Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies with plant cell cultures of Tripterycium wilfordii isolation of metabolites and biotransformation… Han, Kang 1995

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

Item Metadata

Download

Media
831-ubc_1995-982896.pdf [ 30.03MB ]
Metadata
JSON: 831-1.0061612.json
JSON-LD: 831-1.0061612-ld.json
RDF/XML (Pretty): 831-1.0061612-rdf.xml
RDF/JSON: 831-1.0061612-rdf.json
Turtle: 831-1.0061612-turtle.txt
N-Triples: 831-1.0061612-rdf-ntriples.txt
Original Record: 831-1.0061612-source.json
Full Text
831-1.0061612-fulltext.txt
Citation
831-1.0061612.ris

Full Text

STUDIES WITH PLANT CELL CULTURES OF TRIPTERYGIUM WILFORDU ISOLATION OF METABOLITES AND BIOTRANSFORMATION STUDIES by  Kang Han B. Sc, Xuzhou Normal University, 1982 M. S c , Research Institute of Chemical Processing and Utilization of Forest Products, Chinese Academy of Forestry, 1985  A THESIS SUBMITED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in  THE FACULTY OF GRADUATE STUDIES (Department of Chemistry)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA November 1994 © Kang Han, 1994  In presenting this  thesis  in  partial fulfilment  of  the  requirements  for an advanced  degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or  by  his  or  her  representatives.  It  is  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  CAAWAI^J  7" The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  J*-o.^3,  tfff  ii  Abstract  This thesis is concerned with the use of plant cell cultures in combination with organic synthesis to provide efficient routes for the synthesis of novel diterpene analogs of triptolide (Tl) and tripdiolide (Td), the two most active principles of the Chinese medicinal plant Tripterygium wilfordii. The present study is also devoted to providing a better understanding with regard to the biosynthesis of related diterpenes in the plant cell culture line of T. wilfordii designated as TRP4a. The initial phases of the research involved isolation and identification of diterpene metabolites from TRP4a cell cultures. Twenty nine compounds (28 diterpenes) were isolated, of which 26 (including 14 new natural products) were isolated for the first time from this cell culture line. Based on the structural relationships revealed within this series, a biosynthetic pathway from dehydroabietane to Tl and Td is proposed. The synthetic studies involved initial synthesis of isodehydroabietenolide from the readily available dehydroabietic acid and subsequent ring C functionalization to provide a series of ring C "activated" analogs such as isotriptophenolide, triptophenolide and demethyl isoneotriptophenolide.  A new, efficient sequence from isodehydroabietenolide to  triptophenolide was developed and a synthesis of demethyl isoneotriptophenolide was completed. The later stages of the present studies were concerned with biotransformation of the various synthetic precursors in both TRP4a whole cell systems and crude enzyme preparations (cell free extract, or CFE). Incubation of isotriptophenolide in TRP4a cell cultures produced two novel epoxy dienones, (7,8)P-epoxy-19-hydroxy-12-oxo-18(4-»3)afce0-abieta-3,9(ll),13trien-18-oic acid lactone, and (7,8)a-epoxy-19-hydroxy-12-oxo-18(4->3)aZ?eo-abieta3,9(1 l),13-trien-18-oic acid lactone, along with additional hydroxylated compounds. This is the first example in which a "para-alkylated" phenol is enzymatically converted to the corresponding para-epoxy dienone in one step. Older cell cultures (21 day) tend to give higher  iii  yields of these epoxides. Further biotransformation results indicated a possible involvement of a quinone methide intermediate in the formation of these epoxides. Biotransformation of isotriptophenolide with CFE preparations from TRP4a cell cultures gave comparable results as the whole cell system except that the biotransformation proceeded much faster. Certain factors in relation to the yield of the epoxides were also studied. Incubation of triptophenolide in whole cell cultures and CFEs showed only a low level of biotransformation in comparison with isotriptophenolide. A large scale experiment did allow the isolation of some hydroxylated products in low yields. These compounds are structurally comparable with those obtained from biotransformation of isotriptophenolide. Biotransformation of demethyl isoneotriptophenolide in whole cell cultures showed rapid consumption of the precursor. The major product isolated was the hydroxy quinone, 12,19-Dihydroxy-l l,14-dioxo-18(4^3)aibeo-abieta-3,8,12-trien-18-oic acid lactone. Biotransformation of the above ring C "activated" precursors has demonstrated that the position and number of the hydroxy groups in ring C are critical to the fate of the biotransformation.  iv  Table of Contents  Abstract  ii  List of Figures  xiv  List of Schemes  xvii  List of Tables  xx  List of Abbreviations.  xxii  Notes.  xxvi  Acknowledgements.  xxvii  Chapter 1 General Introduction.  1  1.1 1.2 1.3 1.4  The Plant Tripterygium wilfordii and Related Species T. wilfordii as Traditional Chinese Medicine: History, Current Clinical Applications and Interests. Phytochemistry of T. wilfordii and Related Species  1 2 5  Pharmacological Activities of Extracts and Compounds from T. wilfordii and Related Species 21 Antileukemia and Antitumor Activities.  21  Immunosuppressive and Anti-inflammatory Activities.  23  Antifertility Activity.  25  Anti-HIV Activity.  29  Side Effects and Toxicity.  29  1.5  Total Syntheses of Triptolide (1)  31  1.6  Plant Cell Culture Biotechnology.  38  1.7  Challenges and Current Research Trends.  1.8  Objectives and Strategies of the Present Investigation  ,  47 50  Chapter 2 Isolation and Structure Elucidation of Diterpene Metabolites from Plant Cell Cultures of T. wilfordii 52  2.1  Introduction  52  2.2  Results and Discussion  53  T. wilfordii Plant Cell Cultures (TRP4a): Growth and Production  53  Isolation of Diterpene Metabolites from TRP4a Cell Cultures  55  Structure Elucidation of Diterpene Metabolites from TRP4a Cell Cultures.  62  Plausible Biosynthetic Pathways.  118  Chapter 3 Syntheses of Precursors for Biotransformation Studies.  133  3.1  Introduction.  133  3.2  Results and Discussion  135  3.2.1  3.2.2  Syntheses of Isodehydroabietenolide and Related Compounds.  135  Synthesis of Isodehydroabietenolide (193)  135  Synthesis of Isodehydroabietenolide Related Compounds.  143  Syntheses of Isotriptophenolide and Related Compounds.  144  Synthesis of Isotriptophenolide (194)  144  Synthesis of 7-Oxo- and 7P-Hydroxy-isotriptophenolide  147  Attempted Synthesis of 8,12-Quinone Methide from Isotriptophenolide (194) 3.2.3  Synthesis of Triptophenolide (106)  3.2.4 Synthesis of Demethyl Isoneotriptophenolide (DINTP, 288) Chapter 4 Biotransformation Studies.  149 155 174 180  4.1  Introduction.  180  4.2  Results and Discussion  184  4.2.1 4.2.2  Biotransformation of 7-Oxo and 7P-Hydroxy-isodehydroabietenolide with TRP4a Cell Cultures  184  Biotransformation of Isotriptophenolide (194) and Related Compounds with TRP4a Cell Cultures  185  Biotransformation of 194 with Older TRP4a Cell Cultures (Trp#300a).  185  Increase of Starting Material (194) Solubility and Isolation ofNewProducts(Trp#301)  187  Biotransformation of 194 with TRP4a Cell Cultures: Time Course Studies.  202  Influence of Cell Age and Incubation Time.  202  Influence of Starting Material-to-Culture Ratio  210  Biotransformation of 194 with Cell Free Extracts (CFE) from TRP4a Cell Cultures.  212  Preliminary Results from CFE Experiments.  212  Time Course Studies with CFE, Cell Homogenates (CH) and Resuspended Cell Pellet (RCP) from TRP4a Cell Cultures of Different Ages  215  Biotransformation of 194 with CFEs and RCPs Obtained by Varying RCF of Centrifugation  226  Influence of pH on the Yield of the Epoxides (305, 306) in CFE Biotransformations  229  Influence of Starting Material-to-CFE Ratio on the Yield of the Epoxides (305,306)  234  Influence of Equivalents of Hydrogen Peroxide on the Yield of the Epoxides (305,306)  236  Influence of Metal Ions on the Yield of the Epoxides (305,306)  237  Biotransformation of 7-Hydroxy-isotriptophenolide (246) with TRP4a Cell Cultures.  240  Biotransformation of the Epoxides (305, 306) with TRP4a Cell Cultures  244  Biotransformation of 194 with TRP4a Cell Cultures: Large Scale Experiment  248  Chemical Epoxidation of the Epoxides (305,306)  260  Biotransformation of Triptophenolide (106) and Related Compounds  263  Biotransformation of 106 with TRP4a Cell Cultures: Preliminary Experiments.  263  vii  Biotransformation of 106 with TRP4a Cell Cultures of Different  4.2.5  Ages  264  Biotransformation of 106 with CFEs from TRP4a Cell Cultures  264  Biotransformation of Quinone 278 with TRP4a Cell Cultures Biotransformation of 106 with TRP4a Cell Cultures: Large Scale Experiment (Trp#334).  266  Biotransformation of 106 with the Fungus Cunninghamella elegans  278  269  Biotransformation of Demethyl Isoneotriptophenolide (288) with TRP4a Cell Cultures  283  Chapter 5 Conclusion and Outlook  288  Chapter 6 Experimental.  298  6.1  General  298  6.2  Experimental for Chapter 2 Isolation of Diterpene Metabolites from Plant Cell Cultures of T. wilfordii (TRP4a).  299  Growth Conditions of TRP4a Cell Cultures.  299  General Harvesting and Extraction Procedures  300  Isolation of Diterpene Metabolites from TRP4a Cell Cultures: A Typical Procedure (Trp#339). 300 Abieta-8,ll,13-triene(196)(dehydroabietane).  303  Synthesis of 196 from dehydroabietic acid (157)  304  Abieta-8,ll,13-trien-3|3-ol(204)  306  Abieta-8,ll,13-trien-3a-ol(202)  307  Abieta-8,ll,13-trien-3-one(198)  307  18-Norabieta-4,8,ll,13-tetraen-3-one(201)  308  Abieta-8,ll,13-trien-7-one(199)  308  Synthesis of 199 from dehydroabietane (196)  309  viii  Abieta-8,ll,13-trien-14-ol(197)  310  Synthesis of 197 from dehydroabietane (196)  310  14-Hydroxy-abieta-8,ll,13-trien-3-one (116) (triptonoterpene)  313  14-Methoxy-abieta-8,ll,13-trien-3-one(200)  314  1 l-Hydroxy-14-methoxy-abieta-8,ll,13-trien-3-one (117) (triptonoterpene)  314  12-Hydroxy-14-methoxy-abieta-8,ll,13-trien-4-one (118) (neotriptonoterpene).  315  14-Methoxy-abieta-8,ll,13-trien-3p-ol(205)  316  14-Methoxy-abieta-8,ll,13-trien-3a-ol(203)  316  18(4H>3)a&?o-Abieta-3,8,l l,13-tetraen-18-oic acid (210)  317  7-Oxo-18(4^>3)afo?0-abieta-3,8,11,13-tetraen-18-oic acid (211)  317  Squalene (195)  318  1 l,19-Dihydroxy-14-methoxy-18(4-»3)a^o-abieta-3,8,l 1,13tetraen-18-oic acid lactone (109) (neotriptophenolide) 318 14,19-Dihydroxy-7-oxo-18(4->3)a&?o-abieta-3,8,11,13tetraen-18-oic acid lactone (108) (triptonide)  319  7p,14,19-Trihydroxy-18(4->3)a^eo-abieta-3,8,ll,13-tetraen18-oic acid lactone (138a) 319 Synthesis of 138a from 108  320  7P, 19-Dihydroxy- 14-methoxy-18(4->3)a£eo-abieta-3,8,11,13tetraen-18-oic acid lactone (207) 320 Triptolide (1)  321  Tripdiolide (2)  321  (7,8)p,(9,11)P,( 12,13)a-Tris(epoxy)-2p, 19-dihydroxy- 14-oxo18(4->3)a&eo-abieta-3-en-l8-oic acid lactone (212) 322 14P-Acetoxy-(7,8)P,(9,ll)P,(12,13)a-tris(epoxy)-2p,19dihydroxy-18(4—»3)a&eo-abieta-3-en-l8-oic acid lactone (208)  323  (7,8)P,(9,ll)p-Bis(epoxy)-2p,19-dihydroxy-14-oxo-18(4->3) a£eo-abieta-3,12-dien-18-oic acid lactone (209) 323  Experimental for Chapter 3 Syntheses of Precursors. 6.3.1  6.3.2  Synthesis of Isodehydroabietenolide (193)  324 324  18-Norabieta-4(19),8,ll,13-tetraene(228)  325  18,19-Dinorabieta-8,ll,13-trien-4-one(229)  327  3-Dimethylthiomethylene-18,19-dinorabieta-8,11,13-trien-4one(230)  329  19-Hydroxy-18(4->3)aZ*?o-abieta-3,8,11,13-tetraen-18-oic acid lactone (193)  330  Isolation of reaction intermediates and byproducts in the step from 230 to 193  332  Synthesis of 7|3-Hydroxy-isodehydroabietenolide (240)  335  19-Hydroxy-7-oxo-18(4->3)cfeeo-abieta-3,8,11,13-tetraen-18oic acid lactone (206) 335 7p\ 19-Dihydroxy-18(4->3)a&eo-abieta-3,8,11,13-tetraen-18oic acid lactone (240) 6.3.3  Synthesis of Isotriptophenolide (194)  336 337  12-Acetyl-19-hydroxy-18(4->3)a6eo-abieta-3,8,11,13-tetraen18-oic acid lactone (241) 337  6.3.4  12-Acetoxy-19-hydroxy-18(4->3)abeo-abieta-3,8,11,13tetraen-18-oic acid lactone (242)  338  12,19-Dihydroxy-18(4->3)afreo-abieta-3,8,11,13-tetraen-18oic acid lactone (194) (isotriptophenolide).  339  Synthesis of 7(3-Hydroxy-isotriptophenolide (246b)  341  12-Acetoxy-19-hydroxy-7-oxo-18(4->3)afo?0-abieta-3,8,11,13tetraen-18-oic acid lactone (244) 341 7p\ 12,19-Trihydroxy-18(4-»3)a£eo-abieta-3,8,11,13-tetraen18-oic acid lactone (246b) 342 6.3.5  Attempted Synthesis of 8,12-Quinone Methide from Isotriptophenolide (194) 19-Hydroxy-12-oxo-18(4->3)Gt£><?0-abieta-3,5,7,9( 11), 13pentaen-18-oic acid lactone (250)  345 345  7a, 12,19-Trihydroxy-18(4->3)a&eo-abieta-3,8,11,13-tetraen18-oic acid lactone (246a) 346  Synthesis of Triptophenolide (106) Preliminary Experiments 19-Hydroxy- 14-methoxy-18(4->3)a6eo-abieta-3,8,11,13tetraen-18-oic acid lactone (107) Synthesis of Triptophenolide (106) from 193  347 347 351 352  19-Hydroxy-12,14-dinitro-18(4^3)abeo-abieta-3,S,11,13tetraen-18-oic acid lactone (272)  352  12-Amino-19-hydroxy-14-nitro-18(4->3)a&eo-abieta3,8,11,13-tetraen-18-oic acid lactone (274)  354  Synthesis of 106 from 274 using Zn-HOAc in the reduction step 19-Hy droxy-11,14-dioxo-18 (4^3)abeo-abicta-3,8,12-trien18-oic acid lactone (278) Direct deamination of 274 with isoamyl nitrile  355 356 357  19-Hydroxy- 14-nitro-18(4->3)a &eo-abieta-3,8,11,13-tetraen18-oic acid lactone (282) 358 19-Hydroxy- 12-iodo- 14-nitro-18(4->3)afceo-abieta-3,8,11,13tetraen-18-oic acid lactone (275) 359 Reaction of 275 with NaBH4-CuCl  360  14-Amino- 19-hydroxy-18(4-»3)aZ?e<?-abieta-3,8,11,13-tetraen18-oic acid lactone (276) 360 Deiodination of 275 by catalytic hydrogenation  361  Deiodination and reduction of 275 by catalytic hydrogenation  361  14-Amino-19-hydroxy-12-iodo-18(4^3)^eo-abieta3,8,1 l,13-tetraen-18-oic acid lactone (280)  362  14,19-Dihydroxy- 12-iodo-18(4->3)a6eo-abieta-3,8,11,13tetraen-18-oic acid lactone (279)  363  14,19-Dihydroxy-18(4->3)a&?0-abieta-3,8,11,13-tetraen-18oic acid lactone (106) (triptophenolide).  366  Synthesis of 12,14,19-Trihydroxy-18(4->3)a&e<?-abieta-3,8,l 1,13-tetraen18-oic acid lactone (DINTP, 288) 367 14-Acetoxy-12,19-dihydroxy-18(4->3)afoo-abieta-3,8,11,13tetraen-18-oic acid lactone (294) 368 12,14,19-Trihydroxy-18(4-43)afeeo-abieta-3,8,11,13-tetraen-18oic acid lactone (288) 368  Experimental for Chapter 4 Biotransformation Studies  369  6.4.1  General  6.4.2  Attempted Biotransformation of 7-Oxo-isodehydroabietenolide (206) with TRP4a Cell Cultures 373  6.4.3  Biotransformation of 7(3-Hydroxy-isodehydroabietenolide (240) with TRP4a Cell Cultures (Trp#295).  374  Biotransformation of Isotriptophenolide (194) and Related Compounds with TRP4a Cell Cultures  374  Biotransformation of 194 with Older Cell Cultures (Trp#300a)  374  6.4.4  369  19-Hydroxy-12-methoxy-18(4->3)a&eo-abieta-3,8,11,13tetraen-18-oic acid lactone (304) Increase of Starting Material Solubility (Trp#301). 7-Ethoxy-12,19-dihydroxy-18(4^3)abeo-abieta-3,8,11,13tetraen-18-oic acid lactone (307)  375 375 376  5a, 12,19-Trihydroxy-18(4->3)a£eo-abieta-3,8,11,13-tetraen18-oic acid lactone (309) 377 5p,12,19-Trihydroxy-18(4->3)aZ7eo-abieta-3,8,ll,13-tetraen18-oic acid lactone (308)  377  (7,8)p-Epoxy-19-hydroxy- 12-oxo-18(4->3)abeo-abieta3,9(1 l),13-trien-18-oic acid lactone (305)  378  (7,8)0C-Epoxy-19-hydroxy-12-oxo-18(4-»3)a6eo-abieta3,9(1 l),13-trien-18-oic acid lactone (306)  379  Influence of Cell Age and Starting Material-to-Culture Ratio: Time Course Studies (Trp#305-309).  380  Preliminary Biotransformation of 194 with CFE Prepared from TRP4a Cell Cultures (Trp#310) 381 Biotransformation of 194 with CFE, CH and RCP from TRP4a Cell Cultures of Different Ages: Time Course Studies (Trp#311-313).  381  Biotransformation of 194 with CFEs and RCPs Obtained by Varying RCF of Centrifugation (Trp#314).  383  Biotransformation of 194 with CFEs Prepared with Buffers of Different pH(Trp#315, 317)  383  Biotransformation of 194 with CFE at Different Starting Materialto-CFE Ratios (Trp#316).  385  Biotransformation of 194 with CFEs Containing Different Equivalents of Hydrogen Peroxide (Trp#318a). 385 Biotransformation of 194 with CFEs with Different Metal Ions Being Added (Trp#320). 386 Biotransformation of 7-Hydroxy-isotriptophenolide (246) with TRP4a Cell Cultures (Trp#340) 386 Biotransformation of the Epoxides (305,306) with TRP4a Cell Cultures (Trp#338)  387  Biotransformation of 194 with TRP4a Cell Cultures: Large Scale Experiment (Trp#336).  388  (7,8)p,(9,ll)a-Bis(epoxy)-5a,19-dihydroxy-12-oxo-18(4->3) a&£o-abieta-3,13-dien-18-oic acid lactone (311) 391 12,19-Dihydroxy-7-oxo-18(4->3)afoo-abieta-3,5,8,ll,13pentaen-18-oic acid lactone (252) Chemical Epoxidation of the Epoxides (305,306) (7,8)a,(13,14)p-Bis(epoxy)-19-hydroxy-12-oxo-18(4->3) cZ?eo-abieta-3,9(ll)-dien-18-oic acid lactone (312)  391 392 392  Biotransformation of Triptophenolide (106) and Related Compounds with TRP4a cell Cultures.  393  Preliminary Experiments (Trp#322).  393  Biotransformation of 106 with TRP4a Cell Cultures of Different Ages (Trp#323-327) Biotransformation of 106 with CFEs from TRP4a Cell Cultures (Trp#318b).  393  Biotransformation of 106 with CFE from TRP4a Cell Cultures: Increase of CFE-to-Starting Material Ratio (Trp#319).  394 394  Biotransformation of 106 with CFE from TRP4a Cell Cultures: Increase of Reaction Time (T#321).  395  Biotransformation of Quinone 278 with TRP4a Cell Cultures.  396  Preliminary Experiment (Trp#328).  396  Larger Scale Experiment (Trp#332). 5a, 19-Dihydroxy-11,14-dioxo-18(4->3)a&?o-abieta-3,8,12trien-18-oic acid lactone (313)  396 397  xiii  Biotransformation of 106 with TRP4a Cell Cultures: Large Scale Experiment (Trp#334).  397  7a, 14,19-Trihydroxy-18(4-»3)afceo-abieta-3,8,11,13-tetraen18-oic acid lactone (138b) 399 5P, 14,19-Trihydroxy-18(4->3)afo?o-abieta-3,8,11,13-tetraen18-oic acid lactone (315) 399 5a, 14,19-Trihydroxy-18(4->3)a6eo-abieta-3,8,11, 13-tetraen18-oic acid lactone (316) 400 6.4.7  Biotransformation of Triptophenolide (106) with the Fungus Cunninghamella elegans  400  1 l,14,19-Trihydroxy-18(4->3)afo?0-abieta-3,8,l 1,13-tetraen18-oic acid lactone (314) 401 1 l-Acetoxy-14,19-dihydroxy-18(4^3)flfce0-abieta-3,8,l 1,13tetraen-18-oic acid lactone (318) 402 6.4.8  Biotransformation of DINTP (288) with TRP4a Cell Cultures.  403  Preliminary Experiment (Trp#30lb)  403  Large Scale Experiment (Trp#335).  403  12,19-Dihydroxy- 14-methoxy-18(4->3)a&?o-abieta-3,8,11,13tetraen-18-oic acid lactone (319) 405 14,19-Dihydroxy-12-methoxy-18(4->3)a&?o-abieta-3,8,11,13tetraen-18-oic acid lactone (320) 406 12,19-Dihydroxy-ll,14-dioxo-18(4->3)a^o-abieta-3,8,12trien-18-oic acid lactone (321)  406  Reference  408  Appendix  421  xiv  List of Figures  Figure 1.1  Alkylation of Thiols by the Diterpene Triepoxides via Hydroxyl-assisted Epoxide Ring Opening 23  Figure 1.2  Development and Routine Maintenance of the TRP4a Cell Line of T. wilfordii  43  Figure 2.1  General Procedure for Cell Culture Harvesting  54  Figure 2.2  Column Chromatographic Separation of Extract from TRP4a Cell Cultures  56  Figure 2.3  Maj or NOEs Observed in NOE Difference Spectra of Compound 204.  65  Figure 2.4  Expanded COS Y Spectrum of Compound 202  68  Figure 2.5  Major NOEs Observed in NOE Difference Spectra of Compound 202.  69  Figure 2.6  Major NOEs Observed in NOE Difference Spectra of Compound 201.  73  Figure 2.7  Expanded COSY Spectrum of Compound 201.  74  Figure 2.8  Comparison of NOE Results between 117 and 118  83  Figure 2.9  Some NOEs Observed for 205.  85  Figure 2.10  NOE Difference Spectra of Compound 203.  88  Figure 2.11  *H NMR spectra of 203 (400 MHz, aromatic region).  90  Figure 2.12  Two Possible Structures for 210  93  Figure 2.13  NOEs Observed for Compound 210.  94  Figure 2.14  Major COSY and NOE Results Observed for 138a  103  Figure 2.15  COS Y Spectrum of Compound 209.  110  Figure 2.16  NOE Difference Spectra of Compound 209  112  Figure 2.17  HMQC Spectrum of 209  115  Figure 2.18 Figure 2.19  HMBC Spectrum of 209. Proposed Mechanism for Formation of 3(3 alcohol and the Probable Relation with the a Isomer  116  Figure 3.1  *H NMR Spectra of Intermediates in the Reaction from 230 to 193.  141  Figure 4.1  Column Chromatographic Separation of Extracts from TRP#301  189  Figure 4.2  Expanded COSY Spectrum of Compound 306.  194  128  XV  Figure 4.3  NOE Difference Spectra of Compound 306.  195  Figure 4.4  X-ray Structure of 306.  197  Figure 4.5  Comparison of Dihedral Angles between H7 and H6 in Compounds 305 and 306.  199  Figure 4.6  Expanded COS Y Spectrum of Compound of 305.  200  Figure 4.7  NOE Difference Spectra of Compound 305.  201  Figure 4.8  Changes in the Amounts of Isotriptophenolide (ITP, 194), 305 and 306 during the Biotransformation (Trp#305, cell age: 15 day). Changes in the Amounts of Isotriptophenolide (ITP, 194), 305 and 306 during the Biotransformation (Trp#306, cell age: 7day) Changes in the Amounts of Isotriptophenolide (ITP, 194), 305 and 306 during the Biotransformation (Trp#307, cell age: 21 day).  Figure 4.9 Figure 4.10 Figure 4.11  204 206 208  Comparison of the Epoxide Yields in Biotransformations of 194 with TRP4a Cell Cultures of Different Ages.  209  Comparison of the Epoxide Yields in Biotransformations of 194 with TRP4a Cell Cultures in Different Starting Material-to-Culture Ratios (Trp#307, 308, 309, cell age: 21 day).  211  Figure 4.13  General Procedure for Preparation of CFE, CH and RCP.  213  Figure 4.14  Changes in the Amounts of Isotriptophenolide (ITP, 194), epoxides 305 and 306 in the Biotransformation of 194 with CFE Prepared from a 21-day-old Cell Culture  219  Comparison of the Epoxide Yields in Biotransformations of 194 with CFE, CH and RCP Prepared from a 21-day-old Cell Culture.  220  Figure 4.12  Figure 4.15 Figure 4.16  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with CFE, CH and RCP Prepared from a 21-day-old Cell Culture....222  Figure 4.17  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with CFE, CH and RCP Prepared from a 13-day-old Cell Culture....222  Figure 4.18  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with CFE, CH and RCP Prepared from a 7-day-old Cell Culture 223  Figure 4.19  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with CFEs Prepared from Cell Cultures of Different Ages (Trp#311:21day,Trp#312: 13 day,Trp#313: 7 day). 224  Figure 4.20  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with CHs Prepared from Cell Cultures of Different Ages (Trp#311:21day,Trp#312: 13 day,Trp#313: 7 day). 224  XVI  Figure 4.21  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with RCPs Prepared from Cell Cultures of Different Ages (Trp#311: 21 day, Trp#312: 13 day, Trp#313: 7 day). 225  Figure 4.22  Comparison of the Epoxide Yields (Specific Yields) in Biotransformations of 194 with CFEs, CHs and RCPs Prepared from Cell Cultures of Different Ages (Trp#311: 21 day, Trp#312: 13 day, Trp#313: 7 day). 226  Figure 4.23  Preparation of CFEs and RCPs by Varying RCF of Centrifugation  227  Figure 4.24  Comparison of the Epoxide Yields in Biotransformations of 194 with CFEs and RCPs Obtained by Varying RCF of Centrifugation  229  Comparison of the Epoxide Yields in Biotransformations of 194 with CFEs Prepared from Phosphate Buffers of Different pH  233  Comparison of the Epoxide Yields in Biotransformations of 194 with CFEs Prepared from Acetate Buffers of Different pH.  233  Comparison of the Epoxide Yields (based on recovered starting material) in Biotransformations of 194 with CFEs Prepared from Buffers of Different pH  234  Comparison of the Epoxide Yields in Biotransformations of 194 with Different Starting Material-to-CFE Ratios.  235  Comparison of the Epoxide Yields in Biotransformations of 194 with Different Equivalents of Hydrogen Peroxide.  237  Comparison of the Epoxide Yields in Biotransformations of 194 with Different Metal Ions Being Added.  239  Figure 4.25 Figure 4.26 Figure 4.27  Figure 4.28 Figure 4.29 Figure 4.30 Figure 4.31  Column Chromatographic Separation of Extracts from Biotransformation of 194 with TRP4a Cell Cultures (large scale, Trp#336)  250  Figure 4.32  Proton Correlations Observed in the COSY Spectrum of 311.  255  Figure 4.33  Major NOEs Observed in the NOE Difference Spectra of 311.  256  Figure 4.34  Proposed Mechanism for the Formation of the C7, C8 Epoxide Group in 305, 306 and 139.  259  Figure 4.35  Stereochemistry of the Epoxy Group between C13 and C14.  262  Figure 4.36  Column Chromatographic Separation of Biotransformation Products of 106  270  Figure 4.37  Some NOEs Observed in NOE Difference Spectra of 315  274  Figure 5.1  Summary of Diterpenes Metabolites Isolated from TRP4a Cell Cultures  288  List of Schemes  Scheme 1.1  Synthesis of (+) Triptolide (1) via BC->ABC Abietane Construction  33  Scheme 1.2  Synthesis of (+) Triptolide (1) via AB->ABC Abietane Construction  34  Scheme 1.3  Biogenetic-type Synthesis of Triptolide (1)  35  Scheme 1.4  Synthesis of Triptolide (1) from Dehydroabietic Acid (157)  37  Scheme 2.1  Proposed Mass Spectral Fragmentation of 204.  66  Scheme 2.2  Synthesis of 196 and 199 from 157  78  Scheme 2.3  Synthesis of 197 from 196  78  Scheme 2.4  Proposed Biosynthetic Pathway to Td (2) via Dehydroabietic Acid (157).... 119  Scheme 2.5  Proposed Biosynthetic Pathway of Td (2) via 4->3 Methyl Transfer.  120  Scheme 2.6  Proposed Biosynthetic Pathway to Tl (1) and Td (2) by M. Roberts  121  Scheme 2.7  Relationship between Different Groups of Compounds.  125  Scheme 2.8  Proposed Mechanism for Enzymatic Formation of the Butenolide.  126  Scheme 2.9  Proposed Biosynthetic Pathway from Dehydroabietane (196) to the Key Intermediate 106  130  Scheme 2.10 Proposed Biosynthetic Pathway from triptophenolide (106) to Tl (1) and Td(2)) Scheme 2.11 An Alternative Pathway from 139 to Tl (1) Scheme 3.1  132 131  Synthesis of IsodehydroabietenoUde (193) from Dehydroabietic Acid (157)  137  Scheme 3.2  Proposed Reaction Mechanism for the Formation of Butenolide Moiety.  142  Scheme 3.3  Proposed Mechanism for the Formation of 239.  143  Scheme 3.4 Scheme 3.5  Synthesis of Isotriptophenolide (194) from IsodehydroabietenoUde (193)... 145 Synthesis of 7p-Hydroxy-isotriptophenolide (246b) 148  Scheme 3.6  Quinone Methide 249 as a Possible Biotransformation Precursor  150  Scheme 3.7  Reaction of Isotriptophenolide (194) with DDQ in THF.  151  Scheme 3.8 Scheme 3.9  Proposed Mechanism for the Formation of Compounds 253, 246. 154 Synthesis of Isotriptophenolide Acetate (171) via C—»A Functionalizatioo.. 155  Scheme 3.10 Preliminary Attempt to Synthesize 106 from 229.  156  Scheme 3.11 Preliminary Attempt to Synthesize 106 from a Derivative of 229.  157  Scheme 3.12 Syntheses of Ring C "Activated" Compounds via the Common Intermediate 193.  158  Scheme 3.13 Reactions from 274 to 106  161  Scheme 3.14 Comparison of Yields of 106 and 267 with respect to 264 and 268  164  Scheme 3.15 Proposed Sequences for Synthesis of 106  165  Scheme 3.16 Deamination of 274 by Isoamyl Nitrite  166  Scheme 3.17 Reduction of Iodo and Nitro Groups with NaBELi-CuCl System.  167  Scheme 3.18 Selective Reduction of Nitro Group withNa2S204  168  Scheme 3.19 The New Sequence for the Synthesis of Triptophenolide (106)  170  Scheme 3.20 Comparison of Reaction Yields between 268 and 264.  171  Scheme 3.21 Overall Yields for the Synthesis of 106 from 193  174  Scheme 3.22 Oxidation of 194 with LTA and Subsequent Dienone-phenol Rearrangement  176  Scheme 3.23 Proposed Mechanism for the Formation of 295.  178  Scheme 4.1  Biotransformation of Allylic Alcohol 297 with TRP4a Cell Cultures  182  Scheme 4.2  Biotransformation of the Butenolide 193 with TRP4a Cell Cultures.  183  Scheme 4.3  Biotransformation of Isotriptophenolide (194) with TRP4a Cell Cultures... 183  Scheme 4.4  Biotransformation of Alcohol 240 with TRP4a Cell Cultures.  185  Scheme 4.5  Proposed Mechanism for the Formation of Epoxides 305 and 306  241  Scheme 4.6  Biotransformation of 7-Hydroxy-isotriptophenolide (246) in the  Scheme 4.7 Scheme 4.8  TRP4a Cell Culture.  242  Proposed Mechanism for the Formation of 307 and 247.  244  Biotransformation of the Epoxides 305 and 306 in the TRP4a Cell Culture 246 Scheme 4.9 Relationship between the Epoxides (305,306) and 7-Hydroxyisotriptophenolide (246a, 246b) 247 Scheme 4.10 Biotransformation of Isotriptophenolide (194) with TRP4a Cell Cultures... 260  xix  Scheme 4.11 Comparison of 106 and 194 Related Compounds Isolated from Biotransformation and TRP4a Cell Cultures  276  Scheme 4.12 Relationship of TriptophenoUde (106) with Its Products in the Biotransformation with C. elegans  282  Scheme 5.1 Scheme 5.2 Scheme 5.3  Proposed Biosynthetic Pathway from Dehydroabietane (196) to the Key Intermediate 106.  291  Proposed Biosynthetic Pathway from Triptophenolide (106) to Tl (1) andTd(2)  292  Comparison of 106 and 194 Related Compounds Isolated from Biotransformation and TRP4a Cell Cultures  294  XX  List of Tables Table 1.1  Macrocyclic Lactone Alkaloids from Tripterygium Species  7  Table 1.2  Spermidine Alkaloids from Tripterygium Species.  7  Table 1.3  Sesquiterpene Polyesters from Tripterygium Species  8  Table 1.4  Triterpenoids from Tripterygium Species  11  Table 1.5  Diterpenoids from Tripterygium species.  17  Table 1.6  Isolated Compounds and Their Pharmacological Activities.  27  Table 2.1  Compounds Isolated from TRP4a Cell Cultures  60  Table 2.2  *H NMR Spectral Data of Abietane Diterpenes 196,204, 202,198,201 and 199.  76  !H NMR Spectral Data of Abietane Diterpenes 197,116, 200,117,118, 205 and 203  91  Table 2.3 Table 2.4  !H NMR Spectral Data of abeo-Abietane Diterpenes 210, 211,193, 206, 106 and 107 100  Table 2.5  !H NMR Spectral Data of afoo-Abietane Diterpenes 109,108,138a, and 207  106  Table 2.6  *H NMR Spectral Data of Epoxides Tl (1), Td (2), 212, 208, and 209  117  Table 3.1  Comparison of Yields for Conversion of 194 to 288.  179  Table 4.1  Compounds Isolated from Biotransformation of 194 with the TRP4a Cell Culture (Trp#301). 188  Table 4.2  Conditions for Biotransformation of 194 with CFE, CH and RCP Prepared from a 21-day-old TRP4a Cell Culture (Trp#311).  216  Conditions for Biotransformation of 194 with CFE, CH and RCP Prepared from a 13-day-old TRP4a Cell Culture (Trp#312).  218  Conditions for Biotransformation of 194 with CFE, CH and RCP Prepared from a 7-day-old TRP4a Cell Culture (Trp#313).  218  Conditions for Biotransformation of 194 with CFEs and RCPs Prepared from a 15-day-old TRP4a Cell Culture by Varying RCF of Centrifugation(Trp#314).  228  Conditions of Biotransformation of 194 with CFEs Prepared from Phosphate Buffers of Different pH(Trp#315).  231  Conditions for Biotransformation of 194 with CFEs Prepared from Acetate Buffers of Different pH(Trp#317).  232  Table 4.3 Table 4.4 Table 4.5  Table 4.6 Table 4.7  xxi  Table 4.8 Table 4.9  Conditions for Biotransformation of 194 with Different CFE-to-Starting Materials Ratio (Trp#316)  235  Conditions for Biotransformation of 194 with Different Equivalents of Hydrogen Peroxide (Trp#318a).  236  Table 4.10  Conditions for Biotransformation of 194 with Different Metal Ions Being Added (Trp#320). 238  Table 4.11  Compounds Isolated from Biotransformation of 246 with the TRP4a Cell Culture. 241  Table 4.12  Compounds Isolated from Biotransformation of 305,306 with the TRP4a Cell Culture. 245  Table 4.13  Compounds Isolated from Biotransformation of 194 with the TRP4a Cell Culture (large scale, Trp#336).  249  Table 4.14  *H NMR Spectral Data of Epoxides 305,306,311 and 312.....  257  Table 4.15  Compounds Isolated from Biotransformation of Triptophenolide (106) with the TRP4a Cell Culture (Trp#334).  271  Table 4.16  Compounds Isolated from Biotransformation of 106 with C. elegans.  279  Table 4.17  Compounds Isolated from Biotransformation of DINTP (288) with the TRP4a Cell Culture Preparation of CFE, CH and RCP for Trp#311, 312 and 313.  284 382  Table 6.1  List of Abbreviations  K AB q Ac ATCC atm B-5 br brine Bu c °C  specific rotation recorded at t°C using sodium D-line AB quartet acetyl American Type Culture Collection atmosphere standard cell culture medium developed by Gamborg and Eveleigh broad saturated sodium chloride solution butyl concentration (g/100 mL) degree Celsius  c. c. CFE CH cm-1 COSY 8  column chromatography cell free extract cell homogenate wave number l U-lH 2-dimensional Correlated NMR Spectroscopy chemical shift  d dd ddd dddd dec A Av  doublet doublet of doublets doublet of doublet of doublets doublet of doublet of doublet of doublets decomposition heat chemical shift difference  DDQ DINTP DMF DMP DMSO Et  2,3-dichloro-5,6-dicyano-1,4-benzoquinone  ED 50 FMN Fr. g  demethyl isoneotriptophenolide dimethylformamide 3,5-dimethylpyrazole dimethylsulphoxide ethyl median effective dose flavin mononucleotide Fraction gram  XX111  8 GTW h [H] hexanes HMBC HMPA HMQC  hv HPLC HRMS Hz IR /  KB I  gravitation constant a multi-glycoside extract from the plant Tripterygium wilfordii hour reduction generally a mixture of several isomers of hexane (C6H14), predominantly n-hexane, and methylcyclopentane (C6H12) Heteronuclear Multiple Bond Connectivity hexamethylphosphoramide Heteronuclear Multiple-Quantum Coherence light radiation high pressure (performance) liquid chromatography high resolution mass spectrum hertz infrared coupling constant a tissue culture cell line derived from human carcinoma of the nasopharynx wavelength  LD 5 0 LDA loge  liter a tissue culture cell line derived from mouse leukemia lithium aluminum hydride median lethal dose lithium diisopropyl amide the log of extinction coefficient  LTA LRMS(orMS) m M M+ m-CPBA Me mg MHz min mL mmol  lead tetraacetate low resolution mass spectrum multiplet molar molecular ion mtfta-chloroperbenzoic acid methyl milligram megahertz minute milliliter millimole  L L-1210 LAH  mp MSNA0.5K0.5 H m/z V  NBS nD nm NMR NOE [O] P-388 PCC Ph ppm PRD2C0100  PRI2C0100 PRL-4 RCF RCP rpm r.t. s sept septd SINEPT t Td TFA TFAA THF Tl  melting point MS medium of Murashige and Skoog supplemented with naphthaleneacetic acid (NA, 0.5 mg/L) and kinetin (K, 0.5 mg/L) micro (10"6) mass to charge ratio frequency Af-bromosuccinimide refractive index nanometre nuclear magnetic resonance nuclear Overhauser effect oxidation a tissue culture cell line derived from mouse leukemia pyridinium chlorochromate phenyl parts per million PRL-4 medium of Gamborg and Eveleigh supplemented with 2,4dichlorophenoxyacetic acid (D, 2 mg/L) and coconut milk (Co, 100 mL/L) PRL-4 medium of Gamborg and Eveleigh supplemented with indole 3-acetic acid (I, 2 mg/L) and coconut milk (Co, 100 mL/L) standard cell culture medium developed by Gamborg and Eveleigh relative centrifugation force resuspended cell pellet revolutions per minute room temperature singlet septet septet of doublets selective insensitive nucleus enhancement by polarization transfer triplet tripdiolide trifluoroacetic acid trifluoroacetic anhydride tetrahydrofuran triptolide  TLC TMS Tosyl (Ts) Tosylate TRP4a UV  thin layer chromatography tetramethylsilane para-toluenesulfonyl (the abbreviation tosyl is employed in the text while Ts is employed in structures) para-toluenesulfonate a cell line of plant cell culture developed from Tripterygium wilfordii ultraviolet  xxvi  Notes  The numbering system used throughout this thesis is that used by contemporary natural products chemists and is illustrated below.  19  18  abietane  The abietane skeleton is designated as 18-nor if C18 is absent or if a double bond is present at C3 or C4. The rearranged abietane skeleton is designated as lS(4-^3)abeo.  18-nor  18(4->3)aZ>eo  xxvii  Acknowledgements I wish to express my gratitude to my research supervisor, Professor James P. Kutney, for his unending guidance and support throughout the course of my research, and in the preparation of this thesis. I would like to extend my gratitude to Mijo Samija, Drs. Francisco Kuri-Brena, YongHuang Chen, Carles Cirera, Yasuaki Hirai, Masazumi Miyakoshi and Jacques Rouden for their help, invaluable discussions and suggestions during my research and in the preparation of this thesis. Acknowledgements are also due to other members of Professor Kutney's group, past and present, especially Nikolay Stoynov, Radka Milanova and Kai Li, for their help and contributions. I would like to express my thanks to secretary Pat Miyagawa. I am also thankful to Dr. Shi-Chang Miao, Edward Koerp and Fang-Ming Kong for helpful discussions during my study and research. I am very grateful to Gary Hewitt, David Chen and Fay Hutton of the Biological Services for their helpful discussions and for their guidance and assistance in the preparation of the cell cultures. The expertise, help and encouragement from the staff of the NMR services are highly appreciated. The staff of mass spectrometry, Mr. P. Borda of microanalytical services, Dr. Steve Rettig of X-ray diffraction analysis services, the faculty and staff of the Chemistry Department are also appreciated. I am very grateful to Edward Koerp, Dr. Reto Riesen, Mijo Samija, Dr. Jacques Rouden, Gary Hewitt, Ted Herrington and Dr. Phil Gunning for their meticulous proofreading of this thesis and invaluable suggestions. Special thanks are due to Ms Xiao-Shu Zhu (Nanjing, China), for her help in collecting the latest Chinese literature during the preparation of this thesis. Scholarship and financial assistance from the State Education Commission of the People's Republic of China and the University of British Columbia are gratefully acknowledged. Finally, I would like to express my heartfelt gratitude to my mother and late father for their inspiration, support and encouragement in the course of my education. I am also deeply indebted to my wife for her patience, support and encouragement.  1 CHAPTER 1 1.1  GENERAL INTRODUCTION  The Plant Tripterygium wilfordii and Related Species  Tripterygium wilfordii Hook. f. (Hooker filius) belongs to the genus Tripterygium, a group of woody vines and shrubs of the Celastraceae family. Tripterygium has four species in China: 1 " 3 Tripterygium  wilfordii Hook, f., Tripterygium  hypoglaucum  (Level.) Hutch,  Tripterygium forrestii Dicls, and Tripterygium regelii Sprague et Takeda. T. wilfordii is usually found in large mountainous areas of south-eastern and southern China, but mainly in south-eastern China.4,5 As a wild plant, it grows in the shady, humid woods and bushes on the hills, in the valleys and near the streams. It is a perennial twining vine of about 2-3 meters in height. The reddish brown twigs have longitudinal ridges and are covered with fine, brown hairs. Its egg-shaped leaves are alternately grown with small serrated edges. The plant blooms in May and June. The small white flowers have five petals, five stamens and a triangular ovary which later develops into a seed that has three jutting longitudinal wings.4 The plant T. wilfordii is named Lei Gong Teng (Thunder God vine) in traditional Chinese medicine. It is also called Mang Cao (Rank grass), San Leng Hua (ThreeWing flower) or Huang Teng (Yellow vine) among the local people, apparently based on the physical appearance of this plant. T. hypoglaucum appears in the same areas as T. wilfordii, but mainly in southern and south-western China. It is distinguished from T. wilfordii by virtue of larger leaves and flowers, and the leaves have white powder on the back.5 T. forrestii also grows in southwestern China,3 but this species was not recorded in the early botanical books.5'6 T. regelii, is mainly found in the mountainous areas or near the forests in the north-eastern part of China and Japan. It is also called Dong Bei Lei Gong Teng (North-Eastern China Thunder God vine) or Hei Man (Black vine).  2  Due to its vast geographical distribution and potent pharmacological activities, T. wilfordii is the predominant species of Tripterygium that is utilized in Chinese herbal medicine, and of the four species it is also the most studied. T. hypoglaucum is also commonly used in Chinese folk medicine for treatment of some diseases.  1.2  T. wilfordii as Traditional Chinese Medicine: History, Current Clinical Applications and Interests  The traditional Chinese medicine, or Chinese materia medica, classifies medicinal substances into three general groups: those which originate from either botanical, zoological or mineral sources. Because those obtained from botanical sources constitute the majority of the traditional Chinese medicine, traditional Chinese medicine is also referred to as Chinese herbal medicine. The use of herbal medicine in China to treat illness and disease can be traced back thousands of years. Some recordings of herbal medicines and their medicinal properties have been obtained from inscriptions found on unearthed bones and tortoise shells that date back to about 16th-11th century B.C.. However, little is known about the practice of medicine prior to the writing of a book Huang Di Nei Jing {Yellow Emperor's Inner Classic, also known as Inner Classic), compiled by unknown authors between 200 B.C. and 100 A.D.. 7,8 This book is the oldest major Chinese medical text extant, which laid the theoretical and philosophical foundations for traditional Chinese medicine. In addition to its emphasis on theory and philosophy, this book also includes 12 prescriptions, with a total of 28 medicinal substances noted.  However, from the perspective of herbal medicine, the Inner Classic is not a  particularly important document. A more specific, systematic and inclusive work, though still quite primitive from today's point of view, is known as Shen Nong Ben Cao Jing {Divine Husbandman's Classic of the Materia Medica, or The Saint Peasant's Scripture of Materia Medica).  This work was compiled by unknown authors in the Later Han dynasty (25-220  3  A.D.). Legend ascribes the authorship of this book to the mythical figure Shen Nong (Divine Husbandman, or Saint Peasant), who, in addition to introducing agriculture and animal husbandry, was also believed to have tasted "the hundred herbs" himself and thus, is viewed as the founder of Chinese herbal medicine. The writing of this book is generally regarded as the start of the historical tradition of Chinese herbal knowledge. In this book, there are 364 entries of medicinal substances of botanical (252 entries), zoological (67 entries) and mineral (45 entries) origins.8 One of the herbs recorded in Shen Nong Ben Cao Jing is T. wilfordii (as Rank grass). It was noted that T. wilfordii could be used for the treatment of fever, chills, edema and carbuncle. In the early times, Chinese farmers used the powdered root of this plant (the most toxic part) to protect their crops from chewing insects, which earned the plant a name as Cai Chong Yao, meaning a pesticide for vegetable insects. There are still studies on its application as a natural means of pest control in recent years.9,10 It was also used as folk medicine for the treatment of cancer and inflammatory diseases.11 Since the late 1960's, the Chinese medical community has increasingly used this plant's extracts to treat rheumatoid arthritis, chronic hepatitis, chronic nephritis, ankylosing spondylitis and various skin disorders with quite encouraging results.12"14 In Chinese folk medicine the plant is used directly with little preparation. For example, in the treatment of rheumatoid arthritis or skin rashes, the fresh roots and leaves are simply smashed and ground up, and the resulting paste is then applied to troubled areas. In clinical treatments of rheumatoid arthritis and lepra reactions, a decoction is used. Plants are usually collected in summer or early autumn. Once dried, the root xylem is obtained after removal of two layers of cortices which are suspected to contain most of the plant's toxic substances. Isolated xylem is then cut into small pieces and gently boiled in water for several hours. The water extract, or decoction, may then be administered orally for a certain period of time, and the symptoms are usually soon alleviated.4  4  Since the late 1970's, the most common form of T. wilfordii preparations used in the clinical treatment of rheumatoid arthritis and some skin disorders are tablets derived from the refined extract or the so-called total multi-glycosides of T. wilfordii (GTW). The dried root xylem is cut into small pieces and extracted with water. The water extract is then extracted with chloroform, and the resulting chloroform solution is concentrated, and then column chromatographed to produce the so-called GTW. About 25 g of xylem yields 1 mg of GTW and each tablet contains 10 mg of the GTW. The term "multi-glycosides" may not be a technically correct term for this preparation. It only implies that the prepared form contains a number of glycosides in addition to other constituents, and does not mean that the active principle(s) must necessarily be a glycoside(s).11 Although some American agricultural researchers came across this plant while searching for. insecticidal alkaloids in the 1950's (vide infra), Western interests in this plant were not aroused until 1972. In the study of the antileukemic activity of maytansine, an active principle from Maytenus ovatus, a plant of the same Celastraceae family, Kupchan and coworkers isolated diterpenes triptolide (Tl, 1) and tripdiolide (Td, 2) from an extract of T. wilfordii. Triptolide and tripdiolide showed strong antileukemia and antitumor activities in pharmacological study.15 Since then, broader efforts have been made in research related to this plant.  R = H,  triptolide (1)  R = OH, tripdiolide (2)  In recent years in the course of treatment of rheumatoid arthritis, dermatoses and during the ongoing evaluation of its general toxicity, GTW and some extracts of T. wilfordii have been found to have immunosuppressive activity.12 Additionally, reversible antifertility activity in  5  male rats and in men have been observed after oral administration at a dose level that shows no apparent side effects or signs of toxicity.11 This finding, though still under study, may provide an alternative for male contraception, and thus has triggered enormous research interests in China. Most recently, some constituents isolated from this plant have been found to show antiHIV activity.16'17 T. wilfordii, a medicinal plant used as an ancient remedy for thousands of years in China is now attracting greater attention not only in China, but also in other parts of the world.  1.3  Phytochemistry of T. wilfordii and Related Species  As a result of the wide pharmacological activities Tripterygium plants have shown, isolation of the active constituents from these plants started in the early 1930's in China. Chou and Mei isolated a red triterpene, named as tripterine, from the petroleum ether extract of the root bark. 18 The red triterpene later proved to be identical with celastrol isolated from Celastrus scandens L by the American scientist Gisvold.19,20 In searching for the insecticidal components from the root of T. wilfordii, Acree and Haller reported the isolation of an alkaloid, wilfordine, in 1950.21 This alkaloid was soon proved to be a mixture by Beroza,22 and he isolated several more in the following years {vide infra). The most interesting discovery was made by Kupchan and co-workers in 1972, when the antileukemia and antitumor diterpenes, Tl (1) and Td (2), were isolated. In recent years, the attention paid to the immunosuppressive and antifertility activities of Tripterygium plants has generated even greater interest, especially in China, in the isolation of active components from these plants. So far, about 140 natural products, most of them alkaloids, sesquiterpenoids, triterpenoids and diterpenoids, have been isolated from Tripterygium plants.  6  Alkaloids In 1951 and 1952, Beroza isolated five insecticidal alkaloids, wilforine, wilfordine, wilforgine, wilfortrine and wilforzine, from the root of T. wilfordii.23'25  Since then, a total of  fourteen macrocyclic lactone alkaloids (Table 1.1) and three spermidine alkaloids have been isolated from Tripterygium plants (Table 1.2). OAc OAc/ OAc AcCV A I A ^OCOPh  AcO  15  OAc OAc/ OAc AcO,. A A ^OCOPh  17 R = trans -PhCH=CHO 18 R = p-furanoyl 19 R = PhCO 16  7  Table 1.1  Macrocyclic Lactone Alkaloids from Tripterygium Species  Compound  Rl  R2  R3  R4  R5  Mol. formula  Species TW,23 TH 26 TW,23 TH 27 TW,24 TH 26 TW, 24 TH 26  3  wilfordine  benzoyl  OH  Ac  Ac  OH  C43H49NO19  4  wilforine  benzoyl  H  Ac  Ac  OH  C43H49NO18  5  wilforgine  p-furanoyl  H  Ac  Ac  OH  C41H47NO19  6  wilfortrine  pVfuranoyl  OH  Ac  Ac  OH  C41H47NO20  7  wilforzine wilforidine wilformine (euonine) wilfornine wilforjine 1-desacetylwilfordine 1-desacetylwilfortrine neowilforine isowilfordine forrestine  benzoyl H  H Ac Ac  Ac Ac Ac  OH OH OH  C41H47NO17 C36H45NOi8  Ac  H OH H  C38H47NO18  TW 25 TW 28 TW 29  nicotinoyl H benzoyl  H H OH  Ac Ac Ac  Ac Ac H  OH OH OH  C42H48N2O18 C36H45NOi7 C4iH 47 NOi 8  TW 30 TW31 TW 32  P-furanoyl  H  Ac  H  OH  C39H45NO19  TW 32  benzoyl  H  Ac  Ac  H  C43H49NO17 C43H49NO19 C43H49NO19  TW 33 TW 34 TF 35  8 9 10 11 12 13 14 15 16  TW = T. wilfordii; TH = T. hypoglaucum; TF=T. forrestii.  Table 1.2  Spermidine Alkaloids from Tripterygium Species  Compound 17 18 19  celacinnine celafurine celabenzine  TW = T. wilfordii  R  Mol. formula  frans-PhCH=CHCO P-furanoyl  C25H31N3O2 C21H27N3O3  benzoyl  C23H29N3O2  Species T\y36>37 1^36,37 T\y36,37  8  Sesquiterpenoids Since the first isolation of sesquiterpene polyesters from T. wilfordii in 1987 by Y. Takaishi,38 more than forty sesquiterpenes of these types have been isolated. Most of them were obtained from T. wilfordii var. regelii by Takaishi's group. These types of sesquiterpene polyesters contain a dihydroagarofuran skeleton (Table 1.3).  R3  dihydroagarofuran skeleton  Table 1.3  R4  20-24  Sesquiterpene Polyesters from Tripterygium Species3  Compound  Rl  20 triptofordin A H 21 triptofordin B OH 22 triptofordin C-l OAc 23 triptofordin C-2 OAc 24 regilidine ONic  R2  R3  R4  R5  Mol. formula  Species  H H OBz OBz H  OCin OBz  Bz Bz Ac Ac Bz  H H =0 P-OH  C31H36O6 C29H34O7 C33H36O11 C33H34O11  TW v.R38 TW v.R38 TW v.R38 TW v.R38  H  C35H37NO8  OBz OBz OBz  TW v. R = T. wilfordii var. regelii; TR = T. regelii. Ac = acetyl; Bz = benzoyl; Cin = Jrans-cinnamoyl; Nic = nicotinoyl.  OR 2  25-33  OAc  34-43  TR  39  9  Table 1.3  Sesquiterpene Polyesters from Tripterygium Species (continued)  Compound  Rl  R2  R3  R4  triptofordin D-l OAc =o OAc Cin triptofordin D-2 OAc P-OAc OAc Cin triptofordin E OAc =o OAc Bz triptofordin F-l OH a-OAc OAC Cin triptofordin F-2 OH a-OBz OAc Ac triptofordin F-3 OAc a-OAc OAc Bz triptofordin F-4 OH a-OH OAc Cin triptofordinine OAc P- OAc Cin A-l ONic 33 triptofordinine OAc P- OAc c-Cin A-2 ONic  25 26 27 28 29 30 31 32  R5  Mol. formula  Species  H H P-OAc P-OAc P-OAc P-OAc P-OAc  C35H38O11  H  C35H40O12 C41H43O12  TW v.R40 TW v.R40 TW v.R40 TW v.R41 TW v.R41 TW v.R41 TW v.R41 TW v.R42  H  C41H43O12  TW v.R42  C37H42O12 C35H38O13 C37H42O13 C35H40O13 C37H42O14  TW v. R = T. wilfordii var. regelii; TR = T. regelii. Ac = acetyl; Bz = benzoyl; Cin = trans-cinmmoyl; Nic = nicotinoyl; c-Cin = cw-cinnamoyl.  Table 1.3  Sesquiterpene Polyesters from Tripterygium Species (continued)  Compound 34 35 36 37 38 39 40 41 42 43  triptogelinE-1 triptogelin E-2 triptogelin E-3 triptogelin E-4 triptogelin E-5 triptogelin E-6 triptogelin E-7 triptogelin E-8 triptogelin G-l triptogelin G-2  R_| p-0(2-MeBu) P-OAc p-0(2-MeBu) =0 P-OCO(CH2)2CH3 p-OCOCH(CH3)2 p-OCO(CH2)2CH3 p-OCOCH(CH3)2 H H  R2  Mol. formula  Species  Bz Bz Cin Bz Bz Bz Cin Cin Cin Bz  C29H40O7  TW v. R43 TW v. R43 TW v. R43 TW v. R43 TW v. R44 TW v. R44 TWv.R 4 4 TW v. R44 TW v. R43 TW v. R44  C26H34O7 C31H42O7 C24H30O6 C28H38O7 C28H38O7 C30H40O7 C30H40O7 C26H34O5 C24H32O5  TW v. R = T. wilfordii var. regelii. P-0(2-MeBu) = P-OCOCHCH3CH2CH3; Ac = acetyl; Bz = benzoyl; Cin = frara-cinnamoyl.  10  OR4  OR2  R^O  44-50  Table 1.3  51-63  Sesquiterpene Polyesters from Tripterygium Species (continued)  Compound 44 triptogelin D-l 45 triptogelinF-1 46 triptogelin F-2 47 triptogelin C-l 48 triptogelin C-2 49 triptogelin C-3 50 triptogelin C-4 51 triptogelin B-l 52 triptogelin B-2 53 triptogelin A-1 54 triptogelin A-2 55 triptogelin A-3 56 triptogelin A-4 57 triptogelin A-5 58 triptogelin A-6 59 triptogelin A-7 60 triptogelin A-8 61 triptogelin A-9 62 triptogelin A-10 63 triptogelin A-11  Rl  R2  R3  R4  OAc H H OAc OAc a  Bz Bz Cin Bz Bz Bz Bz H  H ONic OAc OAc ONic OAc ONic Bz Bz Bz Bz Bz Bz b  OAc H H H H H H Bz Bz Bz Bz Bz Bz Bz Bz Bz Bz Bz Bz Bz  OH H H P-OBz P-OH P-OH =0 P-OBz P-OBz P-OH P-OH c P-OBz P-OBz  Nic Bz Bz H H Nic Bz H H Bz Nic Bz .  Nic Nic H Nic Bz Ac  Mol. formula  Species  43 C28H36O9 TW v.R 44 C30H35NO7 TW v.R C28H36O7 TW V.R44 C28H36O9 TW v.R45 45 C32H37NO9 TW v.R C31H42O9 TW v.R45 43 C30H35NO8 TW v.R 46 C31H36O8 TW v.R  C37H39NO9 C45H44O11 C38H40O10 C31H36O9 C31H34O9 C42H47NO11 C44H43NO11 C30H35NO9 C24H32O8 C43H49NO11 C44H43NO11 C40H42O11  TW v.R46 TW v R46 TW v R46 TW v R46 TW v.R46 TW v.R47 TW v.R47 TW v.R47 TW v.R47 TW v.R47 TWv.R 45 TW v.R45  TW\'. R = T. wilfordii var. acetyl; Bz = benzoyl; Cin= franj-cinnamoyl; Nic =nicotinoyl. regelii; Ac = 1 b a 49 . R = OCOCHCH3CH2CH3; 57: R3 = COCHCH3CH2CH3; c 61: R1 = OCO(CH2)4CH3.  11  Triterpenoids: Triterpene tripterine, or celastrol, was the first constituent isolated from T. wilfordii by Chou and Mei in 1936.18 Presently, more than thirty triterpenoids have been separated from Tripterygium plants (Table 1.4). Except zeorin, these triterpenoids have friedelane, oleanane, or ursane skeletons. Table 1.4  Triterpenoids from Tripterygium Species  Compound 64 65 66 67 68 69 70  celastrol (tripterine) pristimein 3-hydroxy-25-norfriedel-3,1(10)-dien-2-one30-oic acid 3,24-dioxo-friedelan-29-oic acid polpunonic acid orthosphenic acid salaspermic acid  Mol. formula  Species  Q29H38O4 C30H40O4 C29H42O4  TW,18 TR48 TR 48 TR49  C30H46O4  TW 50 TW 51 TW, 52 TR 53 TW,54 TR53  C30H48O3 C30H48O5 C30H48O4  TW = T. wilfordii; TR = T. regelii; TH = T. hypoglaucum.  COOR  friedelane skeleton %vCOOH  HO  64 R = H 65 R = CH3  +COOH  67 R = CHO 68 R = CH3  12  Table 1.4  Triterpenoids from Tripterygium Species (continued)  Compound  Mol. formula  Species  71  wilforlide A  C30H46O3  TH,55 TW,56 TR 57  72  wilforlide B  C30H44O3  73 74 75 76  regelide 3[3-hydroxy-olean-11,13( 18)-diene  C30H42O3 C30H48O C30H46O3 C30H48O3  77 78 79 80 80-A 80-B 81 82 83 84  triptotriterpenic acid A triptotriterpenic acid B triptodihydroxy acid methyl ether 3-acetoxyoleanolic acid oleanolic acid 3-oxo-olean-12-en-29-oic acid regelindiol B regelin D  C30H48O4 C30H48O4 C31H50O4 C32H50O4 C30H48O3  hypoglauterpenic acid 3(3,15(3-dihydroxy-A12-oleanen-28-oic acid  C30H46O3 C30H48O4  TW, 56 TH, 58 TR 57 TR53 TR53 TW59 TW, 54 TR,53 TH 60 TW,61 TH 50 YW52'62 TW51 TH, 55 TR 53 TH 58 TH 63 TR 64 TR 64 TH 58 TW 65  triptotriterpenoidal lactone A 3-epikatonic acid  TW = T. wilfordii; TR = T. regelii; TH = T. hypoglaucum.  C30H46O4 C31H50O4 C31H48O4  13  ,0 ,0  71 72  oleanane skeleton  73  R = B-OH, a-H R= 0  HO  74  75  COOH  HO—v. .^COOCH3  HO'  76 77 78  R = H2 R = a-OH, B-H R = B-OH, a-H  79  14 ^COOCH 3  "'OH  COOH  80 R = p-OAc 80-A R = p-OH 80-B R = O  81 82  R = (3-OH R= 0  COOH  COOH  HO'  83 Table 1.4  Triterpenoids from Tripterygium Species (continued) Compound  85 86 87 88 89 90 91 92 93 94  84  regelindiol A regelin triptotriterpenic acid C (tripterygic acid A)  Mol. formula  Species  C31H50O4 C31H48O4 C30H48O4  TR,64 TH 60 TR,66 TH 67  regelinol regelin C 2oc,3a,24-trihydroxy-A12-ursene-28-oic acid  C31H48O5  hypoglaulide zeorin glut-5-en-2(3, 28-diol ursan-3|3,5a-diol  C30H44O3 C30H52O2 C30H50O2  TW = T. wilfordii; TR = T. regelii; TH = T. hypoglaucum.  C33H50O6 C30H48O5  C30H52O2  T W j 62,68  TH 67 TR 66 TR 64 TR 49 TH 67 TR 69 TW 70 TW 70  15  COOR  '"OH  ursane skeleton  85  R = B-OH, a-H, Rl = CH 3  86 87  R = O, Rl = CH 3 R = B-OH, a-H, Rl = H  COOCH3  COOH  90  ''/'OH  91  92  16  CH2OH  HO  93  94  Diterpenoids: Some of the most interesting compounds from Tripterygium belong to the diterpene family. In addition to their strong antileukemic and antitumor activities, Tl (1) and Td (2) are the first reported natural products containing the 18(4—>3) abeo-abietane skeleton and the first recognized diterpenoid triepoxides.15 Both Tl (1) and Td (2) were isolated from the plant in 0.001% yields. So far 36 diterpenoids have been isolated from T. wilfordii and from its related species. These diterpenes all contain the abietane-type skeleton except for tripterifordin (Table 1.5). It has been shown that triptriolide (100) and 12-epitriptriolide (102) are not artifacts resulting from the extraction processes, but may have been formed in the plants. 71,72 Tripchlorolide (103) was believed to be an artifact formed during the extraction processes.73'74  18(4->3)a£eo-abietane skeleton  95  R=H R = p-OH R = a-OH  17 Table 1.5  Diterpenoids from Tripterygium species Compound  Mol. formula  1 2 95  triptolide (Tl) tripdiolide (Td) triptonide  C20H24O6 C20H24O7  96 97 98 99 100 101 102 103 104  tripterolide triptolidenol 16-hydroxytriptolide triptetraolide triptriolide isotriptetraolide 12-epitriptriolide tripchlorolide 13,14-epoxide-9,ll,12trihydroxytriptolide tripdioltonide  C20H24O7  105  C20H22O6  C20H24O7 C20H24O7 C20H26O8 C20H26O7 C 2 oH 2 60 8 C20H26O7 C20H25CIO6 C20H26O7 C20H24O6  mp, °C  Species  TW, 15 TH 75 TW, 15 TH 63 TW,15 TH 75 225-228 TH 2 193-194 TW 76 232-234 TW 77 258-260 TW 31 TW 71 260-262 250-252 TW 78 268.5+1.0 TW 72 256-258 TW 73 268-270 TW 79 226-227 210-211 210-211  222-224  TW 79  TW = T. wilfordii; TH = T. hypoglaucum; TR = T. regelii.  OH  97  98  103  99 R1 = OH, R2 = H, R3 = B-OH 100 R'=R 2 = H,R3 = B-OH 101 R1 = H, R2 = R3 = B-OH 102 R1 = R2 = H,R3 = a-OH  18  104  Table 1.5  105  Diterpenoids from Tripterygium species (continued)  Compound  Mol. formula  mp, °C  Species TW,80 TH, 2 TR,2 TF 2 TW 80 TW 81  106  triptophenolide (hypolide)  C20H24O3  232-234  107  triptophenolide methyl ether  C21H26O3  152-154  108 109 110 111 112 113 114 115  triptonolide neotriptophenolide isoneotriptophenolide triptoditerpenic acid hypoglic acid triptonoditerpenic acid triptoquinonoic acid A triptoquinonoic acid C  C20H22O4 C21H26O4  214-215 189-191 185-187  C2lH26)4 C21H28O3 C21H28O4 C21H28O4 C20H24O4 C20H24O5  TW80  TW 82  -  ^£[83,84  219-224 189-191 182-183 202-203  TH 85 TH 86 TR,87 TH 88 TR 89  TW = T. wilfordii; TH = T. hypoglaucum; TR = T. regelii; TF = T. forrestii.  OCH3  106 107  R=H R = CH3  R! = OH, R2 = H R! = H, R2 = OH  19  HOOC  HOOC  111 112 113  Table 1.5  Rl = OCH3, R2 = R3 = H R1 = OCH 3 ,R 2 = H, R3 = OH R 1 = OH, R 2 = OCH3, R3 = H  121 122 123 124 125 126 127 128  R=H R = OH  Diterpenoids from Tripterygium species (continued)  Compound 116 117 118 119 120  114 115  triptonoterpene triptonoterpene methyl ether neotriptonoterpene triptonoterpenol 11 -hydroxy- 14-methoxydehydroabietane triptoquinonoic acid B triptoquinonal triptoquinone G triptoquinonol triptoquinondiol 3 -oxo-triptoquinonol tripterifordin wilforonide  Mol. formula  mp,°C  Species  C20H28O2 C21H30O3 C21H30O3  153-155 209-211 205-207 197-199 194-195  TW 76 TW, 76 TH 85 TW 90 TW,91 TH 92 TR89  212-213 127-128 165-166 183-184  TR87 TR 87 TW v. R93 TR 87 TR87  135-136 255-256 187-189  TR 87 TW, , 6 TH 8 4 TW 82  C21H30O4 C21H32O2 C20H26O4 C20H26O3 C20H26O5 C20H28O3 C20H28O4 C20H26O4 C20H30O3 C13H16O3  TW = T. wilfordii; TR = T. regelii; TW v. R = T. wilfordii var. regelii; TH = T. hypoglaucum.  OR1 abietane skeleton  116 R1 = R2 = R3 = H 117 R1 = CH3, R2 = OH, R3 = H 118 R1 = CH3, R2 = H, R3 = OH  20  HCX "OCH3  OCH3  O'  '—OH  119  121 122 123  120  Rl = COOH, R2 = H R ^ C H O , R2 = H Rl = COOH, R 2 = O H  OH  124 125  R = H2 R = p-OH, a-H  126  R=O  127  128  Other Compounds: Quite a few other types of compound have also been isolated from  Tripterygium  species, including dulcitol, 57,94 euonymine,26 |3-sitosterol, daucosterol, fumaric acid,50 (-)syringaresinol,95 /-epicatechin,94  maytenfolic  acid, 8 8  hydroxymethylanthraquinone,96 fatty acids,97 and carbohydrates.98  l,8-dihydroxy-4-  21  1.4  Pharmacological Activities of Extracts and Compounds from T. wilfordii and Related Species  Antileukemia and antitumor activities  In search of the antileukemia principles from T. wilfordii, Kupchan and co-workers found that the alcoholic extract of T. wilfordii exhibited significant in vivo activity against the L-1210 and P-388 leukemia in mice as well as in vitro activity against cells derived from human carcinoma of the nasopharynx (KB). Guided by the assays against KB, L-1210 and P388, they finally found that diterpene triepoxides Tl (1) and Td (2) were responsible for these activities.15 Tl (1) and Td (2) showed significant life-prolonging effects in mice afflicted with the L-1210 lymphoid leukemia at 0.1 mg/kg level and cytotoxicity (ED50) against KB cell culture at 10"3 (Xg/mL concentration." Triptolide (1) administered intraperitoneally at 0.2 mg/kg per day for six days markedly prolonged the survival time of mice with L-615 leukemia. 100 In addition to its antileukemic activity, Tl (1) also showed strong inhibitory effects on the colony formation of several human breast and stomach cancer cell lines (MCF-7, BT-20, MKN-45, MKN-7 and Kato-III) at magnitudes similar to those of the leukemia cell line (HL-60). This may suggest that Tl (1) might have a potential therapeutic value for some types of solid tumors. 101 A study with human HeLa cells showed that Tl (1) is an agent which is non-specific to cell cycle phases (a character shown by alkylating drugs, vide infra), but more sensitive to S phase cells.102 Triptolide (1) and tripdiolide (2) may be classified as alkylating anti-neoplastic agents based on their mechanism of action103 as shown by a number of studies conducted so far.99,104 Both the epoxide functionality and the oc,|3-unsaturated butenolide moiety have been shown to be important for the tumor-inhibitory activity. It was postulated that the 9,ll-epoxy-14(3hydroxy system is necessary for the antileukemia activity of Tl (1) and Td (2). Biological and chemical data are in support of a mechanism that involves intramolecular catalysis (by a  22  neighboring hydroxyl group on the opening of an epoxide by nucleophiles).105 Some plantderived tumor inhibitors may act via selective alkylation of thiol groups of key enzymes that are responsible for growth regulation. Intramolecular catalysis by the C14-hydroxyl group may assist in selective alkylation of the 9,11-epoxide by biological macromolecules" (Figure 1.1, PrSH = propanethiol). Triptonide (95), differing from (1) solely at C14, by virtue of a ketonic functionality instead of a (3 oriented hydroxyl group, showed no antileukemic activity in doses up to 0.4 mg/kg. 14-Epitriptolide (129), a minor variant of Tl (1) with an a oriented C14 hydroxyl group, and the thiol adducts 130 and 131, also showed no antileukemic activity up to 0.4 mg/kg. Additional evidence indicates that nucleophilic addition of thiols to the a,P-unsaturated system may also be involved in the tumor inhibitory properties of some plant-derived compounds.99,106 G. A. Berchtold et al. reported that an analog of Tl (1), 132, which did not have the a,|3-unsaturated butenolide moiety, failed to show antileukemic activity against P-388 lymphocytic leukemia.107 A more recent study showed that the epoxide ring opening by attack of nucleophiles took place at the less hindered 12,13-epoxide rather than at the more hindered 9,11-epoxide, as previously postulated.  X-ray analysis of the product from the reaction of Tl (1) with  propanethiol under identical conditions as those reported earlier confirmed that the thiol exclusively attacked at C12 from the convex side, thus opening the 12, 13-epoxide.104 Aside from diterpenes such as Tl (1) and Td (2), some triterpenes (regelin (86), regelinol (88) and wilforlide A (71)) have been shown to have antitumor activities.66 Some sesquiterpenes, particularly triptofordin F-2 (29) and triptogelin A-1(32), have recently been found to exhibit remarkable antitumor promoting activities both in vitro10S and in vivo.109  23  130 R = H 131 R = OH  129  Figure 1.1  Alkylation of Thiols by the Diterpene Triepoxides via Hydroxy1-assisted Epoxide Ring Opening  HO  132  Immunosuppressive and Anti-inflammatory Activities  One of the activities of some alkylating drugs is their immunosuppressive action.103 Rheumatoid arthritis and some skin diseases are autoimmune disorders in mechanism and constitute one type of inflammation. The major clinical applications of T. wilfordii as a Chinese herbal medicine are for the treatment of these two diseases. However, which particular  24  compound(s) is the main active principle is still under extensive study. The drug for most of the clinical trials is in a form of extract, such as GTW, and most of the pharmacological studies are also conducted on the extracts. There have been numerous reports of successful clinical treatments of rheumatoid arthritis and skin disorders since 1969, when T. wilfordii started to be rather widely used in China. 110 An early clinical trial with 133 consecutive cases of rheumatoid arthritis and ankylosing spondylitis patients showed that treatment with a tincture of T. wilfordii had remarkably relieved cardinal signs of joint inflammation, pain, swelling, and improved joint function.12 Similar successful results were also shown in the treatment of a wide variety of skin diseases, including allergic diseases and diseases with a mechanism possibly related to allergies.14 Several studies showed that the advantageous therapeutic effects of T. wilfordii included relatively rapid relief of the symptoms with high effective treatment rates. 13 ' 111 The drug was more potent than the conventional non-steroid antirheumatic agents such as salicylates, indomethacin, phenylbutazone, etc. 12 and can be used as a substitute for corticosteroids in the treatment of some skin diseases for patients who are steroid-dependent or steroid-resistant, or who have contraindications to steroids.14 In addition to clinical application of GTW in the treatment of rheumatoid arthritis and skin disorders, there were also reports of satisfactory treatment of patients with asthma.112 The mechanism of action of T. wilfordii , though still not clear, has received more attention in recent years.113"116 Its anti-inflammatory effect was demonstrated by its significant inhibition of acute agar-induced edema and histamine-induced capillary hyperpermeability in animals. 117 Additional inflammatory mediators such as prostaglandin E2 (PGE2), which are secreted by human peripheral blood mononuclear cells, were shown to be reduced by this drug in in vitro experiments. 118 The immunosuppressive activity of the drug was shown by its inhibitory effects on both the humoral and cell mediated immunities as monitored by the hemolysin assay, the splenic cell immuno-specific rosette assay and the active rosette assay.14,117 More detailed studies indicated that in vitro GTW inhibited antigen and mitogen  25  stimulated proliferation of human T (thymus-derived) and B (bone marrow derived) cells, inhibited T cell responses and interleukin-2 (IL-2) production by T cells, and suppressed immunoglobulin (Ig) production by B cells at the 0.08-1 |Xg/mL level. At these concentrations, GTW did not affect IL-2R expression by T cells or IL-1 production and antigen presentation by monocytes. 119 The immunosuppressive effect of GTW was further exhibited by its ability to inhibit the activation of T cells. The suppression of T cell proliferation is due to this drug's inhibitory effects on key facets in the cell cycle of the T cells. 113 ' 114 The above experiments also demonstrated that the immunosuppressive effects were reversible after cessation of exposure to the drug. As an immunosuppressive agent, GTW has significantly prolonged the mean survival time of cardiac allografts in rats, comparable with the well-known anti-rejection agent cyclosporine A. 120 The compounds which have been shown to be responsible for the anti-inflammatory and immunosuppressive activity include alkaloids, triterpenoids and diterpenoids. Isolated compounds and their pharmacological activities are summarized in Table 1.6. Celastrol (64) was shown to inhibit the proliferation of lymph cells,121 to inhibit the activities of IL-1, IL-2, the release of PGE2 and the antibody response. 122 ' 123 Tl (1) and Td (2) have both antiinflammatory and immunosuppressive activities. In animal tests, Tl (1) decreased the effect of the humoral-mediated immunity as monitored by the hemolysin test, but not the cell-mediated immunity as assayed by the graft versus host reaction and the tumor-concomitant immunity.100  Antifertility Activity  During the evaluation of the general toxicity of GTW, it was discovered that in hybrid mice, when both the male and female animals were simultaneously fed a laboratory chow containing GTW for 4.5 months, both the body weights and the birth rates decreased. In clinic practices that utilize crude extracts of T. wilfordii for the treatment of rheumatoid arthritis and other skin disorders, clinicians have observed the fact that a decrease in the testicular volume  26  might take place in a few patients after long term exposure to the drug. In some cases, necrospermia or azoospermia occurred, but libido and potency were normal in all subjects, and the seminal indices were found normal three months after cessation of the treatment.11 Detailed animal tests showed that male rats given GTW by gastric gavage at a dose of 10 mg/kg daily, 6 days a week for 8 weeks, all became infertile. The body weight gain, histology of various organs, serum testosterone level and the mating behavior were found normal. Significant decreases in the density and particularly the motility of the spermatozoa from the cauda epididymis were observed. Interestingly, the fertility of the treated rats began to recover 4 weeks after withdrawal of GTW and was fully restored one more week later. Similar results were also obtained from a comparative retrospective study on male patients with rheumatoid arthritis treated with GTW. These results suggested that the antifertility activity of the drug is very likely reversible. The effective dosage for antifertility is only about one third of that for treatment of arthritis and skin disorders, therefore the side effects could be lower. As to the action of the drug, it was believed that at the dose level, GTW mainly causes damage to the epididymal spermatozoa and to a lesser extent the spermatogenic cells.11  27  Table 1.6  Isolated Compounds and Their Pharmacological Activities  Pharmacological activities Compound  anti tumor  4 6  wilforine124 wilfortrine125  9 29 32 86  euonine (wilformine)125 triptofordin F-2 108 triptogelin A-l 109 regelin66 regelinol66 wilforlide A 66 triptotriterpenic acid A 61 triptotriterpenic acid (tripterygic acid)126 3p,15p-dihydroxy-A12-  88 71 77 87 84  antiimmuno. inflamm. suppres.  anti fertility  antiHIV  + + + +a +b  + + + C  + + +  65  78 76 69 70  oleanen-28-oic acid triptotriterpenic acid B 4 9 , 1 2 6 3-epikatonic acid53 orthosphenic acid126 salaspermic acid17  + +  +  anti-inflamm. = anti-inflammation; immuno suppres. = immunosuppression; +: active, -: not active, blank: data not available. a>° antitumor promotion activity.  The total alkaloids fraction127 and several isolated diterpene epoxides from the plant also showed antifertility activity.128 One of the diterpene epoxides, tripchlorolide (103), was found to have an antifertility potency 200 times stronger than that of GTW. Unlike GTW or Tl (1), which cause relatively broad damage, especially to the sperm head (a potential source of mutation), tripchlorolide (103) only causes damage to the spermatozoa in the epididymis without significantly inducing sperm head anomalies or affecting other related organs.129 This compound is now under more extensive evaluation.  28  Table 1.6  Isolated Compounds and Their Pharmacological Activities (continued)  Pharmacological activities  1 2 95 97 98 100 102 103 106 114 121 122 124 125 126 127  Compound  anti tumor  antiinflamm.  immuno suppres.  anti fertility  triptolide15-130 tripdiolide15,130 triptonide130'131 triptolidenol130 16-hydroxytriptolide130 triptriolide130 12-epitriptriolide72 tripchlorolide130 triptophenolide132 triptoquinonoic acid A (triptoquinone A) 93,133 triptoquinonoic acid B (triptoquinone F) 93 triptoquinonal (triptoquinone E) 93  + +  + + + +  + + + +  + + + +  +  + + + + +  triptoquinonol (triptoquinone D) 93 triptoquinondiol (triptoquinone C) 93 3-oxo-triptoquinonol (triptoquinone B) 93,133 tripterifordin16  •  antiHIV  -  + + +  +  + + + + + +  Another independent study was carried out by a group of British and Chinese scientists. Guided by bioassay-directed preparative high pressure liquid chromatography (HPLC) subfractionation of material extracted from both the GTW tablets and the native plants, this group has found that the active antifertility principles of T. wilfordii are Tl (1), Td (2) and closely related derivatives of this class of diterpene epoxides, including tripchlorolide (103), an artifact formed during processing of the plant material.74  29  Anti-fflV Activity  During the growing worldwide campaign against AIDS, people are again turning to nature for a cure. A team of researchers from the United States and China has found that the ethanol extract of the roots of T. wilfordii shows significant anti-HIV activity. Bioassaydirected fractionation of the active extract has led to the isolation and characterization of a new anti-HIV component, a kaurane-type lactone tripterifordin (127). This compound inhibited HIV replication in H9 lymphocyte cells with an EC50* of 1 |0,g/mL (6 \xM) and did not inhibit uninfected H9 cell growth at 15 uM. 16 This group also found GTW, the total glycoside of T. wilfordii, has significant anti-HIV activity. Bioassay-directed fractionation of GTW has led to isolation and characterization of a triterpene, salaspermic acid (70), as the anti-HIV principle from the chloroform-soluble fraction. Salaspermic acid inhibited HIV replication in H9 lymphocytes with an EC50 value of 5 jLig/mL (10 |iM), and it inhibited uninfected H9 cell growth with an IC50** value of 53 \iM. This compound also showed an inhibitory effect against HIV-1 recombinant reversetranscriptase-associated reverse transcriptase activity.  Salaspermic acid and its related  compounds are now under further anti-HIV evaluation.17  Side Effects and Toxicity  In the treatment of rheumatoid arthritis or dermatitis, the major side effects at regular clinical dose levels, i.e., a decoction from 15-25 g of root xylem per day or GTW 60-90 mg/day (1.0-1.5 mg/kg/day), are gastrointestinal disturbances which include nausea, vomiting, anorexia, epigastric burning sensation, xerostomia, diarrhea and constipation. Most of them will subside in the course of treatment and discontinuation of medication is usually not EC50: medium effective concentration. IC50: medium inhibitory concentration.  30  necessary. Leukopenia or thrombocytopenia may be found in a few patients, but the cell counts recover shortly after withdrawal of the drug. Other side effects include menstrual disturbances, oligospermia, azoospermia and a decrease in the size of the testis.11 The single dose toxicity (LD50) of GTW in mice was about 160 mg/kg. The repeated dose toxicity in rats at doses of 30, 60 or 120 mg/kg/day caused lethargy, and lower body weight growth, particularly in the large dose group with longer treatment time. Histological examination revealed damages in the seminiferous tubules, the endometrium and the myometrium. No other significant changes were found in the tested animals. Dogs fed with the drug at a dose of 10-15 mg/kg/day for 14.5 months decreased their food intake, and exhibited a decrease in their white blood cell count. 134  This demonstrated that the  pharmacologically active dosage is approximately six times less than a harmful dosage. A smaller dosage of 20 mg of GTW per day was sufficient to produce reversible infertility in human males. Immunosuppressive side effects were not seen in human subjects at this level while higher dosages (in treatment of arthritis and dermatitis) had produced an increase in secondary infections.11'12 Triptolide (1) elicits strong antileukemic activity in mice. Triptolide (1) at a 0.1 mg/kg level in vivo,15'99 or intraperitoneally at 0.25 mg/kg  10  ° greatly extended the survival time of  mice injected with leukemic cells. The LD50 for Tl (1) (intravenous injection of mice) has been determined to be 0.8 mg/kg for a single dose and 0.16 mg/kg/day over seven days with concurrent fatal degeneration of heart tissue and bone marrow.11 This shows that the toxic dosage is very close to its therapeutic dosage and thus triptolide has not been applied as an antileukemic pharmaceutical for human patients.104 The immunosuppressive and antifertility activities seem to overlap with each other in GTW and other isolated compounds from pharmacological studies. However, animal tests showed that the effective antifertility thresholds were 40-60 times lower than LD50, 5-28 times lower than the dosages for anti-inflammatory activity and 5-12 times lower than dosages for  31  immunosuppressive activity. Therefore, effective antifertility activity could be achieved without necessarily depressing the immune system.128 In another attempt to further reduce the toxicity caused by oral administration, researchers have attempted to formulate the extract of T. wilfordii into an injection solution for external use in the troubled area to treat rheumatoid arthritis patients.135  1.5  Total Syntheses of Triptolide (1)  Due to the strong antileukemic and antitumor activities shown by Tl (1) and Td (2), and their low yields in the plants (0.001%), numerous efforts have been made in searching for alternatives to obtain these materials in quantities for further pharmacological studies since the isolation of these two compounds in the early 1970's. The two major routes were via synthesis and plant cell culture technology {vide infra). The unique abietane-type triepoxide structure imposed a challenge to synthetic chemists and invoked their interests as well. There has been no report on the total synthesis of Td (2) (which was successfully obtained by plant cell culture technology, vide infra), probably because Tl (1) has similar activities and major structural features, or due to some difficulty in adding the extra hydroxyl group. Although there was some synthetic work on the construction of the butenolide moiety,136"138 or the triepoxide system in ring C 139,140 (the two major challenges towards the synthesis of Tl (1)), respectively, the total synthesis falls into two main classes, the synthesis of racemic Tl from small achiral starting materials and the synthesis of optically pure Tl (1) with the same stereochemistry as the isolated / -triptolide (1). The racemic synthesis adopted by Berchtold et al. was to synthesize the dihydronaphthalenone 133 as a starting material with the B/C ring fragment of the abietane skeleton (Scheme 1.1).107>141 Construction of ring A via alkylation and annulation of the naphthelenone provided a suitably functionalized tricyclic intermediate 136 for the construction of the ring C triepoxide  32  system and the butenolide moiety.  Annulation started by the alkylation of 133 with  iodobutyrolactone 134, followed by an opening of the lactone and oxidation to provide the desired intermediate 135. Aldol condensation then yielded the key tricyclic intermediate 136. Reduction of the aldehyde, acidic hydrolysis and rearrangement of the double bond completed the synthesis of the butenolide ring. Ring C construction was achieved by hydroxylation at C7 and subsequent conversion to the epoxy dienone 139 by periodate oxidation.  Further  epoxidation gave racemic triptonide (95), which, after reduction by sodium borohydride, yielded a 3:1 mixture of racemic 14-epitriptolide (129) and triptolide (1). van Tamelen and co-workers have devised a number of routes to the total synthesis of triptolide (1) and triptonide (95). A synthesis of racemic Tl (l) 1 4 2 involving fewer steps, was comprised of a construction of ring C onto an appropriate AB fragment that was derived from decalone 140 and that possessed a hydroxyl functionality at C14 (Scheme 1.2). An efficient Diels-Alder addition to the furan derivative 142, afforded 143. Ring A butenolide construction proceeded from the alkene intermediate 144 via introduction of a hydroxyl group at C3, rearrangement with thionyl chloride and conversion of the C19 ally lie chloride to the allylic alcohol 145. Addition of dimethylformamide dimethylacetal to the allylic alcohol 145, was followed by a carbene [2,3]-sigmatropic rearrangement to 146. Further elaboration yielded the key intermediate 107. A biogenetic-type synthesis of racemic Tl (1) consisted of the construction of a geranylgeraniol-type intermediate 153, and cyclization to the tricyclic skeleton 154. Appropriate functionalization at C3 and C4 led to facile conversion to the butenolide ring and yielded the key intermediate 107 (Scheme 1.3).143  33  k 1  |— 137 R1 = CH 3 , R2,R3 = O p . 108 R1 = H, R2,R3 = O L* 138 R1 = R 3 = H,R 2 = O H  95  1  R' = OH,R 2 = H  129 R1 = H , R 2 = O H Scheme 1.1  Synthesis of (±) Triptolide (1) via BC-»ABC Abietane Construction  a) NaH, DMF; b) Me2NH; c) C1-O3, pyridine, CH2CI2; d)neutral AI2O3, EtOAc; e) NaBH4, EtOH, 2N HCl; f) m-CPBA, CH2CI2; g) Et3N, CH2CI2; h) 2,4,6-trimethylpyridine, MeSC^Cl, DMF; i) H2, Pd-C, EtOAc; j) Cr03, HOAc; k) BBr3, CH2CI2,0°C; 1) NaBH4, EtOH; m) NaI04, MeOH; n) m-CPBA, CH2CI2; o) NaBH4, EtOH  145  146  *  Scheme 1.2  107  1  Synthesis of (±) Triptolide (1) via AB—»ABC Abietane Construction  a) CS2, 2,6-di-f-Bu-4-Me-C6H20Li, THF, Mel; b) (CH3)2=CH2, NaH, DMSO, -10°C; c) HC1 (aq)MeOH; d) LDA, HMPA, THF, TBDMSC1; e) CH2=CCC>2Me, PhH, 65-70°C; f) 5:1 MeOH-6M HC1; g) Mel, NaH, THF; h) MeLi, THF, -15°C; i) MeSC>2Cl, Et3N, CH2CI2; j) Li, NH3, THF, -78°C; k) mCPBA, CH2CI2; 1) LDA, THF; m) SOCI2, Et20, pyridine, 0°C; n) KOAc, DMSO, 75°C; o) NaOMe, NaOH; p) (MeO)2CHNMe2, xylene, A, 4A sieves; q) m-CPBA, CH2CI2; r) (Me3Si)2NLi, THF, 0°C; s) 1M HC1  35  4/  o  + OMe  o  |— 149 R1 = C0 2 Me,R 2 ,R 3 =0 ' p r 150 R ^ f t R ^ s O C L— 151 R1 = R2 = H,R 3 =OH  147  MeQ 2 C  154  155  OMe Me0 2 C°  ,  156 Scheme 1.3  107  Biogenetic-type Synthesis of Triptolide (1)  a) NaH, THF, 0°C; b) Ba(OH)2, H 2 0-Et20, 90°C; c) LAH, EfcO, 0°C; d) LiBr, PBr3, collidine, Et20, -40°C; e) ZnBr2, E Q O , 0°C; f) LiH, CH 3 COCH2C02Me, DMF, 75°C; g) SnCU, CH2CI2, 0°C; h) MeS02Cl, Et3N, CH2CI2, 0°C; i) wz-CPBA, CH2CI2; j) LDA, -78°C  36  The only non-racemic synthesis of triptonide (95) (which could be further elaborated to / -triptolide (1)) from / -dehydroabietic acid (157) was also devised by van Tamelen and coworkers, 144 employing a strategy of introducing the C14 hydroxyl group first, and then constructing the A ring butenolide moiety followed by further elaboration of the epoxide system on ring C. Nitration of dehydroabietic acid (157) gave the 12,14-dinitro compound 158, which was then subjected to catalytic hydrogenation to give a 12-amino-14-nitro compound 159. Subsequent diazotization, and substitution by iodide led to 160, which was then reduced to the amine 161. Diazotization in trifluoroacetic acid yielded the trifluoroaceate 162, which was then used to construct the ring A butenolide moiety. Curtius degradation of 162 provided the isocyanate 163, which was then reduced to secondary amine with lithium aluminum hydride and converted to tertiary amine 164 by refluxing with formaldehyde and formic acid. Oxidation of 164 with ra-CPBA gave the //-oxide, which underwent a Cope elimination in refluxing chloroform to give olefin 165. Oxidative cleavage afforded ketone 166, which, after quite a few steps, gave the important intermediate 171. Introduction of the hydroxyl group at CI followed by periodide oxidation led to the epoxy dienone 139, which was further elaborated to yield /-triptonide (95). Even though the last step to /-triptolide (1) was not completed in this case, in view of the reported reconstitution of the triptolide system by sodium borohydride reduction of 95, the above synthesis can be considered a synthesis of triptolide (1) as well. The yield from dehydroabietic acid (157)to 162 was reported to be 40%, 145 and the rest of the steps from 162 to 171 were calculated to be in the 0.68% yield range. Thus an overall yield of 0.27% from dehydroabietic acid (157) to 171 was achieved. The overall yield to Tl (1) could be even less.  NH 2  N02  C0 2 H  NCO  162  163  j,k  Scheme 1.4  166  Synthesis of Triptolide (1) from Dehydroabietic Acid (157)  a) HNO3-H2SO4, HOAc; 60%; b) H2, Pd/C, HOAc, CF3CO2H; 80%; c) NaN02, CF3CO2H; KI; d) Zn, HOAc, 65-70°C; e) NaN02, CF3CO2H; 40% from 157; f) SOCI2, PhH, DMF; g) NaN3, acetone, H2O; PhH, 100°C; 90% from 162; h) LAH, THF, refluxing; i) HC02H-aq HCHO; 50% from 163; j) m-CPBA, CHCI3, -20°C; k) CHCI3, refluxing; 80% from 164; 1) Os0 4 , NaI04, HOAc-dioxane-H20; 30%  38  OH  m  t,u  H O  166  |— 167 R ^ H , Rz=CH2OH, X \ X z = 0 ' P = 168 R ^ a R ^ C H j O a X ^ O a X ^ C ^ O C ^ P h q S " L - 169 R ^ A c , R2=CH2OH, X ^ O H , X2=CH2OCH2Ph  - 139  ~-~ 1  OHC  x  Scheme 1.4  |—171 R=Ac,X 1 =X 2 =H F=:172 R=Ac,X 1 ,X 2 =0 L * 1 3 8 R=H, X ! =OH, X2=H  Synthesis of Triptolide (1) from Dehydroabietic Acid (157) (continued)  m) LDA, HCHO, THF, -78°C; 50%; n) MeOC(CH3)=CH2, HOAc; o) PhCH20CH2U THF, -78°C; p) HC1, THF; 70% from 167; q) MeOC(CH3)=CH2, HOAc; r) Ac20, pyridine; s) HC1, MeOH; t) PCC, CH2CI2; u) o-C6H4(NH2)2, PhC02H, EtOH, HC1; 18% from 168; v) NaC102, HOSO2NH2, dioxane-H20; w) H2, Pd/C, EtOH; 100% from 170; x) C1O3, HOAC-H2O, 40°C; 20%; y) KOH, MeOH-H20; z) NaBH^ EtOH; 95% from 172  1.6  Plant Cell Culture Biotechnology  As more Tl (1) and Td (2) are required for detailed pharmacological studies or for potential clinical applications, isolation of these compounds from the plants would not be practical because of their low isolation yields (0.001%). The low isolation yields of Tl (1) and Td (2) and their interesting pharmacological activities were among the driving forces for the development of T. wilfordii cell cultures in the early 1980's by Professor Kutney's research  39  group. This technology provided a feasible alternative to obtain these valuable natural products in quantities much higher than from the plants. 146,147 At the present time, there are usually three general methods employed in the production of these biologically active natural products: 1.  Isolation from the intact plant;  2.  Total synthesis or semi-total synthesis;  3.  Isolation from cell cultures of the plant.  Traditionally (and now still in application), isolation from the intact plant has played an important role in the pharmaceutical industry.148 However, for large scale productions of these biologically active natural products for use as pharmaceuticals, various problems are inherent with the isolation technique. Usually, isolations from plants yield only minute quantities of the desired compounds (such as in the case of Tl (1) and Td (2)), and because of the complexity of the crude extracts, separation from co-occurring materials is often difficult, costly, and timeconsuming. Furthermore, the concentration of the desired compound may vary according to the time of plant harvest. Other constraints are incurred by the fact that the plants may grow very slowly, and environmental concerns (a growing factor in recent years) along with geopolitical constraints may make the required plant species actually unavailable. 149 For example, taxol is extracted from the bark of 50-60 year-old Pacific yew trees and as many as 12 trees are required to produce enough taxol to treat one patient. It has been estimated that there are only about 106 trees left in nature and current requirements are about 105 trees per year.150 Therefore, if the trees were harvested at the requirements we have now, this natural resource would disappear from our planet in 10 years. Developing efficient synthetic strategies and methodology has always been the goal for organic chemists in order to eliminate the dependence on the living plant as the source for those natural products, and great progress has been made to date. Efforts to synthesize Tl (1) have been carried out by a number of research groups, though its total synthesis suffers from the disadvantages of multiple steps involved and overall low yields (vide supra).  Moreover,  40  tripdiolide (2) has not been synthesized yet. This illustrates another frustration facing synthetic chemists and the pharmaceutical industry in general. The structural complexity inherent in these biologically active natural products demands multi-step syntheses, and thus has generated great interests and accounted for enormous development in organic synthesis to date. However, though many elegant syntheses have been accomplished in the laboratory, presently only some of them are finally applied in the commercial production of these compounds on a large scale.151 The methodology of plant cell culture development is not new, but is a rapidly expanding area of research that offers new, unique alternatives in acquiring the desired biologically active natural products.151-152 The interest in development of the science of in vitro tissue culture had been foreseen prior to the middle of the nineteenth century by promoters of the cell theory. One experimental approach was unsuccessfully attempted during the first years of the twentieth century, but success was only reached in 1912 with animal cells, and not until 22 years later with plant tissues. The technique of cell culture was rapidly exploited with animal cells, while in the case of plant tissue cultures, a long period of stagnation followed the initial establishment of basic methodology. After 30 years of relative indifference, thousands of scientists rediscovered this "new field" of plant biology, which currently is undergoing considerable expansion under the new name of biotechnology.152 It should be noted, however, that in the last two or three decades of fast development in this area, a substantial percentage of work has been directed to the use of plant cell cultures to study the more "biological" aspects of such cultures, such as investigating cell growth regulation, cell structure, cytodifferentiation, somatic embryogenesis, morphogenesis, and physiology.151 Fermentation technology with fungal and bacterial cultures has made dramatic advances particularly in the pharmaceutical industry, but similar technology with plant cells has not yet reached this level of application.151 Therefore, there has been a great need to explore  41  the potential of plant cell cultures for the production of plant-derived biologically active compounds. There are several advantages associated with plant cell culture biotechnology:151 1.  The desired compounds can be produced year-round under controlled laboratory conditions, assuring a steady supply of the materials without seasonal fluctuation;  2.  Cloning provides selected cell lines for optimized production of desired agents, and metabolic processes may be regulated or manipulated to maximize the production of the desired compounds in yields higher than in the wild-type plant;  3.  The cell cultures provide an excellent media for biosynthetic studies, biotransformation studies and even enzyme isolation.  These advantages are quite obvious and there are many examples. There have been some successes in the commercial production of secondary metabolites by plant cell cultures such as the red dye shikonin by cell cultures of Lithospermum erythrorhizon and the yellow anti-microbial dye, berberine, by a cell culture of Coptis japonica. There have also been a few notable successes in the elucidation of the biosynthesis of secondary metabolites by using cell culture technology, such as the discovery of the enzymatic steps required to synthesize berberine.150 In 1980, Kutney et al. first reported the production of tripdiolide (2) by a cell culture of T. wilfordii grown in modified B-5 and PRL-4 suspension media.  A yield of 0.003%  tripdiolide (2) was reported based on the dry cell weight, and thin layer chromatography (TLC) evidence for the presence of triptolide (1) was also reported. 146 A thorough program by Kutney et al, aimed at maximization of Td (2) yield by way of cell line selection and media optimization, resulted in the development of a callus cell line from a leaf explant of T. wilfordii, designated as TRP4a. 147 The callus was initiated on PRI2C0100 agar (PRL-4 medium of Gamborg and Eveleigh153 (without casein hydrolysate) which was supplemented with indole-3-  42  acetic acid (I) (2 mg/L) and coconut milk (Co) (100 mL/L)). After initiation, the callus was transferred and maintained on PRD2C0100 agar (PRL-4 medium supplemented with 2,4dichlorophenoxyacetic acid (D) (2 mg/L) and coconut milk (100 mL/L)). The cell line was selected for further development based on TLC and KB cytotoxicity activity analyses and on growth vigour.  Stock suspension cultures of TRP4a were initiated and maintained in  PRD2C0100 broth as a growth medium. Maintenance of the stock culture was carried out by subculturing of a 10% inoculum at 3 week intervals into fresh media. An extensive study of the growth parameters revealed that resuspension from the growth medium into MSNA0.5K0.5 broth (MS medium of Murashige and Skoog 154 supplemented with naphthaleneacetic acid (NA) (0.5 mg/L) and kinetin (K) (0.5 mg/L)) produced the highest levels of tripdiolide. After 35 days of incubation, tripdiolide (2) levels peaked at 4.0 mg/L, a level 36 times greater than that isolated from the whole plant (based on a dry cell weight of 10 mg/mL). Therefore, the development of plant cell culture technology has provided an additional and more effective route for the production of Tl (1) and Td (2). A study on the levels of various nutrients in the media also showed that Td (2) production was highest in MSNA0.5K0.5 medium containing 1650 mg/L of ammonium nitrate, 40 g/L of sucrose, and 880 mg/L of calcium chloride. The method for the production of the TRP4a cell line and growth of cell suspension cultures for Td (2) production is summarized in Figure 1.2. The TRP4a cell line has been a stable cell line since its establishment in the early 1980's and is continuously producing the important diterpene metabolites triptolide (1) and tripdiolide (2). Two other groups have developed suspension cultures of Tripterygium, but the maximum yields of Td (2) were relatively low (below 0.01%).155'156  43  Explant from the leaf  Callus grown on PRLJCOJOO agar  Callus maintenance on PRD2Coi0o a S a r  r  subculture  Stock suspension culture maintained in PRD2Co100 broth routine maintenance Cell resuspended and cultivated in MSNA0 5K() 5 production medium  Cells harvested and extracted at the end of the growth phase  Figure 1.2  Development and Routine Maintenance of the TRP4a Cell Line of T. wilfordii  In addition to triptolide (1) and tripdiolide (2), several other metabolites have been isolated from the TRP4a cell cultures. Most of the metabolites were isolated under the guidance of biological assay. Some were new compounds (174-176,183), while one (78) was later found in the T. wilfordii plant. Others (64, 68, 77, 184) have been isolated from Tripterygium  plants, but the rest have not yet been isolated from the  Tripterygium  plants. I46,157,158 These compounds include quinone methide-type compounds (64,173,173A), friedelene-type triterpenes (68,174), oleanene-type triterpenes (77, 78,175-180), a ursenetype triterpene (181), diterpenes (157,182,183) and phytosterols (184). There has not been detailed isolation work on diterpene metabolites from this cell culture.  44  Compounds isolated from TRP4a cell cultures:  COOH  64  173  COOH  173-A  68  COOH  HO  174  77 78  R = a-OH R = B-OH  „xCOOH  175 176  R = B-OH R = a-OH  177 178  R1 = R2 = H R1 = CH2OH, R2 = H  179  Rl = H,R 2 = OH  COOH  180  181  OH  Me0 2 C  157  dehydroabietic acid  182  OCH3  H<T  184  46  Manipulation of cell cultures to obtain higher production yields of desired secondary metabolites not only can be achieved by changing the media nutrients and growth parameters as we have seen in the development of the TRP4a cell line, but also may be accomplished by adding synthesized biosynthetic intermediates, or precursors, through biotransformations, or by elicitation.150 Early biotransformation studies with TRP4a cell cultures were focused on the formation of the butenolide moiety of the molecule with mixed results. 159,160 M. Roberts used isodehydroabietenolide (193) as a starting material and obtained several products with hydroxyl groups introduced at C7 and/or C2 positions, but no change to the C ring was observed.160 F. Kuri-Brena continued the work and started to use the C ring "activated" precursor, isotriptophenolide (194), for the biotransformation study in an attempt to bring some changes to the C ring. The results showed that only small amounts of the starting material were transformed to its methyl ether while most of the starting material was recovered.161 The detailed results will be given in Chapter 4 of this thesis. On another front, M. Samija of this group succeeded in increasing the yield of some triterpene acids, which are among those with biological activities, by elicitation of the cell cultures with fungal preparations.162  isodehydroabietenolide (193)  isotriptophenolide (194)  47  Biotransformation studies of TRP4a cell cultures by other members of this group showed that the cell cultures were capable of hydroxylating or epoxidizing compounds with totally different structures (e.g., tobacco cembranoids).163 From our previous work related to the TRP4a cell cultures, it can be seen that we have achieved our primary goal of obtaining triptolide (1) and tripdiolide (2) by cell culture in greater yields than from the plants, and we also have acquired some results and experience in biotransformation studies. It is important then, to take advantage of what we have achieved and to advance further.  1.7  Challenges and Current Research Trends  As was previously mentioned, the immunosuppressive and antifertility activities from the compounds isolated from Tripterygium overlap, and their toxicity is still relatively high. These side effects and toxicity, though not very serious, are definitely not desirable. These potential problems restrict the use of these compounds as a long term measure for male contraception, or for prolonged treatment of patients with rheumatoid arthritis or skin disorders, or as potential anti-rejection agents for organ transplantation. Therefore, looking for close analogs of this class of compound but with more defined pharmacological activity and less toxicity has been the focus of current research efforts in recent years. Major efforts that have been made include: 1.  Isolation of new compounds from native plants or their extracts, with a focus on Tl (1) and Td (2) closely related diterpene epoxides;  2.  Structural modifications of relatively readily available diterpene epoxides such as Tl (1); and  3.  Combination of organic synthesis with plant cell culture technology to obtain related novel diterpene epoxide analogs.  48  Several new, related diterpene epoxides have been isolated from extracts of Tripterygium  plants by Chinese researchers in recent years.  These epoxides include  triptolidenol (97), 16-hydroxytriptolide (98), triptetraolide (99), triptriolide (100), isotetraolide (101), 12-epitriptriolide (102), tripchlorolide (103), 13,14-epoxide-9,l 1,12-trihydroxytriptolide (104) and tripdioltonide (105) (structures etc., vide supra). Tripchlorolide (103) was isolated from GTW and was suspected to be an artifact since organochlorine compounds usually do not occur naturally in higher plants. This compound may have been produced from Tl (1) by the action of HC1 at some stage in the extraction process, for instance, during the extraction with chloroform, which usually contains traces of HC1. This hypothesis was confirmed by chemical correlation experiments by different groups. 73 ' 74 Treatment of Tl (1) with HC1 in acetic acid (0.4 N) at 0-4°C gave tripchlorolide (103), which could be converted back to Tl (1) by a brief reflux in ethanol under basic conditions.73  triptolide (1)  tripchlorolide (103)  Interestingly, compared with Tl (1), tripchlorolide (103) exhibited less toxicity without significantly compromising its activities.128,130 Another isolated compound, triptriolide (100), showed very low toxicity (LDso>250 mg/kg) but stronger anti-inflammatory activity.71 These findings have inspired people to modify the triepoxide system in order to find more effective compounds. Yu et al. prepared 9 derivatives (103,185-192) from Tl (1) and studied their structureactivity relationship. They found that the chlorohydrin 103 and bromohydrin (185) had similar  49  activities as Tl (1), but the chlorohydrin 103 was less toxic than Tl (1). However, the rest of the compounds (186-192) all showed much decreased activities. In view of the facile elimination of HC1 or HBr from 103 or 185 to form Tl (1), (note: they obtained 91.8% yield of 103 from Tl (1) and 88.0% yield of Tl (1) from 103 under mild conditions), they suggested that these two compounds might undergo an identical elimination step to form Tl (1) in biological systems, thus showing similar activity as Tl (1). Other derivatives that did not revert back to Tl (1) easily, or had the C14 hydroxyl group blocked (191), showed that their activities were greatly reduced. However, they did not offer any explanations as to why the toxicity of 103 was lower. 104 One fact that is clear from their results is that the C12, C13-epoxide and C14 hydroxyl groups likely play important roles in the biological activities of this class of compounds, and thus closely related analogs could be more effective.  >>*  O--  /k.  J^° > O w° H  0-—' 103 185 186 187 188  R = C1 R = Br R = SPr R = OAc R = OCH3  189 190  Rl == H,R2 = OH Rl == CI, R 2= H  50  OAc  191  192  The combined approach of chemistry and plant cell culture biotechnology by Professor Kutney's group has led to a number of results. The most recent and interesting results will be discussed in the following chapters of this thesis. Other new developments in recent years include the establishment of quantitative HPLC and high resolution gas chromatographic (GC) analyses of Tl (1) in prepared drug formulations or plant specimens.135,164,165 GC-MS analysis of volatile constituents from Tripterygium plants have also been investigated.97  1.8  Objectives and Strategies of the Present Investigation  In view of the challenges and current interests related to Tripterygium,  and our  advantage of having an established T. wilfordii cell culture, the present investigation is aimed at a systematic study in an attempt to: 1.  Explore the possibility of increasing Tl (1) and Td (2) yields by incubating appropriate synthetic precursors;  2.  Obtain Tl (1) and Td (2) analogs by a combination of organic synthesis and biotransformation;  3.  Isolate new diterpene metabolites from the cell cultures in order to understand the metabolism of these diterpenes and the possible link to the biosynthesis of Tl (1) and Td (2).  51  Previous results have shown that there is a great similarity in the spectrum of metabolites isolated thus far from TRP4a cell cultures to those isolated from the whole plant, suggesting that the diterpenes and triterpenes are metabolized in a similar way as they are in whole plants. Therefore, our strategy was to systematically isolate diterpene metabolites from the cell cultures (Chapter 2), combine the information we would obtain from the isolation and the results from previous biotransformation studies to design and synthesize appropriate precursors (Chapter 3), and then carry out biotransformation studies with these synthetic precursors (Chapter 4). By systematic isolation of these diterpenes, we could obtain some information about the metabolism of related diterpenes, which probably would give us some ideas about the biosynthesis of Tl (1) and Td (2). This would also help us in choosing appropriate precursors for biotransformation studies, and at the same time provide information as to the diterpenes naturally present in the cell culture in order to differentiate them from possible biotransformation products. In addition, isolated diterpene metabolites may have some biological activities as well. With respect to biotransformation studies, and at the time when this program was initiated, we were in a position to evaluate ring C "activated" precursors. It was of interest to have access to various ring C "activated" precursors, available by chemical synthesis, since such substances may increase the possibility of enzyme-catalyzed biotransformations to novel ring C functionalized diterpene analogs. These latter compounds may possess interesting pharmacological activities and would provide some information about the biosynthesis of Tl (1) and Td (2).  52  CHAPTER 2  ISOLATION AND STRUCTURE ELUCIDATION OF DITERPENE METABOLITES FROM PLANT CELL CULTURES OF T. WILFORDII  2.1  Introduction  As noted earlier, the unique triepoxide moiety of diterpenoids Tl (1) and Td (2) is the key element to their pharmacological activities. In addition to Tl (1) and Td (2), various other diterpenoids have been isolated from Tripterygium  plants. Some of them have a close  structural relationship with Tl (1) and Td (2), but they are not likely to be the biosynthetic intermediates of Tl (1) and Td (2) {vide supra ). In previous studies, our group had isolated some triterpenes but only a few diterpenes (including Tl (1) and Td (2)) from the TRP4a cell cultures. A main focus of the earlier studies was directed at isolation of pharmacologically active compounds. A detailed investigation aimed at more extensive isolation of diterpene metabolites, in order to obtain some insight into their metabolic processes, and their possible connections to the biosynthesis of Tl (1) and Td (2), was desirable. Cell culture technology provided us with a very powerful tool to accomplish our goal because it produces large quantities of fresh materials with constant quality, while isolations from whole plants are usually done on dried material, in which some components may have been lost or undergone some irreversible changes. As discussed previously, the diterpenoids isolated from T. wilfordii plant sources or the cell cultures are basically abietane or afreo-abietane type diterpenes. Most of them have various functionalizations on rings A and C, while some of them have functional groups on ring B as well. Characteristic signals of these functionalities, observable in lH NMR spectra, are signals of an AM or AMX system* in the aromatic region, a resonance between 8 4.5-5.0 A and M are much more strongly coupled with each other than with X for all relevant compounds discussed in this thesis. In some cases, the spin system may be classified as an AB or an ABX system; however, for the sake of convenience, the term AM or AMX system will still be used in those situations.  53  from methylene protons of the butenolide moiety, signals from protons on epoxide rings, and the readily recognizable isopropyl group signals. These distinct resonances provided a guide to locating these diterpenoid compounds during isolation studies. Extracts from the cell cultures were column chromatographed and all eluates were pooled into fractions based on their major spots on TLC plates. These fractions were briefly analyzed by *H NMR, and those which showed relevant signals were further purified.  2.2  Results and Discussion  T. wilfordii Plant Cell Cultures (TRP4a): Growth and Production  The T. wilfordii plant cell culture was first initiated in the late 1970's, and the TRP4a cell line was selected in the early 1980's, based on its ability to maximize the production of Td (2) and Tl (1). Since then, the culture has been grown and maintained according to the published procedure outlined in Plant Medica (1983).147 A minor change made in the past few years to the original protocol was that the subculturing interval was shortened to around 2 weeks from 3 weeks, because no significant changes in the growth and metabolism of the cell cultures had been observed with the shorter subculturing interval. Thus, the cells are maintained in PRD2C0100 medium (growth medium) with subculturing to new growth medium at 14 day intervals. At the time of subculturing, an inoculum (10%, v/v) of the cells are transferred to MSNA0.5K0.5 medium (production medium) for the production of Tl (1) and Td (2). All cell suspension cultures grown in conical flasks were incubated without illumination at 27 ± 1°C on a rotary shaker with a 7/8" throw and run at 140 rpm. The cell suspension cultures that were used in this study were grown in MSNA0.5K0.5 medium. Previous results showed that the growth of the TRP4a cells includes a lag phase (first 710 days), a growth phase (from 7-10 to 25-35 days) and a stationary phase (after 25-35 days). Td (2) and Tl (1) started to accumulate rapidly during the growth phase and leveled off when  54  cell growth reached the stationary phase. With the isolation of biosynthetic intermediates in mind, the cell culture was harvested near its stationary phase and utilized for the isolation study.  cell suspension cultures filtration through Miracloth™  broth  cells homogenization in EtOAc  saturation with NaCl, extraction with EtOAc  filtration, ,, separation  solvent removal broth extract  EtOAc layer  aqeouslayer saturation with NaCl, extraction with EtOAc '  '  '  solvent removal  cell extract  Figure 2.1  General Procedure for Cell Culture Harvesting  The general procedure for harvesting and extracting the cell culture is outlined in Figure 2.1. The cell cultures (5 x 550 mL) were harvested after growth in MSNA0.5K0.5 medium for 28 days (nD = 1.3330, pH = 6.29). The combined cell suspenion was filtered through Miracloth™, a coarse fibrous cloth which separates solid cell substances from liquid broth. The cells (wet weight 299 g) were frozen until the time of extraction. The broth was saturated with sodium chloride and extracted with ethyl acetate. The ethyl acetate solution was dried,  55  filtered, and the solvent was removed to give a brown colored broth extract (543 mg). The cells were thawed and homogenized in ethyl acetate. The resulting suspension was filtered through Celite and the filtrate was separated into an ethyl acetate layer and an aqueous layer. The aqueous layer was re-extracted with ethyl acetate and the ethyl acetate extracts were combined. Removal of the solvent yielded a dark colored cell extract (1.04 g).  Isolation of Diterpene Metabolites from TRP4a Cell Cultures  TLC analysis (toluene:chloroform:ethyl acetate:formic acid, 105:48:45:3, developed twice) of the broth and cell extracts exhibited similar chromatograms, except that the cell extract showed more of the faster migrating spots (less polar compounds) compared to the broth extract. The broth extract and cell extract were combined and extensive column chromatography on silica gel was performed. The chromatographic separation of the extract is depicted in Figure 2.2. After repeated column chromatography (c. c.) with various solvent systems, 29 compounds were isolated. Chromatographic purification of a previously isolated, crude Td fraction (49 mg) provided additional quantities of 209 and a small amount of 212 (0.2 mg). Since 212 was also suspected to be produced by the cell cultures, it is discussed here together with the other isolated compounds. The compounds isolated from the extract of the TRP4a cell culture are summarized in Table 2.1.  Combined Broth and Cell Extract (1.58 g) c. c. i) hexanes-EtOAc (6:4, 5:5, 3:7, 1:9) ii) EtOAc iii) MeOH  in  iv  i I nvn v  VI  I (264 mg) c. c. CH2Cl2-EtOAc (95:5, 90:5, 1:1)  IB  IA c. c. hexanes  IC  ID  c. c. i) hexanes ii) hexanes-EtOAc (97:3)  195 (2.6 mg)  IB1  IF  IG  c. c. i) CH2Cl2-EtOAc (97:3) ii) hexanes-acetone (9:1) iii) CH2Cl2-EtOAc (95:5)  108 (0.1 mg)  IB2  c. c. i) CH 2 Cl 2 -hexanes(3:7,l:l) ii) CH2C12  c. c. hexanes  I  I  195 (8.5 mg)  196 (17.5 mg)  (IB2ai)2  IE  IB2a  IB2b  c. c. hexanes-acetone (98:2)  •  EB2aii c. c.  hexanesacetone (97:3)  196 (1.5 mg)  ) • • • IBZam 2am c. c.  c. c. hexanesacetone (98:2)  198 (1.5 mg) hexanes-  IB2c c. c.  hexanesacetone (98:2)  199 (1.0 mg)  acetone ! (97:3)  197 (1.0 mg)  discarded  ure 2.2  Column Chromatographic Separation of Extract from TRP4a Cell Cultures  IE (14.9 mg)  IC(17.1mg)  c. c. hexanesEtOAc (85:15,80:20)  c. c.  hexanes-EtOAc (95:5,9:1)  I  I  200 201 202 116 (3.5 mg) (0.9 mg) (1.2 mg) (4.1 mg)  205 (1.2 mg)  c. c. hexanes-EtOAc (9:1,8:2,7:3)  i  c. c. CH2C12EtOAc (98:2)  '  202 (2.3 mg)  rID3  J  ID2  c. c. hexanesEtOAc (9:1)  1 .\  ID2a  193 (1.4 mg)  IE»2b c. c. CH2C12EtOAc (95:5)  2()3 (1.1 mg)  Figure 2.2  T  ID4  117 (12.3 m  117 (9.2 mg)  1  c. c. CH2C12EtOAc ,(95:5) 2134  (2.1 mg)  T  ID5  c. c. hexanesEtOAc (85:15)  ID 2c  IE2 IE3 118 108 (1.5 mg) (5.3 mg)  c. c. CH2C12EtOAc (99:1)  ID (49.2 mg)  IDl  I  IE1  ID6  ID7  c. c. CH2C12EtOAc (99:1)  c. c. CH2C12EtOAc (99:1)  193 (1.4 mg)  1 c. c. hexanesEtOAc (1:1)  107 107 (2.1 mg) (1.4 mg)  IG (8.8 mg) c. c. hexanes-EtOAc , (8:2)  11)9 (3.C)mg)  Column Chromatographic Separation of Extract from TRP4a Cell Cultures (continued)  II(125mg)  III (38.8 mg)  c. c.  c. c. CH2Cl2-EtOAc (95:5,9:1,8:2, 2:1, 1:2)  CH2Cl2-EtOAc (97:3,3:7,9:1,8:2,6:4,1:1)  IIA c. c. hexanesEtOAc (6:4)  im  nc  IID  106 (2.1 mg) 109 206 108 (2.3 mg) (4.0 mg) (2.9 mg)  106 (2.7 mg)  108 106 (l.lmg)(6.3mg)  V(85.1mg)  IV (128 mg)  c. c.  c. c. CH2Cl2-EtOAc (9:1,8:2,6:4, 4:6, 1:4)  IVA IVB IVC 106 138 (1.0 mg) (2.4 mg)  CH2Cl2-EtOAc (95:5,9:1,8:2, 2:1, 1:2)  IVB(11.9mg)  VA  VB Tl(l) (0.6 mg)  c. c.  CH2C12EtOAc (9:1)  207 (0.8 mg)  ure 2.2  Column Chromatographic Separation of Extract from TRP4a Cell Culture (continued)  VI(119mg)  VII (599 mg)  c. c. i) CH2Cl2-EtOAc  c. c. i) hexanes-CHCl3EtOAc-HC0 3 H  (8:2, 6:4, 4:6, 1:4) ii) EtOAc  (110:50:45:3, 110:50:184:3) u) CHCi3-EtOAc-HC02H (50:150:3)  VIA Tl(l) (0.6 mg)  V DB c. c. CH2Cl2-EtOAc (8:2,2:1),  VJ IA 210 (2.9 mg)  VIIB c. c. hexanes-CHCl3EtOAc-HC0 3 H (150:50:45:3, 200:50:45:3)  •  VI]31 V B2 VIB3 Tl (1) Td (2) c. c. hexanes(0.6 mg)(54.8 mg) EtOAc (3:7)  vni3i c. c. hexanesacetone (7:3)  1  1  21 (0.7 mg)  v nJ3ii 209 (1.0 mg)  1  208 (0.6 mg)  ure 2.2  Column Chromatographic Separation of Extract from TRP4a Cell Culture (continued)  60  Table 2.1  Compounds Isolated from TRP4a Cell Cultures (5 x 550 mL)  Compound  mg  Compound  106 107 108 109 116 117 118  12.1  138a  3.5 8.8 7.0 4.1  193 195 196 197 198 199  21.5  1.5  118  mg 2.4 2.8 11.1 19.0  1.0 1.5 1.0  Compound  200 201 202 203 204 205 206  138a  mg 3.5 0.9 3.5 1.1 2.1 1.2 2.9  207 208 209 210 211 212  mg 0.8 0.6 1.0 2.9 0.7 0.2  TI(1)  7.3  Td(2)  54.8  Compound  193  61  N 195  197  196  198  199  HON  200  201  202  OCH3  OCH3  HO°"  HO  203  204  205  OAc  206  207  208  62  Structure Elucidation of Diterpene Metabolites Isolated from TRP4a Cell Cultures Compound 196 was isolated as an optically active ([a]20 D +49.2°), colorless oil with a molecular weight of 270 (M + ). Its IR spectrum indicated aromatic (3070, 1600, 1480 cm"1) and saturated hydrocarbon moieties (2975, 1460 cm -1 ) in the molecule. High resolution mass spectrometry gave a molecular mass which corresponds to a molecular formula of C20H30.  196  63  The *H NMR spectrum of 196 showed signals of an AMX system in the aromatic region (5 6.88, IH, d, J = 1.6 Hz; 8 6.97, IH, dd, J = 8.2, 1.6 Hz; 8 7.16, IH, d, J = 8.2 Hz), indicating an 1, 2, 4 tri-substituted benzene ring. Three tertiary methyl signals were found at 8 0.92, 0.93, 1.17, and two secondary methyl signals were found at 1.21 (6H, d, / = 6.9 Hz), respectively. The 2D ^ - ^ H Correlated SpectroscopY (COSY) spectrum showed that the secondary methyl signal at 8 1.21 was coupled with a signal at 8 2.83 (IH, sept, / = 6.9 Hz), indicating that an isopropyl group was attached to the aromatic ring. The multiplet (2H) at 8 2.89, which was coupled with signals at 8 1.64-1.81 (IH) and 81.86 (IH), was apparently due to benzylic methylene protons (H7). The remaining carbon attached to the benzene ring was therefore a tertiary carbon since no other benzylic proton signals were found in the spectrum. All the spectral information suggested that 196 had a dehydroabietane type skeleton. In view of the fact that dehydroabietic acid (157) had been previously isolated from TRP4a cell cultures by this group, and that this compound has only carbon and hydrogen in the molecule, 196 was tentatively assigned as dehydroabietane. The COSY and Nuclear Overhauser Effect (NOE) results were supportive of such a structure, and the proton signals were assigned accordingly. Comparison of all the spectral data with the literature,166 and finally a synthesis of 196 from dehydroabietic acid (157) (vide infra) confirmed the identity of 196 as (+)-dehydroabietane. Dehydroabietane has been detected from several other cell cultures capable of producing diterpenes with an abietane skeleton.167 Its isolation, however, has never been reported from Tripterygium, and it was also first time that dehydroabietane was identified in TRP4a cell cultures.  Compound 204 was isolated as colorless needles (mp 98-100°C) with a molecular formula of C20H30O. It was optically active ([ocfl +45.6°) and its IR spectrum indicated the presence of a hydroxyl group (3600, 1040, 1020 cm"1) and a benzene ring (3025, 1600, 1500 cm"1).  64  204  The *H NMR spectrum of 204 showed typical dehydroabietane-type AMX system signals at 8 6.88 (1H, br s, H14), 6.97 (1H, br d, / = 8.2 Hz, H12), 7.13 (1H, d, J = 8.2Hz, H l l ) , three tertiary methyl signals at 8 0.88, 1.05, 1.17 and two isopropyl methyl signals at 1.20 (6H, d, / = 6.9 Hz). A signal at 5 3.27 (1H, dd, J = 10.8, 5.2 Hz) indicated a proton geminal to a hydroxyl group. The above spectral information was indicative of a hydroxylated dehydroabietane structure. Signals at 8 2.84 and 2.93 were due to geminally coupled benzylic protons, and assigned to the two C7 protons. The protons at C7 were also coupled with C6 protons in the four-proton multiplet between 8 1.64 and 1.92 (H2 and H6). The proton signal at 8 2.28 (br ddd, J = 13.0, 3.3, 3.3 Hz) showed a strong NOE with the H l l resonance (Figure 2.3). The signal at 8 2.28 was enhanced by 8% when H l l was irradiated, and the HI 1 signal was increased by 20% upon saturation of the signal at 8 2.28. By inspection of a model of the molecule, it is obvious that only Hip has the proximity to H l l , and because of its nearly co-planar relationship to the benzene ring (ring C), the anisotropic deshielding influence of the benzene ring has resulted in a downfield shift of the HI[3 signal as compared with that of its geminal partner, HI a (8 1.52). This particular spatial relationship between HI(3 and H l l has aided to facilitate the structure elucidation of these types of compounds, since, by locating the HI (3 signal (through NOE with H l l ) , assignment of other ring A proton signals would be relatively easy. On the other hand, benzylic proton signals (H7), recognizable by their relatively low field chemical shifts and in some cases through NOEs with the C14 proton, provided a starting point to locate other ring B proton signals in the *H NMR spectra.  65  Figure 2.3  Major NOEs Observed in NOE Difference Spectra of Compound 204  The proton signal of HI(3 showed strong cross peaks to its geminally coupled Hloc signal (5 1.52) in the COSY spectrum, and both were correlated with the H2 signals in the fourproton multiplet between 8 1.64-1.92 (H2, H6). The proton signal at 8 1.31 (dd, / = 12.1, 2.1 Hz) showed cross peaks to the same group of signals as did the H7 signals and thus was assigned to H5. Therefore, the resonance at 8 3.27, which was due to a proton geminal to the hydroxyl group, must arise from the proton at C3. This signal was a doublet of doublets (with cross peaks to H2) with coupling constants of 10.8 and 5.2 Hz, indicating that it was an axial proton.168 On this basis, the hydroxyl group at C3 must be in a p orientation. The results of NOE difference experiments were also consistent with the stereochemistry of this molecule (see Figure 2.3). Irradiation of the H3a proton at 8 3.27 showed an enhancement of the H5 signal (9%) due to their 1,3-diaxial relationship. Saturation of the methyl resonance at 8 1.05 increased the intensity of signals at 8 3.27 (H3a), 1.83-1.89 (H6a) and 1.31 (H5), suggesting that this methyl signal was due to the C18 methyl group. Since the environment of the C20 methyl group is similar to that of the CI6 and C17 methyl groups, their chemical shifts are not expected to vary significantly and should be slightly shifted downfield (deshielded by the benzene ring) compared with the signals of the C18 and C19 methyl groups. Therefore, the signal at 8 0.88 was attributed to the C19 methyl group, and the signal at 8 1.17 was assigned to the C20 methyl group. Irradiation of the C19 methyl resonance showed NOEs to the C20 methyl signal (1,3-diaxial), and H2{3, H6P resonances at 8  66 1.70-1.80. The mass spectrum of 204 showed the expected fragmentation169 at m/z 271 (M CH3), m/z 253 (M - CH3 - H 2 0, base peak), m/z 185 and 159 (Scheme 2.1).  m/z 159  Scheme 2.1  m/z 227  Proposed Mass Spectral Fragmentation of 204  m/z 185  67  J. G. Urones et al. reported the first isolation of compound 204 as a natural product from Nepeta tuberrosa subsp. reticulata in 1988. 170 It has also been synthesized by T. Matsumoto et al. from dehydroabietic acid (157).171 This compound has never been reported from Tripterygium, and the present work documents the initial separation and identification of 204 in TRP4a cell cultures. The IR, *H and  13  C NMR spectral data obtained were identical  with those reported in the literature.  202  Compound 202 was obtained as a colorless, wax-like solid (mp 102-104°C). This optically active compound ([ot]20 D +18.2°) revealed a molecular formula of C20H30O by high resolution mass spectrometry. Its IR spectrum showed bands for the hydroxyl group (3628, 1057 cnr 1 ) and the benzene ring (3004, 1498 cm"1). The !H NMR spectrum of 202 showed comparable features with that of 204, suggesting that 202 was likely to be a similar hydroxylated dehydroabietane compound. The multiplets at 8 2.89 (2H) were due to benzylic C7 protons, which, in a COSY experiment, showed cross peaks (Figure 2.4, d) to a four-proton multiplet between 8 1.65-1.80, suggesting the presence of two C6 protons. Irradiation of the proton at 8 7.15 (Hll) caused an 8% enhancement of the signal at 8 2.01 (1H, br ddd, J = 13.7, 3.5, 3.5 Hz) in NOE difference experiments (Figure 2.5), indicating that the enhanced signal was due to Hlp\ The Hip proton showed a strong geminal coupling with the Hloc proton at 8 1.85 (1H, br ddd, J = 13.7, 13.7, 3.5 Hz) (Figure 2.4, i), and both protons exhibited correlations with signals at 8 2.09 (1H, dddd, / = 13.7, 13.7, 3.5, 2.5 Hz) (Figure 2.4, f, g) and 8 1.65-1.80 (4H, two of which were H6, vide supra) (Figure 2.4, h, j).  68  Thus, the signal at 8 2.09 must be from one of the C2 protons and the other C2 proton signal must be in the multiplet at 8 1.65-1.80.  1.0  1.5  _  2.0  _  2  _  3,  3, PPM  3.5  3.0  2.5  2.0  1.5  1.0  PPM  a, H3p/H2cx; b, H3p/H2P; c, H15/H16.H17; d, H7/H6; e, H2p/H2a; f, H2p/Hla; g, H2p/Hlp; h, Hlp/H2a; i, Hlp/Hla; j , Hla/H2a  Figure 2.4  Expanded COSY Spectrum of Compound 202  69  The proton geminal to the hydroxyl group appeared at 8 3.48 (br dd, J = 2.5, 2.5 Hz) and was coupled to both C2 protons (Figure 2.4, a, b); therefore, it was identified as the proton at C3. The small coupling constants suggested that this proton was in an equatorial position and the hydroxyl group must be a-axially situated, i.e., this compound was an a alcohol. Because of the axial hydroxyl group at C3, the H l a signal was shifted downfield from 8 1.52 in 204 to 1.85 in 202 (1,3-diaxial interaction), and the H5 resonance also shifted downfield from 8 1.31 in 204 to 8 1.65-1.80 (H5 was assigned to the only remaining proton in that multiplet). A comparison of the *H NMR data is provided in Table 2.2 (vide infra). Irradiation of H30 caused NOEs to both C2 protons, and the methyl groups at 8 0.94 and 1.01 (CI8 and C19 methyl groups). Compared with 196 and 204, the methyl signal at 8 1.18 in compound 202 can only arise from the C20 methyl group. Saturation of the C20 methyl resonance increased the intensity of the H2 signal at 8 2.09 by 3%, and that of the Hip signal by a similar percentage, indicating that the H2 signal at 8 2.09 was due to H2p\ The coupling constants of this proton (dddd, J = 13.7,13.7, 3.5, 2.5 Hz) also supported this assignment (geminal coupling constant J = 13.7 Hz; H2p-Hloc coupling constant J = 13.7 Hz). Irradiation of the C20 methyl group also enhanced the methyl resonance at 8 0.94, so the latter signal can only be due to the C19 methyl group (1,3-diaxial relationship with the C20 methyl). The remaining methyl resonance at 8 1.01 was assigned to the CI8 methyl group.  Figure 2.5  Major NOEs Observed in NOE Difference Spectra of Compound 202  70  The  13  C NMR spectrum of 202 revealed the chemical shift of C3 at 8 75.8, while the  chemical shift of C3 in 204 was at 5 78.8. This also provided strong evidence that the hydroxyl group was a (axial) in this compound but P (equatorial) in 204. It is well known that steric interactions (y gauche butane-like effect) shift upfield not only the signal of a carbon with a y substituent, but also the signals of the carbons that are in between ("internal" gauche butane effect).172'173 Since the steric interactions are stronger in the a isomer 202, the chemical shift of the C3 carbon in the a isomer should be at a higher field than that in the P isomer.  j  «^mf^  HO.  1 g a u c n e effect  OH^  202  The mass spectrum of 202 exhibited a similar fragmentation pattern as 204, but the molecular ion and other fragments were relatively weak when compared with the base peak at m/z 253 (M - CH3 - H2O) for 202. Compound 202 is a new compound, its isolation also constitutes its first isolation from T. wilfordii plant cell cultures.  19  18  198 Compound 198 was isolated as an optically active compound ([oc],20 D +31.1°) with a molecular formula of C20H28O. Its IR spectrum displayed absorption bands for a carbonyl (1708 cm"1) and a benzene ring (3010,1500 cm"1).  71  The lH NMR spectrum showed the typical AMX system pattern in the aromatic region, and the protons were assigned accordingly. The two-proton multiplet centered at 8 2.88 was assigned to the benzylic protons (H7), on the basis of their NOEs observed with H14. The methylene protons at C6 were found at 8 1.79 by their cross peaks to H7. A signal at 8 1.90 (1H, dd, J = 11.3, 3.2 Hz) also showed a cross peak to the H6 signal, indicating that this signal must be due to H5. Irradiation of H l l caused an enhancement (11%) of the Hip signal at 8 2.46 (1H, ddd, J = 13.3, 7.5, 4.2 Hz). The COSY spectrum revealed a strong geminal coupling of H i p with the proton at 8 1.94 (1H, m, Hloc), and both showed correlations with protons at 8 2.55 (1H, ddd, J = 15.7, 7.6, 4.2 Hz) and 2.67 (1H, ddd, J = 15.7, 10.0, 7.5 Hz). It meant that the signals at 8 2.55 and 2.67 must belong to the two C2 protons, and the coupling constants of these two protons indicated that the one at 8 2.67 was the axial proton (H2p), which was confirmed by a NOE enhancement upon irradiation of the C20 methyl group. The downfield shift of the H2 signals compared with their counterparts in 196 suggested that the carbonyl was at C3. Irradiation of the methyl group at 8 1.27 (H20) showed NOEs to H l l and the methyl group at 8 1.12 (HI9). The remaining methyl signal at 8 1.14 was assigned to the C18 methyl group. Therefore, compound 198 was assigned as abieta-8,11,13-trien-3-one. Ketone 198 has been isolated from Juniperus sabina114 and Salvia wiedemanniO15  It  has never been isolated from Tripterygium plants, and here we report its initial isolation from TRP4a cell cultures.  Compound 201 was isolated as a pale yellowish oil and showed a specific rotation of +136.9° ([cc]23 D , c =0.0132, MeOH). High resolution mass spectrometry revealed its molecular formula as C19H24O, one carbon less than normal diterpenes. The IR absorptions indicated the presence of benzene (3025, 1500 cnr 1 ) and a,P~unsaturated ketone (1660 (strong), 1620 cm -1 ) functionalities. Strong UV absorptions at 219.0 nm (log e = 3.93) and 243.3 nm (log 8 = 4.03) also supported the presence of an a,P-unsaturated ketone system in this molecule.  72  201  The *H NMR spectrum exhibited the familiar AMX system signals in the aromatic region. Only four methyl group signals were observed, which indicated that loss of a methyl group was probably attributed to the missing carbon in the molecular formula. One methyl signal was shifted downfield to 8 1.82 (3H, br s), suggesting that this methyl group was very likely attached to an unsaturated carbon.168 The isopropyl methyl signals were found at 8 1.22 (6H, d, J = 7.0 Hz), and the H15 signal was found in the three-proton multiplet between 8 2.762.91. Irradiation of H l l enhanced H i p signal at 8 2.35 (1H, ddd, / = 13.2, 5.1, 2.3 Hz) (Figure 2.6). The COSY spectrum showed correlations between Hl|3 and protons at 8 2.04 (1H, br ddd, J = 14.8, 13.2, 4.9 Hz), 2.52 (2H, m), and 2.71 (lH,ddd, J = 17.7, 14.8, 5.1 Hz) (Figure 2.7, k, j , f). Saturation of the Hip resonance increased the intensities of the signal at 8 2.04 by 22% and another one at 8 2.71 by 7%. This suggested that the signal at 8 2.04 was due to proton H l a (spatially nearer to HI(3). As it was expected, irradiation of H l a enhanced the signal of Hl|3 by 14%, and also resulted in a moderate enhancement of the signal at 8 2.52 (about 3%). Based on these observations, it was certain that the proton at 8 2.71 was H2(3, while the other in the two-proton multiplet at 8 2.52 was H2a. The chemical shifts of these C2 protons indicated that the ketone group was at C3. Irradiation of H14 showed NOE enhancements of the signal at 8 2.98 (1H, m), and some proton signals in the three-proton multiplet between 8 2.76 and 2.91 (one of them was HI5, vide supra). This means that the proton at 8 2.98 should be one of the C7 protons, and the  73  other C7 proton must be in that multiplet (the enhancement should include H7 and also HI5 because of their similar relationships with H14). The proton signal at 8 2.98 showed correlations to this multiplet (Figure 2.7, b) and another multiplet at 8 2.52 (2H, one was H2a, vide supra) (Figure 2.7, a), and these two multiplets were also mutually correlated (Figure 2.7, d). It was not difficult to conclude that one of the remaining protons in the multiplet at 8 2.52 must be a C6 proton, and accordingly the only proton left in the multiplet between 8 2.76 and 2.91 must be the other C6 proton. The H7 signal at 8 2.98 was assigned to H7(3 because, by inspection of a model, it can be seen that H7|3 is in a quasi-equatorial orientation, thus more deshielded than H7a by the nearby benzene ring. Irradiation of the methyl group at 8 1.50 (C20 methyl group) increased the signal intensity of H l l at 8 7.20. The downfield shift of this methyl signal compared with its counterpart in 198 (from 8 1.27 to 1.50), and even greater downfield shift of the C6 protons (from 8 1.79 in 198 to 2.52 and 8 2.76-2.91 in this compound) indicated that there must be a double bond between C4 and C5, thus forming the a,p-unsaturated ketone with the carbonyl at C3.  Figure 2.6  Major NOEs Observed in NOE Difference Spectra of Compound 201  The methyl group at 8 1.82 can only be attached to C4 and this was confirmed by a weak COSY correlation (Figure 2.7, h; only showing slight broadening in *H NMR spectrum) between this methyl group and the H6 in the multiplet at 8 2.52 (homoallylic coupling 168 ). This also revealed that the H6 resonance at 8 2.52 was due to the H6P proton, because H6(3 is  74  nearly perpendicular to the C4, C5 double bond plane while the other H6 proton at 8 2.76-2.91 (therefore, must be the H6a) is lying in the plane, thus being more deshielded by C4, C5 double bond. Furthermore, this assignment was justified by the enhancement of H6oc (11%) upon irradiation of the methyl signal at 5 1.82 (H19).  L  1.0  _  2.0  _  3.0  1. 0 PPM a, H7(3/H6(3; b, H7p/H6cc,H7a; c> H15/H16.H17; d, H6a,H7a/H6P: e, H2(3/Hla; f, H2p/Hlp; g, H2p/H2a; h, H6p/H19; i, H2a/Hla; j , H2a/Hlp; k, Hlp/Hla  Figure 2.7  Expanded COSY Spectrum of Compound 201  75  199  Compound 199 had a molecular formula of C20H28O. An IR absorption at 1660 cm -1 suggested the presence of a carbonyl group adjacent to an aromatic ring. The latter was supported by a strong UV absorption at 254.0 nm (log e = 4.03). The *H NMR spectrum displayed some similarity to that of dehydroabietane (196), thus this compound was probably the 7-oxo analog of 196. The aromatic signals were shifted to lower field. The H14 signal was found at 5 7.85 (1H, d, / = 2.1 Hz), which was about 1 ppm lower than its counterpart in compound 196, apparently due to the deshielding effect of a nearby keto group at C7. The two H6 signals, which were also shifted downfield, were found at 8 2.62 (1H, dd, 7=18.1, 13.5 Hz) and 8 2.71 (1H, dd, J = 18.1, 4.4 Hz), respectively. The COSY spectrum indicated that the two C6 protons were only coupled with H5 at 8 1.86 (1H, dd, J = 13.5, 4.4 Hz). The remaining proton signals were assigned as based on the NOE and COSY spectra. All the spectral data were in agreement with the structure assigned to 199, and its identity was confirmed by a synthesis from dehydroabietane (196, vide infra). A comparison of the *H NMR data of 196, 204, 202,198, 201 and 199 is given in Table 2.2. . This compound is a known compound,176,177 but it is the first isolation from TRP4a cell cultures. No isolation from Tripterygium plants has been reported.  18 19 20  0.93, s 0.92, s 1.17,s  7.16, d (8.2) 6.97, dd (8.2,1.6) 6.88, d (1.6) 2.83, sept (6.9) 1.21, d (6.9)  2.89, m  7p  11 12 14 15 16,17  2.89, m  7a  3(3 5 1.65-1.80, m 1.65-1.80, m  1.64-1.92, m  1.05, s 0.88, s 1.17, s  1.01, s 0.94, s 1.18, s  7.15, d (8.2) 6.96, d(8.2) 6.86, s 2.80, sept (7.0) 1.20, d (7.0)  2.89, m  2.89, m  1.65-1.80, m  1.31, dd (12.1, 2.1) 1.64-1.92, m  2.84, m 2.93, dd (17.2, 6.0) 7.13, d (8.2) 6.97, d (8.2) 6.88, s 2.80, sept (6.9) 1.20, d (6.9)  3.48, dd (2.5, 2.5)  -  2.09, dddd (13.7, 13.7, 3.5,2.5)  202 1.85, ddd (13.7, 13.7, 3.5) 2.01, ddd (13.7, 3.5, 3.5) 1.65-1.80, m  -  3.27, dd (10.8, 5.2)  1.21, m  3a  1.46, d (13.2) 1.34, dd (12.4, 2.4) 1.86, dddd (13.2, 6a 7.1,2.4,2.4) 6(3 - 1.64-1.81, m  1.64-1.92, m  1.64-1.81, m  2P  2.28, ddd (13.0, 3.3, 3.3) 1.64-1.92, m  204 1.52, m  1.59, m  196 1.39, ddd (13.1, 12.7, 3.6) 2.26, d (12.7)  1.14, s 1.12, s 1.27, s  7.15, d (8.2) 7.00, d (8.2) 6.90, s 2.81, sept (6.9) 1.21, d (6.9)  2.88, m  2.88, m  1.79, m  1.90, dd (11.3, 3.3) 1.79, m  -  -  2.46, ddd (13.3, 7.5,4.2) 2.55, ddd (15.7, 7.6,4.2) 2.67, ddd (15.7, 10.0, 7.5)  1.94, m  198  -  -  1.52, m 1.86, dd (13.5, 4.4) 2.71, dd (18.1, 4.4) 2.62, dd (18.1, 13.5)  1.75, dddd, (13.7, 13.7,3.3,3.3) 1.27, m  1.66, m  2.31,(1(12.6)  1.52, m  199  7.20, d (8.1) 7.27, d (8.2) 7.08, dd (8.1, 1.5) 7.37, dd (8.2, 2.1) 6.95, d (1.5) 7.85, d (2.1) 2.76-2.91, m 2.90, sept (6.9) 1.225, 1.227, d 1.22, d (7.0) (6.9) 0.98, s 1.82, s 0.91, s 1.50, s 1.215, s  2.98, m  2.76-2.91, m  2.52, m  2.76-2.91, m  -  -  -  2.71, ddd, (17.7, 14.8,5.1)  201 2.04, ddd (14.8, 13.2,4.9) 2.35, ddd (13.2, 5.1,2.3) 2.52, m  *H NMR Spectral Data of Abietane Diterpenes 196,204,202,198,201 and 199 (400 MHz in CDC13, 5 in ppm, J in Hz in parentheses)  2a  1(3  la  Table 2.2  77  197  Compound 197 had a molecular formula of C20H30O. The IR spectrum showed an absorption band for a hydroxyl group at 3620 cm"1. The sharp appearance of this absorption indicated that this hydroxyl group was very likely to be a phenolic hydroxyl group, because this IR spectral feature was frequently observed with phenolic compounds during the course of the present investigation. The *H NMR spectrum showed AM system signals in the aromatic region, suggesting that the phenolic hydroxyl group was attached to CI4. The downfield shift of the H15 signal (5 3.15, sept, / = 6.9 Hz) compared with that (5 2.83) in 196 supported this assumption. The proton resonance at 8 4.59 (IH, br s) was due to a hydroxyl group because irradiation of this signal caused the water peak of the solvent to become negative in NOE difference spectra (saturation transfer178). This irradiation also enhanced the signals of H15, and the protons at 8 2.81 (IH, br dd, J = 16.5, 6.6 Hz) and 2.61 (IH, ddd, J = 16.5, 11.3, 7.9 Hz), which were assigned to C7 protons accordingly. Therefore, the hydroxyl group was situated at C14, i.e., 197 was a C14-hydroxyl analog of compound 196. The rest of the proton signals were assigned according to the data from COSY and NOE spectra. The structures of 196,197 and 199 were finally confirmed by chemical synthesis from dehydroabietic acid (157). Reduction of 157 by lithium aluminum hydride (LAH)179 gave the alcohol 196(i), which was then converted to the tosylate180 and subsequently reduced to give dehydroabietane (196)181 (Scheme 2.2).  78  b |— U  196(i) R = OH * 196(H) R = OTs  199 Scheme 2.2  Syntheses of 196 and 199 from 157  a) LAH, ether; b) p-toluenesulfonyl chloride, pyridine; c) Nal, Zn, HMPA, A; d) GO3  196  Scheme 2.3  z Rx = N0 2 ,R»2_ = N0 2 1 2 c 1— 197(H) R = NH2, R = N0 2 °pl97(iii) R1 = I,R 2 =N0 2 p="197(iv) R1 = I,R 2 = NH2 C '-*197(v) R1 = I,R2=OH  br—197(f)  197  Synthesis of 197 from 196  a) HNO3, H2SO4, MeN02; b) H2, Pt/C, HOAc-EtOH-benzene; c) NaN02, KI, TFA-HOAC-H2O; d) Na2S2Q4, EtOH, A; e) NaNC>2, TFA; HCl, MeOH; f) H2, Pd/C, Na2C03, MeOH-benzene  79  The synthetic sequences leading to 197 and 199 from 196 were based on the results from the synthetic studies of biotransformation precursors (see Chapter 3 for details). Oxidation of 196 with chromium (VI) oxide (Cr03) yielded the 7-oxo compound 199 (Scheme 2.2). A multistep synthesis from dehydroabietane (196) afforded compound 197 (Scheme 2.3). The spectral data of the synthetic compounds were identical to those of the corresponding isolated compounds. Compound 197 has been synthesized before, 182 but its isolation as a natural product has not been found in the literature.  116 Compound 116 was isolated as an optically active ([aJ 3 +117.4°, c = 0.0145, MeOH) white powder (mp 138-140°C) with a molecular formula of C20H28O2. Its IR spectrum revealed the presence of hydroxyl (3620 cnr 1 ), carbonyl (1700 cm -1 ) and benzene ring (3025, 1580, 1500 cm -1 ) functionalities. The similarity between this compound and 197 in the aromatic region (AM system) and the broad singlet at 5 4.65 in the *H NMR spectrum suggested that this compound also had a hydroxyl group at CI4. The signals of the methyl groups and most of the aliphatic protons showed a comparable pattern with that of 198, indicating the location of the carbonyl group at C3. Irradiation of the proton at 8 4.65 (IH, br s) caused signal increases of H15 and the proton at 6 2.90 (IH, br dd, / = 16.6, 5.9 Hz, H70), but depressed the signal from traces of water in the solvent, thereby confirming the presence of a hydroxyl group at CI4. All the spectral information suggested that compound 116 was 14-hydroxyl-abieta-8,ll,13-trien-3-one.  80  This compound has been previously isolated from Tripterygium plants, and our *H NMR data were identical to those reported.89  200  Compound 200 (C21H30O2) was isolated as colorless needles (mp 134-136°C) with a specific rotation of +84.9° ([oc]g\ c = 0.0171, MeOH). The IR absorptions established the presence of a ketone (1700 cm -1 ), a benzene ring (3025, 1500 cm-1) and possibly an aryl alkyl ether (1260, 1040 cm"1). The *H NMR spectrum of 200 showed close resemblance with that of 116. The H l l signal (1H, d, J = 8.3 Hz) was shifted downfield to 5 7.00 (8 6.83 in 116) while H12 (1H, d, J = 8.3 Hz) was less affected at 5 7.06. Meanwhile, the phenolic hydroxyl proton was not observed but a methoxy signal was found at 8 3.71 (3H, s). Therefore, compound 200 was very likely to be the methyl ether of 116. The downfield shifts of resonances due to H15 (8 3.27, septet, J = 6.9 Hz), H7p (8 3.10, br dd, / = 17.4, 5.5 Hz) and H7a (in the two-proton multiplet at 8 2.66) from 8 3.10, 2.90 and near 8 2.60, respectively, compared with their counterparts in compound 116, provided strong supporting evidence for the presence of a methoxy group at C14. Irradiation of the methoxy group at 8 3.71 showed NOE effects with H15 (6%), H7(3 (4%), and H7oc (4%), thus confirming this assignment. Other individual protons were assigned according to coupling constants, and the data obtained from COSY and NOE experiments. This compound has not been reported from the Tripterygium plants.  81  117  Compound 117 was isolated as colorless needles (mp 184-186°C) with an optical activity of +190.2° ( [ a j 3 , c = 0.00736, MeOH). Its molecular formula was determined as C21H30O3. The IR spectrum displayed bands for hydroxyl (3600 cm -1 ), carbonyl (1700 cnr 1 ), and benzene ring (3025, 1490 cm-1) functionalities. The lH NMR spectrum showed a methoxy group at 5 3.66 (3H, s), but only one aromatic proton at 5 6.36 (1H, s). The signal at 8 4.55 (1H, s) changed its shape (from sharp to very broad) as well as its chemical shift position when the concentration of the NMR sample was changed, suggesting that it was a phenolic hydroxyl proton. Irradiation of the proton at 8 6.36 increased the signal intensities of the hydroxyl proton (2%), H15 (1%), and the C16, C17 methyl protons (1%), while saturation of the hydroxyl proton resonance caused NOE enhancements to the signals at 8 6.36 (18%) and 8 3.05 (1H, ddd, / = 13.8, 8.5, 6.3 Hz) (2%). It was obvious that the signal at 8 6.36 was due to H12, which is close to the C16, C17 methyl groups, H15, and any proton in the group attached to CI 1 (Figure 2.8). The NOEs observed upon irradiation of the hydroxyl proton at CI 1 not only confirmed its spatial proximity to H12 but also revealed the location of the H i p signal in the spectrum. The downfield shift of Hlf} signal (8 3.05, compared to 8 2.44 in 200) was another indication of severe crowding caused by introduction of the hydroxyl group at C l l . 1 8 3 Irradiation of the methoxy group at 8 3.66 produced enhancements of the HI 5 signal (4%), and the H7 signals at 8 3.10 (1H, ddd, J = 17.1, 4.7, 1.8 Hz) (3%) and 2.56 (1H, ddd, / = 17.1, 13.0, 6.0 Hz) (3%), respectively. Therefore, the methoxy group was situated at C14. The rest of the protons were assigned based on their coupling constants, COSY, and NOE results. This  82  compound has been isolated from Tripterygium plants. Our lR NMR and specific rotation were consistent with those previously reported.90  118  Compound 118 was obtained as optically active ([a]^ +56.6°, c = 0.145, CHCI3) colorless prisms. It melted at 188-190°C and had a molecular formula of C21H30O3. The IR spectrum showed absorptions of hydroxyl (3600 cm -1 ), keto (1702 cm"1), and benzene (3029, 1609, 1585 cm"1) groups. Its 1 H NMR spectrum showed an aromatic proton at 8 6.38, and a methoxy group at 8 3.67 (3H, s). A signal at 8 4.47 (1H, s) was likely to be a phenolic proton as compared with 117. The isopropyl methine resonance (H15, septet, / = 7.2 Hz) was found shifted downfield to 8 3.41. Also shifted downfield were the two isopropyl methyl signals at 8 1.34 and 1.35 (3H each, both d, J = 7.2 Hz). Saturation of the resonance at 8 6.38 enhanced the signals at both 8 4.47 (6%) and 8 2.33 (1H, ddd, J - 13.1, 7.4, 4.2 Hz) (11%), and slightly increased the methyl resonance at 8 1.25 (3H, s) (Figure 2.8). Upon irradiation of the 8 4.47 signal, NOEs were observed for the proton signal at 8 6.38 (10%) as expected, and the isopropyl methyl signals (1%), but the signal at about 8 1.5 due to traces of water in the solvent was depressed. This indicated that the signal at 8 4.47 was from a hydroxyl proton, and this hydroxyl group was situated at CI2, where it was closer to the CI6, C17 isopropyl methyl groups, rather than at C l l . Accordingly, the resonance at 8 6.38 was due to H l l and the signal at 8 2.33 must arise from Hlp\ This explained why the H15 and the isopropyl signals were shifted downfield, while the HI(3  83  resonance appeared at a chemical shift similar to the one found in 117. A position change of the hydroxyl group from CI 1 to C12 increases steric crowding around the isopropyl group, but in turn releases tension previously directed towards Hip. NOEs were observed between the methoxy group at 8 3.67 and H15, and a signal at 8 2.99 (1H, ddd, J = 17.1, 5.7, 1.6 Hz), as well as one of the multiplet signals between 8 2.50 and 2.71. Therefore, the methoxy group was located at C14, and the proton at 8 2.99 was H70 while the one in the multiplet was H7a.  117  Figure 2.8  118  A Comparison of NOE Results between 117 and 118  The H7p proton of 118 coupled intensively with the proton observed at 8 1.68 (1H, ddd, / = 12.5, 12.5, 5.7 Hz), while H7a signal showed cross peaks to both the signals at 8 1.68 and at 8 1.83 (1H, m) with the coupling being stronger to the latter (8 1.83). Inspections of the molecular model revealed that the required conformation places the C7 and C6 protons in a spatial situation which is in the middle between eclipsed and staggered relationships with each other. Such a relationship existed in virtually all compounds of the dehydroabietane or the abeo dehydroabietane type discussed in this thesis if ring A and B were trans fused and ring C was a benzene ring. The dihedral angle between H7p and H6p is around 30°, thus there is a strong coupling between these two protons, but since the angle between H7|3 and H6oc is near 90°, little or no coupling between them is usually observed. The proton signal at 8 1.68 was thus assigned to H6p\ On the other hand, H7cc is separated by a small dihedral angle with H6a, but an angle  84  close to 180° with H6(3, so it was strongly coupled with both of the C6 protons. The coupling of H7oc may be slightly stronger with one of the two C6 protons depending on the individual compound. This coupling pattern between H6 and H7 was helpful in assigning these protons. Thus, the signal at 8 1.83 was assigned to H6a.  \ -90 degree  Coupling of H5 to both H6 signals was obscured by diagonal crowding, but the appearance of the H5 resonance as a sharp doublet of doublets at 8 1.84 allowed facile recognition. The remainder of the protons were assigned accordingly by COSY and NOE experiments. This compound has been previously reported by B.-N. Zhou et al. from an isolate of T. wilfordii. Their lH NMR data (100 MHz) were comparable with ours.90  1  14  OCH3  205  The optically active compound 205 ([a]^ +34.1°C, c = 0.063, CHCI3) was isolated as a wax-like solid with a molecular formula of C21H32O2. The IR absorptions indicated the presence of a hydroxyl group (3614 cm -1 ), benzene ring (3002, 1480 cm -1 ), and a possible ether (1030 cm -1 ) linkage. The *H NMR spectrum was generally very similar with that of 200  85  from 5 3 ppm to the aromatic region, but resembled that of compound 204 from 8 2.5 ppm to the methyl signal region (at about 8 1 ppm). Irradiation of the signal at 8 6.98 (IH, d, J = 8.4 Hz, HI 1) produced an enhancement of the signal at 8 2.27 (IH, ddd, J = 13.1, 3.5, 3.5 Hz, Hip) (12%). NOEs were observed for the signal at 3.26 (1H, septet, J = 6.9 Hz, H15) and another one at 8 3.02 (1H, br dd, / = 17.6, 5.3 Hz) upon saturation of the methoxy resonance at 8 3.70 (3H, s). Therefore, the methoxy group was again attached to C14, and the resonance at 8 3.02 was assigned to H7(3. The Hip proton showed strong coupling to H l a at 8 1.52 (IH, m), and both were correlated with a two-proton multiplet at 8 1.77 (H2). The C2 protons exhibited cross peaks to a signal at 8 3.28 (IH, dd, J = 10.9, 4.9 Hz), indicating that this proton, which was geminal to a secondary hydroxyl group, was at C3 and a oriented (axial proton with large coupling constants), i.e., compound 205 was a P alcohol ( 13 C NMR showed C3 at 8 78.8, same as for compound 204).  H0^3  Figure 2.9  Some NOEs observed for 205  Figure 2.9 portrays some NOEs observed for 205 with regard to the assignment of the tertiary methyl groups. Irradiation of the methyl signal at 8 1.06 enhanced signals of H3a (8%), H6a (10%), H5 (3%), and of a methyl group at 8 0.88 (slightly), indicating that the irradiated methyl group was the CI8 methyl group, and the enhanced methyl signal could only be due to the C19 methyl group. Saturation of the C19 methyl resonance caused signal  86  intensity increases of the C18 methyl (slightly), and another tertiary methyl group at 8 1.17 (2%), which was then assigned to the C20 methyl group. Therefore, the structure of 205 was determined as shown. Compound 205 is a new compound, and isolated for the first time from TRP4a cell cultures.  203  Compound 203 (C21H32O2) was obtained as colorless needles with a melting point of 157-159°C. It was optically active ( [ a ^ +27.1°) and exhibited IR absorptions of hydroxyl (3626 cm -1 ), benzene ring (3007, 1480 cm-1) and possible ether (1029 cnr 1 ) groups. The XH NMR spectrum indicated some similarities of this compound to 205, but there were some differences. The AM system signals in 205 did not compare favorably with the singlet-like signal (about 5 7, 2H) in aromatic region of 203. The resonance of the proton geminal to the hydroxyl group was slightly shifted downfield (from 8 3.28 to 8 3.49) and showed smaller coupling constants, indicating its equatorial orientation at C3 (i.e., an a alcohol). The methoxy group, isopropyl methine, and the two C7 proton signals, were virtually identical in chemical shifts and coupling patterns to their counterparts in 205. Irradiation of the methoxy group at 8 3.69 enhanced signals of H15 (5%), H7p (2%, assigned in comparison with 205), and H7oc (3%), confirming that this methoxy group was attached to C14 (Figure 2.10b). Irradiation of the singlet-like aromatic protons (2H) centered at 8 7.15 caused enhancements of the isopropyl methyl signals, and a signal at 8 2.00 (1H, br ddd, J = 13.1, 3.8, 3.5 Hz), which was thus assigned to Hip (Figure 2.10a). It seemed likely that the singlet-like aromatic proton signals were due to protons HI 1 and HI2.  87  The COSY spectrum of 203 displayed a cross peak between the H7P signal and a threeproton multiplet from 5 1.62 to 1.77, while H7a resonance showed two cross peaks, one to the above multiplet and the other to a multiplet observed between 8 1.77 and 1.90 (2H). This suggested that H6P was part of the first multiplet and H6oc part of the second one. Proton H i p exhibited strong coupling with part of the multiplet containing H6oc (8 1.77-1.90), indicating that the other proton in this two-proton multiplet must be HI a. Both H l a and Hip signals were also correlated with a signal at 8 2.09 (1H, dddd, / = 14.1, 14.1, 3.8, 2.4 Hz), and with part of the three-proton multiplet containing H6P, indicating that the protons at C2 were in the above locations. The two H2 signals exhibited cross peaks to the signal at 8 3.49 (1H, br dd, J = 3.3, 2.4 Hz), suggesting that this signal, geminal to the hydroxyl group, arose from a p-oriented (equatorial) H3 (thus an oc alcohol at C3). The remaining proton in the three-proton multiplet between 8 1.62 and 1.77 (the other two had been assigned to H2a and H6P, vide infra) was assigned to H5 in comparison with 202, and also because of an enhancement observed for the doublet-like signal (H5) in the multiplet when H7a was irradiated. Accordingly, the methyl signals at 8 0.94, 1.03 and 1.178 were assigned to the CI9, CI8 and C20 methyl groups, respectively, in comparison with compound 202. Irradiation of the C16, C17 and C20 methyl groups (chemical shifts were very close) caused NOEs to HI5, H12 (Hll), Hip, C19 methyl group, as well as H2 at 8 2.09 (thus assigned as H2p) (not shown in Figure 2.10). The 13 C NMR spectrum showed a C3 signal at 8 75.7, upfield shifted as compared with the chemical shift of C3 (8 78.8) in 205, but almost the same as that (8 75.8) in 202. Thus, compound 203 was assigned as 14-methoxy-abieta-8,ll,13-trien-3a-ol.  88  VWfc.P^» > "»V"'' <  •^r  HKW\,*H^M *  fcWl  <fciw^»"•wi"i^».M»N«^»/H^»rf«»W^<^'  M**w-«><ayY*iiv»VV>4AJ^^^  ij^fW^fr^^^  y s , w " » ' i » «» j>^Mfcn^n^W^^*ft**>N^%V> » y * y ^ « i v * » SA «• ^IIW iMiB'Vwgn^^n  *W*W • * > • — ^ S M ^ f * * — J w « " * . ••WlliH<N|H> llll»—•«»—*« • '»*•!—Wfr»«W  ^Y^V»Vv>w*^**''fc**'''*' *^* , *^ , ^^ ts >f*'  r  iMWf^*«—»t«W»  | M * W ^ W V W S ,  wpwy*.  • i^V^*^**»,^^,v*>^^-"»,*^^^*y»A it  ^ y v - V '• *v*~ "WW  f  '^Y^rJVpr--^-  ****/vfav/*iA  HI6 HI7 H20  g  H19  -OCH3  HIS  HU.HI2 Hla  H2a  HteHS /H6P  ^  JV_ 1  I ' ' ' ' I ' 7.5 7.0  • I ' 6.5  I '  s.s  s.s  • I • 5.0  1  4.5  I ' ' ' • 1 4.0 %.': PPH  3.0  ' I • 2.5  -r-, 2.0  , . i i . . , . | , 1.5 1.0  a, {H11,H12}: H i p , H16.H17; b, {CI4-OCH3}: H15, H7p, H7a; c, {H3PJ: H2P, H2a, OH, H18, H19; d, {H7a}: CI4-OCH3, H6a, H5a, (H6p); e, {H18}: H3P, H6a, H5, H19; f, {H19}: H3p, H2p, H6p, H20; g, Off-resonance spectrum of 203 (400MHz, CDCI3). ({irradiated}: enhanced)  Figure 2.10  NOE Difference Spectra of Compound 203  89  203  183  This assignment was contradictory, in the aspect of the location of the methoxy group, to the compound 183 previously isolated from TRP4a cultures by this group, 158 ' 160 because both compounds had identical *H NMR spectral data. In order to clarify the assignment, a *H NMR spectrum was taken on the original sample of 183 dissolved in deuterated benzene (C6D6), and the spectrum was compared with that of the same sample measured in deuterated chloroform (CDCI3). The spectrum recorded in C6D6 unambiguously exhibited typical AM system signals in the aromatic region, showing one proton at 5 7.06 (d, J = 8.3 Hz) and the other at 8 7.10 (d, J = 8.3 Hz) (Figure 2.11). The large coupling constant demonstrated that these two protons must be in ortho positions with respect to each other on the benzene ring, such as HI 1 and H12, rather than in the para positions, such as HI 1 and H14 as was proposed in the early reports (for benzene, J ortho = 6.0-9.0, Jmeta = 1.0-3.0, J para — 0.0-1.0 Hz 168 ). The SINEPT (Selective Insensitive Nucleus Enhancement by Polarization Transfer) experimental results described in the early reports indicated that the methoxy group could not be attached to CI 1, but these SINEPT results cannot differentiate whether the methoxy group is attached t o C 1 2 o r C 1 4 . A close analysis of the spectrum, recorded in CDCI3, revealed that there was an AM system in the aromatic region, but the two doublets were so close to each other that their inside halves almost overlapped, thus appearing as either a broad singlet or two poorly resolved singlets. The outside halves of these two doublets were so weak in intensity that they might be overlooked or mistaken as "spinning sidebands", a phenomenon which may be ruled out by examining the nearby solvent peak which showed no such sidebands. After a careful examination of the spectra, the doublet at 5 7.03 (7 = 8.4 Hz, in CD3CI) was assigned to  90  H12 and the other at 8 7.00 (/ = 8.4 Hz) was assigned to HI 1, in comparison with compound 205. The H12 signal in both compounds appeared shorter and broader than did its neighboring Hll signal, perhaps due to a very small, long range, allylic coupling of H12 with H15. Therefore compound 183 should be assigned as 203. This compound has never been isolated from Tripterygium plants. A comparison of lH NMR data of compounds 197,116, 200,117, 118, 205 and 203 is provided in Table 2.3.  i  i  P  I  1  I  7.5  J  I  7.0  i  I  I  i  1  'I  1  1  Figure 2.11  !  in CDC13  1  1——i  7.0  ppm  ppm a)  ' i  7.5  b)  in C6D6  H NMR spectra of 203 (400 MHz, aromatic region)  1  r-  -  3.23, sept (6.9) 1.15,1.165, d (6.9) 1.163, s 1.12, s 1.27, s 3.66, s  1.70, dddd (12.6, 12.6, 12.6, 5.5) 2.66, m  -  3.27, sept (6.9) 1.19, 1.20, d (6.9) 1.15, s J 1.12, s 1.27, s 3.71, s  -OMe -  7a  -  2.61, ddd (16.5, 2.41-2.64, m 11.3,7.9) 2.81, dd (16.5, 2.90, dd (16.6, 7P 6.6) 5.9) 6.83, d (8.2) 11 6.86, d (8.2) 7.02, d (8.2) 12 7.02, d (8.2) O H , 4.65, s 14 O H , 4.59, s 15 3.15, sept (6.9) 3.10, sept (7.0) 1.22,1.24, d 16,17 1.24, 1.26, d (6.9) (7.0) 1.15, s 18 0.96, s 1.12, s 19 0.94, s 1.30, s 20 1.20, d (0.4)  1.71-1.96, m  1.65-1.81,m  6p  6a  1.99, dd (12.3, 1.8) 1.79, dddd (13.0,6.0, 1.8, 1.8) 1.54, dddd (13.0, 13.0, 12.3,4.7) 2.56, ddd (17.1, 13.0,6.0) 3.10, ddd (17.1, 4.7, 1.8) O H , 4.55, s 6.36, s  1.87, dd (12.6, 1.5) 1.84, m  1.33, dd (12.7, 1.71-1.96, m 2.1) 1.98, dd (13.3, 1.71-1.96, m 7.9)  3.10, dd, (17.4, 5.5) 7.00, d (8.3) 7.06, d (8.3)  -  5  -  -  1.48, d (13.2)  -  3P  -  -  1.21, m  1.60, m  1.94, m  117  3a  2P  1.92, m  200  2.44, ddd (13.2, 3.05, ddd (13.8, 7.5,4.1) 8.5, 6.3) 2.41-2.64, m 2.55, ddd (15.8, 2.66, ddd (15.2, 7.6,4.1) 9.9, 6.3) 2.69, ddd (15.7, 2.66, m 2.43, ddd (15.2, 10.7,7.1) 8.5, 5.9)  1.65-1.81,m  •P  2a  116  la  197 1.52, m  205  203 1.77-1.90, m  3.41, sept (7.2) 1.34, 1.35, d (7.2) 1.14, s 1.11,s 1.25, s 3.67, s  -  2.09, dddd (14.1,14.1,3.8, 2.4)  3.26, sept (6.9) 1.18,1.19, d (6.9) 1.06, s 0.88, s 1.17, s 3.70, s  -  3.26, sept (6.9) 1.180,1.189, d (6.9) 1.03, s 0.94, s 1.178, s 3.69, s  -  2.73, ddd (17.8, 10.6, 8.6) 2.99, dd (17.8, 4.9) 7.00, d (8.4) 7.03, d (8.4)  1.62-1.77, m  3.28, dd (10.9, 4.9) 3.49, dd (3.3, 2.4) 1.28, dd (12.5, 1.62-1.77, m 2.0) 1.91, dd (13.2, 1.77-1.90, m 7.5)  1.77, m  1.68, ddd (12.5, 1.68, dddd 12.5, 5.7) (13.2, 12.5, 11.6,5.3) 2.70, ddd (17.6, 2.50-2.71, m 11.6,7.5) 2.99, ddd (17.1, 3.02, dd (17.6, 5.3) 5.7, 1.6) 6.38, s 6.98, d (8.4) O H , 4.47, s 7.03, d (8.4)  1.84, dd (12.5, 2.1) 1.83,m  -  -  2.50-2.71, m  2.33, ddd (13.1, 2.27, ddd (13.1, 2.00, ddd (13.1, 3.8, 3.5) 7.4,4.2) 3.5,3.5) 2.50-2.71, m 1.62-1.77, m 1.77, m  118 1.91, m  'H NMR Spectral Data of Abietane Diterpenes 197,116, 200,117,118, 205 and 203 (400 MHz in CDC13, 8 in ppm, J in Hz in parentheses)  1.39, ddd (12.8, 1.71-1.96, m 12.8, 3.7) 2.27, d (12.8) 2.41-2.64, m  Table 2.3  92  H02C  210  Compound 210 (C20H26O2) was isolated from the polar fraction of the extract (Fr. VII, eluate of methanol; see Figure 2.2). The IR spectrum showed the presence of hydroxyl (3525 cm - 1 ), benzene ring (3034, 1500 cnr 1 ), and possibly a,p-unsaturated carboxyl groups (1685 cm -1 ). The *H NMR spectrum exhibited the AMX system in the aromatic region. Irradiation of proton H14 caused NOE enhancements to signals at 8 2.95 (2H, m, H7) (2%), the septet at 8 2.83 (1H, J = 7.0 Hz, H15) (7%) and the two isopropyl methyl groups at 8 1.22 (6H, d, / = 7.0 Hz, C16, C17 methyls) (1%) (Figure 2.13). The COSY spectrum showed correlations of these two C7 protons to mutually coupled protons (H6) at 8 2.25 (1H, m) and 8 1.71 (1H, ddd, J = 13.1, 10.1, 8.1 Hz). On the basis of the large coupling constants of the 8 1.71 signal, it was logical to assign this signal to the proton H6P (NOE was also observed upon irradiation of the C20 methyl group), and the signal at 8 2.25 to proton H6cc. Saturation of the HI 1 resonance resulted in NOEs to a methyl signal at 8 1.03 (C20 methyl group) , the H12 signal, and a signal (Hip) in the middle of a two-proton multiplet between 8 2.30 and 2.40. The other proton in this multiplet showed NOE when the two C7 protons were irradiated. Also enhanced were the signals of H14 and the two C6 protons. So the proton signal in the multiplet with HI(3 must be due to H5, since H7a experiences a nearly 1,3-diaxial relationship with H5. Proton H i p was strongly coupled with Hloc at 8 1.61 (1H, br ddd, J = 12.7, 12.7, 7.5 Hz), and both protons were correlated with the multiplet between 8 2.42 and 2.63 (2H, H2). The chemical shift of the methyl resonance at 8 2.09 (3H, br d, J = 1.2 Hz) indicated that this  93  methyl group was attached to an unsaturated carbon. The only position left for unsaturated carbons were C3 and C4. The chemical shifts of H2, H5 and H6 suggested that they were in a proximity of a double bond, thus there was a likelihood of a double bond between C3 and C4. Based on the observations of only four methyl group signals in the ^H NMR spectrum, two oxygen atoms in the molecular formula (a polar compound), and an IR absorption at 1685 cm -1 , the remaining functional group of this compound was probably a carboxyl group. There are two possible positions that the carboxyl group and a methyl may occupy within the alkene linkage in this molecule: one is that the carboxyl group is attached to C3 and the methyl to C4 (Figure 2.12a), and the other is just the reverse (Figure 2.12b).  H02CT  y I  i  ^ C02H  a  Figure 2.12  b  Two Possible Structures for 210  Figure 2.13 illustrates the NOE results in the spectral analyses of 210. Irradiation of the multiplet containing HI(3 and H5 resulted in signal enhancements of H l l (by HI (3), H7 (H7a, by H5), the C20 methyl (by HIP), H l a (by H i p and H5), and of the methyl at 8 2.09 (3% for the methyl, by H5). Saturation of the H6cc resonance showed NOEs to H6p (6%), H7 (4% for both protons), and to the methyl group at 8 2.09 (5% for the methyl). However, irradiation of the C2 multiplet did not produce any enhancement of the 8 2.09 methyl signal. These NOE results demonstrated that the methyl group (CI9 methyl) was attached to C4 (close to H5 and almost co-planar with H6a as well) while the carboxyl group was attached to C3.  94  Figure 2.13  NOEs Observed for Compound 210  The COSY spectrum also displayed small cross peaks of the C19 methyl signal to both the H2 (homoallylic coupling) and H5 (long range coupling) signals, providing additional evidence for a double bond between C3 and C4. A similar compound with a hydroxyl group at C15 (182) was previously isolated from TRP4a cell cultures, and 210 was also obtained as an intermediate in the synthesis of 182 by this group. 157 The spectral data here were comparable with those reported for the synthetic intermediate. Compound 210 is a first time isolate from TRP4a cell cultures.  211  Compound 211 (C20H24O3) was also isolated in the polar fraction of the extract. The IR spectrum suggested the presence of hydroxyl (3530 cm -1 ), benzene ring (3033, 1590 cm -1 ), and possible cc,(3-unsaturated carboxyl and aromatic carbonyl (1682 cm-1) functionalities.  95  Its lH NMR spectrum showed some similarities between this compound and 210, but in the aromatic region, the AMX system signals bore a resemblance to compound 199. The large downfield shift of H14 from 8 6.94 in 210 to 7.90 in this compound suggested that there was a keto group at the C7 position as in compound 199. Irradiation of HI 1 at 5 7.35 showed NOE enhancements to a signal at 8 2.44 (1H, br dd, / = 11.2, 6.0 Hz, Hip1), and a methyl signal at 8 1.13, which was assigned to the C20 methyl group accordingly. The COSY spectrum exhibited cross peaks between Hip and H l a at 8 1.71 (1H, m) (strong coupling). Both protons were coupled with C2 protons at 8 2.51 (1H, m) and 2.68 (1H, m), respectively. Irradiation of H l a caused signal enhancements of, not only Hl(3 and H2a but also a proton signal at 8 2.83 (1H, br d, / = 14.2 Hz). This signal was assigned to H5 because of its 1,3-diaxial relationship with H l a . The H5 proton in turn was strongly coupled with two mutually correlated signals at 8 2.59 (1H, dd, J = 18.1, 14.2 Hz) and 8 3.04 (1H, dd, J = 18.1, 4.4 Hz), which were thus assigned to the two C6 protons. No other protons were coupled with these two C6 protons. In consideration of this fact that there were no C7 protons along with the downfield shift of the H6 and H14 signals, it was concluded that C7 contained a keto group. The rest of the protons were assigned according to NOE and COSY results, as well as in comparison with compounds 210 and 199. Similarly to compound 210, the C19 methyl resonance showed small cross peaks to H5 and to both H2 signals. Saturation of the C19 methyl resonance showed NOEs to H6a (7%) and H5 (5%), while irradiation of either H6oc or H5 reciprocated the enhancement of the C19 methyl signal. These results provided supporting evidence for the assignment of this compound as a 7-oxo analog of 210.  Compound 195 was isolated as a colorless oil. It was obtained during the search for compounds that may be present prior to dehydroabietane (196) in the metabolic process of T. wilfordii.  The overwhelming simplicity displayed by its lH NMR spectrum made us feel  somewhat discouraged, yet at the same time very curious. The IR spectrum exhibited virtually only saturated hydrocarbon absorptions at 2950, 1460 and 1390 cm - 1 .  Upon further  96  examination, a small shoulder at 3050 cm -1 and a weak peak at 1670 cm -1 were observed, suggesting that there may be some double bonds in the molecule.  QJ 195  The *H NMR spectrum showed four sets of signals at 5 5.10 (m), 2.03 (m), 1.66 (br s) and 1.58 (br s) in a ratio of 3:10:3:9. The signals at 8 5.10 were likely due to olefinic protons and were correlated with the signals at 8 2.03. These two groups of signals exhibited slight couplings with the other two groups. The mass spectrum showed a molecular ion peak at m/z 410 and other major ion peaks at m/z 367, 341, 81, and 69 (base peak). The base peak suggested a fragment of isopentene and the molecular weight corresponded to squalene. Squalene is a known biosynthetic precursor of triterpenes and steroids.184 Triterpenoids and phytosterols are two groups of compounds which have been widely isolated from both Tripterygium plants and TRP4a cell cultures. High resolution mass spectrometry confirmed its molecular formula as C30H50 and the spectral data (lH NMR, MS) were in agreement with those in the literature.185  Compound 193 was obtained as a solid with a molecular formula of C20H24O3. Its IR spectrum displayed absorption bands due to a benzene ring (3020, 1600, 1500 cm -1 ) and an a,P-unsaturated ester or a butenolide (1745, 1670 cm -1 ). The *H NMR spectrum exhibited AMX system signals in the aromatic region and an AB quartet centered at 8 4.76. The AB quartet was characteristic of a butenolide functional group, and this functionality has been found in many diterpene metabolites from Tripterygium plants, including triptolide (1) and  97  tripdiolide  (2).  These distinct spectroscopic features drew the attention to  isodehydroabietenolide (193), a compound which had been synthesized and extensively studied by our group (see Chapter 3 for the synthesis of this compound). The *H NMR and mass spectra of isolated 193 were identical with those of the synthetic one.  O—J 19  193 (isodehydroabietenolide)  Compound 193 was isolated for the first time from TRP4a cell cultures even though it had been postulated as one of the intermediates in the biosynthetic pathway. 157 It is likely formed by enzymatic hydroxylation of 210 at CI9, followed by a facile lactonization to form the butenolide moiety in the molecule. Literature sources make no mention of the isolation of such a compound from Tripterygium plants.  206 Compound 206 was isolated as a solid with a molecular formula of C20H22O3. Its IR spectrum revealed the presence of a carbonyl group conjugated to a benzene ring (3010, 1580, 1550, 1490 cm -1 ), and of a butenolide (1750, 1670 cnr 1 ) functional group. A strong UV absorption at 255.8 nm (log e 4.05) also indicated an aromatic ketone.  98  The *H NMR spectrum of 206 displayed extremely similar AMX system signals in the aromatic region to those found in compound 211, indicating a carbonyl group at C7. The aliphatic region of the spectrum exhibited some similarities to that of 193. The lack of the C7 protons, along with additional spectral information mentioned above suggested that this compound was the 7-oxo derivative of 193. The H5 resonance was shifted downfield from 8 2.70 in 193 to 3.21, and the H6 resonances from 5 1.91 to about 2.74 in this compound, apparently caused by the presence of a carbonyl group at C7. Comparison of the spectra of isolated 206 with those of synthetic 206 confirmed that they were identical (see Chapter 3 for the synthesis of this compound). This compound has previously been obtained in the biotransformation of isodehydroabietenolide (193) in TRP4a cell cultures.160  0-~"i9  106 (triptophenolide) 20  Compound 106 was obtained as optically active ([oc]D +37.6°, c = 0.25, MeOH), colorless prisms (mp 228-230°C, dec.) with a molecular formula of C20H24O3. Its IR absorptions indicated the presence of hydroxyl (3620 cm -1 ), benzene ring (3025, 1575, 1500 cm -1 ) and butenolide (1750, 1680 cm-1) functional groups. The lH NMR spectrum showed similar AM system signals in the aromatic region as those of 116, and had some similarities to 193 in the aliphatic region, indicating a 14-hydroxy analog of 193. The signal at 8 4.71 (1H, br s, D2O exchangeable) was due to the hydroxyl group attached to C14 because of its NOEs with H7 and H15. The remainder of the protons were assigned according to the COSY and NOE results. All the spectral data were identical  99  with those from an authentic triptophenolide sample obtained from the T. wilfordii plant.* This compound has been isolated from the four species of Tripterygium  plants, 2 but has not  previously been isolated from TRP4a cell cultures.  107  Compound 107 was obtained as a solid with a molecular formula of C21H26O3. The IR spectrum showed absorption bands for benzene ring (3050, 1600, 1500 cm -1 ), butenolide (1740, 1680 cm"1), and possible ether (1280, 1020 cm"1) groups. The *H NMR spectrum was almost identical with that of 106 except that the signal in the aromatic region appeared to be a singlet at 5 7.10 (2H) and the hydroxyl proton signal in 106 at about 8 4.7 was now a methoxy signal at 5 3.72. In consideration of the similar situation with 203 (methoxy group at C14 made H l l and H12 appear as a "singlet", vide supra), it was evident that this compound was the methyl ether of 106. Irradiation of the singlet-like aromatic signal (2H) resulted in signal enhancements of Hl(3 (6%), the C16, C17 methyl groups (4% for two methyls) and the C20 methyl group (1% for the methyl), confirming that the aromatic signals actually arose from H l l and HI2. Other spectral differences from that of 106 were a small downfield shift of the H15 and H7 resonances. This compound has been isolated from the Tripterygium plants, but this is its initial isolation from the TRP4a cell cultures. The spectral data were identical to those in the literature. A comparison of *H NMR spectral data of the isolated a&eo-abietane diterpenes 210, 211,193, 206,106 and 107 is given in Table 2.4.  The authentic sample was generously provided by Professor B.-N. Zhou, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, the People's Republic of China.  2.30-2.40, m 2.25,m 1.71, ddd (13.1, 10.1,8.1) 2.95, m  2.95, m  5 6a  7p  11 12 14 15 16,17 19 20 -OMe  7a  60  7.35, d (8.2) 7.42, dd (8.2, 2.0) 7.90, d (2.0) 2.93, sept (7.0) 1.25, d (7.0) 2.07, s 1.13, s -  7.22, d (8.1) 7.00, d (8.1) 6.94, s 2.83, sept (7.0) 1.22, d (7.0) 2.09, d (1.2) 1.03, s  -  -  7.38, d (8.0) 7.45, dd (8.0, 1.9) 7.93, d (1.9) 2.95, sept (6.9) 1.25, d (6.9) 4.76, m 1.14, s -  -  -  6.93, d (8.2) 7.05, d (8.2) OH, 4.71, s 3.08, sept (7.0) 1.24, 1.26, d (7.0) 4.77, AB a (17.2) 1.01, s  2.73, m  3.28, sept (6.9) 1.20, 1.22, d (6.9) 4.77, ABq (17.3) 1.01, s 3.72, s  -  2.93, ddd (18.2, 10.3, 2.0) 3.05, ddd (18.2, 7.5, 1.9) 7.10, s 7.10, s 2.73, m  -  1.80-2.00, m  1.90, m  2.738, d (11.3)  2.67, d (13.1) 1.80-2.00, m  2.37, m  107 1.69, ddd (12.2, 12.2, 6.3) 2.44-2.56, m 2.44-2.56, m  2.69, d (13.4) 1.99, m  2.39, m  106 1.68, ddd (12.3, 12.3, 6.5) 2.44-2.57, m 2.44-2.57, m  3.21, m 2.743, d (7.8)  2.43, m  206 1.90, ddd (11.6, 11.6,6.5) 2.58, m 2.58, m  7.25, d (8.1) 7.03, dd (8.1, 1.5) 6.96, d (1.5) 2.82, sept (6.9) 1.22, d (6.9) 4.76, AB a (17.2) 1.01  3.00, m  2.83, d (14.2) 2.70, m 3.04, dd (18.1,4.4) 1.91, m 2.59, dd (18.1, 1.91, m 14.2) 3.00, m  2.37, m  2.51, m  2.42-2.63, m  2a  IP  la  2P  211  193 1.69, ddd (12.4, 12.4, 6.5) 2.44, dd (11.2, 6.0) 2.50, m 2.68, m 2.50, m 1.71, m  J in Hz in parentheses)  'H NMR Spectral Data of afoo-Abietan Diterpenes 210, 211,193, 206,106 and 107 (400 MHz in CDC13, 5 in ppm,  210 1.61, ddd (12.7, 12.7,7.5) 2.30-2.40, m 2.42-2.63, m  Table 2.4  101  109  Compound 109 was isolated as a pale yellowish powder with a molecular formula of C21H26O4. It was optically active ( [ a | 4 +113.4°, c = 0.307, MeOH), and the IR spectrum established the presence of hydroxyl (3600 cm -1 ), butenolide (1745 cm -1 ), benzene ring (3020 cm -1 ), and ether (1240,1030 cm-1) functionalities. The *H NMR spectrum exhibited an aromatic signal at 8 6.39 (s), a hydroxyl proton at 8 4.58 (s, D2O exchangeable) and a methoxy group at 8 3.66, similar to compound 117. Irradiation of the proton at 8 6.39 showed NOEs to the hydroxyl proton at 8 4.58 and the two isopropyl methyl groups at 1.18 (6H, d, J = 6.9 Hz), suggesting that the irradiated proton was H12 and the hydroxyl group was at CI 1 as in 117. Saturation of the methoxy resonance at 8 3.66 resulted in signal enhancements of H15 at 8 3.23 (1H, septet, J = 6.9 Hz) and a proton at 8 3.06 (1H, br dd, / = 17.9, 4.6 Hz), confirming that the methoxy was attached to C14, and the enhanced signal at 8 3.06 was H7(3. The downfield-shifted signal at 8 3.51 (1H, ddd, / = 13.5, 5.4, 2.0 Hz) due to the presence of a hydroxyl group at C l l was assigned to H i p . The remainder of the protons were assigned accordingly as based on COSY and NOE results. The structure of this compound was determined as ll-hydroxyl-14-methoxy-18(4^3)a^eo-abieta3,8,1 l,13-tetraen-18-oic acid lactone (108).  This compound has been isolated from  Tripterygium plants, but its isolation here constitutes the first isolation from TRP4a cell cultures.  102  108  Compound 108 was isolated as optically active ([oc^ -36.6°, c = 1.82, CHCI3), colorless needles (mp 202-204°C) with a molecular formula of C20H22O4. Its UV spectrum showed a strong absorption at 268.4 nm (log £ = 4.02), indicating an aromatic ketone chromophore. The IR spectrum suggested the presence of hydroxyl (3700, 1240 cm -1 ), butenolide (1750, 1670 cm -1 ), and aromatic carbonyl (1620 cm-1) functionalities. The lH NMR spectrum of 108 revealed some similarities with 106, such as the AM system signals in the aromatic region. The H7 signals were not found in the spectrum. This suggested that C7 contained a keto group, which was consistent with the UV and IR spectral data. The close proximity of this carbonyl to the C14 hydroxyl proton allowed for the possibility of a hydrogen bonding, thus this active proton (D2O exchangeable) was found at a very low field at 8 13.3. The H6 signals always appear as sharp peaks around 8 2.7-2.8 when there is a keto group at C7 for compounds of the isodehydroabietenolide series. The doublet at 8 2.77 (1H, J = 7.1 Hz) was assigned to H6oc while the doublet at 8 2.78 (1H, J = 11.7 Hz) to H6P since irradiation of the C20 methyl group greatly enhanced the doublet at 8 2.78. This compound has previously been isolated from the Tripterygium plants, but was obtained for the first time from TRP4a cell cultures. A comparison of the spectral data confirmed its identity.  Compound 138a (structure on next page) was isolated as a solid with a molecular formula of C20H24O4. The presence of hydroxyl (3600, 3400, 1260,1040 cm -1 ), benzene ring (3050, 1630, 1580, 1500 cm 4 ) and butenolide (1745, 1680 cm -1 ) groups was indicated by its IR absorptions.  103  138a  The *H NMR spectrum of 138a was similar to that of 106, but no H7 signals were observed in the normal region around 8 2.8-2.9. However, a new signal, probably due to a benzylic or ally lie proton geminal to a hydroxyl group, appeared at 8 5.22 (IH, ddd, J = 10.0, 7.8, 7.8 Hz). Figure 2.14 depicts the major results from COSY and NOE experiments. The signal at 8 5.22 showed cross peaks to signals at 8 3.14 (IH, br d, J = 7.8 Hz), 2.33 (2H, m), and 1.99 (IH, ddd, / = 13.9, 12.7, 10.0 Hz). The H5 resonance, which was easily located once it was compared with the spectrum of 106, was found at 8 2.76 (IH, br d, J = 13.8 Hz). The strong coupling shown between H5 and the proton at 8 1.99 suggested that this proton was H6p. Both H5 and H6|3 were also coupled with part of the two-proton multiplet at 8 2.33, indicating that one of the protons in this multiplet was H6oc. Because the signal at 8 5.22 was only coupled with two C6 proton, it was assigned to proton at C7.  Figure 2.14  Major COSY and NOE Results Observed for 138a  104  The signal at 8 3.14, which was also coupled with the H7 signal, was assigned to the proton of the hydroxyl group at C7 since its chemical shift changed with the concentration of the sample. Irradiation of H7 at 5 5.22 resulted in NOEs to H5 (5%) and H6a (6%), indicating that H7 was a-oriented, and thus the compound was a C7 p alcohol. Other signals that were enhanced due to irradiation at 8 5.22 were the 7-hydroxyl proton at 8 3.14 and another one at 8.22, the latter was believed to be the proton of the C14-hydroxyl group. Saturation of the C7 hydroxyl resonance showed NOEs to H7a (11%) and H6|3 (4%). These results proved that the assignment of the P hydroxyl group at C7 was correct. The mass spectrum of compound 138a showed a very weak molecular ion (0.5%), 145 but moderate M - 2 and quite strong M - H2O and M - H2O - CH3 ion peaks. The identity of this compound was finally confirmed by comparing it with the reduction product of 108. Compound 108, treated with sodium borohydride in ethanol at room temperature for a couple of hours, gave 138a (p alcohol 145 ) as the major product, and a small amount of mixture containing both 7a (138b) and 7p alcohols.  108  138a 7p-OH 138b 7oc-OH  Compound 207 (structure on next page) was obtained as a solid. High resolution mass spectrometry showed a molecular formula of C21H26O4. The *H NMR spectrum of 207 exhibited great similarities with 138a, with the exception of a methoxy group at 8 3.83 and the lack of a C14 hydroxyl proton. Furthermore, the split between AM system signals in the aromatic region was much smaller than in 138 (one signal at 8 7.20 (1H, d, J = 8.4 Hz) and the  105  other at 5 7.12 (1H, d, J = 8.4 Hz)), suggesting a methoxy group at C14, as compared with 107 and 106. If the hydroxyl group at C7 was methylated, then the H7 signal was expected to move upfield by approximately 0.2-0.4 ppm. 168,186 However, the chemical shift of H7 was virtually unchanged, so this compound was proposed as the C14 hydroxyl methyl ether of 138a and the remaining protons were assigned accordingly.  207  The structural assignment of 207 was supported by NOEs to the methoxy group, H5 at 8 2.72 (1H, br d, J = 14.0 Hz), H6a at 8 2.32 (1H, m), and the hydroxyl group at 8 1.5 (buried in the water signal), caused by irradiation of H7a at 8 5.25 (1H, dd, J = 9.0, 9.0 Hz). Irradiation of the methoxy group increased the intensities of H7a and H15 signals as expected. The lU NMR data for 109,108,138a, and 207 are summarized in Table 2.5.  Tl (1)  Td (2)  Tl (1) and Td (2) were isolated as colorless crystals and their spectra were identical with those obtained from authentic samples.  106  !  Table 2.5  H NMR Spectral Data of afoo-Abietane Diterpenes 109,108,138a and 207 (400 MHz in CDCI3, 8 in ppm, J in Hz in parentheses)  109  138a  108  207  la  1.53, m  1.79, ddd (12.4, 12.4, 5.9)  1.64, ddd (12.6, 12.6, 6.6)  1.65, m  1(3  3.51, ddd (13.5, 5.4, 2.0)  2.51, dd (12.4, 6.0)  2.44, m  2.27-2.53, m  2a  2.38, m  2.57,m  2.44, m  2.27-2.53, m  2(3  2.38, m  2.40, m  2.33, m  2.27-2.53, m  5 6a  2.76, m  3.16, m  2.76, d (13.8)  2.72, d (14.0)  1.77, m  2.77, d (7.1)  2.33, m  2.27-2.53, m  6P  1.77, m  2.78, d (11.7)  1.99, ddd (13.9, 12.7, 10.0)  1.93, ddd (14.0, 13.0, 9.0)  7a  2.76, m  -  5.22, ddd (10.0, 7.8,7.8)  5.25, dd (9.0, 9.0)  7P  3.06, dd (17.9,4.6) -  OH, 3.14, d (7.8)  -  11  OH, 4.58, s  6.86, d (7.9)  6.85, d (8.2)  7.12, d (8.4)  12  6.39, s  7.40, d (7.9)  7.13, d (8.2)  7.20, d (8.4)  14  -  OH, 13.3, s  OH, 8.22, s  -  15  3.23, sept (6.9)  3.33, sept (7.0)  3.28, sept (6.9)  3.27, sept (6.9)  1.18, d (6.9)  1.21, 1.23, d (7.0)  1.20, 1.22, d (6.9)  1.18,1.27, d (6.9)  19  4.77, AB a (17.2)  4.75, AB a (17.1)  4.75, AB a (17.2)  4.77, m  20  1.14, s  1.12, s  1.08, s  1.10, s  -OMe  3.66, s  -  -  3.83, s  16,17  Compound 212 (structure on next page) was isolated as a colorless powder with a molecular formula of C20H22O7. Its IR spectrum indicated the presence of hydroxyl (3270 cm" l  ), butenolide (1757 cm -1 ) and keto (1731 cm -1 ) functional groups. The *H NMR spectrum of 212 displayed a very similar spectrum to that of Td (2), but  the H14 signal was not present and the HI 1, H12 resonances were shifted downfield by 0.15  107  and 0.3 ppm, respectively. The H7 and HI5 signals were also shifted downfield by 0.05 and 0.15 ppm. This was a clear indication of the presence of a keto group at C14. The other proton signals were found at almost the same positions in the spectrum as those of Td (2). The H2oc signal at 8 4.62 (1H, m) was correlated with HI a at 5 1.53, which, in turn, was coupled with H i p at 8 1.88 (1H, br d, J = 14.1 Hz). The HI l a at 8 4.06 (1H, d,J= 2.9 Hz) was coupled with H12p at 8 3.81 (1H, d, J = 2.9 Hz). The H7cc resonance at 8 3.40 (1H, d, J = 5.4 Hz) showed cross peaks to H6oc at 8 2.25 (1H, br ddd, J = 15.0, 5.8, 5.4 Hz), which was coupled with H6p at 8 2.09 (1H, br dd, / = 15.0, 13.4 Hz). Both H6 signals showed correlations with H5 at 8 2.74 (1H, br dd, / = 13.4, 5.8 Hz). The C19 protons appeared at 8 4.76 (2H, m) not as a broad AB quartet as in 106, but as a sharp peak, a feature typical of Td (2) and Tl (1) related compounds.  212  triptonide (95)  Triptonide (95), the "Tl version" of compound 212 which has no hydroxyl group at C2, has been isolated from the Tripterygium plants, but compound 212 has never before been isolated from either the Tripterygium plants or from TRP4a cell cultures. The ! H NMR data were comparable with those of triptonide (95) 187 except for the H2 related signals. This new compound is named triptolonide.  108  208  Compound 208 was obtained as a colorless solid with a molecular formula of C22H26O8. Its IR spectrum displayed absorptions for hydroxyl (3600 cm"1), butenolide (1752, 1680 cm -1 ), and possible ester (1752, 1234 cm-1) functionalities. The *H NMR again exhibited a very similar spectrum to that of Td (2), except that the H14 resonance was not found around 8 3.4, but a new signal appeared at 5 5.06. An acetate methyl signal was found at 8 2.15. Irradiation of the proton at 8 5.06 (1H, s) resulted in signal enhancements of H7a at 8 3.44 (1H, d, / = 5.6 Hz), of the methyl at 8 2.15 and of the two isopropyl methyl groups at 8 0.82 and 0.94 (3H each, both d, / = 7.0 Hz). NOEs were observed between H7a and the proton at 8 5.06, and H6a at 8 2.20 (1H, ddd, J = 14.7, 5.8, 5.6 Hz). The signal at 8 5.06 was also enhanced (together with H15) upon irradiation of the isopropyl methyl signals. These NOE results along with those from molecular model analysis confirmed that the proton at 8 5.06 was the HI4a proton, and it was about 1.7 ppm downfield due to the ester group at C14.168  208  109  The structure elucidation of 208 was further supported by mass spectrometry. The mass spectrum of this compound showed the molecular ion peak at m/z 418(1.1%) and the M - H2O ion peak at m/z 400 (4.1%). Cleavage of either CH3CO or CH3COO group from the molecular ion or from the M - H2O fragment was seen at m/z 375 (1.1%), 359 (1.3%) or 357 (1.3%), 341 (2.0%), respectively. The structure of the new compound 208 was thus assigned as the C14acetate of Td.  209 Compound 209 was isolated as an optically active ([aj^4 -225.9°, c = 0.251, CHCI3) white powder with a molecular formula of C20H22O6. Its IR spectrum showed presence of hydroxyl (3600 cm -1 ), butenolide (1755 cm"1) and a,p-unsaturated keto groups. The UV spectrum had a strong absorption at 255.8 nm (log e = 3.57). Its !H NMR spectrum revealed some similarities with that of Td (2). The signals of the protons on ring A and B were little changed. The C19 protons were found at 8 4.76 (2H, m, sharp), and H2a at 8 4.63 (1H, br d, J = 5.9 Hz). Proton H5 appeared at 8 2.66 (1H, br dd, J = 13.1, 5.7 Hz), and the H6oc and H60 resonances were located at 8 2.24 (1H, ddd, J = 14.6, 5.7, 5.3 Hz) and 2.13 (1H, br dd, J = 14.6, 13.1 Hz), respectively. The differences between 209 and Td (2) were seen by their ring C proton signals. There were only two doublets present in the region from 8 4.0 to 3.3, whereas four doublets can be found for Td (H7, HI 1, H12, and H14). The H15 signal was shifted downfield from 8 2.20 to 2.83 (1H, septet d J = 6.8, 1.0 Hz) as compared with Td (2). A new signal was found at 8 6.91  110  (IH, dd, / = 4.8, 1.0 Hz), which was coupled with a proton at 8 3.82 (IH, d, J = 4.8 Hz) as shown by the COSY spectrum (Figure 2.15, a).  JUL  A_JUL_AA_^/_A  2.0 .  a  «' *  _  4.0  _  6.0  #  • *  *  PM I I r r i r r t p r r i n n r n i r n r r r f t u n n i ITI r m T m fi 1111 n 1111 M I n i n | ri  6.0  4.0 PPH  2.0  a, H12/Hllcc; b, H2a/Hla; c, H19/H5; d, H7a/H6a; e, H15/H16.H17; f, H5/H6a,H6p; g,Hlp/Hla;h,H6a/H6p Figure 2.15  COSY Spectrum of Compound 209  Ill  Irradiation of the proton at 5 6.91 resulted in signal enhancements of the proton at 8 3.82 (5%), H15 (2%) and the isopropyl methyl groups (10% for two methyls) (Figure 2.16a). In addition, irradiation of the proton at 8 3.82 showed NOEs not only to the proton at 6.91, but also to Hip (4%) and H l a (8%) (Figure 2.16c). These two irradiations demonstrated that the signal at 8 6.91, which must be from a proton attached to an unsaturated carbon, was due to H12, while the one at 8 3.82 was H l l . H l l was believed to be an a proton on the basis of biogenesis, and its orientation was subsequently confirmed by the NOE experiments ( H l a was more enhanced than HI(3 when H l l was irradiated). Irradiation of the proton at 8 3.59 showed NOEs to H6oc (5%) and H6|3 (1%), indicating that the irradiated proton was the H7a proton (Figure 2.16d). There was no H14 proton in this compound since irradiation of H7a should have enhanced this H14a signal. Furthermore, in conjunction with information provided by IR and UV spectra, along with the fact that H15 and H7a signals were shifted downfield significantly in comparison with Td (2), it was concluded that C14 was a carbonyl group and thus formed an a,p-unsaturated ketone system with a double bond between C12 and C13. The small coupling (1 Hz) that existed between HI2 and H15 (allylic coupling, not shown in Figure 2.15) also provided additional support for this assignment. In order to assign the carbon signals in the 13 C NMR spectrum, HMBC (Heteronuclear Multiple Bond Connectivity) and HMQC (Heteronuclear Multiple-Quantum Coherence) experiments were performed on a Bruker AMX-500 (500 MHz) instrument. The HMQC spectrum (Figure 2.17) showed cross peaks of isopropyl methyl proton signals to two carbon signals at 8 21.3 and 21.5, indicating that these two carbon signals arose from C16 and C17 (Figure 2.17,1). The angular methyl proton signal shared a cross peak with a carbon signal at 8 15.4, suggesting that the carbon signal was due to C20 (Figure 2.17, k). Both HI proton signals were correlated with a carbon signal at 8 38.6, so this carbon must be CI (Figure 2.17, i, j). Similarly, C2, C5, C6, C7, C l l , C12, C15 and C19 were found at 8 59.43, 41.3, 23.1, 61.9, 52.2, 136.5, 27.1 and 70.1, respectively.  112  »Hl»»»M«)  TV  ^Wfcf * . ^ » > . « w »  OOyfrMffW t ^ ^ ^ i ^ i h j f ^ ^ f r M i M J l V ^ l y  • I.! • • • • I . * * I , V•"•••V  • |«w  mJb  yHViW^*  i  L  1 ^v*!*^**m  —J^V—._H  «»II,WI.«I  ty*..  wtmmrmJ+4***,  UJ  ^4i_ IL-wv^ ^vj^w——  vvr  i.i^U't ,iL«»  .. •  nwiwMto^l «Mmi»i» l 'f IWINMIJM'I *'<*•»"« ^ * ***< " "**"t '"« '"'' ' " y ttyWHI* w>>>iv»W«^M  . i . •i.»-*^ u WUttw*s^^y^\.  4r  -yUS  /4»--v  H20  H16.17  HI9  Htl HI2  H60 H7  H6a| HIP H2a  LJL. ' I ' ' • ' I ' ' ' ' I 7-5 7.t 6.5  I 6.8  S.S  i • i , i i i . • i i • S.i 4.5 4.8 PPM  3.S  Hk  jfaWUOU^ S.I  i , i , , i , , , , , , , , , , , i , , ,„, 2.5 2.8 1.5 1.8 .5  a, {H12}: Hlla, H15, H16,H17; b, {H2a}: OH, H1B, Hla, c, {Hlla}: H12, HlB, Hla; d, {H7a}: H6a, H6B; e, {C2-pOH}: H2a, H20; j , Off-resonance spectrum of 209 (400 MHz, CDCI3). ({irradiated}: enhanced)  Figure 2.16  NOE Difference Spectra of Compound 209 (continued to the next page)  113  .  -J-  I, i ' i l l i  r\r\  VHv>—'"••"  S / U «•.,-,.<•.,  mi >—•^.i,i>i w . r i„ t iwn.^ INH.II ••,»>»>  t>U<MW**> *n0*m*mJ\t*S*mi***iimm*  i^.^i^-rTijtJjA/ir  *--h i, i~ Jt iA"i^ii'r  -••»Wii.AN^  ^"•'r  il  \VVTT^,N*  i***T  A ». » . ) y . I » » I I I > I •.**.«. < * S ^ r * t ^ V  I^  **JWNW^^«*  JpjiMfitr1 r - T i l i  » > . * ^ » I V I . i" «t»i.« »<'•«• i W i  H20  H16.17  HI9  H6P  HII H7 HI2  1  H6a| Hip H2ct  • i • ' • • i • • ' • i • ' ' • I • • ' • I • " • • I ' ' ' " r 7.5 7.» 6.5 6.S S.S S.I 4.5  JUL  H  )  a  InALAiUl  i'• ' ' ' i • ' • • i • 4.0 3.S 3.1 PPH  1  I ' ' ' • I 2.5 2.C  ' I ' ' ' ' I ' ' ' ' I • ' ' ' I  i.s  i.a  .s  f, {Hip}: H2cx, Hlloc, (Hlcc), H20 ; g, {Hla}: H2a, Hlla, H5, Hip; h, {H20}: H6p, Hip; i, {H16.H17}: H12, H15; j , Off-resonance spectrum of 209 (400 MHz, CDCI3). ({irradiated}: enhanced)  Figure 2.16  NOE Difference Spectra of Compound 209 (continued)  The HMBC spectrum was helpful in determination of those quaternary carbon signals. Figure 2.18 highlights correlations between the quaternary carbon resonances and the assigned  114  proton signals. It showed that the signals of angular methyl protons, the two HI, the H6oc and H2a protons were all correlated with a carbon signal at 8 35.1 (Figure 2.18, c, i, j , k, 1). Since the number of bonds over which the hydrogen-carbon long range coupling can be detected is usually two or three in these experiments, the only carbon that has two or three bonds connections with protons H20, HI (two bonds) and with the protons H2a and H6a (three bonds) is CIO. So the carbon signal at 8 35.1 was due to the carbon CIO. Because the geometry between these bonds, i.e., the coupling path, is important for the transfer of this type of long range coupling,172 not all the protons within two or three bonds exhibit cross peaks to this carbon (for instance, H6p\ H5 and HI l a in this case), but all the protons showing cross peaks should be within that 2-3 bond range. The carbon signal at 8 63.9 was correlated with Hip, the angular methyl protons, and H12 (Figure 2.18, o, r, a), suggesting that it was the C9 signal. The C3 resonance was found at 8 127.2 because it had correlations with the Hip, H2oc and the H19 signals. The carbon signal at 8 172.6 was assigned to CI8 since all butenolide compounds of this series show the C18 signal around 8 173 although no cross peaks were observable to any protons in this case. The rest of the carbon signals were assigned readily except for the C8 signal. Of the 20 carbons, only the C8 signal was not easy to locate in the HMBC spectrum. The C2 carbon signal at 8 59.43 exhibited a cross peak to Hip, which is connected to C2 with two bonds. However, C2 seemed to have another cross peak to H6P that is five bonds away. Upon closer examination of the "cross peak" between C2 and H6p (Figure 2.18, m), it was found that this cross peak corresponded to a carbon signal at a slightly higher field than that of C2, suggesting that this cross peak was actually caused by H6P with another carbon identified to be C8 (three bonds). This signal was so weak and close to the C2 signal that it was hard to find by routine 13 C NMR spectra recorded on instruments with low or moderate magnetic field. A 13 C NMR spectrum run on the AMX-500 spectrometer with expansion of the spectrum finally revealed the C8 signal at 8 59.38, right near the foot of the C2 signal at 8 59.43. Since  115  C8 is in the center of the rigid epoxide functions and has fewer protons nearby, it probably has a longer relaxation time (Tj) and thus results in a very weak signal.  Ai  L  JLJULA)U[JL__I  S\0  3-  s  \  sir 8-  0 C Q-  G5 5-  111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 i i 11  ppa  8  6  a, C12/H12; b, C19/H19; c, C2/H2ot; d, Cll/Hll; e, C7/H7; f, C15/H15; g, C5/H5; h, C6/H6; i, Cl/Hlp; j , Cl/Hla; k, C20/H20; 1, C16,C16/H16,H17  Figure 2.17  HMQC Spectrum of 209  ppn  - 50  -100  0 —  d ••  P. f •  si . s ••  h - q.  e ••  •  -150  m  b. 1111111111111•111111111 • 11111111 11111111111111111111 [ 1111111111111111111111111 n 111111 j 1111111  PP«  6  4  2  PPm  0  a, C9/H12; b, C14/H12; c, C10/H2a; d, C3/H19,H2a; e, C4/H19,H2a; f, C13/Hlla; g, C13/H15; h, C4/H5; i, C10/H6a; j , C10/H1P; k, ClO/Hla; I, C10/H20; m, C8/H6P; n, C2/Hlp; o, C9/Hlp; p, C3/Hlp; q, C4/H6p; r, C9/H20; s, C13/H16.H17  Figure 2.18  HMBC Spectrum of 209  2P 5  3.90, d (3.2) 3.49, dd (3.2,0.8) 3.38, s O H , 2.74, s 2.20, m  3.87, d (3.1)  3.49, dd (3.1, 0.8)  3.39, d (10.5)  O H , 2.71, d (10.5)  2.22, sept (7.0)  -  4.65,s 1.09, s -  -  -  -OAc  3.34, d (5.3)  3.34, d (5.5)  0.85,0.98, d (6.9) 4.74, s 1.29, s  19 20  O H , 2.35, s  4.63, d (5.9)  1.97, d (13.9)  -  0.87,0.96, d (6.9) 4.76, m 1.26, s  2.38, sept (6.9)  -  -  3.81, d (2.9)  4.06, d (2.9)  -  3.40, d (5.4)  2.15,s  0.82, 0.94, d (7.0) 4.74, s 1.24, s  1.860, sept (7.0)  -  5.06, s  3.51, d (3.1)  3.83, d (3.1)  -  3.44, d (5.6)  -  2.83, sept d (6.8, 1.0) 0.99, 1.05, d (6.8) 4.76, m 1.30, s  -  -  H12, 6.91, dd (4.8, 1.0)  3.82, d (4.8)  -  3.59, d (5.3)  2.62, ddd (13.1, 6.3, 2.74, dd (13.4, 5.8) 2.60, dd (13.0, 5.8) 2.66, dd (13.1, 5.7) 1.5) 2.15, ddd (14.8, 5.9, 2.20, m 2.25, ddd (15.0,5.8, 2.20, ddd (14.7, 5.8, 2.24, ddd (14.6, 5.7, 5.5) 5.4) 5.3) 5.6) 1.94, dd (14.8, 13.3) 2.05, dd (14.7, 13.1) 2.09, dd (15.0, 13.4) 2.00, dd (14.7, 13.0) 2.13, dd (14.6, 13.1)  2.67, m  -  2.11,m -  4.59, d (5.8)  4.62, m  -  1.862, d (14.1)  209  1.43, dd (14.1, 5.8) 1.52, dd (13.9, 5.9)  208  1.88, d (14.1)  Td(2) 212 1.41, dd (13.8, 5.6) 1.53, m  1.19, ddd (12.4, 12.4, 5.9) 1.53, ddd (12.4, 5.4, 1.83, d (13.8) 1.2) 2.29, d (18.2) 4.58, d (5.6)  11(1)  parentheses)  H NMR Spectral Data of Epoxides Tl (1), Td (2), 212, 208 and 209 (400 MHz in CDC13, 5 in ppm, J in Hz in  l  16, 17 0.86,0.98, d (7.0)  14a 14p 15  12(3  11a  6P 7a 7P  6a  2a  IP  la  Table 2.6  118  All the spectral evidence confirmed the structure of 209 as (7,8)p\(9,ll)P-bis(epoxy)2p\19-dihydroxy-14-oxo-18(4—»3)a&e<9-abieta-3,12-dien-18-oic acid lactone (named triptolenonide), a new compound that has not been reported in the literature. A comparison of *H NMR spectral data of isolated epoxides Tl (1), Td (2), 212, 208, and 209 is provided in Table 2.6. In summary, 29 compounds were isolated from the ethyl acetate extract of the TRP4a cell cultures, 28 of them diterpene metabolites. Of the 29 compounds, 26 were isolated for the first time from the plant cell cultures of T. wilfordii, and 20 were not yet isolated from Tripterygium species. Eight compounds (200,202,205,207,208, 209, 211 and 212) were new compounds and 6 compounds (138a, 193,197, 201, 206 and 210) were isolated for the first time as natural products.  Plausible Biosynthetic Pathways  For chemists, the motivation for the use of plant cell cultures for the production of biologically active compounds is the understanding of the corresponding secondary metabolic processes, so that one may logically and consistently control and manipulate the relevant parameters. Secondary metabolites isolated from a plant or from a plant cell culture are not necessarily biosynthetic intermediates, but as more metabolites are isolated, there is an increasing chance of isolating the true intermediates. A biosynthetic pathway has to be ultimately proven by labeling experiments etc., however a proposed biosynthetic pathway based on isolated metabolites often provides a frame or a basis on which detailed labeling studies may be designed or conducted. As illustrated in the previous section, our efforts in the isolation of diterpene metabolites from cell cultures of T. wilfordii have resulted in addition of many new and related compounds to the list of diterpene metabolites from Tripterygium. Study of these structurally  119  related compounds may provide important information about their metabolic relationships, and their possible role in the biosynthesis of the triepoxides, Tl (1) and Td (2).  CH2OH  geranylgeraniol  labdane-type  abietane-type  pimarane-type  H02C C02H  dehydroabietic acid (157)  210  Td(2)  Scheme 2.4  Proposed Biosynthetic Pathway to Td (2) via Dehydroabietic Acid (157)  The isolation of dehydroabietic acid (157) and the hydroxyl ester, 182, as co-occurring metabolites from TRP4a cell cultures in an earlier study, led to proposals of the biosynthetic  120  pathway of Tl (l). 1 5 9 A biosynthetic pathway from geranylgeraniol through dehydroabietic acid (157) and the acid 210 was proposed (Scheme 2.4). Alternatively, a methyl group transfer was proposed from dehydroabietane (196) followed by oxidation and butenolide formation, prior to oxidation of ring C to give Tl (1) (Scheme 2.5).  196  H02C  210  Td(2) Scheme 2.5  Proposed Biosynthetic Pathway of Td (2) via 4—»3 Methyl Transfer  The latest biosynthetic pathway was proposed by M. Roberts based on his work in this project (Scheme 2.6). 160  We now have successfully isolated some of these proposed  intermediates, such as 210, 193 in Scheme 2.4, 196, 210,193 in Scheme 2.5, and compounds  121  197,106 and 138 in Scheme 2.6. However, with the discovery of additional metabolites, the proposed biosynthetic pathways for Tl (1) and Td (2) production may be amended and expanded to accommodate all current data.  Td (2) *  Scheme 2.6  Tl(l)  Proposed Biosynthetic Pathway to Tl (1) and Td (2) by M. Roberts  The diterpenes isolated from the TRP4a cell cultures may be classified into five groups. The first group consists of only one compound, dehydroabietane (196), which has no other oxygenated functional groups in any part of the molecule. The second group contains  122  compounds that have various degrees of functionalization on ring A. Examples are 198, 202, 204, 210 and 193. The third group includes compounds with functional groups on either ring C or B, such as 197 and 199.  The fourth group is comprised of compounds that are  functionalized on rings A and C (some on ring B as well), but exhibit no epoxide functionalities. Examples are 116, 200,203,205, 211,206,106,107,108 and 138a. The fifth group contains compounds possessing a lactone moiety on ring A and epoxide groups on rings B and C. This group includes compounds like Tl (1), Td (2), and the newly isolated 208, 209 and 212. Examples of Group 2 compounds:  210  Examples of Group 3 compounds:  193  123  Examples of Group 4 compounds:  The biosynthesis of dehydroabietane (196) from geranylgeraniol was not the focus of our interests, but rather, the biosynthesis from dehydroabietane to Tl (1) or Td (2) was more appealing to us. It is both probable and logical that dehydroabietane (196) is the common precursor for nearly all the diterpene metabolites in T. wilfordii. Dehydroabietic acid (157), previously isolated from the TRP4a cell cultures, likely resulted from dehydroabietane (196). In consideration of the lactone moiety on ring A of Tl (1) and Td (2), it is likely that the lactone carbonyl group is derived from the methyl group that has migrated from C4 (as suggested by  124  Scheme 2.5). This kind of methyl migration is a very common occurrence in natural product biosynthesis. 184 Migration of a carboxy group (such as that from C4 to C3 in dehydroabietic acid, Scheme 2.4) is rarely seen in the literature.  Examples of Group 5 compounds:  Isolation of Group 2 and Group 3 compounds reveals that the functionalization of ring A or ring C may possibly be the initial step in this biosynthetic pathway, or perhaps the functionalizations of ring A or ring C may actually take place simultaneously.  The  biosynthesis of Tl (1) and Td (2) at this stage is not a linear process. Isolation of Group 4 compounds gives us evidence that the functionalization of ring A may proceed after the functionalization of ring C and vice versa. No major differences are found between ring A functional groups in Group 2 and 4, and also no major differences occur between ring C functional groups in Group 3 and 4. So dehydroabietane (196) may first be functionalized on ring A then C, or first on ring C and then A, both routes producing Group 4 compounds. Some Group 4 compounds may be the intermediates to Group 5 compounds. The lactone moiety on  125  ring A seems to be necessary for the enzymatic formation of the epoxide functionality on ring C or B, because of all the Group 5 diterpenes isolated thus far from either the plants or from the cell cultures, none of them have an epoxide without already containing the lactone moiety on ring A. Scheme 2.7 outlines the general relationship and transformation between these five groups of compounds.  Group 1  ring A functionalized  dehydroabietane (196)  Group 3  Group 2  Group 4  ringC functionalized  both ring A and C functionalized  •  Group 5  Scheme 2.7  epoxides  Relationship between Different Groups of Compounds  The enzymatic formation of the lactone can start with a hydroxyl group at C3 position (Scheme 2.8). Loss of the hydroxyl group and migration of the methyl from C4 leads to the alkene 213. AUylic hydroxylation at C18 generates the alcohol 214, which is subsequently oxidized to the acid 210. Another allylie hydroxylation at C19 produces the hydroxyl acid 215, which readily lactonizes to form the butenolide 193.  enz.-HO  R=H  R=H  (214 (i))  R = OH (213(B))  R = OH (214(H))  HOOC  (213(i))  HOOC  R=H  (210)  R = OH (210(H))  R = H (215©) R = OH (215(H))  R=H  (193)  R = OH (106)  Scheme 2.8  Proposed Mechanism for Enzymatic Formation of the Butenolide  127  As shown in Scheme 2.8, elimination of the hydroxyl group followed by migration of the methyl from C4 in the first stage may take place in a concerted manner, or in steps. The trans diaxial relationship between the leaving group and the migrating group is favored for such a rearrangement from a chemical point of view. 188 Therefore, only the 3a-hydroxyl isomer is a more plausible precursor of the butenolide. Usually enzymatic reactions are stereospecific,184 therefore the hydroxylation of dehydroabietane should yield only one isomer, such as the 3 a hydroxyl isomer. This compound should then be the preferred substrate for the following enzymatic transformation. However, both 3oc-alcohols (202, 203) and 3p-alcohols (204, 205) were isolated from the cell cultures. One possibility for this product distribution is that the formation of the 3-hydroxyl compound is not so specific and the next step is also not specific, so that the enzyme can convert both alcohols to a same intermediate, which is then proceeded by a methyl migration.  Another postulation is that the enzyme reaction is  stereospecific in hydroxylating dehydroabietane (196) to form only the 3a-alcohol 202, which is then transformed to the butenolide. However, if the rearrangement step is not strictly concerted, it is reasonable that a water molecule from the surroundings can act as a nucleophile and substitute the leaving group, thereby forming the (3 alcohol (Figure 2.19). On the basis of this hypothesis, the 3p-alcohol is formed as a side product prior to the methyl migration. The "wrong" P alcohol isomer may probably remain there, or be oxidized to a ketone (198, 200). The ketone could be a dead end in terms of further biosynthesis, or it could be reduced to the "right" alcohol by certain enzymes in the system and re-enter the biosynthetic process. Functionalization of ring C is probably via a direct hydroxylation through an arene oxide. 184 All the mono hydroxylation occurs at C14. The second phenolic hydroxyl group could enter at either C l l (mostly) or CI2. This suggests that the C14 hydroxyl group is the starting point for further transformations in the C ring. It is interesting to note that in most cases the C14 hydroxyl group is methylated as the methyl ether, and that in most 12,14dihydroxy compounds, it is the C14 hydroxyl group, not the C12 hydroxyl group, that is methylated.  The C14 hydroxyl group is much hindered than the C12 hydroxyl, thus  128  methylation at that site should not be favored. One possibility could be that the methylation of the hydroxyl group at C14 may act as a means of a metabolic regulation by the cells. A free C14 hydroxyl group could be part of the active binding site for the enzyme that is responsible for further transformations (for example, epoxidation); and such a process may be regulated by methylation of this group. The methylated compounds may stay "locked" in a pool, and when needed, cleavage of the methyl provides enzymes with the proper substrate for the next transformation.  products  Me  1,2 methyl shift H,0  Me  Me  3P-alcohol  [o] [H]  to] 3cc-alcohol Figure 2.19  CT Me  3-ketone  Proposed Mechanism for Formation of 3(3-alcohol and the Probable Relation with the a Isomer  Ring A and ring C functionalization seem to be a combination of convergent and ramified paths (metabolic grid 184 ), leading to the key intermediate triptophenolide (106) (Scheme 2.9). As mentioned earlier, epoxidation can only take place after the butenolide  129  moiety on ring A has been constructed. The functionalization on the C ring exclusively involves the introducing of hydroxyl group at C14, indicating that the starting point for epoxidation must be at CI4. Based on the above postulations, compound 106 may likely be the key intermediate in the biosynthesis of Tl (1) and Td (2) since it has the lactone on ring A and its C14 hydroxyl is free. It has been isolated from all four species of Tripterygium plants.2  The appropriate  precursors in Group 2 or Group 3 can only pass through this "gate" to reach their destiny, Tl (1) or Td (2). The first epoxidation probably occurs on ring B (Scheme 2.10). There are two possible mechanistic pathways to the formation of Tl (1) and Td (2). One is that a hydroxylation at C7 takes place first, followed by an Adler-type oxidation to form the epoxy dienone 139 (7-hydroxylation path way). The other possible route suggests that the substrate is first oxidized to an ortho quinone methide 219 and then epoxidation takes place between C7 and C8 to form 139 (quinone methide pathway). Quinone methides are believed to play a very important role as intermediates in the biosynthesis of secondary metabolites, for example, the lignans. 184,189 They are also transient intermediates in many synthetic reactions. 190,191 Compound 139 has not been isolated from the Tripterygium plants or from the cell cultures, but it is likely an intermediate involved in the biosynthesis of Tl (1) and Td (2). Alternatively, an intermediate such as 7-hydroperoxide of 106 may also be possible in the formation of 139. The sequence for the formation of the three epoxide groups in Tl (1) and Td (2) is still not clear, though they are likely formed in such a sequence as C7, C8 -> C9, CI 1 -> C12, C13. The isolation of 209 and 212 provides a supporting evidence for such a postulation. The ketone at C14 is finally reduced to a p hydroxyl group to form Tl (1). It is also possible that the C14 ketone is initially reduced to alcohol 220 and followed by epoxidation (Scheme 2.11), but there is insufficient evidence to support this mechanistic postulation at this stage.  ring A functionalization  Scheme 2.9  ring C functionalization  Proposed Biosynthetic Pathway from Dehydroabietane (196) to the Key Intermediate 106  131  95 R = H 212 R = OH  Tl(l) R = H Td(2) R = OH  Scheme 2.10 Proposed Biosynthetic Pathway from Triptophenolide (106) to Tl (1) and Td (2)  132  Scheme 2.11 An Alternative Pathway from 139 to Tl (1)  Td (2) was thought to be the C2 hydroxylation product of Tl (1). Isolation of 209 implies that C2 hydroxylation could occur prior to completion of the epoxidation, but since no other compounds with a C2 hydroxyl group were isolated without two or three epoxide groups already attached, hydroxylation at C2 appears to take place only when epoxide groups have been incorporated into the molecule. The completed biosynthetic pathway is proposed as in Scheme 2.9 and Scheme 2.10.  133  CHAPTER 3  SYNTHESES OF PRECURSORS FOR BIOTRANSFORMATION STUDIES  3.1  Introduction  As shown earlier, the formation of the epoxide system in Tl (1) and Td (2) likely takes place after the butenolide moiety on ring A, and a hydroxyl group at C14 have been incorporated by the appropriate enzymes. It was considered essential to evaluate appropriate synthetic precursors for the purpose of biotransformation to Tl (1) or Td (2) in the later stage of the biosynthetic process. It was hoped that such experiments would: (a) increase the yields and reduce the time required for metabolite production, (b) generate some new analogs of Tl (1) or Td (2) that might also have some biological activities, and (c) afford additional information about the biosynthesis. Initial biotransformation studies utilizing TRP4a cell cultures focused on the construction of the butenolide on ring A. 159 The various precursors employed in the early studies include dehydroabietic acid (157) and some synthetic compounds with abietane-type structures.159,192 In recent years, our interest has turned to building the epoxide system in the molecule. Investigation in our laboratories led to the syntheses of several precursors, of which isodehydroabietenolide (193) showed some results in biotransformation studies with TRP4a cell culture (e.g., hydroxylations at C2 and/or C7, etc.). 160 However, lack of functionalization on ring C during biotransformation of isodehydroabietenolide (193) suggested that ring C should be "activated" by introduction of hydroxyl groups.  A synthetic sequence to  isotriptophenolide (194), a C12 hydroxylated derivative of isodehydroabietenolide (193), was then developed.161 Application of 194 in the biotransformation studies with the TRP4a cell cultures resulted in a small yield of the methyl ether of 194 with much starting material recovered.161  134  193  194  106  The lack of extensive biotransformation of 194 with TRP4a cell cultures raised a challenge to the idea that ring C "activated" precursors would be transformed by the cells. But there were several ways we might be able to achieve our goals.  First, perform the  biotransformation of 194 under different experimental conditions. Secondly, synthesize other ring C "activated" precursors and carry out the biotransformation studies. Furthermore, it would also be interesting to synthesize some ring B functionalized, or both ring B and C functionalized precursors, to examine the possibility of biotransformations with the TRP4a cell cultures. As the synthetic sequences to isodehydroabietenolide (193) and isotriptophenolide (194) had been fairly well established, the focus of the present work was directed on the development of synthetic sequences for new ring C "activated" precursors, such as triptophenolide (106) and 12,14-dihydroxylated isodehydroabietenolide.  As a whole,  isotriptophenolide (194), triptophenolide (106), and the 12,14-diol would represent a series of compounds with different positions and degrees of hydroxylation on ring C. This situation would provide an interesting opportunity to develop a systematic approach to our biotransformation studies, and could afford synthetic strategies for chemical functionalization of aromatic rings within this diterpene family.  135  3.2  Results and Discussion  3.2.1  Syntheses of Isodehydroabietenolide and Related Compounds  Synthesis of Isodehydroabietenolide (193)  The synthetic sequence to isodehydroabietenolide (193) from dehydroabietic acid (157) was initially developed by M. Roberts.160 A slightly modified sequence, developed by F. KuriBrena,161 was applied in the synthesis of this compound. The starting material, dehydroabietic acid (157), was purchased as a technical grade, crude product (K & K Laboratories). Dehydroabietic acid incorporates most of the carbon skeleton, and most importantly, the chiral centers at C5 and CIO required for Tl (1) and Td (2) related molecules. However, the disadvantage to start with dehydroabietic acid is that "the range of synthetic approaches is limited by the chemistry of the starting material rather than by one's imagination". 145,161 This became more apparent when ring C functionalization was attempted (vide infra). The commercial sample of crude dehydroabietic acid was purified before further synthesis. The yield of partially purified dehydroabietic acid (89-92% by gas chromatography) from the commercial product was about 30%. The remaining impurities, which are isomeric resin acids and difficult to remove at this stage, were readily eliminated after the first step in the synthetic sequence. The partially purified dehydroabietic acid (157) was treated with excess thionyl chloride and a catalytic amount of dimethylformaniide (DMF) in benzene to give the acid chloride 222 (Scheme 3.1). The crude acid chloride was reacted with sodium azide in acetone and the resulting acyl azide 223 was heated in toluene to effect the Curtius rearrangement to afford the isocyanate 224. Reduction of the isocyanate with lithium aluminum hydride (LAH) in tetrahydrofuran (THF) under reflux for about 22 h yielded the secondary methyl amine 225,  136  which was then refluxed with aqueous formaldehyde and formic acid to give the tertiary dimethyl amine 226 (Eschweiler-Clarke methylation). The tertiary amine was converted to the corresponding //-oxide 227 by treatment with raefa-chloroperbenzoic acid (ra-CPB A) at -20°C. The reaction mixture was heated at reflux for 30 min to eliminate dimethyl hydroxyamine, giving exo-olefin 228 as the only isomer. At this stage, the crude exo-olefin was readily purified by column chromatography to afford pure 228 as a colorless oil with an overall yield of 71% from dehydroabietic acid (157). Ozonolysis of exo-olefin 228 carried out in a mixture of methanol-dichloromethane (5:1) at -78°C followed by treatment with dimethyl sulfide, generating the ketone 229 as a colorless oil (90% yield) which crystallized upon cooling. A minor by-product in this reaction was the 4,7-diketone 231. On attempting to perform the reaction on a larger scale (25 g), the reaction time became lengthy and the yield decreased due to an increasing amount of the diketone 231.  O  SMe O  231  230  The attachment of a one carbon unit to the C3 position was achieved in nearly quantitative yield by initial deprotonation at C3 with lithium 4-methyl-2,6-di-*-butyl phenolate, followed by alkylation with carbon disulfide, and treatment with methyl iodide to form the aoxoketene dithioacetal (a-dithiomethylene ketone) 230 as yellow crystals.  , \— 222 R = CI I— 223 R = N 3  232  Scheme 3.1  Synthesis of IsodehydroabietenoUde (193) from Dehydroabietic Acid (157)  a) SOC12, benzene, DMF; b) NaN3, acetone; c) A, toluene; d) IJAIH4, THF; e)HCHO, HCO2H; f) mCPBA; g) A; h) O3, -78°C; Me2S; i) lithium 4-methyl-2,6-di-f-butyl phenolate, THF, CS2; Mel; j) Me2S=CH2, THF, -20°C; k) HC1, MeOH-MeCN  138  Completion of the synthesis of isodehydroabietenolide (193) was accomplished in a two-step, one pot reaction. Trimethylsulfonium iodide was first treated with n-butyl lithium in THF at low temperature to generate the dimethylsulfonium methylid. Ketone 230 was then added to the ylid to give, presumably, an epoxide 193 232 (one isomer or a mixture of both isomers 232a and 232b) which was hydrolyzed without isolation to directly yield the butenolide 193. Recrystallization of the crude product with ethyl acetate and hexanes afforded pure isodehydroabietenolide 193 as colorless needles (mp 98-100°C, 58% yield from 230; 35% overall yield from dehydroabietic acid (157)). The *H NMR spectrum of 193 exhibited signals of an AMX system in the aromatic region due to the ring C protons. One of the characteristic features of this compound was the resonances due to protons HI9, which appeared at 8 4.76 as a broad AB quartet. In the course of this study, some *H NMR spectral characteristics of this series of compounds were noted. Changes in chemical shifts or appearance of these H19 signals (e.g., wider split between the AB quartet, etc.) usually reflected the change in stereochemistry or functionalization at C5. Furthermore, the C20 methyl group signal was located at 8 1.01, while the isopropyl methyl resonances were at 8 1.22 (6H, d, / = 6.9 Hz). It has been found that the C20 methyl signal usually appears at higher field than the isopropyl methyl signals when the A/B ring junction is trans. If the A/B ring is cis, the C20 methyl resonance would shift downfield by about 0.2 to 0.3 ppm, 145 thus often appearing at the low field side of the isopropyl signals. The change in the chemical shifts of these methyl signals usually implicated the change from A/B trans to A/B cis in the molecule. Variable yields encountered in the conversion of 230 to 193 161 provoked the interest to investigate the reaction in more detail. In the reaction of the sulfonium ylid with the dithioacetal 230 (Scheme 3.1, step j), the progress of the reaction was monitored by TLC (hexanes-ethyl acetate, 9:1). Two major intermediates (Rf = 0.58, 0.32, respectively) were observed, and they had been thought to be the two isomeric epoxides (232a, 232b) formed in the reaction.  139  MeS  MeS  SMe  SMe  232a  232b  The reaction intermediates were subjected to a rapid column chromatography using dichloromethane-hexanes (2:1) as eluent to remove unreacted starting material and other polar compounds. The intermediates thus obtained were chromatographed again using hexanes-ethyl acetate (9:1) as eluent to separate the two pure intermediates to which structures 233 (Rf = 0.58) and 234 (Rf = 0.32) were assigned. It was noticed that 234 partially converted to 233 during the column chromatography.  Both intermediates (233 and 234) were able to be  converted to 193 upon treatment with concentrated hydrochloric acid under the same conditions (Scheme 3.1, stepk).  MeS  233  High resolution mass spectrometry of 233 revealed its molecular formula as C21H26OS. The *H NMR spectrum showed some similarities to that of the dithioacetal 230. In the aromatic region, a new signal was found at 8 7.26 (IH, d, J = 1.8 Hz, H19). Only one methylthio group signal was found at 8 2.33. All the spectral analyses of this compound  140  (COSY, NOE,  13  C NMR etc.) confirmed its structure as the furan derivative 233. Similar  reactions had been reported in the synthesis of 3- and 3,4-substituted furans.194  MeS  234  The lH NMR spectrum of 234 was measured immediately after isolation (in acetones ' ) . The spectrum showed a multiplet at 8 4.75, but 1.5 h later the intensity of this signal decreased while a new signal at 8 7.47 appeared. Analysis of this sample one day later revealed that the signal at 8 4.75 had already disappeared and the spectrum was identical to that of 233 (Figure 3.1), indicating a facile transformation of 234 to 233. The chemical shift and the appearance of the signal at 8 4.75 were very similar to the H19 signals in isodehydroabietenolide (193), and the intermediate did show a m/z 374 ion peak in the mass spectrum. On this basis the structure of this intermediate was proposed as 234, rather than epoxides of 232a or 232b, since the chemical shift of protons of these epoxides cannot be at such a low field as 8 4.75. 186 Therefore, in the reaction of dithioacetal 230 with sulfonium ylid, regardless whether one epoxide isomer is formed predominantly over the other (probably this is the case), this epoxide or the mixture of both must have rearranged rapidly to the dihydrofuran derivative 234. Subsequent elimination of methanethiol from 234 leads to 233. Intermediates 233 and 234 are able to be converted to 193 upon hydrolysis. The overall mechanism of butenolide formation is proposed in Scheme 3.2.194-195  141  m«m mmmmf*mlmimmmm*m1*9***m amm***m*mmiMmm*mm+ i»win»wr  »«* wxwu^wi t********^*  x O i K«ihi ihin «»!>•>• inoli > » ' > < I * « « » < » « I I M » » * > * » » » « * I » < ; » * W W « « I I » « I m»i I«II •»<>• n^* VWntnx*n. «»iiri>i ii'i»miii»—MW»4'  c  JUL_*L  IL^A. I  1  1  7.5  1  1  1—1  1  7.0  1  1  1  1  6.5  !  1  1  '  1  6.0  1  '  '  '  1  '  5.5  5 0  '  '  I  '  '  '  4.5  1  ~~1  '  4. U PPM  a, *H NMR spectrum of 234 taken immediately after separation; b, 1.5 h later; c, 1 day later; d, spectrum of 233 (400 MHz, acetone-^)  Figure 3.1  *H NMR Spectra of Intermediates in the Reaction from 230 to 193  Major by-products isolated from the butenolide formation step were 237, 238 and 239. B-Keto ester 237 and B-keto thioester 238 (written in their enolic forms on the basis of ! H NMR analysis) were likely hydrolysis products of unreacted starting material 230.193 The A/B cis structure in both compounds was probably due to epimerization at C5 in the hydrolysis step since C5 readily epimerizes under basic or strong acidic conditions.145  142  .  Me2S=CH2  MeS  MeS  V -MeSH MeS MeS^y  MeS  H^O  s  -MeSH  °Vyi -  236  Scheme 3.2  193  Proposed Reaction Mechanism for the Formation of the Butenolide Moiety  MeO  MeS  237  238  Compound 239 was probably formed from 234 in the hydrolysis step (Scheme 3.3). The thiomethyl group, acting as a nucleophile, substituted the protonated hydroxyl group and finally generated the thiophene moiety in 239.  These results suggested that further  143  improvement of the yield of this reaction may be achieved by optimizing the reaction conditions to reduce or eliminate these byproducts {e.g., by completely converting 230 to 233 and 234; and converting 234 to 233 before hydrolysis, etc.).  H - N u * Me  234  234(i)  MeS  234(H)  H  J  H-Nu: H-SMe, etc.  239 Scheme 3.3  Proposed Mechanism for the Formation of 239  Syntheses of Isodehydroabietenolide Related Compounds  Previous results showed that biotransformation of isodehydroabietenolide (193) in TRP4a cell cultures produced some C7 and C2 hydroxylated compounds. 160 In order to examine whether some of these compounds can be further transformed by the cell culture, 7oxo-isodehydroabietenolide (206) and 7P-hydroxy-isodehydroabietenolide (240) were synthesized. The starting material 193 was treated with CrC>3 to give 206 in 49% yield (based on recovered starting material). Several different conditions and oxidizing agents196"200 (Jones;  144  Cr0 3 -Ac 2 0-HOAc-C6H 6 ; DMP-CrC>3; NBS-hv; PCC-CH2CI2; PCC-C6H6) were attempted with similar or poorer yields.  The ketone 206 was treated with sodium borohydride in ethanol at room temperature, stereoselectively affording 7P-hydroxy-isodehydroabietenolide (240) in 88% yield. Its IR spectrum showed a hydroxyl band at 3600 cm' 1 and the *H NMR spectrum exhibited the H7oc signal at 8 4.99 (1H, dd, J = 8.6, 8.6 Hz), which was enhanced when H5 was irradiated.  3.2.2 Syntheses of Isotriptophenolide and Related Compounds  Synthesis of Isotriptophenolide (194)  Isotriptophenolide (194) has not been isolated from Tripterygium plants or from the TRP4a cell cultures. It is an "artificial" precursor in terms of trying to incorporate it into the biosynthesis of Tl (1) and Td (2). However, it can be readily synthesized in good yield, and its close relationship with triptophenolide (106) may provide some interesting results in biotransformation studies with TRP4a cell cultures. The synthetic sequence was developed by F. Kuri-Brena161, and was adopted with a few changes. The synthesis of isotriptophenolide (194) involved Friedel-Crafts acylation of isodehydroabietenolide 193, followed by Baeyer-Villiger oxidation, and hydrolysis of the resulting ester to the desired phenol (Scheme 3.4). In the acylation reaction, 12-acetyl  145  isodehydroabietenolide (241) was obtained as a white solid. The yields of the crude product (reasonably pure and can be used directly for the next step) ranged from 95% to quantitative. Purification of the crude product with column chromatography afforded the pure product as a white crystalline solid in a yield of 92% from the starting material 193. The acylation occurred exclusively on C12 as shown by the *H NMR spectrum. The H l l signal (IH, s) was shifted downfield to 8 7.44, as compared with 8 7.25 in the starting material, as a result of the introduction of the acetyl group at CI2.  242 Scheme 3.4  Synthesis of Isotriptophenolide (194) from Isodehydroabietenolide (193)  a) acetyl chloride, anh. AICI3, CS2, reflux; b) m-CPBA, CH2CI2; c) HC1, MeOH  The Baeyer-Villiger oxidation of 241 was carried out by treatment of the starting material 241 in dichloromethane with m-CPBA. The crude 12-acetoxy-isodehydroabietenolide (242) was obtained in high yield as a yellowish powder. Purification of the product could be achieved by chromatography on silica gel with isopropyl ether as eluent. However, in the routine preparation of starting material for biotransformations or other purposes, this  146  purification was omitted since the unreacted starting material (in small amounts) was difficult to separate from the product by column chromatography. Also partial hydrolysis of the 12acetoxy compound 242 to the phenol 194 on the column complicated the purification. It was found that as long as these intermediates were sufficiently pure by TLC, it was preferable to purify the final product 194 since different retention times between 194 and the impurities allowed an efficient column separation. The *H NMR spectrum exhibited the HI 1 signal at 8 6.92 as a singlet. Hydrolysis of the acetate 242 was performed in methanol with concentrated hydrochloric acid. The crude phenol was purified by column chromatography with isopropyl ether, and recrystallized in ethyl acetate and isopropyl ether. The pure isotriptophenolide (194) was obtained as a white crystalline solid and in an overall yield of 84% from isodehydroabietenolide (193). The IR spectrum of the phenol showed a hydroxyl absorption at 3400 cm -1 , and its *H NMR spectrum exhibited the hydroxyl proton signal at 8 4.84 (exchangeable with D2O). The HI 1 resonance was found at 8 6.72 (s) and the H14 signal at 8 6.90 (s), on the basis of NOE results. The omission of intermediate purification facilitated the routine preparation of compound 194. However, care should be taken in the work-up stages to make sure that the crude intermediates are free of any inorganic salts or remaining excess reagents. In one case, a by-product was isolated at the end of the synthesis, which was found to be 11-chloroisotriptophenolide (243). The formation of this by-product could be caused by chloration in the last one or two steps due to small amounts of impurities. The mass spectrum of 243 exhibited two molecular ion peaks at m/z 348 (with 37C1) and 346 (with 35C1) in a ratio of about 1 to 3, indicating a chlorine atom in the molecule. Its *H NMR spectrum showed a very similar spectrum to that of 194, but only one aromatic proton signal was found at 8 6.85 (s), which was assigned to H14 due to its NOE effect with protons at C7. Thus the chloro group must be at CI 1, and this was confirmed by the dramatic downfield shift of the Hip signal (from 8 2.282.54 in 194 to 8 3.64), apparently due to the crowding caused by the chloro group at CI 1.  147  243  Syntheses of 7-Oxo- and 7p-Hydroxy-IsotriptophenoIide  The synthesis of 7p-hydroxy-isotriptophenolide was carried out with a similar procedure as in the synthesis of 7P-hydroxy-isodehydroabietenolide (240) (section 3.2.1) (Scheme 3.5). Because direct oxidation of isotriptophenolide (194) with Cr03 could result in undesired oxidation of ring C, 7-oxo-isotriptophenolide (244) was synthesized by oxidation of 12-acetoxy-isotriptophenolide (242) with Cr03 solution in aqueous acetic acid. The oxidation afforded 12-acetoxy-7-oxo-isotriptophenolide (244) as a yellowish powder in a yield of 63% based on recovered starting material. Compound 244 was then hydrolyzed to 7-oxo-isotriptophenolide (245) by treatment with hydrochloric acid in methanol. The resulting product, after recrystallization in methanol and water, was obtained as colorless needles (mp 287-289°C, dec) in a yield of 84%. Reduction of ketone 245 with sodium borohydride in ethanol (Method A) gave the required 7(3hydroxy-isotriptophenolide 246b in 26% yield. The diol 246b can also be synthesized from 244 by direct reduction with sodium borohydride in ethanol (Method B). The acetate group was also cleaved during the reduction. Purification of the crude product gave the diol 246b in 27% yield from 244. This compound was not very stable, and decomposition was observed during workup and column chromatography. The diol 246b decomposed to the C6, C7 olefin 247 in the NMR tube when deuterated chloroform was used as the solvent prior to passage through a short column of alumina. The IR spectrum of 246b showed a band for hydroxyl  148  groups at 3350 cm' 1 and its *H NMR spectrum displayed the H7a signal at 8 4.95 (IH, br dd, J = 8.6, 8.6 Hz). The stereochemistry at C7 was confirmed by observation of NOEs between H7a and H5. OAc  OAc  245 Scheme 3.5  246b  Synthesis of 7P-Hydroxyl-isotriptophenolide (246b)  a) Cr03, HOAc; b) HC1, MeOH; c), d) NaBH4, MeOH  247  246a  149  A very small amount of the other isomer, 7a-hydroxy-isotriptophenolide (246a), could also be seen from the spectrum.* The H14 and HI 1 signals of the a isomer were located at 8 7.17 (IH, s) and 6.73 (IH, s), respectively. Compared with 8 7.35 (IH, s) and 6.69 (IH, s) for the p isomer, this rather large difference in chemical shift of the H14 resonance (0.18 ppm) between these two isomers reflected the fact that the P-hydroxyl group in 246b is in a quasi equatorial orientation and in close proximity to proton H14 in space. On the other hand, the 7a-hydroxyl group in 248 is close to an axial orientation, thus removed from proton H14, so the chemical shift of H14 is less affected. Another difference was the chemical shift of H5. The spectrum of the a isomer exhibited a downfield shifted H5 signal at 8 3.15 compared with 8 2.75 in the p isomer, due to its 1,3-diaxial interaction with the oc-hydroxyl group at C7. These lH NMR spectral features proved very helpful in structure elucidation in the later biotransformation studies.  Attempted Synthesis of 8,12-Quinone Methide from Isotriptophenolide (194)  One of the proposed mechanisms to form the epoxide group between C7 and C8 is via an ortho-qainone methide intermediate 219 (section 2.2). If a synthesis of a para-qainone methide from the available isotriptophenolide (194) could be achieved and then this precursor was incubated with TRP4a cells, the epoxidation between C7 and C8 might be possible (Scheme 3.6). Essentially pure 246a was obtained later in another chemical reaction and also from a biotransformation experiment.  150  OH  TRP4a cell cultures O;  H  o-  Scheme 3.6  194  249  Quinone Methide 249 as a Possible Biotransformation Precursor  Extended quinone methide structures have been found in some natural products and they are believed to be the key function to those compounds' biological activities.189 Despite the long-lasting interests in synthesis of simple para- or o/t&o-quinone methide structures, only a few attempts have been successful and with some limitations such as the requirement of certain groups attached to the ring or the methide carbon.201"205  para-quinone methide  ortho-qainont methide  However, the simple quinone methides are believed to be the intermediates involved in some biological processes and also in many synthetic reactions. 189,206 ' 207 Several methods  151  have been developed in making such structures as reaction intermediates, such as oxidation, elimination and so on. 208,209 As to our particular molecule, oxidation seemed to be the best route to success. Oxidation of isotriptophenolide (194) with oxygen and cuprous chloride (CuCl) 210 led to a complicated mixture. Treatment of 194 with potassium ferricyanide (K3Fe(CN)6)211 also gave a complicated mixture with some unreacted starting material remaining. Dichlorodicyanoquinone (DDQ) was reported capable of effecting such an oxidation from a phenol to a quinone methide.212 Preliminary results showed that the starting material disappeared in about 2-3 h and TLC showed a yellow spot (before spraying) among a moderately complicated mixture. As DDQ seemed more promising than the other reagents, it was subsequently more extensively examined. Among the solvents tested, THF was better than benzene, dichloromethane and dioxane as based on TLC observations. Different conditions for oxidation of isotriptophenolide (194) by DDQ in THF, such as the equivalents of DDQ, temperature and reaction time, showed that 2 equivalents of DDQ in about 2-3 h at room temperature gave better results for the product which showed the yellow spot on TLC.  Scheme 3.7  Reaction of Isotriptophenolide (194) with DDQ in THF  A typical reaction, for example, was carried out by adding DDQ (291 mg, 2 equiv.) to a stirred solution of isotriptophenolide (194, 200 mg, 0.641 mmol) in THF (30 mL), and the mixture was stirred at room temperature for 3 h (Scheme 3.7). The solvent was removed and  152  the crude product was purified by repeated column chromatography. The compound isolated was a dark orange solid (250, 35 mg) in a yield of 18%.  250  The IR spectrum of the product 250 showed the butenolide carbonyl at 1750 cm -1 and typical quinone bands at 1640 and 1600 cm -1 . The molecular formula was determined by high resolution mass spectrometry as C20H20O3, two hydrogen atoms less than the anticipated quinone methide 249. Its *H NMR spectrum displayed four olefinic proton signals at 8 6.36 (IH, d, J = 7.0 Hz), 6.56 (IH, d, / = 1.6 Hz), 6.82 (IH, dd, / = 7.0, 1.6 Hz), and 7.00 (IH, s). Protons at 8 6.36 and 6.82 were coupled with each other with a coupling constant of 7.0 Hz. In addition, the proton at 8 6.82 was also coupled with the proton at 8 6.56 with a small coupling constant of 1.6 Hz. The H19 signals were located at 8 4.98, at a much lower field than that of the corresponding signals in isotriptophenolide 194 (H19, 8 4.76), and with a wider splitting (Av = 0.16 ppm) between the AB quartet, suggesting that there must be some reaction at C5. The C20 methyl signal was shifted downfield to 8 1.29, also suggesting some reaction at C5 (for example, double bond formation). Irradiation of HI 1 at 8 6.56 resulted in an enhancement of the Hip signal at 8 2.33 (IH, br d, / = 13.3 Hz), which was strongly coupled with Hloc at 8 1.88 (IH, m). Both protons were also correlated with a multiplet at 8 2.62 (2H), which were then assigned to the C2 protons. The H15 signal was found at 8 3.14 (IH, septet, / = 6.4 Hz), which showed cross peaks to the two isopropyl methyl signals at 8 1.15. These signals accounted for all the aliphatic proton signals in the spectrum; therefore, the four olefinic proton signals must be due to protons at C6, C7, C l l (assigned) and C14. Irradiation of the C19  153  protons increased the intensity of the signal at 8 6.36, and vice versa, indicating the enhanced signal was due to H6. The proton (6 6.82) which was strongly coupled with H6, therefore, must be H7. Finally, irradiation of the signal at 8 7.00 caused NOE enhancements to H15, two isopropyl methyl groups as well as H7; thus, this proton was assigned to H14. The small coupling between HI 1 and H7 was a long range coupling ( 5 J), which has been found in other similar structures.213 On this basis, the structure of the oxidation product was assigned as 250, an extended quinone methide. Compounds 247, 251, 252 and 245 were isolated as byproducts. OH  247  251  252  245  Unfortunately oxidation of isotriptophenolide (194) with DDQ yielded the extended quinone methide 250 rather than the desired 249, regardless whether the reaction was performed at low temperatures or in short time periods. The structure of 250 suggested that the reaction likely went through the intermediate stage of 249, but this compound may be labile and thus undergo further reactions very rapidly, so it was simply impossible to isolate the  154  intermediate 249. This postulation could be further examined if this intermediate could be trapped. Therefore, an oxidation of isotriptophenolide (194, 100 mg) with DDQ was carried out in methanol rather than in THF. 214 After a short reaction time, the crude product was subjected to a rapid column chromatography, yielding four fractions (Frs. 1-4). Fr. 1 (11 mg) was a mixture of unreacted starting material and compound 247. Fr. 2 (24 mg, 22%) was an 1:1 mixture of 7 a and 7p-methoxy-isotriptophenolide (253), and Fr. 3 (59 mg) was a complicated mixture containing some 250. The final fraction (8 mg) was essentially pure 246a. The lH NMR spectrum of Fr. 2 exhibited great similarity to those of 7a and 7[3-hydroxyisotriptophenolide (246a, 246b). The H14 signal of the 7f3 and a-methoxy compounds was located at 8 7.28 and 7.13, respectively, and the corresponding H7 was found at 8 4.64 (dd, / = 8, 8 Hz) and 4.30 (br s), respectively. The methoxy group for both isomers was located about 8 3.47.  These methoxylated compounds suggested that the reaction proceeded via the  intermediate 249 (Scheme 3.8), and the latter underwent a facile Michael addition type reaction 215 with methanol and water acting as nucleophiles. In the reaction in THF, the intermediate 249 was further oxidized to extended quinone methide 250. OH  -^E^^^OMe  Scheme 3.8  Proposed Mechanism for the Formation of Compounds 253, 246 and 250  155  In conclusion, it appeared difficult to obtain the quinone methide 249 and further investigations were discontinued. However, the conjugated quinone 250 may be of further interest since some natural products with a similar quinone system have shown interesting biological activities.216  3.2.3  Synthesis of Triptophenolide (106)  The synthesis of 106 (in most cases as a racemic mixture), as an intermediate towards the synthesis of Tl (1), has already been reported. 107,142 The synthetic route of more direct relevance to the present study was developed by E. E. van Tamelen et al., in which they employed dehydroabietic acid (157) as a starting material.144 The strategy of the published sequence was to first functionalize ring C and then elaborate the butenolide ring system in the later stages (Scheme 3.9).144  OCOCF3  „  ..  165  30%  166 Scheme 3.9  171  Summary of the Synthesis of Isotriptophenolide Acetate (171) via C—>A Functionalization  156  The problem encountered in this sequence was that ring C was extensively degraded during the oxidative conversion of exo-olefin 165 to ketone 174, thus lowering the overall yield of the sequence. In order to avoid such losses, our preliminary experiments were designed to initially synthesize the ketone 229 (or its protected form), and then attempt the introduction of the C14 hydroxyl group by the established procedure. 145 Ketone 229 was readily available as an intermediate in the synthesis of 193, but it was not a good starting material for the intended sequence because epimerization at C5 occurred in the reactions (probably in the zinc-acetic acid reduction step, according to ^B. NMR results) (Scheme 3.10).  O  229  1— 254 R1 = R2 = N0 2 p 255 R ! = NH 2 ,R 2 = N 0 2 ° ^ 2 5 6 R1 = I,R 2 =N( i2  b  e  .— 257 R = NH 2 I— 258 R = O H and 259 R = OCOCF 3  Scheme 3.10 Preliminary Attempt to Synthesize 106 from 229 a) HNO3, H2SO4, HOAc; b) H2, Pt/C; c) NaN02, KI; d) Zn, HOAc; e) NaNC^, CF3C02H, HOAc  In order to avoid epimerization at C5, the ketone 229 was first reduced to the alcohol and then converted to the corresponding acetate (Scheme 3.11). Thus, the ketone 229 was treated with sodium borohydride in ethanol, and the resulting alcohol 260 (P alcohol according to its *H NMR) was treated with acetic anhydride in pyridine. The acetate 261 was then nitrated, reduced, diazotized, treated with iodide and finally reduced again to afford the phenol trifluoroacetate 266. The latter was then hydrolyzed to give the phenol 267. The phenol was protected as its 14-methyl ether 268 and then the acetate group at C4 was reduced back to  157 alcohol 269. The alcohol was oxidized to ketone 270 by titration with Jones reagent and the ketone was then converted to the dithioacetal 271. Treatment of the dithioacetal 271 with dimethylsulphonium methylid, followed by acid hydrolysis afforded the methyl triptophenolide (107) as colorless prisms in an overall yield of 5.7 % from 229.  229  bC  260 R = OH 261 R = OAc  IT ZT p g p: h p 1 •-*• G  .OH  -269 R= < H ^270 R= 0 H  kf  271  262 263 264 265 266 267 268  n  R1 = R 2 = N 0 2 R1 = NH 2 ,R 2 =N0 2 R] = I,R 2 = N0 2 R1 = H,R 2 = NH2 R1 =H,R2=OCOCF3 R1 = H,R 2 =OH R1 = H,R 2 =OMe  r— 107 R = Me I— 106 R = H  Scheme 3.11 Preliminary Attempt to Synthesize 106 from a Derivative of 229 a) NaBH4, EtOH; b) Ac20, pyridine; c)HN03, H2SO4, HOAc; d)H2, Pt/C; e)NaN02, KI; 0 Zn, HOAc; g) NaN02, CF3CO2H, HOAc; h) HC1, MeOH; i) n-Bu Li, Mel; j) LAH, THF; k) Jone's reagent; 1) potassium 4-methyl-2,6-di-t-butyl phenolate, CS2, Mel; m) Me2S=CH2; HC1; n) HBr, HOAc  Hydrolysis of the methyl ether was achieved by refluxing with hydrobromic acid in acetic acid,217 giving triptophenolide (106) in excellent yield. Ether cleavage by reaction with  158 boron tribromide was not successful,107>218>219 probably due to lack of a neighboring carbonyl group. 220 The synthesized triptophenolide (106) was spectrally identical with authentic and isolated samples. The second sequence was successful in providing triptophenolide (106), but there were still some shortcomings. First, the protection and deprotection of the C4 hydroxyl group and the C14 phenolic hydroxyl group prolonged the already long sequence. Secondly, the overall yield (not optimized), although similar to the reported sequence, was too low for our requirements in terms of preparing larger quantities of 106 for future studies. A possible solution to this problem, and with the syntheses of other ring C activated precursors in mind, was to use a common, "advanced" intermediate, such as the butenolide 193, in which the butenolide ring system has been already elaborated (Scheme 3.12). OH  DHA  I  H  Scheme 3.12 Synthesis of Ring C "Activated" Compounds via the Common Intermediate 193  159  With the butenolide 193 as the common starting material, one could take advantage of the well established sequence to synthesize 193, and then use it to further synthesize ring C hydroxylated compounds, such as isotriptophenolide (194), triptophenolide (106) and others. There had been some concern about the liability of the butenolide moiety in the subsequent reactions to ring C hydroxylated compounds.161 However, high yields of acetylation of 193 in the synthesis of isotriptophenolide (194), and of the hydrolysis of the methyl ether 107 in the preliminary synthesis of triptophenolide (106) were obtained under harsh reaction conditions. These results reflected the fact that the butenolide moiety was much more stable than we had predicted. A series of small scale experiments under some relevant reaction conditions, such as nitration, catalytic hydrogenation, and zinc-acetic acid reduction, confirmed that compound 193 was stable enough under those conditions. Therefore, a sequence using isodehydroabietenolide (193) as starting material to synthesize triptophenolide (106) was investigated. Initially, similar conditions as those in the literature were applied.145 The butenolide 193 was nitrated with a mixture of concentrated nitric acid and sulfuric acid at 0-5°C. The product precipitated as a pale yellow solid. Column chromatography of the crude product gave the 12,14-dinitro compound 272 as a pale yellow solid (mp 112-115°C) in 55% yield and 12-nitro compound 273 in 15% yield.  The IR spectrum of 272 showed the butenolide absorptions at 1740 and 1680 cm -1 , and strong bands for nitro groups at 1530 and 1370 cm-1. The *H NMR spectrum exhibited only one  160  aromatic proton signal at 8 7.62 (1H, s, HI 1), which showed an NOE to the H i p at 5 2.46 (1H, dd, / = 12.3, 6.6 Hz). The remainder of the spectrum was similar to that of 193. The 12-nitro compound 273 was obtained as a pale yellow crystalline solid (mp 176-178°C, dec). Its IR spectrum showed the absorption bands for the nitro group at 1520 and 1370 cm"1. Two aromatic signals were observed as singlets in the *H NMR spectrum. Since the desired 12,14-dinitro compound 272 was only obtained in a moderate yield under this condition, further studies to improve the yield were essential. The nitration was repeated using nitromethane 221 as the solvent (the starting material was more soluble in nitromethane). The reaction went well and afforded the dinitro compound 272 in 86% yield. The reaction was repeated a number of times and the average yield was consistently greater than 86%. The next step was catalytic reduction of the 12-nitro group in 272. The reaction was carried out under hydrogen (about atmospheric pressure) at room temperature with Pt (10% on carbon) as catalyst. The crude product was recrystallized in ethyl acetate to give 14-nitro-12amine 274 as yellow needles (mp 227-229°C, dec). The mother liquor was chromatographed to afford more product, bringing the total yield to 83%.  274  The IR spectrum of 274 showed absorption bands for the amino group at 3500 and 3400 cm-1, and the nitro group at 1520, 1350 cm"1, respectively. The HI 1 signal was found at 8 6.71 in its *H NMR spectrum. The reduction occurred on the C12 nitro group regioselectively, because the C14 nitro group is much more hindered than the C12 nitro group.  161  Compound 274 was then diazotized with sodium nitrite (Scheme 3.13), followed by addition of potassium iodide. The crude product 275 was isolated by filtration and used directly for the next reaction. The crude 275 was treated with zinc powder in warm acetic acid (6570°C) overnight and the resulting crude amine 276 was diazotized again with sodium nitrite in trifluoroacetic acid to form the trifluoroacetate 277. The latter was then subsequently hydrolyzed to the phenol 106 in 33% overall yield from 274, and with a yellow by-product in 5% yield. The spectra of the synthetic triptophenolide (106) were identical with those from the authentic sample obtained in the earlier studies. The yellow by-product 278 showed quinone absorption bands at 1680 and 1640 cm"1. Its *H NMR spectrum exhibited the proton at C12 as a doublet at 8 6.39 (1H, d, / = 1.1 Hz, allylic coupling with H15). All the spectral data indicated that the by-product was the 11,14-quinone 278.  I  H 274  275  OCOCF3  277  276  XT H 106  Scheme 3.13 Reactions from 274 to 106 a) NaN02, TFA, HOAc, H2O; KI; b) Zn, HOAc, A; c) NaN02, TFA; d) HC1, MeOH  162  H0 2 C  278  114  Quinone 278 is structurally related to compound 114 isolated from T. regelii89 and may exhibit similar pharmacological activities (inhibition of interleukin-1),133 but screening data are not yet available. However, in the synthesis of triptophenolide (106), formation of this byproduct would decrease the yield of 106, and thus its formation should be decreased. An attempt was made to examine whether lowering the temperature of the reduction with zinc in acetic acid could decrease the formation of the quinone, because once the C14-amine 276 is formed, some of it may be oxidized to the quinone. Therefore, a reaction was carried out in the similar manner but the reduction was conducted at 30-40°C for 14 h and then 65-75°C for only 2 h. Column chromatography of the crude product from the final step gave triptophenolide (106) in 23% yield, quinone 278 (10% yield), and another compound 279 (9%).  279  The new compound had a molecular formula of C20H23IO3. Its IR spectrum suggested the presence of a hydroxyl group (3625, 1240 cm -1 ). The ! H NMR spectrum showed a similar pattern to that of triptophenolide (106) except that only one aromatic proton signal was found at  163  8 7.42 (1H, s), and the H15 resonance was shifted downfield to 8 3.42. Irradiation of the aromatic proton enhanced HIP at 8 2.43, indicating it to be due to HI 1. Therefore, the new compound was assigned as 12-iodo-triptophenolide (279). It was obvious that lowering the temperature of the reduction step did not decrease the formation of the quinone, instead, incomplete reduction produced another by-product 279. Since the overall yield from 274 to 106 was not very satisfactory, each step of the reaction sequence was then examined individually. Therefore, the 14-nitro-12-amino analog 274 was treated with zinc powder in acetic acid at 65-70°C overnight. The resulting crude product was chromatographed to give a mixture of the C14 amine 276 and the 12-iodo-14-amine 280 in a ratio of about 3.5 to 1 by comparing the integrations of their C20 methyl signals in the *H NMR spectra (8 1.01 and 0.99, respectively). Compound 281 was also isolated.  280  281  Compound 281 had a molecular formula of C20H23NO3 with the IR spectrum revealing an amino group at 3510 and 3350 cm-1. The ! H NMR spectrum was similar to that of 108. Further separation of 280 and 276 was not successful, and thus the mixture (208 mg) was used in the next step to synthesize the phenol. The reactions (diazotization and hydrolysis) were carried out under the same conditions as before, and the final product was chromatographed to give the quinone 278 (18 mg), triptophenolide (106, 38 mg) and 12-iodo-triptophenolide (279, 49 mg). The ratio of 279 to 106 was about 1 to 1 in the products, although the ratio of their corresponding starting materials was about 1 to 3.5 based on lH NMR data. In other  164  words, the 12-iodo- 14-amino analog 280 afforded the corresponding phenol 267 in more than 3 times higher yield than the conversion of the C14 amine 276 to 106 (Scheme 3.14).  r  <  K. 268 (264:268-3.5:1)  267 (106:267-1:1)  Scheme 3.14 Comparison of Yields of 106 and 267 with respect to 264 and 268  The iodo group was originally introduced to improve the yield of indirect deamination at C12, because direct deamination with sodium nitrite and hypophosphorous acid was reported to afford poor yields. 145 In the present investigation, the presence of this iodo group at C12 increased the yield of the subsequent conversion of the C14 amino group to a hydroxyl group. This is a new finding. The corresponding quinone (see 278) but bearing the 12-iodo function was not found in this particular experiment. In order to use the new finding in the effort to improve the yield of 106, and also because of the overall unsatisfactory results obtained with zinc-acid reduction, the synthetic strategy should be re-organized. The possible alternatives in obtaining 106 from the various available synthetic intermediates are summarized in Scheme 3.15.  165 NH2  H  275  H  282  276  OH I 7  I  H  ,*'  H  106  279  Scheme 3.15 Proposed Sequences for Synthesis of 106  Common procedures for the replacement of an aromatic primary amino group by hydrogen involve preliminary diazotization of the aromatic amine followed by reductive substitution by a hydrogen donor.222 Hypophosphorous acid has proven to be a generally effective reducing agent for diazonium salts, and remains the standard reagent for such reactions.  166  However, as hypophosphorous acid was reported not so effective in a very similar situation,145 another alternative was sought. A method of reductive deamination of arylamines by alkyl nitrites (pentyl nitrite, isopentyl nitrite, r-butyl nitrite etc.) came to our attention. Moderate and high yields were obtained for such in situ one-pot conversions of aromatic amines to the corresponding hydrocarbons.223,224 Thus, a small scale experiment was attempted. A solution of the 14-nitro-12-amino compound 274 in DMF was added dropwise to a stirred solution of isoamyl nitrite in DMF at 65°C (Scheme 3.17). The reaction was continued for a short time and then worked up. Column chromatography of the crude product yielded the 14-nitro-compound 282 in 33% yield. A by-product was also isolated.  Scheme 3.16 Deamination of 274 by Isoamyl Nitrite a) isoamyl nitrite, DMF, A  Compound 282 was isolated as pale yellow prisms (mp 231-233°C,dec). The *H NMR spectrum was similar to that of the C14 amine 276, but the HI 1 and H12 signals were slightly shifted downfield to S 7.43 (IH, d, J = 8.3 Hz) and 7.24 (IH, d,J= 8.3 Hz). The by-product had a molecular formula of C20H23NO5. Its IR spectrum suggested the presence of hydroxyl (3700-3600 cm -1 ) and nitro groups (1560, 1370 cm -1 ). The *H NMR spectrum was similar to that of 282, except that only one H7 signal was found at 5 5.73 (IH, dd, J = 5.5, 1.8 Hz) and that the H5 signal was significantly shifted downfield to 8 3.15 (lH,br d, J = 12.2 Hz). This was an indication of the presence of an a hydroxyl group at C7. On the basis of NMR, IR and MS spectra, this by-product was proposed as 283.  167  283  The low yield of this direct deamination turned our attention to reactions that could simultaneously reduce the nitro and iodo groups. Most metal borohydrides are generally unable to reduce aromatic nitro compounds under ordinary conditions. 188 However, it has been reported that aromatic nitro compounds can be reduced with sodium borohydride-transition metal salts systems in good yields.225"230 Interestingly, it was reported that iodo substituents on the aromatic ring were also reduced together with the nitro groups by a potassium borohydridecopper (I) chloride system.230 Therefore, the 12-iodo-14-nitro compound 275 was treated with cuprous chloride and sodium borohydride in methanol (Scheme 3.17). The crude product was chromatographed to afford the C14 amine 276 in about 60 % yield (not optimized). No 12-iodo14-amino compound 280 could be detected from the *H NMR spectrum.  275  276  Scheme 3.17 Reduction of Iodo and Nitro Groups with NaBH4-CuCl System  168  The amine 276 was isolated as a pale pink crystalline solid (mp 186-188°C). The IR spectrum showed the amine bands at 3550 and 3350 cm -1 . The *H NMR spectrum exhibited the H l l and H12 resonances at 5 6.84 (1H, d, / = 8.2 Hz) and 7.04 (1H, d, J = 8.2 Hz), respectively. The remainder of the spectrum was similar to that of triptophenolide (106). Catalytic hydrogenation has also been used to cleave aromatic halogen groups in synthesis. 47,231 In our case, palladium (10% on carbon) was used in an attempt to cleave the iodo group. 232 The 12-Iodo-14-nitro compound 275 in methanol solution was stirred under hydrogen atmosphere with the palladium catalyst for several hours, giving the 14-nitro compound 282 in 48% yield. In order to determine whether the reaction can further reduce the 14-nitro group, a second reaction was run for a prolonged time. The products, after purification, were 282 (23% yield) and the amine 276 (44% yield). The catalytic hydrogenation for the above reaction was slow and the yield was unsatisfactory. Among the methods tried, sodium borohydride-copper (I) chloride seemed promising. While a consideration to pursue a more detailed study with the sodium borohydridecuprous chloride reduction was under way, another experiment with sodium dithionite (Na2S204) 2 3 3 , 2 3 4 showed very encouraging results. Reduction of the 12-iodo-14-nitro compound 275 with sodium dithionite selectively reduced the nitro group without affecting the iodo group, giving a very high yield of the 12-iodo-14-nitro compound 280 (87%, Scheme 3.18).  Scheme 3.18 Selective Reduction of Nitro Group with Na2S204  169  In consideration of the potential high yield from 280 to the corresponding phenol 279, this sequence would be very promising if an appropriate condition for the last step, removal of the iodo group from the phenol 279, could be found. Catalytic dehalogenation was attempted again to remove the iodo group from 279. The reaction was carried out in methanol with palladium on carbon as catalyst.232 An equivalent amount of potassium carbonate was added as the acceptor of hydrogen iodide. The reaction proceeded very well and afforded the desired product in 97% yield. With the excellent yield obtained for the last step, the overall sequence was now efficient. The new sequence is summarized in Scheme 3.19. With the individual sequences from the 14-nitro-12-amine 274 to triptophenolide (106) being established, the formal synthesis of 106 from isodehydroabietenolide (193) was carried out. The 12,14-dinitro compound 272 and 14-nitro-12-amine 274 were synthesized following the procedures we have discussed earlier (vide supra). Then 274 was treated with sodium nitrite followed by potassium iodide to afford the 12-iodo-14-amine 275 in 65% yield. Although a few attempts were made to improve the yield, no significant improvement was achieved. Compound 275 was then treated with sodium dithionite in refluxing ethanol to give the 12-iodo-14-amine 280. The average yield of this reaction was 86-88%. Compound 280 was diazotized in trifluoroacetic acid at -12 to -10°C. The crude product, triptophenolide trifluoroacetate (282), was directly hydrolyzed with concentrated hydrochloric acid in methanol. In a preliminary experiment, an attempt was made to purify the trifluoroacetate by column chromatography. However, partial hydrolysis of the ester was observed during the purification, and since we were only interested in the final product 106, the purification was omitted. The hydrolysis was carried out overnight, and the solid product precipitated out when the reaction was almost completed. Workup of the reaction gave a crude product which, after column chromatography, afforded 12-iodo-triptophenolide (279) in 77% yield. The average yield of the reaction from 280 to 279 was between 77 to 79%.  170  N0 2  NH2  'N0 2  o*  O;  H  o-  193  272  H  o-  274  (—284 R = COCF3 * L— 279 R = H Scheme 3.19 The New Sequence for the Synthesis of Triptophenolide (106) a) HNO3, H2S04, MeN02; b) H2, Pt/C, HOAc; c) NaN02, TFA, HOAc, KI; d) Na2S204, EtOH, A; e) NaN02, TFA; f) HC1, MeOH; g) H2, Pd/C, Na2C03  As a comparison, two sets of the reaction, with the C14 amines 276 and 280 as the starting materials, respectively, were carried out at the same time and under identical conditions. The results showed that the reaction with 280 constantly afforded 12-iodo-triptophenolide (279) in 77-78% isolated yields, while with 276, the reaction only gave 45-46% of 106 after purification. This clearly demonstrated that the iodo group at C12 greatly increased the yield of converting the C14 amino group to a hydroxyl group via diazotization (Scheme 3.20).  171  Scheme 3.20 Comparison of Reaction Yields between 268 and 264 a) NaN02, TFA; b) HC1, MeOH  In order to examine whether zinc or iodide salts were catalysts for the last diazotization step, 145 small amounts of both salts were intentionally added to the reaction mixture. No significant differences were observed as compared with those without the salts. Therefore, the reaction is not catalyzed by these inorganic salts. Another interesting point was that, in contrast to reactions with the amine 276 as starting material, diazotizations with the 12-iodo-14-amine 280 only produced about 1.5% yield of the 12-iodo-quinone 285 as a by-product while the former usually gave about 10% yield of quinone 278. The other by-products isolated in small quantities from diazotization of 280 were 286 and 287. Compound 285 was isolated as orange prisms (mp 186.5-187.5°C, dec). Its IR spectrum showed the quinone absorptions at 1680 and 1650 cm -1 . The *H NMR spectrum was similar to that of 278, except for the absence of the H12 signal.  172  Compound 286 was obtained as pale yellow prisms with a molecular formula of C20H22INO5. The IR spectrum indicated the presence of hydroxyl (3400, 1180 cm -1 ) and nitro (1520, 1360 cm -1 ) groups. Its lH NMR spectrum showed no aromatic signals. All the spectral information suggested the structure as assigned to 286. This compound was probably produced via nitration by small amounts of nitric acid in the reaction mixture.  Compound 287 was isolated as a white powder with a molecular formula of C20H23IO3. The IR spectrum showed hydroxyl bands at 3400 and 1010 cm -1 . Its *H NMR spectrum displayed two aromatic proton signals at 8 7.75 (IH, s) and 6.99 (IH, s). The pattern of H19, H7, H5, H6, H2, HI and the C20 methyl signals was very similar to that of isotriptophenolide (194). However, one of the two methyl signals was found at 5 1.28 (3H, d, J = 6.2 Hz) where usually the two isopropyl methyl signals were located. The characteristic septet signal of the isopropyl methine proton was also not present. Instead, one multiplet appeared at 8 4.09 (IH) and other two at 8 2.73 (IH) and 2.85 (IH), respectively. The COSY spectrum exhibited cross  173  peaks between the proton at 5 4.09 and the other two multiplets, and the methyl signal at 8 1.28. However, no cross peaks were found between the multiplets (8 2.73, 2.85) and the methyl signal. These data showed that the isopropyl group was no longer present. HO  _3' \  COSY  287  The iodo group was probably still at C12 and the proton at 8 4.09 (not in a benzylic position) was likely geminal to a hydroxyl group. Irradiation of the proton at 8 7.75 resulted a signal enhancement of Hip centered at 8 2.45 (doublet of doublets), indicating that the irradiated proton was H l l . Irradiation of the proton at 8 6.99 showed NOE enhancements to the C7 protons at 8 2.94 (2H, m) and the two multiplets at 8 2.73 and 2.85, suggesting that the irradiated proton was H14 and the two multiplets were in a position similar to the original "H15" in the molecule (the chemical shifts were also similar). In consideration of the coupling relationships among the proton at 8 4.09, the two multiplets and the methyl group, the structure was assigned as 287. Because the carbon at C2' was chiral, so the compound was actually a pair of diastereomers, which were supported by the presence of slightly different two sets of signals in lH and  13  C NMR spectra. No further separation of these two diastereomers was  attempted. Further studies as to the formation of 287 were not undertaken though hydride transfer and methyl group migration might be involved in the process.235"237 In summary, the newly developed sequence from isodehydroabietenolide (193) to triptophenolide (106) significantly increased the yield of triptophenolide (106) (30% overall yield from 193) (Scheme 3.21). The overall yield from dehydroabietic acid (157) via 193 to  174  triptophenolide (106) was 11%, which was much higher (about 40 times) than that previously reported (0.27%). 145 Finally, the strategy of the synthesis of ring C activated compounds via isodehydroabietenolide (193) was more logical, and the use of triptophenolide (106) as starting material for biotransformations or other studies was now realistic.  Scheme 3.21 Overall Yields for the Synthesis of 106 from 193  3 . 2 . 4 Synthesis of Demethyl Isoneotriptophenolide (DINTP, 288)  (Me)  demethyl isoneotriptophenolide (DINTP) (288)  175  In planing the synthetic strategy to 288, it was noted that the synthesis of methyl 12,14dihydroxy-dehydroabietate (290) by diazotization of the corresponding 12,14-diamine 289 was reported unsuccessful.238 Consequently, the appropriate 12,14-diamino intermediate which could be made available by reduction of the 12,14-dinitro compound 272, did not appear attractive. NH 2  diazotization X C0 2 Me  289  It is known that phenols can be oxidized by lead tetraacetate (LTA) to ortho-and paraquinol acetates (acetoxycyclohexadienones),239'240 which then can be converted to resorcinol derivatives by a dienone-phenol rearrangement and subsequent hydrolysis.241"243 By using this strategy, M. Shimagaki etal. synthesized methyl 12,14-dihydroxy-dehydroabietate (290) from methyl 12-hydroxy-dehydroabietate (291). 238 Preliminary results* showed that this reaction was also applicable to our synthesized isotriptophenolide (194), so studies along this route were pursued. OH  194  F. Kuri-Brena of this group had done some preliminary work in reactions of 194 with LTA and subsequent rearrangement.  176  Isotriptophenolide (194) was treated with LTA (1.1 equiv.) in acetic acid solution at room temperature.239,244 The resulting crude product (para- and ortho-quinol acetates, 292 and 293) was not purified since both isomers should rearrange to the same product 238 (Scheme 3.22). Thus the crude product was dissolved in trifluoroacetic acid anhydride (TFAA) and stirred at room temperature for 1 day.238 The resulting product was chromatographed to give 14DINTP acetate 294 in 49% yield. Another trial was run under similar conditions but using 2 equivalents of LTA, and the rearrangement step was continued for about 42 h. Chromatography of the crude product afforded 294 in 54% yield. It was apparent that increases in amount of LTA and reaction time provided only a slightly improvement in the yield.  Scheme 3.22 Oxidation of 194 with LTA and Subsequent Dienone-Phenol Rearrangement  The IR spectrum of compound 294 showed a hydroxyl group at 3400 cm -1 and the acetate at 1740 (broad, overlapping with the absorption from the butenolide moiety) and 1220  177  cm-1. The *H NMR spectrum displayed the HI 1 signal at 8 6.66 and the acetate methyl group at 8 2.32. Hydrolysis of the acetate provided some difficulty. The hydrolysis conditions used in the preliminary experiments 238 failed to achieve removal of the acetate function. Reflux of the acetate 294 with moderate amounts of concentrated sulfuric acid in methanol were similarly unsuccessful. Finally, the hydrolysis was achieved in an aqueous sulfuric acid solution under reflux temperature. The yield of crude DINTP (288, one spot on TLC) was essentially quantitative. DINTP (288) was recrystallized in ethyl acetate and dichloromethane to give colorless prisms (mp 198-199°C). Its IR spectrum showed hydroxyl group absorption at 3400 cm -1 . The l  H NMR displayed the H l l (NOE effect with Hip) signal at quite high field (8 6.36), as  compared with isotriptophenolide (194) and triptophenolide (106), but the H15 resonance was remarkably shifted downfield to about 8 3.4 due to the adjacent two hydroxyl groups.  295  One by-product 295 was isolated as a red powder (mp 174-175°C) with a molecular formula of C20H20O4. Its IR spectrum showed a hydroxyl group at 3300 cm -1 and a conjugated enone at 1640, 1590 and 1560 cm -1 . The lH NMR spectrum was similar to that of 250, but the HI(3 signal was shifted downfield to 8 3.51 (IH, ddd, / = 13.6, 4.5, 2.4 Hz), suggesting that there was a group at CI l. 183 At the low field, there were four signals, one proton at 8 7.68 (IH, br s) was exchangeable with D2O, indicating it was a hydroxyl proton. Irradiation of this proton resulted in an enhancement of the Hip signal, suggesting that the hydroxyl group was at CI 1.  178  Irradiation of the signal at 8 6.93 (IH, s, H14) enhanced resonances of H15 (5 3.14, septet, J = 6.9 Hz), the two isopropyl methyls and a proton at 8 6.69 (IH, d, / = 6.6 Hz, H7). The H7 proton was coupled with a proton at 8 6.29 (IH, d,J = 6.6 Hz, H6). Therefore, the structure of this compound was determined as 295. This compound has a similar structure as the antimalarial principles 295-A and 295-B isolated from an African plant Hoslundia opposita Vahl. Our lH NMR data were comparable with those relevant in the literature.245 A possible mechanism for the formation of 295 is presented in Scheme 3.23.  295-A R = benzoyl 295-B R = cinnamoyl  295 (i)  295 (ii)  295 Scheme 3.23 Proposed Mechanism for the Formation of 295  A series of experiments were conducted to examine whether the yield of DINTP could be improved (Table 3.1). The results showed that both acetic acid and chloroform could be used as  179  solvents in the oxidation step. The amount of LTA could range from 2.2 to 1.5 equivalents without significantly changing the yield. In the rearrangement step, zinc chloride seemed to have little effect on the yield.243 Since chloroform can be used at lower temperature, and the workup conditions are less complex, this solvent is preferable to acetic acid.  Table 3.1  Yields for Conversion of 194 to 288  Oxidation entry  a  Hydrolysis  Rearrangement  194  solvent  LTA  (mg)  (35 mL)  1  500  HOAc  2  20  2  500  HOAc  2  3  500  HOAc  4  500  CHC13  time  solvent (mL) ZnCl2 time  (equiv.) (min)  time  overall  (mg)  (h)  (h)  yield (%)  TFAA (8)  -  48  4  53  10  TFAA(8)  -  48  4  54  1.5  20  TFAA (6)  50  50.5  4  56  2.2  30  TFAA (5.6)a  50  50  4  56  additional 0.8 mL TFA was added.  Because the overall yield of 288 from isotriptophenolide (194) was about 56% (three steps), it was acceptable for our purpose and no further studies were pursued.  180  CHAPTER 4  4.1  BIOTRANSFORMATION STUDIES  Introduction  In the study of diterpene metabolites from the plant cell cultures of T. wilfordii, a biosynthetic pathway from dehydroabietane (196) to Tl (1) and Td (2) was proposed (vide supra). In order to provide further information relating to possible biosynthetic intermediates and simultaneously, provide a family of novel diterpene analogs for pharmacological screening, a series of biotransformation experiments with the compounds synthesized in the previous chapter were pursued. Generally speaking, a biotransformation refers to the conversion of a given substrate into another product by biological means such as the use of micro-organisms, cultures of animal and plant cells, or purified and partially purified enzymes. 246,247 In the context of organic synthesis, biotransformations are reactions which utilize these biological catalysts.248'249 Substrates are not limited to those compounds which are normally present in the living organism (the biosynthetic intermediates of secondary metabolism) but instead may be "foreign" compounds which could undergo chemical alternations. 248 The role of biotransformation in organic synthesis is one of support rather than supplantation. Biotransformations should be employed to their advantages when a given reaction step is not easily accomplished by "ordinary" chemical methods.250 The best results are obtained with the combination of these methods.248 In initial biotransformation studies with TRP4a cell cultures, interest was focused on the formation of the butenolide moiety in the Tl (1) or Td (2) skeleton.  Preliminary  biotransformation studies examined radio-labeled dehydroabietic acid (157) and the hydroxy ester 182,159 as well as unlabeled synthetic precursors 210,213(i) and 296-300.192  181  Labeled 297, an allylic alcohol, was shown to be converted to the corresponding aldehyde 301 and acid 299a 160 (Scheme 4.1). This result is in agreement with the proposed mechanism for the formation of the butenolide moiety in the molecule (section 2.2).  182  297  Scheme 4.1  301  299a  Biotransformation of Allylic Alcohol 297 with TRP4a Cell Cultures  Administration of synthetic isodehydroabietenolide (193) as a precursor, one in which the butenolide ring system is intact, showed that TRP4a cell cultures was able to convert this compound to certain hydroxylated compounds; however, no epoxide formation was observed (Scheme 4.2). 160 At the time of the above study, the natural occurrence of 193 and 206 in the cell cultures was unknown. Labeling experiments clearly demonstrated that the hydroxylated products were derived, at least in part, from the exogenous starting material. Biotransformations of the ring C hydroxylated precursor isotriptophenolide (194) by TRP4a cell cultures gave only small amounts of one product, the methyl ether of the starting material (Scheme 4.3). 161 The lack of an obvious biotransformation of 194 contradicted the idea that a ring C activated precursor could be transformed by the cells. However, there were still many opportunities for progress. The work included in this thesis was part of an effort to validate this hypothesis. In the present biotransformation studies, attention was paid to selection of appropriate experimental conditions and to examination of the relevant precursors. In the studies with the plant cell cultures of T. wilfordii, our focus was directed at possible yield improvement of the diterpene triepoxides Tl (1) and Td (2) by incubation of appropriate synthetic precursors, and also at synthesis of new epoxide analogs (through biotransformations). Therefore, the proposed biosynthetic intermediate, triptophenolide (106), and related precursors were used as substrates.  183  Of the substrates prepared, triptophenolide (106) and the 7-oxo-butenolide 206 have been isolated from TRP4a cell cultures, the others are "foreign" precursors.  302  303  Scheme 4.2  Biotransformation of the Butenolide 193 with TRP4a Cell Cultures  Scheme 4.3  Biotransformation of Isotriptophenolide (194) with TRP4a Cell Cultures  Cell cultures were grown and maintained under the conditions described in Chapter 2. Cell cultures grown in MSNA0.5K0.5 medium were used in biotransformation studies. In general, the precursor was dissolved in a certain amount of ethanol, added to the culture, and  184  culture incubation continued under standard conditions. Samples were extracted and subjected to column chromatography to isolate the products. In time course studies, samples were taken at certain intervals and analyzed by TLC and/or high performance liquid chromatography (HPLC).  4.2  Results and Discussion  4.2.1  Biotransformation of 7-Oxo and 7|3-Hydroxy-isodehydroabietenoIide with TRP4a Cell Cultures  Biotransformation of isodehydroabietenolide (193) with TRP4a cell cultures produced hydroxylated compounds and ketones. In order to examine whether some of these compounds could be further transformed, 7-oxo-isodehydroabietenolide (206) and 7|3-hydroxyisodehydroabietenolide (240) were incubated with TRP4a cell cultures. The starting material 206 (100 mg) was dissolved in ethanol (8 mL) and added to a 14day-old culture (2 x 550 mL). A control experiment (4 mL ethanol added in 550 mL culture), and a blank experiment (50 mg starting material in 4 mL ethanol added to 550 mL sterile medium (no cells) were also conducted under the same conditions. Samples were taken after incubation for 1,4 and 6 days, and the culture was harvested after 7 days. The samples were extracted and then analyzed by TLC. Results showed that there were no significant changes to the starting material as well as Tl (1) and Td (2) in the test samples. The biotransformation was repeated under the same conditions except that 50 mg of starting material was dissolved in 50 mL of ethanol and incubated in 550 mL of the cultures. Samples taken after incubation for 1, 5, 7 days also showed no reaction. 7p-Hydroxy-isodehydroabietenolide (240) was also evaluated. The starting material (80 mg) in ethanol (4 mL) was added to a 14-day-old culture (550 mL). A control and a blank experiments were also run under the same conditions. Samples taken after 3 days of incubation  185  showed a spot corresponding to 206 on TLC and some remaining starting material. After the culture was harvested at 7 days, the starting material was almost completely oxidized to ketone 206. The blank sample showed little reaction; therefore, the oxidation of 240 to 206 was caused by the enzymes in the cells, not simply by atmospheric oxygen (Scheme 4.4).  Scheme 4.4  Biotransformation of Alcohol 240 with TRP4a Cell Cultures  The results from biotransformation of 240 and 206 suggested that cell-produced 7-oxo compounds, such as 199, 211, 206 and 108 are likely to be formed in a similar manner, i.e., hydroxylation and further oxidation. However, experiments to confirm this speculation were not performed. These experiments, as a supplement to the early investigation by M. Roberts, 160 completed the biotransformation of the isodehydroabietenolide (193) series. These results demonstrated again the need for the introduction of activating groups in ring C.  4.2.2  Biotransformation of Isotriptophenolide (194) and Related Compounds with TRP4a Cell Cultures  Biotransformation of 194 with Older TRP4a Cell Cultures (Trp#300a)  Previous biotransformation studies with isotriptophenolide (194) in TRP4a cell cultures used a 7-day-old culture and a 7 day incubation.161 Earlier studies had showed that Td (2) and  186  Tl (1) were produced during late growth phase.147 This indicates that the enzymes responsible for the biosynthesis of Tl (1) and Td (2) may only be present or active in older cells. Therefore, biotransformations with older cells were studied. Isotriptophenolide (194, 46.5 mg) dissolved in ethanol (1 mL) was added to a 21-dayold TRP4a cell culture (550 mL). Samples were taken after incubation for 1, 2 and 4 days, respectively. The samples were extracted and then analyzed by TLC (toluene-chloroform-ethyl acetate-formic acid, 105:48:45:3, developed twice). The results showed that the starting material (Rf 0.53) remained in substantial quantities, although a less polar spot (Rf 0.71) appeared after 1 day, and a polar spot (Rf 0.24) appeared after 2 days of incubation. No significant changes to Tl (1) and Td (2) were observed. The culture was harvested after 5 days and the cells and broth were extracted. Column chromatography of the cell and broth extracts gave isotriptophenolide methyl ether (304, 3 mg) and the starting material (194, 17 mg). Attempt of isolation of the polar spot failed due to decomposition on the column (very likely the polar compound was 7-hydroxy-isotriptophenolide, vide infra). Again, with the exception of the possible formation of 7-hydroxy-isotriptophenolide, biotransformation of isotriptophenolide (194) with an older cell culture gave no new products other than the methyl ether 304.  Compound 304 was isolated as a white powder with a molecular formula of C21H26O3. The *H NMR spectrum was identical to that of the authentic compound previously obtained.  187  Increase of Starting Material (194) Solubility and Isolation of New Products (Trp#301)  There are various factors which may result in a lack of incorporation or biotransformation of exogenous precursors:  the specificity of the enzyme system, the  impermeability of the cellular membrane towards the exogenous material, or an induced alteration or destruction of the substrate before reaching the active sites.184 One of the factors which might have played a role in the previous experiments was that the starting material was not highly soluble in the cell culture medium and thus may not pass through the cell membrane. To increase the solubility of the starting material, the amount of the solvent (ethanol) used to dissolve the starting material was increased from 0.2% to 4.5% with respect to the total volume of the culture. Thus, isotriptophenolide (194, 100 mg) was dissolved in ethanol (50 mL) and added evenly to two flasks of TRP4a cell cultures (21 days old, 550 mL each). Samples were taken after incubation for 26 h (Sample 1) and 72 h (Sample 2), and were subsequently extracted with ethyl acetate. TLC (toluene-chloroform-ethyl acetate-formic acid, 105:48:45:3, developed twice) showed that Sample 1 contained a considerable amount of starting material (Rf 0.57), and that the methyl ether was hardly detectable. A new, polar, UV active spot (Rf 0.41) was detected in a small amount by TLC. Another polar spot (Rf 0.22), which had been found in Trp#300a, was also present. Sample 2 showed increased amounts of the polar spots (Rf 0.41 and Rf 0.22) and much residual starting material. No significant changes were observed in the amounts of Tl (1) and Td (2) as determined by TLC. The culture was harvested after incubation for 96 h, and the cells and the broth were extracted to yield a cell extract (257.9 mg) and a broth extract (162.4 mg). Both extracts were subjected to repeated column chromatography (Figure 4.1), yielding 8 products along with recovered starting material (Table 4.1). The total recovery of material was about 89%.  188  Table 4.1  a  Compounds Isolated from Biotransformation of 194 with TRP4a Cell Cultures (Trp#301)  Compound  l94a  246  305 + 306b  309  245  308  307  304  mg  58.6  8.9  15.2  2.6  1.6  1.6  0.7  0.6  %c  _  20  35  6  4  4  1.5  1  Recovered starting material; b Mixture of 305 and 306 (7.9 mg) plus 305 (1.0 mg) and 306 (6.3 mg); c Based on  recovered starting material.  Compound 307 had a molecular formula of C22H28O4. The *H NMR spectrum showed that it was actually a mixture of two isomers, 7a-ethoxy (307a) and 7(3-ethoxyisotriptophenolide (307b), in a ratio of approximately 4 to 1. For the oc isomer, the HI 1 and H14 signals were located at 8 6.70 and 7.11, respectively, whereas for the (3 isomer, these two signals were at 8 6.68 and 7.29, respectively. As in 7-hydroxy-isodehydroabietenolide (see discussion in the synthesis of 246b), the H14 resonance of the P isomer was found at lower field than that of the a isomer. The H7p signal of 307a was located at 5 4.38 (br dd, / = 4.2, 1.6 Hz), at a higher field than that in 307b. The methylene protons of the ethoxy group appeared around 6 3.53-3.37 and the methyl group at 8 1.24.  In agreement with the  assignment, irradiation of the 7{3 proton resulted in signal enhancements of HI4, H6 and the methylene protons of the ethoxy group. The two isomers were not readily separable by column chromatography and they decomposed rapidly; therefore, further separation was not attempted  307a  7a-ethoxy  307b  7p-ethoxy  189  Broth Extract (162.4 mg) c. c. hexanes-EtOAc (6:4); EtOAc  1 II c. c.  I III  IV  VI  c. c. CH2C12EtOAc (8:2)  CH2C12EtOAc (9:1)  c. c.  CH2Cl2-EtOAc (8:2) c. c.  194 307 309 (2.0 mg) (0.7 mg) (2.6 mg)  VII  CH2C12EtOAc 305 (90:5) 1.0 mg  306 245 (6.3 mg) (1.6 mg)  c. c.  hexanes-EtOAc (6:4)  308 246 d - 6 mg) (0.8 mg)  Cell Extract (257.9 mg) c. c.  hexanes-EtOAc (6:4); EtOAc  I + V II  IV + IX  c. c.  c. c  CH2C12EtOAc (90:5)  VI+ X  CH2C12EtOAc (90:5)  304 194 305+306 246 (0.6 mg) (56.6 mg) (7.9 mg) (4.5 mg)  Figure 4.1  c. c.  hexanesEtOAc (5:5)  VIB  VIB + VII c. c.  hexanesEtOAc (5:5)  246 (3.6 mg)  Column Chromatographic Separation of Extracts from TRP#301  190  > Oc:  H  o246a 7a-OH 246b 7J3-OH  309  305  306  308  245 OCH 3  307a 7cc-OEt 307b 7P-OEt  304  Compound 246 was isolated as a mixture of 7cc-hydroxy (246a) and 7J3-hydroxyisotriptophenolide (246b), with 246a as the major component. The two sets of ! H NMR signals were identical with those obtained from the synthetic compounds (section 3.2.2). As mentioned earlier, since these compounds were readily decomposed, further separation of the two isomers was not performed.  191  Compound 309 was isolated as an optically active ( [ a g -37.9°, c = 0.070, CHCI3), white crystalline solid (mp 190-192°C) with a molecular formula of C20H24O4. One additional oxygen atom in the molecule suggested a new hydroxyl group. The IR spectrum showed bands for the hydroxyl group at 3650 and 1040 cm"1. Its !H NMR spectrum displayed some similarity to that of isotriptophenolide (194). The C19 proton signals were slightly shifted downfield to 8 4.89 (2H, br AB q , Av = 0.18 ppm, J = 17.1 Hz) with wider splitting (Av) of the AB quartet. As mentioned in previous chapter, this usually meant a change at C5, for example, hydroxylation or formation of a double bond. The phenolic hydroxyl proton signal was found at 8 4.78, which was exchangeable with D2O. The C7 protons at 8 3.01 (2H, m) were coupled with a proton at 8 1.97 (1H, m) and another in the multiplet between 8 2.04 and 2.26 (3H), indicating they were C6 protons. Protons at 8 2.36 (1H, m) and 2.54 (1H, br d, J = 17.3 Hz) were coupled with each other, and both had cross peaks with C19 protons, indicating that they were C2 protons. Irradiation of HI 1 enhanced signals of the C12 hydroxyl proton at 8 4.78, of another hydroxyl proton at 8 1.90 (D2O exchangeable), and of HI{3 centered at 8 2.19 as part of the three-proton multiplet (one was H6, vide supra) between 8 2.04 and 2.26. The two C2 protons were coupled with Hip, and with the H l a centered at 8 2.12 which was the third proton in the Hip, H6 multiplet. No C5 proton was found in the spectrum; therefore, the new hydroxyl group must be at C5. As the C20 methyl protons (8 1.08) were still at higher field than the isopropyl methyl groups (about 8 1.23), the A and B rings were believed to be in trans junction, in other words, the hydroxyl group at C5 was a oriented. This assignment was supported by the downfield shift of the H l a signal to 8 2.12 from 1.68 as compared with its  192  counterpart in the starting material 194 due to the 1,3-diaxial relation. The 13 C NMR spectrum showed the C5 carbon signal at 8 70.1. The mass spectrum exhibited the molecular ion peak at m/z 328, an M - H2O fragment at 310 and an M - H2O - CH3 fragment at 295, respectively. All these data were consistent with the structure assigned to 309.  Compound 308 was obtained as an optically active ([oc]D = +171.9°, c = 0.281, CHCI3), pale yellow powder (mp 92-95°C). The molecular formula was determined as C20H24O4. Its IR spectrum showed absorption bands for hydroxyl groups at 3600 cm-1 (sharp, phenolic hydroxyl group) and 3470 cm -1 (broad, alkyl hydroxyl group). The ! H NMR spectrum exhibited the C19 protons at 8 4.90 (2H, br AB q , Av = 0.1 ppm, J = 12 Hz), which also showed a wider split between the AB quartet. All the other proton signals were found except the H5 signal. Furthermore, the C20 methyl signal (8 1.32) was at lower field than the isopropyl methyl signals (8 1.21). These data suggested that the new hydroxyl group was at C5, but in a P orientation, thus resulting in an A/B cis junction. Irradiation of the hydroxyl proton signal at 8 2.30 resulted in an enhancement of C20 methyl signal and thereby confirmed that the hydroxyl at C5 was p oriented. All the remaining protons were assigned according to COSY and NOE experimental results. The mixture of 305 and 306 was repeatedly chromatographed on preparative TLC with anhydrous ether to give pure 305 and 306, respectively. Compound 306 was isolated as optically active ( [ a | , = +161.5°, c = 0.117, CHCI3), colorless needles (mp 198-200°C, dec) with a molecular formula of C20H22O4. The UV  193  spectrum displayed absorption maxims at 219.5 nm (log e = 4.17) and 260.2 nm (log 8 = 4.36). Its IR spectrum showed quinone type absorption bands at 1680 and 1650 cm -1 and a possible epoxide band at 920 cm -1 .  2  O—'19  306  The 1 H NMR spectrum showed that 306 was still a butenolide type compound but structural changes were apparent. The two proton signals of ring C were shifted upwards to 6 6.48 (IH, s) and 5.99 (IH, s), respectively, and a new signal appeared at 8 3.82 (IH, d, J = 2.4 Hz) coupled with some aliphatic protons (COSY spectrum, Figure 4.2). Irradiation of the signal at 8 6.48 resulted in signal enhancements of Hlf3 at 8 2.19 (IH, br dd, J = 13.3, 6.1 Hz) and of the C20 methyl group at 8 0.95, thereby indicating that the irradiated proton was H l l (Figure 4.3a). Irradiation of the proton at 8 5.99 showed NOEs to H15 at 8 2.95 (IH, septet d, / = 6.9, 0.9 Hz) and the two isopropyl methyl groups at 1.05 and 1.08 (3H each, both d, / = 6.9 Hz). Also enhanced was the new signal at 8 3.82, suggesting that the irradiated proton was H14 and the new signal was due to a proton at C7 (Figure 4.3b). Compared with the spectra of Tl (1) and Td (2), the chemical shift and the appearance of the new signal from H7 was very similar to those signals from protons attached to the epoxide carbons of Tl (1) and Td (2). The H7 of 306 was coupled with a proton at 8 2.02 (IH, br dd, / = 14.4, 12.8 Hz) and a multiplet at 8 2.30 (2H), indicating that the two C6 protons were in these locations (Figure 4.2, d, e). Also coupled with the C6 protons was a signal at 8 3.23 (IH, br d, J = 12.8 Hz), which suggested that it was due to H5 (Figure 4.2, f, g).  J  L  _A__/s_  JVAA-J* -  1.  _  2.3  3.3  4.3  PPM 4.0  3. 3 PFM  2.0  a: H19/H2P; b: H19/H2cc; c: H19/H5; d: H7P/H6P; e: H7p/H6cc; f: H5/H6p; g: H5/H6cc; h: H15/H16.H17; i: H2a/Hla; j : H2a/H2p, Hip; k: H2p/Hla; 1: H6a/H6p; m: H2p/Hlp; n: Hlp/Hla  Figure 4.2  Expanded COSY Spectrum of Compound 306  "W  • •MVMian^.jrmEA—•«w^*»'i^  k >#||W»**%.*» »!»•**¥ Wf#>l ^ ^ ^ ^ ^ M ^ W * * ^ »•«••« I • • • • V I ' I * * I*^ A M ^ ^ W I W H H W ^ V I M  d  il n w y ^ i —fcM*—*'  ~*F~*ty\f{T'  —ft»*»J 0**+****mJm*AMftm*+**i+  H.wA^  y^^Xt*'—'-*^.*  J  4V^+*m0W**# y . n ^ r " - i  < M ^ n M * M y M iiw—i—w M^^JgLf^ytM»**VW^^i »W|W»M**','•.»•> ^ y i • — < H M W > ^ « « *  W -  iH .,*" " , J t J M ^ r n '  ***~~~~*%(({  ^ i ^ * ^ . * ^ ^hfr*AL*w» w ^ m o w i y A ^ * " ' ! —• M*^VWWV * ,m» *,— • i n ^ ^ y ^ ^ { r ' S i / y l ^ * A r * . • • • ** •  _  ~A-*v,-fr-  i t*i**» m**i^**mi*m*  ^MwV>**«S\fl**i»**Allilfaj%Nftfav^^  *jl*>*^\»*~*t}**SF'*L)lk*~^^  iMJ*W**+»*. H16.17  H20  HI!  1114  H7P  H(n H6D " » HIP  H19  X "  • i •  7.5  7.8  6.5  E.0  5.!  I •  s.e  1  IJL^LIL^LJ r •  4 8 fPH  3.5  5.8  2.5  2.C  —I I.S  a, {Hll}: Hip, H20; b, {H14}: H7P, H15, H16.H17; c, {H7p}: H14, H6a, H6P; d, {H5}: H19, H6a, Hla; e, {Hla}: H5, H2a, HIP; f, {H20}: HI 1, H2p, Hip, H6p; g, off-resonance spectrum (400 MHz, CDCI3). ({irradiated}: enhanced)  Figure 4.3  NOE Difference Spectra of Compound 306  196  The COSY spectrum also showed correlations between H19 signals (5 4.70, 2H, br AB q , Av = 0.08 ppm, J = 17.1 Hz) and the H5 signal (Figure 4.2, c). Also coupled with H19 was a proton at 5 2.47 (1H, br d, J = 17.9 Hz) and another proton in the multiplet at 8 2.30 (2H, one was H6) (Figure 4.2, b, a). This indicated that the signal at 8 2.47 was one of the C2 protons, whereas the other was in the multiplet at 8 2.30. Irradiation of the C20 methyl group resulted in signal enhancements of HI 1, HI B, the H6 at 8 2.02 and the H2 in the multiplet at 8 2.30, which showed that the H6 at 8 2.02 and the H2 in the multiplet (8 2.30) were B oriented (Figure 4.3f). The chemical shift of the C5 proton was at lower field as compared with other related compounds, which indicated that the epoxy group was a oriented and thus had an approximate 1,3-diaxial interaction with H5. Based on all the spectral information, this compound was identified as the (7,8)B-epoxy-dienone 306. Examination of the molecular model showed that the dihedral angles between H7B/H6a and H7B/H6B were of a similar magnitude (less than 90°) while that of H7B/H6a was slightly smaller (Figure 4.5). Therefore, H7p was coupled with both C6 protons but more strongly with H6a, which made the H7B signal appear as a broad doublet. Experience suggested that, if the 7,8-epoxy group was B oriented as in Tl (1) and Td (2), then the C19 protons would appear as a broad singlet. But in the spectrum of 306, the C19 protons were still evident as an AB quartet, which supported the assignment of the epoxide to an a position. All the signals of the 13 C NMR were assigned in accord with the results from HMBC and HMQC experiments. The carbon signals of C7 and C8 were found at 8 60.9 and 54.3, respectively. Finally, an X-ray analysis of the single crystals from a large scale experiment (vide infra) confirmed that the structural assignment was correct (Figure 4.4).  197  02  H4 C2 H3"°  C16  C17  yci5  02  W  1 C13  C12J  02 C14  C8J  Ol  Cll  "STc7 o"  iC9  ciqj  C12  A C6  r\  C5 C4 C19 04<  Figure 4.4  03  X-ray Structure of 306  H13  198  The resemblance between 305 and 306 in the *H NMR spectra suggested that 305 was probably the P isomer. Compound 305 was isolated as colorless plates (mp 230-232°C, dec) with a specific optical rotation of -295.5° ([cc^ , c - 0.0670, CHCI3). The molecular formula was determined as C20H22O4, which was identical to that of 306. Its IR spectrum also showed the quinone absorption bands at 1660 and 1640 cm"1 and an epoxide band at 910 cm -1 .  305  The *H NMR spectrum of 305 displayed the HI 1 and H14 signals at 8 6.42 (IH, s) and 5.99 (IH, d, / = 1.0 Hz), respectively. The C19 protons appeared as a broad singlet at 5 4.68 (2H). The corresponding H7 was located at 5 3.84 (IH, d, J = 5.8 Hz) and was strongly coupled with a proton at 2.26 (IH, ddd, / = 14.5, 5.8, 5.8 Hz) thereby indicating that it was a C6 proton (Figure 4.6, d). This C6 proton then was coupled with its geminal partner at 8 2.13 (IH, br dd, / = 14.5, 13.4 Hz) (Figure 4.6, m) and both were correlated with H5 at 8 2.58 (IH, br d, / = 13.4 Hz) (Figure 4.6, f, g). Irradiation of H l l at 8 6.42 resulted in signal enhancements of H i p at 8 2.10 (IH, br dd, J = 12.9, 5.4 Hz, 11%) and of H l a at 8 1.62 (IH, br ddd, J = 12.9, 12.4, 6.0 Hz, 3%) (Figure 4.7a). The H19 signals showed cross peaks to H5 (Figure 4.6, c), and the two H2 signals at 8 2.30 (IH, m) and 2.47 (IH, br d, J = 18.5 Hz) (Figure 4.6, a, b), respectively. Compared with 306, the relatively higher field of the H5 signal and the lower field of the C20 methyl proton signal (8 1.17, 3H) indicated that the epoxy group between C7 and C8 was p oriented. Irradiation of the C20 methyl group enhanced the signals of the H2 at 8 2.30 and the H6 at 8 2.13, suggesting that these two signals arose from the corresponding P oriented  199  protons (Figure 4.7f). Saturation of the H7 resonance caused NOEs to H14 and the other C6 proton (which must be H6a) at 5 2.26 and confirmed that the C7 proton was an a proton, i.e., the epoxy group was P oriented (Figure 4.7c). Examination of the molecular model revealed that the dihedral angle between H7a and H6oc was very small but the one between H7a and H6P was around 90° (Figure 4.5). This is the reason why the H7oc showed a cross peak only to H6a in the COSY spectrum and appeared as a sharp doublet in the lH NMR spectrum, exactly astheH7inTl(l)orTd(2).  O  H6p  -90 degree H7oc H6a  305  Figure 4.5  The  306  Comparison of Dihedral Angles between H7 and H6 in Compounds 305 and 306  13  C NMR and HMBC, HMQC experiments were performed and all the carbon  signals were assigned accordingly. The carbon signals of the epoxy group (C7 and C8) were located at 5 64.0 and 56.0, respectively. Single crystals of 305 obtained from a large scale experiment (vide infra) were attempted for X-ray analysis; however, the crystals suffered X-ray induced decomposition, thus no X-ray analysis was obtained. Based on the correlation between 305 and 306 in the *H NMR spectra, the confirmation of the structure of 306 was an indirect confirmation of the structure assigned to 305.  _  2  _  3  PPM -r-i—r-f—r—r—r—I—r-r—  4.0 PPM  '! 2.0  1.0  a: H19/H2p; H19/H2oc; c: H19/H5; d: H7cc/H6<x; e: H15/H16.H17; f: H5/H6p; g: H5/H6a; h: H2a/H2p; i: H2a/Hla; j : H2a/Hlp; k: H2p/Hla; 1: H2p/Hlp; m: H6a/H6p; n: Hlp/Hla  Figure 4.6  Expanded COSY Spectram of Compound 305  201  ••>» •  ^^••^tj^^i^na ttMk wwWA»  -*>~  I  V  •H0*«H—»I%.Y>WI  I  f~\r~ J*"*1'*'  •Avi*  l»|M«W« »M"« ill ^ J W W M * i W " l > i i ' > * \ ^ * » * « « « M ^ « ' « M I ^ ' < l ' ^ ' * t > * V«*Ny»ym»-  »  i'  "<'«<J^',»'"'-"'yw.  «»«•  „J,  »'• "VV" Tj  |b«*^«'  tyl—*>*y/»-<Sl  ' ^ i i ^ y u l > ' w > ^ ^ ' ^ ^ w i ^ ^ ^ ^ ^ i i i ^ » y ' i » ^ i / | N i ^ i i U i i i ^ i ^ | | n»ii»%u4i»^>w  J*'*m„m. U . . .  ,»•!  ^tmift^  mm*******  •»AtM  H20  Hll  g  HI4  H16.17  H19 H7a H6a H2pj I HIP  >v 1  A.  _/  , • • ' ' , ' • ' • I ' ' • ' I • ' • • I ' ' ' ' I '" ?.S 7. a E.5 £.8 S.5 5.U  I  ' ' 1 • ' • ' I ' ' ' ' I ' ' • ' I • ' ' ' I ' ' • ' .' • ' ' • ' I • ' ' " I • ' ' ' ) ' 4.5 '.8 J.5 i.» 2.5 2.3 1.5 1.0 .5  PPM  a, {Hll}: Hip, Hla; b, {H14}: H7a, H15, H16.H17; c, {H7a}: H14, H6a; d, {H5}: H6a, Hla; e, {Hip, H6p}: H14, H19, H7a, H2a, H6a, Hla, H20; f, {H20}: H2p, H6P; g, off-resonance spectrum (400 MHz, CDCI3). ({irradiated}: enhanced)  Figure 4.7  NOE Difference Spectra of Compound 305  202  The isolation of the two novel mono-epoxides, 305 and 306, from biotransformation of isotriptophenolide (194) with TRP4a cell cultures was one of the most interesting results obtained in recent years. It strongly demonstrated that the cell culture of T. wilfordii has the ability, as we anticipated, to transform the C ring "activated" precursor to the corresponding epoxide. The combined yield of the epoxides was moderate at 35% based on recovered starting material, but it had potential for improvement. Moreover, this result has expanded our knowledge of the capability of the biological system in this cell culture and has provided a unique venue for the production of such novel diterpene epoxides as 305 and 306. These two compounds may have some biological activities and may also be used as intermediates to synthesize Tl analogs by chemical means or through biotransformations. Significance in this result also existed as it was the first example in which a "para-alkylated" phenol was converted to apara-epoxy dienone in a single step through biotransformation.  Biotransformation of 194 with TRP4a Cell Cultures: Time Course Studies  There are many factors which may affect the yield of the epoxides, 305 and 306. The most important ones are the age of the cells, the duration of the incubation and the starting material-to-culture ratio. A series of biotransformations were carried out in order to investigate these influences and hence to find the appropriate conditions for the optimal formation of the products.  Influence of the Cell Age and Incubation Time  Typically, TRP4a cell cultures grown in MSNA0.5K0.5 have a 7 to 10 day lag phase followed by a rapid growth phase that gives way to a stationary phase by 28-35 days. The production of Td (2) was found highest in the later growth phase.147 To examine the influences  203  of culture age on the yield of the epoxides, three typical ages (7, 15 and 21 days) were chosen for the studies. Isotriptophenolide (194, 200 mg) in ethanol (50 mL) was added to TRP4a cell cultures (1.1 L) of 7, 15 and 21 days old, respectively. The cultures were incubated under standard conditions and samples were taken at regular intervals. The samples were filtered, and the resulting broth and cells were extracted separately with ethyl acetate and analyzed by TLC and HPLC. An experiment with a 15-day-old cell culture was examined first and was followed by experiments with cell cultures of 7 and 21 days old. Figure 4.8 shows the changes of isotriptophenolide (194) and the epoxides 305 and 306 with incubation time in the biotransformation with the 15-day-old TRP4a cell culture (Trp#305). From Figure 4.8a and b, it can be seen that the starting material decreased very rapidly from initial value of 200 mg/L (not shown in the figures) to a total of about 110 mg/L after the first day of incubation, then the decrease slowed down. The epoxides were mainly in the broth and increased significantly during 7 to 9 days of incubation. The starting material was more or less equally distributed between the cells and the broth. The amount "in" the cells may include both the starting material which was in the cells during the biotransformation and that which was only adsorbed on the cell walls or cell debris