solid (202 mg, 98% yield). 1H-NMR (300 MHz, CDCl3): δ 3.74 (3H, s, COOCH3), 3.44 (2H, s, COCH2COOCH3), 2.52 (2H, t, J=7.2 Hz, CH2CO), 1.10-1.70 (50H, br m, aliphatic CH), 0.88 (3H, t, J=6.9 Hz, CH3).  2-Nonacosanone: To a solution of 50 mg (0.104 mmol) methyl 3-oxotriacontanoate in 3 mL dimethylformamide (DMF, Aldrich, ≥99.8%) and 0.5 mL distilled water, 132 mg (2.35 mmol) KOH (Fisher, reagent grade, 86.6% batch assay) and a catalytic amount of tetrabutylammonium bromide (Aldrich, ≥98%) were added. The mixture was refluxed at 120°C for four hours, then quenched with dilute HCl and extracted with chloroform. The combined organic layers were dried over anhydrous Na2SO4, filtered and concentrated under reduced pressure. The residue was taken up in hexane (Fisher, HPLC grade, ≥99.9%) and purified by column chromatography (eluting with hexane:ethyl acetate 20:1) to afford the product as a white solid (31.5 mg, 73%   230 yield). 1H-NMR (300 MHz, CDCl3): δ 2.41 (2H, t, J=7.5 Hz, CH2CO), 2.13 (3H, s, COCH3), 1.10-1.90 (50H, br m, aliphatic CH), 0.88 (3H, t, J=6.9 Hz, CH3).  2-Nonacosanol: To a solution of 15 mg (0.035 mmol) 2-nonacosanone in a mixture of 1 mL CHCl3 and 1 mL THF (Aldrich, ≥99.9%, 250 ppm BHT as inhibitor), 11 mg (0.291 mmol) of NaBH4 (Fisher, ≥98%) were added. After stirring overnight at room temperature, the mixture was quenched with dilute HCl and the product extracted with chloroform. The combined organic layers were dried over Na2SO4, filtered and concentrated under reduced pressure. The residue was taken up in chloroform and purified on a preparative TLC plate (hexane:ethyl acetate 11:1) to afford the product as a white solid (11 mg, 73% yield). 1H-NMR (300 MHz, CDCl3): δ 3.80 (1H, m, CHOH), 1.20-1.50 (52H, br m, aliphatic CH), 1.18 (3H, d, J=6.3 Hz, CH(OH)CH3), 0.88 (3H, t, J=6.9 Hz, CH3).  Methyl 3-hydroxytriacontanoate: To a solution of 48 mg (0.1 mmol) methyl 3-oxotriacontanoate in 4 mL of distilled CHCl3 and 4 mL of distilled THF, 48 mg (1.27 mmol) of NaBH4 were added. The mixture was allowed to stir at room temperature overnight before being quenched with dilute HCl and extracted with fresh chloroform. The combined organic layers were dried over Na2SO4, filtered and concentrated under reduced pressure. The residue was purified by recrystallization from MeOH to yield the product as a white solid (34.5 mg, 72% yield). 1H-NMR (300 MHz, CDCl3): δ 4.01 (1H, m, CHOH), 3.71 (3H, s, COOCH3), 2.52 (1H, dd, J=16.2, 3.0 Hz, CHHCOOCH3), 2.40 (1H, dd, J=16.2, 8.7 Hz, CHHCOOCH3), 1.10-1.70 (52H, br m, aliphatic CH), 0.88 (3H, t, J=6.9 Hz, CH3).    231 3-Hydroxytriacontanoic acid: At room temperature, 2 mL of glacial acetic acid (Merck, ≥99.8%) were added to 20 mg (0.04 mmol) methyl 3-hydroxytriacontanoate placed in a glass vial. The vial was heated mildly to dissolve the solid. Then concentrated HCl was added to the solution until a precipitate started to persist. A few more drops of acetic acid were added to give a slightly opaque solution. The mixture was allowed to stand for three days at room temperature and then the white solid was collected by vacuum filtration. The crude product was taken up in chloroform, washed with NaHCO3 and dried over anhydrous Na2SO4. The drying agent was removed by filtration, and the filtrate was dried under vacuum to afford the product as a white solid (13.8 mg, 71% yield). ). 1H-NMR (300 MHz, CDCl3): δ 4.07 (1H, m, CHOH), 2.56 (1H, dd, J=16.1, 3.0 Hz, CHHCOOH), 2.48 (1H, dd, J=16.1, 8.5 Hz, CHHCOOH), 1.10-1.70 (52H, br m, aliphatic CH), 0.88 (3H, t, J=6.9 Hz, CH3).   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0308612/manifest

Comment

Related Items