Citation
Chemistry of taxanes and Taxus species

Material Information

Title:
Chemistry of taxanes and Taxus species
Creator:
Johnson, James Harvey, 1970-
Publication Date:
Language:
English
Physical Description:
xiv, 146 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Acetates ( jstor )
Alcohols ( jstor )
Esters ( jstor )
Ethers ( jstor )
Nitrate esters ( jstor )
Room temperature ( jstor )
Sodium ( jstor )
Solvents ( jstor )
Sulfates ( jstor )
Taxoids ( jstor )
Antineoplastic Agents, Phytogenic -- chemistry ( mesh )
Antineoplastic Agents, Phytogenic -- isolation & purification ( mesh )
Department of Medicinal Chemistry thesis Ph.D ( mesh )
Dissertations, Academic -- College of Pharmacy -- Department of Medicinal Chemistry -- UF ( mesh )
Paclitaxel -- analogs & derivatives ( mesh )
Paclitaxel -- chemistry ( mesh )
Paclitaxel -- isolation & purification ( mesh )
Plants, Medicinal ( mesh )
Research ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Bibliography: leaves 141-145.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by James Harvey Johnson Jr.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
029903704 ( ALEPH )
51624221 ( OCLC )

Downloads

This item has the following downloads:


Full Text












CHEMISTRY OF TAXANES AND TAXUS SPECIES
















By

JAMES HARVEY JOHNSON JR.














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998


























This work is dedicated to my father whose untimely passing due to the disease that this work addresses on April 28, 1998 has left a deep void in my life that will never be filled. Although he would not be considered an educated man by most standards he taught me more than any textbook or professor ever could. I hoped he could be here when this work was completed but that was not to be. Nevertheless I hope that this accomplishment would make him as proud of me as I was of him. I love and miss you daddy.















ACKNOWLEDGMENTS


There are many people who have helped me along this journey, some through deeds and some through inspiration. Firstly I would like to acknowledge those in academia that have made this event possible. To the late Koppaka V. Rao, my graduate advisor, I would I would like to give much thanks for his patience and understanding and for the wisdom and experience which he has imparted on me. He will certainly be missed but I hope to keep his memory alive through the work I accomplish throughout my career. To the late Maya Ganguli, my undergraduate advisor and mentor, I would like to give thanks for opening my eyes to the exciting world of chemistry. I could not have asked for a teacher more caring and better suited than you to help me get started in my chemistry education. I would also like to acknowledge other faculty members that have given their time and energy so that I could accomplish this goal; John Perrin for accepting the duty of being my graduate advisor upon Dr. Rao's passing, Margaret James, Ken Sloan, and William Dolbier for serving on my thesis committee, and Bob Higgins for guidence and always interesting conversation during my undergraduate studies. Finally I must give thanks to my fellow students along the way both graduate and undergraduate with special thanks to Veronica Hall and Bernice Kidd for their help and friendship during my undergraduate years and to Ravi Orugunty and John Juchurn whom I shared my trials and tribulations with for the last 5 years as lab mates in Dr. Rao's laboratory.


iii









On more personal notes I first must give thanks to my wife, Amy, for all her support through this ordeal. Many times I knew she wanted me to stay home more often, many times I knew she was tired of living the impoverished life of a graduate student but through it all she remained faithful and supportive and for this I owe her a large debt of graditude. I would also like to acknowlegde the family of my wife for all their fmnacial support that was given to us when we needed it. This also holds true for the families of my sisters Melinda and Teresa. I am very grateful to you all for your support. Last of all I would like to acknowledge my parents. It has been said that there is only one point in a persons life in which he is totally at the mercy of fortune, this is when he is born. I must say that I was blessed with parents who knew love and showed it to each other as well as myself, and that they showed this love by instilling In me discipline, integrety, and respect. I certainly could not be here today if it were not for them.


























IV














TABLE OF CONTENTS


page

A CK N O W LED G M EN TS ................................................................................. iii

L IS T O F T A B L E S ................................................................................................ v iii

L IST O F F IG U R E S .................................... .................................................. . ix

A B B R E V IA T IO N S ....................... ................................................................. xii

A B S T R A C T ............................................ ......... ............................................. .. xii

CHAPTERS

1 HISTORY AND BACKGROUD OF PACLITAXEL ...................1...

In tro d u ctio n .............................................................................................. 1
Early W ork and Structural Elucidation ............................................................. 1
M ech an ism o f A ctio n ....................................................................................... 6
T otal S ynth esis ......................................................................................... . 7
Structure-Activity Relationships ...................................... 12

2 ISOLATION OF TAXOID AND NON-TAXOID COMPOUNDS
FR O M TAX U S SPECIE S ..................................................................... 15

L arge Scale Isolation Process ....................................................................... 15
Isolation of Minor Compounds from the Bark of Taxus brevifolia................ 22
Synthesis of T axam airin B ........................................................................... 26
Isolation of Minor Compounds from Taxusfloridana ...................... 35
Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde ..................................... 42
E x p e rim e n ta l ................................. .............. .................................................4 4
Isolation of Minor Compounds from Taxus brevifoa ........................... 44
Acetylation of 1 3-Hydroxy-7-Deacetyl Baccatin I .................................. 47
Isolation of Taxamairin A (38) from Taxus brevifolia ............................ 47
Methylation of Taxamairin A ................................................................ 48
Synthesis of Taxam airin B (39) .............................................................. 49
Isolation of Minor Compounds from Taxusfloridana ............................. 57

v









page

Acetylation of Taxiflorine .................................. ................ 61
O xidation of T axiflorine ......................... ................................................. 61
Acetylation of Taxchinin L (83) ............................................... ............... 62
Synthesis of Trans-2, 6-Dimethoxy Cinnamaldehyde (85) .......................... 62

3 PREPARATION OF NITRATE ESTERS OF PACLITAXEL
AND RELATED TAXANES .................................................... ............. 66

Complete Nitration of Paclitaxel and Related Taxanes ................................... 66
Regioselective Nitrations of Paclitaxel and Related Taxanes ............................ 70
Reaction of Taxanes Nitrate Esters ................................................................74
Complete Reductive Hydrolysis of Nitrate Esters with Zn
and A cetic A cid ...........................................................................74
R eaction w ith N aBH 4 ..............................................................................74
Reaction with Ammonium Sulfide ............. .................... 76
Acetylation of Taxane Nitrate Esters ......................................... .............76
R eaction w ith N aN 3 ................................................................... ............. 77
E x p e rim e n ta l ...................... ...................................................... ....................... 8 6
Com plete N itrations of Taxanes ................................................................ 86
Regioselective Nitration of Paclitaxel ...................................... .............. 88
Regioselective Nitration of 10-Deacetyl Baccatin III .................................. 89
Conversion of 10-Deacetyl Paclitaxel-7-3-Xyloside to
10-Deacetyl Paclitaxel 122 ................................. 91
Regioselective Nitrations of 10-Deacetyl Paclitaxel ................................ 92
Regioselective Nitration of 10-Deacetyl Paclitaxel-7-j3-Xyloside ................ 94
Reductive Denitration of Paclitaxel-7, 2'-Dinitrate Ester ........................... 98
Reaction of Paclitaxel-7, 2'-Dinitrate with NaBH4 .................................. 98
Selective Denitration of Paclitaxel-7, 2'-Dinitrate Ester ...........................99
Acetylation of 10-Deacetyl Paclitaxel-7-03-Xyloside2", 3", 4", 10, 2'-Pentanitrate Ester ............................ ............. 100
Reaction of Paclitaxel-7, 2'-Dinitrate Ester with NaN3 .................. .......... 102
Synthesis of 2'-Oxo-Paclitaxel-7-Nitrate Ester from
Paclitaxel-7, 2'-Dinitrate Ester ............................. 103
Synthesis of 2'-Oxo-Paclitaxel-7-Nitrate Ester from Paclitaxel7-M ononitrate Ester .............. .............................. 103
Acetylation of 2'-Oxo-Paclitaxel-7-Mononitrate Ester ............................ 104

4 SYNTHESIS OF ANALOGUES WITH POTENTIALLY
IMPROVED WATER SOLUBILITY .............................................107

Intro du ctio n ........................................................................... .......... ............... 10 7
Synthesis of Analogues Starting from 10-Deacetyl Paclitaxel-7-Xyloside ......... 108
Attempted Synthesis of Taxane Glycosides ........... .... ............. 115

vi











L1210 Cytotoxicity of Analogues .................................. 125
Experim ental ................................................................................. ....... ........ 126
Oxidation of Xyloside with Periodate ....... ........................................ 127
Condensation of Dialdehyde with Malonic Acid ....... ................ 127
Condensation of Dialdehyde with Nitromethane ........................ .............. 128
Reduction of Dialdehyde to the Diol ........... .................. 128
General Procedure for Reductive Aminations ............................. 129
Synthesis of Glucose Pentaacetate ............ ..................133
Synthesis of loc-Bromo-Tetraacetyl Glucose ................ ................. 134
Synthesis of 1-Hydroxy-Tetraacetyl Glucose ............................ 134
Synthesis of lc-Trichloroacetimidate-Tetraacetyl Glucose ........................ 135
Synthesis of Tetraacetyl Phenyl Thioglucoside ........................... ............. 135
Synthesis of Tetraacetyl Glucose, Phenyl Sulfoxide ................................. 136
Preparation of Tetraacetyl Benzyl P3-Glucoside by the
Koenigs-Knorr method ................... ................... 136
Preparation of Tetraacetyl Benzyl 3-Glucoside by the
Trichloroacetim idate M ethod ........................................................ 137
Preparation of Tetraacetyl 3-Sitosterol P3-Glucoside by the
Koenigs-Knorr M ethod ................................................... ............ 137
Preparation of Tetraacetyl P3-Sitosterol P3-Glucoside by the
Trichloroacetimidate M ethod ....................................... ............... 138
Attempted Glucosylation of 2'-Acetyl Paclitaxel by the
Koenigs-Knorr Method ............. ....................... 138
L1210 Cytotoxicity Assay ....... ... ........ .................. 139

LIST OF REFERENCES ..................... ................. 141

BIOGRAPHICAL SKETCH ............................................. 146


















vii















LIST OF TABLES


Table page

1-1 Physical and Chemical Properties of Paclitaxel ......................... ............... 3

2-1 1H and 13C NMR Values for Related Abeo-Taxanes ................................ 39

3-1 1H and "3C NMR Values for Completely Nitrated Taxanes ......................... 68

4-1 ID50 Values of Paclitaxel and Xylosides in Tubuline Assay ........................ 109

4-2 L1210 Cytotoxicity of Paclitaxel and Analogues ....................................... 126






























viii















LIST OF FIGURES


Figure page

1-1 Structure of P aclitaxel ......................................................................... . 2

1-2 1H NM R Spectrum of Paclitaxel ........................................................... 4

1-3 Halogenated Products of Methanolysis Used for X-Ray Crystallography .... 5 1-4 Neutral and Alkaline Oxidation of Paclitaxel .......................................... 7

1-5 N icolaou Synthesis of Paclitaxel ........................................................... 9

1-6 H olton Synthesis of Paclitaxel ................................................................ 10

1-7 Holton Synthesis of Paclitaxel .................................... 11

1-8 Structure-A ctivity R elationships ............................................................ 14

2-1 Structure of M ajor Taxanes .................................................................... 16

2-2 Reverse-Phase Isolation of Taxanes ....................................................... 19

2-3 Ozonolysis of Cephalomannine/Paclitaxel Mixture ...................... 21

2-4 Compounds from the Bark of Taxus brevifolia ...................................... 23

2-5 Acetylation of I 3-Hydroxy-7-Deacetyl Baccatin I ................................. 23

2-6 Structure of Abeo-Abietane Diterpenoids .............................................. 25

2-7 NOE Correlations of Taxamairin A ....................................................... 25

2-8 Retrosynthetic Analysis of Taxamairin B .............................. 27

2-9 Literature Synthesis of Taxamairin B ..................................................... 28


ix









Figure Page

2-10 Literature Synthesis ofTaxamairin B ........... .... ................. 30

2-11 Synthesis of Taxam airin B ...................................................................... 32

2-12 Reactions of W rong Regioselectivity .......................................................33

2-13 Structures of Taxiflorine and Related Compounds ...................................37

2-14 Oxidation of Taxiflorine .......................................................................... 37

2-15 FHMBC Correlation of Taxiflorine Acetate .......................... 38

2-16 Compounds from the Needles of Taxus floridana .................................... 41

2-17 Synthesis of Trans-2, 6-Dimethoxy Cinnamaldehyde ................................ 43

3-1 N itration of Paclitaxel, 7-OH > 2'-OH ....................................................... 67

3-2 Nitration of 10-Deacetyl Paclitaxel-7- 3-Xyloside 1"-OH, 2"-OH,
3"-OH > 10-OH > 2'-OH ........................................................... 72

3-3 Regioselective Acetylation of 10-Deacetyl Paclitaxel-7- 3-Xyloside ............ 72

3-4 Conversion of 10-Deacetyl Paclitaxel Xyloside to 10-Deacetyl Paclitaxel ... 73 3-5 Reductive Denitration of Paclitaxel ............................ 75

3-6 Hydrolysis of the Side-Chain of Paclitaxel-7, 2'-Dinitrate with NaBH4 ....... 75

3-7 Regioselective Denitration of Paclitaxel-7, 2'-Dinitrate with
Am m onium Sulfide ......................................................................77

3-8 Enol Acetate Formation of Nitrate Esters ................................. ...............78

3-9 Reaction of 2'-Nitrate Ester with N aN3 ................................ ........ ...... 80

3-10 Oxidation of Paclitaxel-7-Nitrate Ester .............. ................... 82

3-11 Acetylation of K eto-Ester ........................................................ ............... 83






x








Figure page

3-12 Mechanism of Keto-Ester Degradation ............................. 85

4-1 Synthesis of Ionizable A nalogues ............................................................... 111

4-2 Condensation w ith M alonic A cid ............................................................... 112

4-3 Condensation w ith N itrom ethane ............................................................... 114

4-4 D ialdehyde R eduction to D iol ................................................................... 116

4-5 R eductiv e A m inations ............................................................................... 117

4-6 Synthesis of G lycosyl D onors .................................................................... 120

4-7 K oenigs-K norr G lycosylation ..................................................................... 122

4-8 Trichloroacetimidate Glycosylation ................................. 123

4-9 Rearrangement of 2'-Acetyl Paclitaxel .................................................. 124




























xi















ABBREVIATIONS


Ac acetate
Bn benzyl
Bz benzoate
CIMS chemical ionization mass spectroscopy DCC dicyclohexylcarbodiimide DDQ 2, 3-dichloro-5, 6-dicyano-1, 4-benzoquinone DMF dimethylformamide DMSO dimethyl sulfoxide EIMS electron impact mass spectroscopy FABMS fast atom bombardment mass spectroscopy HMBC heteronuclear multiple bond correlation HPLC high pressure liquid chromatography LAH lithium aluminum anhydride LDA lithium diisopropylamide NMR nuclear magnetic resonance NOE nuclear Overhauser effect PDC pyridinium dichromate PTSA para-toluene sulphonic acid RaNi rainey nickel TBS tert-butyl dimethyl silyl TES triethyl silyl Tf- triflate
TLC thin layer chromatography TMEDA tetramethylethylenediamine














xii















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

CHEMISTRY OF TAXANES AND TAXUS SPECIES By

James Harvey Johnson Jr.

December 1998



Chairman: Koppaka V. Rao (deceased) Co-Chairman: John Perrin
Major Department: Medicinal Chemistry

The chemistry of both taxane diterpenoids and Taxus species are studied. Several taxanes are isolation from both Taxus brevifolia and Taxusfloridana. In addition to these taxanes several non-taxane compounds are also isolated. One of these, trans-2, 6dimethoxycinnamaldehyde is a novel structure that has been synthesized. Another nontaxane, taxamairin B, which belongs to rare class of diterpenes, is synthesized by a more efficient route than was reported in the literature. Nitrate ester forming reactions are also studied with paclitaxel and closely related analogues. It is shown that this reaction is regioselective in many cases and thus illustrates the potential use of nitrate esters as protecting groups in taxane chemistry. Also several unexpected reactions of these nitrate esters are explored including an unusually rearrangement in which the nitrated paclitaxel



xiii








side chain reacts in mild base to yield the corresponding baccatin III and dibenzamide. Finally, a series of potentially more water soluble paclitaxel analogues are prepared by oxidizing the naturally occurring 10-deacetyl paclitaxel-7-xyloside with periodate and then reacting the resulting dialdehyde with amines and carbon nucleophiles. These compounds are also tested for cytotoxicity in the L1210 assay system. Although these compounds are not as active as paclitaxel most are more active than the xyloside from which they are obtained.




































x1v














CHAPTER 1
HISTORY AND BACKGROUND OF PACLITAXEL Introduction


Paclitaxel (TaxolTM) (1) is a potent antitumor agent that was originally isolated from the Pacific yew tree, Taxus brevifolia, by Wall and Wani in 1971 (Wani et al., 1971). The structure of paclitaxel is that of a tetracyclic diterpene ester (Figure 1-1). Its structure has several unusual features including an oxetane ring, an 8-membered B-ring, a bridgehead double bond, 11 asymmetric centers, and a N-substituted phenylisoserine ester. This structure has provided a great challenge to synthetic organic chemists since its elucidation but was conquered initially by two groups simultaneously and by other groups since (Nicolaou et al., 1994; Holton et al., 1994a; Holton et al., 1994b). The ongoing phytochemical study of Taxus brevifolia as well as other Taxuts sp. has yielded many taxanes other than paclitaxel. Compounds with rearranged ring systems, various types of esters, as well as glycosides have been isolated and several reviews have been published (Kingston et al., 1993; Das et al., 1995; Appendino, 1995).


Early Work and Structural Elucidation


The pioneering work concerning paclitaxel began in the late 1950s when the National Cancer Institute started a screening program of plant extracts using tumor


I






2



0




NH 0 18HO 9 OHc2



OBz Oc



Figure 1-1: Structure of Paclitaxel


systems models in vivo and tumor cell lines. From these studies the stemn bark extract of the Pacific yew tree was shown to display cytotoxicity in the KB assay and also activity against carcinosarcoma in rats and leukemia in mice. In connection with this NCI screening program, Wall and his collaborators studied the in vitro bioassay guided fractionation of the active extract and in 1969 paclitaxel was isolated and shown to be the most active constituent of" the extract. This isolation was carried Out by extracting the dried stem bark with 95% ethanol. The extract was then partitioned between water and 4 1 chloroform : methanol. The organic layer was evaporated to a solid and purified by a 3step Craig countercurrent distribution method which yielded paclitaxel in a yield of 0.004%. As soon as paclitaxel had been isolated in pure form, the structure of the compound was investigated using available spectroscopic methods. Although methods for ultraviolet, infrared, and mass spectrometry were at a reasonably advanced stage in the








late 1960s, NMR was relatively primitive compared to the sophisticated instrumentation and procedures now available. Some of the physical and chemical properties of paclitaxel are shown in Table 1-1 and the 1H NMR spectrum is shown in Figure 1-2.


Tablie 1-1: Physical and Chemical Properties of Paclitaxel
1.) Needles from 50% aqueous methanol or ether
2.) mp 213-216' C
3.) []D2 -49.60 (MeOH)
4.) Unstable towards mineral acid and base
5.) Forms mono and diacetate
6.) Analysis Calcd. for C47H51NO14: C, 66.11; H, 6.20; N, 1.64
Found: C, 65.98; H, 6.10; N, 1.57. Required m/z 853. Found m/z 853
7.) UV 2,rnax (MeOH) 227 nm (F 29,800)


It was evident by this time that paclitaxel probably contained the taxane skeleton. A number of taxane derivatives had been reported in previous literature. It was evident that paclitaxel was more complex than previously reported taxanes since its molecular weight from high resolution mass spectrometry was C47H51NO14, corresponding to a molecular weight of 853. The evidence then indicated that paclitaxel was comprised of a taxane nucleus to which an ester was attached, as preliminary experiments indicated that an ester was easily cleaved from the rest of the molecule. Attempts were made to prepare crystalline halogenated derivatives of paclitaxel, however none had properties suitable for x-ray analysis. Paclitaxel was therefore subjected to a mild base catalyzed methanolysis at 0' C, which yielded a nitrogen containing cx-hydroxy methyl ester, C17H7NO4, a tetraol, C29H36010, and methyl acetate. The methyl ester thus obtained by the mild methanolysis procedure was converted to a parabromobenzoate ester (2) and characterized by x-ray analysis as C24H20BrNO5 with the structure shown in Figure 1-2. The ester may be







4


CL






---------------7=7




























Figure 1-2: 1H NMR Spectrum of Paclitaxel







5

0

NH O OCH3
0

2



Br




I

YO O
0 0 0.11


HO O O...
0


HO
HO OBz OAc


3


Figure 1-3: Halogenated Products of Methanolysis Used for X-Ray Crystallography


regarded as an N-benzoyl derivative of (2R, 3S)-3-phenylisoserine. The tetraol formed by the methanolysis of paclitaxel was converted to a bisiodoacetate (3), C33H3812012, which again received x-ray analysis. The structure is shown in Figure 1-3.






6
Since the ester could have originally been joined to hydroxyl groups at either C-7, C-10, or C--13, it was necessary to establish at which of these hydroxyl moieties the ester had originally been located. When paclitaxel was oxidized with MnO2 under neutral conditions, no reaction occurred. However, MnO2 oxidation of paclitaxel under alkaline conditions smoothly yielded a reaction product (4) with the structure shown in Figure 1-4. It is evident that MnO2 oxidation of paclitaxel under neutral conditions did not effect the hydroxyl groups available for oxidation at C-7 and C-2'. When paclitaxel was oxidized with alkaline MnO2 an analogue of baccatin III with a conjugated carbonyl moiety as shown in Figure 1-4 was obtained. It is well known that MnO2 oxidation of allylic hydroxyl groups under alkaline conditions smoothly forms the corresponding conjugated ketone. This reaction in conjunction with the x-ray structure determination of the structures of the ester and taxane moieties established the structure of paclitaxel (Wani et al., 1971).


Mechanism of Action


Although paclitaxel displayed good activity against human tumor xenographs and murine B 16 melanoma, its cytotoxic properties were not very different from other drugs being tested during the 1970s. Rather, it's attraction to pharmacologists was its unique structure which suggested the possibility of a novel mechanism for an anti-tumor drug. This mechanism was subsequently identified in 1979 by Horwitz and collaborators (Schiff et al., 1979). Paclitaxel proved to be a potent inhibitor of eukaryotic cell replication, blocking cells in the late G2-M phase of the cell cycle. It is an unusual mitotic inhibitor







7

Paclitaxel MnO2 (pH 7) No Reac.
aq. acetone, reflux




Paclitaxel

MnO2 (pH 8) aq. acetone, reflux

AcO O OH





HO OBz OAc
OBz

4

Figure 1-4: Neutral and Alkaline Oxidation of Paclitaxel


because, unlike the vinca alkaloids and colchicine which inhibit microtubule assembly, it promotes the formation of discrete bundles of stable microtubules that result from the reorganization of the microtubule cytoskeleton. The novel characteristic of paclitaxel is its ability to polymerize tubulin in vitro in the absence of guanosine 5'-triphosphate (GTP), which is normally required for tubulin assembly.


Total Synthesis


As mentioned, the total synthesis of paclitaxel has recently been achieved by various groups, however one group alone cannot claim to be the first to accomplish this






8
daunting task as Nicolaou and Holton published their work simultaneously in two separate journals. Their respective works are now briefly described.

Relying on the previous work of other groups as well as his own studies, Nicolaou envisioned the late formation of the oxetane (D) ring, oxygenation/reduction of the C-13 position, and attachment of the side chain. Thus, the problem was perceived as limited to the assembly of paclitaxel's ABC ring system in either its fully functionalized form or a form that would serve as its progenitor. Figure 1-5 displays the retrosynthetic analysis involving the bond disconnections on which the synthetic strategy was based. Thus, in the synthetic direction the following key operations were performed: 1) two fragments (7 and 8) representing precursors to rings A and C were coupled by a Shapiro reaction and a McMurry coupling to assemble the ABC ring skeleton; 2) installment of the oxetane ring; 3) addition of the various substituents around the peripheries of rings B and C; 4) oxygenation at C-13; and 5) esterification to attach the side chain. Both precursors to rings A and C were made possible using the Diels-Alder transform which led to starting materials that were either commercially available or known in the literature.

In contrast to the convergent synthesis by Nicolaou, Holton took a more linear approach. The facile epimerization of paclitaxel at C-7 is well documented, and has been postulated to occur via a retroaldol-aldol process. Holton chose therefore to pursue a synthetic strategy in which this stereocenter would be introduced at an early stage and carried throughout most of the synthesis in the absence of a C-9 carbonyl group, thereby avoiding epimerization. Thus, his route to paclitaxel proceeds retrosynthetically through the C-7 protected baccatin III 13 to the tricyclic ketone 14, which arises from C ring







9

0


O
NHOAcO O OH
,N O A ..,




5 HOOe OBz Ac

1. Esterification 2. Oxygenation
3. McMurry Pinacol Coupling
4. Shapiro Coupling


TESO 0 OTBS OTBS OBn


N -- CA
Ph Bz I'
67 NNHSO2Ar O

6 O
Diels-Alder
Diels-Alder Diels-Alder
IDiels-Alder


Cl OAc 0 O
CN OEt HO
9
10 11

OH 12


Figure 1-5: Nicolaou Synthesis of Paclitaxel


closure of a precursor properly functionalized at C-1, C-2, C-3, C-7, and C-8 (15). Synthesis of this precursor was made possible by conformational control of the eight







10
AcO 0 OH



Paclitaxel H

HO =- OAc
OBz
13



eo Peeo OP




PO 15 PO O

15 14




1.. -----R0
PO PO. ... P ..








16 17




Figure 1-6: Holton Synthesis of Paclitaxel


membered B ring, via ketone 16. Ketone 16 was projected to arise from an aldol condensation of the enolate of ketone 17, the formation of which would be made possible









f11



18 19





0



21 HO 20




TESO



0 TBSO ......

22 OH 23



Figure 1-7: Holton Synthesis of Paclitaxel


by conformational control exerted by the C-10 group. This scheme is shown in Figure 16. The synthesis of this starting ketone begins with camphor (18). Camphor is converted to P3-patchouline (19) and its epoxide also known as "Patchino" (20). Patchino is then converted to epoxide 21 and then rearranged to diol 22 which is followed by epoxy alcohol fragmentation to give the starting ketone 23 as is shown in Figure 1-7.







12

Although there have been many publications concerning the synthesis of the phenyl isoserine side chain, the most common and that which was used in both of these methods is using the f3-lactam 6 in Figure 1-5.


Structure-Activity Relationships


During the past decade a tremendous amount of work has been performed to determine what parts of the structure of paclitaxel are necessary to illicit biological activity and reviews have been published (Commercon et al., 1995; Chen & Farina, 1995; Kingston, 1995; Ojima et al., 1995; Georg et al., 1995). The structure of paclitaxel can be divided into 3 sections when discussing structure-activity relationships and these are: 1) the N-benzoyl phenylisoserine side chain; 2) the southern hemisphere including C-2, C-4, and the oxetane ring; and 3) the northern hemisphere including C-7, C-9, and C-10.

The side chain plays a major role in the biological functions of this antitumor agent and without the side chain the resulting baccatin III is inactive. Protection of the C-2' hydroxyl group results in major loss of activity in tubulin assays but if the group is labile (acetate) then the activity remains in cell culture presumably acting as a prodrug. Structural modifications of the paclitaxel side chain have been reported by several groups. These studies reveal a number of interesting features and important findings include the following: 1) the C-3' amide group is critical although the amide's aryl group may be substituted by other aryl or alkyl groups; 2) the C-3' aryl group is required since replacement by a methyl group reduces activity but, if larger alkyl groups are used, the activity remains; 3 ) the C-3' bound nitrogen can be replaced by an oxygen atom without








significant loss of activity; 4) one of the C-2' or C-3' polar functions can be removed without significant effect, but the removal of both or interchange of their positions causes dramatic loss of activity and; 5) the (2'S, 3'R) naturally occurring isomer is the most active of the four possible isomers.

Concerning the southern hemisphere of the structure, this area is also very important in terms of biological activity. First of all the oxetane ring is necessary for activity. Structural and molecular modeling studies show that this 4-membered ring is involved in a conformational lock of the diterpene skeleton and the C-13 side chain through a pseudo chair conformation of ring C. The C-2 benzoyl group is also necessary for activity as the C-2 debenzoyl paclitaxel showed little in vitro cytotoxicity; however, some groups have shown that modified benzoyl groups or aryl acyl groups do retain the cytotoxicity. The C-4 acetate is not as important as the C-2 benzoate in that if the acetate is removed the activity is reduced only slightly.

The northern hemisphere is the least sensitive part of the paclitaxel structure. The C-7 hydroxyl may be esterified, epimerized, or removed without significant loss of activity. Specifically a xylosyl group at C-7 actually increases the activity in the tubulin binding assay but leads to decreased activity in cell culture. Presumably there is a transport problem associated with the xyloside that causes this decreased activity. The C-9 keto group may be reduced which actually slightly improves activity and the C-10 acetyl or acetoxy group may be removed without significant loss of activity.

It should also be said that contraction of the A ring does not reduce the tubulindisassembly inhibition activity very much in spite of the significant structural change







14
acetyl or acetoxy reduction group may be improves removed without activity N-acyl group significant loss slighty
required of activity may be esterified
0 epimerized or
AcO 0D O removed without
NH 0 significant loss of
activity

oxetane ring
0 2 0 required for
24 H activity
phenyl group/ OBz
or a close free 2'-hydroxyl I,-*
analog required group or a benzoyloxy group removal of acetate
hydrolysable essential; certain reduces activity slight ester thereof substituted groups required have improved
activity



Figure 1-8: Structure-Activity Relationships


implied by this conversion. These analogues, however, are much less cytotoxic in cell

culture which may be due to the instability of the A ring contracted analogues in cell

culture media or to its failure to enter the cell. A summary of structure-activity

relationships is shown in Figure 1-8.















CHAPTER 2
ISOLATION OF TAXOID AND NON-TAXOID COMPOUNDS FROM TAXUS SPECIES


Large Scale Isolation Process


Although paclitaxel is one of the most promising anti-tumor drugs to receive FDA approval in many years, it has been beset with many problems not the least of which is adequate production of the drug. The original, and until recently major source of the compound was the bark of the Pacific yew (Taxus brevifolia), from which paclitaxel was isolated in a yield of 0.01-0.013% on a large scale. Although several related taxanes that can serve as precursors for the semi-synthesis of paclitaxel, for example, 10-deacetyl baccatin III (25), co-occur in the bark with paclitaxel, there are no reports to indicate that these are being isolated from the bark on a large scale. Thus the low yields of paclitaxel realized by the original process, the apparent unavailability of other useful taxanes analogues, and the environmental concerns raised by the need to cut the slow-growing yew trees for harvesting the bark, are some of the reasons why the bark is no longer considered an attractive source for the large scale production of paclitaxel.

Among the alternatives that are being actively studied are the following: 1) isolation of 10-deacetyl baccatin III from the European yew (Taxus baccata) and its semi15







16

HO 0 OH


H O .......
O
HOH
HO OBz OAc
OBz
25


0


O



HO OAc
OBz

26 R1 =H, R2 = Ac, R3 = phenyl 27 R1 = 1-xylosyl, R2 = H, R3 = tiglyl 28 R1 = 3-xylosyl, R2 = H, R3 = phenyl 29 R1 = 3-xylosyl, R2 = H, R3 = n-pentyl 30 R1 = 13-xylosyl, R2 = Ac, R3 = phenyl 31 R1 = H, R2 = H, R3 = phenyl 32 R1 = H, R2 = Ac, R3 = tiglyl


Figure 2-1: Structure of Major Taxanes


synthetic conversion to paclitaxel and 2) large-scale cultivation of the ornamental yew (Taxus media Hicksii) and isolation of paclitaxel from its needles/twigs. Among the future alternatives are total synthesis, of which various schemes have been published, and isolation from large-scale plant cell culture.







17

In recent years this laboratory has developed a large-scale isolation procedure using a single reverse-phase column (Rao, 1993; Rao et al. 1995). This procedure has several advantages over other published procedures some of which are that it is much simpler, gives higher yields of paclitaxel, and yields several other taxanes which can be converted to paclitaxel. Specifically, the following yields are obtained for the major compounds from the bark: paclitaxel (26) (0.04%), 10-deacetyl baccatin III (25) (0.02%), 10-deacetyl paclitaxel-7-03-xyloside (28) (0.1%), 10-deacetyl paclitaxel-C-7- 3-xyloside

(29) (0.04%), 10-deacetyl cephalomannine-7- 3-xyloside (27) (0.006%), paclitaxel-7-3xyloside (30) (0.008%), 10-deacetyl paclitaxel (31) (0.008%), and cephalomannine (32) (0.004%) (Figure 2-1). The procedure for this process is defined below and in Figure 2-2.

Air dried yew bark (200-250 lbs.) was extracted with 100 gallons of methanol in a batch process a total of three times with each extraction lasting one day. The pooled methanol extracts were concentrated under reduced pressure (<300 C) using a semicontinuously operated still until the volume of the concentrate reached 20-25 gallons. Extraction of the concentrated methanolic extract with chloroform was performed in 50100 gallon tanks equipped with an air-driven stirrer. The concentrate was stirred with water and chloroform for about 30 minutes, then 2-14 hours were necessary to allow for any emulsion to clear. The chloroform layer was drained off from the bottom and the water layer was extracted two additional times. The pooled chloroform layers were concentrated under a vacuum to a thick syrup which was poured into glass trays and converted to powder using a vacuum oven at 35-400 C. The powder was obtained in a yield of 18-26 g per kg of the bark.







18

For chromatography, stainless steel columns either 4" x 4' or 6" x 6' were used. The columns were packed with C-i18 bonded silica as a slurry in methanol. Approximately 3-4 kg and 12-13 kg of silica were used with the 4" and 6" columns, respectively. After a thorough wash with methanol, the columns were equilibrated with 25% acetonitrile in water. For running the 6" diameter column, the powder from the chloroform extract (22.5 kg) was dissolved in acetonitrile (5 1) and while this mixture was being stirred with equilibrated C-i 8 silica (1 -2 1), it was diluted with water to make 20 1. The mixture was then allowed to stand for 15-30 minutes and the clear supernatant siphoned off into another container. The slurry was applied to the column, followed by part of the supernatant, after which the column was sealed. The remaining supernatant was pumped into the column using a diaphragm metering pump maintaining a pressure of 30-80 psi. After the sample had been pumped onto the column it was eluted with a step gradient of 35, 40, 45, and 50% acetonitrile in water. The change in solvent was dictated by the results of the TLC and HIPLC of the fractions but usually 40-50 1 of each solvent was used. After this, the column was washed with methanol, followed by a mixture of ethyl acetate and ligroin until the effluent was nearly colorless. Following this, the column was again washed with methanol and equilibrated with 25% acetonitrile in water. The column fractions (about 2 1 each) were allowed to stand at room temperature for 2-8 days, by which time crystals appeared in many. Soon after, the crystals were filtered, analyzed for purity and composition by FIIPLC, and recrystallized from the appropriate solvent.

In terms of the elution sequence of the taxanes, the earliest taxane to appear was lO-deacetyl baccatin 111 (25) which crystallized from the fractions eluted by 35%







19

Methanol Extract






Partition Between Water and Chloroform








Water Chloroform

C18 Column
Acetonitrile/Water



I I



Pure Paclitaxel Fractions Fractions Containing Paclitaxel
(Crystalline) and Cephalomannine


Ozonolysis/
Regular Silica Colum Paclitaxel 0.04%
10-Deacetyl Baccatin III 0.02%
10-Deacetyl Paclitaxel-7-Xyloside 0.1% Paclitaxel
10-Deacetyl Paclitaxel-C-7-Xyloside 0.04%





Figure 2-2: Reverse-Phase Isolation of Taxanes







20

acetonitrile in water. The next group of taxanes to be eluted were the various xylosidic taxanes; 10-deacetyl cephalomannine-7-3-xyloside (27), 10-deacetyl paclitaxel-7-03xyloside (28), 10-deacetyl paclitaxel-C-7-13-xyloside (29), and paclitaxel-7-03-xyloside

(30). Of these the first two were well separated. As the elution of 10-deacetyl paclitaxel-73-xyloside was nearing completion, 10-deacetyl paclitaxel-C-7- 3-xyloside started to elute. Halfway though its elution, paclitaxel-7- 3-xyloside and 10-deacetyl paclitaxel (31) started to co-elute. These last two compounds also crystallized together, however separation was readily achieved by running the mixture through a regular silica column using 0-5% methanol in chloroform as solvent.

Continued elution of the column with 50% acetonitrile in water gave cephalomannine (32), followed closely by paclitaxel (26). The earlier part of the band contained mixtures of the two, but the later fractions contained mostly paclitaxel which could be recrystallized. The fractions that contained the mixture were combined and dried to a solid. This solid was then subjected to ozonolysis at -780 C for 45 minutes. This process converted the cephalomannine to the keto-amide 34 but did not disturb paclitaxel (Figure 2-3). After workup this material was run through a regular silica column with 05% acetone in chloroform and the paclitaxel was isolated. It should be pointed out that this process was necessary because paclitaxel and cephalomannine cannot be separated on regular-phase silica.







21


0
AcO 0 OH
NH 0



OH 0
HO OAc
OBz
33



Paclitaxel


03
CH2CI2/CH30H
room temp.



0 0
AcO 0 OH




OH 0
HO H OAc
34 ()Bz



Paclitaxel


Figure 2-3: Ozonolysis of Cephalomannine/Paclitaxel Mixture







22

Isolation of Minor Compounds from the Bark of Taxus brevifolia


Obviously the above process was only used to isolate major compounds; however, many minor compounds exist in the filtrates or in the in-between fractions. A TLC analysis of the filtrates from the region between 10-deacetyl baccatin III and 10-deacetyl paclitaxel (xyloside fractions) showed many interesting compounds, the identity of which could not be determined by comparison with available standards. This material was therefore concentrated to a solid and rechromatographed on regular-phase silica using an elution system of 0-5% acetone in dichloromethane to 0-5% methanol in 5% acetone/dichloromethane.

The first compound to be eluted was l3-hydroxy baccatin I (35). This compound has been known for quite some time and was first isolated from Taxus baccata in 1970 (Della Casa De Marcano & Halsall, 1970). This compound belongs to the baccatin I subfamily because it contains a C-4-C-20 epoxide as opposed to an oxetane ring. Baccatin VI

(36) was eluted next and is another well known taxane isolated for the first time from Taxus baccata in 1975 (Della Casa De Marcano & Halsall, 1975). This compound is sonamed because it is esterified at C-9 as opposed to paclitaxel/10-deacetyl baccatin III which have a C-9 ketone (Figure 2-4).

The next compound to be eluted was 103-hydroxy-7-deacetyl baccatin I (37). This compound was recently isolated from the needles of Taxus brevifolia (Chu et al., 1993) and has been reported to undergo an acetyl migration from C-9 to C-7 when kept in solution. Indeed, this compound did form another spot on TLC when left in solution;







23


AcO )AcAcO OAc OAc

OAcO


OAc

O~ 0OBzOc
35 36


AOH PH OH



AcO'0



37 38


Figure 2-4: Compounds from the Bark of Taxus brevifolia 37



acetic anhydride/pyridine 18 hours, room temp. 35


Figure 2-5: Acetylation of 1 f3-Hydroxy-7-Deacetyl Baccatin I






24

however, no attempt was made to determine if this new compound was the C-7-acetyl, C9-hydroxy isomer. To further confirm its structure 37 was acetylated with acetic anhydride and pyridine and the product matched 1-hydroxy baccatin I (35) in every way (Figure 25). Finally 9-dihydro-13-acetyl baccatin III (38) was eluted as determined by NMR spectroscopy (Figure 2-4). This compound was isolated earlier from the needles of Taxus canadensis (Gunawardana et al., 1992), but the current isolation is the first from the bark of Taxus brevifolia. This compound has received much attention from Abbott Laboratories as a possible precursor to their own paclitaxel analogue.

In addition to this work on the pre-paclitaxel fractions another interesting compound was isolated while attempting to obtain more paclitaxel from the filtrates of paclitaxel-containing reverse-phase fractions. This crystalline compound was eluted with a solvent mixture of 2% acetone in dichloromethane and was given the name brevixanthane because of its yellow color. Based on 1H and 13C NMR spectra it was quickly concluded that brevixanthane belonged to a rare group of diterpenes known as 9(10-*20)-abeoabietane diterpenoids that have previously been isolated from Taxus species. These diterpenes consist of a 6, 7, 6 tricyclic carbon ring system with ring C being aromatic and are a novel diterpene structural class. The only other members of this group include taxamairin A (38) and B (39) from Taxus chinensis var. Mairei (Liang et al., 1987) and brevitaxin (40) from Taxus brevifolia which contains a C6-C3 side chain (Arslanian et al., 1995) (Figure 2-6). Initially, we thought this compound may be novel based on its 1H NMR spectrum. Brevixanthane contained only one methoxyl which had a chemical shift of

3.89 ppm while the methoxyl of taxamairin A was reported to be at 3.99 ppm. All other






25



O O


CH30 OCH3 HO OC

38 39



O
O




40 O O


\HO OCH3

OH


Figure 2-6: Structure of Abeo-Abietane Diterpenoids



O
O
CH3
-- ~ CH3 \/ H H O OCI
41

Figure 2-7: NOE Correlations of Taxamairin A






26

chemical shifts were almost identical. Thus, it was speculated that the methoxyl/hydroxyl positions may be the reverse of that of taxamairin A. However, it was then observed that the UV spectrum of brevixanthane was identical to that reported for taxamairin A while that of taxamairin B was reported to be completely different. This was confirmed by synthesizing taxamairin B from brevixanthane with dimethylsulfate. Thus, since the UV spectra of taxamairins A and B are so different it stands to reason that the UV spectra of taxamairin A and brevixanthane should also be different if they were different structures. This was not the case. To solve this question of structure 'H NNR NOE experiments were performed. These experiments illustrated that if the methoxyl methyl of brevixanthane is irradiated then the phenolic proton and the isopropyl methyne proton are enhanced proving the proximity of the methoxyl to the isopropyl group. Also when the phenol proton was irradiated the methoxyl protons and the C-4 proton was enhanced proving a close proximity between the phenolic group with and carbon 4 of the B ring (Figure 2-7). Thus, it is concluded that brevixanthane was the same compound as taxamairin A.


Synthesis of Taxamairin B


The total synthesis of taxamairin diterpenes has been accomplished by one other group (Wang & Pan, 1995a; Wang et al., 1995b). The synthetic strategy which was used by this group was derived from the retrosynthetic analysis as outlined in Figure 2-8. Ketone 45 was the key synthetic intermediate because it contains the entire carboncyclic framework of taxamairin B. Thus the A ring precursor (49) was readily obtained from 1,3






27

O
O O



42 43
CH30 OCH3 CH30 OCH3







4 O0O

45 CH30 OCH3
CH30 OCH3








BrCH2 OCH3
46 OCH3
47

Figure 2-8: Retrosynthetic Analysis of Taxamairin B


-cyclohexanedione (48) by azeotropic removal of water from a benzene/hexane/isopropyl alcohol solution with PTSA as catalyst in a yield of 95% (Figure 2-9). The synthesis of the C ring was begun by oxidizing o-vanillin (50) with AgO to give the carboxylic acid. This phenolic acid was then dimethylated using dimethylsulfate to yield the methyl ester (51). This is then treated with two equivalents of methyl magnesium bromide to yield the







28



0 0 0 0
2-propanol, hexane
PTSA

48 49

0 0
1.) AgO, NaOH
-~H H20,700 C, 20min OCH3

OH 2.) dimethylsulfate -~ '. OCH3
0C113 K2C03, acetone 51 0013
50
2 CH3MgBr



OH
20% 112S04 reflux, 6 hrs 01

0C113 53 0013

I 52
H2, Ra Ni

1.) n-butyl lithium, TMEDA, TUN 00C

OC32.) (C1120)1 HOCH2 0013

OCH3 0013 55
54 /PBr3, CH2CI2
room temp., 15 min



BrCH2 003


56

Figure 2-9: Literature Synthesis of Taxamairin B






29

tertiary alcohol (52). Dehydration was then achieved by heating with 20% H2SO4. The resulting olefin (53) was then reduced with RaNi and H2. A hydroxymethyl group was attached ortho to the methoxyl by treating 54 with n-butyl lithium in THI and TMEDA and later addition of paraformaldehyde. The hydroxymethyl function was converted to a bromomethyl function using PBr3 in dichloromethane. This bromo compound (56) then underwent nucleophilic substitution with ketone 49 and LDA as the base to yield ketone 57 (Figure 2-10). Ketone 57 was treated with vinyl magnesium bromide and the B ring was closed by a Friedal-Crafts alkylation using BF3-OEt2. Finally the gem-dimethyls were attached using potassium t-butoxide and methyl iodide to produce the key intermediate 60 in which the olefin bond moved into the cycloheptane ring. Oxidation of the allylic and benzylic methylene group to produce the ketone at C-1 was achieved by using excess aqueous 75% t-butyl hydroperoxide and catalytic amounts of chromic anhydride. Finally, olefination of C-4, 5 and C-6, 7 was achieved by heating with excess DDQ in toluene followed by hydrogenation with 10% Pd-C. This then gave the final product taxamairin B.

Since the above synthesis was the only synthesis for this class of diterpenes it was decided that it would be a noteworthy side project to synthesis this type of diterpene by a simpler method. This method also uses a convergent approach by constructing an A ring precursor and a C ring precursor and then bringing the precursors together to form the B ring. The A ring precursor was synthesized by modifying a known method starting with 1, 3-dicyclohexanedione (63) (Shuzi et al., 1991). This diketone was dimethylated with methyl iodide and K2CO3 in refluxing acetone in give a yield of about 65% after vacuum distillation. The 2, 2-dimethyl-1, 3-cyclohexanedione product (64) was then mono-






30

49 + 56


SLDA, THF


/- CH2=CHMgBr \ / Et2O "
57 58 s
H3CO OCH3 H3CO OCH3


I BF3-Et2O toluene

O
t-BuOK O
SCH3I


H3CO OCH3
H3CO OCH3
t-BuOOH CrO3, CH2C12

O O

1.) DDQ, toluene \ / 2.) H2, 10% Pd-C 61 H3CO OCH3 62 H3CO OCH3


Figure 2-10: Literature Synthesis of Taxamairin B


brominated by slowly adding Br2 in dichloromethane at room temperature while closely monitoring the TLC. This process gave a yield of 60% after purifying the product by column chromatography. Finally, 4-bromo-2, 2-dimethyl-1, 3-cyclohexanedione (65) was






31

dehydrohalogenated by refluxing with excess LiCI in DMF for 2 hours. This process gave a yield of 82% after purification by column chromatography of the A ring precursor 2, 2dimethyl-4-cyclohexene- 1, 3-dione (66) (Figure 2-11).

The formation of the C ring precursor was more problematic. Acetovanillone (67) was used as the starting material and this compound was isopropylated by heating with 90% H2S04 and isopropyl alcohol at 60' C for 36 hours. Unfortunately this reaction could not be pushed beyond 50% conversion based on TLC of the reaction mixture, and after purification by column chromatography gave a yield of 40%. However this process is a better alternative than the 5-step process outlined in the previous synthesis for placing the isopropyl group on the ring. Following this, the phenolic group was methylated using dimethylsulfate and K2C03 in refluxing acetone for 2 hours. This process was nearly quantitative to yield the dimethoxy product (68) (Figure 2-11). At this point a one carbon oxygenated substituent had to be introduced between the acetyl and methoxyl groups. Initially, a Vilsmeier-Haack reaction was attempted, but this gave a variety of products in which the acetyl seemed to undergo some reaction; however, no effort was made to characterize these products. Undesirable reactions also occurred with this method if the acetyl was first reduced to an alcohol group or completely reduced to an ethyl group. Attempts were also made to acetoxymethylate the desired position so the resulting acetate could be hydrolyzed and the alcohol oxidized to the aldehyde. The reaction was performed using 85% H3PO4, acetic anhydride, and paraformaldehyde and the chosen substrate was the reduced ethyl compound (73) which has less steric bulk and is more activated than the







32


0 0
CH31 0*

01::::rO K2C03
63 64
Br2, CH202




LiCl, DMF reflux, 2 hrs. Br*

66 65


0 0
1.) 2-propanol, 90% H2SO4

OH 2.) dimethyl sulfate, OCH3
K2C03 *'-Oc OCH3 68 OCH3
67


1 NaBH4
0 1.) n-BuLi OH
TMEDA CH20, -780C

'OCH3 2.) PDC oxidation OCH3
Oct
0 OCH3 69 OCH3
70
0


66 + 70 )0piperidine 0-::pyridine, reflux 00OHti
71 H3CO

Figure 2-11: Synthesis of Taxamairin B







33

0

NaCNBH3, ZnCl Bw

OCH3
1 7 *33 OCI
OCH3 OCH3

72

AcO 85% H3PO4, AC20
paraformaldehyde
73 30,
OCH3
7 *
OCH3



Br

Br2, CH202
73 30

OCH3 75 OCH3




OH Br

Br2, CH202 69
OCH3
7 7 A
6 OCH3




Figure 2-12: Reactions of Wrong Regioselectivity






34

oxygenated analogues. This substrate was obtained by reducing ketone 72 with NaCNBH3 in the presence of ZnC12 (Figure 2-12). Although acetoxymethylation proceeded quite well the regioselectivity was wrong as was determined by NOE experiments (Figure 2-12). These experiments clearly showed enhancement of one of the methoxy methyls when the aromatic proton was irradiated; likewise when the oxygenated methylene was irradiated the isopropyl methyne signal was enhanced.

At this point lithiation of the aromatic ring and reaction with paraformaldehyde seemed to be a more attractive way of introducing a hydroxymethyl group which could then be oxidized to the aldehyde. Initially, lithium/halogen exchange was the desired process but bromination of the reduced alcohol substrate (69) as well as the totally reduced ethyl substrate (73) yielded a brominated product with the wrong regiochemistry (75, 76) (Figure 2-12). This conclusion was reached following NOE experiments as previously described above.

In light of these results a direct lithiation was attempted with n-butyllithium in diethyl ether and TMEDA at -78' C using the reduced alcohol substrate (69) (Figure 211). It is well know that lithium complexes with methoxy groups and therefore addition usually takes place ortho to a methoxy if one is present on the aromatic ring. About 30-45 minutes after adding the n-butyllithium, paraformaldehyde was added. This reaction proceeded smoothly however yields were only around 50%. In any event, the regioselectivity was as desired based on NOE with the hydroxymethyl group adding ortho to the methoxy.






35

The resulting diol was then oxidized to the keto-aldehyde (70) with the mild oxidizing reagent PDC. This product was coupled to 2, 2-dimethyl-4-cyclohexene-1, 3dione (66) by a tandem cross-aldol reaction in pyridine and piperidine (Figure 2-11). Although other products were produced the desired product was the major one. This pathway was expected since the most acidic carbon would be adding to the most electrophilic carbon first. Once this occurs, the 7-membered ring would be expected to close rather easily by another aldol reaction. The minor products were undoubtedly a variety of other cross-aldol products. This major product matched taxamairin B, which was obtained by methylating taxamairin A, in every way.


Isolation of Minor Compounds from Taxusfloridana


The same reverse-phase chromatography protocol previously described was applied to the needles of the Florida yew (Taxusfloridana) with very good results (Rao et al., 1996a). Although there was some question initially if this protocol would work on needles as well as on bark because of the greater content of waxes and pigments found in the needles, these questions were put to rest as all of the lipophilic material remained on the column while using the appropriate taxane solvent system without clogging up the column. It was found that paclitaxel could be isolated from these needles in a yield of 0.0 1% and 10-deacetyl baccatin III could be obtained in yields of 0.06%. Again, by TLC analysis, this time of the pre-paclitaxel fractions, it was found that many unidentifiable compounds were present in the filtrates. These filtrates were combined and evaporated to






36

dryness and reapplied to a regular-phase silica column using dichloromethane with 0-10% acetone and then 10% acetone with 0-10% methanol in dichloromethane.

A few compounds eluted with straight dichloromethane and thus had to be run on another column. This work will be discussed later. Elution with 2% acetone/dichloromethane gave 1 3-hydroxy baccatin I (35, Figure 2-4) as mentioned before (Della Casa De Marcano & Halsall, 1970). This was followed by a compound that was earlier given the name taxiflorine (77). Taxiflorine is an example of an 11(15->1)abeotaxane meaning that the A ring has contracted to contain only 5 carbons. Taxiflorine was previously isolated by our group but was published with an incorrect structure assignment (78) (Rao et al., 1996a). Taxiflorine itself gives difficult to interpret 'H and 13C NMR spectra for reasons that will be discussed later; however, upon acetylation the spectra are easier to interpret. According to the original structural assignment (78), its acetate should be the same compound as 13-acetyl-13-decinnamoyl taxchinin B (79) previously isolated by another group (Das et al., 1995); however, the spectral properties of these two compounds did not match (Table 2-1). It was then postulated that the correct structure of taxiflorine is one in which the C-10 benzoate and C-9 hydroxy groups are reversed so that the hydroxyl is at the C-10 position. To confirm this idea taxiflorine was oxidized with Jones reagent to the ketone (80, Figure 2-14) and its 3C spectrum was compared with those of some known C-9 and C-10 keto taxanes. The carbonyl signal of 80 seen at 192.2 ppm is consistent with an ai, D3-unsaturated ketone system, and in contrast to the 199-204 ppm signal of C-9 keto taxanes. Similarly, the C-12 signal of 80 is at 156.8






37

HO OBz OAc BzO OH OAc


Ac AcOOc



77 78




HO OAc OAc
77Ac..cO ..




79


Figure 2-13: Structure of Taxiflorine and Related Compounds O _OAc OAc



Jones oxidation 77 acetone, 30 min AcO.
room temp.
HO OAc OAe



Figure 2-14: Oxidation of Taxiflorine ppm which is also consistent with the 3-carbon of an x, 1-unsaturated ketone system. The C-12 signal in C-9 keto compounds is usually around 147.0 ppm. Also, the UV spectrum






38






H






AcO OAc





AcO""1''


HtO' Oz OAc



81


Figure 2-15: HMBC Correlation of Taxiflorine Acetate


of 80 displayed a k. at 232 nm with a shoulder at 253 nm, this shoulder is consistent with an c, 13-unsaturated ketone system. Finally, to confirm the revised structure an HMBC spectrum was taken on the taxiflorine acetate (81, Figure 2-15). From this spectrum the interactions between the ortho protons and benzoate carbonyl carbon were clearly visible as was the interaction between the benzoate carbonyl carbon and the proton at C-9. Likewise the C-19 protons interacted with the C-9 carbon, which in turn interacted with the C-9 protons on the regular HETCOR spectrum. With this information in hand the






39

benzoate could be firmly placed at the C-9 position giving the correct structure of

taxiflorine. This correcting structure was identical to a compound recently isolated from

Taxus chinensis var. Mairei and given the name taxchinin M (Tanaka et al., 1996).

Elution with 5% acetone in dichloromethane yielded (-) rhododendrol (82, Figure

2-16) which has been reported previously in Taxus brevifolia (Chu et al., 1994) and

Betula pendula (Smite et al., 1993). This was followed by 13-deacetyl taxiflorine (83,

Figure 2-16). Like taxiflorine, 83 also exhibited a 1H spectrum which contained very broad

rounded peaks, whereas the spectrum of its acetate was normal. Also this acetate was

identical to the acetate of taxiflorine. It was also discovered that it was possible to get a

better spectrum of both these compounds (taxiflorine and 13-deacetyl taxiflorine) if the

spectra were run at lower temperatures. Therefore 1H and COSY spectra of 83 were


Table 2-1: 1H and 13C NMR Values for Related Abeo-Taxanes H or C # Compd. 81 Compd. 79 Compd. 80
1 ****, 67.3 ****, 68.5 ****, 65.5
2 6.16 d (7.2Hz), 6.17 d (7.9Hz), 67.8 6.22 d (7.5Hz), 68.7
67.8
3 3.00 d (7.5Hz), 3.01 d (7.9Hz), 44.7 3.12 d (7.8Hz), 44.1
43.7
4 ****, 78.6 ****, 79.2 ****, 79.0
5 4.96 d (7.2Hz), 4.99 d (7.6Hz), 84.6 5.00 d (6.0Hz), 84.9
84.6
60s 2.67 m, 34.5 2.60 m, 34.7 2.74 m, 34.3
60 1.77 m, **** 1.91 m, **** 1.84 m, ****
7 5.54 t (7.8Hz), 70.2 5.59 t (8.2Hz), 70.6 5.16 t (7.5Hz), 71.0
8 ****, 43.5 ****, 43.5 ****, 44.5
9 6.32 d (10.8), 77.2 6.21 d (10.9Hz), 76.3 6.32 s, 83.6
10 6.43 d (10.5), 67.6 6.58 d (10.9Hz), 68.8 ****, 192.2
11 ****, 135.8 ****, 135.7 ****, 137.5
12 ****, 146.8 ****, 147.7 ****, 156.8
13 5.61 t (6.9Hz), 78.4 5.62 t (7.7Hz), 78.7 5.72 t (7.2Hz), 78.9
14cc 1.68 m, 36.6 2.50 m, 36.7 1.76 dd (8.1, 14.7Hz), 37.1






40

Table 2-1 --continued
H or C # Compd. 81 Compd.79 Compd.80
1403 2.26 m, 2.00 m, 2.40 dd (7.2, 14.4Hz),

15 75.2 ****, 75.7 76.3
16 1.17s,27.3 1.15s,27.7 1.22s,25.5
17 1.20s,24.9 1.07 s, 25.4 1.18s,27.4
18 1.87 s, 11.5 2.01 s, 11.9 2.08 s, 13.8
19 1.77 s, 13.0 1.68 s, 12.4 1.91 s, 13.6
20a 4.49 d (7.2Hz), 4.50 d (7.7Hz), 74.5 4.56 d (7.2Hz), 74.7
74.3
20f3 4.41 d (7.2Hz), 4.41 d (7.7Hz), 4.45 d (7.2Hz), ****

q-Bz **,129.3 **,129.0 **,129.4
o-Bz 7.92 d (7.2Hz), 7.86 d (7.8Hz), 129.4 8.07 d (7.2Hz), 129.8
129.5
m-Bz 7.43 t (7.8Hz), 7.43 t (7.8Hz), 128.6 7.45 t (8.1Hz), 128.3
128.2
p-Bz 7.56 t (7.5Hz), 7.55 t (7.8Hz), 133.3 7.58 t (7.2Hz), 133.2
133.1
H or C # Compd. 81 Compd. 79 Compd. 80
OCOCH3 1.65 s, 1.82 s, 2.03 1.74 s, 2.02 s, 2.02 s, 2.01 s, 2.05 s, 2.11 s, 2.13
s, 2.13 s, 2.14 s, 2.08 s, 2.12 s, 20.6, s, 20.9, 21.5, 21.6, 21.8
20.5, 21.0, 21.6, 21.3, 21.4, 21.6, 22.0
21.7,21.9
C=O 166.2, 167.8, 163.9, 168.9, 169.0, 166.8, 169.0, 169.6, 170.3,
168.9, 169.7, 169.7, 170.2, 170.4 170.4
170.1, 170.3


obtained at -10' C and the structure was determined to be that which is shown and which

was previously isolated by another group from Taxus chinensis var. Maiei and given the

name taxchinin L (Tanaka et al., 1996). It has been reported in the literature that

abeotaxanes which contain a C-9 benzoate and a C-10 hydroxyl group usually give proton

spectra in which the peaks are broad and rounded (Rao & Juchum, 1998). This is because






41

HO OBz OAc

HO H
HO \ HO-O
H
82 HO OAc OAc

OH 83
HO
OCH3 O

HO H

H OH
HO OCH3



HO
H 84 85
O

O


H AcO OAc
CH30 OA87
HOO








Figure 2-16: Compounds from the Needles of Taxusfioridana


these compounds seem to exist in an equilibrium between two conformers. If the spectrum is taken at low temperature however, two sets of signals can be distinguished.
H0 8V6 OAc OAc
OCH3 87
OH



Figure 2-16: Compounds from the Needles of Taxus floridana


these compounds seem to exist in an equilibrium between two conformers. If the spectrum is taken at low temperature however, two sets of signals can be distinguished.






42

On elution with 10% acetone/dichloromethane additional amounts of 10-deacetyl baccatin III were obtained; and with 2% methanol/10% acetone/dichloromethane the polyhydroxylated steroid ponasterone A (84) was isolated (Figure 2-16). This compound was previously isolated from Taxus brevifolia (Rao et al., 1996b). Finally, with 5-10% methanol/10% acetone/dichloromethane 10-deacetyl paclitaxel-7-3-xyloside was isolated for the first time from Taxusfloridana.

The fractions mentioned earlier that were eluted with dichloromethane were concentrated and put on another silica column using 25% ethyl acetate/ligroin as the starting mobile phase. At 30% ethyl acetate/ligroin trans-2,6-dimethoxy cinnamaldehyde

(85) was eluted; and it will be discussed later. Elution continued with 50% ethyl acetate/ligroin, which gave the lignan c-conidendrin (86) followed by 1-deoxy baccatin IV

(87); both of which have been previously isolated (Figure 2-16) (Miller et al., 1982; Miller, 1980).


Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde


As mentioned above trans-2, 6-dimethoxy cinnamaldehyde was one of the compounds isolated from Taxusfloridana. Although this structure was determined quite easily based on 1H and 3C NMR spectra, di-ortho oxy substituted C6-C3 compounds had previously not been isolated from natural products. Thus 85 was synthesized to verify the structure following Figure 2-17. Thus trans-2, 6-dimethoxy cinnamic acid (88) was methylated with methanol and H2SO4 followed by reduction to the alcohol (89) with LAH in a total yield of about 55%. The alcohol was then oxidized to the aldehyde (90) with






43

CH3 CH3

1.) EtOH, H+ OH 2O
2.) LAH
88 OCH3 89 OCH3


I Jone's oxidation


CH3 O CH3 O

Malonic acid
OH .
Pyridine/piperidine \ i
91 OCH3 90 OCH3

1.) Dimethylsulfate
2.) LAH

CH30 CH3

/OH ~ PDC oxidation
OH H

92 OCH3 93 OCH3



Figure 2-17: Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde


Jones reagent in a yield of 79% and the aldehyde was then condensed with malonic acid to yield the corresponding cinnamic acid (91) in about 85% yield. This acid was again methylated with dimethylsulfate and reduced with LAH to the alcohol (92) to give a total yield of 60%. Finally 92 was oxidized to the desired aldehyde (93) using the mild oxidizing agent PDC, to prevent further oxidation, in a yield of 38% (Figure 2-17). This aldehyde






44

was identical to the natural product in all ways. A thorough search of the literature confirmed that this was a novel compound and that no other 2, 6-dioxy cinnamyl compound has been found in nature.


Experimental


All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of the reaction, All NMR spectra were recorded using either a Varian VXR-300 or a Varian Gemini-300 spectrophotometer using CDC13 as solvent. Infrared spectra were obtained using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolet spectra were obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a Fisher melting point apparatus. Column chromatography was accomplished using 100-200 mesh silica gel.

Isolation of Minor Compounds from Taxus brevifolia

The filtrates from the region between 10-deacetyl baccatin III and 10-deacetyl paclitaxel on the reverse-phase chromatographic separation were concentrated to a syrup (400 g). A 5 g aliquot was applied to a normal-phase silica column (100 g) in dichloromethane, and chromatographed with an elution sequence consisting of 1-5% acetone and then 5% acetone and 1-5% methanol. A total of 200 ml of each solvent mixture was used before progressing to the next solvent system and fractions of about 20 ml were collected and monitored by TLC. The order of elution was as follows: 1f3-






45

hydroxy baccatin I, baccatin VI, 13-hydroxy-7-deacetyl baccatin I, and 9-dihydro-13acetyl baccatin III.

10-Hydroxy baccatin I (35)

The compound was eluted with 2% acetone in dichloromethane to give 281 mg of 35 crystallized from diethyl ether and ligroin. White crystalline powder, mp 259-2610 C, 1H NMR 5: 1.24 (s, 3H, 17-H), 1.25 (s, 3H, 19-H), 1.65 (s, 3H, 16-H), 1.80 (m, 1H, 6H3), 1.90 (m, 1H, 14-H3), 2.00 (s, 3H, OAc), 2.05 (s, 3H, OAc), 2.06 (s, 3H, OAc), 2.06 (s, 3H, OAc), 2.09 (s, 3H, OAc), 2.12 (s, 3H, OAc), 2.18 (m, 1H, 6-Ha), 2.22 (s, 3H, 18H), 2.32 (d, 4.8Hz, 1H, 20-HP3), 2.54 (dd, 9.9, 15.3Hz, 1H, 14-HU), 3.19 (d, 3.6Hz, 1H, 3-H), 3.56 (d, 5.4Hz, 1H, 20-H), 4.22 (br s, 1H, 5-H), 5.49 (m, 2H, 2-H, 7-H), 6.05 (d, 11.1Hz, 1H, 9-H), 6.09 (t, 7.8Hz, 1H, 13-H), 6.22 (d, 11.4Hz, 1H, 10-H). 13C NMR 6: 13.6, 15.4, 20.6, 20.8, 20.9, 21.3, 21.4, 21.6, 21.8, 28.4, 31.0, 38.5, 41.3, 43.2, 46.6,

49.9, 58.3, 68.7, 70.7, 71.1, 72.1, 75.1, 76.0, 77.7, 135.6, 140.3, 169.0, 169.2, 169.3,

169.7, 169.8, 170.1.

Baccatin VI (36)

The compound was eluted with 4% acetone in dichloromethane to give 362 mg of 36 crystallized from diethyl ether and ligroin. White crystalline powder, mp 245-2470 C, 1H NMR 6: 1.23 (s, 3H, 17-H), 1.61 (s, 3H, 19-H), 1.79 (s, 3H, 16-H), 1.87 (m, 1H, 6H3), 2.00 (s, 3H, OAc), 2.03 (s, 3H, 18-H), 2.10 (s, 3H, OAc), 2.11 (s, 3H, OAc), 2.17 (m, 2H, 14-Hxa,3), 2.20 (s, 3H, OAc), 2.29 (s, 3H, OAc), 2.51 (m, 1H, 6-Ha), 3.18 (d, 5.7Hz, 1H, 3-H), 4.13 (d, 8.4Hz, 1H, 20-H3), 4.33 (d, 8.4Hz, 1H, 20-Hu), 4.97 (d, 9.0Hz, 1H, 5-H), 5.87 (d, 6.0Hz, 1H, 2-H), 6.00 (d, 11.4Hz, 1H, 9-H), 6.17 (t, 8.1Hz,







46

1H, 13-H), 6.22 (d, 11.1Hz, 1H, 10-H), 7.48 (t, 7.8Hz, 2H, m-Bz), 7.61 (t, 7.2Hz, 1H, pBz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C NMR 6: 12.8, 14.9, 20.7, 20.9, 21.2, 21.4, 22.3,

22.7, 28.3, 34.5, 35.1, 42.8, 45.8, 47.3, 69.7, 70.4, 71.8, 73.3, 75.0, 76.4, 78.9, 81.5,

83.8, 128.6, 129.3, 130.1, 133.7, 135.6, 141.2, 166.9, 168.9, 169.1, 169.8, 170.1, 170.4.

10j3-Hydroxy-7-deacetyl baccatin I (37)

The compound was eluted with 4% acetone in dichloromethane to give 132 mg of 37 crystallized from diethyl ether and ligroin. White crystalline powder, mp 234-236' C, FABMS m/z: 611 (M+ 1), IH NMR 5: 1.18 (s, 3H, 16-H), 1.24 (s, 3H, 19-H), 1.66 (s, 3H, 17-H), 1.85 (m, 2H, 14-HL,3), 1.92 (m, 2H, 6-Ha,3), 2.04 (s, 3H, OAc), 2.05 (s, 3H, OAc), 2.11 (s, 3H, OAc), 2.13 (s, 3H, OAc), 2.14 (d, 1.2Hz, 3H, 18-H), 2.19 (s, 3H, OAc), 2.32 (d, 5.1Hz, 1H, 20-HO), 2.53 (dd, 9.6, 15.0Hz, 1H, 14-Ha), 3.08 (d, 3.6Hz, 1H, 3-H), 3.52 (d, 5.1Hz, 1H, 20-Ha), 4.21 (t, 3.0Hz, 1H, 5-H), 4.27 (dd, 4.8, 10.8, 1H, 7-H), 5.45 (d, 3.6Hz, 1H, 2-H), 6.07 (t, 7.2Hz, 1H, 13-H), 6.14 (d, 11.1Hz, 1H, 9-H), 6.20 (d, 10.8Hz, 1H, 10-H). 13C NMR 8: 12.5, 15.5, 20.5, 20.8, 20.9, 21.3, 21.6, 21.8,

28.5, 32.3, 38.4, 40.4, 43.3, 47.1, 49.9, 59.2, 69.1, 70.4, 71.8, 72.5, 76.1, 78.0, 78.1,

135.7, 140.7, 168.3, 169.1, 169.5, 169.6, 170.0. 9-Dihydro-13-acetyl baccatin III (38)

The compound was eluted with 5% acetone and 2% methanol in dichloromethane to give 184 mg of 38 crystallized from diethyl ether and ligroin. White crystalline powder, mp 243-2440 C, FABMS m/z: 631 (M + 1), 'H NMR 8: 1.25 (s, 3H, 16-H), 1.68 (s, 3H, 17-H), 1.82 (s, 3H, 19-H), 1.93 (d, 1.2Hz, 3H, 18-H), 1.96 (m, 1H, 6-HP), 2.14 (s, 3H, 10-OAc), 2.19 (s, 3H, 13-OAc), 2.21 (m, 2H, 14-Ha,13), 2.28 (s, 3H, 4-OAc), 2.53 (m,






47

1H, 6-Hu), 3.05 (d, 6.0Hz, 1H, 3-H), 4.16 (d, 8.1Hz, 1H, 20-H3), 4.31 (d, 8.1Hz, 1H, 20-Ho), 4.43 (m, 2H, 7-H, 9-H), 4.95 (d, 8.4Hz, 1H, 5-H), 5.75 (d, 5.7Hz, 1H, 2-H), 6.17 (t, 6.9Hz, 1H, 13-H), 6.19 (d, 10.8, 1H, 10-H), 7.48 (t, 7.8Hz, 2H, m-Bz), 7.61 (t, 7.5Hz, 1H, p-Bz), 8.09 (d, 7.2Hz, 2H, o-Bz). 13C NMR 8: 12.5, 14.8, 21.2, 21.3, 22.6,

22.8, 28.3, 35.5, 38.0, 43.1, 45.0, 47.2, 69.8, 73.3, 73.7, 74.0, 76.9, 77.3, 78.8, 82.2,

84.1, 128.6, 129.4, 130.1, 133.6, 135.0, 139.7, 167.0, 169.3, 170.4.

Acetylation of 103-Hydroxy-7-Deacetyl Baccatin I

A total of 30 mg of 1-hydroxy-7-deacetyl baccatin I (37) was dissolved in 1 ml of acetic anhydride and 1 ml of pyridine. This mixture was stirred at room temperature for 18 hours and then water was added to the mixture. Sodium bicarbonate was added slowly until no further evolution of CO2 was observed. The aqueous mixture was then extracted twice with dichloromethane and the combined organic layers were washed with 0.1 N NaOH, 0.1 N HC1, and water successively and dried with sodium sulfate. The dichloromethane was evaporated and the product was crystallized from diethyl ether and ligroin to yield 22 mg of acetylated 1 f-hydroxy-7-deacetyl baccatin I which was identical in every way to 1 -hydroxy baccatin I (35).

Isolation of Taxamairin A (38) from Taxus brevifolia

A total of 27 g of semi-pure paclitaxel, which had crystallized from reverse-phase fractions, was dissolved in 200 ml of dichloromethane and applied to a silica column with 225 g of 240 mesh silica gel. Solvent was pumped through with an Eldex Laboratories metering pump model B-100-S-4 at a pressure not exceeding 25 psi. The beginning solvent was 2 : 1 dichloromethane : ligroin, then 3 : 1 dichloromethane : ligroin, followed






48

by dichloromethane. At this point the column was eluted with 2-5% acetone in dichloromethane and then 2-5% methanol and 5% acetone in dichloromethane. A total of 500 ml of each solvent mixture was pumped through before switching to the next solvent. Fractions of about 100 ml were collected and monitored by TLC. Taxamairin A was eluted with 2% acetone in dichloromethane and crystallized from the elution solvent. It was recrystallized from dichloromethane to yield 275 mg. Yellow needles, mp 252-253' C, EIMS mlz: 338 (80%, M), 310 (74%), 295 (100%), 267 (63%), 237 (18%), 156 (24%). CIMS: 339 (M + 1). UV Xmax: 211, 255, 385nm IR (KBr): 1672, 1535, 1320, 1195, 1052 cm"1. 1H NMR : 1.33 (d, 6.9Hz, 6H, 19-H, 20-H), 1.46 (s, 6H, 12-H, 13-H), 3.35 (heptet, 6.9Hz, 1H, 18-H), 3.89 (s, 3H, 15-OMe), 6.11 (d, 9.6Hz, 1H, 7-H), 6.65 (s, 1H, 14-OH), 6.95 (s, 1H, 11-H), 7.31 (d, 9.9Hz, 1H, 11-H), 7.77 (s, 1H, 17-H), 7.94 (s, 1H, 4-H). 13C NMR 5: 23.3, 23.4, 26.7, 26.8, 27.4, 50.5, 62.0, 119.3, 120.8, 123.7, 130.1,

131.1, 133.7, 136.6, 146.1, 146.8, 147.8, 148.2, 151.4, 188.2, 200.9.

Methylation of Taxamairin A

Taxamairin A (50 mg) was dissolved in 3 ml acetone and excess K2C03 was added together with 0.25 ml of dimethyl sulfate. This mixture was refluxed for 3 hours and at that point 0.5 ml of concentrated NH4OH was added to the mixture and stirred for 15 minutes. The acetone was partially evaporated and water was added. This aqueous solution was then extracted twice with dichloromethane and the combined organic layers were washed with 0.1 N NaOH and then with water. After drying with sodium sulfate, the solvent was removed and the residue was crystallized from dichloromethane to yield 32 mg of taxamairin B (39). Yellowish white needles, mp >290' C, UV 4,x: 219, 281, 355






49

nm. IR (KBr): 1670, 1621, 1333, 1305, 1038 cm'. 1H NMR 5: 1.30 (d, 6.9Hz, 6H, 19H, 20-H), 1.46 (s, 6H, 12-H, 13-H), 3.41 (heptet, 6.9Hz, 1H, 18-H), 3.98 (s, 6H, 14OMe, 15-OMe), 6.12 (d, 9.6Hz, 1H, 7-H), 6.94 (s, 1H, 11-H), 7.31 (d, 1H, 11-H), 7.87 (s, 1H, 17-H), 7.93 (s, 1H, 4-H). 13C NMR 6: 22.9, 23.0, 26.7, 26.8, 27.8, 50.4, 60.7,

61.2, 123.0., 124.1, 127.3, 128.3, 130.9, 131.5, 133.8, 136.1, 147.8, 150.8, 151.3, 153.9,

188.0, 200.7.

Synthesis of Taxamairin B (39)

2, 2-Dimethyl-1, 3-cyclohexanedione (64)

A total of 10 g of 1, 3-cyclohexanedione (63) and 30.6 g (2.5 eq.) of K2CO3 was added to 40 ml of acetone to which 31.5 g (2.5 eq.) of CH3I had been added. The mixture was refluxed overnight. After cooling the mixture the K2CO3 was filtered and the acetone was removed under vacuum. The residual material was partitioned between water and diethyl ether and the water layer was discarded. The solvent was evaporated to yield a syrup which was poured into a mixture of 20 ml of conc. HC1 and 20 g of ice. This was stirred for 30 minutes to decompose the methyl enol ether which accounts for about 30% of the mixture, then water and diethyl ether were added and partitioned. The organic layer was washed twice with water, then dried with sodium sulfate. After removal of the solvent, the crude liquid product was distilled using a water aspirator vacuum (-15 mm Hg) and the product distilled at 120-1220 C. Upon standing the product crystallized yielding 6.2 g. Colorless crystals, mp 31-320 C, 'H NMR 6: 1.29 (s, 6H, CH3), 1.93 (in, 6.5Hz, 2H, 5-H), 2.67 (t, 6.9Hz, 4H, 4-H, 6-H). 13C NMR 5: 18.1, 22.3, 37.4, 61.8,

210.6.






50

4-Bromo-2, 2-dimethyl-1, 3-cyclohexanedione (65)

A total of 2.0 g of dimethylated diketone 64 was dissolved in 5 ml of CH2C12 and a separate bromine mixture was prepared by adding excess bromine to CH2C12 in a ratio of about 4 drops bromine to 1 ml of CH2C12. The bromine solution was dropwise added with stirring at room temperature to the diketone solution and the TLC was monitored using 12 crystals as an indicator. The reaction was stopped when only a small amount of starting material was observed and the major spot on TLC had a slightly higher Rf value in 30% ethyl acetate in ligroin. Water was added to the reaction mixture and partitioned. The water layer was discarded and the organic layer was washed twice with water, dried with sodium sulfate, and the solvent was removed by evaporation. The residue was then put on a silica column using 20-30% ethyl acetate as the solvent to give 1.14 g of the product as a yellow oil. This material was stored at -50 C and upon storage the product crystallized. Colorless crystals, mp 48-50' C, 'H NMR 8: 1.34 (s, 3H, CH3), 1.45 (s, 3H, CH3), 2.272.60 (m, 2H, 6-H), 2.72-2.97 (m, 2H, 5-H), 4.73 (dd, 4.2, 6.9Hz, 1H, 4-H). 13C NMR 6:

23.1, 24.1, 26.5, 34.3, 48.8, 59.5, 203.2, 208.2. 2, 2-Dimethyl-4-cyclohexene-1, 3-dione (66)

A total of 1.0 g of bromo compound 65 was dissolved in 5 ml of DMF and 1.0 g of LiC1 was also added. This mixture was refluxed for 2 hours at which time the TLC showed the presence of a slower moving product and only a small amount of starting material. Water and diethyl ether were added to the mixture and partitioned. The water layer was partitioned twice more with diethyl ether and all the organic layers were combined and washed with water twice, dried with sodium sulfate, and the solvent was






51

evaporationed. The residue was ran through a silica column using 15-30% ethyl acetate in ligroin as the eluent to give 823 mg of product as a yellow oil. Yellow oil, 'H NMR 6: 1.35 (s, 6H, CH3), 3.35 (dd, 2.4, 3.9Hz, 2H, 6-H), 6.25 (dt, 2.1, 10.2Hz, 1H, 4-H), 7.03 (dt, 3.9, 10.2Hz, 1H, 5-H).

Isopropylation of acetovanillone

A total of 3.0 g of acetovanillone (67) was added to 20 ml of 90% H2SO4 and 2 ml of isopropyl alcohol was added. This mixture was stirred at 600 C for 36 hours. At this point about 50% conversion to the product was observed on TLC. Longer reaction times did not increase conversion and higher temperatures caused decomposition. Thus the mixture was diluted with ice water and diethyl ether and partitioned. The water layer was partitioned once again with ether and the ether layers combined. The organic layer was then partitioned twice with 1.0 N NaOH and the organic layer was discarded. The aqueous layer was acidified with concentrated HCI and extracted twice with diethyl ether. This organic layer was then washed with water twice, dried with sodium sulfate, and concentrated. The residue was chromatographed on a silica column using 20-40% ethyl acetate in ligroin as the eluent to give 1.42 g of product which crystallized on standing. Clear colorless crystals, mp 116-117o C, EIMS m/z: 208 (42%, M), 193 (100%), 163 (17%), 77(12%), 1H NMR 6: 1.27 (d, 6.9Hz, 6H, CH3), 2.57 (s, 3H, CH3), 3.33 (quintet, 6.9Hz, 1H, CH), 3.95 (s, 3H, OCH3), 6.23 (br s, 1H, Ar-OH), 7.39 (d, 1.5Hz, 1H, Ar-H), 7.50 (d, 1.5Hz, 1H, Ar-H). 13C NMR 6: 22.2, 26.2, 27.2, 56.2, 107.4, 121.0, 129.4,

133.7, 146.2, 147.6, 197.1.






52

Methylation of isopropyl acetovanillone

A total of 1.0 g of isopropylated acetovanillone was dissolved in 20 ml of acetone then 3.0 g of K2CO3 and 1 ml of dimethyl sulfate were added. The mixture was refluxed for 3 hours at which time no starting material remained according to TLC. Then 1 ml of conc. NH4OH was added and the mixture was stirred for 30 minutes. The acetone was then partially removed and the residue was partitioned between water and diethyl ether. The organic layer was washed twice with water, dried with sodium sulfate, and concentrated to yield 956 mg of product (68) as a clear colorless oil. Reduction of 68

A total of 1.0 g of 68 was dissolved in 8 ml of methanol with 3 drops of 1.0 N NaOH. To this was added dropwise a solution of NaBH4 in methanol with 1.0 N NaOH and the TLC was monitored. When no starting material remained the mixture was acidified with 0.1 N HC1 and the methanol was partially removed. The residue was partitioned between water and diethyl ether and the organic layer was washed twice with water, dried with sodium sulfate, and concentrated under vacuum. This material was chromatographed on a silica column using 25-50% ethyl acetate in ligroin as eluent to give 856 mg of alcoholic product (69) as a clear colorless oil. Clear colorless oil, EIMS m/z: 224 (97%, M), 209 (100%), 179 (18%), 139 (90%), 124 (33%). 1H NMR 6: 1.21 (d, 6.9Hz, 6H, CH3), 1.48 (d, 6.3Hz, 3H, CH3), 3.34 (heptet, 6.6Hz, IH, CH), 3.80 (s, 3H, OCH3), 3.86 (s, 3H, OCH3), 4.83 (q, 6.3Hz, 1H, CH), 6.81 (s, 2H, Ar-H). 13C NMR 6: 23.4, 25.0, 26.8, 55.6, 60.8, 70.5, 106.6, 115.1, 141.6, 142.2, 145.3, 152.5.






53

Hydroxymethylation of 69

A total of 200 mg of 69 was dissolved in 3 ml of dry diethyl ether, 205 mg (2 eq.) of TMEDA was added, and the mixture was cooled to -78' C under a helium atmosphere. After cooling the mixture, 0.4 ml of 2.5 M n-butyllithium in hexanes (2.2 eq.) was added with stirring. After stirring for 30-45 minutes excess paraformaldehyde was added and the mixture was stirred overnight while warming to room temperature. The mixture was diluted with water and diethyl ether and partitioned. The organic layer was washed twice with water, dried with sodium sulfate, and concentrated. According to TLC about 50% of the product remained. This material was separated on a silica column using 30-50% ethyl acetate in ligroin as eluent and a total of 116 mg of product was isolated as a slightly yellow oil. Yellow oil, 1H NMR 6: 1.22 (d, 6.9Hz, 6H, CH3), 1.56 (d, 6.3Hz, 3H, CH3), 3.32 (quintet, 6.9Hz, 1H, CH), 3.84 (s, 3H, OCH3), 3.88 (s, 3H, OCH3), 4.68 (d, 11.7Hz, IH, CH20), 4.86 (d, 11.7Hz, 1H, CH20), 5.12 (q, 6.6, 12.9Hz, 1H, CH), 7.10 (s, 1H, ArH). 13C NMR 5 23.1, 23.4, 27.0, 33.5, 55.9, 60.6, 61.2, 67.2, 118.7, 129.9, 139.5,

142.6, 149.6, 151.5.

Oxidation of diol to keto-aldehyde 70

A total of 600 mg of Cr03 (1 eq.) was added to a solution of 950 mg of pyridine (2 eq.) in 15 ml of CH2C12 and this was stirred at room temperature for 15 minutes. At this point the PDC solution was added dropwise to a solution of 100 mg of the diol in 3 ml of CH2C12. The TLC of the reaction mixture was always checked about 5 minutes after adding 3-4 drops of the PDC solution using 2, 4-dinitrophenylhydrazine as the indicator. More PDC was added until the reaction was complete according to TLC. At this point the






54

reaction was filtered and the filtrate was washed twice with 0.1 N HCI, twice with 0.1 N NaOH, and twice with water. The organic layer was then dried with Na2SO4 and evaporated under a vacuum to a solid residue. This material was separated on a regular silica column using 20-40% ethyl acetate in ligroin. A total of 62 mg of clear colorless oil was obtained. Clear colorless oil, 'H NMR 8: 1.24 (d, 6.9Hz, 6H, CH3), 2.49 (s, 3H, CH3), 3.37 (m, 1H, CH), 3.91 (s, 3H, OCH3), 3.95 (s, 3H, OCH3), 7.08 (s, 1H, Ar-H), 10.33 (s, 1H, CHO).

Condensation of keto-aldehyde 70 with dione 66

A total of 100 mg of keto-aldehyde 70 was added to 3 ml of pyridine containing

-10 drops of piperidine. To this mixture was added 80 mg of dione 66 and the mixture was refluxed with stirring for 6 hours. At this time TLC analysis showed a product with the same Rf value as the methylated taxamairin A (taxamairin B) along with other products. The mixture was diluted with diethyl ether and washed three times with 0.IN HCl until the water layer was still acidic. The organic layer was then washed twice with water, dried with Na2SO4, and evaporated under a vacuum to a solid residue. The product was isolated using a regular silica column with 20-40% ethyl acetate in ligroin and a total of 46 mg of product was isolated as a yellowish white amorphous solid. All spectral data was identical to that of the methylated taxamairin A. Complete reduction of methyl isopropyl acetovanillone (72)

A total of 500 mg of methylated isopropyl acetovanillone was dissolved in 5 ml of THF and 1.0 g of ZnC12 was added and the mixture was stirred at 60 C. To this was added NaCNBH3 in small portions and the TLC was monitored. When nearly all the






55

starting material had been converted to a faster moving product the mixture was diluted with water and diethyl ether and partitioned. The organic layer was washed with 0. IN HC1 and twice with water, dried over sodium sulfate, and concentrated. The product was purified by silica chromatography using 15-20% ethyl acetate in ligroin as solvent. A total of 387 mg of product (73) was obtained as a slightly yellow oil. Yellow oil, 'H NMR 6: 1.21 (d, 6.9Hz, 6H, CH3), 1.24 (t, 7.5Hz, 3H, CH3), 2.60 (q, 7.8, 15.3Hz, 3H, CH3), 3.33 (quintet, 6.9Hz, 1H, CH), 3.79 (s, 3H, OCH3), 3.85 (s, 3H, OCH3), 6.60 (d, 1.5Hz, 1H, Ar-H), 6.67 (d, 1.5Hz, iH, Ar-H). 13C NMR 6: 15.6, 23.6, 26.7, 28.9, 55.6, 60.9, 109.3,

117.3, 140.0, 142.0, 144.1, 152.3.

Acetoxymethylation of 73

A total of 100 mg of 73 was added to 2.0 ml of 85% H3PO4 and to this was added 200 mg of paraformaldehyde and 0.5 ml of acetic anhydride. The mixture was stirred at room temperature for 3 hours at which point the TLC showed most of the starting material to be gone and a slower moving product had formed. The mixture was partitioned between water and diethyl ether and the organic layer was washed twice with 0.1 N HC1 and twice with water, dried with sodium sulfate, and concentrated. The residue was separated on a silica column using 15-25% ethyl acetate in ligroin as the solvent. A total of 64 mg of the product (74) was isolated as a clear colorless oil. Clear colorless oil, 'H .NMR 6: 1.21 (t, 7.5Hz, 3H, CH3), 1.35 (d, 7.2Hz, 6H, CH3), 2.07 (s, 3H, OAc), 2.68 (q, 7.5, 15.0Hz, 2H, CH2), 3.22 (m, 1H, CH), 3.84 (s, 3H, OCH3), 3.86 (s, 3H, OCH3), 5.14 (s, 2H, OCH2), 6.66 (s, 1H, Ar-H). '3C NMR 6 16.2, 21.0, 21.9, 26.8, 28.9, 55.5, 60.6, 60.7, 110.7, 122.9, 142.1, 146.6, 153.2, 171.2.






56

Bromination of 73

A total of 100 mg of 73 was dissolved in 3.0 ml of CH2C12 and a dilute bromine solution in CH2C12 was added dropwise with stirring at room temperature and the TLC was monitored. The addition was stopped when a small amount of starting material remained and a faster moving product spot was present. Water and more CH2Cl2 was added to the mixture and partitioned. The organic layer was washed twice with 0.1 N HC1 and twice with water, dried with sodium sulfate, and concentrated. The residue was separated on a silica column using 15-25% ethyl acetate in ligroin and 73 mg of product

(75) was isolated as a yellowish oil. Yellow oil, 'H NMR 8: 1.22 (t, 7.5Hz, 3H, CH3), 1.35 (d, 6.6Hz, 6H, CH3), 2.75 (q, 7.5, 15.3Hz, 2H, CH2), 3.72 (m, 1H, CH), 3.83 (s, 3H, OCH3), 3.84 (s, 3H, OCH3), 6.69 (s, 1H, Ar-H). 13C NMR 5: 14.4, 21.0, 30.9, 34.2, 55.7, 60.9, t11.0, 139.0, 140.5, 152.0.

Bromination of 69

A total of 100 mg of 69 was dissolved in 3.0 ml of CH2C2 and a dilute bromine solution in CH2Cl2 was added dropwise with stirring at room temperature and the TLC was monitored. The addition was stopped when only a small amount of starting material remained and a faster moving product spot was present. Water and more CH2Cl2 was added to the mixture and partitioned. The organic layer was washed twice with 0.1 N HC1 and twice with water, dried with sodium sulfate, and concentrated. The residue was separated on a silica column using 15-25% ethyl acetate in ligroin and 58 mg of product

(76) was isolated as a yellowish oil. Yellow oil, 1H NMR 6: 1.34 (t, 6.6Hz, 6H, CH3), 2.02 (d, 6.9Hz, 3H, CH3), 3.71 (m, 1H, CH), 3.86 (s, 3H, OCH3), 3.89 (s, 3H, OCH3),






57

5.78 (q, 6.6, 13.5Hz, 1H, CH), 7.10 (s, 1H, Ar-H). 1C NMR 5: 20.9, 26.7, 35.4, 50.0, 55.8, 60.9, 109.8, 118.0, 140.6, 142.5, 152.5. Isolation of Minor Compounds from Taxusfloridana

The mother liquors of fractions that contained 10-deacetyl baccatin III from the reverse-phase column (25-40% acetonitrile in water) were concentrated to a syrup (10 g). On standing, this syrup became an amorphous solid and 3 g of this was applied to a normal-phase silica column (40 g) using dichloromethane as the starting solvent. Then the column was eluted successively with dichloromethane containing 2, 5, and 10% acetone and then with addition of 2, 5, and 10% methanol. A total of 200 ml of each solvent mixture was passed through the column before the next solvent mixture was started and 20 ml fractions were collected and monitored by TLC. The initial dichloromethane eluent was put aside for further chromatography and the order of elution of the more polar compounds was as follows: 1-hydroxy baccatin I, taxiflorine (taxchinin M),

rhododendrol, taxchinin L, 10-deacetyl baccatin III, ponasterone A, and 10-deacetyl paclitaxel-7- 3-xyloside.

The initial dichloromethane eluent (0.250 mg) was applied to another silica column (3 g) using 25% ethyl acetate in ligroin as the initial solvent and proceeding to 50% ethyl acetate in 10% intervals. A total of 50 ml of each solvent was used before progressing to the next solvent mixture and 5 ml fractions were collected and monitored by TLC. The order of elution was trans-2,6-dimethoxy cinnamaldehyde, co-conidendrin, and 1-deoxy baccatin IV.






58

1-Hydroxy baccatin I (35)

The compound was eluted with 2% acetone in dichloromethane and a total of 189 mg of 35 was crystallized from diethyl ether and ligroin. Its physical and spectral properties are identical to that reported above. Taxiflorine (taxchinin M) (77)

The compound was eluted with 2% acetone in dichloromethane and a total of 129 mg was crystallized from diethyl ether and ligroin. No NMR was reported since the spectrum contained poorly defined peaks.

Rhododendrol (82)

This compound was eluted with 5% acetone in dichloromethane and a total 435 mg was crystallized directly from the eluting solvent. Clear colorless crystals, mp 74-760 C, 1H NMR 8: 1.23 (d, 6.2Hz, 3H, 1-H), 1.73 (m, 2H, 3-H), 2.63 (m, 2H, 4-H), 3.83 (m, 1H, 2-H), 6.74 (m, 2H, m-Bz), 7.04 (m, 2H, o-Bz). 13C NMR 8: 23.5, 31.2, 40.9, 67.7, 115.3, 129.4, 133.9, 153.8.

Taxchinin L (83)

This compound was eluted with 5% acetone in dichloromethane and a total of 162 mg was crystallized from diethyl ether and ligroin. White crystalline powder, mp 264-2660 C. 'H NMR (-100 C) 8: 1.02 (s, 3H, 16-H), 1.22 (s, 3H, 17-H), 1.43 (m, 1H, 14-HU), 1.73 (s, 3H, 19-H), 1.79 (s, 3H, 7-OAc), 1.81 (m, 1H, 6-HP), 1.96 (s, 3H, 18-H), 2.05 (s, 3H, 2-OAc), 2.13 (m, 1H, 14-HC), 2.17 (s, 3H, 4-OAc), 2.61 (m, 1H, 6-Hc), 3.15 (d, 7.4Hz, 1H, 3-H), 4.45 (m, 2H, 13-H, 20-H3), 4.53 (d, 7.4Hz, 1H, 20-Ha), 4.69 (t, 10.3Hz, 1H, 10-H), 4.94 (d, 6.2Hz, 1H, 5-H), 5.46 (dd, 5.6, 8.6Hz, 1H, 7-H), 5.95 (d,






59

7.3Hz, 1H, 2-H), 5.97 (d, 10.1Hz, 1H, 9-H), 7.45 (t, 7.5Hz, 2H, m-Bz), 7.57 (t, 6.9Hz, 1H, p-Bz), 7.99 (d, 7.5Hz, 2H, o-Bz). 13C NMR (-10o C) 5: 11.3, 13.9, 21.6, 21.8, 22.4,

25.2, 27.2, 34.3, 39.2, 42.9, 43.4, 66.3, 66.7, 68.4, 69.9, 74.9, 76.2, 77.5, 79.3, 80.7,

85.3, 128.2, 129.7, 129.9, 133.0, 137.5, 146.0, 167.7, 170.7, 170.8, 171.6.

10-Deacetyl baccatin III (25)

This compound was eluted with 10% acetone in dichloromethane and a total of 112 mg was crystallized from diethyl ether and ligroin. Spectral properties were identical to those reported in the literature (Dennis et al., 1988). Ponasterone A (84)

This compound was eluted with 10% acetone and 2% methanol in

dichloromethane and a total of 83 mg was crystallized directly from the eluting solvent. Spectral properties were identical to those reported in the literature (Miller et al., 1982). 10-Deacetyl paclitaxel-7-3-xyloside (28)

This compound was eluted with 5% methanol and 10% acetone in

dichloromethane and a total of 79 mg was obtained as an amorphous solid. Spectral properties were identical to those reported in the literature (Senilh et al., 1984). Trans-2, 6-dimethoxy cinnamaldehyde (85)

This compound was eluted with 25% ethyl acetate in ligroin and a total of 27 mg was obtained as an glassy solid. UV kmax: 313 nm. IR (KBr): 3100, 3000-2940, 2810, 2740-2700, 1660, 1605-1585, 1475, 1260, 1140, 1100-1080, 970, 840, 725 cm'. EIMS m/z: 192 (30%, M+), 161 (100%), 149 (17%), 91 (15%). 1H NMR 5: 3.90 (s, 6H, OMe), 6.58 (d, 8.4Hz, 2H, m-Ar), 7.17 (dd, 7.8, 16.0Hz, 1H, 2-H), 7.33 (t, 8.4Hz, 1H, p-






60

Ar), 7.93 (d, 16.0Hz, 1H, 3-H), 9.64 (d, 8.1Hz, 1H, 1-H). 13C NMR 8: 55.8 x 2 (OMe), 103.6 (m-Ar), 112.1 (q-Ar), 131.6 (2-C), 132.6 (p-Ar), 144.5 (3-C), 160.5 (o-Ar), 196.4

(1-C).

ct-Conidendrin (86)

This compound was eluted with 50% ethyl acetate in ligroin and a total of 38 mg was crystallized from diethyl ether and ligroin. White crystalline powder, mp 257-2590 C, EIMS m/z: 356 (100%, M+), 255 (13%), 241 (26%), 137 (14%). 'H NMR 5: 2.5 (inm, 2H), 2.7-3.1 (m, 1iH), 3.73 (s, 3H), 3.78 (s, 3H), 3.9-4.2 (m, 4H), 6.26 (s, 1H), 6.53 (d, 2.0Hz, 1H)., 6.58 (d, 8.0Hz, 1H), 6.62 (s, 1H), 6.73 (d, 8.0Hz, 1H). 13C DEPT NMR 6: 29.3 (CH2), 41.9 (CH), 47.5 (CH), 49.9 (CH), 55.9 (CH3), 60.0 (CH3), 71.9 (CH2), 110.1

(CH), 111.3 (CH), 114.6 (CH), 115.1 (CH), 121.5 (CH), 126.3 (C), 131.7 (C), 134.0 (C), 144.2 (C), :144.9 (C), 145.5 (C), 147.0 (C), 177.0 (C). 1-Deoxy baccatin IV (87)

This compound was eluted with 50% ethyl acetate in ligroin and a total of 57 mg was crystallized from diethyl ether and ligroin. White crystalline powder, mp 262-264o C, 'H NMR 6: 1.12 (s, 3H, 17-H), 1.54 (s, 3H, 19-H), 1.78 (m, 1H, 1-H), 1.79 (s, 3H, 16H), 1.88 (m, 1H, 14-H3), 1.91 (m, 1H, 6-H3), 1.97 (s, 3H, 18-H), 2.02 (s, 3H, Oac), 2.05 (s, 3H, Oac), 2.08 (s, 3H, Oac), 2.10 (s, 3H, Oac), 2.16 (s, 3H, Oac), 2.17 (s, 3H, Oac), 2.41 (m, 1H, 14-Ha), 2.50 (s, 1H, 6-Ha), 2.87 (d, 1H, 3-H), 4.19 (d, 7.5Hz, 20-H3), 4.51 (d, 8.1Hz, 1H, 20-Hu), 4.99 (d, 8.7Hz, 1H, 5-H), 5.52 (m, 2H, 2-H, 7-H), 5.91 (m, 2H, 9-H, 13-H), 6.15 (d, 11.4Hz, 1H, 10-H). 13C NMR 6: 12.7, 14.9, 20.7, 20.8, 20.9,






61

21.2, 21.4, 21.6, 22.6, 26.9, 31.3, 34.6, 37.9, 44.5, 45.6, 46.9, 69.0, 71.0, 71.1, 71.9,

75.4, 77.2, 81.0, 83.9, 133.3, 138.9, 169.2, 169.3, 169.7, 170.1, 170.2, 170.4.

Acetylation of Taxiflorine

Taxiflorine (77) 120 mg was dissolved in 2 ml of acetic anhydride and 1 ml of pyridine was added. The mixture was stirred at room temperature for 18 hours and then water was added to the mixture. Sodium bicarbonate was added slowly until no further evolution of CO2 was observed. The aqueous mixture was then extracted twice with dichloromethane and the combined organic layers were washed with 0.1 N NaOH, 0.1 N HCl, and water successively and dried with sodium sulfate. The dichloromethane was evaporated and the product (112 mg) (81) was obtained as a glassy solid. For NMR data see Table 2-1.

Oxidation of Taxiflorine

Taxiflorine (77) 50 mg was dissolved in 2 ml of acetone and a few drops of Jones reagent was added and the mixture was stirred at room temperature for 2 hours. At this time the acetone was partially evaporated and water was added. This aqueous mixture was extracted twice with dichloromethane and the combined organic layers were washed with 0.1 N NaOH and then with water and then dried with sodium sulfate. After the dichloromethane was evaporated the residue was crystallized from diethyl ether and ligroin to yield 36 mg of the ketone product. White crystalline powder. For NMR data see Table 2-1.






62

Acetylation of Taxchinin L (83)

Taxchinin L 100 mg was dissolved in 2 ml of acetic anhydride and 1 ml of pyridine was added. The mixture was stirred at room temperature for 18 hours and then water was added to the mixture. Sodium bicarbonate was added slowly until no further evolution of CO2 was observed. The aqueous mixture was then extracted twice with dichloromethane and the combined organic layers were washed with 0.1 N NaOH, 0.1 N HCl, and water successively and dried with sodium sulfate. The dichloromethane was evaporated and the product (84 mg) was obtained as a glassy solid. The physical and spectral properties of this acetylated product (81) was identical with the acetate of taxiflorine in every way. Synthesis of Trans-2, 6-Dimethoxycinnamaldehyde (85) Methyl 2, 6-dimethoxybenzoate

A total of 5.0 g of 2, 6-dimethoxybenzoic acid (88) was dissolved in 30 ml of methanol and 1.0 ml of concentrated H2SO4 was added. This mixture was refluxed for 48 hours at which time most of the methanol was removed under reduced pressure. The residue was partitioned between water and diethyl ether and the organic layer was partitioned with 0.1 N NaOH to remove the remaining starting material, this was performed three times. Finally the organic layer was washed with water twice and dried with sodium sulfate. Upon removal of the solvent under reduced pressure the methyl ester product crystallized to yield 3.1 g of product.

2, 6-Dimethoxybenzyl alcohol (89)

A total of 3.1 g of the methyl ester was dissolved in 15 ml of THF and this was cooled to 00 C. A total of 581 mg (1 eq.) of LAH was then carefully added to the mixture






63

with stirring. This mixture was then refluxed for 4 hours at which time no starting material remained. Thus about 10 ml of acetone was added to neutralize the remaining LAH and this was stirred for 18 hours. The solvent was then removed by rotovap and the residue was partitioned between water and diethyl ether. The organic layer was washed three times with water and was dried with sodium sulfate. Upon removal of the solvent by evaporation the product crystallized to yield 1.5 g of the product alcohol. Clear colorless crystals, 1H NMR 6: 3.83 (s, 6H, OCH3), 4.70 (s, 2H, CH2), 6.46 (d, 8.4Hz, 2H, m-Ar),

7.13 (t, 8.7Hz, 1H, p-Ar).

2, 6-Dimethoxybenzaldehyde (90)

A total of 1.5 g of the alcohol (89) was dissolved in 10 ml of acetone and Jones reagent was added dropwise while the TLC was monitored using 2, 4dinitrophenylhydrazine as the indicator. Quite suprisingly the aldehyde had a lower Rf value than the alcohol. The reaction was continued until no alcohol starting material was observed on TLC. The acetone was partially removed and the residue was partitioned between water and diethyl ether. The organic layer was washed with water twice and dried with sodium sulfate. Upon removal of the solvent under reduced pressure the product crystallized to yield 1.18 g of the aldehyde product. Clear colorless crystals, 'H NMR 6: 3.85 (s, 6H, OCH3), 6.42 (d, 8.4Hz, 2H, m-Ar), 7.33 (t, 8.6Hz, 1H, p-Ar), 10.35 (s, 1H, CHO).

Trans-2, 6-dimethoxycinnamic acid (91)

A total of 1.18 g of the aldehyde (90) and 1.38 g (2 eq.) of malonic acid was dissolved in 4 ml of pyridine and a few drops of piperidine were added. This mixture was






64

refluxed overnight. The following day the TLC seemed unchanged but 2, 4-dinitrophenyl hydrazine did not give a positive test indicating that no aldehyde remained and the cinnamic acid product had an identical Rf value. The mixture was partitioned between 0. 1 N HCl and diethyl ether and the organic layer was washed twice more with 0. 1 N HCl and then three times with water and dried with sodium sulfate. Upon evaporation of the solvent by evaporation 1.25 g of the cinnamic acid product crystallized. Methyl tranrs-2, 6-dimethoxycinnamate

A total of 1.25 g of the cinnamic acid product (91) was dissolved in 15 Ml. of acetone and 3.0 g of K2C03 was added with 0.5 ml of dimethyl sulfate. The mixture was refluxed for 3 hours at which time no starting material was observed by TLC and thus 2.0 ml of ammonium hydroxide was added to decompose any remaining dimethyl sulfate. After 15 minutes of stirring, the solvent was partially removed under vacuum and the residue was partitioned between water and diethyl ether. The organic layer was washed three times with water and then dried with sodium sulfate. Upon removal of the solvent

1.23 g of the methyl ester product crystallized. Trans-2, 6-dimethoxycinnamyl alcohol (92)

A total of 1.23 g of the methyl ester was dissolved in 15 ml of TI-I and this was cooled to 0' C. A total of 223 mg (1 eq.) of LAH was then carefully added to the mixture with stirring. This mixture was then refluxed for 4 hours at which time no starting material remained so about 10 ml of acetone was added to neutralize the remaining LAH and this was stirred for 18 hours. The solvent was then removed by evaporation and the residue was partitioned between water and diethyl ether. The organic layer was washed three






65

times with water and was dried with sodium sulfate. Since minor impurities were also present in the mixture after workup a silica column was ran using 15-30% ethyl acetate in ligroin as the solvent. A total of 700 mg of the product alcohol was obtained as a clear yellowish liquid. 1H NMR 5: 3.84 (s, 6H, OCH3), 4.32 (d, 5.1Hz, 2H, 1-H), 6.56 (d, 8.4Hz, 2H, m-Ar), 6.75-6.92 (m, 2H, 2-H, 3-H), 7.15 (t, 8.1Hz, 1H, p-Ar).13C NMR 6:

55.6, 65.4, 103.8, 113.9, 121.7, 128.2, 132.6, 158.4. Trans-2, 6-dimethoxycinnamaldehyde (93)

A total of 4.0 g of Cr03 was dissolved in 6.5 ml of pyridine at 0' C and stirred until a reddish orange solid formed (PDC). At that point 561 mg of 92 was added in 1.0 ml of acetone and the reaction was stirred for 4 hours at room temperature. At this point the TLC showed only a small amount of starting material remaining so the reaction was stopped. All physical and chemical properties matched that of the natural product.














CHAPTER 3
PREPARATION OF NITRATE ESTERS OF PACLITAXEL AND RELATED TAXANES


Complete Nitration of Paclitaxel and Related Taxanes


In an attempt to nitrate the phenyl ring of the N-benzoyl phenylisoserine side chain in paclitaxel (Figure 3-1, 94) in order to determine its effect on potency, the compound was subjected to reaction with a 1 : 5 mixture of concentrated nitric acid in acetic anhydride and an equal volume of dichloromethane. The reaction proceeded readily at room temperature in 30 minutes to give a single product which exhibited a much higher Rf on TLC than paclitaxel and even higher than 2', 7-diacetyl paclitaxel (Mellado et al., 1984), thus eliminating the possibility that acid-catalyzed acetylation had occurred. The 1H NMR spectrum showed a considerable downfield shift of the 2' and 7 proton signals (H-2' 4.78 ppm --> 5.69 ppm, H-7 4.40 ppm -- 5.75 ppm). Following characterization by elemental analysis and IR spectroscopy, which exhibited characteristic bands for nitrate esters at 1650, 1270, and 835 cm1, the structure was determined to be that of paclitaxel2', 7-dinitrate ester (Figure 3-1, 95). After a review of the literature it was found that the conditions used are actually standard conditions for nitrate ester synthesis (Boschan et al., 1955). It was surprising that the product showed no evidence of rearrangement of the A66






67


AcO 0 OH
NH 0
~~~1....,o


OH z O
OHH 94 OHO OAc
OBz


5 : 1 / acetic anhydride : HNO 3 CH2C2
room temp., 30 min

O
AcO O ONO2
NH O



ONO2 O HO95 Bz OAc





0
O
AcO O ONO2
5: 1 /Ac20: HNO3 lH O
CH2CI2
94- / 0....
9 o C, 15 min
OH 0
HO OAc
96 OBz


Figure 3=i Nitration of Paelitaxel, 7-OH > 2'-OH


ring or cleavage of the oxetane ring, both of which have been known to occur in the

presence of strong acids (Chen et al., 1993).






68

The reaction was also repeated with three of the natural analogues of paclitaxel,

ie., 10-deacetyl baccatin III, 10-deacetyl paclitaxel, and 10-deacetyl paclitaxel-7-3xyloside. In each case the reaction proceeded to yield the corresponding tri- (Figure 3-2,

97), tri-(Figure 3-3, 101), and penta-nitrate esters (Figure 3-4, 104), respectively. All of

these compounds crystallized readily from ethyl ether after workup without any need for

chromatography to give nearly quantitative yields. In no case was nitration at the sterically

hindered 1-hydroxyl observed. NMR chemical shift values are given in Table 3-1.


Table 3-1: 'H and 13C NMR Values for Completely Nitrated Taxanes H or C# Compd. 95 Compd. 97 Compd. 101 Compd. 104
1 **** 78.5 ****, 78.2 **** 78.5 **** 78.7
2 5.72 d (6.9Hz), 5.64 d (6.6Hz), 5.71 d (6.9Hz), 5.71 d (6.9Hz),
74.3 73.5 74.0 74.5
3 4.02 d (6.9Hz), 3.92 d (6.9Hz), 3.95 d (6.9Hz), 3.81 d (6.6Hz),
47.2 47.7 47.2 45.7
4 ****, 80.7 ****, 80.2 ****, 80.6 ****, 80.6
5 4.99 d (9.0Hz), 4.97 d (8.7Hz), 4.98 d (9.0Hz), 4.88 d (9.0Hz),
83.6 83.5 83.4 83.5
6at 2.69 m, 32.5 2.72 m, 32.5 2.71 m, 32.5 2.81 m, 35.8
603 2.04 m, **** 2.04 m, **** 2.07 m, **** 2.04 m, ****
7 5.75 dd (7.2, 5.78 dd (7.2, 5.75 dd (7.2, 4.17 m, 80.1
10.5 Hz), 79.8 10.5 Hz), 79.9 10.5 Hz), 79.8
8 **** 55.3 55.8 55.6 57.8
9 ****, 200.4 ****, 199.3 ****, 199.5 ****, 199.9
10 6.31 s, 74.5 6.38 s, 81.4 6.36s, 81.6 6.42s, 82.1
11 135.5 **** 134.2 **** 135.4 **** 135.5
12 140.9 ****, 142.2 ****, 144.2 ****, 143.1
13 6.30 t (9.9 Hz), 6.23 t (8.1 Hz), 6.29 t (9.0 Hz), 6.25 t (8.7 Hz),
72.8 78.1 72.5 72.6
14c 2.38 m, 35.4 2.48 dd (9.9, 2.39 m, 35.3 2.42 m, 35.3
15.6 Hz), 34.2
1403 2.38 m, **** 2.36 dd (2.8, 2.39 m, 2.30 m, *
15.9 Hz), ****
15 **,43.3 ****, 42.9 **,43.1 **,42.9
16 1.23 s, 26.5 1.20 s, 26.6 1.20 s, 26.4 1.19 s, 26.2
17 1.17 s, 21.5 1.15 s, 20.3 1.14 s, 21.5 1.13 s, 21.6






69

Table 3-1--continued
i or C# Compd. 95 Compd. 97 Compd. 101 Compd. 104
18 1.94 s, 14.3 2.12 s, 15.0 2.00 s, 14.8 1.97 s, 14.8
19 1.82 s, 11.0 1.82 s, 10.8 1.84 s, 11.0 1.75 s, 10.7
20a 4.35 d (8.7 Hz), 4.36 d (8.7 Hz), 4.37 d (8.4 Hz), 4.35 d (8.7 Hz),
76.2 76.0 76.2 76.5
2003 4.20 d (8.4 Hz), 4.11 d (8.4 Hz), 4.20 d (8.4 Hz), 4.19 d (8.7 Hz),

2' 5.69 d (3.0 Hz), 5.68 d (2.7 Hz), 5.65 d (3.0 Hz),
80.3 80.2 80.3
3' 6.14 dd (2.7, 9.6 6.13 dd (2.7, 9.3 6.10 dd (3.0, 9.3
Hz), 52.1 Hz), 52.1 Hz), 52.2
NH 6.97 d (9.6 Hz), **** 6.90 d (9.3 Hz), 6.82 d (9.3 Hz),

1" 4.76 d (4.5 Hz),
98.9
2" 4.98 dd (4.2, 6.3
Hz), 74.3
3"* 5.26 t (6.3 Hz),
72.9
4" 5.08 dd (5.7, 9.6
Hz), 74.1
5" ax **** 4.23 m, 59.5
5" eq 3.68 dd (5.7,
12.9 Hz), ****
4-Ac 2.50 s, 22.6 2.43 s, 22.3 2.51 s, 22.6 2.48 s, 22.7
10-Ac 2.18 s, 20.6 OBz-1 ****, 130.3 ***128.6 ***130.3 ***129.0
OBz-2,6 8.12 d (7.2 Hz), 8.06 d (6.9 Hz), 8.13 d (7.5 Hz), 8.13 d (7.2 Hz),
130.2 130.1 130.2 130.2
OBz-3,5 7.52 t (7.8 Hz), 7.51 t (7.8 Hz), 7.52, 128.8 7.40-7.55, 128.8
128.8 128.9
OBz-4 7.63 t, 133.8 7.65 t, (7.5 Hz), 7.64 t (7.2 Hz), 7.63 t (6.6 Hz), 131.9 133.9 133.8
NBz-1 131.2 ****, 131.3 **** 131.2
NBz-2,6 7.73 d (7.2 Hz), 7.73 d (7.5 Hz), 7.73 d (6.9 Hz),
127.1 127.1 127.1
NBz-3,5 7.41-7.46, 129.4 7.41-7.46, 129.4 7.40-7.55, 129.4
NBz-4 7.41-7.46, 129.0 **** 7.41-7.46, 129.1 7.40-7.55, 129.1
Ph-i ****, 133.1 ***** 133.0 **** 133.1
Ph-2,6 7.41-7.46, 126.5 **** 7.41-7.46, 126.5 7.40-7.55, 126.5
Ph-3,5 7.41-7.46, 128.8 7.41-7.46, 128.8 7.40-7.55, 128.8
Ph-4 7.49-7.54, 132.3 7.52, 132.4 7.40-7.55, 132.3






70

Table 3-3--continued
H or C# Compd. 95 Compd. 97 Compd. 101 Compd. 104
C=O 170.2, 169.5, 171.1, 166.9 170.4, 167.6, 170.3, 167.3,
167.7, 166.8, 166.7, 166.7 166.8, 166.7
166.7


Regioselective Nitrations of Paclitaxel and Related Taxanes


Based on these results it was decided to run the reaction at 00 C for only 10-15 minutes to determine if the reaction was regioselective. Using this protocol, paclitaxel gave the 7-mononitrate of paclitaxel (96) in -90% yield after crystallization from the crude reaction mixture (Figure 3-1). The position of the nitrate ester was easily determined by 1H NMR and COSY spectroscopy as the nitrate ester causes a downfield shift of 1.3-1.7 ppm on the adjacent proton. This result was interesting since the 2'-hydroxyl is much more reactive than the 7-hydroxyl in acetylation reactions (Mellado et al., 1984). With nitration however the order of reactivity was 7-OH > 2'-OH.

Similar studies were applied to 10-deacetyl baccatin III, 10-deacetyl paclitaxel, and 10-deacetyl paclitaxel-7-03-xyloside. They also displayed some regioselectivity, however because these compounds contain more than 2 reactive hydroxyl groups, silica column chromatography was needed to separate the products. When 10-deacetyl baccatin III was subjected to this protocol three partially nitrated compounds were obtained namely the 10mononitrate (98), the 10, 13-dinitrate (99), and the 7, 10-dinitrate (100), although 98 and 99 were the major compounds. From this it can be concluded that the order of reactivity is 10-OH > 13-OH > 7-OH. This also differs from the acetylation reactivities in which the 7hydroxyl is the most reactive followed by the 10-hydroxyl and the 13-hydroxyl






71

respectively (Gueritte-Voegelein et al., 1986). The reaction of 10-deacetyl paclitaxel likewise gave the 10-mononitrate (102) and the 7, 10-dinitrate (103), again showing the 10-hydroxyl to be the most reactive and the 2'-hydroxyl the least (10-OH > 7-OH > 2'OH). With acetylation the 2'-hydroxyl is the most reactive followed by the 7- and 10hydroxyls respectively (Kingston et al., 1982). Finally, 10-deacetyl paclitaxel-7-P-xyloside was tested and because of the five available hydroxyls many products were observed on TLC and only the major products were isolated. These included the 2"-mononitrate (105), the 3"-mononitrate (106), the 4"-mononitrate (107), and the 2", 3", 4", 10-tetranitrate (108) (Figure 3-2). This result indicates that the sugar hydroxyls are about equally reactive and more reactive than the 10-hydroxyl, with the 2'-hydroxyl being again the least reactive (2", 3", 4'" > 10 > 2'). Although no acetylation studies have been performed on the xylosides, this lab has shown that the 2'-hydroxyl is more reactive in standard acetylation conditions than either the sugar hydroxyls and the 10-hydroxyl is the least reactive (Figure 3-3).

At this point it should also be mentioned that the 10-deacetyl paclitaxel (112) used in this work was not isolated directly from biomass but was actually converted from 10deacetyl paclitaxel-7-3-xyloside (110). This conversion involves oxidizing the xyloside to the dialdehyde (111) and then cleaving the dialdehyde with phenylhydrazine to give the desired product and the corresponding phenylhydrazones (Figure 3-4) (Rao, 1997). In conclusion, it has been shown that nitrate esters of taxanes can be formed under mild conditions and in many cases regioselectivity is shown. In view of this work it is






72

0

NHO 0 2
OR3
OR1
OR5 OAc
HO z OAe
OBz
104 R1 = NO2, R2 = NO2, R3 = NO2, R4= NO2, R5 = NO2
105 R1 = NO2, R2 = H, R3 = H, R4 = H, R5 = H 106 R1 =H,R2=NO2,R3=H,R4=H, R5=H 107 R1=H,R2=H,R3=NO2,R4=H, Rs5=H
108 R1 = NO2, R2 = NO2, R3 = NO2, R4 = NO2, R5 = H

Figure 3-2: Nitration of 10-Deacetyl Paclitaxel-7- P-Xyloside 1"-OH, 2"-OH, 3"-OH > 10-OH > 2'-OH



10-Deacetyl Paclitaxel-7- P-Xyloside

acetic anhydride/pyridine 30 sec., room temp.
O

NH 0 0
F. NH 0 HO n o ....0.- o


OH
OAc 0
H z
HO OAc
OBz
109

Figure 3-3: Regioselective Acetylation of 10-Deacetyl Paclitaxel-7- 3-Xyloside


conceivable that nitrate esters can be used as selective hydroxyl protecting groups when

reactivity differing from acetylation reactivity is desired.







73

0
HO_ 0
NHO0 HOn 0

0OH
OHH

110 OBz


{NaIO4, H +


0
HO_ 0
NHO 0 H O0
H
00
OH0

11OBz c


IPhenyihydrazine!
AcOll and MeOH

0

HOHOO

NH 0


OH0

112 O)Bz


HC-N -NH-C6H5
I + HOCH2CH=N-NH-6H5 HIC N-NH-CH5
114
113


Figure 3-4: Conversion of lO-Deacetyl Paclitaxel Xyloside to lO-Deacetyl Paclitaxel






74

Reactions of Taxane Nitrate Esters


Complete Reductive Hydrolysis of Nitrate Esters with Zn and Acetic Acid

Since it had been shown that nitrate esters may serve as regioselective hydroxyl protecting groups, the next step was to determine under what conditions the nitrate esters may be removed without affecting the remainder of the molecule. Reductive nitration is known to occur under a variety of conditions and reagents including high-pressure catalytic hydrogenation, lithium aluminum hydride, hydrazine, Grignard reagents, metallic sodium, and hydrogen sulfide or ammonium sulfide (Green & Wuts, 1991). However many of these methods would undoubtedly react with the taxane structure also, therefore the method chosen was zinc in acetic acid. This method was not only very simple but also caused no rearrangements, hydrolysis, or any other side reactions. It involved dissolving the nitrate ester in acetic acid and adding zinc dust with stirring at room temperature for 30 minutes to give a quantitative yield of the parent alcohol (Figure 3-5).

At this point it was decided to run a series of reactions using a variety of reagents in order to determine what effect the conditions may have on the nitrate esters and/or taxanes. In all cases either paclitaxel-7, 2'-dinitrate or 10-deacetyl paclitaxel-7-3-xyloside2", 3", 4", 10, 2'-pentanitrate was used in these reactions. A discussion of these reactions follows.

Reaction with NaBH4

Paclitaxel-dinitrate was dissolved in methanol and an excess of NaBH4 was added and stirred at room temperature. After 10 minutes TLC confirmed that the reaction was complete and two major products had formed. These products were determined by






75

O
AcO O ONO2
NH O



0N00
ONO2 Of
HOH
115 Oz OAc
OBz
IZn AcOH
room temp., 30 min

Paclitaxel

Figure 3-5: Reductive Denitration of Paclitaxel









Paclitaxel-7, 2'-Dinitrate


I NaBH4 / CH3OH room temp., 10 min O AcO O ONO2

NH
+ H Ol ......
OH 'OH

N02HO H
OBz
116 117


Figure 3-6: Hydrolysis of the Side-Chain of Pacitaxel-7, 2'Dinitrate with NaBH4






76

NMR spectroscopy to be baccatin III-7-nitrate (117) and the side chain alcohol nitrate ester (116). Apparently the NaBH4 serves only to quickly reduce the side chain ester (Figure 3-6)

Reaction with Ammonium Sulfide

Paclitaxel-dinitrate (118) was dissolved in acetonitrile and ammonium sulfide was added. After 2 minutes TLC showed a major slower moving product had formed. After isolation this product was determined to be paclitaxel-7-nitrate (119, Figure 3-7). This result indicates that the nitrate esters may be selectivity removed at least under some conditions however this line of research was not studied further due to time constraints. Acetylation of Taxane Nitrate Esters

In an effort to acetylate the 1-hydroxyl of 10-deacetyl paclitaxel-7-xylosidepentanitrate (120), this compound was dissolved in acetic anhydride and a small amount of DMAP was added. This was stirred at room temperature overnight. After workup the TLC of the reaction mixture displayed two major products (121, 122) and essentially no starting material (Figure 3-8). After separation and isolation it was concluded by NMR spectroscopy that the faster moving product contained two additional acetates while the other product contained one additional acetate. FAB mass spectroscopy was used to determine that the molecular weight of the first compound was 1205 and that of the other was 1163, a difference of 42 or one acetate. The proton spectrum of 122 did not display any signal for H-2', H-3' or N-H. While the proton spectrum of 121 did not display any signal for H-2' or H-3', it did contain a far downfield signal at 11.41 ppm which could indicate a very acidic amide. In view of this information the structure of 121 and 122 were






77

O
0 AcO O ONO2
NH O
0 1.......



ON02 O
H
ON02 HO OAc
118 OBz Ac

20% (NH4)2S, CH3CN room temp., 2 min

0
AcO O ONO2




OH 0 HH
HOOBz OAc
OBz
119
Figure 3-7: Regioselective Denitration of Paclitaxel-7, 2'-Dinitrate with Ammonium Sulfide


assigned as shown (Figure 3-8). Apparently, an enol acetate initially forms between the 2'and 3'- positions to give compound 121. This can then form another enol-like acetate because of the increased acidity of the amide nitrogen to yield compound 122. At this point however it was not understood how the initial enol acetate was formed. Reaction with NaN3

In an attempt to displace a nitrate group, paclitaxel-dinitrate (123) was dissolved in acetonitrile and NaN3 was added with stirring at room temperature. After two hours most of the starting material was no longer present and a product with similiar Rf (124) was






78

0
02NO 0 0 0

NH O 02O OO O 02
ONHON02

ONO2
ON02 O
HO OAc
OBz
120


Acetic Anhydride / DMAP
CH2CI2, room temp., overnight


O
NH 0 02N| 0 O 2
ON2
O002



OAc 0 ONO2
HO OAc
OBz
121

+
OAc
02NO OO OON02

O- H ON02
OAc 0
HO OAc
OBz
122


Figure 3-8: Enol Acetate Formation of Nitrate Esters






79

present on TLC (Figure 3-9). The reaction continued overnight and on the following day it was found that the initial product was no longer present and two other products (125, 126) were now formed, one faster (125) than the "intermediate" product which exhibited strong UV absorbance but did not char with H2S04 and a slower product (126) which exhibited UV absorbance as well as charring with acid on TLC. These two products were isolated by column chromatography and determined by NMR spectroscopy to be baccatin III-7-nitrate (slower product) (126) and dibenzamide (faster product) (125). Dibenzamide was also synthesized from benzamide and benzoyl chloride in the presence of NaH to insure that the reaction product was indeed dibenzamide. It was later found that in DMF as solvent the reaction proceeded much faster and if it were stopped after only 10 minutes the "intermediate" was the major constituent of the reaction mixture. This product was isolated and determined to have a molecular weight of 896. The 1H NMR of this compound was very unusual in that many of the signals seemed to be in duplicate. Signals for the N-H and H-3' were present however there was no signal for the H-2'. The duplication of peaks was similar to what one may find in a racemic mixture presenting the possibility that the stereochemistry at one of the asymmetric carbons had been scrambled. After reviewing the literature concerning nitrate esters it was found that it is quite normal for a nitrate ester to undergo alpha-elimination in the presence of strong base to yield a carbonyl (Boschan et al., 1955). In this case the C-2' is already acidic due to its proximity to the C-l' ester carbonyl, therefore it was quite conceivable that even a weak base such as NaN3 may cause alpha-elimination. With this information in hand the structure of 124 was determined to be as shown with the stereochemistry at C-3' racemic. It was also






80

O
N O AcO 0 ONO2
NH O


ONO2 0

123 OBz

SNaN3, DMF room temp., 10 min
0
NHO AcO O ONO2
NH 0
Y ' ....... .

0 0
HO
HO OAc
124 OBz

Continued reaction AcO O ONO2


N+ HO ......
H O
HH 76 125 + OBz Ac

Glycolic Acid ? 126


Figure 3-9: Reaction of 2'-Nitrate Ester with NaN 3


decided to confirm this structure by producing this compound by a more typical route. Thus paclitaxel-7-nitrate (127) was oxidized with Jones reagent to yield a product (128)






81

possessing the same Rf value as the intermediate keto-ester (Figure 3-10). However this oxidation product only showed one set of signals on the 1H NMR spectrum and this set of peaks matched one of the sets of peaks in the intermediate keto-ester 'H NMR spectra. Presumably C-3' does not racemize in the acidic conditions of the Jones oxidation.

The formation of this keto-ester under basic conditions explains how the enol acetate and dienol acetate can form in the presence of acetic anhydride and pyridine (Figure 3-8). Once the keto-ester forms the H-3' can then be abstracted by the base to form the enolate which then undergoes O-acetylation and once the first enol acetate is formed the formation of the second proceeds as mentioned before. Indeed it was shown that if the paclitaxel keto-ester (129) was treated with pyridine in acetic anhydride the monoenol acetate (130) was the major product (Figure 3-11). This compound was analogues to 121 (Figure 3-8) without the xylose. If this compound was reacted further under these same conditions a slightly faster moving product was formed that was probably the dienol acetate however it was not isolated due to time constraints. Also a second product was also obtained from the keto-ester acetylation that appears to be an enol acetate-enol. This conclusion was arrived at because unlike enol acetate 130 this compound does not show a down field N-H signal yet it has the same mass and the same number of acetates as 130. On TLC however this compound has a much lower Rf than 130, thus it is concluded that this compound may be either 131 or 132 (Figure 3-1 1).

One aspect of this that was not clear however was how the keto-ester breaks down under basic conditions to give dibenzamide and 10-deacetyl baccatin III-7-nitrate ester and what happens to the C-i' and C-2' carbons. It should be stated that this reaction proceeds






82

O
AcO 0 ON02
NH O

/ Oil OH O

127 HO OBz OAc

Jones oxidation
60o C, 6 hrs
O"

xAcO O ONO2
NH O


aH
0.0 128 HO OBz OAc


Figure 3-10: Oxidation of Paclitaxel-7-Nitrate Ester


in DMF or acetonitrile with all bases tried including NaN3, NaOAc, NaOBz, triethylamine, and hydroxide, with hydroxide giving the fastest reaction. However this reaction did not take place when using dichloromethane as solvent with NaN3. No other intermediate products were observed on TLC as loss of the intermediate keto-ester seemed to coincide with formation of the final products. In order to determine that this reaction was base catalyzed the intermediate keto-ester was subjected to three conditions; acetonitrile with dilute hydroxide added, neet acetonitrile, and acetonitrile with dilute HCl added. This study showed that after 24 hours the basic solution was 85-90% decomposed to the final







83

0
AcO 0 ONO, NHO 0



00

129 HO OBz OAc

Sacetic anhydride, DMAP
room temp., 18 hrs

0
AcO 0 ONO, NHO0



OAc n
130 HO OBz OAc


OH
AcO 0 0N02
N 0



OAc0
131 HO z OAc

GAc OR
AcO 0 0N02
N 0OiHO
132 O~zOAc



Figure 3-11: Acetylation of Keto-Ester






84

products, the neutral solution was about 40% decomposed, and the acid solution still contained almost all keto-ester.

Concerning the C-l' and C-2' fragment, it was assumed that C-l' exist as a carboxyl in its final form however it is unclear concerning the C-2'. Thus it can be concluded that this two carbon fragment may exist as acetic acid, glycolic acid, or glyoxalic acid in its final form. A failed attempt was made to derivatize the acid function by treating the reaction mixture with DCC and aniline to produce a UV active amide that could be isolated and characterized.

Unfortunately because of time constraints a mechanism for this rearrangement could not be conclusively established however a possible mechanism has been formulated and is presented below and in Figure 3-12. It has already been established that the ketoester (133) can enolize in the presence of base to form the enolate and thus the enol (134). The amide proton in this enol is subsequently made quite acidic and can also be abstracted by base and after electron migration the imine alcohol can form (135). This conjugated imine is thus a reactive Michael-type adduct which can be attacked by hydroxide with the glycolic enolate serving as the leaving group. The imine can then be rearranged to form dibenzamide (138) while the C-I' ester is hydrolyzed giving glycolic acid (140) and baccatin III-7-nitrate ester (141). This hydrolysis has been shown to occur last because the ester enolate would presumably serve as a better leaving group than the acid enolate, however hydrolysis of the side chain may occur first. It would be interesting to test the hypothesis by subjecting the keto-acid to these conditions to see if the rearrangement still takes place. One could also alkylate the amide to a tertiary amide and determine if the






85

0 BNH 0 H+ H
ORO
HOR OR
OH
B:-13
133 134



H

O

w OR
O-H
0OH
136 HO+ 135
0


OH 137
H+ R = Baccatin III-7-Nitrate Ester
H

OR
OH 137 R =Baccatin 111-7-Nitrate Ester




0 0

N H
138

0 -OH
O -OH HO + ROH
HO O OH
OR 141
139 140

Figure 3-12: Mechanism of Keto-Ester Degradation







86

rearrangement takes place without this available amide proton. In any event time did not allow this to be studied further.


Experimental


All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of the reaction. All NMR spectra were recorded using either a Varian VIXR-300 or a Varian Gemini-300 spectrophotometer using CDC13 as solvent. Infrared spectra were obtained using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolent spectra were obtained using a Shimadzu. UV160U recording spectrophotometer. Mass spectra were recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a Fisher melting point apparatus. Column chromatography was used in conjunction with 100-200 mesh silica gel.

Complete Nitrations of Taxanes

Paclitaxel-7, 2'-dinitrate ester (95)

Paclitaxel 500 mg was dissolved in 6 ml of CH2Cl2 and a mixture of 5 ml of acetic anhydride and 1 mlA of concentrated nitric acid was added slowly. The mixture was prepared by slowly adding the nitric acid to ice cold acetic anhydride so that the mixture does not get too hot. The reaction mixture was allowed to stir at room temperature for 30 minutes. At this point 20 ml of water was added and while stirring NaHCO3 was slowly added until no further frothing was observed. Additional CH2Cl2 was added and the water layer was extracted 3 times with CH2C12. The organic layer was dried with Na2SO4 and the solvent was evaporated. The product was crystallized with diethyl ether and ligroin to




Full Text
23
OAc
OAc
Figure 2-4: Compounds from the Bark of Taxus brevifolia
37
acetic anhydride/pyridine
18 hours, room temp.
Y
35
Figure 2-5: Acetylation of 1 P-Hydroxy-7-Deacetyl Baccatin I


17
In recent years this laboratory has developed a large-scale isolation procedure
using a single reverse-phase column (Rao, 1993; Rao et al. 1995). This procedure has
several advantages over other published procedures some of which are that it is much
simpler, gives higher yields of paclitaxel, and yields several other taxanes which can be
converted to paclitaxel. Specifically, the following yields are obtained for the major
compounds from the bark: paclitaxel (26) (0.04%), 10-deacetyl baccatin III (25) (0.02%),
10-deacetyl paclitaxel-7-(3-xyloside (28) (0.1%), 10-deacetyl paclitaxel-C-7-P-xyloside
(29) (0.04%), 10-deacetyl cephalomannine-7-p-xyloside (27) (0.006%), paclitaxel-7-p-
xyloside (30) (0.008%), 10-deacetyl paclitaxel (31) (0.008%), and cephalomannine (32)
(0.004%) (Figure 2-1). The procedure for this process is defined below and in Figure 2-2.
Air dried yew bark (200-250 lbs.) was extracted with 100 gallons of methanol in a
batch process a total of three times with each extraction lasting one day. The pooled
methanol extracts were concentrated under reduced pressure (<30 C) using a semi-
continuously operated still until the volume of the concentrate reached 20-25 gallons.
Extraction of the concentrated methanolic extract with chloroform was performed in 50-
100 gallon tanks equipped with an air-driven stirrer. The concentrate was stirred with
water and chloroform for about 30 minutes, then 2-14 hours were necessary to allow for
any emulsion to clear. The chloroform layer was drained off from the bottom and the
water layer was extracted two additional times. The pooled chloroform layers were
concentrated under a vacuum to a thick syrup which was poured into glass trays and
converted to powder using a vacuum oven at 35-40 C. The powder was obtained in a
yield of 18-26 g per kg of the bark.


36
dryness and reapplied to a regular-phase silica column using dichloromethane with 0-10%
acetone and then 10% acetone with 0-10% methanol in dichloromethane.
A few compounds eluted with straight dichloromethane and thus had to be run on
another column. This work will be discussed later. Elution with 2%
acetone/dichloromethane gave 1 (3-hydroxy baccatin I (35, Figure 2-4) as mentioned before
(Della Casa De Marcano & Halsall, 1970). This was followed by a compound that was
earlier given the name taxiflorine (77). Taxiflorine is an example of an 11(15>1)-
abeotaxane meaning that the A ring has contracted to contain only 5 carbons. Taxiflorine
was previously isolated by our group but was published with an incorrect structure
assignment (78) (Rao et al., 1996a). Taxiflorine itself gives difficult to interpret 'H and ljC
NMR spectra for reasons that will be discussed later; however, upon acetylation the
spectra are easier to interpret. According to the original structural assignment (78), its
acetate should be the same compound as 13-acetyl-13-decinnamoyl taxchinin B (79)
previously isolated by another group (Das et al., 1995); however, the spectral properties
of these two compounds did not match (Table 2-1). It was then postulated that the correct
structure of taxiflorine is one in which the C-10 benzoate and C-9 hydroxy groups are
reversed so that the hydroxyl is at the C-10 position. To confirm this idea taxiflorine was
oxidized with Jones reagent to the ketone (80, Figure 2-14) and its 1jC spectrum was
compared with those of some known C-9 and C-10 keto taxanes. The carbonyl signal of
80 seen at 192.2 ppm is consistent with an a, (3-unsaturated ketone system, and in contrast
to the 199-204 ppm signal of C-9 keto taxanes. Similarly, the C-12 signal of 80 is at 156.8


CHAPTER 3
PREPARATION OF NITRATE ESTERS OF PACLITAXEL AND RELATED
TAXANES
Complete Nitration of Paclitaxel and Related Taxanes
In an attempt to nitrate the phenyl ring of the N-benzoyl phenylisoserine side chain
in paclitaxel (Figure 3-1, 94) in order to determine its effect on potency, the compound
was subjected to reaction with a 1 : 5 mixture of concentrated nitric acid in acetic
anhydride and an equal volume of dichloromethane. The reaction proceeded readily at
room temperature in 30 minutes to give a single product which exhibited a much higher Rf
on TLC than paclitaxel and even higher than 2, 7-diacetyl paclitaxel (Mellado et al.,
1984), thus eliminating the possibility that acid-catalyzed acetylation had occurred. The
NMR spectrum showed a considerable downfield shift of the 2 and 7 proton signals (H-2
4.78 ppm > 5.69 ppm, H-7 4.40 ppm > 5.75 ppm). Following characterization by
elemental analysis and IR spectroscopy, which exhibited characteristic bands for nitrate
esters at 1650, 1270, and 835 cm'1, the structure was determined to be that of paclitaxel-
2, 7-dinitrate ester (Figure 3-1, 95). After a review of the literature it was found that the
conditions used are actually standard conditions for nitrate ester synthesis (Boschan et al.,
1955). It was surprising that the product showed no evidence of rearrangement of the A-
66


105
Padtaxel-7-mononitrate ester-2-3-enol acetate (130)
White crystalline powder, mp 175-178 C, UV (CH3OH): 233 nm, FABMS
m/z: 939 (39%, M+l), 614 (31%), 554 (59%), 326 (27%), 308 (100%), 266 (79%), 237
(35%), 204 (83%). H NMR 5: 1.18 (s, 3H, 17-H), 1.26 (s, 3H, 16-H), 1.82 (s, 3H, 19-
H), 1.97 (s, 3H, 18-H), 2.05 (m, 3H, 6-H(3), 2.10 (m, 1H, 14-Hp), 2.11 (s, 3H, 2~OAc),
2.21 (s, 3H, 10-0Ac), 2.41 (s, 3H, 4-OAc), 2.59 (m, 1H, 14-Ha), 2.70 (m, 1H, 6-Ha),
4.02 (d, 6.6Hz, 1H, 2-H), 4.15 (d, 8.7Hz, 1H, 20-Hp), 4.34 (d, 8.7Hz, 1H, 20-Ha), 4.98
(d, 9.3Hz, 1H, 5-H), 5,67 (d, 6.9Hz, 1H, 2-H), 5.75 (dd, 7.2, 10.5Hz, 1H, 7-H), 6.09 (t,
8.4Hz, 1H, 13-H), 6.33 (s, 1H, 10-H), 7.37-7.63 (m, 11H, m,p-Bz, o,m,p-Ph, m,p-NBz),
7.94 (d, 7.2Hz, 2H, o-NBz), 8.06 (d, 6.9Hz, 2H, o-Bz), 11.43 (br s, 1H, N-H). 13C NMR
5: 10.9, 14.8, 20.3, 20.6, 20.8, 21.8, 26.5, 32.6, 36.3, 43.1, 47.4, 55.4, 71.8, 74.0, 74.7,
76.2, 79.1, 80.0, 80.4, 83.5, 121.1, 127.6, 127.7, 128.2, 128.7, 128.9, 129.0, 129.5,
130.0, 131.7, 132.8, 132.9, 133.1, 133.9, 141.0, 147.4, 164.8, 165.0, 166.8, 169.4, 169.7,
170.4, 200.5.
PaclitaxeI-7-nitrate ester enol (131,132)
White crystalline powder, mp 176-179 C, FABMS m/z: 939 (23%, M+l), 614
(18%), 554 (29%), 460 (25%), 410 (52%), 308 (29%), 266 (38%), 136 (27%), 105
(100%). ]H NMR 6: 1.09 (s, 3H, 17-H), 1.15 (s, 3H, 16-H), 1.28 (s, 3H, 19-H), 1.77 (s,
3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.17 (s, 3H, OAc), 2.23 (m, 1H, 14-HP) 2.28 (s, 3H,
OAc), 2.40 (m, 1H, 14-Ha), 2.41 (s, 3H, OAc), 2.63 (m, 1H, 6-Ha), 3.92 (d, 7.2Hz, 1H,
3-H), 4.15 (d, 8.4Hz, 1H, 20-Hp), 4.28 (d, 8.4Hz, 1H, 20-Ha), 4.92 (d, 8.4Hz, 1H, 5-H),
5.59 (d, 7.2Hz, 1H, 2-H), 5.71 (dd, 7.2, 10.8Hz, 1H, 7-H), 5.91 (t, 9.6Hz, 1H, 13-H),


26
chemical shifts were almost identical. Thus, it was speculated that the methoxyl/hydroxyl
positions may be the reverse of that of taxamairin A. However, it was then observed that
the UV spectrum of brevixanthane was identical to that reported for taxamairin A while
that of taxamairin B was reported to be completely different. This was confirmed by
synthesizing taxamairin B from brevixanthane with dimethylsulfate. Thus, since the UV
spectra of taxamairins A and B are so different it stands to reason that the UV spectra of
taxamairin A and brevixanthane should also be different if they were different structures.
This was not the case. To solve this question of structure !H NMR NOE experiments
were performed. These experiments illustrated that if the methoxyl methyl of
brevixanthane is irradiated then the phenolic proton and the isopropyl methyne proton are
enhanced proving the proximity of the methoxyl to the isopropyl group. Also when the
phenol proton was irradiated the methoxyl protons and the C-4 proton was enhanced
proving a close proximity between the phenolic group with and carbon 4 of the B ring
(Figure 2-7). Thus, it is concluded that brevixanthane was the same compound as
taxamairin A.
Synthesis of Taxamairin B
The total synthesis of taxamairin diterpenes has been accomplished by one other
group (Wang & Pan, 1995a; Wang et al., 1995b). The synthetic strategy which was used
by this group was derived from the retro synthetic analysis as outlined in Figure 2-8.
Ketone 45 was the key synthetic intermediate because it contains the entire carboncyclic
framework of taxamairin B. Thus the A ring precursor (49) was readily obtained from 1,3


102
Bz), 11.41 (s, 1H, N-H). 13C NMR 8: 10.5, 15.3,20.3,21.1,21.9,26.2,35.8,36.2,42.8,
45.8, 57.9, 59.1, 71.5, 72.3, 73.7, 73.8, 74.1, 76.4, 79.0, 80.3, 80.4, 82.2, 83.3, 98.6,
120.9, 127.6, 127.7, 128.2, 128.7, 128.8, 128.9, 129.6, 130.0, 130.9, 131.7, 132.8, 133.0,
133.9, 143.2, 147.7, 164.6, 165.0, 166.9, 169.6, 170.7, 200.0.
Reaction of Paclitaxel-7, 2-Dinitrate Ester with NaN3
Paclitaxel-7, 2-dinitrate ester 200 mg was dissolved in 4 ml of acetonitrile and 200
mg of sodium azide was added. This mixture was stirred at room temperature overnight.
At that point the acetonitrile was partially evaporated and the residue was partitioned
between water and dichloromethane. The organic layer was removed and the water layer
was partitioned twice more with dichloromethane. The combined organic layers were
washed once with water and then dried with sodium sulfate. TLC analysis showed no
starting material and two products. The faster moving product was UV active but did not
char with 1 N H2SO4 while the slower product did give a positive result in both cases. The
solvent was removed and the residue was put on a regular silica column with 5-15%
acetone in benzene. Both products were cleanly isolated and the slower product
crystallized from diethyl ether and ligroin while the faster compound crystallized upon
evaporation of the fraction solvents. The faster spot was determined to be dibenzamide
(125) (20 mg) and the slower product was baccatin III-7-nitrate ester (126) (152 mg).
Dibenzamide (125)
Colorless needles, mp 150-151 C, UV Lmax (CH3OH): 242 nm, IR (KBr): 3240,
1770, 1475, 1225, 1115, 705 cm'1, HNMR6: 7.51 (t, 7.5Hz, 4H, m-NBz), 7.61 (t,


73
O
HC=NNHQH5
HC=NNHC6H5
113
HOCH2CH=NNHC6H5
114
Figure 3-4: Conversion of 10-Deacetyl Paclitaxel
Xyloside to 10-Deacetyl Paclitaxel


27
Figure 2-8: Retrosynthetic Analysis of Taxamairin B
-cyclohexanedione (48) by azeotropic removal of water from a benzene/hexane/isopropyl
alcohol solution with PTSA as catalyst in a yield of 95% (Figure 2-9). The synthesis of the
C ring was begun by oxidizing o-vanillin (50) with AgO to give the carboxylic acid. This
phenolic acid was then dimethylated using dimethylsulfate to yield the methyl ester (51).
This is then treated with two equivalents of methyl magnesium bromide to yield the


21
+
Paclitaxel
03
ch2ci2/ch3oh
room temp.
Paclitaxel
Figure 2-3: Ozonolysis of Cephalomannine/Paclitaxel Mixture


48
by dichloromethane. At this point the column was eluted with 2-5% acetone in
dichloromethane and then 2-5% methanol and 5% acetone in dichloromethane. A total of
500 ml of each solvent mixture was pumped through before switching to the next solvent.
Fractions of about 100 ml were collected and monitored by TLC. Taxamairin A was
eluted with 2% acetone in dichloromethane and crystallized from the elution solvent. It
was recrystallized from dichloromethane to yield 275 mg. Yellow needles, mp 252-253 C,
EIMS m/z: 338 (80%, M), 310 (74%), 295 (100%), 267 (63%), 237 (18%), 156 (24%).
CIMS: 339 (M + 1). UV 211, 255, 385nm. IR (KBr): 1672, 1535, 1320, 1195,
1052 cm'1. HNMR5: 1.33 (d, 6.9Hz, 6H, 19-H, 20-H), 1.46 (s, 6H, 12-H, 13-H), 3.35
(heptet, 6.9Hz, 1H, 18-H), 3.89 (s, 3H, 15-OMe), 6.11 (d, 9.6Hz, 1H, 7-H), 6.65 (s, 1H,
14-OH), 6.95 (s, 1H, 11-H), 7.31 (d, 9.9Hz, 1H, 11-H), 7.77 (s, 1H, 17-H), 7.94 (s, 1H,
4-H). 13C NMR 5: 23.3, 23.4, 26.7, 26.8, 27.4, 50.5, 62.0, 119.3, 120.8, 123.7, 130.1,
131.1, 133.7, 136.6, 146.1, 146.8, 147.8, 148.2, 151.4, 188.2, 200.9.
Methylation of Taxamairin A
Taxamairin A (50 mg) was dissolved in 3 ml acetone and excess K2CO3 was added
together with 0.25 ml of dimethyl sulfate. This mixture was refluxed for 3 hours and at
that point 0.5 ml of concentrated NH4OH was added to the mixture and stirred for 15
minutes. The acetone was partially evaporated and water was added. This aqueous
solution was then extracted twice with dichloromethane and the combined organic layers
were washed with 0.1 N NaOH and then with water. After drying with sodium sulfate, the
solvent was removed and the residue was crystallized from dichloromethane to yield 32
mg of taxamairin B (39). Yellowish white needles, mp >290 C, UV 'kmax: 219, 281, 355


7
Mn02 (pH 7) T
Pachtaxel f ->- No Reac.
aq. acetone, reflux
Paclitaxel
MnQ2 (pH 8)
aq. acetone,
reflux
O
Figure 1-4: Neutral and Alkaline Oxidation of Paclitaxel
because, unlike the vinca alkaloids and colchicine which inhibit microtubule assembly, it
promotes the formation of discrete bundles of stable microtubules that result from the
reorganization of the microtubule cytoskeleton. The novel characteristic of paclitaxel is its
ability to polymerize tubulin in vitro in the absence of guanosine 5-triphosphate (GTP),
which is normally required for tubulin assembly.
Total Synthesis
As mentioned, the total synthesis of paclitaxel has recently been achieved by
various groups, however one group alone cannot claim to be the first to accomplish this


83
O
OH
Figure 3-11: Acetylation ofKeto-Ester


LIST OF REFERENCES
Ali, S. M.; Hoemann, M. Z.; Aube, J.; Mitscher, L. A.; Georg, G. I. Novel Cytotoxic 3-
(tert-Butyl) 3-Dephenyl Analogs of Paclitaxel and Docetaxel. J. Med. Chem.
1995, 35, 3821-3828.
Appendino, G. The Phytochemistry of the Yew Tree. Nat. Prod. Rep. 1995, 72, 349-360.
Arslanian, R. L.; Bailey, D. T.; Kent, M. C.; Richheimer, S. L.; Thornburg, K. R.;
Timmons, D. W.; Zheng, Q. Y. Brevitaxin, A New Diterpenolignan from the Bark
of Taxus brevifolia. J. Nat. Prod. 1995, 55, 583-585.
Boschan, R.; Merrow, R. T.; Van Dolah, R. W. The Chemistry of Nitrate Esters. Chem.
Rev. 1955, 55, 485-510.
Chen, S. H.; Farina, V. Paclitaxel Structure-Activity Relationships and Core Skeletal
Rearrangements. Taxane Anticancer Agents, Basic Science and Current Status,
ACS Symposium Series 583, 1995, 247-261.
Chen, S. H.; Huang, S.; Wei, J.; Farrina, V. The Chemistry of Taxanes-Reaction of Taxol
and Baccatin Derivatives with Lewis-Acids in Aprotic and Protic Media.
Tetrahedron 1993, 49, 2805-2828.
Chu, A.; Furlan, M.; Davin, L. B.; Zajicek, J.; Towers, G. H. N.; Soucy-Breau, C. M.;
Rettig, S. J.; Croteau, R.; Lewis, G. H. Phenylbutanoid and Taxane-Like
Metabolites from Needles of Taxus brevifolia. Phytochem. 1994, 36, 975-985.
Chu, A.; Davin, L. B.; Zajicek, J.; Lewis, N. G.; Croteau, R. Intramolecular Migrations in
Taxanes form Taxus brevifolia. Phytochem. 1993, 34, 473-476.
Commercon, A.; Bourzat, J. D.; Didier, E.; Lavelle, F. Practical Semisynthesis and
Antimitotic Activity of Docetaxel and Side-Chain Analogues. Taxane Anticancer
Agents, Basic Science and Current Status, ACS Symposium Series 583, 1995,
233-246.
Das, B.; Rao, S. P.; Das, R. Naturally Occuring Rearranged Taxoids. Planta Med. 1995,
61, 393-397.
141


Figure page
3-12 Mechanism of Keto-Ester Degradation 85
4-1 Synthesis of Ionizadle Analogues Ill
4-2 Condensation with Malonic Acid 112
4-3 Condensation with Nitromethane 114
4-4 Dialdehyde Reduction to Diol 116
4-5 Reductive Aminations 117
4-6 Synthesis of Glycosyl Donors 120
4-7 Koenigs-Knorr Glycosylation 122
4-8 Trichloroacetimidate Glycosylation 123
4-9 Rearrangement of 2-Acetyl Paclitaxel 124
xi


6
Since the ester could have originally been joined to hydroxyl groups at either C-7,
C-10, or C--13, it was necessary to establish at which of these hydroxyl moieties the ester
had originally been located. When paclitaxel was oxidized with MnC>2 under neutral
conditions, no reaction occurred. However, Mn02 oxidation of paclitaxel under alkaline
conditions smoothly yielded a reaction product (4) with the structure shown in Figure 1-4.
It is evident that Mn02 oxidation of paclitaxel under neutral conditions did not effect the
hydroxyl groups available for oxidation at C-7 and C-2. When paclitaxel was oxidized
with alkaline MnC>2 an analogue of baccatin III with a conjugated carbonyl moiety as
shown in Figure 1-4 was obtained. It is well known that MnCh oxidation of allylic
hydroxyl groups under alkaline conditions smoothly forms the corresponding conjugated
ketone. This reaction in conjunction with the x-ray structure determination of the
structures of the ester and taxane moieties established the structure of paclitaxel (Wani et
al 1971).
Mechanism of Action
Although paclitaxel displayed good activity against human tumor xenographs and
murine B16 melanoma, its cytotoxic properties were not very different from other drugs
being tested during the 1970s. Rather, its attraction to pharmacologists was its unique
structure which suggested the possibility of a novel mechanism for an anti-tumor drug.
This mechanism was subsequently identified in 1979 by Horwitz and collaborators (Schiff
et al., 1979). Paclitaxel proved to be a potent inhibitor of eukaryotic cell replication,
blocking cells in the late G2-M phase of the cell cycle. It is an unusual mitotic inhibitor


74
Reactions of Taxane Nitrate Esters
Complete Reductive Hydrolysis of Nitrate Esters with Zn and Acetic Acid
Since it had been shown that nitrate esters may serve as regioselective hydroxyl
protecting groups, the next step was to determine under what conditions the nitrate esters
may be removed without affecting the remainder of the molecule. Reductive nitration is
known to occur under a variety of conditions and reagents including high-pressure
catalytic hydrogenation, lithium aluminum hydride, hydrazine, Grignard reagents, metallic
sodium, and hydrogen sulfide or ammonium sulfide (Green & Wuts, 1991). However
many of these methods would undoubtedly react with the taxane structure also, therefore
the method chosen was zinc in acetic acid. This method was not only very simple but also
caused no rearrangements, hydrolysis, or any other side reactions. It involved dissolving
the nitrate ester in acetic acid and adding zinc dust with stirring at room temperature for
30 minutes to give a quantitative yield of the parent alcohol (Figure 3-5).
At this point it was decided to run a series of reactions using a variety of reagents
in order to determine what effect the conditions may have on the nitrate esters and/or
taxanes. In all cases either paclitaxel-7, 2-dinitrate or 10-deacetyl paclitaxel-7-P-xyloside-
2, 3, 4, 10, 2-pentanitrate was used in these reactions. A discussion of these
reactions follows.
Reaction with NaBH4
Paclitaxel-dinitrate was dissolved in methanol and an excess of NaBH4 was added
and stirred at room temperature. After 10 minutes TLC confirmed that the reaction was
complete and two major products had formed. These products were determined by


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
LIST OF TABLES viii
LIST OF FIGURES ix
ABBREVIATIONS xii
ABSTRACT xiii
CHAPTERS
1 HISTORY AND BACKGROUD OF PACLITAXEL 1
Introduction I
Early Work and Structural Elucidation 1
Mechanism of Action 6
Total Synthesis 7
Structure-Activity Relationships 12
2 ISOLATION OF TAXOID AND NON-TAXOID COMPOUNDS
FROM TAXI 'S SPECIES 15
Large Scale Isolation Process 15
Isolation of Minor Compounds from the Bark of Taxus brevifolia 22
Synthesis of Taxamairin B 26
Isolation of Minor Compounds from Taxus floridana 35
Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde 42
Experimental 44
Isolation of Minor Compounds from Taxus brevifolia 44
Acetylation of 1 (3-Hydroxy-7-Deacetyl Baccatin I 47
Isolation of Taxamairin A (38) from Taxus brevifolia 47
Methylation of Taxamairin A 48
Synthesis of Taxamairin B (39) 49
Isolation of Minor Compounds from Taxus floridana 57
v


38
Figure 2-15: HMBC Correlation of Taxiflorine Acetate
of 80 displayed a A,x at 232 nm with a shoulder at 253 nm, this shoulder is consistent
with an a, (3-unsaturated ketone system. Finally, to confirm the revised structure an
HMBC spectrum was taken on the taxiflorine acetate (81, Figure 2-15). From this
spectrum the interactions between the ortho protons and benzoate carbonyl carbon were
clearly visible as was the interaction between the benzoate carbonyl carbon and the proton
at C-9. Likewise the C-19 protons interacted with the C-9 carbon, which in turn interacted
with the C-9 protons on the regular HETCOR spectrum. With this information in hand the


62
Acetylation of Taxchinin L (83)
Taxchinin L 100 mg was dissolved in 2 ml of acetic anhydride and 1 ml of pyridine
was added. The mixture was stirred at room temperature for 18 hours and then water was
added to the mixture. Sodium bicarbonate was added slowly until no further evolution of
C02 was observed. The aqueous mixture was then extracted twice with dichloromethane
and the combined organic layers were washed with 0.1 N NaOH, 0.1 N HC1, and water
successively and dried with sodium sulfate. The dichloromethane was evaporated and the
product (84 mg) was obtained as a glassy solid. The physical and spectral properties of
this acetylated product (81) was identical with the acetate of taxiflorine in every way.
Synthesis of Trans-2, 6-Dimethoxycinnamaldehyde (85)
Methyl 2, 6-dimethoxybenzoate
A total of 5.0 g of 2, 6-dimethoxybenzoic acid (88) was dissolved in 30 ml of
methanol and 1.0 ml of concentrated H2SO4 was added. This mixture was refluxed for 48
hours at which time most of the methanol was removed under reduced pressure. The
residue was partitioned between water and diethyl ether and the organic layer was
partitioned with 0.1 N NaOH to remove the remaining starting material, this was
performed three times. Finally the organic layer was washed with water twice and dried
with sodium sulfate. Upon removal of the solvent under reduced pressure the methyl ester
product crystallized to yield 3.1 g of product.
2, 6-Dimethoxybenzyl alcohol (89)
A total of 3.1 g of the methyl ester was dissolved in 15 ml of THF and this was
cooled to 0 C. A total of 581 mg (1 eq.) of LAH was then carefully added to the mixture


30
49 + 56
BF3-Et20
toluene
t-BuOOH
Ci o3, CH2C12
y
Figure 2-10: Literature Synthesis of Taxamairin B
brominated by slowly adding Br2 in dichloromethane at room temperature while closely
monitoring the TLC. This process gave a yield of 60% after purifying the product by
column chromatography. Finally, 4-bromo-2, 2-dimethyl-l, 3-cyclohexanedione (65) was


70
Table 3-3continued
H or C# Coimpd. 95
Compd. 97
Compd. 101
Compd. 104
C=0 170.2, 169.5,
171.1, 166.9
170.4, 167.6,
170.3, 167.3,
167.7, 166.8,
166.7, 166.7
166.8, 166.7
166.7
Regioselective Nitrations of Paclitaxel and Related Taxanes
Based on these results it was decided to run the reaction at 0 C for only 10-15
minutes to determine if the reaction was regioselective. Using this protocol, paclitaxel
gave the 7-mononitrate of paclitaxel (96) in -90% yield after crystallization from the crude
reaction mixture (Figure 3-1). The position of the nitrate ester was easily determined by
'H NMR and COSY spectroscopy as the nitrate ester causes a downfteld shift of 1.3-1.7
ppm on the adjacent proton. This result was interesting since the 2-hydroxyl is much
more reactive than the 7-hydroxyl in acetylation reactions (Mellado et al., 1984). With
nitration however the order of reactivity was 7-OH > 2-OH.
Similar studies were applied to 10-deacetyl baccatin III, 10-deacetyl paclitaxel, and
10-deacetyl paclitaxel-7-p-xyloside. They also displayed some regioselectivity, however
because these compounds contain more than 2 reactive hydroxyl groups, silica column
chromatography was needed to separate the products. When 10-deacetyl baccatin III was
subjected to this protocol three partially nitrated compounds were obtained namely the 10-
mononitrate (98), the 10, 13-dinitrate (99), and the 7, 10-dinitrate (100), although 98 and
99 were the major compounds. From this it can be concluded that the order of reactivity is
10-OH > 13-OH > 7-OH. This also differs from the acetylation reactivities in which the 7-
hydroxyl is the most reactive followed by the 10-hydroxyl and the 13-hydroxyl


98
13C NMR 5: 10.8, 14.2, 20.6, 22.6, 26.7, 35.3, 35.4, 43.1, 46.8, 54.9, 56.8, 61.6, 71.4,
72.4, 73.1, 74.5, 74.7, 76.6, 78.7, 79.7, 81.1, 81.7, 84.0, 104.4, 127.0, 127.1, 128.3,
128.7, 128.8, 128.9, 129.0, 130.3, 132.0, 133.6, 133.7, 136.1, 137.9, 138.1, 166.8, 167.5,
170.6, 173.0, 210.
Reductive Denitration of Paclitaxel-7, 2-Dinitrate Ester
Paclitaxel-7, 2-dinitrate 200 mg was dissolved in 3 ml of acetic acid and 500 mg
of zinc powder was added. The mixture was stirred at room temperature for 30 minutes
and then filtered through a celite bed to remove the zinc powder. Water was added to the
acetic acid and the acid was neutralized with bicarbonate. The aqueous layer was
extracted with dichloromethane twice and the organic layer was dried with sodium sulfate.
After removal of the solvent 174 mg of paclitaxel was crystallized from diethyl ether and
ligroin. All NMR spectra matched those of an authentic sample.
Reaction of PaclitaxeI-7-2-Dinitrate with NaBH4
Paclitaxel-7-2-dinitrate ester 200 mg was dissolved in 2 ml of methanol and
excess NaBH4 was added. This was stirred for 10 minutes at room temperature and the
reaction was quenched with 1 N HC1. The methanol was partially removed and water and
dichloromethane was added and partitioned. The organic layer was removed and the water
layer was extracted twice more with dichloromethane. The combined water layers were
washed with water once and then dried with sodium sulfate. TLC analysis showed two
products, the faster was UV active but did not char with H2S04 while the slower product
gave a positive test in both cases. The evaporated residue was applied to a regular silica
column and eluted with 5-15% acetone in benzene. Both compounds were isolated and


112
O
+
O o
X X
ll(X
151
pyridine/piperidine
reflux, 3 hrs.
y
O
Figure 4-2: Condensation with Malonic Acid
with 2.5 eq. of nitromethane and stirring at room temperature for overnight. Although
essentially all the starting material disappeared the reaction was not a clean one being that
at least three products were formed. This result was somewhat expected since the reaction


19
Methanol Extract
T
Partition Between Water and Chloroform
Water Chloroform
C18 Column
Acetonitrile/Water
Pure Paclitaxel Fractions Fractions Containing Paclitaxel
(Crystalline) and Cephalomannine
Paclitaxel 0.04%
10-Deacetyl Baccatin III 0.02%
10-Deacetyl PaclitaxeI-7-Xyloside 0.1%
10-Deacetyl Paclitaxel-C-7-Xyloside 0.04%
Ozonolysis/
Regular Silica Colum
T
Paclitaxel
Figure 2-2: Reverse-Phase Isolation of Taxanes


126
amination products (160-164) the only compounds which is less active than the xyloside is
the m-aminosalicylic acid product (162) which has a substituent ortho to the amino group.
The p-aminosalicylic acid product (161) however is much more active, thus indicating that
ortho substituents are not well tolerated. However more work needs to be performed to
verify this.
Table 4-2: L1210 Cytotoxicity of Paclitaxel and Analogues
Compd.
ICso (ppm)
IC50 (pM)
Compd./Paclitaxel
Paclitaxel
0.0089
0.01
s¡ 5|< >i<
10-Deacetyl
7-Xyloside
Paclitaxel- 0.37
0.39
39
152
3.31
3.47
347
155
0.89
0.92
92
160
0.048
0.047
4.7
161
0.095
0.092
9.2
162
2.00
1.94
194
163
0.071
0.071
7.1
164
0.13
0.12
12
157
0.93
1.02
102
Experimental
All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of
the reaction. All NMR spectra were recorded using either a Varan VXR-300 or a Varan
Gemini-300 spectrophotometer using CDC13 as solvent. Infrared spectra were obtained
using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolent spectra were
obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were
recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a
Fisher melting point apparatus. Column chromatography was used in conjunction with
100-200 mesh silica gel.


75
O
Figure 3-5: Reductive Denitration of Paditaxel
Paclitaxel-7, 2'-Dinitrate
NaBH4 / CH3OH
room temp., 10 min
Figure 3-6: Hydrolysis of the Side-Chain of Pacitaxel-7, 2'-
Dinitrate with NaBH4


18
For chromatography, stainless steel columns either 4 x 4 or 6 x 6 were used.
The columns were packed with C-18 bonded silica as a slurry in methanol. Approximately
3-4 kg and 12-13 kg of silica were used with the 4 and 6 columns, respectively. After a
thorough wash with methanol, the columns were equilibrated with 25% acetonitrile in
water. For running the 6 diameter column, the powder from the chloroform extract (2-
2.5 kg) was dissolved in acetonitrile (5 1) and while this mixture was being stirred with
equilibrated C-18 silica (1-2 1), it was diluted with water to make 20 1. The mixture was
then allowed to stand for 15-30 minutes and the clear supernatant siphoned off into
another container. The slurry was applied to the column, followed by part of the
supernatant, after which the column was sealed. The remaining supernatant was pumped
into the column using a diaphragm metering pump maintaining a pressure of 30-80 psi.
After the sample had been pumped onto the column it was eluted with a step gradient of
35, 40, 45, and 50% acetonitrile in water. The change in solvent was dictated by the
results of the TLC and HPLC of the fractions but usually 40-50 1 of each solvent was
used. After this, the column was washed with methanol, followed by a mixture of ethyl
acetate and ligroin until the effluent was nearly colorless. Following this, the column was
again washed with methanol and equilibrated with 25% acetonitrile in water. The column
fractions (about 2 1 each) were allowed to stand at room temperature for 2-8 days, by
which time crystals appeared in many. Soon after, the crystals were filtered, analyzed for
purity and composition by HPLC, and recrystallized from the appropriate solvent.
In terms of the elution sequence of the taxanes, the earliest taxane to appear was
10-deacetyl baccatin III (25) which crystallized from the fractions eluted by 35%


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
(jju
Do
Donghai Wu
Assistant Professor of Medicinal
Chemistry
This dissertation was submitted to the Graduate Faculty of the College of
Pharmacy and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
December 1998
Dean, College of Pharr
Dean, Graduate School


BIOGRAPHICAL SKETCH
The author was bom to James Harvey Johnson and Lawilla McLamb
Johnson on September 30, 1970 in Fayetteville, North Carolina. His father worked
for the civil service at Fort Bragg Army Reservation and his mother was a
homemaker. He grew up with two older sisters, Melinda and Teresa. He attended
and graduated in June 1988 from Cape Fear High School in Vander, North
Carolina near Fayetteville. In January 1989 he enrolled at Fayetteville State
University on a full academic scholarship. It was at this time that he began to
develop a love for the discipline of chemistry. After graduating with a B.S. in
chemistry with honors in May 1993 he enrolled in the Department of Medicinal
Chemistry at the University of Florida during August 1993. It was there that this
doctoral work was completed under the supervision of Koppaka V. Rao. The
author was married to Amy Raye Hardy on April 30, 1994 and has one child,
James Harvey Johnson III (Trey) born September 23, 1995.
146


page
L1210 Cytotoxicity of Analogues 125
Experimental 126
Oxidation of Xyloside with Periodate 127
Condensation of Dialdehyde with Malonic Acid 127
Condensation of Dialdehyde with Nitromethane 128
Reduction of Dialdehyde to the Diol 128
General Procedure for Reductive Aminations 129
Synthesis of Glucose Pentaacetate 133
Synthesis of la-Bromo-Tetraacetyl Glucose 134
Synthesis of 1-Hydroxy-Tetraacetyl Glucose 134
Synthesis of la-Trichloroacetimidate-Tetraacetyl Glucose 135
Synthesis of Tetraacetyl Phenyl Thioglucoside 135
Synthesis of Tetraacetyl Glucose, Phenyl Sulfoxide 136
Preparation of Tetraacetyl Benzyl P-Glucoside by the
Koenigs-Knorr method 136
Preparation of Tetraacetyl Benzyl p-Glucoside by the
Trichloroacetimidate Method 137
Preparation of Tetraacetyl P-Sitosterol P-Glucoside by the
Koenigs-Knorr Method 137
Preparation of Tetraacetyl p-Sitosterol P-Glucoside by the
Trichloroacetimidate Method 138
Attempted Glucosylation of 2-Acetyl Paclitaxel by the
Koenigs-Knorr Method 138
L1210 Cytotoxicity Assay 139
LIST OF REFERENCES 141
BIOGRAPHICAL SKETCH 146
vii


68
The reaction was also repeated with three of the natural analogues of paclitaxel,
ie., 10-deacetyl baccatin III, 10-deacetyl paclitaxel, and 10-deacetyl paclitaxel-7-(3-
xyloside. In each case the reaction proceeded to yield the corresponding tri- (Figure 3-2,
97), tri-(Figure 3-3, 101), and penta-nitrate esters (Figure 3-4, 104), respectively. All of
these compounds crystallized readily from ethyl ether after workup without any need for
chromatography to give nearly quantitative yields. In no case was nitration at the sterically
hindered 1-hydroxyl observed. NMR chemical shift values are given in Table 3-1.
Table 3-1: FI and ljC NMR Values for Completely Nitrated Taxanes
Hor C#
Compd. 95
Compd. 97
Compd. 101
Compd. 104
1
4444 yg g
**** 78 2
4444 yg 5
4444 yg y
2
5.72 d (6.9Hz),
74.3
5.64 d (6.6Hz),
73.5
5.71 d(6.9FIz),
74.0
5.71 d (6.9Hz),
74.5
3
4.02 d (6.9Hz),
47.2
3.92 d (6.9Hz),
47.7
3.95 d(6.9Hz),
47.2
3.81 d (6.6Hz),
45.7
4
4444 gQ y
4444 gQ y
4444 gQ g
4444 gQ g
5
4.99 d (9.0Hz),
83.6
4.97 d (8.7Hz),
83.5
4.98 d (9.0Hz),
83.4
4.88 d (9.0Hz),
83.5
6a
2.69 m, 32.5
2.72 m, 32.5
2.71 m, 32.5
2.81 m, 35.8
6P
2.04 m, ****
2.04 m, ****
2.07 m, ****
2.04 m, ****
7
5.75 dd (7.2,
10.5 Hz), 79.8
5.78 dd (7.2,
10.5 Hz), 79.9
5.75 dd (7.2,
10.5 Hz), 79.8
4.17 m, 80.1
8
4444 gg g
****, 55.8
**** 55 6
4444 5y 8
9
****, 200.4
4444 299 3
4444 299 5
4444 299 9
10
6.31 s, 74.5
6.38 s, 81.4
6.36 s, 81.6
6.42 s, 82.1
11
****, 135.5
4444 \242
****, 135.4
****, 135.5
12
**** |409
Hi 4= * \a,2 2
4444 244 2
4444 243 2
13
6.30 t (9.9 Hz),
72.8
6.23 t (8.1 Hz),
78.1
6.29 t (9.0 Hz),
72.5
6.25 t (8.7 Hz),
72.6
14a
2.38 m, 35.4
2.48 dd (9.9,
15.6 Hz), 34.2
2.39 m, 35.3
2.42 m, 35.3
140
2.38 m, ****
2.36 dd (2.8,
15.9 Hz), ****
2.39 m, ****
2.30 m, ****
15
**** 43 2
**** qy 9
444 4 q.3 2
4444, 42 9
16
1.23 s, 26.5
1.20 s, 26.6
1.20 s, 26.4
1.19 s, 26.2
17
1.17 s, 21.5
1.15 s, 20.3
1.14 s, 21.5
1.13 s, 21.6


25
och3
Figure 2-6: Structure of Abeo-Abietane Diterpenoids
Figure 2-7: NOE Correlations of Taxamairin A


20
acetonitrile in water. The next group of taxanes to be eluted were the various xylosidic
taxanes; 10-deacetyl cephalomannine-7-p-xyloside (27), 10-deacetyl paclitaxel-7-P-
xyloside (28), 10-deacetyl paclitaxel-C-7-P-xyloside (29), and paclitaxel-7-p-xyloside
(30). Of these the first two were well separated. As the elution of 10-deacetyl paclitaxel-7-
P-xyloside was nearing completion, 10-deacetyl paclitaxel-C-7-P-xyloside started to elute.
Halfway though its elution, paclitaxel-7-P-xyloside and 10-deacetyl paclitaxel (31) started
to co-elute. These last two compounds also crystallized together, however separation was
readily achieved by running the mixture through a regular silica column using 0-5%
methanol in chloroform as solvent.
Continued elution of the column with 50% acetonitrile in water gave
cephalomannine (32), followed closely by paclitaxel (26). The earlier part of the band
contained mixtures of the two, but the later fractions contained mostly paclitaxel which
could be recrystallized. The fractions that contained the mixture were combined and dried
to a solid. This solid was then subjected to ozonolysis at -78 C for 45 minutes. This
process converted the cephalomannine to the keto-amide 34 but did not disturb paclitaxel
(Figure 2-3). After workup this material was run through a regular silica column with 0-
5% acetone in chloroform and the paclitaxel was isolated. It should be pointed out that
this process was necessary because paclitaxel and cephalomannine cannot be separated on
regular-phase silica.


101
3.70 (dd, 5.1, 12.9Hz, 1H, 5-H), 3.87 (d, 6.9Hz, 1H, 3-H), 4.18 (d, 8.4Hz, 1H, 20-H(3),
4.23 (m, 1H, 7-H), 4.29 (d, 8.4Hz, 1H, 20-Ha), 4.80 (d, 8.7Hz, 1H, 5-H), 4.81 (d, 3.8Hz,
1H, 1 -H), 4.97 (dd, 4.2, 5.7Hz, 1H, 2-H), 5.07 (dd, 4.5, 8.7Hz, 1H, 4-H), 5.26 (t,
5.7Hz, 1H, 3-H), 5.65 (d, 7.2Hz, 1H, 2-H), 6.06 (t, 7.8Hz, 1H, 13-H), 6.44 (s, 1H, 10-
H), 6.73 (d, 7.2Hz, 2H, o-NBz), 7.17 (t, 7.8Hz, 2H, m-NBz) 7.23-7.28 (m, 3H, o-Ph, p-
NBz), 7.52-7.63 (m, 4H, m,p-Bz, p-Ph), 8.21 (d, 7.5Hz, 2H, o-Bz). 13C NMR 5: 10.7,
15.2, 20.3, 21.1, 21.7, 26.1, 26.5, 35.7, 36.0, 42.7, 46.0, 57.9, 59.3, 71.4, 72.7, 73.9,
74.1, 74.5, 76.4, 79.4, 80.0, 80.3, 82.2, 83.5, 98.6, 128.1, 128.3, 128.4, 128.5, 128.8,
129.2, 129.8, 130.4, 131.4, 132.6, 133.6, 133.7, 133.8, 135.0, 137.3, 142.7, 161.6, 167.0,
170.1, 172.0, 172.1, 174.6, 200.0.
10-Deacetyl paclitaxel -7-P-xyIoside-2, 3, 4, 10-tetranitrate-mono-eno! acetate
(121)
White crystalline powder, mp 178-180 C, UV Amax (CH3OH): 232 nm, FABMS
m/z: 1164 (M + 1), 326, 308. NMR 5: 1.15 (s, 3H, 17-H), 1.22 (s, 3H, 16-H), 1.77
(s, 3H, 19-H), 2.01 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.10 (s, 3H, OAc), 2.11 (m, 1H,
14-HP), 2.39 (s, 3H, OAc), 2.56 (m, 1H, 14-Ha), 2.83 (m, 1H, 6-Ha), 3.72 (dd, 5.1,
12.9Hz, 1H, 5-Heq), 3.81 (d, 7.2Hz, 1H, 3-H), 4,16 (d, 8.7Hz, 1H, 20-Hp), 4.16 (m,
1H, 7-H), 4.23 (dd, 3.6, 15.6Hz, 1H, 5-Hax), 4.33 (d, 8.7Hz, 1H, 20-Ha), 4.83 (d,
3.6Hz, 1H, 1-H), 4.87 (d, 8.7Hz, 1H, 5-H), 4.97 (dd, 3.9, 6.0Hz, 1H, 2-H), 5.07 (dd,
5.1, 9.3Hz, 1H, 4-H), 5.26 (t, 5.7Hz, 1H, 3-H), 5.69 (d, 6.9Hz, 1H, 2-H), 6.10 (t,
8.1Hz, 1H, 13-H), 6.44 (s, 1H, 10-H), 7.39-7.52 (m, 9H, m-Bz, o,m,p-Bz, m-NBz), 7.57
(m, 1H, p-NBz), 7.61 (m, 1H, p-Bz), 7.94 (d, 7.2Hz, 2H, o-NBz), 8.06 (d, 7.2Hz, 2H, o-


16
O
26 Ri = H, R2 = Ac, R3 phenyl
27 Ri = p-xylosyl, R2 = H, R3 = tiglyl
28 Ri = P-xylosyl, R2 = H, R3 = phenyl
29 Ri = p-xylosyl, R2 = H, R3 = n-pentyl
30 Rj = P-xylosyl, R2 = Ac, R3 = phenyl
31 Ri = H, R2 = H, R3 = phenyl
32 Rj = H, R2 = Ac, R3 = tiglyl
Figure 2-1: Structure of Major Taxanes
synthetic conversion to paclitaxel and 2) large-scale cultivation of the ornamental yew
(Taxus media Hicksii) and isolation of paclitaxel from its needles/twigs. Among the future
alternatives are total synthesis, of which various schemes have been published, and
isolation from large-scale plant cell culture.


142
Das, B.; Rao, S. P.; Srinivas, K. V. N. S.; Yadav, I. S.; Das, R. A Taxoid from Needles of
Himalayan Tctxus baccata. Phytochem. 1995, 38, 671-674.
Della Casa De Marcano, D. P.; Halsall, T. G. The Structure of Diterpenoid Baccatin-I, the
4(3, 20-Epoxide of 2a, 5a, 7(3, 9a, 10(3, 13a-Hexa-acetoxytaxa-4(20), 11-Diene.
Chem. Comm. 1970, 1381-1382.
Della Casa De Marcano, D. P.; Halsall, T. G. Structures of Some Taxane Diterpenoids,
Baccatins-III, -IV, -VI and -VII and 1-Dehydroxybaccatin-IV, Possessing an
Oxetane Ring. Chem. Comm. 1975, 365-367.
Dennis, J. N.; Green, A. E.; Guenard, D.; Voegelein, F. G.; Managatal, L.; Poitier, P. A
Highly Efficient, Practical Approach to Natural Taxol. J. Am. Chem. Soc. 1988,
110, 5917-5919.
Georg, G. I.; Harriman, G. C. B.; Velde, D. G. V.; Boge, T. C.; Cheruvallath, Z. S.;
Datta, A.; Hepperle, M.; Park, H.; Himes, R. H; Jayasinghe, L. Medicinal
Chemistry of Paclitaxel; Chemistry, Structure-Activity Relationships, and
Conformational Analysis. Taxane Anticancer Agents, Basic Science and Current
Status, ACS Symposium Series 583, 1995, 217-232.
Green, T. W ; Wuts, P. G. M. Protective Groups in Organic Synthesis 1991, 417-420.
Gueritte-Voegelein, F.; Senilh, V.; David, B.; Guenard, D.; Potier, P. Tetrahedron 1986,
42, 4451-4457.
Gunawardana, G. P.; Premachandran, U.; Burres, N. S.; Whittern, D. N.; Henry, R.;
Spanton, S. Isolation of 9-Dihydro-13-Acetylbaccatin-III from Taxus canadensis.
J. Nat. Prod. 1992, 55, 1686-1689.
Holton, R. A.; Somoza, C.; Kim, H. B.; Liang, F.; Biediger, R. J.; Boatman, P. D.;
Shindo, M.; Smith, C. C.; Kim, S.; Nadizadeh, H.; Suzuki, Y.; Tao, C.; Vu, P.;
Tang, S.; Zhang, P.; Murthi, K. K.; Gentile, L. N.; Liu, J. H. First Total Synthesis
of Taxol. 1. Functionalization of the B Ring. J. Am. Chem. Soc. 1994a, 116, 1597-
1598.
Holton, R. A.; Somoza, C.; Kim, H. B.; Liang, F.; Biediger, R. J.; Boatman, P. D.;
Shindo, M.; Smith, C. C.; Kim, S.; Nadizadeh, H.; Suzuki, Y.; Tao, C.; Vu, P.;
Tang, S.; Zhang, P.; Murthi, K. K.; Gentile, L. N.; Liu, J. H. First Total Synthesis
of Taxol. 2. Completion of the C and D Rings. J. Am. Chem. Soc. 1994b, 116,
1599-1600.
Kahne, D.; Walker, S.; Cheng, Y.; Engen D. V. Glycosylation of Unreactive Substrates. J.
Am. Chem. Soc. 1989, 111, 6881-6882.


122
Hg(CN)2, Z11CI2, orAgN03
CH2CI2, room temp.
Figure 4-7: Koenigs-Knorr Glycosylation
order to take advantage of less hindered hydroxyls, 10-deacetyl paclitaxel-7-xyloside was
used as the substrate hoping the reaction would occur on one of the xylose hydroxyls.
However as before no reaction took place with the weaker Lewis acids. In an attempt to


76
NMR spectroscopy to be baccatin III-7-nitrate (117) and the side chain alcohol nitrate
ester (116). Apparently the NaBH4 serves only to quickly reduce the side chain ester
(Figure 3-6)
Reaction with Ammonium Sulfide
Paclitaxel-dinitrate (118) was dissolved in acetonitrile and ammonium sulfide was
added. After 2 minutes TLC showed a major slower moving product had formed. After
isolation this product was determined to be paclitaxel-7-nitrate (119, Figure 3-7). This
result indicates that the nitrate esters may be selectivity removed at least under some
conditions however this line of research was not studied further due to time constraints.
Acetylation of Taxane Nitrate Esters
In an effort to acetylate the 1-hydroxyl of 10-deacetyl paclitaxel-7-xyloside-
pentanitrate (120), this compound was dissolved in acetic anhydride and a small amount of
DMAP was added. This was stirred at room temperature overnight. After workup the
TLC of the reaction mixture displayed two major products (121, 122) and essentially no
starting material (Figure 3-8). After separation and isolation it was concluded by NMR
spectroscopy that the faster moving product contained two additional acetates while the
other product contained one additional acetate. FAB mass spectroscopy was used to
determine that the molecular weight of the first compound was 1205 and that of the other
was 1163, a difference of 42 or one acetate. The proton spectrum of 122 did not display
any signal for H-2, H-3 or N-H. While the proton spectrum of 121 did not display any
signal for H-2 or H-3, it did contain a far downfield signal at 11.41 ppm which could
indicate a very acidic amide. In view of this information the structure of 121 and 122 were


31
dehydrohalogenated by refluxing with excess LiCl in DMF for 2 hours. This process gave
a yield of 82% after purification by column chromatography of the A ring precursor 2, 2-
dimethyl-4-cyclohexene-l, 3-dione (66) (Figure 2-11).
The formation of the C ring precursor was more problematic. Acetovanillone (67)
was used as the starting material and this compound was isopropylated by heating with
90% H2S04 and isopropyl alcohol at 60 C for 36 hours. Unfortunately this reaction could
not be pushed beyond 50% conversion based on TLC of the reaction mixture, and after
purification by column chromatography gave a yield of 40%. However this process is a
better alternative than the 5-step process outlined in the previous synthesis for placing the
isopropyl group on the ring. Following this, the phenolic group was methylated using
dimethylsulfate and K2C03 in refluxing acetone for 2 hours. This process was nearly
quantitative to yield the dimethoxy product (68) (Figure 2-11). At this point a one carbon
oxygenated substituent had to be introduced between the acetyl and methoxyl groups.
Initially, a Vilsmeier-Haack reaction was attempted, but this gave a variety of products in
which the acetyl seemed to undergo some reaction; however, no effort was made to
characterize these products. Undesirable reactions also occurred with this method if the
acetyl was first reduced to an alcohol group or completely reduced to an ethyl group.
Attempts were also made to acetoxymethylate the desired position so the resulting acetate
could be hydrolyzed and the alcohol oxidized to the aldehyde. The reaction was performed
using 85% H3PO4, acetic anhydride, and paraformaldehyde and the chosen substrate was
the reduced ethyl compound (73) which has less steric bulk and is more activated than the


69
Table 3-1continued
HorC#
Compd. 95
Compd. 97
Compd. 101
Compd. 104
18
1.94 s, 14.3
2.12 s, 15.0
2.00 s, 14.8
1.97 s, 14.8
19
1.82 s, 11.0
1.82 s, 10.8
1.84 s, 11.0
1.75 s, 10.7
20a
4.35 d (8.7 Hz),
76.2
4.36 d (8.7 Hz),
76.0
4.37 d (8.4 Hz),
76.2
4.35 d (8.7 Hz),
76.5
20p
4.20 d (8.4 Hz),
****
4.11 d (8.4 Hz),
****
4.20 d (8.4 Hz),
****
4.19 d (8.7 Hz),
Hi * *
V
5.69 d (3.0 Hz),
80.3
****
5.68 d (2.7 Hz),
80.2
5.65 d (3.0 Hz),
80.3
3
6.14 dd (2.7, 9.6
Hz), 52.1
****
6.13 dd (2.7, 9.3
Hz), 52.1
6.10 dd (3.0, 9.3
Hz), 52.2
NH
6.97 d (9.6 Hz),
* * *
****
6.90 d (9.3 Hz),
****
6.82 d (9.3 Hz),
****
1
* Hi %
****
****
4.76 d (4.5 Hz),
98.9
2
Hi Hi Hi*
****
* * *
4.98 dd (4.2, 6.3
Hz), 74.3
3
Hi***
****
****
5.26 t (6.3 Hz),
72.9
4
H< Hi *
****
****
5.08 dd (5.7, 9.6
Hz), 74.1
5 ax
****
****
****
4.23 m, 59.5
5 eq
****
****
****
3.68 dd (5.7,
12.9 Hz), ****
4-Ac
2.50 s, 22.6
2.43 s, 22.3
2.51 s, 22.6
2.48 s, 22.7
10-Ac
2.18 s, 20.6
****
****
OBz-1
**** 220 3
****, 128.6
**** |20 3
**** 2290
OBz-2,6
8.12 d (7.2 Hz),
130.2
8.06 d (6.9 Hz),
130.1
8.13 d (7.5 Hz),
130.2
8.13 d (7.2 Hz),
130.2
OBz-3,5
7.52 t (7.8 Hz),
128.8
7.51 t (7.8 Hz),
128.9
7.52, 128.8
7.40-7.55, 128.8
OBz-4
7.63 t, 133.8
7.65 t, (7.5 Hz),
131.9
7.64 t (7.2 Hz),
133.9
7.63 t (6.6 Hz),
133.8
NBz-1
****, 131.2
****
**** 12] 3
He*** 2
NBz-2,6
7.73 d (7.2 Hz),
127.1
****
7.73 d (7.5 Hz),
127.1
7.73 d (6.9 Hz),
127.1
NBz-3,5
7.41-7.46, 129.4
****
7.41-7.46, 129.4
7.40-7.55, 129.4
NBz-4
7.41-7.46, 129.0
Hi***
7.41-7.46, 129.1
7.40-7.55, 129.1
Ph-1
****, 133.1
****
****, 133.0
**** 233 1
Ph-2,6
7.41-7.46, 126.5
****
7.41-7.46, 126.5
7.40-7.55, 126.5
Ph-3,5
7.41-7.46, 128.8
****
7.41-7.46, 128.8
7.40-7.55, 128.8
Ph-4
7.49-7.54, 132.3
****
7.52, 132.4
7.40-7.55, 132.3


53
Hydroxymetfiylation of 69
A total of 200 mg of 69 was dissolved in 3 ml of dry diethyl ether, 205 mg (2 eq.)
of TMEDA was added, and the mixture was cooled to -78 C under a helium atmosphere.
After cooling the mixture, 0.4 ml of 2.5 M n-butyllithium in hexanes (2.2 eq.) was added
with stirring. After stirring for 30-45 minutes excess paraformaldehyde was added and the
mixture was stirred overnight while warming to room temperature. The mixture was
diluted with water and diethyl ether and partitioned. The organic layer was washed twice
with water, dried with sodium sulfate, and concentrated. According to TLC about 50% of
the product remained. This material was separated on a silica column using 30-50% ethyl
acetate in ligroin as eluent and a total of 116 mg of product was isolated as a slightly
yellow oil. Yellow oil, lU NMR 8: 1.22 (d, 6.9Hz, 6H, CH3), 1.56 (d, 6.3Hz, 3H, CH3),
3.32 (quintet, 6.9Hz, 1H, CH), 3.84 (s, 3H, OCH3), 3.88 (s, 3H, OCH3), 4.68 (d, 11.7Hz,
1H, CH20), 4.86 (d, 11.7Hz, 1H, CH20), 5.12 (q, 6.6, 12.9Hz, 1H, CH), 7.10 (s, 1H, Ar-
H). 13C NMR 8: 23.1, 23.4, 27.0, 33.5, 55.9, 60.6, 61.2, 67.2, 118.7, 129.9, 139.5,
142.6, 149.6, 151.5.
Oxidation of diol to keto-aldehyde 70
A total of 600 mg of Cr03 (1 eq.) was added to a solution of 950 mg of pyridine
(2 eq.) in 15 ml of CH2CI2 and this was stirred at room temperature for 15 minutes. At this
point the PDC solution was added dropwise to a solution of 100 mg of the diol in 3 ml of
CH2CI2. The TLC of the reaction mixture was always checked about 5 minutes after
adding 3-4 drops of the PDC solution using 2, 4-dinitrophenylhydrazine as the indicator.
More PDC was added until the reaction was complete according to TLC. At this point the


118
involves using a suitable protected glycosyl bromide or chloride as the glycosyl donor. The
reaction is facilitated by using various heavy metal salts as activating agents. These metals
technically do not function as catalyst since an equivalent amount is needed. Some of the
more common heavy metal salts include; AgOTf, Ag20, AgC104, AgNOs, Hg(CN)2, and
HgBr2. The silver salts are the strongest activators with AgOTf being the strongest. In
some cases an acid scavenger may also be used and examples of these include HgO,
CdC03, and s-collidine.
The trichloroacetimidate-mediated glycosylation was announced in 1980 as an
alternative method to the classical Koenigs-Knorr procedure and now appears to be one of
the most ideal glycosylation protocol (Toshima & Tatsuta, 1993). This method involves
using a suitably protected anomeric trichloroacetimidate as the glycosyl donor. This donor
is prepared by condensing a protected 1-hydroxy sugar with trichloroacetonitrile in the
presence of base. Depending on which base is used, one can prepare the a or (3 epimer
using kinetic or thermodynamic control. This is not possible with halides. This reaction is
smoothly promoted by the catalytic use of BF3-Et20, TMSOTf, CCI3CHO, PTSA, and
ZnBr2.
The sulphoxide method is the newest of the three methods being first reported in
the literature in 1989 (Kahne et al., 1989). This method involves the use of a suitably
protected glycosyl sulphoxide as the glycosyl donor (usually a phenyl sulphoxide), an
equimolar amount of triflic anhydride as the activator, and 2, 6-di-tert-butyl-4-
methylpyridine as an acid scavenger. The stereochemistry of the resulting glycoside can be
controlled by varying the reaction solvent. This method has been shown to be very


Ill
H2, cat.
v
Figure 4-1: Synthesis of Ionizable Analogues


65
times with water and was dried with sodium sulfate. Since minor impurities were also
present in the mixture after workup a silica column was ran using 15-30% ethyl acetate in
ligroin as the solvent. A total of 700 mg of the product alcohol was obtained as a clear
yellowish liquid. !H NMR S: 3.84 (s, 6H, OCH3), 4.32 (d, 5.1Hz, 2H, 1-H), 6.56 (d,
8.4Hz, 2H, m-Ar), 6.75-6.92 (m, 2H, 2-H, 3-H), 7.15 (t, 8.1Hz, 1H, p-Ar).13C NMR 5:
55.6, 65.4, 103.8, 113.9, 121.7, 128.2, 132.6, 158.4.
Trans-2, 6-dimethoxycinnamaldehyde (93)
A total of 4.0 g of Cr03 was dissolved in 6.5 ml of pyridine at 0 C and stirred until
a reddish orange solid formed (PDC). At that point 561 mg of 92 was added in 1.0 ml of
acetone and the reaction was stirred for 4 hours at room temperature. At this point the
TLC showed only a small amount of starting material remaining so the reaction was
stopped. All physical and chemical properties matched that of the natural product.


108
been the focus of much research and various methods have been developed. A variety of
water-soluble analogues have been developed which contain esterase or phosphatase-
cleavable pro-moieties. However, these prodrugs are liable to exhibit unstable efficacy
because of variation in the enzymatic activity amoung patients. Therefore it would be very
advantageous to develop non-prodrugs of paclitaxel with satisfactory stability in vivo, high
water-solubility, and potent antitumor activity. This laboratory has studied two possible
approaches to this type of analogue and they are discussed below.
Synthesis of Analogues Starting from
10-Deacetyl Paclitaxel-7-Xyloside
It was discussed in Chapter 2 that 10-deacetyl paclitaxel-7-xyloside is actually the
most predominate taxane found in the bark of Taxus brevifolia occurring in a yield of
0.1% which is 2.5 times as much as paclitaxel. Other xylosides (10-deacetyl paclitaxel-C-
7-xyloside and 10-deacetyl cephalomannine-7-xyloside) are also found in high yields and
can be converted into 10-deacetyl paclitaxel-7-xyloside by modification of the amide
function. Although these xylosides have not been reported to a great extent in other Taxus
species, this does not mean they are not present since most published isolation procedures
are not ideal for obtaining these xylosides. It has also been reported that these xylosides
inhibit the in vitro disassembly of microtubules from mammalian brain at lower
concentrations than paclitaxel (Lataste et al., 1984). This is shown in Table 4-1. In light of
this fact and the assumption that these xylosides would undoubtedly have greater water-
solubility than paclitaxel, it is interesting that no reports have seriously examined the
possibility of using the xylosides as alternatives to paclitaxel clinically. One reason for this


94
(d, 7.2Hz, 2H, o-NBz), 8.10 (d, 6.9Hz, 2H, o-Bz). 13C NMR 5: 10.9, 14.9, 21.2, 22.4,
26.4, 32.5, 35.5, 43.0, 47.3, 55.2, 55.6, 72.0, 73.1, 73.8, 76.2, 78.5, 79.9, 80.5, 81.7,
83.4, 127.0, 127.1, 128.4, 128.7, 128.8, 129.1, 130.1, 130.2, 131.2, 132.1, 133.5, 133.9,
137.7, 144.2, 166.7, 167.4, 170.9, 172.6, 199.5.
Regioselective Nitrations of 10-Deacetyl Paclitaxel-7-p-Xyloside
10-Deacetyl paclitaxel-7-(3-xyloside 1.0 g was dissolved in 12 ml of
dichloromethane and was cooled to 0 C. To this was added 6 ml of a 5 : 1 mixture of
acetic anhydride and concentrated nitric acid which was also cooled to 0 C and the
reaction mixture was stirred in an ice bath for 5-10 minutes. The reaction was worked up
in the normal way and analyzed by TLC. This analysis showed many product spots
however some seemed to be more predominate than others. Initially, a crude silica column
was ran on this mixture using 0% methanol and 5% acetone in dichloromethane - 5%
methanol and 15% acetone in dichloromethane as the solvent. The fractions from this
column were combined into three groups; fast, medium, and slow in elution order. The
fast moving group contained some completely nitrated product, one major product that
was slower than the completely nitrated one, and a couple of minor products. This
material was ran on a silica column using 40% ethyl acetate in ligroin 60% ethyl acetate
in ligroin as the solvent. A total of 116 mg of the major product was obtained and
crystallized from dichloromethane. This product was determined to be 10-deacetyl
paclitaxel-7-(3-xyloside-2, 3, 4, 10-tetranitrate ester (108). The middle group from the
initial crude column contained many minor products and further separation was not
attempted. The slowest group from the initial column contained three major products as


136
Synthesis of Tetraacetyl Glucose, Phenyl Sulfoxide
A total of 250 mg of the phenyl thioglucoside was dissolved in 2 ml of CH2CI2 at
0 C and an equivalent of mCPBA was added and the mixture was stirred at 0 C for 1
hour. The mixture was then diluted with 0.1 N NaOH and CH2C12 and partioned. The
organic layer was washed with water, dried with Na2S04, and evaporated at which time
the product crystallized. After drying under reduced pressure the yield was 226 mg.
Yellowish white crystalline powder. The product existed as a almost equal mixture of a
and (3 anomers with anomeric carbon signals at 89.8 and 92.2 ppm.
Preparation of Tetraacetyl Benzyl P-Glucoside by the Koenigs-Knorr Method
A total of 300 mg of acetobromoglucose and 80 mg of benzyl alcohol was
dissolved in 2 ml of CH2C12 and excess ZnCl2 was added (100 mg). The reaction was
stirred at room temperature for 18 hours, the solid was filtered off and the filtrate was
diluted with water and CH2C12 and partitioned. The organic layer was washed with water,
dried with Na2S04, evaporated, and the residue was ran on a silica column using 30-40%
ethyl acetate in ligroin. A total of 216 mg of crystalline product was upon fraction
evaporation. White crystalline powder, 'HNMR8: 2.00 (s, 3H, OAc), 2.01 (s, 3H, OAc),
2.02 (s, 3H, OAc), 2.11 (s, 3H, OAc), 3.68 (m, 1H, 5-H), 4.18 (dd, 2.1, 12.0Hz, 1H, 6-
H), 4.28 (4.8, 12.0Hz, 1H, 6-H), 4.55 (d, 7.5Hz, 1H, 1-H), 4.63 (d, 12.3Hz, 1H,
OCH2Ar), 4 90 (d, 12.3Hz, 1H, OCH2Ar), 5.04-5.21 (m, 3H, 2-, 3-, 4-H). 13C NMR 5:
20.5, 20.6, 61.9, 68.4, 70.7, 71.2, 71.8, 72.8, 99.2, 127.7, 128.0, 128.4, 136.6, 169.2,
169.3, 170.2, 170.6.


40
Table 2-1 continued
HorC#
Compd. 81
Compd.79
Compd.80
14p
2.26 m, ****
2.00 m, ****
2.40 dd (7.2, 14.4Hz),
15
****; 75 2
Hi*** 7^ 7
**** 752
16
1.17 s, 27.3
1.15 s, 27.7
1.22 s, 25.5
17
1.20 s, 24.9
1.07 s, 25.4
1.18 s, 27.4
18
1.87 s, 11.5
2.01 s, 11.9
2.08 s, 13.8
19
1.77 s, 13.0
1.68 s, 12.4
1.91 s, 13.6
20a
4.49 d (7.2Hz),
74.3
4.50 d (7.7Hz), 74.5
4.56 d (7.2Hz), 74.7
20(3
4.41 d (7.2Hz),
4.41 d (7.7Hz), ****
4.45 d (7.2Hz), ****
q-Bz
**** 129 3
**** 229 0
**** 229 4
o-Bz
7.92 d (7.2Hz),
129.5
7.86 d (7.8Hz), 129.4
8.07 d (7.2Hz), 129.8
m-Bz
7.43 t (7.8Hz),
128.2
7.43 t (7.8Hz), 128.6
7.45 t (8.1Hz), 128.3
p-Bz
7.56 t (7.5Hz),
133.1
7.55 t (7.8Hz), 133.3
7.58 t (7.2Hz), 133.2
HorC#
Compd. 81
Compd. 79
Compd. 80
OCOCH3
1.65 s, 1.82 s, 2.03
s, 2.13 s, 2.14 s,
20.5, 21.0, 21.6,
21.7, 21.9
1.74 s, 2.02 s, 2.02 s,
2.08 s, 2.12 s, 20.6,
21.3, 21.4, 21.6, 22.0
2.01 s, 2.05 s, 2.11 s, 2.13
s, 20.9, 21.5, 21.6, 21.8
c=o
166.2, 167.8,
168.9, 169.7,
170.1, 170.3
163.9, 168.9, 169.0,
169.7, 170.2, 170.4
166.8, 169.0, 169.6, 170.3,
170.4
obtained at -10 C and the structure was determined to be that which is shown and which
was previously isolated by another group from Taxus chinensis var. Maiei and given the
name taxchinin L (Tanaka et al., 1996). It has been reported in the literature that
abeotaxanes which contain a C-9 benzoate and a C-10 hydroxyl group usually give proton
spectra in which the peaks are broad and rounded (Rao & Juchum, 1998). This is because


47
1H, 6-Ha), 3.05 (d, 6.0Hz, 1H, 3-H), 4.16 (d, 8.1Hz, 1H, 20-Hp), 4.31 (d, 8.1Hz, 1H,
20-Ha), 4.43 (m, 2H, 7-H, 9-H), 4.95 (d, 8.4Hz, 1H, 5-H), 5.75 (d, 5.7Hz, 1H, 2-H),
6.17 (t, 6.9Hz, 1H, 13-H), 6.19 (d, 10.8, 1H, 10-H), 7.48 (t, 7.8Hz, 2H, m-Bz), 7.61 (t,
7.5Hz, 1H, p-Bz), 8.09 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 12.5, 14.8, 21.2, 21.3, 22.6,
22.8, 28.3, 35.5, 38.0, 43.1, 45.0, 47.2, 69.8, 73.3, 73.7, 74.0, 76.9, 77.3, 78.8, 82.2,
84.1, 128.6, 129.4, 130.1, 133.6, 135.0, 139.7, 167.0, 169.3, 170.4.
Acetylation of ip-Hydroxy-7-Deacetyl Baccatin I
A total of 30 mg of ip-hydroxy-7-deacetyl baccatin I (37) was dissolved in 1 ml of
acetic anhydride and 1 ml of pyridine. This mixture was stirred at room temperature for 18
hours and then water was added to the mixture. Sodium bicarbonate was added slowly
until no further evolution of C02 was observed. The aqueous mixture was then extracted
twice with dichloromethane and the combined organic layers were washed with 0.1 N
NaOH, 0.1 N HC1, and water successively and dried with sodium sulfate. The
dichloromethane was evaporated and the product was crystallized from diethyl ether and
ligroin to yield 22 mg of acetylated ip-hydroxy-7-deacetyl baccatin I which was identical
in every way to lp-hydroxy baccatin I (35).
Isolation of Taxamairin A (38) from Taxus brevifolia
A total of 27 g of semi-pure paclitaxel, which had crystallized from reverse-phase
fractions, was dissolved in 200 ml of dichloromethane and applied to a silica column with
225 g of 240 mesh silica gel. Solvent was pumped through with an Eldex Laboratories
metering pump model B-100-S-4 at a pressure not exceeding 25 psi. The beginning
solvent was 2 : 1 dichloromethane : ligroin, then 3 : 1 dichloromethane : ligroin, followed


71
respectively (Gueritte-Voegelein et al., 1986). The reaction of 10-deacetyl paclitaxel
likewise gave the 10-mononitrate (102) and the 7, 10-dinitrate (103), again showing the
10-hydroxyl to be the most reactive and the 2-hydroxyl the least (10-OH > 7-OH > 2-
OH). With acetylation the 2-hydroxyl is the most reactive followed by the 7- and 10-
hydroxyls respectively (Kingston et al., 1982). Finally, 10-deacetyl paclitaxel-7-(3-xyloside
was tested and because of the five available hydroxyls many products were observed on
TLC and only the major products were isolated. These included the 2-mononitrate (105),
the 3-mononitrate (106), the 4-mononitrate (107), and the 2, 3, 4, 10-tetranitrate
(108) (Figure 3-2). This result indicates that the sugar hydroxyls are about equally reactive
and more reactive than the 10-hydroxyl, with the 2-hydroxyl being again the least reactive
(2, 3, 4 > 10 > 2). Although no acetylation studies have been performed on the
xylosides, this lab has shown that the 2-hydroxyl is more reactive in standard acetylation
conditions than either the sugar hydroxyls and the 10-hydroxyl is the least reactive (Figure
3-3).
At this point it should also be mentioned that the 10-deacetyl paclitaxel (112) used
in this work was not isolated directly from biomass but was actually converted from 10-
deacetyl paclitaxel-7-(3-xyloside (110). This conversion involves oxidizing the xyloside to
the dialdehyde (111) and then cleaving the dialdehyde with phenylhydrazine to give the
desired product and the corresponding phenylhydrazones (Figure 3-4) (Rao, 1997). In
conclusion, it has been shown that nitrate esters of taxanes can be formed under mild
conditions and in many cases regioselectivity is shown. In view of this work it is


82
O
Figure 3-10: Oxidation of Paclitaxel-7-Nitrate Ester
in DMF or acetonitrile with all bases tried including NaN3, NaOAc, NaOBz, triethylamine,
and hydroxide, with hydroxide giving the fastest reaction. However this reaction did not
take place when using dichloromethane as solvent with NaN3. No other intermediate
products were observed on TLC as loss of the intermediate keto-ester seemed to coincide
with formation of the final products. In order to determine that this reaction was base
catalyzed the intermediate keto-ester was subjected to three conditions; acetonitrile with
dilute hydroxide added, neet acetonitrile, and acetonitrile with dilute HC1 added. This
study showed that after 24 hours the basic solution was 85-90% decomposed to the final


114
O
+
CH3-NO2
154
CH2CI2, Et3N
T
O
Figure 4-3: Condensation with Nitromethane
When using aromatic amines however this problem was not encountered and a
smooth reaction occurred. The dialdehyde (158) was added with 5 eq. of amine in 2 : 1
CH3OH and acetic acid and an excess of NaCNBH3 was added and stirred for 1.5 hours.


59
7.3Hz, 1H, 2-H), 5.97 (d, 10.1Hz, 1H, 9-H), 7.45 (t, 7.5Hz, 2H, m-Bz), 7.57 (t, 6.9Hz,
1H, p-Bz), 7.99 (d, 7.5Hz, 2H, o-Bz). 13C NMR (-10 C) 5: 11.3, 13.9, 21.6, 21.8, 22.4,
25.2, 27.2, 34.3, 39.2, 42.9, 43.4, 66.3, 66.7, 68.4, 69.9, 74.9, 76.2, 77.5, 79.3, 80.7,
85.3, 128.2, 129.7, 129.9, 133.0, 137.5, 146.0, 167.7, 170.7, 170.8, 171.6.
10-Deacetyil baccatin III (25)
This compound was eluted with 10% acetone in dichloromethane and a total of
112 mg was crystallized from diethyl ether and ligroin. Spectral properties were identical
to those reported in the literature (Dennis et al., 1988).
Ponasterone A (84)
This compound was eluted with 10% acetone and 2% methanol in
dichloromethane and a total of 83 mg was crystallized directly from the eluting solvent.
Spectral properties were identical to those reported in the literature (Miller et al., 1982).
10-Deacetyl paclitaxeI-7-J3-xyloside (28)
This compound was eluted with 5% methanol and 10% acetone in
dichloromethane and a total of 79 mg was obtained as an amorphous solid. Spectral
properties were identical to those reported in the literature (Senilh et al., 1984).
Trans-2, 6-cIimethoxy cinnamaldehyde (85)
This compound was eluted with 25% ethyl acetate in ligroin and a total of 27 mg
was obtained as an glassy solid. UV A,max: 313 nm. IR (KBr): 3100, 3000-2940, 2810,
2740-2700, 1660, 1605-1585, 1475, 1260, 1140, 1100-1080, 970, 840, 725 cm'1. EIMS
m/z: 192 (30%, M+), 161 (100%), 149 (17%), 91 (15%). :H NMR 6: 3.90 (s, 6H,
OMe), 6.58 (d, 8.4Hz, 2H, m-Ar), 7.17 (dd, 7.8, 16.0Hz, 1H, 2-H), 7.33 (t, 8.4Hz, 1H, p-


87
yield 492 mg. White crystalline powder, mp 166-168 C, IR (KBr) 3450, 1725, 1650,
1365, 1270, 1230, 1065, 835, 700 cm'', Anal. Calc, for C47H5iN3018: C 58.69; H 5.31; N
4.37. Fd. C 58.83, H 5.15, N 4.11. 'H and 13C NMR see Table 1.
10-Deacetyl baccatin ni-7, 10, 13-trinitrate ester (97)
This compound was prepared starting with 500 mg of 10-deacetyl baccatin III and
following a procedure identical to that of paclitaxel-7, 2~dinitrate. In this case however,
the starting material does not initially dissolve in dichloromethane, but after reacting for a
few minutes all material goes into solution. A total of 480 mg were crystallized from
diethyl ether and ligroin. White crystalline powder, mp 159-161 C, Anal. Calc, for
C29H33N3O16: C 51.25; H 4.90; N 6.18. Fd. C 51.63; H 5.25; N 5.83. *H and I3C NMR
see Table 1.
10-Deacetyl paclitaxel-7, 10, 2-trinitrate ester (101)
This compound was prepared starting with 500 mg of 10-deacetyl paclitaxel and
following a procedure identical to that of paclitaxel 7, 2-dinitrate. A total of 506 mg of
product was crystallized from diethyl ether and ligroin. White crystalline powder, mp 159-
162 C, Anal. Calc, for C45H46N4O19 + H20: C 56.02; H 5.01; N 5.81. Fd. C 56.07; H
4.91; N 5.64. 'Hand 13C NMR see Table 1.
10-Deacetyl paditaxeI-7-P-xyloside-l, 2, 3, 10, 2-pentanitrate ester (104)
This compound was prepared starting with 500 mg of 10-deacetyl paclitaxel-7-p-
xyloside and following a procedure identical to that of paclitaxel-7, 2-dinitrate. As with
10-deacetyl baccatin III, all material went into solution only after the reaction had
proceeded for a few minutes. A total of 512 mg of product was crystallized from diethyl


60
At), 7.93 (d, 16.0Hz, 1H, 3-H), 9.64 (d, 8.1Hz, 1H, 1-H). 13C NMR 5: 55.8 x 2 (OMe),
103.6 (m-Ar), 112.1 (q-Ar), 131.6 (2-C), 132.6 (p-Ar), 144.5 (3-C), 160.5 (o-Ar), 196.4
(1-C).
a-Conidendrin (86)
This compound was eluted with 50% ethyl acetate in ligroin and a total of 38 mg
was crystallized from diethyl ether and ligroin. White crystalline powder, mp 257-259 C,
EIMS m/z: 356 (100%, M+), 255 (13%), 241 (26%), 137 (14%). H NMR 5: 2.5 (m,
2H), 2.7-3.1 (m, 1H), 3.73 (s, 3H), 3.78 (s, 3H), 3.9-4.2 (m, 4H), 6.26 (s, 1H), 6.53 (d,
2.0Hz, 1H), 6.58 (d, 8.0Hz, 1H), 6.62 (s, 1H), 6.73 (d, 8.0Hz, 1H). 13C DEPT NMR 5:
29.3 (CH2), 41.9 (CH), 47.5 (CH), 49.9 (CH), 55.9 (CH3), 60.0 (CH3), 71.9 (CH2), 110.1
(CH), 111.3 (CH), 114.6 (CH), 115.1 (CH), 121.5 (CH), 126.3(C), 131.7(C), 134.0(C),
144.2 (C), 144.9 (C), 145.5 (C), 147.0 (C), 177.0 (C).
1-Deoxy baccatin IV (87)
This compound was eluted with 50% ethyl acetate in ligroin and a total of 57 mg
was crystallized from diethyl ether and ligroin. White crystalline powder, mp 262-264 C,
'HNMR8: 1.12 (s, 3H, 17-H), 1.54 (s, 3H, 19-H), 1.78 (m, 1H, 1-H), 1.79 (s, 3H, 16-
H), 1.88 (m, 1H, 14-HJ3), 1.91 (m, 1H, 6-Hp), 1.97 (s, 3H, 18-H), 2.02 (s, 3H, Oac), 2.05
(s, 3H, Oac), 2.08 (s, 3H, Oac), 2.10 (s, 3H, Oac), 2.16 (s, 3H, Oac), 2.17 (s, 3H, Oac),
2.41 (m, 1H, 14-Ha), 2.50 (s, 1H, 6-Ha), 2.87 (d, 1H, 3-H), 4.19 (d, 7.5Hz, 20-HP),
4.51 (d, 8.1Hz, 1H, 20-Ha), 4.99 (d, 8.7Hz, 1H, 5-H), 5.52 (m, 2H, 2-H, 7-H), 5.91 (m,
2H, 9-H, 13-H), 6.15 (d, 11.4Hz, 1H, 10-H). 13C NMR 5: 12.7, 14.9, 20.7, 20.8, 20.9,


125
A few reactions of the trichloroacetimidate type were also investigated with
paclitaxel and various analogues but no favorable reactions occurred. Unfortunately due to
time constraints and other factors (death of professor) this area of research was ceased.
The author does believe that if more controlled conditions were used (dry solvents,
nitrogen atmosphere, low temperatures, less amounts of Lewis acids, etc.) these reactions
could be successful. Also the sulphoxide method may be better than the two that were
investigated and deserves attention that the author could not give it.
L1210 Cytotoxicity of Analogues
The LI 210 assay is commonly used to test compounds for their cytotoxicity and is
a screen for compounds with anticancer activity. These murine leukemia cells have a very
rapid doubling time of about 12 hours. Thus, they provide a convenient and relatively
reliable means for the determination of the cytotoxicity of many compounds (Thayer et al.,
1971).
The following compounds were chosen to be tested on the L1210 assay: paclitaxel,
10-deacetyl paclitaxel-7-xyloside, malonic acid condensation product 152, nitromethane
condensation product 155, reductive amination products 160-164, and the reduced diol
157. The ionizable compounds were tested in their neutral form. Table 4-2 shows the
results of the assay.
As Table 4-2 shows, there was a wide range of activity in this series of
compounds. Although none of the compounds tested is nearly as active as paclitaxel, 4 of
these showed greater activity than 10-deacetyl paclitaxel-7-xyloside. Of the reductive


113
generates three stereocenters thus a variety of isomers are possible. The major product
was isolated and based on the mass spectrum and LlC NMR it was concluded that this was
the desired product with unknown stereochemistry (155). Overlapping peaks in the 'H
NMR spectrum made it impossible to determine cis/trans configuration based on coupling
constants. It is not unreasonable to assume that the isolated major product is the all-trans
equatorial conformer since that would be the thermodynamically favored product.
Aside from having a greater number of nucleophiles to choose from, the reductive
amination procedure also offers the advantage of not generating asymmetric carbons.
Indeed this reaction actually converts two asymmetric carbons into methylene carbons.
The reducing agent chosen for these reactions was sodium cyanoborohydride. This
reagent has the advantage over sodium borohydride since it can be used in mildly acidic
conditions therefore no loss of the side chain occurs as with sodium borohydride. Also this
reagent has been shown to perform ideally in reductive amination reactions.
Initially the goal was to use a variety of amines, both aromatic and aliphatic, as
well as amino acids in these reactions. Attachment of di- and tripeptides was envisioned as
well. Unfortunately aliphatic amines including common amino acids did not condense with
the dialdehyde under the conditions used. Instead, the dialdehyde (156) function was
reduced to the corresponding diol (157) (Figure 4-4), and this occurrence was probably
due to the increased steric hindrance at the sp" a-carbon as opposed to the sp2 a-carbon
of an aromatic amine. Although this problem may have been overcome by varying the
conditions, time did not allow for this investigation.


11
Figure 1-7: Holton Synthesis of Paclitaxel
by conformational control exerted by the C-10a group. This scheme is shown in Figure 1-
6. The synthesis of this starting ketone begins with camphor (18). Camphor is converted
to P-patchouline (19) and its epoxide also known as Patchino (20). Patchino is then
converted to epoxide 21 and then rearranged to diol 22 which is followed by epoxy
alcohol fragmentation to give the starting ketone 23 as is shown in Figure 1-7.


130
p-Aminobenzoic acid product (160)
Grayish white crystalline powder, 478 mg, mp 198-200 C, UV Aax (CH3OH):
228, 300 nm, FABMS m/z: 1017 (M+l), 794, 754, 715, 714, 610, 206, 105. *H NMR 5
(some DMSO): 1.11 (s, 3H, 17-H), 1.21 (s, 3H, 16-H), 1.80 (s, 3H, 19-H), 1.81 (m, 1H,
14-Hp), 1.89 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.38 (m, 1H, 14-Ha), 2.43 (s, 3H, 4-
OAc), 2.76 (m, 1H, 6-Ha), 3.07 (dd, 4.5, 12.0Hz, 1H, 2-H), 3.27 (m, 2H, 2-, 4-H),
3.56 (br s, 1H, 4-H), 3.69 (m, 1H, 5-H), 3.90 (d, 6.6Hz, 1H, 3-H), 4.02 (m, 1H, 5-
H), 4.18-4.29 (m, 3H, 7-, 20a-, 20P-H), 4.56 (br s, 1H, 1-H), 4.70 (br s, 1H, 2-H),
4.93 (d, 9.3Hz, 1H, 5-H), 5.14 (s, 1H, 10-H), 5.65 (d, 6.6Hz, 1H, 2-H), 5.76 (d, 8.4Hz,
1H, 3-H), 6.22 (t, 8.1Hz, 1H, 13-H), 6.85 (d, 8.1Hz, 2H, o-Ar), 7.29-7.61 (m, 11H, m,p-
Bz, m,p-NBz, o,m,p-Ph), 7.86 (d, 7.2Hz, 2H, o-NBz), 7.93 (d, 8.4Hz, 2H, m-Ar), 8.12
(d, 7.2Hz, 2H, o-Bz), 8.13 (d, 8.4Hz, 1H, N-H). 13C NMR 5 (some DMSO): 10.3, 13.8,
20.5, 22.2, 26.2, 35.2, 35.3, 42.7, 45.6, 46.1, 50.4, 55.2, 56.5, 61.0, 70.9, 73.5, 74.0,
74.5, 75.9, 77.6, 80.0, 80.2, 83.7, 98.3, 113.2, 120.4, 126.5, 127.0, 127.2, 127.9, 128.1,
128.1, 129.4, 129.6, 130.9, 131.1, 132.9, 133.9, 135.7, 137.9, 138.5, 153.0, 165.9, 166.7,
167.9, 169.9, 172.2, 209.6
p-Salicylic acid product (161)
Grayish brown crystalline powder, 378 mg, mp 182-184 C, UV \max (CH3OH):
230, 311 nm, FABMS m/z: 1034 (M + 2), 1033 (M + 1), 748, 730, 222, 204, 105. *H
NMR 5 (some DMSO): 1.11 (s, 3H, 17-H), 1.20 (s, 3H, 16-H), 1.79 (s, 3H, 19-H), 1.80
(m, 1H, 14-Hp), 2.02 (m, 1H, 6-HP), 2.25 (m, 1H, 14-Ha), 2.43 (s, 3H, 4-OAc), 2.74 (m,
1H, 6-Ha), 3.08 (dd, 4.8, 12.6Hz, 1H, 2-H), 3.26 (m, 3H, 2-, 4-, 4-H), 3.69 (m,


LIST OF TABLES
Table page
1-1 Physical and Chemical Properties of Paclitaxel 3
2-1 'H and 13C NMR Values for Related Abeo-Taxanes 39
3-1 *14 and L"C NMR Values for Completely Nitrated Taxanes 68
4-1 ID50 Values of Paclitaxel and Xylosides in Tubuline Assay 109
4-2 L1210 Cytotoxicity of Paclitaxel and Analogues 126
viii


late 1960s, NMR was relatively primitive compared to the sophisticated instrumentation
and procedures now available. Some of the physical and chemical properties of paclitaxel
are shown in Table 1-1 and the *H NMR spectrum is shown in Figure 1-2.
Table 1-1: Physical and Chemical Properties of Paclitaxel
1.) Needles from 50% aqueous methanol or ether
2.) mp 213-216 C
3.) [q]D20 -49,6 (MeOH)
4.) Unstable towards mineral acid and base
5.) Forms mono and diacetate
6.) Analysis Caled, for C47H51NOi4: C, 66.11; H, 6.20; N, 1.64
Found: C, 65.98; H, 6,10; N, 1.57. Required m/z 853. Found m/z 853
7.) UV ^ax (MeOH) 227 nm (s 29,800)
It was evident by this time that paclitaxel probably contained the taxane skeleton.
A number of taxane derivatives had been reported in previous literature. It was evident
that paclitaxel was more complex than previously reported taxanes since its molecular
weight from high resolution mass spectrometry was C^HsiNOu, corresponding to a
molecular weight of 853. The evidence then indicated that paclitaxel was comprised of a
taxane nucleus to which an ester was attached, as preliminary experiments indicated that
an ester was easily cleaved from the rest of the molecule. Attempts were made to prepare
crystalline halogenated derivatives of paclitaxel, however none had properties suitable for
x-ray analysis. Paclitaxel was therefore subjected to a mild base catalyzed methanolysis at
0 C, which yielded a nitrogen containing a-hydroxy methyl ester, C17H17NCX1, a tetraol,
C29H36Oio, and methyl acetate. The methyl ester thus obtained by the mild methanolysis
procedure was converted to a parabromobenzoate ester (2) and characterized by x-ray
analysis as C24H2oBrN05 with the structure shown in Figure 1-2. The ester may be


LIST OF FIGURES
Figure Page
1-1 Structure of Paclitaxel 2
1-2 11 NMR Spectrum of Paclitaxel 4
1 -3 Halogenated Products of Methanolysis Used for X-Ray Crystallography .... 5
1-4 Neutral and Alkaline Oxidation of Paclitaxel 7
1-5 Nicolaou Synthesis of Paclitaxel 9
1-6 Holton Synthesis of Paclitaxel 10
1-7 Holton Synthesis of Paclitaxel 11
1-8 Structure-Activity Relationships 14
2-1 Structure of Major Taxanes 16
2-2 Reverse-Phase Isolation of Taxanes 19
2-3 Ozonolysis of Cephalomannine/Paclitaxel Mixture 21
2-4 Compounds from the Bark of Taxus brevifolia 23
2-5 Acetylation of lJ3-Hydroxy-7-Deacetyl Baccatin I 23
2-6 Structure of Abeo-Abietane Diterpenoids 25
2-7 NOE Correlations of Taxamairin A 25
2-8 Retro synthetic Analysis of Taxamairin B 27
2-9 Literature Synthesis of Taxamairin B 28
ix


134
determined by 'H and ljC NMR spectroscopy. Selected 13C NMR signals 5: 61.4, 67.7,
67.8, 69.1, 69.7, 70.2, 72.6, 72.7, 89.0, 91.6.
Synthesis of la-Bromo-Tetraacetyl Glucose
A total of 3.0 g of glucose pentaacetate was dissolved in 6 ml of CH2CI2 and then
8 ml of 30% HBr in acetic acid was added and this was stirred at room temperature for 2
hours. At that point the mixture was diluted with 30 ml of CH2CI2 and 75 ml of ice water
and partitioned. The organic layer was separated and the aqueous layer was partitioned
again with CH2CI2. The combined organic layers were washed with ice cold NaHCO;,
solution three times and then dried with Na2S04. The solvent was removed under reduced
pressure and the residue was taken up in diethyl ether and ligroin and put in the freezer for
crystallization. After overnight crystals had formed and they were filtered, washed with
ligroin, and dried under reduced pressure. The yield was 2.20 g and this material was kept
in the freezer to avoid decomposition.
Synthesis of 1-Hydroxy-Tetraacetyl Glucose
A total of 2.0 g of acetobromoglucose was dissolved in 5 ml of acetonitrile and 1
ml of water was added as excess Hg(CN)2 and the mixture was stirred at room
temperature for 30 minutes at which time the reaction was complete. After filtering off the
Hg(CN)2 the acetonitrile was removed under reduced pressure and the residue was diluted
with water and CH2C12 and partitioned. The organic layer was separated and the water
layer was partitioned again with CH2CI2. The combined organic layers were then washed
with water, dried with Na2S04, and evaporated to yield 1.7 g of a syrup. Clear, colorless
syrup, product existed as a mixture of a and (3 anomers, but the a anomer is predominate.


100
solid (167 mg) which was determined to be paclitaxel-7-mononitrate ester (119). All NMR
spectra matched those of an authentic sample.
Acetylation of 10-Deacetyl Paclitaxel-7-p-Xyloside-2, 3, 4, 10, 2,-Pentanitrate
Ester
10-Deacetyl paclitaxel-7-p-xyloside-2, 3, 4, 10, 2-tetranitrate 600 mg was
dissolved in 25 ml of acetic anhydride and 120 mg of DMAP was added and the reaction
was stirred at room temperature for overnight. Water was added as was NaHC03 with
stirring until no further frothing was observed. The aqueous mixture was then extracted
three times with dichloromethane and this organic layer was washed with a saturated NaCl
solution, with water, and dried with Na2S04. The TLC analysis of the organic layer
showed two major products which were slightly faster moving than the starting material.
A couple of minor products were present but no starting material was seen. This material
was ran on a regular silica column using 25% 40% ethyl acetate in ligroin. A total of
172 mg of the faster spot was isolated and crystallized from acetone and ligroin. This
product was determined to be the di-enol acetate (122). A total of 146 mg of the slower
spot was isolated and crystallized from dichloromethane. This product was determined to
be the mono-enol acetate (121).
10-Deacetyl paditaxel -7-P-xyloside-2, 3, 4, 10-tetranitrate-di-enol acetate (122)
Clear colorless needles, mp 193-195C, UV (CH3OH): 229 nm, FABMS m/z:
1206 (M+ 1), 703, 613, 522. 'HNMRS: 1.13 (s, 3H, 17-H), 1.19 (s, 3H, 16-H), 1.77 (s,
3H, 19-H), 2.03 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.21 (s, 3H, OAc), 2.29 (m, 1H, 14-
HP), 2.30 (s, 3H, OAc), 2.52 (m, 1H, 14-Ha), 2.53 (s, 3H, OAc), 2.79 (m, 1H, 6-Ha),


93
10-Deacetyl paclitaxel-10-mononitrate ester (102)
White amorphous powder, Anal. Calc, for C45H48N2O15: C 63.08; H 5.65; N 3.27.
Fd. C 62.75; H 5.97; N 3.90. XH NMR 5: 1.14 (s, 3H, 17-H), 1.19 (s, 3H, 16-H), 1.61 (s,
3H, 18-H), 1.83 (s, 3H, 18-H), 1.98 (m, 1H, 6-Hp), 2.32 (m, 2H, 14-Ha,p), 2.38 (s, 3H,
4-OAc), 2.56 (m, 1H, 6-Ha), 3.75 (d, 6.9Hz, 1H, 3-H), 4.18, (d, 8.7Hz, 1H, 20-Hp),
4.21 (dd, 3.6, 5.4Hz, 1H, 7-H), 4.30 (d, 8.1Hz, 1H, 20-Ha), 4.79 (br s, 1H, 2-H), 4.90
(d, 8.7Hz, 1H, 5-H), 5.68 (d, 7.2Hz, 1H, 2-H), 5.75 (dd, 2.1, 8.7Hz, 1H, 3-H), 6.19 (t,
8.4Hz, 1H, 13-H), 6.41 (s, 1H, 10-H), 7.09 (d, 9.0Hz, 1H, NH), 7.35-7.53 (m, 10H, m-
Bz, o,m,p-Ph, m,p-NBz), 7.61 (t, 7.2Hz, 1H, p-Bz), 7.73 (d, 6.9Hz, 2H, o-NBz), 8.11 (d,
7.2Hz, 2H, o-Bz). 13C NMR 5: 9.6, 15.0, 21.6, 22.5, 26.4, 35.5, 36.6, 43.0, 46.2, 55.2,
58.4, 68.2, 71.5, 72.1, 73.1, 74.5, 76.5, 78.6, 81.0, 82.2, 84.1, 127.0, 127.1, 128.4, 128.7,
128.8, 129.0, 130.2, 130.7, 130.9, 132.0, 133.5, 133.8, 137.8, 143.8, 166.8, 167.3, 170.5,
172.6, 202.6.
10-Deacetyl paclitaxel-7,10-dinitrate ester (103)
White crystalline powder, mp 170-172 C, Anal. Calc, for C45H47N3O18: C 58.76;
H 5.37; N 4.57. Fd. C 59.11; H 5.27; N 4.49. !H NMR 5: 1.13 (s, 3H, 17-H), 1.20 (s,
3H, 16-H), 1.83 (s, 3H, 19-H), 1.87 (s, 3H, 18-H), 2.06 (m, 1H, 6-Hp), 2.36 (m, 2H, 14-
Ha, p), 2.41 (s, 3H, 4-OAc), 2.70 (m, 1H, 6-Ha), 3.91 (d, 6.9Hz, 1H, 3-H), 4.17 (d,
8.4Hz, 1H, 20-Hp), 4.34 (d, 8.4Hz, 1H, 20-Ha), 4.80 (br s, 1H, 2-H), 4.95 (d, 9.0Hz,
1H, 5-H), 5.67 (d, 6.9Hz, 1H, 2-H), 5.71 (dd, 7.2, 10.5Hz, 1H, 7-H), 5.77 (dd, 2.1,
8.7Hz, 1H, 3-H), 6.20 (t, 9.0Hz, 1H, 13-H), 6.33 (s, 1H, 10-H), 7.07 (d, 9.0Hz, 1H,
NH), 7.36-7.63 (m, 10-H, m-Bz, o,m,p-Ph, m,p-NBz,), 7.63 (t, 7.5Hz, 1H, p-Bz), 7.73


13
significant loss of activity; 4) one of the C-2 or C-3 polar functions can be removed
without significant effect, but the removal of both or interchange of their positions causes
dramatic loss of activity and; 5) the (2S, 3R) naturally occurring isomer is the most
active of the four possible isomers.
Concerning the southern hemisphere of the structure, this area is also very
important in terms of biological activity. First of all the oxetane ring is necessary for
activity. Structural and molecular modeling studies show that this 4-membered ring is
involved in a conformational lock of the diterpene skeleton and the C-13 side chain
through a pseudo chair conformation of ring C. The C-2 benzoyl group is also necessary
for activity as the C-2 debenzoyl paclitaxel showed little in vitro cytotoxicity; however,
some groups have shown that modified benzoyl groups or aryl acyl groups do retain the
cytotoxicity. The C-4 acetate is not as important as the C-2 benzoate in that if the acetate
is removed the activity is reduced only slightly.
The northern hemisphere is the least sensitive part of the paclitaxel structure. The
C-7 hydroxyl may be esterified, epimerized, or removed without significant loss of
activity. Specifically a xylosyl group at C-7 actually increases the activity in the tubulin
binding assay but leads to decreased activity in cell culture. Presumably there is a transport
problem associated with the xyloside that causes this decreased activity. The C-9 keto
group may be reduced which actually slightly improves activity and the C-10 acetyl or
acetoxy group may be removed without significant loss of activity.
It should also be said that contraction of the A ring does not reduce the tubulin-
disassembly inhibition activity very much in spite of the significant structural change


63
with stirring. This mixture was then refluxed for 4 hours at which time no starting material
remained. Thus about 10 ml of acetone was added to neutralize the remaining LAH and
this was stirred for 18 hours. The solvent was then removed by rotovap and the residue
was partitioned between water and diethyl ether. The organic layer was washed three
times with water and was dried with sodium sulfate. Upon removal of the solvent by
evaporation the product crystallized to yield 1.5 g of the product alcohol. Clear colorless
crystals, *H NMR 5: 3.83 (s, 6H, OCH3), 4.70 (s, 2H, CH2), 6.46 (d, 8.4Hz, 2H, m-Ar),
7.13 (t, 8.7Hz, 1H, p-Ar).
2, 6-Dimetlioxybenzaldehyde (90)
A total of 1.5 g of the alcohol (89) was dissolved in 10 ml of acetone and Jones
reagent was added dropwise while the TLC was monitored using 2, 4-
dinitrophenylhydrazine as the indicator. Quite suprisingly the aldehyde had a lower Rf
value than the alcohol. The reaction was continued until no alcohol starting material was
observed on TLC. The acetone was partially removed and the residue was partitioned
between water and diethyl ether. The organic layer was washed with water twice and dried
with sodium sulfate. Upon removal of the solvent under reduced pressure the product
crystallized to yield 1.18 g of the aldehyde product. Clear colorless crystals, H NMR 5:
3.85 (s, 6H, OCH3), 6.42 (d, 8.4Hz, 2H, m-Ar), 7.33 (t, 8.6Hz, 1H, p-Ar), 10.35 (s, 1H,
CHO).
Trans-2, 6-dimethoxycinnamic acid (91)
A total of 1.18 g of the aldehyde (90) and 1.38 g (2 eq.) of malonic acid was
dissolved in 4 ml of pyridine and a few drops of piperidine were added. This mixture was


135
Selected a anomer L,C NMR signals 5: 61.9, 67.0, 68.4, 69.8, 71.1, 90.0. Selected p
anomer 1jC NMR signals 8: 60.4, 68.4, 71.9, 72.2, 73.0, 95.4.
Synthesis of la-Trichloroacetimidate-Tetraacetyl Glucose
A total of 2.2 g of 1-OH-tetraacetyl glucose, 4.14 g (5 eq) of trichloroacetonitrile,
and 4.4 g (1.1 eq) of K2C03 was added to 4 ml of CH2C12 and stirred at room temperature
for 3 days. The mixture was then diluted with water and CH2C12 and partitioned. The
organic layer was separated and the water layer was partitioned again with CH2C12. The
combined organic layers were washed with water, dried with Na2S04, and evaporated to a
syrup. The product was sufficiently pure for further reactions and the yield was 1.96 g.
Clear colorless syrup, 13C NMR 8: 20.5, 20.6, 20.6, 20.7, 61.9, 67.1, 68.5, 69.9, 71.1,
90.0, 163.8, 169.7, 170.2, 170.3, 170.9.
Synthesis of Tetraacetyl Phenyl Thioglucoside
A total of 500 mg of acetobromoglucose, 200 mg (1.5 eq) of thiophenol, and 1.25
eq of KOH were mixed in methanol with the acetobromoglucose added last. The mixture
was stirred at room temperature for 20 minutes at which time TLC showed that the
reaction was complete. The product crystallized upon ceasing the stirring and the crystals
were filtered, washed with aqueous methanol, and dried under reduced pressure. The yield
was 410 mg White crystalline powder, mp 'H NMR 8: 1.99 (s, 3H, OAc), 2.02 (s, 3H,
OAc), 2.08 (s, 3H, OAc), 2.09 (s, 3H, OAc), 3.73 (m, 1H, 5-H), 4.20 (m, 2H, 6-H), 4.72
(d, 6.6Hz, 1H, 1-H), 4.98 (t, 10.2Hz, 1H, 2-H), 5.05 (t, 9.9Hz, 1H, 4-H), 5.23 (t, 9.3Hz,
1H, 3-H), 7.32 (m, 3H, o,p-Ar), 7.50 (m, 2H, rn-Ar). 13C NMR 8: 20.5, 20.7, 62.1, 68.2,
70.0, 74.0, 75.8, 85.7, 128.4, 128.9, 131.6, 133.1, 169.2, 169.3, 170.1, 170.5.


121
alcohol reaction giving the best results. These reactions were conducted at room
temperature in CH2C12 (Figure 4-7). The (3 orientation was determined based on the
anomeric carbon and proton chemical shifts as well as comparison to the acetylated
naturally occurring P-sitosterol-p-glucoside. Of the activators that were available,
Hg(CN)2, AgN03, and ZnCl2 were the best. None of the stronger silver Lewis acids were
available. The trichloroacetimidate method was also attempted with both of these
substrates using PTSA and BF3-Et20 as the activators and CH2C12 as solvent. The
reactions were performed at room temperature (Figure 4-8). Although a substantial
amount of hydrolysis of the donor did occur, it was obvious that this method was superior
giving greater yields in a shorter amount of time. The products formed also contained the
P orientation.
Since these results were encouraging, similar reactions were attempted with
paclitaxel. Unfortunately none of these reactions were successful. As mentioned earlier
paclitaxel and related taxanes are quite unstable to acids, both mineral and Lewis. Under
these conditions a variety of rearrangements have been documented. When the weaker
Lewis acids such as Hg(CN)2 were used in conjunction with the Koenigs-Knorr method
no reaction took place at all. This was probably due to the steric hindrance involved at the
C-2 and C-7 hydroxyls of paclitaxel. Therefore faced with this problem the use of
stronger Lewis acids would be required; however, when this was attempted with AgNO:,
and ZnCl2 rearrangements did occur. In most cases the rearranged product was not
identified but in the case of AgNCF the product was identified as the rearranged taxane
183 when paclitaxel-2-acetate (182) was used as the aglycone (Figure 4-9). Therefore, in


131
1H, 5-H), 3.90 (d, 7.2Hz, 1H, 3-H), 4.01 (m, 1H, 5-H), 4.16-4.29 (m, 3H, 7-, 20a-,
20J3-H), 4.54 (br s, 1H, 1-H), 4.70 (d, 3.0Hz, 1H, 2-H), 4.92 (d, 8.7Hz, 1H, 5-H), 5.14
(s, 1H, 10-H), 5.65 (d, 6.9Hz, 1H, 2-H), 5.76 (dd, 3.0, 8.7Hz, 1H, 6-Ar), 6.21 (t, 8.4Hz,
1H, 13-H), 6.29 (d, 2.1Hz, 1H, 2-Ar), 6.38 (dd, 2.1, 9.0Hz, 1H, 3-H), 7.29-7.55 (m,
10H, m-OBz, m,p-NBz, o,m,p-Ph), 7.61 (t, 7.2Hz, 1H, p-OBz), 7.71 (d, 9.0Hz, 1H, N-
H), 7.86 (d, 7.2Hz, 2H, o-NBz), 8.12 (d, 7.5Hz, 2H, o-OBz), 8.17 (d, 9.0Hz, 1H, 5-Ar).
13C NMR 6 (some DMSO): 10.3, 13.8, 20.5, 22.1, 26.1, 35.2, 35.3, 42.7, 45.6, 45.7,
49.9, 55.1, 56.5, 60.8, 70.9, 73.5, 74.0, 74.4, 75.9, 77.2, 80.0, 80.2, 83.6, 98.1, 100.1,
103.1, 105.5, 126.5, 127.0, 127.2, 127.9, 128.0, 128.1, 129.3, 129.6, 131.1, 131.2, 132.9,
133.8, 135.6, 137.8, 138.4, 155.1, 163.0, 165.8, 166.8, 169.9, 171.9, 172.2, 209.5.
m-Salicylic acid product (162)
Reddish brown crystalline powder, 392 mg, mp 173-175 C (dec.), UV 7,max
(CH3OH): 224, 343 nm, FABMS m/z: 1034 (M + 2), 1032 (M), 222, 204, 105, *H NMR
5: 1.10 (s, 3H, 17-H), 1.18 (s, 3H, 16-H), 1.82 (s, 3H, 19-H), 1.82 (m, 1H, 14-Hp), 1.83
(s, 3H, 18-H), 2.07 (m, 1H, 6-Hp), 2.31 (m, 1H, 14-Ha), 2.36 (s, 3H, 4-OAc), 2.72 (m,
1H, 2-H), 2.79 (m, 1H, 4-H), 2.86 (m, 1H, 6-Ha), 2.97 (m, 1H, 2-H), 3.00 (m, 1H,
4-H), 3.72 (m, 1H, 5-H), 3.89 (d, 5.7Hz, 1H, 3-H), 3.99 (m, 1H, 5-H), 4.20 (m, 1H,
7-H), 4.20 (d, 7.5Hz, 1H, 20-Hp), 4.29 (d, 7.5Hz, 1H, 20-Ha), 4.61 (br s, 1H, 1-H),
4.79 (br s, 1H, 2-H), 4.93 (d, 8.7Hz, 1H, 5-H), 5.28 (s, 1H, 10-H), 5.65 (d, 6.0Hz, 1H,
2-H), 5.78 (d, 9.0Hz, 1H, 3-H), 6.18 (br s, 1H, 13-H), 6.80 (d, 8.7Hz, 1H, o-Ar), 7.08
(d, 8.4Hz, 1H, m-Ar), 7.21 (br s, 1H, p-Ar), 7.37 (m, 1H, N-H), 7.33-7.49 (m, 10H, m-
Bz, m,p-NBz, o,m,p-Ph), 7.58 (t, 7.2Hz, 1H, p-Bz), 7.74 (d, 7.5Hz, 2H, o-NBz), 8.08 (d,


61
21.2, 21.4, 21.6, 22.6, 26.9, 31.3, 34.6, 37.9, 44.5, 45.6, 46.9, 69.0, 71.0, 71.1, 71.9,
75.4, 77.2, 81.0, 83.9, 133.3, 138.9, 169.2, 169.3, 169.7, 170.1, 170.2, 170.4.
Acetylation of Taxiflorine
Taxiflorine (77) 120 mg was dissolved in 2 ml of acetic anhydride and 1 ml of
pyridine was added. The mixture was stirred at room temperature for 18 hours and then
water was added to the mixture. Sodium bicarbonate was added slowly until no further
evolution of C02 was observed. The aqueous mixture was then extracted twice with
dichloromethane and the combined organic layers were washed with 0.1 N NaOH, 0.1 N
HC1, and water successively and dried with sodium sulfate. The dichloromethane was
evaporated and the product (112 mg) (81) was obtained as a glassy solid. For NMR data
see Table 2-1.
Oxidation of Taxiflorine
Taxiflorine (77) 50 mg was dissolved in 2 ml of acetone and a few drops of Jones
reagent was added and the mixture was stirred at room temperature for 2 hours. At this
time the acetone was partially evaporated and water was added. This aqueous mixture was
extracted twice with dichloromethane and the combined organic layers were washed with
0.1 N NaOH and then with water and then dried with sodium sulfate. After the
dichloromethane was evaporated the residue was crystallized from diethyl ether and ligroin
to yield 36 mg of the ketone product. White crystalline powder. For NMR data see Table
2-1.


103
7.2Hz, 2H, p-NBz), 7.87 (d, 7.8Hz, 4H, o-NBz), 8.99 (br s, 1H, N-H). 13C NMR 5:
127.9, 128.9, 133.1, 133.3, 166.4.
Baccatin IU-7-nitrate ester (126)
see compound 117
Synthesis of 2-Oxo Paditaxel-7-Nitrate Ester from Paclitaxel-7, 2-Dinitrate Ester
Paclitaxel-7, 2-dinitrate ester 200 mg was dissolved in 3 ml of DMF and 200 mg
of NaN3 was added. Immediately the solution turned a deep pink color while stirring at
room temperature. After 10 minutes water was added and the color dissipated and a white
solid precipitated from the solution. This solid was filtered and dried to yield 182 mg of
>95% keto-ester (124). White amorphous powder, UV X* (CH3OH): 230 nm, IR
(KBr): 2960, 1725, 1640, 1270, 1225, 1060, 1020, 830, 700 cm'1. FABMS: 897 (82%,
M+l), 614 (13%), 554 (35%), 307 (16%), 284 (55%), 210 (100%). For NMR see
below compound.
Synthesis of 2-Oxo Paclitaxel-7-Nitrate Ester from PaclitaxeI-7-Mononitrate Ester
Paclitaxel-7-mononitrate ester 200 mg was dissolved in 3 ml of acetone and a few
drops of 3 N Jones reagent was added. This solution was stirred at 60 C and checked by
TLC every 30-60 minutes and more Jones reagent was added as needed. After about 6
hours the TLC showed about a 40% conversion and the reaction didnt seem to proceed
any further so the acetone was evaporated and the residue was taken up in water. The
water layer was extracted with dichloromethane twice and the combined organic layers
were washed with water once. The dried organic layer was evaporated to dryness and the
residue was put on a silica column and eluted with 5-15% acetone in benzene. A total of


42
On elution with 10% acetone/dichloromethane additional amounts of 10-deacetyl
baccatin III were obtained; and with 2% methanol/10% acetone/dichloromethane the
polyhydroxylated steroid ponasterone A (84) was isolated (Figure 2-16). This compound
was previously isolated from Taxus brevifolia (Rao et al., 1996b). Finally, with 5-10%
methanol/10% acetone/dichloromethane 10-deacetyl paclitaxel-7-(3-xyloside was isolated
for the first time from Taxus floridana.
The fractions mentioned earlier that were eluted with dichloromethane were
concentrated and put on another silica column using 25% ethyl acetate/ligroin as the
starting mobile phase. At 30% ethyl acetate/ligroin trans-2,6-dimethoxy cinnamaldehyde
(85) was eluted; and it will be discussed later. Elution continued with 50% ethyl
acetate/ligroin, which gave the lignan a-conidendrin (86) followed by 1 -deoxy baccatin IV
(87); both of which have been previously isolated (Figure 2-16) (Miller et al., 1982;
Miller, 1980).
Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde
As mentioned above trans-2, 6-dimethoxy cinnamaldehyde was one of the
compounds isolated from Taxus floridana. Although this structure was determined quite
easily based on ]H and 13C NMR spectra, di-ortho oxy substituted C6-C3 compounds had
previously not been isolated from natural products. Thus 85 was synthesized to verify the
structure following Figure 2-17. Thus trans-2, 6-dimethoxy cinnamic acid (88) was
methylated with methanol and H2S04 followed by reduction to the alcohol (89) with LAH
in a total yield of about 55%. The alcohol was then oxidized to the aldehyde (90) with


55
starting material had been converted to a faster moving product the mixture was diluted
with water and diethyl ether and partitioned. The organic layer was washed with 0. IN HC1
and twice with water, dried over sodium sulfate, and concentrated. The product was
purified by silica chromatography using 15-20% ethyl acetate in ligroin as solvent. A total
of 387 mg of product (73) was obtained as a slightly yellow oil. Yellow oil, !H NMR 5:
1.21 (d, 6.9Hz, 6H, CH3), 1.24 (t, 7.5Hz, 3H, CH3), 2.60 (q, 7.8, 15.3Hz, 3H, CH3), 3.33
(quintet, 6.9Hz, 1H, CH), 3.79 (s, 3H, OCH3), 3.85 (s, 3H, OCH3), 6.60 (d, 1.5Hz, 1H,
Ar-H), 6.67 (d, 1.5Hz, 1H, Ar-H). 13C NMR 5. 15.6, 23.6, 26.7, 28.9, 55.6, 60.9, 109.3,
117.3, 140.0, 142.0, 144.1, 152.3.
Acetoxymethylation of 73
A total of 100 mg of 73 was added to 2.0 ml of 85% H3PO4 and to this was added
200 mg of paraformaldehyde and 0.5 ml of acetic anhydride. The mixture was stirred at
room temperature for 3 hours at which point the TLC showed most of the starting
material to be gone and a slower moving product had formed. The mixture was partitioned
between water and diethyl ether and the organic layer was washed twice with 0.1 N HC1
and twice with water, dried with sodium sulfate, and concentrated. The residue was
separated on a silica column using 15-25% ethyl acetate in ligroin as the solvent. A total
of 64 mg of the product (74) was isolated as a clear colorless oil. Clear colorless oil, H
NMR 5: 1.21 (t, 7.5Hz, 3H, CH3), 1.35 (d, 7.2Hz, 6H, CH3), 2.07 (s, 3H, OAc), 2.68 (q,
7.5, 15.0Hz, 2H, CH2), 3.22 (m, 1H, CH), 3.84 (s, 3H, OCH3), 3.86 (s, 3H, OCH3), 5.14
(s, 2H, OCH2), 6.66 (s, 1H, Ar-H). 13C NMR 5: 16.2, 21.0, 21.9, 26.8, 28.9, 55.5, 60.6,
60.7, 110.7, 122.9, 142.1, 146.6, 153.2, 171.2.


2
O
Figure 1-1: Structure of Paclitaxel
systems models in vivo and tumor cell lines. From these studies the stem bark extract of
the Pacific yew tree was shown to display cytotoxicity in the KB assay and also activity
against carcinosarcoma in rats and leukemia in mice. In connection with this NCI
screening program, Wall and his collaborators studied the in vitro bioassay guided
fractionation of the active extract and in 1969 paclitaxel was isolated and shown to be the
most active constituent of the extract. This isolation was carried out by extracting the
dried stem bark with 95% ethanol. The extract was then partitioned between water and 4 :
1 chloroform : methanol. The organic layer was evaporated to a solid and purified by a 3-
step Craig countercurrent distribution method which yielded paclitaxel in a yield of
0.004%, As soon as paclitaxel had been isolated in pure form, the structure of the
compound was investigated using available spectroscopic methods. Although methods for
ultraviolet, infrared, and mass spectrometry were at a reasonably advanced stage in the


This work is dedicated to my father whose untimely passing due to the disease that this
work addresses on April 28, 1998 has left a deep void in my life that will never be filled.
Although he would not be considered an educated man by most standards he taught me
more than any textbook or professor ever could. I hoped he could be here when this work
was completed but that was not to be. Nevertheless I hope that this accomplishment
would make him as proud of me as I was of him. I love and miss you daddy.


12
Although there have been many publications concerning the synthesis of the phenyl
isoserine side chain, the most common and that which was used in both of these methods
is using the P-lactam 6 in Figure 1-5.
Structure-Activity Relationships
During the past decade a tremendous amount of work has been performed to
determine what parts of the structure of paclitaxel are necessary to illicit biological activity
and reviews have been published (Commercon et ah, 1995; Chen & Farina, 1995;
Kingston, 1995; Ojima et ah, 1995; Georg et ah, 1995). The structure of paclitaxel can be
divided into 3 sections when discussing structure-activity relationships and these are: 1)
the N-benzoyl phenylisoserine side chain; 2) the southern hemisphere including C-2, C-4,
and the oxetane ring; and 3) the northern hemisphere including C-7, C-9, and C-10.
The side chain plays a major role in the biological functions of this antitumor agent
and without the side chain the resulting baccatin III is inactive. Protection of the C-2
hydroxyl group results in major loss of activity in tubulin assays but if the group is labile
(acetate) then the activity remains in cell culture presumably acting as a prodrug.
Structural modifications of the paclitaxel side chain have been reported by several groups.
These studies reveal a number of interesting features and important findings include the
following: 1) the C-3 amide group is critical although the amides aryl group may be
substituted by other aryl or alkyl groups; 2) the C-3 aryl group is required since
replacement by a methyl group reduces activity but, if larger alkyl groups are used, the
activity remains; 3) the C-3 bound nitrogen can be replaced by an oxygen atom without


50
4-Bromo-2, 2-dimethyl-l, 3-cyclohexanedione (65)
A total of 2.0 g of dimethylated diketone 64 was dissolved in 5 ml of CH2C12 and a
separate bromine mixture was prepared by adding excess bromine to CH2C12 in a ratio of
about 4 drops bromine to 1 ml of CH2C12. The bromine solution was dropwise added with
stirring at room temperature to the diketone solution and the TLC was monitored using I2
crystals as an indicator. The reaction was stopped when only a small amount of starting
material was observed and the major spot on TLC had a slightly higher Rf value in 30%
ethyl acetate in ligroin. Water was added to the reaction mixture and partitioned. The
water layer was discarded and the organic layer was washed twice with water, dried with
sodium sulfate, and the solvent was removed by evaporation. The residue was then put on
a silica column using 20-30% ethyl acetate as the solvent to give 1.14 g of the product as
a yellow oil. This material was stored at -5 C and upon storage the product crystallized.
Colorless crystals, mp 48-50 C, [H NMR 6: 1.34 (s, 3H, CH3), 1.45 (s, 3H, CH3), 2.27-
2.60 (m, 2H, 6-H), 2.72-2.97 (m, 2H, 5-H), 4.73 (dd, 4.2, 6.9Hz, 1H, 4-H). 13C NMR 5:
23.1, 24.1, 26.5, 34.3, 48.8, 59,5, 203.2, 208.2.
2, 2-Dimethyl-4-cyclohexene-l, 3-dione (66)
A total of 1.0 g of bromo compound 65 was dissolved in 5 ml of DMF and 1.0 g
of LiCl was also added. This mixture was refluxed for 2 hours at which time the TLC
showed the presence of a slower moving product and only a small amount of starting
material. Water and diethyl ether were added to the mixture and partitioned. The water
layer was partitioned twice more with diethyl ether and all the organic layers were
combined and washed with water twice, dried with sodium sulfate, and the solvent was


Figure Page
2-10 Literature Synthesis ofTaxamairinB 30
2-11 Synthesis of Taxamairin B 32
2-12 Reactions of Wrong Regioselectivity 33
2-13 Structures of Taxiflorine and Related Compounds 37
2-14 Oxidation of Taxiflorine 37
2-15 HMBC Correlation of Taxiflorine Acetate 38
2-16 Compounds from the Needles of Taxus floridana 41
2-17 Synthesis of Trans-2, 6-Dimethoxy Cinnamaldehyde 43
3-1 Nitration of Paclitaxel, 7-OH > 2-OH 67
3-2 Nitration of 10-Deacetyl Paclitaxel-7-P-Xyloside l-OH, 2-OH,
3-OH > 10-OH > 2-OH 72
3-3 Regioselective Acetylation of 10-Deacetyl Paclitaxel-7-P-Xyloside 72
3-4 Conversion of 10-Deacetyl Paclitaxel Xyloside to 10-Deacetyl Paclitaxel... 73
3-5 Reductive Denitration of Paclitaxel 75
3-6 Hydrolysis of the Side-Chain of Paclitaxel-7, 2-Dinitrate with NaBH4 75
3-7 Regioselective Denitration of Paclitaxel-7, 2-Dinitrate with
Ammonium Sulfide 77
3-8 Enol Acetate Formation of Nitrate Esters 78
3-9 Reaction of 2-Nitrate Ester with NaNs 80
3-10 Oxidation of Paclitaxel-7-Nitrate Ester 82
3-11 Acetylation of Keto-Ester 83
x


CHAPTER 2
ISOLATION OF TAXOID AND NON-TAXOID COMPOUNDS
FROM TAXUS SPECIES
Large Scale Isolation Process
Although paclitaxel is one of the most promising anti-tumor drugs to receive FDA
approval in many years, it has been beset with many problems not the least of which is
adequate production of the drug. The original, and until recently major source of the
compound was the bark of the Pacific yew (Tcixits brevifolia), from which paclitaxel was
isolated in a yield of 0.01-0.013% on a large scale. Although several related taxanes that
can serve as precursors for the semi-synthesis of paclitaxel, for example, 10-deacetyl
baccatin III (25), co-occur in the bark with paclitaxel, there are no reports to indicate that
these are being isolated from the bark on a large scale. Thus the low yields of paclitaxel
realized by the original process, the apparent unavailability of other useful taxanes
analogues, and the environmental concerns raised by the need to cut the slow-growing
yew trees for harvesting the bark, are some of the reasons why the bark is no longer
considered an attractive source for the large scale production of paclitaxel.
Among the alternatives that are being actively studied are the following: 1)
isolation of 10-deacetyl baccatin III from the European yew (Tcixus baccatd) and its semi-
15


143
Kingston, D. G. I. Recent Advances in the Chemistry and Structure-Activity Relationships
of Paclitaxel. Taxane Anticancer Agents, Basic Science and Current Status, ACS
Symposium Series 583, 1995, 203-216.
Kingston, D. G. I.; Molinero, A. A.; Rimoldi, J. M. The Taxane Diterpenoids. Progress in
the Chemistry of Organic Natural Products 1993, 61, 1-206.
Kingston, D. G. I.; Hawkins, D. R.; Ovinton, L. New Taxanes from Taxus brevifolia.
J. Nat. Prod. 1982, 45, 466-470.
Lataste, H ; Senilh, V.; Wright, M.; Guenard, D.; Potier, P. Relationships Between the
Structures of Taxol and Baccatin III Derivatives and their in vitro Action on the
Disassembly of Mammalian Brain and Physarum Amoebal Microtubules. Proc.
Natl. Acad. Sci. 1984, 81, 4090-4094.
Liang, J.; Min, Z.; Iinuma, M.; Tanaka, T.; Mizuno, M. Two New Antineoplastic
Diterpenes from Taxus mairei. Chern. Pharm. Bull. 1987, 35, 2613-2614.
Mellado, W.; Magri, N. F.; Kingston, D. G. I.; Garcia-Arenas, R.; Orr, G. A.; Horwitz, S.
B. Preparation and Biological Activity of Taxol Acetates. Biophys. Res. Comm.
1984., 124, 329-336.
Miller, R. W A Brief Survey of Taxus Alkaloids and Other Taxane Derivatives. J. Nat.
Prod. 1980, 43, 425-437.
Miller, R. W ; McLaughlin, J. L.; Powell, R. G.; Plattner, R. D ; Weisleder, D.; Smith, C.
R. Nat. Prod. 1982, 47, 131-138.
Nicolaou, K. C.; Yang, Z.; Liu, J. J.; Ueno, H.; Nantermet, P. G.; Guy, R. K.; Claiborne,
C. F., Renaud, J.; Couladouros, E. A.; Paulvannan, K.; Sorensen, E. J. Total
Synthesis of Taxol. Nature 1994, 367, 630-634.
Ojima, I.; Park, Y. H.; Fenoglio, I.; Duelos, O.; Sun, C. M.; Kuduk, S. D.; Zueco, M.;
Appendino, G.; Pera, P.; Veith, J. M.; Bernacki, R. J.; Bissery, M. C.; Combeau,
C.; Vrignaud, P.; Riou, J. F.; Lavelle, F. Synthesis and Structure-Activity
Relationships of New Taxoids. Taxane Anticancer Agents, Basic Science and
Current Status, ACS Symposium Series 583, 1995, 262-275.
Paradis, R.; Page, M. New Active Paclitaxel Glucuronide Derivative with Improved Water
Solubility. Int. J. Oncol. 1998, 12, 391-394.
Rao, K. V. Taxol and Related Taxanes. I. Taxanes of Taxus brevifolia Bark. Pharm. Res.
1993. 10, 521-524.


110
nitroalkanes The dialdehyde could also undergo reductive amination in the presence of an
amine and a reducing agent. The latter method would result is the formation of
morpholino analogues and is well documented, while the former reactions would result in
the formation of a tetrahydropyran ring. In each case an ionizable group could be
incorporated into the function by using these methods assuming reduction of the nitro
group to an amine following the condensation (Figure 4-1). Indeed each of these methods
were exploited in this work however, because of the unavailability of a large number of (3-
dicarbonyls and nitroalkanes and the apply supply of amines, the reductive amination
procedure received the most attention. It should also be mentioned that only 10-deacetyl
paclitaxel-7-xyloside was used in this work because on the large supply of this compound
on hand.
Concerning condensation with (3-dicarbonyls, only one reaction was attempted and
this was condensation of the dialdehyde (150) with malonic acid (151) (Figure 4-2). This
reaction proceeded smoothly by refluxing the dialdehyde with 1.5 eq. of malonic acid in
pyridine/piperidine. However, the expected diacid was not the product isolated. Instead,
based on NMR and mass spectroscopy it was concluded that decarboxylation and loss of
H20 had taken place thus resulting in the a, (3-unsaturated acid (152) as the product
(Figure 4-2). The stereochemistry at C-2 was not determined.
Condensation with a nitroalkane was also only attempted with one such reagent
and this was nitromethane (154) (Figure 4-3). This reaction proceeded by using a 1 : 1
mixture of CH2CI2 and Et3N as solvent/base and dissolving the dialdehyde (153) along


8
daunting task as Nicolaou and Holton published their work simultaneously in two separate
journals. Their respective works are now briefly described.
Relying on the previous work of other groups as well as his own studies, Nicolaou
envisioned the late formation of the oxetane (D) ring, oxygenation/reduction of the C-13
position, and attachment of the side chain. Thus, the problem was perceived as limited to
the assembly of paclitaxels ABC ring system in either its fully functionalized form or a
form that would serve as its progenitor. Figure 1-5 displays the retrosynthetic analysis
involving the bond disconnections on which the synthetic strategy was based. Thus, in the
synthetic direction the following key operations were performed: 1) two fragments (7 and
8) representing precursors to rings A and C were coupled by a Shapiro reaction and a
McMurry coupling to assemble the ABC ring skeleton; 2) installment of the oxetane ring;
3) addition of the various substituents around the peripheries of rings B and C; 4)
oxygenation at C-13; and 5) esterification to attach the side chain. Both precursors to
rings A and C were made possible using the Diels-Alder transform which led to starting
materials that were either commercially available or known in the literature.
In contrast to the convergent synthesis by Nicolaou, Holton took a more linear
approach. The facile epimerization of paclitaxel at C-7 is well documented, and has been
postulated to occur via a retroaldol-aldol process. Holton chose therefore to pursue a
synthetic strategy in which this stereocenter would be introduced at an early stage and
carried throughout most of the synthesis in the absence of a C-9 carbonyl group, thereby
avoiding epimerization. Thus, his route to paclitaxel proceeds retrosynthetically through
the C-7 protected baccatin III 13 to the tricyclic ketone 14, which arises from C ring


81
possessing the same Rf value as the intermediate keto-ester (Figure 3-10). However this
oxidation product only showed one set of signals on the *H NMR spectrum and this set of
peaks matched one of the sets of peaks in the intermediate keto-ester *H NMR spectra.
Presumably C-3 does not racemize in the acidic conditions of the Jones oxidation.
The formation of this keto-ester under basic conditions explains how the enol
acetate and dienol acetate can form in the presence of acetic anhydride and pyridine
(Figure 3-8). Once the keto-ester forms the H-3 can then be abstracted by the base to
form the enolate which then undergoes O-acetylation and once the first enol acetate is
formed the formation of the second proceeds as mentioned before. Indeed it was shown
that if the paclitaxel keto-ester (129) was treated with pyridine in acetic anhydride the
monoenol acetate (130) was the major product (Figure 3-11). This compound was
analogues to 121 (Figure 3-8) without the xylose. If this compound was reacted further
under these same conditions a slightly faster moving product was formed that was
probably the dienol acetate however it was not isolated due to time constraints. Also a
second product was also obtained from the keto-ester acetylation that appears to be an
enol acetate-enol. This conclusion was arrived at because unlike enol acetate 130 this
compound does not show a down field N-H signal yet it has the same mass and the same
number of acetates as 130. On TLC however this compound has a much lower Rf than
130, thus it is concluded that this compound may be either 131 or 132 (Figure 3-11).
One aspect of this that was not clear however was how the keto-ester breaks down
under basic conditions to give dibenzamide and 10-deacetyl baccatin III-7-nitrate ester and
what happens to the C-l and C-2 carbons. It should be stated that this reaction proceeds


99
crystallized from diethyl ether and ligroin. The faster spot was determined to be the side
chain alcohol nitrate ester (116) (34 mg) and the slower product was baccatin III-7-nitrate
ester (117) (148 mg).
Side chain alcohol nitrate ester (116)
Colorless needles, *H NMR 6: 3.68 (t, 8.7Hz, 1H, O-H), 3.87 (d, 14.1Hz, 2H, 1-
H) 5.20 (m, 1H, 2-H), 5.80 (dd, 3.0, 9.6Hz, 1H, 3-H), 6.77 (d, 9.0Hz, 1H, N-H), 7.36-
7.60 (m, 8H, o,m,p-Ph, m,p-NBz), 7.81 (d, 6.9Hz, 2H, o-NBz). 13C NMR 5: 52.0, 59.2,
83.7, 126.6, 127.2, 128.4, 128.6, 128.9, 129.3, 132.5, 136.7, 168.6.
Baccatin IIlt-7-nitrate ester (117)
White crystalline powder, UV (CH3OH): 231 nm, *H NMR 8: 1.10 (s, 3H,
17-H), 1.13 (s, 3H, 16-H), 1.80 (s, 3H, 19-H), 2.03 (m, 1H, 6-H(3), 2.10 (s, 3H, 18-H),
2.20 (s, 3H, 10-0Ac) 2.24 (m, 2H, 14-Ha,(3), 2.32 (s, 3H, 4-OAc), 2.71 (m, 1H, 6-Ha),
4.07 (d, 6.9Hz, 1H, 3-H), 4.14 (d, 8.7Hz, 1H, 2O-H0), 4.34 (d, 8.7Hz, 1H, 20-Ha), 4.89
(t, 8.1Hz, 1H, 13-H), 5.00 (d, 8.7Hz, 1H, 5-H), 5.62 (d, 7.2Hz, 1H, 2-H), 5.80 (dd, 7.2,
10.5Hz, 1H, 7-H), 6.34 (s, 1H, 10-H), 7.49 (t, 7.8Hz, 2H, m-Bz), 7.62 (t, 7.5Hz, 1H, p-
Bz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 10.9, 15.2, 20.2, 20.7, 22.5, 26.8, 32.6,
38.4, 42.7, 47.8, 55.4, 67.8, 74.1, 75.4, 76.2, 78.7, 80.2, 80.3, 83.6, 128.7, 129.1, 130.1,
131.7, 133.8, 145.3, 166.9, 169.5, 171.0, 200.9.
Selective Denitration of Paclitaxel-7, 2-Dinitrate Ester
Paclitaxel-7, 2-dinitrate 200 mg was dissolved in 2 ml of acetonitrile and 0.1 ml of
20% ammonium sulfide and stirred at room temperature for 2 minutes. Water was added
and the mixture was extracted with diethyl ether twice. Removal of the ether left a white


104
63 mg of product (128) was obtained. White amorphous powder, *H NMR 5: 1.21 (s,
3H, 17-H), 1.26 (s, 3H, 16-H), 1.79 (s, 3H, 19-H), 1.99 (s, 3H, 18-H), 2.05 (m, 1H, 6-
Hj3), 2.12 (m, 1H, 14-H(3), 2.18 (s, 3H, 4-OAc), 2.20 (s, 3H, 10-OAc), 2.34 (m, 1H, 14-
Ha), 2.70 (m, 1H, 6-Ha), 4.01 (d, 6.6Hz, 1H, 3-H), 4.11 (d, 8.4Hz, 1H, 20-H(3), 4.31 (d,
8.7Hz, 1H, 20-Ha), 4.97 (d, 8.7Hz, 1H, 5-H), 5.64 (d, 6.9Hz, 1H, 2-H), 5.79 (dd, 7.5,
10.8Hz, 1H, 7-H), 6.20 (t, 8.1Hz, 1H, 13-H), 6.31 (s, 1H, 10-H), 6.42 (d, 5.4Hz, 1H, 3-
H), 7.14 (d, 5.4Hz, 1H, N-H), 7.43-7.51 (m, 10H, m-Bz, o,m,p-Ph, m,p-NBz), 7.62 (t,
6.3Hz, 1H, p-Bz), 7.84 (d, 7.2Hz, 2H, o-NBz), 8.03 (d, 7.2Hz, 2H, o-Bz).
Acetylation of 2-Oxo-PacIitaxel-7-Mononitrate Ester
A total of 300 mg of 2-oxo-paclitaxel-7-mononitrate ester was dissolved in 2 ml
of acetic anhydride and 50 mg of DMAP was added and the solution was stirred at room
temperature for 18 hours. Water was then added to the solution and sodium bicarbonate
was added slowly with stirring. After the release of CO2 stopped the solution was
extracted twice with dichloromethane. The combined organic layers were washed with 0.1
N NaOH, 0.1 N HC1, and water successively, dried with sodium sulfate and evaporated to
a solid residue. This residue was put on a silica column and eluted with 0-10% acetone in
dichloromethane. Two major products were eluted, the faster being the C-2-C-3 mono-
enol acetate (130) which was crystallized from diethyl ether and ligroin to yield 175 mg,
and the slower product was determined to probably be one of two possible enols (131,
132). This was also crystallized from diethyl ether and ligroin to yield 38 mg.


77
O
Figure 3-7: Regioselective Denitration of PacIitaxeI-7, 2-Dinitrate
with Ammonium Sulfide
assigned as shown (Figure 3-8). Apparently, an enol acetate initially forms between the 2-
and 3- positions to give compound 121. This can then form another enol-like acetate
because of the increased acidity of the amide nitrogen to yield compound 122. At this
point however it was not understood how the initial enol acetate was formed.
Reaction with NaN3
In an attempt to displace a nitrate group, paclitaxel-dinitrate (123) was dissolved in
acetonitrile and NaN3 was added with stirring at room temperature. After two hours most
of the starting material was no longer present and a product with similiar Rf (124) was


39
benzoate could be firmly placed at the C-9 position giving the correct structure of
taxiflorine. This correcting structure was identical to a compound recently isolated from
Taxus chinensis var. Mairei and given the name taxchinin M (Tanaka et al., 1996).
Elution with 5% acetone in dichloromethane yielded (-) rhododendrol (82, Figure
2-16) which has been reported previously in Taxus brevifolia (Chu et al., 1994) and
Betula pndula (Smite et al., 1993). This was followed by 13-deacetyl taxiflorine (83,
Figure 2-16). Like taxiflorine, 83 also exhibited a !H spectrum which contained very broad
rounded peaks, whereas the spectrum of its acetate was normal. Also this acetate was
identical to the acetate of taxiflorine. It was also discovered that it was possible to get a
better spectrum of both these compounds (taxiflorine and 13-deacetyl taxiflorine) if the
spectra were run at lower temperatures. Therefore !H and COSY spectra of 83 were
Table 2-1: *H and LC NMR Values for Related Abeo-Taxanes
HorC#
Compd. 81
Compd. 79
Compd. 80
1
****, 67.3
****, 68.5
****, 65.5
2
6.16 d (7.2Hz),
67.8
6.17 d (7.9Hz), 67.8
6.22 d (7.5Hz), 68.7
3
3.00 d (7.5Hz),
43.7
3.01 d (7.9Hz), 44.7
3.12 d (7.8Hz), 44.1
4
**** 73 g
**** yc) 2
**** 'jg q
5
4.96 d (7.2Hz),
84.6
4.99 d (7.6Hz), 84.6
5.00 d (6.0Hz), 84.9
6a
2.67 m, 34.5
2.60 m, 34.7
2.74 m, 34.3
6(3
1.77 m, ****
1.91 m, ****
1 84 m ****
7
5.54 t (7.8Hz), 70.2
5.59 t (8.2Hz), 70.6
5.16 t (7.5Hz), 71.0
8
**** 42 5
****, 43.5
**** 44 2
9
6.32 d (10.8), 77.2
6.21 d (10.9Hz), 76.3
6.32 s, 83.6
10
6.43 d (10.5), 67.6
6.58 d (10.9Hz), 68.8
**** 2922
11
****, 135.8
****, 135.7
****, 137.5
12
****, 146.8
****, 147.7
****, 156.8
13
5.61 t (6.9Hz), 78.4
5.62 t (7.7Hz), 78.7
5.72 t (7.2Hz), 78.9
14a
1.68 m, 36.6
2.50 m, 36.7
1.76 dd (8.1, 14.7Hz), 37.1


45
hydroxy baccatin I, baccatin VI, ip-hydroxy-7-deacetyl baccatin I, and 9-dihydro-13-
acetyl baccatin III.
13-Hydroxy baccatin I (35)
The compound was eluted with 2% acetone in dichloromethane to give 281 mg of
35 crystallized from diethyl ether and ligroin. White crystalline powder, mp 259-261 C,
HNMR5: 1.24 (s, 3H, 17-H), 1.25 (s, 3H, 19-H), 1.65 (s, 3H, 16-H), 1.80 (m, 1H, 6-
H3), 1.90 (m, 1H, 14-H3), 2.00 (s, 3H, OAc), 2.05 (s, 3H, OAc), 2.06 (s, 3H, OAc), 2.06
(s, 3H, OAc), 2.09 (s, 3H, OAc), 2.12 (s, 3H, OAc), 2.18 (m, 1H, 6-Ha), 2.22 (s, 3H, 18-
H), 2.32 (d, 4.8Hz, 1H, 20-Hp), 2.54 (dd, 9.9, 15.3Hz, 1H, 14-Ha), 3.19 (d, 3.6Hz, 1H,
3-H), 3.56 (d, 5.4Hz, 1H, 20-Ha), 4.22 (br s, 1H, 5-H), 5.49 (m, 2H, 2-H, 7-H), 6.05 (d,
11.1Hz, 1H, 9-H), 6.09 (t, 7.8Hz, 1H, 13-H), 6.22 (d, 11.4Hz, 1H, 10-H). 13C NMR 5:
13.6, 15.4, 20.6, 20.8, 20.9, 21.3, 21.4, 21.6, 21.8, 28.4, 31.0, 38.5, 41.3, 43.2, 46.6,
49.9, 58.3, 68.7, 70.7, 71.1, 72.1, 75.1, 76.0, 77.7, 135.6, 140.3, 169.0, 169.2, 169.3,
169.7, 169.8, 170.1.
Baccatin VI (36)
The compound was eluted with 4% acetone in dichloromethane to give 362 mg of
36 crystallized from diethyl ether and ligroin. White crystalline powder, mp 245-247 C,
*H NMR 8: 1.23 (s, 3H, 17-H), 1.61 (s, 3H, 19-H), 1.79 (s, 3H, 16-H), 1.87 (m, 1H, 6-
HP), 2.00 (s, 3H, OAc), 2.03 (s, 3H, 18-H), 2.10 (s, 3H, OAc), 2.11 (s, 3H, OAc), 2.17
(m, 2H, 14-Ha,p), 2.20 (s, 3H, OAc), 2.29 (s, 3H, OAc), 2.51 (m, 1H, 6-Ha), 3.18 (d,
5.7Hz, 1H, 3-H), 4.13 (d, 8.4Hz, 1H, 20-Hp), 4.33 (d, 8.4Hz, 1H, 20-Ha), 4.97 (d,
9.0Hz, 1H, 5-H), 5.87 (d, 6.0Hz, 1H, 2-H), 6.00 (d, 11.4Hz, 1H, 9-H), 6.17 (t, 8.1Hz,


24
however, no attempt was made to determine if this new compound was the C-7-acetyl, C-
9-hydroxy isomer. To further confirm its structure 37 was acetylated with acetic anhydride
and pyridine and the product matched 1 [3-hydroxy baccatin I (35) in every way (Figure 2-
5). Finally 9-dihydro-13-acetyl baccatin III (38) was eluted as determined by NMR
spectroscopy (Figure 2-4). This compound was isolated earlier from the needles of Taxns
canadensis (Gunawardana et al., 1992), but the current isolation is the first from the bark
of Taxns brevifolia. This compound has received much attention from Abbott
Laboratories as a possible precursor to their own paclitaxel analogue.
In addition to this work on the pre-paclitaxel fractions another interesting
compound was isolated while attempting to obtain more paclitaxel from the filtrates of
paclitaxel-containing reverse-phase fractions. This crystalline compound was eluted with a
solvent mixture of 2% acetone in dichloromethane and was given the name brevixanthane
because of its yellow color. Based on 'H and L"C NMR spectra it was quickly concluded
that brevixanthane belonged to a rare group of diterpenes known as 9( 10>20)-abeo-
abietane diterpenoids that have previously been isolated from Taxns species. These
diterpenes consist of a 6, 7, 6 tricyclic carbon ring system with ring C being aromatic and
are a novel diterpene structural class. The only other members of this group include
taxamairin A (38) and B (39) from Taxns chinensis var. Mairei (Liang et al., 1987) and
brevitaxin (40) from Taxns brevifolia which contains a C6-C3 side chain (Arslanian et al.,
1995) (Figure 2-6). Initially, we thought this compound may be novel based on its 'H
NMR spectrum. Brevixanthane contained only one methoxyl which had a chemical shift of
3.89 ppm while the methoxyl of taxamairin A was reported to be at 3.99 ppm. All other


34
oxygenated analogues. This substrate was obtained by reducing ketone 72 with NaCNBH3
in the presence of ZnCl2 (Figure 2-12). Although acetoxymethylation proceeded quite well
the regioselectivity was wrong as was determined by NOE experiments (Figure 2-12).
These experiments clearly showed enhancement of one of the methoxy methyls when the
aromatic proton was irradiated; likewise when the oxygenated methylene was irradiated
the isopropyl methyne signal was enhanced.
At this point lithiation of the aromatic ring and reaction with paraformaldehyde
seemed to be a more attractive way of introducing a hydroxymethyl group which could
then be oxidized to the aldehyde. Initially, lithium/halogen exchange was the desired
process but bromination of the reduced alcohol substrate (69) as well as the totally
reduced ethyl substrate (73) yielded a brominated product with the wrong regiochemistry
(75, 76) (Figure 2-12). This conclusion was reached following NOE experiments as
previously described above.
In light of these results a direct lithiation was attempted with n-butyllithium in
diethyl ether and TMEDA at -78 C using the reduced alcohol substrate (69) (Figure 2-
11). It is well know that lithium complexes with methoxy groups and therefore addition
usually takes place ortho to a methoxy if one is present on the aromatic ring. About 30-45
minutes after adding the n-butyllithium, paraformaldehyde was added. This reaction
proceeded smoothly however yields were only around 50%. In any event, the
regioselectivity was as desired based on NOE with the hydroxymethyl group adding ortho
to the methoxy.


78
Acetic A-nh!'dri<1eem,pDOT^rnight
CH2Ch>room te p
+
Figure 3-8:
Enol Acetate Formation
of Nitrate Esters


page
Acetylation of Taxiflorine 61
Oxidation of Taxiflorine 61
Acetylation of Taxchinin L (83) 62
Synthesis ofTrans-2, 6-Dimethoxy Cinnamaldehyde (85) 62
3 PREPARATION OF NITRATE ESTERS OF PACLITAXEL
AND RELATED TAXANES 66
Complete Nitration of Paclitaxel and Related Taxanes 66
Regioselective Nitrations of Paclitaxel and Related Taxanes 70
Reaction of Taxanes Nitrate Esters 74
Complete Reductive Hydrolysis of Nitrate Esters with Zn
and Acetic Acid 74
Reaction with NaBH4 74
Reaction with Ammonium Sulfide 76
Acetylation of Taxane Nitrate Esters 76
Reaction with NaN3 77
Experimental 86
Complete Nitrations of Taxanes 86
Regioselective Nitration of Paclitaxel 88
Regioselective Nitration of 10-Deacetyl Baccatin III 89
Conversion of 10-Deacetyl Paclitaxel-7-(3-Xyloside to
10-Deacetyl Paclitaxel 122 91
Regioselective Nitrations of 10-Deacetyl Paclitaxel 92
Regioselective Nitration of 10-Deacetyl Paclitaxel-7-P-Xyloside 94
Reductive Denitration of Paclitaxel-7, 2-Dinitrate Ester 98
Reaction of Paclitaxel-7, 2-Dinitrate withNaBEL 98
Selective Denitration of Paclitaxel-7, 2-Dinitrate Ester 99
Acetylation of 10-Deacetyl Paclitaxel-7-(3-Xyloside-
2, 3, 4, 10, 2-Pentanitrate Ester 100
Reaction of Paclitaxel-7, 2-Dinitrate Ester with NaN3 102
Synthesis of 2-Oxo-Paclitaxel-7-Nitrate Ester from
Paclitaxel-7, 2-Dinitrate Ester 103
Synthesis of 2-Oxo-Paclitaxel-7-Nitrate Ester from Paclitaxel-
7-Mononitrate Ester 103
Acetylation of 2-Oxo-Paclitaxel-7-Mononitrate Ester 104
4 SYNTHESIS OF ANALOGUES WITH POTENTIALLY
IMPROVED WATER SOLUBILITY 107
Introduction 107
Synthesis of Analogues Starting from 10-Deacetyl Paclitaxel-7-Xyloside 108
Attempted Synthesis of Taxane Glycosides 115
vi


138
Preparation of Tetraacetyl P-Sitosterol P-GIucoside by the Trichloroacetimidate
Method
A total of 100 mg of tetraacetyl glucosyl trichloroacetimidate and 100 mg of P-
sitosterol was dissolved in 3 ml of CH2CI2 and 36 mg of PTSA was added. The reaction
mixture was stirred at room temperature for 1 hour and the mixture was diluted with
water and CH2CI2 and partioned. The organic layer was washed with water, dried with
Na2S04, and evaporated. The residue was separated on a silica column using 30-40%
ethyl acetate in ligroin as the solvent. A total of 62 mg of amorphous solid product was
obtained from the evaporating fractions. White amorphous powder. !H NMR spectrum
chemical shifts were identical with those reported above.
Attempted Glucosylation of 2-Acetyl Paclitaxel by the Koenigs-Knorr Method
A total of 200 mg of 2-acetyl paclitaxel and 200 mg of acetobromoglucose was
dissolved in 3 ml of CH2CI2 and 100 mg of AgNOs was added. This mixture was stirred
overnight at room temperature. At this time the mixture was filtered and the filtrate was
diluted with water and CH2CI2 and partitioned. The organic layer was washed with water,
dried with Na2S04, and evaporated. The residue was separated on a silica column using
15-30% acetone in benzene as the mobile phase. A total of 132 mg of product was
obtained as an amorphous solid. I3C NMR 8: 10.6, 11.5, 19.8, 20.4, 25.7, 27,2, 32.6,
36.2, 44.1, 53.3, 57.6, 63.5, 68.6, 69,3, 71.1, 71.2, 71.3, 72.8, 74.5, 75.1, 82.0, 126.4,
127.0, 128.2, 128.4, 128.7, 128.8, 129,6, 129.8, 131.9, 133.1, 134.2, 136.9, 137.5, 145.9,
166.3, 167.3, 167.6, 169.4, 169.6, 169.7, 202.0.


44
was identical to the natural product in all ways. A thorough search of the literature
confirmed that this was a novel compound and that no other 2, 6-dioxy cinnamyl
compound has been found in nature.
Experimental
All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of
the reaction. All NMR spectra were recorded using either a Varan VXR-300 or a Varan
Gemini-300 spectrophotometer using CDCI3 as solvent. Infrared spectra were obtained
using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolet spectra were
obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were
recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a
Fisher melting point apparatus. Column chromatography was accomplished using 100-200
mesh silica gel.
Isolation of Minor Compounds from Taxus brevifolia
The filtrates from the region between 10-deacetyl baccatin III and 10-deacetyl
paclitaxel on the reverse-phase chromatographic separation were concentrated to a syrup
(400 g). A 5 g aliquot was applied to a normal-phase silica column (100 g) in
dichloromethane, and chromatographed with an elution sequence consisting of 1-5%
acetone and then 5% acetone and 1-5% methanol. A total of 200 ml of each solvent
mixture was used before progressing to the next solvent system and fractions of about 20
ml were collected and monitored by TLC. The order of elution was as follows: ip-


124
O
acetobromoglucose
AgNC>3
CH2CI2 room temp.
Y
O
Figure 4-9: Rearrangement of 2'-AcetyI Paclitaxel
subjected to the Koenigs-Knorr reaction the main product was loss of the diol function
due to hydrolysis to give 10-deacetyl paclitaxel.


64
refluxed overnight. The following day the TLC seemed unchanged but 2, 4-dinitrophenyl
hydrazine did not give a positive test indicating that no aldehyde remained and the
cinnamic acid product had an identical Rf value. The mixture was partitioned between 0.1
N HC1 and diethyl ether and the organic layer was washed twice more with 0.1 N HC1 and
then three times with water and dried with sodium sulfate. Upon evaporation of the
solvent by evaporation 1.25 g of the cinnamic acid product crystallized.
Methyl trams-2, 6-dimethoxycinnamate
A total of 1.25 g of the cinnamic acid product (91) was dissolved in 15 ml of
acetone and 3.0 g of K2CO3 was added with 0.5 ml of dimethyl sulfate. The mixture was
refluxed for 3 hours at which time no starting material was observed by TLC and thus 2.0
ml of ammonium hydroxide was added to decompose any remaining dimethyl sulfate.
After 15 minutes of stirring, the solvent was partially removed under vacuum and the
residue was partitioned between water and diethyl ether. The organic layer was washed
three times with water and then dried with sodium sulfate. Upon removal of the solvent
1.23 g of the methyl ester product crystallized.
Trans-2, 6-dirmethoxycinnamyl alcohol (92)
A total of 1.23 g of the methyl ester was dissolved in 15 ml of THF and this was
cooled to 0 C. A total of 223 mg (1 eq.) of LAH was then carefully added to the mixture
with stirring This mixture was then refluxed for 4 hours at which time no starting material
remained so about 10 ml of acetone was added to neutralize the remaining LAH and this
was stirred for 18 hours. The solvent was then removed by evaporation and the residue
was partitioned between water and diethyl ether. The organic layer was washed three


46
1H, 13-H), 6.22 (d, 11.1Hz, 1H, 10-H), 7.48 (t, 7.8Hz, 2H, m-Bz), 7.61 (t, 7.2Hz, 1H, p-
Bz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 12.8, 14.9, 20.7, 20.9, 21.2, 21.4, 22.3,
22.7, 28.3, 34.5, 35.1, 42.8, 45.8, 47.3, 69.7, 70.4, 71.8, 73.3, 75.0, 76.4, 78.9, 81.5,
83.8, 128.6, 129.3, 130.1, 133.7, 135.6, 141.2, 166.9, 168.9, 169.1, 169.8, 170.1, 170.4.
l{3-Hydroxy-7-deacetyl baccatin I (37)
The compound was eluted with 4% acetone in dichloromethane to give 132 mg of
37 crystallized from diethyl ether and ligroin. White crystalline powder, mp 234-236 C,
FABMSm/z: 611 (M + 1),'H NMR 5: 1.18 (s, 3H, 16-H), 1.24 (s, 3H, 19-H), 1.66 (s,
3H, 17-H), 1.85 (m, 2H, 14-Ha,(3), 1.92 (m, 2H, 6-Ha,(3), 2.04 (s, 3H, OAc), 2.05 (s,
3H, OAc), 2.11 (s, 3H, OAc), 2.13 (s, 3H, OAc), 2.14 (d, 1.2Hz, 3H, 18-H), 2.19 (s, 3H,
OAc), 2.32 (d, 5.1Hz, 1H, 20-H|3), 2.53 (dd, 9.6, 15.0Hz, 1H, 14-Ha), 3.08 (d, 3.6Hz,
1H, 3-H), 3.52 (d, 5.1Hz, 1H, 20-Ha), 4.21 (t, 3.0Hz, 1H, 5-H), 4.27 (dd, 4.8, 10.8, 1H,
7-H), 5.45 (d, 3.6Hz, 1H, 2-H), 6.07 (t, 7.2Hz, 1H, 13-H), 6.14 (d, 11.1Hz, 1H, 9-H),
6.20 (d, 10.8Hz, 1H, 10-H). 13C NMR 6: 12.5, 15.5, 20.5, 20.8, 20.9, 21.3, 21.6, 21.8,
28.5, 32.3, 38.4, 40.4, 43.3, 47.1, 49.9, 59.2, 69.1, 70.4, 71.8, 72.5, 76.1, 78.0, 78.1,
135.7, 140.7, 168.3, 169.1, 169.5, 169.6, 170.0.
9-Dihydro-13-acetyl baccatin IH (38)
The compound was eluted with 5% acetone and 2% methanol in dichloromethane
to give 184 mg of 38 crystallized from diethyl ether and ligroin. White crystalline powder,
mp 243-244 C, FABMS m/z: 631 (M + 1), !H NMR 5: 1.25 (s, 3H, 16-H), 1.68 (s, 3H,
17-H), 1.82 (s, 3H, 19-H), 1.93 (d, 1.2Hz, 3H, 18-H), 1.96 (m, 1H, 6-HJ3), 2.14 (s, 3H,
10-0Ac), 2.19 (s, 3H, 13-OAc), 2.21 (m, 2H, 14-Ha,(3), 2.28 (s, 3H, 4-OAc), 2.53 (m,


85
Figure 3-12: Mechanism of Keto-Ester Degradation


80
O
Glycolic Acid ? 126
Figure 3-9: Reaction of 2'-Nitrate Ester with NaN 3
decided to confirm this structure by producing this compound by a more typical route.
Thus paclitaxel-7-nitrate (127) was oxidized with Jones reagent to yield a product (128)


132
7.5Hz, o-Bz). 13C NMR 5: 10.8, 14.2, 20.5, 22.5, 26.7, 35.6, 35.7, 43.1, 46.5, 49.9, 53.6,
55.2, 57.1, 62.7, 72.3, 73.2, 74.3, 74.6, 77.2, 78.5, 80.7, 81.0, 84.2, 99.4, 111.8, 117.2,
118.2, 126.9, 127.0, 127.1, 128.3, 128.6, 128.7, 128.9, 129.1, 130.1, 132.0, 133.5, 133.7,
135.9, 137.8, 138.5, 142.7, 156.7, 166.8, 167.6, 170.6, 172.2, 172.6, 209.9.
p-Nitroaniline product (163)
Yellow crystalline powder, 423 mg, mp 177-179 C, UV Aax (CH3OH): 230, 380
nm, FABMSm/z: 1003 (M), 987, 701, 638, 579, 207, 105. lU NMR 5: 1.10 (s, 3H, 17-
H), 1.20 (s, 3H, 16-H), 1.77 (s, 3H, 19-H), 1.80 (s, 3H, 18-H), 1.80 (m, 1H, 14-H0), 2.02
(m, 1H, 6-H(3), 2.29 (m, 1H, 14-Ha), 2.38 (s, 3H, 4-OAc), 2.74 (m, 1H, 6-Ha), 3.28 (m,
2H, 2-H), 3.34 (m, 1H, 4-H), 3.40 (m, 1H, 4-H), 3.67 (m, 1H, 5-H), 3.89 (d,
6.6Hz, 1H, 3-H), 4.01 (m, 1H, 5-H), 4.18 (m, 1H, 7-H), 4.19 (d, 9.0Hz, 1H, 20-Hp),
4.29 (d, 8.4Hz, 1H, 20-Ha), 4.58 (br s, 1H, 1-H), 4.78 (d, 2.7Hz, 1H, 2-H), 4.90 (d,
8.4Hz, 1H, 5-H), 5.13 (s, 1H, 10-H), 5.65 (d, 6.9Hz, 1H, 2-H), 5.77 (dd, 2.7, 9.0Hz, 1H,
3-H), 6.19 (t, 8.4Hz, 1H, 13-H), 6.78 (d, 9.3Hz, 2H, o-Ar), 7.21 (d, 8.7Hz, 1H, N-H),
7.33-7.51 (m, 10H, m-Bz, m,p-NBz, o,m,p-Ph), 7.60 (t, 7.5Hz, 1H, p-Bz), 7.75 (d,
7.2Hz, 2H, o-NBz), 8.10 (d, 7.5Hz, 2H, o-Bz), 8.14 (d, 9.3Hz, 2H, m-Ar). 13C NMR 6:
10.7, 14.3, 20.6, 22.5, 26.7, 35.7, 35.9, 43.0, 46.3, 50.2, 55.2, 57.2, 60.7, 72.3, 73.3,
74.4, 74.6, 76.5, 78.6, 80.6, 80.8, 84.1, 98.2, 112.7, 125.9, 126.9, 127.0, 127.0, 127.1,
128.3, 128.6, 128.7, 128.9, 129.2, 130.1, 131.9, 133.6, 136.2, 137.9, 138.2, 138.8, 154.3,
166.8, 167.2, 170.6, 172.5, 210.2.


88
ether and ligroin. White crystalline powder, mp 187-188 C, Anal. Calc, for C50H52N6O27:
C 51.38; H 4.48; N 7.19. Fd. C 51.53; H 4.59; N 6.82. *H and 13C NMR see Table 1.
Regioselective Nitration of Paclitaxe!
Paclitaxel 500 mg was dissolved in 6 ml of dichloromethane and cooled to 0 C
with an ice bath. A mixture of 5 ml of acetic anhydride and 1 ml of concentrated nitric acid
also cooled to 0 C was added and the total was stirred in an ice bath for 15 minutes. At
this point the reaction was worked up in the same manner as paclitaxel-7, 2-dinitrate.
Although a small amount of paclitaxel-7, 2-dinitrate was seen on TLC, the product was
sufficiently pure to be crystallized directly from diethyl ether and ligroin to give 435 mg of
product (96). White crystalline powder, mp 163-165 C, Anal. Calc, for C47H50N2O16 +
H20: C 61.57; H 5.72; N 3.06. Fd. C 61.95; H 6.10; N 2.91. *H NMR 5: 1.16 (s,3H, 17-
H), 1.22 (s, 3H, 16-H), 1.81 (s, 3H, 19-H), 1.83 (s, 3H, 18-H), 2.04 (m, 1H, 6-HP), 2.19
(s, 3H, 10-0Ac), 2.36 (m, 2H, 14-Ha,p), 2.41 (s, 3H, 4-OAc), 2.68 (m, 1H, 6-Ha), 3.98
(d, 6.6Hz, 1H, 3-H), 4.18 (d, 8.4Hz, 1H, 20-Hp), 4.33 (d, 8.7Hz, 1H, 20-Ha), 4.81 (br s,
1H, 2-H), 4.96 (d, 8.4Hz, 1H, 5-H), 5.68 (d, 6.9Hz, 1H, H-2), 5.74 (dd, 3.3, 10.5Hz,
1H, H-7), 5.79 (dd, 2.4, 9.0Hz, 1H, 3-H), 6.21 (t, 8.1Hz, 1H, 13-H), 6.28 (s, 1H, 10-H),
7.10 (d, 9.0Hz, 1H, NH), 7.35-7.54 (m, 10H, m-Bz, o,m,p-Ph, m,p-NBz), 7.63 (t, 7.2Hz,
1H, p-Bz), 7.75 (d, 7.2Hz, 2H, o-NBz), 8.11 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 11.0,
14.5, 20.6, 21.1, 22.5, 26.6, 32.6, 35.6, 43.2, 47.3, 55.1, 55.3, 72.2, 73.1, 74.1, 74.8,
76.2, 78.5, 79.9, 80.6, 83.6, 127,0, 128,3, 128.7, 129.0, 130.1, 132.0, 133.0, 133.6,
133.8, 137.9, 141.0, 166.7, 167.3, 169.5, 170.7, 172.7, 200.4.


43
Jone's oxidation
Malonic acid
Pyridine/piperidine ,
90 OCII3
1.) Dimethylsulfate
2.) LAH
Figure 2-17: Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde
Jones reagent in a yield of 79% and the aldehyde was then condensed with malonic acid to
yield the corresponding cinnamic acid (91) in about 85% yield. This acid was again
methylated with dimethylsulfate and reduced with LAH to the alcohol (92) to give a total
yield of 60%. Finally 92 was oxidized to the desired aldehyde (93) using the mild oxidizing
agent PDC, to prevent further oxidation, in a yield of 38% (Figure 2-17). This aldehyde


CHEMISTRY OF TAXANES AND TAXUS SPECIES
By
JAMES HARVEY JOHNSON JR.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1998


140
The IC50 for each case was determined from the plot of log [concentration] versus
percent inhibition (not shown). The percent inhibition for each concentration was
determined by the following equation:
% Inhibition = [1 (Td T0 / Tc T0)] 100
Where Td is the number of cells per ml of the drug treated wells, T0 is the number of cells
at the start of the test, and Tc is the average number of cells per ml in the control wells.
The average of four readings for each concentration was used to calculate the IC50 for
each compound.


On more personal notes I first must give thanks to my wife, Amy, for all her
support through this ordeal. Many times I knew she wanted me to stay home more often,
many times I knew she was tired of living the impoverished life of a graduate student but
through it all she remained faithful and supportive and for this I owe her a large debt of
graditude. I would also like to acknowlegde the family of my wife for all their finacial
support that was given to us when we needed it. This also holds true for the families of my
sisters Melinda and Teresa. I am very grateful to you all for your support. Last of all I
would like to acknowledge my parents. It has been said that there is only one point in a
persons life in which he is totally at the mercy of fortune, this is when he is born. I must
say that I was blessed with parents who knew love and showed it to each other as well as
myself, and that they showed this love by instilling in me discipline, integrety, and respect.
I certainly could not be here today if it were not for them.
IV


109
may be the fact that although the xylosides are more active in vitro, they have been
reported to be less active in cell culture toxicity assays (Rao, 1993). Obviously the xylose
unit is either causing a decrease in uptake into the cell or to the site of action or the xylose
unit is serving as a site for metabolism which serves to inactivate the compound.
Table 4-1: ID50 Values of Paclitaxel and Xylosides in Tubuline Assay
Compound
ID50 for disassembly of microtubules, p,M
Paclitaxel
0.5
Paclitaxel-7-Xyloside
0.2
10-Deacetyl Paclitaxel-7-Xyloside
0.3
Cephalomauinine-7-Xyloside
0.25
With this information in hand, the goal was then to devise a scheme in which the
xylose unit would be changed in such a way as to hopefully increase the cell culture
cytotoxicity over the parent xyloside as well as retain an assumed water-solubility
advantage over paclitaxel. As discussed in Chapter 3, the xylose unit in these xylosides can
be easily oxidized with periodate to the dialdehyde. It should be mentioned that this
reaction occurs without interfering with any other part of the molecule including the a-
hydroxy ketone function at C-9 and C-10 of 10-deacetyl paclitaxel-7-xyloside since this
general function type is usually cleaved with periodate. The dialdehyde that is formed
upon oxidation and loss of one carbon actually exist as an equilibrium between 3 different
structures (143, 144, 145) but since it is the dialdehyde (144) which serves as the
electrophile the oxidation product will be referred to as the dialdehyde (Figure 4-1).
Once this dialdehyde is obtained by stirring the xyloside in an aqueous acidic
solution also containing NaI04, it can then be condensed with a variety of nucleophiles.
Among these are carbon nucelophiles such as P-dicarbonyl compounds as well as


91
10-Deacetyl baccatin III-7,10-dinitrate ester (100)
White amorphous solid, Anal Calc, for C29H34N2O14: C 54.88; H 5.40; N 4.41. Fd.
C 55.01; H 5.63; N 4.29. *H NMR 5: 1.08 (s,3H, 17-H), 1.10 (s,3H, 16-H), 1.82 (s, 3H,
19-H), 2.05 (m, 1H, 6-H(3), 2.16 (s, 3H, 18-H), 2.20 (m, 2H, 14-Ha,(3), 2.32 (s, 3H, 4-
OAc), 2.72 (m, 1H, 6-Ha), 3.99 (d, 6.9Hz, 1H, 3-H), 4.13 (d, 8.4Hz, 1H, 20-Hp), 4.35
(d, 8.4Hz, 1H, 20-Ha), 4.91 (t, 8.4Hz, 1H, 13-H), 4.99 (d, 9.0Hz, 1H, 5-H), 5.62 (d,
6.9Hz, 1H, 2-H), 5.80 (dd, 7.2, 10.5Hz, 1H, 7-H), 6.40 (s, 1H, 10-H), 7.50 (t, 7.5Hz, 2H,
m-Bz), 7.63 (t, 7.5Hz, 1H, p-Bz), 8.09 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 10.8, 15.6,
20.4, 22.4, 26.6, 32.5, 38.4, 42.5, 47.8, 55.7, 60.4, 67.8, 73.8, 76.1, 78.5, 80.1, 80.2,
82.4, 83.5, 128.7, 128.9, 129.0, 130.0, 133.9, 148.9, 166.8, 171.1, 200.1.
Conversion of 10-Deacetyl Paditaxel-7-P-XyIoside to 10-Deacetyl Paclitaxel (112)
10-Deacetyl paclitaxel-7-p-xyloside 1.0 g was dissolved in 10 ml of 1 : 1 THF and
water and 2 ml of 1 N H2S04 was added. This was followed by 0.71 g of NaI04 and the
mixture was stirred overnight at room temperature. The mixture was diluted with water
and extracted three times with dichloromethane. The organic layer was evaporated to
dryness to yield 0.95 g of a white powder. The material was then dissolved in 20 ml of
methanol and 0.5 ml of phenylhydrazine and 3 ml of acetic acid was added. This mixture
was heated at 60 C for 3 hours at which point TLC analysis showed almost complete
conversion to 10-deacetyl paclitaxel. The mixture was diluted with water, acidified and
extracted with dichloromethane three times and the organic portion was evaporated to a
dark red oil The concentrate was partitioned in a countercurrent fashion, between 3 : 2
methanol/water and 4: 1 benzene/ligroin as the two phases, and using 3 separatory funnels.


119
applicable to unreactive and hindered nucleophiles and has been the basis for developing a
combinatorial process involving oligosaccharides.
Acetyl protected glycosyl donors for each of these methods were prepared using
D-glucose as the sugar as shown in Figure 4-6. Initially D-glucose (165) is acetylated with
acetic anhydride and pyridine to give the pentaacetate (166) in near quantitative yields.
Next, bromine was introduced at the anomeric carbon by treating the pentaacetate with
30% HBr in acetic acid which gave yields around 80%. This product (167) was used as
the Koenigs-Knorr donor and also as a starting material for the other two donors. To
synthesize the trichloroacetimidate donor the bromide was hydrolyzed in aqueous
acetonitrile with Hg(CN)2 added. This gave a quantitative yield. The a-
trichloroacetimidate (169) was produced from this by adding trichloroacetonitrile and
using K2CO3 as base in CH2CI2 in a yield of about 75%. The sulphoxide donor was
prepared by first by making the phenylthioglycoside (170) using the bromide, thiophenol,
and KOH as base. This reaction gave near quantitative yields. The sulfide was then
oxidized to the sulphoxide (171) with mCPBA to give near quantitative yields (Figure 4-
6).
Once these donors were in hand a series of test glycosylations were performed
using (3-sitosterol (172) and benzyl alcohol (173) as aglycones to determine the best
conditions. Unfortunately the sulphoxide method was never attempted due to a lack of the
triflic anhydride activator. However both the Koenigs-Knorr and trichloroacetimidate
methods were attempted and meet with success. In terms of the Koenigs-Knorr method,
(3-glycosides were formed with both of these aglycones in moderate yields with the benzyl


58
lP-Hydroxy baccatin I (35)
The compound was eluted with 2% acetone in dichloromethane and a total of 189
mg of 35 was crystallized from diethyl ether and ligroin. Its physical and spectral
properties are identical to that reported above.
Taxiflorine (taxchinin M) (77)
The compound was eluted with 2% acetone in dichloromethane and a total of 129
mg was crystallized from diethyl ether and ligroin. No NMR was reported since the
spectrum contained poorly defined peaks.
Rhododendrol (82)
This compound was eluted with 5% acetone in dichloromethane and a total 435
mg was crystallized directly from the eluting solvent. Clear colorless crystals, mp 74-76
C, 'H NMR 5: 1.23 (d, 6.2Hz, 3H, 1-H), 1.73 (m, 2H, 3-H), 2.63 (m, 2H, 4-H), 3.83 (m,
1H, 2-H), 6.74 (m, 2H, m-Bz), 7.04 (m, 2H, o-Bz). 13C NMR 6: 23.5, 31.2, 40.9, 67.7,
115.3, 129.4, 133.9, 153.8.
Taxchinin L (83)
This compound was eluted with 5% acetone in dichloromethane and a total of 162
mg was crystallized from diethyl ether and ligroin. White crystalline powder, mp 264-266
C. lK NMR (-10 C) 5: 1.02 (s, 3H, 16-H), 1.22 (s, 3H, 17-H), 1.43 (m, 1H, 14-Ha),
1.73 (s, 3H, 19-H), 1.79 (s, 3H, 7-OAc), 1.81 (m, 1H, 6-Hp), 1.96 (s, 3H, 18-H), 2.05 (s,
3H, 2-OAc), 2.13 (m, 1H, 14-Ha), 2.17 (s, 3H, 4-OAc), 2.61 (m, 1H, 6-Ha), 3.15 (d,
7.4Hz, 1H, 3-H), 4.45 (m, 2H, 13-H, 20-Hp), 4.53 (d, 7.4Hz, 1H, 20-Ha), 4.69 (t,
10.3Hz, 1H, 10-H), 4.94 (d, 6.2Hz, 1H, 5-H), 5.46 (dd, 5.6, 8.6Hz, 1H, 7-H), 5.95 (d,


145
Wang, X.; Pan, X. A Convergent Synthesis of Taxamairin A, B Precursor I. 8-(2, 3-
Dimethoxy-4-isopropylbenzyl)-l, 4-dioxaspiro[4, 5]decan-7-one. J. Indian Chem.
Soc. lL995a, 73, 497-498.
Wang, X.; Pan, X.; Zhang, C.; Chen, Y. The Total Synthesis of Taxamairin B. Synth.
Comm. 1995b, 25, 3413-3419.
Wani, M. C.; Taylor, H. L.; Wall, M. E.; Coggon, P.; McPhail, A. T. Plant Antitumor
Agents. VI. The Isolation and Structure of Taxol, a Novel Antileukemic and
Antitumor Agent from Taxus brevifolia. J. Am. Chem. Soc. 1971, 93, 2325-2327.


128
Bz). 13C NMR 5 (some DMSO): 10.6, 14.1, 20.6, 22.5, 26.5, 35.1, 35.6, 43.0, 46.5,
55.2, 56.9, 60.7, 63.2, 72.0, 73.3, 74.4, 74.7, 76.5, 78.4, 80.8, 80.9, 84.0, 101.9, 127.0,
127.1, 127.7, 128.0, 128.2, 128.2, 128.5, 128.6, 128.7, 129.3, 130.1, 131.8, 133.5, 133.6,
135.9, 138.1, 138.3, 139.6, 166.6, 167.8, 168.3, 170.4, 172.7, 209.4.
Condensation of Dialdehyde with Nitromethane
A total of 200 mg of dialdehyde and 35 mg of nitromethane was dissolved in a 1 :
1 mixture of benzene and triethylamine and this mixture was stirred for 4 days at room
temperature. At this point the reaction mixture was evaporated to a residue uder reduced
pressure and this residue was ran on a silica column using 10-30% acetone in benzene as
the mobile phase. A total of 45 mg of amorphous product was obtained as the major
product. White amorphous powder, UV A,max (CH3OH): 228 nm, FABMS m/z: 974 (M +
2), 956, 913, 689, 670, 603, 286, 154, 136, 105, 81. 13C NMR 5 (acetonitrile): 11.4,
14.5, 21.4, 23.1, 27.0, 35.6, 36.8, 43.9, 47.3, 56.9, 57.6, 62.8, 68.3, 70.3, 72.2, 74.6,
75.6, 75.7, 77.0, 79.0, 81.3, 81.8, 84.4, 92.8, 99.9, 128.2, 128.3, 128.7, 129.4, 129.5,
129.6, 130.9, 132.5, 134.4, 135.2, 137.0, 139.0, 139.9, 166.8, 168.1, 171.5, 173.6, 212.0.
Reduction of Dialdehyde to the Diol
A total of 100 mg of dialdehyde was dissolved in a 1 : 1 mixture of methanol and
acetic acid and 100 mg of NaCNBTh was added and this mixture was stirred at room
temperature for 1 hour. At this point the mixture was diluted with ethyl acetate and water
and partitioned. The organic layer was washed twice with saturated NaHCOj solution and
twice with water. The organic layer was then dried with Na2S04 and evaporated under
reduced pressure. The solid residue was purified by running a quick silica column with


120
Ac20

pyridine
Hg(CN)2
<
aq. CH3CN
167 OAc
CI3CCN, K2CO3,
CH2C12
room temp., 72 hrs.
KOH/
thiophenol
V
Figure 4-6: Synthesis of Glycosyl Donors


33
Figure 2-12: Reactions of Wrong Regioselectivity


4
Figure 1-2: 'H NMR Spectrum of Paclitaxel
vu;


106
6 19 (s, 1H, 10-H), 7.47-7.63 (m, 11H, m,p-Bz, o,m,p-Ph, m,p-NBz), 7.84 (d, 7.2Hz, 2H,
o-NBz), 8.12 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 11.1, 13.3, 20.6, 20.7, 21.3, 22.7, 26.1,
32.4, 35.8, 42.9, 47.2, 55.1, 71.9, 74.3, 74.6, 76.2, 79.4, 79.8, 79.9, 83.7, 127.5, 128.6,
128.7, 129.0, 129.1, 129.3, 129.4, 130.3, 130.7, 132.2, 132.7, 132.8, 133.5, 133.7, 133.8,
141.4, 163.4, 163.7, 166.7, 168.5, 169.4, 171.5, 200.6.


37
AcO
Figure 2-13: Structure of Taxiflorine and Related Compounds
Figure 2-14: Oxidation of Taxiflorine
ppm which is also consistent with the ¡3-carbon of an a, (3-unsaturated ketone system. The
C-12 signal in C-9 keto compounds is usually around 147.0 ppm. Also, the UV spectrum


97
14-Ha,p), 2 39 (s, 3H, 4-OAc), 2.71 (m, 1H, 6-Ha), 3.35 (m, 1H, 5-Heq), 3.38 (m,
1H, 2-H), 3.75 (br s, 1H, 4-H), 3.88 (d, 6.3Hz, 1H, 3-H), 3.95 (m, 1H, 5-Hax), 4.08
(t, 8.1Hz, 1H, 7-H), 4.15 (d, 6.6Hz, 1H, 1-H), 4.20 (d, 8.1Hz, 1H, 20-HP), 4.31 (d,
8.1Hz, 1H, 20-Ha), 4.79 (br s, 1H, 2-H), 4.90 (d, 9.0Hz, 1H, 5-H), 5.03 (t, 8.4Hz, 1H,
3-H), 5.18 (s, 1H, 10-H), 5.65 (d, 6.6Hz, 1H, 2-H), 5.77 (d, 7.8Hz, 1H, 3-H), 6.19 (t,
7.8Hz, 1H, 13-H), 7.15 (d, 7.5Hz, 1H, N-H), 7.40-7.53 (m, 10H, m-Bz, o,m,p-Ph, m,p-
NBz), 7.62 (t, 6.9Hz, 1H, p-Bz), 7.75 (d, 7.5Hz, 2H, o-NBz), 8.12 (d, 7.2Hz, 2H, o-Bz).
13C NMR 8: 10.1, 13.6, 20.3, 22.0, 25.9, 34.9, 35.0, 42.6, 46.0, 55.1, 55.8, 65.0, 66.2,
69.6, 70.7, 73.4, 74.1, 74.3, 75.6, 78.5, 80.2, 81.3, 83.4, 86.2, 104.3, 126.4, 126.9, 127.7,
127.8, 127.9, 128.0, 129.3, 129.4, 130.8, 132.6, 133.8, 135.6, 137.0, 138.5, 165.6, 166.5,
169.6, 172.0, 208.5.
10-Deacetyl pacIitaxel-7-3-xyloside-4-mononitrate ester (107)
White crystalline powder, mp 193-195 C, Anal. Calc, for C50H56N2O19 + H20: C
59.64; H 5.81; N 2.78. Fd.C 59.96; H 6.18; N 2.56. 'HNMR5: 1.12 (s,3H, 17-H), 1.25
(s, 3H, 16-H), 1.79 (s, 3H, 19-H), 1.81 (s, 3H, 18-H), 2.09 (m, 1H, 6-HP), 2.31 (m, 2H,
14-Ha,p), 2.34 (s, 3H, 4-OAc), 2.64 (m, 1H, 6-Ha), 3.01 (br s, 1H, 2-H), 3.28 (t,
9.9Hz, 1H, 5-Heq), 3.51 (t, 8.7Hz, 1H, 3-H), 3.83 (d, 6.3Hz, 1H, 3-H), 3.98 (d,
7.5Hz, 1H, 5-Hax), 4.11 (m, 1H, 1-H), 4.12 (m, 1H, 7-H), 4.19 (d, 7.8Hz, 1H, 20-
Hp), 4.30 (d, 8.4Hz, 1H, 20-Ha), 4.82 (br s, 1H, 2-H), 4.90 (m, 1H, 5-H), 4.91 (m, 1H,
4-H), 5.28 (s, 1H, 10-H), 5.64 (d, 6.6Hz, 1H, 2-H), 5.77 (d, 9.3Hz, 1H, 3-H), 6.17 (t,
7.8Hz, 1H, 13-H), 7.21 (d, 9.3Hz, 1H, N-H), 7.31-7.53 (m, 10-H, m-Bz, o,m,p-Ph, m,p-
NBz), 7.61 (t, 7.5Hz, 1H, p-Bz), 7.72 (d, 7.5Hz, 2H, o-NBz), 8.14 (d, 7.5Hz, 2H, o-Bz).


51
evaporationed. The residue was ran through a silica column using 15-30% ethyl acetate in
ligroin as the eluent to give 823 mg of product as a yellow oil. Yellow oil, *H NMR 5:
1.35 (s, 6H, CH3), 3.35 (dd, 2.4, 3.9Hz, 2H, 6-H), 6.25 (dt, 2.1, 10.2Hz, 1H, 4-H), 7.03
(dt, 3.9, 10.2Hz, 1H, 5-H).
Isopropylation of acetovanillone
A total of 3.0 g of acetovanillone (67) was added to 20 ml of 90% H2S04 and 2 ml
of isopropyl alcohol was added. This mixture was stirred at 60 C for 36 hours. At this
point about 50% conversion to the product was observed on TLC. Longer reaction times
did not increase conversion and higher temperatures caused decomposition. Thus the
mixture was diluted with ice water and diethyl ether and partitioned. The water layer was
partitioned once again with ether and the ether layers combined. The organic layer was
then partitioned twice with 1.0 N NaOH and the organic layer was discarded. The
aqueous layer was acidified with concentrated HC1 and extracted twice with diethyl ether.
This organic layer was then washed with water twice, dried with sodium sulfate, and
concentrated. The residue was chromatographed on a silica column using 20-40% ethyl
acetate in ligroin as the eluent to give 1.42 g of product which crystallized on standing.
Clear colorless crystals, mp 116-117 C, EIMS m/z: 208 (42%, M), 193 (100%), 163
(17%), 77 (12%), >H NMR 5: 1.27 (d, 6.9Hz, 6H, CH3), 2.57 (s, 3H, CH3), 3.33 (quintet,
6.9Hz, 1H, CH), 3.95 (s, 3H, OCH3), 6.23 (br s, 1H, Ar-OH), 7.39 (d, 1.5Hz, 1H, Ar-H),
7.50 (d, 1.5Hz, 1H, Ar-H). 13C NMR 5: 22.2, 26.2, 27.2, 56.2, 107.4, 121.0, 129.4,
133.7, 146.2, 147.6, 197.1.


89
Regioselective Nitration of 10-Deacetyl Baccatin III
10-Deacetyl baccatin III 500 mg was dissolved in 6 ml of dichloromethane and
cooled to 0 C with an ice bath. A mixture of 5 ml of acetic anhydride and 1 ml of
concentrated nitric acid also cooled to 0 C was added and the total was stirred in an ice
bath for 10 minutes. At this point the reaction was worked up in the same manner as
paclitaxel-7, 2-dinitrate. The TLC of the reaction mixture showed 4 compounds, two of
which were a small amount of 10-deacetyl baccatin III (low Rf) starting material and a
small amount of 10-deacetyl baccatin III-7, 10, 13-trinitrate (high Rf). A silica column
was ran to isolate the two intermediate products using 0-15% acetone in dichloromethane
as solvent. The more non-polar of the two products actually turned out to be a mixture of
two compounds while the more polar product was determined by NMR spectroscopy to
be 10-deacetyl baccatin III-10-mononitrate (98). The more non-polar two product
mixture was then ran on another silica column using 30-50% ethyl acetate in ligroin as
solvent, which separated the two compounds quite well. The product with the high Rf was
determined to be 10-deacetyl baccatin III-10, 13-dinitrate (99), and the other to be 10-
deacetyl baccatin III-7, 10-dinitrate (100). Product yields were as follows; 10-deacetyl
baccatin III-10-mononitrate 180 mg crystallized from diethyl ether and ligroin, 10-deacetyl
baccatin III-10, 13-dinitrate 115 mg crystallized from diethyl ether and ligroin, 10-deacetyl
baccatin III-7, 10-dinitrate 62 mg.


CHAPTER 1
HISTORY AND BACKGROUND OF PACLITAXEL
Introduction
Paclitaxel (Taxol1M) (1) is a potent antitumor agent that was originally isolated
from the Pacific yew tree, Taxus brevifolia, by Wall and Wani in 1971 (Wani et al., 1971).
The structure of paclitaxel is that of a tetracyclic diterpene ester (Figure 1-1). Its structure
has several unusual features including an oxetane ring, an 8-membered B-ring, a
bridgehead double bond, 11 asymmetric centers, and a N-substituted phenylisoserine ester.
This structure has provided a great challenge to synthetic organic chemists since its
elucidation but was conquered initially by two groups simultaneously and by other groups
since (Nicolaou et al., 1994; Holton et al., 1994a; Holton et al., 1994b). The ongoing
phytochemical study of Taxus brevifolia as well as other Taxus sp. has yielded many
taxanes other than paclitaxel Compounds with rearranged ring systems, various types of
esters, as well as glycosides have been isolated and several reviews have been published
(Kingston et al., 1993; Das et al., 1995; Appendino, 1995).
Early Work and Structural Elucidation
The pioneering work concerning paclitaxel began in the late 1950s when the
National Cancer Institute started a screening program of plant extracts using tumor
I


79
present on TLC (Figure 3-9). The reaction continued overnight and on the following day it
was found that the initial product was no longer present and two other products (125,
126) were now formed, one faster (125) than the intermediate product which exhibited
strong UV absorbance but did not char with H2S04 and a slower product (126) which
exhibited UV absorbance as well as charring with acid on TLC. These two products were
isolated by column chromatography and determined by NMR spectroscopy to be baccatin
III-7-nitrate (slower product) (126) and dibenzamide (faster product) (125). Dibenzamide
was also synthesized from benzamide and benzoyl chloride in the presence of NaH to
insure that the reaction product was indeed dibenzamide. It was later found that in DMF
as solvent the reaction proceeded much faster and if it were stopped after only 10 minutes
the intermediate was the major constituent of the reaction mixture. This product was
isolated and determined to have a molecular weight of 896. The 'H NMR of this
compound was very unusual in that many of the signals seemed to be in duplicate. Signals
for the N-H and H-3 were present however there was no signal for the H-2. The
duplication of peaks was similar to what one may find in a racemic mixture presenting the
possibility that the stereochemistry at one of the asymmetric carbons had been scrambled.
After reviewing the literature concerning nitrate esters it was found that it is quite normal
for a nitrate ester to undergo alpha-elimination in the presence of strong base to yield a
carbonyl (Boschan et al., 1955). In this case the C-2 is already acidic due to its proximity
to the C-L ester carbonyl, therefore it was quite conceivable that even a weak base such
as NaNs may cause alpha-elimination. With this information in hand the structure of 124
was determined to be as shown with the stereochemistry at C-3 racemic. It was also


129
using 40-50% acetone in benzene as the mobile phase. Evaporation of the appropriate
fractions yielded 182 mg of amorphous solid as the product. UV Aax (CH3OH): 229 nm,
FABMS m/z: 916 (M+l), 661, 653, 551, 509, 176, 154, 136, 105, 81. *HNMR5: 1.07
(s, 3H, 17-H), 1.16 (s, 3H, 16-H), 1.76 (s, 3H, 19-H), 1.82 (s, 3H, 18-H), 1.82 (m, 1H,
14-Hp), 1.96 (m, 1H, 6-Hp), 2.24 (m, 1H, 14-Ha), 2.35 (s, 3H, 4-OAc), 2.69 (m, 1H, 6-
Ha), 3.38-3 55 (m, 6H, 2-, 4-, 5-H), 3.84 (d, 6.3Hz, 1H, 3-H), 4.13 (m, 1H, 7-H),
4.14 (d, 8.1Hz, 1H, 20-Hp), 4.28 (d, 8.1Hz, 1H, 20-Ha), 4.56 (br s, 1H, 1-H), 4.80 (br
s, 1H, 2-H), 4.91 (d, 9.0Hz, 1H, 5-H), 5.23 (s, 1H, 10-H), 5.63 (d, 6.6Hz, 1H, 2-H),
5.74 (d, 6.9Hz, 1H, 3-H), 6.17 (t, 8.4Hz, 1H, 13-H), 7.30-7.52 (m, 11H, m-OBz, m,p-
NBz, o,m,p-Ph, N-H), 7.61 (t, 7.2Hz, 1H, p-OBz), 7.73 (d, 7.5Hz, 2H, o-NBz), 8.10 (d,
7.2Hz, 2H, o-OBz). 13C NMR 6: 10.7, 14.2, 20.7, 22.5, 26.6, 35.5, 35.7, 43.1, 46.6,
55.2, 57.0, 61.8, 62.3, 67.7, 72.1, 73.2, 74.6, 74.7, 76.5, 78.6, 78.8, 80.9, 84.0, 103.9,
127.0, 127.1, 128.2, 128.3, 128.6, 128.7, 128.9, 129.2, 130.1, 131.9, 133.7, 136.0, 138.0,
138.2, 166.8, 167.5, 170.7, 172.9, 210.3.
General Procedure for Reductive Animations
A total of 500 mg of the dialdehyde and 500 mg of the amine were dissolved in 6
ml of 2 : 1 methanol and acetic acid and excess (200 mg) of NaCNBH3 was added. The
mixture was stirred at room temperature for 1.5 2.0 hours and then diluted with water
and extracted three times with ethyl acetate. The organic layer was then washed with
NaHC03 twice and with water twice. The organic layer was then dried with Na2S4 and
evaporated. The product was separated by putting the residue on a silica column and
eluted with 20-40% acetone in benzene.


123
CC13
PTSA or BF3-Et20
v CH2CI2, room temp.
Figure 4-8: Trichloroacetimidate Glycosylation
produce a primary hydroxyl which should be easily glycosylated, 10-deacetyl paclitaxel-7-
xyloside was oxidized to the dialdehyde as before and this dialdehyde was subsequently
reduced with NaCNBH3 to the diol (Figure 4-4). Unfortunately when this substrate was


CHEMISTRY OF TAXANES AND TAXUS SPECIES
By
JAMES HARVEY JOHNSON JR.
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1998

This work is dedicated to my father whose untimely passing due to the disease that this
work addresses on April 28, 1998 has left a deep void in my life that will never be filled.
Although he would not be considered an educated man by most standards he taught me
more than any textbook or professor ever could. I hoped he could be here when this work
was completed but that was not to be. Nevertheless I hope that this accomplishment
would make him as proud of me as I was of him. I love and miss you daddy.

ACKNOWLEDGMENTS
There are many people who have helped me along this journey, some through
deeds and some through inspiration. Firstly I would like to acknowledge those in academia
that have made this event possible. To the late Koppaka V. Rao, my graduate advisor, I
would I would like to give much thanks for his patience and understanding and for the
wisdom and experience which he has imparted on me. He will certainly be missed but I
hope to keep his memory alive through the work I accomplish throughout my career. To
the late Maya Ganguli, my undergraduate advisor and mentor, I would like to give thanks
for opening my eyes to the exciting world of chemistry. I could not have asked for a
teacher more caring and better suited than you to help me get started in my chemistry
education. I would also like to acknowledge other faculty members that have given their
time and energy so that I could accomplish this goal; John Perrin for accepting the duty of
being my graduate advisor upon Dr. Raos passing, Margaret James, Ken Sloan, and
William Dolbier for serving on my thesis committee, and Bob Higgins for guidence and
always interesting conversation during my undergraduate studies. Finally I must give
thanks to my fellow students along the way both graduate and undergraduate with special
thanks to Veronica Hall and Bernice Kidd for their help and friendship during my
undergraduate years and to Ravi Orugunty and John Juchum whom I shared my trials and
tribulations with for the last 5 years as lab mates in Dr. Raos laboratory.

On more personal notes I first must give thanks to my wife, Amy, for all her
support through this ordeal. Many times I knew she wanted me to stay home more often,
many times I knew she was tired of living the impoverished life of a graduate student but
through it all she remained faithful and supportive and for this I owe her a large debt of
graditude. I would also like to acknowlegde the family of my wife for all their finacial
support that was given to us when we needed it. This also holds true for the families of my
sisters Melinda and Teresa. I am very grateful to you all for your support. Last of all I
would like to acknowledge my parents. It has been said that there is only one point in a
persons life in which he is totally at the mercy of fortune, this is when he is born. I must
say that I was blessed with parents who knew love and showed it to each other as well as
myself, and that they showed this love by instilling in me discipline, integrety, and respect.
I certainly could not be here today if it were not for them.
IV

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
LIST OF TABLES viii
LIST OF FIGURES ix
ABBREVIATIONS xii
ABSTRACT xiii
CHAPTERS
1 HISTORY AND BACKGROUD OF PACLITAXEL 1
Introduction I
Early Work and Structural Elucidation 1
Mechanism of Action 6
Total Synthesis 7
Structure-Activity Relationships 12
2 ISOLATION OF TAXOID AND NON-TAXOID COMPOUNDS
FROM TAXI 'S SPECIES 15
Large Scale Isolation Process 15
Isolation of Minor Compounds from the Bark of Taxus brevifolia 22
Synthesis of Taxamairin B 26
Isolation of Minor Compounds from Taxus floridana 35
Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde 42
Experimental 44
Isolation of Minor Compounds from Taxus brevifolia 44
Acetylation of 1 (3-Hydroxy-7-Deacetyl Baccatin I 47
Isolation of Taxamairin A (38) from Taxus brevifolia 47
Methylation of Taxamairin A 48
Synthesis of Taxamairin B (39) 49
Isolation of Minor Compounds from Taxus floridana 57
v

page
Acetylation of Taxiflorine 61
Oxidation of Taxiflorine 61
Acetylation of Taxchinin L (83) 62
Synthesis ofTrans-2, 6-Dimethoxy Cinnamaldehyde (85) 62
3 PREPARATION OF NITRATE ESTERS OF PACLITAXEL
AND RELATED TAXANES 66
Complete Nitration of Paclitaxel and Related Taxanes 66
Regioselective Nitrations of Paclitaxel and Related Taxanes 70
Reaction of Taxanes Nitrate Esters 74
Complete Reductive Hydrolysis of Nitrate Esters with Zn
and Acetic Acid 74
Reaction with NaBH4 74
Reaction with Ammonium Sulfide 76
Acetylation of Taxane Nitrate Esters 76
Reaction with NaN3 77
Experimental 86
Complete Nitrations of Taxanes 86
Regioselective Nitration of Paclitaxel 88
Regioselective Nitration of 10-Deacetyl Baccatin III 89
Conversion of 10-Deacetyl Paclitaxel-7-(3-Xyloside to
10-Deacetyl Paclitaxel 122 91
Regioselective Nitrations of 10-Deacetyl Paclitaxel 92
Regioselective Nitration of 10-Deacetyl Paclitaxel-7-P-Xyloside 94
Reductive Denitration of Paclitaxel-7, 2-Dinitrate Ester 98
Reaction of Paclitaxel-7, 2-Dinitrate withNaBEL 98
Selective Denitration of Paclitaxel-7, 2-Dinitrate Ester 99
Acetylation of 10-Deacetyl Paclitaxel-7-(3-Xyloside-
2, 3, 4, 10, 2-Pentanitrate Ester 100
Reaction of Paclitaxel-7, 2-Dinitrate Ester with NaN3 102
Synthesis of 2-Oxo-Paclitaxel-7-Nitrate Ester from
Paclitaxel-7, 2-Dinitrate Ester 103
Synthesis of 2-Oxo-Paclitaxel-7-Nitrate Ester from Paclitaxel-
7-Mononitrate Ester 103
Acetylation of 2-Oxo-Paclitaxel-7-Mononitrate Ester 104
4 SYNTHESIS OF ANALOGUES WITH POTENTIALLY
IMPROVED WATER SOLUBILITY 107
Introduction 107
Synthesis of Analogues Starting from 10-Deacetyl Paclitaxel-7-Xyloside 108
Attempted Synthesis of Taxane Glycosides 115
vi

page
L1210 Cytotoxicity of Analogues 125
Experimental 126
Oxidation of Xyloside with Periodate 127
Condensation of Dialdehyde with Malonic Acid 127
Condensation of Dialdehyde with Nitromethane 128
Reduction of Dialdehyde to the Diol 128
General Procedure for Reductive Aminations 129
Synthesis of Glucose Pentaacetate 133
Synthesis of la-Bromo-Tetraacetyl Glucose 134
Synthesis of 1-Hydroxy-Tetraacetyl Glucose 134
Synthesis of la-Trichloroacetimidate-Tetraacetyl Glucose 135
Synthesis of Tetraacetyl Phenyl Thioglucoside 135
Synthesis of Tetraacetyl Glucose, Phenyl Sulfoxide 136
Preparation of Tetraacetyl Benzyl P-Glucoside by the
Koenigs-Knorr method 136
Preparation of Tetraacetyl Benzyl p-Glucoside by the
Trichloroacetimidate Method 137
Preparation of Tetraacetyl P-Sitosterol P-Glucoside by the
Koenigs-Knorr Method 137
Preparation of Tetraacetyl p-Sitosterol P-Glucoside by the
Trichloroacetimidate Method 138
Attempted Glucosylation of 2-Acetyl Paclitaxel by the
Koenigs-Knorr Method 138
L1210 Cytotoxicity Assay 139
LIST OF REFERENCES 141
BIOGRAPHICAL SKETCH 146
vii

LIST OF TABLES
Table page
1-1 Physical and Chemical Properties of Paclitaxel 3
2-1 'H and 13C NMR Values for Related Abeo-Taxanes 39
3-1 *14 and L"C NMR Values for Completely Nitrated Taxanes 68
4-1 ID50 Values of Paclitaxel and Xylosides in Tubuline Assay 109
4-2 L1210 Cytotoxicity of Paclitaxel and Analogues 126
viii

LIST OF FIGURES
Figure Page
1-1 Structure of Paclitaxel 2
1-2 11 NMR Spectrum of Paclitaxel 4
1 -3 Halogenated Products of Methanolysis Used for X-Ray Crystallography .... 5
1-4 Neutral and Alkaline Oxidation of Paclitaxel 7
1-5 Nicolaou Synthesis of Paclitaxel 9
1-6 Holton Synthesis of Paclitaxel 10
1-7 Holton Synthesis of Paclitaxel 11
1-8 Structure-Activity Relationships 14
2-1 Structure of Major Taxanes 16
2-2 Reverse-Phase Isolation of Taxanes 19
2-3 Ozonolysis of Cephalomannine/Paclitaxel Mixture 21
2-4 Compounds from the Bark of Taxus brevifolia 23
2-5 Acetylation of lJ3-Hydroxy-7-Deacetyl Baccatin I 23
2-6 Structure of Abeo-Abietane Diterpenoids 25
2-7 NOE Correlations of Taxamairin A 25
2-8 Retro synthetic Analysis of Taxamairin B 27
2-9 Literature Synthesis of Taxamairin B 28
ix

Figure Page
2-10 Literature Synthesis ofTaxamairinB 30
2-11 Synthesis of Taxamairin B 32
2-12 Reactions of Wrong Regioselectivity 33
2-13 Structures of Taxiflorine and Related Compounds 37
2-14 Oxidation of Taxiflorine 37
2-15 HMBC Correlation of Taxiflorine Acetate 38
2-16 Compounds from the Needles of Taxus floridana 41
2-17 Synthesis of Trans-2, 6-Dimethoxy Cinnamaldehyde 43
3-1 Nitration of Paclitaxel, 7-OH > 2-OH 67
3-2 Nitration of 10-Deacetyl Paclitaxel-7-P-Xyloside l-OH, 2-OH,
3-OH > 10-OH > 2-OH 72
3-3 Regioselective Acetylation of 10-Deacetyl Paclitaxel-7-P-Xyloside 72
3-4 Conversion of 10-Deacetyl Paclitaxel Xyloside to 10-Deacetyl Paclitaxel... 73
3-5 Reductive Denitration of Paclitaxel 75
3-6 Hydrolysis of the Side-Chain of Paclitaxel-7, 2-Dinitrate with NaBH4 75
3-7 Regioselective Denitration of Paclitaxel-7, 2-Dinitrate with
Ammonium Sulfide 77
3-8 Enol Acetate Formation of Nitrate Esters 78
3-9 Reaction of 2-Nitrate Ester with NaNs 80
3-10 Oxidation of Paclitaxel-7-Nitrate Ester 82
3-11 Acetylation of Keto-Ester 83
x

Figure page
3-12 Mechanism of Keto-Ester Degradation 85
4-1 Synthesis of Ionizadle Analogues Ill
4-2 Condensation with Malonic Acid 112
4-3 Condensation with Nitromethane 114
4-4 Dialdehyde Reduction to Diol 116
4-5 Reductive Aminations 117
4-6 Synthesis of Glycosyl Donors 120
4-7 Koenigs-Knorr Glycosylation 122
4-8 Trichloroacetimidate Glycosylation 123
4-9 Rearrangement of 2-Acetyl Paclitaxel 124
xi

ABBREVIATIONS
Ac acetate
Bn benzyl
Bz benzoate
CIMS chemical ionization mass spectroscopy
DCC dicyclohexylcarbodiimide
DDQ 2, 3-dichloro-5, 6-dicyano-l, 4-benzoquinone
DMF dimethylformamide
DMSO dimethyl sulfoxide
EIMS electron impact mass spectroscopy
FABMS fast atom bombardment mass spectroscopy
HMBC heteronuclear multiple bond correlation
HPLC high pressure liquid chromatography
LAH lithium aluminum anhydride
LDA lithium diisopropylamide
NMR nuclear magnetic resonance
NOE nuclear Overhauser effect
PDC pyridinium dichromate
PTSA para-toluene sulphonic acid
RaNi rainey nickel
TBS tert-butyl dimethyl silyl
TES triethyl silyl
Tf triflate
TLC thin layer chromatography
TMEDA tetramethylethylenediamine
xii

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
CHEMISTRY OF TAXANES AND TAXUS SPECIES
By
James Harvey Johnson Jr.
December 1998
Chairman: Koppaka V. Rao (deceased)
Co-Chairman: John Perrin
Major Department: Medicinal Chemistry
The chemistry of both taxane diterpenoids and Taxus species are studied. Several
taxanes are isolation from both Taxus brevifolia and Taxus floridana. In addition to these
taxanes several non-taxane compounds are also isolated. One of these, trans-2, 6-
dimethoxycinnamaldehyde is a novel structure that has been synthesized. Another non-
taxane, taxamairin B, which belongs to rare class of diterpenes, is synthesized by a more
efficient route than was reported in the literature. Nitrate ester forming reactions are also
studied with paclitaxel and closely related analogues. It is shown that this reaction is
regioselective in many cases and thus illustrates the potential use of nitrate esters as
protecting groups in taxane chemistry. Also several unexpected reactions of these nitrate
esters are explored including an unusually rearrangement in which the nitrated paclitaxel

side chain reacts in mild base to yield the corresponding baccatin III and dibenzamide.
Finally, a series of potentially more water soluble paclitaxel analogues are prepared by
oxidizing the naturally occurring 10-deacetyl paclitaxel-7-xyloside with periodate and then
reacting the resulting dialdehyde with amines and carbon nucleophiles. These compounds
are also tested for cytotoxicity in the L1210 assay system. Although these compounds are
not as active as paclitaxel most are more active than the xyloside from which they are
obtained.
xiv

CHAPTER 1
HISTORY AND BACKGROUND OF PACLITAXEL
Introduction
Paclitaxel (Taxol1M) (1) is a potent antitumor agent that was originally isolated
from the Pacific yew tree, Taxus brevifolia, by Wall and Wani in 1971 (Wani et al., 1971).
The structure of paclitaxel is that of a tetracyclic diterpene ester (Figure 1-1). Its structure
has several unusual features including an oxetane ring, an 8-membered B-ring, a
bridgehead double bond, 11 asymmetric centers, and a N-substituted phenylisoserine ester.
This structure has provided a great challenge to synthetic organic chemists since its
elucidation but was conquered initially by two groups simultaneously and by other groups
since (Nicolaou et al., 1994; Holton et al., 1994a; Holton et al., 1994b). The ongoing
phytochemical study of Taxus brevifolia as well as other Taxus sp. has yielded many
taxanes other than paclitaxel Compounds with rearranged ring systems, various types of
esters, as well as glycosides have been isolated and several reviews have been published
(Kingston et al., 1993; Das et al., 1995; Appendino, 1995).
Early Work and Structural Elucidation
The pioneering work concerning paclitaxel began in the late 1950s when the
National Cancer Institute started a screening program of plant extracts using tumor
I

2
O
Figure 1-1: Structure of Paclitaxel
systems models in vivo and tumor cell lines. From these studies the stem bark extract of
the Pacific yew tree was shown to display cytotoxicity in the KB assay and also activity
against carcinosarcoma in rats and leukemia in mice. In connection with this NCI
screening program, Wall and his collaborators studied the in vitro bioassay guided
fractionation of the active extract and in 1969 paclitaxel was isolated and shown to be the
most active constituent of the extract. This isolation was carried out by extracting the
dried stem bark with 95% ethanol. The extract was then partitioned between water and 4 :
1 chloroform : methanol. The organic layer was evaporated to a solid and purified by a 3-
step Craig countercurrent distribution method which yielded paclitaxel in a yield of
0.004%, As soon as paclitaxel had been isolated in pure form, the structure of the
compound was investigated using available spectroscopic methods. Although methods for
ultraviolet, infrared, and mass spectrometry were at a reasonably advanced stage in the

late 1960s, NMR was relatively primitive compared to the sophisticated instrumentation
and procedures now available. Some of the physical and chemical properties of paclitaxel
are shown in Table 1-1 and the *H NMR spectrum is shown in Figure 1-2.
Table 1-1: Physical and Chemical Properties of Paclitaxel
1.) Needles from 50% aqueous methanol or ether
2.) mp 213-216 C
3.) [q]D20 -49,6 (MeOH)
4.) Unstable towards mineral acid and base
5.) Forms mono and diacetate
6.) Analysis Caled, for C47H51NOi4: C, 66.11; H, 6.20; N, 1.64
Found: C, 65.98; H, 6,10; N, 1.57. Required m/z 853. Found m/z 853
7.) UV ^ax (MeOH) 227 nm (s 29,800)
It was evident by this time that paclitaxel probably contained the taxane skeleton.
A number of taxane derivatives had been reported in previous literature. It was evident
that paclitaxel was more complex than previously reported taxanes since its molecular
weight from high resolution mass spectrometry was C^HsiNOu, corresponding to a
molecular weight of 853. The evidence then indicated that paclitaxel was comprised of a
taxane nucleus to which an ester was attached, as preliminary experiments indicated that
an ester was easily cleaved from the rest of the molecule. Attempts were made to prepare
crystalline halogenated derivatives of paclitaxel, however none had properties suitable for
x-ray analysis. Paclitaxel was therefore subjected to a mild base catalyzed methanolysis at
0 C, which yielded a nitrogen containing a-hydroxy methyl ester, C17H17NCX1, a tetraol,
C29H36Oio, and methyl acetate. The methyl ester thus obtained by the mild methanolysis
procedure was converted to a parabromobenzoate ester (2) and characterized by x-ray
analysis as C24H2oBrN05 with the structure shown in Figure 1-2. The ester may be

4
Figure 1-2: 'H NMR Spectrum of Paclitaxel
vu;

5
O
I
3
Figure 1-3: Halogenated Products of Methanolysis
Used for X-Ray Crystallography
regarded as an N-benzoyl derivative of (2R, 3S)-3-phenylisoserine. The tetraol formed by
the methanolysis of paclitaxel was converted to a bisiodoacetate (3), C33H38I2O12, which
again received x-ray analysis. The structure is shown in Figure 1-3.

6
Since the ester could have originally been joined to hydroxyl groups at either C-7,
C-10, or C--13, it was necessary to establish at which of these hydroxyl moieties the ester
had originally been located. When paclitaxel was oxidized with MnC>2 under neutral
conditions, no reaction occurred. However, Mn02 oxidation of paclitaxel under alkaline
conditions smoothly yielded a reaction product (4) with the structure shown in Figure 1-4.
It is evident that Mn02 oxidation of paclitaxel under neutral conditions did not effect the
hydroxyl groups available for oxidation at C-7 and C-2. When paclitaxel was oxidized
with alkaline MnC>2 an analogue of baccatin III with a conjugated carbonyl moiety as
shown in Figure 1-4 was obtained. It is well known that MnCh oxidation of allylic
hydroxyl groups under alkaline conditions smoothly forms the corresponding conjugated
ketone. This reaction in conjunction with the x-ray structure determination of the
structures of the ester and taxane moieties established the structure of paclitaxel (Wani et
al 1971).
Mechanism of Action
Although paclitaxel displayed good activity against human tumor xenographs and
murine B16 melanoma, its cytotoxic properties were not very different from other drugs
being tested during the 1970s. Rather, its attraction to pharmacologists was its unique
structure which suggested the possibility of a novel mechanism for an anti-tumor drug.
This mechanism was subsequently identified in 1979 by Horwitz and collaborators (Schiff
et al., 1979). Paclitaxel proved to be a potent inhibitor of eukaryotic cell replication,
blocking cells in the late G2-M phase of the cell cycle. It is an unusual mitotic inhibitor

7
Mn02 (pH 7) T
Pachtaxel f ->- No Reac.
aq. acetone, reflux
Paclitaxel
MnQ2 (pH 8)
aq. acetone,
reflux
O
Figure 1-4: Neutral and Alkaline Oxidation of Paclitaxel
because, unlike the vinca alkaloids and colchicine which inhibit microtubule assembly, it
promotes the formation of discrete bundles of stable microtubules that result from the
reorganization of the microtubule cytoskeleton. The novel characteristic of paclitaxel is its
ability to polymerize tubulin in vitro in the absence of guanosine 5-triphosphate (GTP),
which is normally required for tubulin assembly.
Total Synthesis
As mentioned, the total synthesis of paclitaxel has recently been achieved by
various groups, however one group alone cannot claim to be the first to accomplish this

8
daunting task as Nicolaou and Holton published their work simultaneously in two separate
journals. Their respective works are now briefly described.
Relying on the previous work of other groups as well as his own studies, Nicolaou
envisioned the late formation of the oxetane (D) ring, oxygenation/reduction of the C-13
position, and attachment of the side chain. Thus, the problem was perceived as limited to
the assembly of paclitaxels ABC ring system in either its fully functionalized form or a
form that would serve as its progenitor. Figure 1-5 displays the retrosynthetic analysis
involving the bond disconnections on which the synthetic strategy was based. Thus, in the
synthetic direction the following key operations were performed: 1) two fragments (7 and
8) representing precursors to rings A and C were coupled by a Shapiro reaction and a
McMurry coupling to assemble the ABC ring skeleton; 2) installment of the oxetane ring;
3) addition of the various substituents around the peripheries of rings B and C; 4)
oxygenation at C-13; and 5) esterification to attach the side chain. Both precursors to
rings A and C were made possible using the Diels-Alder transform which led to starting
materials that were either commercially available or known in the literature.
In contrast to the convergent synthesis by Nicolaou, Holton took a more linear
approach. The facile epimerization of paclitaxel at C-7 is well documented, and has been
postulated to occur via a retroaldol-aldol process. Holton chose therefore to pursue a
synthetic strategy in which this stereocenter would be introduced at an early stage and
carried throughout most of the synthesis in the absence of a C-9 carbonyl group, thereby
avoiding epimerization. Thus, his route to paclitaxel proceeds retrosynthetically through
the C-7 protected baccatin III 13 to the tricyclic ketone 14, which arises from C ring

9
O
y
1. Esterification
2. Oxygenation
3. McMurry Pinacol Coupling
4. Shapiro Coupling
TESO O
-N
Ph
Bz
6
/ Diels-AIder
/
OTBS OBn
Diels-AIder
Figure 1-5: Nicolaou Synthesis of Paclitaxel
closure of a precursor properly functionalized at C-l, C-2, C-3, C-7, and C-8 (15).
Synthesis of this precursor was made possible by conformational control of the eight

10
Figure 1-6: Holton Synthesis of Paclitaxel
membered B ring, via ketone 16. Ketone 16 was projected to arise from an aldol
condensation of the enolate of ketone 17, the formation of which would be made possible

11
Figure 1-7: Holton Synthesis of Paclitaxel
by conformational control exerted by the C-10a group. This scheme is shown in Figure 1-
6. The synthesis of this starting ketone begins with camphor (18). Camphor is converted
to P-patchouline (19) and its epoxide also known as Patchino (20). Patchino is then
converted to epoxide 21 and then rearranged to diol 22 which is followed by epoxy
alcohol fragmentation to give the starting ketone 23 as is shown in Figure 1-7.

12
Although there have been many publications concerning the synthesis of the phenyl
isoserine side chain, the most common and that which was used in both of these methods
is using the P-lactam 6 in Figure 1-5.
Structure-Activity Relationships
During the past decade a tremendous amount of work has been performed to
determine what parts of the structure of paclitaxel are necessary to illicit biological activity
and reviews have been published (Commercon et ah, 1995; Chen & Farina, 1995;
Kingston, 1995; Ojima et ah, 1995; Georg et ah, 1995). The structure of paclitaxel can be
divided into 3 sections when discussing structure-activity relationships and these are: 1)
the N-benzoyl phenylisoserine side chain; 2) the southern hemisphere including C-2, C-4,
and the oxetane ring; and 3) the northern hemisphere including C-7, C-9, and C-10.
The side chain plays a major role in the biological functions of this antitumor agent
and without the side chain the resulting baccatin III is inactive. Protection of the C-2
hydroxyl group results in major loss of activity in tubulin assays but if the group is labile
(acetate) then the activity remains in cell culture presumably acting as a prodrug.
Structural modifications of the paclitaxel side chain have been reported by several groups.
These studies reveal a number of interesting features and important findings include the
following: 1) the C-3 amide group is critical although the amides aryl group may be
substituted by other aryl or alkyl groups; 2) the C-3 aryl group is required since
replacement by a methyl group reduces activity but, if larger alkyl groups are used, the
activity remains; 3) the C-3 bound nitrogen can be replaced by an oxygen atom without

13
significant loss of activity; 4) one of the C-2 or C-3 polar functions can be removed
without significant effect, but the removal of both or interchange of their positions causes
dramatic loss of activity and; 5) the (2S, 3R) naturally occurring isomer is the most
active of the four possible isomers.
Concerning the southern hemisphere of the structure, this area is also very
important in terms of biological activity. First of all the oxetane ring is necessary for
activity. Structural and molecular modeling studies show that this 4-membered ring is
involved in a conformational lock of the diterpene skeleton and the C-13 side chain
through a pseudo chair conformation of ring C. The C-2 benzoyl group is also necessary
for activity as the C-2 debenzoyl paclitaxel showed little in vitro cytotoxicity; however,
some groups have shown that modified benzoyl groups or aryl acyl groups do retain the
cytotoxicity. The C-4 acetate is not as important as the C-2 benzoate in that if the acetate
is removed the activity is reduced only slightly.
The northern hemisphere is the least sensitive part of the paclitaxel structure. The
C-7 hydroxyl may be esterified, epimerized, or removed without significant loss of
activity. Specifically a xylosyl group at C-7 actually increases the activity in the tubulin
binding assay but leads to decreased activity in cell culture. Presumably there is a transport
problem associated with the xyloside that causes this decreased activity. The C-9 keto
group may be reduced which actually slightly improves activity and the C-10 acetyl or
acetoxy group may be removed without significant loss of activity.
It should also be said that contraction of the A ring does not reduce the tubulin-
disassembly inhibition activity very much in spite of the significant structural change

14
N-acyl group
required
\
acetyl or acetoxy
group may be
removed without
significant loss
of activity
phenyl group
or a close
analog required
free 2'-hydroxyl
group or a
hydrolysable
ester thereof
required
benzoyloxy group
essential; certain
substituted groups
have improved
activity
may be esterified
epimerized or
removed without
significant loss of
activity
oxetane ring
required for
activity
removal of acetate
reduces activity slight
Figure 1-8: Structure-Activity Relationships
implied by this conversion. These analogues, however, are much less cytotoxic in cell
culture which may be due to the instability of the A ring contracted analogues in cell
culture media or to its failure to enter the cell. A summary of structure-activity
relationships is shown in Figure 1-8.

CHAPTER 2
ISOLATION OF TAXOID AND NON-TAXOID COMPOUNDS
FROM TAXUS SPECIES
Large Scale Isolation Process
Although paclitaxel is one of the most promising anti-tumor drugs to receive FDA
approval in many years, it has been beset with many problems not the least of which is
adequate production of the drug. The original, and until recently major source of the
compound was the bark of the Pacific yew (Tcixits brevifolia), from which paclitaxel was
isolated in a yield of 0.01-0.013% on a large scale. Although several related taxanes that
can serve as precursors for the semi-synthesis of paclitaxel, for example, 10-deacetyl
baccatin III (25), co-occur in the bark with paclitaxel, there are no reports to indicate that
these are being isolated from the bark on a large scale. Thus the low yields of paclitaxel
realized by the original process, the apparent unavailability of other useful taxanes
analogues, and the environmental concerns raised by the need to cut the slow-growing
yew trees for harvesting the bark, are some of the reasons why the bark is no longer
considered an attractive source for the large scale production of paclitaxel.
Among the alternatives that are being actively studied are the following: 1)
isolation of 10-deacetyl baccatin III from the European yew (Tcixus baccatd) and its semi-
15

16
O
26 Ri = H, R2 = Ac, R3 phenyl
27 Ri = p-xylosyl, R2 = H, R3 = tiglyl
28 Ri = P-xylosyl, R2 = H, R3 = phenyl
29 Ri = p-xylosyl, R2 = H, R3 = n-pentyl
30 Rj = P-xylosyl, R2 = Ac, R3 = phenyl
31 Ri = H, R2 = H, R3 = phenyl
32 Rj = H, R2 = Ac, R3 = tiglyl
Figure 2-1: Structure of Major Taxanes
synthetic conversion to paclitaxel and 2) large-scale cultivation of the ornamental yew
(Taxus media Hicksii) and isolation of paclitaxel from its needles/twigs. Among the future
alternatives are total synthesis, of which various schemes have been published, and
isolation from large-scale plant cell culture.

17
In recent years this laboratory has developed a large-scale isolation procedure
using a single reverse-phase column (Rao, 1993; Rao et al. 1995). This procedure has
several advantages over other published procedures some of which are that it is much
simpler, gives higher yields of paclitaxel, and yields several other taxanes which can be
converted to paclitaxel. Specifically, the following yields are obtained for the major
compounds from the bark: paclitaxel (26) (0.04%), 10-deacetyl baccatin III (25) (0.02%),
10-deacetyl paclitaxel-7-(3-xyloside (28) (0.1%), 10-deacetyl paclitaxel-C-7-P-xyloside
(29) (0.04%), 10-deacetyl cephalomannine-7-p-xyloside (27) (0.006%), paclitaxel-7-p-
xyloside (30) (0.008%), 10-deacetyl paclitaxel (31) (0.008%), and cephalomannine (32)
(0.004%) (Figure 2-1). The procedure for this process is defined below and in Figure 2-2.
Air dried yew bark (200-250 lbs.) was extracted with 100 gallons of methanol in a
batch process a total of three times with each extraction lasting one day. The pooled
methanol extracts were concentrated under reduced pressure (<30 C) using a semi-
continuously operated still until the volume of the concentrate reached 20-25 gallons.
Extraction of the concentrated methanolic extract with chloroform was performed in 50-
100 gallon tanks equipped with an air-driven stirrer. The concentrate was stirred with
water and chloroform for about 30 minutes, then 2-14 hours were necessary to allow for
any emulsion to clear. The chloroform layer was drained off from the bottom and the
water layer was extracted two additional times. The pooled chloroform layers were
concentrated under a vacuum to a thick syrup which was poured into glass trays and
converted to powder using a vacuum oven at 35-40 C. The powder was obtained in a
yield of 18-26 g per kg of the bark.

18
For chromatography, stainless steel columns either 4 x 4 or 6 x 6 were used.
The columns were packed with C-18 bonded silica as a slurry in methanol. Approximately
3-4 kg and 12-13 kg of silica were used with the 4 and 6 columns, respectively. After a
thorough wash with methanol, the columns were equilibrated with 25% acetonitrile in
water. For running the 6 diameter column, the powder from the chloroform extract (2-
2.5 kg) was dissolved in acetonitrile (5 1) and while this mixture was being stirred with
equilibrated C-18 silica (1-2 1), it was diluted with water to make 20 1. The mixture was
then allowed to stand for 15-30 minutes and the clear supernatant siphoned off into
another container. The slurry was applied to the column, followed by part of the
supernatant, after which the column was sealed. The remaining supernatant was pumped
into the column using a diaphragm metering pump maintaining a pressure of 30-80 psi.
After the sample had been pumped onto the column it was eluted with a step gradient of
35, 40, 45, and 50% acetonitrile in water. The change in solvent was dictated by the
results of the TLC and HPLC of the fractions but usually 40-50 1 of each solvent was
used. After this, the column was washed with methanol, followed by a mixture of ethyl
acetate and ligroin until the effluent was nearly colorless. Following this, the column was
again washed with methanol and equilibrated with 25% acetonitrile in water. The column
fractions (about 2 1 each) were allowed to stand at room temperature for 2-8 days, by
which time crystals appeared in many. Soon after, the crystals were filtered, analyzed for
purity and composition by HPLC, and recrystallized from the appropriate solvent.
In terms of the elution sequence of the taxanes, the earliest taxane to appear was
10-deacetyl baccatin III (25) which crystallized from the fractions eluted by 35%

19
Methanol Extract
T
Partition Between Water and Chloroform
Water Chloroform
C18 Column
Acetonitrile/Water
Pure Paclitaxel Fractions Fractions Containing Paclitaxel
(Crystalline) and Cephalomannine
Paclitaxel 0.04%
10-Deacetyl Baccatin III 0.02%
10-Deacetyl PaclitaxeI-7-Xyloside 0.1%
10-Deacetyl Paclitaxel-C-7-Xyloside 0.04%
Ozonolysis/
Regular Silica Colum
T
Paclitaxel
Figure 2-2: Reverse-Phase Isolation of Taxanes

20
acetonitrile in water. The next group of taxanes to be eluted were the various xylosidic
taxanes; 10-deacetyl cephalomannine-7-p-xyloside (27), 10-deacetyl paclitaxel-7-P-
xyloside (28), 10-deacetyl paclitaxel-C-7-P-xyloside (29), and paclitaxel-7-p-xyloside
(30). Of these the first two were well separated. As the elution of 10-deacetyl paclitaxel-7-
P-xyloside was nearing completion, 10-deacetyl paclitaxel-C-7-P-xyloside started to elute.
Halfway though its elution, paclitaxel-7-P-xyloside and 10-deacetyl paclitaxel (31) started
to co-elute. These last two compounds also crystallized together, however separation was
readily achieved by running the mixture through a regular silica column using 0-5%
methanol in chloroform as solvent.
Continued elution of the column with 50% acetonitrile in water gave
cephalomannine (32), followed closely by paclitaxel (26). The earlier part of the band
contained mixtures of the two, but the later fractions contained mostly paclitaxel which
could be recrystallized. The fractions that contained the mixture were combined and dried
to a solid. This solid was then subjected to ozonolysis at -78 C for 45 minutes. This
process converted the cephalomannine to the keto-amide 34 but did not disturb paclitaxel
(Figure 2-3). After workup this material was run through a regular silica column with 0-
5% acetone in chloroform and the paclitaxel was isolated. It should be pointed out that
this process was necessary because paclitaxel and cephalomannine cannot be separated on
regular-phase silica.

21
+
Paclitaxel
03
ch2ci2/ch3oh
room temp.
Paclitaxel
Figure 2-3: Ozonolysis of Cephalomannine/Paclitaxel Mixture

22
Isolation of Minor Compounds from the Bark of Taxus brevifolia
Obviously the above process was only used to isolate major compounds; however,
many minor compounds exist in the filtrates or in the in-between fractions. A TLC analysis
of the filtrates from the region between 10-deacetyl baccatin III and 10-deacetyl paclitaxel
(xyloside fractions) showed many interesting compounds, the identity of which could not
be determined by comparison with available standards. This material was therefore
concentrated to a solid and rechromatographed on regular-phase silica using an elution
system of 0-5% acetone in dichloromethane to 0-5% methanol in 5%
acetone/dichloromethane.
The first compound to be eluted was 1 (3-hydroxy baccatin I (35). This compound
has been known for quite some time and was first isolated from Taxus baccata in 1970
(Della Casa De Marcano & Halsall, 1970). This compound belongs to the baccatin I sub
family because it contains a C-4-C-20 epoxide as opposed to an oxetane ring. Baccatin VI
(36) was eluted next and is another well known taxane isolated for the first time from
Taxus baccata in 1975 (Della Casa De Marcano & Halsall, 1975). This compound is so-
named because it is esterified at C-9 as opposed to paclitaxel/ 10-deacetyl baccatin III
which have a C-9 ketone (Figure 2-4).
The next compound to be eluted was l(3-hydroxy-7-deacetyl baccatin I (37). This
compound was recently isolated from the needles of Taxus brevifolia (Chu et al., 1993)
and has been reported to undergo an acetyl migration from C-9 to C-7 when kept in
solution. Indeed, this compound did form another spot on TLC when left in solution;

23
OAc
OAc
Figure 2-4: Compounds from the Bark of Taxus brevifolia
37
acetic anhydride/pyridine
18 hours, room temp.
Y
35
Figure 2-5: Acetylation of 1 P-Hydroxy-7-Deacetyl Baccatin I

24
however, no attempt was made to determine if this new compound was the C-7-acetyl, C-
9-hydroxy isomer. To further confirm its structure 37 was acetylated with acetic anhydride
and pyridine and the product matched 1 [3-hydroxy baccatin I (35) in every way (Figure 2-
5). Finally 9-dihydro-13-acetyl baccatin III (38) was eluted as determined by NMR
spectroscopy (Figure 2-4). This compound was isolated earlier from the needles of Taxns
canadensis (Gunawardana et al., 1992), but the current isolation is the first from the bark
of Taxns brevifolia. This compound has received much attention from Abbott
Laboratories as a possible precursor to their own paclitaxel analogue.
In addition to this work on the pre-paclitaxel fractions another interesting
compound was isolated while attempting to obtain more paclitaxel from the filtrates of
paclitaxel-containing reverse-phase fractions. This crystalline compound was eluted with a
solvent mixture of 2% acetone in dichloromethane and was given the name brevixanthane
because of its yellow color. Based on 'H and L"C NMR spectra it was quickly concluded
that brevixanthane belonged to a rare group of diterpenes known as 9( 10>20)-abeo-
abietane diterpenoids that have previously been isolated from Taxns species. These
diterpenes consist of a 6, 7, 6 tricyclic carbon ring system with ring C being aromatic and
are a novel diterpene structural class. The only other members of this group include
taxamairin A (38) and B (39) from Taxns chinensis var. Mairei (Liang et al., 1987) and
brevitaxin (40) from Taxns brevifolia which contains a C6-C3 side chain (Arslanian et al.,
1995) (Figure 2-6). Initially, we thought this compound may be novel based on its 'H
NMR spectrum. Brevixanthane contained only one methoxyl which had a chemical shift of
3.89 ppm while the methoxyl of taxamairin A was reported to be at 3.99 ppm. All other

25
och3
Figure 2-6: Structure of Abeo-Abietane Diterpenoids
Figure 2-7: NOE Correlations of Taxamairin A

26
chemical shifts were almost identical. Thus, it was speculated that the methoxyl/hydroxyl
positions may be the reverse of that of taxamairin A. However, it was then observed that
the UV spectrum of brevixanthane was identical to that reported for taxamairin A while
that of taxamairin B was reported to be completely different. This was confirmed by
synthesizing taxamairin B from brevixanthane with dimethylsulfate. Thus, since the UV
spectra of taxamairins A and B are so different it stands to reason that the UV spectra of
taxamairin A and brevixanthane should also be different if they were different structures.
This was not the case. To solve this question of structure !H NMR NOE experiments
were performed. These experiments illustrated that if the methoxyl methyl of
brevixanthane is irradiated then the phenolic proton and the isopropyl methyne proton are
enhanced proving the proximity of the methoxyl to the isopropyl group. Also when the
phenol proton was irradiated the methoxyl protons and the C-4 proton was enhanced
proving a close proximity between the phenolic group with and carbon 4 of the B ring
(Figure 2-7). Thus, it is concluded that brevixanthane was the same compound as
taxamairin A.
Synthesis of Taxamairin B
The total synthesis of taxamairin diterpenes has been accomplished by one other
group (Wang & Pan, 1995a; Wang et al., 1995b). The synthetic strategy which was used
by this group was derived from the retro synthetic analysis as outlined in Figure 2-8.
Ketone 45 was the key synthetic intermediate because it contains the entire carboncyclic
framework of taxamairin B. Thus the A ring precursor (49) was readily obtained from 1,3

27
Figure 2-8: Retrosynthetic Analysis of Taxamairin B
-cyclohexanedione (48) by azeotropic removal of water from a benzene/hexane/isopropyl
alcohol solution with PTSA as catalyst in a yield of 95% (Figure 2-9). The synthesis of the
C ring was begun by oxidizing o-vanillin (50) with AgO to give the carboxylic acid. This
phenolic acid was then dimethylated using dimethylsulfate to yield the methyl ester (51).
This is then treated with two equivalents of methyl magnesium bromide to yield the

28
2-propanol, hexane
PTSA
48
49
O
1.) AgO, NaOH
H2O, 70 C, 20 min
K2CO3, acetone 51
50
20% H2S04
<
reflux, 6 hrs
H2, Ra Ni
O
'OCH3
OCH3
OCH3
2 CH3MgBr
Et20
OCH3
52
OCH3
1.) n-butyl lithium,
TMEDA, THF
O C
2.) (CH20)n HocH2
OCH3
54
PBr3, CH2CI2
room temp., 15 min
Figure 2-9: Literature Synthesis of Taxamairin B

29
tertiary alcohol (52). Dehydration was then achieved by heating with 20% H2SO4. The
resulting olefin (53) was then reduced with RaNi and H2. A hydroxymethyl group was
attached ortho to the methoxyl by treating 54 with n-butyl lithium in THF and TMEDA
and later addition of paraformaldehyde. The hydroxymethyl function was converted to a
bromomethyl function using PBr3 in dichloromethane. This bromo compound (56) then
underwent nucleophilic substitution with ketone 49 and LDA as the base to yield ketone
57 (Figure 2-10). Ketone 57 was treated with vinyl magnesium bromide and the B ring
was closed by a Friedal-Crafts alkylation using BF3-OEt2. Finally the gem-dimethyls were
attached using potassium t-butoxide and methyl iodide to produce the key intermediate 60
in which the olefin bond moved into the cycloheptane ring. Oxidation of the allylic and
benzylic methylene group to produce the ketone at C-l was achieved by using excess
aqueous 75% t-butyl hydroperoxide and catalytic amounts of chromic anhydride. Finally,
olefination of C-4, 5 and C-6, 7 was achieved by heating with excess DDQ in toluene
followed by hydrogenation with 10% Pd-C. This then gave the final product taxamairin B.
Since the above synthesis was the only synthesis for this class of diterpenes it was
decided that it would be a noteworthy side project to synthesis this type of diterpene by a
simpler method. This method also uses a convergent approach by constructing an A ring
precursor and a C ring precursor and then bringing the precursors together to form the B
ring. The A ring precursor was synthesized by modifying a known method starting with 1,
3-dicyclohexanedione (63) (Shuzi et al., 1991). This diketone was dimethylated with
methyl iodide and K2C03 in refluxing acetone in give a yield of about 65% after vacuum
distillation. The 2, 2-dimethyl-1, 3-cyclohexanedione product (64) was then mono-

30
49 + 56
BF3-Et20
toluene
t-BuOOH
Ci o3, CH2C12
y
Figure 2-10: Literature Synthesis of Taxamairin B
brominated by slowly adding Br2 in dichloromethane at room temperature while closely
monitoring the TLC. This process gave a yield of 60% after purifying the product by
column chromatography. Finally, 4-bromo-2, 2-dimethyl-l, 3-cyclohexanedione (65) was

31
dehydrohalogenated by refluxing with excess LiCl in DMF for 2 hours. This process gave
a yield of 82% after purification by column chromatography of the A ring precursor 2, 2-
dimethyl-4-cyclohexene-l, 3-dione (66) (Figure 2-11).
The formation of the C ring precursor was more problematic. Acetovanillone (67)
was used as the starting material and this compound was isopropylated by heating with
90% H2S04 and isopropyl alcohol at 60 C for 36 hours. Unfortunately this reaction could
not be pushed beyond 50% conversion based on TLC of the reaction mixture, and after
purification by column chromatography gave a yield of 40%. However this process is a
better alternative than the 5-step process outlined in the previous synthesis for placing the
isopropyl group on the ring. Following this, the phenolic group was methylated using
dimethylsulfate and K2C03 in refluxing acetone for 2 hours. This process was nearly
quantitative to yield the dimethoxy product (68) (Figure 2-11). At this point a one carbon
oxygenated substituent had to be introduced between the acetyl and methoxyl groups.
Initially, a Vilsmeier-Haack reaction was attempted, but this gave a variety of products in
which the acetyl seemed to undergo some reaction; however, no effort was made to
characterize these products. Undesirable reactions also occurred with this method if the
acetyl was first reduced to an alcohol group or completely reduced to an ethyl group.
Attempts were also made to acetoxymethylate the desired position so the resulting acetate
could be hydrolyzed and the alcohol oxidized to the aldehyde. The reaction was performed
using 85% H3PO4, acetic anhydride, and paraformaldehyde and the chosen substrate was
the reduced ethyl compound (73) which has less steric bulk and is more activated than the

32
ch3i
K¡coT
63
O
1.) n-BuLi
TMEDA
CH20, -780C
OCH3
69 OCH3
66 + 70
Figure 2-11: Synthesis of Taxamairin B

33
Figure 2-12: Reactions of Wrong Regioselectivity

34
oxygenated analogues. This substrate was obtained by reducing ketone 72 with NaCNBH3
in the presence of ZnCl2 (Figure 2-12). Although acetoxymethylation proceeded quite well
the regioselectivity was wrong as was determined by NOE experiments (Figure 2-12).
These experiments clearly showed enhancement of one of the methoxy methyls when the
aromatic proton was irradiated; likewise when the oxygenated methylene was irradiated
the isopropyl methyne signal was enhanced.
At this point lithiation of the aromatic ring and reaction with paraformaldehyde
seemed to be a more attractive way of introducing a hydroxymethyl group which could
then be oxidized to the aldehyde. Initially, lithium/halogen exchange was the desired
process but bromination of the reduced alcohol substrate (69) as well as the totally
reduced ethyl substrate (73) yielded a brominated product with the wrong regiochemistry
(75, 76) (Figure 2-12). This conclusion was reached following NOE experiments as
previously described above.
In light of these results a direct lithiation was attempted with n-butyllithium in
diethyl ether and TMEDA at -78 C using the reduced alcohol substrate (69) (Figure 2-
11). It is well know that lithium complexes with methoxy groups and therefore addition
usually takes place ortho to a methoxy if one is present on the aromatic ring. About 30-45
minutes after adding the n-butyllithium, paraformaldehyde was added. This reaction
proceeded smoothly however yields were only around 50%. In any event, the
regioselectivity was as desired based on NOE with the hydroxymethyl group adding ortho
to the methoxy.

35
The resulting diol was then oxidized to the keto-aldehyde (70) with the mild
oxidizing reagent PDC. This product was coupled to 2, 2-dimethyl-4-cyclohexene-l, 3-
dione (66) by a tandem cross-aldol reaction in pyridine and piperidine (Figure 2-11).
Although other products were produced the desired product was the major one. This
pathway was expected since the most acidic carbon would be adding to the most
electrophilic carbon first. Once this occurs, the 7-membered ring would be expected to
close rather easily by another aldol reaction. The minor products were undoubtedly a
variety of other cross-aldol products. This major product matched taxamairin B, which
was obtained by methylating taxamairin A, in every way.
Isolation of Minor Compounds from Taxus floridana
The same reverse-phase chromatography protocol previously described was
applied to the needles of the Florida yew (Taxus floridana) with very good results (Rao et
al., 1996a). Although there was some question initially if this protocol would work on
needles as well as on bark because of the greater content of waxes and pigments found in
the needles, these questions were put to rest as all of the lipophilic material remained on
the column while using the appropriate taxane solvent system without clogging up the
column. It was found that paclitaxel could be isolated from these needles in a yield of
0.01% and 10-deacetyl baccatin III could be obtained in yields of 0.06%. Again, by TLC
analysis, this time of the pre-paclitaxel fractions, it was found that many unidentifiable
compounds were present in the filtrates. These filtrates were combined and evaporated to

36
dryness and reapplied to a regular-phase silica column using dichloromethane with 0-10%
acetone and then 10% acetone with 0-10% methanol in dichloromethane.
A few compounds eluted with straight dichloromethane and thus had to be run on
another column. This work will be discussed later. Elution with 2%
acetone/dichloromethane gave 1 (3-hydroxy baccatin I (35, Figure 2-4) as mentioned before
(Della Casa De Marcano & Halsall, 1970). This was followed by a compound that was
earlier given the name taxiflorine (77). Taxiflorine is an example of an 11(15>1)-
abeotaxane meaning that the A ring has contracted to contain only 5 carbons. Taxiflorine
was previously isolated by our group but was published with an incorrect structure
assignment (78) (Rao et al., 1996a). Taxiflorine itself gives difficult to interpret 'H and ljC
NMR spectra for reasons that will be discussed later; however, upon acetylation the
spectra are easier to interpret. According to the original structural assignment (78), its
acetate should be the same compound as 13-acetyl-13-decinnamoyl taxchinin B (79)
previously isolated by another group (Das et al., 1995); however, the spectral properties
of these two compounds did not match (Table 2-1). It was then postulated that the correct
structure of taxiflorine is one in which the C-10 benzoate and C-9 hydroxy groups are
reversed so that the hydroxyl is at the C-10 position. To confirm this idea taxiflorine was
oxidized with Jones reagent to the ketone (80, Figure 2-14) and its 1jC spectrum was
compared with those of some known C-9 and C-10 keto taxanes. The carbonyl signal of
80 seen at 192.2 ppm is consistent with an a, (3-unsaturated ketone system, and in contrast
to the 199-204 ppm signal of C-9 keto taxanes. Similarly, the C-12 signal of 80 is at 156.8

37
AcO
Figure 2-13: Structure of Taxiflorine and Related Compounds
Figure 2-14: Oxidation of Taxiflorine
ppm which is also consistent with the ¡3-carbon of an a, (3-unsaturated ketone system. The
C-12 signal in C-9 keto compounds is usually around 147.0 ppm. Also, the UV spectrum

38
Figure 2-15: HMBC Correlation of Taxiflorine Acetate
of 80 displayed a A,x at 232 nm with a shoulder at 253 nm, this shoulder is consistent
with an a, (3-unsaturated ketone system. Finally, to confirm the revised structure an
HMBC spectrum was taken on the taxiflorine acetate (81, Figure 2-15). From this
spectrum the interactions between the ortho protons and benzoate carbonyl carbon were
clearly visible as was the interaction between the benzoate carbonyl carbon and the proton
at C-9. Likewise the C-19 protons interacted with the C-9 carbon, which in turn interacted
with the C-9 protons on the regular HETCOR spectrum. With this information in hand the

39
benzoate could be firmly placed at the C-9 position giving the correct structure of
taxiflorine. This correcting structure was identical to a compound recently isolated from
Taxus chinensis var. Mairei and given the name taxchinin M (Tanaka et al., 1996).
Elution with 5% acetone in dichloromethane yielded (-) rhododendrol (82, Figure
2-16) which has been reported previously in Taxus brevifolia (Chu et al., 1994) and
Betula pndula (Smite et al., 1993). This was followed by 13-deacetyl taxiflorine (83,
Figure 2-16). Like taxiflorine, 83 also exhibited a !H spectrum which contained very broad
rounded peaks, whereas the spectrum of its acetate was normal. Also this acetate was
identical to the acetate of taxiflorine. It was also discovered that it was possible to get a
better spectrum of both these compounds (taxiflorine and 13-deacetyl taxiflorine) if the
spectra were run at lower temperatures. Therefore !H and COSY spectra of 83 were
Table 2-1: *H and LC NMR Values for Related Abeo-Taxanes
HorC#
Compd. 81
Compd. 79
Compd. 80
1
****, 67.3
****, 68.5
****, 65.5
2
6.16 d (7.2Hz),
67.8
6.17 d (7.9Hz), 67.8
6.22 d (7.5Hz), 68.7
3
3.00 d (7.5Hz),
43.7
3.01 d (7.9Hz), 44.7
3.12 d (7.8Hz), 44.1
4
**** 73 g
**** yc) 2
**** 'jg q
5
4.96 d (7.2Hz),
84.6
4.99 d (7.6Hz), 84.6
5.00 d (6.0Hz), 84.9
6a
2.67 m, 34.5
2.60 m, 34.7
2.74 m, 34.3
6(3
1.77 m, ****
1.91 m, ****
1 84 m ****
7
5.54 t (7.8Hz), 70.2
5.59 t (8.2Hz), 70.6
5.16 t (7.5Hz), 71.0
8
**** 42 5
****, 43.5
**** 44 2
9
6.32 d (10.8), 77.2
6.21 d (10.9Hz), 76.3
6.32 s, 83.6
10
6.43 d (10.5), 67.6
6.58 d (10.9Hz), 68.8
**** 2922
11
****, 135.8
****, 135.7
****, 137.5
12
****, 146.8
****, 147.7
****, 156.8
13
5.61 t (6.9Hz), 78.4
5.62 t (7.7Hz), 78.7
5.72 t (7.2Hz), 78.9
14a
1.68 m, 36.6
2.50 m, 36.7
1.76 dd (8.1, 14.7Hz), 37.1

40
Table 2-1 continued
HorC#
Compd. 81
Compd.79
Compd.80
14p
2.26 m, ****
2.00 m, ****
2.40 dd (7.2, 14.4Hz),
15
****; 75 2
Hi*** 7^ 7
**** 752
16
1.17 s, 27.3
1.15 s, 27.7
1.22 s, 25.5
17
1.20 s, 24.9
1.07 s, 25.4
1.18 s, 27.4
18
1.87 s, 11.5
2.01 s, 11.9
2.08 s, 13.8
19
1.77 s, 13.0
1.68 s, 12.4
1.91 s, 13.6
20a
4.49 d (7.2Hz),
74.3
4.50 d (7.7Hz), 74.5
4.56 d (7.2Hz), 74.7
20(3
4.41 d (7.2Hz),
4.41 d (7.7Hz), ****
4.45 d (7.2Hz), ****
q-Bz
**** 129 3
**** 229 0
**** 229 4
o-Bz
7.92 d (7.2Hz),
129.5
7.86 d (7.8Hz), 129.4
8.07 d (7.2Hz), 129.8
m-Bz
7.43 t (7.8Hz),
128.2
7.43 t (7.8Hz), 128.6
7.45 t (8.1Hz), 128.3
p-Bz
7.56 t (7.5Hz),
133.1
7.55 t (7.8Hz), 133.3
7.58 t (7.2Hz), 133.2
HorC#
Compd. 81
Compd. 79
Compd. 80
OCOCH3
1.65 s, 1.82 s, 2.03
s, 2.13 s, 2.14 s,
20.5, 21.0, 21.6,
21.7, 21.9
1.74 s, 2.02 s, 2.02 s,
2.08 s, 2.12 s, 20.6,
21.3, 21.4, 21.6, 22.0
2.01 s, 2.05 s, 2.11 s, 2.13
s, 20.9, 21.5, 21.6, 21.8
c=o
166.2, 167.8,
168.9, 169.7,
170.1, 170.3
163.9, 168.9, 169.0,
169.7, 170.2, 170.4
166.8, 169.0, 169.6, 170.3,
170.4
obtained at -10 C and the structure was determined to be that which is shown and which
was previously isolated by another group from Taxus chinensis var. Maiei and given the
name taxchinin L (Tanaka et al., 1996). It has been reported in the literature that
abeotaxanes which contain a C-9 benzoate and a C-10 hydroxyl group usually give proton
spectra in which the peaks are broad and rounded (Rao & Juchum, 1998). This is because

41
O
Figure 2-16: Compounds from the Needles of Taxus floridana
these compounds seem to exist in an equilibrium between two conformers. If the spectrum
is taken at low temperature however, two sets of signals can be distinguished.

42
On elution with 10% acetone/dichloromethane additional amounts of 10-deacetyl
baccatin III were obtained; and with 2% methanol/10% acetone/dichloromethane the
polyhydroxylated steroid ponasterone A (84) was isolated (Figure 2-16). This compound
was previously isolated from Taxus brevifolia (Rao et al., 1996b). Finally, with 5-10%
methanol/10% acetone/dichloromethane 10-deacetyl paclitaxel-7-(3-xyloside was isolated
for the first time from Taxus floridana.
The fractions mentioned earlier that were eluted with dichloromethane were
concentrated and put on another silica column using 25% ethyl acetate/ligroin as the
starting mobile phase. At 30% ethyl acetate/ligroin trans-2,6-dimethoxy cinnamaldehyde
(85) was eluted; and it will be discussed later. Elution continued with 50% ethyl
acetate/ligroin, which gave the lignan a-conidendrin (86) followed by 1 -deoxy baccatin IV
(87); both of which have been previously isolated (Figure 2-16) (Miller et al., 1982;
Miller, 1980).
Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde
As mentioned above trans-2, 6-dimethoxy cinnamaldehyde was one of the
compounds isolated from Taxus floridana. Although this structure was determined quite
easily based on ]H and 13C NMR spectra, di-ortho oxy substituted C6-C3 compounds had
previously not been isolated from natural products. Thus 85 was synthesized to verify the
structure following Figure 2-17. Thus trans-2, 6-dimethoxy cinnamic acid (88) was
methylated with methanol and H2S04 followed by reduction to the alcohol (89) with LAH
in a total yield of about 55%. The alcohol was then oxidized to the aldehyde (90) with

43
Jone's oxidation
Malonic acid
Pyridine/piperidine ,
90 OCII3
1.) Dimethylsulfate
2.) LAH
Figure 2-17: Synthesis of Trans 2, 6-Dimethoxy Cinnamaldehyde
Jones reagent in a yield of 79% and the aldehyde was then condensed with malonic acid to
yield the corresponding cinnamic acid (91) in about 85% yield. This acid was again
methylated with dimethylsulfate and reduced with LAH to the alcohol (92) to give a total
yield of 60%. Finally 92 was oxidized to the desired aldehyde (93) using the mild oxidizing
agent PDC, to prevent further oxidation, in a yield of 38% (Figure 2-17). This aldehyde

44
was identical to the natural product in all ways. A thorough search of the literature
confirmed that this was a novel compound and that no other 2, 6-dioxy cinnamyl
compound has been found in nature.
Experimental
All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of
the reaction. All NMR spectra were recorded using either a Varan VXR-300 or a Varan
Gemini-300 spectrophotometer using CDCI3 as solvent. Infrared spectra were obtained
using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolet spectra were
obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were
recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a
Fisher melting point apparatus. Column chromatography was accomplished using 100-200
mesh silica gel.
Isolation of Minor Compounds from Taxus brevifolia
The filtrates from the region between 10-deacetyl baccatin III and 10-deacetyl
paclitaxel on the reverse-phase chromatographic separation were concentrated to a syrup
(400 g). A 5 g aliquot was applied to a normal-phase silica column (100 g) in
dichloromethane, and chromatographed with an elution sequence consisting of 1-5%
acetone and then 5% acetone and 1-5% methanol. A total of 200 ml of each solvent
mixture was used before progressing to the next solvent system and fractions of about 20
ml were collected and monitored by TLC. The order of elution was as follows: ip-

45
hydroxy baccatin I, baccatin VI, ip-hydroxy-7-deacetyl baccatin I, and 9-dihydro-13-
acetyl baccatin III.
13-Hydroxy baccatin I (35)
The compound was eluted with 2% acetone in dichloromethane to give 281 mg of
35 crystallized from diethyl ether and ligroin. White crystalline powder, mp 259-261 C,
HNMR5: 1.24 (s, 3H, 17-H), 1.25 (s, 3H, 19-H), 1.65 (s, 3H, 16-H), 1.80 (m, 1H, 6-
H3), 1.90 (m, 1H, 14-H3), 2.00 (s, 3H, OAc), 2.05 (s, 3H, OAc), 2.06 (s, 3H, OAc), 2.06
(s, 3H, OAc), 2.09 (s, 3H, OAc), 2.12 (s, 3H, OAc), 2.18 (m, 1H, 6-Ha), 2.22 (s, 3H, 18-
H), 2.32 (d, 4.8Hz, 1H, 20-Hp), 2.54 (dd, 9.9, 15.3Hz, 1H, 14-Ha), 3.19 (d, 3.6Hz, 1H,
3-H), 3.56 (d, 5.4Hz, 1H, 20-Ha), 4.22 (br s, 1H, 5-H), 5.49 (m, 2H, 2-H, 7-H), 6.05 (d,
11.1Hz, 1H, 9-H), 6.09 (t, 7.8Hz, 1H, 13-H), 6.22 (d, 11.4Hz, 1H, 10-H). 13C NMR 5:
13.6, 15.4, 20.6, 20.8, 20.9, 21.3, 21.4, 21.6, 21.8, 28.4, 31.0, 38.5, 41.3, 43.2, 46.6,
49.9, 58.3, 68.7, 70.7, 71.1, 72.1, 75.1, 76.0, 77.7, 135.6, 140.3, 169.0, 169.2, 169.3,
169.7, 169.8, 170.1.
Baccatin VI (36)
The compound was eluted with 4% acetone in dichloromethane to give 362 mg of
36 crystallized from diethyl ether and ligroin. White crystalline powder, mp 245-247 C,
*H NMR 8: 1.23 (s, 3H, 17-H), 1.61 (s, 3H, 19-H), 1.79 (s, 3H, 16-H), 1.87 (m, 1H, 6-
HP), 2.00 (s, 3H, OAc), 2.03 (s, 3H, 18-H), 2.10 (s, 3H, OAc), 2.11 (s, 3H, OAc), 2.17
(m, 2H, 14-Ha,p), 2.20 (s, 3H, OAc), 2.29 (s, 3H, OAc), 2.51 (m, 1H, 6-Ha), 3.18 (d,
5.7Hz, 1H, 3-H), 4.13 (d, 8.4Hz, 1H, 20-Hp), 4.33 (d, 8.4Hz, 1H, 20-Ha), 4.97 (d,
9.0Hz, 1H, 5-H), 5.87 (d, 6.0Hz, 1H, 2-H), 6.00 (d, 11.4Hz, 1H, 9-H), 6.17 (t, 8.1Hz,

46
1H, 13-H), 6.22 (d, 11.1Hz, 1H, 10-H), 7.48 (t, 7.8Hz, 2H, m-Bz), 7.61 (t, 7.2Hz, 1H, p-
Bz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 12.8, 14.9, 20.7, 20.9, 21.2, 21.4, 22.3,
22.7, 28.3, 34.5, 35.1, 42.8, 45.8, 47.3, 69.7, 70.4, 71.8, 73.3, 75.0, 76.4, 78.9, 81.5,
83.8, 128.6, 129.3, 130.1, 133.7, 135.6, 141.2, 166.9, 168.9, 169.1, 169.8, 170.1, 170.4.
l{3-Hydroxy-7-deacetyl baccatin I (37)
The compound was eluted with 4% acetone in dichloromethane to give 132 mg of
37 crystallized from diethyl ether and ligroin. White crystalline powder, mp 234-236 C,
FABMSm/z: 611 (M + 1),'H NMR 5: 1.18 (s, 3H, 16-H), 1.24 (s, 3H, 19-H), 1.66 (s,
3H, 17-H), 1.85 (m, 2H, 14-Ha,(3), 1.92 (m, 2H, 6-Ha,(3), 2.04 (s, 3H, OAc), 2.05 (s,
3H, OAc), 2.11 (s, 3H, OAc), 2.13 (s, 3H, OAc), 2.14 (d, 1.2Hz, 3H, 18-H), 2.19 (s, 3H,
OAc), 2.32 (d, 5.1Hz, 1H, 20-H|3), 2.53 (dd, 9.6, 15.0Hz, 1H, 14-Ha), 3.08 (d, 3.6Hz,
1H, 3-H), 3.52 (d, 5.1Hz, 1H, 20-Ha), 4.21 (t, 3.0Hz, 1H, 5-H), 4.27 (dd, 4.8, 10.8, 1H,
7-H), 5.45 (d, 3.6Hz, 1H, 2-H), 6.07 (t, 7.2Hz, 1H, 13-H), 6.14 (d, 11.1Hz, 1H, 9-H),
6.20 (d, 10.8Hz, 1H, 10-H). 13C NMR 6: 12.5, 15.5, 20.5, 20.8, 20.9, 21.3, 21.6, 21.8,
28.5, 32.3, 38.4, 40.4, 43.3, 47.1, 49.9, 59.2, 69.1, 70.4, 71.8, 72.5, 76.1, 78.0, 78.1,
135.7, 140.7, 168.3, 169.1, 169.5, 169.6, 170.0.
9-Dihydro-13-acetyl baccatin IH (38)
The compound was eluted with 5% acetone and 2% methanol in dichloromethane
to give 184 mg of 38 crystallized from diethyl ether and ligroin. White crystalline powder,
mp 243-244 C, FABMS m/z: 631 (M + 1), !H NMR 5: 1.25 (s, 3H, 16-H), 1.68 (s, 3H,
17-H), 1.82 (s, 3H, 19-H), 1.93 (d, 1.2Hz, 3H, 18-H), 1.96 (m, 1H, 6-HJ3), 2.14 (s, 3H,
10-0Ac), 2.19 (s, 3H, 13-OAc), 2.21 (m, 2H, 14-Ha,(3), 2.28 (s, 3H, 4-OAc), 2.53 (m,

47
1H, 6-Ha), 3.05 (d, 6.0Hz, 1H, 3-H), 4.16 (d, 8.1Hz, 1H, 20-Hp), 4.31 (d, 8.1Hz, 1H,
20-Ha), 4.43 (m, 2H, 7-H, 9-H), 4.95 (d, 8.4Hz, 1H, 5-H), 5.75 (d, 5.7Hz, 1H, 2-H),
6.17 (t, 6.9Hz, 1H, 13-H), 6.19 (d, 10.8, 1H, 10-H), 7.48 (t, 7.8Hz, 2H, m-Bz), 7.61 (t,
7.5Hz, 1H, p-Bz), 8.09 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 12.5, 14.8, 21.2, 21.3, 22.6,
22.8, 28.3, 35.5, 38.0, 43.1, 45.0, 47.2, 69.8, 73.3, 73.7, 74.0, 76.9, 77.3, 78.8, 82.2,
84.1, 128.6, 129.4, 130.1, 133.6, 135.0, 139.7, 167.0, 169.3, 170.4.
Acetylation of ip-Hydroxy-7-Deacetyl Baccatin I
A total of 30 mg of ip-hydroxy-7-deacetyl baccatin I (37) was dissolved in 1 ml of
acetic anhydride and 1 ml of pyridine. This mixture was stirred at room temperature for 18
hours and then water was added to the mixture. Sodium bicarbonate was added slowly
until no further evolution of C02 was observed. The aqueous mixture was then extracted
twice with dichloromethane and the combined organic layers were washed with 0.1 N
NaOH, 0.1 N HC1, and water successively and dried with sodium sulfate. The
dichloromethane was evaporated and the product was crystallized from diethyl ether and
ligroin to yield 22 mg of acetylated ip-hydroxy-7-deacetyl baccatin I which was identical
in every way to lp-hydroxy baccatin I (35).
Isolation of Taxamairin A (38) from Taxus brevifolia
A total of 27 g of semi-pure paclitaxel, which had crystallized from reverse-phase
fractions, was dissolved in 200 ml of dichloromethane and applied to a silica column with
225 g of 240 mesh silica gel. Solvent was pumped through with an Eldex Laboratories
metering pump model B-100-S-4 at a pressure not exceeding 25 psi. The beginning
solvent was 2 : 1 dichloromethane : ligroin, then 3 : 1 dichloromethane : ligroin, followed

48
by dichloromethane. At this point the column was eluted with 2-5% acetone in
dichloromethane and then 2-5% methanol and 5% acetone in dichloromethane. A total of
500 ml of each solvent mixture was pumped through before switching to the next solvent.
Fractions of about 100 ml were collected and monitored by TLC. Taxamairin A was
eluted with 2% acetone in dichloromethane and crystallized from the elution solvent. It
was recrystallized from dichloromethane to yield 275 mg. Yellow needles, mp 252-253 C,
EIMS m/z: 338 (80%, M), 310 (74%), 295 (100%), 267 (63%), 237 (18%), 156 (24%).
CIMS: 339 (M + 1). UV 211, 255, 385nm. IR (KBr): 1672, 1535, 1320, 1195,
1052 cm'1. HNMR5: 1.33 (d, 6.9Hz, 6H, 19-H, 20-H), 1.46 (s, 6H, 12-H, 13-H), 3.35
(heptet, 6.9Hz, 1H, 18-H), 3.89 (s, 3H, 15-OMe), 6.11 (d, 9.6Hz, 1H, 7-H), 6.65 (s, 1H,
14-OH), 6.95 (s, 1H, 11-H), 7.31 (d, 9.9Hz, 1H, 11-H), 7.77 (s, 1H, 17-H), 7.94 (s, 1H,
4-H). 13C NMR 5: 23.3, 23.4, 26.7, 26.8, 27.4, 50.5, 62.0, 119.3, 120.8, 123.7, 130.1,
131.1, 133.7, 136.6, 146.1, 146.8, 147.8, 148.2, 151.4, 188.2, 200.9.
Methylation of Taxamairin A
Taxamairin A (50 mg) was dissolved in 3 ml acetone and excess K2CO3 was added
together with 0.25 ml of dimethyl sulfate. This mixture was refluxed for 3 hours and at
that point 0.5 ml of concentrated NH4OH was added to the mixture and stirred for 15
minutes. The acetone was partially evaporated and water was added. This aqueous
solution was then extracted twice with dichloromethane and the combined organic layers
were washed with 0.1 N NaOH and then with water. After drying with sodium sulfate, the
solvent was removed and the residue was crystallized from dichloromethane to yield 32
mg of taxamairin B (39). Yellowish white needles, mp >290 C, UV 'kmax: 219, 281, 355

49
nm. IR(KBr): 1670, 1621, 1333, 1305, 1038 cm'1. *H NMR 5: 1.30 (d, 6.9Hz, 6H, 19-
H, 20-H), 1.46 (s, 6H, 12-H, 13-H), 3.41 (heptet, 6.9Hz, 1H, 18-H), 3.98 (s, 6H, 14-
OMe, 15-OMe), 6.12 (d, 9.6Hz, 1H, 7-H), 6.94 (s, 1H, 11-H), 7.31 (d, 1H, 11-H), 7.87
(s, 1H, 17-H), 7.93 (s, 1H, 4-H). 13C NMR 5: 22.9, 23.0, 26.7, 26.8, 27.8, 50.4, 60.7,
61.2, 123.0, 124.1, 127.3, 128.3, 130.9, 131.5, 133.8, 136.1, 147.8, 150.8, 151.3, 153.9,
188.0, 200.7.
Synthesis of Taxamairin B (39)
2, 2-Dimethyl-l, 3-cyclohexanedione (64)
A total of 10 g of 1, 3-cyclohexanedione (63) and 30.6 g (2.5 eq.) of K2CO3 was
added to 40 ml of acetone to which 31.5 g (2.5 eq.) of CH3I had been added. The mixture
was refluxed overnight. After cooling the mixture the K2CO3 was filtered and the acetone
was removed under vacuum. The residual material was partitioned between water and
diethyl ether and the water layer was discarded. The solvent was evaporated to yield a
syrup which was poured into a mixture of 20 ml of cone. HC1 and 20 g of ice. This was
stirred for 30 minutes to decompose the methyl enol ether which accounts for about 30%
of the mixture, then water and diethyl ether were added and partitioned. The organic layer
was washed twice with water, then dried with sodium sulfate. After removal of the
solvent, the crude liquid product was distilled using a water aspirator vacuum (~15 mm
Hg) and the product distilled at 120-122 C. Upon standing the product crystallized
yielding 6.2 g. Colorless crystals, mp 31-32 C, 'H NMR 6: 1.29 (s, 6H, CH3), 1.93 (m,
6.5Hz, 2H, 5-H), 2.67 (t, 6.9Hz, 4H, 4-H, 6-H). 13C NMR 5: 18.1, 22.3, 37.4, 61.8,
210.6.

50
4-Bromo-2, 2-dimethyl-l, 3-cyclohexanedione (65)
A total of 2.0 g of dimethylated diketone 64 was dissolved in 5 ml of CH2C12 and a
separate bromine mixture was prepared by adding excess bromine to CH2C12 in a ratio of
about 4 drops bromine to 1 ml of CH2C12. The bromine solution was dropwise added with
stirring at room temperature to the diketone solution and the TLC was monitored using I2
crystals as an indicator. The reaction was stopped when only a small amount of starting
material was observed and the major spot on TLC had a slightly higher Rf value in 30%
ethyl acetate in ligroin. Water was added to the reaction mixture and partitioned. The
water layer was discarded and the organic layer was washed twice with water, dried with
sodium sulfate, and the solvent was removed by evaporation. The residue was then put on
a silica column using 20-30% ethyl acetate as the solvent to give 1.14 g of the product as
a yellow oil. This material was stored at -5 C and upon storage the product crystallized.
Colorless crystals, mp 48-50 C, [H NMR 6: 1.34 (s, 3H, CH3), 1.45 (s, 3H, CH3), 2.27-
2.60 (m, 2H, 6-H), 2.72-2.97 (m, 2H, 5-H), 4.73 (dd, 4.2, 6.9Hz, 1H, 4-H). 13C NMR 5:
23.1, 24.1, 26.5, 34.3, 48.8, 59,5, 203.2, 208.2.
2, 2-Dimethyl-4-cyclohexene-l, 3-dione (66)
A total of 1.0 g of bromo compound 65 was dissolved in 5 ml of DMF and 1.0 g
of LiCl was also added. This mixture was refluxed for 2 hours at which time the TLC
showed the presence of a slower moving product and only a small amount of starting
material. Water and diethyl ether were added to the mixture and partitioned. The water
layer was partitioned twice more with diethyl ether and all the organic layers were
combined and washed with water twice, dried with sodium sulfate, and the solvent was

51
evaporationed. The residue was ran through a silica column using 15-30% ethyl acetate in
ligroin as the eluent to give 823 mg of product as a yellow oil. Yellow oil, *H NMR 5:
1.35 (s, 6H, CH3), 3.35 (dd, 2.4, 3.9Hz, 2H, 6-H), 6.25 (dt, 2.1, 10.2Hz, 1H, 4-H), 7.03
(dt, 3.9, 10.2Hz, 1H, 5-H).
Isopropylation of acetovanillone
A total of 3.0 g of acetovanillone (67) was added to 20 ml of 90% H2S04 and 2 ml
of isopropyl alcohol was added. This mixture was stirred at 60 C for 36 hours. At this
point about 50% conversion to the product was observed on TLC. Longer reaction times
did not increase conversion and higher temperatures caused decomposition. Thus the
mixture was diluted with ice water and diethyl ether and partitioned. The water layer was
partitioned once again with ether and the ether layers combined. The organic layer was
then partitioned twice with 1.0 N NaOH and the organic layer was discarded. The
aqueous layer was acidified with concentrated HC1 and extracted twice with diethyl ether.
This organic layer was then washed with water twice, dried with sodium sulfate, and
concentrated. The residue was chromatographed on a silica column using 20-40% ethyl
acetate in ligroin as the eluent to give 1.42 g of product which crystallized on standing.
Clear colorless crystals, mp 116-117 C, EIMS m/z: 208 (42%, M), 193 (100%), 163
(17%), 77 (12%), >H NMR 5: 1.27 (d, 6.9Hz, 6H, CH3), 2.57 (s, 3H, CH3), 3.33 (quintet,
6.9Hz, 1H, CH), 3.95 (s, 3H, OCH3), 6.23 (br s, 1H, Ar-OH), 7.39 (d, 1.5Hz, 1H, Ar-H),
7.50 (d, 1.5Hz, 1H, Ar-H). 13C NMR 5: 22.2, 26.2, 27.2, 56.2, 107.4, 121.0, 129.4,
133.7, 146.2, 147.6, 197.1.

52
Methylation of isopropyl acetovanillone
A total of 1.0 g of isopropylated acetovanillone was dissolved in 20 ml of acetone
then 3.0 g of K2CO3 and 1 ml of dimethyl sulfate were added. The mixture was refluxed
for 3 hours at which time no starting material remained according to TLC. Then 1 ml of
cone. NH4OH was added and the mixture was stirred for 30 minutes. The acetone was
then partially removed and the residue was partitioned between water and diethyl ether.
The organic layer was washed twice with water, dried with sodium sulfate, and
concentrated to yield 956 mg of product (68) as a clear colorless oil.
Reduction of 68
A total of 1.0 g of 68 was dissolved in 8 ml of methanol with 3 drops of 1.0 N
NaOH. To this was added dropwise a solution of NaBH4 in methanol with 1.0 N NaOH
and the TLC was monitored. When no starting material remained the mixture was
acidified with 0.1 N HC1 and the methanol was partially removed. The residue was
partitioned between water and diethyl ether and the organic layer was washed twice with
water, dried with sodium sulfate, and concentrated under vacuum. This material was
chromatographed on a silica column using 25-50% ethyl acetate in ligroin as eluent to give
856 mg of alcoholic product (69) as a clear colorless oil. Clear colorless oil, EIMS m/z:
224 (97%, M), 209 (100%), 179 (18%), 139 (90%), 124 (33%). H NMR 5: 1.21 (d,
6.9Hz, 6H, CH3), 1.48 (d, 6.3Hz, 3H, CH3), 3.34 (heptet, 6.6Hz, 1H, CH), 3.80 (s, 3H,
OCH3), 3.86 (s, 3H, OCH3), 4.83 (q, 6.3Hz, 1H, CH), 6.81 (s, 2H, Ar-H). 13C NMR 5:
23.4, 25.0, 26.8, 55.6, 60.8, 70.5, 106.6, 115.1, 141.6, 142.2, 145.3, 152.5.

53
Hydroxymetfiylation of 69
A total of 200 mg of 69 was dissolved in 3 ml of dry diethyl ether, 205 mg (2 eq.)
of TMEDA was added, and the mixture was cooled to -78 C under a helium atmosphere.
After cooling the mixture, 0.4 ml of 2.5 M n-butyllithium in hexanes (2.2 eq.) was added
with stirring. After stirring for 30-45 minutes excess paraformaldehyde was added and the
mixture was stirred overnight while warming to room temperature. The mixture was
diluted with water and diethyl ether and partitioned. The organic layer was washed twice
with water, dried with sodium sulfate, and concentrated. According to TLC about 50% of
the product remained. This material was separated on a silica column using 30-50% ethyl
acetate in ligroin as eluent and a total of 116 mg of product was isolated as a slightly
yellow oil. Yellow oil, lU NMR 8: 1.22 (d, 6.9Hz, 6H, CH3), 1.56 (d, 6.3Hz, 3H, CH3),
3.32 (quintet, 6.9Hz, 1H, CH), 3.84 (s, 3H, OCH3), 3.88 (s, 3H, OCH3), 4.68 (d, 11.7Hz,
1H, CH20), 4.86 (d, 11.7Hz, 1H, CH20), 5.12 (q, 6.6, 12.9Hz, 1H, CH), 7.10 (s, 1H, Ar-
H). 13C NMR 8: 23.1, 23.4, 27.0, 33.5, 55.9, 60.6, 61.2, 67.2, 118.7, 129.9, 139.5,
142.6, 149.6, 151.5.
Oxidation of diol to keto-aldehyde 70
A total of 600 mg of Cr03 (1 eq.) was added to a solution of 950 mg of pyridine
(2 eq.) in 15 ml of CH2CI2 and this was stirred at room temperature for 15 minutes. At this
point the PDC solution was added dropwise to a solution of 100 mg of the diol in 3 ml of
CH2CI2. The TLC of the reaction mixture was always checked about 5 minutes after
adding 3-4 drops of the PDC solution using 2, 4-dinitrophenylhydrazine as the indicator.
More PDC was added until the reaction was complete according to TLC. At this point the

54
reaction was filtered and the filtrate was washed twice with 0.1 N HC1, twice with 0.1 N
NaOH, and twice with water. The organic layer was then dried with Na2S04 and
evaporated under a vacuum to a solid residue. This material was separated on a regular
silica column using 20-40% ethyl acetate in ligroin. A total of 62 mg of clear colorless oil
was obtained. Clear colorless oil, 'H NMR 5: 1.24 (d, 6.9Hz, 6H, CH3), 2.49 (s, 3H,
CH3), 3.37 (m, 1H, CH), 3.91 (s, 3H, OCH3), 3.95 (s, 3H, OCH3), 7.08 (s, 1H, Ar-H),
10.33 (s, 1H, CHO).
Condensation of keto-aldehyde 70 with dione 66
A total of 100 mg of keto-aldehyde 70 was added to 3 ml of pyridine containing
~10 drops of piperidine. To this mixture was added 80 mg of dione 66 and the mixture
was refluxed with stirring for 6 hours. At this time TLC analysis showed a product with
the same Rf value as the methylated taxamairin A (taxamairin B) along with other
products. The mixture was diluted with diethyl ether and washed three times with 0. IN
HC1 until the water layer was still acidic. The organic layer was then washed twice with
water, dried with Na2S04, and evaporated under a vacuum to a solid residue. The product
was isolated using a regular silica column with 20-40% ethyl acetate in ligroin and a total
of 46 mg of product was isolated as a yellowish white amorphous solid. All spectral data
was identical to that of the methylated taxamairin A.
Complete reduction of methyl isopropyl acetovanillone (72)
A total of 500 mg of methylated isopropyl acetovanillone was dissolved in 5 ml of
THF and 1.0 g of ZnCl2 was added and the mixture was stirred at 60 C. To this was
added NaCNBH3 in small portions and the TLC was monitored. When nearly all the

55
starting material had been converted to a faster moving product the mixture was diluted
with water and diethyl ether and partitioned. The organic layer was washed with 0. IN HC1
and twice with water, dried over sodium sulfate, and concentrated. The product was
purified by silica chromatography using 15-20% ethyl acetate in ligroin as solvent. A total
of 387 mg of product (73) was obtained as a slightly yellow oil. Yellow oil, !H NMR 5:
1.21 (d, 6.9Hz, 6H, CH3), 1.24 (t, 7.5Hz, 3H, CH3), 2.60 (q, 7.8, 15.3Hz, 3H, CH3), 3.33
(quintet, 6.9Hz, 1H, CH), 3.79 (s, 3H, OCH3), 3.85 (s, 3H, OCH3), 6.60 (d, 1.5Hz, 1H,
Ar-H), 6.67 (d, 1.5Hz, 1H, Ar-H). 13C NMR 5. 15.6, 23.6, 26.7, 28.9, 55.6, 60.9, 109.3,
117.3, 140.0, 142.0, 144.1, 152.3.
Acetoxymethylation of 73
A total of 100 mg of 73 was added to 2.0 ml of 85% H3PO4 and to this was added
200 mg of paraformaldehyde and 0.5 ml of acetic anhydride. The mixture was stirred at
room temperature for 3 hours at which point the TLC showed most of the starting
material to be gone and a slower moving product had formed. The mixture was partitioned
between water and diethyl ether and the organic layer was washed twice with 0.1 N HC1
and twice with water, dried with sodium sulfate, and concentrated. The residue was
separated on a silica column using 15-25% ethyl acetate in ligroin as the solvent. A total
of 64 mg of the product (74) was isolated as a clear colorless oil. Clear colorless oil, H
NMR 5: 1.21 (t, 7.5Hz, 3H, CH3), 1.35 (d, 7.2Hz, 6H, CH3), 2.07 (s, 3H, OAc), 2.68 (q,
7.5, 15.0Hz, 2H, CH2), 3.22 (m, 1H, CH), 3.84 (s, 3H, OCH3), 3.86 (s, 3H, OCH3), 5.14
(s, 2H, OCH2), 6.66 (s, 1H, Ar-H). 13C NMR 5: 16.2, 21.0, 21.9, 26.8, 28.9, 55.5, 60.6,
60.7, 110.7, 122.9, 142.1, 146.6, 153.2, 171.2.

56
Brominatic n of 73
A total of 100 mg of 73 was dissolved in 3.0 ml of CH2CI2 and a dilute bromine
solution in CH2C12 was added dropwise with stirring at room temperature and the TLC
was monitored. The addition was stopped when a small amount of starting material
remained and a faster moving product spot was present. Water and more CH2C12 was
added to the mixture and partitioned. The organic layer was washed twice with 0.1 N HC1
and twice with water, dried with sodium sulfate, and concentrated. The residue was
separated on a silica column using 15-25% ethyl acetate in ligroin and 73 mg of product
(75) was isolated as a yellowish oil. Yellow oil, 'H NMR 5: 1.22 (t, 7.5Hz, 3H, CH3),
1.35 (d, 6.6Hz, 6H, CH3), 2.75 (q, 7.5, 15.3Hz, 2H, CH2), 3.72 (m, 1H, CH), 3.83 (s, 3H,
OCH3), 3.84 (s, 3H, OCH3), 6.69 (s, 1H, Ar-H). 13C NMR 5: 14.4, 21.0, 30.9, 34.2,
55.7, 60.9, 111.0, 139.0, 140.5, 152.0.
Brominaticn of 69
A total of 100 mg of 69 was dissolved in 3.0 ml of CH2C12 and a dilute bromine
solution in CH2C12 was added dropwise with stirring at room temperature and the TLC
was monitored. The addition was stopped when only a small amount of starting material
remained and a faster moving product spot was present. Water and more CH2C12 was
added to the mixture and partitioned. The organic layer was washed twice with 0.1 N HC1
and twice with water, dried with sodium sulfate, and concentrated. The residue was
separated on a silica column using 15-25% ethyl acetate in ligroin and 58 mg of product
(76) was isolated as a yellowish oil. Yellow oil, 'H NMR 5: 1.34 (t, 6.6Hz, 6H, CH3),
2.02 (d, 6.9Hz, 3H, CH3), 3.71 (m, 1H, CH), 3.86 (s, 3H, OCH3), 3.89 (s, 3H, OCH3),

57
5.78 (q, 6.6, 13.5Hz, 1H, CH), 7.10 (s, 1H, Ar-H). 13C NMR 6: 20.9, 26.7, 35.4, 50.0,
55.8, 60.9, 109.8, 118.0, 140.6, 142.5, 152.5.
Isolation of Minor Compounds from Taxus floridana
The mother liquors of fractions that contained 10-deacetyl baccatin III from the
reverse-phase column (25-40% acetonitrile in water) were concentrated to a syrup (10 g).
On standing, this syrup became an amorphous solid and 3 g of this was applied to a
normal-phase silica column (40 g) using dichloromethane as the starting solvent. Then the
column was eluted successively with dichloromethane containing 2, 5, and 10% acetone
and then with addition of 2, 5, and 10% methanol. A total of 200 ml of each solvent
mixture was passed through the column before the next solvent mixture was started and
20 ml fractions were collected and monitored by TLC. The initial dichloromethane eluent
was put aside for further chromatography and the order of elution of the more polar
compounds was as follows: ip-hydroxy baccatin I, taxiflorine (taxchinin M),
rhododendrol, taxchinin L, 10-deacetyl baccatin III, ponasterone A, and 10-deacetyl
paclitaxel-7-P-xyloside.
The initial dichloromethane eluent (0.250 mg) was applied to another silica column
(3 g) using 25% ethyl acetate in ligroin as the initial solvent and proceeding to 50% ethyl
acetate in 10% intervals. A total of 50 ml of each solvent was used before progressing to
the next solvent mixture and 5 ml fractions were collected and monitored by TLC. The
order of elution was trans-2,6-dimethoxy cinnamaldehyde, a-conidendrin, and 1-deoxy
baccatin IV.

58
lP-Hydroxy baccatin I (35)
The compound was eluted with 2% acetone in dichloromethane and a total of 189
mg of 35 was crystallized from diethyl ether and ligroin. Its physical and spectral
properties are identical to that reported above.
Taxiflorine (taxchinin M) (77)
The compound was eluted with 2% acetone in dichloromethane and a total of 129
mg was crystallized from diethyl ether and ligroin. No NMR was reported since the
spectrum contained poorly defined peaks.
Rhododendrol (82)
This compound was eluted with 5% acetone in dichloromethane and a total 435
mg was crystallized directly from the eluting solvent. Clear colorless crystals, mp 74-76
C, 'H NMR 5: 1.23 (d, 6.2Hz, 3H, 1-H), 1.73 (m, 2H, 3-H), 2.63 (m, 2H, 4-H), 3.83 (m,
1H, 2-H), 6.74 (m, 2H, m-Bz), 7.04 (m, 2H, o-Bz). 13C NMR 6: 23.5, 31.2, 40.9, 67.7,
115.3, 129.4, 133.9, 153.8.
Taxchinin L (83)
This compound was eluted with 5% acetone in dichloromethane and a total of 162
mg was crystallized from diethyl ether and ligroin. White crystalline powder, mp 264-266
C. lK NMR (-10 C) 5: 1.02 (s, 3H, 16-H), 1.22 (s, 3H, 17-H), 1.43 (m, 1H, 14-Ha),
1.73 (s, 3H, 19-H), 1.79 (s, 3H, 7-OAc), 1.81 (m, 1H, 6-Hp), 1.96 (s, 3H, 18-H), 2.05 (s,
3H, 2-OAc), 2.13 (m, 1H, 14-Ha), 2.17 (s, 3H, 4-OAc), 2.61 (m, 1H, 6-Ha), 3.15 (d,
7.4Hz, 1H, 3-H), 4.45 (m, 2H, 13-H, 20-Hp), 4.53 (d, 7.4Hz, 1H, 20-Ha), 4.69 (t,
10.3Hz, 1H, 10-H), 4.94 (d, 6.2Hz, 1H, 5-H), 5.46 (dd, 5.6, 8.6Hz, 1H, 7-H), 5.95 (d,

59
7.3Hz, 1H, 2-H), 5.97 (d, 10.1Hz, 1H, 9-H), 7.45 (t, 7.5Hz, 2H, m-Bz), 7.57 (t, 6.9Hz,
1H, p-Bz), 7.99 (d, 7.5Hz, 2H, o-Bz). 13C NMR (-10 C) 5: 11.3, 13.9, 21.6, 21.8, 22.4,
25.2, 27.2, 34.3, 39.2, 42.9, 43.4, 66.3, 66.7, 68.4, 69.9, 74.9, 76.2, 77.5, 79.3, 80.7,
85.3, 128.2, 129.7, 129.9, 133.0, 137.5, 146.0, 167.7, 170.7, 170.8, 171.6.
10-Deacetyil baccatin III (25)
This compound was eluted with 10% acetone in dichloromethane and a total of
112 mg was crystallized from diethyl ether and ligroin. Spectral properties were identical
to those reported in the literature (Dennis et al., 1988).
Ponasterone A (84)
This compound was eluted with 10% acetone and 2% methanol in
dichloromethane and a total of 83 mg was crystallized directly from the eluting solvent.
Spectral properties were identical to those reported in the literature (Miller et al., 1982).
10-Deacetyl paclitaxeI-7-J3-xyloside (28)
This compound was eluted with 5% methanol and 10% acetone in
dichloromethane and a total of 79 mg was obtained as an amorphous solid. Spectral
properties were identical to those reported in the literature (Senilh et al., 1984).
Trans-2, 6-cIimethoxy cinnamaldehyde (85)
This compound was eluted with 25% ethyl acetate in ligroin and a total of 27 mg
was obtained as an glassy solid. UV A,max: 313 nm. IR (KBr): 3100, 3000-2940, 2810,
2740-2700, 1660, 1605-1585, 1475, 1260, 1140, 1100-1080, 970, 840, 725 cm'1. EIMS
m/z: 192 (30%, M+), 161 (100%), 149 (17%), 91 (15%). :H NMR 6: 3.90 (s, 6H,
OMe), 6.58 (d, 8.4Hz, 2H, m-Ar), 7.17 (dd, 7.8, 16.0Hz, 1H, 2-H), 7.33 (t, 8.4Hz, 1H, p-

60
At), 7.93 (d, 16.0Hz, 1H, 3-H), 9.64 (d, 8.1Hz, 1H, 1-H). 13C NMR 5: 55.8 x 2 (OMe),
103.6 (m-Ar), 112.1 (q-Ar), 131.6 (2-C), 132.6 (p-Ar), 144.5 (3-C), 160.5 (o-Ar), 196.4
(1-C).
a-Conidendrin (86)
This compound was eluted with 50% ethyl acetate in ligroin and a total of 38 mg
was crystallized from diethyl ether and ligroin. White crystalline powder, mp 257-259 C,
EIMS m/z: 356 (100%, M+), 255 (13%), 241 (26%), 137 (14%). H NMR 5: 2.5 (m,
2H), 2.7-3.1 (m, 1H), 3.73 (s, 3H), 3.78 (s, 3H), 3.9-4.2 (m, 4H), 6.26 (s, 1H), 6.53 (d,
2.0Hz, 1H), 6.58 (d, 8.0Hz, 1H), 6.62 (s, 1H), 6.73 (d, 8.0Hz, 1H). 13C DEPT NMR 5:
29.3 (CH2), 41.9 (CH), 47.5 (CH), 49.9 (CH), 55.9 (CH3), 60.0 (CH3), 71.9 (CH2), 110.1
(CH), 111.3 (CH), 114.6 (CH), 115.1 (CH), 121.5 (CH), 126.3(C), 131.7(C), 134.0(C),
144.2 (C), 144.9 (C), 145.5 (C), 147.0 (C), 177.0 (C).
1-Deoxy baccatin IV (87)
This compound was eluted with 50% ethyl acetate in ligroin and a total of 57 mg
was crystallized from diethyl ether and ligroin. White crystalline powder, mp 262-264 C,
'HNMR8: 1.12 (s, 3H, 17-H), 1.54 (s, 3H, 19-H), 1.78 (m, 1H, 1-H), 1.79 (s, 3H, 16-
H), 1.88 (m, 1H, 14-HJ3), 1.91 (m, 1H, 6-Hp), 1.97 (s, 3H, 18-H), 2.02 (s, 3H, Oac), 2.05
(s, 3H, Oac), 2.08 (s, 3H, Oac), 2.10 (s, 3H, Oac), 2.16 (s, 3H, Oac), 2.17 (s, 3H, Oac),
2.41 (m, 1H, 14-Ha), 2.50 (s, 1H, 6-Ha), 2.87 (d, 1H, 3-H), 4.19 (d, 7.5Hz, 20-HP),
4.51 (d, 8.1Hz, 1H, 20-Ha), 4.99 (d, 8.7Hz, 1H, 5-H), 5.52 (m, 2H, 2-H, 7-H), 5.91 (m,
2H, 9-H, 13-H), 6.15 (d, 11.4Hz, 1H, 10-H). 13C NMR 5: 12.7, 14.9, 20.7, 20.8, 20.9,

61
21.2, 21.4, 21.6, 22.6, 26.9, 31.3, 34.6, 37.9, 44.5, 45.6, 46.9, 69.0, 71.0, 71.1, 71.9,
75.4, 77.2, 81.0, 83.9, 133.3, 138.9, 169.2, 169.3, 169.7, 170.1, 170.2, 170.4.
Acetylation of Taxiflorine
Taxiflorine (77) 120 mg was dissolved in 2 ml of acetic anhydride and 1 ml of
pyridine was added. The mixture was stirred at room temperature for 18 hours and then
water was added to the mixture. Sodium bicarbonate was added slowly until no further
evolution of C02 was observed. The aqueous mixture was then extracted twice with
dichloromethane and the combined organic layers were washed with 0.1 N NaOH, 0.1 N
HC1, and water successively and dried with sodium sulfate. The dichloromethane was
evaporated and the product (112 mg) (81) was obtained as a glassy solid. For NMR data
see Table 2-1.
Oxidation of Taxiflorine
Taxiflorine (77) 50 mg was dissolved in 2 ml of acetone and a few drops of Jones
reagent was added and the mixture was stirred at room temperature for 2 hours. At this
time the acetone was partially evaporated and water was added. This aqueous mixture was
extracted twice with dichloromethane and the combined organic layers were washed with
0.1 N NaOH and then with water and then dried with sodium sulfate. After the
dichloromethane was evaporated the residue was crystallized from diethyl ether and ligroin
to yield 36 mg of the ketone product. White crystalline powder. For NMR data see Table
2-1.

62
Acetylation of Taxchinin L (83)
Taxchinin L 100 mg was dissolved in 2 ml of acetic anhydride and 1 ml of pyridine
was added. The mixture was stirred at room temperature for 18 hours and then water was
added to the mixture. Sodium bicarbonate was added slowly until no further evolution of
C02 was observed. The aqueous mixture was then extracted twice with dichloromethane
and the combined organic layers were washed with 0.1 N NaOH, 0.1 N HC1, and water
successively and dried with sodium sulfate. The dichloromethane was evaporated and the
product (84 mg) was obtained as a glassy solid. The physical and spectral properties of
this acetylated product (81) was identical with the acetate of taxiflorine in every way.
Synthesis of Trans-2, 6-Dimethoxycinnamaldehyde (85)
Methyl 2, 6-dimethoxybenzoate
A total of 5.0 g of 2, 6-dimethoxybenzoic acid (88) was dissolved in 30 ml of
methanol and 1.0 ml of concentrated H2SO4 was added. This mixture was refluxed for 48
hours at which time most of the methanol was removed under reduced pressure. The
residue was partitioned between water and diethyl ether and the organic layer was
partitioned with 0.1 N NaOH to remove the remaining starting material, this was
performed three times. Finally the organic layer was washed with water twice and dried
with sodium sulfate. Upon removal of the solvent under reduced pressure the methyl ester
product crystallized to yield 3.1 g of product.
2, 6-Dimethoxybenzyl alcohol (89)
A total of 3.1 g of the methyl ester was dissolved in 15 ml of THF and this was
cooled to 0 C. A total of 581 mg (1 eq.) of LAH was then carefully added to the mixture

63
with stirring. This mixture was then refluxed for 4 hours at which time no starting material
remained. Thus about 10 ml of acetone was added to neutralize the remaining LAH and
this was stirred for 18 hours. The solvent was then removed by rotovap and the residue
was partitioned between water and diethyl ether. The organic layer was washed three
times with water and was dried with sodium sulfate. Upon removal of the solvent by
evaporation the product crystallized to yield 1.5 g of the product alcohol. Clear colorless
crystals, *H NMR 5: 3.83 (s, 6H, OCH3), 4.70 (s, 2H, CH2), 6.46 (d, 8.4Hz, 2H, m-Ar),
7.13 (t, 8.7Hz, 1H, p-Ar).
2, 6-Dimetlioxybenzaldehyde (90)
A total of 1.5 g of the alcohol (89) was dissolved in 10 ml of acetone and Jones
reagent was added dropwise while the TLC was monitored using 2, 4-
dinitrophenylhydrazine as the indicator. Quite suprisingly the aldehyde had a lower Rf
value than the alcohol. The reaction was continued until no alcohol starting material was
observed on TLC. The acetone was partially removed and the residue was partitioned
between water and diethyl ether. The organic layer was washed with water twice and dried
with sodium sulfate. Upon removal of the solvent under reduced pressure the product
crystallized to yield 1.18 g of the aldehyde product. Clear colorless crystals, H NMR 5:
3.85 (s, 6H, OCH3), 6.42 (d, 8.4Hz, 2H, m-Ar), 7.33 (t, 8.6Hz, 1H, p-Ar), 10.35 (s, 1H,
CHO).
Trans-2, 6-dimethoxycinnamic acid (91)
A total of 1.18 g of the aldehyde (90) and 1.38 g (2 eq.) of malonic acid was
dissolved in 4 ml of pyridine and a few drops of piperidine were added. This mixture was

64
refluxed overnight. The following day the TLC seemed unchanged but 2, 4-dinitrophenyl
hydrazine did not give a positive test indicating that no aldehyde remained and the
cinnamic acid product had an identical Rf value. The mixture was partitioned between 0.1
N HC1 and diethyl ether and the organic layer was washed twice more with 0.1 N HC1 and
then three times with water and dried with sodium sulfate. Upon evaporation of the
solvent by evaporation 1.25 g of the cinnamic acid product crystallized.
Methyl trams-2, 6-dimethoxycinnamate
A total of 1.25 g of the cinnamic acid product (91) was dissolved in 15 ml of
acetone and 3.0 g of K2CO3 was added with 0.5 ml of dimethyl sulfate. The mixture was
refluxed for 3 hours at which time no starting material was observed by TLC and thus 2.0
ml of ammonium hydroxide was added to decompose any remaining dimethyl sulfate.
After 15 minutes of stirring, the solvent was partially removed under vacuum and the
residue was partitioned between water and diethyl ether. The organic layer was washed
three times with water and then dried with sodium sulfate. Upon removal of the solvent
1.23 g of the methyl ester product crystallized.
Trans-2, 6-dirmethoxycinnamyl alcohol (92)
A total of 1.23 g of the methyl ester was dissolved in 15 ml of THF and this was
cooled to 0 C. A total of 223 mg (1 eq.) of LAH was then carefully added to the mixture
with stirring This mixture was then refluxed for 4 hours at which time no starting material
remained so about 10 ml of acetone was added to neutralize the remaining LAH and this
was stirred for 18 hours. The solvent was then removed by evaporation and the residue
was partitioned between water and diethyl ether. The organic layer was washed three

65
times with water and was dried with sodium sulfate. Since minor impurities were also
present in the mixture after workup a silica column was ran using 15-30% ethyl acetate in
ligroin as the solvent. A total of 700 mg of the product alcohol was obtained as a clear
yellowish liquid. !H NMR S: 3.84 (s, 6H, OCH3), 4.32 (d, 5.1Hz, 2H, 1-H), 6.56 (d,
8.4Hz, 2H, m-Ar), 6.75-6.92 (m, 2H, 2-H, 3-H), 7.15 (t, 8.1Hz, 1H, p-Ar).13C NMR 5:
55.6, 65.4, 103.8, 113.9, 121.7, 128.2, 132.6, 158.4.
Trans-2, 6-dimethoxycinnamaldehyde (93)
A total of 4.0 g of Cr03 was dissolved in 6.5 ml of pyridine at 0 C and stirred until
a reddish orange solid formed (PDC). At that point 561 mg of 92 was added in 1.0 ml of
acetone and the reaction was stirred for 4 hours at room temperature. At this point the
TLC showed only a small amount of starting material remaining so the reaction was
stopped. All physical and chemical properties matched that of the natural product.

CHAPTER 3
PREPARATION OF NITRATE ESTERS OF PACLITAXEL AND RELATED
TAXANES
Complete Nitration of Paclitaxel and Related Taxanes
In an attempt to nitrate the phenyl ring of the N-benzoyl phenylisoserine side chain
in paclitaxel (Figure 3-1, 94) in order to determine its effect on potency, the compound
was subjected to reaction with a 1 : 5 mixture of concentrated nitric acid in acetic
anhydride and an equal volume of dichloromethane. The reaction proceeded readily at
room temperature in 30 minutes to give a single product which exhibited a much higher Rf
on TLC than paclitaxel and even higher than 2, 7-diacetyl paclitaxel (Mellado et al.,
1984), thus eliminating the possibility that acid-catalyzed acetylation had occurred. The
NMR spectrum showed a considerable downfield shift of the 2 and 7 proton signals (H-2
4.78 ppm > 5.69 ppm, H-7 4.40 ppm > 5.75 ppm). Following characterization by
elemental analysis and IR spectroscopy, which exhibited characteristic bands for nitrate
esters at 1650, 1270, and 835 cm'1, the structure was determined to be that of paclitaxel-
2, 7-dinitrate ester (Figure 3-1, 95). After a review of the literature it was found that the
conditions used are actually standard conditions for nitrate ester synthesis (Boschan et al.,
1955). It was surprising that the product showed no evidence of rearrangement of the A-
66

67
94
O
5:1/ acetic anhydride : HNO 3
CH2C12
room temp., 30 min
Figur 3=1} Nitration of Paelitaxei, 7-OH > 2* 011
ring or cleavage of the oxetane ring, both of which have been known to occur in the
presence of strong acids (Chen et al., 1993).

68
The reaction was also repeated with three of the natural analogues of paclitaxel,
ie., 10-deacetyl baccatin III, 10-deacetyl paclitaxel, and 10-deacetyl paclitaxel-7-(3-
xyloside. In each case the reaction proceeded to yield the corresponding tri- (Figure 3-2,
97), tri-(Figure 3-3, 101), and penta-nitrate esters (Figure 3-4, 104), respectively. All of
these compounds crystallized readily from ethyl ether after workup without any need for
chromatography to give nearly quantitative yields. In no case was nitration at the sterically
hindered 1-hydroxyl observed. NMR chemical shift values are given in Table 3-1.
Table 3-1: FI and ljC NMR Values for Completely Nitrated Taxanes
Hor C#
Compd. 95
Compd. 97
Compd. 101
Compd. 104
1
4444 yg g
**** 78 2
4444 yg 5
4444 yg y
2
5.72 d (6.9Hz),
74.3
5.64 d (6.6Hz),
73.5
5.71 d(6.9FIz),
74.0
5.71 d (6.9Hz),
74.5
3
4.02 d (6.9Hz),
47.2
3.92 d (6.9Hz),
47.7
3.95 d(6.9Hz),
47.2
3.81 d (6.6Hz),
45.7
4
4444 gQ y
4444 gQ y
4444 gQ g
4444 gQ g
5
4.99 d (9.0Hz),
83.6
4.97 d (8.7Hz),
83.5
4.98 d (9.0Hz),
83.4
4.88 d (9.0Hz),
83.5
6a
2.69 m, 32.5
2.72 m, 32.5
2.71 m, 32.5
2.81 m, 35.8
6P
2.04 m, ****
2.04 m, ****
2.07 m, ****
2.04 m, ****
7
5.75 dd (7.2,
10.5 Hz), 79.8
5.78 dd (7.2,
10.5 Hz), 79.9
5.75 dd (7.2,
10.5 Hz), 79.8
4.17 m, 80.1
8
4444 gg g
****, 55.8
**** 55 6
4444 5y 8
9
****, 200.4
4444 299 3
4444 299 5
4444 299 9
10
6.31 s, 74.5
6.38 s, 81.4
6.36 s, 81.6
6.42 s, 82.1
11
****, 135.5
4444 \242
****, 135.4
****, 135.5
12
**** |409
Hi 4= * \a,2 2
4444 244 2
4444 243 2
13
6.30 t (9.9 Hz),
72.8
6.23 t (8.1 Hz),
78.1
6.29 t (9.0 Hz),
72.5
6.25 t (8.7 Hz),
72.6
14a
2.38 m, 35.4
2.48 dd (9.9,
15.6 Hz), 34.2
2.39 m, 35.3
2.42 m, 35.3
140
2.38 m, ****
2.36 dd (2.8,
15.9 Hz), ****
2.39 m, ****
2.30 m, ****
15
**** 43 2
**** qy 9
444 4 q.3 2
4444, 42 9
16
1.23 s, 26.5
1.20 s, 26.6
1.20 s, 26.4
1.19 s, 26.2
17
1.17 s, 21.5
1.15 s, 20.3
1.14 s, 21.5
1.13 s, 21.6

69
Table 3-1continued
HorC#
Compd. 95
Compd. 97
Compd. 101
Compd. 104
18
1.94 s, 14.3
2.12 s, 15.0
2.00 s, 14.8
1.97 s, 14.8
19
1.82 s, 11.0
1.82 s, 10.8
1.84 s, 11.0
1.75 s, 10.7
20a
4.35 d (8.7 Hz),
76.2
4.36 d (8.7 Hz),
76.0
4.37 d (8.4 Hz),
76.2
4.35 d (8.7 Hz),
76.5
20p
4.20 d (8.4 Hz),
****
4.11 d (8.4 Hz),
****
4.20 d (8.4 Hz),
****
4.19 d (8.7 Hz),
Hi * *
V
5.69 d (3.0 Hz),
80.3
****
5.68 d (2.7 Hz),
80.2
5.65 d (3.0 Hz),
80.3
3
6.14 dd (2.7, 9.6
Hz), 52.1
****
6.13 dd (2.7, 9.3
Hz), 52.1
6.10 dd (3.0, 9.3
Hz), 52.2
NH
6.97 d (9.6 Hz),
* * *
****
6.90 d (9.3 Hz),
****
6.82 d (9.3 Hz),
****
1
* Hi %
****
****
4.76 d (4.5 Hz),
98.9
2
Hi Hi Hi*
****
* * *
4.98 dd (4.2, 6.3
Hz), 74.3
3
Hi***
****
****
5.26 t (6.3 Hz),
72.9
4
H< Hi *
****
****
5.08 dd (5.7, 9.6
Hz), 74.1
5 ax
****
****
****
4.23 m, 59.5
5 eq
****
****
****
3.68 dd (5.7,
12.9 Hz), ****
4-Ac
2.50 s, 22.6
2.43 s, 22.3
2.51 s, 22.6
2.48 s, 22.7
10-Ac
2.18 s, 20.6
****
****
OBz-1
**** 220 3
****, 128.6
**** |20 3
**** 2290
OBz-2,6
8.12 d (7.2 Hz),
130.2
8.06 d (6.9 Hz),
130.1
8.13 d (7.5 Hz),
130.2
8.13 d (7.2 Hz),
130.2
OBz-3,5
7.52 t (7.8 Hz),
128.8
7.51 t (7.8 Hz),
128.9
7.52, 128.8
7.40-7.55, 128.8
OBz-4
7.63 t, 133.8
7.65 t, (7.5 Hz),
131.9
7.64 t (7.2 Hz),
133.9
7.63 t (6.6 Hz),
133.8
NBz-1
****, 131.2
****
**** 12] 3
He*** 2
NBz-2,6
7.73 d (7.2 Hz),
127.1
****
7.73 d (7.5 Hz),
127.1
7.73 d (6.9 Hz),
127.1
NBz-3,5
7.41-7.46, 129.4
****
7.41-7.46, 129.4
7.40-7.55, 129.4
NBz-4
7.41-7.46, 129.0
Hi***
7.41-7.46, 129.1
7.40-7.55, 129.1
Ph-1
****, 133.1
****
****, 133.0
**** 233 1
Ph-2,6
7.41-7.46, 126.5
****
7.41-7.46, 126.5
7.40-7.55, 126.5
Ph-3,5
7.41-7.46, 128.8
****
7.41-7.46, 128.8
7.40-7.55, 128.8
Ph-4
7.49-7.54, 132.3
****
7.52, 132.4
7.40-7.55, 132.3

70
Table 3-3continued
H or C# Coimpd. 95
Compd. 97
Compd. 101
Compd. 104
C=0 170.2, 169.5,
171.1, 166.9
170.4, 167.6,
170.3, 167.3,
167.7, 166.8,
166.7, 166.7
166.8, 166.7
166.7
Regioselective Nitrations of Paclitaxel and Related Taxanes
Based on these results it was decided to run the reaction at 0 C for only 10-15
minutes to determine if the reaction was regioselective. Using this protocol, paclitaxel
gave the 7-mononitrate of paclitaxel (96) in -90% yield after crystallization from the crude
reaction mixture (Figure 3-1). The position of the nitrate ester was easily determined by
'H NMR and COSY spectroscopy as the nitrate ester causes a downfteld shift of 1.3-1.7
ppm on the adjacent proton. This result was interesting since the 2-hydroxyl is much
more reactive than the 7-hydroxyl in acetylation reactions (Mellado et al., 1984). With
nitration however the order of reactivity was 7-OH > 2-OH.
Similar studies were applied to 10-deacetyl baccatin III, 10-deacetyl paclitaxel, and
10-deacetyl paclitaxel-7-p-xyloside. They also displayed some regioselectivity, however
because these compounds contain more than 2 reactive hydroxyl groups, silica column
chromatography was needed to separate the products. When 10-deacetyl baccatin III was
subjected to this protocol three partially nitrated compounds were obtained namely the 10-
mononitrate (98), the 10, 13-dinitrate (99), and the 7, 10-dinitrate (100), although 98 and
99 were the major compounds. From this it can be concluded that the order of reactivity is
10-OH > 13-OH > 7-OH. This also differs from the acetylation reactivities in which the 7-
hydroxyl is the most reactive followed by the 10-hydroxyl and the 13-hydroxyl

71
respectively (Gueritte-Voegelein et al., 1986). The reaction of 10-deacetyl paclitaxel
likewise gave the 10-mononitrate (102) and the 7, 10-dinitrate (103), again showing the
10-hydroxyl to be the most reactive and the 2-hydroxyl the least (10-OH > 7-OH > 2-
OH). With acetylation the 2-hydroxyl is the most reactive followed by the 7- and 10-
hydroxyls respectively (Kingston et al., 1982). Finally, 10-deacetyl paclitaxel-7-(3-xyloside
was tested and because of the five available hydroxyls many products were observed on
TLC and only the major products were isolated. These included the 2-mononitrate (105),
the 3-mononitrate (106), the 4-mononitrate (107), and the 2, 3, 4, 10-tetranitrate
(108) (Figure 3-2). This result indicates that the sugar hydroxyls are about equally reactive
and more reactive than the 10-hydroxyl, with the 2-hydroxyl being again the least reactive
(2, 3, 4 > 10 > 2). Although no acetylation studies have been performed on the
xylosides, this lab has shown that the 2-hydroxyl is more reactive in standard acetylation
conditions than either the sugar hydroxyls and the 10-hydroxyl is the least reactive (Figure
3-3).
At this point it should also be mentioned that the 10-deacetyl paclitaxel (112) used
in this work was not isolated directly from biomass but was actually converted from 10-
deacetyl paclitaxel-7-(3-xyloside (110). This conversion involves oxidizing the xyloside to
the dialdehyde (111) and then cleaving the dialdehyde with phenylhydrazine to give the
desired product and the corresponding phenylhydrazones (Figure 3-4) (Rao, 1997). In
conclusion, it has been shown that nitrate esters of taxanes can be formed under mild
conditions and in many cases regioselectivity is shown. In view of this work it is

72
O
104 Rx = N02, R2 = N02, R3 = N02, R4 = N02, R5 = N02
105 Rx = N02, R2 = H, R3 = H, R4 = H, R5 = H
106 Rx = H, R2 = N02, R3 = H, R4 = H, R5 = II
107 Rx = H, R2 = H, R3 = N02, R4 = H, R5 = H
108 Rj = N02, R2 = N02, R3 = N02, R4 = N02, R5 = H
Figure 3-2: Nitration of 10-Deacetyl Paclitaxel-7- p-Xyloside
l"-OH, 2"-OH, 3"-OH > 10-OH > 2'-OH
10-Deacetyl Paclitaxel-7- p-Xyloside
acetic anhydride/pyridine
30 sec., room temp.
T
Figure 3-3: Regioselective Acetylation of
10-Deacetyl Paclitaxel-7-p-Xyloside
conceivable that nitrate esters can be used as selective hydroxyl protecting groups when
reactivity differing from acetylation reactivity is desired.

73
O
HC=NNHQH5
HC=NNHC6H5
113
HOCH2CH=NNHC6H5
114
Figure 3-4: Conversion of 10-Deacetyl Paclitaxel
Xyloside to 10-Deacetyl Paclitaxel

74
Reactions of Taxane Nitrate Esters
Complete Reductive Hydrolysis of Nitrate Esters with Zn and Acetic Acid
Since it had been shown that nitrate esters may serve as regioselective hydroxyl
protecting groups, the next step was to determine under what conditions the nitrate esters
may be removed without affecting the remainder of the molecule. Reductive nitration is
known to occur under a variety of conditions and reagents including high-pressure
catalytic hydrogenation, lithium aluminum hydride, hydrazine, Grignard reagents, metallic
sodium, and hydrogen sulfide or ammonium sulfide (Green & Wuts, 1991). However
many of these methods would undoubtedly react with the taxane structure also, therefore
the method chosen was zinc in acetic acid. This method was not only very simple but also
caused no rearrangements, hydrolysis, or any other side reactions. It involved dissolving
the nitrate ester in acetic acid and adding zinc dust with stirring at room temperature for
30 minutes to give a quantitative yield of the parent alcohol (Figure 3-5).
At this point it was decided to run a series of reactions using a variety of reagents
in order to determine what effect the conditions may have on the nitrate esters and/or
taxanes. In all cases either paclitaxel-7, 2-dinitrate or 10-deacetyl paclitaxel-7-P-xyloside-
2, 3, 4, 10, 2-pentanitrate was used in these reactions. A discussion of these
reactions follows.
Reaction with NaBH4
Paclitaxel-dinitrate was dissolved in methanol and an excess of NaBH4 was added
and stirred at room temperature. After 10 minutes TLC confirmed that the reaction was
complete and two major products had formed. These products were determined by

75
O
Figure 3-5: Reductive Denitration of Paditaxel
Paclitaxel-7, 2'-Dinitrate
NaBH4 / CH3OH
room temp., 10 min
Figure 3-6: Hydrolysis of the Side-Chain of Pacitaxel-7, 2'-
Dinitrate with NaBH4

76
NMR spectroscopy to be baccatin III-7-nitrate (117) and the side chain alcohol nitrate
ester (116). Apparently the NaBH4 serves only to quickly reduce the side chain ester
(Figure 3-6)
Reaction with Ammonium Sulfide
Paclitaxel-dinitrate (118) was dissolved in acetonitrile and ammonium sulfide was
added. After 2 minutes TLC showed a major slower moving product had formed. After
isolation this product was determined to be paclitaxel-7-nitrate (119, Figure 3-7). This
result indicates that the nitrate esters may be selectivity removed at least under some
conditions however this line of research was not studied further due to time constraints.
Acetylation of Taxane Nitrate Esters
In an effort to acetylate the 1-hydroxyl of 10-deacetyl paclitaxel-7-xyloside-
pentanitrate (120), this compound was dissolved in acetic anhydride and a small amount of
DMAP was added. This was stirred at room temperature overnight. After workup the
TLC of the reaction mixture displayed two major products (121, 122) and essentially no
starting material (Figure 3-8). After separation and isolation it was concluded by NMR
spectroscopy that the faster moving product contained two additional acetates while the
other product contained one additional acetate. FAB mass spectroscopy was used to
determine that the molecular weight of the first compound was 1205 and that of the other
was 1163, a difference of 42 or one acetate. The proton spectrum of 122 did not display
any signal for H-2, H-3 or N-H. While the proton spectrum of 121 did not display any
signal for H-2 or H-3, it did contain a far downfield signal at 11.41 ppm which could
indicate a very acidic amide. In view of this information the structure of 121 and 122 were

77
O
Figure 3-7: Regioselective Denitration of PacIitaxeI-7, 2-Dinitrate
with Ammonium Sulfide
assigned as shown (Figure 3-8). Apparently, an enol acetate initially forms between the 2-
and 3- positions to give compound 121. This can then form another enol-like acetate
because of the increased acidity of the amide nitrogen to yield compound 122. At this
point however it was not understood how the initial enol acetate was formed.
Reaction with NaN3
In an attempt to displace a nitrate group, paclitaxel-dinitrate (123) was dissolved in
acetonitrile and NaN3 was added with stirring at room temperature. After two hours most
of the starting material was no longer present and a product with similiar Rf (124) was

78
Acetic A-nh!'dri<1eem,pDOT^rnight
CH2Ch>room te p
+
Figure 3-8:
Enol Acetate Formation
of Nitrate Esters

79
present on TLC (Figure 3-9). The reaction continued overnight and on the following day it
was found that the initial product was no longer present and two other products (125,
126) were now formed, one faster (125) than the intermediate product which exhibited
strong UV absorbance but did not char with H2S04 and a slower product (126) which
exhibited UV absorbance as well as charring with acid on TLC. These two products were
isolated by column chromatography and determined by NMR spectroscopy to be baccatin
III-7-nitrate (slower product) (126) and dibenzamide (faster product) (125). Dibenzamide
was also synthesized from benzamide and benzoyl chloride in the presence of NaH to
insure that the reaction product was indeed dibenzamide. It was later found that in DMF
as solvent the reaction proceeded much faster and if it were stopped after only 10 minutes
the intermediate was the major constituent of the reaction mixture. This product was
isolated and determined to have a molecular weight of 896. The 'H NMR of this
compound was very unusual in that many of the signals seemed to be in duplicate. Signals
for the N-H and H-3 were present however there was no signal for the H-2. The
duplication of peaks was similar to what one may find in a racemic mixture presenting the
possibility that the stereochemistry at one of the asymmetric carbons had been scrambled.
After reviewing the literature concerning nitrate esters it was found that it is quite normal
for a nitrate ester to undergo alpha-elimination in the presence of strong base to yield a
carbonyl (Boschan et al., 1955). In this case the C-2 is already acidic due to its proximity
to the C-L ester carbonyl, therefore it was quite conceivable that even a weak base such
as NaNs may cause alpha-elimination. With this information in hand the structure of 124
was determined to be as shown with the stereochemistry at C-3 racemic. It was also

80
O
Glycolic Acid ? 126
Figure 3-9: Reaction of 2'-Nitrate Ester with NaN 3
decided to confirm this structure by producing this compound by a more typical route.
Thus paclitaxel-7-nitrate (127) was oxidized with Jones reagent to yield a product (128)

81
possessing the same Rf value as the intermediate keto-ester (Figure 3-10). However this
oxidation product only showed one set of signals on the *H NMR spectrum and this set of
peaks matched one of the sets of peaks in the intermediate keto-ester *H NMR spectra.
Presumably C-3 does not racemize in the acidic conditions of the Jones oxidation.
The formation of this keto-ester under basic conditions explains how the enol
acetate and dienol acetate can form in the presence of acetic anhydride and pyridine
(Figure 3-8). Once the keto-ester forms the H-3 can then be abstracted by the base to
form the enolate which then undergoes O-acetylation and once the first enol acetate is
formed the formation of the second proceeds as mentioned before. Indeed it was shown
that if the paclitaxel keto-ester (129) was treated with pyridine in acetic anhydride the
monoenol acetate (130) was the major product (Figure 3-11). This compound was
analogues to 121 (Figure 3-8) without the xylose. If this compound was reacted further
under these same conditions a slightly faster moving product was formed that was
probably the dienol acetate however it was not isolated due to time constraints. Also a
second product was also obtained from the keto-ester acetylation that appears to be an
enol acetate-enol. This conclusion was arrived at because unlike enol acetate 130 this
compound does not show a down field N-H signal yet it has the same mass and the same
number of acetates as 130. On TLC however this compound has a much lower Rf than
130, thus it is concluded that this compound may be either 131 or 132 (Figure 3-11).
One aspect of this that was not clear however was how the keto-ester breaks down
under basic conditions to give dibenzamide and 10-deacetyl baccatin III-7-nitrate ester and
what happens to the C-l and C-2 carbons. It should be stated that this reaction proceeds

82
O
Figure 3-10: Oxidation of Paclitaxel-7-Nitrate Ester
in DMF or acetonitrile with all bases tried including NaN3, NaOAc, NaOBz, triethylamine,
and hydroxide, with hydroxide giving the fastest reaction. However this reaction did not
take place when using dichloromethane as solvent with NaN3. No other intermediate
products were observed on TLC as loss of the intermediate keto-ester seemed to coincide
with formation of the final products. In order to determine that this reaction was base
catalyzed the intermediate keto-ester was subjected to three conditions; acetonitrile with
dilute hydroxide added, neet acetonitrile, and acetonitrile with dilute HC1 added. This
study showed that after 24 hours the basic solution was 85-90% decomposed to the final

83
O
OH
Figure 3-11: Acetylation ofKeto-Ester

84
products, the neutral solution was about 40% decomposed, and the acid solution still
contained almost all keto-ester.
Concerning the C-l and C-2 fragment, it was assumed that C-l exist as a
carboxyl in its final form however it is unclear concerning the C-2. Thus it can be
concluded that this two carbon fragment may exist as acetic acid, glycolic acid, or
glyoxalic acid in its final form. A failed attempt was made to derivatize the acid function
by treating the reaction mixture with DCC and aniline to produce a UV active amide that
could be isolated and characterized.
Unfortunately because of time constraints a mechanism for this rearrangement
could not be conclusively established however a possible mechanism has been formulated
and is presented below and in Figure 3-12. It has already been established that the keto-
ester (133) can enolize in the presence of base to form the enolate and thus the enol (134).
The amide proton in this enol is subsequently made quite acidic and can also be abstracted
by base and after electron migration the imine alcohol can form (135). This conjugated
imine is thus a reactive Michael-type adduct which can be attacked by hydroxide with the
glycolic enolate serving as the leaving group. The imine can then be rearranged to form
dibenzamide (138) while the C-l ester is hydrolyzed giving glycolic acid (140) and
baccatin III-7-nitrate ester (141). This hydrolysis has been shown to occur last because the
ester enolate would presumably serve as a better leaving group than the acid enolate,
however hydrolysis of the side chain may occur first. It would be interesting to test the
hypothesis by subjecting the keto-acid to these conditions to see if the rearrangement still
takes place. One could also alkylate the amide to a tertiary amide and determine if the

85
Figure 3-12: Mechanism of Keto-Ester Degradation

86
rearrangement takes place without this available amide proton. In any event time did not
allow this to be studied further.
Experimenta!
All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of
the reaction. All NMR spectra were recorded using either a Varan VXR-300 or a Varan
Gemini-300 spectrophotometer using CDC13 as solvent. Infrared spectra were obtained
using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolent spectra were
obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were
recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a
Fisher melting point apparatus. Column chromatography was used in conjunction with
100-200 mesh silica gel.
Complete Nitrations of Taxanes
Paclitaxel-7, 2~dinitrate ester (95)
Paclitaxel 500 mg was dissolved in 6 ml of CH2CI2 and a mixture of 5 ml of acetic
anhydride and 1 ml of concentrated nitric acid was added slowly. The mixture was
prepared by slowly adding the nitric acid to ice cold acetic anhydride so that the mixture
does not get too hot. The reaction mixture was allowed to stir at room temperature for 30
minutes. At this point 20 ml of water was added and while stirring NaHC03 was slowly
added until no further frothing was observed. Additional CH2C12 was added and the water
layer was extracted 3 times with CH2C12. The organic layer was dried with Na2S04 and the
solvent was evaporated. The product was crystallized with diethyl ether and ligroin to

87
yield 492 mg. White crystalline powder, mp 166-168 C, IR (KBr) 3450, 1725, 1650,
1365, 1270, 1230, 1065, 835, 700 cm'', Anal. Calc, for C47H5iN3018: C 58.69; H 5.31; N
4.37. Fd. C 58.83, H 5.15, N 4.11. 'H and 13C NMR see Table 1.
10-Deacetyl baccatin ni-7, 10, 13-trinitrate ester (97)
This compound was prepared starting with 500 mg of 10-deacetyl baccatin III and
following a procedure identical to that of paclitaxel-7, 2~dinitrate. In this case however,
the starting material does not initially dissolve in dichloromethane, but after reacting for a
few minutes all material goes into solution. A total of 480 mg were crystallized from
diethyl ether and ligroin. White crystalline powder, mp 159-161 C, Anal. Calc, for
C29H33N3O16: C 51.25; H 4.90; N 6.18. Fd. C 51.63; H 5.25; N 5.83. *H and I3C NMR
see Table 1.
10-Deacetyl paclitaxel-7, 10, 2-trinitrate ester (101)
This compound was prepared starting with 500 mg of 10-deacetyl paclitaxel and
following a procedure identical to that of paclitaxel 7, 2-dinitrate. A total of 506 mg of
product was crystallized from diethyl ether and ligroin. White crystalline powder, mp 159-
162 C, Anal. Calc, for C45H46N4O19 + H20: C 56.02; H 5.01; N 5.81. Fd. C 56.07; H
4.91; N 5.64. 'Hand 13C NMR see Table 1.
10-Deacetyl paditaxeI-7-P-xyloside-l, 2, 3, 10, 2-pentanitrate ester (104)
This compound was prepared starting with 500 mg of 10-deacetyl paclitaxel-7-p-
xyloside and following a procedure identical to that of paclitaxel-7, 2-dinitrate. As with
10-deacetyl baccatin III, all material went into solution only after the reaction had
proceeded for a few minutes. A total of 512 mg of product was crystallized from diethyl

88
ether and ligroin. White crystalline powder, mp 187-188 C, Anal. Calc, for C50H52N6O27:
C 51.38; H 4.48; N 7.19. Fd. C 51.53; H 4.59; N 6.82. *H and 13C NMR see Table 1.
Regioselective Nitration of Paclitaxe!
Paclitaxel 500 mg was dissolved in 6 ml of dichloromethane and cooled to 0 C
with an ice bath. A mixture of 5 ml of acetic anhydride and 1 ml of concentrated nitric acid
also cooled to 0 C was added and the total was stirred in an ice bath for 15 minutes. At
this point the reaction was worked up in the same manner as paclitaxel-7, 2-dinitrate.
Although a small amount of paclitaxel-7, 2-dinitrate was seen on TLC, the product was
sufficiently pure to be crystallized directly from diethyl ether and ligroin to give 435 mg of
product (96). White crystalline powder, mp 163-165 C, Anal. Calc, for C47H50N2O16 +
H20: C 61.57; H 5.72; N 3.06. Fd. C 61.95; H 6.10; N 2.91. *H NMR 5: 1.16 (s,3H, 17-
H), 1.22 (s, 3H, 16-H), 1.81 (s, 3H, 19-H), 1.83 (s, 3H, 18-H), 2.04 (m, 1H, 6-HP), 2.19
(s, 3H, 10-0Ac), 2.36 (m, 2H, 14-Ha,p), 2.41 (s, 3H, 4-OAc), 2.68 (m, 1H, 6-Ha), 3.98
(d, 6.6Hz, 1H, 3-H), 4.18 (d, 8.4Hz, 1H, 20-Hp), 4.33 (d, 8.7Hz, 1H, 20-Ha), 4.81 (br s,
1H, 2-H), 4.96 (d, 8.4Hz, 1H, 5-H), 5.68 (d, 6.9Hz, 1H, H-2), 5.74 (dd, 3.3, 10.5Hz,
1H, H-7), 5.79 (dd, 2.4, 9.0Hz, 1H, 3-H), 6.21 (t, 8.1Hz, 1H, 13-H), 6.28 (s, 1H, 10-H),
7.10 (d, 9.0Hz, 1H, NH), 7.35-7.54 (m, 10H, m-Bz, o,m,p-Ph, m,p-NBz), 7.63 (t, 7.2Hz,
1H, p-Bz), 7.75 (d, 7.2Hz, 2H, o-NBz), 8.11 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 11.0,
14.5, 20.6, 21.1, 22.5, 26.6, 32.6, 35.6, 43.2, 47.3, 55.1, 55.3, 72.2, 73.1, 74.1, 74.8,
76.2, 78.5, 79.9, 80.6, 83.6, 127,0, 128,3, 128.7, 129.0, 130.1, 132.0, 133.0, 133.6,
133.8, 137.9, 141.0, 166.7, 167.3, 169.5, 170.7, 172.7, 200.4.

89
Regioselective Nitration of 10-Deacetyl Baccatin III
10-Deacetyl baccatin III 500 mg was dissolved in 6 ml of dichloromethane and
cooled to 0 C with an ice bath. A mixture of 5 ml of acetic anhydride and 1 ml of
concentrated nitric acid also cooled to 0 C was added and the total was stirred in an ice
bath for 10 minutes. At this point the reaction was worked up in the same manner as
paclitaxel-7, 2-dinitrate. The TLC of the reaction mixture showed 4 compounds, two of
which were a small amount of 10-deacetyl baccatin III (low Rf) starting material and a
small amount of 10-deacetyl baccatin III-7, 10, 13-trinitrate (high Rf). A silica column
was ran to isolate the two intermediate products using 0-15% acetone in dichloromethane
as solvent. The more non-polar of the two products actually turned out to be a mixture of
two compounds while the more polar product was determined by NMR spectroscopy to
be 10-deacetyl baccatin III-10-mononitrate (98). The more non-polar two product
mixture was then ran on another silica column using 30-50% ethyl acetate in ligroin as
solvent, which separated the two compounds quite well. The product with the high Rf was
determined to be 10-deacetyl baccatin III-10, 13-dinitrate (99), and the other to be 10-
deacetyl baccatin III-7, 10-dinitrate (100). Product yields were as follows; 10-deacetyl
baccatin III-10-mononitrate 180 mg crystallized from diethyl ether and ligroin, 10-deacetyl
baccatin III-10, 13-dinitrate 115 mg crystallized from diethyl ether and ligroin, 10-deacetyl
baccatin III-7, 10-dinitrate 62 mg.

90
10-Deacetyl baccatin III- 10-mononitrate ester (98)
White crystalline powder, mp 169-171 C, Anal. Calc, for C29H35NO12: C 59.08,
H 5.98, N 2.38. Fd. C 58.90, H 6.28, N 2.19. !H NMR 5: 1.09 (s, 3H, 17-H), 1.13 (s,
3H, 16-H), 1.70 (s, 3H, 19-H), 1.85 (m, 1H, 6-Hp), 2.13 (s, 3H, 18-H), 2.16 (m, 2H, 14-
Ha,P), 2.30 (s, 3H, 4-OAc), 2.62 (m, 1H, 6-Ha), 3.85 (d, 6.9Hz, 1H, 3-H), 4.15 (d,
8.4Hz, 1H, 20-Hp), 4.32 (d, 8.1Hz, 1H, 20-Ha), 4.41 (dd, 6.6, 10.2Hz, 1H, 7-H), 4.91 (t,
7.8Hz, 1H, 13-H), 4.97 (d, 9.6Hz, 1H, 5-H), 5.65 (d, 7.2Hz, 1H, 2-H), 6.49 (s, 1H, 10-
H), 7.49 (t, 7.8Hz, 2H, m-Bz), 7.62 (t, 7.5Hz, 1H, p-Bz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C
NMR 6: 9.5, 15.8, 20.8, 22.5, 26.6, 36.6, 38.5, 42.5, 46.6, 58.5, 67.8, 71.6, 74.6, 76.5,
78.8, 80.7, 82.9, 84.1, 128.7, 129.2, 129.5, 130.1, 133.8, 148.3, 167.0, 170.8, 203.1.
10-Deacetyl baccatin HI-10, 13-dinitrate ester (99)
White crystalline powder, mp 202-204 C, Anal. Calc, for C29H34N2O14: C 54.89;
H 5.40; N 4.41. Fd. C 55.17; H 5.77; N 4.07. NMR 8: 1.16 (s, 3H, 17-H), 1.18 (s, 3H,
16-H), 1.69 (s, 3H, 19-H), 1.87 (m, 1H, 6-Hp), 2.08 (s, 3H, 18-H), 2.34 (m, 1H, 14-Hp),
2.38 (s, 3H, 4-OAc), 2.46 (m, 1H, 14-Ha), 2.60 (m, 1H, 6-Ha), 3.77 (d, 7.2Hz, 1H, 3-
H), 4.12 (d, 8.1Hz, 1H, 20-HP), 4.32 ( d, 8.4Hz, 1H, 20-Ha), 4.40 (dd, 6.6, 10.8Hz, 1H,
7-H), 4.94 (d, 9.0Hz, 1H, 5-H), 5.65 (d, 7.2Hz, 1H, 2-H), 6.21 (t, 8.1Hz, 1H, 13-H), 6.47
(s, 1H, 10-H), 7.50 (t, 8.1Hz, 2H, m-Bz), 7.64 (t, 7.2Hz, 1H, p-Bz), 8.06 (d, 7.2Hz, 2H,
o-Bz). 13C NMR 5: 9.4, 15.1, 20.6, 22.2, 26.5, 34.2, 36.6, 42.8, 46.6, 58.4, 60.4, 71.4,
74.2, 76.2, 78.2, 78.3, 80.7, 81.9, 84.1, 128.7, 128.8, 130.0, 132.4, 133.9, 141.6, 166.9,
170.1, 202.0.

91
10-Deacetyl baccatin III-7,10-dinitrate ester (100)
White amorphous solid, Anal Calc, for C29H34N2O14: C 54.88; H 5.40; N 4.41. Fd.
C 55.01; H 5.63; N 4.29. *H NMR 5: 1.08 (s,3H, 17-H), 1.10 (s,3H, 16-H), 1.82 (s, 3H,
19-H), 2.05 (m, 1H, 6-H(3), 2.16 (s, 3H, 18-H), 2.20 (m, 2H, 14-Ha,(3), 2.32 (s, 3H, 4-
OAc), 2.72 (m, 1H, 6-Ha), 3.99 (d, 6.9Hz, 1H, 3-H), 4.13 (d, 8.4Hz, 1H, 20-Hp), 4.35
(d, 8.4Hz, 1H, 20-Ha), 4.91 (t, 8.4Hz, 1H, 13-H), 4.99 (d, 9.0Hz, 1H, 5-H), 5.62 (d,
6.9Hz, 1H, 2-H), 5.80 (dd, 7.2, 10.5Hz, 1H, 7-H), 6.40 (s, 1H, 10-H), 7.50 (t, 7.5Hz, 2H,
m-Bz), 7.63 (t, 7.5Hz, 1H, p-Bz), 8.09 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 10.8, 15.6,
20.4, 22.4, 26.6, 32.5, 38.4, 42.5, 47.8, 55.7, 60.4, 67.8, 73.8, 76.1, 78.5, 80.1, 80.2,
82.4, 83.5, 128.7, 128.9, 129.0, 130.0, 133.9, 148.9, 166.8, 171.1, 200.1.
Conversion of 10-Deacetyl Paditaxel-7-P-XyIoside to 10-Deacetyl Paclitaxel (112)
10-Deacetyl paclitaxel-7-p-xyloside 1.0 g was dissolved in 10 ml of 1 : 1 THF and
water and 2 ml of 1 N H2S04 was added. This was followed by 0.71 g of NaI04 and the
mixture was stirred overnight at room temperature. The mixture was diluted with water
and extracted three times with dichloromethane. The organic layer was evaporated to
dryness to yield 0.95 g of a white powder. The material was then dissolved in 20 ml of
methanol and 0.5 ml of phenylhydrazine and 3 ml of acetic acid was added. This mixture
was heated at 60 C for 3 hours at which point TLC analysis showed almost complete
conversion to 10-deacetyl paclitaxel. The mixture was diluted with water, acidified and
extracted with dichloromethane three times and the organic portion was evaporated to a
dark red oil The concentrate was partitioned in a countercurrent fashion, between 3 : 2
methanol/water and 4: 1 benzene/ligroin as the two phases, and using 3 separatory funnels.

92
The benzene/ligroin layers which contained the phenylhydrazones were separated and
concentrated to dryness. The combined methanol/water layer was concentrated partially
and extracted three times with dichloromethane and the organic layer was concentrated to
yield crude 10-deacetyl paclitaxel (0.8g). This material was clean enough for further work.
All NMR spetra matched that of an authentic sample.
Regioselective Nitrations of 10-Deacetyl Paclitaxel
10-Deacetyl Paclitaxel 500 mg was dissolved in 6 ml dichloromethane and cooled
to 0 C. Next 3 ml of a 5 : 1 mixture of acetic anhydride and concentrated nitric acid
which was also cooled to 0 C was added and the reaction mixture was stirred in an ice
bath for 5 minutes. The reaction was worked up as with paclitaxel-7, 2-nitrate ester.
Analysis of the reaction mixture by TLC revealed that practically all the starting material
was gone and four products were present with two being major products. The fastest
moving product was the completely nitrated 10-deacetyl paclitaxel and was in minor
amounts. The next was a major product followed by two minor products and then by
another major product. A silica column was used to separate this products with 0 15%
acetone in dichloromethane as the solvent. From this column 175 mg of the faster major
product (103) was obtained and crystallized from diethyl ether while 126 mg of the slower
major product (102) was isolated but contained impurities. This was put on another
column of the same type and solvent and 87 mg of the product was obtained. This material
however would not crystallize. After analysis by NMR the faster product was determined
to be 10-deacetyl paclitaxel-7, 10-dinitrate ester (103) and the slower product was the 10-
mononitrate ester (102).

93
10-Deacetyl paclitaxel-10-mononitrate ester (102)
White amorphous powder, Anal. Calc, for C45H48N2O15: C 63.08; H 5.65; N 3.27.
Fd. C 62.75; H 5.97; N 3.90. XH NMR 5: 1.14 (s, 3H, 17-H), 1.19 (s, 3H, 16-H), 1.61 (s,
3H, 18-H), 1.83 (s, 3H, 18-H), 1.98 (m, 1H, 6-Hp), 2.32 (m, 2H, 14-Ha,p), 2.38 (s, 3H,
4-OAc), 2.56 (m, 1H, 6-Ha), 3.75 (d, 6.9Hz, 1H, 3-H), 4.18, (d, 8.7Hz, 1H, 20-Hp),
4.21 (dd, 3.6, 5.4Hz, 1H, 7-H), 4.30 (d, 8.1Hz, 1H, 20-Ha), 4.79 (br s, 1H, 2-H), 4.90
(d, 8.7Hz, 1H, 5-H), 5.68 (d, 7.2Hz, 1H, 2-H), 5.75 (dd, 2.1, 8.7Hz, 1H, 3-H), 6.19 (t,
8.4Hz, 1H, 13-H), 6.41 (s, 1H, 10-H), 7.09 (d, 9.0Hz, 1H, NH), 7.35-7.53 (m, 10H, m-
Bz, o,m,p-Ph, m,p-NBz), 7.61 (t, 7.2Hz, 1H, p-Bz), 7.73 (d, 6.9Hz, 2H, o-NBz), 8.11 (d,
7.2Hz, 2H, o-Bz). 13C NMR 5: 9.6, 15.0, 21.6, 22.5, 26.4, 35.5, 36.6, 43.0, 46.2, 55.2,
58.4, 68.2, 71.5, 72.1, 73.1, 74.5, 76.5, 78.6, 81.0, 82.2, 84.1, 127.0, 127.1, 128.4, 128.7,
128.8, 129.0, 130.2, 130.7, 130.9, 132.0, 133.5, 133.8, 137.8, 143.8, 166.8, 167.3, 170.5,
172.6, 202.6.
10-Deacetyl paclitaxel-7,10-dinitrate ester (103)
White crystalline powder, mp 170-172 C, Anal. Calc, for C45H47N3O18: C 58.76;
H 5.37; N 4.57. Fd. C 59.11; H 5.27; N 4.49. !H NMR 5: 1.13 (s, 3H, 17-H), 1.20 (s,
3H, 16-H), 1.83 (s, 3H, 19-H), 1.87 (s, 3H, 18-H), 2.06 (m, 1H, 6-Hp), 2.36 (m, 2H, 14-
Ha, p), 2.41 (s, 3H, 4-OAc), 2.70 (m, 1H, 6-Ha), 3.91 (d, 6.9Hz, 1H, 3-H), 4.17 (d,
8.4Hz, 1H, 20-Hp), 4.34 (d, 8.4Hz, 1H, 20-Ha), 4.80 (br s, 1H, 2-H), 4.95 (d, 9.0Hz,
1H, 5-H), 5.67 (d, 6.9Hz, 1H, 2-H), 5.71 (dd, 7.2, 10.5Hz, 1H, 7-H), 5.77 (dd, 2.1,
8.7Hz, 1H, 3-H), 6.20 (t, 9.0Hz, 1H, 13-H), 6.33 (s, 1H, 10-H), 7.07 (d, 9.0Hz, 1H,
NH), 7.36-7.63 (m, 10-H, m-Bz, o,m,p-Ph, m,p-NBz,), 7.63 (t, 7.5Hz, 1H, p-Bz), 7.73

94
(d, 7.2Hz, 2H, o-NBz), 8.10 (d, 6.9Hz, 2H, o-Bz). 13C NMR 5: 10.9, 14.9, 21.2, 22.4,
26.4, 32.5, 35.5, 43.0, 47.3, 55.2, 55.6, 72.0, 73.1, 73.8, 76.2, 78.5, 79.9, 80.5, 81.7,
83.4, 127.0, 127.1, 128.4, 128.7, 128.8, 129.1, 130.1, 130.2, 131.2, 132.1, 133.5, 133.9,
137.7, 144.2, 166.7, 167.4, 170.9, 172.6, 199.5.
Regioselective Nitrations of 10-Deacetyl Paclitaxel-7-p-Xyloside
10-Deacetyl paclitaxel-7-(3-xyloside 1.0 g was dissolved in 12 ml of
dichloromethane and was cooled to 0 C. To this was added 6 ml of a 5 : 1 mixture of
acetic anhydride and concentrated nitric acid which was also cooled to 0 C and the
reaction mixture was stirred in an ice bath for 5-10 minutes. The reaction was worked up
in the normal way and analyzed by TLC. This analysis showed many product spots
however some seemed to be more predominate than others. Initially, a crude silica column
was ran on this mixture using 0% methanol and 5% acetone in dichloromethane - 5%
methanol and 15% acetone in dichloromethane as the solvent. The fractions from this
column were combined into three groups; fast, medium, and slow in elution order. The
fast moving group contained some completely nitrated product, one major product that
was slower than the completely nitrated one, and a couple of minor products. This
material was ran on a silica column using 40% ethyl acetate in ligroin 60% ethyl acetate
in ligroin as the solvent. A total of 116 mg of the major product was obtained and
crystallized from dichloromethane. This product was determined to be 10-deacetyl
paclitaxel-7-(3-xyloside-2, 3, 4, 10-tetranitrate ester (108). The middle group from the
initial crude column contained many minor products and further separation was not
attempted. The slowest group from the initial column contained three major products as

95
well as some starting material and minor products. A clean separation of the three major
products was not possible using only one solvent system thus a first column was ran on
this material using 40% > 80% ethyl acetate in ligroin. This eluted 84 mg of the slowest
of the three products in pure form and this compound was determined to be 10-deacetyl
paclitaxel-7-P-xyloside-2-mononitrate ester (105). This product did not crystallize. The
mixture of the remaining two products was separated on a silica column using 30% >
50% acetone in benzene as solvent. The faster of these two products was the 3-
mononitrate ester (106) and was crystallized from dichloromethane (76 mg). The slower
product was determined to be the 4-mononitrate ester (107) (101 mg) and was not
crystallized.
10-Deacetyl paclitaxel-7-(3-xyloside-2, 3, 4, 10-tetranitrate ester (108)
White crystalline powder, mp 182-184 C, Anal. Calc, for C50H53N5O25: C 53.43;
H 4.75; N 6.23. Fd. C 53.67; H 4.82; N 5.87. >H NMR 5: 1.12 (s, 3H, 17-H), 1.18 (s,
3H, 16-H), 1.75 (s, 3H, 19-H), 1.86 (s, 3H, 18-H), 2.03 (m, 1H, 6-H{3), 2.34 (m, 2H, 14-
Ha,(3), 2.39 (s, 3H, 4-OAc), 2.80 (m, 1H, 6-Ha), 3.67 (dd, 5.4, 12.9Hz, 1H, 5-Heq),
3.77 (d, 6.9Hz, 1H, 3-H), 4.14 (m, 1H, 5-Hax), 4.20 (d, 8.1Hz, 1H, 20-Hp), 4.21 (m,
1H, 7-H), 4.32 (d, 8.1Hz, 1H, 20-Ha), 4.76 (d, 4.2Hz, 1H, 1-H), 4.78 (d, 2.7Hz, 1H,
2-H), 4.85 (d, 8.1Hz, 1H, 5-H), 4.97 (dd, 2.1, 3.9Hz, 1H, 2-H), 5.07 (dd, 4.2, 5.4Hz,
1H, 4-H), 5.26 (t, 6.0Hz, 1H, 3-H), 5.69 (d, 6.9Hz, 1H, 2-H), 5.75 (dd, 2.1, 8.4Hz,
1H, 3-H), 6.18 (t, 9.0Hz, 1H, 13-H), 6.39 (s, 1H, 10-H), 7.05 (d, 9.0Hz, 1H, N-H), 7.36-
7.53 (m, 10H, m-Bz, o,m,p-Ph, m,p-NBz), 7.62 (t, 7.2Hz, 1H, p-Bz), 7.72 (d, 7.2Hz, 2H,
o-NBz), 8.11 (d, 7.5Hz, 2H, o-Bz). 13C NMR 5: 10.6, 14.9, 21.4, 22.5, 26.2, 35.5, 35.8,

96
42.9, 45.7, 55.3, 57.8, 59.4, 72.0, 72.7, 73.2, 74.0, 74.2, 74.4, 76.4, 78.5, 80.2, 80.5,
82.2, 83.5, 98.8, 127.0, 127.1, 128.4, 128.7, 128.8, 129.0, 129.1, 130.1, 131.1, 132.1,
133.5, 133.8, 137.7, 143.2, 166.8, 167.4, 170.9, 172.7, 200.0.
10-Deacetyl paclitaxeI-7-p-xyIoside-2-mononitrate ester (105)
White amorphous powder, Anal. Calc, for C50H56N2O19 + H20: C 59.64; H 5.81;
N 2.78. Fd. C 60.02; H 6.18; N 2.40. HNMR5: 1.17 (s, 3H, 17-H), 1.26 (s, 3H, 16-H),
1.74 (s, 3H, 19-H), 1.81 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.29 (m, 2H, 14-Ha,p), 2.36
(s, 3H, 4-OAc), 2.72 (m, 1H, 6-Ha), 3.17 (t, 10.8, 1H, 5-Heq), 3.50 (t, 8.7Hz, 1H, 3-
H), 3.66 (dd, 7.5, 12.3Hz, 1H, 4-H), 3.84 (d, 6.6Hz, 1H, 3-H), 3.91 (dd, 4.5, 11.4Hz,
1H, 5-Hax), 4.04 (m, 1H, 7-H), 4.18 (m, 1H, 1-H), 4.19 (d, 8.7Hz, 1H, 20-HP), 4.28
(d, 8.7Hz, 1H, 20-Ha), 4.79 (m, 1H, 2-H), 4.81 (m, 1H, 2-H), 4.88 (d, 9.0Hz, 1H, 5-
H), 5.07 (s, 1H, 10-H), 5.61 (d, 6.6Hz, 1H, 2-H), 5.73 (dd, 2.1, 8.7Hz, 1H, 3-H), 6.15 (t,
7.8Hz, 1H, 13-H), 7.23 (d, 9.0Hz, 1H, N-H), 7.34-7.53 (m, 10H, m-Bz, o,m,p-Ph, m,p-
NBz), 7.61 (t, 7.2Hz, 1H, p-Bz), 7.74 (d, 7.8Hz, 2H, o-NBz), 8.09 (d, 7.5Hz, 2H, o-Bz).
13C NMR 5: 10.6, 14.2, 20.4, 22.5, 26.6, 35.7, 35.8, 43.0, 46.5, 55.3, 57.0, 65.1, 69.9,
72.4, 73.2, 73.6, 74.4, 74.6, 76.5, 78.6, 80.6, 80.9, 81.8, 84.0, 101.5, 127.1, 127.2, 128.4,
128.7, 128.8, 129.0, 129.1, 130.2, 132.1, 133.6, 133.7, 136.1, 137.8, 137.9, 166.8, 167.4,
170.8, 172.7, 209.2.
10-Deacetyl paclitaxel-7-P-xyloside-3-mononitrate ester (106)
White crystalline powder, mp 209-211 C, Anal. Calc, for C50H56N2O19 + H20: C
59.64; H 5.81; N 2.78. Fd. C 59.27; H 5.87; N 2.89. ^NMRS: 1.09 (s,3H, 17-H), 1.19
(s, 3H, 16-H), 1.80 (s, 3H, 19-H), 1.81 (s, 3H, 18-H), 2.02 (m, 1H, 6-Hp), 2.28 (m, 2H,

97
14-Ha,p), 2 39 (s, 3H, 4-OAc), 2.71 (m, 1H, 6-Ha), 3.35 (m, 1H, 5-Heq), 3.38 (m,
1H, 2-H), 3.75 (br s, 1H, 4-H), 3.88 (d, 6.3Hz, 1H, 3-H), 3.95 (m, 1H, 5-Hax), 4.08
(t, 8.1Hz, 1H, 7-H), 4.15 (d, 6.6Hz, 1H, 1-H), 4.20 (d, 8.1Hz, 1H, 20-HP), 4.31 (d,
8.1Hz, 1H, 20-Ha), 4.79 (br s, 1H, 2-H), 4.90 (d, 9.0Hz, 1H, 5-H), 5.03 (t, 8.4Hz, 1H,
3-H), 5.18 (s, 1H, 10-H), 5.65 (d, 6.6Hz, 1H, 2-H), 5.77 (d, 7.8Hz, 1H, 3-H), 6.19 (t,
7.8Hz, 1H, 13-H), 7.15 (d, 7.5Hz, 1H, N-H), 7.40-7.53 (m, 10H, m-Bz, o,m,p-Ph, m,p-
NBz), 7.62 (t, 6.9Hz, 1H, p-Bz), 7.75 (d, 7.5Hz, 2H, o-NBz), 8.12 (d, 7.2Hz, 2H, o-Bz).
13C NMR 8: 10.1, 13.6, 20.3, 22.0, 25.9, 34.9, 35.0, 42.6, 46.0, 55.1, 55.8, 65.0, 66.2,
69.6, 70.7, 73.4, 74.1, 74.3, 75.6, 78.5, 80.2, 81.3, 83.4, 86.2, 104.3, 126.4, 126.9, 127.7,
127.8, 127.9, 128.0, 129.3, 129.4, 130.8, 132.6, 133.8, 135.6, 137.0, 138.5, 165.6, 166.5,
169.6, 172.0, 208.5.
10-Deacetyl pacIitaxel-7-3-xyloside-4-mononitrate ester (107)
White crystalline powder, mp 193-195 C, Anal. Calc, for C50H56N2O19 + H20: C
59.64; H 5.81; N 2.78. Fd.C 59.96; H 6.18; N 2.56. 'HNMR5: 1.12 (s,3H, 17-H), 1.25
(s, 3H, 16-H), 1.79 (s, 3H, 19-H), 1.81 (s, 3H, 18-H), 2.09 (m, 1H, 6-HP), 2.31 (m, 2H,
14-Ha,p), 2.34 (s, 3H, 4-OAc), 2.64 (m, 1H, 6-Ha), 3.01 (br s, 1H, 2-H), 3.28 (t,
9.9Hz, 1H, 5-Heq), 3.51 (t, 8.7Hz, 1H, 3-H), 3.83 (d, 6.3Hz, 1H, 3-H), 3.98 (d,
7.5Hz, 1H, 5-Hax), 4.11 (m, 1H, 1-H), 4.12 (m, 1H, 7-H), 4.19 (d, 7.8Hz, 1H, 20-
Hp), 4.30 (d, 8.4Hz, 1H, 20-Ha), 4.82 (br s, 1H, 2-H), 4.90 (m, 1H, 5-H), 4.91 (m, 1H,
4-H), 5.28 (s, 1H, 10-H), 5.64 (d, 6.6Hz, 1H, 2-H), 5.77 (d, 9.3Hz, 1H, 3-H), 6.17 (t,
7.8Hz, 1H, 13-H), 7.21 (d, 9.3Hz, 1H, N-H), 7.31-7.53 (m, 10-H, m-Bz, o,m,p-Ph, m,p-
NBz), 7.61 (t, 7.5Hz, 1H, p-Bz), 7.72 (d, 7.5Hz, 2H, o-NBz), 8.14 (d, 7.5Hz, 2H, o-Bz).

98
13C NMR 5: 10.8, 14.2, 20.6, 22.6, 26.7, 35.3, 35.4, 43.1, 46.8, 54.9, 56.8, 61.6, 71.4,
72.4, 73.1, 74.5, 74.7, 76.6, 78.7, 79.7, 81.1, 81.7, 84.0, 104.4, 127.0, 127.1, 128.3,
128.7, 128.8, 128.9, 129.0, 130.3, 132.0, 133.6, 133.7, 136.1, 137.9, 138.1, 166.8, 167.5,
170.6, 173.0, 210.
Reductive Denitration of Paclitaxel-7, 2-Dinitrate Ester
Paclitaxel-7, 2-dinitrate 200 mg was dissolved in 3 ml of acetic acid and 500 mg
of zinc powder was added. The mixture was stirred at room temperature for 30 minutes
and then filtered through a celite bed to remove the zinc powder. Water was added to the
acetic acid and the acid was neutralized with bicarbonate. The aqueous layer was
extracted with dichloromethane twice and the organic layer was dried with sodium sulfate.
After removal of the solvent 174 mg of paclitaxel was crystallized from diethyl ether and
ligroin. All NMR spectra matched those of an authentic sample.
Reaction of PaclitaxeI-7-2-Dinitrate with NaBH4
Paclitaxel-7-2-dinitrate ester 200 mg was dissolved in 2 ml of methanol and
excess NaBH4 was added. This was stirred for 10 minutes at room temperature and the
reaction was quenched with 1 N HC1. The methanol was partially removed and water and
dichloromethane was added and partitioned. The organic layer was removed and the water
layer was extracted twice more with dichloromethane. The combined water layers were
washed with water once and then dried with sodium sulfate. TLC analysis showed two
products, the faster was UV active but did not char with H2S04 while the slower product
gave a positive test in both cases. The evaporated residue was applied to a regular silica
column and eluted with 5-15% acetone in benzene. Both compounds were isolated and

99
crystallized from diethyl ether and ligroin. The faster spot was determined to be the side
chain alcohol nitrate ester (116) (34 mg) and the slower product was baccatin III-7-nitrate
ester (117) (148 mg).
Side chain alcohol nitrate ester (116)
Colorless needles, *H NMR 6: 3.68 (t, 8.7Hz, 1H, O-H), 3.87 (d, 14.1Hz, 2H, 1-
H) 5.20 (m, 1H, 2-H), 5.80 (dd, 3.0, 9.6Hz, 1H, 3-H), 6.77 (d, 9.0Hz, 1H, N-H), 7.36-
7.60 (m, 8H, o,m,p-Ph, m,p-NBz), 7.81 (d, 6.9Hz, 2H, o-NBz). 13C NMR 5: 52.0, 59.2,
83.7, 126.6, 127.2, 128.4, 128.6, 128.9, 129.3, 132.5, 136.7, 168.6.
Baccatin IIlt-7-nitrate ester (117)
White crystalline powder, UV (CH3OH): 231 nm, *H NMR 8: 1.10 (s, 3H,
17-H), 1.13 (s, 3H, 16-H), 1.80 (s, 3H, 19-H), 2.03 (m, 1H, 6-H(3), 2.10 (s, 3H, 18-H),
2.20 (s, 3H, 10-0Ac) 2.24 (m, 2H, 14-Ha,(3), 2.32 (s, 3H, 4-OAc), 2.71 (m, 1H, 6-Ha),
4.07 (d, 6.9Hz, 1H, 3-H), 4.14 (d, 8.7Hz, 1H, 2O-H0), 4.34 (d, 8.7Hz, 1H, 20-Ha), 4.89
(t, 8.1Hz, 1H, 13-H), 5.00 (d, 8.7Hz, 1H, 5-H), 5.62 (d, 7.2Hz, 1H, 2-H), 5.80 (dd, 7.2,
10.5Hz, 1H, 7-H), 6.34 (s, 1H, 10-H), 7.49 (t, 7.8Hz, 2H, m-Bz), 7.62 (t, 7.5Hz, 1H, p-
Bz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 10.9, 15.2, 20.2, 20.7, 22.5, 26.8, 32.6,
38.4, 42.7, 47.8, 55.4, 67.8, 74.1, 75.4, 76.2, 78.7, 80.2, 80.3, 83.6, 128.7, 129.1, 130.1,
131.7, 133.8, 145.3, 166.9, 169.5, 171.0, 200.9.
Selective Denitration of Paclitaxel-7, 2-Dinitrate Ester
Paclitaxel-7, 2-dinitrate 200 mg was dissolved in 2 ml of acetonitrile and 0.1 ml of
20% ammonium sulfide and stirred at room temperature for 2 minutes. Water was added
and the mixture was extracted with diethyl ether twice. Removal of the ether left a white

100
solid (167 mg) which was determined to be paclitaxel-7-mononitrate ester (119). All NMR
spectra matched those of an authentic sample.
Acetylation of 10-Deacetyl Paclitaxel-7-p-Xyloside-2, 3, 4, 10, 2,-Pentanitrate
Ester
10-Deacetyl paclitaxel-7-p-xyloside-2, 3, 4, 10, 2-tetranitrate 600 mg was
dissolved in 25 ml of acetic anhydride and 120 mg of DMAP was added and the reaction
was stirred at room temperature for overnight. Water was added as was NaHC03 with
stirring until no further frothing was observed. The aqueous mixture was then extracted
three times with dichloromethane and this organic layer was washed with a saturated NaCl
solution, with water, and dried with Na2S04. The TLC analysis of the organic layer
showed two major products which were slightly faster moving than the starting material.
A couple of minor products were present but no starting material was seen. This material
was ran on a regular silica column using 25% 40% ethyl acetate in ligroin. A total of
172 mg of the faster spot was isolated and crystallized from acetone and ligroin. This
product was determined to be the di-enol acetate (122). A total of 146 mg of the slower
spot was isolated and crystallized from dichloromethane. This product was determined to
be the mono-enol acetate (121).
10-Deacetyl paditaxel -7-P-xyloside-2, 3, 4, 10-tetranitrate-di-enol acetate (122)
Clear colorless needles, mp 193-195C, UV (CH3OH): 229 nm, FABMS m/z:
1206 (M+ 1), 703, 613, 522. 'HNMRS: 1.13 (s, 3H, 17-H), 1.19 (s, 3H, 16-H), 1.77 (s,
3H, 19-H), 2.03 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.21 (s, 3H, OAc), 2.29 (m, 1H, 14-
HP), 2.30 (s, 3H, OAc), 2.52 (m, 1H, 14-Ha), 2.53 (s, 3H, OAc), 2.79 (m, 1H, 6-Ha),

101
3.70 (dd, 5.1, 12.9Hz, 1H, 5-H), 3.87 (d, 6.9Hz, 1H, 3-H), 4.18 (d, 8.4Hz, 1H, 20-H(3),
4.23 (m, 1H, 7-H), 4.29 (d, 8.4Hz, 1H, 20-Ha), 4.80 (d, 8.7Hz, 1H, 5-H), 4.81 (d, 3.8Hz,
1H, 1 -H), 4.97 (dd, 4.2, 5.7Hz, 1H, 2-H), 5.07 (dd, 4.5, 8.7Hz, 1H, 4-H), 5.26 (t,
5.7Hz, 1H, 3-H), 5.65 (d, 7.2Hz, 1H, 2-H), 6.06 (t, 7.8Hz, 1H, 13-H), 6.44 (s, 1H, 10-
H), 6.73 (d, 7.2Hz, 2H, o-NBz), 7.17 (t, 7.8Hz, 2H, m-NBz) 7.23-7.28 (m, 3H, o-Ph, p-
NBz), 7.52-7.63 (m, 4H, m,p-Bz, p-Ph), 8.21 (d, 7.5Hz, 2H, o-Bz). 13C NMR 5: 10.7,
15.2, 20.3, 21.1, 21.7, 26.1, 26.5, 35.7, 36.0, 42.7, 46.0, 57.9, 59.3, 71.4, 72.7, 73.9,
74.1, 74.5, 76.4, 79.4, 80.0, 80.3, 82.2, 83.5, 98.6, 128.1, 128.3, 128.4, 128.5, 128.8,
129.2, 129.8, 130.4, 131.4, 132.6, 133.6, 133.7, 133.8, 135.0, 137.3, 142.7, 161.6, 167.0,
170.1, 172.0, 172.1, 174.6, 200.0.
10-Deacetyl paclitaxel -7-P-xyIoside-2, 3, 4, 10-tetranitrate-mono-eno! acetate
(121)
White crystalline powder, mp 178-180 C, UV Amax (CH3OH): 232 nm, FABMS
m/z: 1164 (M + 1), 326, 308. NMR 5: 1.15 (s, 3H, 17-H), 1.22 (s, 3H, 16-H), 1.77
(s, 3H, 19-H), 2.01 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.10 (s, 3H, OAc), 2.11 (m, 1H,
14-HP), 2.39 (s, 3H, OAc), 2.56 (m, 1H, 14-Ha), 2.83 (m, 1H, 6-Ha), 3.72 (dd, 5.1,
12.9Hz, 1H, 5-Heq), 3.81 (d, 7.2Hz, 1H, 3-H), 4,16 (d, 8.7Hz, 1H, 20-Hp), 4.16 (m,
1H, 7-H), 4.23 (dd, 3.6, 15.6Hz, 1H, 5-Hax), 4.33 (d, 8.7Hz, 1H, 20-Ha), 4.83 (d,
3.6Hz, 1H, 1-H), 4.87 (d, 8.7Hz, 1H, 5-H), 4.97 (dd, 3.9, 6.0Hz, 1H, 2-H), 5.07 (dd,
5.1, 9.3Hz, 1H, 4-H), 5.26 (t, 5.7Hz, 1H, 3-H), 5.69 (d, 6.9Hz, 1H, 2-H), 6.10 (t,
8.1Hz, 1H, 13-H), 6.44 (s, 1H, 10-H), 7.39-7.52 (m, 9H, m-Bz, o,m,p-Bz, m-NBz), 7.57
(m, 1H, p-NBz), 7.61 (m, 1H, p-Bz), 7.94 (d, 7.2Hz, 2H, o-NBz), 8.06 (d, 7.2Hz, 2H, o-

102
Bz), 11.41 (s, 1H, N-H). 13C NMR 8: 10.5, 15.3,20.3,21.1,21.9,26.2,35.8,36.2,42.8,
45.8, 57.9, 59.1, 71.5, 72.3, 73.7, 73.8, 74.1, 76.4, 79.0, 80.3, 80.4, 82.2, 83.3, 98.6,
120.9, 127.6, 127.7, 128.2, 128.7, 128.8, 128.9, 129.6, 130.0, 130.9, 131.7, 132.8, 133.0,
133.9, 143.2, 147.7, 164.6, 165.0, 166.9, 169.6, 170.7, 200.0.
Reaction of Paclitaxel-7, 2-Dinitrate Ester with NaN3
Paclitaxel-7, 2-dinitrate ester 200 mg was dissolved in 4 ml of acetonitrile and 200
mg of sodium azide was added. This mixture was stirred at room temperature overnight.
At that point the acetonitrile was partially evaporated and the residue was partitioned
between water and dichloromethane. The organic layer was removed and the water layer
was partitioned twice more with dichloromethane. The combined organic layers were
washed once with water and then dried with sodium sulfate. TLC analysis showed no
starting material and two products. The faster moving product was UV active but did not
char with 1 N H2SO4 while the slower product did give a positive result in both cases. The
solvent was removed and the residue was put on a regular silica column with 5-15%
acetone in benzene. Both products were cleanly isolated and the slower product
crystallized from diethyl ether and ligroin while the faster compound crystallized upon
evaporation of the fraction solvents. The faster spot was determined to be dibenzamide
(125) (20 mg) and the slower product was baccatin III-7-nitrate ester (126) (152 mg).
Dibenzamide (125)
Colorless needles, mp 150-151 C, UV Lmax (CH3OH): 242 nm, IR (KBr): 3240,
1770, 1475, 1225, 1115, 705 cm'1, HNMR6: 7.51 (t, 7.5Hz, 4H, m-NBz), 7.61 (t,

103
7.2Hz, 2H, p-NBz), 7.87 (d, 7.8Hz, 4H, o-NBz), 8.99 (br s, 1H, N-H). 13C NMR 5:
127.9, 128.9, 133.1, 133.3, 166.4.
Baccatin IU-7-nitrate ester (126)
see compound 117
Synthesis of 2-Oxo Paditaxel-7-Nitrate Ester from Paclitaxel-7, 2-Dinitrate Ester
Paclitaxel-7, 2-dinitrate ester 200 mg was dissolved in 3 ml of DMF and 200 mg
of NaN3 was added. Immediately the solution turned a deep pink color while stirring at
room temperature. After 10 minutes water was added and the color dissipated and a white
solid precipitated from the solution. This solid was filtered and dried to yield 182 mg of
>95% keto-ester (124). White amorphous powder, UV X* (CH3OH): 230 nm, IR
(KBr): 2960, 1725, 1640, 1270, 1225, 1060, 1020, 830, 700 cm'1. FABMS: 897 (82%,
M+l), 614 (13%), 554 (35%), 307 (16%), 284 (55%), 210 (100%). For NMR see
below compound.
Synthesis of 2-Oxo Paclitaxel-7-Nitrate Ester from PaclitaxeI-7-Mononitrate Ester
Paclitaxel-7-mononitrate ester 200 mg was dissolved in 3 ml of acetone and a few
drops of 3 N Jones reagent was added. This solution was stirred at 60 C and checked by
TLC every 30-60 minutes and more Jones reagent was added as needed. After about 6
hours the TLC showed about a 40% conversion and the reaction didnt seem to proceed
any further so the acetone was evaporated and the residue was taken up in water. The
water layer was extracted with dichloromethane twice and the combined organic layers
were washed with water once. The dried organic layer was evaporated to dryness and the
residue was put on a silica column and eluted with 5-15% acetone in benzene. A total of

104
63 mg of product (128) was obtained. White amorphous powder, *H NMR 5: 1.21 (s,
3H, 17-H), 1.26 (s, 3H, 16-H), 1.79 (s, 3H, 19-H), 1.99 (s, 3H, 18-H), 2.05 (m, 1H, 6-
Hj3), 2.12 (m, 1H, 14-H(3), 2.18 (s, 3H, 4-OAc), 2.20 (s, 3H, 10-OAc), 2.34 (m, 1H, 14-
Ha), 2.70 (m, 1H, 6-Ha), 4.01 (d, 6.6Hz, 1H, 3-H), 4.11 (d, 8.4Hz, 1H, 20-H(3), 4.31 (d,
8.7Hz, 1H, 20-Ha), 4.97 (d, 8.7Hz, 1H, 5-H), 5.64 (d, 6.9Hz, 1H, 2-H), 5.79 (dd, 7.5,
10.8Hz, 1H, 7-H), 6.20 (t, 8.1Hz, 1H, 13-H), 6.31 (s, 1H, 10-H), 6.42 (d, 5.4Hz, 1H, 3-
H), 7.14 (d, 5.4Hz, 1H, N-H), 7.43-7.51 (m, 10H, m-Bz, o,m,p-Ph, m,p-NBz), 7.62 (t,
6.3Hz, 1H, p-Bz), 7.84 (d, 7.2Hz, 2H, o-NBz), 8.03 (d, 7.2Hz, 2H, o-Bz).
Acetylation of 2-Oxo-PacIitaxel-7-Mononitrate Ester
A total of 300 mg of 2-oxo-paclitaxel-7-mononitrate ester was dissolved in 2 ml
of acetic anhydride and 50 mg of DMAP was added and the solution was stirred at room
temperature for 18 hours. Water was then added to the solution and sodium bicarbonate
was added slowly with stirring. After the release of CO2 stopped the solution was
extracted twice with dichloromethane. The combined organic layers were washed with 0.1
N NaOH, 0.1 N HC1, and water successively, dried with sodium sulfate and evaporated to
a solid residue. This residue was put on a silica column and eluted with 0-10% acetone in
dichloromethane. Two major products were eluted, the faster being the C-2-C-3 mono-
enol acetate (130) which was crystallized from diethyl ether and ligroin to yield 175 mg,
and the slower product was determined to probably be one of two possible enols (131,
132). This was also crystallized from diethyl ether and ligroin to yield 38 mg.

105
Padtaxel-7-mononitrate ester-2-3-enol acetate (130)
White crystalline powder, mp 175-178 C, UV (CH3OH): 233 nm, FABMS
m/z: 939 (39%, M+l), 614 (31%), 554 (59%), 326 (27%), 308 (100%), 266 (79%), 237
(35%), 204 (83%). H NMR 5: 1.18 (s, 3H, 17-H), 1.26 (s, 3H, 16-H), 1.82 (s, 3H, 19-
H), 1.97 (s, 3H, 18-H), 2.05 (m, 3H, 6-H(3), 2.10 (m, 1H, 14-Hp), 2.11 (s, 3H, 2~OAc),
2.21 (s, 3H, 10-0Ac), 2.41 (s, 3H, 4-OAc), 2.59 (m, 1H, 14-Ha), 2.70 (m, 1H, 6-Ha),
4.02 (d, 6.6Hz, 1H, 2-H), 4.15 (d, 8.7Hz, 1H, 20-Hp), 4.34 (d, 8.7Hz, 1H, 20-Ha), 4.98
(d, 9.3Hz, 1H, 5-H), 5,67 (d, 6.9Hz, 1H, 2-H), 5.75 (dd, 7.2, 10.5Hz, 1H, 7-H), 6.09 (t,
8.4Hz, 1H, 13-H), 6.33 (s, 1H, 10-H), 7.37-7.63 (m, 11H, m,p-Bz, o,m,p-Ph, m,p-NBz),
7.94 (d, 7.2Hz, 2H, o-NBz), 8.06 (d, 6.9Hz, 2H, o-Bz), 11.43 (br s, 1H, N-H). 13C NMR
5: 10.9, 14.8, 20.3, 20.6, 20.8, 21.8, 26.5, 32.6, 36.3, 43.1, 47.4, 55.4, 71.8, 74.0, 74.7,
76.2, 79.1, 80.0, 80.4, 83.5, 121.1, 127.6, 127.7, 128.2, 128.7, 128.9, 129.0, 129.5,
130.0, 131.7, 132.8, 132.9, 133.1, 133.9, 141.0, 147.4, 164.8, 165.0, 166.8, 169.4, 169.7,
170.4, 200.5.
PaclitaxeI-7-nitrate ester enol (131,132)
White crystalline powder, mp 176-179 C, FABMS m/z: 939 (23%, M+l), 614
(18%), 554 (29%), 460 (25%), 410 (52%), 308 (29%), 266 (38%), 136 (27%), 105
(100%). ]H NMR 6: 1.09 (s, 3H, 17-H), 1.15 (s, 3H, 16-H), 1.28 (s, 3H, 19-H), 1.77 (s,
3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.17 (s, 3H, OAc), 2.23 (m, 1H, 14-HP) 2.28 (s, 3H,
OAc), 2.40 (m, 1H, 14-Ha), 2.41 (s, 3H, OAc), 2.63 (m, 1H, 6-Ha), 3.92 (d, 7.2Hz, 1H,
3-H), 4.15 (d, 8.4Hz, 1H, 20-Hp), 4.28 (d, 8.4Hz, 1H, 20-Ha), 4.92 (d, 8.4Hz, 1H, 5-H),
5.59 (d, 7.2Hz, 1H, 2-H), 5.71 (dd, 7.2, 10.8Hz, 1H, 7-H), 5.91 (t, 9.6Hz, 1H, 13-H),

106
6 19 (s, 1H, 10-H), 7.47-7.63 (m, 11H, m,p-Bz, o,m,p-Ph, m,p-NBz), 7.84 (d, 7.2Hz, 2H,
o-NBz), 8.12 (d, 7.2Hz, 2H, o-Bz). 13C NMR 5: 11.1, 13.3, 20.6, 20.7, 21.3, 22.7, 26.1,
32.4, 35.8, 42.9, 47.2, 55.1, 71.9, 74.3, 74.6, 76.2, 79.4, 79.8, 79.9, 83.7, 127.5, 128.6,
128.7, 129.0, 129.1, 129.3, 129.4, 130.3, 130.7, 132.2, 132.7, 132.8, 133.5, 133.7, 133.8,
141.4, 163.4, 163.7, 166.7, 168.5, 169.4, 171.5, 200.6.

CHAPTER 4
SYNTHESIS OF ANALOGUES WITH POTENTIALLY
IMPROVED WATER SOLUBILITY
Introduction
In spite of paclitaxels great promise in treatment of refractory and untreatable
human neoplasms it is afflicted with formulation and systemic administration problems.
These problems stem from its extreme low solubility in water which has been reported as
low as 0.25 pg/ml (Ali et al., 1995). Consequently, special formulations requiring
excipients such as Cremeophore EL have been necessary for intravenous administration.
In the case of paclitaxel the amount of Cremophore EL required to administer the
therapeutic dose (135-200 mg/m2) represents the highest amount ever to be used with any
drug. Exposure to the large amounts of Cremophore EL has produced major
hypersensitivity reactions in patients. It is perceived that such adverse effects are vehicle
related, since it is well documented that Cremophore EL alone causes hypotension and
histamine release in dogs. The high incidences and severity of these side effects to
paclitaxel almost led to the termination of some earlier Phase I clinical trials. However,
prolonged infusions and prophylactic medications with antihistamines and corticosteroids
have avoided the adverse episodes and allowed continuation of clinical use of Cremophore
EL formulation. Still, because of these problems and the additional medications needed,
an equally potent but more water-soluble analogue of paclitaxel is desirable. This area has
107

108
been the focus of much research and various methods have been developed. A variety of
water-soluble analogues have been developed which contain esterase or phosphatase-
cleavable pro-moieties. However, these prodrugs are liable to exhibit unstable efficacy
because of variation in the enzymatic activity amoung patients. Therefore it would be very
advantageous to develop non-prodrugs of paclitaxel with satisfactory stability in vivo, high
water-solubility, and potent antitumor activity. This laboratory has studied two possible
approaches to this type of analogue and they are discussed below.
Synthesis of Analogues Starting from
10-Deacetyl Paclitaxel-7-Xyloside
It was discussed in Chapter 2 that 10-deacetyl paclitaxel-7-xyloside is actually the
most predominate taxane found in the bark of Taxus brevifolia occurring in a yield of
0.1% which is 2.5 times as much as paclitaxel. Other xylosides (10-deacetyl paclitaxel-C-
7-xyloside and 10-deacetyl cephalomannine-7-xyloside) are also found in high yields and
can be converted into 10-deacetyl paclitaxel-7-xyloside by modification of the amide
function. Although these xylosides have not been reported to a great extent in other Taxus
species, this does not mean they are not present since most published isolation procedures
are not ideal for obtaining these xylosides. It has also been reported that these xylosides
inhibit the in vitro disassembly of microtubules from mammalian brain at lower
concentrations than paclitaxel (Lataste et al., 1984). This is shown in Table 4-1. In light of
this fact and the assumption that these xylosides would undoubtedly have greater water-
solubility than paclitaxel, it is interesting that no reports have seriously examined the
possibility of using the xylosides as alternatives to paclitaxel clinically. One reason for this

109
may be the fact that although the xylosides are more active in vitro, they have been
reported to be less active in cell culture toxicity assays (Rao, 1993). Obviously the xylose
unit is either causing a decrease in uptake into the cell or to the site of action or the xylose
unit is serving as a site for metabolism which serves to inactivate the compound.
Table 4-1: ID50 Values of Paclitaxel and Xylosides in Tubuline Assay
Compound
ID50 for disassembly of microtubules, p,M
Paclitaxel
0.5
Paclitaxel-7-Xyloside
0.2
10-Deacetyl Paclitaxel-7-Xyloside
0.3
Cephalomauinine-7-Xyloside
0.25
With this information in hand, the goal was then to devise a scheme in which the
xylose unit would be changed in such a way as to hopefully increase the cell culture
cytotoxicity over the parent xyloside as well as retain an assumed water-solubility
advantage over paclitaxel. As discussed in Chapter 3, the xylose unit in these xylosides can
be easily oxidized with periodate to the dialdehyde. It should be mentioned that this
reaction occurs without interfering with any other part of the molecule including the a-
hydroxy ketone function at C-9 and C-10 of 10-deacetyl paclitaxel-7-xyloside since this
general function type is usually cleaved with periodate. The dialdehyde that is formed
upon oxidation and loss of one carbon actually exist as an equilibrium between 3 different
structures (143, 144, 145) but since it is the dialdehyde (144) which serves as the
electrophile the oxidation product will be referred to as the dialdehyde (Figure 4-1).
Once this dialdehyde is obtained by stirring the xyloside in an aqueous acidic
solution also containing NaI04, it can then be condensed with a variety of nucleophiles.
Among these are carbon nucelophiles such as P-dicarbonyl compounds as well as

110
nitroalkanes The dialdehyde could also undergo reductive amination in the presence of an
amine and a reducing agent. The latter method would result is the formation of
morpholino analogues and is well documented, while the former reactions would result in
the formation of a tetrahydropyran ring. In each case an ionizable group could be
incorporated into the function by using these methods assuming reduction of the nitro
group to an amine following the condensation (Figure 4-1). Indeed each of these methods
were exploited in this work however, because of the unavailability of a large number of (3-
dicarbonyls and nitroalkanes and the apply supply of amines, the reductive amination
procedure received the most attention. It should also be mentioned that only 10-deacetyl
paclitaxel-7-xyloside was used in this work because on the large supply of this compound
on hand.
Concerning condensation with (3-dicarbonyls, only one reaction was attempted and
this was condensation of the dialdehyde (150) with malonic acid (151) (Figure 4-2). This
reaction proceeded smoothly by refluxing the dialdehyde with 1.5 eq. of malonic acid in
pyridine/piperidine. However, the expected diacid was not the product isolated. Instead,
based on NMR and mass spectroscopy it was concluded that decarboxylation and loss of
H20 had taken place thus resulting in the a, (3-unsaturated acid (152) as the product
(Figure 4-2). The stereochemistry at C-2 was not determined.
Condensation with a nitroalkane was also only attempted with one such reagent
and this was nitromethane (154) (Figure 4-3). This reaction proceeded by using a 1 : 1
mixture of CH2CI2 and Et3N as solvent/base and dissolving the dialdehyde (153) along

Ill
H2, cat.
v
Figure 4-1: Synthesis of Ionizable Analogues

112
O
+
O o
X X
ll(X
151
pyridine/piperidine
reflux, 3 hrs.
y
O
Figure 4-2: Condensation with Malonic Acid
with 2.5 eq. of nitromethane and stirring at room temperature for overnight. Although
essentially all the starting material disappeared the reaction was not a clean one being that
at least three products were formed. This result was somewhat expected since the reaction

113
generates three stereocenters thus a variety of isomers are possible. The major product
was isolated and based on the mass spectrum and LlC NMR it was concluded that this was
the desired product with unknown stereochemistry (155). Overlapping peaks in the 'H
NMR spectrum made it impossible to determine cis/trans configuration based on coupling
constants. It is not unreasonable to assume that the isolated major product is the all-trans
equatorial conformer since that would be the thermodynamically favored product.
Aside from having a greater number of nucleophiles to choose from, the reductive
amination procedure also offers the advantage of not generating asymmetric carbons.
Indeed this reaction actually converts two asymmetric carbons into methylene carbons.
The reducing agent chosen for these reactions was sodium cyanoborohydride. This
reagent has the advantage over sodium borohydride since it can be used in mildly acidic
conditions therefore no loss of the side chain occurs as with sodium borohydride. Also this
reagent has been shown to perform ideally in reductive amination reactions.
Initially the goal was to use a variety of amines, both aromatic and aliphatic, as
well as amino acids in these reactions. Attachment of di- and tripeptides was envisioned as
well. Unfortunately aliphatic amines including common amino acids did not condense with
the dialdehyde under the conditions used. Instead, the dialdehyde (156) function was
reduced to the corresponding diol (157) (Figure 4-4), and this occurrence was probably
due to the increased steric hindrance at the sp" a-carbon as opposed to the sp2 a-carbon
of an aromatic amine. Although this problem may have been overcome by varying the
conditions, time did not allow for this investigation.

114
O
+
CH3-NO2
154
CH2CI2, Et3N
T
O
Figure 4-3: Condensation with Nitromethane
When using aromatic amines however this problem was not encountered and a
smooth reaction occurred. The dialdehyde (158) was added with 5 eq. of amine in 2 : 1
CH3OH and acetic acid and an excess of NaCNBH3 was added and stirred for 1.5 hours.

115
These conditions gave clean products which were easily purified by column
chromatography. The following amines were used: p-aminobenzoic acid, p-aminosalicylic
acid, m-aminosalicylic acid, p-nitroaniline, and p-aminobenzenesulfonamide which was
obtained by reducing p-nitrobenzenesulfonamide with Sn and HC1. These amines
generated the corresponding morpholino analogues (160-164) (Figure 4-5).
Attempted Synthesis of Taxane Glycosides
An alternative method of preparing analogues with increased water-solubility was
also studied in this work. This method involved attaching sugar units to the taxane moiety
and thereby forming glycosides. This would certainly increase the water-solubility and
quite possibly have a positive effect on the activity of the drug since it is well known that
carbohydrates play a vital role in molecular recognition. The only taxane glycoside that
had been studied was the naturally occurring xylosides and although they displayed very
good activity in tubuline assays but lost activity in cell culture as mentioned, it was hoped
that attaching different sugars would overcome this problem. Also the possibility existed
that if a di- or trisaccharide could be attached to paclitaxel or one of its close analogues
then possibly it would become orally available. This type of behavior is seen with the
cardiac glycosides in which if the sugars are removed the compound in no longer orally
active. As of this writing two papers have been published in which a taxane has been
linked with a carbohydrate unit to increase the water-solubility (Paradis & Page, 1998;
Takashi et ah, 1998).

116
O
+
aliphatic amine
NaCNBH3
2 : 1 CH3OH : AcOH
O
Figure 4-4: Dialdehyde Reduction to Diol
Three methods of glycosylation were to be studied in this work are these are; 1) the
Koenigs-Knorr method, 2) the trichloroacetimidate method, and 3) the sulphoxide
method. These will now be briefly discussed.
The Koenigs-Knorr method was first introduced in 1901 and has since been a
mainstay in glycosylation chemistry (Toshima & Tatsuta, 1993). In its classical form it

117
NaCNBH3
2 : 1 CH3OH : AcOH
Y room temp., 1.5 hrs
O
160 Ri = COOH, R2 = H, R3 = H
161 Ri COOH, R2 = OH, R3 = H
162 Rx = H, R2 = COOH, R3 = OH
163 Rj = N02, R2 = H, R3 = H
164 Rj = S02NH2, R2 H, R3 = H
Figure 4-5: Reductive Aminations

118
involves using a suitable protected glycosyl bromide or chloride as the glycosyl donor. The
reaction is facilitated by using various heavy metal salts as activating agents. These metals
technically do not function as catalyst since an equivalent amount is needed. Some of the
more common heavy metal salts include; AgOTf, Ag20, AgC104, AgNOs, Hg(CN)2, and
HgBr2. The silver salts are the strongest activators with AgOTf being the strongest. In
some cases an acid scavenger may also be used and examples of these include HgO,
CdC03, and s-collidine.
The trichloroacetimidate-mediated glycosylation was announced in 1980 as an
alternative method to the classical Koenigs-Knorr procedure and now appears to be one of
the most ideal glycosylation protocol (Toshima & Tatsuta, 1993). This method involves
using a suitably protected anomeric trichloroacetimidate as the glycosyl donor. This donor
is prepared by condensing a protected 1-hydroxy sugar with trichloroacetonitrile in the
presence of base. Depending on which base is used, one can prepare the a or (3 epimer
using kinetic or thermodynamic control. This is not possible with halides. This reaction is
smoothly promoted by the catalytic use of BF3-Et20, TMSOTf, CCI3CHO, PTSA, and
ZnBr2.
The sulphoxide method is the newest of the three methods being first reported in
the literature in 1989 (Kahne et al., 1989). This method involves the use of a suitably
protected glycosyl sulphoxide as the glycosyl donor (usually a phenyl sulphoxide), an
equimolar amount of triflic anhydride as the activator, and 2, 6-di-tert-butyl-4-
methylpyridine as an acid scavenger. The stereochemistry of the resulting glycoside can be
controlled by varying the reaction solvent. This method has been shown to be very

119
applicable to unreactive and hindered nucleophiles and has been the basis for developing a
combinatorial process involving oligosaccharides.
Acetyl protected glycosyl donors for each of these methods were prepared using
D-glucose as the sugar as shown in Figure 4-6. Initially D-glucose (165) is acetylated with
acetic anhydride and pyridine to give the pentaacetate (166) in near quantitative yields.
Next, bromine was introduced at the anomeric carbon by treating the pentaacetate with
30% HBr in acetic acid which gave yields around 80%. This product (167) was used as
the Koenigs-Knorr donor and also as a starting material for the other two donors. To
synthesize the trichloroacetimidate donor the bromide was hydrolyzed in aqueous
acetonitrile with Hg(CN)2 added. This gave a quantitative yield. The a-
trichloroacetimidate (169) was produced from this by adding trichloroacetonitrile and
using K2CO3 as base in CH2CI2 in a yield of about 75%. The sulphoxide donor was
prepared by first by making the phenylthioglycoside (170) using the bromide, thiophenol,
and KOH as base. This reaction gave near quantitative yields. The sulfide was then
oxidized to the sulphoxide (171) with mCPBA to give near quantitative yields (Figure 4-
6).
Once these donors were in hand a series of test glycosylations were performed
using (3-sitosterol (172) and benzyl alcohol (173) as aglycones to determine the best
conditions. Unfortunately the sulphoxide method was never attempted due to a lack of the
triflic anhydride activator. However both the Koenigs-Knorr and trichloroacetimidate
methods were attempted and meet with success. In terms of the Koenigs-Knorr method,
(3-glycosides were formed with both of these aglycones in moderate yields with the benzyl

120
Ac20

pyridine
Hg(CN)2
<
aq. CH3CN
167 OAc
CI3CCN, K2CO3,
CH2C12
room temp., 72 hrs.
KOH/
thiophenol
V
Figure 4-6: Synthesis of Glycosyl Donors

121
alcohol reaction giving the best results. These reactions were conducted at room
temperature in CH2C12 (Figure 4-7). The (3 orientation was determined based on the
anomeric carbon and proton chemical shifts as well as comparison to the acetylated
naturally occurring P-sitosterol-p-glucoside. Of the activators that were available,
Hg(CN)2, AgN03, and ZnCl2 were the best. None of the stronger silver Lewis acids were
available. The trichloroacetimidate method was also attempted with both of these
substrates using PTSA and BF3-Et20 as the activators and CH2C12 as solvent. The
reactions were performed at room temperature (Figure 4-8). Although a substantial
amount of hydrolysis of the donor did occur, it was obvious that this method was superior
giving greater yields in a shorter amount of time. The products formed also contained the
P orientation.
Since these results were encouraging, similar reactions were attempted with
paclitaxel. Unfortunately none of these reactions were successful. As mentioned earlier
paclitaxel and related taxanes are quite unstable to acids, both mineral and Lewis. Under
these conditions a variety of rearrangements have been documented. When the weaker
Lewis acids such as Hg(CN)2 were used in conjunction with the Koenigs-Knorr method
no reaction took place at all. This was probably due to the steric hindrance involved at the
C-2 and C-7 hydroxyls of paclitaxel. Therefore faced with this problem the use of
stronger Lewis acids would be required; however, when this was attempted with AgNO:,
and ZnCl2 rearrangements did occur. In most cases the rearranged product was not
identified but in the case of AgNCF the product was identified as the rearranged taxane
183 when paclitaxel-2-acetate (182) was used as the aglycone (Figure 4-9). Therefore, in

122
Hg(CN)2, Z11CI2, orAgN03
CH2CI2, room temp.
Figure 4-7: Koenigs-Knorr Glycosylation
order to take advantage of less hindered hydroxyls, 10-deacetyl paclitaxel-7-xyloside was
used as the substrate hoping the reaction would occur on one of the xylose hydroxyls.
However as before no reaction took place with the weaker Lewis acids. In an attempt to

123
CC13
PTSA or BF3-Et20
v CH2CI2, room temp.
Figure 4-8: Trichloroacetimidate Glycosylation
produce a primary hydroxyl which should be easily glycosylated, 10-deacetyl paclitaxel-7-
xyloside was oxidized to the dialdehyde as before and this dialdehyde was subsequently
reduced with NaCNBH3 to the diol (Figure 4-4). Unfortunately when this substrate was

124
O
acetobromoglucose
AgNC>3
CH2CI2 room temp.
Y
O
Figure 4-9: Rearrangement of 2'-AcetyI Paclitaxel
subjected to the Koenigs-Knorr reaction the main product was loss of the diol function
due to hydrolysis to give 10-deacetyl paclitaxel.

125
A few reactions of the trichloroacetimidate type were also investigated with
paclitaxel and various analogues but no favorable reactions occurred. Unfortunately due to
time constraints and other factors (death of professor) this area of research was ceased.
The author does believe that if more controlled conditions were used (dry solvents,
nitrogen atmosphere, low temperatures, less amounts of Lewis acids, etc.) these reactions
could be successful. Also the sulphoxide method may be better than the two that were
investigated and deserves attention that the author could not give it.
L1210 Cytotoxicity of Analogues
The LI 210 assay is commonly used to test compounds for their cytotoxicity and is
a screen for compounds with anticancer activity. These murine leukemia cells have a very
rapid doubling time of about 12 hours. Thus, they provide a convenient and relatively
reliable means for the determination of the cytotoxicity of many compounds (Thayer et al.,
1971).
The following compounds were chosen to be tested on the L1210 assay: paclitaxel,
10-deacetyl paclitaxel-7-xyloside, malonic acid condensation product 152, nitromethane
condensation product 155, reductive amination products 160-164, and the reduced diol
157. The ionizable compounds were tested in their neutral form. Table 4-2 shows the
results of the assay.
As Table 4-2 shows, there was a wide range of activity in this series of
compounds. Although none of the compounds tested is nearly as active as paclitaxel, 4 of
these showed greater activity than 10-deacetyl paclitaxel-7-xyloside. Of the reductive

126
amination products (160-164) the only compounds which is less active than the xyloside is
the m-aminosalicylic acid product (162) which has a substituent ortho to the amino group.
The p-aminosalicylic acid product (161) however is much more active, thus indicating that
ortho substituents are not well tolerated. However more work needs to be performed to
verify this.
Table 4-2: L1210 Cytotoxicity of Paclitaxel and Analogues
Compd.
ICso (ppm)
IC50 (pM)
Compd./Paclitaxel
Paclitaxel
0.0089
0.01
s¡ 5|< >i<
10-Deacetyl
7-Xyloside
Paclitaxel- 0.37
0.39
39
152
3.31
3.47
347
155
0.89
0.92
92
160
0.048
0.047
4.7
161
0.095
0.092
9.2
162
2.00
1.94
194
163
0.071
0.071
7.1
164
0.13
0.12
12
157
0.93
1.02
102
Experimental
All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of
the reaction. All NMR spectra were recorded using either a Varan VXR-300 or a Varan
Gemini-300 spectrophotometer using CDC13 as solvent. Infrared spectra were obtained
using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolent spectra were
obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were
recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a
Fisher melting point apparatus. Column chromatography was used in conjunction with
100-200 mesh silica gel.

127
Oxidation of Xyloside with Periodate
10-Deacetyl paclitaxel -7-(3-xyloside (1.0 g) was dissolved in 10 ml of 1 : 1 THF
and water and 2 ml of 1 N H2S04 was added. This was followed by 0.71 g of NaI04 and
the mixture was stirred overnight at room temperature. The mixture was diluted with
water and the solid which precipitated (920 mg) was filtered and dried.
Condensation of Dialdehyde with Malonic Acid
A total of 200 mg of dialdehyde was dissolved in 3 ml of pyridine and 5 drops of
piperidine was also added. To this mixture was added 20 mg of malonic acid and the
mixture was refluxed for 3 hours. At this point the mixture was diluted with CH2CI2 and
washed with 0.1 N HC1 3 times and with water twice. The organic layer was then dried
with Na2S04 and evaporated to a residue. The residue was separated on a silica column
using 30-50% ethyl acetate in ligroin with a few drops of acetic acid as the mobile phase.
A total of 123 mg of product was crystallized from diethyl ether, ligroin, and acetone.
Yellowish white crystalline powder, mp 172-174 C (dec.), UV ^max (CH3OH): 225 nm,
FABMS m/z: 954 (M + 1), 795, 669, 651, 105. *H NMR 5 (some DMSO): 1.09 (s, 3H,
17-H), 1.18 (s, 3H, 16-H), 1.75 (s, 3H, 19-H), 1.85 (s, 3H, 18-H), 1.87 (m, 1H, 14-H(3),
2.04 (m, 1H, 6-HP), 2.30 (m, 1H, 4-Ha), 2.39 (s, 3H, 4-OAc), 2.78 (m, 1H, 6-Ha), 3.89
(d, 6.3Hz, 1H, 3-H), 4.05 (s, 1H, 2-H), 4.14-4.30 (m, 5H, 7-, 20a-, 20(3-, 5ax-, 5eq-
H), 4.73 (s, 1H, 1-H), 4.83 (d, 3.0Hz, 1H, 2-H), 4.94 (d, 8.7Hz, 1H, 5-H), 5.25 (s, 1H,
10-H), 5.65 (d, 6.9Hz, 1H, 2-H), 5.80 (dd, 2.1, 8.7Hz, 1H, 3-H), 6.21 (t, 8.1Hz, 1H, 13-
H), 7.04 (s, 1H, 4-H), 7.31-7.52 (m, 10H, m-Bz, m,p-NBz, o,m,p-Ph), 7.59 (t, 7.5Hz,
1H, p-Bz), 7.60 (d, 9.3Hz, 1H, N-H), 7.77 (d, 7.2Hz, 2H, o-NBz), 8.10 (d, 7.5Hz, 2H, o-

128
Bz). 13C NMR 5 (some DMSO): 10.6, 14.1, 20.6, 22.5, 26.5, 35.1, 35.6, 43.0, 46.5,
55.2, 56.9, 60.7, 63.2, 72.0, 73.3, 74.4, 74.7, 76.5, 78.4, 80.8, 80.9, 84.0, 101.9, 127.0,
127.1, 127.7, 128.0, 128.2, 128.2, 128.5, 128.6, 128.7, 129.3, 130.1, 131.8, 133.5, 133.6,
135.9, 138.1, 138.3, 139.6, 166.6, 167.8, 168.3, 170.4, 172.7, 209.4.
Condensation of Dialdehyde with Nitromethane
A total of 200 mg of dialdehyde and 35 mg of nitromethane was dissolved in a 1 :
1 mixture of benzene and triethylamine and this mixture was stirred for 4 days at room
temperature. At this point the reaction mixture was evaporated to a residue uder reduced
pressure and this residue was ran on a silica column using 10-30% acetone in benzene as
the mobile phase. A total of 45 mg of amorphous product was obtained as the major
product. White amorphous powder, UV A,max (CH3OH): 228 nm, FABMS m/z: 974 (M +
2), 956, 913, 689, 670, 603, 286, 154, 136, 105, 81. 13C NMR 5 (acetonitrile): 11.4,
14.5, 21.4, 23.1, 27.0, 35.6, 36.8, 43.9, 47.3, 56.9, 57.6, 62.8, 68.3, 70.3, 72.2, 74.6,
75.6, 75.7, 77.0, 79.0, 81.3, 81.8, 84.4, 92.8, 99.9, 128.2, 128.3, 128.7, 129.4, 129.5,
129.6, 130.9, 132.5, 134.4, 135.2, 137.0, 139.0, 139.9, 166.8, 168.1, 171.5, 173.6, 212.0.
Reduction of Dialdehyde to the Diol
A total of 100 mg of dialdehyde was dissolved in a 1 : 1 mixture of methanol and
acetic acid and 100 mg of NaCNBTh was added and this mixture was stirred at room
temperature for 1 hour. At this point the mixture was diluted with ethyl acetate and water
and partitioned. The organic layer was washed twice with saturated NaHCOj solution and
twice with water. The organic layer was then dried with Na2S04 and evaporated under
reduced pressure. The solid residue was purified by running a quick silica column with

129
using 40-50% acetone in benzene as the mobile phase. Evaporation of the appropriate
fractions yielded 182 mg of amorphous solid as the product. UV Aax (CH3OH): 229 nm,
FABMS m/z: 916 (M+l), 661, 653, 551, 509, 176, 154, 136, 105, 81. *HNMR5: 1.07
(s, 3H, 17-H), 1.16 (s, 3H, 16-H), 1.76 (s, 3H, 19-H), 1.82 (s, 3H, 18-H), 1.82 (m, 1H,
14-Hp), 1.96 (m, 1H, 6-Hp), 2.24 (m, 1H, 14-Ha), 2.35 (s, 3H, 4-OAc), 2.69 (m, 1H, 6-
Ha), 3.38-3 55 (m, 6H, 2-, 4-, 5-H), 3.84 (d, 6.3Hz, 1H, 3-H), 4.13 (m, 1H, 7-H),
4.14 (d, 8.1Hz, 1H, 20-Hp), 4.28 (d, 8.1Hz, 1H, 20-Ha), 4.56 (br s, 1H, 1-H), 4.80 (br
s, 1H, 2-H), 4.91 (d, 9.0Hz, 1H, 5-H), 5.23 (s, 1H, 10-H), 5.63 (d, 6.6Hz, 1H, 2-H),
5.74 (d, 6.9Hz, 1H, 3-H), 6.17 (t, 8.4Hz, 1H, 13-H), 7.30-7.52 (m, 11H, m-OBz, m,p-
NBz, o,m,p-Ph, N-H), 7.61 (t, 7.2Hz, 1H, p-OBz), 7.73 (d, 7.5Hz, 2H, o-NBz), 8.10 (d,
7.2Hz, 2H, o-OBz). 13C NMR 6: 10.7, 14.2, 20.7, 22.5, 26.6, 35.5, 35.7, 43.1, 46.6,
55.2, 57.0, 61.8, 62.3, 67.7, 72.1, 73.2, 74.6, 74.7, 76.5, 78.6, 78.8, 80.9, 84.0, 103.9,
127.0, 127.1, 128.2, 128.3, 128.6, 128.7, 128.9, 129.2, 130.1, 131.9, 133.7, 136.0, 138.0,
138.2, 166.8, 167.5, 170.7, 172.9, 210.3.
General Procedure for Reductive Animations
A total of 500 mg of the dialdehyde and 500 mg of the amine were dissolved in 6
ml of 2 : 1 methanol and acetic acid and excess (200 mg) of NaCNBH3 was added. The
mixture was stirred at room temperature for 1.5 2.0 hours and then diluted with water
and extracted three times with ethyl acetate. The organic layer was then washed with
NaHC03 twice and with water twice. The organic layer was then dried with Na2S4 and
evaporated. The product was separated by putting the residue on a silica column and
eluted with 20-40% acetone in benzene.

130
p-Aminobenzoic acid product (160)
Grayish white crystalline powder, 478 mg, mp 198-200 C, UV Aax (CH3OH):
228, 300 nm, FABMS m/z: 1017 (M+l), 794, 754, 715, 714, 610, 206, 105. *H NMR 5
(some DMSO): 1.11 (s, 3H, 17-H), 1.21 (s, 3H, 16-H), 1.80 (s, 3H, 19-H), 1.81 (m, 1H,
14-Hp), 1.89 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.38 (m, 1H, 14-Ha), 2.43 (s, 3H, 4-
OAc), 2.76 (m, 1H, 6-Ha), 3.07 (dd, 4.5, 12.0Hz, 1H, 2-H), 3.27 (m, 2H, 2-, 4-H),
3.56 (br s, 1H, 4-H), 3.69 (m, 1H, 5-H), 3.90 (d, 6.6Hz, 1H, 3-H), 4.02 (m, 1H, 5-
H), 4.18-4.29 (m, 3H, 7-, 20a-, 20P-H), 4.56 (br s, 1H, 1-H), 4.70 (br s, 1H, 2-H),
4.93 (d, 9.3Hz, 1H, 5-H), 5.14 (s, 1H, 10-H), 5.65 (d, 6.6Hz, 1H, 2-H), 5.76 (d, 8.4Hz,
1H, 3-H), 6.22 (t, 8.1Hz, 1H, 13-H), 6.85 (d, 8.1Hz, 2H, o-Ar), 7.29-7.61 (m, 11H, m,p-
Bz, m,p-NBz, o,m,p-Ph), 7.86 (d, 7.2Hz, 2H, o-NBz), 7.93 (d, 8.4Hz, 2H, m-Ar), 8.12
(d, 7.2Hz, 2H, o-Bz), 8.13 (d, 8.4Hz, 1H, N-H). 13C NMR 5 (some DMSO): 10.3, 13.8,
20.5, 22.2, 26.2, 35.2, 35.3, 42.7, 45.6, 46.1, 50.4, 55.2, 56.5, 61.0, 70.9, 73.5, 74.0,
74.5, 75.9, 77.6, 80.0, 80.2, 83.7, 98.3, 113.2, 120.4, 126.5, 127.0, 127.2, 127.9, 128.1,
128.1, 129.4, 129.6, 130.9, 131.1, 132.9, 133.9, 135.7, 137.9, 138.5, 153.0, 165.9, 166.7,
167.9, 169.9, 172.2, 209.6
p-Salicylic acid product (161)
Grayish brown crystalline powder, 378 mg, mp 182-184 C, UV \max (CH3OH):
230, 311 nm, FABMS m/z: 1034 (M + 2), 1033 (M + 1), 748, 730, 222, 204, 105. *H
NMR 5 (some DMSO): 1.11 (s, 3H, 17-H), 1.20 (s, 3H, 16-H), 1.79 (s, 3H, 19-H), 1.80
(m, 1H, 14-Hp), 2.02 (m, 1H, 6-HP), 2.25 (m, 1H, 14-Ha), 2.43 (s, 3H, 4-OAc), 2.74 (m,
1H, 6-Ha), 3.08 (dd, 4.8, 12.6Hz, 1H, 2-H), 3.26 (m, 3H, 2-, 4-, 4-H), 3.69 (m,

131
1H, 5-H), 3.90 (d, 7.2Hz, 1H, 3-H), 4.01 (m, 1H, 5-H), 4.16-4.29 (m, 3H, 7-, 20a-,
20J3-H), 4.54 (br s, 1H, 1-H), 4.70 (d, 3.0Hz, 1H, 2-H), 4.92 (d, 8.7Hz, 1H, 5-H), 5.14
(s, 1H, 10-H), 5.65 (d, 6.9Hz, 1H, 2-H), 5.76 (dd, 3.0, 8.7Hz, 1H, 6-Ar), 6.21 (t, 8.4Hz,
1H, 13-H), 6.29 (d, 2.1Hz, 1H, 2-Ar), 6.38 (dd, 2.1, 9.0Hz, 1H, 3-H), 7.29-7.55 (m,
10H, m-OBz, m,p-NBz, o,m,p-Ph), 7.61 (t, 7.2Hz, 1H, p-OBz), 7.71 (d, 9.0Hz, 1H, N-
H), 7.86 (d, 7.2Hz, 2H, o-NBz), 8.12 (d, 7.5Hz, 2H, o-OBz), 8.17 (d, 9.0Hz, 1H, 5-Ar).
13C NMR 6 (some DMSO): 10.3, 13.8, 20.5, 22.1, 26.1, 35.2, 35.3, 42.7, 45.6, 45.7,
49.9, 55.1, 56.5, 60.8, 70.9, 73.5, 74.0, 74.4, 75.9, 77.2, 80.0, 80.2, 83.6, 98.1, 100.1,
103.1, 105.5, 126.5, 127.0, 127.2, 127.9, 128.0, 128.1, 129.3, 129.6, 131.1, 131.2, 132.9,
133.8, 135.6, 137.8, 138.4, 155.1, 163.0, 165.8, 166.8, 169.9, 171.9, 172.2, 209.5.
m-Salicylic acid product (162)
Reddish brown crystalline powder, 392 mg, mp 173-175 C (dec.), UV 7,max
(CH3OH): 224, 343 nm, FABMS m/z: 1034 (M + 2), 1032 (M), 222, 204, 105, *H NMR
5: 1.10 (s, 3H, 17-H), 1.18 (s, 3H, 16-H), 1.82 (s, 3H, 19-H), 1.82 (m, 1H, 14-Hp), 1.83
(s, 3H, 18-H), 2.07 (m, 1H, 6-Hp), 2.31 (m, 1H, 14-Ha), 2.36 (s, 3H, 4-OAc), 2.72 (m,
1H, 2-H), 2.79 (m, 1H, 4-H), 2.86 (m, 1H, 6-Ha), 2.97 (m, 1H, 2-H), 3.00 (m, 1H,
4-H), 3.72 (m, 1H, 5-H), 3.89 (d, 5.7Hz, 1H, 3-H), 3.99 (m, 1H, 5-H), 4.20 (m, 1H,
7-H), 4.20 (d, 7.5Hz, 1H, 20-Hp), 4.29 (d, 7.5Hz, 1H, 20-Ha), 4.61 (br s, 1H, 1-H),
4.79 (br s, 1H, 2-H), 4.93 (d, 8.7Hz, 1H, 5-H), 5.28 (s, 1H, 10-H), 5.65 (d, 6.0Hz, 1H,
2-H), 5.78 (d, 9.0Hz, 1H, 3-H), 6.18 (br s, 1H, 13-H), 6.80 (d, 8.7Hz, 1H, o-Ar), 7.08
(d, 8.4Hz, 1H, m-Ar), 7.21 (br s, 1H, p-Ar), 7.37 (m, 1H, N-H), 7.33-7.49 (m, 10H, m-
Bz, m,p-NBz, o,m,p-Ph), 7.58 (t, 7.2Hz, 1H, p-Bz), 7.74 (d, 7.5Hz, 2H, o-NBz), 8.08 (d,

132
7.5Hz, o-Bz). 13C NMR 5: 10.8, 14.2, 20.5, 22.5, 26.7, 35.6, 35.7, 43.1, 46.5, 49.9, 53.6,
55.2, 57.1, 62.7, 72.3, 73.2, 74.3, 74.6, 77.2, 78.5, 80.7, 81.0, 84.2, 99.4, 111.8, 117.2,
118.2, 126.9, 127.0, 127.1, 128.3, 128.6, 128.7, 128.9, 129.1, 130.1, 132.0, 133.5, 133.7,
135.9, 137.8, 138.5, 142.7, 156.7, 166.8, 167.6, 170.6, 172.2, 172.6, 209.9.
p-Nitroaniline product (163)
Yellow crystalline powder, 423 mg, mp 177-179 C, UV Aax (CH3OH): 230, 380
nm, FABMSm/z: 1003 (M), 987, 701, 638, 579, 207, 105. lU NMR 5: 1.10 (s, 3H, 17-
H), 1.20 (s, 3H, 16-H), 1.77 (s, 3H, 19-H), 1.80 (s, 3H, 18-H), 1.80 (m, 1H, 14-H0), 2.02
(m, 1H, 6-H(3), 2.29 (m, 1H, 14-Ha), 2.38 (s, 3H, 4-OAc), 2.74 (m, 1H, 6-Ha), 3.28 (m,
2H, 2-H), 3.34 (m, 1H, 4-H), 3.40 (m, 1H, 4-H), 3.67 (m, 1H, 5-H), 3.89 (d,
6.6Hz, 1H, 3-H), 4.01 (m, 1H, 5-H), 4.18 (m, 1H, 7-H), 4.19 (d, 9.0Hz, 1H, 20-Hp),
4.29 (d, 8.4Hz, 1H, 20-Ha), 4.58 (br s, 1H, 1-H), 4.78 (d, 2.7Hz, 1H, 2-H), 4.90 (d,
8.4Hz, 1H, 5-H), 5.13 (s, 1H, 10-H), 5.65 (d, 6.9Hz, 1H, 2-H), 5.77 (dd, 2.7, 9.0Hz, 1H,
3-H), 6.19 (t, 8.4Hz, 1H, 13-H), 6.78 (d, 9.3Hz, 2H, o-Ar), 7.21 (d, 8.7Hz, 1H, N-H),
7.33-7.51 (m, 10H, m-Bz, m,p-NBz, o,m,p-Ph), 7.60 (t, 7.5Hz, 1H, p-Bz), 7.75 (d,
7.2Hz, 2H, o-NBz), 8.10 (d, 7.5Hz, 2H, o-Bz), 8.14 (d, 9.3Hz, 2H, m-Ar). 13C NMR 6:
10.7, 14.3, 20.6, 22.5, 26.7, 35.7, 35.9, 43.0, 46.3, 50.2, 55.2, 57.2, 60.7, 72.3, 73.3,
74.4, 74.6, 76.5, 78.6, 80.6, 80.8, 84.1, 98.2, 112.7, 125.9, 126.9, 127.0, 127.0, 127.1,
128.3, 128.6, 128.7, 128.9, 129.2, 130.1, 131.9, 133.6, 136.2, 137.9, 138.2, 138.8, 154.3,
166.8, 167.2, 170.6, 172.5, 210.2.

133
p-Aminobenzenesulfonamide product (164)
Light brown crystalline powder, 418 mg, mp 184-186 C, UV /W (CH3OH):
221, 273 nm, FABMS m/z: 1053 (M + 2), 1020, 767, 734, 241, 105. *H NMR 5 (some
DMSO): 1.09 (s, 3H, 17-H), 1.17 (s, 3H, 16-H), 1.75 (s, 3H, 19-H), 1.76 (m, 1H, 14-
Hp), 1.87 (s, 3H, 18-H), 1.98 (m, 1H, 6-HP), 2.18 (m, 1H, 14-Ha), 2.40 (s, 3H, 4-Oac),
2.72 (m, 1H, 6-Ha), 3.10 (m, 2H, 2-, 4-H), 3.25 (m, 2H, 2-, 4-H), 3.67 (m, 1H,
5-H), 3.87 (br s, 1H, 3-H), 4.02 (m, 1H, 5-H), 4.05 (m, 1H, 7-H), 4.08-4.22 (m, 2H,
20a-, 20P-H), 4.57 (br s, 1H, 1-H), 4.66 (br s, 1H, 2-H), 4.92 (br s, 1H, 5-H), 5.13 (s,
1H, 10-H), 5.61 (br s, 1H, 2-H), 5.73 (d, 7.5Hz, 1H, 3-H), 6.17 (br s, 1H, 13-H), 6.89
(d, 8.7Hz, 2H, o-Ar), 7.28-7.61 (m, 11H, m,p-OBz, m,p-NBz, o,m,p-Ph), 7.75 (d, 8.7Hz,
2H, m-Ar), 7.87 (d, 6.9Hz, o-NBz), 8.10 (br s, 2H, o-OBz), 8.31 (d, 8.7Hz, 1H, N-H).
13C NMR 5 (some DMSO): 9.9, 13.3, 20.1, 21.8, 25.8, 34.7, 34.8, 42.3, 45.1, 45.8, 49.9,
55.0, 56.0, 60.3, 70.2, 73.2, 73.6, 74.1, 75.4, 76.5, 79.7, 83.2, 97.7, 113.1, 126.2, 126.6,
126.7, 126.8, 127.4, 127.6, 127.7, 129.2, 130.6, 132.1, 132.4, 133.5, 135.4, 137.1, 138.3,
151.8, 165.2, 166.2, 169.4, 171.8, 209.0.
Synthesis of Glucose Pentaacetate
A total of 1.0 g of d-glucose was added to a mixture of 5.0 g (4.6 ml) of acetic
anhydride and 6.5 g (6.65 ml) of pyridine at 0 C and this was stirred overnight while
allowing the mixture to warm to room temperature. At this point about 20 ml of ice water
was added and the product crystallized out of solution. These crystals were filtered,
washed with water, and dried under reduced pressure. A total of 2.12 g of product was
obtained. White crystalline powder, product existed as a mixture of a and P anomers as

134
determined by 'H and ljC NMR spectroscopy. Selected 13C NMR signals 5: 61.4, 67.7,
67.8, 69.1, 69.7, 70.2, 72.6, 72.7, 89.0, 91.6.
Synthesis of la-Bromo-Tetraacetyl Glucose
A total of 3.0 g of glucose pentaacetate was dissolved in 6 ml of CH2CI2 and then
8 ml of 30% HBr in acetic acid was added and this was stirred at room temperature for 2
hours. At that point the mixture was diluted with 30 ml of CH2CI2 and 75 ml of ice water
and partitioned. The organic layer was separated and the aqueous layer was partitioned
again with CH2CI2. The combined organic layers were washed with ice cold NaHCO;,
solution three times and then dried with Na2S04. The solvent was removed under reduced
pressure and the residue was taken up in diethyl ether and ligroin and put in the freezer for
crystallization. After overnight crystals had formed and they were filtered, washed with
ligroin, and dried under reduced pressure. The yield was 2.20 g and this material was kept
in the freezer to avoid decomposition.
Synthesis of 1-Hydroxy-Tetraacetyl Glucose
A total of 2.0 g of acetobromoglucose was dissolved in 5 ml of acetonitrile and 1
ml of water was added as excess Hg(CN)2 and the mixture was stirred at room
temperature for 30 minutes at which time the reaction was complete. After filtering off the
Hg(CN)2 the acetonitrile was removed under reduced pressure and the residue was diluted
with water and CH2C12 and partitioned. The organic layer was separated and the water
layer was partitioned again with CH2CI2. The combined organic layers were then washed
with water, dried with Na2S04, and evaporated to yield 1.7 g of a syrup. Clear, colorless
syrup, product existed as a mixture of a and (3 anomers, but the a anomer is predominate.

135
Selected a anomer L,C NMR signals 5: 61.9, 67.0, 68.4, 69.8, 71.1, 90.0. Selected p
anomer 1jC NMR signals 8: 60.4, 68.4, 71.9, 72.2, 73.0, 95.4.
Synthesis of la-Trichloroacetimidate-Tetraacetyl Glucose
A total of 2.2 g of 1-OH-tetraacetyl glucose, 4.14 g (5 eq) of trichloroacetonitrile,
and 4.4 g (1.1 eq) of K2C03 was added to 4 ml of CH2C12 and stirred at room temperature
for 3 days. The mixture was then diluted with water and CH2C12 and partitioned. The
organic layer was separated and the water layer was partitioned again with CH2C12. The
combined organic layers were washed with water, dried with Na2S04, and evaporated to a
syrup. The product was sufficiently pure for further reactions and the yield was 1.96 g.
Clear colorless syrup, 13C NMR 8: 20.5, 20.6, 20.6, 20.7, 61.9, 67.1, 68.5, 69.9, 71.1,
90.0, 163.8, 169.7, 170.2, 170.3, 170.9.
Synthesis of Tetraacetyl Phenyl Thioglucoside
A total of 500 mg of acetobromoglucose, 200 mg (1.5 eq) of thiophenol, and 1.25
eq of KOH were mixed in methanol with the acetobromoglucose added last. The mixture
was stirred at room temperature for 20 minutes at which time TLC showed that the
reaction was complete. The product crystallized upon ceasing the stirring and the crystals
were filtered, washed with aqueous methanol, and dried under reduced pressure. The yield
was 410 mg White crystalline powder, mp 'H NMR 8: 1.99 (s, 3H, OAc), 2.02 (s, 3H,
OAc), 2.08 (s, 3H, OAc), 2.09 (s, 3H, OAc), 3.73 (m, 1H, 5-H), 4.20 (m, 2H, 6-H), 4.72
(d, 6.6Hz, 1H, 1-H), 4.98 (t, 10.2Hz, 1H, 2-H), 5.05 (t, 9.9Hz, 1H, 4-H), 5.23 (t, 9.3Hz,
1H, 3-H), 7.32 (m, 3H, o,p-Ar), 7.50 (m, 2H, rn-Ar). 13C NMR 8: 20.5, 20.7, 62.1, 68.2,
70.0, 74.0, 75.8, 85.7, 128.4, 128.9, 131.6, 133.1, 169.2, 169.3, 170.1, 170.5.

136
Synthesis of Tetraacetyl Glucose, Phenyl Sulfoxide
A total of 250 mg of the phenyl thioglucoside was dissolved in 2 ml of CH2CI2 at
0 C and an equivalent of mCPBA was added and the mixture was stirred at 0 C for 1
hour. The mixture was then diluted with 0.1 N NaOH and CH2C12 and partioned. The
organic layer was washed with water, dried with Na2S04, and evaporated at which time
the product crystallized. After drying under reduced pressure the yield was 226 mg.
Yellowish white crystalline powder. The product existed as a almost equal mixture of a
and (3 anomers with anomeric carbon signals at 89.8 and 92.2 ppm.
Preparation of Tetraacetyl Benzyl P-Glucoside by the Koenigs-Knorr Method
A total of 300 mg of acetobromoglucose and 80 mg of benzyl alcohol was
dissolved in 2 ml of CH2C12 and excess ZnCl2 was added (100 mg). The reaction was
stirred at room temperature for 18 hours, the solid was filtered off and the filtrate was
diluted with water and CH2C12 and partitioned. The organic layer was washed with water,
dried with Na2S04, evaporated, and the residue was ran on a silica column using 30-40%
ethyl acetate in ligroin. A total of 216 mg of crystalline product was upon fraction
evaporation. White crystalline powder, 'HNMR8: 2.00 (s, 3H, OAc), 2.01 (s, 3H, OAc),
2.02 (s, 3H, OAc), 2.11 (s, 3H, OAc), 3.68 (m, 1H, 5-H), 4.18 (dd, 2.1, 12.0Hz, 1H, 6-
H), 4.28 (4.8, 12.0Hz, 1H, 6-H), 4.55 (d, 7.5Hz, 1H, 1-H), 4.63 (d, 12.3Hz, 1H,
OCH2Ar), 4 90 (d, 12.3Hz, 1H, OCH2Ar), 5.04-5.21 (m, 3H, 2-, 3-, 4-H). 13C NMR 5:
20.5, 20.6, 61.9, 68.4, 70.7, 71.2, 71.8, 72.8, 99.2, 127.7, 128.0, 128.4, 136.6, 169.2,
169.3, 170.2, 170.6.

137
Preparation of Tetraacetyl Benzyl (3-Glucoside by the Trichloroacetimidate Method
A total of 100 mg of tetraacetyl glucosyl trichloroacetimidate and 28 mg of benzyl
alcohol was dissolved in 2 ml of CH2CI2 and 36 mg of PTSA was added. The reaction
mixture was stirred at room temperature for 1 hour and the mixture was diluted with
water and CH2C12 and partitioned. The organic layer was washed with water, dried with
Na2S04, and evaporated. The residue was separated on a silica column using 30-40%
ethyl acetate in ligroin as the solvent. A total of 56 mg of product crystallized from the
evaporating fractions. White crystalline powder, and k'C NMR spectra chemical shifts
were identical with those reported above.
Preparation of Tetraacetyl P-Sitosterol P-GIucoside by the Koenigs-Knorr Method
A total of 300 mg of acetobromoglucose and 300 mg of p-sitosterol was dissolved
in 3 ml of CH2C12 and excess ZnCl2 was added (100 mg). The reaction was stirred at room
temperature for 18 hours, the solid was filtered off and the filtrate was diluted with water
and CH2C12 and partioned. The organic layer was washed with water, dried with Na2S04,
evaporated, and the residue was ran on a silica column using 30-40% ethyl acetate in
ligroin. A total of 168 mg of amorphous solid was upon fraction evaporation. White
amorphous solid, 'H NMR 5 (only downfield signals are listed): 3.49 (m, 1H, 3-H), 3.67
(m, 1H, 5-H), 4.11 (d, 11.7Hz, 1H, 6-H), 4.26 (dd, 4.5, 11.7Hz, 1H, 6-H), 4.59 (d,
8.1Hz, 1H, l-H), 4.95 (t, 9.3Hz, 1H, 2-H), 5.07 (t, 9.6Hz, 1H, 4-H), 5.21 (t, 9.6Hz, 1H,
3-H), 5.36 (d, 4.2Hz, 1H, 6-H).

138
Preparation of Tetraacetyl P-Sitosterol P-GIucoside by the Trichloroacetimidate
Method
A total of 100 mg of tetraacetyl glucosyl trichloroacetimidate and 100 mg of P-
sitosterol was dissolved in 3 ml of CH2CI2 and 36 mg of PTSA was added. The reaction
mixture was stirred at room temperature for 1 hour and the mixture was diluted with
water and CH2CI2 and partioned. The organic layer was washed with water, dried with
Na2S04, and evaporated. The residue was separated on a silica column using 30-40%
ethyl acetate in ligroin as the solvent. A total of 62 mg of amorphous solid product was
obtained from the evaporating fractions. White amorphous powder. !H NMR spectrum
chemical shifts were identical with those reported above.
Attempted Glucosylation of 2-Acetyl Paclitaxel by the Koenigs-Knorr Method
A total of 200 mg of 2-acetyl paclitaxel and 200 mg of acetobromoglucose was
dissolved in 3 ml of CH2CI2 and 100 mg of AgNOs was added. This mixture was stirred
overnight at room temperature. At this time the mixture was filtered and the filtrate was
diluted with water and CH2CI2 and partitioned. The organic layer was washed with water,
dried with Na2S04, and evaporated. The residue was separated on a silica column using
15-30% acetone in benzene as the mobile phase. A total of 132 mg of product was
obtained as an amorphous solid. I3C NMR 8: 10.6, 11.5, 19.8, 20.4, 25.7, 27,2, 32.6,
36.2, 44.1, 53.3, 57.6, 63.5, 68.6, 69,3, 71.1, 71.2, 71.3, 72.8, 74.5, 75.1, 82.0, 126.4,
127.0, 128.2, 128.4, 128.7, 128.8, 129,6, 129.8, 131.9, 133.1, 134.2, 136.9, 137.5, 145.9,
166.3, 167.3, 167.6, 169.4, 169.6, 169.7, 202.0.

139
L1210 Cytotoxicity Assay
The cells are maintained and subcultured in RPMI medium that was prepared in a
sterile manner in the laboratory. The cell population was maintained between 150,000 and
600,000 cells/ml. At the time of the assay, the cell suspension was diluted to contain
150,000 cells/ml. This solution (2 ml) was placed in each well of a Becton-Dickinson
deep-well plate (24 wells/plate). The test compounds were weighed and dissolved in
sufficient DMSO to make 2 mg/ml. Then several dilutions of these stock solutions were
made and tested. To the wells that contained the L1210 cells were added 10 pi aliquots of
the DMSO solutions so that the final concentration was known in parts per million. For
paclitaxel the concentrations used were 0.1, 0.05, 0.025, 0.0125, 0.00625, and 0.003125
ppm, and for all other compounds it was 10, 2, 0.2, 0.1, and 0.05 ppm. Controls were also
used in which 10 pi of DMSO were added to the well. Each concentration was tested in
quadruplicate (4 wells).
After addition of the compounds the plates were incubated for 48 hours. The
plates were removed and the cells in each well were counted to determine the number of
cells per ml. The contents of each was thoroughly mixed using a sterile 2 ml pipette, then,
1 ml of the cell suspension was transferred to a clean test tube and diluted with 1 ml of
trypan blue stain which only dyes the dead cells. The viable cell (unstained) were then
counted by shaking and placing 0.1 ml of this suspension on a Fisher hemacytometer and
counting the cells found in the five gridded areas of the hemcytometer. This number was
then multiplied by 4000 which gave the number of cell/ml.

140
The IC50 for each case was determined from the plot of log [concentration] versus
percent inhibition (not shown). The percent inhibition for each concentration was
determined by the following equation:
% Inhibition = [1 (Td T0 / Tc T0)] 100
Where Td is the number of cells per ml of the drug treated wells, T0 is the number of cells
at the start of the test, and Tc is the average number of cells per ml in the control wells.
The average of four readings for each concentration was used to calculate the IC50 for
each compound.

LIST OF REFERENCES
Ali, S. M.; Hoemann, M. Z.; Aube, J.; Mitscher, L. A.; Georg, G. I. Novel Cytotoxic 3-
(tert-Butyl) 3-Dephenyl Analogs of Paclitaxel and Docetaxel. J. Med. Chem.
1995, 35, 3821-3828.
Appendino, G. The Phytochemistry of the Yew Tree. Nat. Prod. Rep. 1995, 72, 349-360.
Arslanian, R. L.; Bailey, D. T.; Kent, M. C.; Richheimer, S. L.; Thornburg, K. R.;
Timmons, D. W.; Zheng, Q. Y. Brevitaxin, A New Diterpenolignan from the Bark
of Taxus brevifolia. J. Nat. Prod. 1995, 55, 583-585.
Boschan, R.; Merrow, R. T.; Van Dolah, R. W. The Chemistry of Nitrate Esters. Chem.
Rev. 1955, 55, 485-510.
Chen, S. H.; Farina, V. Paclitaxel Structure-Activity Relationships and Core Skeletal
Rearrangements. Taxane Anticancer Agents, Basic Science and Current Status,
ACS Symposium Series 583, 1995, 247-261.
Chen, S. H.; Huang, S.; Wei, J.; Farrina, V. The Chemistry of Taxanes-Reaction of Taxol
and Baccatin Derivatives with Lewis-Acids in Aprotic and Protic Media.
Tetrahedron 1993, 49, 2805-2828.
Chu, A.; Furlan, M.; Davin, L. B.; Zajicek, J.; Towers, G. H. N.; Soucy-Breau, C. M.;
Rettig, S. J.; Croteau, R.; Lewis, G. H. Phenylbutanoid and Taxane-Like
Metabolites from Needles of Taxus brevifolia. Phytochem. 1994, 36, 975-985.
Chu, A.; Davin, L. B.; Zajicek, J.; Lewis, N. G.; Croteau, R. Intramolecular Migrations in
Taxanes form Taxus brevifolia. Phytochem. 1993, 34, 473-476.
Commercon, A.; Bourzat, J. D.; Didier, E.; Lavelle, F. Practical Semisynthesis and
Antimitotic Activity of Docetaxel and Side-Chain Analogues. Taxane Anticancer
Agents, Basic Science and Current Status, ACS Symposium Series 583, 1995,
233-246.
Das, B.; Rao, S. P.; Das, R. Naturally Occuring Rearranged Taxoids. Planta Med. 1995,
61, 393-397.
141

142
Das, B.; Rao, S. P.; Srinivas, K. V. N. S.; Yadav, I. S.; Das, R. A Taxoid from Needles of
Himalayan Tctxus baccata. Phytochem. 1995, 38, 671-674.
Della Casa De Marcano, D. P.; Halsall, T. G. The Structure of Diterpenoid Baccatin-I, the
4(3, 20-Epoxide of 2a, 5a, 7(3, 9a, 10(3, 13a-Hexa-acetoxytaxa-4(20), 11-Diene.
Chem. Comm. 1970, 1381-1382.
Della Casa De Marcano, D. P.; Halsall, T. G. Structures of Some Taxane Diterpenoids,
Baccatins-III, -IV, -VI and -VII and 1-Dehydroxybaccatin-IV, Possessing an
Oxetane Ring. Chem. Comm. 1975, 365-367.
Dennis, J. N.; Green, A. E.; Guenard, D.; Voegelein, F. G.; Managatal, L.; Poitier, P. A
Highly Efficient, Practical Approach to Natural Taxol. J. Am. Chem. Soc. 1988,
110, 5917-5919.
Georg, G. I.; Harriman, G. C. B.; Velde, D. G. V.; Boge, T. C.; Cheruvallath, Z. S.;
Datta, A.; Hepperle, M.; Park, H.; Himes, R. H; Jayasinghe, L. Medicinal
Chemistry of Paclitaxel; Chemistry, Structure-Activity Relationships, and
Conformational Analysis. Taxane Anticancer Agents, Basic Science and Current
Status, ACS Symposium Series 583, 1995, 217-232.
Green, T. W ; Wuts, P. G. M. Protective Groups in Organic Synthesis 1991, 417-420.
Gueritte-Voegelein, F.; Senilh, V.; David, B.; Guenard, D.; Potier, P. Tetrahedron 1986,
42, 4451-4457.
Gunawardana, G. P.; Premachandran, U.; Burres, N. S.; Whittern, D. N.; Henry, R.;
Spanton, S. Isolation of 9-Dihydro-13-Acetylbaccatin-III from Taxus canadensis.
J. Nat. Prod. 1992, 55, 1686-1689.
Holton, R. A.; Somoza, C.; Kim, H. B.; Liang, F.; Biediger, R. J.; Boatman, P. D.;
Shindo, M.; Smith, C. C.; Kim, S.; Nadizadeh, H.; Suzuki, Y.; Tao, C.; Vu, P.;
Tang, S.; Zhang, P.; Murthi, K. K.; Gentile, L. N.; Liu, J. H. First Total Synthesis
of Taxol. 1. Functionalization of the B Ring. J. Am. Chem. Soc. 1994a, 116, 1597-
1598.
Holton, R. A.; Somoza, C.; Kim, H. B.; Liang, F.; Biediger, R. J.; Boatman, P. D.;
Shindo, M.; Smith, C. C.; Kim, S.; Nadizadeh, H.; Suzuki, Y.; Tao, C.; Vu, P.;
Tang, S.; Zhang, P.; Murthi, K. K.; Gentile, L. N.; Liu, J. H. First Total Synthesis
of Taxol. 2. Completion of the C and D Rings. J. Am. Chem. Soc. 1994b, 116,
1599-1600.
Kahne, D.; Walker, S.; Cheng, Y.; Engen D. V. Glycosylation of Unreactive Substrates. J.
Am. Chem. Soc. 1989, 111, 6881-6882.

143
Kingston, D. G. I. Recent Advances in the Chemistry and Structure-Activity Relationships
of Paclitaxel. Taxane Anticancer Agents, Basic Science and Current Status, ACS
Symposium Series 583, 1995, 203-216.
Kingston, D. G. I.; Molinero, A. A.; Rimoldi, J. M. The Taxane Diterpenoids. Progress in
the Chemistry of Organic Natural Products 1993, 61, 1-206.
Kingston, D. G. I.; Hawkins, D. R.; Ovinton, L. New Taxanes from Taxus brevifolia.
J. Nat. Prod. 1982, 45, 466-470.
Lataste, H ; Senilh, V.; Wright, M.; Guenard, D.; Potier, P. Relationships Between the
Structures of Taxol and Baccatin III Derivatives and their in vitro Action on the
Disassembly of Mammalian Brain and Physarum Amoebal Microtubules. Proc.
Natl. Acad. Sci. 1984, 81, 4090-4094.
Liang, J.; Min, Z.; Iinuma, M.; Tanaka, T.; Mizuno, M. Two New Antineoplastic
Diterpenes from Taxus mairei. Chern. Pharm. Bull. 1987, 35, 2613-2614.
Mellado, W.; Magri, N. F.; Kingston, D. G. I.; Garcia-Arenas, R.; Orr, G. A.; Horwitz, S.
B. Preparation and Biological Activity of Taxol Acetates. Biophys. Res. Comm.
1984., 124, 329-336.
Miller, R. W A Brief Survey of Taxus Alkaloids and Other Taxane Derivatives. J. Nat.
Prod. 1980, 43, 425-437.
Miller, R. W ; McLaughlin, J. L.; Powell, R. G.; Plattner, R. D ; Weisleder, D.; Smith, C.
R. Nat. Prod. 1982, 47, 131-138.
Nicolaou, K. C.; Yang, Z.; Liu, J. J.; Ueno, H.; Nantermet, P. G.; Guy, R. K.; Claiborne,
C. F., Renaud, J.; Couladouros, E. A.; Paulvannan, K.; Sorensen, E. J. Total
Synthesis of Taxol. Nature 1994, 367, 630-634.
Ojima, I.; Park, Y. H.; Fenoglio, I.; Duelos, O.; Sun, C. M.; Kuduk, S. D.; Zueco, M.;
Appendino, G.; Pera, P.; Veith, J. M.; Bernacki, R. J.; Bissery, M. C.; Combeau,
C.; Vrignaud, P.; Riou, J. F.; Lavelle, F. Synthesis and Structure-Activity
Relationships of New Taxoids. Taxane Anticancer Agents, Basic Science and
Current Status, ACS Symposium Series 583, 1995, 262-275.
Paradis, R.; Page, M. New Active Paclitaxel Glucuronide Derivative with Improved Water
Solubility. Int. J. Oncol. 1998, 12, 391-394.
Rao, K. V. Taxol and Related Taxanes. I. Taxanes of Taxus brevifolia Bark. Pharm. Res.
1993. 10, 521-524.

144
Rao, K. V. Semi-Synthesis of 10-Deacetyl Paclitaxel form 10-Deacetyl Paclitaxel-7-
Xyloside. J. Heter. Chem. 1997, 34, 675-680.
Rao, K. V.; Juchum, J. Taxanes form The Bark of Taxus brevifolia. Phytochem. 1998, 47,
1315-1324.
Rao, K. V.; Bhakuni, R. S.; Juchum, J.; Davies, R. M. A Large Scale Process for
Paclitaxel and Other Taxanes from the Needles of Taxus media Hicksii and Taxus
floridana Using Reverse Phase Column Chromatography. J. Liq. Chrom. & Rel.
Technol. 1996a, 19, 427-447.
Rao, K. V.; Bhakuni, R. S.; Hanuman, J. B.; Davies, R.; Johnson, J. Taxanes from the
Bark of Taxus brevifolia. Phytochem. 1996b, 41, 863-866.
Rao, K. V.; Hanuman, J. B.; Alverez, C.; Stoy, M; Juchum, J.; Davies, R. M.; Baxley, R.
A New Large-Scale Process for Taxol and Related Taxanes from Taxus brevifolia.
Pharm. Res. 1995, 12, 1003-1010.
Schiff, P. B.; Fant, J.; Horwitz, S. B. Promotion of Microtubule Assembly irt vitro by
Taxol. Nature 1979, 277, 665-667.
Senilh, V.; Blechert, S.; Picot, F.; Pother, P.; Varenne, P. Mise en Evidence de Nouveaux
Analogues du Taxol Extraits de Taxus baccata. J. Nat. Prod. 1984, 47, 131-137.
Shuzi, T; Xinzhong, S.; Caiyu, W.; Wenbin, S.; Weiyi, H. Synthesis of 2, 2-Dimethyl-4-
Cyclohexene-1, 3-Dione. J. China Pharm. Univ. 1991, 22, 173-174.
Smite, E.; Lundgren, L. N.; Andersson, R. Arylbutanoid and Diarylheptanoid Glycosides
from Inner Bark of Betulapndula. Phytochem. 1993, 32, 365-369.
Takashi, T.; Tsukamoto, H.; Yamada, H. Design and Synthesis of a Water-Soluble Taxol
Analogue: Taxol-Sialyl Conjugate. Bioorg. Med. Chem. Lett. 1998, 8, 113-116.
Tanaka, K.; Fuji, K.; Yokoi, T.; Shingu, T.; Li, B.; Sun, H. Structures of Taxchinins L and
M, Two New Diterpenoids from Taxus chinensis var. Mairei. Chem. Pharm.
Bull. 1996, 44, 1770-1774.
Thayer, P. S.; Himmelfarb, P.; Watts, G. L. Cytotoxicity Assays with L1210 Cells in vitro:
Comparison with L1210 in vivo and KB Cells in vitro. Cancer Chemother. Rep.
1971, 2, 1-10.
Toshima, K.; Tatsuta, K. Recent Progress in O-Glycosylation Methods and Its Application
to Natural Products Synthesis. Chem. Rev. 1993, 93, 1503-1531.

145
Wang, X.; Pan, X. A Convergent Synthesis of Taxamairin A, B Precursor I. 8-(2, 3-
Dimethoxy-4-isopropylbenzyl)-l, 4-dioxaspiro[4, 5]decan-7-one. J. Indian Chem.
Soc. lL995a, 73, 497-498.
Wang, X.; Pan, X.; Zhang, C.; Chen, Y. The Total Synthesis of Taxamairin B. Synth.
Comm. 1995b, 25, 3413-3419.
Wani, M. C.; Taylor, H. L.; Wall, M. E.; Coggon, P.; McPhail, A. T. Plant Antitumor
Agents. VI. The Isolation and Structure of Taxol, a Novel Antileukemic and
Antitumor Agent from Taxus brevifolia. J. Am. Chem. Soc. 1971, 93, 2325-2327.

BIOGRAPHICAL SKETCH
The author was bom to James Harvey Johnson and Lawilla McLamb
Johnson on September 30, 1970 in Fayetteville, North Carolina. His father worked
for the civil service at Fort Bragg Army Reservation and his mother was a
homemaker. He grew up with two older sisters, Melinda and Teresa. He attended
and graduated in June 1988 from Cape Fear High School in Vander, North
Carolina near Fayetteville. In January 1989 he enrolled at Fayetteville State
University on a full academic scholarship. It was at this time that he began to
develop a love for the discipline of chemistry. After graduating with a B.S. in
chemistry with honors in May 1993 he enrolled in the Department of Medicinal
Chemistry at the University of Florida during August 1993. It was there that this
doctoral work was completed under the supervision of Koppaka V. Rao. The
author was married to Amy Raye Hardy on April 30, 1994 and has one child,
James Harvey Johnson III (Trey) born September 23, 1995.
146

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Koppaka V. Rao, Chair (deceased)
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Johpi
Profs
Perrin, Co-Chair
sor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
0, .Jc
Margaret'O. James
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy. /
William R. Dolbier Jr. /
Professor of Chemistry { /

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
(jju
Do
Donghai Wu
Assistant Professor of Medicinal
Chemistry
This dissertation was submitted to the Graduate Faculty of the College of
Pharmacy and to the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
December 1998
Dean, College of Pharr
Dean, Graduate School



144
Rao, K. V. Semi-Synthesis of 10-Deacetyl Paclitaxel form 10-Deacetyl Paclitaxel-7-
Xyloside. J. Heter. Chem. 1997, 34, 675-680.
Rao, K. V.; Juchum, J. Taxanes form The Bark of Taxus brevifolia. Phytochem. 1998, 47,
1315-1324.
Rao, K. V.; Bhakuni, R. S.; Juchum, J.; Davies, R. M. A Large Scale Process for
Paclitaxel and Other Taxanes from the Needles of Taxus media Hicksii and Taxus
floridana Using Reverse Phase Column Chromatography. J. Liq. Chrom. & Rel.
Technol. 1996a, 19, 427-447.
Rao, K. V.; Bhakuni, R. S.; Hanuman, J. B.; Davies, R.; Johnson, J. Taxanes from the
Bark of Taxus brevifolia. Phytochem. 1996b, 41, 863-866.
Rao, K. V.; Hanuman, J. B.; Alverez, C.; Stoy, M; Juchum, J.; Davies, R. M.; Baxley, R.
A New Large-Scale Process for Taxol and Related Taxanes from Taxus brevifolia.
Pharm. Res. 1995, 12, 1003-1010.
Schiff, P. B.; Fant, J.; Horwitz, S. B. Promotion of Microtubule Assembly irt vitro by
Taxol. Nature 1979, 277, 665-667.
Senilh, V.; Blechert, S.; Picot, F.; Pother, P.; Varenne, P. Mise en Evidence de Nouveaux
Analogues du Taxol Extraits de Taxus baccata. J. Nat. Prod. 1984, 47, 131-137.
Shuzi, T; Xinzhong, S.; Caiyu, W.; Wenbin, S.; Weiyi, H. Synthesis of 2, 2-Dimethyl-4-
Cyclohexene-1, 3-Dione. J. China Pharm. Univ. 1991, 22, 173-174.
Smite, E.; Lundgren, L. N.; Andersson, R. Arylbutanoid and Diarylheptanoid Glycosides
from Inner Bark of Betulapndula. Phytochem. 1993, 32, 365-369.
Takashi, T.; Tsukamoto, H.; Yamada, H. Design and Synthesis of a Water-Soluble Taxol
Analogue: Taxol-Sialyl Conjugate. Bioorg. Med. Chem. Lett. 1998, 8, 113-116.
Tanaka, K.; Fuji, K.; Yokoi, T.; Shingu, T.; Li, B.; Sun, H. Structures of Taxchinins L and
M, Two New Diterpenoids from Taxus chinensis var. Mairei. Chem. Pharm.
Bull. 1996, 44, 1770-1774.
Thayer, P. S.; Himmelfarb, P.; Watts, G. L. Cytotoxicity Assays with L1210 Cells in vitro:
Comparison with L1210 in vivo and KB Cells in vitro. Cancer Chemother. Rep.
1971, 2, 1-10.
Toshima, K.; Tatsuta, K. Recent Progress in O-Glycosylation Methods and Its Application
to Natural Products Synthesis. Chem. Rev. 1993, 93, 1503-1531.


137
Preparation of Tetraacetyl Benzyl (3-Glucoside by the Trichloroacetimidate Method
A total of 100 mg of tetraacetyl glucosyl trichloroacetimidate and 28 mg of benzyl
alcohol was dissolved in 2 ml of CH2CI2 and 36 mg of PTSA was added. The reaction
mixture was stirred at room temperature for 1 hour and the mixture was diluted with
water and CH2C12 and partitioned. The organic layer was washed with water, dried with
Na2S04, and evaporated. The residue was separated on a silica column using 30-40%
ethyl acetate in ligroin as the solvent. A total of 56 mg of product crystallized from the
evaporating fractions. White crystalline powder, and k'C NMR spectra chemical shifts
were identical with those reported above.
Preparation of Tetraacetyl P-Sitosterol P-GIucoside by the Koenigs-Knorr Method
A total of 300 mg of acetobromoglucose and 300 mg of p-sitosterol was dissolved
in 3 ml of CH2C12 and excess ZnCl2 was added (100 mg). The reaction was stirred at room
temperature for 18 hours, the solid was filtered off and the filtrate was diluted with water
and CH2C12 and partioned. The organic layer was washed with water, dried with Na2S04,
evaporated, and the residue was ran on a silica column using 30-40% ethyl acetate in
ligroin. A total of 168 mg of amorphous solid was upon fraction evaporation. White
amorphous solid, 'H NMR 5 (only downfield signals are listed): 3.49 (m, 1H, 3-H), 3.67
(m, 1H, 5-H), 4.11 (d, 11.7Hz, 1H, 6-H), 4.26 (dd, 4.5, 11.7Hz, 1H, 6-H), 4.59 (d,
8.1Hz, 1H, l-H), 4.95 (t, 9.3Hz, 1H, 2-H), 5.07 (t, 9.6Hz, 1H, 4-H), 5.21 (t, 9.6Hz, 1H,
3-H), 5.36 (d, 4.2Hz, 1H, 6-H).


9
O
y
1. Esterification
2. Oxygenation
3. McMurry Pinacol Coupling
4. Shapiro Coupling
TESO O
-N
Ph
Bz
6
/ Diels-AIder
/
OTBS OBn
Diels-AIder
Figure 1-5: Nicolaou Synthesis of Paclitaxel
closure of a precursor properly functionalized at C-l, C-2, C-3, C-7, and C-8 (15).
Synthesis of this precursor was made possible by conformational control of the eight


29
tertiary alcohol (52). Dehydration was then achieved by heating with 20% H2SO4. The
resulting olefin (53) was then reduced with RaNi and H2. A hydroxymethyl group was
attached ortho to the methoxyl by treating 54 with n-butyl lithium in THF and TMEDA
and later addition of paraformaldehyde. The hydroxymethyl function was converted to a
bromomethyl function using PBr3 in dichloromethane. This bromo compound (56) then
underwent nucleophilic substitution with ketone 49 and LDA as the base to yield ketone
57 (Figure 2-10). Ketone 57 was treated with vinyl magnesium bromide and the B ring
was closed by a Friedal-Crafts alkylation using BF3-OEt2. Finally the gem-dimethyls were
attached using potassium t-butoxide and methyl iodide to produce the key intermediate 60
in which the olefin bond moved into the cycloheptane ring. Oxidation of the allylic and
benzylic methylene group to produce the ketone at C-l was achieved by using excess
aqueous 75% t-butyl hydroperoxide and catalytic amounts of chromic anhydride. Finally,
olefination of C-4, 5 and C-6, 7 was achieved by heating with excess DDQ in toluene
followed by hydrogenation with 10% Pd-C. This then gave the final product taxamairin B.
Since the above synthesis was the only synthesis for this class of diterpenes it was
decided that it would be a noteworthy side project to synthesis this type of diterpene by a
simpler method. This method also uses a convergent approach by constructing an A ring
precursor and a C ring precursor and then bringing the precursors together to form the B
ring. The A ring precursor was synthesized by modifying a known method starting with 1,
3-dicyclohexanedione (63) (Shuzi et al., 1991). This diketone was dimethylated with
methyl iodide and K2C03 in refluxing acetone in give a yield of about 65% after vacuum
distillation. The 2, 2-dimethyl-1, 3-cyclohexanedione product (64) was then mono-


10
Figure 1-6: Holton Synthesis of Paclitaxel
membered B ring, via ketone 16. Ketone 16 was projected to arise from an aldol
condensation of the enolate of ketone 17, the formation of which would be made possible


92
The benzene/ligroin layers which contained the phenylhydrazones were separated and
concentrated to dryness. The combined methanol/water layer was concentrated partially
and extracted three times with dichloromethane and the organic layer was concentrated to
yield crude 10-deacetyl paclitaxel (0.8g). This material was clean enough for further work.
All NMR spetra matched that of an authentic sample.
Regioselective Nitrations of 10-Deacetyl Paclitaxel
10-Deacetyl Paclitaxel 500 mg was dissolved in 6 ml dichloromethane and cooled
to 0 C. Next 3 ml of a 5 : 1 mixture of acetic anhydride and concentrated nitric acid
which was also cooled to 0 C was added and the reaction mixture was stirred in an ice
bath for 5 minutes. The reaction was worked up as with paclitaxel-7, 2-nitrate ester.
Analysis of the reaction mixture by TLC revealed that practically all the starting material
was gone and four products were present with two being major products. The fastest
moving product was the completely nitrated 10-deacetyl paclitaxel and was in minor
amounts. The next was a major product followed by two minor products and then by
another major product. A silica column was used to separate this products with 0 15%
acetone in dichloromethane as the solvent. From this column 175 mg of the faster major
product (103) was obtained and crystallized from diethyl ether while 126 mg of the slower
major product (102) was isolated but contained impurities. This was put on another
column of the same type and solvent and 87 mg of the product was obtained. This material
however would not crystallize. After analysis by NMR the faster product was determined
to be 10-deacetyl paclitaxel-7, 10-dinitrate ester (103) and the slower product was the 10-
mononitrate ester (102).


72
O
104 Rx = N02, R2 = N02, R3 = N02, R4 = N02, R5 = N02
105 Rx = N02, R2 = H, R3 = H, R4 = H, R5 = H
106 Rx = H, R2 = N02, R3 = H, R4 = H, R5 = II
107 Rx = H, R2 = H, R3 = N02, R4 = H, R5 = H
108 Rj = N02, R2 = N02, R3 = N02, R4 = N02, R5 = H
Figure 3-2: Nitration of 10-Deacetyl Paclitaxel-7- p-Xyloside
l"-OH, 2"-OH, 3"-OH > 10-OH > 2'-OH
10-Deacetyl Paclitaxel-7- p-Xyloside
acetic anhydride/pyridine
30 sec., room temp.
T
Figure 3-3: Regioselective Acetylation of
10-Deacetyl Paclitaxel-7-p-Xyloside
conceivable that nitrate esters can be used as selective hydroxyl protecting groups when
reactivity differing from acetylation reactivity is desired.


57
5.78 (q, 6.6, 13.5Hz, 1H, CH), 7.10 (s, 1H, Ar-H). 13C NMR 6: 20.9, 26.7, 35.4, 50.0,
55.8, 60.9, 109.8, 118.0, 140.6, 142.5, 152.5.
Isolation of Minor Compounds from Taxus floridana
The mother liquors of fractions that contained 10-deacetyl baccatin III from the
reverse-phase column (25-40% acetonitrile in water) were concentrated to a syrup (10 g).
On standing, this syrup became an amorphous solid and 3 g of this was applied to a
normal-phase silica column (40 g) using dichloromethane as the starting solvent. Then the
column was eluted successively with dichloromethane containing 2, 5, and 10% acetone
and then with addition of 2, 5, and 10% methanol. A total of 200 ml of each solvent
mixture was passed through the column before the next solvent mixture was started and
20 ml fractions were collected and monitored by TLC. The initial dichloromethane eluent
was put aside for further chromatography and the order of elution of the more polar
compounds was as follows: ip-hydroxy baccatin I, taxiflorine (taxchinin M),
rhododendrol, taxchinin L, 10-deacetyl baccatin III, ponasterone A, and 10-deacetyl
paclitaxel-7-P-xyloside.
The initial dichloromethane eluent (0.250 mg) was applied to another silica column
(3 g) using 25% ethyl acetate in ligroin as the initial solvent and proceeding to 50% ethyl
acetate in 10% intervals. A total of 50 ml of each solvent was used before progressing to
the next solvent mixture and 5 ml fractions were collected and monitored by TLC. The
order of elution was trans-2,6-dimethoxy cinnamaldehyde, a-conidendrin, and 1-deoxy
baccatin IV.


22
Isolation of Minor Compounds from the Bark of Taxus brevifolia
Obviously the above process was only used to isolate major compounds; however,
many minor compounds exist in the filtrates or in the in-between fractions. A TLC analysis
of the filtrates from the region between 10-deacetyl baccatin III and 10-deacetyl paclitaxel
(xyloside fractions) showed many interesting compounds, the identity of which could not
be determined by comparison with available standards. This material was therefore
concentrated to a solid and rechromatographed on regular-phase silica using an elution
system of 0-5% acetone in dichloromethane to 0-5% methanol in 5%
acetone/dichloromethane.
The first compound to be eluted was 1 (3-hydroxy baccatin I (35). This compound
has been known for quite some time and was first isolated from Taxus baccata in 1970
(Della Casa De Marcano & Halsall, 1970). This compound belongs to the baccatin I sub
family because it contains a C-4-C-20 epoxide as opposed to an oxetane ring. Baccatin VI
(36) was eluted next and is another well known taxane isolated for the first time from
Taxus baccata in 1975 (Della Casa De Marcano & Halsall, 1975). This compound is so-
named because it is esterified at C-9 as opposed to paclitaxel/ 10-deacetyl baccatin III
which have a C-9 ketone (Figure 2-4).
The next compound to be eluted was l(3-hydroxy-7-deacetyl baccatin I (37). This
compound was recently isolated from the needles of Taxus brevifolia (Chu et al., 1993)
and has been reported to undergo an acetyl migration from C-9 to C-7 when kept in
solution. Indeed, this compound did form another spot on TLC when left in solution;


95
well as some starting material and minor products. A clean separation of the three major
products was not possible using only one solvent system thus a first column was ran on
this material using 40% > 80% ethyl acetate in ligroin. This eluted 84 mg of the slowest
of the three products in pure form and this compound was determined to be 10-deacetyl
paclitaxel-7-P-xyloside-2-mononitrate ester (105). This product did not crystallize. The
mixture of the remaining two products was separated on a silica column using 30% >
50% acetone in benzene as solvent. The faster of these two products was the 3-
mononitrate ester (106) and was crystallized from dichloromethane (76 mg). The slower
product was determined to be the 4-mononitrate ester (107) (101 mg) and was not
crystallized.
10-Deacetyl paclitaxel-7-(3-xyloside-2, 3, 4, 10-tetranitrate ester (108)
White crystalline powder, mp 182-184 C, Anal. Calc, for C50H53N5O25: C 53.43;
H 4.75; N 6.23. Fd. C 53.67; H 4.82; N 5.87. >H NMR 5: 1.12 (s, 3H, 17-H), 1.18 (s,
3H, 16-H), 1.75 (s, 3H, 19-H), 1.86 (s, 3H, 18-H), 2.03 (m, 1H, 6-H{3), 2.34 (m, 2H, 14-
Ha,(3), 2.39 (s, 3H, 4-OAc), 2.80 (m, 1H, 6-Ha), 3.67 (dd, 5.4, 12.9Hz, 1H, 5-Heq),
3.77 (d, 6.9Hz, 1H, 3-H), 4.14 (m, 1H, 5-Hax), 4.20 (d, 8.1Hz, 1H, 20-Hp), 4.21 (m,
1H, 7-H), 4.32 (d, 8.1Hz, 1H, 20-Ha), 4.76 (d, 4.2Hz, 1H, 1-H), 4.78 (d, 2.7Hz, 1H,
2-H), 4.85 (d, 8.1Hz, 1H, 5-H), 4.97 (dd, 2.1, 3.9Hz, 1H, 2-H), 5.07 (dd, 4.2, 5.4Hz,
1H, 4-H), 5.26 (t, 6.0Hz, 1H, 3-H), 5.69 (d, 6.9Hz, 1H, 2-H), 5.75 (dd, 2.1, 8.4Hz,
1H, 3-H), 6.18 (t, 9.0Hz, 1H, 13-H), 6.39 (s, 1H, 10-H), 7.05 (d, 9.0Hz, 1H, N-H), 7.36-
7.53 (m, 10H, m-Bz, o,m,p-Ph, m,p-NBz), 7.62 (t, 7.2Hz, 1H, p-Bz), 7.72 (d, 7.2Hz, 2H,
o-NBz), 8.11 (d, 7.5Hz, 2H, o-Bz). 13C NMR 5: 10.6, 14.9, 21.4, 22.5, 26.2, 35.5, 35.8,


116
O
+
aliphatic amine
NaCNBH3
2 : 1 CH3OH : AcOH
O
Figure 4-4: Dialdehyde Reduction to Diol
Three methods of glycosylation were to be studied in this work are these are; 1) the
Koenigs-Knorr method, 2) the trichloroacetimidate method, and 3) the sulphoxide
method. These will now be briefly discussed.
The Koenigs-Knorr method was first introduced in 1901 and has since been a
mainstay in glycosylation chemistry (Toshima & Tatsuta, 1993). In its classical form it


49
nm. IR(KBr): 1670, 1621, 1333, 1305, 1038 cm'1. *H NMR 5: 1.30 (d, 6.9Hz, 6H, 19-
H, 20-H), 1.46 (s, 6H, 12-H, 13-H), 3.41 (heptet, 6.9Hz, 1H, 18-H), 3.98 (s, 6H, 14-
OMe, 15-OMe), 6.12 (d, 9.6Hz, 1H, 7-H), 6.94 (s, 1H, 11-H), 7.31 (d, 1H, 11-H), 7.87
(s, 1H, 17-H), 7.93 (s, 1H, 4-H). 13C NMR 5: 22.9, 23.0, 26.7, 26.8, 27.8, 50.4, 60.7,
61.2, 123.0, 124.1, 127.3, 128.3, 130.9, 131.5, 133.8, 136.1, 147.8, 150.8, 151.3, 153.9,
188.0, 200.7.
Synthesis of Taxamairin B (39)
2, 2-Dimethyl-l, 3-cyclohexanedione (64)
A total of 10 g of 1, 3-cyclohexanedione (63) and 30.6 g (2.5 eq.) of K2CO3 was
added to 40 ml of acetone to which 31.5 g (2.5 eq.) of CH3I had been added. The mixture
was refluxed overnight. After cooling the mixture the K2CO3 was filtered and the acetone
was removed under vacuum. The residual material was partitioned between water and
diethyl ether and the water layer was discarded. The solvent was evaporated to yield a
syrup which was poured into a mixture of 20 ml of cone. HC1 and 20 g of ice. This was
stirred for 30 minutes to decompose the methyl enol ether which accounts for about 30%
of the mixture, then water and diethyl ether were added and partitioned. The organic layer
was washed twice with water, then dried with sodium sulfate. After removal of the
solvent, the crude liquid product was distilled using a water aspirator vacuum (~15 mm
Hg) and the product distilled at 120-122 C. Upon standing the product crystallized
yielding 6.2 g. Colorless crystals, mp 31-32 C, 'H NMR 6: 1.29 (s, 6H, CH3), 1.93 (m,
6.5Hz, 2H, 5-H), 2.67 (t, 6.9Hz, 4H, 4-H, 6-H). 13C NMR 5: 18.1, 22.3, 37.4, 61.8,
210.6.


133
p-Aminobenzenesulfonamide product (164)
Light brown crystalline powder, 418 mg, mp 184-186 C, UV /W (CH3OH):
221, 273 nm, FABMS m/z: 1053 (M + 2), 1020, 767, 734, 241, 105. *H NMR 5 (some
DMSO): 1.09 (s, 3H, 17-H), 1.17 (s, 3H, 16-H), 1.75 (s, 3H, 19-H), 1.76 (m, 1H, 14-
Hp), 1.87 (s, 3H, 18-H), 1.98 (m, 1H, 6-HP), 2.18 (m, 1H, 14-Ha), 2.40 (s, 3H, 4-Oac),
2.72 (m, 1H, 6-Ha), 3.10 (m, 2H, 2-, 4-H), 3.25 (m, 2H, 2-, 4-H), 3.67 (m, 1H,
5-H), 3.87 (br s, 1H, 3-H), 4.02 (m, 1H, 5-H), 4.05 (m, 1H, 7-H), 4.08-4.22 (m, 2H,
20a-, 20P-H), 4.57 (br s, 1H, 1-H), 4.66 (br s, 1H, 2-H), 4.92 (br s, 1H, 5-H), 5.13 (s,
1H, 10-H), 5.61 (br s, 1H, 2-H), 5.73 (d, 7.5Hz, 1H, 3-H), 6.17 (br s, 1H, 13-H), 6.89
(d, 8.7Hz, 2H, o-Ar), 7.28-7.61 (m, 11H, m,p-OBz, m,p-NBz, o,m,p-Ph), 7.75 (d, 8.7Hz,
2H, m-Ar), 7.87 (d, 6.9Hz, o-NBz), 8.10 (br s, 2H, o-OBz), 8.31 (d, 8.7Hz, 1H, N-H).
13C NMR 5 (some DMSO): 9.9, 13.3, 20.1, 21.8, 25.8, 34.7, 34.8, 42.3, 45.1, 45.8, 49.9,
55.0, 56.0, 60.3, 70.2, 73.2, 73.6, 74.1, 75.4, 76.5, 79.7, 83.2, 97.7, 113.1, 126.2, 126.6,
126.7, 126.8, 127.4, 127.6, 127.7, 129.2, 130.6, 132.1, 132.4, 133.5, 135.4, 137.1, 138.3,
151.8, 165.2, 166.2, 169.4, 171.8, 209.0.
Synthesis of Glucose Pentaacetate
A total of 1.0 g of d-glucose was added to a mixture of 5.0 g (4.6 ml) of acetic
anhydride and 6.5 g (6.65 ml) of pyridine at 0 C and this was stirred overnight while
allowing the mixture to warm to room temperature. At this point about 20 ml of ice water
was added and the product crystallized out of solution. These crystals were filtered,
washed with water, and dried under reduced pressure. A total of 2.12 g of product was
obtained. White crystalline powder, product existed as a mixture of a and P anomers as


5
O
I
3
Figure 1-3: Halogenated Products of Methanolysis
Used for X-Ray Crystallography
regarded as an N-benzoyl derivative of (2R, 3S)-3-phenylisoserine. The tetraol formed by
the methanolysis of paclitaxel was converted to a bisiodoacetate (3), C33H38I2O12, which
again received x-ray analysis. The structure is shown in Figure 1-3.


84
products, the neutral solution was about 40% decomposed, and the acid solution still
contained almost all keto-ester.
Concerning the C-l and C-2 fragment, it was assumed that C-l exist as a
carboxyl in its final form however it is unclear concerning the C-2. Thus it can be
concluded that this two carbon fragment may exist as acetic acid, glycolic acid, or
glyoxalic acid in its final form. A failed attempt was made to derivatize the acid function
by treating the reaction mixture with DCC and aniline to produce a UV active amide that
could be isolated and characterized.
Unfortunately because of time constraints a mechanism for this rearrangement
could not be conclusively established however a possible mechanism has been formulated
and is presented below and in Figure 3-12. It has already been established that the keto-
ester (133) can enolize in the presence of base to form the enolate and thus the enol (134).
The amide proton in this enol is subsequently made quite acidic and can also be abstracted
by base and after electron migration the imine alcohol can form (135). This conjugated
imine is thus a reactive Michael-type adduct which can be attacked by hydroxide with the
glycolic enolate serving as the leaving group. The imine can then be rearranged to form
dibenzamide (138) while the C-l ester is hydrolyzed giving glycolic acid (140) and
baccatin III-7-nitrate ester (141). This hydrolysis has been shown to occur last because the
ester enolate would presumably serve as a better leaving group than the acid enolate,
however hydrolysis of the side chain may occur first. It would be interesting to test the
hypothesis by subjecting the keto-acid to these conditions to see if the rearrangement still
takes place. One could also alkylate the amide to a tertiary amide and determine if the


35
The resulting diol was then oxidized to the keto-aldehyde (70) with the mild
oxidizing reagent PDC. This product was coupled to 2, 2-dimethyl-4-cyclohexene-l, 3-
dione (66) by a tandem cross-aldol reaction in pyridine and piperidine (Figure 2-11).
Although other products were produced the desired product was the major one. This
pathway was expected since the most acidic carbon would be adding to the most
electrophilic carbon first. Once this occurs, the 7-membered ring would be expected to
close rather easily by another aldol reaction. The minor products were undoubtedly a
variety of other cross-aldol products. This major product matched taxamairin B, which
was obtained by methylating taxamairin A, in every way.
Isolation of Minor Compounds from Taxus floridana
The same reverse-phase chromatography protocol previously described was
applied to the needles of the Florida yew (Taxus floridana) with very good results (Rao et
al., 1996a). Although there was some question initially if this protocol would work on
needles as well as on bark because of the greater content of waxes and pigments found in
the needles, these questions were put to rest as all of the lipophilic material remained on
the column while using the appropriate taxane solvent system without clogging up the
column. It was found that paclitaxel could be isolated from these needles in a yield of
0.01% and 10-deacetyl baccatin III could be obtained in yields of 0.06%. Again, by TLC
analysis, this time of the pre-paclitaxel fractions, it was found that many unidentifiable
compounds were present in the filtrates. These filtrates were combined and evaporated to


I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Koppaka V. Rao, Chair (deceased)
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Johpi
Profs
Perrin, Co-Chair
sor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
0, .Jc
Margaret'O. James
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor
Professor of Medicinal Chemistry
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy. /
William R. Dolbier Jr. /
Professor of Chemistry { /


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
CHEMISTRY OF TAXANES AND TAXUS SPECIES
By
James Harvey Johnson Jr.
December 1998
Chairman: Koppaka V. Rao (deceased)
Co-Chairman: John Perrin
Major Department: Medicinal Chemistry
The chemistry of both taxane diterpenoids and Taxus species are studied. Several
taxanes are isolation from both Taxus brevifolia and Taxus floridana. In addition to these
taxanes several non-taxane compounds are also isolated. One of these, trans-2, 6-
dimethoxycinnamaldehyde is a novel structure that has been synthesized. Another non-
taxane, taxamairin B, which belongs to rare class of diterpenes, is synthesized by a more
efficient route than was reported in the literature. Nitrate ester forming reactions are also
studied with paclitaxel and closely related analogues. It is shown that this reaction is
regioselective in many cases and thus illustrates the potential use of nitrate esters as
protecting groups in taxane chemistry. Also several unexpected reactions of these nitrate
esters are explored including an unusually rearrangement in which the nitrated paclitaxel


117
NaCNBH3
2 : 1 CH3OH : AcOH
Y room temp., 1.5 hrs
O
160 Ri = COOH, R2 = H, R3 = H
161 Ri COOH, R2 = OH, R3 = H
162 Rx = H, R2 = COOH, R3 = OH
163 Rj = N02, R2 = H, R3 = H
164 Rj = S02NH2, R2 H, R3 = H
Figure 4-5: Reductive Aminations


127
Oxidation of Xyloside with Periodate
10-Deacetyl paclitaxel -7-(3-xyloside (1.0 g) was dissolved in 10 ml of 1 : 1 THF
and water and 2 ml of 1 N H2S04 was added. This was followed by 0.71 g of NaI04 and
the mixture was stirred overnight at room temperature. The mixture was diluted with
water and the solid which precipitated (920 mg) was filtered and dried.
Condensation of Dialdehyde with Malonic Acid
A total of 200 mg of dialdehyde was dissolved in 3 ml of pyridine and 5 drops of
piperidine was also added. To this mixture was added 20 mg of malonic acid and the
mixture was refluxed for 3 hours. At this point the mixture was diluted with CH2CI2 and
washed with 0.1 N HC1 3 times and with water twice. The organic layer was then dried
with Na2S04 and evaporated to a residue. The residue was separated on a silica column
using 30-50% ethyl acetate in ligroin with a few drops of acetic acid as the mobile phase.
A total of 123 mg of product was crystallized from diethyl ether, ligroin, and acetone.
Yellowish white crystalline powder, mp 172-174 C (dec.), UV ^max (CH3OH): 225 nm,
FABMS m/z: 954 (M + 1), 795, 669, 651, 105. *H NMR 5 (some DMSO): 1.09 (s, 3H,
17-H), 1.18 (s, 3H, 16-H), 1.75 (s, 3H, 19-H), 1.85 (s, 3H, 18-H), 1.87 (m, 1H, 14-H(3),
2.04 (m, 1H, 6-HP), 2.30 (m, 1H, 4-Ha), 2.39 (s, 3H, 4-OAc), 2.78 (m, 1H, 6-Ha), 3.89
(d, 6.3Hz, 1H, 3-H), 4.05 (s, 1H, 2-H), 4.14-4.30 (m, 5H, 7-, 20a-, 20(3-, 5ax-, 5eq-
H), 4.73 (s, 1H, 1-H), 4.83 (d, 3.0Hz, 1H, 2-H), 4.94 (d, 8.7Hz, 1H, 5-H), 5.25 (s, 1H,
10-H), 5.65 (d, 6.9Hz, 1H, 2-H), 5.80 (dd, 2.1, 8.7Hz, 1H, 3-H), 6.21 (t, 8.1Hz, 1H, 13-
H), 7.04 (s, 1H, 4-H), 7.31-7.52 (m, 10H, m-Bz, m,p-NBz, o,m,p-Ph), 7.59 (t, 7.5Hz,
1H, p-Bz), 7.60 (d, 9.3Hz, 1H, N-H), 7.77 (d, 7.2Hz, 2H, o-NBz), 8.10 (d, 7.5Hz, 2H, o-


32
ch3i
K¡coT
63
O
1.) n-BuLi
TMEDA
CH20, -780C
OCH3
69 OCH3
66 + 70
Figure 2-11: Synthesis of Taxamairin B


ACKNOWLEDGMENTS
There are many people who have helped me along this journey, some through
deeds and some through inspiration. Firstly I would like to acknowledge those in academia
that have made this event possible. To the late Koppaka V. Rao, my graduate advisor, I
would I would like to give much thanks for his patience and understanding and for the
wisdom and experience which he has imparted on me. He will certainly be missed but I
hope to keep his memory alive through the work I accomplish throughout my career. To
the late Maya Ganguli, my undergraduate advisor and mentor, I would like to give thanks
for opening my eyes to the exciting world of chemistry. I could not have asked for a
teacher more caring and better suited than you to help me get started in my chemistry
education. I would also like to acknowledge other faculty members that have given their
time and energy so that I could accomplish this goal; John Perrin for accepting the duty of
being my graduate advisor upon Dr. Raos passing, Margaret James, Ken Sloan, and
William Dolbier for serving on my thesis committee, and Bob Higgins for guidence and
always interesting conversation during my undergraduate studies. Finally I must give
thanks to my fellow students along the way both graduate and undergraduate with special
thanks to Veronica Hall and Bernice Kidd for their help and friendship during my
undergraduate years and to Ravi Orugunty and John Juchum whom I shared my trials and
tribulations with for the last 5 years as lab mates in Dr. Raos laboratory.


14
N-acyl group
required
\
acetyl or acetoxy
group may be
removed without
significant loss
of activity
phenyl group
or a close
analog required
free 2'-hydroxyl
group or a
hydrolysable
ester thereof
required
benzoyloxy group
essential; certain
substituted groups
have improved
activity
may be esterified
epimerized or
removed without
significant loss of
activity
oxetane ring
required for
activity
removal of acetate
reduces activity slight
Figure 1-8: Structure-Activity Relationships
implied by this conversion. These analogues, however, are much less cytotoxic in cell
culture which may be due to the instability of the A ring contracted analogues in cell
culture media or to its failure to enter the cell. A summary of structure-activity
relationships is shown in Figure 1-8.


90
10-Deacetyl baccatin III- 10-mononitrate ester (98)
White crystalline powder, mp 169-171 C, Anal. Calc, for C29H35NO12: C 59.08,
H 5.98, N 2.38. Fd. C 58.90, H 6.28, N 2.19. !H NMR 5: 1.09 (s, 3H, 17-H), 1.13 (s,
3H, 16-H), 1.70 (s, 3H, 19-H), 1.85 (m, 1H, 6-Hp), 2.13 (s, 3H, 18-H), 2.16 (m, 2H, 14-
Ha,P), 2.30 (s, 3H, 4-OAc), 2.62 (m, 1H, 6-Ha), 3.85 (d, 6.9Hz, 1H, 3-H), 4.15 (d,
8.4Hz, 1H, 20-Hp), 4.32 (d, 8.1Hz, 1H, 20-Ha), 4.41 (dd, 6.6, 10.2Hz, 1H, 7-H), 4.91 (t,
7.8Hz, 1H, 13-H), 4.97 (d, 9.6Hz, 1H, 5-H), 5.65 (d, 7.2Hz, 1H, 2-H), 6.49 (s, 1H, 10-
H), 7.49 (t, 7.8Hz, 2H, m-Bz), 7.62 (t, 7.5Hz, 1H, p-Bz), 8.10 (d, 7.2Hz, 2H, o-Bz). 13C
NMR 6: 9.5, 15.8, 20.8, 22.5, 26.6, 36.6, 38.5, 42.5, 46.6, 58.5, 67.8, 71.6, 74.6, 76.5,
78.8, 80.7, 82.9, 84.1, 128.7, 129.2, 129.5, 130.1, 133.8, 148.3, 167.0, 170.8, 203.1.
10-Deacetyl baccatin HI-10, 13-dinitrate ester (99)
White crystalline powder, mp 202-204 C, Anal. Calc, for C29H34N2O14: C 54.89;
H 5.40; N 4.41. Fd. C 55.17; H 5.77; N 4.07. NMR 8: 1.16 (s, 3H, 17-H), 1.18 (s, 3H,
16-H), 1.69 (s, 3H, 19-H), 1.87 (m, 1H, 6-Hp), 2.08 (s, 3H, 18-H), 2.34 (m, 1H, 14-Hp),
2.38 (s, 3H, 4-OAc), 2.46 (m, 1H, 14-Ha), 2.60 (m, 1H, 6-Ha), 3.77 (d, 7.2Hz, 1H, 3-
H), 4.12 (d, 8.1Hz, 1H, 20-HP), 4.32 ( d, 8.4Hz, 1H, 20-Ha), 4.40 (dd, 6.6, 10.8Hz, 1H,
7-H), 4.94 (d, 9.0Hz, 1H, 5-H), 5.65 (d, 7.2Hz, 1H, 2-H), 6.21 (t, 8.1Hz, 1H, 13-H), 6.47
(s, 1H, 10-H), 7.50 (t, 8.1Hz, 2H, m-Bz), 7.64 (t, 7.2Hz, 1H, p-Bz), 8.06 (d, 7.2Hz, 2H,
o-Bz). 13C NMR 5: 9.4, 15.1, 20.6, 22.2, 26.5, 34.2, 36.6, 42.8, 46.6, 58.4, 60.4, 71.4,
74.2, 76.2, 78.2, 78.3, 80.7, 81.9, 84.1, 128.7, 128.8, 130.0, 132.4, 133.9, 141.6, 166.9,
170.1, 202.0.


28
2-propanol, hexane
PTSA
48
49
O
1.) AgO, NaOH
H2O, 70 C, 20 min
K2CO3, acetone 51
50
20% H2S04
<
reflux, 6 hrs
H2, Ra Ni
O
'OCH3
OCH3
OCH3
2 CH3MgBr
Et20
OCH3
52
OCH3
1.) n-butyl lithium,
TMEDA, THF
O C
2.) (CH20)n HocH2
OCH3
54
PBr3, CH2CI2
room temp., 15 min
Figure 2-9: Literature Synthesis of Taxamairin B


CHAPTER 4
SYNTHESIS OF ANALOGUES WITH POTENTIALLY
IMPROVED WATER SOLUBILITY
Introduction
In spite of paclitaxels great promise in treatment of refractory and untreatable
human neoplasms it is afflicted with formulation and systemic administration problems.
These problems stem from its extreme low solubility in water which has been reported as
low as 0.25 pg/ml (Ali et al., 1995). Consequently, special formulations requiring
excipients such as Cremeophore EL have been necessary for intravenous administration.
In the case of paclitaxel the amount of Cremophore EL required to administer the
therapeutic dose (135-200 mg/m2) represents the highest amount ever to be used with any
drug. Exposure to the large amounts of Cremophore EL has produced major
hypersensitivity reactions in patients. It is perceived that such adverse effects are vehicle
related, since it is well documented that Cremophore EL alone causes hypotension and
histamine release in dogs. The high incidences and severity of these side effects to
paclitaxel almost led to the termination of some earlier Phase I clinical trials. However,
prolonged infusions and prophylactic medications with antihistamines and corticosteroids
have avoided the adverse episodes and allowed continuation of clinical use of Cremophore
EL formulation. Still, because of these problems and the additional medications needed,
an equally potent but more water-soluble analogue of paclitaxel is desirable. This area has
107


86
rearrangement takes place without this available amide proton. In any event time did not
allow this to be studied further.
Experimenta!
All reactions were monitored by silica gel 60 HF254 TLC to ensure completion of
the reaction. All NMR spectra were recorded using either a Varan VXR-300 or a Varan
Gemini-300 spectrophotometer using CDC13 as solvent. Infrared spectra were obtained
using a Perkin-Elmer 1420 ratio recording spectrophotometer. Ultraviolent spectra were
obtained using a Shimadzu UV160U recording spectrophotometer. Mass spectra were
recorded on a Finnigan Mat 950 Q spectrometer. Melting points were obtained by using a
Fisher melting point apparatus. Column chromatography was used in conjunction with
100-200 mesh silica gel.
Complete Nitrations of Taxanes
Paclitaxel-7, 2~dinitrate ester (95)
Paclitaxel 500 mg was dissolved in 6 ml of CH2CI2 and a mixture of 5 ml of acetic
anhydride and 1 ml of concentrated nitric acid was added slowly. The mixture was
prepared by slowly adding the nitric acid to ice cold acetic anhydride so that the mixture
does not get too hot. The reaction mixture was allowed to stir at room temperature for 30
minutes. At this point 20 ml of water was added and while stirring NaHC03 was slowly
added until no further frothing was observed. Additional CH2C12 was added and the water
layer was extracted 3 times with CH2C12. The organic layer was dried with Na2S04 and the
solvent was evaporated. The product was crystallized with diethyl ether and ligroin to


54
reaction was filtered and the filtrate was washed twice with 0.1 N HC1, twice with 0.1 N
NaOH, and twice with water. The organic layer was then dried with Na2S04 and
evaporated under a vacuum to a solid residue. This material was separated on a regular
silica column using 20-40% ethyl acetate in ligroin. A total of 62 mg of clear colorless oil
was obtained. Clear colorless oil, 'H NMR 5: 1.24 (d, 6.9Hz, 6H, CH3), 2.49 (s, 3H,
CH3), 3.37 (m, 1H, CH), 3.91 (s, 3H, OCH3), 3.95 (s, 3H, OCH3), 7.08 (s, 1H, Ar-H),
10.33 (s, 1H, CHO).
Condensation of keto-aldehyde 70 with dione 66
A total of 100 mg of keto-aldehyde 70 was added to 3 ml of pyridine containing
~10 drops of piperidine. To this mixture was added 80 mg of dione 66 and the mixture
was refluxed with stirring for 6 hours. At this time TLC analysis showed a product with
the same Rf value as the methylated taxamairin A (taxamairin B) along with other
products. The mixture was diluted with diethyl ether and washed three times with 0. IN
HC1 until the water layer was still acidic. The organic layer was then washed twice with
water, dried with Na2S04, and evaporated under a vacuum to a solid residue. The product
was isolated using a regular silica column with 20-40% ethyl acetate in ligroin and a total
of 46 mg of product was isolated as a yellowish white amorphous solid. All spectral data
was identical to that of the methylated taxamairin A.
Complete reduction of methyl isopropyl acetovanillone (72)
A total of 500 mg of methylated isopropyl acetovanillone was dissolved in 5 ml of
THF and 1.0 g of ZnCl2 was added and the mixture was stirred at 60 C. To this was
added NaCNBH3 in small portions and the TLC was monitored. When nearly all the


52
Methylation of isopropyl acetovanillone
A total of 1.0 g of isopropylated acetovanillone was dissolved in 20 ml of acetone
then 3.0 g of K2CO3 and 1 ml of dimethyl sulfate were added. The mixture was refluxed
for 3 hours at which time no starting material remained according to TLC. Then 1 ml of
cone. NH4OH was added and the mixture was stirred for 30 minutes. The acetone was
then partially removed and the residue was partitioned between water and diethyl ether.
The organic layer was washed twice with water, dried with sodium sulfate, and
concentrated to yield 956 mg of product (68) as a clear colorless oil.
Reduction of 68
A total of 1.0 g of 68 was dissolved in 8 ml of methanol with 3 drops of 1.0 N
NaOH. To this was added dropwise a solution of NaBH4 in methanol with 1.0 N NaOH
and the TLC was monitored. When no starting material remained the mixture was
acidified with 0.1 N HC1 and the methanol was partially removed. The residue was
partitioned between water and diethyl ether and the organic layer was washed twice with
water, dried with sodium sulfate, and concentrated under vacuum. This material was
chromatographed on a silica column using 25-50% ethyl acetate in ligroin as eluent to give
856 mg of alcoholic product (69) as a clear colorless oil. Clear colorless oil, EIMS m/z:
224 (97%, M), 209 (100%), 179 (18%), 139 (90%), 124 (33%). H NMR 5: 1.21 (d,
6.9Hz, 6H, CH3), 1.48 (d, 6.3Hz, 3H, CH3), 3.34 (heptet, 6.6Hz, 1H, CH), 3.80 (s, 3H,
OCH3), 3.86 (s, 3H, OCH3), 4.83 (q, 6.3Hz, 1H, CH), 6.81 (s, 2H, Ar-H). 13C NMR 5:
23.4, 25.0, 26.8, 55.6, 60.8, 70.5, 106.6, 115.1, 141.6, 142.2, 145.3, 152.5.


139
L1210 Cytotoxicity Assay
The cells are maintained and subcultured in RPMI medium that was prepared in a
sterile manner in the laboratory. The cell population was maintained between 150,000 and
600,000 cells/ml. At the time of the assay, the cell suspension was diluted to contain
150,000 cells/ml. This solution (2 ml) was placed in each well of a Becton-Dickinson
deep-well plate (24 wells/plate). The test compounds were weighed and dissolved in
sufficient DMSO to make 2 mg/ml. Then several dilutions of these stock solutions were
made and tested. To the wells that contained the L1210 cells were added 10 pi aliquots of
the DMSO solutions so that the final concentration was known in parts per million. For
paclitaxel the concentrations used were 0.1, 0.05, 0.025, 0.0125, 0.00625, and 0.003125
ppm, and for all other compounds it was 10, 2, 0.2, 0.1, and 0.05 ppm. Controls were also
used in which 10 pi of DMSO were added to the well. Each concentration was tested in
quadruplicate (4 wells).
After addition of the compounds the plates were incubated for 48 hours. The
plates were removed and the cells in each well were counted to determine the number of
cells per ml. The contents of each was thoroughly mixed using a sterile 2 ml pipette, then,
1 ml of the cell suspension was transferred to a clean test tube and diluted with 1 ml of
trypan blue stain which only dyes the dead cells. The viable cell (unstained) were then
counted by shaking and placing 0.1 ml of this suspension on a Fisher hemacytometer and
counting the cells found in the five gridded areas of the hemcytometer. This number was
then multiplied by 4000 which gave the number of cell/ml.


56
Brominatic n of 73
A total of 100 mg of 73 was dissolved in 3.0 ml of CH2CI2 and a dilute bromine
solution in CH2C12 was added dropwise with stirring at room temperature and the TLC
was monitored. The addition was stopped when a small amount of starting material
remained and a faster moving product spot was present. Water and more CH2C12 was
added to the mixture and partitioned. The organic layer was washed twice with 0.1 N HC1
and twice with water, dried with sodium sulfate, and concentrated. The residue was
separated on a silica column using 15-25% ethyl acetate in ligroin and 73 mg of product
(75) was isolated as a yellowish oil. Yellow oil, 'H NMR 5: 1.22 (t, 7.5Hz, 3H, CH3),
1.35 (d, 6.6Hz, 6H, CH3), 2.75 (q, 7.5, 15.3Hz, 2H, CH2), 3.72 (m, 1H, CH), 3.83 (s, 3H,
OCH3), 3.84 (s, 3H, OCH3), 6.69 (s, 1H, Ar-H). 13C NMR 5: 14.4, 21.0, 30.9, 34.2,
55.7, 60.9, 111.0, 139.0, 140.5, 152.0.
Brominaticn of 69
A total of 100 mg of 69 was dissolved in 3.0 ml of CH2C12 and a dilute bromine
solution in CH2C12 was added dropwise with stirring at room temperature and the TLC
was monitored. The addition was stopped when only a small amount of starting material
remained and a faster moving product spot was present. Water and more CH2C12 was
added to the mixture and partitioned. The organic layer was washed twice with 0.1 N HC1
and twice with water, dried with sodium sulfate, and concentrated. The residue was
separated on a silica column using 15-25% ethyl acetate in ligroin and 58 mg of product
(76) was isolated as a yellowish oil. Yellow oil, 'H NMR 5: 1.34 (t, 6.6Hz, 6H, CH3),
2.02 (d, 6.9Hz, 3H, CH3), 3.71 (m, 1H, CH), 3.86 (s, 3H, OCH3), 3.89 (s, 3H, OCH3),


96
42.9, 45.7, 55.3, 57.8, 59.4, 72.0, 72.7, 73.2, 74.0, 74.2, 74.4, 76.4, 78.5, 80.2, 80.5,
82.2, 83.5, 98.8, 127.0, 127.1, 128.4, 128.7, 128.8, 129.0, 129.1, 130.1, 131.1, 132.1,
133.5, 133.8, 137.7, 143.2, 166.8, 167.4, 170.9, 172.7, 200.0.
10-Deacetyl paclitaxeI-7-p-xyIoside-2-mononitrate ester (105)
White amorphous powder, Anal. Calc, for C50H56N2O19 + H20: C 59.64; H 5.81;
N 2.78. Fd. C 60.02; H 6.18; N 2.40. HNMR5: 1.17 (s, 3H, 17-H), 1.26 (s, 3H, 16-H),
1.74 (s, 3H, 19-H), 1.81 (s, 3H, 18-H), 2.04 (m, 1H, 6-Hp), 2.29 (m, 2H, 14-Ha,p), 2.36
(s, 3H, 4-OAc), 2.72 (m, 1H, 6-Ha), 3.17 (t, 10.8, 1H, 5-Heq), 3.50 (t, 8.7Hz, 1H, 3-
H), 3.66 (dd, 7.5, 12.3Hz, 1H, 4-H), 3.84 (d, 6.6Hz, 1H, 3-H), 3.91 (dd, 4.5, 11.4Hz,
1H, 5-Hax), 4.04 (m, 1H, 7-H), 4.18 (m, 1H, 1-H), 4.19 (d, 8.7Hz, 1H, 20-HP), 4.28
(d, 8.7Hz, 1H, 20-Ha), 4.79 (m, 1H, 2-H), 4.81 (m, 1H, 2-H), 4.88 (d, 9.0Hz, 1H, 5-
H), 5.07 (s, 1H, 10-H), 5.61 (d, 6.6Hz, 1H, 2-H), 5.73 (dd, 2.1, 8.7Hz, 1H, 3-H), 6.15 (t,
7.8Hz, 1H, 13-H), 7.23 (d, 9.0Hz, 1H, N-H), 7.34-7.53 (m, 10H, m-Bz, o,m,p-Ph, m,p-
NBz), 7.61 (t, 7.2Hz, 1H, p-Bz), 7.74 (d, 7.8Hz, 2H, o-NBz), 8.09 (d, 7.5Hz, 2H, o-Bz).
13C NMR 5: 10.6, 14.2, 20.4, 22.5, 26.6, 35.7, 35.8, 43.0, 46.5, 55.3, 57.0, 65.1, 69.9,
72.4, 73.2, 73.6, 74.4, 74.6, 76.5, 78.6, 80.6, 80.9, 81.8, 84.0, 101.5, 127.1, 127.2, 128.4,
128.7, 128.8, 129.0, 129.1, 130.2, 132.1, 133.6, 133.7, 136.1, 137.8, 137.9, 166.8, 167.4,
170.8, 172.7, 209.2.
10-Deacetyl paclitaxel-7-P-xyloside-3-mononitrate ester (106)
White crystalline powder, mp 209-211 C, Anal. Calc, for C50H56N2O19 + H20: C
59.64; H 5.81; N 2.78. Fd. C 59.27; H 5.87; N 2.89. ^NMRS: 1.09 (s,3H, 17-H), 1.19
(s, 3H, 16-H), 1.80 (s, 3H, 19-H), 1.81 (s, 3H, 18-H), 2.02 (m, 1H, 6-Hp), 2.28 (m, 2H,


67
94
O
5:1/ acetic anhydride : HNO 3
CH2C12
room temp., 30 min
Figur 3=1} Nitration of Paelitaxei, 7-OH > 2* 011
ring or cleavage of the oxetane ring, both of which have been known to occur in the
presence of strong acids (Chen et al., 1993).


41
O
Figure 2-16: Compounds from the Needles of Taxus floridana
these compounds seem to exist in an equilibrium between two conformers. If the spectrum
is taken at low temperature however, two sets of signals can be distinguished.


side chain reacts in mild base to yield the corresponding baccatin III and dibenzamide.
Finally, a series of potentially more water soluble paclitaxel analogues are prepared by
oxidizing the naturally occurring 10-deacetyl paclitaxel-7-xyloside with periodate and then
reacting the resulting dialdehyde with amines and carbon nucleophiles. These compounds
are also tested for cytotoxicity in the L1210 assay system. Although these compounds are
not as active as paclitaxel most are more active than the xyloside from which they are
obtained.
xiv


115
These conditions gave clean products which were easily purified by column
chromatography. The following amines were used: p-aminobenzoic acid, p-aminosalicylic
acid, m-aminosalicylic acid, p-nitroaniline, and p-aminobenzenesulfonamide which was
obtained by reducing p-nitrobenzenesulfonamide with Sn and HC1. These amines
generated the corresponding morpholino analogues (160-164) (Figure 4-5).
Attempted Synthesis of Taxane Glycosides
An alternative method of preparing analogues with increased water-solubility was
also studied in this work. This method involved attaching sugar units to the taxane moiety
and thereby forming glycosides. This would certainly increase the water-solubility and
quite possibly have a positive effect on the activity of the drug since it is well known that
carbohydrates play a vital role in molecular recognition. The only taxane glycoside that
had been studied was the naturally occurring xylosides and although they displayed very
good activity in tubuline assays but lost activity in cell culture as mentioned, it was hoped
that attaching different sugars would overcome this problem. Also the possibility existed
that if a di- or trisaccharide could be attached to paclitaxel or one of its close analogues
then possibly it would become orally available. This type of behavior is seen with the
cardiac glycosides in which if the sugars are removed the compound in no longer orally
active. As of this writing two papers have been published in which a taxane has been
linked with a carbohydrate unit to increase the water-solubility (Paradis & Page, 1998;
Takashi et ah, 1998).


ABBREVIATIONS
Ac acetate
Bn benzyl
Bz benzoate
CIMS chemical ionization mass spectroscopy
DCC dicyclohexylcarbodiimide
DDQ 2, 3-dichloro-5, 6-dicyano-l, 4-benzoquinone
DMF dimethylformamide
DMSO dimethyl sulfoxide
EIMS electron impact mass spectroscopy
FABMS fast atom bombardment mass spectroscopy
HMBC heteronuclear multiple bond correlation
HPLC high pressure liquid chromatography
LAH lithium aluminum anhydride
LDA lithium diisopropylamide
NMR nuclear magnetic resonance
NOE nuclear Overhauser effect
PDC pyridinium dichromate
PTSA para-toluene sulphonic acid
RaNi rainey nickel
TBS tert-butyl dimethyl silyl
TES triethyl silyl
Tf triflate
TLC thin layer chromatography
TMEDA tetramethylethylenediamine
xii